Ankur's PhD Thesis - TSpace

Influence of Crystalline Microstructure on OpticalResponse of Single ZnSe Nanowires

by

Ankur Saxena

A thesis submitted in conformity with the requirementsfor the degree of Doctor of Philosophy

Graduate Department of Materials Science and EngineeringUniversity of Toronto

Copyright © 2012 by Ankur Saxena

Abstract

Influence of Crystalline Microstructure on Optical Response of Single ZnSe Nanowires

Ankur Saxena

Doctor of Philosophy

Graduate Department of Materials Science and Engineering

University of Toronto

2012

Semiconductor nanowires (NWs) are anticipated to play a crucial role in future electronic

and optoelectronic devices. Their practical applications remain hindered by an urging

need for feasible strategies to tailor their optical and electronic properties. Strategies

based on strain and alloying are limited by issues such as defects, interface broadening

and alloy scattering. In this thesis, a novel method to engineer the optoelectronic prop-

erties based on strain-free periodic structural modulations in chemically homogeneous

Nanowire Twinning Superlattices (NTSLs) is experimentally demonstrated. NTSLs are

an emerging new class of nanoscale material, composed of periodically arranged rotation

twin-planes along the length of NWs. The main objective of this thesis is to establish

the relationship between the electronic energy band gap (Eg) and the twin-plane spac-

ing (d) in NTSLs, quantified using a periodicity parameter, based on ZnSe. ZnSe was

chosen because of its excellent luminescence properties, and potential in fabrication of

optoelectronic devices in the near-UV and blue region of the spectrum.

A prerequisite to establishing this correspondence is a prior knowledge of the photolu-

minescence (PL) response and the nature of fundamental optical transitions in defect-free

single crystal ZnSe NWs with ZB and wurtzite (WZ) crystal structures. There has been

no systematic work done yet on understanding these fundamental optical processes, par-

ticularly on single NWs and in relation to their crystalline microstructure. Therefore, the

secondary objective of this thesis is to study the influence of native point defects on the

ii

optical response of single ZnSe NWs in direct relation to their crystalline microstructure.

The PL response from single ZB and WZ NWs was determined unambiguously, and

excitonic emission linewidths close to 1 meV were observed, which are the narrowest

reported linewidths thus far on ZnSe NWs. Based on this and extensive optical and

structural characterization on individual NTSLs, a linear variation in Eg is shown through

a monotonic shift in PL peak position from ZnSe NTSLs as a function of d, with Eg’s

that lie between those of ZB and WZ crystal structures. This linear variation in Eg was

also validated by ab Initio electronic structure calculations. This establishes NTSLs as

new nanoscale polytypes advantageous for applications requiring tunable band gaps.

iii

Dedication

To my loving grandmother

Smt. Katori Devi

and grandfather

Sri Raj Narayan Saxena

iv

Acknowledgements

It is my distinct pleasure to thank all those who helped me in one way or another in

making this thesis possible. First and foremost, my sincerest gratitude is reserved for

my thesis supervisor, Dr. Harry E. Ruda, for his constant support, expert guidance and

consistent encouragement. I am most grateful to him for the flexibility in research he

gave me, which allowed me to perform the experimental work that I enjoyed the most.

I extend my sincere thanks to Mitsuru Sugisaki, an expert in single quantum dot

spectroscopy and a faculty member at Osaka University in Japan, who spent some

time with me in the optical lab on his brief trip to Toronto. In those few hours, I

learned immensely about the design of experimental arrangement for confocal micro-

photoluminescence spectroscopy, and ways to attain precision in optical alignment. I am

deeply indebted to my former colleague and a close friend, Usha Philipose, for providing

me with the ZnSe nanowire samples for my work. I also would like to thank Qi Pan

(Eric), a former undergraduate thesis student in my research group, for providing me

with ZnSe nanowire samples after Usha joined as a faculty member at the University of

North Texas. But for the expertise of Usha and tireless work by Eric in synthesizing

ZnSe nanowires, my research work would not have been possible.

I am grateful to Selva V. Nair, a senior scientist in my research group, who enthusi-

astically agreed to perform the electronic structure calculations on Nanowire Twinning

Superlattices. These were crucially helpful in verifying the experimental data. I am

also thankful to him for stimulating discussions at several occasions, which helped me in

my understanding of the optical processes in semiconductors. I sincerely thank Christina

Souza, and Carlos Fernandes, both senior scientists, for proof-reading my thesis and their

comments. I am very grateful to Christina for always answering my questions related to

research work, optics and instruments inside the optical lab, troubleshooting and admin-

istrative in nature. I am also thankful to Carlos for his help every time the chiller used

to break down in the optical lab.

v

The electron microscopy work described in this thesis would not have been possible

without the technical assistance from Fred Pearson and Carmen Andrei at the Canadian

Centre for Electron Microscopy (CCEM). I am thankful to both of them for doing their

best in accommodating my requests for time on the instruments, considering that these

were always booked well in advance. In particular, I am grateful to Fred for his persistence

and patience while working on my NW samples, when at times it used to take hours in

tilting the specimen holder to get the right diffraction conditions. Those familiar with

the Ti:Sapphire lasers know that they are far from being plug-and-play systems. I would

like to thank Suxia Yang who taught me how to use the Ti:Sapphire and other lasers

inside the optical lab in the first few years of my project, when I was a novice.

I would like to thank my supervisory committee members, Prof. Glenn Hibbard,

Prof. Francis Dawson and Prof. Z. H. Lu, for a careful reading of my thesis and their

invaluable inputs.

I am thankful to all graduate students in my group, former and present, in creating

a healthy work environment conducive for research. In particular, I am thankful to Joe

Salfi, my colleague for almost the entire duration of my project and a very good friend,

for his extremely useful suggestions with regards to trouble-shooting in the lab, and

frequent interesting discussions. I would also like to thank Millie Morris, our business

and administrative officer, who was always friendly and efficiently handled the matters

related to purchases and returns.

It would be an act of injustice if I do not express my gratefulness to my wife, Girija

Dharmaraj, for her immense patience, constant personal support, and encouragement at

times when experiments would not go as I had hoped. I would also like to acknowledge

her help in providing me with the MATLAB scripts at times. I am also most grateful to

my parents and my sisters, Shanu and Ankita, for their continued encouragement and

support for the duration of my research work.

I gratefully acknowledge financial support from Eleanor and Burnett Thall - Ontario

vi

Graduate Scholarship in Science and Technology (OGSST), J. Edgar McAllister Grad-

uate Award, University of Toronto Open fellowships, funding from Dr. Ruda and the

Department of Materials Science and Engineering at the University of Toronto. The

electron microscopy research described in this thesis was performed at the CCEM, which

is supported by NSERC and other government agencies.

vii

Contents

Abstract ii

List of Tables xii

List of Figures xvi

List of Acronyms xxxi

List of Symbols xxxv

Glossary of terms xliii

1 Background-Current state of the art 1

1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1.1 II-VI compound semiconductors . . . . . . . . . . . . . . . . . . . 1

1.1.2 ZnSe for opto-electronic devices . . . . . . . . . . . . . . . . . . . 2

1.2 Semiconductor nanowires . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.2.1 ZnSe nanowires . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

1.3 Semiconductor nanowire heterostructures . . . . . . . . . . . . . . . . . . 6

1.3.1 Nanowire Twinning Superlattices (NTSLs) . . . . . . . . . . . . . 7

1.4 Growth of semiconductor nanowires . . . . . . . . . . . . . . . . . . . . . 10

1.4.1 Growth and structural characterization of ZnSe nanowires . . . . 12

1.5 Luminescence characterization of ZnSe single crystals and thin films . . . 13

viii

1.5.1 Characteristic features of ZnSe LTPL spectrum . . . . . . . . . . 15

1.5.2 Excitonic emission in ZnSe . . . . . . . . . . . . . . . . . . . . . . 17

1.6 Optical response of ZnSe nanowires . . . . . . . . . . . . . . . . . . . . . 19

1.6.1 Distinction from ZnSe single crystals and thin films . . . . . . . . 19

Distinguishing features . . . . . . . . . . . . . . . . . . . . . . . . 20

Confinement effects in nanowires . . . . . . . . . . . . . . . . . . 21

1.6.2 Luminescence characterization of ZnSe nanowires . . . . . . . . . 22

1.6.3 Influence of heat treatment . . . . . . . . . . . . . . . . . . . . . . 25

1.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

2 Motivation and Objectives 30

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

2.2 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

2.3 Gaps in the understanding of fundamental optical transitions in ZnSe NWs

in literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

2.3.1 Role of native point defects . . . . . . . . . . . . . . . . . . . . . 35

2.3.2 Lack of luminescence studies on single ZnSe nanowires . . . . . . 37

2.4 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

3 Experimental Methodology 44

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

3.2 Rationale behind the choice of experimental techniques . . . . . . . . . . 46

3.2.1 Advantages of PL spectroscopy . . . . . . . . . . . . . . . . . . . 47

3.2.2 Limitations of PL spectroscopy . . . . . . . . . . . . . . . . . . . 49

3.3 Growth of ZnSe nanowires and NTSLs . . . . . . . . . . . . . . . . . . . 50

3.4 Obstacles in combined optical and structural characterization of the same

individual nanowire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

ix

3.4.1 Bottleneck - incompatible sample preparation methods for different

techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

3.5 Photoluminescence (PL) spectroscopy . . . . . . . . . . . . . . . . . . . . 56

3.6 Confocal Micro-photoluminescence (µ-PL) Spectroscopy . . . . . . . . . 61

3.7 Time-resolved Photoluminescence (TRPL) . . . . . . . . . . . . . . . . . 66

3.8 Transmission Electron Microscopy (TEM) . . . . . . . . . . . . . . . . . 69

3.9 Acquisition and analysis of PL results . . . . . . . . . . . . . . . . . . . . 70

3.9.1 Data acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

Integration time ti . . . . . . . . . . . . . . . . . . . . . . . . . . 71

Slit-width dslit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72

Excitation Intensity Iexc . . . . . . . . . . . . . . . . . . . . . . . 73

3.9.2 Spectral correction and normalization of PL spectra . . . . . . . . 74

Spectral correction . . . . . . . . . . . . . . . . . . . . . . . . . . 74

Normalization of PL spectra . . . . . . . . . . . . . . . . . . . . . 75

3.9.3 Data analysis using curve-fitting . . . . . . . . . . . . . . . . . . . 77

3.10 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

4 Influence of stoichiometry on optical response 82

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

4.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

4.2.1 Low-temperature PL spectra . . . . . . . . . . . . . . . . . . . . . 83

4.2.2 Excitation intensity dependence of PL spectra . . . . . . . . . . . 85

4.3 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

4.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

5 Single zinc-blende ZnSe nanowires 99

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

5.2 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

x

5.2.1 Characterization of single ZB ZnSe NWs . . . . . . . . . . . . . . 100

HR-TEM characterization . . . . . . . . . . . . . . . . . . . . . . 101

Low-temperature PL spectra . . . . . . . . . . . . . . . . . . . . . 101

5.2.2 Comparison with an ensemble of ZB NWs . . . . . . . . . . . . . 111

5.2.3 Dependence of PL on temperature . . . . . . . . . . . . . . . . . 115

5.2.4 Dependence of LTPL on Iexc . . . . . . . . . . . . . . . . . . . . . 126

5.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126

6 Single wurtzite ZnSe nanowires 130

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

6.2 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

6.2.1 Characterization of single WZ nanowires . . . . . . . . . . . . . . 131

HR-TEM characterization . . . . . . . . . . . . . . . . . . . . . . 132

Low-temperature PL spectra . . . . . . . . . . . . . . . . . . . . . 134

6.2.2 Dependence of PL on temperature . . . . . . . . . . . . . . . . . 141

6.2.3 Dependence of LTPL on Iexc . . . . . . . . . . . . . . . . . . . . . 145

6.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148

7 ZnSe Nanowire Twinning Superlattices 150

7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

7.1.1 Periodicity parameter (γ) . . . . . . . . . . . . . . . . . . . . . . 151

7.1.2 Excitation intensity dependence of LTPL . . . . . . . . . . . . . . 157

7.2 Variation in band gap of NTSLs . . . . . . . . . . . . . . . . . . . . . . . 159

7.3 Time-resolved photoluminescence . . . . . . . . . . . . . . . . . . . . . . 162

7.4 Dependence of PL on temperature . . . . . . . . . . . . . . . . . . . . . . 166

7.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170

8 Conclusions and Outlook 173

8.1 Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182

xi

A Bound-exciton lines in ZnSe films 184

B Survey of luminescence studies on ZnSe nanostructures 187

C Photoluminescence (PL) Spectroscopy 194

C.1 Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194

C.2 Theoretical models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202

C.2.1 Excitation intensity dependence of PL . . . . . . . . . . . . . . . 202

Dependence of IPL on Iexc . . . . . . . . . . . . . . . . . . . . . . 203

Dependence of peak energy on Iexc . . . . . . . . . . . . . . . . . 204

C.2.2 Temperature dependence of PL . . . . . . . . . . . . . . . . . . . 205

Dependence of peak energy on T . . . . . . . . . . . . . . . . . . 205

Dependence of HWHM on T . . . . . . . . . . . . . . . . . . . . . 207

Dependence of IPL on T . . . . . . . . . . . . . . . . . . . . . . . 208

D Time-resolved Photoluminescence (TRPL) spectroscopy 211

E Electronic band structures 214

E.1 Energy band diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214

E.2 Band structure calculations for Nanowire Twinning Superlattices . . . . . 217

F Softwares and Programs used 221

G Supplementary TEM micrographs 223

G.1 Chapter 5 - ZB ZnSe nanowires . . . . . . . . . . . . . . . . . . . . . . . 223

G.2 Chapter 6 - WZ ZnSe nanowires . . . . . . . . . . . . . . . . . . . . . . . 225

G.3 Chapter 7 - ZnSe NTSLs . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

Bibliography 240

Author Index 279

xii

List of Tables

4.1 Table showing the values of different fitting parameters of Equation 4.3 . 93

5.1 Experimental and calculated ratios of the intensity of the m-th order LO-

phonon replica (Im) to that of the ZPL (I(m=0)) for the Id1 and Id′

1 lines.

The calculated ratios were found using Equation 5.2. Nph are the values

in columns (2) and (4) corresponding to m = 1. The intensity ratios are

listed for ZB-NW-1, 2 and 3, and a bundle of ZB NWs. These ratios

reported by Jiang et al. are also listed [287]. . . . . . . . . . . . . . . . . 109

5.2 Values of the fit parameters of the Varshni’s equation (Equation 5.3) which

describes the temperature dependence of the Eg. (See Figure 5.10). The

values reported previously by other authors are also listed for comparison.

A fitting error of 0 indicates error of the order of 10−5 eV. . . . . . . . . 117

5.3 Values of the fit parameters of the Bose-Einstein type expression (Equa-

tion 5.4) which describes the temperature dependence of the direct Eg.

(See Figures 5.10 and 5.11 (a,b)). The values reported by Malikova et

al. [263] are also listed for comparison. A fitting error of 0 indicates error

of the order of 10−5 eV. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120

5.4 Values of the fit parameters of the Equation 5.5 which describes the tem-

perature dependence of the exciton linewidth (HWHM). (See Figures 5.12

and 5.13 (a,b)). The values previously reported by other authors are also

listed for comparison. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

xiii

6.1 Phonon energies of the acoustic and optical phonons in WZ ZnSe, as given

by Liang and Yoffe [305]. . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

6.2 Proposed assignments of the PL peaks observed in the LTPL spectrum

from WZ-NW-1. The assignment given by Liang and Yoffe [305] are also

given as a reference for the common peaks observed, and I1 and I2 peaks

(their notation). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138

6.3 Experimental and calculated ratios of the intensity of the m-th order LO-

phonon replica (Im) to that of the ZPL (I(m=0)) for the DAP-related tran-

sitions in the LTPL from WZ-NW-1. The calculated ratios were found

using Equation 5.2. Nph are the values in columns (3) and (6) correspond-

ing to m = 1. The values of these ratios given by Liang and Yoffe [305]

are also given for comparison. . . . . . . . . . . . . . . . . . . . . . . . . 139

6.4 Values of the fit parameters of the Bose-Einstein type expression (Equa-

tion 5.4) which describes the temperature dependence of the direct Eg.

(See Figure 6.8). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143

6.5 Values of the fit parameters of the Equation 5.5 which describes the tem-

perature dependence of the exciton linewidth (HWHM). (See Figure 6.10).

The values obtained for the Id1 line for ZB-NW-3 are also listed for com-

parison. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148

7.1 Twin-plane spacing d, error in d (δd) (s.e.m.) and dispersion in twin-plane

spacing ∆d (s.d.) for NTSLs 1-5 whose distributions of segment widths

are shown in Figure 7.2. NS is the sample size. Also listed are the results

of Kolmogorov Smirnov (K-S) normality test (for an alpha level of 0.05)

for the distributions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

7.2 Values of the fit parameters of Equations 5.3 (Varshni’s equation) and 5.4

(Bose-Einstein type expression) which describe the temperature depen-

dence of the Eg. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168

xiv

7.3 Values of the fit parameters of Equation 5.5 which describes the temper-

ature dependence of the exciton linewidth (HWHM). . . . . . . . . . . . 170

8.1 Summary of the recombination lines identified in ZB and WZ ZnSe NWs.

The exciton binding energies (EBX) for donor and acceptor bound exci-

tons, ionization energies of the donors (ED) and acceptors (EA) are also

listed. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180

8.2 Summary of the values of the fit parameters determined using dependene

of PL on temperature for single ZB and WZ NWs and a single NTSL. . . 181

A.1 Donors and donor-bound excitons in ZnSe films. Line positions at 4.2

K. EBX(D0X) exciton-donor binding energy, ED donor ionization energy,

dfilm film thickness. Reprinted with permission from Ref. [7] © 1990 John

Wiley & Sons. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185

A.2 Acceptors and acceptor-exciton complexes in ZnSe films. Line positions at

4.2 K. EBX(A0X) exciton-acceptor binding energy, EA acceptor ionization

energy, dfilm film thickness. Reprinted with permission from Ref. [7] ©

1990 John Wiley & Sons. . . . . . . . . . . . . . . . . . . . . . . . . . . . 186

B.1 Summary of room-temperature and low-temperature luminescence (PL

and CL) studies on ZnSe nanostructures. Also included are reported crys-

tal structures and morphologies of nanostructures. . . . . . . . . . . . . . 187

B.2 Summary of energy positions of the near band-edge (NBE) peak and deep-

level (DL) emission band reported for ZnSe nanostructures using their

RTPL characterization. . . . . . . . . . . . . . . . . . . . . . . . . . . . 188

B.3 Summary of LTPL studies on ZnSe nanowires. The energy position(s) of

the PL line(s), their assignment(s)a, ionization energies of the donors (ED)

and acceptors (EA), temperature of the experiments, and any intentional

doping carried out are listed. . . . . . . . . . . . . . . . . . . . . . . . . . 189

xv

List of Figures

1.1 (a) A schematic illustration of a photodetector device based on an indi-

vidual single crystal ZnSe nanobelt whose SEM image is shown in (b). (c)

SEM image of a photodetector based on a single ZnSe NW. The image

shows a four-terminal contact structure. The scale bar is 4 µm. Reprinted

with permission from (a,b) Ref. [121] © 2009 John Wiley & Sons, and (c)

Ref. [71] © 2006 American Institute of Physics. . . . . . . . . . . . . . . 5

1.2 (a,b) Overview and (c) high-resolution TEM images of InP NWs of differ-

ent diameters. Scale bars in (a) 100 nm, (b) 50 nm and (c) 5 nm. There is

no fixed viewing direction because of the different orientations of NWs. (d)

TEM image of an InAs NW viewed along the 〈110〉 direction showing the

periodically arranged twin-planes. (e) An atomic model of a section of a

NTSL with the locations of twin-planes marked by arrows. The twin-plane

spacing (d) is the spacing between two consecutive twin-planes, e.g. the

distance between the planes marked by A and C. Reprinted with permis-

sion from (a,b,c) Ref. [132], (d) Ref. [124] © 2008 Macmillan Publishers

Ltd., and (e) Ref. [122] © 2011 IEEE. . . . . . . . . . . . . . . . . . . . 8

1.3 Schematic illustrating the principles of growth of NWs based on a Vapour-

Liquid-Solid (VLS) growth mechanism. See text for details. . . . . . . . . 11

xvi

1.4 Near band-edge luminescence at 1.6 K from Ga-doped ZnSe layer grown

epitaxially on GaAs (100) substrate. General features of the spectrum are

explained in the text. Reprinted with permission from Ref. [7] © 1990

John Wiley & Sons. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

1.5 (a) RTCL spectrum from an individual ZnSe NW showing two peaks re-

lated to NBE and DL emission at 2.68 eV and 1.96 eV, respectively. (b)

LTPL spectrum (10 K) from an array of Ag-doped ZnSe NWs. Inset shows

the near band-edge region enlarged. See text for details. Reprinted with

permission from (a) Ref. [102] © 2011 Elsevier, and (b) Ref. [106] © 2011

American Institute of Physics. . . . . . . . . . . . . . . . . . . . . . . . . 24

1.6 RTPL spectrum from ZnSe NWs: (a) grown under Zn-rich conditions

showing stronger NBE emission compared to the weaker DL emission,

and (b) grown under Se-rich conditions showing strong DL emission while

NBE emission is absent. Insets show dominating (a) blue and (b) red

luminescence from single ZnSe NWs. Reprinted with permission from

Ref. [72] © 2006 American Institute of Physics. . . . . . . . . . . . . . . 26

3.1 A schematic of the experimental arrangement used for PL spectroscopy.

(M=mirrors, FM=flip-mirrors (mirrors on flip mounts), BS=beam-splitters,

LS=laser spectrometer, AC=auto-correlator) . . . . . . . . . . . . . . . . 58

3.2 Photograph of a section of the experimental arrangement for PL spec-

troscopy. The optical fibre is placed close to the microscope objective. . . 59

3.3 Photograph of the experimental arrangement for confocal µ-PL spectroscopy.

Inset in the lower right corner shows the confocal part with a pin-hole in

the focal plane. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

3.4 Schematic of the confocal µ-PL spectroscopy setup. Inset: Area on an

as-grown sample excited by the laser. The laser spot is elliptical due to

the oblique incidence of the exciting laser beam. . . . . . . . . . . . . . . 64

xvii

3.5 A simplified schematic of the detection of time-resolved photoluminescence

TRPL using a streak camera. . . . . . . . . . . . . . . . . . . . . . . . . 67

4.1 (a) LTPL spectrum from ZnSe NWs grown under excess Zn growth condi-

tions. The PL was taken at 3.1 K with average Iexc=5 W/cm2. (b) Near

band-edge region of the spectrum shown in (a), showing three peaks and

a shoulder near the lowest energy peak. These peaks can be resolved into

five different emission peaks. . . . . . . . . . . . . . . . . . . . . . . . . . 83

4.2 (a) LTPL spectrum from ZnSe NWs grown under excess Se growth condi-

tions. The PL was taken at 3.1 K with average Iexc=2.5 W/cm2. (b) Near

band-edge region of the spectrum shown in (a), showing a broad peak.

This peak can be resolved to show two different emission peaks. . . . . . 84

4.3 Dependence of PL from ZnSe NWs grown with excess Zn on Iexc. The

average Iexc was varied from 50 mW/cm2 to 25 W/cm2. All spectra were

recorded at 3.1 K. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

4.4 Change of integrated PL intensity (IPL) for individual emission peaks

shown in Figure 4.1 (b) with Iexc (log-log scale). The plots have been

offset for clarity. The solid line is a fit to Equation (4.1). See text for

more details. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

4.5 Variation of peak energy with Iexc for emission peak B at 2.714 eV (shown

in Figure 4.1 (b)). The solid line is a fit to the Equation 4.3. . . . . . . . 90

4.6 Variation of peak energy with Iexc for emission peak C at 2.686 eV (shown

in Figure 4.1 (b)). The solid line is a fit to the Equation 4.3. . . . . . . . 91

5.1 Structural characterization of ZB-NW-1: (a) Overview TEM micrograph,

(b) HR-TEM micrograph from an area of ZB-NW-1, and (c) Indexed SAD

pattern. The indexed spots correspond to a ZB crystal structure. The

viewing direction for (b) and (c) is 〈110〉. . . . . . . . . . . . . . . . . . . 102

xviii

5.2 (a) HR-TEM micrograph from an area of ZB-NW-1, (b) Indexed FFT

pattern taken from the area squared (in red) in (a). The indexed spots

correspond to a ZB crystal structure. The viewing direction corresponds

to 〈110〉. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

5.3 Excitonic region of the LTPL spectrum (3.5 K) from ZB-NW-1 acquired

under the conditions of high-throughput. (Iexc and dslit are as indicated).

The inset shows a confocal PL image of ZB-NW-1 under laser excitation

with an average Iexc=12.7 W/cm2. The diameter of pin-hole in the PL

image is 5 µm. The LTPL spectrum was fit to individual emission peaks

using Lorentzians as shown. . . . . . . . . . . . . . . . . . . . . . . . . . 104

5.4 Excitonic region of the LTPL spectrum (3.6 K) acquired from ZB-NW-2

under the conditions of (a) high-throughput, and (b) high-resolution. (Iexc

and dslit are as indicated). The inset in (a) shows a confocal PL image

of ZB-NW-2 under laser excitation with average Iexc=12.7 W/cm2. The

LTPL spectra were fit to individual emission peaks using Lorentzians as

shown. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

5.5 Excitonic region of the LTPL spectrum (3.8 K) acquired from ZB-NW-

3 under the conditions of (a) high-throughput, and (b) high-resolution.

(Iexc and dslit are as indicated). The LTPL spectra were fit to individual

emission peaks using Lorentzians as shown. . . . . . . . . . . . . . . . . . 107

5.6 Excitonic region of the LTPL spectrum (3.7 K) from ZB-NW-1 acquired

under the conditions of high-resolution. (Iexc and dslit are as indicated).

The inset shows a confocal PL image of ZB-NW-1 under laser excitation

with an average Iexc=12.7 W/cm2. The diameter of pin-hole in the PL

image is 5 µm. The LTPL spectrum was fit to individual emission peaks

using Lorentzians as shown. . . . . . . . . . . . . . . . . . . . . . . . . . 110

xix

5.7 Excitonic region of the LTPL spectrum (3.9 K) from a bundle of ZB

NWs acquired under the conditions of (a) high-throughput, and (b) high-

resolution. (Iexc and dslit are as indicated). The inset in (a) shows a

confocal PL image of the bundle of ZB NWs under laser excitation with

average Iexc=12.7 W/cm2. The LTPL spectra were fit to individual emis-

sion peaks using Lorentzians as shown. . . . . . . . . . . . . . . . . . . . 112

5.8 Excitonic region of the LTPL spectrum (4.0 K) acquired from ZB-NW-4

with (a) dslit=50 µm and (b) dslit=100 µm. The LTPL spectra were fit to

individual emission peaks (only the Id1 line is shown in each case). The Γp

of the Id1 lines are (a) 1.5 meV and (b) 1.6 meV. . . . . . . . . . . . . . . 114

5.9 Temperature dependent PL spectra from ZB-NW-3. All spectra were ac-

quired at a fixed average Iexc=31.8 W/cm2, are normalized by the integra-

tion times (ti) and displayed offset for clarity. The LTPL spectra taken at

temperatures above 115 K are not shown. . . . . . . . . . . . . . . . . . 116

5.10 Variation in PL peak energy of the FX emission (Ep=2.800 eV) (for ZB-

NW-3) with temperature. The solid and dashed lines are fits to the Equa-

tion 5.3 and 5.4, respectively. . . . . . . . . . . . . . . . . . . . . . . . . 118

5.11 Variation in PL peak energies for the (a) Id1 and (b) Id′

1 lines (for ZB-NW-3)

with temperature. The solid line are fits to the Equation 5.4. . . . . . . . 119

5.12 Change in HWHM for the FX emission (Ep=2.800 eV) with temperature

(for ZB-NW-3). The solid line is a fit to the Equation 5.5. . . . . . . . . 121

5.13 Change in HWHM for the (a) Id1 and (b) Id′

1 lines with temperature (for

ZB-NW-3). The solid lines are fits to the Equation 5.5. . . . . . . . . . . 123

5.14 Change in integrated PL intensity (IPL) for the FX emission (Ep=2.800

eV) with temperature (for ZB-NW-3). The dashed and solid lines are fits

to the Equations 5.6 and 5.7, respectively. Note the log scale for IPL. . . 124

xx

5.15 Change in integrated PL intensity (IPL) for the (a) Id1 and (b) Id′

1 lines with

temperature (for ZB-NW-3). The solid lines are fits to the Equation 5.6.

Note the log scale for IPL. . . . . . . . . . . . . . . . . . . . . . . . . . . 125

5.16 Iexc-dependent LTPL spectra from ZB-NW-1. All spectra were acquired

at a constant temperature (3.9 K), are normalized to unity and displayed

offset for clarity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

6.1 Structural characterization of WZ-NW-1: (a) Overview TEM micrograph,

and (b) TEM micrograph at a magnification higher than in (a). The circle

denotes the area from where the SAD pattern (shown in Figure 6.3 (b))

was taken. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132

6.2 (a) HR-TEM micrograph from an area of WZ-NW-1, (b) Indexed FFT

pattern taken from the area squared (in red) in (a). The indexed spots

correspond to a WZ crystal structure. The viewing direction for (a) and

(b) is 〈1100〉. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

6.3 (a) HR-TEM micrograph from another area, and (b) SAD pattern acquired

from a larger area (shown in Figure 6.1(b)) of WZ-NW-1 . The indexed

spots correspond to a WZ crystal structure. The viewing direction for (a)

and (b) is 〈1100〉. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

6.4 Excitonic region of the LTPL spectrum (3.8 K) from WZ-NW-1 acquired

under the conditions of high-throughput. (Iexc and dslit are as indicated).

The inset shows a confocal PL image of WZ-NW-1 under laser excita-

tion with an average Iexc=12.7 W/cm2. The LTPL spectrum was fit to

individual emission peaks using Lorentzians as shown. . . . . . . . . . . . 136

xxi

6.5 (a,b) Excitonic region of the LTPL spectrum (3.8 K) from WZ-NW-1

acquired under the conditions of high-resolution. (Iexc and dslit are as

indicated). The I1 line with a Γp=2.8 meV is shown in (a). The LTPL

spectrum was fit to individual emission peaks using Lorentzians as shown

in (b). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140

6.6 Excitonic region of the LTPL spectrum (4.0 K) acquired from WZ-NW-2

under the conditions of (a) high-throughput, and (b) high-resolution. (Iexc

and dslit are as indicated). The inset in (a) shows a confocal PL image

of WZ-NW-2 under laser excitation with average Iexc=12.7 W/cm2. The

LTPL spectra in both cases were fit to individual emission peaks using

Lorentzians as shown. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

6.7 Temperature dependent PL spectra from WZ-NW-1. All spectra were

acquired at a fixed average Iexc=38.2 W/cm2, are normalized by the inte-

gration times (ti) and displayed offset for clarity. The LTPL spectra taken

at temperatures above 70 K are not shown. . . . . . . . . . . . . . . . . . 142

6.8 Variation in PL peak energy of the I1 line (Ep=2.841 eV) (for WZ-NW-1)

with temperature. The solid and dashed lines are fits to the Equation 5.4

for temperatures above 30 K and 40 K, respectively. . . . . . . . . . . . . 144

6.9 Change in integrated PL intensity (IPL) for the I1 line with temperature

(for WZ-NW-1). The solid line is a fit to the Equation 5.6. Note the log

scale for IPL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145

6.10 Change in HWHM for the I1 line (Ep=2.841 eV) with temperature (for

WZ-NW-1). The solid line is a fit to the Equation 5.5. . . . . . . . . . . 146

6.11 Iexc-dependent LTPL spectra from WZ-NW-1. All spectra were acquired

at a constant temperature (3.8 K), are normalized to unity and displayed

offset for clarity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147

xxii

7.1 Structural characterization of representative ZnSe NTSLs: HR-

TEM micrographs of ZnSe (a) NTSL-1(γ=0.100), (b) NTSL-2(γ=0.077)

and (c) NTSL-3(γ=0.059) viewed along the 〈110〉 zone-axis of NTSLs

showing the periodically arranged twin-planes. (d) Normal percentile dis-

tribution of segment widths (nm) of ZnSe NSTL-3(0.059) with d=5.63±0.07

nm (s.e.m.) and ∆d=0.65 nm (s.d.). A narrow distribution indicates a

high degree of periodicity of twin-plane spacings in NTSL-3(0.059). Nor-

mality tests are described in Table 7.1. (e) Indexed spots in the power

spectrum taken from the area highlighted in (c) confirm that the indi-

vidual domains are 〈111〉-ZB ZnSe with 〈111〉-direction coinciding with

the growth direction of NTSL-3(0.059). (f) Indexed SAD pattern from an

area of ZnSe NTSL-3(0.059), double diffraction spots in the SAD pattern

further demonstrate the excellent periodicity of twin-plane spacing d. . . 153

7.2 Histograms representing the distributions of segment widths (nm) for NTSLs

1-5. The dashed curves are normal approximations to the distributions.

The corresponding twin-plane spacing d, error in d (δd) (s.e.m.) and dis-

persion in twin-plane spacing ∆d (s.d.) are listed in Table 7.1 along with

the results of the normality tests. . . . . . . . . . . . . . . . . . . . . . . 154

xxiii

7.3 Optical and structural characterization of NTSL-4(γ=0.012): (a)

LTPL spectrum (4K) from NTSL-4(0.012) with dominant free (FX) and

bound exciton (D0X, A0X) related emission peaks. Multiple order phonon

replicas of the A0X peak are also observed. (b) HR-TEM micrograph

viewed along the 〈110〉 direction of NTSL-4(0.012) showing the periodically

arranged twin-planes. (c) Spectrally and temporally-resolved PL from

ZnSe NTSL-4(0.012) showing the decay in time of individual emissions.

Streak images corresponding to TRPL from (d) NTSL-4(0.012) and (e)

single ZB NW (ZB-NW-1). TRPL decay curves extracted from streak

images are shown in Figure 7.8. . . . . . . . . . . . . . . . . . . . . . . . 156

7.4 Variation in LTPL spectra obtained from NTSL-5(0.019) as a function of

excitation intensity Iexc. All spectra were acquired at a constant temper-

ature (4 K), normalized to unity and are displayed offset for clarity. . . . 159

7.5 Change in peak energies as a function of Iexc for two individual emission

peaks observed in the LTPL from NTSL-5(0.019). Each spectrum shown

in Figure 7.4 was fit to individual emission peaks using Lorentzians, and

the peak energies thus obtained are shown with varying Iexc. The solid

lines are guide to the eye. . . . . . . . . . . . . . . . . . . . . . . . . . . 160

xxiv

7.6 Variation in PL spectra as a function of periodicity parameter γ:

(a) LTPL (4K) spectra obtained from five different NTSLs with varying

periodicity parameters γ. Periodicity parameter γ and the order of the

polytype (2/γ)H (or 2NH) are indicated in the figure. Also shown are

the experimentally obtained PL spectra from single ZB (γ=0), and single

WZ(2H) (γ=1) NWs. There is a monotonic blue-shift in LTPL spectra

with increasing γ indicating the increase in energy band gap for polytypes

of ZnSe with maximum band gap for WZ ZnSe. (b) HR-TEM micrographs

corresponding to the NTSLs(γ), ZB and WZ NWs whose LTPL spectra

are displayed in (a). The HR-TEM micrographs were taken at similar

magnification factors along 〈110〉 zone-axis. . . . . . . . . . . . . . . . . . 163

7.7 Variation in energy band gaps of ZnSe NTSLs as a function of

the periodicity parameter γ: Projected and measured band gaps (red)

for six different NTSLs with varying periodicity parameters γ, ZB and

WZ NWs are compared with band gaps obtained from band-structure

calculations (blue) for ZB and WZ ZnSe, and higher order polytypes. The

band gaps obtained from electronic structure calculations for all structures

were scissor-shifted [310] by 1.642 eV to match the ZB ZnSe band gap

with its experimental value. Electronic structures obtained from ab Initio

calculations for 2H, 4H (γ=0.5) and 6H (γ=0.333) polytypes of ZnSe are

shown in the Figure E.3 while Figure E.4 compares the calculated band-

structures for 2H and 4H polytypes. . . . . . . . . . . . . . . . . . . . . . 164

xxv

7.8 Low-temperature photoluminescence (4 K) decay from NTSL-4(0.012) (red)

and single ZB NW (ZB-NW-1 from Chapter 5) (blue) extracted from the

streak images shown in Figures 7.3 (d) and (e), respectively, with an en-

ergy window of 20 meV centred around the acceptor-bound exciton peak

(A0X) (the strongest peak). Both decay curves are characterized by a

bi-exponential decay with comparable lifetimes as shown. . . . . . . . . . 165

7.9 Temperature dependent PL spectra from NTSL-7(0.014). All spectra were

acquired at a fixed average Iexc=31.2 W/cm2, are normalized by the inte-

gration times (ti) and displayed offset for clarity. The LTPL spectra taken

at temperatures above 150 K are not shown. . . . . . . . . . . . . . . . . 167

7.10 Variation in PL peak energy of the emission corresponding to the band gap

recombination (red, for NTSL-7(0.014)) with temperature. Change in PL

peak energy for the FX emission for ZB-NW-3 (green, experimental) and

energies corresponding to the FX absorption for WZ ZnSe [238] (blue)

are also shown. The solid and dashed lines in each case are fits to the

Equations 5.3 and 5.4, respectively. . . . . . . . . . . . . . . . . . . . . . 169

7.11 Change in integrated PL intensity (IPL) for the emission corresponding

to the band gap recombination with temperature for NTSL-7(0.014). The

solid line is a fit to the Equation 5.6. Note the log scale for IPL. . . . . . 170

7.12 Change in HWHM for the band gap recombination with temperature for

NTSL-7(0.014). The solid line is a fit to the Equation 5.5. . . . . . . . . 171

xxvi

C.1 (a) Relative positions of donor and acceptor levels in the simplified band

diagram; (b) Radiative and non-radiative transitions: (A) Free exciton

recombination (FX), (B) and (C) radiative recombination of donor- and

acceptor-bound excitons (D0X,A0X), (D) donor-acceptor pair (DAP) re-

combination (D0A0), (E) radiative recombination of a free electron and

a neutral acceptor (eA0), (F) radiative recombination of a free hole and

a neutral donor (D0h), and (G) and (H) are non-radiative transitions of

free electrons and holes to ionized donors and acceptors, respectively. [(b)

adapted from Ref [251]] . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196

D.1 Components and principle of operation of a streak camera . . . . . . . . 212

E.1 First BZ of (a) ZB and (b) WZ crystal lattices [316]. . . . . . . . . . . . 216

E.2 Calculated band structures of ZnSe in (a) ZB and (b) WZ crystal struc-

tures. Reprinted with permission from Ref. [317] © 1994 American Phys-

ical Society. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217

E.3 Band-structure of ZnSe NTSLs calculated using ab Initio pseudopotential

method within the local density approximation (LDA). The energy bands

of the NH structure are closely related to that of the wurtzite structure

(2H) with bands folded at 2/N(0, 0, 2π/c) along the Γ-A direction. This

can be clearly seen by comparing the band-structures for the 2H (a) and

4H (b), for example. It is noted that there are some important deviations

such as splitting of some degeneracies that cannot be accounted for by

a simple folding. However, the bands near the direct band gap at the Γ

point are very similar in all the structures apart from a gradual shift in the

band gap with N . The horizontal line indicates the position of the valence

band maximum (VBM). Also shown is the calculated band-structure for

6H-polytype (c). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219

xxvii

E.4 Comparison of the calculated band-structures for wurtzite (2H) (red) and

4H-polytype (blue) structures. A and A1 denote the Brillouin zone bound-

ary along the growth direction for 2H and 4H structures, respectively. The

band gap for 4H-poytype is reduced compared to that of wurtzite (2H)

structure. The horizontal line (green) indicates the position of the valence

band maximum (VBM). . . . . . . . . . . . . . . . . . . . . . . . . . . . 220

G.1 Structural characterization of ZB-NW-2: (a) Overview TEM micrograph,

and (b) TEM micrograph at a magnification higher than in (a). . . . . . 224

G.2 (a) HR-TEM micrograph from an area of ZB-NW-2, and (b) Indexed FFT

pattern acquired from an area (squared in red) in (a). The viewing direc-

tion for (a) and (b) corresponds to 〈110〉. . . . . . . . . . . . . . . . . . . 224

G.3 (a) HR-TEM micrograph from an area of ZB-NW-2, and (b) Indexed SAD

pattern from a larger area (shown in Figure G.1 (b)) of ZB-NW-2. The

viewing direction for (a) and (b) corresponds to 〈110〉. . . . . . . . . . . 225

G.4 Structural characterization of WZ-NW-2: (a) Overview TEM micrograph,

and (b) TEM micrograph at a magnification higher than in (a). . . . . . 226

G.5 (a) HR-TEM micrograph from an area of WZ-NW-2, (b) Indexed FFT

pattern acquired from an area (squared in red) in (a). The viewing direc-

tion for (a) and (b) is 〈1100〉. . . . . . . . . . . . . . . . . . . . . . . . . 226

G.6 Structural characterization of NTSL-1(0.100): (a) Overview TEM micro-

graph, and (b) HR-TEM micrograph from an area of NTSL-1(0.100). The

viewing direction for (a) and (b) is 〈110〉. . . . . . . . . . . . . . . . . . . 227

G.7 HR-TEM micrograph from an area of NTSL-1(0.100). The periodic twin-

planes as atomically sharp interfaces can be seen. The viewing direction

is 〈110〉. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228

xxviii

G.8 (a) HR-TEM micrograph from an area of NTSL-1(0.100), and (b) Indexed

FFT pattern taken from the area (squared in red) in (a). The viewing

direction is 〈110〉. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229


graph, and (b) HR-TEM micrograph from an area of NTSL-2(0.077). The

viewing direction for (a) and (b) is 〈110〉. . . . . . . . . . . . . . . . . . . 229

G.10 HR-TEM micrograph from an area of NTSL-2(0.077). The viewing direc-

tion is 〈110〉. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230



direction is 〈110〉. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231


graph from an area of NTSL-3(0.059), and (b) a HR-TEM micrograph

from an area of NTSL-3(0.059). The viewing direction for (a) and (b) is

〈110〉. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232

G.13 A composite image of the HR-TEM micrographs of NTSL-3(0.059), taken

at successive locations along its length. . . . . . . . . . . . . . . . . . . . 233



direction is 〈110〉. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234


SAD pattern from a larger area (shown in Figure G.12 (a)) of NTSL-

3(0.059). The viewing direction for (a) and (b) is 〈110〉. . . . . . . . . . . 235


graph from an area of NTSL-4(0.012), and (b) a TEM micrograph from

an area of NTSL-4(0.012), at a magnification higher than in (a) . The

viewing direction for (b) is 〈110〉. . . . . . . . . . . . . . . . . . . . . . . 235

xxix



direction is 〈110〉. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236

G.18 (a) HR-TEM micrograph from another area of NTSL-4(0.012), and (b)

Indexed SAD pattern from a larger area of NTSL-4(0.012). The viewing

direction for (a) and (b) corresponds to 〈110〉. . . . . . . . . . . . . . . . 236


graph from an area of NTSL-5(0.019), and (b) a HR-TEM micrograph

from an area of NTSL-5(0.019). The viewing direction for (a) and (b) is

〈110〉. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237



direction is 〈110〉. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237

G.21 (a) HR-TEM micrograph from another area of NTSL-5(0.019), and (b)

Indexed SAD pattern from a larger area (shown in Figure G.19 (a)) of

NTSL-5(0.019). The viewing direction for (a) and (b) is 〈110〉. . . . . . . 238


graph of NTSL-6(0.333), which extends on both sides of the grid bar la-

belled as shown, and (b) TEM micrograph of NTSL-6(0.333) on the side

labelled ‘Top of the grid bar’. . . . . . . . . . . . . . . . . . . . . . . . . 238

G.23 On the side labelled ‘Below the grid bar’: (a) Overview TEM micrograph

from an area of NTSL-6(0.333), and (b) a HR-TEM micrograph from an

area of NTSL-6(0.333). The viewing direction for (a) and (b) is 〈110〉. . . 239

G.24 On the side labelled ‘Top of the grid bar’: (a) Overview TEM micrograph

from an area of NTSL-6(0.333), and (b) a HR-TEM micrograph from an

area of NTSL-6(0.333). The viewing direction for (a) and (b) is 〈110〉. . . 239

xxx

List of Acronyms

1D One-dimensional

BBO Beta Barium Borate

BEC bound exciton complex

BX Bound exciton

BZ Brillouin Zone

CCD Charge-Coupled Device

CL cathodoluminescence

COV coefficient of variation

CBM conduction band minimum

CVD chemical vapour deposition

DAP Donor-acceptor pair

DL Deep level

ED Electron diffraction

EDS Energy Dispersive X-ray Spectroscopy; see also EDXS

EDXS Energy Dispersive X-ray Spectroscopy; see also EDS

EM Electron microscope (or microscopy); see also SEM, STEM, TEM,

HR-TEM

FCC face-centered cubic

FET Field Effect Transistor

FIB Focused ion beam

FFT Fast-Fourier Transform

FX Free exciton

xxxi

FWHM Full-width at half-maximum

HPD-TA High-Performance Digital Temporal Analyser

HR-TEM High-resolution Transmission Electron Microscope (or Microscopy); see

also SEM, STEM, TEM

HWHM Half-width at half-maximum

LA longitudinal-acoustic

LED Light Emitting Diode; see also BLED

LDA local density approximation

LO longitudinal-optical

LPE liquid-phase epitaxy

LTCL low-temperature cathodoluminescence; see also RTCL

LTPL low-temperature photoluminescence; see also RTPL

MBE Molecular Beam Epitaxy

MOCVD Metal-Organic Chemical Vapor Deposition

NA Numerical aperture

NBE near band-edge

NTSL Nanowire Twinning Superlattice; see also RTNW

NW nanowire

PDF Powder diffraction file

PL photoluminescence

PMT photo-multiplier tube

ppm parts per million

ppb parts per billion

xxxii

QTH quartz-tungsten-halogen

RT room-temperature

RTCL room-temperature cathodoluminescence; see also LTCL

RTNW randomly twinned nanowire; see also NTSL

RTPL room-temperature photoluminescence; see also LTPL

SA Self-activated

SAD selected-area diffraction pattern

s.d. standard deviation

s.e.m. standard error in mean

SHG Second Harmonic Generator

SNW Single Nanowire

SEM Scanning Electron Microscope (or Microscopy); see also STEM, TEM,

HR-TEM

SNR signal-to-noise ratio

STEM Scanning and Transmission Electron Microscope (or Microscopy); see

also SEM, TEM, HR-TEM

SX Surface exciton

TA transverse-acoustic

TO transverse-optical

TEM Transmission Electron Microscope (or Microscopy); see also SEM,

STEM, HR-TEM

TRPL Time-resolved photoluminescence

TSL Twinning Superlattice; see also NTSL

VBM valence band maximum

xxxiii

VLS Vapor-Liquid-Solid

VND Variable Neutral Density

UV ultra-violet

WZ Wurtzite

XPS X-ray photoemission spectroscopy

XRD X-ray diffraction

ZB Zinc-blende

ZB-WZ Zinc-blende-wurtzite

ZPL zero-phonon line

xxxiv

List of Symbols

α a fitting parameter in Equation 5.3, representing dEg/dT at high temperatures

αB strength of the exciton (electon)-average pnohon interaction in Equation 5.4

β a fitting parameter in Equation 5.3 considered to be equivalent to θD

χ2-test a statistical test to compare the observed experimental data with that expected

from a model

∆µ standard error in mean (s.e.m.) in the distribution of individual segment widths

of NTSLs

∆d dispersion in d, (s.d.) (=σ)

δd error in d, (s.e.m.) (=∆µ)

ε dielectric constant

η efficiency of PL emission

γ Periodicity parameter used to characterize the twin-plane spacing d in NTSLs

Γ-point centre of the Brillouin zone (BZ) with k=0

Γp linewidths of luminescence peaks in terms of the full-width at half-maximum

(FWHM)

Γimp a parameter in Equation C.24 representing contribution to linewidths due to scat-

tering by fully ionized impurities

Γinh a parameter representing inhomogeneous broadening in Equation 5.5

xxxv

ΓLO a parameter in Equation 5.5 representing the strength of the exciton (electron)-LO

phonon interaction

γth a parameter in Equation 5.5 representing the strength of the exciton (electron)-

acoustic phonon interaction

~ωLO energy of the LO-phonon replica

~ reduced Planck’s constant=h/2π=6.58212 ×10−16 eV.sec

B notation for a twin-plane in NTSLs

µ mean of the distribution of individual segment widths of NTSLs

σ standard deviation of the distribution of individual segment widths of NTSLs

τ PL lifetime

τ1, τ2 PL lifetimes in bi-exponential decay processes

τAX ,τnrAX radiative and non-radiative lifetimes of acceptor-bound excitons

τDX ,τnrDX radiative and non-radiative lifetimes of donor-bound excitons

τFX ,τnrFX radiative and non-radiative lifetimes of free excitons

θB temperature corresponding to an average phonon energy in Equation 5.4

θD Debye temperature for a crystal

θLO temperature corresponding to LO-phonon energy

a lattice constant of a material in ZB crystal structure

a′, b′, ..., l′ coefficients in the Equation C.14

A− ionized acceptor

A−X excitons bound to ionized acceptors

A0 neutral acceptor

A0X excitons bound to neutral acceptors

xxxvi

aB exciton Bohr radius

aGaAs lattice constant of GaAs

avdW effective van der Waals coefficient for the interaction between a neutral donor and

a neutral acceptor

aZnSe lattice constant of ZnSe (in zinc-blende crystal structure)

AsSe As atom on the site of a Se atom in ZnSe

C,C1, C2 coefficients in Equations C.30 and C.31

CuZn substitutional Cu atom on the site of a Zn atom in ZnSe

d Twin-plane spacing (or) the distance between successive twin-planes in NTSLs

D+ ionized donor

D+X excitons bound to ionized donors

D0 neutral donor

D0X excitons bound to neutral donors

D0h donor-to-hole transition

df ilm film thickness

d<111> The interplanar spacing in the 〈111〉 direction in ZB-ZnSe

dslit Slit-width of the spectrometer

e electronic charge

E(r) term representing Coulombic interaction energy between a donor and an acceptor

e− an electron

EA ionization energy of an acceptor

Ea thermal activation energy for a non-radiative recombination process

EB Coulombic interaction energy between a donor and an acceptor separated by RB.

xxxvii

ED ionization energy of a donor

Eg electronic energy band gap

Ek is the kinetic energy of the free electron (=kBT )

Em energy of emitted photons in a DAP-related recombination

Ep energy position of a PL peak in a PL spectrum

Ex notation for the free exciton emission used in the literature

Ea1, Ea2 thermal activation energies for non-radiative recombination processes, subscripts

denote different mechanisms

EBX binding energy of the exciton bound to donors or acceptors

EDAP (r) energy position of the DAP emission, a function of the distance r between

associated donor and acceptor

EDL energy position of the deep-level emission band in the RTPL spectrum of ZnSe

nanostructures

EFB energy position of PL peak related to free-to-bound transitions

EFX binding energy of free excitons

Eg,WZ Eg of ZnSe in the WZ crystal structure

Eg,ZB Eg of ZnSe in the ZB crystal structure

Eimp average binding energy of the impurities in Equation C.24

ENBE energy position of the near band -edge peaks in the RTPL spectrum of ZnSe

nanostructures

Eph energy of photons of light

eA0 electron-to-acceptor transition

GaZn Ga atom on the site of a Zn atom in ZnSe

h Planck’s constant=4.13567 ×10−15 eV.sec

xxxviii

hνB emitted photon energy of a DAP separated by a shallow impurity Bohr radius

hνm emission band peak energy in a DAP recombination (=Em)

hν∞ photon energy corresponding to infinitely distant donor-acceptor pair in Equa-

tion 4.3

hνBE,m energy position of the m-th order phonon-replica of the bound exciton emission

hνBE energy position of the bound exciton emission

hνDAP energy position of the DAP emission

hνFX,m energy position of the m-th order phonon-replica of the free exciton emission

hνFX energy position of the free exciton emission

h+ a hole

i index used to represent NTSLs

I − VZn deep centre in I-doped ZnSe related to VZn and I

I1 notation for PL lines related to neutral acceptor bound excitons

I∗Ag1 notation for a PL line related to excitons bound to AgiAgZn acceptor complex.

see Table B.3

Id′

1 notation for a PL line related to excitons bound to (VZn-complex) deep neutral

acceptors related

Ideep1 , Id1 notation for PL lines related to excitons bound to deep neutral acceptors

I2 notation for PL lines related to neutral donor bound excitons

Ih2 notation for a PL line due to excitons bound to ionized acceptors in the wurtzite

crystal structure, see Table B.3

I∗2 alternate notation for a donor-bound exciton

Im integrated intensity of the m-th order phonon-replica

Ix notation for a PL line at 2.795 eV in ZnSe

xxxix

IY integrated PL intensity of the D0X-line

ID0X integrated PL intensity of the Y-line

Iexc,0 a constant of proportionality in Equation 4.3

Iexc excitation intensity (for excitation of PL)

IPL integrated PL intensity for a PL peak in the PL spectrum

k wave-vector related to the direction of motion of an electron in the crystal

kB Boltzmann constant=8.61733 ×10−5 eV/K

m order of the phonon-replica

m∗ effective mass of a carrier (electron or hole)

m∗e electron effective mass

m∗h hole effective mass

m∗r reduced effective mass of the free excitons

N Number of monolayers between successive twin-planes in NTSLs

NA concentration of acceptors

ND concentration of donors

NS sample size in the distribution of individual segment widths of NTSLs

NA0 concentration of neutral acceptors

nAX concentration of acceptor-bound excitons

ND0 concentration of neutral donors

nDX concentration of donor-bound excitons

nFE concentration of free excitons

nFX principal quantum number of the excited states of the free excitons

NLO Bose-Einstein statistical factor

xl

Nph average number of phonons emitted in simultaneous emission of photons and

phonons in a transition

p coefficient in the power law relation between IPL and Iexc (Equation C.20)

Pr probability of a radiative transition

Pnr,0, Pnr1,0, Pnr2,0 prefactors in Equations C.26 and C.31

Pnr1, Pnr2 probabilities of non-radiative transitions, subscripts denote different mecha-

nisms

Pnr probability of a non-radiative transition

R volume fraction of ZnSe NWs on ZnSe epilayers

r separation between donors and acceptors associated with a DAP recombination

RB shallow impurity Bohr radius

RH Bohr radius for hydrogen

RY ratio of the integrated PL intensities of the D0X and the Y-line (=ID0X/IY )

Sei Interstitial Se atom

SeZn Antisite defect with a Se atom in place of a Zn atom

T Temperature

ti integration time used in the acquisition of a PL spectrum

Vepi total volume excited by the laser beam of ZnSe epilayers

Vnw total volume excited by the laser beam of ZnSe NWs

VSe Vacancies of Se

V xSe alternate notation for neutral Se vacancies

VZn Vacancies of Zn

VZn/GaZn complex defect centre in ZnSe related to VZn and GaZn

xli

Y emission band in ZnSe related to structural defects

Zni Interstitial Zn atom

ZnSe Antisite defect with a Zn atom in place of a Se atom

ABC atomic planes in the stacking sequence in crystals, each letter denotes a bilayer

xlii

Glossary of terms

acoustic phonons: modes of lattice vibrations in crystals composed of more than one

atoms, where nearby atoms of different kinds vibrate in the same direction with a

periodically varying amplitude

anti-site defect: a type of point defect in compounds when atoms of different elements

exchange their lattice positions, e.g., a Zn atom on a site of Se atom in ZnSe (ZnSe)

binary compound semiconductors: compound semiconductors consisting of two

elements only

Brillouin zone: the wavevector (k) values for a particular energy band are said to

form a Brillouin zone

excitons: a quasi-particle of an electron-hole pair bound by Coulomb interaction

interstitials: a type of point defect where a crystal contains an extra atom at a position

which is not a regular lattice site

longitudinal (LO and LA) phonons: phonons with atomic displacements parallel

to the propagation direction. These generally have higher energy (compared to the

transverse phonons)

native point defects: point defects that involve atoms of the elements forming the

crystal, so called to distinguish from point defects that involve foreign impurities

optical phonons: modes of lattice vibrations in crystals composed of more than one

atoms, where nearby atoms of different kinds vibrate in opposite directions with

amplitudes that are inversely proportional to the atomic masses

xliii

phonon-replicas: subsidiary peaks in a PL spectrum which are a result of simultaneous

emission of a photon and a phonon. These are equally spaced with separation energy

equal to the phonon energy

phonons: collective vibrational modes of the atoms forming a crystal. These modes are

quantized where each quanta of lattice vibrations corresponds to a phonon energy

planar defects: a disruption in a homogeneous material that extends across a plane,

e.g., a stacking fault

point defect: crystallographic defects in a crystal which extend only a few interatomic

distances in any direction (or) defects localized at or around one lattice site only

in a crystal

polytypes: different modifications of a material which are chemically identical but

differ in atomic arrangement

rotation twin-plane: in a frequently occurring mode of twinning where one domain

of the crystal is a mirror image of the other part of the crystal along a certain

crystallographic plane called as twin-plane. A rotation twin-plane involves the

rotation of two domains of the crystal by 60◦ or its odd multiples

stacking faults: a type of planar defect, which is an interruption of one or more layers

in the stacking sequence of atomic planes in crystals

transverse (LA and TA) phonons: phonons with atomic displacements perpendic-

ular to the propagation direction. These generally have lower energy (compared to

the longitudinal phonons)

twinning: twinning refers to the occurrence of more than one domains in a crystal which

are oriented with respect to each other according to some symmetrical relationship

xliv

substitutional impurity: a type of point defect where atoms of foreign elements

replace an atom in the host lattice, e.g. GaZn where Ga replaces a Zn atom

vacancies: a type of point defect in which an atom is missing from a lattice site

xlv

Chapter 1

Background-Current state of the art

1.1 Introduction

1.1.1 II-VI compound semiconductors

II-VI compounds are formed from the elements of group II and VI of the periodic table.

They include sulphides, selenides, tellurides and oxides of zinc (Zn), cadmium (Cd),

mercury (Hg) and alkaline earth metals, and their alloys. II-VI semiconductors such as

CdS, CdSe, CdTe, ZnO, ZnS, ZnSe, ZnTe possess wide direct band gaps, Eg, with the

minima of the conduction band and maxima of the valence band occurring at the Γ-point

(k=0), the centre of the Brillouin zone (BZ). The band structure notation is described

in Appendix E. While there is no formal definition of ”wide band gap”, a semiconductor

may be referred to as having a wide band gap if Eg is about twice or three times that

of Si, the most commonly used semiconductor Si [1], with Eg=1.1 eV [2]. Neumark

somewhat arbitrarily specifies wide band gap as an energy gap exceeding ∼1.5 eV [3],

and this definition of wide band gap is implied in this thesis. Materials other than II-

VI compounds may also possess wide band gaps. For example, GaN (Eg,GaN= 3.39 eV

at room-temperature (RT) [2]) from the III-V family of compounds and SiC (Eg,SiC=

2.9 eV at RT [2]) with an indirect band gap are among the prominently studied wide

1

Chapter 1. Background 2

band gap semiconductors. In an indirect band gap semiconductor, band extrema occur at

different k’s in the BZ. Applications of SiC, GaN and ZnSe in devices have been reviewed

by Morkoc et al. [2], while recent progress in their application towards ultra-violet (UV)

photodetection have been reviewed by Monroy et al. [4]. Wide band gap II-VI compound

semiconductors have been a subject of intense research during the last three decades,

which can be attributed to their promising optical and electronic properties [5–7]. They

have attracted much attention as prospective UV and visible LED (light-emitting diode)

materials owing to their excellent luminescence properties. Progress towards realization

of luminescent devices based on II-VI compounds has been reviewed by Park et al. [8]

and Aven et al. [9], among others.

1.1.2 ZnSe for opto-electronic devices

ZnSe is a typical wide band gap compound semiconductor with a direct band gap of

2.7 eV at RT [6, 7] and corresponding band-edge separation (of the band extrema) in

the blue spectral region. ZnSe is considered an excellent optical material for blue lasers

and optical devices in the short wavelength range [2, 5, 6, 8, 9]. Another advantage of

using ZnSe in device applications is its complete miscibility with other group II and VI

elements which allow for alloy compounds [10, 11]. ZnSe has a zinc-blende (ZB) crystal

structure [5] which consists of two inter-penetrating face-centred cubic (FCC) lattices,

which are offset by a quarter of the unit cell body-diagonal [12]. ZnSe also occurs in

the Wurtzite (WZ) crystal structure which is meta-stable (for ZnSe) at RT [13]. For

device fabrication using ZnSe, one of the important requirements is the realization of p-n

junctions based on ZnSe, which requires successful incorporation of both n- and p- type

impurities in the semiconductor. ZnSe may easily be doped n-type, but not p-type [2,

8, 9, 14]. In the past, ZnSe faced problems of low purity and strong self-compensation,

and fabrication of p-n junctions used to be very difficult [5, 8]. With the advancements

in the growth techniques such as metal-organic chemical vapour deposition (MOCVD)


and molecular beam epitaxy (MBE), growth of high-purity ZnSe became feasible. At the

same time, continued research led to the possibility of p-type doping [5]. Applications of

ZnSe in optoelectronic devices have been discussed previously [2, 4–6, 8, 9] while Luo et

al. [15] and Kolodzieski et al. [16] have focused exclusively on developments in the area

of ZnSe-based laser-diodes. Characterization of II-VI compound semiconductors in the

past relied heavily upon the techniques of optical spectroscopy such as luminescence [14],

which in simple terms is the emission of light by a material under laser excitation. In

fact, most of the fundamental optical processes were first identified and studied in II-VI

compounds, and later recognized in other materials [14]. A brief overview of luminescence

spectroscopy applied to ZnSe single crystals and epilayers is given in Section 1.5. The

epilayers here refer to ZnSe layers grown on a crystalline substrate with a lattice constant

closely matched to that of ZnSe. The substrate for growth of ZnSe epilayers has mostly

been GaAs [2, 7].

1.2 Semiconductor nanowires

The last decade has seen an upsurge of interest in the growth, characterization and

technological applications of one-dimensional (1D) structures made of a wide variety of

materials [17]. Depending on their dimensions and the particular material, the motion

of carriers may be confined in two directions in such 1D structures while the third is

unconfined and capable of electrical conduction. In such cases, structures typically will

have nanometer scale confining dimensions and micron scale lengths, corresponding to

large aspect ratios. These 1D structures are thus called 1D-nanostructures, and are

commonly referred to as nanowires, nanoribbons, nanobelts, nanotubes etc based on

their particular geometry. 1D nanostructures of various II-VI [18] and III-V [19] binary

compound semiconductors, metallic and semiconducting elements (e.g. Ag [20], Si [21,

22]) and other inorganic compounds (e.g. SnO2 [23]) have been synthesized. An overview


of nanowire synthesis is given in Section 1.4. Interest in such 1D nanostructures follows

the trend towards miniaturization of opto-electronic devices [24], and the fact that current

lithographic patterning technology for miniaturization is reaching its limits. It is believed

that 1D-nanostructures, henceforth termed as nanowires (NWs) are ideal candidates for

exploiting the dependence of electrical, optical and mechanical properties on reduced

dimensionality [24,25]. The electrical and optical properties of semiconductor NWs may

be significantly affected by quantum confinement, whereby the energy band gap of a

semiconductor increases as NW diameter reduces [25]. Since the developments in research

on NWs has occurred concurrently for a large variety of materials, the current literature

on NWs is vast and beyond the scope of this thesis. Yang et al. [26] provided a critical

view of the progress in the past decade of research on NWs. The following are a few

important areas where NWs have shown promise for practical and commercially viable

applications: biological sensors [27–29], chemical sensors [30], NW photonics [31], nano-

electronics [17,32–34], NW photovoltaics [35,36], photodetectors [37], and thermoelectric

applications [38–40]. Due to their small size, NWs also find applications in unconventional

areas where bulk crystals and thin-film structures cannot be efficiently used, e.g., single

cell endoscopy [41] and drug delivery to target cells inside living body [42].

An important application based on the optical response of semiconductor NWs is in

single-NW (SNW) lasers. This is based on the salient characteristics of NWs such as their

single-crystalline structure, high index of refraction and smooth surfaces, which allow

their end-facets to serve as partial mirrors, forming a so-called Fabry-Perot cavity. The

NW itself acts as an optical gain medium. SNW lasers based on GaN [43] and ZnO [44]

NWs have been demonstrated, while a ZnO nanolaser based on electrical injection was

realized recently [45]. In fact, NW-based nanolasers have become an established sub-field

of study, with reviews by Mariano et al. [46], Ning [47] and Vanmaekelbergh et al. [48].


1.2.1 ZnSe nanowires

There have been considerable advances in the growth of II-VI compound semiconductor

NWs [18], with the majority of the work focussed on ZnO [44,49,50], ZnS [51], CdSe [52],

CdS [53,54] and ZnSe [55]. ZnSe NWs have been extensively studied with a large number

of works devoted to their synthesis and structural characterization. Section 1.6.2 presents

a survey of optical studies of ZnSe NWs [55–120] aimed at understanding their optical

properties. Demonstrated applications of ZnSe NWs include photo-detectors [71, 121],

humidity sensors [113], and field-effect transistors (FETs). [119] For example, Fig-

ures 1.1 (a,b) show a schematic (a), and a scanning electron microscope (SEM) image

(b), of a photodetector device based on a single ZnSe nanobelt. Figure 1.1 (c) shows

a four-terminal contact structure configured as a photodetector based on a single ZnSe

NW. See references [71,121] for more details.

c)

Figure 1.1: (a) A schematic illustration of a photodetector device based on an individual

single crystal ZnSe nanobelt whose SEM image is shown in (b). (c) SEM image of a

photodetector based on a single ZnSe NW. The image shows a four-terminal contact

structure. The scale bar is 4 µm. Reprinted with permission from (a,b) Ref. [121] ©

2009 John Wiley & Sons, and (c) Ref. [71] © 2006 American Institute of Physics.

Another unique aspect of NWs, discussed below, is that they offer the possibility for

materials engineering at the nanoscale through careful control over their crystal structure.


1.3 Semiconductor nanowire heterostructures

Significant progress has been made in the field of NWs in general, both fundamentally

and technologically. This has led to further developments, which are hitherto not pos-

sible on bulk crystals, epilayers or nanoparticles. One of the emerging areas is that of

crystal structure engineering [122,123]. This is based on a precise control over the crystal

structure, and realization of heterostructures with desired properties. A semiconductor

heterostructure means a structure composed of two or more different semiconductors with

different Eg’s, which may arise due to modulations in composition (e.g., due to differ-

ent materials or alloying), crystal structures (e.g. ZB or WZ), or strain. A periodically

modulated heterostructure is also called a superlattice. Considerable and remarkable

progress has been made in the area of crystal structure engineering in III-V compound

semiconductor NWs, a few examples of which include control over design of ZB-WZ

superlattices [124, 125], quantum dots defined by alternate crystal structures precisely

placed in NWs [126], and precisely controlled semiconductor heterostructures [127]. Fur-

ther details on these developments may be found in a recent review by Caroff et al. [122].

It may be noted that the optical and electronic properties of these crystal structure en-

gineered heterostructures are not fully understood yet, which is reasonable given that

the success in their fabrication has been recent. This progress in crystal structure engi-

neering has almost eluded II-VI semiconductor NWs where such studies have not been

reported thus far. However, II-VI compounds being miscible in each other are excellent

for compositional NW heterostructures and superlattices, and have been exploited aptly

towards the synthesis of these structures. The progress in band gap modulation based on

compositional engineering of NWs has been summarized by Pan et al. [128] and Zhuang

et al. [129]. The optical properties of compositional NW heterostructures have been stud-

ied and expectedly resemble those of bulk alloys and conventional heterostructures. An

important issue adversely affecting the optical properties of alloys has been the disorder

introduced by alloy fluctuations and alloy scattering. The crystal potential in alloys is


perturbed as a result of the microscopic and statistical variations of the alloy composi-

tion. These variations strongly affect the linewidths of luminescence lines. A detailed

discussion on this can be found in a review by Pavesi et al. [130]. A new kind of structure

unique to NWs, called a Nanowire Twinning Superlattice (NTSL) was observed earlier

[104,131], and will be discussed in more detail in the next section.

1.3.1 Nanowire Twinning Superlattices (NTSLs)

A Nanowire Twinning Superlattice (NTSL) is composed of periodically arranged rotation

twin-planes in a NW along its length. A rotation twin-plane in a ZB crystal involves

rotation of two domains of a crystal with respect to each other. A NTSL is composed

of a series of periodic rotation twin-planes, each of which separate two domains of NWs

rotated with respect to each other by 60◦ or odd multiples thereof. Figures 1.2 (a-c) show

Transmission Electron Microscope (TEM) images of NTSLs realized on InP NWs, while

(d) shows a InAs NW with periodic twin-planes. Figure 1.2 (e) shows an atomic model

of a section of a NTSL with the locations of rotation twin-planes shown by arrows. The

twin-plane spacing or twin-periodicity d, is defined as the spacing between two consecutive

twin-planes. This is shown as the distance between the planes marked by A and C in

Figure 1.2 (e). The structure of a NTSL can be considered as follows. The bonds in the

〈111〉-direction in a NW crystal are cut, and one half of the crystal is rotated by 60◦ about

the bond axis. All the bonds in the two halves of the crystal are then reconnected after the

rotation. The rotation twin-planes are coherent defect-free interfaces. This means that

there is no change in the composition at a twin-plane and no strain is induced because

all the original bonds, bond lengths and bond angles are preserved. Therefore, NTSLs

are compositionally uniform and fully strain-relaxed superlattices. Strictly speaking, a

NTSL is not a heterostructure because there is no variation in composition or crystal

structure along its length, and consequently neither of its band gap. The implications of

periodic rotation twin-planes in NTSLs can be understood as follows.


d e

B C AB C B A

C B A B C A B C B A C B A

100 nm

d

d<111>

Figure 1.2: (a,b) Overview and (c) high-resolution TEM images of InP NWs of different

diameters. Scale bars in (a) 100 nm, (b) 50 nm and (c) 5 nm. There is no fixed viewing

direction because of the different orientations of NWs. (d) TEM image of an InAs NW

viewed along the 〈110〉 direction showing the periodically arranged twin-planes. (e) An

atomic model of a section of a NTSL with the locations of twin-planes marked by arrows.

The twin-plane spacing (d) is the spacing between two consecutive twin-planes, e.g. the

distance between the planes marked by A and C. Reprinted with permission from (a,b,c)

Ref. [132], (d) Ref. [124] © 2008 Macmillan Publishers Ltd., and (e) Ref. [122] © 2011

IEEE.


In binary compound semiconductors with a ZB crystal structure, a pair of atoms

of the two elements defines one layer in the stacking sequence, also called a bilayer.

The stacking sequence (or the arrangement) of atomic planes for a material with a ZB

crystal structure is ...ABCABC..., while for a material with a wurtzite (WZ) crystal

structure, the sequence is ...ABAB..., where each letter represents a bilayer. In a

ZB material, the stacking sequence is reversed (...ABCABACBA...) at a rotation-

twin-plane (denoted by B). See Figure 1.2 (e) for an atomic model of a section of a

NTSLs where twin-planes C, A and C are marked by an arrow, and the twin-plane

spacing (d) is shown as the distance between planes C and A. Many authors believe

that the ABA sequence around the twin-plane in NTSLs constitutes one monolayer of

the WZ material [132, 133]. They consider NTSLs as ZB-WZ superlattices consisting of

alternating regions of the ZB and WZ material. Caroff et al. [122,124] on the other hand

have stated that two consecutive twin-planes (i.e. ...ABCABABC...) are required to

define a segment of the WZ structure (ABAB), with a minimum of four bilayers. They

maintained that a single twin-plane does not define a layer of the WZ material, and

postulated that the NTSLs with periodic twin-planes can be considered as polytypes

of the parent material. A polytype refers to the different forms of a material which

are identical in chemical composition but differ in their crystalline arrangements, and

ZB-WZ polytypism is probably the most common example [134]. This means that for

materials which exist in ZB and WZ crystal structures, their NTSLs would have an

intermediate character in terms of their physical properties. Various terms have been

used in the literature to refer to NTSLs including periodically twinned NWs [131, 135],

twinning modulations [77], periodically faceted NWs [136], twin-plane superlattices [124]

and twinning superlattices [132], but they all refer to the same class of structures. The

use of the term ‘NTSL’ is specific to this thesis.

Twinning Superlattices (TSLs) were predicted to exist [137] and possess unusual elec-

tronic and optical properties, including a direct Eg in indirect Eg semiconductors such


as Ge and GaP [137,138]. NTSLs possess long-range order in one dimension (the growth

direction) with periodic atomically sharp interfaces. The issues of alloying and strain

encountered in conventional heterostructured superlattices which adversely affect their

optical properties are avoided in NTSLs. They are also expected to lead to interesting

phenomena including miniband formation [122, 138] with clear advantages such as be-

ing chemically simpler (composed of one compound only) and suitable for cost-effective

fabrication. NTSLs are much more versatile than polytypic (WZ-ZB) and compositional

superlattices, as a wider variety of materials exhibit twinning than those that show poly-

typism or lattice compatibility. Further, synthesis of NTSLs can also be achieved in an

inexpensive natural self-organized way. NTSLs are interesting for thermoelectric appli-

cations as well since the side-facets resulting from periodic twinning (see Figure 1.2 (d),

for example) lead to phonon back-scattering, as shown by recent experiments [139] on

randomly twinned InAs NWs, and simulations [136] for Si NWs. NTSLs are therefore

promising new structures for materials engineering at the nanoscale. NTSLs are unar-

guably among the least studied structures (among NWs) for their fundamental optical

and electronic properties. NTSLs are revisited in Section 2.2 in Chapter 2 where this

lack of understanding of their fundamental properties is discussed in detail.

1.4 Growth of semiconductor nanowires

The growth of 1D structures has been observed for more than 50 years [140], and these

were referred to as whiskers in the past as opposed to the currently used term NWs.

Later, Westwater et al. [141] and Duan et al. [142] renewed the interest in NWs with

demonstrations of the growth of semiconductor NWs using vapour-liquid-solid (VLS)

growth mechanism. A wide variety of growth techniques have been used in the synthesis

of NWs. These techniques have met with varying degrees of success in control over the

growth of NWs. A general overview of these strategies has been given by Dresselhaus


et al. [24], Xia et al. [143], Law et al. [144], and Barth et al. [145]. Some of the reviews

cited in the previous sections also briefly discuss the growth of NWs. The particular

method which relies on metal-catalyst assisted growth of NWs following a VLS growth

mechanism, tends to be the most favoured growth approach. Fortuna et al. [146] have

discussed recent progress in the growth of semiconductor NWs via the metal-catalyzed

VLS growth process. The principles of the VLS growth mechanism, being relevant to

this thesis, are briefly summarized below.

Wagner and Ellis [147] are credited for proposing the growth mechanism known as the

Vapour-Liquid-Solid (VLS) growth. Methods for growth of NWs that employ the VLS

mechanism are the most popular because they yield large quantities of NWs with single

crystal structures. In a typical VLS growth method, a substrate (e.g., crystalline Si, after

Au#

Zn#&#Se#Source#

Figure 1.3: Schematic illustrating the principles of growth of NWs based on a Vapour-

Liquid-Solid (VLS) growth mechanism. See text for details.

appropriate cleaning procedures) is used with a metal film (typically Au) deposited on it

which serves as the catalyst material. At elevated temperatures, the catalyst film breaks

into liquid droplets. The size of these droplets determine the diameter of the NWs. A

vapour of the appropriate source material is introduced into the growth apparatus and

supplies the constituents (or reactants) for growth of NWs. These constituent elements

dissolve in the liquid catalyst droplets and upon their supersaturation, nucleation and

growth of single crystal NWs occurs. This process is shown schematically in Figure 1.3.


The term VLS describes reactants (vapour), catalyst droplet alloyed with constituent

elements (liquid) and growing crystalline NW (solid). In a typical scenario, the presence

of catalyst particles (also called a ‘tip’) at the end of NWs is considered as evidence of

the VLS growth process [59, 148]. The simplicity of this method however is misleading

in the sense that careful control over growth conditions is required in order to obtain a

high yield of NWs with single crystal structure.

1.4.1 Growth and structural characterization of ZnSe nanowires

Growth of ZnSe NWs has been achieved through a variety of growth methods. These in-

clude chemical vapour deposition (CVD) [60,68,70], MBE [67,105,110], MOCVD [57,59,

64, 65], laser-ablation [61], solvo-thermal methods [62, 63,80], solution-based growth [69,

79,83], and electrochemical methods [86]. The most common route to prepare ZnSe NWs

has been the metal-catalyst assisted VLS growth mechanism [146,149]. In metal-catalyst

assisted growth of ZnSe NWs mainly Au was used as the catalyst [59,61,65], while other

metals have also been used in some cases - e.g., Ag [57, 106, 120], Fe [105], Sn [68] and

Ga [120]. Likewise, Si ((100) [57,59,65] or (111) [72,78]) has been used as the substrate

for growth in most studies while other substrates such as GaAs (100) [64, 66, 105], GaP

(111) [150], alumina [104] and sapphire [91] have also been used. Most reports on ZnSe

1-D and quasi-1D nanostructures describe their growth in the NW (cylindrical) form.

Other structural morphologies (e.g., a flat belt-like shape referred to as nanoribbons) of

ZnSe nanostructures have also been reported. Structural morphology refers to the shape

of the nanostructures irrespective of their crystal structure. These include nanobelts and

nanoribbons [57, 61, 64, 73], nanorings [70], nanowheels [76], nanoflowers [89], tetrapod

branched ZnSe nanorods [68], nanodonuts [151] and nanopeas [152] with their names

suggestive of their morphologies. The most common techniques used for the structural

characterization of ZnSe NWs have been X-ray diffraction (XRD) [57–59], SEM [57–59],

TEM (including high-resolution TEM (HR-TEM) and selected area electron diffraction


(SAD)) [56–59]. Compositional analysis has been mainly carried out using energy disper-

sive X-ray spectroscopy (EDXS or EDS) [56–59], with X-ray photoemission spectroscopy

(XPS) [63] also being used in certain cases, especially for doped ZnSe NWs [80]. These

references are only a few examples of the use of these structural characterization tech-

niques, while one or more of them have been used in almost all reports. ZnSe NWs

with single crystal structure may be termed as homogeneous NWs, indicating uniform

microstructure throughout their length.

1.5 Luminescence characterization of ZnSe single crys-

tals and thin films

Luminescence spectroscopy has been the primary method of optical characterization of

most II-VI compound semiconductors in the past, including ZnSe. The reasons for

this will become apparent in Section 3.2. The principles of photoluminescence (PL)

spectroscopy are described in Appendix C.1, along with the related concepts of opti-

cal transitions which invariably involve excitons, donor-acceptor pairs (DAP), phonons

(longitudinal optical (LO) and transverse acoustic (TA)), and phonon-replicas. In brief,

PL involves the excitation of a semiconductor by light with photon energy equal to or

greater than the Eg of the semiconductor. The electron-hole pairs are created as a result

of this excitation, and recombine radiatively via different processes, e.g., excitonic and

DAP-related. The energy of the emitted photons depends on the recombination process

involved. The resulting PL spectrum provides a finger-print of the impurities in the semi-

conductor and is used for their identification. PL studies on ZnSe crystals date back to

the work performed by Bube et al. [153]. The first detailed study of low-temperature pho-

toluminescence (LTPL) from ZnSe single crystals was reported by Reynolds et al. [154],

and thereafter Dean and Merz [155] reported the first observation of donor-acceptor pairs

(DAP) in PL at 4.2 K from ZnSe thin films grown by vapour-phase epitaxy. Since then


many researchers have studied optical emission from ZnSe crystals and epilayers in order

to establish the growth conditions required for the growth of optical quality ZnSe crystals

and films [7,156–160]. Along with the impurity-related PL of ZnSe crystals, excitonic PL

has also been an important research topic, especially in high-purity ZnSe crystals and

epilayers. Most of the available literature concerning ZnSe epilayers is about ZnSe layers

grown on GaAs substrates. Optical characterization studies of ZnSe single crystals and

epilayers have been reviewed on several occasions, including reference [7]. When phonons

as well as the photons are emitted in a PL transition, subsidiary peaks appear in the

PL spectrum known as phonon-replicas. The energy of these peaks is lower than that of

the main transition peak (zero-phonon line (ZPL)) by an amount equal to the phonon

energy.

An obvious requirement for device fabrication using ZnSe is the growth of ZnSe ma-

terial with high-purity, without any unintentional intrinsic (native point defects) and

extrinsic (residual impurities) defects. The native point defects (sometimes also referred

to as native defects or point defects) include vacancies, interstitials and antisites. In the

case of ZnSe, the native point defects consist of Zn-vacancies (VZn), Se-vacancies (VSe),

Zn-interstitials (Zni), Se-interstitials (Sei), Zn and Se based antisites (ZnSe and SeZn),

and complexes of these point defects. A defect complex consists of at least two single

point defects situated at neighbouring lattice sites. These single defects are bound to-

gether by a large binding energy, and therefore are difficult to separate. The two bound

defects then act as a single defect. Residual impurities refer to the diffusion of trace

amounts of other elements present in the source material into the crystals or epilayers

being grown. Various terms have been used in the literature to refer to them includ-

ing residual impurities, unintentional impurities, unintentional doping and background

impurities. The term unintentional is used here to distinguish from intentional doping

whereby donors and acceptors are intentionally introduced using suitable elements. For

intentional doping of ZnSe, suitable acceptor elements are the group-I elements such as


Li, Na or Cu on Zn site, or group-V elements such as N, P, or As on Se site. A suitable

choice for donor elements are group-III elements, such as Al, Ga or In on Zn site, or

group-VII elements such as F, Cl or I on Se site. It has also been reported that the de-

fect VSe can act as a donor, while VZn may serve as an acceptor [7]. The intrinsic (native

point defects) and extrinsic (unintentional or intentional) impurities which contribute to

the optical emission spectra are also collectively referred to as optically active centres,

luminescent centres, or simply as centres in the literature. Most commonly observed

features in a typical PL spectrum from ZnSe are summarized below.

1.5.1 Characteristic features of ZnSe LTPL spectrum

Figure 1.4 shows the typical near band-edge luminescence at 1.6 K of a lightly Ga-doped

ZnSe layer grown epitaxially on a GaAs (100) substrate (ZnSe/GaAs). (Near band-edge

means at energies close to and slightly lower than the band gap energy). The spectrum

contains several peaks as shown, and these are commonly observed in LTPL spectra from

ZnSe. Such LTPL spectra are also sometimes referred to as complex PL spectra due to

contributions from a large number of radiative centres [5]. Free and bound exciton peaks

(or lines) are observed near 2.8 eV. A typical LTPL spectrum also contains recombination

lines due to the neutral donor-bound excitons (D0X), usually labelled as I2 lines, and

due to the neutral acceptor-bound excitons (A0X) generally labelled as I1 lines. The

origin of this nomenclature for PL lines is attributed to Thomas and Hopfield [161, 162]

who used them first. It needs to be noted that there have been variations in the use

of these labels. For example, labels I2 and I1 may have symbols of elements or defects

as superscripts or subscripts in which case they mean that the line is related to that

particular impurity (element or defect). Further, these labels may also involve other

superscripts or subscripts whose usage is not established in literature but is rather a

choice of authors, for example Ih2 and I∗Ag1 [106]. Their meaning, however, is usually clear

from the context. The quantity of interest is always the peak position of the particular


Figure 1.4: Near band-edge luminescence at 1.6 K from Ga-doped ZnSe layer grown

epitaxially on GaAs (100) substrate. General features of the spectrum are explained in

the text. Reprinted with permission from Ref. [7] © 1990 John Wiley & Sons.

peak. The binding energies of common (D0X) and (A0X) usually lie between 2 and 20

meV, and majority of bound-exciton lines can be found in a small energy range below

the free-exciton line. At the lower energy side of the spectrum, lines due to different

donor-acceptor pair (DAP) emissions can be found, which are usually accompanied by

their LO-phonon replicas (i.e., DA-LO, DA-2LO shown in Figure 1.4). Different DAP

emissions in ZnSe are also sometimes labelled Q, P and R, although the use of these

labels is not prevalent anymore.

Bound-exciton and DAP luminescence lines occur due to recombination in the pres-

ence of shallow impurities, and in most cases can be related to a certain point defect

or dopant in the material. Shallow impurities refer to impurities that require an energy

of the order of the thermal energy to cause their ionization. The emission lines in the

PL spectrum which are observed at even lower energies are generally called deep-level

(DL) emission(s). DL emissions are mostly related to the non-optimum growth or doping


conditions which give rise to undesirable defects or impurity-defect levels. DL emissions

in ZnSe can be further distinguished as:

• Y-line emission at 2.605 eV, associated with structural and extended defects [163];

• Lines at 2.5 eV, referred to as S or M lines in the literature. These are often

accompanied by LO-phonon replicas and are correlated with a kind of DAP that

involves shallow donors and distant deep acceptors [163];

• A broad-band emission centred at about 2.3 or 2.4 eV, sometimes also referred to

as “Cu-green” emission. This is normally attributed to unintentional incorporation

of Cu during the growth, and this line is usually accompanied by a specific Ideep1

(or Id1 ) emission line at around 2.78 eV. Ideep1 (or Id1 ) line is considered to be due

to a Cu-acceptor-exciton complex. Ideep1 is also sometimes related to an exciton at

a VZn acceptor centre [164].

• Self-activated (SA) luminescence between 2.14 eV and 1.97 eV, which is related to

VZn vacancies [165] or a VZn/GaZn complex defect centre [7].

1.5.2 Excitonic emission in ZnSe

One of the ways to assess the quality of a semiconductor single crystal is to consider its PL

response. When the PL spectrum is dominated by strong near band-edge features such

as strong and narrow excitonic peaks due to bound excitons and donor-acceptor pairs,

and the intensity of DL emissions is greatly suppressed, this is taken as an indication of

high-purity growth, high crystalline quality, and successful doping (if attempted) without

undesired defects [163, 166, 167]. There have been a large number of reports on PL of

undoped ZnSe crystals and epilayers grown by a variety of methods. The bound exciton

lines observed in the PL spectra of these ZnSe materials have been assigned to different

centres by different authors, and the origin of some of these lines is still questionable.


Doping of ZnSe by different dopant elements has been studied; e.g. Ga [168,169], Cl [170],

Al [171], I [172], and In [173]. By increasing the concentration of a dopant, the donor-

bound exciton line specific to that donor is expected to be enhanced, and the line can be

assigned to that particular dopant. But this assignment is based on the assumption that

the dopant material acts as a donor-like simple substitutional impurity atom. Often,

a more complex donor centre or even an acceptor centre is formed due to the donor

material. This can be observed when the donor-bound exciton line is enhanced, and

other emission lines (especially DL emissions) are affected as well. For example, there

has been a report on rising SA emission for increasing I doping which was explained

by the formation of deep (I − VZn) centres [172]. The binding energies determined for

different donor-bound excitons, and the donor ionization energies in ZnSe films are listed

in Table A.1 [7] in Appendix B.

ZnSe has been doped with N [174], Li [173], P [175] and Na [173] in order to obtain

p-type doping. The most successful method of p-type doping of ZnSe has been the

doping by nitrogen, in particular ZnSe doped by a flux of atomic nitrogen. The difficulty

in obtaining p-type ZnSe is quite often attributed to the self-compensation of acceptor

impurities. A possible explanation of this phenomenon put forward is that Se vacancies

(VSe), which act as donors, are activated as a consequence of the doping process, and

result in a compensation of the acceptor impurities [7]. The binding energies for different

acceptor-bound excitons and the ionization energies for different acceptors in ZnSe films

are listed in Table A.2 [7] in Appendix B. Acceptors introduced in ZnSe can also form

defect complexes, similar to the case of donors and defect complexes of donors. Despite

a large number of reports on doping studies of ZnSe, the values of donor and acceptor

ionization energies and bound-exciton binding energies vary appreciably as shown in

Tables A.1 and A.2, even for ZnSe films grown by the same method. The identification

of donor and acceptor bound exciton lines can be considered to have a wide range of

reports available, but the reports on defect complexes formed by these impurities with


native defects are too few to offer substantial and reliable information on them.

1.6 Optical response of ZnSe nanowires

1.6.1 Distinction from ZnSe single crystals and thin films

The natural questions that arise are, in what ways are the optical properties of ZnSe NWs

(i) similar to and (ii) different from, those of bulk ZnSe and ZnSe epilayers. Dresselhaus et

al. [24] have noted that NWs with diameters larger than a few nm have crystal structure

similar to that of the bulk parent material. This is because the quantum confinement

effects which arise due to confinement of carriers in two-dimensions in NWs do not appear

until the confinement in these dimensions is of the order of the exciton Bohr radius

(aB). aB is the approximate distance between an electron and a hole in an exciton. For

NWs with diameters considerably larger than aB (e.g. d & 10aB) have essentially the

same electronic band-structure as bulk material. A brief introduction to electronic band-

structures is given in Appendix E. Optical transitions in a semiconductor are intimately

connected with their band-structure. The nature of optical transitions in NWs of large

diameters will therefore be similar to those in bulk single crystals. This means that the

peak positions observed in the luminescence spectra of NWs can be directly compared

with corresponding peaks in spectra from their bulk form. It is for this reason that the

PL spectra from NWs of a wide range of materials, particularly those of wide band gap

semiconductors [55,103–109,111–114,119,120], have been explained by comparing them

with the bulk luminescence spectra. The concept of surface excitons (SX) has also been

used to explain PL results in certain cases (e.g. ZnO [176], GaN [177]). Surface excitons

refer to excitons localized at the surface of NWs. In case of ZnSe NWs, there have been

no peaks reported thus far which cannot be explained by assuming bulk-like behaviour.


Distinguishing features

However, there are important differences in optical response of NWs. distinction. The

nature of recombination centres in NWs is quite different due to differences in how these

structures are prepared. In the case of ZnSe, bulk single crystals are grown using high-

temperature melt-grown methods. Even the highest purity source materials used for

their growth contains trace elements (impurities) of varying concentrations (typically

∼1-100 ppm, ppm=parts per million)which are incorporated into the bulk ZnSe crystals

during growth, and are detectable by PL spectroscopy due to its high sensitivity. Crystal

growth of ZnSe from the melt was reviewed by Rudolph et al. [178]. This issue of

residual impurity incorporation is far less significant for low-temperature vapour phase

growth of ZnSe NWs. The issues for growth of ZnSe epilayers originate with the lack

of a suitable substrate. GaAs has almost exclusively been used as a substrate to grow

ZnSe epilayers (e.g., by MBE and MOCVD) leading to inter-diffusion of impurities from

the substrate into the films. The key impurity in this case is Ga, a donor in ZnSe [7],

which appears as one of the dominant features in the PL spectra of ZnSe epilayers [7].

Another important issue with ZnSe epilayers on GaAs substrates relates to the biaxial

strain which arises from the lattice mismatch between ZnSe and GaAs (aZnSe=5.6693

A [179] and aGaAs=5.6533 A [179] giving a misfit strain ∼0.25% [178], ∼0.27 % [7]).

Further, the different thermal expansion coefficients for ZnSe (7.0 × 10−6 K−1 [178])

and GaAs (5.8 × 10−6 K−1 [178]) also contribute to this strain. This strain leads to a

splitting of the valence band into heavy-hole and light-hole sub-bands, and PL spectra

acquired from ZnSe epilayers contain PL peaks from these two bands. The top of the

valence band in ZB-ZnSe is doubly-degenerate which means the two bands have the same

energy. This degeneracy is lifted by strain and the band is split into heavy-hole and light-

hole bands. These issues are discussed by Gutowski et al. in their review [7] on optical

properties of ZnSe epilayers and films. ZnSe NWs are different from ZnSe epilayers in

that their optical properties are not affected by the issues of Ga incorporation or strain.


An important question that needs to be answered for ZnSe NWs is then, in the absence

of intentional doping and strain, which optical processes dictate their optical properties?

The above discussion was presented to clarify the need for optical characterization of

ZnSe NWs in order to understand the fundamental processes that govern their optical

properties. This information is crucial in order to be able to fabricate optical devices

based on single ZnSe NWs. It is clear that the above arguments are specific to ZnSe only,

and do not necessarily apply to NWs based on other materials.

Confinement effects in nanowires

Another important aspect of semiconductor NWs relates to the effects of quantum con-

finement and dielectric confinement. In a semiconductor NW, the photoexcited carriers

may be two-dimensionally confined, and free to move along the length of the wire. This

results in a overlap between the electron and hole wavefunctions, thereby increasing the

binding energy of the excitons. Hence, the exciton binding energy can be tuned by chang-

ing the size of the NW, but this tuning can only provide exciton binding energies a few

times their bulk value. Another way to realise the same effect is the dielectric confine-

ment effect [180, 181]. The exciton binding energy may be enhanced when the electric

field lines from a hole to an electron, pass through the insulator when a semiconductor

NW is surrounded by an insulator with smaller dielectric constant. Anisotropy in optical

absorption and emission for semiconductor NWs embedded in dielectric have been dis-

cussed recently [182]. The exciton Bohr radius of bulk ZnSe is aB = 3.30 nm [183]. ZnSe

NWs of diameter < aB would be expected to show quantum confinement effects, and this

has been shown in narrow solvent-grown ZnSe NWs [69]. The fabrication of NWs with

size < aB is challenging using the vapour-phase growth technique, but there is a possibil-

ity to realise dielectric confinement effects, as has been shown for ZnS nanorods as large

as 100 nm in diameter, embedded in polycarbonate membrane pores [184]. Since the ex-

citon binding energy of ZnSe is only 21 meV, dielectric enhancement of exciton binding


energy holds the potential and promise for the room-temperature excitonic devices.

1.6.2 Luminescence characterization of ZnSe nanowires

This section summarizes the recent progress in optical characterization of ZnSe NWs.

PL (along with cathodoluminescence (CL)) has been routinely used for the optical char-

acterization of ZnSe NWs. CL is a technique similar to PL where an electron beam is

used for the excitation of luminescence, and is almost exclusively carried out inside an

electron microscope (EM). The discussion on optical characterization of ZnSe NWs can

be divided into three broad categories and corresponding sub-categories as follows:

• Based on characterization method(s): room-temperature PL (or CL) and low-

temperature PL (or CL) characterization

• Based on crystal structure: ZB, WZ, a mixture of NWs of these two crystal struc-

tures, and NWs with a mixture of two crystal structures (bicrystalline).

• Based on morphology: NWs, nanobelts and nanoribbons, and other morphologies

Table B.1 in Appendix B provides a summary of the reported studies on the optical

characterization of ZnSe NWs according within the above categories. The crystal struc-

tures listed in Table B.1 are as reported by the authors. In cases where crystal structure

was not reported, it was determined from other works by the same authors, and in cases

where other studies were not available, crystal structure was taken to be ZB because

ZnSe predominantly crystallizes in the ZB form [5]. Below, room-temperature (RT) opti-

cal characterization is discussed first, followed by a brief description of low-temperature

optical studies on ZnSe NWs. Discussion on the variations in crystal structure and mor-

phologies is embedded within these sections for brevity and clarity. The discussion is

aimed to provide an overview of the studies while no comments are made on the dis-

crepancies and shortcomings in these results, which appear in Chapter 2. ZB crystal

structure and cylindrical morphology (NW) are implied, unless noted otherwise.


RTPL spectroscopy is unarguably the most common technique that has been em-

ployed for the optical characterization of ZnSe NWs. Room-temperature cathodolumi-

nescence (RTCL) has also been used [88,90,93] but to a much lesser extent than RTPL.

In most cases, and in particular for studies whose main focus was to describe the growth

of ZnSe NWs, RTPL has been used as a complementary characterization method. RTPL

was used in order to judge the crystalline quality of ZnSe NWs that were synthesized.

A typical RTPL (or RTCL) spectrum from ZnSe NWs consists of two main regions: (i)

one narrow peak (∼53 meV) [57, 66] around 463 nm (2.68 eV), and (ii) a broad band

(∼650 meV) [72, 77, 78, 81] around 620 nm (2.0 eV). These are shown in Figure 1.5 (a).

The narrow peak is related to the near band-edge (NBE) luminescence, while the broad

band is related to DL emission. Authors who observed the NBE peak have related it to

the excellent crystalline quality of their NWs, in most cases regardless of the existence

of DL emission band or the ratio of NBE peak to DL emission band [63,93].

RTPL has invariably been performed on an array (collection) of NWs, while RTCL

being inside an EM has been performed on single NWs as well as their array. Several

authors have observed the growth of NWs and nanoribbons together in their samples [57,

61,64], and both have been studied collectively in RTPL experiments on arrays. Similarly,

authors have also observed growth of ZnSe NWs in ZB and WZ crystal structures in the

same growth samples [60, 67, 73, 79], and they have been studied together using RTPL.

Widely varying values have been reported for the NBE peak and the DL emission band

by different authors, and different explanations have been used to justify their energy

positions. Table B.2 lists the values reported for the energy positions of the NBE peak

and the DL emission band which are divided according to the range of their values.

Table B.3 in Appendix B is a comprehensive summary of all LTPL and low-temperature

CL (LTCL) studies reported so far on ZnSe NWs. This summary includes LTPL and

LTCL peak positions, their respective assignments and labels as reported by different

authors. Figure 1.5 (b) shows a LTPL spectrum (10 K) from Ag-doped ZnSe NWs. As


(a) (b)

Figure 1.5: (a) RTCL spectrum from an individual ZnSe NW showing two peaks related

to NBE and DL emission at 2.68 eV and 1.96 eV, respectively. (b) LTPL spectrum (10 K)

from an array of Ag-doped ZnSe NWs. Inset shows the near band-edge region enlarged.

See text for details. Reprinted with permission from (a) Ref. [102] © 2011 Elsevier, and

(b) Ref. [106] © 2011 American Institute of Physics.

was noted earlier LTPL spectra from ZnSe are usually complex due to a large number

of peaks. The LTPL spectrum shown in Figure 1.5 (b) [106] is likely the most complex

PL spectra reported so far on ZnSe NWs. The LTPL peaks were related to the bound

excitons in WZ-ZnSe and to bound excitons related to Ag impurities in ZB-ZnSe. Peaks

due to phonon-replicas and DAP related recombination line due to Ag were also noted.

The LTPL peak positions and their details are listed in Table B.3. Similar to the case of

RTPL, most LTPL studies have been performed on an array of NWs, and LTCL on sin-

gle NWs and their arrays. It has been argued that metal-catalyst diffuses in ZnSe NWs,

and this diffusion influences the optical properties of ZnSe NWs. For example, Xiang et

al. [55] and Jin et al. [76] have related their LTPL spectra from ZnSe NWs to diffusion

of Au, while Liang et al. [120] using Ag and Ga catalysts have claimed that Ag and Ga

act as dopants in ZnSe NWs and that the LTPL peaks from their ZnSe NWs arise due

to Ag and Ga. But these claims are based on little evidence, and not supported by other


studies where Au or other metal-catalysts were used for the growth of ZnSe NWs. It may

also be noted from Table B.3 that there have been no LTPL studies on role of native

point defects in the optical properties of ZnSe NWs. Incidentally, Zhang et al. [103] and

Liu et al. [108] studied ZnSe NWs grown on GaAs substrates, and related LTPL lines

to substitutional point defects (GaZn and AsSe). Similar to the case of RTPL charac-

terization, LTPL characterization has also been performed on mixtures of ZnSe NWs

and nanoribbons [106], mixed NWs of ZB and WZ crystal structures [106, 107, 117] and

bi-crystalline NWs [106, 110, 111, 115, 116]. However, in case of RTPL discrepancies in-

troduced by studying such mixtures are not that significant because RTPL spectra are

much broader and different contributions (e.g. from NWs and nanoribbons) cannot be

easily distinguished. These differences are much more pronounced at low-temperatures,

and cannot be resolved by studying mixtures of NWs. These details are discussed in

Chapter 2.

1.6.3 Influence of heat treatment

One of the methods used to study the role of native point defects on the optical prop-

erties of ZnSe involves modulating the stoichiometry, and studying the resulting optical

response from ZnSe. Stoichiometry refers to the occupancy of Zn and Se lattice sites

in the crystalline material. These procedures may involve, for example: (i) growth of

ZnSe (single crystals, films or NWs) with an extra source of one of the constituent el-

ements (Zn or Se), (ii) heating ZnSe in contact with molten Zn, a process known as

Zn-extraction [185], (iii) post-growth treatment such as annealing in vapours of one of

the constituent elements. These methods are collectively referred to as heat treatment,

because they involve elevated temperatures. There have been studies on the effects of

native defects on optical emission properties, but by far they have been fewer compared

to the extensive work on foreign impurities [7]. Even though a fair number of reports

are available on the effects of native point defects on optical properties of ZnSe, they


are not devoid of controversy. The exact nature of intrinsic defects, and the role they

play in complex low-temperature optical spectra of ZnSe are not completely understood

yet. This can be justifiably attributed to the overlap of PL lines due to some foreign

impurities and the native defects, which makes the unambiguous identification of lines

difficult. The optical properties and quality of ZnSe NWs, similar to those of single

crystals, can be strongly affected by the presence of native defects, arising from the sto-

ichiometric deviations even in high-purity growth conditions. The presence of native

defects adversely affects the optical and electrical properties of ZnSe. For example, na-

tive defects and self-compensation have been implicated in the difficulty in p-doping and

control of conductivity in ZnSe single crystals. It has been established in the literature

that Zni and VSe act as donors [7, 174, 186–188], while Sei and VZn act as acceptors in

ZnSe [7, 174,187,189,190].

(a) (b)

Figure 1.6: RTPL spectrum from ZnSe NWs: (a) grown under Zn-rich conditions showing

stronger NBE emission compared to the weaker DL emission, and (b) grown under Se-

rich conditions showing strong DL emission while NBE emission is absent. Insets show

dominating (a) blue and (b) red luminescence from single ZnSe NWs. Reprinted with

permission from Ref. [72] © 2006 American Institute of Physics.

Philipose et al. [72] have earlier reported on the RTPL experiments on NWs grown

under stoichiometric and non-stoichiometric growth conditions. This growth method


is similar to the one used for the growth of ZnSe NWs studied in this thesis. They

showed that the optical emission characteristics of these NWs can be tuned to obtain

emission in the desired near band-edge (NBE) region or low-energy self-activated (SA)

emission region. This is shown in Figure 1.6(a) and (b), where in (a) the emission

corresponds to the NBE (blue) emission, while in (b) the emission is related to the SA

(red) emission. In particular, they showed using post-growth annealing treatments in Zn

and Se atmospheres, that the near band-edge luminescence can be observed and enhanced

with Zn-treatment, and that this luminescence can be quenched with Se-treatment. They

also showed that this effect is reversible, i.e., the luminescence in either case can be almost

fully recovered following a (Zn-Se-Zn-...) annealing cycle. The reversibility effect that

they observed (in their RTPL measurements) clearly shows that the origin of the near

band-edge and SA-related emissions can be attributed to native defects in the crystal

lattice, and that the concentration of residual impurities in these NWs is very low. This

reversibility effect observed by Philipose et al. [72] on ZnSe NWs is consistent with the

earlier work performed by Roppischer et al. [191, 192] on intrinsic defects in ZB, n-type

ZnSe bulk crystals using LTPL. They observed that the edge emission spectra of ZnSe

undergoes significant change when the crystals were heat treated in Zn or Se atmospheres,

and that this effect was reversible over a few annealing cycles. They assigned a position

0.19 eV above the valance band to the acceptor centre, and associated this to vacancy

of zinc VZn, and a position of 10 meV below the conduction band to the donor centre,

vacancy of Se (VSe) responsible for these effects. In another report by Sekoguchi et

al. [193], they observed a change in intensity of PL lines (Ix = 2.795eV and I2 = 2.797eV )

from hetero-epitaxially grown MOCVD ZnSe layers onto GaAs and Ge substrates after

heat treatment, although they did not assign the centres responsible for the emission.

They also observed appearance of a new line Id1 = 2.780eV after the heat treatment, which

can be associated to the VZn acceptors. The process of using thermal annealing in order to

increase the intensity of luminescence, and in particular enhance blue luminescence, has


been used in the past [186–188,190,194,195], and it has been known that the concentration

of zinc vacancies responsible for the self-activated luminescence can be controlled by

treatment in gaseous or liquid Zn [193–195]. Beyond the work performed by Philipose et

al. [72] using RTPL, there have been no attempts to understand the influence of native

point defects on optical response of ZnSe NWs.

1.7 Summary

An overview of the current state of the art in the field of optical characterization of ZnSe

NWs was given in this chapter. First, a general introduction to II-VI compounds and

ZnSe was presented. This was followed by a summary of recent developments in the

general area of semiconductor NWs, ZnSe NWs and their importance in practical appli-

cations. NTSLs were introduced as an emerging new class of nanomaterials promising

for future nanoscale optoelectronic and photonic devices, and their unique structure was

explained. Thereafter, a general overview of the growth of NWs with emphasis on VLS

growth mechanism was discussed. Optical characterization studies on ZnSe crystals and

epilayers were used to discuss donor and acceptor-bound exciton PL emission in ZnSe.

A typical PL spectrum from ZnSe epilayers was used as an example to highlight the

characteristic features of low-temperature PL spectrum from ZnSe. The distinguishing

features of the luminescence properties of ZnSe NWs were explained in contrast with

the dominating optical processes in ZnSe crystals and epilayers, which do not apply to

NWs. A summary of RTPL and LTPL characterization studies on ZnSe NWs reported

so far in the literature was given with remarks on their general features. The influence

of heat treatment and the role of native point defects in optical properties of ZnSe was

described, and use of heat treatment as a tool to identify native point defects in ZnSe

was also discussed. However, no attempt was made to highlight the gaps in the funda-

mental understanding of optical response of ZnSe NTSLs and NWs. These are discussed


in detail in Chapter 2.

Chapter 2

Motivation and Objectives

2.1 Introduction

A general overview of the progress in optical characterization of ZnSe NWs and the unique

properties of NTSLs were described in Chapter 1. The current chapter outlines the the-

sis motivation and identifies the primary and secondary objectives of this thesis. This

chapter is organized as follows. First, the motivation behind the work is explained based

on the current lack of understanding of the optical and electronic properties of NTSLs.

NTSLs based on ZnSe were chosen as a platform for establishing their fundamental prop-

erties, such as the electronic band gap (Eg,NTSL), using optical spectroscopy as a probe.

It is explained that this cannot be achieved without an unambiguous determination of

the nature of optical transitions in homogeneous (i.e. with no twin-planes) ZnSe NWs

first. Section 2.3 identifies clear deficiencies in the understanding of the fundamental op-

tical properties of homogeneous ZnSe NWs. The importance of the role of native point

defects in the optical response of ZnSe NWs is explained, and a lack of systematic studies

involving native point defects in the literature is highlighted in Section 2.3.1. The opti-

cal characterization of ZnSe NWs using LTPL has invariably been performed on arrays

of NWs. The information obtained from these experiments represents the response of

30

Chapter 2. Motivation and Objectives 31

an ensemble of NWs as opposed to the required knowledge of the optical properties of

single NWs. The limitations and drawbacks associated with the experiments on an array

of NWs are discussed, and the importance of optical characterization of single NWs is

explained in Section 2.3.2 using relevant examples from the literature. Remarks are also

made on the interpretation of existing results on LTPL of ZnSe NWs. It is also empha-

sized that the uncertainties in experimental results can be largely avoided by combining

techniques of optical and structural characterization on the same individual NWs. The

specific primary and secondary objectives of this thesis are outlined in Section 2.4.

2.2 Motivation

NTSLs were introduced in Section 1.3.1 as an emerging new class of electronic materials.

NTSLs have alternating equally spaced homogeneous domains of ZB crystal structure

along the growth direction rotated by 60◦ or odd multiples thereof at a rotation twin-

plane. Their unusual properties and potential in future applications was also discussed

therein. The existence of TSLs was predicted by Ikonic [137] based on the theoretical

calculations of their electronic properties. TSLs did not attract much attention from

researchers initially as an experimental realization of TSLs was not possible at that time.

This changed after the developments in the field of NWs, when experimental synthesis of

NTSLs became a reality. Incidentally, ZnSe was among the first materials on which the

growth of NTSLs was experimentally reported [104]. However, in this work Li et al. [104]

observed a self-organised growth of NTSLs and did not report on any specific conditions

for realization. Since then NTSLs have been synthesised in several materials including

InP [132,196], InAs [124], ZnS [131,135], GaP [196], and SiC [197,198]. There have also

been several reports on the growth of ZnSe NTSLs [77,87,88,96,104,120]. Various models

have been proposed for the growth of NTSLs (e.g. [104, 199, 199, 200]). For example, Li

et al. [104] proposed a model for the occurrence of periodic twins based on the release


of stored elastic energy resulting from the side-wall facets. In their case, the twin-plane

spacing (d) was linearly proportional to the diameter of NTSLs. However, their model

fails to describe the non-linear dependence observed by other authors (e.g. [124,132]), who

suggested different models for the diameter dependence of the periodicity of twin-planes.

In most of the above reports, the NTSLs were obtained by their natural (spontaneous)

growth with no control over their periodicity. Remarkable progress was demonstrated by

Algra et al. [132] and Caroff et al. [124] where they showed the experimental synthesis

of NTSLs with precise control over their periodicity. Despite this recent progress in the

growth of NTSLs with controlled twin-plane spacings [124,132], the understanding of their

electronic structure and optical properties remains unclear. Because the experimental

evidence is lacking, the role of periodically arranged twin-planes in influencing the optical

and electronic properties of NTSLs has not been understood yet. Some authors are

also of the opinion that the twin-planes in NWs are detrimental to their optoelectronic

performance [201–204], and extensive efforts are being devoted to eliminating twin-planes

in NWs [202,203,205].

A few authors have studied the optical properties of NTSLs using RTPL [77,78,87,88,

96]. Woo et al. [204] found that the RTPL from an array of InP NWs containing rotation

twin-planes at random positions along the length of the NW (i.e. not periodic) showed

weaker luminescence than from those without twin-planes. Philipose et al. [78, 81] have

considered the twin-planes in NWs as structural defects, and related the strong deep-

level (DL) emission they observed from periodically twinned NWs to the presence of

twin-planes. They were however careful in pointing out that the DL emission could also

be a result of the stoichiometric deviations in their samples, and not related to periodic

twinning. Li et al. [104] and Liang et al. [120] have studied the optical properties of

arrays of NTSLs (consisting of NTSLs of different periodicities) using low-temperature

photoluminescence (LTPL) and found that the PL spectrum shows characteristic features

similar to those found in a PL spectrum from ZB-ZnSe. This was explained by comparing


the PL spectrum from ZnSe NTSLs with that from ZB-ZnSe single crystals and thin films,

as the homogeneous segments in NTSLs are ZB-ZnSe separated by the twin-planes. Li

et al. [104] also observed a weak emission at 2.836 eV (higher than the band gap of

ZB-ZnSe) and related it to the small amount of WZ-ZnSe present in their sample. The

LTPL studies on ZnSe NTSLs performed so far have failed to reveal any information

that connects the periodicity of twin-planes (d) to their optical properties. This is also

true for NTSLs based on other materials [135] which were studied using RTPL. LTPL

characterization has not been reported on NTSLs of other materials.

Attempts have been made to calculate the electronic structure of NTSLs [206, 207],

but were restricted to the diameter dependence of electronic structure of unrealistically

thin NWs; i.e., NWs with diameters ∼1 nm which are very rarely obtained using estab-

lished growth methods. They are also unsuitable for device fabrication using standard

lithographic techniques. For large diameter NTSLs, empirical models fail to predict any

significant change in the electronic structure. In spite of the lack of any experimental and

theoretical evidence, NTSLs have persistently been conjectured to be strong candidates

for band-structure engineering through control over their twin-plane spacing between

successive domains [124,132,133].

As noted previously, the optical characterization of NTSLs based on ZnSe and other

compound semiconductors reported so far has failed to provide any evidence of the role

that twin-plane spacing (d) in NTSLs plays in their optical properties. This lack of

meaningful experimental results on NTSLs can be understood by considering the tech-

nical challenges involved in the nanoscale mapping of the characteristics of individual

NTSLs. This mapping is required to faithfully understand their properties. A discussion

of these challenges is presented in Chapter 3. The information obtained from LTPL on

optical transitions is fundamentally related to the electronic structure of the material.

An unambiguous determination of the origin and nature of the optical transitions in ZnSe

NWs of ZB and WZ crystal structures is a prerequisite before these can be established


for NTSLs. This is particularly important in view of the current interpretations that

NTSLs are ZB-WZ superlattices or polytypes of the parent material (i.e. intermediate

between ZB and WZ). The origin of the optical transitions in a semiconductor however

is ultimately related to the native point defects. In other words, the recombination lines

mainly arise due to native point defects rather than planar structural defects (except

for the Y-line emission in ZnSe which has been attributed to the structural defects).

Therefore, knowledge of the role of native point defects in ZnSe NWs is required to

fully understand optical transitions in NTSLs. The nature of optical transitions in ZnSe

NWs is unclear yet due to a lack of their extensive LTPL characterization and unclear

interpretation of existing LTPL results.

2.3 Gaps in the understanding of fundamental opti-

cal transitions in ZnSe NWs in literature

A major shortcoming in luminescence studies of ZnSe NWs has been the lack of their

low-temperature optical characterization. As evident from Table B.1 in Appendix B,

the number of reports on ZnSe NWs using RTPL far exceeds those using LTPL for

their optical characterization. Existing reports that describe LTPL from ZnSe NWs are

listed in Table B.1, and their general overview was given in Section 1.6.2. Table B.3 in

Appendix B lists the LTPL peak positions and their respective assignments described

in these studies. In cases where they were reported, the binding energies of donors and

acceptors, and linewidths of emission lines are also listed in Table B.3. Further, any

doping of NWs employed in these works is also indicated. In the following, first the

role of native point defects in the luminescence spectrum from ZnSe NWs is discussed.

This is followed by a discussion on the role of inhomogeneities in as-grown NW samples

that influence the experimentally obtained PL spectrum and render their interpretation

ambiguous.


2.3.1 Role of native point defects

It was explained in Section 1.6.3 that the optical properties of ZnSe NWs, unlike those

of melt-grown ZnSe single crystals, are not influenced by the presence of unintentional

residual impurities. It was also emphasized that the issues of strain and inter-diffusion

of impurities from the growth substrate are not encountered in the case of ZnSe NWs.

It is therefore important to identify the optical processes and recombination centres that

dictate the optical properties of ZnSe NWs. It is well established in the literature that

deviations from stoichiometry during the growth of ZnSe crystals result in native point

defects [178], which influence their optical and electronic properties [178]. Similarly, the

optical properties of ZnSe NWs can be strongly affected by the presence of native point

defects arising from stoichiometric deviations [72]. In order to precisely identify the

native point defects responsible for their respective PL lines in ZnSe NWs, it is necessary

to carry out a systematic analysis of the LTPL experiments on them, taking into account

the earlier studies that have been reported on annealing and thermal treatment of ZnSe

crystals. Below, it is explained that such systematic studies have not been performed,

and indeed, preliminary studies suffer from incorrect interpretations.

Philipose et al. [72] studied the role of native point defects in optical properties of ZnSe

NWs based on their RTPL experiments and post-growth treatment of ZnSe NWs. They

showed that the PL from ZnSe NWs can be tuned to obtain emission in the near band-

edge (NBE) emission region or the DL emission region by annealing treatments in Zn and

Se vapours, respectively. This is shown in Figure 1.6 and discussed in Section 1.6.3 in

Chapter 1. They related the broad DL emission band to a DAP type recombination where

donors and acceptors were identified as interstitials and vacancies of Zn, respectively. The

experiments described by Philipose et al. [72, 78] using RTPL are significant in showing

that the native point defects play an important role in the optical response of ZnSe NWs.

Besides these works by Philipose et al. [72,78], the native point defects in ZnSe NWs have

not been studied using their LTPL characterization. This is also evident from Table B.3


where it can be noticed that authors have not related the LTPL peaks to the native point

defects in ZnSe. This can be ascribed to two reasons.

In many cases, the LTPL spectra reported are not the typical complex PL spectra [55,

108,117,118], which are expected from ZnSe. In these cases, the spectra show a few peaks

in the NBE region (actually at energies lower than what is considered NBE region) and

broad DL emission band. In several cases excitonic peaks were not even observed at low-

temperatures [55, 110, 111, 113, 116–118]. This can either be due to the poor crystalline

quality of NW samples, or due to inadequate experimental arrangement used for their

LTPL characterization. In Chapter 3 it is explained that the implementation of PL is

simple for arrays of NWs, but needs to be performed carefully so that the PL signal can

be collected with a reasonable collection efficiency. In cases where the sample quality is

poor and the experimental arrangement is of low collection efficiency, it is not surprising

that the LTPL spectra may not resemble the typical complex PL spectra.

In a few cases, the LTPL spectra from undoped ZnSe NWs are similar to the complex

LTPL spectra from ZnSe single crystals and thin films. But in most of these works,

the LTPL peaks have been related to extrinsic impurities, and justified based on various

reasons. These include, for example, (i) presence of GaZn and AsSe due to the use of

GaAs substrates [103, 108] (DAP peaks), (ii) presence of Al due to the use of alumina

substrate [104], (iii) residual impurities [111], and (iv) Na and Cu from residual impuri-

ties [113]. In other reports on undoped ZnSe NWs, the LTPL peaks have been assigned

to bound excitons, but no attempts were made to identify the impurities responsible

for binding these excitons [103, 104, 107, 109, 110, 115–117]. These facts can also be no-

ticed from Table B.3 in Appendix B. In another example, Ohno et al. [105] regarded the

emission peak at 2.795 eV as due to the excitons bound to VZn. This is an incorrect

assignment because the LTPL peak near 2.795 eV is due to the excitons bound to VSe,

as can also be noticed from Table A.1 in Appendix B. This is despite the fact that the

reference that Ohno et al. [105] have used [208] also regards this peak as due to VSe.


It is important to emphasize that the growth of ZnSe NWs (or NWs of any material

in general) usually does not result in identical NWs on the growth substrate. In other

words, the NWs can have different concentrations of native point defects depending on

their location on the substrate, and more importantly if the flow of constituent vapours

is affected by other NWs growing in their vicinity. This issue of inhomogeneities on

the growth sample is quite serious which necessitates that optical characterization be

performed on individual NWs. In the next section, the importance of optical studies on

single NWs is explained.

2.3.2 Lack of luminescence studies on single ZnSe nanowires

Most experimental work on LTPL from ZnSe NWs has been on arrays of NWs. There are

only a few reports describing LTPL [110, 111, 115, 116] and LTCL [108, 116] from single

ZnSe NWs. It is important to note that a majority of these reports [110, 111, 115, 116]

originate from the same authors and their main focus was to study the optical properties

of CdSe quantum dots inside ZnSe NWs). These works on single NWs have not revealed

any new information on ZnSe NWs which was not already available from reports on

arrays of NWs. In fact, a few LTPL studies [103, 104, 106, 107, 109] on arrays have

reported more detailed spectra than those on single ZnSe NWs. The low-temperature

optical characterization at the scale of single NWs is important for the following reasons:

1. Area-dependent inhomogeneities on the growth sample: As grown NW sam-

ples are inhomogeneous in that the density of NWs is not uniform on the sample. Also,

the NWs in different regions of the as grown sample may contain native point defects of

different types, and of varying concentrations. For example, Tribu et al. [110] have men-

tioned that in their ZnSe NW samples grown by MBE, many NWs showed practically no

LTPL emission while the others showed strong LTPL. Philipose et al. [77] also noted that

in their ZnSe NWs grown by the vapour-phase transport growth method, different PL

intensities in different areas of the sample. Liu et al. [108] also found variations in LTCL


from different NWs. In order to determine the intrinsic properties of NWs, it is therefore

important to perform the LTPL experiments on single NWs so that such inhomogeneities

do not influence the results.

2. Ensemble broadening of PL lines: Since the PL signal is collected from a large

number of NWs when studying an array, the LTPL lines appear broader than they really

are; i.e., their measured linewidths (taken to be the full-width at half-maximum, FWHM)

are rather large. This broadening can be termed inhomogeneous broadening, in contrast

to the homogeneous broadening which results from the natural broadening of PL lines.

The importance of the narrow linewidths in terms of homogeneous and inhomogeneous

broadening is explained in Section 3.9.3. Briefly, in general, narrow linewidths indicate

better crystalline and optical quality of semiconductors whereas broadening in PL peaks

is related to disorder in the material which limits optical performance. PL linewidths are

large for an array because different NWs contributing to the PL have different diameters.

In many cases, NWs and nanoribbons exist together in the sample [57, 61, 64, 81, 84, 88,

90, 106, 209], and variations in morphology also add to this inhomogeneous broadening.

In other words, when studying an array of NWs, the optical response of the ensemble

is obtained, while the interest is mainly in intrinsic properties of individual NWs. This

inhomogeneous broadening can be excluded only by studying single NWs. For example,

Wishmeier et al. [210] compared the linewidths of LTPL from an ensemble and from

single ZnO NWs (for many NWs), and found that the linewidths from single NWs are

0.3-0.5 meV smaller than those from an ensemble of NWs. In one particular case, they

found the linewidth for excitonic emission of 1.9 meV for single NWs, and 2.1 meV for

an ensemble of NWs (comprised of about 100 NWs). They related this larger linewidth

in the ensemble to inhomogeneous broadening, relating to the variations in diameters,

morphology and crystalline quality of the NWs in the ensemble. They also confirmed

these variations in different NWs by using SEM imaging. Further, they also found that

the luminescence spectra from single and ensemble ZnO NWs are quite comparable (e.g.


in terms of peak positions) except for the difference in their linewidths. The narrowest

linewidth of excitonic PL lines reported on NWs is 300 µeV for a single GaN NW by

Brandt et al. [177]. The variation in PL from different NWs was also observed by Brandt

et al. [177]. They found that most individual NWs show different PL spectra, with many

individual NWs exhibiting linewidths larger than 20 meV. In most work on single NWs

or their arrays, the reported linewidths of excitonic PL lines are larger than in their bulk

form. Brandt et al. [177] have also noted that this fact has been largely ignored in the

literature. The narrowest linewidth reported so far for exciton PL lines in ZnSe NWs is

10 meV, which is almost 10 ten times larger than the narrowest linewidths reported on

NWs of other materials (∼1 meV) [211–216]. In most reports on LTPL from ZnSe NWs,

the linewidths of excitonic lines were not reported. Other reported values of linewidths

include 40 meV by Liu et al. [108] (for a peak at 2.783 eV which they related to the

electron-hole plasma in their NWs), and a large value of 165 meV by Hsiao et al. [118]

which was related to the DAP recombination. It is therefore critical that single ZnSe NWs

be studied in order to reveal the intrinsic linewidths of excitonic emissions in their LTPL

spectrum excluding the effects of ensemble inhomogeneous broadening. It is important to

note that by studying single NWs only the effects of ensemble inhomogeneous broadening

can be eliminated, not necessarily those of other forms of inhomogeneous broadening.

3. Undergrowth on the sample substrate: Another issue with experiments on as-

grown substrates relates to the growth of poly-crystalline crystallites of ZnSe and/or

deposition of other materials (e.g. precipitates of Zn or Se) on the substrate surface,

beneath the NWs. Cooley et al. [114] concluded that the PL response from ZnSe NWs

in their samples originated from the undergrowth of deformed ZnSe NWs on the sam-

ple substrate, rather than from the straight NWs which were of main interest. Shan

et al. [107] have studied the optical response of WZ-ZnSe NWs. In their LTPL spec-

trum, they related only two peaks (at 2.841 eV and 2.746 eV) to the WZ-ZnSe, while

the other five peaks were related to the underlying ZB-ZnSe crystallites on the growth


substrate. Aichele et al. [115] have also assigned one of the peaks in their LTPL spec-

tra from ZnSe NWs to bulk ZnSe that they suggested formed simultaneously with the

growth of ZnSe NWs on the sample floor. Ohno et al. [105] studied ZnSe NWs grown

on a buffer layer of ZnSe using LTPL spectroscopy. They explained their results on

the temperature dependence of PL by taking into account the volume fraction of ZnSe

NWs, R = Vnw/(Vnw + Vepi), where Vnw and Vepi are the total volume excited by the

laser beam of the NWs and the epilayer, respectively. Further, Liu et al. [108] have also

observed that the ZnSe layer underneath the NWs also emits light, and contributes to

the LTCL signal even if by a small amount. These examples clearly illustrate that the

optical properties of ZnSe NWs can not be completely understood when they are stud-

ied using LTPL experiments on as grown arrays of NWs. It is important that NWs be

isolated from the growth substrate and studied individually to exclude the effects of the

underlying material on the growth substrate.

4. Mixture of nanowires and nanoribbons: In many cases of LTPL experiments

on arrays of ZnSe NWs, the NWs were studied together with the ZnSe nanoribbons

(e.g. [106]). Apart from contributing to the broadening of PL lines, as explained above,

these variations in morphologies of nanostructures may also result in variations in PL

spectrum (e.g. different peak positions). For example, Zhang et al. [103] observed a peak

at energy position higher than the band gap of ZB ZnSe in their LTPL work on ensemble

ZnSe nanoribbons. It was in this work that a 10 meV linewidth of an excitonic line was

reported as noted above. In their work on Ag-doped ZnSe NWs, Zhang et al. [106] have

related one peak (at energy position higher than the band gap of ZB ZnSe) to regions

of WZ crystal structure in bi-crystalline (i.e. consisting of both ZB and WZ regions)

nanoribbons. They suggested the existence of ZnSe nanoribbons in their samples along

with that of ZnSe NWs. It is therefore important to study the optical properties of ZnSe

NWs separately from those of ZnSe nanoribbons.


5. Mixture of nanowires of different crystal structures: Several authors have

observed the existence of NWs in two crystal structures in the same sample [104,107,110,

111,115–117]. Luminescence spectra from arrays of NWs from these samples would clearly

include contributions from NWs with different crystal structures. This would lead to

uncertainties in interpretation of experimental results. A few examples are worth noting

in this regard. Liu et al. [109] have reported the crystal structure of their ZnSe nanobelts

as WZ, but they have related all peaks (seven in the NBE region) to the ZB-ZnSe. In

a similar example, Li et al. [119] reported WZ crystal structure for their ZnSe NWs

but assigned all LTPL peaks to ZB structured ZnSe. The latter two examples point to

inconsistencies in the explanation of LTPL results on ZnSe NWs. Further, there have been

several other reports where authors have studied mixtures of NWs of ZB and WZ crystal

structures [104,115,117]. In other works on LTPL from single ZnSe NWs [110,111,116],

the NWs themselves consisted of regions of different crystal structures. It is clear from the

above, that studying a mixture of NWs of two crystal structures or NWs of mixed crystal

structures, fails to provide information on optical transitions in ZnSe NWs in either

crystal structure. More specifically, the identification of PL lines and their assignments in

such cases remain ambiguous. These uncertainties can be avoided by optically studying

single NWs along with the determination of their crystal structure. It also becomes

apparent from the above discussion that the crystal structure determination of the same

individual NWs which are studied optically can be used to understand the origin of

radiative transitions in relation to the electronic structure.

Based on the above discussion it is evident that the optical response from an array

of NWs is highly convoluted because of several factors. It is also important to note that

more than one of these factors contribute simultaneously to the PL spectrum of ZnSe

NWs and result in broadening of PL lines and uncertainties in results. This explains the

need for optical characterization and associated structural characterization of individual

ZnSe NWs.


2.4 Objectives

In view of the discussion in this chapter so far, it is evident that:

• NTSLs have emerged as new materials with potential for materials engineering at the

nanoscale. However, an understanding of their fundamental optical and electronic prop-

erties is lacking. This can be attributed to the technical challenges in their experimental

characterization at the scale of single NTSLs.

• In the case of ZnSe, efforts to understand the fundamental optical transitions in LTPL

spectra from NTSLs require prior knowledge of these transitions in homogeneous (i.e.

defect-free) ZnSe NWs in ZB and WZ crystal structures. These are also lacking in the

literature. This can be ascribed to the experimental work on PL from ZnSe NWs being

on a mixture of NWs of both crystal structures. The existing reports which have optically

studied arrays of NWs, suffer from ambiguities in their interpretation.

• The optical transitions in ZnSe NWs are ultimately related to the native point defects,

which have not yet been identified in the literature.

The primary and the secondary objectives of this thesis are as follows:

1. Primary objectives: To determine

(a) The relationship between the electronic band gap (Eg,NTSL) and twin period-

icity (d) of NTSLs: Eg,NTSL = Eg,NTSL(d). The relationship between Eg,ZB,

Eg,WZ , and Eg,NTSL is expected to emerge from Eg,NTSL(d).

(b) The optical performance of NTSLs based on luminescence linewidths (), and

luminescence lifetime τ

2. Secondary objectives: To unambiguously identify

(a) The optical transitions in single ZB ZnSe and WZ ZnSe NWs with determina-

tion of exciton binding energy for donors (EBX (D0X)) and acceptors (EBX


(A0X)), ionization energies for donors (ED) and acceptors (EA), and ioniza-

tion energies for donors (ED) and acceptors (EA) in a donor-acceptor pair

related transitions.

(b) Physical properties of single ZB-ZnSe and WZ-ZnSe from LTPL spectra, in-

cluding temperature dependence of the electronic band gap (dEg

dT) and strength

of exciton-phonon interactions.

(c) Native point defects responsible for the optical emission in ZnSe NWs, with

determination of EBX (for D0X and A0X), ED and EA for bound exciton

emissions, and ED and EA for DAP-related emissions.

Chapter 3

Experimental Methodology

3.1 Introduction

The work described in this thesis is principally experimental in nature, and existing the-

oretical models were used for the interpretation and analysis of experimental results to

understand fundamental optical properties of ZnSe NWs and NTSLs. The results pre-

sented in Chapter 7 on ZnSe NTSLs however, besides being experimentally novel, pro-

vide new fundamental insights which can not be understood based on existing theoretical

frameworks. These results are supported by the electronic structure calculations which

are briefly described in Appendix E.2. This chapter explains, in detail, the methodology

used for performing the experimental work described in this thesis, while the theoretical

models are presented in Appendix C.2. The use of existing theoretical models in the

case of ZnSe NWs is justified because, as explained in Chapter 1 (in Section 1.6.1), the

nature of optical transitions in ZnSe NWs are not expected to be fundamentally different

from those of ZnSe single crystals and epilayers. The difference lies in the luminescent

centres contributing to these transitions being different in NWs. This is also supported

by the fact that these models have been extensively used in the literature for describing

the optical properties of NWs [105, 106, 108, 217]. Three experimental techniques were

44

Chapter 3. Experimental Methodology 45

used for the characterization of ZnSe NWs and NTSLs, namely (i) PL spectroscopy, (ii)

Time-resolved PL (TRPL) spectroscopy, and (iii) TEM (including related techniques of

HR-TEM and SAD). The fundamentals of PL spectroscopy and TRPL spectroscopy are

described in Appendices C and D respectively, while this chapter is mainly concerned

with their experimental procedures (instrumentation). The details on fundamentals and

principles, instrumentation and operation of a TEM can be found in standard text-books

on electron microscopy [218, 219]. PL spectroscopy was used in two different modifi-

cations referred to as: (i) PL spectroscopy (similar to conventional PL spectroscopy),

and (ii) confocal micro-photoluminescence (µ-PL) spectroscopy. Gustafsson et al. [220]

have pointed out that despite being widely used, the definition of µ-PL spectroscopy is

rather vague. In a traditional sense, when a microscope (or a microscope objective) is

used for the excitation of luminescence in the experimental arrangement resulting in an

excitation spot of size ∼1 µm on the sample, the method is referred to as µ-PL spec-

troscopy [220,221]. This is different from (macro-)PL spectroscopy where the excitation

spot sizes can be larger. Commercially available microscopes are commonly utilized in

µ-PL spectroscopy arrangements [220]. In a broader sense, any PL arrangement with a

spatial resolution of the order of ∼1 µm can be termed µ-PL setup [220]. Spatial resolu-

tion here refers to the size of the excited spot or of the area from where the luminescence

is collected. Confocal µ-PL spectroscopy refers to the particular case when a confocal

arrangement of lenses is used in the µ-PL spectroscopy experimental setup [221–223].

This can be achieved either by using a commercially available confocal microscope or

by designing such an arrangement for specific application(s) in the laboratory, both be-

ing almost equally popular. In the present case, the experimental arrangement was

custom-designed, described in greater detail in Section 3.6. It is to be noted that strictly

speaking, TRPL spectroscopy is not a spectroscopic technique in that the result obtained

is neither a spectrum nor is it plotted as a function of wavelength (or equivalent energy).

Regardless, TRPL and TRPL spectroscopy are used interchangeably in the literature to


refer to the same technique. The term ‘low-temperature(s)’ has been used in this thesis

to imply temperatures close to the liquid-helium (He) temperature (4 K), with actual

temperatures noted at appropriate places.

The rationale behind the choice of these three characterization techniques is described

first in this chapter. This is followed by a brief description of the method used for

the growth of ZnSe NWs and NTSLs. Next, the technical challenges involved in the

characterization of the same individual NWs and NTSLs using these three experimental

techniques is explained based on the difference in sample preparation methods for PL

and TEM methods. The details of the experimental arrangement for PL spectroscopy are

presented next, and its limitations in optical characterization at the scale of single NWs

are highlighted. This is followed by the descriptions of the experimental arrangements

for confocal µ-PL spectroscopy and TRPL. The determination of crystal structure of

NWs and periodicity of NTSLs using TEM is also described briefly. Thereafter, different

experimental conditions used for the acquisition of PL spectra are described together

with the remarks on the normalization of PL results. These also include details on fitting

of the PL spectra to different peak functions. It needs to be clarified that these fitting

procedures do not refer to the fitting of PL spectra to theoretical models. Instead, they

refer to fitting of spectra to one or more peak functions to obtain peak positions and

linewidths more accurately.

3.2 Rationale behind the choice of experimental tech-

niques

PL spectroscopy and TRPL are powerful and well-established techniques for characteri-

zation of optical properties of semiconductors, described in more detail in Appendices C

and D, respectively. In brief, PL and TRPL involve excitation of a semiconductor ma-

terial by light whose photon energy is equal to or greater than the band gap Eg of the


semiconductor. These photons are absorbed in the material and generate electron-hole

pairs. These electron-hole pairs then radiatively recombine emitting photons of energy

corresponding to the optical processes involved in this recombination. In PL, a spectrum

is obtained for intensity of emitted light (actually the collected light, both being pro-

portional to the number of photons emitted) at each wavelength in a certain wavelength

range of interest. These wavelengths can be converted to energy to yield a spectrum

in terms of energy, although in practice the spectra can be presented either in terms of

the wavelength (e.g. [55,111,117]) or energy (e.g. [106,109]) with no loss of information.

See also, for example, Figure 1.5 in Section 1.6.2. The experimental methodology for

PL therefore involves two main steps of excitation and collection (sometimes also called

detection). TRPL differs from PL in that recombination lifetimes are measured which

correspond to the time taken by an electron and a hole to recombine, or equivalently

the time between absorption and recombination. A PL spectrum can also be recorded in

TRPL depending on the instrumentation, and this is quite often the case.

3.2.1 Advantages of PL spectroscopy

PL has been used for the experimental work described in this thesis for three main

reasons:

1. High sensitivity: PL is extremely sensitive to the concentration of native point

defects and impurities in a semiconductor material. In practice, the sensitivity of PL is

limited by the instrumentation used in that not all light emitted by the sample is collected

and only a small fraction of it is measured. This fraction depends on the collection

efficiency of the experimental arrangement, which in turn depends on the accuracy of

the experimental setup, losses (due to reflection and absorption of emitted light) by the

instruments and optical components, and sensitivity of the detector used. It is therefore

important to design experimental setups with maximum achievable collection efficiency.

In general, for carefully designed experimental setups, a sensitivity of 1012–1014 radiative


centres/cm3 can be achieved, corresponding to 10−7–10−9% (considering 1023 atoms/cm3

in a typical semiconductor material). It is worth mentioning that no other semiconductor

characterization technique that can be used on single NWs has sensitivity close to that

of PL, except TRPL which was also used for the experiments.

2. Non-destructive: PL is a non-destructive technique in that there is no damage to

the sample while performing PL, and they can be re-used with other characterization

methods or for further PL experiments. Since in PL a sample is irradiated with light,

photons of light do not cause any damage to the sample. When laser radiation is used

for excitation, it is clear that this is true only for excitation intensities up to a certain

limit. It is not surprising that extremely high laser intensities can even melt the sample,

but such conditions are almost never used in PL experiments. The typical excitation

intensities used are much below the damage threshold of the material. For example,

the damage threshold for ZnSe is 1.2±0.3 GW/cm2 [224] (for laser radiation), while

the excitation intensities typically used are ∼few kW/cm2 (e.g. 0.8 kW/cm2 [106], 5

kW/cm2 [72]), six orders of magnitude smaller. It is crucial for experiments described

in this thesis that optical characterization be non-destructive because single NWs which

were studied optically were also characterized later by TEM and HR-TEM. Also, several

PL experiments were needed to be performed on the same single NW or array samples,

and this was only possible due to the non-destructive nature of PL. Further, no special

methods are required for sample preparation for PL experiments. For study of NWs

attached to a substrate, the substrate can be conveniently used. The situation is however

more complex for experiments involving single NWs, and discussed in detail in Section 3.4.

Note that this is not a limitation of PL, but a demand imposed by the objectives of the

experiments.

3. Contact-less: A significant advantage in using PL is that it is a contact-less method

- i.e., metal contacts are not required for PL experiments. The process of fabricating


metal contacts to NWs is not only tedious, time-consuming and expensive, but would

also render the sample containing NWs unusable for structural characterization using

TEM. This is because the sample containing NWs with metal contacts cannot be easily

used inside a TEM.

Since PL and TRPL are similar in excitation part, and different only in the detection

part, the above advantages also apply equally to TRPL. The advantages mentioned

above are also among the reasons which make these techniques widely used methods for

characterization of semiconductors.

3.2.2 Limitations of PL spectroscopy

PL also has certain limitations as discussed below:

1. Cost: (applicable only to work in this thesis) In general, PL is considered a cost-

effective characterization technique because once the experimental arrangement is made,

the operating costs are minimal especially for room-temperature PL (RTPL) experiments.

However, since all the PL and TRPL experiments described in this thesis were performed

at low-temperatures, a considerable expense was incurred on liquid He required to cool the

sample. Further, the major modification of the experimental setup for PL spectroscopy

to that for confocal µ-PL spectroscopy also involved additional costs.

2. Semi-quantitative: PL is a semi-quantitative characterization technique in that it

is difficult to obtain quantitative information based on a PL spectrum alone. In general,

a large number of measurements are required with a varying parameter, e.g., excitation

intensity, temperature, etc., in order to yield quantitative information. Because PL is

non-destructive, in principle it can be performed an unlimited number of times. The

resulting disadvantage is that more time is required for these additional experiments.

Also, large number of additional low-temperature experiments described in this thesis

also added to the cost (liquid He).


3. Radiative recombination: The information obtained from PL is specific to radiative

recombinations only. Since luminescence light from radiative recombinations is collected

in PL, information on non-radiative recombinations which do not result in light emission

is not contained in a PL spectrum.

The advantages of high sensitivity and PL being a non-destructive and contact-less

method clearly outweigh the disadvantages of extra cost and time. Since TRPL can

be used to obtain information on non-radiative recombinations, combined use of PL

and TRPL provides the required information. TEM, HR-TEM and SAD were used for

structural characterization of ZnSe NWs presented in this thesis because there is no

other structural characterization method that can be applied to single NWs to yield

information on their crystal structure and periodicity of twin-planes in NTSLs. The next

section is a summary of methods used for the growth of ZnSe NWs and NTSLs.

3.3 Growth of ZnSe nanowires and NTSLs

The method used for the growth of ZnSe NWs and NTSLs is based on a Au-catalyzed

vapour-liquid-solid (VLS) growth mechanism. The principles of VLS growth mechanism

were described earlier in Section 1.4. A vapour-phase growth method using a tube furnace

apparatus was used for the growth of NWs and NTSLs on a Si(111) (or Si(100)) substrate.

This is also referred to as chemical vapour deposition CVD method. The methodology

is similar to that described previously by Philipose et al. [72, 81]. The Si substrate was

first cleaned and a thin layer of Au (20-50 nm) was deposited by thermal evaporation.

The source material used was ZnSe powder (Alpha Aesar and SPi Inc., 99.999% purity).

The growth was carried out at a fixed substrate temperature under a constant flow of

Ar carrier gas. No intentional doping was carried out. The growth of ZnSe NWs in

conditions of excess Zn (Se) was carried out with an extra source of Zn (Se) with the

source material.


Philipose et al. [78] and Wang et al. [77] have earlier reported that at growth tem-

peratures higher than 650 ◦C, the NWs obtained contain periodic twinning (i.e. NTSLs).

For the growth of NTSLs, substrate temperatures in the range of 650 ◦-750 ◦C were used.

Apart from the technical challenges involved in the characterization of single NTSLs (de-

scribed in the next section), another difficulty stems from the lack of control over the

twin-plane spacing during their growth using this method. This means that NTSLs with

desired periodicities were not obtained on the growth substrate. This is because the

minor fluctuations in source and substrate temperatures and the flux of vapours in the

vapour-phase growth method (tube-furnace based) result in growth of mixture of NWs.

This mixture contains homogeneous NWs, NTSLs with different twin-plane spacings and

randomly twinned NWs (RTNWs). (A RTNW contains rotation twin-planes at random

positions along its length, in contrast to those in NTSLs which are periodic). In other

words, there was no control over the twin-plane spacings during the growth of NTSLs by

the method used, and NTSLs with varying twin-plane spacings were self-organized on the

growth substrate. This lack of control was compensated by performing a sequence of com-

bined optical and structural characterization experiments on several individual NTSLs,

and NTSLs with a range of twin-plane spacings were then selected. The next section

explains the obstacles that needed to be overcome before this combined characterization

could be performed.

3.4 Obstacles in combined optical and structural char-

acterization of the same individual nanowire

In view of the objectives outlined in Chapter 2, a successful correlation between optical

properties and twin-plane spacing (d) is required in order to establish their correspon-

dence. This, in turn, requires the structural characterization on individual NTSLs to

determine their periodicity (d), and optical characterization on the same NTSL to deter-


mine their fundamental properties, such as the electronic band gap Eg. It is crucial that

experiments on NTSLs be performed on individual NTSLs since the determination of

optical response from an ensemble of NTSLs is not of any significance. The importance

of studies on single homogeneous NWs in contrast to those on an array of NWs has

also been discussed at length in Section 2.3.2 in the previous chapter (Chapter 2). The

requirement of optical characterization on single NWs in direct relation with their micro-

structural characteristics was also emphasized therein. Characterization at the scale of

single NTSL and NWs with combined use of two characterization techniques is extremely

difficult. This combination increases the complexity of the experiments, to the extent

that such studies have not been reported on single homogeneous NWs or NTSLs based on

any material. It is equally important to highlight that the challenges in characterization

of single NWs are compounded as the number of characterization techniques to be used

on the ‘same’ NW is increased. Accordingly, there have been fewer studies that describe

both LTPL and TRPL on the same NW (e.g. InP NWs [216]) than those that describe

LTPL alone. Optical and structural characterization on the same NW is far more chal-

lenging and so far, there have only been three recent reports that describe LTPL and

HR-TEM characterization on the same NW [133, 225, 226], and none on NTSLs. It is

of utmost importance to point out that there has been no study reported so far that

describes LTPL, TRPL and HR-TEM characterization on the same NW (homogeneous

and inhomogeneous). Considering that this is true for NWs of all semiconductor materi-

als being researched by an enormously large research community globally, the magnitude

of the challenge can be understood. The details of these challenges and the required

modifications to the experimental methodology are described below. More specifically,

the next section explains why optical and structural characterization using three differ-

ent techniques cannot be performed on the same NW using routine methodology. It is

re-iterated here that in the following discussion optical characterization refers to LTPL

characterization only, and that RTPL does not yield the information of interest.


3.4.1 Bottleneck - incompatible sample preparation methods

for different techniques

The most important hurdle in optical and structural characterization on the same individ-

ual NW is the incompatibility of the sample preparation and sample mounting methods

used for different techniques. Thus, it is extremely important to find a method that allows

optical and structural characterization to be performed on the same NW. The most com-

mon methods of LTPL characterization involve either immersing the sample in liquid He,

which has been a common method in the past [227–229], or mounting the sample on the

cold-finger of an optical microscopy cryostat. In a typical procedure of synthesis of NWs

using the VLS growth mechanism, a substrate of suitable size (with a thin metal film for

catalyst-assisted growth processes) is used inside the growth apparatus. These include,

for example, a tube-furnace for vapour-phase transport method [60, 68], a high-vacuum

chamber for MBE [67, 105] and MOCVD methods [57, 59]. A variety of materials have

been used as substrates (referred hereafter to as as-grown substrates), including but not

limited to, Si(001) [57,59], Si(111) [72,78], sapphire [91,230], quartz [231], GaAs [64,66],

etc. The NWs grow on the surface(s) of the substrates exposed to vapours and are ready

to be used with optical, structural and electrical characterization methods, typically af-

ter a brief inspection with a SEM or an optical microscope (OM). For low-temperature

optical characterization purposes, two methods are most commonly used for mounting

the sample containing NWs to the cryostat. One is to use the as-grown substrate inside

the cryostat [112], the other is to transfer the NWs to another clean bare substrate ei-

ther mechanically [210] or by sonication in a solution [210] (using a suitable alcohol e.g.

methanol [72]) and then putting a drop of the solution on a substrate. Both methods

result in an opaque sample which is then mounted on the cold-finger of a cryostat using

a cryogenic vacuum grease. Other methods to secure the sample to the cryostat can also

be used, but are less common e.g., use of a liquid Indium seal. The use of vacuum grease


is important to make a good thermal contact between the sample and the cold-finger,

otherwise the sample may actually not be at the temperature shown by the temperature

controller used for monitoring the temperature during the experiments. The opaque sam-

ple thus obtained cannot be used inside a TEM chamber because operation of a TEM is

based on a transmitted beam of electrons. The issue is then making this opaque sample

so thin (at least locally near the site of the NW to be studied) that it can be used inside

a TEM.

This issue has commonly been faced in cross-sectional TEM characterization of semi-

conductor devices previously. Szot et al. [232] proposed a new method of sample prepa-

ration for cross-sectional TEM based on the focused ion beam (FIB) machining of semi-

conductor devices to obtain thin foils for TEM characterization. This method has been

widely used since then for preparing samples for TEM characterization out of semicon-

ductor devices [233,234]. In the simplest cases, the sample containing NW(s) of interest

can be used inside a TEM by using a FIB sputtering of the sample substrate, leaving the

NW to be studied intact. However, the use of this method for sample preparation of NW

samples for TEM characterization has been rare. This is because TEM characterization

of NWs dispersed on TEM grids (described in the next paragraph) is straightforward,

and FIB methods are not needed in general for sample preparation. FIB methods have

been used for specimen preparation for samples containing NWs, however it was not in

connection with optical spectroscopy [235]. In reports on optical characterization of sin-

gle NWs, different NWs have been studied through optical spectroscopy and TEM, i.e.

it is not the same NW which is studied by both methods. Further, this method (FIB)

is impractical when combined optical spectroscopy and electron microscopy are to be

performed because (i) it is very expensive given the high cost of FIB instrument, (ii) it is

time-consuming because FIB sputtering would require at least a few hours for each NW,

and (iii) there is a high risk of damage by the FIB procedure to the NW to be studied.

The structural characterization of NWs using a TEM requires an electron transpar-


ent conducting substrate thin enough for transmission of electrons or a substrate which

allows the NWs to be exposed through holes. This is almost exclusively carried out using

commercially available TEM grids of suitable mesh size and made of a compatible mate-

rial, copper being least expensive and thus most commonly used. These TEM grids are

typically 3 (or 3.05) mm in diameter compatible with most commercially available TEM

sample-holders, and 20-50 microns thick. These TEM grids cannot be used inside the

optical cryostat for low-temperature characterization because being so thin they cannot

be mounted on the optical cryostat using vacuum grease, as that would result in their

destruction. Even if they are not destroyed, they cannot be used inside a TEM chamber

because of the contamination by the hydrocarbons in the grease. Room temperature

optical characterization is certainly possible as there is no need for a good thermal con-

tact and this was done earlier by Zapien et al. [236] for single ZnO nanoribbons. In the

present case however, as explained earlier the objective is to perform low-temperature

optical characterization. There are at least three reports of characterization involving PL

and TEM on the same NW, concurrent with the work described in this thesis. These are

on InP NWs [133], heterostructrure GaAs NWs [225] and heterostructure GaAs/GaAsSb

NWs [226]. Bao et al. [133] used a new kind of TEM grid in their work on InP NWs

which uses a silicon substrate with a window covered by thin (20 nm-30 nm) silicon ni-

tride membrane. They also fabricated gold patterns lithographically on these grids for

dissipation of heat generated by the laser beam. There are however a few issues with us-

ing such grids: (i) the thin silicon nitride membrane is very fragile making the transfer of

NWs on the grid difficult. Even if this is achieved, the grid cannot be used several times

given its fragility, (ii) the problem again is that of mounting the Si substrate (grid) on the

optical cryostat, and (iii) designing metal patterns for heat dissipation is time-consuming

and expensive.

To overcome these issues, a sample holder was custom-designed to hold the TEM

grids, and the sample holder was used inside the optical microscopy cryostat for low-


temperature optical characterization. The NWs were mechanically transferred on the

TEM grids (TedPella Inc.) by bringing the as-grown substrate and grids in close contact.

The sample holder had a circular groove of diameter exactly equal to the diameter of the

grid, and a cover-ring was used above the grid to secure it to the sample holder. Different

sample holders were designed for grids of different diameters (e.g. 3 mm and 3.05 mm).

The height of the sample holder was smaller than the distance from the cryostat window

to the cold-finger of the optical cryostat (1 mm). The sample holder and the cover-ring

were made of Cu. Since Cu is a good conductor of heat, no further modifications (metal

patterns etc) were required to ensure efficient heat dissipation. The cover-ring was useful

not only in securing the TEM grid so that it does not fall out of the sample holder while

handling, but also in facilitating a good contact of the grid with the sample holder, which

in turn would ensure a good thermal contact with the cryostat. It was found that on TEM

grids without a reference scheme, keeping track of the same NW during the experiments

is quite difficult. The grids used were reference-finder type (TedPella Inc., 79750), where

reference marks (letters or symbols) can be used as a coordinate system to locate the

same NW during the experiments. In what follows, the experimental methodology which

enabled different characterization techniques to be implemented on the same NWs and

NTSLs is described in detail. The next section describes the details of the experimental

arrangement for PL spectroscopy.

3.5 Photoluminescence (PL) spectroscopy

The experimental arrangement described in this section was used for the LTPL experi-

ments on arrays of ZnSe NWs using the as-grown substrates. A diode-pumped solid state

laser (Coherent Verdi-V8/V10) with output wavelength of 532 nm was used to optically

pump the Ti:Sapphire oscillator (Coherent MIRA-900F). The output power of Verdi-

V8/V10 can be adjusted up to 10 W, while 8 W was routinely used for the experiments.


The output wavelength of Ti:Sapphire oscillator can be tuned from 700 nm to 990 nm

and the output power, which depends on its output wavelength and pump power from

Verdi-V8/V10 among other factors, can be up to 1.75 W. These other factors include, rel-

ative humidity in the oscillator cavity, absorption at certain wavelengths by water, chiller

temperature and cleanliness of the optics inside the system. The output laser radiation

from the Ti:Sapphire oscillator is pulsed with a pulse-width of 200 fs (1 fs=1 × 10−15

sec) and a frequency of 76 MHz. The output from the Ti:Sapphire oscillator is, in turn,

used to pump a commercial second harmonic generator (SHG) (INRAD Model 5-050),

which uses a BBO (beta-barium borate) crystal (Casix). The laser output from SHG has

similar pulse-width and frequency as the input pump laser with correspondingly half the

wavelength.

The fundamental band-gap Eg of ZnSe is 2.68 eV at room-temperature [72, 102],

Eg,ZB=2.822 eV at 4 K for zinc-blende (ZB) ZnSe [155, 237] and Eg,WZ=2.874 eV at 4

K for wurtzite (WZ) ZnSe [238], corresponding to wavelengths of 463 nm, 439.3 nm and

431.4 nm, respectively. The appropriate wavelength for excitation therefore needs to be

shorter than 431 nm for above band gap excitation of ZnSe. A convenient wavelength

for excitation of ZnSe is 400 nm (photon energy, Eph=3.1 eV) which corresponds to the

output wavelength from the Ti:Sapphire oscillator of 800 nm. This is also close to the

wavelength where the output power from the Ti:Sapphire oscillator is maximum (785

nm). Further, the influence of the exciting laser beam in the luminescence spectrum are

avoided at a wavelength of 400 nm.

The laser beam from SHG is directed to the sample using a set of mirrors and lenses. A

schematic of the experimental set-up is shown in Figure 3.1. The NW sample (as-grown

substrate) is placed inside a continuous-flow type liquid He cryostat (Janis Research,

ST-500) and temperatures down to 3 K can be typically reached, provided the relative

humidity is not too high in the atmosphere. (It should be noted that the lowest temper-

atures reached, also called the base temperature, are not always equal, but are close to


Experimental Setup PL spectroscopy

Verd

i Solid

-

Sta

te L

aser

Ti:Sapphire

Coherent MIRA-900F

LS

AC

Second

Harmonic

Generator

Pin Diode

BS1

BS2

BS3

FM1

FM2

M1 M2

M3

M4

M5

M6

M7

M9

M8

M10

Hamamatsu

C-5680

Streak

Camera

Spectra-

Pro

Spectro

meter

Lens

Cryostat

Optical fibre

Figure 3.1: A schematic of the experimental arrangement used for PL spectroscopy.

(M=mirrors, FM=flip-mirrors (mirrors on flip mounts), BS=beam-splitters, LS=laser

spectrometer, AC=auto-correlator)

the liquid He temperature 4±1 K). The temperature is monitored using a temperature

controller (Lakeshore, Model 320) with an accuracy of ±1 K. The temperature controller

is equipped with a heater which can be used to vary the sample temperature during the

experiments. When required, the excitation intensity (Iexc, excitation power per unit

area) of the laser can be adjusted by using a variable neutral density (VND) filter. It is

a customary practice to note the Iexc or the range of Iexc used, similar to the case of tem-

perature (T ), together with the results. The luminescence signal from the NW sample is

collected using a long-working distance objective (Mitutoyo, [M Plan Apo SL 50]), and

guided to the spectrometer using an optical fibre placed close to the microscope objective,


Spectrometer, Triax-320

Liq. He Cryostat

Optical fibre

Microscope objective

Streak camera

Figure 3.2: Photograph of a section of the experimental arrangement for PL spectroscopy.

The optical fibre is placed close to the microscope objective.

as shown in Figure 3.2. (An optical alignment without using the optical fibre can also

be used instead). The numerical aperture NA of the microscope objective is 0.42. The

PL signal is dispersed by a spectrometer (Jobin-Yvon, TRIAX320). The spectrometer

has three diffraction gratings, and the two most commonly used gratings had 1200 and

1800 grooves/mm. The spectrometer resolution is 0.1 nm at 365 nm and 546 nm. When

converted to energy, this corresponds to a resolution of 0.93 meV and 0.42 meV at these

wavelengths, respectively. The dispersed signal is detected by a liquid nitrogen cooled

charge-coupled device (CCD) detector (Jobin-Yvon, CCD3000). It is to be noted that

the resolution is limited not only by the spectrometer, but also by the detector. The ex-

perimental setup includes other accessories and diagnostics for the laser beam, such as a


laser spectrometer (Pulse-scope, Ape-Berlin) for measuring the wavelength, and an auto-

correlator (Auto-Correlator Mini, Ape-Berlin) to measure the pulse-width. The entire

experimental arrangement involves few other equipment (e.g. oscilloscopes, mechanical

pump(s) etc) and optical elements (e.g. VNDs, optical filters, mirrors, lenses etc) for

the purposes of attenuating and steering the laser beam and the PL signal, for example.

Figure 3.2 shows an image of the detection part of this experimental arrangement.

During LTPL experiments on an array of NWs, an area on the sample is selected

where a large number of NWs (also referred to as an ensemble or a bundle of NWs)

are studied collectively after optical excitation. The PL signal from an array of NWs

is more intense than that from a single NW. This is simply because a large number of

NWs contribute to the PL signal in an array, as opposed to the case of a single NW.

The stronger PL signal from an array leads to the acquisition of PL spectra with high

signal-to-noise ratios (SNR), thereby also reducing the inaccuracies in their collection

and interpretation. Further, characterization of single NWs also requires imaging capa-

bilities in the experimental arrangement so that single NWs can be located for the LTPL

experiments. These factors together explain the lack of reports on single ZnSe NWs using

LTPL spectroscopy. This is, however, in contrast to the NWs of other materials which

have been studied extensively using single NWs [177,210,216,239]. (However, in general,

the number of experimental works on arrays of NWs far exceed those on single NWs for

any material).

The experimental setup for PL spectroscopy described above is highly efficient, es-

pecially for experiments on an array of NWs. The results described in Chapter 4 were

obtained using this experimental setup. However, it showed limitations in functionality

regarding experiments on single NWs. This is because the luminescence signal was cou-

pled directly to the optical fibre and imaging of single NWs was not possible, which is an

important requirement for their characterization individually. It is a necessary step to

locate the single NW to be studied on the sample before the actual PL experiment. This


can be done by taking images of the sample before the PL experiment. More importantly,

the PL signal from a single NW is much weaker in intensity (or signal strength) than that

from an array of NWs which contains a high density of NWs. The collection efficiency

of the setup described above was not optimum for characterization at the scale of single

NWs; i.e., the experiments would have suffered from low signal-to-noise ratios (SNR).

TRPL experiments on single NWs were also not possible on this setup. This is because

the luminescence signal could not be used for TRPL detection since the optical fibre car-

rying the signal was coupled directly to the microscope objective at one end and to the

spectrometer at the other end. TRPL in general is more difficult to detect because the

luminescence signal is resolved both spectrally (in wavelength) and temporally (in time).

Further, the detectors used in the TRPL detection scheme employed in this thesis suffer

from poor sensitivity [240]. This refers to the detector used inside the streak camera

which was used for the TRPL experiments. In other words, a PL spectrum can always

be acquired (if permitted by the instrumentation) whenever TRPL can be measured, but

not vice-versa. For example, in their report on PL from single InP NWs, Bao et al. [133]

have noted that they could not measure TRPL because of the weak luminescence signal

from single NWs.

3.6 Confocal Micro-photoluminescence (µ-PL) Spec-

troscopy

The experimental setup required for optical characterization of single NWs is more in-

volved than that needed for an array of NWs. LTPL studies on single NWs have been

reported on NWs of different materials (e.g. InN [239], ZnO [210], GaN [177], and

InP [216]). It was noted in Chapter 2 that there has been almost no progress in optical

characterization of single ZnSe NWs (when compared with the NWs of other semicon-

ductor materials) in that (i) very few LTPL studies have been reported thus far on single


SHG

Spectrometer

Pin-hole

Microscope objective

Optical fibre

Figure 3.3: Photograph of the experimental arrangement for confocal µ-PL spectroscopy.

Inset in the lower right corner shows the confocal part with a pin-hole in the focal plane.

ZnSe NWs [111,116], (ii) there have been no TRPL studies on ZnSe NWs (not even on an

array of NWs). It is because of this, that the references used below in the discussion are

to NWs of other semiconductor materials. It is important to note that the experimental

modifications described hereafter are not exclusive to characterization of ZnSe NWs, but

can be used with comparable performance for single NWs of any semiconductor material

with necessary changes according to the spectral region of interest. The discussion can

therefore now be expanded to include NWs of all semiconductor materials to provide a

broader perspective.

In the experimental arrangement for PL spectroscopy described in the previous section

the optical fibre for collection of PL signal was placed close to the microscope objective.


In the modified confocal µ-PL arrangement, the optical fibre is placed after the confo-

cal assembly. The experimental arrangement designed for confocal µ-PL spectroscopy

is shown in Figure 3.3 and a simple schematic is shown in Figure 3.4. The extent of

modifications is evident by comparing the position of microscope objective and optical

fibre in Figures 3.2 and 3.3. The instruments used for the excitation and collection of

luminescence are similar in both cases, while the major difference lies in the procedure for

mounting the NW sample. The sample (a TEM grid secured inside the sample-holder)

was mounted on the cold-finger of an optical microscopy continuous-flow type cryostat

using a cryogenic vacuum grease. The sample chamber was evacuated using a mechani-

cal pump before the transfer of liquid He starts in pull-mode using the same mechanical

pump. A quartz-tungsten-halogen (QTH) lamp was positioned to illuminate the sample

with white (broadband) light in normal incidence for imaging purposes. The lumines-

cence and reflected white light from the sample was collected in a reflection geometry

using a long working distance microscope objective lens (Mitutoyo, NA=0.42). After the

cool-down process is complete and a temperature around 4 K is reached, the NW of in-

terest is located using imaging with white light. The sample image obtained at the image

plane lacks good contrast. The white light was passed through a polariser to illuminate

the sample, and the back-reflected light was passed through another polariser (an anal-

yser) before being imaged by a CCD camera (Hitachi KP-32N). In this way images with

good contrast were obtained whose brightness can be adjusted by adjusting the crossed

polarisers. One advantage of this method is that anisotropic crystals (e.g. with a WZ

crystal structure) are easily identified as bright objects, but this also results in extinction

of brightness of the isotropic crystals (e.g. with a ZB crystal structure). This is because

anisotropic crystals rotate the polarization of incident light [241] and appear bright. The

light reflected from isotropic crystals with a certain polarization does not pass through

the analyser with orthogonal polarization, thus making them look dark or invisible (ex-

tinction). The analyser is removed during the LTPL experiments after initial imaging so


He

-Ne

Las

er

Halogen lamp (Illumination)

Streak camera

Pin-hole (200 μm)

Linear translation stage

Excitation

Objective

CCD camera

Fibre

Translation stage

To TV

To detector

VND

BS

Sample

Focal plane

Image plane

Z

X Y

Figure 3.4: Schematic of the confocal µ-PL spectroscopy setup. Inset: Area on an as-

grown sample excited by the laser. The laser spot is elliptical due to the oblique incidence

of the exciting laser beam.

as not to attenuate or otherwise influence the PL signal. The laser excitation was used

in an oblique-incidence mode. This results in a large area of the excited spot which is

elliptical in shape (instead of a circular spot obtained with normal incidence), shown in

the lower left corner of Figure 3.4. The area of this spot is approximately 100 µm ×

50 µm, and this area has been used to calculate excitation intensities (from excitation

powers). The large area of the excited spot is advantageous in the case of experiments

presented in this thesis because of the use of a pulsed excitation with high peak powers.

Due to this large area, Iexc remains nominal despite high peak powers of the pulsed laser.

The PL signal can also be imaged at the image plane to obtain a PL image of the


NW being studied. Through this approach the NWs which are efficient emitters can be

chosen and experiments focused on their studies. It is important to select the NWs which

lie in the open or exposed areas of the grid (i.e. not lying on the grid bars) for subsequent

TEM characterization. In experiments described in this thesis, the NWs studied were

almost always in the open areas. Due to the large mesh sizes (of reference-finder grids)

used, resulting in large open areas, NWs attached to the grid bar at one or both ends

were chosen. However, due to the issue of charging by the electron beam inside the TEM,

several NWs in this configuration were rendered unusable.

The PL signal passes through a periscope assembly, and through a confocal arrange-

ment. The periscope assembly was used to compensate for the height. It does not affect

the characteristics or the performance of the experimental setup in any way and should

not be needed in general. A pin-hole is maintained at the focal plane of the confocal

arrangement, and is used to select the region of the NW of interest. The pin-hole and

the confocal assembly are shown in the inset of Figure 3.3. The focal plane and the

image plane are shown in Figure 3.4. The pin-hole was mounted on a three-dimensional

translation stage to allow adjustment in the focal plane, and to select an area of the

NW (or the whole NW) in the other two planes. The use of a pin-hole is one of the

important distinguishing factors from other reports on LTPL of single NWs. In these

works a pin-hole was not used. In terms of the spatial resolution, in the experiments

described in this thesis the resolution is limited by the size of the pin-hole, not the area

of the excited spot which is large. A spatial resolution of 5 µm can be obtained by using

a pin-hole of diameter 200 µm. The spatial resolution can be changed by using pin-holes

of different sizes (e.g. the spatial resolution is 3.75 µm and 2.5 µm with pin-holes of

diameters 150 µm and 100 µm, respectively). However, as the size of the pin-hole is

reduced, the amount of the signal collected is also reduced. In most cases, a pin-hole

with a diameter of 200 µm was used, while the use of other pin-holes was rare. In LTPL

studies reported in literature without the use of a pin-hole, the resolution is limited by


the area of the excited spot (e.g. 1.5 µm (ZnO [210]), 2 µm (InP [213] and GaN [211]),

3 µm (GaN [177] and InP [216])).

3.7 Time-resolved Photoluminescence (TRPL)

A part of the signal (≈ 50%) is sent towards the entrance slit of a second spectrometer

(ACTON PI 300i), passing through another confocal arrangement without a pin-hole,

as shown in the Figure 3.3. The use of a second pin-hole here would require another

CCD camera at the image plane and would also require adjustments in X- and Y-planes

to synchronize the image with that of the first pin-hole, and in the Z-plane to adjust

the position of the pin-hole in the focal plane. The use of a second pin-hole can be

avoided because it does not improve the spatial resolution or the collection efficiency of

the entire setup. This spectrometer also has three diffraction gratings, and a grating

blazed at 500 nm with 600 grooves/mm was used for the experiments. The efficiency

of the grating is maximum at the blaze wavelength. The spectrometer has a manual

vertical slit whose width can be adjusted from 0 to 3 mm using a micrometer knob.

This second spectrometer is coupled to a streak camera (Hamamatsu C-5680), equipped

with a CCD camera at the detection plane. Figure 3.5 shows a simple schematic of time-

resolved detection using a streak camera. The operating principles of a streak camera are

described in more detail in Appendix D. The photons of incoming luminescence signal

are converted to photoelectrons using a photo-cathode, and are swept (dispersed) in time

using a time-varying high-voltage. These electrons are then imaged by a detector (a CCD

camera in this case) to form an image (the streak image), which contains information

about PL lifetimes. The input optics of the streak camera is placed at one of the exit

ports (with no slits) of the spectrometer. Other accessories used in conjunction with the

streak camera are a delay unit (for adjusting the delay time), a pin-diode (to synchronize

the frequency with that of the laser signal), and a chiller (RTE-4). The streak camera


Incident signal

Time varying high voltage

Streak image

Figure 3.5: A simplified schematic of the detection of time-resolved photoluminescence

TRPL using a streak camera.

has a manual adjustable horizontal slit.

The PL signal enters the spectrometer through a aspheric lens with a high numerical

aperture which helps to obtain a diffraction limited spot at the entrance slit of the

spectrometer and that of the streak camera. Since the PL signal goes through a pin-hole

and two orthogonal slits, the signal will not always be at the entrance slit of the streak

camera and needs to be adjusted before entering the spectrometer. No adjustments can be

made after the entrance slit of the spectrometer. The easiest way to obtain this is to use

the aspheric lens (which couples the signal to the spectrometer) on a XYZ-translation lens

mount, or a XY-translation lens mount which is positioned on a translation stage. The

latter assembly was used in the present experiments. The Z-translation of the translation

stage was used to focus the PL signal at the entrance slit of the spectrometer, and it was

left unchanged. This translation in Z-direction is optional, and should not be needed in


general. It was used here because of space constraints on the optical bench. The XY-

translation of the lens mount was used to move the lens in accordance with the movement

of the pin-hole. This was done by monitoring the laser signal in focus mode (signal is not

swept in time) of the streak camera with minimal gain applied. Despite the high-stability

of the Ti:Sapphire laser, the wavelength of excitation can easily fluctuate by ±0.1 nm.

The SHG output wavelength was recorded using the first spectrometer (described in

Section 3.5), and then the X-translation of the lens mount was adjusted to obtain the

image of the laser signal centred at the same wavelength. This also ensures that the

movement of the lens does not affect the accuracy of the peak positions being recorded.

The Y-axis is then adjusted to maximize the intensity of the signal in the vertical (time)

direction. The delay time is adjusted on the program (HPD-TA) (described further in

Appendix F), and then locked to remove problems of jitter as much as possible. It

was observed that jitter is negligible for the time-range of 3 (1500 ps) and 4 (2200 ps),

appreciable for the time-range 2 (800 ps) and very difficult to control for the time-range

1 (150 ps), which is reasonable for instruments not equipped with jitter-correction optics.

The jitter arises due to two factors, (i) the fluctuations in the frequency of the laser, and

(ii) the electronics inside the streak camera.

The methodology for LTPL and TRPL experiments on the same NWs described

above was simultaneous and performed at the same time on two different computers.

This gives an added crucial advantage in removing the ambiguities that may arise from

fluctuations if the experiments were carried out sequentially (or on different days). These

fluctuations can result from fluctuations in the laser power, sample temperature etc and

also uncontrollable factors such as relative humidity in the atmosphere, heating up of the

mechanical pump, detectors etc.


3.8 Transmission Electron Microscopy (TEM)

For the TEM experiments, the TEM grid was simply taken out of the sample holder and

was used inside a TEM using routine procedures. A JEOL-2010F STEM operating at

200 kV and an aberration-corrected Titan-80-300 (FEI) STEM operating at 300 kV were

used for the experiments. HR-TEM characterization was performed on both instruments.

Selected-area electron diffraction (SAD) patterns were acquired on JEOL-2010F. The

TEM micrographs were acquired, processed and analysed using the DigitalMicrograph

program (GATAN) (listed in Appendix F). The program can also be used to acquire power

spectra (fast-Fourier transform, (FFT) pattern) by selecting an area on the micrographs.

The FFT patterns can be used as diffraction patterns for areas which are too small

to be selected for the acquisition of SADs (or when SAD patterns were not available).

TEM is a powerful tool for characterization of microstructure of materials. However,

its use in this thesis was limited to the determination of crystal structure of NWs and

periodicity of twin-planes in NTSLs. Crystal structure can be determined by using the

information on lattice spacings from HR-TEM micrographs. Also, indexed spots in SAD

patterns can be used for the same purpose. The latter method is more deterministic

than the former because different crystal structures may have similar lattice spacings in

different directions. In this thesis, both HR-TEM micrographs and SAD patterns are

presented, the conclusion about the crystal structure was based on the indexed SAD

patterns. The typical method used for indexing of SAD patterns involves measuring the

distances between sets of diffraction spots, and angles between the lines joining these

spots [218,219]. The ratio of these distances and angles are used to determine the crystal

structure. Since the SAD patterns were calibrated in the present case and only two

crystal structures are involved, their indexing was quite straightforward. The distances

between different sets of spots and angles between different directions were measured.

The distances were compared with the information in crystallographic database (Powder


diffraction file, PDF#00-037-1463 for ZB ZnSe and PDF#00-015-0105 for WZ ZnSe),

and orientation of spots was determined.

3.9 Acquisition and analysis of PL results

The above sections describe the experimental methodology in terms of the arrangement

and application of instruments, and specific procedures that allowed collection of the data

(PL spectra and PL lifetimes). But nothing was said about the details of their acquisition

and later analysis. These include parameters that can be adjusted during the acquisition

of data, and procedures used for normalization of PL spectra and curve-fitting to different

peak functions. This section briefly describes these details, and can be sub-divided into

three sections dealing with (i) data acquisition, (ii) spectral correction and normalization

of PL spectra, and (iii) data analysis using curve-fitting. This discussion is important in

understanding why different PL spectra appear differently, and reasoning behind fitting

them to different peak functions.

3.9.1 Data acquisition

The PL signal is eventually detected and measured by a detector. The most commonly

used detectors include photo-diodes, photomultiplier tubes (PMTs) and CCD based de-

tectors. These detectors differ in their sensitivity, spectral response, cost and usable

lifetimes. All PL spectra described in this thesis were collected using a Si based CCD

detector, and the discussion below is specific for such detectors. The principles in gen-

eral apply to all detection systems, it should be clarified that some of these details apply

slightly differently for each of them. The use of the term ‘detection system’ henceforth

includes all instruments and optical components placed after the sample in the experi-

mental arrangements described earlier. The PL signal from a sample of a semiconductor

material ultimately reflects the quality and characteristics of the sample, but for a given


sample there are three factors that can be easily controlled during the acquisition of PL

spectra and may affect their interpretation. These factors determine the amount of signal

collected, which in turn is related to the spectral resolution and accuracy of the data.

These three factors are discussed next. In this discussion, spectral resolution means the

ability of the detection system to distinguish closely spaced PL lines with linewidths

limited only by the sample, not the instrument. Accuracy of the detection system means

that the individual peaks are recognizable with reasonable signal-to-noise ratios (SNR).

This is important because LTPL spectra from II-VI semiconductors in general contain

several peaks resulting from different optical processes, but their intensities (or signal

strengths) can vary depending upon the contribution of each process to the entire PL

signal. While the stronger peaks are easily recognized, in some cases even without any

fitting, due to their high intensities, the weaker ones with small SNR are not always

reliably interpreted. In literature, these indistinct peaks with small SNR are referred to

as shoulders. In general, both high spectral resolution and high peak intensities cannot

be obtained together, as explained below.

For discussion, the LTPL spectra with high spectral resolution can be termed high-

resolution spectra, distinguished from LTPL spectra with large peak intensities for PL

peaks which can be labelled high-throughput spectra. Different conditions are required

in each case and these can be understood in terms of the three factors mentioned above

and described below.

Integration time ti

The easiest parameter to understand of these is the integration time ti, which is simply the

amount of time over which the signal is collected. Large integration times mean collection

of proportionately more signal which helps to increase the SNR in the spectrum. Weak

PL signal, especially from single NWs, can be collected more reliably with high SNR

using larger ti. However, using large ti has a disadvantage that it tends to reduce the


spectral resolution (i.e. leads to broadening of PL lines) because of the effects of spectral

diffusion [242]. Spectral diffusion is a change in the energy position of a recombination

line due to random fluctuations in the environment of the related recombination centre.

These small spectral shifts are related to the fluctuating local electric fields produced by

the movement of charge carriers in the vicinity of the recombination centres. Thus, high-

resolution LTPL spectra are collected with as small ti as possible to obtain narrowest

possible linewidth of the PL peak of interest which should also be clearly resolved. This

also means that the other peaks in the spectrum (especially the weakest ones) might

not be distinctly recorded as they will be merged with the background noise and appear

only as a shoulder. Further, in general, a change in ti enhances or reduces the signal

proportionately. For example, a particular peak in a LTPL spectrum acquired with ti=2

seconds will be almost twice as strong when compared to a similar measurement with

ti=1 second. In practice, ti cannot be made arbitrarily large as that would slow down

the experiments by the same factor.

Slit-width dslit

Slit-width (dslit), which denotes the width of the slit or entrance aperture of the spectrom-

eter, can also be varied during the acquisition of LTPL spectra. The spectral resolution

is inversely related to dslit, i.e. as dslit is decreased the resolution is increased. This is

because narrower dslit results in less divergence of the beam that enters the spectrome-

ter. This in turn reduces the dispersion in wavelengths inside the spectrometer. Larger

dslit allows more signal to be collected but the change in signal strength is not directly

proportional to the change in dslit, unlike the case of ti. In fact, the amount of signal

varies as d2slit. Therefore, high-resolution LTPL spectra are acquired with the smallest

dslit that allows a clear peak to be recorded. Similar to the case of varying ti, not all

peaks in LTPL spectra can be recorded under conditions of high-resolution. In the case

of high-throughput spectra, larger dslit are used and different peaks can be recorded with


good SNR, but the PL lines are broader and appear to be merged when they are closely

spaced. In general, dslit cannot be made too large as it was observed that several PL

peaks are merged into one with very large dslit.

Excitation Intensity Iexc

The third factor that affects the spectral resolution and throughput is excitation intensity

Iexc. It is clear that larger number of electron-hole pairs are generated when Iexc is

increased, which in general correspond to an increase in signal strength. However, as

discussed in Appendix C.2, this change in signal strength is not directly proportional to

the change in Iexc, but depends on the optical processes involved. Even if the change

is not linear, signal strength can be increased by increasing Iexc (at least in the low-

injection regime). But this also results in loss of resolution because processes such as

exciton-exciton, exciton-electron, and electron-electron interactions which result from

higher Iexc tend to broaden the PL lines. These are essentially many-body processes

which require specialized treatment and are outside the scope of this thesis. Thus, high-

resolution spectra are acquired using minimum Iexc which also results in some PL peaks

not being clearly recorded. High-throughput spectra, on the other hand, are recorded

with nominal Iexc resulting in PL peaks appearing with a good SNR along with some

broadening of these lines.

Based on the above discussion it can be concluded that different conditions required

for acquisition of LTPL spectra are: (i) high resolution: smallest ti, dslit and Iexc, and

(ii) high-throughput: large ti, and nominal dslit and Iexc. Further, it can also be inferred

that high-resolution LTPL spectra will contain fewer PL peaks with noisier background,

in contrast to high-throughput LTPL spectra with larger number of recognizable peaks

and larger linewidths. It should also be noted that in the literature the intensities of

PL peaks have no actual units, and they are reported in arbitrary units. This can

be understood by noting that different authors use different instruments in collecting


PL spectra, and different conditions of acquisition. There is no direct correspondence

between the intensity recorded by the detection system and actual photon counts, other

than that they are proportional. This correspondence can be established by correcting

the PL spectra by spectral response of the detection system, but this is almost never

done. This is discussed further in the next section (Section 3.9.2).

In general, all LTPL spectra presented in this thesis were acquired with different

combinations of varying ti, dslit and Iexc. Specifically, each spectra was acquired about six

times on an average, only the ones with highest resolution compatible with good SNR are

presented in this thesis. In a few cases spectra are acquired under both conditions (high-

resolution and high-throughput) and this is noted where these spectra are presented.

This was done to record PL peaks with sample-limited linewidths.

3.9.2 Spectral correction and normalization of PL spectra

Spectral correction

The PL spectra presented in this thesis are as-acquired (or raw) spectra; i.e., they were not

averaged or otherwise modified. This is because the averaging procedure almost always

influences the peak positions and linewidths of PL peaks, even if this influence is marginal.

This issue is more important in the case of high-resolution LTPL spectra. For example, a

commonly used method of averaging (also called smoothing) is adjacent-averaging where

each data-point in the spectrum is replaced by an average of user-defined number of data

points around the original data-point. As this user-defined number of data-points, labeled

averaging width, is increased the SNR is improved but is accompanied by a reduction

in peak intensity and broadening of the linewidth. Moreover, since most PL spectra in

this thesis were later fit to peak functions (Gaussian or Lorentzian), averaging the data

makes the assessment of goodness of fit based on residuals difficult because the averaged

noise may be mixed with the actual signal. Also, the spectra have not been corrected for


instrument response. This is a common practice in PL spectroscopy of semiconductors as

noted earlier. The reason behind this is that, in general, the instrument spectral response

function is not known or it is not easily available to perform the correction procedure.

This can be further understood by taking into account the number of instruments and

optical elements (lenses, mirrors etc) in an experimental arrangement. For example, in

a typical µ-PL spectroscopy arrangement, these would include: inside the spectrometer

(i) diffraction or prism grating(s), (ii) reflecting mirrors (minimum 2, identical), (iii)

collimating mirrors (minimum 2, identical), and outside the spectrometer (iv) CCD (or

other) detector, (v) microscope objective(s), (vi) optical fibre (when used), (vii) all optical

elements such as neutral density filters, optical filters, cryostat windows, mirrors, lenses

etc. In practice, each of them has a certain characteristic response in the wavelength

region of interest (except possibly for VND where neutral in its name is due to a flat

response over a wide spectral range). The instrument response function in such situations

is a convolution of spectral response functions for all elements, many of which are not

accurately known. Therefore, data corrected by instrument response is more error-prone

than without. Datta et al. [243] have criticized this practice and attributed quantitative

discrepancies and irreproducibility of luminescence spectra in the literature to this culture

of not correcting for instrumental spectral response.

Normalization of PL spectra

Since the PL intensity in a LTPL spectrum is in arbitrary units, normalizing the PL

spectrum by any arbitrary factor has no effect on the peak positions and linewidths of

individual PL peaks. Clearly, the integrated PL intensity (IPL) is changed as a result

of any normalization. PL spectra presented in this thesis have been normalized in some

cases for clarity, especially when a comparison between two or more spectra is made.

These normalization procedures are summarized below.

1. Normalization to unity implies the entire spectrum is divided by the highest value of


PL intensity in the spectrum. The strongest peak in the resulting spectrum then has a

peak height of one (in arbitrary units).

2. Normalization to integration time ti means the entire spectrum is divided by ti, re-

sulting in a spectrum corresponding to a collection time of one unit of ti (which in most

cases is one second). This is useful when comparing PL spectra acquired with different ti.

As discussed earlier in Section 3.9.1, the change in PL intensity is directly proportional

to the change in ti, so normalization with ti does not affect the characteristics of the

PL spectrum. It was also mentioned therein that change in PL intensity is not directly

proportional to changes in dslit or Iexc. Therefore, no PL spectrum has been normalized

to either dslit or Iexc.

3. In general for two or more PL spectra: (i) when peak intensities are compared, no nor-

malization has been performed, (ii) when peak positions or peak linewidths are compared,

peak intensities (heights) are adjusted to be equal for all PL spectra.

4. In experiments to study dependence of PL on Iexc, Iexc was varied but care was taken

to keep dslit and temperature T fixed, and ti was varied as needed. In graphs comparing

Iexc-dependent PL spectra, normalization to unity has been used. When these PL spectra

were fit to individual peaks, normalization to ti was used.

5. In temperature-dependent PL experiments T was varied but dslit and Iexc were kept

fixed while ti was adjusted as needed. In graphs comparing temperature-dependent PL

spectra, normalization to ti has been used. Normalization to ti was also used when these

PL spectra were fit to individual peaks. (The fits to Iexc-dependent and T -dependent PL

spectra are not shown because there are enormously large in number. These are typically

never shown, only the obtained parameters of the fits are presented).


3.9.3 Data analysis using curve-fitting

The peak positions in a LTPL spectra can be determined in two ways: (i) by manually

finding the peak positions from the spectrum (e.g., using a graphing program or image-

processing program), and (ii) by fitting the peaks to different peak functions. Both

these methods have been used in the literature. For example, in the case of LTPL from

ZnSe NWs, ther peak positions in most reports were reported without fitting to any

peak functions [55,103,105,106,108,109,111,113,116–120]. Ohno et al. [105] studied the

dependence of linewidth of excitonic emission on temperature but did not mention any

fitting of PL spectra to obtain the linewidths. It should be assumed that they used a

fitting procedure which was not mentioned. When fitting of PL spectra to peak functions

is used, either a Gaussian or a Lorentzian lineshape is used depending on the broadening

mechanism involved. These can be understood as follows.

The lineshape of free and bound excitons is described by a Gaussian or a Lorentzian

lineshape if they are due to inhomogeneous and homogeneous broadening, respectively [244].

A Gaussian lineshape is expected when the linewidth broadening is a result of crystal

imperfections, impurities or strain, called inhomogeneous broadening. In these cases, the

broadening is dominated by the randomly distributed local variations of the associated

electronic states in the sample. The transition energy of these states vary from one point

to another in the sample resulting in observed broadening. Lorentzian lineshape is ex-

pected when the broadening mechanism is homogeneous broadening related to the lifetime

of the associated excited states. Further, in processes where exciton-phonon interactions

are involved, a Gaussian lineshape is obtained for strong exciton-phonon interaction, and

it is Lorentzian for a weak exciton-phonon interaction. This is because in the case of

strong interactions, the lifetime of every individual exciton created by the absorption of

light is reduced by collisions with phonons and result in lifetime broadening (Gaussian).

In the case of weak interactions, the lifetimes are not significantly affected and homoge-

neous broadening dominates (Lorentzian). It should be pointed out that linewidths due


to homogeneous broadening are rarely observed in practise, and this is evident by the

rather large linewidths reported for excitonic emission in NWs. The reported linewidths

of excitonic emissions from NWs ∼1 meV (discussed in Section 2.3.2) are not entirely

due to homogeneous broadening, but also include contributions of inhomogeneous broad-

ening. This is because crystals and crystalline NWs almost always contain impurities

(intrinsic or extrinsic), and often also contain structural imperfections. In general, the

broadening of linewidths includes contributions from both homogeneous and inhomoge-

neous broadening mechanisms, and it is not possible to accurately determine whether

the broadening observed is entirely homogeneous or inhomogeneous. Because of this, in

the use of peak functions to fit the PL spectra, both Gaussian [104, 107, 217, 245–247]

and Lorentzian [148, 231, 248, 249] lineshapes have been used by different authors to fit

excitonic PL lines. The use of a Gaussian lineshape by Brandt et al. [177] to fit an

exciton peak (linewidth <300 µeV) is debatable in that a Lorentizan lineshape may have

been a better choice. This is because a linewidth of <300 µeV is quite likely to be a

result of purely homogeneous broadening. It needs to be emphasized that the use of one

or the other peak function for curve-fitting in the literature does not result in discrep-

ancies. This is because a fitting procedure using peak functions is only used to obtain

peak positions and linewidths more accurately, and the error in such fitting is much less

than the experimental errors. In other words, the error in a particular peak position (or

linewidth) obtained by using different peak functions (Gaussian or Lorentzian) is more

likely to be due to experimental error rather than the use of different functions for fitting.

(It is clear that the choice of peak function is critically important when study is related

to the broadening mechanisms or exciton-phonon interactions).

A choice of appropriate peak function for curve-fitting can be made by using both

functions at first to fit the curves, and then making a choice based on the goodness of fit in

each case. This was done, for example, by Chen at el. [244]. Similarly, for the PL spectra

presented in this thesis, both function were used first (for a few spectra only). It was


found that the fitting method using Lorentizan lineshapes reproduces the experimental

spectra better than that using Gaussian lineshapes. This points to a greater contribution

of homogeneous broadening in the linewidths and weak exciton-phonon interactions. A

Lorentzian lineshape was therefore used to fit PL spectra in the near band-edge NBE

region (excitonic emissions are located in this region). Further, almost all PL spectra

presented in this contain multiple peaks, the number of which is not always obvious. In

the fitting procedure, the multi-peak spectra is first fit to the number of peaks which are

obvious from the PL spectrum. This number is increased gradually until the χ2 error is

minimized. Statistically, minimization of χ2 means finding the values of the fit parameters

such that the sum of the squares of the deviations of the theoretical curve(s) from the

experimental data points for a range of independent variables are minimum. This is a

standard procedure for fitting of multi-peak spectra [250]. The deep-level (DL) emission

appear as broad symmetrical bands, and they are described by a Gaussian lineshape. A

Gaussian lineshape was used for fitting of DL emission band(s) in the PL spectra. It is a

customary practice to note the function used for curve-fitting and the number of peaks

obtained from such a fitting procedure along with the results. This has been followed in

describing the results in this thesis.

3.10 Summary

In this chapter the experimental methodology was described in detail. The entire ex-

perimental procedure consists of two optical characterization techniques namely, LTPL

spectroscopy and TRPL, and one structural characterization technique of TEM (includ-

ing HR-TEM and SAD). First the rationale behind using these three characterization

techniques was described. The advantages and limitations of PL and TRPL methods

were also discussed. This was followed by a brief description of the growth method used

for the synthesis of ZnSe NTSLs and NWs. The technical difficulties involved in a com-


bined use of three characterization techniques on the same individual NTSL and NWs

were described in detail. The lack of any experimental work so far that describes the use

of these three characterization methods on the same NW (independent of the material)

was also emphasized. A new methodology was proposed based on the design of a cost-

effective sample-holder, and implementation of confocal µ-PL spectroscopy and TRPL

techniques on a purpose-built confocal optical microscope setup. The experimental ar-

rangement for PL spectroscopy was described which was used for the characterization

of array of ZnSe NWs. The limitations of this experimental arrangement were also

pointed out. The experimental arrangements designed for confocal µ-PL spectroscopy

and TRPL were described in detail such that the experiments can be reproduced by

others. The structural characterization of ZnSe NTSLs and NWs using TEM was also

briefly discussed. The experimental conditions used in the acquisition of LTPL spectra

were discussed. The variations in signal-to-noise ratios (SNR) in different spectra were

related to integration times (ti), slit-width (dslit) and excitation intensity (Iexc). Nor-

malization procedures used for normalizing the PL spectra were described, while the use

of peak functions in curve-fitting was discussed in relation to the analysis of PL spec-

tra. The experimental methodology described in this chapter may be considered as an

important contribution since it can be utilized to study individual NWs (homogeneous,

heterostructures, NTSLs) of any material by combined luminescence spectroscopy and

electron microscopy. Further, even more characterization techniques (e.g. absorption

and Raman spectroscopy) can also be realized in this experimental arrangement.

All characterization experiments described in this thesis were designed by the author.

All the LTPL and TRPL experiments presented in this thesis were performed by the

author. Structural characterization utilizing TEM and HR-TEM was performed at an

outside facility (Canadian Centre of Electron Microscopy) dedicated for research using

electron microscopy techniques. These were done with the assistance of technical staff at

this facility, the inputs specific to the experiments were those of the author. Exceptions


to the authors’s contributions in experimental work are as follows. The growth of ZnSe

NWs studied in this thesis was performed in collaboration with other researchers at the

author’s facility. The electronic structure calculations of NTSLs, which were helpful

in validating the experimental results were carried out in collaboration at the author’s

facility.

Chapter 4

Influence of stoichiometry on optical

response

4.1 Introduction

It has been discussed in detail in Chapter 1 how deviations from stoichiometry in ZnSe

single crystals and films influence their optical and electronic properties. Details on the

procedures for heat treatment and earlier observations on stoichiometric effects were also

presented therein. In Chapter 2 it was explained that the role of native point defects on

the optical response of ZnSe NWs has not been studied. This chapter presents the results

on the detailed PL characterization and their systematic analysis for ZnSe NWs. PL from

vapor-phase grown undoped ZnSe NWs grown under excess Zn and Se growth conditions

was studied. In particular, the dependence of PL on excitation intensity (Iexc) for ZnSe

NWs grown under the conditions of excess Zn was analyzed. Luminescence spectra are

characterized by strong near band-edge luminescence with negligible deep-level emission.

PL spectrum from ZnSe NWs grown under the conditions of excess Se shows contrasting

features, and is dominated by deep-level emission with negligible emission in the near

band-edge region. Henceforth, the use of term NWs implies ZnSe NWs, unless noted

82

Chapter 4. Influence of stoichiometry on optical response 83

otherwise.

4.2 Results

In this section the results for PL measurements on ZnSe NWs are presented, followed by

their discussion in the following section1.

4.2.1 Low-temperature PL spectra

1 . 6 1 . 8 2 . 0 2 . 2 2 . 4 2 . 6 2 . 8 2 . 3 2 . 4 2 . 5 2 . 6 2 . 7 2 . 8

P L : Z n S e N W s T e m p . = 3 . 1 K I e x c = 5 W / c m 2

( a ) ( b )

E n e r g y ( e V )

PL In

tensity

(arb.

units)

A

E

DC

P L : Z n S e N W s I n d i v i d u a l p e a k s S u m o f i n d i v i d u a l p e a k s

E n e r g y ( e V )

B

Figure 4.1: (a) LTPL spectrum from ZnSe NWs grown under excess Zn growth conditions.

The PL was taken at 3.1 K with average Iexc=5 W/cm2. (b) Near band-edge region of

the spectrum shown in (a), showing three peaks and a shoulder near the lowest energy

peak. These peaks can be resolved into five different emission peaks.

Figure 4.1 (a) shows the LTPL spectrum for excess Zn NWs at liquid He temperature

using average Iexc=5 W/cm2. The LTPL spectrum shows dominant emission in the

1Contents of this chapter partially appear in Ref. [112] (author’s work). Reprinted with permissionfrom Ref. [112] © 2008 American Institute of Physics.


1 . 6 1 . 8 2 . 0 2 . 2 2 . 4 2 . 6 2 . 8

PL In

tensity

(arb.

units)

E n e r g y ( e V )

P L Z n S e N W s : E x c e s s S e S u m o f i n d i v i d u a l p e a k s I n d i v i d u a l p e a k s

( a )

2 . 5 0 2 . 5 5 2 . 6 0 2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5

P L Z n S e N W s : E x c e s s S e S u m o f i n d i v i d u a l p e a k s I n d i v i d u a l p e a k s

( N B E r e g i o n )

( b )

E n e r g y ( e V )

Figure 4.2: (a) LTPL spectrum from ZnSe NWs grown under excess Se growth conditions.

The PL was taken at 3.1 K with average Iexc=2.5 W/cm2. (b) Near band-edge region of

the spectrum shown in (a), showing a broad peak. This peak can be resolved to show

two different emission peaks.

near band-edge region, with three peaks on the high-energy side of the spectrum, and

a shoulder near the lowest energy peak. This was accompanied by a very weak broad

emission at lower energies. For the purposes of this chapter, the focus will only be on

the near band-edge part of the spectrum. The spectrum can be fitted to five different

emission peaks with Lorentzian lineshapes, as shown in the Figure 4.1 (b). The peaks in

the near band-edge region were centred at (A) 2.794 eV, (B) 2.714 eV, (C) 2.686 eV, (D)

2.603 eV, and (E) 2.580 eV, respectively (see Figure 4.1 (b)). Details of these peaks in the

PL spectrum will be discussed below. The PL spectrum from excess Se NWs taken at 3.1

K with an average Iexc=2.5 W/cm2 is presented in the Figure 4.2. The LTPL spectrum

shows two broad bands, shown in Figure 4.2 (a), at 2.27 eV and 1.96 eV. The higher

energy side of the spectrum has relatively weak emission peaks with energy positions of

2.748 eV and 2.714 eV, as shown in the Figure 4.2 (b).


4.2.2 Excitation intensity dependence of PL spectra

The Iexc dependence of the PL spectrum of a semiconductor can be analyzed to yield

detailed information on the radiative recombination mechanisms. In particular, the type

of transition and the ionization energies of donor and acceptor impurities for a donor-

acceptor pair (DAP) transition can be determined. The variation of integrated intensity

of PL emission band with Iexc gives information about the type of radiative transi-

tion [251] (although this information is not exhaustive as will be described below). The

dependence of peak energy of a DAP emission peak on Iexc provides information about

the ionization energies of the involved donor and acceptor levels introduced by impurities

and intrinsic defects.

PL measurements were made at varying Iexc and all spectra for Iexc-dependent PL

were taken at 3.1 K using an excitation wavelength of 391 nm (3.17 eV). The laser

Iexc was varied between 50 mW/cm2 and 25 W/cm2 using a neutral density filter. The

parameters being studied as a function of laser Iexc are integrated PL intensity (IPL)

and peak energy (Ep) for individual emission peaks. Figure 4.3 shows the PL spectra

for different Iexc in the near band-edge region of the spectrum. Some spectra taken at

intermediate values of Iexc have been omitted for clarity. The PL spectra at all Iexc show

three peaks as described above. Each spectrum was fitted to five different emission peaks

and the fit parameters (Ep and IPL) were used for the analysis described below. The

theoretical models used in this analysis are described briefly here, while further details

on them are given in the Appendix C.2.1.

The variation of IPL with laser Iexc can be used to designate the underlying recombina-

tion processes and to evaluate the crystal quality from that information. This evaluation

involves the observation of PL peaks related to bound excitons and DAP-related transi-

tions in the near band-edge region of the LTPL spectrum. Schmidt et al. [251] considered

the different transition mechanisms involved in the recombination processes in a semicon-

ductor material. They analytically solved a set of simultaneous rate equations for these


2 . 3 2 . 4 2 . 5 2 . 6 2 . 7 2 . 8

2 5 W / c m 2

1 0 W / c m 2

5 W / c m 2

0 . 5 W / c m 2

0 . 0 5 W / c m 2

T e m p . = 3 . 1 K

PL In

tensity

(arb.

units)

E n e r g y ( e V )

Figure 4.3: Dependence of PL from ZnSe NWs grown with excess Zn on Iexc. The average

Iexc was varied from 50 mW/cm2 to 25 W/cm2. All spectra were recorded at 3.1 K.

processes with assumptions on the excitation, as well as on radiative and non-radiative

recombinations mechanisms. They showed that the variation of IPL with Iexc can be

expressed by a power law relation:

IPL ∝ Ipexc (4.1)

where p is a coefficient. They ascertained that for free-exciton and bound-exciton emission

the value of p lies between 1 and 2, while p < 1 for free-to-bound and donor-acceptor pair


recombinations. [251]. These values of p can also be understood as follows. Since the

DAP-related transitions are limited by the density of defects, their PL intensity saturates

as Iexc is increased. Since excitons form independently of the defects, their p values are

larger than 1.

Figure 4.4 shows a log(IPL) − log(Iexc) plot for all the peaks shown in Figure 4.1

(b), and for all Iexc used in the experiments (some can be seen in Figure 4.3). Based

on the model described above given by Schmidt et al. [251], p was found to be 1.19 for

emission peak A (2.794 eV), indicating that this is an excitonic emission. Using the

band-gap of ZnSe (Eg = 2.822 at 4 K) [155, 237], and the binding energy of the free

exciton (EFX = 21 meV) [252], the binding (localization) energy of this exciton to the

impurity is found to be EBX = Eg − EFX − Ep = 2.822 − 0.021 − 2.794 = 7 meV. (Ep

refers to the energy position of the peak under consideration). Using the empirical rule

of Halsted and Aven [253], the binding (ionization) energy of the impurity is 35 meV, if

it is a donor, and 70 meV if an acceptor. p values for B (2.714 eV) and C (2.686 eV)

emissions were found to be 0.76 and 0.68, respectively, implying that the recombination

process is either free-to-bound or DAP recombination. It will be seen later that these

are actually DAP-related transitions. The D (2.603 eV) and E (2.580 eV) emission peaks

show a sub-linear dependence on Iexc with p=0.66 and 0.94, respectively.

In LTPL from DAP-related recombinations, and for sufficiently distant donor and

acceptor pairs, the energy of the emitted photon Em is given by [254]:

Em = Eg − (EA + ED)− E(r) (4.2)

where EA and ED are the acceptor and donor ionization energies, E(r) = −e2/εr is the

Coulombic interaction energy of the donor-acceptor pair separated by a distance r, and ε

is the low-frequency dielectric constant. Pair recombination is often comprised of a series

of closely packed lines corresponding to different values of r, and in some cases, like here,

this series of lines results in a broad band. It has been observed that the band maximum

energy (peak energy) shifts to higher energies as the Iexc is increased. For low enough


- 1 . 5 - 1 . 0 - 0 . 5 0 . 0 0 . 5 1 . 0 1 . 5l o g 1 0 ( I e x c ) ( W / c m 2 )

log10

(I PL) (a

.u.)

A ( p = 1 . 1 9 ) B ( p = 0 . 7 6 ) C ( p = 0 . 6 8 ) D ( p = 0 . 6 6 ) E ( p = 0 . 9 4 )

Figure 4.4: Change of integrated PL intensity (IPL) for individual emission peaks shown

in Figure 4.1 (b) with Iexc (log-log scale). The plots have been offset for clarity. The

solid line is a fit to Equation (4.1). See text for more details.

temperatures to suppress thermal ionization of donors and acceptors, the emitted photon

energy can be related to the laser Iexc as [237]:

Iexc = Iexc,0(hνm − hν∞)3

hνB + hν∞ − 2hνmexp

(− 2(hνB − hν∞)

hνm − hν∞

)(4.3)

where hνm(= Em) is the emission band peak energy, hν∞(= Eg − (EA + ED)) is the

photon energy corresponding to a infinitely distant donor-acceptor pair, hνB(= Eg −


(EA + ED) + e2/εRB) is the emitted photon energy of a donor-acceptor pair separated

by a shallow impurity Bohr radius (RB), and Iexc,0 is a constant of proportionality. The

values of hν∞, hνB and Iexc,0 are determined using a non-linear least square fit to the

experimental data. As the laser Iexc is increased, closely-lying donor-acceptor pairs are

favoured, and the transition energy moves to higher energy according to Equation (4.2).

This relation assumes that the generation rate of neutral pairs and their annihilation by

recombination rate are equal. Given that pulsed excitation was used in measurements

described in this chapter, an assumption has been made that the relation holds for the

experiments described here. Since the direction of energy shift is opposite to that due to

the temperature dependence of the Eg, this method provides an easy way to identify the

donor-acceptor pairs. Since the peak energy for peaks B (2.714 eV) and C (2.686 eV)

increases with increasing Iexc, these are associated with DAP-related transitions, rather

than to free-to-bound transitions.

Figures 4.5 and 4.6 show the dependence of peak energy on Iexc for peak B (Ep=2.714

eV) and peak C (Ep=2.686 eV), respectively. The solid lines correspond to the Equa-

tion 4.3 above. The peaks B and C shift to higher energy as the Iexc is increased, and

Equation 4.3 can be used to model this behavior. The value of the fitting parameter

Iexc,0 includes a constant of proportionality factor, and hence is irrelevant to the dis-

cussion here. The fit parameters and their detailed analysis are presented in the next

section.

4.3 Discussion

Previously, Roppischer et al. [191, 192] studied the intrinsic defects in ZnSe crystals

following annealing in Zn and Se rich atmospheres. In their earlier work [191], they

reported that after Zn-treatment, a new line I∗2 appears in the PL spectrum, and ascribed

it to the intrinsic donor V xSe (neutral Se vacancy) with a binding energy of 2 meV.


2 . 7 0 6 2 . 7 0 7 2 . 7 0 8 2 . 7 0 9 2 . 7 1 0 2 . 7 1 1 2 . 7 1 2

0

5

1 0

1 5

2 0

2 5I ex

c (W/cm

2 )

P e a k E n e r g y ( e V )

Figure 4.5: Variation of peak energy with Iexc for emission peak B at 2.714 eV (shown

in Figure 4.1 (b)). The solid line is a fit to the Equation 4.3.

The binding energy of the donor was reported to be 10 meV below the bottom of the

conduction band. In their work [191], the samples were excited using the UV-filtered

high-pressure mercury lamp. In their later work [192], where the samples were excited

using the laser radiation, they did not observe this line. Instead, they observed another

line (referred to by them as “D-line”) at 443.8 nm (2.7937 eV), and they related this

to the exciton-donor interaction. Since this line is at a similar energy position as the

A (2.794 eV) line, and based on the other evidence reported below, this 443.8 nm line


2 . 6 7 8 2 . 6 7 9 2 . 6 8 0 2 . 6 8 1 2 . 6 8 2 2 . 6 8 3 2 . 6 8 4 2 . 6 8 5 2 . 6 8 6

0

5

1 0

1 5

2 0

2 5I ex

c (W/cm

2 )

P e a k E n e r g y ( e V )

Figure 4.6: Variation of peak energy with Iexc for emission peak C at 2.686 eV (shown

in Figure 4.1 (b)). The solid line is a fit to the Equation 4.3.

observed by them could actually be the A (2.794 eV) line due to an exciton bound to a

neutral donor, and not the “D-line” as assigned by them. In fact, it is conjectured that

the I∗2 line in Reference [191] and the “D-line” in Reference [192] are the same.

Jeong et al. [208, 255, 256] have done extensive work on the ZnSe epilayers grown

by hot-wall epitaxy and have shown the influence of over-pressure of the constituent

components on the optical properties of ZnSe epilayers. They have observed a peak

at 2.794 eV in their samples, similar to the present results, and have related it to the


bound excitons at a neutral donor (D0X). Even though they have observed the peak

at same energy position in the spectrum as reported here, they found different binding

energies for the bound exciton and the donor because of the different values of the Eg

and exciton binding energy that they were using. They assigned the peak to excitons

bound to VSe at a neutral donor, and have found EBX = 14 meV, and the binding energy

to be ED = 70 meV. In the present case, the peak at 2.794 eV is assigned to the same

origin, corresponding to EBX = 7 meV, and a donor binding energy of ED = 35 meV.

This appears reasonable as it was shown by Morimoto [174] that in their Zn-rich films,

the Se-vacancy (VSe) is the main donor. Also, based on their analysis, Roppischer have

ruled out the possibility of Zni as the responsible centre [191].

Two donor-acceptor pair related recombinations were observed in presently studied

samples at 2.714 eV and 2.686 eV. These observed DA pair bands are similar to those

identified earlier by Dean and Merz [155] (labelled as R and Q bands, respectively), and

also by Swaminathan and Greene [159]. According to Iida [257] and Dean and Merz [155],

the increase in energy is 2.0 meV and 3.5 meV per decade of increase in Iexc, based on

their PL measurements, while Ikeda et al. [258] have reported 1.13 meV increase per

decade increase in electrical injection. In ZnSe NW samples studied here, it was noticed

that the change in energy is smaller at weak Iexc, and increases as the Iexc is increased.

This change is ≈1 meV at the lowest Iexc to ≈4 meV at the highest Iexc. Table 4.1 lists

the fit parameters obtained from a fit of experimental data to Equation 4.3. In Table 4.1,

EB = hνB − hν∞(= e2/εRB). As above, the value of the band-gap of ZnSe is Eg = 2.822

eV [155, 237] and an exciton binding energy of EFX = 21 meV [252] was used for these

calculations.

Zacks and Halperin [237] calculated hν∞ = 2.691 eV and hνB = 2.750 eV, and the

values of hν∞ = 2.693 eV and hνB = 2.746 eV found here are in very good agreement

with their values. They used the model they developed [237] (Equation 4.3) to analyze

the experimental data for the R band spectra of Dean and Merz [155]. Also, the present


Table 4.1: Table showing the values of different fitting parameters of Equation 4.3

Ep (eV) hν∞ (eV) hνB (eV) ED + EA (meV) EB (meV)

2.714 2.693 2.746 129 53

2.686 2.656 2.739 166 83

value for the sum of ionization energies of the donor and acceptor, for emission band B

(2.714 eV) of 129 meV, is very similar to their value of 131 meV. Dean and Merz [155]

themselves gave a value of EA + ED of 141 meV, with hν∞ = 2.681 eV. Iida [257] has

given a value of ED = 26 ± 3 meV and EA = 100 ± 1 meV, giving ED + EA of 126 ± 4

meV, which is similar to the value of 129 meV reported here. Ikeda et al. [258] have

proposed a value of ED + EA > 137 meV, which is not in agreement with the values

found by Zacks and Halperin [237], and Iida [257], but agrees well with the value found

by Dean and Merz [155].

An estimate of the shallow hydrogenic donor energy can be made by using ED =

(13.6/ε)(RH/RB) [237], where RH = 0.529 A is the Bohr radius for hydrogen and RB in

turn can be obtained from EB. For peak B, EB = 53 meV gives RB of 31.3 A, and for

peak C, EB = 83 meV gives RB = 20.0 A. According to the hydrogenic model, donor

binding energy ED = (m∗ee4/2~2ε2) = (13.6/ε2)(m∗e/m0), where (m∗e/m0) is the relative

effective mass of electron at the donor. Using (m∗e/m0) = 0.17 [252,259] and ε = 8.66 [252]

at liquid He temperature, ED = 30.8 meV. Similarly, the hydrogenic acceptor binding

energy is 136 meV using m∗h/m0 = 0.75 [252]. For peak B at 2.714 eV, ED = 26.5 meV,

or within the accuracy of experiments, 27 ± 1 meV, which is close to the hydrogenic

energy of 30.8 meV. Also, the present value of ED = 27± 1 meV is similar to that found

by Iida [257] (26± 1 meV) and Zacks and Halperin [237] (30 meV). The binding energy

of this donor is similar to the binding energies for group-III and group-VII donors, such

as Al(26.3 meV), Cl(26.9 meV), Ga(27.9 mev), In(28.9 meV) and F(29.3 meV) [158]. It

is unlikely that group-III and group-VII elements are involved, as excitonic lines related


to any group-III or group-VII element were not observed. Furthermore, post-growth

annealing treatments on NW samples similar to the ones studied here have shown [72]

that the near band-edge luminescence at room temperature almost completely disappears

when the NWs were annealed in a Se-atmosphere. The near band-edge luminescence is

recovered when the samples are annealed in a Zn-atmosphere. This suggests that the

donor can not be a group-III or group-VII element, as annealing in a Se-atmosphere

would not change the luminescence due to this donor. It is proposed that the shallow

donor is a simple native defect of Zn or Se, or a complex defect of the native defects.

Recently, Ivanova et al. [260] have reported on the interaction of intrinsic defects with Al

donors, and it is a possibility that in present ZnSe NWs, the donors and acceptors are

related to the complex defects of native defects with impurities. The acceptor binding

energy is 102 meV, which is again in good agreement with the values found by Iida [257]

(100 ± 1 meV) and Zacks and Halperin [237] (101 meV). The present value of acceptor

binding energy agrees well with the values reported in literature [7].

For peak C at 2.686 eV, RB = 20.0 A giving ED = 41.5 meV, which does not

match the shallow donor hydrogenic energy. The value for the acceptor energy is thus

EA = 124.5 eV. However, the value of ED +EA = 166 meV is similar to the value found

by Dean and Merz [155] for their Q band (150 meV). One possibility is that the acceptor

is a hydrogenic impurity, while the other possibility is that neither of donor and acceptor

are hydrogenic impurities. The value of acceptor energy can be found by measuring PL

at higher temperatures, since the DA pair band evolves towards the free-to-bound (FB)

recombination as the temperature is increased. The energy for the emitted photon in a

FB transition, EFB, is given by [261]

EFB = Eg − EA + Ek (4.4)

where Ek = kBT is the kinetic energy of the free electron. kB is the Boltzmann constant

and T is the temperature. PL measurements were performed at 40 K and 75 K, and

EFB was found to be EFB/40K = 2.682 eV and EFB/75K = 2.679 eV. Ek corresponding to


these temperatures is Ek/40K = 3.4 meV and Ek/75K = 6.5 meV, giving EA/40K = 139.4

meV and EA/75K = 138.5 meV. Eg at 40 K and 75 K were found using Varshni’s [262]

semiempirical relation with α = 7.3 × 10−4 eV/K and β = 295 K [263]. These results

indicate that the acceptor binding energy for band C is about 139 meV, which gives

a donor binding energy of 27 meV. This value of EA is close to the shallow acceptor

hydrogenic energy, and the value of ED is also in very good agreement with the shallow

donor hydrogenic energy. Since the EA values agree for two different measurements at

different temperatures, it is believed that the correct values are EA = 139 meV and

ED = 27 meV. The donor and acceptor must thus be assigned again to the defects,

simple or complex, of the native defects of the crystal, based on the earlier arguments.

Morimoto [264] have observed two FB peaks at 2.698 eV and 2.680 eV in PL at 77 K of

their MOCVD grown ZnSe samples, and have reported EA values of 110-114 meV and

128-130 meV, respectively. It is suggested that these two FB peaks are actually bands B

and C, respectively, observed in ZnSe NW samples studied here, and the reported values

of EA are similar to their values.

The peak D at 2.603 eV was first observed by Dean [163] in their MOCVD grown ZnSe

samples. They related this peak to the recombination at extended defects, where strong

delocalization of binding potential takes place. This perturbation in the potential due

to extended defects is associated with the emission at 2.605 eV, labelled as Y-line [163].

Taguchi et al. [265] observed the Y-line at 2.61 eV and its phonon replica at 2.58 eV. In

the present results, the Y-line emission at 2.603 eV would give the energy of first-phonon

replica to be 2.572 eV. The error in peak fitting procedures for peaks A, B, C and D is

about 0.1 meV, while this error for peak E is 4.8 meV because of its weak intensity. Peak

E appears only as a shoulder at the low-energy side of the spectrum in Figure 4.1 (b),

and it is reasonable to think that this is the first-LO phonon replica of the Y-line peak D.

Y-line emission has been discussed in the literature on ZnSe epilayers grown by different

methods [266–270], and recently in NWs as well [103, 106]. It has been suggested [267]


that this line is strong in intensity for samples with low background impurities and high

density of structural defects. In the present samples, the strength of Y-line with respect

to the donor-bound exciton line (RY = ID0X/IY ) changes with Iexc from 0.1 to 3.57, and

follows a super-linear dependence with the Iexc with a scaling coefficient of 1.19. ID0X

and IY are the integrated PL intensities of the D0X emission and the Y-line, respectively.

In present samples, it is believed that the concentration of impurities is very low, so that

the cause of Y-line emission would be the structural defects. It is also to be noted that

this emission is not seen from all areas of the sample, but only at certain areas. This

indicates that the sample is inhomogeneous, and that the structural defects are localized

at certain areas only, i.e. not all NWs. The NWs themselves are of good crystalline

quality [72,77,81], but there can be sources of these defects in the samples. For example,

the residue (or precipitates) deposited on the substrate surface can consist of structural

defects. Also, it is quite possible that in the ensemble of NWs, there is a distribution

of NWs with different crystalline quality or different concentration of structural defects

that give rise to the Y-line emission.

The LTPL spectrum for the NWs grown in conditions of excess Se (shown in Figure 4.2

(a)) shows features in contrast to those in the LTPL spectrum from NWs grown in the

conditions of excess Zn (shown in Figures 4.1 (a,b)). The LTPL spectrum from excess

Se NWs show that the PL is dominated by deep-level (DL) luminescence at 2.27 eV and

1.96 eV, while no exciton-related peaks were observed in the near band-edge region of

the spectrum (Figure 4.2 (b)). In earlier works by Roppischer et al. [191], they found

that when ZnSe crystals were annealed in a Se atmosphere, the donor-bound exciton

peaks related to VSe disappeared and new exciton-related peaks related to VZn appeared.

These VZn-related recombination lines were associated with the excitons bound to deep

neutral acceptors where VZn acts as an acceptor. The DL emission can be related to

the vacancies of Zn and this assignment can be understood as follows. For ZnSe NWs

prepared under excess Zn, the concentration of VZn is low, while those of VSe (which act


as donors) is higher, leading to a donor-bound exciton line. A low intensity of the DL

emissions in this case shows that VSe donors do not participate in these broad bands. In

NWs with excess Se, the concentration of VZn is higher and that of VSe is low. This results

in a strong intensity of the exciton lines related to VZn-acceptors, while the VSe-related

emission gets weaker or completely disappears [191]. In earlier works [271,272], the VZn-

related acceptor bound emission was found to be accompanied by the DL emissions. In

the present case however, the VZn-related exciton emission was not observed in NWs with

excess Se while strong DL emissions were seen. This is attributed to the participation of

VZn in the DL emissions. The absence of VZn-related exciton emission can be ascribed

to the un-optimum growth of NWs in conditions of excess Se, which leads to a poor

crystalline quality. Philipose et al [77, 78] have shown that NWs grown in conditions of

excess Se contain a high density of structural defects such as stacking faults.

4.4 Conclusions

In this chapter the results were presented on the PL from ZnSe NWs synthesized using

vapor-phase growth method under conditions of excess Zn and Se. Iexc-dependence of

PL from NWs grown under the conditions of excess Zn was analyzed in detail. A strong

peak related to an exciton bound to a neutral donor at VSe was observed at 2.794 eV.

The binding energy of the exciton to the neutral donor was found to be 7 meV, and

the binding energy of the neutral donor was found to be 35 meV. Two peaks related to

donor-acceptor pair recombination were observed at 2.714 eV and 2.686 eV. The binding

energies of both the donors were 27 ± 1 meV, while those of the acceptors were 102.5

and 139 meV respectively. These donors and acceptors were assigned to the complexes

formed by single native defects such as VSe, VZn, Sei and Zni with other native defects

or foreign impurities. Y-line emission and its LO-phonon replica were also observed. The

LTPL spectrum from ZnSe NWs grown under conditions of excess Se was also compared


with that from NWs grown under excess of Zn. It was found that the donor-bound

exciton line related to VSe in NWs with excess Zn was absent in the NWs with excess Se,

and no other exciton-related line was observed. The PL spectrum from NWs with excess

Se was dominated by deep-level emissions at 2.27 eV and 1.96 eV. These were related to

the presence of VZn in the NWs.

Chapter 5

Single zinc-blende ZnSe nanowires

5.1 Introduction

As discussed in Chapters 1 and 2, considerable effort has been devoted to the growth

of ZnSe NWs and understanding their fundamental optical properties. Despite this, de-

tailed optical characterization in terms of the identification of recombination centres in

ZnSe NWs is still lacking. Reports on the growth of ZnSe NWs usually include their

optical characterization using RTPL, mainly as a preliminary indicator of their optical

quality. There has been limited progress in understanding the optical response of ZnSe

NWs, particularly on individual NWs, using LTPL which can provide the information

needed to understand the underlying recombination mechanisms. ZnSe primarily crys-

tallizes in the ZB crystal structure, and ZB NWs obtained using VLS growth mechanism

are single crystalline. ZnSe NWs also form in the WZ crystal structure and in many

cases co-exist with the ZB NWs in the same as-grown samples. It was also discussed in

Chapter 2 that ZnSe NWs usually occur with other 1D-nanostructures such as nanorib-

bons, nanosaws, nanobelts, etc, in the same sample. RTPL and LTPL characterization

of ZnSe NWs reported in the literature has been on arrays of ZnSe NWs. These ev-

idently include contributions from NWs of different diameters, crystal structures, and

99

Chapter 5. Single zinc-blende ZnSe nanowires 100

other 1D-nanostructures. There has been no report so far correlating the recombination

mechanisms in ZB NWs with their single-crystalline microstructure.

This chapter presents the results on the optical characterization of individual ZB

NWs with a determination of their crystal structure using TEM, HR-TEM and SAD,

and is organised as follows. First, the LTPL spectrum from a single ZB NW is presented,

and the origin of the different peaks observed is discussed. The optical response of other

single ZB NWs measured using LTPL is also presented. The peak positions (Ep) and

linewidths (Γp) (defined as the FWHM) of peaks in the PL spectrum from a single ZB

NW are then compared with those from a bundle of ZB NWs. The identification of the PL

peaks is further supported by using the dependence of PL on temperature and excitation

intensity. Quantitative information on the energy band gap, and recombination lines is

also obtained using the analysis of the dependence of PL on temperature. Details on the

theoretical models used in this analysis are given in the Appendix C.2.2.

5.2 Results and Discussion

5.2.1 Characterization of single ZB ZnSe NWs

Six individual ZB NWs were characterized by both optical and structural characterization

methods. Five of these were selected from the same as-grown ZnSe NW sample. In this

section the structural characterization results are presented on one ZB NW (labelled as

ZB-NW-1). The results on optical characterization using LTPL are presented for ZB-

NW-1 and three other single NWs (labelled ZB-NW-2, 3 and 4). For the determination

of the crystal structure, HR-TEM micrographs were taken at several locations along the

length of the NWs. The SAD patterns were acquired from larger areas at a few locations

along the length of the NWs. It was found that the majority of ZB NWs were defect-free

and single-crystalline.


HR-TEM characterization

Figure 5.1 (a) shows an overview TEM micrograph of ZB-NW-1 on the TEM grid, and

Figure 5.1 (b) shows a HR-TEM micrograph from an area of ZB-NW-1. Figure 5.1 (c)

shows a SAD pattern from an area of ZB-NW-1. The indexed diffraction spots in the SAD

pattern correspond to the ZB crystal structure. Indexed SAD patterns from other areas

of ZB-NW-1 also confirmed its ZB crystal structure. Figure 5.2 (a) shows a HR-TEM

micrograph from another area of ZB-NW-1. The FFT pattern shown in the Figure 5.2

(b) was acquired from the area highlighted (in red) in (a). The indexed spots in the FFT

pattern further highlight that the area selected for acquiring the FFT pattern on the

HR-TEM micrograph of ZB-NW-1 is single crystalline with a ZB crystal structure. The

TEM and HR-TEM micrographs, and indexed SAD and FFT patterns for ZB-NW-2 are

shown in Figures G.1, G.2 and G.3 in Appendix G which confirm a ZB crystal structure

for ZB-NW-2.

Low-temperature PL spectra

The LTPL spectrum (3.5 K) from ZB-NW-1 near the excitonic region is shown in Figure

5.3 acquired under the conditions of high throughput. Iexc and dslit are as indicated in the

Figure 5.3. The inset to Figure 5.3 shows a confocal PL image of ZB-NW-1 under laser

excitation with Iexc=12.7 W/cm2. The LTPL spectrum was fit to individual emission

peaks using Lorentzians, as shown. Two dominant peaks are observed at 2.785 eV and

2.780 eV, and their origin will be discussed first. In the literature on ZnSe single crystals

and epilayers, a recombination line near 2.783 eV has been reported on several occasions,

and labelled as Id1 [255,256,273–283] or Ideep1 [284–291] line. The lines related to excitons

bound to neutral acceptors, have usually been labelled as I1 lines. The Id1 (or Ideep1 ) line

is believed to originate from a deep neutral acceptor. This line has been assigned to two

origins in the past. Roppischer et al. [191] and Satoh and Igaki [271] related the Id1 line

to the vacancies of Zn (VZn), while Dean [272] and Dean et al. [292] suggested that the


ZB-NW-1

ab

c

111

111

220

220

002

002

111

111

Fig

ure

5.1:

Str

uct

ura

lch

arac

teri

zati

onof

ZB

-NW

-1:

(a)

Ove

rvie

wT

EM

mic

rogr

aph,

(b)

HR

-TE

Mm

icro

grap

hfr

oman

area

of

ZB

-NW

-1,

and

(c)

Index

edSA

Dpat

tern

.T

he

index

edsp

ots

corr

esp

ond

toa

ZB

cryst

alst

ruct

ure

.T

he

vie

win

gdir

ecti

onfo

r

(b)

and

(c)

is〈1

10〉.


a b

111

111

220

220

002

002

111

111

Figure 5.2: (a) HR-TEM micrograph from an area of ZB-NW-1, (b) Indexed FFT pattern

taken from the area squared (in red) in (a). The indexed spots correspond to a ZB crystal

structure. The viewing direction corresponds to 〈110〉.

acceptor involved is CuZn (substitutional Cu atom at a Zn site). Huang et al. [286] later

confirmed that both assignments were correct, and that the peak energies of the Id line

due to these two origins differ by 0.07 meV. They used controlled deviations from the

stoichiometry by heat treatments in molten Zn, saturation in an atmosphere of Se, and

low level doping by Cu, to arrive at this conclusion. Isshiki and Masumoto [277] later

confirmed the results of Huang et al. [286] by heat treatment and incorporation of Cu

impurities in ZnSe single crystals.

A procedure that has been used to confirm the role of VZn acceptors in the appearance

of the Id1 line is heat treatment under excess Zn, or excess Se. This was first used by

Roppischer et al. [191] for their assignment of the Id1 line to the VZn. It is to be noted,

that at the time of their experiments, the terminology Id1 or Ideep1 was not in use, and

they labelled this line as the I1 line [191]. This can be understood as follows. Under

conditions of excess Zn (e.g. by heat treatment in molten Zn or higher Zn pressure),


2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5

I 1d’ -3LO

I 1d -3LO I 1d’ -2L

O I 1d -2LO

I 1d’ -1LO

L T P L - Z B - N W - 1 I 1

d l i n e I 1

d ’ l i n e S u m o f i n d i v i d u a l p e a k s

T e m p . = 3 . 5 K I e x c = 1 2 . 7 W / c m 2

d s l i t = 5 0 0 µm

PL In

tensity

(arb.

units)

E n e r g y ( e V )

4 . 5 m e V

I 1d -1LO

FX

Figure 5.3: Excitonic region of the LTPL spectrum (3.5 K) from ZB-NW-1 acquired

under the conditions of high-throughput. (Iexc and dslit are as indicated). The inset

shows a confocal PL image of ZB-NW-1 under laser excitation with an average Iexc=12.7

W/cm2. The diameter of pin-hole in the PL image is 5 µm. The LTPL spectrum was fit

to individual emission peaks using Lorentzians as shown.

the concentration of VZn is reduced, and the intensity of the Id1 line is therefore also

reduced [191, 276, 278–281, 284]. In many cases it was found that the Id1 line completely

disappears after Zn-treatment [191,255,274,275,277,284,286,288]. When the heat treat-

ment is carried out under conditions of excess Se, the concentration of VZn increases,


and results in an increase in the intensity of the Id1 line [191, 271, 280, 282, 284, 288, 289],

or it reappears (if disappeared due to a prior Zn-treatment) [191, 255, 277, 284]. The

disappearance and reappearance of the Id1 line under different heat treatments has been

observed by several authors [191, 255, 277, 280, 284], and based on this observation, they

related the Id1 line to VZn acceptors [191, 277, 280, 284]. Any involvement of CuZn in the

Id1 line has also been effectively ruled out based on these observations. For example,

based on their LTPL experiments on samples annealed in Zn and Se atmospheres, Pohl

et al. [280] concluded that any role of Cu in the Id1 line can be excluded. Several au-

thors have also argued that the Id1 line is due either to VZn or complexes containing VZn

(VZn-complex) [271,280,284]. However, a clear distinction has not yet been established.

It is important to note that a wide range of energy positions of the Id1 line has been

reported in the literature, from 2.779 eV [188] to 2.7858 eV [273]. In some cases the

authors identified the line as the Id1 line without noting the peak position [274,275,277].

This is because of the characteristic strong LO-phonon coupling of the Id1 line. The Id1

line is known to occur with several LO-phonon replicas [191, 255, 274, 278, 279, 284, 286,

287, 289, 290]. Phonon-replicas of up to six orders have been reported for this line [291]

indicative of a strong LO-phonon coupling. Skromme et al. [293], for example, observed

a recombination line at 2.7823 eV, which is common to most reported energy positions of

the Id1 line. But they did not relate it to the Id1 line as they did not see any LO-phonon

replicas of this line in the PL spectrum.

Figures 5.4 (a,b) show the LTPL spectra (3.6 K) from ZB-NW-2 acquired under the

conditions of (a) high-throughput, and (b) high-resolution. The LTPL spectra (3.8 K)

taken using similar conditions from ZB-NW-3 are shown in the Figures 5.5 (a,b). It

is seen that the PL peaks similar to those in the LTPL spectrum from ZB-NW-1 are

observed for both ZB-NW-2 and 3. In the present case, the absence of Cu in ZnSe NWs

(grown using the vapor-phase transport method) was confirmed by their growth under

conditions of excess Zn and Se and subsequent LTPL experiments (see Chapter 4). The


2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5

( b )PL

Inten

sity (a

rb. un

its)

E n e r g y ( e V )

L T P L - Z B - N W - 2 I 1

d l i n e I n d i v i d u a l p e a k s S u m o f i n d i v i d u a l p e a k s

T e m p . = 3 . 6 K I e x c = 1 2 . 7 W / c m 2

d s l i t = 5 0 0 µm

4 . 4 m e V

( a )

2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5

L T P L - Z B - N W - 2 I 1


T e m p . = 3 . 6 K I e x c = 1 2 . 7 W / c m 2

d s l i t = 1 0 0 µm

E n e r g y ( e V )

3 m e V

Figure 5.4: Excitonic region of the LTPL spectrum (3.6 K) acquired from ZB-NW-2

under the conditions of (a) high-throughput, and (b) high-resolution. (Iexc and dslit are

as indicated). The inset in (a) shows a confocal PL image of ZB-NW-2 under laser

excitation with average Iexc=12.7 W/cm2. The LTPL spectra were fit to individual

emission peaks using Lorentzians as shown.

two dominant peaks observed in the LTPL spectrum from ZB-NW-1, 2 and 3 at 2.785

eV and 2.780 eV can be related to VZn acceptors, and acceptors involving complexes

of VZn, respectively. The fact that these are two different peaks, and that the lower

energy peak at 2.780 eV is not somehow related to the 2.785 eV peak is confirmed by

two observations. Firstly, in the LTPL spectrum from ZB-NW-3 (Figure 5.5), these can

be seen as two separate peaks. Secondly, the LO-phonon replicas of these two peaks can

be easily identified in the LTPL spectra shown in Figures 5.3, 5.4 and 5.5. The line with

Ep=2.785 eV is labelled as the Id1 line, while the other line with Ep=2.780 eV is labelled

as the Id′

1 line. The Id1 line is assigned to a neutral acceptor (VZn), while the Id′

1 line is

ascribed to another neutral acceptor related to (VZn-complex).


2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5

PL In

tensity

(arb.

units)

E n e r g y ( e V )

L T P L - Z B - N W - 3 I 1


T e m p . = 3 . 8 K I e x c = 1 2 . 7 W / c m 2

d s l i t = 5 0 0 µm

4 . 5 m e V3 . 8 m e V

( a ) ( b )

2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5

L T P L - Z B - N W - 3 I 1


T e m p . = 3 . 8 K I e x c = 6 . 4 W / c m 2

d s l i t = 1 0 0 µm

E n e r g y ( e V )

Figure 5.5: Excitonic region of the LTPL spectrum (3.8 K) acquired from ZB-NW-3


as indicated). The LTPL spectra were fit to individual emission peaks using Lorentzians

as shown.

The energy position of the LO-phonon replica of the m-th order (Ep,m) is given by

Ep,m = Ep −m~ωLO (5.1)

where Ep is the energy position of the zero-phonon line (ZPL) (m = 0), and ~ωLO is the

LO-phonon energy (∼31 meV in ZnSe [294]). The peaks at 2.755 eV, 2.723 eV and 2.692

are Id1 -1LO, Id1 -2LO and Id1 -3LO, respectively. Similarly, the peaks at 2.748 eV, 2.715

eV and 2.685 eV are Id′

1 -1LO, Id′

1 -2LO and Id′

1 -3LO, respectively. These are indicated by

arrows in Figure 5.3. The calculated intensity ratio of the intensity of the m-th order

LO-phonon replica (Im) to that of the ZPL (I(m=0)) is given by Hopfield’s relation [295]

as:

Im/I(m=0) = Nmph/m! (5.2)

where Nph is the average number of phonons emitted. The experimental and calculated


intensity ratios of the LO-phonon replicas to the ZPL for the Id1 and Id′

1 lines are listed in

the Table 5.1 for ZB-NW-1, 2 and 3. (Table 5.1 also includes these ratios for LTPL from

a bundle of ZB NWs, to be discussed in the next section). The values in the columns

(2) and (4) corresponding to m=1 represent Nph in each case. As mentioned earlier,

several authors have reported on the observation of LO-phonon replicas of the Id1 line,

but their intensity ratios have not been reported in most cases. The values of intensity

ratios reported by Jiang et al. [287] are also included in the Table 5.1 for comparison.

Jiang et al. [287] found Nph of 0.45 for the Id1 line. It may be seen from the Table 5.1

that Nph for the Id1 is larger than that for the Id′

1 line. Also, the Nph are not identical for

the three NWs and the NW bundle which indicates different strengths of the LO-phonon

coupling in these NWs. These values are 0.53 < Nph < 0.7 for the Id1 line, and 0.25

< Nph < 0.42 for the Id′

1 line.

The peak with Ep=2.800 eV is related to the emission of free excitons (FX), and

labelled as FX in the Figure 5.3. The FX emission was also observed by various au-

thors in as-grown ZnSe samples [188,255,284] and in ZnSe samples after heat treatment

in the conditions of excess Zn [188]. This peak has been labelled in many cases as

Ex [255,271,274,275,284]. This assignment of the peak at 2.800 eV to FX is further con-

firmed using the dependence of PL on temperature later in this chapter. These results

indicate that the optical response of ZnSe NWs grown under stoichiometric conditions

(i.e. using stoichiometric ZnSe source and no further heat treatment) is dominated by

recombination related to native point defects (VZn and VZn-complex), and that the un-

intentional impurities play no role. The ZnSe NWs are therefore intrinsically Se-rich (or

Zn-deficient). Further, the Y-line (Ep=2.603 eV) which was observed in the PL from

an array of ZnSe NWs in the as-grown samples (see Chapter 4) was not observed in

the case of single ZB NWs. This can be seen in the Figures 5.3, 5.4 and 5.5 on LTPL

from ZB-NW-1, 2 and 3, respectively. This shows that the origin of the Y-line is due to

structural defects such as dislocations present in the undergrowth of polycrystalline ZnSe


Table 5.1: Experimental and calculated ratios of the intensity of the m-th order LO-

phonon replica (Im) to that of the ZPL (I(m=0)) for the Id1 and Id′

1 lines. The calculated

ratios were found using Equation 5.2. Nph are the values in columns (2) and (4) corre-

sponding to m = 1. The intensity ratios are listed for ZB-NW-1, 2 and 3, and a bundle

of ZB NWs. These ratios reported by Jiang et al. are also listed [287].

Id1 Id′

1

(1) (2) (3) (4)

Sample m Expt. Calculated Expt. Calculated

ZB-NW-1 1 0.525 0.53 0.472 0.42

2 0.170 0.140 0.076 0.084

3 0.021 0.024 0.005 0.012

ZB-NW-2 1 0.620 0.610 0.242 0.28

2 0.182 0.186 0.047 0.039

3 0.061 0.038 0.014 0.004

ZB-NW-3 1 0.692 0.69 0.247 0.25

2 0.244 0.238 0.03 0.031

3 0.085 0.055

ZB-NW Bundle 1 0.755 0.700 0.295 0.350

2 0.252 0.245 0.070 0.061

3 0.058 0.057 0.043 0.007

Ref. [287] 1 0.46 ± 0.02 0.45

2 0.092 ± 0.01 0.101

3 0.015 ± 0.004 0.015

4 0.0020 ± 0.0005 0.0017


2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5

L T P L - Z B - N W - 1 I 1


T e m p . = 3 . 7 K I e x c = 3 . 2 W / c m 2

d s l i t = 1 0 0 µmPL

Inten

sity (a

rb. un

its)

E n e r g y ( e V )

2 . 4 m e V

Figure 5.6: Excitonic region of the LTPL spectrum (3.7 K) from ZB-NW-1 acquired

under the conditions of high-resolution. (Iexc and dslit are as indicated). The inset

shows a confocal PL image of ZB-NW-1 under laser excitation with an average Iexc=12.7

W/cm2. The diameter of pin-hole in the PL image is 5 µm. The LTPL spectrum was fit


on the sample substrate, and is not related to the NWs at all. This is also to be expected

given the defect-free single-crystalline structure of the NWs which was confirmed by the

TEM, HR-TEM and SAD.

The binding energy of the excitons, EBX , related to the Id1 and Id′

1 lines can be

determined using EBX = Eg − EFX − Ep. The Ep for the FX line, equal to Eg − EFX ,

being experimentally known, can be used here. For the Id1 line, EBX = 2.800 − 2.785


eV=15 meV. The ionization energy of the neutral acceptor (EA), using the rule of Halsted

and Aven [253], is then 150 meV. Similarly, for the Id′

1 line, EBX = 2.800− 2.780 eV=20

meV, and EA=200 meV. The Γp of the Id1 line is 4.5 meV in the LTPL spectrum acquired

under the conditions of high-throughput (shown in Figure 5.3). The LTPL spectrum from

ZB-NW-1 acquired under the conditions of high-resolution is shown in Figure 5.6. The

spectrum in this case was fit to six individual emission peaks using Lorentzians. The Γp

of the Id1 line at 2.785 eV is 2.4 meV, which is significantly narrower than that shown in

Figure 5.3 (4.5 meV). It may also be noticed that only the first order LO-phonon replica

of the Id1 and Id′

1 lines can be seen in the Figure 5.6, and the reasons for this have been

explained in detail in Chapter 3 in context of the conditions used for the acquisition of

PL spectra with high-resolution.

5.2.2 Comparison with an ensemble of ZB NWs

A bundle of ZB NWs on the TEM grid was selected using an iris (a variable aperture

pin-hole) for the LTPL experiments. A bundle of NWs is distinguished from an array

of NWs in that the bundle contains far fewer number of NWs than the array. This is

because the area selected on the sample (i.e. the TEM grid) with an iris, and used for

the collection of PL signal is much smaller than that of an array of NWs without the

use of an iris. It is estimated that about 5-10 NWs contribute to the PL signal in this

bundle. This can be seen by comparing the confocal PL image of the NW bundle (shown

in the inset of Figure 5.7) with that of the ZB-NW-1 and 2 (inset of Figures 5.3 and

5.4) (all at the same magnification). Since no structural characterization was possible

on this bundle of NWs lying on the TEM grid bar, it was assumed that it consists of

ZB NWs, possibly including 1D-nanostructures other than NWs such as nanoribbons.

The assumption that the NWs in this bundle are all ZB is reasonable since other NWs

studied from this sample (and from the same TEM grid) were all single-crystalline ZB,

including ZB-NW-2 and ZB-NW-3.


2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5

PL In

tensity

(arb.

units)

E n e r g y ( e V )

L T P L - Z B - N W - B u n d l e I 1


T e m p . = 3 . 9 K I e x c = 3 . 2 W / c m 2

d s l i t = 5 0 0 µm

5 . 3 m e V

2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5

L T P L - Z B - N W - B u n d l e I 1


T e m p . = 3 . 9 K I e x c = 3 . 2 W / c m 2

d s l i t = 1 0 0 µm

E n e r g y ( e V )

4 . 3 m e V

( a )

( b )

Figure 5.7: Excitonic region of the LTPL spectrum (3.9 K) from a bundle of ZB NWs

acquired under the conditions of (a) high-throughput, and (b) high-resolution. (Iexc and

dslit are as indicated). The inset in (a) shows a confocal PL image of the bundle of ZB

NWs under laser excitation with average Iexc=12.7 W/cm2. The LTPL spectra were fit


The excitonic region of the LTPL spectrum (3.9 K) from this bundle of ZB NWs

is shown in Figure 5.7 acquired under the conditions of (a) high-throughput, and (b)

high-resolution. It may be observed that the conditions of acquisition (Iexc and dslit) are

similar to those for ZB-NW-1 (Iexc is different in (a)). Both spectra were fit to individual

emission peaks using Lorentzians, as shown. Inset in Figure 5.7 shows the confocal PL

image of the bundle of NWs selected using an iris under laser excitation with average Iexc

of 3.2 W/cm2. The Ep for the different peaks observed were found to be similar to those

in the LTPL spectra from ZB-NW-1, 2 and 3, and no additional peaks were noticed.

This further confirms the assumption that all NWs in the bundle are single-crystalline

ZB. The Γp of the Id1 line obtained are 5.3 meV and 4.3 meV under the conditions of

high-throughput and high-resolution, respectively. These are larger than the Γp of the


Id1 line in LTPL from ZB-NW-1, 2 and 3 under similar conditions. It was discussed in

Chapter 3 that the Γp observed in the PL spectra from an array of NWs are larger than

those in the PL from single NWs. This is similar to the observation by Wischmeier et

al. [210] where they observed the Γp from an array of ZnO NWs larger by 0.2-0.5 meV

than those from single ZnO NWs for excitonic emission lines. The broadening of the Γp

in the present case is however larger than they reported, i.e. 0.8 and 1.9 meV for the

two conditions (when comparing Γp of the Id1 line in the case of the bundle of ZB NWs

and ZB-NW-1). Wischmeier et al. [210] related the broadening of the exciton emission

lines in PL from an array of NWs to the ensemble averaging of the PL signal over 100

NWs with varying diameters, morphologies and crystalline quality. The same argument

can be used to explain the broadening of the Id1 line in the LTPL from the bundle of ZB

ZnSe NWs. The Γp of the Id1 line in the LTPL spectra from ZB-NW-2 and ZB-NW-3

acquired under the conditions of high-resolution are 3 meV and 3.5 meV, respectively

(see Figures 5.4 and 5.5). These are larger than the Γp of the Id1 line from ZB-NW-1 (2.4

meV under similar conditions). This indicates that different NWs in the sample exhibit

slightly different Γp of the exciton emission lines (Id1 line in this case). Based on this,

another possible explanation for the observed broadening of the Id1 line in PL spectra

from the bundle of NWs could be that the measured Γp is limited by the NW with the

largest Γp of the Id1 line.

However, even narrower emission lines were observed in the PL from other ZnSe

NWs. These were not taken from the same sample as for ZB-NW-1, 2 and 3. Figures 5.8

(a,b) show the LTPL spectra from ZB-NW-4 (whose structural characterization was not

possible because it was on a Si substrate). Both spectra were fit to individual emission

peaks and the Id1 line is shown in the Figures 5.8 (a,b). The Γp of the Id1 lines were found

to be (a) 1.5 and (b) 1.6 meV. The Γp of the excitonic emission lines close to 1 meV

were also observed in the LTPL from single ZnSe nanoribbons. These are the narrowest

Γp’s reported so far for exciton emission lines in ZnSe NWs. These are also significantly


2 . 6 0 2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5

PL In

tensity

(arb.

units)

E n e r g y ( e V )

L T P L - Z B - N W - 4 I 1

d l i n e T e m p . = 4 . 0 K d s l i t = 5 0 µm

1 . 5 m e V

2 . 6 0 2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5

( b )

1 . 6 m e V

L T P L - Z B - N W - 4 I 1

d l i n e T e m p . = 4 . 0 K d s l i t = 1 0 0 µm

E n e r g y ( e V )

( a )

Figure 5.8: Excitonic region of the LTPL spectrum (4.0 K) acquired from ZB-NW-4 with

(a) dslit=50 µm and (b) dslit=100 µm. The LTPL spectra were fit to individual emission

peaks (only the Id1 line is shown in each case). The Γp of the Id1 lines are (a) 1.5 meV

and (b) 1.6 meV.

narrower than the previously reported Γp of 10 meV for the I2 (2.798 eV) and I1 (2.784

eV) lines in ZnSe nanoribbons grown by MOCVD [103]. The growth of semiconductor

NWs using MOCVD (and also MBE) is considered a high-purity growth method which

offers almost complete control over the growth process. These methods yield NWs of

high purity, but are slow and expensive. This is in contrast to the growth of NWs using

CVD method, which is considered a fast and inexpensive growth method with limited

control and resulting in NWs of lower purity and inferior crystalline quality. The fact

that the Γp of the exciton emission lines close to 1 meV are observed in the ZnSe NWs

grown using CVD method (the present case) is remarkable. This also shows that the VLS

growth of ZnSe NWs using CVD method can yield NWs of optical quality comparable

(or even better) to those from expensive methods such as MBE and MOCVD. This has

important technological implications in that the inexpensive CVD growth method can be


used to synthesise NWs with excellent optical quality on a large industrial scale. With

respect to the literature on other II-VI and III-V compound semiconductor NWs, Γp

close to 1-2 meV reported here compare favourably with the narrowest Γp reported (∼1

meV) [211–216]. (See also Section 2.3.2).

5.2.3 Dependence of PL on temperature

PL from ZB-NW-3 was recorded as a function of the sample temperature. Figure 5.9

shows the change in PL as the temperature is increased from 4 K to RT. Each of these

spectra were fit to individual PL peaks, and values obtained for the Ep, Γp and IPL

are discussed below. It was observed that the intensity of the Id1 and Id′

1 exciton lines

diminish rapidly as the temperature is increased (also referred to as quenching). At

a temperature around 50 K, the Id1 and Id′

1 emissions become quite weak. At the same

time, the FX emission becomes the dominant emission as the temperature is increased. At

higher temperatures FX emission merges with the emission related to the recombination

of electrons and holes in the conduction and the valence bands, respectively (free carrier

recombination).

The Ep of the FX emission can be used to determine the dependence of the Eg on

temperature. This is because the FX emission energy follows the Eg as the temperature is

increased, up to higher temperatures before the free excitons dissociate into free carriers

by the thermal energy. Figure 5.10 shows the change in Ep of the FX emission (for

ZB-NW-3) with the temperature. The dependence of Eg on temperature is described in

terms of the semi-empirical Varshni’s equation [262] given by:

Eg(T ) = Eg(0)− αT 2

β + T(5.3)

where α and β are the fitting constants. α can be considered to represent the high-

temperature limit of the temperature gradient (dEg/dT ) of Eg, and β has been related

to the Debye temperature θD for the crystal [262]. The solid line in Figure 5.10 is a fit to


2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 4 K 8 K 2 0 K 3 0 K 4 0 K 5 0 K 6 0 K 7 0 K 8 0 K 9 0 K 1 0 0 K 1 1 5 K

PL In

tensity

(arb.

units)

E n e r g y ( e V )

Z B - N W - 3I e x c = 3 1 . 8 W / c m 2

Figure 5.9: Temperature dependent PL spectra from ZB-NW-3. All spectra were acquired

at a fixed average Iexc=31.8 W/cm2, are normalized by the integration times (ti) and

displayed offset for clarity. The LTPL spectra taken at temperatures above 115 K are

not shown.


the Equation 5.3, and the parameters obtained from the fit are listed in Table 5.2. The

same parameters previously reported by other authors on ZnSe single crystals and thin-

films are also listed in the Table 5.2 for comparison. The peak positions of the PL peaks

for which these parameters were reported are also listed in the Table 5.2 under Eg(0).

The parameters α and β are found to be in excellent agreement with the previously

reported values. This confirms the assignment of the PL peak at 2.800 eV to the free

excitons. Further, this also shows that the temperature gradient (dEg/dT ) (high-T limit)

in ZB NWs is similar to that in ZnSe single crystals and thin-films.

Table 5.2: Values of the fit parameters of the Varshni’s equation (Equation 5.3) which

describes the temperature dependence of the Eg. (See Figure 5.10). The values reported

previously by other authors are also listed for comparison. A fitting error of 0 indicates

error of the order of 10−5 eV.

Eg(0) α β Source

(eV) (10−4 eV/K) (K)

FX, 2.799 ± 0 8.9 ± 0.7 361.5 ± 44.8 Expt.

2.794 8.59 350.0 [296]

2.797 8.59 350.0 [297]

2.804 eV 8.59 405 [283,298]

2.800 ± 0.005 7.3 ± 0.4 295 ± 35 [263]

The temperature dependence of the direct Eg is also described by a Bose-Einstein

type expression [263,299]:

Eg(T ) = Eg(0)− 2αB[exp

(θBT

)− 1

] (5.4)

where αB represents the strength of the exciton-average phonon interaction, and θB

corresponds to the average phonon temperature. The dashed line in Figure 5.10 is a fit to


0 5 0 1 0 0 1 5 0 2 0 0 2 5 02 . 7 0

2 . 7 2

2 . 7 4

2 . 7 6

2 . 7 8

2 . 8 0 Z B - N W - 3 - E x p t .

E p = 2 . 8 0 0 e V F i t F i t

Energ

y (eV

)

T e m p e r a t u r e ( K )

Figure 5.10: Variation in PL peak energy of the FX emission (Ep=2.800 eV) (for ZB-

NW-3) with temperature. The solid and dashed lines are fits to the Equation 5.3 and

5.4, respectively.

Equation 5.4 and the obtained fit parameters are listed in the Table 5.3. These parameters

as reported by Malikova et al. [263] are also listed in the Table 5.3 for comparison. The

agreement between the two is good. The change in Ep for the Id1 and Id′

1 lines at 2.785 eV

and 2.780 eV, respectively, were also fit to these relations. However, a fit to the Varshni’s

equation (Equation 5.3) did not converge in both cases (unless one parameter was kept

fixed). This can be attributed to the smaller temperature range for which the values

of the Ep were available, because the Id1 emission was not observed above 150 K and

the Id′

1 line disappeared near 80 K. The Varshni’s equation describes the temperature


0 2 0 4 0 6 0 8 0 1 0 0 1 2 0 1 4 0 1 6 02 . 7 4

2 . 7 5

2 . 7 6

2 . 7 7

2 . 7 8

2 . 7 9En

ergy (

eV)

E n e r g y ( e V )

Z B - N W - 3 - E x p t . E p = 2 . 7 8 5 e V

F i t

0 2 0 4 0 6 0 8 0

( b )

Z B - N W - 3 - E x p t . E p = 2 . 7 8 0 e V

F i t

E n e r g y ( e V )

( a )

Figure 5.11: Variation in PL peak energies for the (a) Id1 and (b) Id′

1 lines (for ZB-NW-3)

with temperature. The solid line are fits to the Equation 5.4.

dependence of the Eg in a temperature range from low-temperatures to the RT. The

fit to this equation did not converge because the experimental data was in a smaller

temperature range only. The change in Ep for the Id1 and Id′

1 lines with the temperature

are shown in the Figures 5.11 (a) and (b), respectively. The Ep for both lines were fit to

the Equation 5.4 (solid lines), and the fit parameters are listed in Table 5.3. In the case

of the Id1 line, the parameters αB and θB are in good agreement to those obtained for

the FX emission. However, for the Id′

1 line the agreement for these parameters with the

FX emission and the Id1 lines is not so good, the fit itself is however as good as for the

Id1 line (based on the residuals). This shows different behaviour for the Id1 and Id′

1 lines

with temperature.

The temperature dependence of the Γp/2 (half-width at half-maximum, HWHM) of

the exciton emission in direct Eg semiconductors is expressed by the relation [300]:


Table 5.3: Values of the fit parameters of the Bose-Einstein type expression (Equation 5.4)

which describes the temperature dependence of the direct Eg. (See Figures 5.10 and 5.11

(a,b)). The values reported by Malikova et al. [263] are also listed for comparison. A

fitting error of 0 indicates error of the order of 10−5 eV.

Eg(0) αB θB Source

(eV) (meV) (K)

FX emission

2.797 ± 0 56.5 ± 4.7 209.3 ± 11.6 Expt.

2.800 ± 0.005 73 ± 4 260 ± 10 [263]

Id1 and Id′

1 emissions

Id1 , 2.785 ± 0 59.4 ± 11.1 201.9 ± 20.1 Expt.

Id′

1 , 2.779 ± 0 41.1 ± 21.4 141.4 ± 34.0 Expt.

Γp(T )/2 = Γinh + γthT +ΓLO[

exp

(θLOT

)− 1

] (5.5)

where each term on the right in the expression represents a different broadening mech-

anism. The first term (Γinh) is due to the intrinsic effects and represents the inhomo-

geneous broadening, including the instrumental broadening. It is to be noted that Γinh

is independent of the temperature. The second term is related to the lifetime broad-

ening due to exciton-acoustic phonon interactions, where γth is a constant representing

the acoustic phonon coupling. The third term corresponds to the exciton-LO phonon

interaction, where ΓLO is the strength of the exciton-LO-phonon coupling and θLO is

the LO-phonon temperature (or the temperature corresponding to LO-phonon energy).

Figure 5.12 shows the change in HWHM of the FX emission line as a function of the

temperature, and the solid line is a fit to the Equation 5.5. The values of the parameters

in Equation 5.5 obtained from the fit and those reported previously by other authors


0 5 0 1 0 0 1 5 0 2 0 0 2 5 0

5

1 0

1 5

2 0

2 5

HWHM

(meV

)


Z B - N W - 3 H W H M - E x p t . E p = 2 . 8 0 0 e V

F i t

Figure 5.12: Change in HWHM for the FX emission (Ep=2.800 eV) with temperature

(for ZB-NW-3). The solid line is a fit to the Equation 5.5.

on ZnSe crystals and thin films are listed in the Table 5.4. Figures 5.13 (a) and (b)

show the change in HWHM with the temperature for the Id1 and Id′

1 lines, respectively,

and the solid lines are fits to the Equation 5.5. A much smaller value of the Γinh for

the three lines (FX, Id1 and Id′

1 ) compared to those reported by other authors indicate a

small contribution of the inhomogeneous broadening to Γp, and this is consistent with

the narrow Γp of the exciton lines discussed earlier. The Γinh for the Id1 line, in particular,

is much smaller than the other values listed in the Table 5.4. A smaller value of Γinh is

also suggestive of high purity and excellent crystalline quality of ZB-NW-3.

The scattering of excitons due to acoustic phonons play a dominant role in the exciton-


Table 5.4: Values of the fit parameters of the Equation 5.5 which describes the temper-

ature dependence of the exciton linewidth (HWHM). (See Figures 5.12 and 5.13 (a,b)).

The values previously reported by other authors are also listed for comparison.

Γinh ΓLO θLO γth Source

(meV) (meV) (K) (µeV/K)

Ep=2.800 eV (FX) 2.6 ± 1.0 24.3 ± 9.6 360a 60.6 ± 14.0 Expt.

Eg 6.5 ± 2.5 24 ± 8 360a 2.0a [263]

I2 (2.797 eV) 11.1 38.3 362 25 [301]

I2 (2.795 eV) 7 ± 1 20 ± 5 359.75a 30 ± 10 [105]

I2 (2.795 eV) 8 ± 1 32 ± 5 359.75a 30 ± 10 [105]

Ep=2.785 eV (Id1 ) 1.6 ± 1 153.0 ± 62.7 360a 103.3 ± 25.0 Expt.

Ep=2.779 eV (Id′

1 ) 4.5 ± 0.4 913.1 ± 92.2 360a 48.1 ± 13.5 Expt.

a Parameter fixed.

phonon interaction processes, and at low-temperatures the broadening of the exciton Γp is

mainly due to the exciton-acoustic phonon interaction. As the temperature increases, the

population of the LO-phonons increases, and the exciton-LO phonon interaction becomes

stronger. The contribution of this interaction to Γp becomes increasingly significant as

the temperature is increased. The value of γth for the FX emission is higher than that

reported by others, and indicates a stronger FX-acoustic phonon interaction. Note that

Malikova et al. [263] fixed the value of γth in their results as 2.0 µeV/K, which is quite

smaller than the values reported by the others, see Table 5.4. The γth values for the

Id1 and Id′

1 lines are different from that for the FX emission corresponding to different

strengths of the exciton-acoustic phonon interaction for these lines. The value of ΓLO for

the FX emission agrees well with those reported by other authors, while they are much

larger for the Id1 and Id′

1 lines. Such large values of ΓLO for these lines indicate their strong

exciton-LO phonon coupling, and this is in complete agreement with the phonon-replicas


0 2 0 4 0 6 0 8 0 1 0 0 1 2 002468

1 01 21 41 61 82 02 2

( b )( a )

HWHM

(meV

)



F i t

0 2 0 4 0 6 0 8 0


F i t


Figure 5.13: Change in HWHM for the (a) Id1 and (b) Id′

1 lines with temperature (for

ZB-NW-3). The solid lines are fits to the Equation 5.5.

up to 3rd order observed in the LTPL spectrum from ZB-NW-3, for both Id1 and Id′

1 lines.

The thermal quenching of the IPL is described by a relation of the form [302]:

IPL(T ) =IPL(0)

1 + C exp (−Ea/kBT )(5.6)

for quenching that involves one mechanism (or one-step process), where Ea is the ac-

tivation energy of the mechanism and C is a coefficient. For thermal quenching of the

recombination lines involving two mechanisms (or two-step process), the equation that

represents the temperature dependence of IPL is modified to [303]

IPL(T ) =IPL(0)

1 + C1 exp (−Ea1/kBT ) + C2 exp (−Ea2/kBT )(5.7)

where Ea1 and Ea2 are the activation energies of the two mechanisms, and C1 and C2 are

coefficients. Figure 5.14 shows the change in IPL for the FX emission with inverse tem-

perature. The dashed and the solid lines are fits to Equations 5.6 and 5.7, respectively.

The fit to a one-step process gives an Ea of 24 meV, which is close to the exciton binding


5 1 0 1 5 2 0 2 5

Integ

rated

PL In

tensity

(I PL, a

rb. un

its)

1 0 0 0 / T ( K - 1 )

Z B - N W - 3 I P L - E x p t . E p = 2 . 8 0 0 e V

F i t t o t w o - s t e p p r o c e s s E a 1 = 2 0 m e V , E a 2 = 1 3 6 m e V

F i t t o o n e - s t e p p r o c e s s E a = 2 4 m e V

Figure 5.14: Change in integrated PL intensity (IPL) for the FX emission (Ep=2.800 eV)

with temperature (for ZB-NW-3). The dashed and solid lines are fits to the Equations 5.6

and 5.7, respectively. Note the log scale for IPL.

energy (of the free excitons) in ZnSe (EFX=21 meV [252]). The fit to a two-step process

however is better, and yields Ea1 and Ea2 of 20 meV and 136 meV, respectively. The Ea1

of 20 meV is close to the exciton binding energy in ZnSe, while the origin of the second

mechanism (Ea2=136 meV) is uncertain. Zhang et al. [106] observed a similar quenching

mechanism with Ea=136 meV for the IPL of the exciton emission lines in Ag-doped ZnSe

NWs. They did not specify the nature of this mechanism however. It is suggested that

this process may correspond to a non-radiative recombination mechanism, which could be

a result of the surface recombination in NWs. However, this needs further investigation.


0 5 0 1 0 0 1 5 0 2 0 0 2 5 0

( b )

Integ

rated

PL In

tensity

(I PL, a

rb. un

its)

1 0 0 0 / T ( K - 1 )


F i t t o o n e - s t e p p r o c e s s E a = 8 m e V

( a )

0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0


F i t t o o n e - s t e p p r o c e s s E a = 1 0 m e V

1 0 0 0 / T ( K - 1 )

Figure 5.15: Change in integrated PL intensity (IPL) for the (a) Id1 and (b) Id′

1 lines with

temperature (for ZB-NW-3). The solid lines are fits to the Equation 5.6. Note the log

scale for IPL.

Figures 5.15 (a) and (b) show the variation in IPL for the (a) Id1 and (b) Id′

1 lines

with the inverse temperature. The solid lines are fits to the Equation 5.6 for a one-step

quenching process. The Ea for the Id1 line is found to be 8 meV, while for the Id′

1 line,

Ea=10 meV. These are smaller than the binding energies of the neutral acceptor bound

excitons responsible for these lines. A fast thermal quenching of the Id1 line was also

observed by Zhong et al. [188], Shirakawa et al. [283] and Tournie et al. [291], where they

found that the Id1 emission disappears at temperatures around 50 K. Zimmermann et

al. [304] have attempted to address this issue of fast thermal quenching of the acceptor

bound excitons in semiconductors despite their large binding energies. They suggested

this fast quenching to be a result of the phonon-induced non-radiative recombinations.

In the present case, these are the LO-phonons, and this argument gains support from the

strong exciton LO-phonon coupling of the Id1 and Id′

1 lines, as evidenced by the observation

of the LO-phonon replicas in the PL spectrum and large values of the ΓLO for these lines.

Further, large values of the average number of phonons emitted (Nph) add support to

this argument. This means that the phonons emitted during the recombination of the


acceptor bound excitons (which are known for their strong LO-phonon coupling) are

the cause of their faster disappearance as the temperature is increased compared to the

donor-bound excitons, despite their larger binding energies than those of the donor bound

excitons.

5.2.4 Dependence of LTPL on Iexc

LTPL from ZB-NW-1 was measured as a function of the Iexc. Figure 5.16 shows the

change in LTPL as Iexc is increased from 0.32 W/cm2 to 236.8 W/cm2. As can be seen

from Figure 5.16, the FX emission at 2.800 eV is red-shifted to 2.797 eV at the highest

Iexc. This is due to the laser heating of the ZB-NW-1. As the Iexc is increased, it becomes

more difficult to dissipate heat owing to the small volume of the NWs and lack of a good

thermal contact with the TEM grid. Bao et al. [133] observed much larger red-shifts

in PL from single InP NWs at increasing Iexc which they considered to be caused by

the laser heating. It should be noted that large red-shifts seen by Bao et al. [133] were

due to the very high Iexc they used. As the temperature of the ZB-NW-1 increases due

to the laser heating, the band gap and the energy positions of different lines also red-

shift. The analysis described in Chapter 4 using the dependence of PL on Iexc cannot

accurately describe the recombination mechanisms in this case. This is however not a

hindrance in the interpretation of the recombination lines because the dependence of PL

on temperature was useful in obtaining this information.

5.3 Conclusions

In this chapter, the results were presented on the optical characterization of single ZB

ZnSe NWs using LTPL. The structural characterization performed using TEM, HR-TEM

and SAD confirmed the single crystalline ZB microstructure of the same ZB NWs that

were studied using LTPL. The optical response was found to be dominated by emissions


2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5

Z B - N W - 1PL

Inten

sity (a

rb. un

its)

E n e r g y ( e V )

7 4 0 P 0

6 0 0 P 0

4 0 0 P 03 0 0 P 02 0 0 P 0

1 0 0 P 0

4 0 P 0

2 0 P 0

1 0 P 0

4 P 0

2 P 0

P 0

P 0 = 0 . 3 2 W / c m 2

T e m p . = 3 . 5 K

Figure 5.16: Iexc-dependent LTPL spectra from ZB-NW-1. All spectra were acquired at

a constant temperature (3.9 K), are normalized to unity and displayed offset for clarity.


related to the native point defects. No evidence was found of the role of unintentional

impurities in the optical properties of ZB NWs. The recombination lines at 2.785 eV

and 2.780 were assigned to the neutral acceptor bound excitons related to VZn and (VZn-

complex), respectively. Another recombination line at 2.800 eV was related to the free

excitons (FX). The binding energies of the excitons related to the Id1 and Id′

1 were found to

be 15 meV and 20 meV, respectively. The ionization energies of the acceptors responsible

for these lines were 150 meV and 200 meV, respectively. The Id1 and Id′

1 lines were

accompanied by LO-phonon replicas up to 3 orders. The average number of emitted

phonons was found to be 0.53 < Nph < 0.7 for the Id1 line, and 0.25 < Nph < 0.42 for the

Id′

1 line. There have been no reports thus far of the Id1 and Id′

1 lines and their phonon-

replicas in ZnSe NWs. Linewidths of 1.5 meV for the Id1 line were observed. These are the

narrowest linewidths reported so far on ZnSe NWs. The optical response was compared

for the single ZB NWs and a bundle of ZB NWs. For a bundle of NWs, the peak positions

of the recombination lines are similar to those from single NWs. The Y-line emission

related to structural defects was not detected. This confirms that the Y-line does not

originate from single-crystalline NWs, but rather from the polycrystalline undergrowth

on the sample. Further, the linewidths in the case of bundle of NWs were found to be

larger than in single NWs. This is related to the effects of ensemble broadening. The

linewidths exhibited by different NWs for similar conditions of acquisition of PL spectrum

were found to be slightly different, and the ensemble broadening could also be limited by

the NW which shows the largest linewidths.

The PL from single NWs was studied as a function of temperature. The temperature

dependence of the FX emission peak energy confirmed its assignment to the free exci-

tons. The high-temperature limit of the temperature gradient dEg/dT was determined

and found to be in excellent agreement with the reported values on bulk ZnSe. The

temperature dependence of the Id1 and Id′

1 lines was used to determine the parameters of

the Bose-Einstein type expression of the temperature dependence of the band gap. The


dependence of HWHM for the FX emission, and Id1 and Id′

1 lines revealed the small con-

tribution of the inhomogeneous broadening to the linewidths. A strong exciton-acoustic

phonon and exciton-LO phonon coupling was found, which is consistent with the obser-

vation of LO-phonon replicas and the large average number of emitted phonons. The

integrated PL intensity of the FX emission line was found to be quenched with temper-

ature by two mechanisms. The first mechanism with an activation energy of 20 meV is

related to the dissociation of free excitons, while the other with an activation energy of

136 meV could be due to non-radiative recombination related to surface recombination.

The Id1 and Id′

1 lines were quenched with temperature with activation energies smaller

than their binding energies. These are proposed to be due to non-radiative recombination

induced by scattering of excitons with LO-phonons simultaneously emitted.

Chapter 6

Single wurtzite ZnSe nanowires

6.1 Introduction

It was mentioned in Chapter 1 that for bulk ZnSe, the WZ crystal structure is meta-

stable and is therefore difficult to obtain. This explains the availability of a vast amount

of literature on luminescence characterization of ZnSe single crystals and thin films in

the stable ZB phase, but only a few reports are available on the same for WZ bulk ZnSe

single crystals. However, unlike the bulk case, ZnSe NWs can be obtained in the WZ

crystal structure. Although the number of reports on ZB NWs still exceed those on

WZ NWs, there is a clear evidence that ZnSe NWs can also crystallize in the WZ crystal

structure. Existing reports available on WZ NWs have focused on describing their growth

and structural properties, while their optical characterization has not been adequately

described. LTPL from WZ NWs has been reported by a few authors, but these were

performed on an array of NWs which also included ZB NWs. These were discussed in

greater detail in Chapter 2. The optical response of ZB NWs was unambiguously related

to their single crystalline microstructure, and these results were presented in Chapter 5.

It is of particular importance to note that applying a similar strategy is crucial for the

case of WZ NWs, because optical transitions in WZ ZnSe have not been studied in

130

Chapter 6. Single wurtzite ZnSe nanowires 131

detail previously. The work by Liang and Yoffe [305] several decades ago is still the only

authoritative report on luminescence characterization of WZ ZnSe.

This chapter presents the results on the optical response from individual WZ NWs

with their crystal structure determined using TEM, HR-TEM and SAD, and is organ-

ised as follows. First, the results on the structural and optical characterization for an

individual WZ NW (labelled as WZ-NW-1) are presented. The origin of recombination

lines in the LTPL spectrum from WZ-NW-1 is discussed in detail with reference to the

works by Liang and Yoffe [305] and on ZB NWs. Results on luminescence characteri-

zation of another single WZ NW (labelled as WZ-NW-2) are presented thereafter. PL

from WZ-NW-1 was also studied as a function of temperature and Iexc, which are then

described. Quantitative information is extracted from the dependence of PL on temper-

ature, using the models that were used earlier in Chapter 5, and described in more detail

in Appendix C.2.2.

6.2 Results and Discussion

6.2.1 Characterization of single WZ nanowires

Six individual WZ NWs were characterized by both optical and structural characteriza-

tion techniques, and all of them were selected from the same as-grown sample. For the

determination of the crystal structure, similar to the case of ZB NWs, HR-TEM micro-

graphs were taken at several locations along the length of the NWs. The SAD patterns

were also taken at multiple locations along the length of the NWs to arrive at a conclusion

about the crystal structure. Unlike the case of ZB NWs, where most NWs were defect-

free and single-crystalline, a fraction of the WZ NWs contained stacking faults. These

are not included in the results here. In the following, results from structural characteri-

zation are presented for a single-crystalline individual WZ NW (labelled as WZ-NW-1),

and from LTPL characterization on WZ-NW-1 and another individual WZ NW (labelled


as WZ-NW-2).

a b

WZ-NW-1

Location of SAD

Figure 6.1: Structural characterization of WZ-NW-1: (a) Overview TEM micrograph,

and (b) TEM micrograph at a magnification higher than in (a). The circle denotes the

area from where the SAD pattern (shown in Figure 6.3 (b)) was taken.

HR-TEM characterization

An overview TEM micrograph of WZ-NW-1 is shown in Figure 6.1 (a). A higher mag-

nification TEM image of the same NW is shown in Figure 6.1 (b). Figure 6.2 (a) shows

a HR-TEM micrograph from an area of WZ-NW-1. The FFT pattern presented in Fig-

ure 6.2 (b) was taken from an area squared in (a), and the indexed spots in the FFT

pattern show a WZ crystal structure of the selected area. A HR-TEM micrograph from

another area of WZ-NW-1 is presented in Figure 6.3 (a). The SAD pattern shown in

Figure 6.3(b) was taken from the area shown in Figure 6.1 (b). The indexed diffrac-

tion spots confirm that WZ-NW-1 is single crystalline with WZ crystal structure. The

TEM and HR-TEM micrographs, and indexed FFT pattern for WZ-NW-2 are shown in

Figures G.4 and G.5 in Appendix G which confirm a WZ crystal structure for WZ-NW-2.


a b

1102

1101

1101

1102

0002

00021100

1102

1101

1101

1102

1100

Figure 6.2: (a) HR-TEM micrograph from an area of WZ-NW-1, (b) Indexed FFT pattern

taken from the area squared (in red) in (a). The indexed spots correspond to a WZ crystal

structure. The viewing direction for (a) and (b) is 〈1100〉.

a b

11021101

1101

1102

0002

0002

1100

1102

1101

11011102

1100

Figure 6.3: (a) HR-TEM micrograph from another area, and (b) SAD pattern acquired

from a larger area (shown in Figure 6.1(b)) of WZ-NW-1 . The indexed spots correspond

to a WZ crystal structure. The viewing direction for (a) and (b) is 〈1100〉.


Low-temperature PL spectra

Figure 6.4 shows the excitonic region of the LTPL spectrum (3.8 K) from WZ-NW-

1 acquired under the conditions of high throughput. The LTPL spectrum was fit to

individual emission peaks using Lorentzians as shown. The LTPL spectrum from WZ-

NW-1 is more complex than those from ZB NWs (in Chapter 5). This means that the

number of emission peaks is larger in the LTPL from WZ-NW-1, and their interpretation

more difficult. In the case of ZB NWs, the identification of recombination lines was

facilitated by the existing literature on LTPL from ZnSe single crystals and thin films.

This is however not the case for WZ NWs, since there exist only a few reports on LTPL

from WZ ZnSe (strictly speaking just one by Liang and Yoffe [305]). The identification of

recombination lines in the case of WZ-NW-1 will be discussed in terms of the identification

proposed by Liang and Yoffe [305], and by reference to the identification described for

ZB NWs.

In their work on LTPL from WZ ZnSe crystals, Liang and Yoffe [305] observed several

peaks which they related to the free and bound excitons in WZ ZnSe and their satellite

peaks (replicas) due to the simultaneous emission of acoustic and optical phonons. They

gave the energies of these phonons, as listed in the Table 6.1. However, no attempts

were made by them to identify the recombination centres responsible for these lines.

In particular, they observed peaks related to excitons bound to neutral donors, ionized

acceptors and neutral acceptors, which they labelled as I1, I2, and I3, respectively. Note,

that these labels are different from those used conventionally for bound exciton lines, i.e.

for excitons bound to neutral donors (I2), neutral acceptors (I1) and to ionized acceptors

(I3). Some of the peaks due to the emission of phonons were found to be overlapping for

these lines. They gave the binding energy of excitons bound to neutral acceptors (I3) as

0.0098 ± 0.0004 eV, and the ionization energy of the acceptors as EA=0.118 eV.

They also reported an important experiment, which they described only briefly. This

will soon be found to be very important in the current discussion on LTPL from WZ


Table 6.1: Phonon energies of the acoustic and optical phonons in WZ ZnSe, as given by

Liang and Yoffe [305].

Phonon Symbol Energy (eV)

longitudinal optical LO 0.0318 ± 0.0004

transverse optical TO 0.0256 ± 0.0005

longitudinal acoustic LA 0.0098 ± 0.0005

transverse acoustic TA 0.0064 ± 0.0005

NWs. They found that when the WZ ZnSe crystals were implanted with Mn ions, the

crystal structure locally changed to ZB. They concluded this based on the LTPL from

implanted WZ ZnSe crystals where they observed PL peaks similar to those observed

from ZB crystals. In particular, they noticed peaks at 2.7969 eV (which they labelled as

L1) and 2.7834 eV (L3), and related these lines to the excitons bound to neutral donors

and neutral acceptors in ZB ZnSe, respectively. The peak at 2.7834 eV is similar in energy

position to the Id1 line in ZB ZnSe, and in their implanted ZnSe crystals, this line can now

be assigned to the neutral acceptors related to VZn. This assignment is further supported

by the phonon-replicas of 2 orders which they reported for this line (L3-LO and L3-2LO).

Recall from Chapter 5, that a strong-LO phonon coupling is a characteristic feature of

the Id1 line. Since the WZ ZnSe crystals they studied were grown by the vapour-phase

transport method, it can be said that the concentration of unintentional impurities which

usually result in melt-grown crystals would be low, and the LTPL spectrum would be

influenced mainly by the native point defects. Note also, that the work of Liang and

Yoffe [305] pre-dates the work by Roppischer et al. [191], which may be considered to

be the first report relating the Id1 line to the VZn acceptors. This explains the reason

that these lines in LTPL from WZ ZnSe and implanted WZ ZnSe were not assigned by

Liang and Yoffe [305]. They also observed DAP-related emission in WZ ZnSe crystals,

and in implanted WZ ZnSe crystals with local phase transformation to ZB ZnSe. These


2 . 5 5 2 . 6 0 2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5 2 . 9 0

DAP-3

LO DAP-2

LO

DAP-1

LO

L T P L - W Z - N W - 1 I 1 l i n e I 1


T e m p . = 3 . 8 K I e x c = 3 . 8 W / c m 2

d s l i t = 3 0 0 µmPL

Inten

sity (a

rb. un

its)

E n e r g y ( e V )

I 1d -1LO

I 1-1LODA

P

5 . 8 m e V

Figure 6.4: Excitonic region of the LTPL spectrum (3.8 K) from WZ-NW-1 acquired

under the conditions of high-throughput. (Iexc and dslit are as indicated). The inset

shows a confocal PL image of WZ-NW-1 under laser excitation with an average Iexc=12.7

W/cm2. The LTPL spectrum was fit to individual emission peaks using Lorentzians as

shown.

DAP-related transitions were accompanied by LO-phonon replicas in both cases.

In the LTPL spectrum from WZ-NW-1 (shown in Figure 6.4), two dominant lines are

observed at 2.841 eV and 2.833 eV. These are also of approximately the same strengths

(when comparing the peak heights). The binding energy of the excitons can be found

using EBX = Eg −EFX −Ep, where EBX is the binding energy of the donor or acceptor

bound excitons. In the case of WZ ZnSe, Eg and EFX are 2.874 eV [238] and 22 meV [238],

respectively. These give for the line with Ep=2.841 eV, EBX=11 meV, and for the line


with Ep=2.833 eV, EBX=19 meV. The EBX of the line at 2.841 eV is similar to that

reported by Liang and Yoffe [305] for the acceptor bound excitons line (9.8 meV). This

line at 2.841 eV can therefore be related to acceptor bound excitons. According to the

empirical rule of Halsted and Aven [253], the ionization energy of the acceptor is 110

meV, which is similar to value reported by Liang and Yoffe [305] mentioned earlier (118

meV). The other line at 2.833 eV can be related to a deep acceptor, whose ionization

energy is 190 meV. Since in their Mn-ion implanted WZ ZnSe samples, a line similar to

the Id1 line was observed, the neutral acceptor responsible the L3 line (their notation)

can be related to VZn. The observation of these two lines in the LTPL from WZ-NW-1

is similar to the case of ZB NWs where two dominating lines (Id1 and Id′

1 ) were observed.

Id1 line was assigned to the excitons bound to deep neutral acceptors related to VZn,

whereas Id′

1 was ascribed to neutral acceptors related to the VZn-complex. In the case

of WZ-NW-1, therefore, VZn act as shallow acceptors and are responsible for the line at

2.841 eV (labelled in the present case as I1). The other line at 2.833 eV, with reference to

the ZB NWs, can be ascribed to the VZn-complex acceptors which act as deep acceptors.

This line is labelled as the Id1 line. These lines and their 1st LO-phonon replica are

indicated on the Figure 6.4. The other lines can be assigned to the phonon-replicas of

the I1 and Id1 lines, similar to the assignment proposed by Liang and Yoffe [305]. The

energy positions of these lines and suggested assignments based on the present work are

listed in Table 6.2. The energy positions and suggested assignments for the lines given by

Liang and Yoffe [305] are also listed in Table 6.2. Only the lines common to the present

work, and I1 and I2 lines (their notation) are listed. Note that there are no common

assignments except for the case of the I1 line at 2.841 eV. This is the first time since the

work of Liang and Yoffe [305] that a recombination scheme has been proposed for the

optical transitions in WZ ZnSe.

The peak with Ep=2.741 eV is assigned to the DAP-related transition, similar to the

previous assignment by Liang and Yoffe [305] for their PL peak at 2.7384 eV. The peaks


Table 6.2: Proposed assignments of the PL peaks observed in the LTPL spectrum from

WZ-NW-1. The assignment given by Liang and Yoffe [305] are also given as a reference

for the common peaks observed, and I1 and I2 peaks (their notation).

WZ-NW-1 WZ ZnSe Bulk (Ref. [305])

Energy (eV) Proposed assignment Energy (eV) Line & its assignment

2.8452 ± 0.0002 I1, (D0X)

2.8424 ± 0.0002 I2, (A−X)

2.841 I1 2.8383 ± 0.0002 I3, (A0X)

2.833 Id1 2.8313 ± 0.0010 I3-TA

2.823 Id1 -LA 2.8203 ± 0.0006 I3-TO

2.809 I1-LO 2.8094 ± 0.0006 I2-LO

Id1 -TO

2.800 Id1 -LO

2.790 Id1 -LO-LA

2.782 I1-LO-TO

Id1 -2TO

2.776 I1-2LO

2.771 Id1 -2LO

2.760 Id1 -2LO-LA

2.752 Id1 -2LO-LA-TA

with Ep=2.709 eV, 2.678 eV and 2.650 eV being LO-phonon energy apart are related to

the LO-phonon replicas of the DAP-related peak at 2.741 eV. The average number of

phonons emitted was calculated for this transition using Equation 5.2, and the ratios of

the intensity of the m-th order LO-phonon replica to that of the ZPL(Ep=2.741 eV) are

listed in the Table 6.3. The values in columns (3) and (6) corresponding to m=1 represent

Nph. The values of these ratios given by Liang and Yoffe [305] are also included in the


Table 6.3: Experimental and calculated ratios of the intensity of the m-th order LO-

phonon replica (Im) to that of the ZPL (I(m=0)) for the DAP-related transitions in the

LTPL from WZ-NW-1. The calculated ratios were found using Equation 5.2. Nph are

the values in columns (3) and (6) corresponding to m = 1. The values of these ratios

given by Liang and Yoffe [305] are also given for comparison.

WZ-NW-1 WZ ZnSe Bulk (Ref. [305])

m Energy (eV) Expt. Calculated Energy (eV) Expt. Calculated

(1) (2) (3) (4) (5) (6)

DAP 0 2.741 1.00 1.00 2.7384 1.00 1.00

DAP-1LO 1 2.709 0.707 0.690 2.7067 0.69 0.69

DAP-2LO 2 2.678 0.251 0.238 2.6748 0.29 0.24

DAP-2LO 3 2.650 - 0.055 2.6424 0.07 0.05

DAP-4LO 4 - - 0.009 2.612 0.016 0.01

- indicates values were not available

Table 6.3. The Nph was found to be identical to that reported by Liang and Yoffe [305].

The Γp of the I1 line at 2.841 eV in Figure 6.4 is 5.8 meV where the LTPL spectrum

was acquired under the conditions of high-throughput. Figure 6.5 (a) shows a LTPL

spectrum from WZ-NW-1 acquired under the conditions of high-resolution. The LTPL

spectrum was fit to individual emission peaks using Lorentzians, and these are shown

in the Figure 6.5 (b). Figure 6.5 (a) also shows the I1 line with a Γp of 2.8 meV. This

indicates excellent optical quality and crystalline structure of WZ-NW-1, as confirmed

through the TEM and HR-TEM characterization shown in Figures 6.2 and 6.3. These Γp

are similar to those obtained in the LTPL spectra from ZB NWs under similar conditions.

However, Γp close to 1.5 meV observed in the LTPL from ZB-NW-4 could not be obtained

in LTPL from WZ NWs. Since there are no other reports available on the LTPL from WZ

NWs, it is understood that these are the narrowest Γp reported on WZ NWs, indicative of


2 . 5 5 2 . 6 0 2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5 2 . 9 0

PL In

tensity

(arb.

units)

E n e r g y ( e V )

L T P L - W Z - N W - 1 I 1 l i n e

T e m p . = 3 . 8 K I e x c = 3 . 8 W / c m 2

d s l i t = 1 0 0 µm

2 . 8 m e V

2 . 5 5 2 . 6 0 2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5 2 . 9 0

( b ) L T P L - W Z - N W - 1 I 1 l i n e I n d i v i d u a l p e a k s S u m o f i n d i v i d u a l p e a k s

E n e r g y ( e V )

( a )

Figure 6.5: (a,b) Excitonic region of the LTPL spectrum (3.8 K) from WZ-NW-1 acquired

under the conditions of high-resolution. (Iexc and dslit are as indicated). The I1 line with

a Γp=2.8 meV is shown in (a). The LTPL spectrum was fit to individual emission peaks

using Lorentzians as shown in (b).

high optical quality of WZ ZnSe NWs. Figures 6.6 (a,b) show the LTPL spectrum from

WZ-NW-2 acquired using the conditions of (a) high-throughput and (b) high-resolution.

The inset to Figure 6.6 (a) shows a confocal PL image of WZ-NW-2 taken using average

Iexc=12.7 W/cm2. These spectra show general features similar to those from WZ-NW-1,

i.e. a large number of peaks related to bound exciton emissions, DAP-related emission

and their phonon-replicas. There is an important difference however. The Γp of the I1 line

is quite large, even in the LTPL spectrum taken under the conditions of high-resolution

(∼9 meV). Also, the strengths of the I1 and Id1 lines (in terms of their peak heights) are

not equal in this case. This indicates an unequal concentration of the acceptor centres

responsible for the I1 and Id1 lines in the WZ NWs. Further, different Γp for the I1 line

in the LTPL spectra from different NWs also show that their optical response is similar

in terms of the energy positions of different peaks observed, but not in terms of the Γp.


2 . 5 5 2 . 6 0 2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5 2 . 9 0 2 . 9 5

( b )

PL In

tensity

(arb.

units)

E n e r g y ( e V )

L T P L - W Z - N W - 2 I 1 l i n e I n d i v i d u a l p e a k s S u m o f i n d i v i d u a l p e a k s

T e m p . = 4 . 0 K I e x c = 3 1 . 8 W / c m 2

d s l i t = 5 0 0 µm

1 0 m e V

( a )

2 . 5 5 2 . 6 0 2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5 2 . 9 0

L T P L - W Z - N W - 2 I 1 l i n e I n d i v i d u a l p e a k s S u m o f i n d i v i d u a l p e a k s

T e m p . = 4 . 0 K I e x c = 3 . 2 W / c m 2

d s l i t = 1 0 0 µm

E n e r g y ( e V )

9 m e V

Figure 6.6: Excitonic region of the LTPL spectrum (4.0 K) acquired from WZ-NW-2


as indicated). The inset in (a) shows a confocal PL image of WZ-NW-2 under laser

excitation with average Iexc=12.7 W/cm2. The LTPL spectra in both cases were fit to

individual emission peaks using Lorentzians as shown.

6.2.2 Dependence of PL on temperature

PL from WZ-NW-1 was measured as a function of the sample temperature. The variation

in PL from WZ-NW-1 as the temperature is increased from 4 K to the RT is shown in

Figure 6.7. It was observed that the intensity of PL emission is quenched rapidly as the

temperature is increased, and no PL was detected at temperatures above 130 K. Further,

it can also be seen from Figure 6.7 that the excitonic (I1 and Id1 ) PL lines diminish as the

temperature is increased, while the DAP recombinations persist at higher temperatures

(up to 100 K). Figure 6.8 shows the change in energy position of the I1 line with the

temperature. It is seen that the peak energy first increases as the temperature is increased

to 40 K, and then decreases with increasing temperature. This can be understood as

follows. The excitons bound to the VZn responsible for the I1 line dissociate from the


2 . 5 6 2 . 6 4 2 . 7 2 2 . 8 0 2 . 8 8 4 K 1 0 K 2 0 K 3 0 K 4 0 K 5 0 K 6 0 K 7 0 K

PL In

tensity

(arb.

units)

E n e r g y ( e V )

W Z - N W - 1I e x c = 3 8 . 2 W / c m 2

Figure 6.7: Temperature dependent PL spectra from WZ-NW-1. All spectra were ac-

quired at a fixed average Iexc=38.2 W/cm2, are normalized by the integration times (ti)

and displayed offset for clarity. The LTPL spectra taken at temperatures above 70 K are

not shown.


VZn centres, and become free excitons (FX). The energy position of the FX emission

changes as the band gap shrinks with increasing temperatures and reflects this change.

The solid and dashed lines in Figure 6.8 are fits to the Equation 5.4 (Bose-Einstein type

expression for the temperature dependence of the Eg) at temperatures above 30 K and

40 K, respectively. This means that Eg(0) in Equation 5.4 corresponds to Eg(30) and

Eg(40) for the solid and dashed lines, respectively. These are found to be 2.845 eV in both

cases. The parameters obtained from the fits are listed in Table 6.4. These parameters

have not been reported in the literature previously, so they cannot be compared with

other works. Similar to the case of the Id1 and Id′

1 lines (ZB-NW-3 in Chapter 5), a fit

of the energy positions of the I1 line at different temperatures to the Varshni’s equation

(Equation 5.3) did not converge. This can be explained by the same reasoning as given

earlier in Chapter 5.

This process of dissociation of bound excitons to free excitons is also supported by

the fit of IPL at different temperatures to the thermal quenching process involving one

mechanism, described by the Equation 5.6. Figure 6.9 shows the change in IPL with

temperature, and the solid line is a fit to the Equation 5.6. The activation energy of the

thermal quenching process was found to be 9 meV, which is in good agreement with the

binding energy of the neutral acceptor bound excitons (11 meV) related to the I1 line.

Table 6.4: Values of the fit parameters of the Bose-Einstein type expression (Equation 5.4)

which describes the temperature dependence of the direct Eg. (See Figure 6.8).

Eg(0) αB θB Source

(eV) (meV) (K)

2.845 ± 0 69.8 ± 9.2 250.4 ± 14.6 Expt.

2.845 ± 0 64.6 ± 9.6 241.0 ± 16.8 Expt.

Figure 6.10 shows the change in the Γp/2 (HWHM) of the I1 exciton emission line with

the temperature for WZ-NW-1. The change in HWHM with temperature is described


0 2 0 4 0 6 0 8 0 1 0 0 1 2 02 . 8 2 0

2 . 8 2 5

2 . 8 3 0

2 . 8 3 5

2 . 8 4 0

2 . 8 4 5

Energ

y (eV

)


W Z - N W - 1 - E x p t . E p = 2 . 8 4 1 e V

F i t F i t

Figure 6.8: Variation in PL peak energy of the I1 line (Ep=2.841 eV) (for WZ-NW-1) with

temperature. The solid and dashed lines are fits to the Equation 5.4 for temperatures

above 30 K and 40 K, respectively.

by the Equation 5.5, and the solid line in Figure 6.10 is a fit to this equation. The

parameters obtained from the fit are listed in Table 6.5. These parameters from the

literature cannot be listed as they have not been reported before for WZ ZnSe. These

are however listed for the Id1 line in LTPL from ZB-NW-3 for comparison. The value of

Γinh which represents the inhomogeneous broadening term is larger than for the Id1 line.

This can be understood by considering that the Γp for the WZ-NW-1 is somewhat larger

than that for the Id1 line. A large value of ΓLO indicates a strong LO-phonon coupling,

but less stronger than that for the Id1 line in ZB-NW-1. This is explained by the 2 orders


0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0

Integ

rated

PL In

tensity

(I PL, a

rb. un

its)

1 0 0 0 / T ( K - 1 )

W Z - N W - 1 I P L - E x p t . E p = 2 . 8 4 1 e V E a = 9 m e V

F i t t o o n e - s t e p p r o c e s s

Figure 6.9: Change in integrated PL intensity (IPL) for the I1 line with temperature (for

WZ-NW-1). The solid line is a fit to the Equation 5.6. Note the log scale for IPL.

of phonon-replicas observed in the LTPL from WZ-NW-1, in contrast to the case of ZB-

NW-3 where replicas up to 3 orders were seen. A much smaller value of the γth indicates

weaker strength of the exciton-acoustic phonon interaction in the case of WZ-NW-1.

6.2.3 Dependence of LTPL on Iexc

Figure 6.11 shows the variation in LTPL spectra obtained from WZ-NW-1 as the Iexc

is increased from 0.32 W/cm2 to 252.2 W/cm2. The Iexc was increased by almost three

orders of magnitude, which is sufficiently high to induce laser heating effects in single

NWs, as can be seen by the slight red-shits in peak positions at higher Iexc. This effect of


0 2 0 4 0 6 0 8 0 1 0 0 1 2 0

4

6

8

1 0

1 2

1 4

HWHM

(meV

)


W Z - N W - 1 H W H M - E x p t . E p = 2 . 8 4 1 e V

F i t

Figure 6.10: Change in HWHM for the I1 line (Ep=2.841 eV) with temperature (for

WZ-NW-1). The solid line is a fit to the Equation 5.5.

laser heating was also observed in ZB-NW-1 (discussed in Chapter 5), which precluded the

accurate determination of the quantitative information about the different recombination

mechanisms based on the dependence of PL on Iexc. However, it may be noted from

Figure 6.11 that the higher energy emission lines I1 and Id1 at 2.841 eV and 2.832 eV,

respectively, which are due to the shallow-acceptor bound excitons, continue to increase

in intensity as the Iexc is increased. The DAP-related emission peaks at 2.741 eV and

below tend to saturate at higher Iexc. This provides support to the assignment of these

peaks discussed earlier.


2 . 5 5 2 . 6 0 2 . 6 5 2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5 2 . 9 0

7 8 8 P 06 0 0 P 0

4 0 0 P 0

3 0 0 P 0

2 0 0 P 0

1 0 0 P 0

4 0 P 0

2 0 P 0

1 0 P 0

4 P 0

2 P 0

W Z - N W - 1P 0 = 0 . 3 2 W / c m 2

T e m p . = 3 . 8 K

P 0

PL In

tensity

(arb.

units)

E n e r g y ( e V )Figure 6.11: Iexc-dependent LTPL spectra from WZ-NW-1. All spectra were acquired at

a constant temperature (3.8 K), are normalized to unity and displayed offset for clarity.


Table 6.5: Values of the fit parameters of the Equation 5.5 which describes the tem-

perature dependence of the exciton linewidth (HWHM). (See Figure 6.10). The values

obtained for the Id1 line for ZB-NW-3 are also listed for comparison.



WZ-NW-1

Ep=2.841 eV (I1) 4.0 ± 0.2 110.6 ± 10.0 360a 32.7 ± 4.1 Expt.

ZB-NW-3

Ep=2.785 eV (Id1 ) 1.6 ± 1 153.0 ± 62.7 360a 103.3 ± 25.0 Expt.

a Parameter fixed.

6.3 Conclusions

Characteristics of the optical response of single WZ ZnSe NWs were presented in this

chapter in direct relation to their crystal structure. The crystal structure for NWs were

found to be WZ based on TEM, HR-TEM and SAD experiments with their single crys-

talline microstructure. The recombination centres responsible for the PL peaks in the

LTPL spectra from single WZ NWs were identified based on the work of Liang and

Yoffe [305] and by reference to the identification of PL lines for ZB NWs. The optical re-

sponse was found to be dominated by native point defects in WZ ZnSe. A recombination

line at 2.841 eV (I1) was related to the excitons bound to neutral acceptors related to

VZn, which act as deep neutral acceptors in ZB NWs. The binding energy for this exciton

was found to be 11 meV, and the ionization energy of the acceptor 110 meV. Another

recombination line at 2.833 eV (Id1 ) was related to the excitons bound to the deep neutral

acceptors related to VZn-complexes by reference to their observation in ZB NWs. The

binding energy of this exciton and the ionization energy of the acceptor were found to be

19 meV and 190 meV, respectively. Other transitions observed in the LTPL spectra from


WZ NWs were related to the simultaneous emission of acoustic and optical phonons, and

a recombination scheme was proposed. This is the first time since the work of Liang and

Yoffe [305] that a transition scheme has been proposed, and acceptors responsible for

binding the excitons identified. DAP-related transitions were also observed with their

LO-phonon replicas, and the average number of phonons emitted was found to be 0.69 in

agreement with the results of Liang and Yoffe [305]. The linewidths for the I1 line of 2.8

meV were obtained, which indicate excellent optical and crystalline quality of the WZ

NWs. It was also noted that such narrow linewidths have not been reported thus far on

WZ ZnSe NWs, suggestive of their high optical quality. The parameters αB and θB for the

dependence of PL peak energy on temperature were deduced from the dependence of PL

on temperature. The other parameters that were extracted include those related to the

inhomogeneous broadening of the I1 line (Γinh=4.0±0.2), strength of the exciton-acoustic

phonon (γth=32.7±4.1 µeV/K) and exciton-LO phonon (ΓLO=110.6±10.0) interactions.

The activation energy for the thermal quenching of the I1 line was found to be 9 meV in

agreement with the binding energy of the excitons responsible for the I1 lines. This con-

stitutes the first report on detailed characterization and understanding of luminescence

properties of single WZ NWs. This understanding is crucial for design of reliable and

efficient optoelectronic devices with their desired characteristics.

Chapter 7

ZnSe Nanowire Twinning

Superlattices

7.1 Introduction

The distinct properties of emerging nanoscale structures, referred to as NTSLs, based

on their unique structure was discussed in Chapter 1. Their potential in future opto-

electronic and photonic devices, and in materials engineering at the nanoscale was also

discussed therein. It was described in Chapter 2 that the electronic properties of NTSLs

are not understood yet. Further, it was also emphasized that the determination of the

electronic properties of NTSLs as a function of their twin-plane spacing (d) is hindered

by the technical difficulties involved in nanoscale mapping of the physical properties of

the NTSLs (e.g. Eg) with their structure. In this chapter, results are presented on the

extensive optical and structural characterization of individual NTSLs, organized as fol-

lows. First, a periodicity parameter is defined as being inversely proportional to d in

NTSLs, and structural characterization results of a few representative NTSLs are pre-

sented. This is followed by a description of the LTPL spectrum from an individual NTSL,

and the observation of Eg larger than that of ZB ZnSe is discussed in terms of the poly-

150

Chapter 7. ZnSe Nanowire Twinning Superlattices 151

type character of NTSLs. The dependence of LTPL on Iexc is then presented to confirm

this polytype character of NTSLs. The experimental demonstration of variation in Eg

of NTSLs is then presented with a brief description of supporting ab Initio electronic

structure calculations. This variation is expressed as a function of the periodicity pa-

rameters (related to d), and related to Eg for ZB and WZ ZnSe. The TRPL decay from

an individual NTSL is compared with that from a single homogeneous ZB NW to show

that the PL lifetimes in the two cases are similar. The dependence of PL on temperature

for an individual NTSL is then presented, and the temperature gradient dEg/dT of a

NTSL is compared with that of ZB NWs and WZ ZnSe (bulk). The activation energies

of the thermal quenching of PL emission from a NTSL are deduced from its dependence

on temperature, and the parameters related to electron-phonon coupling are obtained.

7.1.1 Periodicity parameter (γ)

As discussed in Chapter 1, the twin-planes are atomically sharp interfaces separating

the two ZB ZnSe domains rotated with respect to each other by 60◦ (or odd multi-

ples thereof). Twin-plane spacing d in NTSLs is equal to d〈111〉N , where d〈111〉 is the

inter-planar spacing in 〈111〉-ZB direction and N is the number of monolayers between

successive twin-planes. d and d〈111〉 are shown in the schematic in Figure 1.2. In order to

characterize NTSLs, a periodicity parameter γ=d〈111〉/d=1/N is defined, such that each

NTSL can be represented as NTSL(γ). All NTSLs have been labelled as NTSL-i(γ),

where index i =1-7 corresponds to the 7 NTSLs presented in this chapter. Figures 7.1

(a)-(c) show HR-TEM micrographs of three representative NTSLs with different twin-

plane spacings d (=Nd〈111〉). The three NTSLs in Figures 7.1 (a)-(c) are NTSL-1(0.100),

NTSL-2(0.077) and NTSL-3(0.059). Figure 7.1 (d) shows a normal percentile distribution

of segment widths in NTSL-3(0.059). The individual domains are 〈111〉-ZB ZnSe with

〈111〉-direction coinciding with the growth direction of NTSL. This is confirmed by the

indexed spots in power spectrum (shown in Figure 7.1 (e)) taken at the area highlighted


in Figure 7.1 (c) of NTSL-3(0.059). Double (twin) spots in the SAD pattern are charac-

teristic of periodic twinning, and double spots seen in the SAD pattern from a larger area

of NTSL-3(0.059) (shown in Figure 7.1 (f)) further highlight the excellent periodicity of

twin-planes. The SAD pattern shows the diffraction spots corresponding to two rotated

ZB ZnSe domains, and these are shown as directions with and without a subscript T (de-

notes twinning) in Figure 7.1 (f). It is to be noted that most reported SADs from NTSLs

show streaks instead of double spots, thus indicating their compromised periodicity.

It was observed that there is a dispersion in the distribution of twin-plane spacings d

of an individual NTSL. The distribution of individual segment widths of NTSLs matches

a normal distribution which would be expected if the variation in d is random. The mean

(µ), standard error in mean (s.e.m.) (∆µ) and standard deviation (σ) of the distribution

are taken to represent d, error in d (=δd) and dispersion (∆d), respectively. The segment

widths were measured by counting the number of monolayers on HR-TEM images (for

smaller d’s), and also by measuring the widths of entire segments from TEM images (for

larger d’s) in which case they were rounded to the nearest multiple of d〈111〉. A small

value of ∆d=0.65 nm which translates to a remarkably low fluctuations of ±2 monolayers

in the normal percentile distribution of segment widths of NTSL-3(0.059) (shown in

Figure 7.1 (d)) indicates excellent periodicity of twin-planes, and compares favourably

with all previously reported growths of NTSLs [77, 104, 124, 132, 201]. Distributions of

segment widths for all NTSLs presented in this chapter are shown as histograms in Figure

7.2, and their statistical parameters described above are listed in Table 7.1. Effectively,

for small values of ∆d the NW is considered a NTSL with periodically arranged twin-

planes, while for large ∆d or large values of coefficient of variation, (COV=σ/µ) the

NW is labelled a randomly twinned NW (RTNW), where twin-planes occur at random

positions along the length of the NW with no observable periodicity. In samples studied

here, occurrence of both NTSLs and RTNWs were expectedly found and both were

characterized similarly.


4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5

0.01

1

10

40

70

95

99.5

Normal Probability Plot of NTSL-3 (0.059)

d = 5.635 nm d = 0.654 nm

Norm

al P

erc

entile

s

Width of segments (nm)

Percentiles

Reference Line

a

e

d

c

b

f

111

002 111

T 111 002

T

002 111

111

T

002T

<111>

111

111

220

111

002

131

131

002

111

220

222 222

131

220T

131

220

131T

220 220T

131T

111

Figure 7.1: Structural characterization of representative ZnSe NTSLs: HR-

TEM micrographs of ZnSe (a) NTSL-1(γ=0.100), (b) NTSL-2(γ=0.077) and (c) NTSL-

3(γ=0.059) viewed along the 〈110〉 zone-axis of NTSLs showing the periodically arranged

twin-planes. (d) Normal percentile distribution of segment widths (nm) of ZnSe NSTL-

3(0.059) with d=5.63±0.07 nm (s.e.m.) and ∆d=0.65 nm (s.d.). A narrow distribution

indicates a high degree of periodicity of twin-plane spacings in NTSL-3(0.059). Normality

tests are described in Table 7.1. (e) Indexed spots in the power spectrum taken from

the area highlighted in (c) confirm that the individual domains are 〈111〉-ZB ZnSe with

〈111〉-direction coinciding with the growth direction of NTSL-3(0.059). (f) Indexed SAD

pattern from an area of ZnSe NTSL-3(0.059), double diffraction spots in the SAD pattern

further demonstrate the excellent periodicity of twin-plane spacing d.


0 5 1 0 1 5 2 0 2 5 3 0 3 5 4 00

5

1 0

1 5

2 0

2 5

3 0

3 5

4 0 N T S L - 1 ( 0 . 1 0 0 ) N T S L - 2 ( 0 . 0 7 7 ) N T S L - 3 ( 0 . 0 5 9 ) N T S L - 4 ( 0 . 0 1 2 ) N T S L - 5 ( 0 . 0 1 9 )

No. o

f cou

nts

W i d t h o f s e g m e n t s ( n m )

Figure 7.2: Histograms representing the distributions of segment widths (nm) for NTSLs

1-5. The dashed curves are normal approximations to the distributions. The correspond-

ing twin-plane spacing d, error in d (δd) (s.e.m.) and dispersion in twin-plane spacing

∆d (s.d.) are listed in Table 7.1 along with the results of the normality tests.

Figure 7.3 (a) shows LTPL spectrum from NTSL-4(0.012), whose HR-TEM micro-

graph is shown in Figure 7.3 (b). As discussed previously in Chapter 1, the LTPL

spectrum from ZnSe, similar to most II-VI compound semiconductors, is usually com-

plex and dominated by emissions related to free and bound excitons, free-to-bound and

donor-acceptor pair (DAP) related transitions, as well as their phonon replicas [7]. The

accepted energy band gap of ZB ZnSe Eg,ZB is 2.822 eV (± 2-3 meV) [7], and associ-

ated LTPL peaks are well documented [7]. LTPL spectrum from a single ZB NW was

experimentally obtained (Chapter 5). The LTPL spectrum in Figure 7.3 (a) from NTSL-

4(0.012) shows features similar to those observed in LTPL from a ZB-NW-1, with free


NTSL Id. γ d (nm) δd (nm) ∆d (nm) Normality test NS

NTSL-1 0.100 3.218 0.148 1.152 Passed 61

NTSL-2 0.077 4.290 0.139 1.184 Passed 72

NTSL-3 0.059 5.635 0.073 0.654 Failed 81

NTSL-4 0.012 27.598 0.478 4.630 Passed 94

NTSL-5 0.019 16.908 0.234 1.843 Passed 62

Table 7.1: Twin-plane spacing d, error in d (δd) (s.e.m.) and dispersion in twin-plane

spacing ∆d (s.d.) for NTSLs 1-5 whose distributions of segment widths are shown in

Figure 7.2. NS is the sample size. Also listed are the results of Kolmogorov Smirnov

(K-S) normality test (for an alpha level of 0.05) for the distributions.

exciton emission peak blue-shifted by 1 meV. This blue-shift represents an increase in the

energy band gap of NTSL-4(0.012) relative to that for ZB ZnSe. The LTPL spectrum

from NTSL-4(0.012) also shows two bound exciton emission peaks related to a shallow

donor-bound exciton (D0X) (VSe-related) and an acceptor-bound exciton (A0X) (VZn-

related), and phonon replicas of the (A0X) peak up to 3rd order. The linewidths of the

exciton emission peaks are quite narrow (4 meV) and similar to the linewidths of corre-

sponding exciton peaks in LTPL spectrum from a single ZB NW acquired under similar

conditions. The integrated PL intensities (IPL) for the two cases are also comparable.

The linewidths and IPL of excitonic peaks observed in other individual NTSLs studied

in this work are similar to those from single ZB NWs. Since these linewidths are among

the narrowest reported so far on ZnSe NWs including the ones synthesised with highest

purity methods, these results illustrate that periodically arranged twin-planes have no

deleterious influence on the optical properties of NTSLs. The increase in Eg of NTSLs


2.65 2.70 2.75 2.80 2.85

D0X

FX

A0X

A0X

-1LO

PL (4K)

ZnSe NTSL-4 ( =0.012)

PL

In

ten

sity (

arb

. u

nits)

Energy (eV)

A0X

-2LO

A0X

-3LO

2.70 2.72 2.74 2.76 2.78 2.80 2.82

500

1000

1500

2000

Tim

e (

ps)

Energy (eV)

2.70 2.72 2.74 2.76 2.78 2.80 2.82

500

1000

1500

2000

Tim

e (

ps)

Energy (eV)

a

c

b

d

e

Figure 7.3: Optical and structural characterization of NTSL-4(γ=0.012): (a)

LTPL spectrum (4K) from NTSL-4(0.012) with dominant free (FX) and bound exciton

(D0X, A0X) related emission peaks. Multiple order phonon replicas of the A0X peak

are also observed. (b) HR-TEM micrograph viewed along the 〈110〉 direction of NTSL-

4(0.012) showing the periodically arranged twin-planes. (c) Spectrally and temporally-

resolved PL from ZnSe NTSL-4(0.012) showing the decay in time of individual emissions.

Streak images corresponding to TRPL from (d) NTSL-4(0.012) and (e) single ZB NW

(ZB-NW-1). TRPL decay curves extracted from streak images are shown in Figure 7.8.

is discussed below in terms of their polytype character.

As has been described previously in the context of the uniqe structure of NTSLs

(Chapter 1), the 〈111〉 direction in ZB crystal structures corresponds to the close-packed


arrangement of atomic planes in the sequence . . .ABCABC. . ., while for WZ crystal

structure (hexagonal close-packed (hcp)), the sequence is . . .ABAB. . . with each letter

representing a bilayer. The periodically arranged twin-planes in NTSLs perturb this

stacking sequence in one dimension (growth direction) which results in an apparent mirror

symmetry at the twin-plane and the observed side-facets [124, 132]. The result is a

periodic structure with stacking sequence . . .ABCABACBA. . ., with B indicating the

twin-plane where stacking sequence is mirrored. This structure qualifies as a polytype

structure [306] with a repeat unit encompassing one twin-plane and two domains with N

monolayers in each. Thus, a ZnSe NTSL with twin-plane spacing d(= Nd〈111〉) represents

a unique polytype of ZnSe, which can be denoted as 2NH-polytype or equivalently as

(2/γ)H-polytype, and it is clear that the number of such polytypes is infinite. A ZB

structure with no twin boundaries is represented by γ=0, while a WZ structure can be

considered as a stacking with a twin-plane at every atomic plane in 〈111〉-ZB direction

and is denoted as γ=1. NTSL-1(0.100), for example, represents a 20H polytype, and

similarly other higher order O(N) (or O(γ)) polytypes such as 4H (γ=0.5), 6H (γ=1/3),

etc. can be constructed. As another example, in the schematic in Figure 1.2 (e), the

repeat unit corresponds to the two twin-planes labelled C and separated by a distance

2d. In this case N=5 and the NTSL represents a 10H polytype. Such polytypes would

possess unique properties individually, and gradually varying overall with γ between two

extremes represented by ZB and WZ ZnSe. In particular, the Eg is expected to vary

between ZB and WZ band gaps with varying γ and this variation Eg(γ) is demonstrated

experimentally and supported through band-structure calculations below.

7.1.2 Excitation intensity dependence of LTPL

LTPL was also studied as a function of Iexc for individual NTSLs. Bao et al [133] have

previously studied LTPL from randomly twinned InP NWs (RTNWs) with varying Iexc

and observed a blue-shift in PL peak position with increasing Iexc. They assumed the


existence of one monolayer of WZ InP arising due to . . .ABA. . . stacking sequence at

the twin-plane and related their observed blue-shift to band filling effect arising due to

staggered band-alignment between ZB and presumed WZ segments. This interpretation

that twinning leads to a type-II band alignment at the twin-plane has been accepted by

many authors [132,206]. However, the perturbation in potential introduced by the twin-

plane (B) in an otherwise extended . . .ABCABCABACBACBA. . . stacking sequence

is too weak to localize electrons or holes required for spatially indirect PL transitions.

The argument of type-II (staggered) band-alignment is however valid for a true WZ-ZB

superlattice structure consisting of extended ZB and WZ domains, and would lead to a

blue-shift in PL when Iexc is increased which has been observed experimentally by Pe-

masiri et al [127] on InP WZ NWs containing ZB segments. Caroff et al [122, 124] have

also dismissed Bao et al’s [133] interpretation and argued that NTSLs be considered as

periodic twinning based polytype structures. Thus, the PL transitions in NTSLs are

direct, and no blue-shift in PL peak position(s) is expected with varying Iexc (excluding

the high excitation regime). LTPL spectra from NTSL-6(0.019) with increasing Iexc and

the corresponding change in peak energies for individual emission peaks are shown in

Figures 7.4 and 7.5, respectively. No blue-shift was observed in any emission peak. Fur-

ther, no blue-shift was observed in PL peak positions in a similar Iexc-dependent LTPL

experiment on a RTNW. Thus, based on Iexc-dependent LTPL experiments, it is clear

that NTSLs behave as polytype structures with no evidence of staggered band-alignment.

This is also supported by ab Initio calculations discussed briefly below. Figure 7.4 shows

excitation intensity (Iexc) dependent PL spectra from NTSL-6(γ=0.019), where Iexc was

varied from 6.25 W/cm2 to 500 W/cm2. All spectra were recorded at a constant temper-

ature (4± 1 K), and were fitted to individual emission peaks using Lorentzians. Change

in peak position for two such emission peaks is shown in Figure 7.5 with varying Iexc.

The PL peaks showed a slight red-shift, not exceeding 4 meV, related to the unavoidable

laser heating under ultra-fast optical excitation conditions and irrelevant to the discussion


here.

2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5 2 . 9 0 2 . 9 5

P 0 = 6 . 2 5 W / c m 2

P 0

2 P 0

4 P 0

PL In

tensity

(arb.

units)

E n e r g y ( e V )

2 0 P 0

4 0 P 0

6 0 P 0

8 0 P 0

8 P 0

Figure 7.4: Variation in LTPL spectra obtained from NTSL-5(0.019) as a function of

excitation intensity Iexc. All spectra were acquired at a constant temperature (4 K),

normalized to unity and are displayed offset for clarity.

7.2 Variation in band gap of NTSLs

The variation in electronic structure and optical properties of NTSLs with varying γ is

discussed below. The electronic band gap of ZB ZnSe is 2.822 eV at 0K [7]. Liang and


0 1 0 0 2 0 0 3 0 0 4 0 0 5 0 02 . 8 0 0

2 . 8 0 5

2 . 8 1 0

2 . 8 1 5

2 . 8 2 0

Peak

energ

y (eV

)

E x c i t a t i o n I n t e n s i t y ( W / c m 2 )

Figure 7.5: Change in peak energies as a function of Iexc for two individual emission peaks

observed in the LTPL from NTSL-5(0.019). Each spectrum shown in Figure 7.4 was fit

to individual emission peaks using Lorentzians, and the peak energies thus obtained are

shown with varying Iexc. The solid lines are guide to the eye.


Yoffe [238] have reported the energy band gap of WZ ZnSe as Eg,WZ=2.874 eV, and

this value has been widely accepted in the literature. Similar to the case of ZB NWs,

the LTPL spectrum from a single WZ NW was experimentally obtained (Chapter 6),

and the Eg was found to agree well with the value reported by Liang and Yoffe [238].

Figure 7.6 (a) shows the variation in PL spectra for NTSLs with different periodicity

parameters (γ) representing polytypes of different orders (also indicated), ZB and WZ

NWs. The corresponding HR-TEM micrographs are shown in Figure 7.6 (b) taken at

similar magnification factors. Note, that the LTPL spectra and HR-TEM micrographs

in Figure 7.6 for ZB (γ=0) and WZ NWs (γ=1) are those of ZB-NW-1 and WZ-NW-1

presented in Chapters 5 and 6, respectively. It is noted that the NTSLs denoted by ∗ in

Figure 7.6 (a) showed PL peaks corresponding to the emission from the band-edges of

NTSLs. This is due to the influence of hole-traps in ZnSe which are also responsible for

the phenomenon of persistent photoconductivity in ZnSe, combined with the ultra-fast

excitation conditions. This phenomenon has been observed experimentally for heavily n-

doped ZnSe epilayers [307–309], and discussed further using dependence of LTPL spectra

and LTPL peak positions on temperature later in this chapter. Thus, it is observed that

the PL peak positions from NTSLs monotonically shift to higher energies with increasing

periodicity parameter γ, but not exceeding that of the band gap of WZ ZnSe. For NTSLs

exhibiting free exciton emission the band gaps were estimated using an exciton-binding

energy EFX of 21 meV [7]. This is meaningful since the EFX in WZ ZnSe has been shown

to be similar to that of ZB ZnSe (difference=1 meV) [238], and thus a change in EFX

for NTSLs would be <1 meV. This may not be true for NTSLs based on other III-V and

II-VI compound semiconductors. The variation in band gaps of NTSLs with γ is shown

in Figure 7.7. NTSLs with γ values up to 0.33 were studied. The end-points with γ=0

and 1 correspond to the pure ZB and WZ crystal structures of ZnSe, respectively. The

variation in band gap is near linear.

The electronic structures for NTSLs were calculated using ab Initio pseudo-potential


method, the details of which can be found in the Appendix E.2. As is well-known,

this method of electronic structure calculations significantly underestimates the band

gaps [310]. For comparison with the experimental values, a constant shift (known as

a scissors-shift) of 1.642 eV was applied uniformly to all structures so as to align the

calculated band gap of ZB ZnSe with the experimental value. The band gaps for all

NTSLs were found to be direct, and folded in k-space as required by the periodicity. See

Figures E.3 and E.4 in the Appendix E.2, for the calculated electronic structures of 2H,

4H and 6H structures. As shown in Figure 7.7, the calculated band-gap as a function

of γ shows a near linear variation with the band-gap gradually increasing from pure

ZB structure to pure WZ structure by about 63 meV. These features are in excellent

agreement with the experiments and provide support for the assignment of the LTPL

peaks discussed above. Based on the above discussion, the energy band gap of NTSLs as

a function of γ can be expressed as

Eg(γ) = Eg,ZB + (Eg,WZ − Eg,ZB)γ (7.1)

7.3 Time-resolved photoluminescence

Time-resolved PL (TRPL) provides useful information regarding the radiative and non-

radiative recombination channels. Bao et al. [133] have further speculated in support

of their hypothesis that the PL lifetimes in RTNWs would be longer than homogeneous

ZB or WZ NWs due to the spatial separation of electrons and holes at the twin-plane

interfaces. It was argued that this spatial separation of electrons and holes would require

more time for them to recombine, and hence result in longer PL lifetimes. This was as-

cribed to the monolayer WZ sections, assumed to exist at the twin boundaries, localizing

the holes, while electrons being localized in the extended ZB domains. This change in PL

lifetimes was experimentally observed in heterostructure NWs with extended ZB and WZ

domains in InP NWs [127]. However, as explained earlier, the perturbation in potential


2.75 2.80 2.85 2.90

2 H

104 H

168 H

34 H

26 H

PL

In

ten

sity (

arb

. u

nits)

Energy (eV)

=0

=0.077*

=0.100*

=1

=0.059*

=0.019*

=0.012

20 H

a b

Fig

ure

7.6:

Vari

ati

on

inP

Lsp

ect

raas

afu

nct

ion

of

peri

odic

ity

para

mete

rγ:

(a)

LT

PL

(4K

)sp

ectr

aob

tain

edfr

om

five

diff

eren

tN

TSL

sw

ith

vary

ing

per

iodic

ity

par

amet

ersγ

.P

erio

dic

ity

par

amet

erγ

and

the

order

ofth

ep

olyty

pe

(2/γ

)H(o

r

2NH

)ar

ein

dic

ated

inth

efigu

re.

Als

osh

own

are

the

exp

erim

enta

lly

obta

ined

PL

spec

tra

from

singl

eZ

B(γ

=0)

,an

dsi

ngl

e

WZ

(2H

)(γ

=1)

NW

s.T

her

eis

am

onot

onic

blu

e-sh

ift

inLT

PL

spec

tra

wit

hin

crea

singγ

indic

atin

gth

ein

crea

sein

ener

gyban

d

gap

for

pol

yty

pes

ofZ

nSe

wit

hm

axim

um

ban

dga

pfo

rW

ZZ

nSe.

(b)

HR

-TE

Mm

icro

grap

hs

corr

esp

ondin

gto

the

NT

SL

s(γ

),

ZB

and

WZ

NW

sw

hos

eLT

PL

spec

tra

are

dis

pla

yed

in(a

).T

he

HR

-TE

Mm

icro

grap

hs

wer

eta

ken

atsi

milar

mag

nifi

cati

on

fact

ors

alon

g〈1

10〉

zone-

axis

.


0 . 0 0 . 2 0 . 4 0 . 6 0 . 8 1 . 0

2 . 8 2

2 . 8 4

2 . 8 6

2 . 8 8 B a n d - g a p : E x p e r i m e n t a l

Energ

y ban

d gap

(eV)

P e r i o d i c i t y p a r a m e t e r γ

B a n d - g a p : C a l c u l a t i o n s

Figure 7.7: Variation in energy band gaps of ZnSe NTSLs as a function of the

periodicity parameter γ: Projected and measured band gaps (red) for six different

NTSLs with varying periodicity parameters γ, ZB and WZ NWs are compared with band

gaps obtained from band-structure calculations (blue) for ZB and WZ ZnSe, and higher

order polytypes. The band gaps obtained from electronic structure calculations for all

structures were scissor-shifted [310] by 1.642 eV to match the ZB ZnSe band gap with

its experimental value. Electronic structures obtained from ab Initio calculations for

2H, 4H (γ=0.5) and 6H (γ=0.333) polytypes of ZnSe are shown in the Figure E.3 while

Figure E.4 compares the calculated band-structures for 2H and 4H polytypes.


0 5 0 0 1 0 0 0 1 5 0 0 2 0 0 0

τ1 = 1 4 7 p s , τ2 = 6 6 1 p s

T R P L D e c a y N T S L - 4 ( 0 . 0 1 2 )

PL In

tensity

(arb.

units)

T i m e ( p s )

τ1 = 2 4 p s , τ2 = 6 0 6 p s

T R P L D e c a y Z B - Z n S e N W

Figure 7.8: Low-temperature photoluminescence (4 K) decay from NTSL-4(0.012) (red)

and single ZB NW (ZB-NW-1 from Chapter 5) (blue) extracted from the streak images

shown in Figures 7.3 (d) and (e), respectively, with an energy window of 20 meV centred

around the acceptor-bound exciton peak (A0X) (the strongest peak). Both decay curves

are characterized by a bi-exponential decay with comparable lifetimes as shown.

at twin-planes in NTSLs is insufficient to localize holes, and thus such an increase in PL

lifetimes is not expected. In fact, the PL lifetimes from NTSLs and RTNWs would be

comparable to those from the homogeneous NWs, dictated primarily by the oscillator

strengths of relevant transitions. As discussed earlier, the calculated electronic structure

indeed shows that the transition dipole moments and hence the oscillator strengths for

inter-band transitions are nearly independent of the periodicity. Figure 7.3 (c) shows


spectrally and temporally resolved PL decay from NTSL-4(0.012) with distinguishable

decay channels for different exciton emissions. Figures 7.3 (d) and (e) are the streak

images representing PL decay obtained from NTSL-4(0.012) and single ZB NW (ZB-

NW-1), respectively. It is readily observed that the PL decay is similar in both cases and

TRPL decay curves extracted from the two streak images yield similar PL lifetimes as

shown in the Figure 7.8.

7.4 Dependence of PL on temperature

PL from NTSL-7(γ=0.014) was measured as a function of temperature, and the variation

in PL with temperature is shown in the Figure 7.9 for temperatures up to 150 K. Each

LTPL spectrum was fit to the individual emission peaks, and the fit parameters obtained

are discussed next. Figure 7.10 shows the change in peak energy for the highest energy

peak (2.82 eV) at different temperatures for NTSL-7(0.014) (red). This peak is related

to the band gap of NTSL-7 corresponding to the recombination of free carriers. The solid

and dashed lines are fits to Equations 5.3 (Varshni’s equation) and 5.4 (Bose-Einstein

type expression), respectively. The parameters obtained from these fits are listed in

Table 7.2, where α represents the temperature gradient (dEg/dT ) of the band gap of

NTSL-7. Experimentally obtained peak positions of the free exciton peak at different

temperatures for ZB NW are also shown in Figure 7.10, with solid and dashed lines

representing the fits to Equations 5.3 and 5.4, respectively. The fit parameters are also

listed in Table 7.2. (These were determined earlier in Chapter 5 for ZB-NW-4). It

was explained in Chapter 6 that the temperature gradient of WZ NWs could not be

determined based on the dependence of PL on temperature, since the FX peak was not

observed in WZ NWs. However, Liang and Yoffe [305] have reported the energy positions

of free exciton absorption at four temperatures based on their absorption experiments

on WZ ZnSe single crystals. These values are shown in Figure 7.10 along with the fits to


2 . 7 0 2 . 7 5 2 . 8 0 2 . 8 5

N T S L - 7 ( γ= 0 . 0 1 4 )I e x c = 3 1 . 8 W / c m 2

4 K 1 0 K 2 0 K 3 0 K 4 0 K 5 0 K 6 0 K 7 0 K 8 0 K 9 0 K 1 0 0 K 1 1 5 K 1 3 0 K 1 5 0 K

PL In

tensity

(arb.

units)

E n e r g y ( e V )Figure 7.9: Temperature dependent PL spectra from NTSL-7(0.014). All spectra were

acquired at a fixed average Iexc=31.2 W/cm2, are normalized by the integration times

(ti) and displayed offset for clarity. The LTPL spectra taken at temperatures above 150

K are not shown.


Table 7.2: Values of the fit parameters of Equations 5.3 (Varshni’s equation) and 5.4

(Bose-Einstein type expression) which describe the temperature dependence of the Eg.

Process Eg(0) α β Source

(eV) (10−4 eV/K) (K)

WZ ZnSe FX 2.853 ± 0 5.0 ± 0.8 139.7 ± 68.7 Ref. [238]

NTSL-7(0.014) Eg 2.820 ± 0 6.5 ± 0.4 222.6 ± 27.8 Expt.

ZB-NW-3 FX 2.799 ± 0 8.9 ± 0.7 361.5 ± 44.8 Expt. (Chapter 5)

Process Eg(0) αB θB Source

(eV) (meV) (K)

WZ ZnSe FX 2.852 ± 0 30.9 ± 2.0 143.8 ± 7.3 Ref. [305]

NTSL-7(0.014) Eg 2.818 ± 0 45.4 ± 1.5 183.4 ± 4.4 Expt.

ZB-NW-3 FX 2.797 ± 0 56.5 ± 4.7 209.3 ± 11.6 Expt. (Chapter 5)

Equations 5.3 (solid line) 5.4 (dashed line). The fit parameters from these are also listed

in Table 7.2. It is important to note that the dEg/dT for NTSL-7 lies between those of ZB

and WZ ZnSe. This further supports the claim that ZnSe NTSLs are polytype structures

of ZnSe in which case their physical properties would be expected to lie between those

of the ZB and WZ ZnSe.

The change in IPL with temperature for NTSL-7(0.014) is shown in Figures 7.11 (a,b)

for the (a) free-carrier recombination and (b) VSe-related donor bound exciton emission,

respectively. The solid lines are fits to the one process thermal quenching model (Equa-

tion 5.6). The activation energy for thermal quenching for free-carrier recombination

was found to be 20 meV. For the (D0X) emission, the activation energy was found to

be 7 meV which is identical to the binding energy of (D0X) excitons (discussed in detail

in Chapter 4). The change in linewidth (HWHM) with temperature for the band gap

recombination from NTSL-7(0.014) is shown in Figure 7.12, where the solid line is a fit

to the expression in Equation 5.5. Note that in this case the second and third terms


0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 02 . 7 0

2 . 7 2

2 . 7 4

2 . 7 6

2 . 7 8

2 . 8 0

2 . 8 2

2 . 8 4

2 . 8 6

Energ

y (eV

)


W Z - Z n S e B u l k ( F X ) - R e f . N T S L - 7 γ ( 0 . 0 1 4 ) E g - E x p t . Z B - N W - 3 ( F X ) - E x p t . F i t F i t

Figure 7.10: Variation in PL peak energy of the emission corresponding to the band gap

recombination (red, for NTSL-7(0.014)) with temperature. Change in PL peak energy

for the FX emission for ZB-NW-3 (green, experimental) and energies corresponding to

the FX absorption for WZ ZnSe [238] (blue) are also shown. The solid and dashed lines

in each case are fits to the Equations 5.3 and 5.4, respectively.

on the right side in Equation 5.5 represent the interaction of electron with acoustic and

optical phonons, respectively. The parameters obtained from the fit are listed in Ta-

ble 7.3. For comparison these parameters obtained for the FX emission from ZB-NW-4

(in Chapter 5) are also listed. The term representing inhomogeneous broadening (Γinh) is

larger for NTSL-7 compared to that for ZB-NW-4, and this is due to the phenomenon of

screening of excitons. The value of ΓLO is comparable in both cases, while γth is almost

half for NTSL-7 compared to that for ZB-NW-4.


0 5 0 1 0 0 1 5 0 2 0 0 2 5 0

( b )

1 0 0 0 / T ( K - 1 )

N T S L - 7 γ( 0 . 0 1 4 ) I P L - E x p t . F i t t o o n e s t e p - p r o c e s s

E p = 2 . 8 1 8 e V E a = 2 0 m e V

Integ

rated

PL In

tensity

(I PL, a

rb. un

its)

1 0 0 0 / T ( K - 1 )

( a )

0 5 0 1 0 0 1 5 0 2 0 0 2 5 0

N T S L - 7 γ( 0 . 0 1 4 ) I P L - E x p t . F i t t o o n e - s t e p p r o c e s s

E p = 2 . 7 9 5 e V E a = 7 m e V

Figure 7.11: Change in integrated PL intensity (IPL) for the emission corresponding to

the band gap recombination with temperature for NTSL-7(0.014). The solid line is a fit

to the Equation 5.6. Note the log scale for IPL.

Table 7.3: Values of the fit parameters of Equation 5.5 which describes the temperature

dependence of the exciton linewidth (HWHM).



NTSL-7(0.014) Eg 6.0 ± 0.5 20.9 ± 5.3 360a 32.8 ± 7.7 Expt.

ZB-NW-3 FX 2.6 ± 1.0 24.3 ± 9.6 360a 60.6 ± 14.0 Expt. (Chapter 5)

a Parameter fixed.

7.5 Conclusions

In this chapter, optical results supported by electronic structure calculations, for nanoscale

electronic structure engineering using ZnSe NTSLs were presented. A linear dependence

of electronic band gap on twin-plane spacing using PL peak positions is shown, with


0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 05

1 0

1 5

2 0

2 5

3 0

HWHM

(meV

)


N T S L - 7 γ ( 0 . 0 1 4 ) H W H M - E x p t . F i t

Figure 7.12: Change in HWHM for the band gap recombination with temperature for

NTSL-7(0.014). The solid line is a fit to the Equation 5.5.

band gap that varies between their values in ZB and WZ crystal structures. NTSLs are

thus unique polytypes of the parent material and possess electronic and optical prop-

erties governed only by the twin-plane spacing. Since NWs have the unique ability to

crystallize as NTSLs a similar variation in band gaps of NTSLs based on other III-V

and II-VI compound semiconductors is expected thus adding a new dimension to NW

functionality. The optical performance of NTSLs was shown to be comparable to those

of homogeneous NWs, based on a comparison of their Γp and IPL. The experimental

evidence of band-structure engineering using periodically arranged twin-planes combined

with their cost-effective synthesis presented here should open the possibility of opto-

electronic device fabrication using nanostructured materials with on-demand engineered

properties.


Dependence of PL on temperature was also studied for a single NTSL, and parameters

α, β, αB, and θB related to the dependence of Eg on temperature were determined. These

were found to be intermediate between those for ZB and WZ crystal structures. The

other parameters that were extracted include those related to inhomogeneous broadening

(Γinh=6.0±0.5), strength of the exciton-acoustic phonon (γth=32.8±7.7) and exciton-LO

phonon (ΓLO=20.9±5.3) interactions. The activation energy for thermal quenching of

band gap luminescence was found to be 20 meV, while that for donor-bound exciton

emission as 7 meV.

Chapter 8

Conclusions and Outlook

Experimental demonstration of the nanoscale band-structure engineering was presented

using extensive optical and structural characterization on individual ZnSe based NTSLs,

supported by ab Initio band-structure calculations. It was shown that the Eg and the

electronic band structure of ZnSe NTSLs systematically vary with their twin-plane spac-

ing, and that NTSLs represent a new class of nanoscale polytype structures. Each of

these essentially infinite number of polytypes, similar to bulk polytypes [311], possesses

characteristic physical properties governed by their twin-plane spacing (d). A linear vari-

ation in Eg was shown through a monotonic shift in PL peak position from ZnSe NTSLs

as a function of their twin-plane spacing, which was quantified by using a periodicity pa-

rameter γ. It was shown that the Eg of a ZnSe NTSL lies between the Eg of ZnSe in ZB

and WZ crystal structures, and depends only on γ. It was also established that a widely

used model that TSLs can be interpreted as ZB-WZ superlattices is not appropriate for

describing the electronic structure of NTSLs. The linewidths and integrated PL intensi-

ties in PL from NTSLs were shown to be comparable to those from homogeneous NWs,

thus demonstrating that the periodic twin-planes in NTSLs have no adverse effect on the

optical response of NTSLs. The periodic twin-planes in NTSLs lead to new materials

with characteristic optical properties in contrast to the view of twin-planes as defects

173

Chapter 8. Conclusions and Outlook 174

that should be avoided. NTSLs studied in this work were synthesized via a cost-effective

method which is a significant advantage in terms of their potential applications.

Temperature dependence of PL from a single NTSL was used to determine the pa-

rameters in Varshni’s equation and Bose-Einstein type expression, which describe the

temperature dependence of Eg. These were found to lie between those of ZB and WZ

crystal structures. The dependence of linewidth of band gap related luminescence was

used to determine the parameters related to broadening of luminescence lines. The

constants for coupling between exciton-acoustic phonon and exciton-LO phonon were

deduced. The band gap luminescence was quenched with increasing temperatures with

an activation energy of 20 meV. The activation energy for thermal quenching of neutral

donor-bound exciton emission at 2.795 eV was 7 meV, in agreement with the binding

energy of the donor-bound excitons. The identification of recombination lines in NTSLs

was based on an unambiguous pre-determination of the recombination centres in NWs

with varying stoichiometry, and for ZB and WZ NWs in direct relation to their crystal

structure, as summarized below.

LTPL from NWs was studied as a function of modulation of their stoichiometry. This

was achieved by their growth in the conditions of excess of Zn and excess Se. LTPL

from excess Zn NWs was also studied with varying Iexc. For NWs with excess Zn, strong

emission related to excitons bound to neutral donors was observed at 2.794 eV. The

binding energy of the exciton to the neutral donor, identified as VSe, was determined to

be 7 meV, and the ionization energy of the donor as 35 meV. Two peaks at 2.714 eV

and 2.686 eV due to DAP recombinations were also observed. Based on the dependence

of LTPL on Iexc, the binding energies for both the donors associated with these DAP

recombinations were calculated to be 27±1 meV, whereas those of the acceptors as 102.5

meV and 139 meV, respectively. These donors and acceptors were ascribed to the defect

complexes formed by single native point defects such as VSe, VZn, Zni and Sei with other

native defects. Y-line emission, which is related to the presence of structural defects such


as stacking faults and dislocations in ZnSe was also observed. The LTPL from NWs

with excess Se was compared with that from NWs with excess Zn. The emission line

related to the excitons bound to neutral donors at VSe, observed in the case of NWs with

excess Zn, was absent in the case of NWs with excess Se. No other exciton-related line

was observed in the latter case, and the LTPL spectrum was dominated by DL emission

bands at 2.27 eV and 1.96 eV. These DL emissions were related to the presence of VZn

in the NWs with excess Se. These observations confirm that the optical transitions in

ZnSe NWs are primarily dictated by the native point defects.

The optical response from single ZB ZnSe NWs was studied using LTPL. The crystal

structure of these NWs was confirmed to be ZB, with a single crystalline microstructure

using TEM, HR-TEM and SAD. The optical emission was found to be dominated by the

native point defects, while no emission peaks related to the unintentional impurities were

detected. This confirms the observation noted above that the optical properties of ZnSe

NWs are governed only by the native point defects. Two strong PL lines were observed

at 2.785 eV and 2.780 eV. Both of these were assigned to the excitons bound to deep

neutral acceptors related to VZn and (VZn-complex), respectively. These were labelled

as Id1 and Id′

1 lines, respectively. The binding energies of the excitons responsible for the

Id1 and Id′

1 lines are 15 meV and 20 meV, respectively, and the ionization energy of the

associated deep acceptors are 150 meV and 200 meV, respectively. The Id1 and Id′

1 lines

were accompanied by their LO-phonon replicas, which are characteristic of the excitons

bound to deep neutral acceptors. The phonon-replicas of up to 3 orders were observed for

both lines, and the average number of emitted phonons was found to be 0.53 < Nph < 0.7

for the Id1 line, and 0.25 < Nph < 0.42 for the Id′

1 line. Another recombination peak at

2.800 eV was also observed, and related to the free-exciton (FX) emission in ZnSe. These

findings on the Id1 and Id′

1 lines, their phonon-replicas as well as the number of emitted

LO phonons have not been reported thus far in ZnSe NWs.

Excitonic emission linewidths of ∼1.5 meV for the Id1 line were observed in single ZB


NWs, which are the narrowest excitonic linewidths reported so far on ZnSe NWs. The

narrowest linewidth previously reported is 10 meV for ZnSe NWs grown by MOCVD

method. The observation of excitonic lines with linewidths close to 1 meV from NWs

grown by CVD growth method in the present case shows that the crystalline quality and

the optical performance of NWs synthesized by the CVD method are comparable to those

from other expensive and time-consuming methods. These narrow linewidths are also

comparable to the narrowest reported linewidths in PL from NWs of other semiconductor

materials. The optical response from a single NW was compared with that from a bundle

of NWs. It was found that the energy position of recombination lines are identical in

both cases, and the Y-line emission related to structural defects was not observed in

either case. This confirms that the Y-line emission actually does not originate from

homogeneous NWs, but rather from the growth of crystallites of ZnSe on the sample

substrate beneath the layer of NWs. The linewidths of the excitonic Id1 emission from

the bundle of NWs were slightly larger than those from single NWs, due to the ensemble

broadening. The linewidths of the Id1 line in different NWs were also slightly different,

and the ensemble broadening effect can also be understood by considering that in the

case of a bundle of NWs, the broadening is limited by the NW which exhibits the largest

linewidths.

PL from single ZB NWs was measured at varying temperatures. The assignment of

the emission peak at 2.800 eV to FX was confirmed by the temperature dependence of

the peak energy of this emission. This was also used to determine the high-temperature

limit of dEg/dT , which was found to be in excellent agreement with the values reported

in the literature for bulk ZnSe. The parameters of the Bose-Einstein type expression

were determined for the FX emission, and found to be in good agreement with those

previously reported for bulk ZnSe. These parameters were also deduced for the Id1 and

Id′

1 lines using dependence of their peak energy on temperature. The change in HWHM

with temperature for the FX emission, Id1 and Id′

1 lines revealed a small contribution of


the inhomogeneous broadening to the linewidths. This is consistent with the observation

of narrow linewidths for the Id1 and Id′

1 lines. A strong exciton-acoustic phonon and

exciton-LO phonon coupling was found for these lines in accord with the observation of

LO-phonon replicas in the PL spectra, and large average number of emitted phonons for

these lines. The integrated PL intensity of the FX emission was quenched with increasing

temperatures by two mechanisms. The first mechanism with an activation energy of 20

meV is related to the dissociation of free excitons with a binding energy of 21 meV. The

other mechanism with an activation energy of 136 meV could be a result of non-radiatve

recombination related to surface recombination. The Id1 and Id′

1 line were quenched

with increasing temperatures by mechanisms with activation energies smaller than their

binding energies. These are proposed to be due to non-radiative recombination induced

by scattering of excitons responsible for these lines with the simultaneously emitted LO-

phonons.

The optical response of single WZ NWs was studied using LTPL, and similar to

the case of ZB NWs, their crystal structure and single crystalline microstructure was

confirmed using structural characterization techniques. The centres responsible for the

recombination lines were identified based on the previous work by Liang and Yoffe [305]

and by reference to the identification of PL lines in ZB NWs. Similar to the ZB NWs, the

optical properties were dictated by the native point defects in WZ NWs. Two lines related

to the excitons bound to neutral acceptors were observed, but only one of them was due

to the deep acceptors. A recombination line at 2.841 eV, labelled as I1, was assigned to

VZn as neutral acceptors. The binding energy of this exciton is 11 meV, and the ionization

energy of the acceptor 110 meV. Another recombination line at 2.833 eV, labelled as Id1 ,

was ascribed to the excitons bound to deep neutral acceptors related to (VZn-complex),

by reference to their observation in ZB NWs. The binding energy of the exciton and the

ionization of the acceptor are 19 meV and 190 meV, respectively. DAP related transitions

were also observed in the LTPL spectra from WZ NWs, along with their phonon-replicas


of up to 3 orders. The average number of phonons was found to be 0.69, identical to that

reported by Liang and Yoffe [305]. Other transitions observed in the LTPL spectra were

assigned to the simultaneous emission of acoustic and optical phonons, and a transition

scheme was proposed. This recombination scheme is an extension of the one proposed

by Liang and Yoffe [305] for different peaks. This is the first time since the report on PL

from WZ ZnSe by Liang and Yoffe [305] that a recombination scheme has been proposed.

Linewidths of ∼2.8 meV were obtained for the I1 line, and these are the narrowest

linewidths reported for WZ ZnSe NWs. PL from single WZ NWs was also measured

as a function of temperature. The parameters of the Bose-Einstein type expression for

the dependence of Eg on temperature were determined for the I1 and Id1 lines. The

parameters related to inhomogeneous broadening of the exciton lines, strengths of the

exciton-acoustic phonon and exciton-LO phonon coupling were deduced for the I1 line

from the dependence of HWHM on temperature. The integrated PL intensity of the I1

line was quenched by a mechanism with an activation energy of 9 meV, similar to the

binding energy of the excitons responsible for the I1 line. The centres responsible for

recombination lines identified in this thesis, binding energies of the excitons, and ioniza-

tion energies of the donors and acceptors are summarized in Table 8.1. The parameters

determined using the dependence of PL on temperature for single ZB and WZ NWs and

a single NTSL are summarized in Table 8.2. These parameters for the ZB and WZ NWs

were compared in Chapter 6 for lines due to similar recombination mechanisms.

It needs to be emphasized that the above work would not have been possible without

the implementation of a new experimental methodology, which is also an important

contribution of this thesis. The new approach was based on the cost-effective design

of a sample-holder to bridge the techniques of optical characterization (PL and TRPL)

and structural characterization (TEM). The sample-holder allows one to overcome the

critical challenge of incompatibility in different sample preparation methods required for

these characterization techniques. The use of these techniques independently is quite


common for semiconductor characterization, but their combined use provides a powerful

means to remove unambiguities in the interpretation of optical emission spectra. The

sample-holder was used to hold the TEM grids during the optical characterization, which

were removed from it for the TEM characterization. The sample holder allows the use

of the same TEM grid multiple times, with careful handling of the fragile grids. This is

useful, e.g., (i) if more NTSLs or NWs are to be studied on the same grid, and (ii) for

further PL and TRPL experiments.

In a broader perspective, it was shown that:

1. NWs of semiconductor materials can crystallize in unique variations, NTSLs, which

are not known to occur in the bulk form for most materials, including ZnSe. NTSLs are

based on a periodic arrangement of twin-planes in NWs along their growth direction,

which is 〈111〉 in ZB structures.

2. NTSLs are new nanoscale polytypes of their parent materials, essentially infinite in

number, with physical properties, such as Eg, intermediate to those of the parent material

in the ZB and WZ crystal structures. Since the strategies employed thus far, e.g. by a

change in the diameter, strain and composition, proved ineffective in yielding NWs with

tunable optoelectronic properties and desired performance, NTSLs provide an attractive

alternative to achieving this tunability.

3. Further to the above, the optical performance of NTSLs was shown to be at par with

those of homogeneous NWs. This is important because, but for this parity, the tunable

properties offered by NTSLs would have little significance in practical applications.

4. The NWs synthesized using a fast and low-cost CVD method, following a VLS growth

mechanism, show excellent crystallinity and optical performance, comparable to NWs

obtained using expensive and time-consuming methods such as MBE and MOCVD. This

is an important result which should open the possibility of a large-scale synthesis of NWs

for device applications.


Table 8.1: Summary of the recombination lines identified in ZB and WZ ZnSe NWs.

The exciton binding energies (EBX) for donor and acceptor bound excitons, ionization

energies of the donors (ED) and acceptors (EA) are also listed.

Ep (eV) Identification Donor/acceptor EBX ED,EA

(meV) (meV)

ZB ZnSe NWs

2.800 FX free excitons

2.794 D0X VSe(a) 7 35

2.785 (Id1 ) A0X VZn(b) 15 EA=150

2.780 (Id′

1 ) A0X (VZn-complex)(b) 20 EA=200

2.714 DAP defect-complex ED=27±1

EA=102.5

2.686 DAP defect-complex ED=27±1

EA=139

2.603 Y-line extended defects

WZ ZnSe NWs

2.841 (I1) A0X VZn(c) 11 EA=110

2.833 (Id1 ) A0X (VZn-complex)(b) 19 EA=190

2.741 DAP

(a): neutral donors, (b): deep neutral acceptors, (c): neutral acceptors


Table 8.2: Summary of the values of the fit parameters determined using dependene of

PL on temperature for single ZB and WZ NWs and a single NTSL.

NW/NTSL PL line Eg(0) (eV) α (10−4 eV/K) β (K)

ZB-NW-3 FX 2.799 ± 0 8.9 ± 0.7 361.5 ± 44.8

NTSL-7(0.014) Eg 2.820 ± 0 6.5 ± 0.4 222.6 ± 27.8

Eg(0) (eV) αB (meV) θB (K)

ZB-NW-3 FX 2.797 ± 0 56.5 ± 4.7 209.3 ± 11.6

Id1 2.785 ± 0 59.4 ± 11.1 201.9 ± 20.1

Id′

1 2.779 ± 0 41.1 ± 21.4 141.4 ± 34.0

WZ-NW-1 I1 2.845 ± 0 69.8 ± 9.2 250.4 ± 14.6

I1 2.845 ± 0 64.6 ± 9.6 241.0 ± 16.8

NTSL-7(0.014) Eg 2.818 ± 0 45.4 ± 1.5 183.4 ± 4.4

Γinh (meV) ΓLO (meV) θLO (K) γth (µeV/K)

ZB-NW-3 FX 2.6 ± 1.0 24.3 ± 9.6 360a 60.6 ± 14.0

Id1 1.6 ± 1 153.0 ± 62.7 360a 103.3 ± 25.0

Id′

1 4.5 ± 0.4 913.1 ± 92.2 360a 48.1 ± 13.5

WZ-NW-1 I1 4.0 ± 0.2 110.6 ± 10.0 360a 32.7 ± 4.1

NTSL-7(0.014) Eg 6.0 ± 0.5 20.9 ± 5.3 360a 32.8 ± 7.7

a Parameter fixed.

A fitting error of 0 indicates error of the order of 10−5 eV.


5. The ZnSe NWs occur in both ZB and a wider band gap WZ crystal structure, and

exhibit excellent luminescence properties, no inferior to those of NWs based on any other

compound semiconductor, and the best in the blue-region of the spectrum. Further, the

range of optical emission can be tuned as desired using ZnSe NTSLs, which can be easily

obtained by a low-cost synthesis method. ZnSe NWs and NTSLs are therefore ideal

candidates for blue light-emitting optical and optoelectronic devices.

8.1 Outlook

1. The NTSLs studied in this thesis were obtained using their spontaneous growth, i.e.,

with limited control over their occurrence and no control over their periodicities. A precise

control over the periodicty of twin-planes in NTSLs has been demonstrated [124, 132],

but by methods such as MOCVD. Achievement of this control over d in NTSLs with a

CVD growth method would represent a significant advance, and further experiments in

this direction are of practical importance.

2. The strength of electron-LO phonon coupling for NTSL was discussed in this the-

sis. The phonon modes in NTSLs may be further studied using optical techniques such

as Raman spectroscopy. The information presented in this thesis, combined with that

obtained from further experiments including Raman, can be quantitatively related to

phonon back-scattering in NTSLs. This should prove very useful in the design of ther-

moelectric applications based on NTSLs.

3. Further to the above, the experimental methodology employed in this thesis can be

extended to include other techniques of optical characterization such as Raman spec-

troscopy with minor modifications. Note that use of a pulsed laser is not suggested for

Raman spectroscopy because ultra-short pulses result in a broadening of the exciting

laser beam. Other techniques, such as absorption spectroscopy and two-photon absorp-

tion spectroscopy, can be implemented in the experimental arrangement described in this


thesis with some modifications.

4. It remains an interesting challenge to find new practical strategies, i.e., beyond the

existing ones mentioned above to widen the range of Eg which NTSLs can assume. Based

on the present work, this range is limited to the difference in Eg between the ZB and WZ

crystal structure of the concerned material. A wider range of Eg’s available for tuning is

an obvious practical advantage.

5. Needless to say, realization of practical devices based on NTSLs, even if on a lab-

oratory scale, is one of the most important advances to look forward to. These may

include applications such as photodetectors and single NW lasers, which have already

been demonstrated for NWs based on ZnSe and other II-VI compound semiconductors.

Appendix A

Bound-exciton lines in ZnSe films

The determined binding energies of the different donor-bound exciton complexes, EBX(D0X),

and the ionization energies of the donors, ED, in ZnSe films are summarized in Table A.1.

These quantities for the acceptor-bound exciton complexes, EBX(A0X) and EA are sum-

marized in Table A.2. These tables are as summarized by Gutowski et al. [7].

184

Appendix A. Bound-exciton lines in ZnSe films 185

Table A.1: Donors and donor-bound excitons in ZnSe films. Line positions at 4.2 K.EBX(D0X) exciton-donor binding energy, ED donor ionization energy, dfilm film thick-ness. Reprinted with permission from Ref. [7] © 1990 John Wiley & Sons.

line EBX(D0X) donor ED dfilm method substrate ref.a

(eV) (meV) (meV) (µm)- Al 26.3 1 to 2 MOCVD ZnSe [130]- Al 30 3 to 4 MOCVD GaAs [136]2.7974 I

′2 Ga 27 to 27.9 1 to 2 MOCVD ZnSe [130]

2.7969 I20 3.8 Ga 0.2 MBE GaAs [118]2.7971 IGa20 4.3(TG = 365◦ C) Ga 1.5 MBE GaAs [109]2.7944 IGa−20 TG = 400◦C Ga 1.5 MBE GaAs [109]2.7966 IGa+20 TG = 400◦C Ga 1.5 MBE GaAs [109]2.7967 IGa20 5.0 Ga 20 MOCVD GaAs [119]2.7950 Ix 5.0 Ga(∗) 1 to 1.5 MBE GaAs [40]2.7969 IGa20 3.9 Ga 0.5 MBE GaAs [122]2.7953 I2 4.7 Ga? 2 to 4 MBE GaAs [44]2.7966 I

′2 3.4 strain-split

Ga 20 3 MBE GaAs [58]2.7949 I2 8.4 Ga 4.9 MBE GaAs [21]- In 28.1 to 28.4 1 to 2 MOCVD ZnSe [130]2.7950 Ix 5.0 In(∗) 1 to 1.5 MBE GaAs [40]- F 29.3 1 to 2 MOCVD [130]- Cl 26.1 to 26.9 1 to 2 MOCVD [130]- Cl 25 2 MBE GaAs [135]2.7954 Ix 4.8 Cl 8 to 20 MBE GaAs [57]2.7965 I20 3.0 Cl? ≥ 2 MBE GaAs [127]2.7945 Ix 5.0 doublet- I 26 3.5 MOCVD GaAs [138]2.7954 Ix 5.5 VSe 3.1 MBE GaAs [121]2.7954 Ix 5.5 VSe? 3 MBE GaAs [94]2.796 Ix 3.5 ? 32 3 MBE GaAs [58]2.7968 I12 3.2 ? 1.4 to 6 MBE GaAs [28]2.7948 I22 5.2 strain-split2.796 I2 ≈ 3 ? 1 MOCVD GaAs [92]2.794 Ix ≈ 5 doublet2.797 I20 3.2 ? 8 to 20 MBE GaAs [57]2.800 Ix 5.4 ? 0.20 MBE GaAs [17](∗) Author claims Ix to be related to Ga or In.a References are those in Reference [7].

Appendix A. Bound-exciton lines in ZnSe films 186

Table A.2: Acceptors and acceptor-exciton complexes in ZnSe films. Line positions at4.2 K. EBX(A0X) exciton-acceptor binding energy, EA acceptor ionization energy, dfilmfilm thickness. Reprinted with permission from Ref. [7] © 1990 John Wiley & Sons.

line EBX(A0X) acceptor EA dfilm method substrate ref.a

(eV) (meV) (meV) (µm)2.7891 - Li 115 2 to 4 MBE GaAs [163]2.7908 ILi1 doublet2.7921 ILi1 10.6 Li 114 - LPE ZnSe [157]2.7918 ILi1 10.9 doublet≈2.792 I∗1 (Li) Li 114 ≥ 10 LPE ZnSe [148]- Li 114 ≈1 MOCVD GaAs [158]2.789 IS1 10.0 Li 118 1 MOCVD GaAs [92]2.7913 IL1 12.0 Li 4.9 MBE GaAs [21]2.792 ILi1 triplet Li - - LPE ? [129]≈ 2.79 - doublet Li - - MBE GaAs [212]2.793 - Na 111.5 3 to 13 LPE ZnSe [147]≈2.793 Iy1 (Na) Na 125 ≥ 10 MBE ZnSe [148](77 K) I1 Na 128 to 130 4 MOCVD GaAs [128]- Na 90 > 3 MBE GaAs [94]

2.783 Ideep1 Cu - LPE [129]2.7828 Id1 18.9 Cu 1 to 1.5 VPE GaAs [168]

2.78293 Ideep1 Cu - LPE ZnSe [157]2.780 ID1 Cu or VZn - MBE GaAs [212]2.792 IN1 N 80 - LPE ? [129]- N 90 ≥ 10 LPE ZnSe [148]2.790 I1 11.0 N 110 1 to 1.5 MBE GaAs [149]2.7917 IN1 N 109 0.5 MBE GaAs [156]- N 112 - MOCVD GaAs [80]- I1? N 110 to 114 4 MOCVD GaAs [128]2.790 I11 11.0 N 110 1.5 to 2 MBE GaAs [154,

150]2.792 Ih1 9.0 doublet N 0.05 to 5 MBE GaAs [150]2.791 I1 11.0 N 110 4 to 6 MOCVD GaAs [155]2.7916 IN1 9.9 N 100 3 to 6 MOCVD GaAs [152]2.792 IN1 10.7 N 111 - LPE ZnSe [157]2.790 IS1 N 102 to 110 2 MBE GaAs [219]2.796 IP1 5.2 P 80 - MBE GaAs [11]- P 90 - LPE ? [129]- P 84 ≥ 10 LPE ZnSe [148]2.789 I∗1 10.5 P 80 to 92 > 3 MBE GaAs [164]2.791 IS1 8.5 doublet- As 110 - LPE ? [129]- As 60 ? MOCVD ? [210]2.7914 I1 As 4 MOCVD GaAs [162]2.7920 Ic 10.7 C? - MOCVD GaAs [174]2.7973 I1 O 80 2 MBE GaAs [173]2.7805 ID1 19.0 VZn ≥ 2 MBE GaAs [127]2.7840 I1 16.4 VZn 2 MBE GaAs [173]2.782 ID1 18.2 VZn 8 to 20 MOCVD GaAs [57]2.7815 Id 19.4 VZn - A 3.1 MBE GaAs [121]

3 MBE GaAs [94]2.7981 Ia 7.3 ? 0.17 to 0.25 MBE GaAs [17]2.7887 Ib 16.7 ?a References are those in Reference [7].

Appendix B

Survey of luminescence studies on

ZnSe nanostructures

Table B.1: Summary of room-temperature and low-temperature luminescence (PL and

CL) studies on ZnSe nanostructures. Also included are reported crystal structures and

morphologies of nanostructures.

RTPL/RTCL LTPL/LTCL

ZB WZ ZB WZ

Nanowiresa

PL [55,58–60,64–

67,71–73,75,77,78,

81,84–86,91,96–99,

104],

CL [57,88,93,102]

PL [56,61–63,67,

69,73,79,83,92,

104,119,209],

CL [90,96,98]

PL

[55,104–107,110–

113,115–118,120],

CL [108,116]

PL

[104,106,107,110,

111,115–117,119],

CL [116]

Nanoribbonsb

PL

[73,80,82,95,103]

PL [84,94,209],

CL [90]

PL [103,106] PL [106,109]

Other

morpholo-

gies

PL

[73,87,89,91,100]

PL [68,87,209],CL

[70,76]

- -

a includes morphologies which have been referred to as nanorods and nanoneedles

b includes flat belt-like morphologies, also referred to as nanobelts

187

Appendix B. Literature survey 188

Table B.2: Summary of energy positions of the near band-edge (NBE) peak and deep-level

(DL) emission band reported for ZnSe nanostructures using their RTPL characterization.

NBE peak Reference(s) DL emission band Reference(s)

energy (ENBE (eV)) energy (EDL (eV))

ENBE < 2.5 [86] EDL < 1.9 [57,64]

2.5 ≤ ENBE < 2.6 -

2.6 ≤ ENBE < 2.7 [57,59,63–66,68,72,

77,81,82,84–86,91,

93,96,100,102,103]

1.9 ≤ EDL < 2.2 [56,61,63,65,66,

68,70,72,73,76,

78,92–97,102,104,

119]2.7 ≤ ENBE < 2.8 [56,58,60,70,73,73,

76,83,86,88,104,

119,209]

2.8 ≤ ENBE < 3.0 [63,69,73,75,79,83,

86,87,90,95,96,98,

99,101,209]

2.2 ≤ EDL [55, 59,64,69,70,

72,73,79,80,82,

83,86,88,96,99–

102,209]3.0 ≤ ENBE [62, 79,92,98]


Table B.3: Summary of LTPL studies on ZnSe nanowires. The energy position(s) of the

PL line(s), their assignment(s)a, ionization energies of the donors (ED) and acceptors

(EA), temperature of the experiments, and any intentional doping carried out are listed.

PL Line (eV) Assignment Donor/Acceptor

(ED/EA)/Comments

Reference (Temp.,

Doping?)

2.46 medium-deep

acceptor

Au (acceptor) [55], (10 K, No)

2.849/2.845 bound exciton

(WZ)

-

[103], (10 K, No)

2.816 (2.849)-1LO -

I2, I1 bound excitons ∆E ≈ 10 meV

BA band-to-acceptor EA=70 meV

2.721 DAP GaZn (D), AsSe (A)

2.602 Y-Line structural defects

2.3 deep-level

emission

self-activated defects

2.836 bound exciton

(WZ)

-

[104], (10 K, No)2.778 Id1 , deep donor ED=42 meV

2.751 BA,

band-to-acceptor

EA=70 meV

2.722 DAP ED=28 meV

2.384 - Al? (alumina

substrate)

1.998 - self-activated defects

2.795 I2 VZn [105], (20 K, No)

- continued on next page


Table B.3 – continued from previous page


(ED/EA)/Comments

Reference (Temp.,

Doping?)

2.842 Ih2 , bound

excitons

ionized acceptors

[106], (10 K, Ag)

2.810 Ih2 -LO -

2.796 IAg2 , bound

exciton

Agi interstitial donor

2.779 IAg1 , bound

exciton

AgZn, EA(AgZn)=0.23

eV

2.747 I∗Ag1 , bound

exciton

AgiAgZn acceptor

complex

2.712 I∗∗Ag1 Ag-impurity associates

2.60 Y-line -

2.682, 2.652 I∗∗Ag1 -LO,

I∗∗Ag1 -2LO

-

2.568, 2.537 Y-LO, Y-2LO -

2.273 Ag-DAP overlap of DAP

luminescence from ZB

and WZ ZnSe

2.841 I1 NBE emission,

WZ-ZnSe

[107], (10 K, No)

2.791 I2 underlying ZB-ZnSe

2.746 I3 shallow centres

(WZ-ZnSe)

I4 DAP (ZB-ZnSe) -

2.67 I5 I4-LO

2.65 I6 I4-2LO





(ED/EA)/Comments

Reference (Temp.,

Doping?)

2.61 I7 I4-3LO

2.783 EHP electron-hole plasma,

∆E=40 meV[108], (4.5 K, No)

2.683 DAP GaZn (D), AsSe (A)

2.2 DL1 deep-level emission

1.95 DL2 deep-level emission

2.806 I2 bound exciton

[109], (10 K, No)

2.792 I1 free exciton

2.762 - band-to-acceptor

2.705 - DAP

2.675, 2.645,

2.615

DAP-LO,

DAP-2LO,

DAP-3LO

-

2.3-2.4 green band vacancies of Zn in

ZnSe

2.48-2.07 - excitons located at

stacking faults[110], (4-5 K, No)

- CdSe quantum dots in

ZnSe nanowires

2.638 - DAP and their

phonon replicas from

residual impurities

[111], (4-5 K, No)





(ED/EA)/Comments

Reference (Temp.,

Doping?)

2.48-2.07 - excitons localized at

the defect zones in

NW

2.794 A neutral donor bound

exciton, VSe

[112], (3.1 K, No)2.714 B DAP, ED=27±1 meV,

EA=102 meV

2.686 C DAP, ED=27±1 meV,

EA=139 meV

2.603 D Y0 line

2.580 E Y0-LO

2.68 - DAP, Na doping

[113], (10 K, No)2.27 - deep-level, Cu doping

1.95 - deep-level,

self-activated peak

2.818 band-edge bulk ZnSe

[115], 5 K, No2.48-2.76 - -

2.07-2.48 - excitons localized at

the defect zones in

nanoneedles


ZnSe nanowires

2.48-2.76 - bunch of spectral lines

[116], (5 K, No)2.07-2.48 - excitons localized at

the defect zones





(ED/EA)/Comments

Reference (Temp.,

Doping?)


ZnSe nanowires

2.655 - exciton-related

luminescence[117], (20 K, No)

2.60 - red-shifted due to

poor crystal quality

2.655 - shallow donor

deep-acceptor pairs,

∆E=165 meV

[118], (20 K, No)

2.8 Ex free exciton

[119], (10 K, P)2.787 IS1 acceptor-bound

exciton, P acceptor,

EA0X=13 meV,

EA=130 meV

- I2 neutral donor-bound

exciton

2.791 P1 shallow donor (Ga)

[120], (10 K, Ga, Ag)2.778 P2 AgZn acceptor

complex

2.700 - DAP

a notation of PL lines, being author-specific, are not included in the List of Symbols

Appendix C

Photoluminescence (PL)

Spectroscopy

C.1 Principles

Photoluminescence (PL) is a spectroscopic technique used for the characterization of

both intrinsic and extrinsic properties of a semiconductor. PL is the luminescence from

a semiconductor when it is excited by light (photo-excitation). When a semiconductor

absorbs a photon of light with energy greater than the Eg (or in the case of resonant

excitation, equal to the Eg) of the semiconductor, an electron is excited to the conduc-

tion band leaving behind a hole (empty electron site) in the valence band. It is to be

noted that Eg refers to the energy band gap at the measurement temperature. The

electron in the conduction band and hole in the valence band are collectively referred

to as an electron-hole pair. The electron and hole, although may be created above the

band minima, relax into a quasi-equilibrium distribution to the bottom of the conduction

band and to the top of the valence band, respectively, via fast phonon-assisted processes

followed by their radiative recombination emitting light of photon energy equal to Eg.

194

Appendix C. PL spectroscopy 195

This process involving free electrons and holes is also referred to as the ‘band-to-band

radiative recombination’ or ’free-carrier recombination’ in a direct Eg semiconductor. In

an indirect Eg semiconductor, where the conduction and valence band edges are sepa-

rated from each other in the momentum (k) space, a phonon must participate in the

recombination process to conserve momentum, and this recombination is called ‘indirect

transition’. An electron-hole pair can also recombine non-radiatively in a so-called non-

radiative recombination, where phonons are emitted instead of photons. Note that both

photons and phonons are emitted in some transitions. PL spectroscopy essentially in-

volves a study of the energy dependence (spectral distribution) of the number of photons

emitted (related to the intensity of PL) during radiative transitions. The features of a PL

spectrum, such as its intensity and spectral distribution, are closely related to the elec-

tronic states of the luminescent centres involved, and can be used to identify them. The

radiative transitions observed in a PL spectrum can be distinguished as intrinsic (band-

to-band), extrinsic (impurity-related) and excitonic. The band-to-band recombination

is the simplest of a variety of pathways (or channels) for the radiative recombination of

electron-hole pairs. A common feature to these different pathways is that the electrons in

the material move into the allowed excited states after photo-excitation. These discrete

energy states between the band edges are often introduced by the impurities and/or na-

tive defects present in semiconductors. The electrons return to their equilibrium states

via radiative or non-radiative transitions. In case of radiative transitions, the energy

of the emitted photons is related to the difference in the energy levels between the two

electron states involved in the transition, i.e. between the excited and the equilibrium

states. Figure C.1 (a) shows a simplified energy band diagram showing most commonly

observed energy levels in a semiconductor, relative to the conduction and valence band

edges. The other recombination channels are discussed below, assuming an above Eg

excitation. These are also shown in Figure C.1 (b).

Free-excitons


- - -

+ + +

A0 A0 A0

+ +

Conduction

Band

Free Exciton

Donors

Valance Band

Acceptors

A0X

D0X

(a) (b)

A-

D0 D0 D0 D+

A B C D E F G H

- - -

Complexes

D0h eA0 D0A0 A0X D0X FE

Figure C.1: (a) Relative positions of donor and acceptor levels in the simplified band dia-

gram; (b) Radiative and non-radiative transitions: (A) Free exciton recombination (FX),

(B) and (C) radiative recombination of donor- and acceptor-bound excitons (D0X,A0X),

(D) donor-acceptor pair (DAP) recombination (D0A0), (E) radiative recombination of a

free electron and a neutral acceptor (eA0), (F) radiative recombination of a free hole and

a neutral donor (D0h), and (G) and (H) are non-radiative transitions of free electrons

and holes to ionized donors and acceptors, respectively. [(b) adapted from Ref [251]]


At low temperatures in a reasonably pure semiconductor, the photo-excited electron

and hole are bound by Coulomb interaction, and the electron-hole pair is then called

an exciton. Excitons are usually classified into: Frenkel excitons and Wannier excitons,

distinguished by the length scale of interaction between electron and hole. The Frenkel

exciton is tightly bound and highly localized, and the distance between an electron and

a hole is about few angstroms. These excitons have a large binding energy and are

typically found in insulators and organic materials. The Wannier exciton on the other

hand, has a small binding energy with the electron-hole separation of the order of 10-

100 A. The Wannier excitons are often found in inorganic materials, and move freely in

the material. Since the excitons in ZnSe are Wannier excitons, the following discussion

related to excitons pertains to Wannier excitons.

If the exciton is not localized about an impurity or defect site in the lattice, it is free

to move in the crystal lattice and is called a free exciton. Free excitons form a series

of separated states near the conduction band-edge, and the emission energy from these

states is given by:

hνFX = Eg − EFX (C.1)

where EFX is the binding energy or the energy states of the free exciton associated with

the Coulomb interaction in the presence of crystal potential and hνFX is the emission

energy corresponding to the free exciton emission. The binding energy of the free exciton

is taken relative to the conduction band. In an exciton, the electron orbits the hole

similar to the electron in a hydrogen atom, thus a hydrogenic model is used to calculate

their binding energy. According to the hydrogenic model for an interacting positive and

negative charge, the binding energy of the free excitons is described as:

EFX =1

n2FX

m∗re4

2~2ε2(C.2)

where ε is the dielectric constant, ~ the reduced Planck’s constant (=h/2π, where h is

the Planck’s constant) and e is the electronic charge. The integer nFX is the exciton


principal quantum number. The ground state of the free exciton corresponds to n=1 in

the Equation C.2. The reduced effective mass of the exciton, m∗r, is given by

1

m∗r=

1

m∗e+

1

m∗h(C.3)

where m∗e is the electron effective mass and m∗h is the hole effective mass. The excitonic

emission spectrum is expected to be like a δ-function, or if the exciton life-time broadening

is included, like a Lorentzian. This is unlike the other recombination transitions which

result in broader emission spectra.

Bound excitons

In a typical semiconductor material, point defects (intrinsic and extrinsic) exist which act

as donors or acceptors (neutral or charged). The shallow donor impurity levels lie in the

energy gap, slightly below the conduction band, see Figure C.1 (a). When these donors

are ionized, their electrons are excited into the conduction band. Similarly, the shallow

acceptor levels lie in the energy gap, just above the edge of the valence band. An electron

is excited from the valence band to the acceptor level when an acceptor is ionized. The

free excitons can be trapped at the sites of these defects via dipole-dipole interactions

(in the case of neutral defects) or dipole-monopole interactions (charged defects). This

trapping of excitons at defect sites is referred to as the binding of excitons resulting in

bound exciton complex(es) (BEC). The excitons in such cases are referred to as bound

excitons, and these may be of the following types:

1. Exciton bound to a neutral donor (D0X): � (⊕)

2. Exciton bound to an ionized donor (D+X): �(⊕)

3. Exciton bound to a neutral acceptor (A0X): ⊕� (⊕)

4. Exciton bound to an ionized acceptor (A−X): �(⊕)

where � and � represent the ionized donor and acceptor, respectively, and and ⊕

are the electron and hole, respectively. An excitons is shown as (⊕). The (D0X)


bound exciton is considered to consist of a donor ion, two electrons and a hole whereas

(D+X) complex consists of a donor ion, an electron and a hole. A similar argument holds

for acceptor-bound excitons (A0X) and (A−X). Note that an ionized acceptor usually

does not bind an exciton since a neutral acceptor and a free electron are energetically

more favourable. This is because the hole mass is usually considerably larger than the

reduced mass of an electron and a hole. The bound excitons give PL emission at slightly

lower energies than the free exciton transitions, with narrower linewidths due to their

increased spatial localization. This is because the bound excitons are spatially localized

at the defect sites, and are not free to move in the crystal. This results in their much

smaller kinetic energy compared to free-excitons. The emission from bound excitons has

a much narrower linewidth (� kBT ) than that of the free exciton (∼ kBT ). Here, kB

is the Boltzmann constant and T is the temperature. The emission energy, or the peak

position in PL spectrum for a bound exciton emission, hνBE, is:

hνBE = Eg − EFX − EBX (C.4)

where EBX is the binding energy of the exciton to the impurity or defect. The peak energy

(hνBE) is specific to a given impurity or defects and its charge state in a semiconductor,

and this is the basis for the identification of impurities and defects in a semiconductor

through PL. The ionization energy of shallow donors, ED, and acceptors, EA, according

to the hydrogenic model are given by:

ED = (m∗ee4/2~2ε2) (C.5)

EA = (m∗he4/2~2ε2) (C.6)

Donor-acceptor pair

Both donor and acceptor impurities typically exist in a semiconductor material. When

such a material is irradiated with light, the photo-generated electron-hole pairs neutralize

the shallow ionized donors and acceptors in the material. This may be represented as


e− + D+ → D0 and h+ + A− → A0. An electron bound to a donor can recombine with

a hole bound to an acceptor, giving rise to a donor-acceptor pair (DAP) emission.

D0 + A0 → D+ + A− + hνDAP (C.7)

In low-temperature PL from DAP recombinations, the recombination energy is given

by [312]:

EDAP (r) = Eg − (EA + ED) +e2

εr−(e2

ε

)(avdWr

)6

(C.8)

where avdW is the effective van der Waals coefficient for the interaction between a neutral

donor and a neutral acceptor. The term (e2/ε)(avdW/r)6 in Equation C.8 takes into

account the closer separations between the pairs, but is somewhat empirical [313], and

negligible for distant pairs. The term (e2/εr) is the energy due to the Coulomb attraction

between the ionized donor and acceptor. For sufficiently distant pairs, Equation C.8 for

the energy of the emitted photon EDAP (r) reduces to [254]:

EDAP (r) = Eg − (EA + ED) +e2

εr(C.9)

The emission energy of a DAP recombination is therefore a function of the separation

r between the donor and acceptor in a donor-acceptor pair (also referred to simply as

pair). Pair recombination often has a series of distinct but closely packed sharp lines

corresponding to different allowed discrete values of r. These discrete values of r arise

because donors and acceptors must be at the lattice sites, and therefore only r values

permitted by the crystal symmetry are allowed. In some cases, this series of lines results

in a broad band [237].

Free-to-bound

Apart from the band-to-band, excitonic and DAP transitions, there is another type of

transition referred to as free-to-bound (FB) transitions, so called because they involve

a free carrier in the conduction or the valence band, and a bound hole or electron. A


free-to-bound or donor-to-hole (D0h) is a recombination of an electron bound to a neutral

donor with a hole in the valence band. The corresponding transition energy for (D0h)

transition is given by:

ED0h = Eg − ED (C.10)

Similarly, an electron from the conduction band recombining with a hole bound to a neu-

ral acceptor is an electron-to-acceptor (eA0) (or bound-to-free) transition, whose transi-

tion energy is:

EeA0 = Eg − EA + kBT (C.11)

where and the last term (kBT ) accounts for the electron kinetic energy. As the temper-

ature increases, the DAP band evolves to a free-to-bound transition. This is due to the

thermal ionization of the shallow impurity and therefore an increase in the free carrier

concentration. Figure C.1 (b) shows the transitions described above schematically (A-F).

Figure C.1 (b) also includes the non-radiative transitions from free electrons and holes

to ionized donors (G) and acceptors (H), respectively.

Phonons are the quantized modes of lattice vibrations due to the vibrations of atoms

forming the crystal. In crystals with more than one kind of atoms, these are of two types:

acoustic and optical. The acoustic modes result when nearby atoms in a crystal vibrate

in the same direction with a periodically varying amplitude. These are similar to sound

waves, and hence the name acoustic. The optical phonons are modes when nearby atoms

vibrate in opposite directions. Each of these are of further two types. The phonons with

atomic displacements parallel to the direction of propagation are longitudinal phonons,

while those with displacements perpendicular to the direction of propagation are trans-

verse phonons. When a PL transition is accompanied by a simultaneous emission of

phonons, subsidiary peaks appear in the PL spectrum on the lower energy side of the

this transition, also referred to as the main transition. These peaks on the lower en-

ergy side are referred to as phonon-replicas, whose energy is lower than that of the

main transition by an energy equal to the phonon energy. For example, in the case of


longitudinal-optical (LO) phonon-replicas of the free exciton emission, the energy of the

main transition and its phonon-replicas is given by:

hνFX,m = Eg − EFX −m~ωLO (C.12)

where integer m is the order of the LO-phonon-replica, and ~ωLO is the LO-phonon

energy. hνFX,m is the energy position of the PL peak corresponding to the m-th or-

der LO-phonon replica. hνFX is the transition energy of the zero-phonon line (ZPL)

corresponding to m=0. Similarly, in case of bound exciton emission accompanied by

LO-phonon emission, Equation C.4 becomes:

hνBE,m = Eg − EFX − EBX −m~ωLO (C.13)

C.2 Theoretical models

One of the disadvantages of using PL spectroscopy is that it is a semi-quantitative char-

acterization technique. In order to extract quantitative information from the PL spectra,

the dependence of PL with different varying quantities (typically excitation intensity

(Iexc) and temperature (T )) is used. The simple theoretical models commonly used to

obtain information on recombination mechanisms through the use of these varying quan-

tities are described below.

C.2.1 Excitation intensity dependence of PL

The dependence of the PL spectrum from a semiconductor material on Iexc can be used to

identify the recombination processes involved, and to calculate the ionization energies of

the donors and acceptors involved in DAP-related transitions. The change in integrated

PL intensity (IPL) with Iexc yields information about the type of the optical transition

(e.g. excitonic versus DAP-related), whereas the variation in peak energy with Iexc for


a DAP type transition provides information about the ionization energies of the donor

and acceptors involved, introduced by the impurities (extrinsic and intrinsic) and native

point defects. These models are discussed next.

Dependence of IPL on Iexc

As the Iexc is changed, the IPL of the PL peak reflects this change which is usually

characteristic of the type of the transition mechanism involved for a particular emission.

In order to elucidate the dependence of IPL on Iexc, Schmidt et al. [251] solved a set of

rate equations for different recombinations including (A) free-exciton (FX), (B) donor-

bound excitons (D0X), (C) acceptor-bound excitons (A0X), (D) DAP, (E) free electron

to a neutral acceptor, (F) free hole to a neutral donor and (G) non-radiative transitions

of free electrons and hole to neutral donors and acceptors respectively. These transitions

(A)-(H) are shown schematically in the Figure C.1 (b). These rate equations are listed

below, and their solution is described in the work of Schmidt et al. [251].

dn

dt= i′L− a′n2 − g′n(ND −ND0)− e′nNA0 (C.14)

dnFEdt

= a′n2 + j′L−[

1

τFE+

1

τnrFE

]nFE − b′nFEND0 − c′nFENA0 , (C.15)

dnDXdt

= b′nFEND0 −[

1

τDX+

1

τnrDX

]nDX , (C.16)

dnAXdt

= c′nFENA0 −[

1

τAX+

1

τnrAX

]nAX , (C.17)

dNA0

dt= h′(NA −NA0)n+ i′(NA −NA0)L− c′nFENA0 +

[1

τAX+

1

τnrAX

]nAX

−d′ND0NA0 , (C.18)

dND0

dt= g′(ND −ND0)n− k′ND0L− b′nFEND0 +

[1

τDX+

1

τnrDX

]nDX

−d′ND0NA0 − f ′ND0n (C.19)

The symbols are as follows: ND (NA) are the concentrations of donors (acceptors),

ND0 (NA0) are the concentrations of neutral donors (acceptors), the radiative and non-


radiative lifetimes of free excitons (τFX and τnrFX), donor-bound excitons (τDX and τnrDX),

and acceptor-bound excitons (τAX and τnrAX). nFE and nDX (nAX) are the concentra-

tions of free and donor-bound (acceptor-bound) excitons, respectively. The coefficients

a′, b′, ..., h′ are the transitions rates of the processes (A-H) indicated in Figure C.1 (b).

The coefficients i′, ...l′ are the transition rates of the following processes: (I) photo-

excitation of the electron-hole pairs by light with photon energy greater than Eg, (J)

formation of the free excitons resonantly by light with photon energy ≈Eg and excitation

of electrons from (K) neutral donors and (L) ionized acceptors by the excitation light,

respectively.

These rate equations were analytic solved using assumptions on excitation, and radia-

tive and non-radiative recombination channels [251]. Schmidt et al. [251] mathematically

expressed the dependence of IPL on Iexc by a power law relation:

IPL ∝ Ipexc (C.20)

The variation of IPL with Iexc can therefore be used to identify the underlying recombi-

nation processes. While this relation can be used to distinguish excitonic emission from

other types, it fails to separate free-to-bound and DAP transitions.

Dependence of peak energy on Iexc

As discussed above, in low-temperature PL from donor-acceptor pair recombinations,

and for sufficiently distant pairs, the energy of the emitted photon EDAP (r) is given by

Equation C.9 [254]. A model was proposed by Zacks and Halperin [237] that described

the dependence of the emission peak energy on Iexc. In their derivation, it was assumed

that the temperature of the semiconductor material is low enough so that the thermal

ionization of donors and acceptors can be neglected. It was also assumed that all photo-

excited electron-hole pairs are captured by ionized impurities, which then become neutral.


The model proposed by them is expressed as [237]:

Iexc = Iexc,0(hνm − hν∞)3

hνB + hν∞ − 2hνmexp

(− 2(hνB − hν∞)

hνm − hν∞

)(C.21)

where hνm(= Em) is the emission band peak energy, hν∞(= Eg − (EA + ED)) is the

photon energy corresponding to infinitely distant donor-acceptor pair, hνB(= Eg−(EA+

ED) + e2/εRB) is the emitted photon energy of a donor-acceptor pair separated by a

shallow impurity Bohr radius (RB), and Iexc,0 is a constant of proportionality. A non-

linear least square fit to the experimental data is used to determine the values of hν∞,

hνB and Iexc,0. The change in peak energy with increasing Iexc can be understood by

considering that as the Iexc is increased, closely-lying donor-acceptor pairs are favoured,

and the transition energy moves to higher energy according to Equation C.9.

C.2.2 Temperature dependence of PL

The variation in PL characteristics with a change in temperature provides important

information about the radiative and non-radiative recombination mechanisms involved.

It has been experimentally observed that as the temperature is increased: (i) the spectral

peak position shifts to lower energies, (ii) the PL peak broadens, and (iii) the integrated

intensity of the PL peak reduces. These have been explained using models as discussed

below.

Dependence of peak energy on T

The red-shift in PL peaks with increasing temperatures is mainly linked to the shrinkage

of the Eg at higher temperatures. The temperature dependence of Eg is most commonly

described by a semi-empirical equation proposed by Varshni [262]:

Eg(T ) = Eg(0)− αT 2

β + T(C.22)

where Eg is the direct or indirect energy band gap, Eg(0) is the value of Eg at 0 K and α

and β are the fitting parameters characteristic of a given material. α can be considered


to be the temperature gradient (dEg/dT ) at high temperatures, and β to be proportional

to the Debye temperature θD for the material. This equation was proposed based on the

two experimental facts that the change in the relative positions of the conduction and

valence bands (which determine the Eg):

1. due to thermal expansion of the crystal lattice is linear at high temperatures and

non-linear in the low-temperature regime. This change in the Eg due to the ex-

pansion of the lattice however only accounts for a quarter of the total change in Eg

with the temperature.

2. due to the electron-lattice interactions is quadratic for T � θD and linear for T �

θD. A major contribution to the total change in the Eg is due to this interaction.

This equation was fit to the data available for Eg at different temperatures for several

materials [262], and the fit was found to be satisfactory. Despite being a widely used

equation to describe the change in Eg with temperature, its weak theoretical basis is

also acknowledged [314]. Another expression with a more microscopic basis, but still

semi-empirical, to model this dependence has also been in use and expressed as [315]:

Eg(T ) = Eg(0)− 2αB[exp

(θBT

)− 1

] (C.23)

where αB is the strength of the exciton-average phonon interaction, and θB is a temper-

ature corresponding to an average phonon energy. This model takes into account the

Bose-Einstein statistical factor NLO (=1/[exp(θLO/T )− 1]) for the phonon emission and

absorption, where θLO is the temperature corresponding to the LO-phonon energy. This

model is based on the consideration that both acoustic and optical phonons contribute

to the red-shift of the Eg via electron-lattice interactions. Note that the Equations C.22

and C.23 describe the dependence of only Eg on temperature, and not necessarily the

dependence of PL peak positions in a PL spectrum corresponding to different processes

on temperature. Since the FX emission follows the Eg for a wide temperature range,


the parameters obtained from the fits of FX emission peak position with temperature to

these equations are representative of the dependence of Eg on temperature. For the PL

peaks related to bound excitons these parameters can be, and usually are, different from

those for the FX emission and Eg.

Dependence of HWHM on T

The dependence of linewidths (Γp) of exciton lines on temperature provides important

information about the exciton-phonon interactions and exciton-impurity scattering. The

Γp of an excitonic line in PL from a semiconductor includes contributions from homo-

geneous and inhomogeneous broadening. The temperature dependence of the excitonic

linewidths (in terms of the HWHM, Γp/2) in semiconductors is represented as a sum of

different contributions to it,

Γp(T )/2 = Γinh + γthT +ΓLO[

exp

(θLO

T

)− 1

] + Γimp exp(−Eimp/kBT ) (C.24)

where each term on the right side of the expression is discussed next. Γinh is a tem-

perature independent inhomogeneous broadening contribution term which includes the

effects of intrinsic lifetimes, carrier-carrier scattering, surface scattering as well as those

of instrumental broadening. The most common mechanism that contributes to the ho-

mogeneous broadening of the Γp is due to the process of phonon scattering. This implies

the collisions of excitons with phonons which result in their dissociation, and both acous-

tic and optical phonons are involved in these scattering processes. The contributions of

acoustic and optical phonons to the homogeneous broadening of Γp are proportional to

their respective population density. In the low temperature range, the density of acoustic

phonons is relatively high and acoustic-phonon scattering dominates. This is represented

by the second term in the Equation C.24, and γth represents the strength of the exciton-

acoustic phonon coupling. The population of optical phonons follows the Bose-Einstein

statistical factor (NLO) and becomes significant at higher temperatures. The third term


represents the homogeneous broadening due to exciton-optical phonon scattering, and

ΓLO is the strength of exciton-optical phonon coupling. The last term in Equation C.24

represents the inhomogeneous broadening due to scattering by ionized impurities, and

proportional to the density of ionized impurity scattering centres. Γimp is contribution to

Γp due to scattering by fully ionized impurities, and Eimp is the average binding energy

of the impurities. The experimentally obtained values of the Γp at different temperatuers

is fit to the Equation C.24, and relevant parameters are obtained. It is to be noted that

the last term in Equation C.24 (exciton-ionized impurity scattering process) is usually

neglected because of its negligible contribution. The resulting equation still contains four

fitting parameters, of which θLO is usually kept fixed as the temperature corresponding

to the LO-phonon energy.

Dependence of IPL on T

The efficiency of PL emission is given by

η =Pr

Pr + Pnr(C.25)

where Pr is temperature independent probability of a radiative transition, and Pnr is the

probability for a non-radiative transition whose dependence on temperature is given by:

Pnr = Pnr,0 exp (−Ea/kBT ) (C.26)

In this equation, Pnr,0 is a temperature independent prefactor, and Ea is the thermal

activation energy which can be understood as the height of a barrier around the re-

combination centre induced by perturbations in the band structure. Only the carriers

with energies similar to or higher than Ea can overcome this barrier and recombine

non-radiatively. The probability of non-radiative transitions therefore increases with the


temperature. The PL emission efficiency becomes

η =1

1 + PnrPr

(C.27)

=1

1 +Pnr,0Pr

exp(−Ea/kBT )(C.28)

=1

1 + C exp (−Ea/kBT )(C.29)

where C = Pnr,0/Pr is a coefficient. The dependence of integrated PL intensity (IPL) on

temperature can therefore be expressed as

IPL(T ) =IPL(0)

1 + C exp (−Ea/kBT )(C.30)

The above equations were based on a single thermally activated non-radiative tran-

sition mechanism. For recombinations accompanied by two such mechanisms, the Equa-

tions C.25 to C.30 are modified to:

η =Pr

Pr + Pnr1 + Pnr2(C.31)

Pnr1 = Pnr1,0 exp (−Ea1/kBT ) (C.32)

Pnr2 = Pnr2,0 exp (−Ea2/kBT ) (C.33)

η =1

1 + Pnr1Pr

+ Pnr2Pr

(C.34)

=1

1 +Pnr1,0Pr

exp(−Ea1/kBT ) +Pnr2,0Pr

exp(−Ea2/kBT )(C.35)

=1

1 + C1 exp(−Ea1/kBT ) + C2 exp(−Ea2/kBT )(C.36)

IPL(T ) =IPL(0)

1 + C1 exp(−Ea1/kBT ) + C2 exp(−Ea2/kBT )(C.37)

where subscripts 1 and 2 refer to the two non-radiative recombination mechanisms. Ex-

tension of these equations to include more than two non-radiative transitions are straight-

forward. These thermal activation energies (Ea, Ea1, Ea2) are interpreted in context of


the behaviour of the observed peaks with the temperature in the temperature-dependent

PL spectra.

Appendix D

Time-resolved Photoluminescence

(TRPL) spectroscopy

Time-resolved spectroscopy techniques are a powerful means of studying a material giving

information about the nature of the excitation, energy transfer, molecular motion and

molecular environment. Time-resolved photoluminescence (TRPL) spectroscopy is an

extension of the time-integrated (or conventional) PL spectroscopy, and facilitates the

study of the dynamics of photo-excited carriers in a semiconductor. In TRPL, ultra-

short laser pulses excite the semiconductor generating a large number of electrons and

holes, and the resulting luminescence from the semiconductor sample is collected and

dispersed by a spectrometer, and later detected by a detector. The principle is similar

to that of the PL spectroscopy, the main difference being that the laser excitation needs

to be pulsed, and the detection device should have a very fast time response. In the

present case, the experimental setup for TRPL spectroscopy is similar to that for PL

spectroscopy, with the exception that the spectrometer and CCD detector are replaced

by another spectrometer (ACTON PI 300i), and a streak camera (Hamamatsu C5680) is

used for time-resolved detection. This is the excite-and-probe technique for time-resolved

211

Appendix D. TRPL spectroscopy 212

Input slit

Input optics

Trigger signal

Photocathode

Multi-channel

Plate (MCP)

Time

Streak image

Incident

luminescence

signal

Deflection

electrodes

Sweep circuit

Accelerating

electrode Phosphor

screen

Wavelength

Figure D.1: Components and principle of operation of a streak camera

spectroscopy, distinguished from the pump-probe methods of time-resolved spectroscopy.

The main components and the operating principle of streak camera are described next.

The PL intensity can be measured as a function of time after the excitation pulse

was incident on the semiconductor using pulsed excitation. The carrier recombinations

in a semiconductor take place at a very fast rate, typically of the order of pico-seconds

(ps), and conventional detectors having longer response times prove to be inadequate

for measuring such fast photo-responses from the semiconductors. A more sophisticated

detector, called a streak camera is used to measure such short carrier lifetimes. The main

components and the operating principle of a streak camera are shown in Figure D.1.

The photons of the emitted luminescence signal entering the system through the input

slit are converted to electrons by a photocathode. The electrons then travel between two

electrodes to which a time-varying electric field is applied synchronously with the incident

light. The sweep rate of the electric field is varied from zero to a certain maximum value

Appendix D. TRPL spectroscopy 213

with rates as high as 2000 V/ns. The electron are deflected by this electric field, and

during this deflection process, the electrons arriving first are deflected less than the ones

arriving later. The electrons then hit a phosphor screen (or CCD array in some cases)

which is placed on the other side of the electrodes. The electrons arriving first hit the

screen in the top part, while the electrons arriving later hit the bottom part of the screen.

The vertical axis of the mapped image of PL intensity on the phosphor screen, also called

the streak image (since the incoming electrons leave a streak on the phosphor screen), is

thus the time axis. The time-resolution of the streak camera is limited by the sensitivity

of the voltage applied to the electrodes, and this can be as low as 100 V/cm. When

a streak camera is used with a spectrometer, the PL is also spectrally resolved before

entering the detector, and the horizontal axis of the phosphor screen (streak image) then

serves as the wavelength axis. The streak camera is used in conjunction with the delay

generator and trigger for synchronization with the ultra-fast laser. The stated resolution

of the streak camera system is 2 ps. During the TRPL measurements, the PL intensity

is recorded as a function of decay time, and the resulting curve can be fitted to yield the

value(s) of carrier recombination lifetime(s) τ .

Appendix E

Electronic band structures

E.1 Energy band diagrams

In isolated atoms, the allowed energy states of the electrons are discrete. The periodic

potential of the atoms (ion-cores and the bound electrons) in a crystal affects the energy

states of the electrons. The electronic energy bands in solids result from the overlapping

of the orbitals of atoms that form the solid. This overlapping depends on the geometric

arrangement of the atoms (crystal lattice), the difference of energy between the atomic

orbitals in the solid, and the number of valence electrons in the atoms. The energy bands

are the ranges of allowed electron energies when atoms are brought together in forming a

crystal. The ranges with no allowed energies of electrons are the forbidden gaps or band

gaps. The highest energy band which contains electrons at temperatures above absolute

zero is referred to as the conduction band. The valence band is the next lower-lying band,

mostly filled with electrons, and separated from the conduction band by an energy gap

(∼eV) in semiconductors. The wavevector k is related to the direction of motion of an

electron in a crystal. The energy band diagrams, also referred to as E−k plots or energy

dispersion curves or electronic band structures, are the informational plots of allowed

214

Appendix E. Electronic band structures 215

values of the electron energy as a function of k. The combined effect of periodic potential

of atoms in a crystal is incorporated by adjusting the mass of an electron to a different

value, known as the effective mass (m∗). The effective mass is inversely proportional to

the curvature of an E − k curve. The optical, electrical and magnetic properties of a

solid are determined by its electronic band structure.

The k values associated with a given energy band are said to form a Brillouin zone

(BZ). The BZ’s are numbered consecutively beginning with the lowest energy band, which

forms the first BZ. The first BZ of ZB crystal lattices is shown in Figure E.1 (a). It is an

octahedron, truncated by the {100} planes at a distance 2π/a from the zone-centre, where

a is the lattice constant of the ZB crystal. Γ identifies the centre of the first BZ at k=0.

For BZs of other crystal lattices, Γ denotes the zone centre. Other symbols in Figure E.1

(a) denote the points and directions of high symmetry, also called critical points. As a

convention, the points and directions inside the BZ are labelled by Greek letters, while

those on the surface of the BZ by Roman letters. Some of these high-symmetry points,

and the three high-symmetry directions are:

• X denotes the zone-end along the 〈100〉 direction

• L denotes the zone-end along the 〈111〉 direction

• K denotes the zone-end along the 〈110〉 direction

• 〈100〉: Γ→ X is denoted as ∆

• 〈111〉: Γ→ L is denoted as Λ

• 〈110〉: Γ→ K is denoted as Σ

The first BZ for a WZ crystal lattice with hexagonal symmetry is shown in Figure E.1

(b). The critical points in the first BZ of a hexagonal lattice are:

• A denotes the zone-end along the 〈001〉 direction

• M denotes the zone-end along the 〈100〉 direction


Fig

ure

E.1

:F

irst

BZ

of(a

)Z

Ban

d(b

)W

Zcr

yst

alla

ttic

es[3

16].


• 〈001〉: Γ→ A is denoted as ∆

• 〈100〉: Γ→ M is denoted as Σ

Figure E.2 shows the electronic structures for ZnSe in the ZB and WZ crystal struc-

tures, calculated using LDA (local density approximation) method [317]. The conduction

band minima (CBM) and valence band maxima (VBM) both occur at k=0.

(a) (b)

Ene

rgy

(eV

)

Figure E.2: Calculated band structures of ZnSe in (a) ZB and (b) WZ crystal structures.

Reprinted with permission from Ref. [317] © 1994 American Physical Society.

E.2 Band structure calculations for Nanowire Twin-

ning Superlattices

Since the NWs and NTSLs in present samples are rather thick with the smallest being 50

nm in diameter, effect of quantum confinement on the electronic structure is negligible.


Therefore, these structures may be modeled as one-dimensional superlattices. The band

structures of NTSLs were calculated using ab Initio pseudopotentials within the local-

density approximation as implemented in the ABINIT package [318]. Trullier-Martins

pseudopotentials were used that include d-orbitals in the valence states for Zn [319].

Inclusion of d-orbitals increases the computational complexity but is essential for an ac-

curate description of structural properties. The lattice constant for ZB structure was

determined by minimization of the total energy to be 5.531 A which is only 1% smaller

than the experimental value. The geometry for all structures were obtained by stacking

layers with ideal lattice spacing of the ZB structure. For wurzite structure (2H) use of

fully optimized geometry reduces the band-gap by 5 meV. For NH structures with N > 2

deviation from the assumed ideal geomerty is expected to change the band-gap by an

even smaller amount. An energy cutoff of 80 Rydberg was used so that the band gap

energy converged to within a few meV. As is well-known, LDA significantly underesti-

mates the band-gap [320], so for comparison with experiments, a constant scissors shift

of 1.642 eV (chosen to align the bandgap of ZB structure with experiments) was applied

uniformly to all structures [320]. The top of the valence band which is 3-fold degener-

ate in the cubic (ZB) structure splits into two in all other structures. This crystal field

splitting is largest in the WZ (2H) structure with a value of 71 meV gradually decreasing

to zero as N increases. In all cases the band-gap is direct (at k = 0) and with inter-

band transition dipole moments only weakly varying with the periodicity. This is not

surprising as the deviation from ideal ZB structure is small and, to a good approxima-

tion, the band-structure may be viewed as that of the ZB structure folded in k-space as

demanded by the periodicity. Figure E.3 presents the electronic structures obtained from

ab Initio calculations for (a) 2H, (b) 4H (γ=0.5) and (c) 6H (γ=0.333) polytypes of ZnSe

with relevant details given in the figure captions. Figure E.4 compares the calculated

band structures for 2H and 4H polytypes where the reduction in direct band gap for 4H

structure in comparison to that of 2H structure can be noticed.


-6

-4

-2

0

2

4

A Γ K M G

Ener

gy (e

V)

-6

-4

-2

0

2

4

A Γ K M G

Ener

gy (e

V)

-6

-4

-2

0

2

4

A Γ K M G

Ener

gy (e

V)

a2H

b4H

c6H

VBM

VBM

VBM

Figure E.3: Band-structure of ZnSe

NTSLs calculated using ab Initio pseu-

dopotential method within the local

density approximation (LDA). The en-

ergy bands of the NH structure are

closely related to that of the wurtzite

structure (2H) with bands folded at

2/N(0, 0, 2π/c) along the Γ-A direction.

This can be clearly seen by compar-

ing the band-structures for the 2H (a)

and 4H (b), for example. It is noted

that there are some important devia-

tions such as splitting of some degenera-

cies that cannot be accounted for by a

simple folding. However, the bands near

the direct band gap at the Γ point are

very similar in all the structures apart

from a gradual shift in the band gap

with N . The horizontal line indicates

the position of the valence band max-

imum (VBM). Also shown is the cal-

culated band-structure for 6H-polytype

(c).


-3

-2

-1

0

1

2

A A1 Γ K

Ener

gy (e

V)

VBM

Figure E.4: Comparison of the calculated band-structures for wurtzite (2H) (red) and

4H-polytype (blue) structures. A and A1 denote the Brillouin zone boundary along the

growth direction for 2H and 4H structures, respectively. The band gap for 4H-poytype

is reduced compared to that of wurtzite (2H) structure. The horizontal line (green)

indicates the position of the valence band maximum (VBM).

Appendix F

Softwares and Programs used

The following softwares and computer programs were used for the acquisition of data,

and their analysis.

• SpectraMax for Windows Version 3.2 (Horiba Scientific) was used to control

the spectrometer (Jobin-Yvon, TRIAX320) and the CCD detector (Jobin-Yvon,

CCD3000). The same program was used to acquire the PL spectra from NWs.

The parameters of acquisition such as ti, dslit and centre wavelength were adjusted

using this program before the acquisition of each PL spectrum. A background

subtraction feature built into this program was used, which automatically subtracts

the background (dark) signal before the acquisition of a PL spectrum.

• HPD-TA (High-Performance Digital Temporal Analyzer) Version 8.4.0 pf5 07.04.2010

(Hamamtsu Photonics Deutschland) was used to acquire TRPL streak images. The

second spectrometer (ACTON PI 300i) and other accessories for the streak cam-

era e.g., the delay generator, were also controlled by this program. The centre

wavelength for the spectrometer, dynamic time range (1-4), delay time, and other

parameters for the acquisition of streak images were controlled using HPD-TA. In

this case, the equivalent of ti is the number of frames to be acquired, which can be

221

Appendix F. Softwares and Programs used 222

adjusted using this program. The slits for both the streak camera and the second

spectrometer were manual, and not adjustable on the program.

• DigitalMicrograph Version 2.02.800.0 (GATAN) was used to acquire the TEM

and HR-TEM microgrpahs on Titan-80-300 (FEI) and JEOL-2010F STEMs. It was

also used for the acquisition of SAD patterns on JEOL-2010F. The FFT patterns

were generated by selecting an area on the HR-TEM micrographs using DigitalMi-

crograph. For the indexing of SAD and FFT patterns, the distances between the

diffraction spots were measured using this program. For counting the number of

monolayers and measuring the widths of the individual segments on TEM and

HR-TEM micrographs from NTSLs, a profile from a selected area was used in

DigitalMicrograph, which displays the intensity of the pixels in the selected area.

• Origin Version 7.0 (OriginLab) was used to analyze the PL spectra. This includes

fitting of the PL spectra to peaks functions, e.g., Lorentzians and Gaussians. The

parameters, e.g., Ep, Γp and IPL obtained from the above fitting were fit to the

theoretical models using the same program. The same program was used to generate

all the PL plots and plots with data i.e., excluding the TEM, HR-TEM micrographs,

SAD patterns and electronic-structure plots, presented in this thesis.

• Others These include:

– ImageJ (Open Source) for image-processing

– MATLAB (MathWorks) for data analysis (in addition to Origin)

– the drivers and accompanying image-acquisition programs for the frame-grabber

boards (Data Translation, DT3120 and 1stVision Inc., PC2Vision) used to

capture images from CCD cameras (Hitachi KP-32N and analog CCD camera

inside the streak camera)

– LibreOffice (Open Source) for labelling the indexed SAD patterns

Appendix G

Supplementary TEM micrographs

TEM and HR-TEM micrographs, and SAD and FFT patterns corresponding to different

ZnSe NWs and NTSLs are presented below.

G.1 Chapter 5 - ZB ZnSe nanowires

Figure G.1 (a) shows an overview TEM micrograph of ZB-NW-2 with its higher magnifi-

cation TEM image shown in Figure G.1 (b). Figure G.2 (a) shows a HR-TEM micrograph

from an area of ZB-NW-2. Figure G.2 (b) shows an indexed FFT pattern acquired from

an area selected in (a) (shown in red). The indexed spots in the FFT pattern confirm

that the selected area of ZB-NW-2 is single crystalline with a ZB crystal structure. Fig-

ure G.3 (a) shows another HR-TEM image from an area of ZB-NW-2. The ZB crystal

structure of ZB-NW-2 is also demonstrated by the indexed SAD pattern shown in Fig-

ure G.3 (b) taken from an area shown in Figure G.1 (b). These confirm that ZB-NW-2

is single crystalline with a ZB crystal structure. The features in the LTPL spectrum of

ZB-NW-2 described in Chapter 5 can therefore be unambiguously related to this single

crystalline ZB structure of ZB-NW-2.

223

Appendix G. Supplementary TEM micrographs 224

a b

ZB-NW-2

Location of SAD

Figure G.1: Structural characterization of ZB-NW-2: (a) Overview TEM micrograph,

and (b) TEM micrograph at a magnification higher than in (a).

a b

111

111

220

220

002

002

111

111

Figure G.2: (a) HR-TEM micrograph from an area of ZB-NW-2, and (b) Indexed FFT

pattern acquired from an area (squared in red) in (a). The viewing direction for (a) and

(b) corresponds to 〈110〉.


a b

111

111

220

220

002

002111

111

Figure G.3: (a) HR-TEM micrograph from an area of ZB-NW-2, and (b) Indexed SAD

pattern from a larger area (shown in Figure G.1 (b)) of ZB-NW-2. The viewing direction

for (a) and (b) corresponds to 〈110〉.

G.2 Chapter 6 - WZ ZnSe nanowires

Figure G.4 (a) shows an overview TEM micrograph of WZ-NW-2 indicating the location

of WZ-NW-2 on the TEM grid, with a higher magnification TEM image shown in Fig-

ure G.4 (b). Figure G.5 (a) shows a HR-TEM micrograph from an area of WZ-NW-2.

Figure G.5 (b) shows a FFT pattern taken from an area squared (in red) in (a). The

indexed spots in the FFT pattern confirm that the crystal structure of WZ-NW-2 is WZ.

G.3 Chapter 7 - ZnSe NTSLs

Figure G.6 (a) shows an overview TEM micrograph of an area of NTSL-1(0.100). A

TEM micrograph from an area of NTSL-1(0.100), at a magnification higher than in (a),

is shown in Figure G.6 (b). Figure G.7 shows a HR-TEM micrograph from another


a b

WZ-NW-2

Figure G.4: Structural characterization of WZ-NW-2: (a) Overview TEM micrograph,

and (b) TEM micrograph at a magnification higher than in (a).

a b

1101

11010002

0002

1100

1101

1101

1100

Figure G.5: (a) HR-TEM micrograph from an area of WZ-NW-2, (b) Indexed FFT

pattern acquired from an area (squared in red) in (a). The viewing direction for (a) and

(b) is 〈1100〉.


area of NTSL-1(0.100), where atomically sharp interfaces at the twin-planes can be seen.

Figure G.8 (b) shows an indexed FFT pattern taken from an area highlighted in the

HR-TEM micrograph of NTSL-1(0.100), shown in Figure G.8 (a). The indexed spots

in the FFT pattern in (b) confirm that the individual domains in NTSL-1(0.100) are

〈111〉-ZB, with 〈111〉-direction coinciding with the growth direction of NTSL. Similar

TEM and HR-TEM micrographs, and FFT pattern for NTSL-2(0.077) are shown in

Figures G.9 (a,b), G.10 and G.11(a,b), respectively. Note that the diffraction spots in

the FFT pattern in Figure G.8 (b) are not as clear as those in Figure G.11 (b). This is

because a much smaller area was used in the former case for acquiring the FFT pattern.

a b

Figure G.6: Structural characterization of NTSL-1(0.100): (a) Overview TEM micro-

graph, and (b) HR-TEM micrograph from an area of NTSL-1(0.100). The viewing di-

rection for (a) and (b) is 〈110〉.


Figure G.7: HR-TEM micrograph from an area of NTSL-1(0.100). The periodic twin-

planes as atomically sharp interfaces can be seen. The viewing direction is 〈110〉.


a b

111

111

220

220

002002

111

111

Figure G.8: (a) HR-TEM micrograph from an area of NTSL-1(0.100), and (b) Indexed

FFT pattern taken from the area (squared in red) in (a). The viewing direction is 〈110〉.

a b


graph, and (b) HR-TEM micrograph from an area of NTSL-2(0.077). The viewing di-

rection for (a) and (b) is 〈110〉.


Figure G.10: HR-TEM micrograph from an area of NTSL-2(0.077). The viewing direction

is 〈110〉.


ab

111

111

220

220

002

002

111

111

Fig

ure

G.1

1:(a

)H

R-T

EM

mic

rogr

aph

from

anar

eaof

NT

SL

-2(0

.077

),an

d(b

)In

dex

edF

FT

pat

tern

take

nfr

omth

ear

ea

(squar

edin

red)

in(a

).T

he

vie

win

gdir

ecti

onis〈1

10〉.


Figure G.12 (a) shows an overview TEM micrograph of NTSL-3(0.059), and (b) shows

a HR-TEM micrograph from an area of NTSL-3(0.059). Figure G.13 shows a composite

image of the HR-TEM micrographs of NTSL-3(0.059). Figure G.14 (a) shows a HR-TEM

micrograph from another area of NTSL-3(0.059). The FFT pattern shown in Figure G.14

(b) was acquired from the area squared (in red) in Figure G.14 (a). The indexed spots in

the FFT pattern in (b) show that the individual domains in NTSL-3(0.059) are 〈111〉-ZB,

with 〈111〉-direction coinciding with the growth direction of NTSL. Figure G.15 (a) shows

a HR-TEM micrograph from another area of NTSL-3(0.059) and (b) shows an indexed

SAD pattern from an area of NTSL-3(0.059), shown in Figure G.12 (a). The diffraction

spots corresponding to two rotated ZB domains are shown with labels (with and without

a letter ‘T’). Double spots, characteristic of periodic twinning, seen in the SAD pattern

from an area of NTSL-3(0.059) highlight the excellent periodicity of twin-planes.

a b

Location of SAD


graph from an area of NTSL-3(0.059), and (b) a HR-TEM micrograph from an area of

NTSL-3(0.059). The viewing direction for (a) and (b) is 〈110〉.

Figures G.16 (a,b) show (a) an overview TEM micrograph and (b) a higher magnifica-

tion TEM micrograph from an area of NTSL-4(0.012). Figure G.17 (a) shows a HR-TEM

micrograph from an area of NTSL-4(0.012). The FFT pattern shown in Figure G.17 (b)


a

b

c

d

Fig

ure

G.1

3:A

com

pos

ite

imag

eof

the

HR

-TE

Mm

icro

grap

hs

ofN

TSL

-3(0

.059

),ta

ken

atsu

cces

sive

loca

tion

sal

ong

its

lengt

h.


was acquired from the area squared (in red) in Figure G.17 (a). The indexed spots in the

FFT pattern in (b) confirm that the individual domains in NTSL-4(0.012) are 〈111〉-ZB,

with 〈111〉-direction coinciding with the growth direction of NTSL. This also confirms

that the individual domains separated by twin-planes are 〈111〉-ZB ZnSe for all NTSLs.

Figure G.18 (a) shows a HR-TEM micrograph from another area of NTSL-4(0.012). Fig-

ure G.18 (b) shows an indexed SAD pattern from an area of NTSL-4(0.012). TEM and

HR-TEM micrographs, and FFT and SAD patterns for NTSL-5(0.019) are shown in Fig-

ures G.19 (a,b), G.20 (a,b) and G.21 (a,b). Figure G.22 (a) shows an overview TEM

micrograph of NTSL-6(0.333) which extended on both sides of the TEM grid bar as

shown, with its TEM micrograph on one side shown in (b). Figures G.23 (a,b) show (a)

a higher magnification TEM and (b) HR-TEM micrograph for NTSL-6(0.333) on the side

labelled as ‘Below the grid bar’, while Figures G.24 (a) and (b) show the corresponding

micrographs for side labelled ‘Top of the grid bar’. Both sides were studied to ensure the

same periodicity along the length of NTSL-6(0.333).

a b

111

111

220

220

002

002

111

111




a b

11111

1

220 T

220

002 T

111 T

111 T

131 T

111

111

002

002 T

131

131 T

220

002

220 T

131

222

222


SAD pattern from a larger area (shown in Figure G.12 (a)) of NTSL-3(0.059). The

viewing direction for (a) and (b) is 〈110〉.

a b


graph from an area of NTSL-4(0.012), and (b) a TEM micrograph from an area of

NTSL-4(0.012), at a magnification higher than in (a) . The viewing direction for (b) is

〈110〉.


a b

111

111

220

220

002

002

111 111



a b

111

111 220T

220T

002

111111

131

111T

111T

002T

002

131

131T

220

002T

220

131T

222

222

Figure G.18: (a) HR-TEM micrograph from another area of NTSL-4(0.012), and (b)

Indexed SAD pattern from a larger area of NTSL-4(0.012). The viewing direction for (a)

and (b) corresponds to 〈110〉.


a b

Location of SAD


graph from an area of NTSL-5(0.019), and (b) a HR-TEM micrograph from an area of

NTSL-5(0.019). The viewing direction for (a) and (b) is 〈110〉.

a b

111

111

220

220

002

002111

111




a b

111

111

220T

220T

002

111

111

131

111T

111T

002T

002

131

131T

220

002T

220131

T

222

222

Figure G.21: (a) HR-TEM micrograph from another area of NTSL-5(0.019), and (b)

Indexed SAD pattern from a larger area (shown in Figure G.19 (a)) of NTSL-5(0.019).

The viewing direction for (a) and (b) is 〈110〉.

a b

Top of the

grid bar

Below the

grid bar

NTSL-2


graph of NTSL-6(0.333), which extends on both sides of the grid bar labelled as shown,

and (b) TEM micrograph of NTSL-6(0.333) on the side labelled ‘Top of the grid bar’.


a b

Figure G.23: On the side labelled ‘Below the grid bar’: (a) Overview TEM micrograph

from an area of NTSL-6(0.333), and (b) a HR-TEM micrograph from an area of NTSL-

6(0.333). The viewing direction for (a) and (b) is 〈110〉.

a b

Figure G.24: On the side labelled ‘Top of the grid bar’: (a) Overview TEM micrograph

from an area of NTSL-6(0.333), and (b) a HR-TEM micrograph from an area of NTSL-

6(0.333). The viewing direction for (a) and (b) is 〈110〉.

Bibliography

[1] K.F. Brennan. Introduction to Semiconductor Devices: For Computing and

Telecommunications Applications. Cambridge University Press, 2005.

[2] H. Morkoc, S. Strite, G.B. Gao, M.E. Lin, B. Sverdlov, and M. Burns. Large-band-

gap SiC, III-V nitride, and II-VI ZnSe-based semiconductor device technologies. J.

Appl. Phys., 76(3):1363–1398, 1994.

[3] G.F. Neumark. Defects in wide band gap II-VI crystals. Materials Science and

Engineering R: Reports, 21(1):iii–46, 1997.

[4] E Monroy, F Omnes, and F Calle. Wide-bandgap semiconductor ultraviolet pho-

todetectors. Semiconductor Science and Technology, 18(4):R33, 2003.

[5] D.R. Vij and N. Singh. Luminescence and Related Properties of II-VI Semiconduc-

tors. Nova Science Publishers Inc., 1998.

[6] H.E. Ruda. Widegap II-VI Compounds for Opto-electronic Applications. Chapman

and Hall, 1992.

[7] J. Gutowski, N. Presser, and G. Kudlek. Optical Properties of ZnSe Epilayers and

Films. physica status solidi (a), 120(1):11–59, 1990.

240

Bibliography 241

[8] Y.S. Park and B.K. Shin. Recent Advances in Injection Luminescence in II-VI

Compounds. In J. Pankove, editor, Electroluminescence, volume 17 of Topics in

Applied Physics, pages 133–170. Springer Berlin / Heidelberg, 1977.

[9] M. Aven and J.Z. Devine. Advances in Injection Luminescence of II-VI Compounds.

Journal of Luminescence, 7:195–212, 1973.

[10] W.M. Yim, J.P. Dismukes, E.J. Stofko, and R.J. Ulmer. Miscibility between ZnSe

and CdTe. physica status solidi (a), 13(1):K57–K61, 1972.

[11] R P Vijayalakshmi, R Venugopal, D R Reddy, and B K Reddy. Structural and

bandgap studies of ZnSxCdSe1−x thin films. Semiconductor Science and Technology,

9(5):1062–1068, 1994.

[12] R.W.G. Wyckoff. Crystal structures. Interscience Publishers, 1963.

[13] H. Hartmann. On the Growth of Hexagonal ZnSe Single Crystals. Kristall und

Technik, 5(4):527–534, 1970.

[14] B.J. Skromme. Luminescence as a Diagnostic of Wide-Gap II-VI Compound Semi-

conductor Materials. Annual Review of Materials Science, 25:601–646, 1995.

[15] H Luo and J K Furdyna. The II-VI semiconductor blue-green laser: challenges and

solution. Semiconductor Science and Technology, 10(8):1041–1048, 1995.

[16] L.A. Kolodziejski, R.L. Gunshor, and A.V. Nurmikko. Wide-Bandgap II-VI Het-

erostructures for Blue/Green Optical Sources: Key Materials Issues. Annual Review

of Materials Science, 25:711–753, 1995.

[17] Guihua Yu and Charles M. Lieber. Assembly and integration of semiconductor

nanowires for functional nanosystems. Pure and Applied Chemistry, 82(12):2295–

2314, 2010.

Bibliography 242

[18] Jiansheng Jie, Wenjun Zhang, Igor Bello, Chun-Sing Lee, and Shuit-Tong Lee.

One-dimensional II-VI nanostructures: Synthesis, properties and optoelectronic

applications. Nano Today, 5(4):313–336, 2010.

[19] Kimberly A. Dick. A review of nanowire growth promoted by alloys and non-

alloying elements with emphasis on Au-assisted III-V nanowires. Progress in Crystal

Growth and Characterization of Materials, 54(3-4):138–173, 2008.

[20] Xiaoxiao Li, Li Wang, and Guangqing Yan. Review: Recent research progress

on preparation of silver nanowires by soft solution method and their applications.

Crystal Research and Technology, 46(5):427–438, 2011.

[21] S. Picraux, Shadi Dayeh, Pradeep Manandhar, Daniel Perea, and Sukgeun Choi.

Silicon and Germanium Nanowires: Growth, Properties, and Integration. JOM

Journal of the Minerals, Metals and Materials Society, 62(4):35–43, 2010.

[22] Volker Schmidt, Joerg V. Wittemann, Stephan Senz, and Ulrich Gosele. Silicon

Nanowires: A Review on Aspects of their Growth and their Electrical Properties.

Advanced Materials, 21(25-26):2681–2702, 2009.

[23] J. Pan, H. Shen, and S. Mathur. One-dimensional SnO2 nanostructures: Synthesis

and Applications. Journal of Nanotechnology, 2012, 2012.

[24] Mildred S. Dresselhaus, Yu-Ming Lin, Oded Rabin, Marcie R. Black, Jing Kong,

and Gene Dresselhaus. Nanowires. In Bharat Bhushan, editor, Springer Handbook

of Nanotechnology, chapter 4, pages 119–167. Springer Berlin Heidelberg, 2010.

[25] H. Yu, J. Li, R.A. Loomis, L.-W. Wang, and W.E. Buhro. Two-versus three-

dimensional quantum confinement in indium phosphide wires and dots. Nature

Materials, 2(8):517–520, 2003.

Bibliography 243

[26] Peidong Yang, Ruoxue Yan, and Melissa Fardy. Semiconductor Nanowire: What’s

Next? Nano Letters, 10(5):1529–1536, 2010.

[27] Kwak J.-W. Park J.W. Choi, Y.-E. Nanotechnology for Early Cancer Detection.

Sensors, 10(1):428–455, 2010.

[28] Kannan and Balasubramanian. Challenges in the use of 1D nanostructures for

on-chip biosensing and diagnostics: A review. Biosensors and Bioelectronics,

26(4):1195–1204, 2010. Selected Papers from the World Congress on Biosen-

sors,Glasgow, Scotland, UK May 26-28, 2010.

[29] Xiaoqin Chi, Dengtong Huang, Zhenghuan Zhao, Zijian Zhou, Zhenyu Yin, and

Jinhao Gao. Nanoprobes for invitro diagnostics of cancer and infectious diseases.

Biomaterials, 33(1):189–206, 2012.

[30] Niranjan S. Ramgir, Yang Yang, and Margit Zacharias. Nanowire-Based Sensors.

Small, 6(16):1705–1722, 2010.

[31] R. Yan, D. Gargas, and P. Yang. Nanowire photonics. Nature Photonics, 3(10):569–

576, 2009.

[32] W. Lu and C.M. Lieber. Nanoelectronics from the bottom up. Nature Materials,

6(11):841–850, 2007.

[33] Wei Lu, Ping Xie, and C.M. Lieber. Nanowire Transistor Performance Limits and

Applications. Electron Devices, IEEE Transactions on, 55(11):2859–2876, 2008.

[34] C.M. Lieber. Semiconductor nanowires: A platform for nanoscience and nanotech-

nology. MRS Bulletin, 36(12):1052–1063, 2011.

[35] Bozhi Tian, Thomas J. Kempa, and Charles M. Lieber. Single nanowire photo-

voltaics. Chem. Soc. Rev., 38(1):16–24, 2009.

Bibliography 244

[36] Ke Sun, A. Kargar, Namsoek Park, K.N. Madsen, P.W. Naughton, T. Bright,

Yi Jing, and Deli Wang. Compound Semiconductor Nanowire Solar Cells. Selected

Topics in Quantum Electronics, IEEE Journal of, 17(4):1033–1049, 2011.

[37] L. VJ, J. Oh, A.P. Nayak, A.M. Katzenmeyer, K.H. Gilchrist, S. Grego, N.P.

Kobayashi, S.-Y. Wang, A.A. Talin, N.K. Dhar, and M.S. Islam. A Perspective on

Nanowire Photodetectors: Current Status, Future Challenges, and Opportunities.

Selected Topics in Quantum Electronics, IEEE Journal of, 17(4):1002–1032, 2011.

[38] D.O. Demchenko, P.D. Heinz, and B. Lee. Determining factors of thermoelectric

properties of semiconductor nanowires. Nanoscale Research Letters, 6(1):502, 2011.

[39] W Kim. Thermal transport in individual thermoelectric nanowires: a review. Ma-

terials Research Innovations, 15(6):375–385, 2011-11-01T00:00:00.

[40] Christopher J. Vineis, Ali Shakouri, Arun Majumdar, and Mercouri G. Kanatzidis.

Nanostructured Thermoelectrics: Big Efficiency Gains from Small Features. Ad-

vanced Materials, 22(36):3970–3980, 2010.

[41] Park J.-H.-Choi Y. Heo C.-J. Yang S.-M. Lee L.P. Yang P. Yan, R. Nanowire-based

single-cell endoscopy. Nature Nanotechnology, 7(3):191–196, 2011.

[42] Nam Hoon Kwon, Miles F. Beaux, Chad Ebert, Lidong Wang, Brian E. Lassiter,

Yong Ho Park, David N. McIlroy, Carolyn J. Hovde, and Gregory A. Bohach.

Nanowire-Based Delivery of Escherichia coli O157 Shiga Toxin 1 A Subunit into

Human and Bovine Cells. Nano Letters, 7(9):2718–2723, 2007.

[43] J.C. Johnson, H.-J. Choi, K.P. Knutsen, R.D. Schaller, P. Yang, and R.J. Saykally.

Single gallium nitride nanowire lasers. Nature Materials, 1(2):106–110, 2002.

Bibliography 245

[44] Michael H. Huang, Samuel Mao, Henning Feick, Haoquan Yan, Yiying Wu, Hannes

Kind, Eicke Weber, Richard Russo, and Peidong Yang. Room-Temperature Ultra-

violet Nanowire Nanolasers. Science, 292(5523):1897–1899, 2001.

[45] Wang G.-Zhou-W. Lin Y. Chernyak-L. Zhao J.-Kong J. Li L. Ren J. Liu J. Chu, S.

Electrically pumped waveguide lasing from ZnO nanowires. Nature Nanotechnology,

6(8):506–510, 2011.

[46] Mariano A Zimmler, Federico Capasso, Sven Muller, and Carsten Ronning. Opti-

cally pumped nanowire lasers: invited review. Semiconductor Science and Tech-

nology, 25(2):024001, 2010.

[47] C.Z. Ning. Semiconductor nanolasers. physica status solidi (b), 247(4):774–788,

2010.

[48] Daniel Vanmaekelbergh and Lambert K. van Vugt. ZnO nanowire lasers. Nanoscale,

3(7):2783–2800, 2011.

[49] P. Yang, H. Yan, S. Mao, R. Russo, J. Johnson, R. Saykally, N. Morris, J. Pham,

R. He, and H.-J. Choi. Controlled Growth of ZnO Nanowires and Their Optical

Properties. Advanced Functional Materials, 12(5):323–331, 2002.

[50] Zhong Lin Wang. Zinc oxide nanostructures: growth, properties and applications.

Journal of Physics: Condensed Matter, 16(25):R829–R858, 2004.

[51] Xuchuan Jiang, Yi Xie, Jun Lu, Liying Zhu, Wei He, and Yitai Qian. Simultaneous

In Situ Formation of ZnS Nanowires in a Liquid Crystal Template by γ-Irradiation.

Chemistry of Materials, 13(4):1213–1218, 2001.

[52] Dongsheng Xu, Xuesong Shi, Guolin Guo, Linlin Gui, and Youqi Tang. Electro-

chemical Preparation of CdSe Nanowire Arrays. The Journal of Physical Chemistry

B, 104(21):5061–5063, 2000.

Bibliography 246

[53] L.V. Titova, Thang Ba Hoang, H.E. Jackson, L.M. Smith, J.M. Yarrison-Rice, J.L.

Lensch, and L.J. Lauhon. Low-temperature photoluminescence imaging and time-

resolved spectroscopy of single CdS nanowires. Appl. Phys. Lett., 89(5):053119,

2006.

[54] Tianyou Zhai, Xiaosheng Fang, Liang Li, Yoshio Bando, and Dmitri Golberg.

One-dimensional CdS nanostructures: synthesis, properties, and applications.

Nanoscale, 2(2):168–187, 2010.

[55] B. Xiang, H.Z. Zhang, G.H. Li, F.H. Yang, F.H. Su, R.M. Wang, J. Xu, G.W. Lu,

X.C. Sun, Q. Zhao, and D.P. Yu. Green-light-emitting ZnSe nanowires fabricated

via vapor phase growth. Appl. Phys. Lett., 82(19):3330–3332, 2003.

[56] Ying-Chun Zhu and Yoshio Bando. Preparation and photoluminescence of single-

crystal zinc selenide nanowires. Chemical Physics Letters, 377(3-4):367–370, 2003.

[57] X.T. Zhang, Z. Liu, Y.P. Leung, Quan Li, and S.K. Hark. Growth and lumines-

cence of zinc-blende-structured ZnSe nanowires by metal-organic chemical vapor

deposition. Appl. Phys. Lett., 83(26):5533–5535, 2003.

[58] Ruitao Lv, Chuanbao Cao, Huazhang Zhai, Dazhi Wang, Siyuan Liu, and Hesun

Zhu. Growth and characterization of single-crystal ZnSe nanorods via surfactant

soft-template method. Solid State Communications, 130(3-4):241–245, 2004.

[59] X.T. Zhang, Z. Liu, K.M. Ip, Y.P. Leung, Quan Li, and S.K. Hark. Luminescence

of ZnSe nanowires grown by metalorganic vapor phase deposition under different

pressures. J. Appl. Phys., 95(10):5752–5755, 2004.

[60] C. Ye, X. Fang, Y. Wang, P. Yan, J. Zhao, and L. Zhang. Structural characteriza-

tion of long ZnSe nanowires. Applied Physics A: Materials Science & Processing,

79(1):113–115, 2004.

Bibliography 247

[61] Yang Jiang, Xiang-Min Meng, Wing-Ching Yiu, Ji Liu, Jun-Xian Ding, Chun-Sing

Lee, and Shuit-Tong Lee. Zinc Selenide Nanoribbons and Nanowires. The Journal

of Physical Chemistry B, 108(9):2784–2787, 2004.

[62] Minghai Chen and Lian Gao. Synthesis and characterization of wurtzite ZnSe one-

dimensional nanocrystals through molecular precursor decomposition by solvother-

mal method. Materials Chemistry and Physics, 91(2-3):437–441, 2005.

[63] S. Xiong, J. Shen, Q. Xie, Y. Gao, Q. Tang, and Y.T. Qian. A Precursor-Based

Route to ZnSe Nanowire Bundles. Advanced Functional Materials, 15(11):1787–

1792, 2005.

[64] Y.P. Leung, Z. Liu, and S.K. Hark. Changes in morphology and growth rate of

quasi-one-dimensional ZnSe nanowires on GaAs (100) substrates by metalorganic

chemical vapor deposition. Journal of Crystal Growth, 279(3-4):248–257, 2005.

[65] K M Ip, Z Liu, C M Ng, and S K Hark. Effects of passivation and ambient gases on

the photoluminescence of ZnSe nanowires. Nanotechnology, 16(8):1144–1147, 2005.

[66] X. Zhang, Z. Liu, Q. Li, Y. Leung, K. Ip, and S. Hark. Routes to Grow Well-Aligned

Arrays of ZnSe Nanowires and Nanorods. Advanced Materials, 17(11):1405–1410,

2005.

[67] A Colli, S Hofmann, A C Ferrari, F Martelli, S Rubini, C Ducati, A Franciosi, and

J Robertson. Selective growth of ZnSe and ZnCdSe nanowires by molecular beam

epitaxy. Nanotechnology, 16(5):S139–S142, 2005.

[68] Junqing Hu, Yoshio Bando, and Dmitri Golberg. Sn-Catalyzed Thermal Evapora-

tion Synthesis of Tetrapod-Branched ZnSe Nanorod Architectures. Small, 1(1):95–

99, 2005.

Bibliography 248

[69] A.B. Panda, S. Acharya, and S. Efrima. Ultranarrow ZnSe Nanorods and

Nanowires: Structure, Spectroscopy, and One-Dimensional Properties. Advanced

Materials, 17(20):2471–2474, 2005.

[70] Y.P. Leung, Wallace C.H. Choy, I. Markov, G.K.H. Pang, H.C. Ong, and T.I. Yuk.

Synthesis of wurtzite ZnSe nanorings by thermal evaporation. Appl. Phys. Lett.,

88(18):183110, 2006.

[71] J. Salfi, U. Philipose, C.F. de Sousa, S. Aouba, and H.E. Ruda. Electrical proper-

ties of Ohmic contacts to ZnSe nanowires and their application to nanowire-based

photodetection. Appl. Phys. Lett., 89(26):261112, 2006.

[72] U. Philipose, T. Xu, S. Yang, Ping Sun, Harry E. Ruda, Y.Q. Wang, and K.L.

Kavanagh. Enhancement of band edge luminescence in ZnSe nanowires. J. Appl.

Phys., 100(8):084316, 2006.

[73] Hui Song, Dazhi Wang, Xichen Ma, Yafeng Cao, and Shufen Zhang. Soft ZnSe

nanobelts: Characterization and feasibility for assembling tubular ZnSe nanostruc-

tures. Solid State Communications, 139(8):430–433, 2006.

[74] Xia Dong-Yan, Dai Lun, Xu Wan-Jin, You Li-Ping, Zhang Bo-Rui, Ran Guang-

Zhao, and Qin Guo-Gang. Synthesis and PL Properties of ZnSe Nanowires with

Zincblende and Wurtzite Structures. Chinese Physics Letters, 23(5):1317, 2006.

[75] Hongzhi Fu, Huanyong Li, Wanqi Jie, and Lan Yang. The growth and characteri-

zation of ZnSe nanoneedles by a simple chemical vapor deposition method. Journal

of Crystal Growth, 289(2):440–444, 2006.

[76] Lei Jin, Wallace C.H. Choy, Yee P. Leung, Tong I. Yuk, Hock C. Ong, and Jian

bo Wang. Synthesis and analysis of abnormal wurtzite ZnSe nanowheels. J. Appl.

Phys., 102(4):044302, 2007.

Bibliography 249

[77] Y.Q. Wang, U. Philipose, T. Xu, H.E. Ruda, and K.L. Kavanagh. Twinning mod-

ulation in ZnSe nanowires. Semiconductor Science and Technology, 22(3):175–178,

2007.

[78] U. Philipose, S. Yang, T. Xu, and Harry E. Ruda. Origin of the red lumines-

cence band in photoluminescence spectra of ZnSe nanowires. Appl. Phys. Lett.,

90(6):063103, 2007.

[79] S. Acharya, A.B. Panda, S. Efrima, and Y. Golan. Polarization Properties

and Switchable Assembly of Ultranarrow ZnSe Nanorods. Advanced Materials,

19(8):1105–1108, 2007.

[80] Yuan-yuan Zhang, Jia-fu Chen, Xu-sheng Zheng, and Xue-hui Dong. Preparation

of Manganese-doped ZnSe Precursor Nanoribbon Bundles and Investigation of its

Magneto-optical Properties. Chinese Journal of Chemical Physics, 20(5):607, 2007.

[81] U. Philipose, Ping Sun, Tao Xu, Harry E. Ruda, L. Yang, and K.L. Kavanagh.

Structure and photoluminescence of ZnSe nanostructures fabricated by vapor phase

growth. J. Appl. Phys., 101(1):014326, 2007.

[82] Feifei Wang, Zhihua Zhang, Ruibin Liu, Xiao Wang, Xing Zhu, Anlian Pan, and

Bingsuo Zou. Structure and stimulated emission of ZnSe nanoribbons grown by

thermal evaporation. Nanotechnology, 18(30):305705, 2007.

[83] Asit Baran Panda, Somobrata Acharya, Shlomo Efrima, and Yuval Golan. Syn-

thesis, Assembly, and Optical Properties of Shape- and Phase-Controlled ZnSe

Nanostructures. Langmuir, 23(2):765–770, 2007.

[84] Z.D. Hu, X.F. Duan, M. Gao, Q. Chen, and L.-M. Peng. ZnSe Nanobelts and

Nanowires Synthesized by a Closed Space Vapor Transport Technique. The Journal

of Physical Chemistry C, 111(7):2987–2991, 2007.

Bibliography 250

[85] U Philipose, Ankur Saxena, Harry E Ruda, P J Simpson, Y Q Wang, and K L

Kavanagh. Defect studies of ZnSe nanowires. Nanotechnology, 19(21):215715, 2008.

[86] Sivakumar Ramanathan, Sridhar Patibandla, Supriyo Bandyopadhyay, John An-

derson, and Jarrod D Edwards. Fluorescence spectroscopy of electrochemically

self-assembled ZnSe and Mn:ZnSe nanowires. Nanotechnology, 19(19):195601, 2008.

[87] Junli Wang and Qing Yang. One-Dimensional Angle-Shaped ZnSe Nanostructures:

Synthesis and Formation Mechanism. Crystal Growth & Design, 8(2):660–664,

2008.

[88] Xia Fan, Xiang Min Meng, Xiao Hong Zhang, Ming Liang Zhang, Jian Sheng Jie,

Wen Jun Zhang, Chun Sing Lee, and Shuit Tong Lee. Formation and Photoelectric

Properties of Periodically Twinned ZnSe/SiO2 Nanocables. The Journal of Physical

Chemistry C, 113(3):834–838, 2009.

[89] Bo Hou, Yanjuan Li, Yongjun Liu, Bo Yuan, Mingfen Jia, and Fengzhi Jiang. A

simple way of shape-controlled synthesis of ZnSe nanocrystals: nanodots, nanoflow-

ers, and nanotubes. CrystEngComm, 11(9):1789–1792, 2009.

[90] Liang Shi, Yeming Xu, and Quan Li. Controlled Fabrication of ZnSe Arrays of

Well-Aligned Nanorods, Nanowires, and Nanobelts with a Facile Template-Free

Route. The Journal of Physical Chemistry C, 113(5):1795–1799, 2009.

[91] C. Lee, S. Park, and J. Jun. Enhanced Photoluminescence of ZnSe Nanowires with

Branches. Journal of the Korean Physical Society, 55(2):554–557, 2009.

[92] Patrick T.K. Chin, Jan W. Stouwdam, and Rene A.J. Janssen. Highly Luminescent

Ultranarrow Mn Doped ZnSe Nanowires. Nano Letters, 9(2):745–750, 2009.

Bibliography 251

[93] Chung-Liang Cheng and Yang-Fang Chen. Low temperature synthesis of ZnSe

nanowires by self-catalytic liquid-solid growth. Materials Chemistry and Physics,

115(1):158–160, 2009.

[94] Haitao Wang, Tian Tian, Shancheng Yan, Ningping Huang, and Zhongdang Xiao.

Large-scale synthesis of ZnSe nanoribbons on zinc substrate. Journal of Crystal

Growth, 311(14):3787–3791, 2009.

[95] Jun Liu and Dongfeng Xue. Solution-based route to semiconductor film: Well-

aligned ZnSe nanobelt arrays. Thin Solid Films, 517(17):4814–4817, 2009.

[96] Lihui Zhang and Heqing Yang. The Ag+ induced solution-liquid-solid growth,

photoluminescence and photocatalytic activity of twinned ZnSe nanowires. Applied

Physics A: Materials Science & Processing, 98(4):801–810, 2010.

[97] Yinxiao Du and Qing-Xin Yuan. Facile synthesis and photoluminescence of ZnSe

nanowires. Journal of Alloys and Compounds, 492(1-2):548–551, 2010.

[98] T.K. Das, R. Bhattacharya, A. Manna, and S. Saha. Role of reducing agent in

the formation of ZnSe nanorods by chemical reduction method. European Physical

Journal Applied Physics, 51:30605, 2010.

[99] Dongmei Han, Chunfeng Song, and Xiaoyu Li. Synthesis and Fluorescence Property

of Mn-Doped ZnSe Nanowires. Journal of Nanomaterials, 2010:290763, 2010.

[100] M. Huang. Synthesis and photoluminescence properties of washboard belt-like ZnSe

nanostructures. Optoelectronics and Advanced Materials-Rapid Communications,

4(12):2076–2079, 2010.

[101] Tingting Yao, Qing Zhao, Zhengping Qiao, Feng Peng, Hongjuan Wang, Hao

Yu, Cheng Chi, and Jian Yang. Chemical Synthesis, Structural Characteriza-

Bibliography 252

tion, Optical Properties, and Photocatalytic Activity of Ultrathin ZnSe Nanorods.

Chemistry–A European Journal, 17(31):8663–8670, 2011.

[102] H.W. Zhu, P.G. Li, M. Lei, L.H. Li, S.L. Wang, and W.H. Tang. Sublimation sand-

wich route to ultralong zinc-blende ZnSe nanowires and the cathodoluminescence

properties of individual nanowires. Journal of Alloys and Compounds, 509(7):3306–

3309, 2011.

[103] X.T. Zhang, K.M. Ip, Z. Liu, Y.P. Leung, Quan Li, and S.K. Hark. Structure and

photoluminescence of ZnSe nanoribbons grown by metal organic chemical vapor

deposition. Appl. Phys. Lett., 84(14):2641–2643, 2004.

[104] Q. Li, X. Gong, C. Wang, J. Wang, K. Ip, and S. Hark. Size-Dependent Periodically

Twinned ZnSe Nanowires. Advanced Materials, 16(16):1436–1440, 2004.

[105] Yutaka Ohno, Takeo Shirahama, Seiji Takeda, Atsushi Ishizumi, and Yoshihiko

Kanemitsu. Fe-catalytic growth of ZnSe nanowires on a ZnSe(001) surface at low

temperatures by molecular-beam epitaxy. Appl. Phys. Lett., 87(4):043105, 2005.

[106] X.T. Zhang, K.M. Ip, Quan Li, and S.K. Hark. Photoluminescence of Ag-doped

ZnSe nanowires synthesized by metalorganic chemical vapor deposition. Appl. Phys.

Lett., 86(20):203114, 2005.

[107] C X Shan, Z Liu, X T Zhang, C C Wong, and S K Hark. Wurtzite ZnSe nanowires:

growth, photoluminescence, and single-wire Raman properties. Nanotechnology,

17(22):5561–5564, 2006.

[108] Z Liu and S K Hark. Localized cathodoluminescence of individual ZnSe nanorods.

Nanotechnology, 17(5):1355–1358, 2006.

Bibliography 253

[109] S.Y. Liu, Wallace C.H. Choy, L. Jin, Y.P. Leung, G.P. Zheng, Jianbo Wang, and

A.K. Soh. Triple-Crystal Zinc Selenide Nanobelts. The Journal of Physical Chem-

istry C, 111(26):9055–9059, 2007.

[110] Adrien Tribu, Gregory Sallen, Thomas Aichele, Regis Andre, Jean-Philippe Poizat,

Catherine Bougerol, Serge Tatarenko, and Kuntheak Kheng. A High-Temperature

Single-Photon Source from Nanowire Quantum Dots. Nano Letters, 8(12):4326–

4329, 2008.

[111] Thomas Aichele, Adrien Tribu, Catherine Bougerol, Kuntheak Kheng, Regis Andre,

and Serge Tatarenko. Defect-free ZnSe nanowire and nanoneedle nanostructures.

Appl. Phys. Lett., 93(14):143106, 2008.

[112] Ankur Saxena, Suxia Yang, U. Philipose, and Harry E. Ruda. Excitonic and pair-

related photoluminescence in ZnSe nanowires. J. Appl. Phys., 103(5):053109, 2008.

[113] Y.P. Leung, Wallace C.H. Choy, and T.I. Yuk. Linearly resistive humidity sensor

based on quasi one-dimensional ZnSe nanostructures. Chemical Physics Letters,

457(1-3):198–201, 2008.

[114] B.J. Cooley, T.E. Clark, B.Z. Liu, C.M. Eichfeld, E.C. Dickey, S.E. Mohney,

S.A. Crooker, and N. Samarth. Growth of Magneto-optically Active (Zn,Mn)Se

Nanowires. Nano Letters, 9(9):3142–3146, 2009.

[115] Thomas Aichele, Adrien Tribu, Gregory Sallen, Juanita Bocquel, Edith Bellet-

Amalric, Catherine Bougerol, Jean-Philippe Poizat, Kuntheak Kheng, Regis Andre,

Serge Tatarenko, and Henri Mariette. CdSe quantum dots in ZnSe nanowires as effi-

cient source for single photons up to 220 K. Journal of Crystal Growth, 311(7):2123–

2127, 2009. International Conference on Molecular Beam Epitaxy (MBE-XV)-The

15th International Conference on Molecular Beam Epitaxy (MBE-XV).

Bibliography 254

[116] Thomas Aichele, Adrien Tribu, Catherine Bougerol, Kuntheak Kheng, Fabrice Do-

natini, Le Si Dang, Regis Andre, and Serge Tatarenko. Growth and properties of

defect-free ZnSe nanowires and nanoneedles. physica status solidi (b), 246(4):812–

815, 2009.

[117] C.H. Hsiao, S.J. Chang, S.B. Wang, S.C. Hung, S.P. Chang, T.C. Li, W.J. Lin,

and B.R. Huang. MBE growth of ZnSe nanowires on oxidized silicon substrate.

Superlattices and Microstructures, 46(4):572–577, 2009.

[118] C.H. Hsiao, S.J. Chang, S.B. Wang, S.P. Chang, T.C. Li, W.J. Lin, C.H. Ko,

T.M. Kuan, and B.R. Huang. ZnSe Nanowire Photodetector Prepared on Oxidized

Silicon Substrate by Molecular-Beam Epitaxy. Journal of The Electrochemical

Society, 156(4):J73–J76, 2009.

[119] Shanying Li, Yang Jiang, Binbin Wang, Di Wu, Junwei Li, Yugang Zhang, Ben

Yang, Xianan Ding, Hongyang Zhou, and Honghai Zhong. Synthesis of p-type

ZnSe nanowires by atmosphere compensating technique. Micro & Nano Letters,

6(6):459–462, 2011.

[120] Yao Liang, Yin Tao, and S.K. Hark. Fabrication and optical properties of ver-

tically aligned ZnSe nanowire arrays catalyzed by Ga and Ag. CrystEngComm,

13(19):5751–5754, 2011.

[121] Xiaosheng Fang, Shenglin Xiong, Tianyou Zhai, Yoshio Bando, Meiyong Liao,

Ujjal K. Gautam, Yasuo Koide, Xiaogang Zhang, Yitai Qian, and Dmitri Gol-

berg. High-Performance Blue/Ultraviolet-Light-Sensitive ZnSe-Nanobelt Photode-

tectors. Advanced Materials, 21(48):5016–5021, 2009.

[122] P. Caroff, J. Bolinsson, and J. Johansson. Crystal Phases in III-V Nanowires: From

Random Toward Engineered Polytypism. Selected Topics in Quantum Electronics,

IEEE Journal of, 17(4):829–846, 2011.

Bibliography 255

[123] Xiaosheng Fang, Yoshio Bando, Ujjal K. Gautam, Tianyou Zhai, Silvija Gradecak,

and Dmitri Golberg. Heterostructures and superlattices in one-dimensional

nanoscale semiconductors. J. Mater. Chem., 19(32):5683–5689, 2009.

[124] P. Caroff, K.A. Dick, J. Johansson, M.E. Messing, K. Deppert, and L. Samuel-

son. Controlled polytypic and twin-plane superlattices in III-V nanowires. Nature

Nanotech., 4:50–55, 2009.

[125] Jessica Bolinsson, Philippe Caroff, Bernhard Mandl, and Kimberly A Dick.

Wurtzite–zincblende superlattices in InAs nanowires using a supply interruption

method. Nanotechnology, 22(26):265606, 2011.

[126] N. Akopian, G. Patriarche, L. Liu, J.-C. Harmand, and V. Zwiller. Crystal Phase

Quantum Dots. Nano Letters, 10(4):1198–1201, 2010.

[127] Kuranananda Pemasiri, Mohammad Montazeri, Richard Gass, Leigh M. Smith,

Howard E. Jackson, Jan Yarrison-Rice, Suriati Paiman, Qiang Gao, H. Hoe Tan,

Chennupati Jagadish, Xin Zhang, and Jin Zou. Carrier Dynamics and Quan-

tum Confinement in type II ZB-WZ InP Nanowire Homostructures. Nano Letters,

9(2):648–654, 2009.

[128] A. Pan, P.L. Nichols, and C.Z. Ning. Semiconductor Alloy Nanowires and

Nanobelts With Tunable Optical Properties. Selected Topics in Quantum Elec-

tronics, IEEE Journal of, 17(4):808–818, 2011.

[129] Xiujuan Zhuang, C.Z. Ning, and Anlian Pan. Composition and Bandgap-Graded

Semiconductor Alloy Nanowires. Advanced Materials, 24(1):13–33, 2012.

[130] Lorenzo Pavesi and Mario Guzzi. Photoluminescence of AlxGa1−xAs alloys. J.

Appl. Phys., 75(10):4779–4842, 1994.

Bibliography 256

[131] Yufeng Hao, Guowen Meng, Zhong Lin Wang, Changhui Ye, and Lide Zhang.

Periodically Twinned Nanowires and Polytypic Nanobelts of ZnS: The Role of

Mass Diffusion in Vapor-Liquid-Solid Growth. Nano Letters, 6(8):1650–1655, 2006.

[132] R.E. Algra, M.A. Verheijen, M.T. Borgstrom, L.-F. Feiner, G. Immink, W.J.P.

Van Enckevort, E. Vlieg, and E.P.A.M. Bakkers. Twinning superlattices in indium

phosphide nanowires. Nature, 456(7220):369–372, 2008.

[133] Jiming Bao, David C. Bell, Federico Capasso, Jakob B. Wagner, Thomas

Martensson, Johanna Tragardh, and Lars Samuelson. Optical Properties of Ro-

tationally Twinned InP Nanowire Heterostructures. Nano Letters, 8(3):836–841,

2008.

[134] Chin-Yu Yeh, Z.W. Lu, S. Froyen, and Alex Zunger. Zinc-blende–wurtzite poly-

typism in semiconductors. Phys. Rev. B, 46(16):10086–10097, 1992.

[135] B.Y. Geng, X.W. Liu, Q.B. Du, X.W. Wei, and L.D. Zhang. Structure and optical

properties of periodically twinned ZnS nanowires. Appl. Phys. Lett., 88(16):163104,

2006.

[136] Frederic Sansoz. Surface Faceting Dependence of Thermal Transport in Silicon

Nanowires. Nano Letters, 11(12):5378–5382, 2011.

[137] Z. Ikonic, G.P. Srivastava, and J.C. Inkson. Twinning superlattices. Solid State

Commun., 86(12):799–802, 1993.

[138] Z. Ikonic, G.P. Srivastava, and J.C. Inkson. Optical properties of twinning su-

perlattices in diamond-type and zinc-blende-type semiconductors. Phys. Rev. B,

52(19):14078–14085, 1995.

[139] Sajal Dhara, Hari S. Solanki, Arvind Pawan R., Vibhor Singh, Shamashis Sengupta,

B.A. Chalke, Abhishek Dhar, Mahesh Gokhale, Arnab Bhattacharya, and Man-

Bibliography 257

dar M. Deshmukh. Tunable thermal conductivity in defect engineered nanowires

at low temperatures. Phys. Rev. B, 84(12):121307(R), 2011.

[140] P.L. Edwards and R.J. Happel Jr. Beryllium Oxide Whiskers and Platelets. J.

Appl. Phys., 33(3):943–948, 1962.

[141] J. Westwater, D.P. Gosain, S. Tomiya, S. Usui, and H. Ruda. Growth of silicon

nanowires via gold/silane vapor–liquid–solid reaction. Journal of Vacuum Science

& Technology B: Microelectronics and Nanometer Structures, 15(3):554–557, 1997.

[142] X. Duan and C.M. Lieber. General Synthesis of Compound Semiconductor

Nanowires. Advanced Materials, 12(4):298–302, 2000.

[143] Y. Xia, P. Yang, Y. Sun, Y. Wu, B. Mayers, B. Gates, Y. Yin, F. Kim, and H. Yan.

One-Dimensional Nanostructures: Synthesis, Characterization, and Applications.

Advanced Materials, 15(5):353–389, 2003.

[144] Matt Law, Joshua Goldberger, and Peidong Yang. Semiconductor Nanowires and

Nanotubes. Annual Review of Materials Research, 34(1):83–122, 2004.

[145] Sven Barth, Francisco Hernandez-Ramirez, Justin D. Holmes, and Albert Romano-

Rodriguez. Synthesis and applications of one-dimensional semiconductors. Progress

in Materials Science, 55(6):563–627, 2010.

[146] Seth A Fortuna and Xiuling Li. Metal-catalyzed semiconductor nanowires: a re-

view on the control of growth directions. Semiconductor Science and Technology,

25(2):024005, 2010.

[147] R.S. Wagner and W.C. Ellis. Vapor-liquid-solid mechanism of single crystal growth.

Appl. Phys. Lett., 4(5):89–90, 1964.

Bibliography 258

[148] Qihua Xiong, G. Chen, J.D. Acord, X. Liu, J.J. Zengel, H.R. Gutierrez, J.M.

Redwing, L.C. Lew Yan Voon, B. Lassen, and P.C. Eklund. Optical Properties of

Rectangular Cross-sectional ZnS Nanowires. Nano Letters, 4(9):1663–1668, 2004.

[149] J. Hu, T.W. Odom, and C.M. Lieber. Chemistry and Physics in One Dimension:

Synthesis and Properties of Nanowires and Nanotubes. Accounts of Chemical Re-

search, 32(5):435–445, 1999.

[150] Y.F. Chan, X.F. Duan, S.K. Chan, I.K. Sou, X.X. Zhang, and N. Wang. ZnSe

nanowires epitaxially grown on GaP(111) substrates by molecular-beam epitaxy.

Appl. Phys. Lett., 83(13):2665–2667, 2003.

[151] Liang Shi, Yeming Xu, and Quan Li. Synthesis of ZnSe nanodonuts via a surfactant-

assisted process. Solid State Communications, 146(9-10):384–386, 2008.

[152] L Shi, Y M Xu, and Quan Li. Fabrication of nanopeas with ZnSe-filled SiO2

nanotube/nanowire configuration. Nanotechnology, 16(10):2100–2103, 2005.

[153] Richard H. Bube. Temperature Dependence of the Width of the Band Gap in

Several Photoconductors. Phys. Rev., 98(2):431–433, 1955.

[154] D.C. Reynolds, L.S. Pedrotti, and O.W. Larson. Edge Emission in Zinc Selenide

Single Crystals. J. Appl. Phys., 32(10):2250–2254, 1961.

[155] P.J. Dean and J.L. Merz. Pair Spectra and “Edge Emission” in Zinc Selenide.

Phys. Rev., 178(3):1310–1318, 1969.

[156] Koh Era and Dietrich W. Langer. Luminescence of ZnSe near the band edge under

strong laser light excitation. Journal of Luminescence, 1-2:514–527, 1970.

[157] J.L. Merz, H. Kukimoto, K. Nassau, and J.W. Shiever. Optical Properties of

Substitutional Donors in ZnSe. Phys. Rev. B, 6(2):545–556, 1972.

Bibliography 259

[158] J.L. Merz, K. Nassau, and J.W. Shiever. Pair Spectra and the Shallow Acceptors

in ZnSe. Phys. Rev. B, 8(4):1444–1452, 1973.

[159] V. Swaminathan and L.C. Greene. Pair spectra, edge emission, and the shallow

acceptors in melt-grown ZnSe. Phys. Rev. B, 14(12):5351–5363, 1976.

[160] R.N. Bhargava, R.J. Seymour, B.J. Fitzpatrick, and S.P. Herko. Donor-acceptor

pair bands in ZnSe. Phys. Rev. B, 20(6):2407–2419, 1979.

[161] J.J. Hopfield and D.G. Thomas. Fine Structure and Magneto-Optic Effects in the

Exciton Spectrum of Cadmium Sulfide. Phys. Rev., 122(1):35–52, 1961.

[162] D.G. Thomas and J.J. Hopfield. Optical Properties of Bound Exciton Complexes

in Cadmium Sulfide. Phys. Rev., 128(5):2135–2148, 1962.

[163] P.J. Dean. Comparison of MOCVD-Grown with Conventional II-VI Materials Pa-

rameters for EL Thin Films. physica status solidi (a), 81(2):625–646, 1984.

[164] Katsuhiro Akimoto, Takao Miyajima, and Yoshifumi Mori. Photoluminescence

spectra of oxygen-doped ZnSe grown by molecular-beam epitaxy. Phys. Rev. B,

39(5):3138–3144, 1989.

[165] Akihiko Yoshikawa, Keiji Tanaka, Shigeki Yamaga, and Haruo Kasai. Growth of

High-Quality ZnSe Films by Low-Pressure Metalorganic Chemical Vapor Deposi-

tion. Jpn. J. Appl. Phys., 23(6):L424–L426, 1984.

[166] P.J. Dean, A.D. Pitt, P.J. Wright, M.L. Young, and B. Cockayne. The opto-

electronic properties of donors in organo-metallic grown zinc selenide. Physica B,

116(1-3):508–513, 1982.

[167] Takafumi Yao, Mutsuo Ogura, Seiichi Matsuoka, and Toshihide Morishita. Elec-

trical and Photo-luminescence Properties of ZnSe Thin Films Grown by Molecular

Bibliography 260

Beam Epitaxy: Substrate Temperature Effect. Jpn. J. Appl. Phys., 22(3):L144–

L146, 1983.

[168] J.L. de Miguel, S.M. Shibli, M.C. Tamargo, and B.J. Skromme. Planar doping

with gallium of molecular beam epitaxial ZnSe. Appl. Phys. Lett., 53(21):2065–

2067, 1988.

[169] B.J. Skromme, S.M. Shibli, J.L. de Miguel, and M.C. Tamargo. Photoluminescence

characterization of ZnSe doped with Ga by bulk and planar doping techniques in

molecular-beam epitaxy. J. Appl. Phys., 65(10):3999–4005, 1989.

[170] Atsushi Kamata, Tsutomu Uemoto, Keijiro Hirahara, and Tatsuro Beppu. The zinc

pressure effect in chlorine-doped ZnSe grown by atmospheric metalorganic chemical

vapor deposition. J. Appl. Phys., 65(6):2561–2563, 1989.

[171] W. Stutius. Conduction mechanism in low-resistivity n-type ZnSe prepared by

organometallic chemical vapor deposition. J. Appl. Phys., 53(1):284–291, 1982.

[172] Akihiko Yoshikawa, Hiroshi Nomura, Shigeki Yamaga, and Haruo Kasai. Controlled

conductivity in iodine-doped ZnSe films grown by metalorganic vapor-phase epi-

taxy. J. Appl. Phys., 65(3):1223–1229, 1989.

[173] C. Werkhoven, B.J. Fitzpatrick, S.P. Herko, R.N. Bhargava, and P.J. Dean. High-

purity ZnSe grown by liquid phase epitaxy. Appl. Phys. Lett., 38(7):540–542, 1981.

[174] Keizo Morimoto. Influence of growth temperature on stoichiometry in ZnSe. J.

Appl. Phys., 66(9):4206–4212, 1989.

[175] K. Kosai, B.J. Fitzpatrick, H.G. Grimmeiss, R.N. Bhargava, and G.F. Neumark.

Shallow acceptors and p-type ZnSe. Appl. Phys. Lett., 35(2):194–196, 1979.

[176] L Wischmeier, T Voss, I Ruckmann, and J Gutowski. Correlations between surface-

excitonic emission bands in ZnO nanowires. Nanotechnology, 19(13):135705, 2008.

Bibliography 261

[177] Oliver Brandt, Carsten Pfuller, Caroline Cheze, Lutz Geelhaar, and Henning

Riechert. Sub-meV linewidth of excitonic luminescence in single GaN nanowires:

Direct evidence for surface excitons. Phys. Rev. B, 81(4):045302, 2010.

[178] P. Rudolph and N. Schafer and T. Fukuda. Crystal growth of ZnSe from the melt.

Materials Science and Engineering R: Reports, 15(3):85–133, 1995.

[179] K. Mohammed, D.A. Cammack, R. Dalby, P. Newbury, B.L. Greenberg,

J. Petruzzello, and R.N. Bhargava. Effect of lattice mismatch in ZnSe epilayers

grown on GaAs by molecular beam epitaxy. Appl. Phys. Lett., 50(1):37–39, 1987.

[180] L. V. Keldysh. Excitons in Semiconductor-Dielectric Nanostructures. physica status

solidi (a), 164(1):3–12, 1997.

[181] E.A. Muljarov, E.A. Zhukov, V.S. Dneprovskii, and Yasuaki Masumoto. Dielec-

trically enhanced excitons in semiconductor-insulator quantum wires: Theory and

experiment. Phys. Rev. B, 62(11):7420–7432, 2000.

[182] H.E. Ruda and A. Shik. Polarization-sensitive optical phenomena in thick semi-

conducting nanowires. J. Appl. Phys., 100(2):024314, 2006.

[183] Yoichi Kawakami, Mitsuru Funato, Shizuo Fujita, Shigeo Fujita, Yoichi Yamada,

and Yasuaki Masumoto. Localized excitons in cubic Zn1−xCdxS lattice matched to

GaAs. Phys. Rev. B, 50(19):14655–14658, 1994.

[184] S.K. Mandal, A.R. Mandal, S. Das, and B. Bhattacharjee. Strong excitonic confine-

ment effect in ZnS and ZnS:Mn nanorods embedded in polycarbonate membrane

pores. J. Appl. Phys., 101(11):114315, 2007.

[185] M. Aven and H.H. Woodbury. Purification of II-VI Compounds By Solvent Ex-

traction. Appl. Phys. Lett., 1(3):53–54, 1962.

Bibliography 262

[186] A.C. Papadopoulo, A.M. Jean-Louis, and J. Charil. Properties of homogeneously

doped ZnSe single crystals obtained by a new growth method. Journal of Crystal

Growth, 44(5):587–592, 1978.

[187] G.N. Ivanova, D.D. Nedeoglo, A.V. Simashkevich, and K.D. Sushkevich. Photo-

luminescence of thermally treated zinc selenide crystals. Journal of Applied Spec-

troscopy, 30(3):318–321, 1979.

[188] G Z Zhong and F J Bryant. Blue emission of cathodoluminescence in zinc selenide

crystals. Journal of Physics C: Solid State Physics, 15(15):3411–3423, 1982.

[189] Phil Won Yu and Y.S. Park. p-type conduction in undoped ZnSe. Appl. Phys.

Lett., 22(7):345–346, 1973.

[190] T Ido, M Kato, A Yoshida, and M Ieda. Heat treatment effects on blue electrolu-

minescence in ZnSe. Journal of Physics D: Applied Physics, 11(1):L5–L7, 1978.

[191] H. Roppischer, J. Jacobs, and B.V. Novikov. The influence of Zn and Se Heat

Treatment on the Exciton Spectra of ZnSe Single Crystals. physica status solidi

(a), 27(1):123–127, 1975.

[192] B.V. Novikov, H. Roppischer, and J. Jacobs. Ionization of Donors by Excitons in

ZnSe Single Crystals at High Excitation Levels. physica status solidi (a), 31(1):K5–

K7, 1975.

[193] Maki Sekoguchi, Shunichi Hayamizu, Takashi Murase, Tsunemasa Taguchi, and

Akio Hiraki. Effect of heat treatment on photoluminescence and interface properties

in MOCVD-grown ZnSe onto GaAs and Ge. Journal of Luminescence, 40-41:838–

839, 1988.

[194] Yu.F. Vaksman, N.V. Malushin, V.M. Skobeeva, S. Agitera Morales, and V.V.

Serdyuk. The influence of annealing in liquid zinc on the photoluminescence spec-

Bibliography 263

trum of single crystals of ZnSe. Journal of Applied Spectroscopy, 21(2):1105–1107,

1974.

[195] Yu.F. Vaksman, N.V. Malushin, and V.V. Serdyuk. Investigations of the photo-

luminescence spectra of ZnSe monocrystals alloyed with aluminum. Journal of

Applied Spectroscopy, 25(5):1391–1394, 1976.

[196] Qihua Xiong, J. Wang, and P.C. Eklund. Coherent Twinning Phenomena: To-

wards Twinning Superlattices in III-V Semiconducting Nanowires. Nano Letters,

6(12):2736–2742, 2006.

[197] Dong-Hua Wang, Di Xu, Qing Wang, Ya-Juan Hao, Guo-Qiang Jin, Xiang-

Yun Guo, and K N Tu. Periodically twinned SiC nanowires. Nanotechnology,

19(21):215602, 2008.

[198] Hyun Woo Shim, Yongfeng Zhang, and Hanchen Huang. Twin formation during

SiC nanowire synthesis. J. Appl. Phys., 104(6):063511, 2008.

[199] Z.H. Zhang, F.F. Wang, and X.F. Duan. Formation mechanism of pseudoperiodical

multi-twinning nanostructures. Journal of Crystal Growth, 303(2):612–615, 2007.

[200] Ying-Teng Zhai and Xin-Gao Gong. Understanding periodically twinned structure

in nano-wires. Phys. Lett. A, 375(18):1889–1892, 2011.

[201] Jonas Johansson, Lisa S. Karlsson, C. Patrik T Svensson, Thomas Martensson,

Brent A. Wacaser, Knut Deppert, Lars Samuelson, and Weifert Seifert. Structural

properties of <111>B-oriented III-V nanowires. Nature Mater., 5(7):574–580, 2006.

[202] Hannah J. Joyce, Qiang Gao, H. Hoe Tan, Chennupati Jagadish, Yong Kim, Xin

Zhang, Yanan Guo, and Jin Zou. Twin-Free Uniform Epitaxial GaAs Nanowires

Grown by a Two-Temperature Process. Nano Letters, 7(4):921–926, 2007.

Bibliography 264

[203] S. Perera, M.A. Fickenscher, H.E. Jackson, L.M. Smith, J.M. Yarrison-Rice, H.J.

Joyce, Q. Gao, H.H. Tan, C. Jagadish, X. Zhang, and J. Zou. Nearly intrinsic exci-

ton lifetimes in single twin-free GaAs/AlGaAs core-shell nanowire heterostructures.

Appl. Phys. Lett., 93(5):053110, 2008.

[204] Robyn L. Woo, Rui Xiao, Yoji Kobayashi, Li Gao, Niti Goel, Mantu K. Hudait,

Thomas E. Mallouk, and R.F. Hicks. Effect of Twinning on the Photoluminescence

and Photoelectrochemical Properties of Indium Phosphide Nanowires Grown on

Silicon (111). Nano Letters, 8(12):4664–4669, 2008.

[205] Xin-Yu Bao, Cesare Soci, Darija Susac, Jon Bratvold, David P. R. Aplin, Wei

Wei, Ching-Yang Chen, Shadi A. Dayeh, Karen L. Kavanagh, and Deli Wang.

Heteroepitaxial Growth of Vertical GaAs Nanowires on Si (111) Substrates by

Metal-Organic Chemical Vapor Deposition. Nano Letters, 8(11):3755–3760, 2008.

[206] Toru Akiyama, Tomoki Yamashita, Kohji Nakamura, and Tomonori Ito. Band

Alignment Tuning in Twin-Plane Superlattices of Semiconductor Nanowires. Nano

Letters, 10(11):4614–4618, 2010.

[207] Dengfeng Li, Zhiguo Wang, and Fei Gao. First-principles study of the electronic

properties of wurtzite, zinc-blende, and twinned InP nanowires. Nanotechnology,

21(50):505709, 2010.

[208] T.S. Jeong, P.Y. Yu, Y.J. Shin, C.J. Youn, H.K. Shin, T.S. Kim, H. Lee, T.S. Lee,

and K.J. Hong. The characterization of ZnSe/GaAs epilayers grown by hot wall

epitaxy. Journal of Crystal Growth, 172:89–96, 1997.

[209] Baojuan Xi, Shenglin Xiong, Dechen Xu, Jingfa Li, Hongyang Zhou, Jun Pan,

Jiangying Li, and Yitai Qian. Tetraethylenepentamine-Directed Controllable Syn-

thesis of Wurtzite ZnSe Nanostructures with Tunable Morphology. Chemistry–A

European Journal, 14(31):9786–9791, 2008.

Bibliography 265

[210] L. Wischmeier, T. Voss, S. Borner, and W. Schade. Comparison of the optical

properties of as-grown ensembles and single ZnO nanowires. Applied Physics A:

Materials Science & Processing, 84(1-2):111–116, 2006.

[211] X.J. Chen, B. Gayral, D. Sam-Giao, C. Bougerol, C. Durand, and J. Eymery.

Catalyst-free growth of high-optical quality GaN nanowires by metal-organic vapor

phase epitaxy. Appl. Phys. Lett., 99(25):251910, 2011.

[212] J.E. Van Nostrand, K.L. Averett, R. Cortez, J. Boeckl, C.E. Stutz, N.A. Sanford,

A.V. Davydov, and J.D. Albrecht. Molecular beam epitaxial growth of high-quality

GaN nanocolumns. Journal of Crystal Growth, 287(2):500–503, 2006. The 16th

American Conference on Crystal Growth and Epitaxy - ACCGE 16 & The 12th

Biennial Workshop on OMVPE - OMVPE 12.

[213] Linus C. Chuang, Michael Moewe, Shanna Crankshaw, and Connie Chang-Hasnain.

Optical properties of InP nanowires on Si substrates with varied synthesis param-

eters. Appl. Phys. Lett., 92(1):013121, 2008.

[214] Florian Furtmayr, Martin Vielemeyer, Martin Stutzmann, Andreas Laufer,

Bruno K. Meyer, and Martin Eickhoff. Optical properties of Si- and Mg-doped

gallium nitride nanowires grown by plasma-assisted molecular beam epitaxy. J.

Appl. Phys., 104(7):074309, 2008.

[215] E. Calleja, M.A. Sanchez-Garcıa, F.J. Sanchez, F. Calle, F.B. Naranjo, E. Munoz,

U. Jahn, and K. Ploog. Luminescence properties and defects in GaN nanocolumns

grown by molecular beam epitaxy. Phys. Rev. B, 62(24):16826–16834, 2000.

[216] S. Reitzenstein, S. Munch, C. Hofmann, A. Forchel, S. Crankshaw, L.C. Chuang,

M. Moewe, and C. Chang-Hasnain. Time resolved microphotoluminescence stud-

ies of single InP nanowires grown by low pressure metal organic chemical vapor

deposition. Appl. Phys. Lett., 91(9):091103, 2007.

Bibliography 266

[217] Rui Chen, Dehui Li, Bo Liu, Zeping Peng, Gagik G. Gurzadyan, Qihua Xiong, and

Handong Sun. Optical and Excitonic Properties of Crystalline ZnS Nanowires:

Toward Efficient Ultraviolet Emission at Room Temperature. Nano Letters,

10(12):4956–4961, 2010.

[218] D.B. Williams and C.B. Carter. Transmission Electron Microscopy: A Textbook

for Materials Science. Plenum Press, 1996.

[219] Brent Fultz and James M. Howe. Transmission electron microscopy and diffrac-

tometry of materials. Advanced Texts in Physics. Springer, 2002.

[220] Anders Gustafsson, Mats-Erik Pistol, Lars Montelius, and Lars Samuelson. Lo-

cal probe techniques for luminescence studies of low-dimensional semiconductor

structures. J. Appl. Phys., 84(4):1715–1775, 1998.

[221] Jun-ichi Kasai and Yoshifumi Katayama. Low-temperature micro-

photoluminescence using confocal microscopy. Review of Scientific Instruments,

66(7):3738–3743, 1995.

[222] C.F. Klingshirn. Semiconductor optics. Advanced texts in physics. Springer, 2005.

[223] E. Dekel, D. Gershoni, E. Ehrenfreund, D. Spektor, J.M. Garcia, and P.M. Petroff.

Optical spectroscopy of a single self-assembled quantum dot. Physica E: Low-

dimensional Systems and Nanostructures, 2(1-4):694–700, 1998.

[224] D.W. Fradin and D.P. Bua. Laser-induced damage in ZnSe. Appl. Phys. Lett.,

24(11):555–557, 1974.

[225] Martin Heiss, Sonia Conesa-Boj, Jun Ren, Hsiang-Han Tseng, Adam Gali, Andreas

Rudolph, Emanuele Uccelli, Francesca Peiro, Joan Ramon Morante, Dieter Schuh,

Elisabeth Reiger, Efthimios Kaxiras, Jordi Arbiol, and Anna Fontcuberta i Morral.

Bibliography 267

Direct correlation of crystal structure and optical properties in wurtzite/zinc-blende

GaAs nanowire heterostructures. Phys. Rev. B, 83(4):045303, 2011.

[226] J Todorovic, A F Moses, T Karlberg, P Olk, D L Dheeraj, B O Fimland, H Weman,

and A T J van Helvoort. Correlated micro-photoluminescence and electron mi-

croscopy studies of the same individual heterostructured semiconductor nanowires.

Nanotechnology, 22(32):325707, 2011.

[227] H. Tews, H. Venghaus, and P.J. Dean. Excited states of shallow acceptors in ZnSe.

Phys. Rev. B, 19(10):5178–5184, 1979.

[228] Xi-Min Huang and Kenzo Igaki. Growth and exciton luminescence of ZnSe and

ZnSxSe1−x single crystals. Journal of Crystal Growth, 78(1):24–30, 1986.

[229] Y.H. Shing and D. Walsh. Band-edge photoluminescence of ZnSe using far-below

band-gap excitations. J. Appl. Phys., 51(3):1842–1843, 1980.

[230] M. Lei, X.L. Fu, P.G. Li, and W.H. Tang. Growth and photoluminescence of

zinc blende ZnS nanowires via metalorganic chemical vapor deposition. Journal of

Alloys and Compounds, 509(19):5769–5772, 2011.

[231] S. Ozaki, T. Tsuchiya, Y. Inokuchi, and S. Adachi. Photoluminescence and pho-

tomodulated transmittance spectroscopy of ZnO nanowires. physica status solidi

(a), 202(7):1325–1335, 2005.

[232] J. Szot, R. Hornsey, T. Ohnishi, and S. Minagawa. Focused ion beam microma-

chining for transmission electron microscopy specimen preparation of semiconduc-

tor laser diodes. Journal of Vacuum Science & Technology B: Microelectronics and

Nanometer Structures, 10(2):575–579, 1992.

[233] R. Hull, D. Bahnck, F.A. Stevie, L.A. Koszi, and S.N.G. Chu. Microscopic studies

of semiconductor lasers utilizing a combination of transmission electron microscopy,

Bibliography 268

electroluminescence imaging, and focused ion beam sputtering. Appl. Phys. Lett.,

62(26):3408–3410, 1993.

[234] C.W. Snyder, M.R. Frei, D. Bahnck, L. Hopkins, R. Hull, L. Harriott, T.Y. Chiu,

T. Fullowan, and B. Tseng. Combining transmission electron microscopy with

focused ion beam sputtering for microstructural investigations of AlGaAs/GaAs

heterojunction bipolar transistors. Journal of Vacuum Science & Technology B:

Microelectronics and Nanometer Structures, 13(4):1514–1518, 1995.

[235] T.J. Prosa, R.A. Alvis, L. Tsakalakos, and V.S. Smentkowski. Analysis of Silicon

Nanowires by Laser Atom Probe Tomography Prepared by a Protected Lift-Out

Processing Technique. Microscopy and Microanalysis, 14(Suppl. 2):456–457, 2008.

[236] J.A. Zapien, Y. Jiang, X.M. Meng, W. Chen, F.C.K. Au, Y. Lifshitz, and S.T. Lee.

Room-temperature single nanoribbon lasers. Appl. Phys. Lett., 84(7):1189–1191,

2004.

[237] Eliam Zacks and A. Halperin. Dependence of the Peak Energy of the Pair-

Photoluminescence Band on Excitation Intensity. Phys. Rev. B, 6(8):3072–3075,

1972.

[238] W.Y. Liang and A.D. Yoffe. Exciton Structure and Absorption edge in Hexagonal

Zinc Selenide. Proc. R. Soc. London Ser. A, 300(1462):326–336, 1967.

[239] Yi-Lu Chang, Zetian Mi, and Feng Li. Photoluminescence Properties of a Nearly

Intrinsic Single InN Nanowire. Advanced Functional Materials, 20(23):4146–4151,

2010.

[240] Nikolai V. Tkachenko. Optical Spectroscopy: Methods and Instrumentations. Else-

vier, 2006.

Bibliography 269

[241] G.R. Fowles. Introduction to Modern Optics. Dover Books on Physics. Dover

Publications, 1989.

[242] G. Sallen, A. Tribu, T. Aichele, R. Andre, L. Besombes, C. Bougerol, M. Richard,

S. Tatarenko, K. Kheng, and J.-Ph. Poizat. Subnanosecond spectral diffusion mea-

surement using photon correlation. Nature Photonics, 4(10):696–699, 2010.

[243] S. Datta, B.G. Yacobi, and D.B. Holt. Scanning electron microscope studies of local

variations in cathodoluminescence in striated ZnS platelets. Journal of Materials

Science, 12(12):2411–2420, 1977.

[244] G.D. Chen, M. Smith, J.Y. Lin, H.X. Jiang, M. Asif Khan, and C.J. Sun. Neutral-

donor-bound exciton recombination dynamics in GaN grown by metalorganic chem-

ical vapor deposition. Appl. Phys. Lett., 67(12):1653–1655, 1995.

[245] Xinlong Xu, Yanyuan Zhao, Edbert Jarvis Sie, Yunhao Lu, Bo Liu, Sandy Adhitia

Ekahana, Xiao Ju, Qike Jiang, Jianbo Wang, Handong Sun, Tze Chien Sum, Cheng

Hon Alfred Huan, Yuan Ping Feng, and Qihua Xiong. Dynamics of Bound Exciton

Complexes in CdS Nanobelts. ACS Nano, 5(5):3660–3669, 2011.

[246] C.J. Summers, W. Tong, T.K. Tran, W. Ogle, W. Park, and B.K. Wagner. Pho-

toluminescence properties of ZnS epilayers grown by metalorganic molecular beam

epitaxy. Journal of Crystal Growth, 159(1-4):64–67, 1996. Proceedings of the sev-

enth international conference on II-VI compounds and devices.

[247] A. Pawlis, M. Panfilova, D.J. As, K. Lischka, K. Sanaka, T.D. Ladd, and Y. Ya-

mamoto. Lasing of donor-bound excitons in ZnSe microdisks. Phys. Rev. B,

77(15):153304, 2008.

[248] Zengxing Zhang, Jianxiong Wang, Huajun Yuan, Yan Gao, Dongfang Liu, Li Song,

Yanjuan Xiang, Xiaowei Zhao, Lifeng Liu, Shudong Luo, Xinyuan Dou, Shicheng

Bibliography 270

Mou, Weiya Zhou, and Sishen Xie. Low-Temperature Growth and Photolumi-

nescence Property of ZnS Nanoribbons. The Journal of Physical Chemistry B,

109(39):18352–18355, 2005.

[249] K Kornitzer, M Grehl, K Thonke, R Sauer, C Kirchner, V Schwegler, M Kamp,

M Leszczynski, I Grzegory, and S Porowski. High-resolution PL spectra of donor-

and acceptor-bound excitons in homoepitaxial GaN-layers. Physica B: Condensed

Matter, 273-274(0):66–69, 1999.

[250] Giuliano Coli, K.K. Bajaj, J. Li, J.Y. Lin, and H.X. Jiang. Linewidths of excitonic

luminescence transitions in AlGaN alloys. Appl. Phys. Lett., 78(13):1829–1831,

2001.

[251] T. Schmidt, K. Lischka, and W. Zulehner. Excitation-power dependence of the

near-band-edge photoluminescence of semiconductors. Phys. Rev. B, 45(16):8989–

8994, 1992.

[252] G.E. Hite, D.T.F. Marple, M. Aven, and B. Segall. Excitons and the Absorption

Edge in ZnSe. Phys. Rev., 156(3):850–859, 1967.

[253] R.E. Halsted and M. Aven. Photoluminescence of Defect-Exciton Complexes in

II-VI Compounds. Phys. Rev. Lett., 14(3):64–65, 1965.

[254] F.E. Williams. Theory of the energy levels of donor-acceptor pairs. Journal of

Physics and Chemistry of Solids, 12(3-4):265–275, 1960.

[255] K.J. Hong, S.H. You, T.S. Jeong, C.J. Youn, M.S. Hong, J.S. Park, and C.S. Park.

Point defects of ZnSe epilayers grown by hot wall epitaxy. Journal of Crystal

Growth, 240(1-2):57–63, 2002.

Bibliography 271

[256] T.S. Jeong, P.Y. Yu, K.J. Hong, T.S. Kim, C.J. Youn, Y.D. Choi, K.S. Lee, B. O,

and M.Y. Yoon. Optical properties of the ZnSe/GaAs heteroepitaxial layers grown

by hot wall epitaxy. Journal of Crystal Growth, 249(1-2):9–14, 2003.

[257] Seishi Iida. Edge and Self-Activated Emissions in Zinc Selenide. Journal of the

Physical Society of Japan, 25(1):177–184, 1968.

[258] K. Ikeda, K. Uchida, and Y. Hamakawa. Observation of a donor-acceptor pair

recombination in the edge emission of ZnSe crystal by electro-luminescence. Journal

of Physics and Chemistry of Solids, 34(11):1985–1991, 1973.

[259] D.T.F. Marple. Electron Effective Mass in ZnSe. J. Appl. Phys., 35(6):1879–1882,

1964.

[260] G.N. Ivanova, D.D. Nedeoglo, N.D. Negeoglo, V.P. Sirkeli, I.M. Tiginyanu, and

V.V. Ursaki. Interaction of intrinsic defects with impurities in Al doped ZnSe

single crystals. J. Appl. Phys., 101(6):063543, 2007.

[261] K. Colbow. Free-to-Bound and Bound-to-Bound Transitions in CdS. Phys. Rev.,

141(2):742–749, 1966.

[262] Y.P. Varshni. Temperature dependence of the energy gap in semiconductors. Phys-

ica, 34(1):149–154, 1967.

[263] L. Malikova, Wojciech Krystek, Fred H. Pollak, N. Dai, A. Cavus, and M.C.

Tamargo. Temperature dependence of the direct gaps of ZnSe and Zn0.56Cd0.44Se.

Phys. Rev. B, 54:1819–1824, 1996.

[264] Keizo Morimoto. Origin of donor and acceptor species in undoped ZnSe grown by

low-pressure metalorganic chemical vapor deposition. J. Appl. Phys., 64(10):4951–

4956, 1988.

Bibliography 272

[265] Tsunemasa Taguchi, Takashi Kusao, and Akio Hiraki. Growth of high-purity ZnSe

by sublimation THM and the characteristics of the Y and Z deep-level emission

lines. Journal of Crystal Growth, 72(1-2):46–50, 1985.

[266] S Myhajlenko, J L Batstone, H J Hutchinson, and J W Steeds. Luminescence

studies of individual dislocations in II-VI (ZnSe) and III-V (InP) semiconductors.

Journal of Physics C: Solid State Physics, 17(35):6477–6492, 1984.

[267] K. Shahzad, J. Petruzzello, D.J. Olego, D.A. Cammack, and J.M. Gaines. Correla-

tion between radiative transitions and structural defects in zinc selenide epitaxial

layers. Appl. Phys. Lett., 57(23):2452–2454, 1990.

[268] Q Liu, H Lakner, C Mendorf, W Taudt, M Heuken, K Heime, and E Kubalek.

The influence of structural defects in ZnSe/GaAs heterostructures on luminescence

properties. Journal of Physics D: Applied Physics, 31(19):2421–2425, 1998.

[269] H.T. Lin, D.H. Rich, and D.B. Wittry. Cathodoluminescence study of domains,

defects, and interdiffusion in ZnSe/GaAs(100). J. Appl. Phys., 75(12):8080–8084,

1994.

[270] Noriyoshi Shibata, Akira Ohki, Sakae Zembutsu, and Akinori Katsui. Photolumi-

nescence Due to Lattice-Mismatch Defects in High-Purity ZnSe Layers Grown by

Metalorganic Vapor Phase Epitaxy. Jpn. J. Appl. Phys., 27(3):L441–L443, 1988.

[271] Shiro Satoh and Kenzo Igaki. Effect of Native Defect Concentration on Free-Exciton

Luminescence in Zinc Selenide. Jpn. J. Appl. Phys., 19(10):1953–1958, 1980.

[272] P.J. Dean. Recent developments in the optical spectroscopy of II - VI compound

semiconductors. Czechoslovak Journal of Physics, B 30:272–287, 1980.

Bibliography 273

[273] Takafumi Yao, Yunosuke Makita, and Shigeru Maekawa. Photoluminescence Prop-

erties of ZnSe Thin Films Grown by Molecular Beam Epitaxy. Jpn. J. Appl. Phys.,

20(10):L741–L744, 1981.

[274] M. Isshiki, T. Yoshida, T. Tomizono, S. Satoh, and K. Igaki. Preparation of high

purity zinc selenide single crystals and evaluation through photoluminescence spec-

tra. Journal of Crystal Growth, 73(2):221–225, 1985.

[275] M. Isshiki, T. Yoshida, K. Igaki, W. Uchida, and S. Suto. Photoluminescence

spectra of high purity zinc selenide single crystals. Journal of Crystal Growth,

72(1-2):162–166, 1985.

[276] H. Cheng, S.K. Mohapatra, J.E. Potts, and T.L. Smith. Effects of beam pressure

ratios on film quality in MBE growth of ZnSe. Journal of Crystal Growth, 81(1-

4):512–517, 1987.

[277] M Isshiki and K Masumoto. The change in the photoluminescence spectra of Cu-

doped ZnSe single crystals caused by heat treatment. Journal of Physics C: Solid

State Physics, 21(14):2771–2777, 1988.

[278] W. Shan, J.M. Hays, X.H. Yang, J.J. Song, E. Cantwell, and J. Aldridge. Bound

exciton luminescence in ZnSe under hydrostatic pressure. Appl. Phys. Lett.,

60(6):736–738, 1992.

[279] J M Hays, W Shan, X H Yang, J J Song, and E Cantwell. Optical investigations

of ZnSe crystals grown by the seeded vapour phase transport technique. Semicon-

ductor Science and Technology, 7(11):1407–1415, 1992.

[280] U.W. Pohl, G.H. Kudlek, A. Klimakow, and A. Hoffmann. Shallow impurity- and

defect-related complexes in undoped ZnSe crystals. Journal of Crystal Growth,

138(1-4):385–390, 1994.

Bibliography 274

[281] T. Fukuda, K. Umetsu, P. Rudolph, H.J. Koh, S. Iida, H. Uchiki, and N. Tsuboi.

Growth and characterization of twin-free ZnSe single crystals by the vertical Bridg-

man method. Journal of Crystal Growth, 161(1-4):45–50, 1996. Conference: Pu-

rification, Doping and Defects in II-VI Materials.

[282] Tsunemasa Taguchi and Takafumi Yao. Effect of Se vapor pressure and thermal

dissociation process on excitonic-emission lines in molecular beam epitaxially grown

high-purity zinc selenide. J. Appl. Phys., 56(10):3002–3005, 1984.

[283] Yasuhiro Shirakawa and Hiroshi Kukimoto. Near-band-edge photoluminescence in

ZnSe grown from indium solution. J. Appl. Phys., 51(4):2014–2019, 1980.

[284] Shiro Satoh and Kenzo Igaki. Bound Exciton Emission of Zinc Selenide. Jpn. J.

Appl. Phys., 20(10):1889–1895, 1981.

[285] R.N. Bhargava. The role of impurities in refined ZnSe and other II-VI semiconduc-

tors. Journal of Crystal Growth, 59(1-2):15–26, 1982.

[286] Shi-Min Huang, Yasuo Nozue, and Kenzo Igaki. Bound-Exciton Luminescence of

Cu-Doped ZnSe. Jpn. J. Appl. Phys., 22(Part 2, No. 7):L420–L422, 1983.

[287] Xue Jin Jiang, Takayuki Hisamune, Yasuo Nozue, and Takenari Goto. Optical

Properties of the Ideep1 Bound Exciton in ZnSe. Journal of the Physical Society of

Japan, 52(11):4008–4013, 1983.

[288] Edith D. Bourret, Franz X. Zach, Kin Man Yu, and James M. Walker. Growth

and characterization of ZnSe grown by organometallic vapor phase epitaxy using

diisopropyl selenide and diethyl zinc. Journal of Crystal Growth, 147(1-2):47–54,

1995.

[289] R. Triboulet, J.O. Ndap, A. Tromson-Carli, P. Lemasson, C. Morhain, and G. Neu.

Growth by solid phase recrystallization and assessment of large ZnSe crystals of

Bibliography 275

high purity and structural perfection. Journal of Crystal Growth, 159(1-4):156–160,

1996. Proceedings of the seventh international conference on II-VI compounds and

devices.

[290] Ching-Hua Su, S. Feth, Ling Jun Wang, and S.L. Lehoczky. Photoluminescence

studies of ZnSe starting materials and vapor grown bulk crystals. Journal of Crystal

Growth, 224(1-2):32–40, 2001.

[291] E. Tournie, C. Morhain, G. Neu, M. Laugt, C. Ongaretto, J.-P. Faurie, R. Triboulet,

and J.O. Ndap. Structural and optical characterization of ZnSe single crystals

grown by solid-phase recrystallization. J. Appl. Phys., 80(5):2983–2989, 1996.

[292] P.J. Dean, A.D. Pitt, M.S. Skolnick, P.J. Wright, and B. Cockayne. Optical prop-

erties of undoped organometallic grown ZnSe and ZnS. Journal of Crystal Growth,

59(1-2):301–306, 1982. Proceedings of the International Conference on II-VI Com-

pounds.

[293] B.J. Skromme, S.M. Shibli, J.L. de Miguel, and M.C. Tamargo. Photoluminescence

characterization of ZnSe doped with Ga by bulk and planar doping techniques in

molecular-beam epitaxy. J. Appl. Phys., 65(10):3999–4005, 1989.

[294] J.L. LaCombe and J.C. Irwin. The temperature dependence of the optical phonon

linewidths and frequencies in ZnSe and ZnTe. Solid State Communications,

8(18):1427–1431, 1970.

[295] J.J. Hopfield. A theory of edge-emission phenomena in CdS, ZnS and ZnO. Journal

of Physics and Chemistry of Solids, 10(2-3):110–119, 1959.

[296] S.Z. Wang and S.F. Yoon. Dynamic mechanisms in photoluminescence of gallium-

doped ZnSe grown by molecular-beam epitaxy. J. Appl. Phys., 93(9):5091–5094,

2003.

Bibliography 276

[297] Shanzhong Wang, Shengwu Xie, Qianjun Pang, Hang Zheng, Yuxing Xia, Rongbin

Ji, Yan Wu, Li He, Zuoming Zhu, Guohua Li, and Zhaoping Wang. Temperature

quenching mechanisms for photoluminescence of MBE-grown chlorine-doped ZnSe

epilayers. Journal of Crystal Growth, 220(4):548–553, 2000.

[298] K. Ohkawa and T. Mitsuyu. p-type ZnSe homoepitaxial layers grown by molecular

beam epitaxy with nitrogen radical doping. J. Appl. Phys., 70(1):439–442, 1991.

[299] P. Lautenschlager, M. Garriga, S. Logothetidis, and M. Cardona. Interband critical

points of GaAs and their temperature dependence. Phys. Rev. B, 35:9174–9189,

1987.

[300] S. Rudin, T.L. Reinecke, and B. Segall. Temperature-dependent exciton linewidths

in semiconductors. Phys. Rev. B, 42:11218–11231, 1990.

[301] Y H Hwang, Y H Um, and J K Furdyna. Temperature dependence of the band-edge

photoluminescence of Zn1−xMnxSe films. Semiconductor Science and Technology,

19(5):565–570, 2004.

[302] J.I. Pankove. Optical Processes in Semiconductors. Dover books on physics. Dover,

1971.

[303] J.D. Lambkin, L. Considine, S. Walsh, G.M. O’Connor, C.J. McDonagh, and T.J.

Glynn. Temperature dependence of the photoluminescence intensity of ordered and

disordered In0.48Ga0.52P. Appl. Phys. Lett., 65(1):73–75, 1994.

[304] H Zimmermann, R Boyn, and K Piel. Thermal quenching of bound exciton emission

due to phonon-induced non-radiative transitions: experimental data for CdTe and

InP. Journal of Physics: Condensed Matter, 4(3):859–868, 1992.

[305] W.Y. Liang and A.D. Yoffe. Luminescence in Hexagonal Zinc Selenide Crystals.

Philos. Mag., 16(144):1153–1166, 1967.

Bibliography 277

[306] A. Guinier (Chairman), G.B. Bokij, K. Boll-Dornberger, J.M. Cowley, S. Durovic,

H. Jagodzinski, P. Krishna, P.M. de Wolff, B.B. Zvyagin, D.E. Cox, P. Goodman,

Th. Hahn, K. Kuchitsu, and S.C. Abrahams. Nomenclature of Polytype Struc-

tures. Report of the International Union of Crystallography Ad-Hoc Committee

on the Nomenclature of Disordered, Modulated and Polytype Structures. Acta

Crystallographica Section A, 40(4):399–404, 1984.

[307] Ziqiang Zhu, Hiroshi Mori, and Takafumi Yao. Extremely low resistivity, high

electron concentration ZnSe grown by planar-doping method. Appl. Phys. Lett.,

61(23):2811–2813, 1992.

[308] Ziqiang Zhu, Takafumi Yao, and Hiroshi Mori. Selective Doping of N-type ZnSe

Layers with Chlorine Grown by Molecular Beam Epitaxy. J. Electron. Mater.,

22:463–466, 1993.

[309] K. Ohkawa, T. Mitsuyu, and O. Yamazaki. Characteristics of Cl-doped ZnSe layers

grown by molecular-beam epitaxy. J. Appl. Phys., 62(8):3216–3221, 1987.

[310] G.A. Baraff and M. Schluter. Migration of interstitials in silicon. Phys. Rev. B,

30:3460–3469, 1984.

[311] A.R. Verma and P. Krishna. Polymorphism and polytypism in crystals. Wiley

monographs in crystallography. Wiley, 1966.

[312] J.J. Hopfield, D.G. Thomas, and M. Gershenzon. Pair Spectra in GaP. Phys. Rev.

Lett., 10:162–164, 1963.

[313] D.G. Thomas, M. Gershenzon, and F.A. Trumbore. Pair Spectra and “Edge”

Emission in Gallium Phosphide. Phys. Rev., 133:A269–A279, 1964.

[314] K.P. O’Donnell and X. Chen. Temperature dependence of semiconductor band

gaps. Appl. Phys. Lett., 58(25):2924–2926, 1991.

Bibliography 278

[315] L. Vina, S. Logothetidis, and M. Cardona. Temperature dependence of the dielec-

tric function of germanium. Phys. Rev. B, 30:1979–1991, 1984.

[316] Umesh K. Mishra and Jasprit Singh. Electronic levels in semiconductors. In Semi-

conductor Device Physics and Design, pages 28–91. Springer Netherlands, 2008.

[317] Oleg Zakharov, Angel Rubio, X. Blase, Marvin L. Cohen, and Steven G. Louie.

Quasiparticle band structures of six II-VI compounds: ZnS, ZnSe, ZnTe, CdS,

CdSe, and CdTe. Phys. Rev. B, 50:10780–10787, 1994.

[318] The ABINIT code is a common project of the Universite Catholique de Louvain,

Corning Incorporated, and other contributors (URL http://www.abinit.org).

[319] N. Troullier and Jose Luıs Martins. Efficient pseudopotentials for plane-wave cal-

culations. Phys. Rev. B, 43:1993–2006, 1991.

[320] R.M. Martin. Electronic Structure: Basic Theory and Practical Methods. Cam-

bridge University Press, 2004.

Bibliography 279

Author Index

Abrahams, S.C. 157

Acharya, S. 5, 12, 21, 23, 187, 188

Acharya, Somobrata 5, 12, 187, 188

Acord, J.D. 12, 78

Adachi, S. 53, 78

Aichele, T. 72

Aichele, Thomas 5, 12, 19, 25, 36, 37, 40,

41, 47, 62, 77, 187, 191, 192

Akimoto, Katsuhiro 17

Akiyama, Toru 33, 158

Akopian, N. 6

Albrecht, J.D. 39, 115

Aldridge, J. 101, 104, 105

Algra, R.E. xvi, 8, 9, 31–33, 152, 157, 158,

182

Alvis, R.A. 54

Anderson, John 5, 12, 187, 188

Andre, R. 72

Andre, Regis 5, 19, 25, 36, 37, 40, 41, 47,

62, 77, 187, 191, 192

Aouba, S. xvi, 5, 187

Aplin, David P. R. 32

Arbiol, Jordi 52, 55

As, D.J. 78

Au, F.C.K. 55

Aven, M. 2, 3, 25, 87, 92, 93, 111, 124, 137

Averett, K.L. 39, 115

Bahnck, D. 54

Bajaj, K.K. 79

Bakkers, E.P.A.M. xvi, 8, 9, 31–33, 152,

157, 158, 182

Balasubramanian 4

Bando, Yoshio xvi, 5, 6, 12, 13, 53, 187,

188

Bandyopadhyay, Supriyo 5, 12, 187, 188

Bao, Jiming 9, 33, 52, 55, 61, 126, 157,

158, 162

Bao, Xin-Yu 32

Baraff, G.A. xxv, 162, 164

Barth, Sven 11

Batstone, J L 95

Beaux, Miles F. 4

Bell, David C. 9, 33, 52, 55, 61, 126, 157,

158, 162

Bellet-Amalric, Edith 5, 25, 36, 37, 40, 41,

187, 192

Bello, Igor 3, 5

Beppu, Tatsuro 18

Besombes, L. 72

Bhargava, R.N. 14, 18, 20, 101

Bhattacharjee, B. 21

Bhattacharya, Arnab 10

Bhattacharya, R. 5, 187, 188

Black, Marcie R. 4, 11, 19

Blase, X. xxvii, 217

Bo-Rui, Zhang 5

bo Wang, Jian 5, 12, 24, 187, 188

Bocquel, Juanita 5, 25, 36, 37, 40, 41, 187,

192

Boeckl, J. 39, 115

Bohach, Gregory A. 4

Bokij, G.B. 157

Bolinsson, J. xvi, 6, 8–10, 158

Bolinsson, Jessica 6

Bibliography 280

Boll-Dornberger, K. 157

Borgstrom, M.T. xvi, 8, 9, 31–33, 152,

157, 158, 182

Borner, S. 38, 53, 60, 61, 66, 113

Bougerol, C. 39, 66, 72, 115

Bougerol, Catherine 5, 12, 19, 25, 36, 37,

40, 41, 47, 62, 77, 187, 191, 192

Bourret, Edith D. 101, 104, 105

Boyn, R 125

Brandt, Oliver 19, 39, 60, 61, 66, 78

Bratvold, Jon 32

Brennan, K.F. 1

Bright, T. 4

Bryant, F J 26, 28, 105, 108, 125

Bua, D.P. 48

Bube, Richard H. 13

Buhro, W.E. 4

Burns, M. 1–3

Calle, F. 39, 115

Calleja, E. 39, 115

Cammack, D.A. 20, 95

Cantwell, E 101, 104, 105

Cao, Chuanbao 5, 12, 13, 187, 188

Cao, Yafeng 5, 12, 23, 187, 188

Capasso, Federico 4, 9, 33, 52, 55, 61, 126,

157, 158, 162

Cardona, M. 117, 206

Caroff, P. xvi, 6, 8–10, 158

Caroff, Philippe 6

Carter, C.B. 45, 69

Cavus, A. xiii, 95, 117, 118, 120, 122

Chalke, B.A. 10

Chan, S.K. 12

Chan, Y.F. 12

Chang-Hasnain, C. 39, 52, 60, 61, 66, 115

Chang-Hasnain, Connie 39, 66, 115

Chang, S.J. 5, 25, 36, 39, 41, 47, 77, 187,

193

Chang, S.P. 5, 25, 36, 39, 41, 47, 77, 187,

193

Chang, Yi-Lu 60, 61

Charil, J. 26, 28

Chen, Ching-Yang 32

Chen, G. 12, 78

Chen, G.D. 77, 78

Chen, Jia-fu 5, 12, 13, 187, 188

Chen, Minghai 5, 12, 187, 188

Chen, Q. 5, 38, 187, 188

Chen, Rui 44, 78

Chen, W. 55

Chen, X. 206

Chen, X.J. 39, 66, 115

Chen, Yang-Fang 5, 23, 187, 188

Cheng, Chung-Liang 5, 23, 187, 188

Cheng, H. 101, 104

Cheze, Caroline 19, 39, 60, 61, 66, 78

Chi, Cheng 5, 188

Chi, Xiaoqin 4

Chin, Patrick T.K. 5, 187, 188

Chiu, T.Y. 54

Choi, H.-J. 4, 5

Choi, Kwak J.-W. Park J.W., Y.-E. 4

Choi, Sukgeun 3

Choi, Y.D. 91, 101

Choy, Wallace C.H. 5, 12, 19, 24, 36, 37,

41, 47, 77, 187, 188, 191, 192

Chu, S.N.G. 54

Bibliography 281

Chu, Wang G.-Zhou-W. Lin Y.

Chernyak-L. Zhao J.-Kong J. Li L. Ren

J. Liu J., S. 4

Chuang, L.C. 39, 52, 60, 61, 66, 115

Chuang, Linus C. 39, 66, 115

Clark, T.E. 5, 19, 39

Cockayne, B. 17, 101

Cohen, Marvin L. xxvii, 217

Colbow, K. 94

Coli, Giuliano 79

Colli, A 5, 12, 23, 53, 187

Conesa-Boj, Sonia 52, 55

Considine, L. 123

Cooley, B.J. 5, 19, 39

Cortez, R. 39, 115

Cowley, J.M. 157

Cox, D.E. 157

Crankshaw, S. 39, 52, 60, 61, 66, 115

Crankshaw, Shanna 39, 66, 115

Crooker, S.A. 5, 19, 39

Dai, N. xiii, 95, 117, 118, 120, 122

Dalby, R. 20

Dang, Le Si 5, 25, 36, 37, 41, 62, 77, 187,

192

Das, S. 21

Das, T.K. 5, 187, 188

Datta, S. 75

Davydov, A.V. 39, 115

Dayeh, Shadi 3

Dayeh, Shadi A. 32

de Miguel, J.L. 18, 105

de Sousa, C.F. xvi, 5, 187

de Wolff, P.M. 157

Dean, P.J. 13, 17, 18, 53, 57, 87, 92–95, 97,

101

Dekel, E. 45

Demchenko, D.O. 4

Deppert, K. xvi, 6, 8, 9, 31–33, 152, 157,

158, 182

Deppert, Knut 32, 152

Deshmukh, Mandar M. 10

Devine, J.Z. 2, 3

Dhar, Abhishek 10

Dhar, N.K. 4

Dhara, Sajal 10

Dheeraj, D L 52, 55

Dick, K.A. xvi, 6, 8, 9, 31–33, 152, 157,

158, 182

Dick, Kimberly A 6

Dickey, E.C. 5, 19, 39

Ding, Jun-Xian 5, 12, 23, 38, 187, 188

Ding, Xianan 5, 19, 41, 77, 187, 188, 193

Dismukes, J.P. 2

Dneprovskii, V.S. 21

Donatini, Fabrice 5, 25, 36, 37, 41, 62, 77,

187, 192

Dong, Xue-hui 5, 12, 13, 187, 188

Dong-Yan, Xia 5

Dou, Xinyuan 78

Dresselhaus, Gene 4, 11, 19

Dresselhaus, Mildred S. 4, 11, 19

Du, Q.B. 9, 31, 33

Du, Yinxiao 5, 187, 188

Duan, X. 10

Duan, X.F. 5, 12, 31, 38, 187, 188

Ducati, C 5, 12, 23, 53, 187

Durand, C. 39, 66, 115

Bibliography 282

Durovic, S. 157

Ebert, Chad 4

Edwards, Jarrod D 5, 12, 187, 188

Edwards, P.L. 10

Efrima, S. 5, 12, 21, 23, 187, 188

Efrima, Shlomo 5, 12, 187, 188

Ehrenfreund, E. 45

Eichfeld, C.M. 5, 19, 39

Eickhoff, Martin 39, 115

Ekahana, Sandy Adhitia 78

Eklund, P.C. 12, 31, 78

Ellis, W.C. 11

Era, Koh 14

Eymery, J. 39, 66, 115

Fan, Xia 5, 23, 31, 32, 38, 187, 188

Fang, X. 5, 12, 23, 53, 187, 188

Fang, Xiaosheng xvi, 5, 6

Fardy, Melissa 4

Faurie, J.-P. 101, 105, 125

Feick, Henning 4, 5

Feiner, L.-F. xvi, 8, 9, 31–33, 152, 157,

158, 182

Feng, Yuan Ping 78

Ferrari, A C 5, 12, 23, 53, 187

Feth, S. 101, 105

Fickenscher, M.A. 32

Fimland, B O 52, 55

Fitzpatrick, B.J. 14, 18

Fontcuberta i Morral, Anna 52, 55

Forchel, A. 39, 52, 60, 61, 66, 115

Fortuna, Seth A 11, 12

Fowles, G.R. 63

Fradin, D.W. 48

Franciosi, A 5, 12, 23, 53, 187

Frei, M.R. 54

Froyen, S. 9

Fu, Hongzhi 5, 187, 188

Fu, X.L. 53

Fujita, Shigeo 21

Fujita, Shizuo 21

Fukuda, T. 101, 104

Fullowan, T. 54

Fultz, Brent 45, 69

Funato, Mitsuru 21

Furdyna, J K 3, 122

Furtmayr, Florian 39, 115

Gaines, J.M. 95

Gali, Adam 52, 55

Gao, Fei 33

Gao, G.B. 1–3

Gao, Jinhao 4

Gao, Li 32

Gao, Lian 5, 12, 187, 188

Gao, M. 5, 38, 187, 188

Gao, Q. 32

Gao, Qiang 6, 32, 158, 162

Gao, Y. 5, 12, 13, 23, 187, 188

Gao, Yan 78

Garcia, J.M. 45

Gargas, D. 4

Garriga, M. 117

Gass, Richard 6, 158, 162

Gates, B. 11

Gautam, Ujjal K. xvi, 5, 6

Gayral, B. 39, 66, 115

Geelhaar, Lutz 19, 39, 60, 61, 66, 78

Bibliography 283

Geng, B.Y. 9, 31, 33

Gershenzon, M. 200

Gershoni, D. 45

Gilchrist, K.H. 4

Glynn, T.J. 123

Goel, Niti 32

Gokhale, Mahesh 10

Golan, Y. 5, 12, 23, 187, 188

Golan, Yuval 5, 12, 187, 188

Golberg, Dmitri xvi, 5, 6, 12, 53, 187, 188

Goldberger, Joshua 11

Gong, X. 5, 7, 12, 19, 31–33, 36, 37, 41,

78, 152, 187–189

Gong, Xin-Gao 31

Goodman, P. 157

Gosain, D.P. 10

Gosele, Ulrich 3

Goto, Takenari xiii, 101, 105, 108, 109

Gradecak, Silvija 6

Greenberg, B.L. 20

Greene, L.C. 14, 92

Grego, S. 4

Grehl, M 78

Grimmeiss, H.G. 18

Grzegory, I 78

Guang-Zhao, Ran 5

Gui, Linlin 5

Guinier (Chairman), A. 157

Gunshor, R.L. 3

Guo-Gang, Qin 5

Guo, Guolin 5

Guo, Xiang-Yun 31

Guo, Yanan 32

Gurzadyan, Gagik G. 44, 78

Gustafsson, Anders 45

Gutierrez, H.R. 12, 78

Gutowski, J 19

Guzzi, Mario 7

Hahn, Th. 157

Halperin, A. 57, 87, 88, 92–94, 200, 204,

205

Halsted, R.E. 87, 111, 137

Hamakawa, Y. 92, 93

Han, Dongmei 5, 187, 188

Hao, Ya-Juan 31

Hao, Yufeng 7, 9, 31

Hark, S. 5, 7, 12, 19, 23, 31–33, 36, 37, 41,

53, 78, 152, 187–189

Hark, S K 5, 12, 19, 25, 36, 37, 39–41, 44,

77, 78, 187, 188, 190, 191

Hark, S.K. xvii, 5, 12, 13, 15, 19, 23–25,

31, 32, 36–38, 40, 44, 47, 48, 53, 77, 95,

114, 124, 187–190, 193

Harmand, J.-C. 6

Harriott, L. 54

Hartmann, H. 2

Hayamizu, Shunichi 27, 28

Hays, J M 101, 104, 105

Hays, J.M. 101, 104, 105

He, Li 117

He, R. 5

He, Wei 5

Heime, K 95

Heinz, P.D. 4

Heiss, Martin 52, 55

Herko, S.P. 14, 18

Hernandez-Ramirez, Francisco 11

Bibliography 284

Heuken, M 95

Hicks, R.F. 32

Hirahara, Keijiro 18

Hiraki, Akio 27, 28, 95

Hisamune, Takayuki xiii, 101, 105, 108,

109

Hite, G.E. 87, 92, 93, 124

Hoang, Thang Ba 5

Hoffmann, A. 101, 104, 105

Hofmann, C. 39, 52, 60, 61, 66, 115

Hofmann, S 5, 12, 23, 53, 187

Holmes, Justin D. 11

Holt, D.B. 75

Hong, K.J. 36, 91, 101, 104, 105, 108

Hong, M.S. 91, 101, 104, 105, 108

Hopfield, J.J. 15, 107, 200

Hopkins, L. 54

Hornsey, R. 54

Hou, Bo 5, 12, 187

Hovde, Carolyn J. 4

Howe, James M. 45, 69

Hsiao, C.H. 5, 25, 36, 39, 41, 47, 77, 187,

193

Hu, J. 12

Hu, Junqing 5, 12, 53, 187, 188

Hu, Z.D. 5, 38, 187, 188

Huan, Cheng Hon Alfred 78

Huang, B.R. 5, 25, 36, 39, 41, 47, 77, 187,

193

Huang, Dengtong 4

Huang, Hanchen 31

Huang, M. 5, 187, 188

Huang, Michael H. 4, 5

Huang, Ningping 5, 187, 188

Huang, Shi-Min 101, 103–105

Huang, Xi-Min 53

Hudait, Mantu K. 32

Hull, R. 54

Hung, S.C. 5, 25, 36, 41, 47, 77, 187, 193

Hutchinson, H J 95

Hwang, Y H 122

Ido, T 26, 28

Ieda, M 26, 28

Igaki, K. 101, 104, 105, 108

Igaki, Kenzo 53, 97, 101, 103–105, 108

Iida, S. 101, 104

Iida, Seishi 92–94

Ikeda, K. 92, 93

Ikonic, Z. 9, 10, 31

Immink, G. xvi, 8, 9, 31–33, 152, 157, 158,

182

Inkson, J.C. 9, 10, 31

Inokuchi, Y. 53, 78

Ip, K. 5, 7, 12, 19, 23, 31–33, 36, 37, 41,

53, 78, 152, 187–189

Ip, K M 5, 12, 187, 188

Ip, K.M. xvii, 5, 12, 13, 15, 19, 24, 25,

36–38, 40, 44, 47, 48, 53, 77, 95, 114,

124, 187–190

Irwin, J.C. 107

Ishizumi, Atsushi 5, 12, 19, 36, 40, 44, 53,

77, 122, 187, 189

Islam, M.S. 4

Isshiki, M 101, 103–105

Ito, Tomonori 33, 158

Ivanova, G.N. 26, 28, 94

Bibliography 285

Jackson, H.E. 5, 32

Jackson, Howard E. 6, 158, 162

Jacobs, J. 27, 89–92, 96, 97, 101, 103–105,

135

Jagadish, C. 32

Jagadish, Chennupati 6, 32, 158, 162

Jagodzinski, H. 157

Jahn, U. 39, 115

Janssen, Rene A.J. 5, 187, 188

Jean-Louis, A.M. 26, 28

Jeong, T.S. 36, 91, 101, 104, 105, 108

Ji, Rongbin 117

Jia, Mingfen 5, 12, 187

Jiang, Fengzhi 5, 12, 187

Jiang, H.X. 77–79

Jiang, Qike 78

Jiang, Xuchuan 5

Jiang, Xue Jin xiii, 101, 105, 108, 109

Jiang, Y. 55

Jiang, Yang 5, 12, 19, 23, 38, 41, 77, 187,

188, 193

Jie, Jian Sheng 5, 23, 31, 32, 38, 187, 188

Jie, Jiansheng 3, 5

Jie, Wanqi 5, 187, 188

Jin, Guo-Qiang 31

Jin, L. 5, 19, 36, 37, 41, 47, 77, 187, 191

Jin, Lei 5, 12, 24, 187, 188

Jing, Yi 4

Johansson, J. xvi, 6, 8–10, 158

Johansson, Jonas 32, 152

Johnson, J. 5

Johnson, J.C. 4

Joyce, Hannah J. 32

Joyce, H.J. 32

Jr., R.J. Happel 10

Ju, Xiao 78

Jun-ichi Kasai and Yoshifumi Katayama

45

Jun, J. 5, 12, 53, 187, 188

Kamata, Atsushi 18

Kamp, M 78

Kanatzidis, Mercouri G. 4

Kanemitsu, Yoshihiko 5, 12, 19, 36, 40, 44,

53, 77, 122, 187, 189

Kannan 4

Kargar, A. 4

Karlberg, T 52, 55

Karlsson, Lisa S. 32, 152

Kasai, Haruo 17, 18

Kato, M 26, 28

Katsui, Akinori 95

Katzenmeyer, A.M. 4

Kavanagh, K L 5, 187, 188

Kavanagh, Karen L. 32

Kavanagh, K.L. xvii, 5, 9, 12, 23, 26–28,

31, 32, 35, 37, 38, 48, 50, 51, 53, 57, 94,

96, 97, 152, 187, 188

Kawakami, Yoichi 21

Kaxiras, Efthimios 52, 55

Keldysh, L. V. 21

Kempa, Thomas J. 4

Khan, M. Asif 77, 78

Kheng, K. 72

Kheng, Kuntheak 5, 12, 19, 25, 36, 37, 40,

41, 47, 62, 77, 187, 191, 192

Kim, F. 11

Kim, T.S. 36, 91, 101

Bibliography 286

Kim, W 4

Kim, Yong 32

Kind, Hannes 4, 5

Kirchner, C 78

Klimakow, A. 101, 104, 105

Klingshirn, C.F. 45

Knutsen, K.P. 4

Ko, C.H. 5, 36, 39, 77, 187, 193

Kobayashi, N.P. 4

Kobayashi, Yoji 32

Koh, H.J. 101, 104

Koide, Yasuo xvi, 5

Kolodziejski, L.A. 3

Kong, Jing 4, 11, 19

Kornitzer, K 78

Kosai, K. 18

Koszi, L.A. 54

Krishna, P. 157, 173

Krystek, Wojciech xiii, 95, 117, 118, 120,

122

Kuan, T.M. 5, 36, 39, 77, 187, 193

Kubalek, E 95

Kuchitsu, K. 157

Kudlek, G. xv, xvii, 2, 3, 14–18, 20, 25, 26,

94, 154, 159, 161, 184–186

Kudlek, G.H. 101, 104, 105

Kukimoto, H. 14

Kukimoto, Hiroshi 101, 117, 125

Kusao, Takashi 95

Kwon, Nam Hoon 4

LaCombe, J.L. 107

Ladd, T.D. 78

Lakner, H 95

Lambkin, J.D. 123

Langer, Dietrich W. 14

Larson, O.W. 13

Lassen, B. 12, 78

Lassiter, Brian E. 4

Laufer, Andreas 39, 115

Laugt, M. 101, 105, 125

Lauhon, L.J. 5

Lautenschlager, P. 117

Law, Matt 11

Lee, B. 4

Lee, C. 5, 12, 53, 187, 188

Lee, Chun Sing 5, 23, 31, 32, 38, 187, 188

Lee, H. 36, 91

Lee, K.S. 91, 101

Lee, Shuit Tong 5, 23, 31, 32, 38, 187, 188

Lee, S.T. 55

Lee, T.S. 36, 91

Lehoczky, S.L. 101, 105

Lei, M. xvii, 5, 24, 53, 57, 187, 188

Lemasson, P. 101, 105

Lensch, J.L. 5

Leszczynski, M 78

Leung, Y. 5, 12, 23, 53, 187, 188

Leung, Yee P. 5, 12, 24, 187, 188

Leung, Y.P. 5, 12, 13, 19, 23, 25, 36–38, 40,

41, 47, 53, 77, 95, 114, 187–189, 191, 192

Lew Yan Voon, L.C. 12, 78

Li, Dehui 44, 78

Li, Dengfeng 33

Li, Feng 60, 61

Li, G.H. 5, 19, 24, 36, 47, 77, 187–189

Li, Guohua 117

Li, Huanyong 5, 187, 188

Bibliography 287

Li, J. 4, 79

Li, Jiangying 38, 187, 188

Li, Jingfa 38, 187, 188

Li, Junwei 5, 19, 41, 77, 187, 188, 193

Li, L.H. xvii, 5, 24, 57, 187, 188

Li, Liang 5

Li, P.G. xvii, 5, 24, 53, 57, 187, 188

Li-Ping, You 5

Li, Q. 5, 7, 12, 19, 23, 31–33, 36, 37, 41,

53, 78, 152, 187–189

Li, Quan xvii, 5, 12, 13, 15, 19, 23–25,

36–38, 40, 44, 47, 48, 53, 77, 95, 114,

124, 187–190

Li, Shanying 5, 19, 41, 77, 187, 188, 193

Li, T.C. 5, 25, 36, 39, 41, 47, 77, 187, 193

Li, Xiaoxiao 3

Li, Xiaoyu 5, 187, 188

Li, Xiuling 11, 12

Li, Yanjuan 5, 12, 187

Liang, W.Y. xiv, xxvi, 57, 131, 134–139,

148, 149, 161, 166, 168, 169, 177, 178

Liang, Yao 5, 12, 19, 24, 31, 32, 77, 187,

193

Liao, Meiyong xvi, 5

Lieber, Charles M. 3, 4

Lieber, C.M. 4, 10, 12

Lifshitz, Y. 55

Lin, H.T. 95

Lin, J.Y. 77–79

Lin, M.E. 1–3

Lin, W.J. 5, 25, 36, 39, 41, 47, 77, 187, 193

Lin, Yu-Ming 4, 11, 19

Lischka, K. xxvii, 78, 85, 87, 196, 203, 204

Liu, Bo 44, 78

Liu, B.Z. 5, 19, 39

Liu, Dongfang 78

Liu, Ji 5, 12, 23, 38, 187, 188

Liu, Jun 5, 187, 188

Liu, L. 6

Liu, Lifeng 78

Liu, Q 95

Liu, Ruibin 5, 187, 188

Liu, Siyuan 5, 12, 13, 187, 188

Liu, S.Y. 5, 19, 36, 37, 41, 47, 77, 187, 191

Liu, X. 12, 78

Liu, X.W. 9, 31, 33

Liu, Yongjun 5, 12, 187

Liu, Z 5, 12, 19, 25, 36, 37, 39–41, 44, 77,

78, 187, 188, 190, 191

Logothetidis, S. 117, 206

Loomis, R.A. 4

Louie, Steven G. xxvii, 217

Lu, G.W. 5, 19, 24, 36, 47, 77, 187–189

Lu, Jun 5

Lu, W. 4

Lu, Wei 4

Lu, Yunhao 78

Lu, Z.W. 9

Lun, Dai 5

Luo, H 3

Luo, Shudong 78

Lv, Ruitao 5, 12, 13, 187, 188

Ma, Xichen 5, 12, 23, 187, 188

Madsen, K.N. 4

Maekawa, Shigeru 101, 105

Majumdar, Arun 4

Makita, Yunosuke 101, 105

Bibliography 288

Malikova, L. xiii, 95, 117, 118, 120, 122

Mallouk, Thomas E. 32

Malushin, N.V. 28

Manandhar, Pradeep 3

Mandal, A.R. 21

Mandal, S.K. 21

Mandl, Bernhard 6

Manna, A. 5, 187, 188

Mao, S. 5

Mao, Samuel 4, 5

Mariette, Henri 5, 25, 36, 37, 40, 41, 187,

192

Markov, I. 5, 12, 187, 188

Marple, D.T.F. 87, 92, 93, 124

Martelli, F 5, 12, 23, 53, 187

Martin, R.M. 218

Martins, Jose Luıs 218

Masumoto, K 101, 103–105

Masumoto, Yasuaki 21

Mathur, S. 3

Matsuoka, Seiichi 17

Mayers, B. 11

McDonagh, C.J. 123

McIlroy, David N. 4

Mendorf, C 95

Meng, Guowen 7, 9, 31

Meng, Xiang Min 5, 23, 31, 32, 38, 187,

188

Meng, X.M. 55

Merz, J.L. 13, 14, 57, 87, 92–94

Messing, M.E. xvi, 6, 8, 9, 31–33, 152, 157,

158, 182

Meyer, Bruno K. 39, 115

Mi, Zetian 60, 61

Minagawa, S. 54

Mishra, Umesh K. xxvii, 216

Mitsuyu, T. 117, 161

Miyajima, Takao 17

Moewe, M. 39, 52, 60, 61, 66, 115

Moewe, Michael 39, 66, 115

Mohammed, K. 20

Mohapatra, S.K. 101, 104

Mohney, S.E. 5, 19, 39

Monroy, E 2, 3

Montazeri, Mohammad 6, 158, 162

Montelius, Lars 45

Morales, S. Agitera 28

Morante, Joan Ramon 52, 55

Morhain, C. 101, 105, 125

Mori, Hiroshi 161

Mori, Yoshifumi 17

Morimoto, Keizo 18, 26, 92, 95

Morishita, Toshihide 17

Morkoc, H. 1–3

Morris, N. 5

Moses, A F 52, 55

Mou, Shicheng 78

Martensson, Thomas 9, 32, 33, 52, 55, 61,

126, 152, 157, 158, 162

Munoz, E. 39, 115

Muljarov, E.A. 21

Muller, Sven 4

Munch, S. 39, 52, 60, 61, 66, 115

Murase, Takashi 27, 28

Myhajlenko, S 95

Nakamura, Kohji 33, 158

Naranjo, F.B. 39, 115

Bibliography 289

Nassau, K. 14, 93

Naughton, P.W. 4

Nayak, A.P. 4

Ndap, J.O. 101, 105, 125

Nedeoglo, D.D. 26, 28, 94

Negeoglo, N.D. 94

Neu, G. 101, 105, 125

Neumark, G.F. 1, 18

Newbury, P. 20

Ng, C M 5, 12, 187, 188

Nichols, P.L. 6

Ning, C.Z. 4, 6

Nomura, Hiroshi 18

Nostrand, J.E. Van 39, 115

Novikov, B.V. 27, 89–92, 96, 97, 101,

103–105, 135

Nozue, Yasuo xiii, 101, 103–105, 108, 109

Nurmikko, A.V. 3

O, B. 91, 101

O’Connor, G.M. 123

Odom, T.W. 12

O’Donnell, K.P. 206

Ogle, W. 78

Ogura, Mutsuo 17

Oh, J. 4

Ohkawa, K. 117, 161

Ohki, Akira 95

Ohnishi, T. 54

Ohno, Yutaka 5, 12, 19, 36, 40, 44, 53, 77,

122, 187, 189

Olego, D.J. 95

Olk, P 52, 55

Omnes, F 2, 3

Ong, H.C. 5, 12, 187, 188

Ong, Hock C. 5, 12, 24, 187, 188

Ongaretto, C. 101, 105, 125

Ozaki, S. 53, 78

P. Rudolph and N. Schafer and T. Fukuda

20, 35

Paiman, Suriati 6, 158, 162

Pan, A. 6

Pan, Anlian 5, 6, 187, 188

Pan, J. 3

Pan, Jun 38, 187, 188

Panda, A.B. 5, 12, 21, 23, 187, 188

Panda, Asit Baran 5, 12, 187, 188

Panfilova, M. 78

Pang, G.K.H. 5, 12, 187, 188

Pang, Qianjun 117

Pankove, J.I. 123

Papadopoulo, A.C. 26, 28

Park, C.S. 91, 101, 104, 105, 108

Park, J.S. 91, 101, 104, 105, 108

Park, Namsoek 4

Park, S. 5, 12, 53, 187, 188

Park, W. 78

Park, Yong Ho 4

Park, Y.S. 2, 3, 26

Patibandla, Sridhar 5, 12, 187, 188

Patriarche, G. 6

Pavesi, Lorenzo 7

Pawlis, A. 78

Pedrotti, L.S. 13

Peiro, Francesca 52, 55

Pemasiri, Kuranananda 6, 158, 162

Peng, Feng 5, 188

Bibliography 290

Peng, L.-M. 5, 38, 187, 188

Peng, Zeping 44, 78

Perea, Daniel 3

Perera, S. 32

Petroff, P.M. 45

Petruzzello, J. 20, 95

Pfuller, Carsten 19, 39, 60, 61, 66, 78

Pham, J. 5

Philipose, U. xvi, xvii, 5, 9, 12, 19, 23,

26–28, 31, 32, 35, 37, 38, 48, 50, 51, 53,

57, 83, 94, 96, 97, 152, 187, 188, 192

Picraux, S. 3

Piel, K 125

Pistol, Mats-Erik 45

Pitt, A.D. 17, 101

Ploog, K. 39, 115

Pohl, U.W. 101, 104, 105

Poizat, J.-Ph. 72

Poizat, Jean-Philippe 5, 12, 25, 36, 37, 40,

41, 187, 191, 192

Pollak, Fred H. xiii, 95, 117, 118, 120, 122

Porowski, S 78

Potts, J.E. 101, 104

Presser, N. xv, xvii, 2, 3, 14–18, 20, 25, 26,

94, 154, 159, 161, 184–186

Prosa, T.J. 54

Qian, Yitai xvi, 5, 38, 187, 188

Qian, Y.T. 5, 12, 13, 23, 187, 188

Qiao, Zhengping 5, 188

R., Arvind Pawan 10

Rabin, Oded 4, 11, 19

Ramanathan, Sivakumar 5, 12, 187, 188

Ramgir, Niranjan S. 4

Reddy, B K 2

Reddy, D R 2

Redwing, J.M. 12, 78

Reiger, Elisabeth 52, 55

Reinecke, T.L. 119

Reitzenstein, S. 39, 52, 60, 61, 66, 115

Ren, Jun 52, 55

Reynolds, D.C. 13

Rich, D.H. 95

Richard, M. 72

Riechert, Henning 19, 39, 60, 61, 66, 78

Robertson, J 5, 12, 23, 53, 187

Romano-Rodriguez, Albert 11

Ronning, Carsten 4

Roppischer, H. 27, 89–92, 96, 97, 101,

103–105, 135

Rubini, S 5, 12, 23, 53, 187

Rubio, Angel xxvii, 217

Ruckmann, I 19

Ruda, H. 10

Ruda, Harry E. xvii, 5, 12, 19, 23, 26–28,

32, 35, 38, 48, 50, 51, 53, 57, 83, 94, 96,

97, 187, 188, 192

Ruda, H.E. xvi, 2, 3, 5, 9, 21, 23, 31, 32,

37, 51, 96, 97, 152, 187, 188

Rudin, S. 119

Rudolph, Andreas 52, 55

Rudolph, P. 101, 104

Russo, R. 5

Russo, Richard 4, 5

Saha, S. 5, 187, 188

Salfi, J. xvi, 5, 187

Sallen, G. 72

Bibliography 291

Sallen, Gregory 5, 12, 25, 36, 37, 40, 41,

187, 191, 192

Sam-Giao, D. 39, 66, 115

Samarth, N. 5, 19, 39

Samuelson, L. xvi, 6, 8, 9, 31–33, 152, 157,

158, 182

Samuelson, Lars 9, 32, 33, 45, 52, 55, 61,

126, 152, 157, 158, 162

Sanaka, K. 78

Sanchez, F.J. 39, 115

Sanchez-Garcıa, M.A. 39, 115

Sanford, N.A. 39, 115

Sansoz, Frederic 9, 10

Satoh, S. 101, 104, 105, 108

Satoh, Shiro 97, 101, 104, 105, 108

Sauer, R 78

Saxena, Ankur 5, 19, 53, 83, 187, 188, 192

Saykally, R. 5

Saykally, R.J. 4

Schade, W. 38, 53, 60, 61, 66, 113

Schaller, R.D. 4

Schluter, M. xxv, 162, 164

Schmidt, T. xxvii, 85, 87, 196, 203, 204

Schmidt, Volker 3

Schuh, Dieter 52, 55

Schwegler, V 78

Segall, B. 87, 92, 93, 119, 124

Seifert, Weifert 32, 152

Sekoguchi, Maki 27, 28

Sengupta, Shamashis 10

Senz, Stephan 3

Serdyuk, V.V. 28

Seymour, R.J. 14

Shahzad, K. 95

Shakouri, Ali 4

Shan, C X 5, 19, 25, 36, 37, 39, 41, 78,

187, 190

Shan, W 101, 104, 105

Shen, H. 3

Shen, J. 5, 12, 13, 23, 187, 188

Shi, L 12

Shi, Liang 5, 12, 23, 38, 187, 188

Shi, Xuesong 5

Shibata, Noriyoshi 95

Shibli, S.M. 18, 105

Shiever, J.W. 14, 93

Shik, A. 21

Shim, Hyun Woo 31

Shin, B.K. 2, 3

Shin, H.K. 36, 91

Shin, Y.J. 36, 91

Shing, Y.H. 53

Shirahama, Takeo 5, 12, 19, 36, 40, 44, 53,

77, 122, 187, 189

Shirakawa, Yasuhiro 101, 117, 125

Sie, Edbert Jarvis 78

Simashkevich, A.V. 26, 28

Simpson, P J 5, 187, 188

Singh, Jasprit xxvii, 216

Singh, N. 2, 3, 15, 22

Singh, Vibhor 10

Sirkeli, V.P. 94

Skobeeva, V.M. 28

Skolnick, M.S. 101

Skromme, B.J. 18, 105

Smentkowski, V.S. 54

Smith, Leigh M. 6, 158, 162

Bibliography 292

Smith, L.M. 5, 32

Smith, M. 77, 78

Smith, T.L. 101, 104

Snyder, C.W. 54

Soci, Cesare 32

Soh, A.K. 5, 19, 36, 37, 41, 47, 77, 187, 191

Solanki, Hari S. 10

Song, Chunfeng 5, 187, 188

Song, Hui 5, 12, 23, 187, 188

Song, J J 101, 104, 105

Song, J.J. 101, 104, 105

Song, Li 78

Sou, I.K. 12

Spektor, D. 45

Srivastava, G.P. 9, 10, 31

Steeds, J W 95

Stevie, F.A. 54

Stofko, E.J. 2

Stouwdam, Jan W. 5, 187, 188

Strite, S. 1–3

Stutius, W. 18

Stutz, C.E. 39, 115

Stutzmann, Martin 39, 115

Su, Ching-Hua 101, 105

Su, F.H. 5, 19, 24, 36, 47, 77, 187–189

Sum, Tze Chien 78

Summers, C.J. 78

Sun, C.J. 77, 78

Sun, Handong 44, 78

Sun, Ke 4

Sun, Ping xvii, 5, 12, 23, 26–28, 32, 35, 38,

48, 50, 53, 57, 94, 96, 187, 188

Sun, X.C. 5, 19, 24, 36, 47, 77, 187–189

Sun, Y. 11

Susac, Darija 32

Sushkevich, K.D. 26, 28

Suto, S. 101, 104, 105, 108

Svensson, C. Patrik T 32, 152

Sverdlov, B. 1–3

Swaminathan, V. 14, 92

Szot, J. 54

Taguchi, Tsunemasa 27, 28, 95, 101, 105

Takeda, Seiji 5, 12, 19, 36, 40, 44, 53, 77,

122, 187, 189

Talin, A.A. 4

Tamargo, M.C. xiii, 18, 95, 105, 117, 118,

120, 122

Tan, H. Hoe 6, 32, 158, 162

Tan, H.H. 32

Tanaka, Keiji 17

Tang, Q. 5, 12, 13, 23, 187, 188

Tang, W.H. xvii, 5, 24, 53, 57, 187, 188

Tang, Youqi 5

Tao, Yin 5, 12, 19, 24, 31, 32, 77, 187, 193

Tatarenko, S. 72

Tatarenko, Serge 5, 12, 19, 25, 36, 37, 40,

41, 47, 62, 77, 187, 191, 192

Taudt, W 95

Tews, H. 53

Thomas, D.G. 15, 200

Thonke, K 78

Tian, Bozhi 4

Tian, Tian 5, 187, 188

Tiginyanu, I.M. 94

Titova, L.V. 5

Tkachenko, Nikolai V. 61

Todorovic, J 52, 55

Bibliography 293

Tomiya, S. 10

Tomizono, T. 101, 104, 105, 108

Tong, W. 78

Tournie, E. 101, 105, 125

Tragardh, Johanna 9, 33, 52, 55, 61, 126,

157, 158, 162

Tran, T.K. 78

Triboulet, R. 101, 105, 125

Tribu, A. 72

Tribu, Adrien 5, 12, 19, 25, 36, 37, 40, 41,

47, 62, 77, 187, 191, 192

Tromson-Carli, A. 101, 105

Troullier, N. 218

Trumbore, F.A. 200

Tsakalakos, L. 54

Tseng, B. 54

Tseng, Hsiang-Han 52, 55

Tsuboi, N. 101, 104

Tsuchiya, T. 53, 78

Tu, K N 31

Uccelli, Emanuele 52, 55

Uchida, K. 92, 93

Uchida, W. 101, 104, 105, 108

Uchiki, H. 101, 104

Uemoto, Tsutomu 18

Ulmer, R.J. 2

Um, Y H 122

Umetsu, K. 101, 104

Ursaki, V.V. 94

Usui, S. 10

Vaksman, Yu.F. 28

Van Enckevort, W.J.P. xvi, 8, 9, 31–33,

152, 157, 158, 182

van Helvoort, A T J 52, 55

van Vugt, Lambert K. 4

Vanmaekelbergh, Daniel 4

Varshni, Y.P. 95, 115, 205, 206

Venghaus, H. 53

Venugopal, R 2

Verheijen, M.A. xvi, 8, 9, 31–33, 152, 157,

158, 182

Verma, A.R. 173

Vina, L. 206

Vielemeyer, Martin 39, 115

Vij, D.R. 2, 3, 15, 22

Vijayalakshmi, R P 2

Vineis, Christopher J. 4

VJ, L. 4

Vlieg, E. xvi, 8, 9, 31–33, 152, 157, 158,

182

Voss, T. 38, 53, 60, 61, 66, 113

Wacaser, Brent A. 32, 152

Wagner, B.K. 78

Wagner, Jakob B. 9, 33, 52, 55, 61, 126,

157, 158, 162

Wagner, R.S. 11

Walker, James M. 101, 104, 105

Walsh, D. 53

Walsh, S. 123

Wan-Jin, Xu 5

Wang, Binbin 5, 19, 41, 77, 187, 188, 193

Wang, C. 5, 7, 12, 19, 31–33, 36, 37, 41,

78, 152, 187–189

Wang, Dazhi 5, 12, 13, 23, 187, 188

Wang, Deli 4, 32

Wang, Dong-Hua 31

Bibliography 294

Wang, Feifei 5, 187, 188

Wang, F.F. 31

Wang, Haitao 5, 187, 188

Wang, Hongjuan 5, 188

Wang, J. 5, 7, 12, 19, 31–33, 36, 37, 41, 78,

152, 187–189

Wang, Jianbo 5, 19, 36, 37, 41, 47, 77, 78,

187, 191

Wang, Jianxiong 78

Wang, Junli 5, 31, 32, 187, 188

Wang, L.-W. 4

Wang, Li 3

Wang, Lidong 4

Wang, Ling Jun 101, 105

Wang, N. 12

Wang, Qing 31

Wang, R.M. 5, 19, 24, 36, 47, 77, 187–189

Wang, S.-Y. 4

Wang, S.B. 5, 25, 36, 39, 41, 47, 77, 187,

193

Wang, Shanzhong 117

Wang, S.L. xvii, 5, 24, 57, 187, 188

Wang, S.Z. 117

Wang, Xiao 5, 187, 188

Wang, Y. 5, 12, 23, 53, 187, 188

Wang, Y Q 5, 187, 188

Wang, Y.Q. xvii, 5, 9, 12, 23, 26–28, 31,

32, 35, 37, 48, 50, 51, 53, 57, 94, 96, 97,

152, 187, 188

Wang, Zhaoping 117

Wang, Zhiguo 33

Wang, Zhong Lin 5, 7, 9, 31

Weber, Eicke 4, 5

Wei, Wei 32

Wei, X.W. 9, 31, 33

Weman, H 52, 55

Werkhoven, C. 18

Westwater, J. 10

Williams, D.B. 45, 69

Williams, F.E. 87, 200, 204

Wischmeier, L. 38, 53, 60, 61, 66, 113

Wittemann, Joerg V. 3

Wittry, D.B. 95

Wong, C C 5, 19, 25, 36, 37, 39, 41, 78,

187, 190

Woo, Robyn L. 32

Woodbury, H.H. 25

Wright, P.J. 17, 101

Wu, Di 5, 19, 41, 77, 187, 188, 193

Wu, Y. 11

Wu, Yan 117

Wu, Yiying 4, 5

Wyckoff, R.W.G. 2

Xi, Baojuan 38, 187, 188

Xia, Y. 11

Xia, Yuxing 117

Xiang, B. 5, 19, 24, 36, 47, 77, 187–189

Xiang, Yanjuan 78

Xiao, Rui 32

Xiao, Zhongdang 5, 187, 188

Xie, Ping 4

Xie, Q. 5, 12, 13, 23, 187, 188

Xie, Shengwu 117

Xie, Sishen 78

Xie, Yi 5

Xiong, Qihua 12, 31, 44, 78

Xiong, S. 5, 12, 13, 23, 187, 188

Bibliography 295

Xiong, Shenglin xvi, 5, 38, 187, 188

Xu, Dechen 38, 187, 188

Xu, Di 31

Xu, Dongsheng 5

Xu, J. 5, 19, 24, 36, 47, 77, 187–189

Xu, T. xvii, 5, 9, 12, 23, 26–28, 31, 32, 35,

37, 48, 50, 51, 53, 57, 94, 96, 97, 152,

187, 188

Xu, Tao 5, 23, 32, 38, 50, 96, 187, 188

Xu, Xinlong 78

Xu, Y M 12

Xu, Yeming 5, 12, 23, 38, 187, 188

Xue, Dongfeng 5, 187, 188

Yacobi, B.G. 75

Yamada, Yoichi 21

Yamaga, Shigeki 17, 18

Yamamoto, Y. 78

Yamashita, Tomoki 33, 158

Yamazaki, O. 161

Yan, Guangqing 3

Yan, H. 5, 11

Yan, Haoquan 4, 5

Yan, P. 5, 12, 23, 53, 187, 188

Yan, Park J.-H.-Choi Y. Heo C.-J. Yang

S.-M. Lee L.P. Yang P., R. 4

Yan, R. 4

Yan, Ruoxue 4

Yan, Shancheng 5, 187, 188

Yang, Ben 5, 19, 41, 77, 187, 188, 193

Yang, F.H. 5, 19, 24, 36, 47, 77, 187–189

Yang, Heqing 5, 31, 32, 187, 188

Yang, Jian 5, 188

Yang, L. 5, 23, 32, 38, 50, 96, 187, 188

Yang, Lan 5, 187, 188

Yang, P. 4, 5, 11

Yang, Peidong 4, 5, 11

Yang, Qing 5, 31, 32, 187, 188

Yang, S. xvii, 5, 12, 23, 26–28, 32, 35, 48,

50, 51, 53, 57, 94, 96, 97, 187, 188

Yang, Suxia 5, 19, 53, 83, 187, 192

Yang, X H 101, 104, 105

Yang, X.H. 101, 104, 105

Yang, Yang 4

Yao, Takafumi 17, 101, 105, 161

Yao, Tingting 5, 188

Yarrison-Rice, Jan 6, 158, 162

Yarrison-Rice, J.M. 5, 32

Ye, C. 5, 12, 23, 53, 187, 188

Ye, Changhui 7, 9, 31

Yeh, Chin-Yu 9

Yim, W.M. 2

Yin, Y. 11

Yin, Zhenyu 4

Yiu, Wing-Ching 5, 12, 23, 38, 187, 188

Yoffe, A.D. xiv, xxvi, 57, 131, 134–139,

148, 149, 161, 166, 168, 169, 177, 178

Yoon, M.Y. 91, 101

Yoon, S.F. 117

Yoshida, A 26, 28

Yoshida, T. 101, 104, 105, 108

Yoshikawa, Akihiko 17, 18

You, S.H. 91, 101, 104, 105, 108

Youn, C.J. 36, 91, 101, 104, 105, 108

Young, M.L. 17

Yu, D.P. 5, 19, 24, 36, 47, 77, 187–189

Yu, Guihua 3, 4

Yu, H. 4

Bibliography 296

Yu, Hao 5, 188

Yu, Kin Man 101, 104, 105

Yu, Phil Won 26

Yu, P.Y. 36, 91, 101

Yuan, Bo 5, 12, 187

Yuan, Huajun 78

Yuan, Qing-Xin 5, 187, 188

Yuk, T.I. 5, 12, 19, 36, 77, 187, 188, 192

Yuk, Tong I. 5, 12, 24, 187, 188

Zach, Franz X. 101, 104, 105

Zacharias, Margit 4

Zacks, Eliam 57, 87, 88, 92–94, 200, 204,

205

Zakharov, Oleg xxvii, 217

Zapien, J.A. 55

Zembutsu, Sakae 95

Zengel, J.J. 12, 78

Zhai, Huazhang 5, 12, 13, 187, 188

Zhai, Tianyou xvi, 5, 6

Zhai, Ying-Teng 31

Zhang, H.Z. 5, 19, 24, 36, 47, 77, 187–189

Zhang, L. 5, 12, 23, 53, 187, 188

Zhang, L.D. 9, 31, 33

Zhang, Lide 7, 9, 31

Zhang, Lihui 5, 31, 32, 187, 188

Zhang, Ming Liang 5, 23, 31, 32, 38, 187,

188

Zhang, Shufen 5, 12, 23, 187, 188

Zhang, Wen Jun 5, 23, 31, 32, 38, 187, 188

Zhang, Wenjun 3, 5

Zhang, X. 5, 12, 23, 32, 53, 187, 188

Zhang, X T 5, 19, 25, 36, 37, 39, 41, 78,

187, 190

Zhang, Xiao Hong 5, 23, 31, 32, 38, 187,

188

Zhang, Xiaogang xvi, 5

Zhang, Xin 6, 32, 158, 162

Zhang, X.T. xvii, 5, 12, 13, 15, 19, 23–25,

36–38, 40, 44, 47, 48, 53, 77, 95, 114,

124, 187–190

Zhang, X.X. 12

Zhang, Yongfeng 31

Zhang, Yuan-yuan 5, 12, 13, 187, 188

Zhang, Yugang 5, 19, 41, 77, 187, 188, 193

Zhang, Zengxing 78

Zhang, Z.H. 31

Zhang, Zhihua 5, 187, 188

Zhao, J. 5, 12, 23, 53, 187, 188

Zhao, Q. 5, 19, 24, 36, 47, 77, 187–189

Zhao, Qing 5, 188

Zhao, Xiaowei 78

Zhao, Yanyuan 78

Zhao, Zhenghuan 4

Zheng, G.P. 5, 19, 36, 37, 41, 47, 77, 187,

191

Zheng, Hang 117

Zheng, Xu-sheng 5, 12, 13, 187, 188

Zhong, G Z 26, 28, 105, 108, 125

Zhong, Honghai 5, 19, 41, 77, 187, 188, 193

Zhou, Hongyang 5, 19, 38, 41, 77, 187,

188, 193

Zhou, Weiya 78

Zhou, Zijian 4

Zhu, Hesun 5, 12, 13, 187, 188

Zhu, H.W. xvii, 5, 24, 57, 187, 188

Zhu, Liying 5

Zhu, Xing 5, 187, 188

Bibliography 297

Zhu, Ying-Chun 5, 13, 187, 188

Zhu, Ziqiang 161

Zhu, Zuoming 117

Zhuang, Xiujuan 6

Zhukov, E.A. 21

Zimmermann, H 125

Zimmler, Mariano A 4

Zou, Bingsuo 5, 187, 188

Zou, J. 32

Zou, Jin 6, 32, 158, 162

Zulehner, W. xxvii, 85, 87, 196, 203, 204

Zunger, Alex 9

Zvyagin, B.B. 157

Zwiller, V. 6

Ankur's PhD Thesis - TSpace

Documents