Au 2 S x /CdS Nanorods by Cation Exchange: Mechanistic Insights into the Competition Between Cation-Exchange and Metal Ion Reduction

1© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim wileyonlinelibrary.com

Au 2 S x /CdS Nanorods by Cation Exchange: Mechanistic Insights into the Competition Between Cation-Exchange and Metal Ion Reduction

Subhajit Kundu , Paromita Kundu , Gustaaf Van Tendeloo , and N. Ravishankar *

electron microscopy (TEM) techniques, we confi rm that

cation exchange does indeed take place in the CdS/Au

system leading to the formation of Au 2 S x ( x = 1 or 3). The

stability of the product, however, depends on the external

driving force available. Under conditions where the system

is able to overcome the kinetic barrier (under electron beam

irradiation, for instance), we observe the conversion of the

metastable Au 2 S x to stable Au. The observations and analysis

have interesting and important implications for the design

and choice of systems for the formation of multifunctional

nanoscale heterostructures using cation exchange.

CdS nanorods used for the study were prepared by a

modifi ed literature method. [ 24 ] The as-grown nanorods have

a uniform morphology and are wurtzite in phase with an

average diameter of ≈10 nm and a length of ≈100 nm. To

study the feasibility of cation exchange, different amounts of

these CdS nanorods were dispersed in 20 mL of de-ionized

water and 10 mL (1.8 m m ) HAuCl 4 solution was added. An

immediate change of color of the CdS powder from bright

yellow to brownish yellow was observed with progressive

darkening of the yellow color over an hour. To investigate

the change, UV–visible absorbance spectra ( Figure 1 a) of

the centrifuged supernatant of reaction were acquired after

an hour that shows a signifi cant drop in the concentration

of HAuCl 4 . For continuous monitoring of the change in the

supernatant, the consumption of HAuCl 4 was studied as a

function of time (inset). The decay kinetics of the 284 nm

peak that corresponds to the ligand-to-metal charge transfer

peak of HAuCl 4 was monitored [ 25,26 ] which indicates a signifi -

cant reduction in the rate of decay with time. To facilitate a

faster completion of reaction, the product was irradiated with

a microwave at 800 W for 3 min that results in a fall in the

HAuCl 4 concentration below the detectable limit as indi-

cated by the absorbance plot.

Figure 1 b shows the SEM-EDS data of bare CdS and

the resulting product after reaction with HAuCl 4 showing

the S, Cd and Au atomic percent in each case. The S:Cd ratio

in as-synthesized CdS is nearly 1:1 indicating stochiometric

CdS nanorods. The S:Cd ratio is higher after reaction with

HAuCl 4 but the S:(Cd+Au) ratio is maintained at 1:1. This

observation indicates that Au is incorporated in the CdS lat-

tice replacing Cd indicating a possible cation exchange reac-

tion. Energy-dispersive X-ray analysis in the SEM of the

supernatant on drying shows the Cd:Cl:S atomic percent

ratio to be 30:62:8. The Cd:Cl ratio indicates the formation of

Heterostructures

DOI: 10.1002/smll.201400524

S. Kundu, Prof. N. Ravishankar Materials Research Centre Indian Institute of Science C.V. Raman Avenue Bangalore 560012 , India E-mail: [email protected]

Dr. P. Kundu, Prof. G. V. Tendeloo Electron Microscopy for Materials Research (EMAT) University of Antwerp Groenenborgerlaan 171, 2020 , Antwerp , Belgium

Semiconductor-based heterostructures fi nd applications in

photovoltaics, hydrogen generation, catalysis, labeling and

sensing. [ 1–11 ] While several routes are available for the syn-

thesis of such heterostructures, cation exchange is one of the

most favored methods. [ 12–15 ] The primary advantage of cation

exchange is that the morphology of the parent phase may

be retained leading to possibility of synthesis of anisotropic

shapes that do not conform to the point group of the mate-

rial. Additionally, kinetically controlled conditions extend

its utility to create materials with metastable composition

and phases. [ 12,16 ] Preservation of anionic framework during

cation exchange [ 17,18 ] enables synthesis of multicomponent

heterostructures with enhanced level of control over mor-

phology and chemistry. [ 19 ] Site specifi city of the exchange

reaction gives rise to several interesting possibilities of nano-

structuring including Ag 2 S bar-coding in CdS, [ 20 ] formation

of Cu 2 S on the tip of CdS nanorods [ 21,22 ] and multipodal

structures. [ 23 ]

In spite of its signifi cant advantages, one of the pri-

mary limitations of cation exchange is that the metals with

higher electron affi nity tend to undergo reduction rather

than cation exchange, thus making the process very system

specifi c. [ 20 ] This competition between cation exchange and

reduction is not well understood. Using CdS/Au as a model

system, we show that cation exchange is feasible even in the

case of metals with higher electron affi nity like Au. We use

thermodynamic arguments to show that the process depends

on a delicate balance between the formation energy of the

product sulfi de and the by-products of cation exchange as

compared to the free energy for the reduction of the metal.

While thermodynamics can predict that the sulfi de can

indeed form, the experimental realization depends on the

kinetics of the process. By employing advanced transmission

small 2014, DOI: 10.1002/smll.201400524

S. Kundu et al.

2 www.small-journal.com

communications

© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

CdCl 2 further supporting the replacement of Cd in the CdS.

Expulsion of a small amount of S from the CdS lattice is also

seen. In addition to the Na 2 S solution, the colorless superna-

tant yields a yellow CdS precipitate indicating the removal of

the Cd 2+ ions from the rods, further supporting the hypoth-

esis that Cd ions in the nanorods are exchanged by Au ions.

XRD data of the samples after reaction show the pres-

ence of peaks corresponding to the wurtzite phase of CdS.

Additional peaks due to the formation of Au 2 S are not

clearly seen (Figure 1 c) in samples with low loading of Au

(≈10%). However at higher loading (≈22% and ≈39%) of

Au the peak due to Au 2 S [ 27 ] becomes more prominent con-

fi rming cation-exchange.

To confi rm the formation of sulfi de of Au, core-level

X-ray photoelectron spectroscopy (XPS) was carried out.

Au4f spectra (Figure 1 d) show the presence of Au 1+ [ 28 ] and

Au 3+ in addition to Au 0 . The absence of a signifi cant peak in

the survey around 200 eV due to Cl2p eliminates any pos-

sibility of chloride of Au. EDS observations clearly indicate a

1:1 replacement of Cd ions by Au. Based on the charge bal-

ance, it is clear that 2 Cd 2+ ions are replaced by one Au 1+ and

one Au 3+ ion leading to both site and charge balance. Based

on the XPS analysis, we observe that the ratio of (Au 0 +

Au 1+ ):Au 3+ is 1:1 indicating that some of the Au 1+ ions on the

surface are converted to the Au 0 state. Au 2 S 3 formation is not

detected in XRD due to its amorphous nature as reported

earlier. [ 29 ] Cd3d spectra of CdS (Supporting Information,

Figure S1e) shows the presence of S 2− and OH − type species.

An additional peak around 409 eV and 415 eV appears (Sup-

porting Information, Figure S1f) for the hybrid sample which

may be due to the emergence of dangling bonds of Cd as has

been reported previously. [ 30 ] This is possibly due to the pres-

ence of an excess amount of Au 3+ at the surface. Two Au 3+

ions can bond with three S 2− ions replacing two Cd 2+ ions

while the third Cd 2+ bond remains unsatisfi ed. This is also

consistent with the fact that some of Au 1+ ions on the surface

are converted to Au 0 .

The morphology of the nanorods remains intact after

cation exchange as shown by the bright-fi eld TEM image

( Figure 2 a); it also shows the presence of Au nanoparticles

of ≈2–3 nm on the surface. However on careful observation,

we fi nd that more of these particles form under the electron

small 2014, DOI: 10.1002/smll.201400524

Figure 1. a) UV–visible absorbance spectra of supernatant, after addition of HAuCl 4 (HA) to an aqueous dispersion of CdS nanorods, recorded before and after irradiation with microwave (MW). It shows a sharp decrease in concentration of HA in the supernatant on mixing which further drops after MW irradiation. Inset shows the kinetics of the decay in concentration of HA. b) SEM-EDS of the hybrid samples with different loadings of gold, indicating a possible cation exchange. c) XRD data of the CdS-Au 2 S x showing peaks for cubic Au 2 S phase. d) XPS confi rming the presence of +1 and +3 oxidation state of gold along with Au 0 .

Mechanistic Insights into the Competition Between Cation-Exchange and Metal Ion Reduction

3www.small-journal.com© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

beam. In case of low loading, the Au 2 S x formation is mostly

at the surface of the CdS rods and the nucleation burst takes

place due to reduction under the electron beam, which results

in faceted fi ne particles of Au as is evident from the high-

resolution high angle annular dark fi eld (HAADF) image

acquired in scanning transmission (STEM) mode (Figure 2 b).

Such beam effect is highly accentuated(Supporting Informa-

tion, Movie 1,2) in the case of higher loading where the meta-

stable Au 2 S x formation is across the depth of the CdS rods

(HAADF micrograph in Figure 2 c,d). The possible reason for

non-uniform distribution, size and shape of the Au particles

could be due to the diffusion of Au 0 from the bulk of the rod

to the surface leading to uncontrolled growth of the formed

nuclei as well as migration and aggregation of the juxtaposed

particles on the surface. To further confi rm the presence and

distribution of Au 2 S/Au 2 S 3 phase in the CdS nanorods energy

dispersive X-ray (EDS) mapping in the STEM mode has

been carried out on the ultra-high loading sample as given

in Figure 2 c. It is evident that S is distributed uniformly

throughout the rod; Cd is concentrated mostly on the places

of lighter contrast in correspondence to the HAADF image

and Au is specifi cally present in the brighter contrast region.

Therefore this implies that under high percentage replace-

ment the anionic sites remain conserved and only the cation

gets replaced, as reported earlier. [ 17 ] The effect of the elec-

tron beam has also been studied for these samples; the same

specimen region was exposed for a few minutes under a

focused electron beam. Figure 2 d shows the low magnifi ca-

tion HAADF-STEM image and the elemental maps acquired

immediately after exposing which confi rms the growth of the

Au particle at the tip of the rod as marked in the fi gure. A

similar transformation has been shown at different times in

Figure 3 .

Using high resolution HAADF-STEM, heavier and

lighter atomic columns can be clearly distinguished which

makes the images directly interpretable which is otherwise

diffi cult by HRTEM. [ 31,32 ] Therefore, we exploit this mode

of electron microscopy to reveal the phase distribution. The

HAADF-STEM image in Figure 4 a clearly shows three dif-

ferent regions with different contrast as bound by the dotted

lines. The HRSTEM image in Figure 4 b confi rms that the

lighter contrast region corresponds to CdS and the brighter

contrast region corresponds to the cubic Au 2 S phase imaged

along the [110] zone. It shows that a portion of rod remains

small 2014, DOI: 10.1002/smll.201400524

Figure 2. a) BF-TEM image of a low-loading CdS-Au 2 S x sample. b) HRSTEM image shows a faceted Au particle, bound by {111} and {001} planes, on CdS rod in [002] growth direction. c) HAADF-STEM image and elemental maps of ultra-high loading sample reveals the presence of Au 2 S x region along with CdS and Au regions. d) Same region after exposing to electron beam shows the transformation of Au 2 S x to Au as marked by the red circles.

S. Kundu et al.

4 www.small-journal.com

communications

© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

intact as CdS with [002] growth direction and a part is con-

verted to cubic Au 2 S, grown in the [110] direction along the

rod. The Au 2 S phase gets reduced to Au under focused elec-

tron beam exposure as evident from the HAADF-STEM

image (Supporting Information, Figure S2a,b).

The free energy of formation of gold

sulfi des by cation exchange reaction as

a function of temperature is shown in

Figure 5 a (details provided in Supporting

Information). The Gibbs free energy

change of formation (ΔG) for both Au 2 S/

Au 2 S 3 is positive which indicates that the

formation of these sulfi des is not thermo-

dynamically feasible. However, due to a

highly favorable formation energy of the

by-product CdCl 2 on cation exchange, the

overall reaction is favorable in this case.

According to calculations, the forma-

tion of Au 2 S 3 is thermodynamically more

favorable than the formation of Au 2 S. To

compare this with the competing reaction

of metal ion reduction, we also evaluated

the free energy change for the formation of Au with water

acting as the reducing agent in the medium by a method

reported in literature. [ 33–35 ] We note that this free energy

change is negative indicating that water can indeed reduce

Au salt to metal at room temperature. However, our control

small 2014, DOI: 10.1002/smll.201400524

Figure 3. Transformation of the CdS-Au 2 S x sample under converged electron beam.

Figure 4. a) HAADF-STEM image clearly shows the presence of CdS, Au 2 S and Au phase as indicated by regions of different contrast bound by blue dotted lines. b) HRSTEM of a selected portion (marked in “a”) showing a signifi cant conversion of hexagonal CdS to cubic Au 2 S x , as imaged in [ 110 110 ] zone orientation. c) A perfect crystal lattice of simple cubic Au 2 S consisting of few unit cells where Au forms a fcc lattice confi guration. d) View of the lattice in [ 110 110 ] zone orientation matching with the Au atom positions in the HRSTEM image.

Mechanistic Insights into the Competition Between Cation-Exchange and Metal Ion Reduction

5www.small-journal.com© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

experiments indicate that there could be signifi cant kinetic

barriers and thus higher temperatures and/or a lowering of

nucleation barrier by lowering the interfacial energy for heter-

ogeneous nucleation may be needed for the formation of Au [ 36 ]

(also see Supporting Information). Our calculations show that

the formation of Au 2 S 3 /Au 2 S is much more favorable than the

formation of Au by reduction. As noted above, although ther-

modynamic calculations indicate the possibilities of formation,

the actual formation of the product under experimental condi-

tions is limited by the kinetic barriers for nucleation. It is very

likely that the barrier for cation exchange where the product

phase forms coherently in the parent is likely to have a lower

interfacial energy and barrier for nucleation as compared to

the formation of a metal with a heterointerface. Once formed,

the Au 2 S/Au 2 S 3 phase is possibly stabilized by the kinetic bar-

rier for the decomposition reaction. Under ambient conditions

and even in the presence of a strong reducing agent, the sulfi de

phase is stable and intact in bulk of the rod as the reduction (if

any) takes place only at the surface of the nanorods in con-

tact with the reducing agent. However, the electron beam can

interact intensely with a signifi cant depth of the material and

thus we observe that Au 2 S/Au 2 S 3 decomposes in situ during

microscopy to give more stable elemental form of the constit-

uent elements.

The reaction conditions signifi cantly affect the compe-

tition between reduction and exchange. For instance, it has

been reported earlier [ 37–39 ] that for the same system viz.,

AuCl 3 and CdS, cation exchange is not favorable when the

reaction was carried out in an organic medium in the pres-

ence of surfactants. However, both our experiments and ther-

modynamic calculations clearly show that cation exchange

does indeed take place in an aqueous medium. The tem-

perature dependence of Δ G predicts that the increment in

driving force per unit change in temperature is more for

reduction than for cation exchange. But the Δ G value for

exchange is much more than that for reduction in the tem-

perature range of interest indicating that still cation exchange

is dominant. The EDS data of a CdS-Au 2 S x sample before

and after microwave (MW) show a sharp increase in the Au

content from 2 to 4 atomic percent indicating that the MW

has a signifi cant role in driving the reac-

tion to completion in a short period of

time. More importantly XPS shows that

the Au 0 :Au 0/1+/3+ peak area ratio increases

slightly on doing MW as predicted by the

thermodynamic calculations. A change in

concentration of HAuCl 4 has a negligible

change on the driving force of both reduc-

tion and exchange (Supporting Informa-

tion, Figure S4). Thus, we conclude that in

the temperature and concentration range

of interest cation exchange is favored with

a relatively weak dependence on reaction

parameters unlike reduction.

In summary, we have demonstrated

that Au, inspite of having a high electron

affi nity, may undergo cation-exchange with

Cd ions in CdS. The formation of a Au 2 S

phase in an exchange reaction with CdS,

has been pointed out clearly for the fi rst time using electron

microscopy. Thermodynamic calculations on the possibility of

cation exchange consolidate the mechanistic understanding.

This predictability is directly applicable to other system com-

binations and provides insights on the competition between

cation exchange and reduction process.

Experimental Section

Synthesis of CdS Nanorods : Synthesis of CdS nanorods have been carried out by solvothermal reaction of a mixture of Cd(NO 3 ) 2 (247 mg), thiourea (122 mg), deionized water (20 mL) and ethylen-ediamine (20 mL) at 160 °C for 8 h. Cleaning was done by repeated sonication and centrifugation with ethanol and water. The product was dried and characterized before further use.

Synthesis of Au 2 S x -CdS : For synthesis of Au 2 S x -CdS hybrid, CdS (20 mg) was dispersed in deionized water (20 mL). HAuCl 4 solution (4 mg in 10 mL deionized water) was added to vigorously stirring CdS colloid. The mixture was stirred for ≈5 min and then irradiated with microwave at 800 W for 3 min. The product was cleaned by repeated sonication and centrifugation with water and dried for characterization. The loading(till 10%) was varied by varying the initial amount of CdS keeping the HAuCl 4 concentration same. However for synthesizing ultra-high loading sample(22%, 39%, and 52% Au) CdS amount was kept fi xed while the concentration of HAuCl 4 solution was varied.

Supporting Information

Supporting Information is available from the Wiley Online Library or from the author.

Acknowledgements

N.R. acknowledges Department of Science and Technology (DST) India for fi nancial support. P.K. and G.V.T. acknowledge ERC grant

small 2014, DOI: 10.1002/smll.201400524

Figure 5. a) Plot of Δ G as a function of temperature for cation-exchange (c.e.) and reduction (red.). Plots for individual formation of Au 2 S and Au 2 S 3 (half-fi lled circles) are also shown. b) Schematic of the thermodynamic and kinetic scenario indicating that cation-exchange is more favorable than reduction under the experimental conditions.

S. Kundu et al.

6 www.small-journal.com

communications

© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim small 2014, DOI: 10.1002/smll.201400524

COUNTATOMS. Tecnai F30 and T20 TEM are a part of Advanced Microscopy and Microanalysis (AFMM) at Indian Institute of Sci-ence (IISc). The authors acknowledge the use of XPS ECSA system which is a part of IISc. The XPS and EDS facility in CENSE is also acknowledged.

[1] M. Gratzel , Nature 2001 , 414 , 338 – 344 . [2] Y. Zhang , Y. Tang , X. Liu , Z. Dong , H. H. Hng , Z. Chen , T. C. Sum ,

X. Chen , Small 2013 , 9 , 996 – 1002 . [3] P. Thiyagarajan , H.-J Ahn , J.-S. Lee , J.-C. Yoon , J.-H. Jang , Small

2013 , 9 , 2341 – 2347 . [4] A. Leelavathi , B. Mukherjee , C. Nethravathi , S. Kundu , M. Dhivya ,

N. Ravishankar , G. Madras , RSC Adv. 2013 , 3 , 20970 – 20977 . [5] R. Costi , A. E. Saunders , E. Elmalem , A. Salant , U. Banin , Nano

Lett. 2008 , 8 , 637 – 641 . [6] J. Tian , Y. Sang , Z. Zhao , W. Zhou , D. Wang , X. Kang , H. Liu ,

J. Wang , S. Chen , H. Cai , H. Huang , Small 2013 , 9 , 3864 – 3872 . [7] P. Kundu , P. A. Deshpande , G. Madras , N. Ravishankar , J. Mater.

Chem. 2011 , 21 , 4209 – 4216 . [8] Y. Yang , M. Y. Gao , Adv. Mater. 2005 , 17 , 2354 – 2357 . [9] M. Bruchez , M. Moronne , P. Gin , S. Weiss , A. P. Alivisatos , Science

1998 , 281 , 2013 – 2016 . [10] K. Liu , M. Sakurai , M. Aono , Small 2012 , 8 , 3599 – 3604 . [11] H. Wang , Z. Sun , Q. Lu , F. Zeng , D. Su , Small 2012 , 8 , 1167 –

1172 . [12] J. B. Rivest , P. K. Jain , Chem. Soc. Rev. 2012 , 42 , 89 – 96 . [13] X. Wu , Y. Yu , Y. Liu , Y. Xu , C. Liu , B. Zhang , Angew. Chem. Int. Ed.

2012 , 51 , 3211 – 3215 . [14] S. Gupta , S. V. Kershaw , A. L. Rogach , Adv. Mater. 2013 , 25 ,

6923 – 6944 . [15] S. Deka , K. Miszta , D. Dorfs , A. Genovese , G. Bertoni , L. Manna ,

Nano Lett. 2010 , 10 , 3770 – 3776 . [16] H. Li , M. Zanella , A. Genovese , M. Povia , A. Falqui , C. Giannini ,

L. Manna , Nano Lett. 2011 , 11 , 4964 – 4970 . [17] P. K. Jain , L. Amirav , S. Aloni , A. P. Alivisatos , J. Am. Chem. Soc.

2010 , 132 , 9997 – 9999 . [18] D. H. Son , S. M. Hughes , Y. Yin , A. P. Alivisatos , Science 2004 ,

306 , 1009 – 1012 . [19] H. Li , R. Brescia , R. Krahne , G. Bertoni , M. J. P. Alcocer ,

C. D’Andrea , F. Scotognella , F. Tassone , M. Zanella , M. De Giorgi , L. Manna , ACS Nano 2012 , 6 , 1637 – 1647 .

[20] R. D. Robinson , B. Sadtler , D. O. Demchenko , C. K. Erdonmez , L.-W. Wang , A. P. Alivisatos , Science 2007 , 317 , 355 – 358 .

[21] J. B. Rivest , S. L. Swisher , L.-K. Fong , H. Zheng , A. P. Alivisatos , ACS Nano 2011 , 5 , 3811 – 3816 .

[22] B. Sadtler , D. O. Demchenko , H. Zheng , S. M. Hughes , M. G. Merkle , U. Dahmen , L.-W. Wang , A. P. Alivisatos , J. Am. Chem. Soc. 2009 , 131 , 5285 – 5293 .

[23] K. Miszta , D. Dorfs , A. Genovese , M. R. Kim , L. Manna , ACS Nano 2011 , 5 , 7176 – 7183 .

[24] F. W. Guicun Li , Z. Zhang , J. Nanopart. Res. 2005 , 7 , 685 – 689 . [25] G. Jiang , L. Wang , W. Chen , Mater. Lett. 2007 , 61 , 278 – 283 . [26] J. Won , K. J. Ihn , Y. S. Kang , Langmuir 2002 , 18 , 8246 – 8249 . [27] K. Ishikawa , T. Isonaga , S. Wakita , Y. Suzuki , Solid State Ionics

1995 , 79 , 60 – 66 . [28] Y. Mikhlin , M. Likhatski , A. Karacharov , V. Zaikovski , A. Krylov ,

Phys. Chem. Chem. Phys. 2009 , 11 , 5445 – 5454 . [29] M. Kristl , M. Drofenik , Inorg. Chem. Commun. 2003 , 6 ,

1419 – 1422 . [30] C.-P. Huang , C.-F. Chao , M.-Y. Shen , T.-M. Chen , Y.-K. Li , Chem.

Asian J. 2012 , 7 , 2848 – 2853 . [31] S. J. Pennycook , M. F. Chisholm , A. R. Lupini , M. Varela ,

A. Y. Borisevich , M. P. Oxley , W. D. Luo , K. van Benthem , S. H. Oh , D. L. Sales , S. I. Molina , J. Garcia-Barriocanal , C. Leon , J. Santamaria , S. N. Rashkeev , S. T. Pantelides , Philos. Trans. R. Soc. London, A 2009 , 367 , 3709 – 3733 .

[32] J. Liu , Microsc. Microanal. 2004 , 10 , 55 – 76 . [33] B. Viswanath , P. Kundu , B. Mukherjee , N. Ravishankar , Nanotech-

nology 2008 , 19 , 195603 . [34] B. Viswanath , P. Kundu , N. Ravishankar , J. Coll. Interf. Sci. 2009 ,

330 , 211 – 219 . [35] P. Kundu , N. Singhania , G. Madras , N. Ravishankar , Dalton Trans.

2012 , 41 , 8762 – 8766 . [36] E. A. Anumol , P. Kundu , P. A. Deshpande , G. Madras ,

N. Ravishankar , ACS Nano 2011 , 5 , 8049 – 8061 . [37] A. E. Saunders , I. Popov , U. Banin , J. Phys. Chem. B 2006 , 110 ,

25421 – 25429 . [38] G. Menagen , J. E. Macdonald , Y. Shemesh , I. Popov , U. Banin , J.

Am. Chem. Soc. 2009 , 131 , 17406 – 17411 . [39] R. Costi , A. E. Saunders , U. Banin , Angew. Chem. Int. Ed. 2010 ,

49 , 4878 – 4897 .

Received: February 26, 2014 Revised: May 8, 2014 Published online: