Formation, Morphology and Optical Properties of SiC Nanopowder

Formation, Morphology and Optical Properties of SiC Nanopowder

T. Nychyporuk 1, a, O. Marty 2, b, J.M. Bluet 1, c, V. Lysenko 1, d, R. Perrin 1, e, G. Guillot 1, f and D. Barbier 1, g

1 Materials Physics Laboratory (LPM), CNRS UMR-5511, INSA de Lyon, 7, av. Jean Capelle, Bat. Blaise Pascal, 69621 Villeurbanne, France

2 Laboratory on Electronics, Nanotechnologies, Sensors (LENAC), Claude Bernard University of Lyon, Bat. Brillouin, 8, André-Marie Ampère Str., 69622 Villeurbanne, France

a [email protected], b [email protected], c [email protected], d [email protected], e [email protected], f [email protected],

g [email protected]

Keywords: nanopowder, nanoparticles, TEM, photoluminescence

Abstract. SiC nanopowder has been formed using an original technological approach based on

grinding of bulk porous SiC nanostructures. The initial porous SiC nanostructures were obtained by

anodization of n+-type 4H-SiC substrate in HF/Ethanol solution under UV illumination. Large

single SiC nanoparticles (~ 30 nm in diameter) constituting the nanopowder have a porous structure

which can be clearly visible. On the other hand, small single SiC nanoparticles (~ 4 nm in diameter)

exhibit a clear crystalline structure. A broad and very intense luminescence band (400 – 900 nm)

provided from the nanopowder corresponds to the radiative processes involving nanoparticle

surface states. A smaller photoluminescence peak centred at 358 nm may correspond to radiative

recombination of the photogenerated excitons confined in the individual and spatially separated 4H-

SiC nanoparticles.

Introduction

Pioneer publications on porous SiC formation go back to 1994 [1, 2]. Since that time many works

performed on this material have been oriented towards a better understanding of its fundamental

properties [3, 4] as well as towards a variety of applications [5-8]. A special interest of such porous

structures is the fabrication of nanoporous SiC layers consisting in numerous interconnected SiC

nanoparticles forming a continuous low-dimensional network in volume in which quantum and

meso-scale phenomena are widely present. As in the case of Si nanostructures, optical

measurements are often used for study of these phenomena. For example, an evidence of quantum

confinement in 3C-SiC nanostructures has been recently observed by using photoluminescent

spectroscopy [9]. However, total 3D charge confinement which is present in an individual

nanoparticle can be strongly broken down in the case of the porous nanostructures due to the

numerous interconnections present between the nanoparticles. The main aim of this work was to

obtain 4H-SiC nanopowder non-embedded in any complex matrix and having strongly reduced

number of the nanoparticle interconnections as well as to study its optical properties.

Experimental

A n-type 4H-SiC substrate (Nd = 4.1018 cm

-3) was used. A thin polycrystalline 3C-SiC layer (1 µm)

was present on its back face due to a previous use of the substrate for another study. The initial

porous SiC nanostructures were obtained by anodization of the substrate in HF/Ethanol solution

under UV illumination. Each porous SiC layer occupies an area of about 0.8 cm2. The thickness of

each porous layer is around 200 micrometers. Then, a mechanical grinding of the bulk porous SiC

nanostructures was performed to decompose the porous nanostructure down to the separated

elementary nanoparticles. By this way, SiC nanopowder constituted by numerous nanoparticles was

Materials Science Forum Vols. 527-529 (2006) pp 763-766Online available since 2006/Oct/15 at www.scientific.net© (2006) Trans Tech Publications, Switzerlanddoi:10.4028/www.scientific.net/MSF.527-529.763

All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of TTP,www.ttp.net. (ID: 134.214.84.38, SCD Doc' INSA - INSA de Lyon, Villeurbanne, France-27/04/14,17:32:45)

obtained. Photoluminescence (PL) properties

of the nanopowder were investigated at low

temperature (8 K) using a 244 nm laser as

optical excitation source.

Results and discussion

First of all, structural properties (crystalline

nature and chemical composition) of the SiC

nanopowder have been investigated by TEM.

Chemical analysis of the formed nanopowder

confirms its composition based on Si and C

atoms. Fig. 1 represents a sub-microscopic

photo of the nanopowder. Numerous isolated

and slightly interconnected nanoparticles of

different sizes as well as big nano-clusters can

be observed. Two kinds of nanoparticles are mainly present. Firstly, a large single porous SiC

nanoparticle (~ 30 nm in diameter) is shown on Fig. 2-a). Being clearly distinguishable, a porous

structure is quite comparable with previously observed single porous silicon nanoparticles which

have a fractal structure [10]. Secondly, small SiC single nanoparticle (~ 4 nm in diameter) shown on

Fig. 2-b) are also present in the nanopowder mixture and exhibits a clear crystalline nature. A low

temperature PL spectrum obtained on the back side of the original bulk 4H-SiC substrate is shown

on Fig. 3. As a 244 nm wavelength was used, the excitation mainly occurs in the polycrystalline

3C-SiC layer due to the high absorption coefficient of SiC at this wavelength. Consequently the PL

spectrum is dominated by (i) nitrogen bound exciton-phonon replicas and (ii) peaks corresponding

to the N-Al donor-acceptor pairs (DAP) in 3C-SiC. The 4H N-Al DAP band emerges slightly at

480 nm. In Figure 4, a low temperature PL spectrum of the SiC nanopowder containing mainly 4H

polytype and a small amount of 3C polytype (less than 5%) is compared to the PL spectrum of the

original bulk SiC substrate. A broad asymmetric peak corresponding to the SiC nanopowder is

observed. It is centred at 663 nm (1.87 eV) and extends all over the visible range from UV (400 nm)

to near infrared (900 nm). As it can be clearly seen in the figure, the PL signal obtained from the

nanopowder is really huge. Its integrated PL intensity is 190 times higher than the signal from the

bulk crystalline substrate. A similar PL signal was also observed under excitation of the

nanopowder with 514.5 nm line of an argon laser. The dominant red luminescence may be due to

a) b)

Figure 2 : High resolution TEM photographs of single large porous (a) and single small

crystalline (b) SiC nanoparticles constituting the SiC nanopowder.

500

nm Figure 1 : TEM picture of the SiC nanopowder

nm

764 Silicon Carbide and Related Materials 2005

an amorphous part of the SiC nanopowder as

proposed by Kassiba et al. [11]. Even if

crystalline nanoparticles are observed by

TEM, the presence of an amorphous sheet

around them or around the nanoparticle

interconnections in the nano-clusters and/or

in the large porous nanoparticles can be

responsible for the obtained intense PL

signal. The high value of the amorphous-like

specific surface in the nanoporous powder

can easily explain the observed high

luminescence intensity. The large width of

the PL peak is due to the broad distribution

of electronic states related to the amorphous

low-dimensional nanoparticles. However,

besides this broad intensive band, a small

peak centered at 358 nm (3.46 eV) can be

also observed. Spectral position of this peak is found to be above the 4H SiC band gap. Therefore, it

can be attributed to the quantum confinement in the 4H-SiC nanoparticles present in the

nanopowder mixture.

In Figure 5, the PL spectra of nanoporous SiC and Si as well as of SiC nanopowder are

compared. Firstly, the same intensive red luminescence band is clearly observed from these three

different nanostructured materials. This experimental fact leads us to put forward an hypothesis of a

similar electronic mechanism responsible for such PL spectra in the nanoporous IV-IV

semiconductors – radiative states related to the amorphous-like specific surface of these

nanostructures, for example. The second noticeable point is the asymmetric shape of this

luminescence band in the case of SiC nanopowder. Indeed, a shoulder at the UV wing of the

spectrum extended from 550 nm down to 400 nm can be easily distinguished. This part of the PL

spectrum may be, for example, (i) due to quantum confinement or (ii) due to some surface states

present in spatially separated individual 3C-SiC nanoparticles. The hypothesis of the quantum

confinement in 3C nanoparticles is consistent with the possible interpretation of the 358 nm peak

(quantum confinement in the individual and spatially separated 4H-SiC nanoparticles) discussed

above. A similar smaller peak is also observed in the bulk nanoporous SiC layer. However, (i) its

intensity is lower and (ii) its energy is closer to the energy gap of the bulk crystalline material.

4000 5000 6000 7000 80000

2

4

6

8

10

12

14

16

18

PL

Inte

nsity

(a.

u.)

Wavelength (Å)

4H N-AlDAP

N-BE in 3C3C N-Al

DAP

4000 5000 6000 7000 80000

2

4

6

8

10

12

14

16

18

PL

Inte

nsity

(a.

u.)

Wavelength (Å)

4H N-AlDAP

N-BE in 3C3C N-Al

DAP

Figure 3 : Low temperature PL spectrum of the

original n+-type 4H-SiC substrate.

Figure 4 : Low temperature PL spectra of the

nanopowder and of the bulk SiC substrate

(notice that the latter is multiplied by 10).

3000 4000 5000 6000 7000 8000 90000

200

400

600

800

1000

1200

1400

1600 Nanoparticles

Bulkx10

PL

Inte

nsi

ty (

a.u

.)

Wavelength (Å)

Figure 5 : Low temperature PL for

nanoporous SiC (diamonds), nanoporous Si

(down triangles) and SiC nanopowder

(circles). The spectra are normalized for

easier comparison. A zoom in the UV

range is presented in the upper left corner.

3000 4000 5000 6000 7000 8000 90000

250

500

750

1000

1250

1500

1750

PL

In

ten

sity

(a

.u.)

Wavelength (Å)

3250 3500 3750 4000

020406080

100120

3000 4000 5000 6000 7000 8000 90000

250

500

750

1000

1250

1500

1750

PL

In

ten

sity

(a

.u.)

Wavelength (Å)

3250 3500 3750 4000

020406080

100120

Materials Science Forum Vols. 527-529 765

Considering that this small peak may be due to recombination of excitons confined in the

nanoparticles, it is logical to find it to be (i) smaller and (ii) red-shifted in comparison with the peak

corresponding to the SiC nanopowder because (i) of the smaller PL quantum issue and (ii) of the less

efficient confinement in the case of the bulk nanoporous material in which numerous

interconnections between the nanoparticles broking down 3D confinement of photogenerated

excitons are present.

Conclusion

Large nanoporous and small crystalline SiC nanoparticles non-embedded in any matrix and

constituting a nanopowder mixture were obtained from grinding of bulk nanoporous SiC. The

resulting nanopowder contains (i) numerous isolated, (ii) slightly interconnected nanoparticles of

different sizes as well as (iii) big nano-clusters. Our photoluminescence analysis of the nanopowder

reveals the presence of two photoluminescence bands: (i) intense broad red band attributed to the

nanoparticle surface states and (ii) small UV band which may be related to quantum confinement in

4H SiC nanoparticles. Nevertheless the UV band could also be attributed to oxide at particle surface

for instance. Consequently, beyond these first results, further experiments will be performed in order

to obtain a clear assignment of the 358 nm peak origin.

References

[1] J.S. Shor, L. Bemis, A.D. Kurtz, I. Grimberg, B.Z. Weiss, M.F. Mac Millan and W.J. Choyke: J.

Appl. Phys. 76 (1994), p. 4045.

[2] A.O. Konstantinov, C.I. Harris and E. Janzén: Appl. Phys. Lett. 65 (1994), p. 2699.

[3] T.V. Torchynska, A. Vivas Hernandez, A. Diaz Cano, S. Jimenez-Sandoval, S. Ostapenko and

M. Mynbaeva: J. Appl. Phys. 97 (2005), p. 033507-1

[4] Y. Shishkin, W.J. Choyke and R.P. Devaty: Mat. Sci. Forum 457-460 (2004), p. 1467.

[5] M. Mynbaeva, S.E. Saddow, G. Melnychuk, I. Nikitina, M. Sheglov, A. Sitnikova, N.

Kuznetsov, K. Mynbaev and V. Dmitriev: Appl. Phys. Lett. 78 (2001), p. 117.

[6] J. Bai, G. Dhanaraj, P. Gouma, M.Dudley and M. Mynbaeva: Mat. Sci. Forum 457-460 (2004),

p. 1479.

[7] A.J. Rosenbloom, Y. Shishkin, D.M. Sipe, Y. Ke, R.P. Devaty and W.J. Choyke: Mat. Sci.

Forum 457-460 (2004), p. 1463.

[8] W.T. Hsieh, Y.K. Fang, W.J. Lee, C.W. Ho, K.H Wu and J.J. Ho: Electr. Lett. 36 (2000), p.

1869.

[9] X.L. Wu, J.Y. Fan, T. Qiu, X. Yang, G.G. Siu and Paul K. Chu: Phys. Rev. Lett. 94 (2005), p.

026102-1.

[10] T. Nychyporuk, V. Lysenko, and D. Barbier: Phys. Rev. B 71 (2005), p. 115402.

[11] A. Kassiba, M. Makowska-Janusik, J. Bouclé, J.F. Bardeau, A. Bulou and N. Herlin-Boime:

Phys. Rev. B 66 (2002), p. 155317.

766 Silicon Carbide and Related Materials 2005

Silicon Carbide and Related Materials 2005 10.4028/www.scientific.net/MSF.527-529 Formation, Morphology and Optical Properties of SiC Nanopowder 10.4028/www.scientific.net/MSF.527-529.763

DOI References

[1] J.S. Shor, L. Bemis, A.D. Kurtz, I. Grimberg, B.Z. Weiss, M.F. Mac Millan and W.J. Choyke: J. ppl.

Phys. 76 (1994), p. 4045.

doi:10.1063/1.357352 [2] A.O. Konstantinov, C.I. Harris and E. Janzén: Appl. Phys. Lett. 65 (1994), p. 2699.

doi:10.1063/1.112610 [4] Y. Shishkin, W.J. Choyke and R.P. Devaty: Mat. Sci. Forum 457-460 (2004), p. 1467.

doi:10.4028/www.scientific.net/MSF.457-460.1467 [5] M. Mynbaeva, S.E. Saddow, G. Melnychuk, I. Nikitina, M. Sheglov, A. Sitnikova, N. uznetsov, K.

Mynbaev and V. Dmitriev: Appl. Phys. Lett. 78 (2001), p. 117.

doi:10.1063/1.1337628 [6] J. Bai, G. Dhanaraj, P. Gouma, M.Dudley and M. Mynbaeva: Mat. Sci. Forum 457-460 (2004), . 1479.

doi:10.4028/www.scientific.net/MSF.457-460.1479 [7] A.J. Rosenbloom, Y. Shishkin, D.M. Sipe, Y. Ke, R.P. Devaty and W.J. Choyke: Mat. Sci. orum 457-460

(2004), p. 1463.

doi:10.4028/www.scientific.net/MSF.457-460.1463 [9] X.L. Wu, J.Y. Fan, T. Qiu, X. Yang, G.G. Siu and Paul K. Chu: Phys. Rev. Lett. 94 (2005), p. 26102-1.

doi:10.1103/PhysRevLett.94.026102 [10] T. Nychyporuk, V. Lysenko, and D. Barbier: Phys. Rev. B 71 (2005), p. 115402.

doi:10.1103/PhysRevB.71.115402 [1] J.S. Shor, L. Bemis, A.D. Kurtz, I. Grimberg, B.Z. Weiss, M.F. Mac Millan and W.J. Choyke: J. Appl.

Phys. 76 (1994), p. 4045.

doi:10.1063/1.357352 [2] A.O. Konstantinov, C.I. Harris and E. Janzn: Appl. Phys. Lett. 65 (1994), p. 2699.

doi:10.1063/1.112610 [3] T.V. Torchynska, A. Vivas Hernandez, A. Diaz Cano, S. Jimenez-Sandoval, S. Ostapenko and M.

Mynbaeva: J. Appl. Phys. 97 (2005), p. 033507-1

doi:10.1063/1.1840095 [5] M. Mynbaeva, S.E. Saddow, G. Melnychuk, I. Nikitina, M. Sheglov, A. Sitnikova, N. Kuznetsov, K.

Mynbaev and V. Dmitriev: Appl. Phys. Lett. 78 (2001), p. 117.

doi:10.1063/1.1337628 [6] J. Bai, G. Dhanaraj, P. Gouma, M.Dudley and M. Mynbaeva: Mat. Sci. Forum 457-460 (2004), p. 1479.

doi:10.4028/www.scientific.net/MSF.457-460.1479 [7] A.J. Rosenbloom, Y. Shishkin, D.M. Sipe, Y. Ke, R.P. Devaty and W.J. Choyke: Mat. Sci. Forum 457-

460 (2004), p. 1463.

doi:10.4028/www.scientific.net/MSF.457-460.1463 [9] X.L. Wu, J.Y. Fan, T. Qiu, X. Yang, G.G. Siu and Paul K. Chu: Phys. Rev. Lett. 94 (2005), p. 026102-1.

doi:10.1103/PhysRevLett.94.026102 [11] A. Kassiba, M. Makowska-Janusik, J. Boucl, J.F. Bardeau, A. Bulou and N. Herlin-Boime: Phys. Rev. B

66 (2002), p. 155317.

doi:10.1103/PhysRevB.66.155317