- Email: [email protected]
Fabrication, spectroscopy, and dynamics of highly luminescent core–shell InP@ZnSe quantum dots
Journal of Colloid and Interface Science 350 (2010) 5–9
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Fabrication, spectroscopy, and dynamics of highly luminescent core–shell InP@ZnSe quantum dots Mee Rahn Kim, Jae Hun Chung, Mihee Lee, Seonghoon Lee, Du-Jeon Jang * School of Chemistry, Seoul National University, NS60, Seoul 151-742, Republic of Korea
a r t i c l e
i n f o
Article history: Received 18 November 2009 Accepted 12 June 2010 Available online 19 June 2010 Keywords: Lifetime Nanoparticle Photoluminescence Semiconductor Trap emission
a b s t r a c t InP quantum dots of 3 nm in diameter have been prepared using a dehalosilylation reaction and passivated with ZnSe to enhance photoluminescence by 6.8 times. Core–shell InP@ZnSe quantum dots dispersed in n-hexane have then been investigated using time-resolved spectroscopy to understand their photoluminescence dynamics. The observed decay times of 0.1, 7, and 1100 ns have been attributed to the relaxation times of electrons in the conduction band, trap sites, and surface states. The surface-state luminescence of core–shell InP@ZnSe quantum dots having the maximum at 760 nm has been distinguished spectrally and dynamically from their band-edge emission having the maximum at 620 nm or from their trap-site emission having the maximum at 660 nm. Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction The design and the controlled fabrication of nanostructured materials having functional properties have been extensively studied [1–11]. In particular, the chemical synthesis and the sizedependent properties of IIVI and IIIV semiconductor nanocrystals are of great interest owing to their potential application in electrical and optoelectronic devices [8–11]. IIIV semiconductor nanocrystals possess superior electronic and optoelectronic properties in comparison to IIVI semiconductor nanomaterials. The main advantages offered by IIIV semiconductor nanocrystals lie in the robustness of the covalent bond versus the ionic bond of IIVI semiconductors and in the reduced toxicity of compounds [12–16]. Therefore, IIIV quantum dots are potential candidates for biological applications such as bioimaging and photodynamic therapy. Among them, InP nanoparticles have been extensively studied particularly because they are hardly toxic and have a broad photoluminescence color range in the visible spectrum [13–16]. InP is a direct band-gap semiconductor and has a bulk band-gap energy of 1.35 eV. Many groups have spent considerable time and effort on the solution phase synthesis of InP nanoparticles using diverse methods such as a dehalosilylation reaction method or a hot injection technique [17–20]. However, InP nanocrystals synthesized in organic solutions show quite low band-edge photoluminescence due to surface traps, dangling bonds, stacking faults, and a high activation barrier for carrier detrapping [21–23]. Because of large surface-to-volume ratios of small nanoparticles, * Corresponding author. Fax: +82 2 889 1568. E-mail address: [email protected] (D.-J. Jang). 0021-9797/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2010.06.037
photoexcited electrons in their conduction bands are extensively trapped into surface states. It is well-known that defect sites of luminescent semiconductor nanocrystals, such as trap sites and surface states, act as photodegradation and luminescent-quenching sites. The defect sites trap electrons and/or holes to induce the non-radiative recombination of the charge carriers, leading to the reduction of luminescence efficiency. Quantum dots exhibit narrow photoluminescence profiles that can be tuned by treating their defect sites with the elimination of both anionic and cationic dangling bonds at their surfaces. The surfaces of nanoparticles can be tailored physically via thermal treatment or chemically via organic or inorganic capping to enhance photoluminescence profoundly [21–27]. Thus, to improve the photoluminescence of InP quantum dots, surface phosphorous atoms lying at the origins of trap sites were removed by being etched with HF or nanocrystals were passivated with shell materials that have wider band gaps [17,24–27]. The type-I core–shell semiconductor, where the shell material having a wider band gap is over grown onto the core material having a narrower band gap, reduces the non-radiative recombination effectively by confining the wave function of an electron–hole pair to the interior of the core crystal [28–30]. In such a core–shell nanocrystal, the shell provides a physical barrier between the optically active core and the surrounding medium, thus elevating stability against various chemical and physical changes. In this paper, we report that highly luminescent core–shell InP@ZnSe quantum dots have been fabricated via a hot injection method and characterized by time-resolved luminescence spectroscopy. The photoluminescence of core–shell InP@ZnSe nanoparticles is enhanced by 6.8 times compared with that of bare InP nanoparticles. In addition, the surface-state photoluminescence
6
M.R. Kim et al. / Journal of Colloid and Interface Science 350 (2010) 5–9
of core–shell InP@ZnSe nanoparticles has been well separated spectrally and dynamically from their band-edge emission or from trap-site emission. 2. Materials and methods 2.1. Materials For the synthesis of InP quantum dots, trioctylphosphine oxide (TOPO) of 10 g and chloroindium oxalate of 1.0 g were degassed in a five-neck flask at 180 °C under 200 m torr of N2 (g) for 30 min. After degassing, the reaction flask was maintained at 270 °C under 1 atm of N2 (g). Tris(trimethylsilyl)phosphine of 0.60 mL and trioctylphosphine (TOP) of 6.0 mL were mixed in a glove box at room temperature, and the mixture was rapidly injected into the reaction flask using a microsyringe, and then temperature of the reaction flask was temporally dropped by the injection of the mixture. The reaction flask was then maintained at 280 °C for several days. Butanol was added to the final reaction mixture, whose color then became brown. The dark brown solution was centrifuged to separate the residual products of InP nanoparticles, which were subsequently added with methanol, precipitated, and separated again by centrifugation. For the synthesis of core–shell InP@ZnSe nanoparticles, both diethylzinc and trioctylphosphine selenide were saturated in TOP in a glove box. The mixture solution was taken in a microsyringe and injected into a hot TOPO solution of 6.0 mL containing InP nanoparticles, which had already been heated to 200 °C under 1 atm of N2 (g), at a rate of six drops per 10 min under vigorous stirring. Product nanoparticles in the reaction flask were annealed at 90 °C for a day under vigorous stirring, and separated by centrifugation. Colloidal samples were prepared by suspending InP nanoparticles or core–shell InP@ZnSe nanoparticles in n-hexane immediately prior to use. 2.2. Methods High-resolution transmission electron microscopy (HRTEM) images were measured employing a microscope (JEOL, JEM-3000F) while high-resolution X-ray diffraction (HRXRD) patterns were recorded with a diffractometer (Bruker, D8 DISCOVER) using Cu Ka radiation (k = 0.154178 nm). Absorption spectra were obtained with a UV/vis spectrometer (Scinco, S-3100), and photoluminescence spectra were detected using a home-built fluorimeter consisting of a Xe lamp of 75 W (ARC, XS432) and monochromators of 0.15 m and 0.30 m (ARC, Spectropro 150 and 300, respectively). Picosecond photoluminescence kinetic profiles were measured employing a streak camera of 10 ps (Hamamatsu, C2839) attached with a CCD (Princeton Instruments, RTE128H) with excitation using a modelocked Nd:YAG laser of 25 ps (Quantel, YG501). In addition, nanosecond photoluminescence kinetic profiles were measured with a photomultiplier tube (Hamamatsu, R928) connected an oscilloscope of 1 GHz (Lecroy, Wavepro 950) with excitation using a Q-switched Nd:YAG laser of 6 ns (Quantel, Brilliant). Kinetic constants were extracted by fitting measured kinetic profiles to computer-simulated exponential curves convoluted with instrument response functions. Nanosecond time-resolved luminescence spectra were measured with an intensified CCD (Princeton Instruments, ICCD576G), having a gating resolution of 2 ns and attached to a 0.5 m spectrometer (Acton Research, Spectrapro 500), with excitation from a Q-switched Nd:YAG laser (Quantel, Brilliant) of 6 ns. Both the laser and the intensified CCD were triggered by a pulse/delay generator (Stanford Research Systems, DG535). 3. Results and discussion The HRTEM image of Fig. 1 shows that our fabricated core–shell InP@ZnSe quantum dots have diameters of about 3 nm. Every
Fig. 1. HRTEM images of core–shell InP@ZnSe quantum dots. All three nanoparticles enclosed in white dashed circles having diameters of 3 nm show a latticefringe distance of 0.34 nm.
nanocrystal in white dashed circles shows a lattice-fringe distance of 0.34 nm, which agrees well with the spacing of 0.3388 nm between the (1 1 1) planes of the zincblende InP structure. This suggests that the ZnSe shells of the core–shell nanocrystals are very thin. We have estimated that the thickness of the ZnSe shells is approximately one monolayer. The nanoparticles have been investigated with powder HRXRD. The HRXRD patterns of bare InP nanocrystals and core–shell InP@ZnSe nanocrystals are shown in Fig. 2. On one hand, the HRXRD patterns of core–shell InP@ZnSe nanocrystals (with dotted lines as eye guidances in order to show how well peaks match with bulk bar patterns), bulk zincblende InP (denoted as bars on the bottom), and bulk zinc–blende ZnSe (denoted as bars on the bottom) are shown in Fig. 2a. The peaks of {(1 1 1), (2 2 0), and (3 1 1)} coming from the core InP and the nearby peaks of {(1 1 1), (2 2 0), and (3 1 1) marked with arrows} due to the shell ZnSe show up definitely. Thus, this indicates clearly that our InP@ZnSe nanocrystals have a core–shell structure. On the other hand, the HRXRD patterns of InP nanocrystals and bulk zincblende InP (denoted as bars on the bottom) are shown in Fig. 2b. The characteristic peaks of InP nanoparticles agree well with those of the bulk zincblende InP. We consider that the passivation of the core InP nanocrystals with the ZnSe shells has been carried out uniformly with a minor atomic level reconstruction of the surface because the lattice mismatch of the lattice parameters between zincblende InP and zincblende ZnSe is insignificant [29,31]. The mean crystallite diameter, d, of the small nanocrystals can be determined from the line width of a HRXRD spectrum by the Scherrer’s formula: hdi = 0.94k/(Bcos hB), where k is the employed X-ray wavelength, B is the full width at half maximum of the diffraction peak (radian), and hB is the half angle of the diffraction peak on the 2h scale. The mean crystalline diameter of InP quantum dots has been estimated to be 2.2 nm, which agrees approximately with the mean core diameter of core–shell InP@ZnSe quantum dots observed with Fig. 1. On one hand, Fig. 3 indicates that the absorption spectra of bare InP nanoparticles and core–shell InP@ZnSe nanoparticles dispersed in n-hexane show characteristic excitonic transition peaks at 600 nm. The absorption band edge of core–shell InP@ZnSe quantum dots seems to be shifted slightly to the blue compared to that of bare InP quantum dots because passivation has increased spectral broadening and scattering significantly [25,26]. On the other
M.R. Kim et al. / Journal of Colloid and Interface Science 350 (2010) 5–9
Fig. 2. (a) HRXRD patterns of core–shell InP@ZnSe nanocrystals (red), bulk zinc– blende InP (green), and bulk zinc–blende ZnSe (orange). (b) HRXRD patterns of InP nanocrystals (blue) and bulk zinc–blende InP (green). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
7
ences potential confinement and tends to localize in the core nanocrystal because the conduction and the valence bands of the core material are lower and higher in energy than the respective ones of the shell material. However, the photoluminescence of our prepared core–shell InP@ZnSe nanoparticles is blue-shifted slightly because the photoluminescence of core–shell InP@ZnSe nanoparticles is less contributed by defect emission than that of bare InP nanoparticles is. The narrow band-edge emission of core–shell InP@ZnSe at 620 nm suggests that the core InP nanoparticles are almost monodisperse in size. The passivation of InP nanocrystals with ZnSe improves the emission quantum yields of the core nanocrystals enormously to increase the band-edge photoluminescence by 6.8 times by reducing defect sites at the surfaces of InP quantum dots. Note that the passivation of InP nanocrystals with ZnS has been reported to increase the quantum yield of photoluminescence by a factor of 2 [17]. The defect emission of core–shell InP@ZnSe nanoparticles at 700–750 nm is shifted on a large scale to the blue from that of bare InP nanoparticles at 750–800 nm [32–36]. Furthermore, the ratio of defect emission to band-edge emission in core–shell nanocrystals is much smaller than that in bare nanocrystals. These also indicate that passivation with ZnSe has blocked efficiently the ensnaring of electrons in the conduction band of a InP nanocrystal into trap sites to enhance band-edge emission with decreasing defect emission. We have deconvoluted each photoluminescence spectrum with two Gaussian functions to understand the nature of defect emission in both bare and core–shell nanocrystals. Whereas the deconvoluted defect emission of bare InP quantum dots has the maximum at 760 nm, that of core–shell InP@ZnSe quantum dots shows the peak at 710 nm. This also supports that passivation with ZnSe reduces the trap sites and surface states of InP quantum dots. Fig. 4 reveals that the photoluminescence kinetic profiles of core–shell InP@ZnSe nanoparticles suspended in n-hexane show three exponential-decay components of 0.1 ns (28%), 0.8 ns (38%), and 7 ns (34%) as indicated in Table 1. The decay time of 0.1 ns is considered to result from the capturing time of photoexcited electrons at the conduction band into shallow traps. While the time constant of 0.8 ns is ascribed to arise mainly from the ensnaring time of electrons in shallow traps into deep traps, the decay time of 7 ns is assigned to the trapping time of electrons in deep traps
Fig. 3. Absorption (dotted) and photoluminescence spectra (solid) of bare InP nanoparticles (blue) and core–shell InP@ZnSe nanoparticles (red) suspended in nhexane. Samples were excited at 500 nm for the photoluminescence spectra, and each photoluminescence spectrum was deconvoluted with two Gaussian functions (green). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
hand, the photoluminescence spectra of bare InP and core–shell InP@ZnSe nanoparticles in n-hexane show band-edge emission peaks at 630 nm and 620 nm, respectively. In general, the electrons and the holes of a type-I core–shell semiconductor material experi-
Fig. 4. Picosecond luminescence kinetic profiles, measured in short (a) and long time windows (b), of core–shell InP@ZnSe nanoparticles dispersed in n-hexane. The sample was excited at 532 nm and monitored at 600 nm.
8
M.R. Kim et al. / Journal of Colloid and Interface Science 350 (2010) 5–9
Table 1 Luminescence kinetic time constants of core–shell InP@ZnSe nanoparticles in n-hexane, extracted from Figs. 4 to 5.
a b
Figures
kmon (nm)a
Rise time (ns)
Decay time (ns)
Fig. 4 Fig. 5
600 800
Instant 60
0.1 (28%)b + 0.8 (38%) + 7 (34%) 1100
Monitored luminescence wavelength. Initial amplitude percentage of each component.
into even deeper traps. The photoluminescence of zincblende InP nanowires has been reported to have a decay time of 5.0–8.4 ns [37], which is similar to the relaxation time of electrons in the deep trap sites of core–shell InP@ZnSe quantum dots. The nanosecond luminescence kinetic profiles of Fig. 5, monitored at 800 nm, show that the surface-state photoluminescence of InP@ZnSe nanoparticles rises on the time scale of 60 ns and decays exponentially on the time scale of 1100 ns (Table 1). It is noteworthy that the rise kinetics of photoluminescence at trap sites or surface states is very difficult to be measured by a time-resolved spectroscopic method in general. The rise time of 60 ns, attributed to the trapping time of electrons at deeper traps into surface states, suggests that diverse trap sites such as deeper traps exist at interfaces between InP cores and ZnSe shells. Finally, electrons ensnared at surface trap states decay radiatively into the valence band of InP core quantum dots by electron–hole recombination on the slow time scale of 1100 ns. The time-resolved luminescence spectra of InP@ZnSe nanoparticles dispersed in n-hexane obtained at different time delays after excitation (Fig. 6) have led us to measure a trap-site emission spectrum and a surface-state emission spectrum separately. However, we could not obtain a time-resolved spectrum of exclusive bandedge luminescence with our current temporal resolution of 6 ns because photoexcited electrons at the conduction band are captured into trap sites on the time scale of 0.1 ns (Fig. 4). The emission spectrum of Fig. 6a reveals that emission measured at the short time delay after excitation is governed by the recombination of electrons in trap sites with holes in the valence band. Thus, the
Fig. 6. Trap-site (a) and surface-state luminescence spectra (b) of core–shell InP@ZnSe quantum dots dispersed in n-hexane, measured at time delays of 15 ns and 480 ns, respectively, after sample excitation at 532 nm using a Q-switched Nd:YAG laser of 6 ns.
emission spectrum of Fig. 6a having the maximum at 660 nm hardly contains either band-edge emission or surface-state emission. Trap-site emission is shifted extensively (by 40 nm) to the red from the band-edge emission spectrum of Fig. 3. Furthermore, the band width of trap-site emission is much broader (by 1.3 times) than that of band-edge emission. The broad band was considered to associate with the trap-to-band recombination of weakly coupled electron–hole pairs, where electrons were trapped at phosphorous vacancies and holes were located at the valence band. Thus, we suggest that the trap-site emission spectrum of Fig. 6a arises from the radiative relaxation of electrons at diversely different deep and deeper trap sites [33–36]. The emission spectrum of Fig. 6b having the maximum at 760 nm results from the relaxation of electrons at surface states into the valence band. It is noteworthy that the wavelength at the maximum of Fig. 6b is almost the same as the wavelength at the maximum of the deconvoluted trap emission of bare InP quantum dots in Fig. 3. The enormously large red shift (100 nm) of Fig. 6b from Fig. 6a suggests that the surface states of InP quantum dots are extremely different energetically and environmentally from the trap sites of InP quantum dots. We consider that this has enabled us to measure the formation kinetics of the surface-state photoluminescence of core–shell InP@ZnSe nanocrystals. 4. Summary
Fig. 5. Nanosecond luminescence kinetic profiles, measured in short (a) and long time windows (b), of core–shell InP@ZnSe nanoparticles dispersed in n-hexane. The sample was excited at 532 nm and monitored at 800 nm.
InP quantum dots of 3 nm in diameter have been prepared via hot injection synthesis and passivated with ZnSe to enhance photoluminescence by 6.8 times. Core–shell InP@ZnSe quantum dots dispersed in n-hexane have then been investigated using time-resolved spectroscopy to understand their photoluminescence dynamics. Fig. 7 summarizes the dynamics and spectroscopy of the photoluminescence of core–shell InP@ZnSe nanoparticles dispersed in n-hexane. Photoexcited electrons in the conduction band are trapped into shallow traps within 0.1 ns. On one hand, while deep traps capture electrons from shallow traps in 0.8 ns, their electrons are ensnared into deeper traps on the time scale of 7 ns. On the other hand, the photoluminescence of surface states rises in 60 ns and decays on the time scale of 1100 ns. Combining
M.R. Kim et al. / Journal of Colloid and Interface Science 350 (2010) 5–9
Fig. 7. Schematic diagram for the luminescence relaxation kinetics of core–shell InP@ZnSe quantum dots dispersed in n-hexane. All the time constants are given in units of nanoseconds.
time-resolved photoluminescence spectra with a static luminescence spectrum, we have separated the individual spectra of trap-site emission at 660 nm and surface-state luminescence at 760 nm from the spectrum of band-edge photoluminescence at 620 nm. Acknowledgments This work was supported by research Grants through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science, and Technology (2009-0071184 and 20090082846). D.J.J. is also thankful to the SRC program of NRF (2010-0001635) while M.R.K. acknowledges a BK21 scholarship as well. S.L. also acknowledges the NANO Systems Institute, National Core Research Center, KIER, and NRF for artificial atoms research. References [1] Z. Tang, Z. Zhang, Y. Wang, S.C. Glotzer, N.A. Kotov, Science 314 (2006) 274.
9
[2] S.Y. Park, A.K.R. Lytton-Jean, B. Lee, S. Weigand, G.C. Schatz, C.A. Mirkin, Nature 451 (2008) 553. [3] M.R. Kim, D.-J. Jang, Chem. Commun. (2008) 5218. [4] J.-Y. Kim, S. Lee, K.-H. Yoo, D.-J. Jang, Appl. Phys. Lett. 94 (2009) 153301. [5] M.R. Kim, S.Y. Heo, D.-J. Jang, J. Colloid. Interface Sci. 339 (2009) 217. [6] M.A. El-Sayed, Acc. Chem. Res. 37 (2004) 326. [7] W.K. Bae, J. Kwak, J.W. Park, K. Char, C. Lee, S. Lee, Adv. Mater. 21 (2009) 1690. [8] M.S. Gudiksen, L.J. Lauhon, J. Wang, D.C. Smith, C.M. Lieber, Nature 415 (2002) 617. [9] X. Peng, J. Wickham, A.P. Alivisatos, J. Am. Chem. Soc. 120 (1998) 5343. [10] S. Mokkapati, C. Jagadish, Mater. Today 12 (2009) 22. [11] M.C. Beard, G.M. Turner, J.E. Murphy, O.I. Micic, M.C. Hanna, A.J. Nozik, C.A. Schmuttenmaer, Nano Lett. 3 (2003) 1695. [12] S. Hussain, N. Won, J. Nam, J. Bang, H. Chung, S. Kim, ChemPhysChem 10 (2009) 1466. [13] S. Kim, W. Shim, H. Seo, J.H. Bae, J. Sung, S.H. Choi, W.K. Moon, G. Lee, B. Lee, S.W. Kim, Chem. Commun. (2009) 1267. [14] D.W. Lucey, D.J. MacRae, M. Furis, Y. Sahoo, A.N. Cartwright, P.N. Prasad, Chem. Mater. 17 (2005) 3754. [15] R. Xie, D. Battaglia, X. Peng, J. Am. Chem. Soc. 129 (2007) 15432. [16] J. Wang, M.S. Gudiksen, X. Duan, Y. Cui, C.M. Lieber, Science 293 (2001) 1455. [17] S. Xu, J. Ziegler, T. Nann, J. Mater. Chem. 18 (2008) 2653. [18] D.V. Talapin, N. Gaponic, H. Horchert, A.L. Rogach, M. Haase, H. Weller, J. Phys. Chem. B 106 (2002) 12659. [19] R.L. Woo, L. Gao, N. Goel, M.K. Hudait, K.L. Wang, S. Kodambaka, R.F. Hicks, Nano Lett. 9 (2009) 2207. [20] S. Xu, S. Kumar, T. Nann, J. Am. Chem. Soc. 128 (2006) 1054. [21] S. Adam, D.V. Talapin, H. Borchert, A. Lobo, C. McGinley, A.R.B. de Castro, M. Haase, H. Weller, T. Möller, J. Chem. Phys. 123 (2005) 084706. [22] A. Narayanaswamy, L.F. Feiner, P.J. van der Zaag, J. Phys. Chem. C 112 (2008) 6775. [23] K. Adlkofer, E.F. Duijs, F. Findeis, M. Bichler, A. Zrenner, E. Sackmann, G. Abstreiter, M. Tanaka, Phys. Chem. Chem. Phys. 4 (2002) 785. [24] L. Li, P. Reiss, J. Am. Chem. Soc. 130 (2008) 11588. [25] E. Ryu, S. Kim, E. Jang, S. Jun, H. Jang, B. Kim, S.-W. Kim, Chem. Mater. 21 (2009) 573. [26] S. Haubold, M. Haase, A. Kornowski, H. Weller, ChemPhysChem 2 (2001) 331. [27] C. Li, M. Ando, H. Enomoto, N. Murase, J. Phys. Chem. C 112 (2008) 20190. [28] M.R. Kim, Y.-M. Kang, D.-J. Jang, J. Phys. Chem. C 111 (2007) 18507. [29] P. Reiss, M. Protière, L. Li, Small 5 (2009) 154. [30] J.J. Li, Y.A. Wang, W. Guo, J.C. Keay, T.D. Mishima, M.B. Johnson, X. Peng, J. Am. Chem. Soc. 125 (2003) 12567. [31] S.Zh. Karazhanov, L.C. Lew Yan Voon, Semiconductors 39 (2005) 161. [32] A.V. Maleev, I.V. Ignatiev, I.Ya. Gerlovin, I.E. Kozin, Y. Masumoto, Phys. Rev. B 71 (2005) 195323. [33] J.L. Blackburn, D.C. Selmarten, R.J. Ellingson, M. Jones, O. Micic, A.J. Nozik, J. Phys. Chem. B 109 (2005) 2625. [34] S.-H. Kim, R.H. Wolters, J.R. Heath, J. Chem. Phys. 105 (1996) 7957. [35] Y. Zhao, Z. Dong, S. Miao, A. Deng, J. Yang, B. Wang, J. Appl. Phys. 100 (2006) 123519. [36] L. Langof, L. Fradkin, E. Ehrenfreund, E. Lifshitz, O.I. Micic, A.J. Nozik, Chem. Phys. 297 (2004) 93. [37] K. Pemasiri, M. Montazeri, R. Gass, L.M. Smith, H.E. Jackson, J. Yarrison-Rice, S. Painam, Q. Gao, H.H. Tan, C. Jagadish, X. Zhang, J. Zou, Nano Lett. 9 (2009) 648.
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science www.elsevier.com/locate/jcis
Fabrication, spectroscopy, and dynamics of highly luminescent core–shell InP@ZnSe quantum dots Mee Rahn Kim, Jae Hun Chung, Mihee Lee, Seonghoon Lee, Du-Jeon Jang * School of Chemistry, Seoul National University, NS60, Seoul 151-742, Republic of Korea
a r t i c l e
i n f o
Article history: Received 18 November 2009 Accepted 12 June 2010 Available online 19 June 2010 Keywords: Lifetime Nanoparticle Photoluminescence Semiconductor Trap emission
a b s t r a c t InP quantum dots of 3 nm in diameter have been prepared using a dehalosilylation reaction and passivated with ZnSe to enhance photoluminescence by 6.8 times. Core–shell InP@ZnSe quantum dots dispersed in n-hexane have then been investigated using time-resolved spectroscopy to understand their photoluminescence dynamics. The observed decay times of 0.1, 7, and 1100 ns have been attributed to the relaxation times of electrons in the conduction band, trap sites, and surface states. The surface-state luminescence of core–shell InP@ZnSe quantum dots having the maximum at 760 nm has been distinguished spectrally and dynamically from their band-edge emission having the maximum at 620 nm or from their trap-site emission having the maximum at 660 nm. Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction The design and the controlled fabrication of nanostructured materials having functional properties have been extensively studied [1–11]. In particular, the chemical synthesis and the sizedependent properties of IIVI and IIIV semiconductor nanocrystals are of great interest owing to their potential application in electrical and optoelectronic devices [8–11]. IIIV semiconductor nanocrystals possess superior electronic and optoelectronic properties in comparison to IIVI semiconductor nanomaterials. The main advantages offered by IIIV semiconductor nanocrystals lie in the robustness of the covalent bond versus the ionic bond of IIVI semiconductors and in the reduced toxicity of compounds [12–16]. Therefore, IIIV quantum dots are potential candidates for biological applications such as bioimaging and photodynamic therapy. Among them, InP nanoparticles have been extensively studied particularly because they are hardly toxic and have a broad photoluminescence color range in the visible spectrum [13–16]. InP is a direct band-gap semiconductor and has a bulk band-gap energy of 1.35 eV. Many groups have spent considerable time and effort on the solution phase synthesis of InP nanoparticles using diverse methods such as a dehalosilylation reaction method or a hot injection technique [17–20]. However, InP nanocrystals synthesized in organic solutions show quite low band-edge photoluminescence due to surface traps, dangling bonds, stacking faults, and a high activation barrier for carrier detrapping [21–23]. Because of large surface-to-volume ratios of small nanoparticles, * Corresponding author. Fax: +82 2 889 1568. E-mail address: [email protected] (D.-J. Jang). 0021-9797/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jcis.2010.06.037
photoexcited electrons in their conduction bands are extensively trapped into surface states. It is well-known that defect sites of luminescent semiconductor nanocrystals, such as trap sites and surface states, act as photodegradation and luminescent-quenching sites. The defect sites trap electrons and/or holes to induce the non-radiative recombination of the charge carriers, leading to the reduction of luminescence efficiency. Quantum dots exhibit narrow photoluminescence profiles that can be tuned by treating their defect sites with the elimination of both anionic and cationic dangling bonds at their surfaces. The surfaces of nanoparticles can be tailored physically via thermal treatment or chemically via organic or inorganic capping to enhance photoluminescence profoundly [21–27]. Thus, to improve the photoluminescence of InP quantum dots, surface phosphorous atoms lying at the origins of trap sites were removed by being etched with HF or nanocrystals were passivated with shell materials that have wider band gaps [17,24–27]. The type-I core–shell semiconductor, where the shell material having a wider band gap is over grown onto the core material having a narrower band gap, reduces the non-radiative recombination effectively by confining the wave function of an electron–hole pair to the interior of the core crystal [28–30]. In such a core–shell nanocrystal, the shell provides a physical barrier between the optically active core and the surrounding medium, thus elevating stability against various chemical and physical changes. In this paper, we report that highly luminescent core–shell InP@ZnSe quantum dots have been fabricated via a hot injection method and characterized by time-resolved luminescence spectroscopy. The photoluminescence of core–shell InP@ZnSe nanoparticles is enhanced by 6.8 times compared with that of bare InP nanoparticles. In addition, the surface-state photoluminescence
6
M.R. Kim et al. / Journal of Colloid and Interface Science 350 (2010) 5–9
of core–shell InP@ZnSe nanoparticles has been well separated spectrally and dynamically from their band-edge emission or from trap-site emission. 2. Materials and methods 2.1. Materials For the synthesis of InP quantum dots, trioctylphosphine oxide (TOPO) of 10 g and chloroindium oxalate of 1.0 g were degassed in a five-neck flask at 180 °C under 200 m torr of N2 (g) for 30 min. After degassing, the reaction flask was maintained at 270 °C under 1 atm of N2 (g). Tris(trimethylsilyl)phosphine of 0.60 mL and trioctylphosphine (TOP) of 6.0 mL were mixed in a glove box at room temperature, and the mixture was rapidly injected into the reaction flask using a microsyringe, and then temperature of the reaction flask was temporally dropped by the injection of the mixture. The reaction flask was then maintained at 280 °C for several days. Butanol was added to the final reaction mixture, whose color then became brown. The dark brown solution was centrifuged to separate the residual products of InP nanoparticles, which were subsequently added with methanol, precipitated, and separated again by centrifugation. For the synthesis of core–shell InP@ZnSe nanoparticles, both diethylzinc and trioctylphosphine selenide were saturated in TOP in a glove box. The mixture solution was taken in a microsyringe and injected into a hot TOPO solution of 6.0 mL containing InP nanoparticles, which had already been heated to 200 °C under 1 atm of N2 (g), at a rate of six drops per 10 min under vigorous stirring. Product nanoparticles in the reaction flask were annealed at 90 °C for a day under vigorous stirring, and separated by centrifugation. Colloidal samples were prepared by suspending InP nanoparticles or core–shell InP@ZnSe nanoparticles in n-hexane immediately prior to use. 2.2. Methods High-resolution transmission electron microscopy (HRTEM) images were measured employing a microscope (JEOL, JEM-3000F) while high-resolution X-ray diffraction (HRXRD) patterns were recorded with a diffractometer (Bruker, D8 DISCOVER) using Cu Ka radiation (k = 0.154178 nm). Absorption spectra were obtained with a UV/vis spectrometer (Scinco, S-3100), and photoluminescence spectra were detected using a home-built fluorimeter consisting of a Xe lamp of 75 W (ARC, XS432) and monochromators of 0.15 m and 0.30 m (ARC, Spectropro 150 and 300, respectively). Picosecond photoluminescence kinetic profiles were measured employing a streak camera of 10 ps (Hamamatsu, C2839) attached with a CCD (Princeton Instruments, RTE128H) with excitation using a modelocked Nd:YAG laser of 25 ps (Quantel, YG501). In addition, nanosecond photoluminescence kinetic profiles were measured with a photomultiplier tube (Hamamatsu, R928) connected an oscilloscope of 1 GHz (Lecroy, Wavepro 950) with excitation using a Q-switched Nd:YAG laser of 6 ns (Quantel, Brilliant). Kinetic constants were extracted by fitting measured kinetic profiles to computer-simulated exponential curves convoluted with instrument response functions. Nanosecond time-resolved luminescence spectra were measured with an intensified CCD (Princeton Instruments, ICCD576G), having a gating resolution of 2 ns and attached to a 0.5 m spectrometer (Acton Research, Spectrapro 500), with excitation from a Q-switched Nd:YAG laser (Quantel, Brilliant) of 6 ns. Both the laser and the intensified CCD were triggered by a pulse/delay generator (Stanford Research Systems, DG535). 3. Results and discussion The HRTEM image of Fig. 1 shows that our fabricated core–shell InP@ZnSe quantum dots have diameters of about 3 nm. Every
Fig. 1. HRTEM images of core–shell InP@ZnSe quantum dots. All three nanoparticles enclosed in white dashed circles having diameters of 3 nm show a latticefringe distance of 0.34 nm.
nanocrystal in white dashed circles shows a lattice-fringe distance of 0.34 nm, which agrees well with the spacing of 0.3388 nm between the (1 1 1) planes of the zincblende InP structure. This suggests that the ZnSe shells of the core–shell nanocrystals are very thin. We have estimated that the thickness of the ZnSe shells is approximately one monolayer. The nanoparticles have been investigated with powder HRXRD. The HRXRD patterns of bare InP nanocrystals and core–shell InP@ZnSe nanocrystals are shown in Fig. 2. On one hand, the HRXRD patterns of core–shell InP@ZnSe nanocrystals (with dotted lines as eye guidances in order to show how well peaks match with bulk bar patterns), bulk zincblende InP (denoted as bars on the bottom), and bulk zinc–blende ZnSe (denoted as bars on the bottom) are shown in Fig. 2a. The peaks of {(1 1 1), (2 2 0), and (3 1 1)} coming from the core InP and the nearby peaks of {(1 1 1), (2 2 0), and (3 1 1) marked with arrows} due to the shell ZnSe show up definitely. Thus, this indicates clearly that our InP@ZnSe nanocrystals have a core–shell structure. On the other hand, the HRXRD patterns of InP nanocrystals and bulk zincblende InP (denoted as bars on the bottom) are shown in Fig. 2b. The characteristic peaks of InP nanoparticles agree well with those of the bulk zincblende InP. We consider that the passivation of the core InP nanocrystals with the ZnSe shells has been carried out uniformly with a minor atomic level reconstruction of the surface because the lattice mismatch of the lattice parameters between zincblende InP and zincblende ZnSe is insignificant [29,31]. The mean crystallite diameter, d, of the small nanocrystals can be determined from the line width of a HRXRD spectrum by the Scherrer’s formula: hdi = 0.94k/(Bcos hB), where k is the employed X-ray wavelength, B is the full width at half maximum of the diffraction peak (radian), and hB is the half angle of the diffraction peak on the 2h scale. The mean crystalline diameter of InP quantum dots has been estimated to be 2.2 nm, which agrees approximately with the mean core diameter of core–shell InP@ZnSe quantum dots observed with Fig. 1. On one hand, Fig. 3 indicates that the absorption spectra of bare InP nanoparticles and core–shell InP@ZnSe nanoparticles dispersed in n-hexane show characteristic excitonic transition peaks at 600 nm. The absorption band edge of core–shell InP@ZnSe quantum dots seems to be shifted slightly to the blue compared to that of bare InP quantum dots because passivation has increased spectral broadening and scattering significantly [25,26]. On the other
M.R. Kim et al. / Journal of Colloid and Interface Science 350 (2010) 5–9
Fig. 2. (a) HRXRD patterns of core–shell InP@ZnSe nanocrystals (red), bulk zinc– blende InP (green), and bulk zinc–blende ZnSe (orange). (b) HRXRD patterns of InP nanocrystals (blue) and bulk zinc–blende InP (green). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
7
ences potential confinement and tends to localize in the core nanocrystal because the conduction and the valence bands of the core material are lower and higher in energy than the respective ones of the shell material. However, the photoluminescence of our prepared core–shell InP@ZnSe nanoparticles is blue-shifted slightly because the photoluminescence of core–shell InP@ZnSe nanoparticles is less contributed by defect emission than that of bare InP nanoparticles is. The narrow band-edge emission of core–shell InP@ZnSe at 620 nm suggests that the core InP nanoparticles are almost monodisperse in size. The passivation of InP nanocrystals with ZnSe improves the emission quantum yields of the core nanocrystals enormously to increase the band-edge photoluminescence by 6.8 times by reducing defect sites at the surfaces of InP quantum dots. Note that the passivation of InP nanocrystals with ZnS has been reported to increase the quantum yield of photoluminescence by a factor of 2 [17]. The defect emission of core–shell InP@ZnSe nanoparticles at 700–750 nm is shifted on a large scale to the blue from that of bare InP nanoparticles at 750–800 nm [32–36]. Furthermore, the ratio of defect emission to band-edge emission in core–shell nanocrystals is much smaller than that in bare nanocrystals. These also indicate that passivation with ZnSe has blocked efficiently the ensnaring of electrons in the conduction band of a InP nanocrystal into trap sites to enhance band-edge emission with decreasing defect emission. We have deconvoluted each photoluminescence spectrum with two Gaussian functions to understand the nature of defect emission in both bare and core–shell nanocrystals. Whereas the deconvoluted defect emission of bare InP quantum dots has the maximum at 760 nm, that of core–shell InP@ZnSe quantum dots shows the peak at 710 nm. This also supports that passivation with ZnSe reduces the trap sites and surface states of InP quantum dots. Fig. 4 reveals that the photoluminescence kinetic profiles of core–shell InP@ZnSe nanoparticles suspended in n-hexane show three exponential-decay components of 0.1 ns (28%), 0.8 ns (38%), and 7 ns (34%) as indicated in Table 1. The decay time of 0.1 ns is considered to result from the capturing time of photoexcited electrons at the conduction band into shallow traps. While the time constant of 0.8 ns is ascribed to arise mainly from the ensnaring time of electrons in shallow traps into deep traps, the decay time of 7 ns is assigned to the trapping time of electrons in deep traps
Fig. 3. Absorption (dotted) and photoluminescence spectra (solid) of bare InP nanoparticles (blue) and core–shell InP@ZnSe nanoparticles (red) suspended in nhexane. Samples were excited at 500 nm for the photoluminescence spectra, and each photoluminescence spectrum was deconvoluted with two Gaussian functions (green). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
hand, the photoluminescence spectra of bare InP and core–shell InP@ZnSe nanoparticles in n-hexane show band-edge emission peaks at 630 nm and 620 nm, respectively. In general, the electrons and the holes of a type-I core–shell semiconductor material experi-
Fig. 4. Picosecond luminescence kinetic profiles, measured in short (a) and long time windows (b), of core–shell InP@ZnSe nanoparticles dispersed in n-hexane. The sample was excited at 532 nm and monitored at 600 nm.
8
M.R. Kim et al. / Journal of Colloid and Interface Science 350 (2010) 5–9
Table 1 Luminescence kinetic time constants of core–shell InP@ZnSe nanoparticles in n-hexane, extracted from Figs. 4 to 5.
a b
Figures
kmon (nm)a
Rise time (ns)
Decay time (ns)
Fig. 4 Fig. 5
600 800
Instant 60
0.1 (28%)b + 0.8 (38%) + 7 (34%) 1100
Monitored luminescence wavelength. Initial amplitude percentage of each component.
into even deeper traps. The photoluminescence of zincblende InP nanowires has been reported to have a decay time of 5.0–8.4 ns [37], which is similar to the relaxation time of electrons in the deep trap sites of core–shell InP@ZnSe quantum dots. The nanosecond luminescence kinetic profiles of Fig. 5, monitored at 800 nm, show that the surface-state photoluminescence of InP@ZnSe nanoparticles rises on the time scale of 60 ns and decays exponentially on the time scale of 1100 ns (Table 1). It is noteworthy that the rise kinetics of photoluminescence at trap sites or surface states is very difficult to be measured by a time-resolved spectroscopic method in general. The rise time of 60 ns, attributed to the trapping time of electrons at deeper traps into surface states, suggests that diverse trap sites such as deeper traps exist at interfaces between InP cores and ZnSe shells. Finally, electrons ensnared at surface trap states decay radiatively into the valence band of InP core quantum dots by electron–hole recombination on the slow time scale of 1100 ns. The time-resolved luminescence spectra of InP@ZnSe nanoparticles dispersed in n-hexane obtained at different time delays after excitation (Fig. 6) have led us to measure a trap-site emission spectrum and a surface-state emission spectrum separately. However, we could not obtain a time-resolved spectrum of exclusive bandedge luminescence with our current temporal resolution of 6 ns because photoexcited electrons at the conduction band are captured into trap sites on the time scale of 0.1 ns (Fig. 4). The emission spectrum of Fig. 6a reveals that emission measured at the short time delay after excitation is governed by the recombination of electrons in trap sites with holes in the valence band. Thus, the
Fig. 6. Trap-site (a) and surface-state luminescence spectra (b) of core–shell InP@ZnSe quantum dots dispersed in n-hexane, measured at time delays of 15 ns and 480 ns, respectively, after sample excitation at 532 nm using a Q-switched Nd:YAG laser of 6 ns.
emission spectrum of Fig. 6a having the maximum at 660 nm hardly contains either band-edge emission or surface-state emission. Trap-site emission is shifted extensively (by 40 nm) to the red from the band-edge emission spectrum of Fig. 3. Furthermore, the band width of trap-site emission is much broader (by 1.3 times) than that of band-edge emission. The broad band was considered to associate with the trap-to-band recombination of weakly coupled electron–hole pairs, where electrons were trapped at phosphorous vacancies and holes were located at the valence band. Thus, we suggest that the trap-site emission spectrum of Fig. 6a arises from the radiative relaxation of electrons at diversely different deep and deeper trap sites [33–36]. The emission spectrum of Fig. 6b having the maximum at 760 nm results from the relaxation of electrons at surface states into the valence band. It is noteworthy that the wavelength at the maximum of Fig. 6b is almost the same as the wavelength at the maximum of the deconvoluted trap emission of bare InP quantum dots in Fig. 3. The enormously large red shift (100 nm) of Fig. 6b from Fig. 6a suggests that the surface states of InP quantum dots are extremely different energetically and environmentally from the trap sites of InP quantum dots. We consider that this has enabled us to measure the formation kinetics of the surface-state photoluminescence of core–shell InP@ZnSe nanocrystals. 4. Summary
Fig. 5. Nanosecond luminescence kinetic profiles, measured in short (a) and long time windows (b), of core–shell InP@ZnSe nanoparticles dispersed in n-hexane. The sample was excited at 532 nm and monitored at 800 nm.
InP quantum dots of 3 nm in diameter have been prepared via hot injection synthesis and passivated with ZnSe to enhance photoluminescence by 6.8 times. Core–shell InP@ZnSe quantum dots dispersed in n-hexane have then been investigated using time-resolved spectroscopy to understand their photoluminescence dynamics. Fig. 7 summarizes the dynamics and spectroscopy of the photoluminescence of core–shell InP@ZnSe nanoparticles dispersed in n-hexane. Photoexcited electrons in the conduction band are trapped into shallow traps within 0.1 ns. On one hand, while deep traps capture electrons from shallow traps in 0.8 ns, their electrons are ensnared into deeper traps on the time scale of 7 ns. On the other hand, the photoluminescence of surface states rises in 60 ns and decays on the time scale of 1100 ns. Combining
M.R. Kim et al. / Journal of Colloid and Interface Science 350 (2010) 5–9
Fig. 7. Schematic diagram for the luminescence relaxation kinetics of core–shell InP@ZnSe quantum dots dispersed in n-hexane. All the time constants are given in units of nanoseconds.
time-resolved photoluminescence spectra with a static luminescence spectrum, we have separated the individual spectra of trap-site emission at 660 nm and surface-state luminescence at 760 nm from the spectrum of band-edge photoluminescence at 620 nm. Acknowledgments This work was supported by research Grants through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science, and Technology (2009-0071184 and 20090082846). D.J.J. is also thankful to the SRC program of NRF (2010-0001635) while M.R.K. acknowledges a BK21 scholarship as well. S.L. also acknowledges the NANO Systems Institute, National Core Research Center, KIER, and NRF for artificial atoms research. References [1] Z. Tang, Z. Zhang, Y. Wang, S.C. Glotzer, N.A. Kotov, Science 314 (2006) 274.
9
[2] S.Y. Park, A.K.R. Lytton-Jean, B. Lee, S. Weigand, G.C. Schatz, C.A. Mirkin, Nature 451 (2008) 553. [3] M.R. Kim, D.-J. Jang, Chem. Commun. (2008) 5218. [4] J.-Y. Kim, S. Lee, K.-H. Yoo, D.-J. Jang, Appl. Phys. Lett. 94 (2009) 153301. [5] M.R. Kim, S.Y. Heo, D.-J. Jang, J. Colloid. Interface Sci. 339 (2009) 217. [6] M.A. El-Sayed, Acc. Chem. Res. 37 (2004) 326. [7] W.K. Bae, J. Kwak, J.W. Park, K. Char, C. Lee, S. Lee, Adv. Mater. 21 (2009) 1690. [8] M.S. Gudiksen, L.J. Lauhon, J. Wang, D.C. Smith, C.M. Lieber, Nature 415 (2002) 617. [9] X. Peng, J. Wickham, A.P. Alivisatos, J. Am. Chem. Soc. 120 (1998) 5343. [10] S. Mokkapati, C. Jagadish, Mater. Today 12 (2009) 22. [11] M.C. Beard, G.M. Turner, J.E. Murphy, O.I. Micic, M.C. Hanna, A.J. Nozik, C.A. Schmuttenmaer, Nano Lett. 3 (2003) 1695. [12] S. Hussain, N. Won, J. Nam, J. Bang, H. Chung, S. Kim, ChemPhysChem 10 (2009) 1466. [13] S. Kim, W. Shim, H. Seo, J.H. Bae, J. Sung, S.H. Choi, W.K. Moon, G. Lee, B. Lee, S.W. Kim, Chem. Commun. (2009) 1267. [14] D.W. Lucey, D.J. MacRae, M. Furis, Y. Sahoo, A.N. Cartwright, P.N. Prasad, Chem. Mater. 17 (2005) 3754. [15] R. Xie, D. Battaglia, X. Peng, J. Am. Chem. Soc. 129 (2007) 15432. [16] J. Wang, M.S. Gudiksen, X. Duan, Y. Cui, C.M. Lieber, Science 293 (2001) 1455. [17] S. Xu, J. Ziegler, T. Nann, J. Mater. Chem. 18 (2008) 2653. [18] D.V. Talapin, N. Gaponic, H. Horchert, A.L. Rogach, M. Haase, H. Weller, J. Phys. Chem. B 106 (2002) 12659. [19] R.L. Woo, L. Gao, N. Goel, M.K. Hudait, K.L. Wang, S. Kodambaka, R.F. Hicks, Nano Lett. 9 (2009) 2207. [20] S. Xu, S. Kumar, T. Nann, J. Am. Chem. Soc. 128 (2006) 1054. [21] S. Adam, D.V. Talapin, H. Borchert, A. Lobo, C. McGinley, A.R.B. de Castro, M. Haase, H. Weller, T. Möller, J. Chem. Phys. 123 (2005) 084706. [22] A. Narayanaswamy, L.F. Feiner, P.J. van der Zaag, J. Phys. Chem. C 112 (2008) 6775. [23] K. Adlkofer, E.F. Duijs, F. Findeis, M. Bichler, A. Zrenner, E. Sackmann, G. Abstreiter, M. Tanaka, Phys. Chem. Chem. Phys. 4 (2002) 785. [24] L. Li, P. Reiss, J. Am. Chem. Soc. 130 (2008) 11588. [25] E. Ryu, S. Kim, E. Jang, S. Jun, H. Jang, B. Kim, S.-W. Kim, Chem. Mater. 21 (2009) 573. [26] S. Haubold, M. Haase, A. Kornowski, H. Weller, ChemPhysChem 2 (2001) 331. [27] C. Li, M. Ando, H. Enomoto, N. Murase, J. Phys. Chem. C 112 (2008) 20190. [28] M.R. Kim, Y.-M. Kang, D.-J. Jang, J. Phys. Chem. C 111 (2007) 18507. [29] P. Reiss, M. Protière, L. Li, Small 5 (2009) 154. [30] J.J. Li, Y.A. Wang, W. Guo, J.C. Keay, T.D. Mishima, M.B. Johnson, X. Peng, J. Am. Chem. Soc. 125 (2003) 12567. [31] S.Zh. Karazhanov, L.C. Lew Yan Voon, Semiconductors 39 (2005) 161. [32] A.V. Maleev, I.V. Ignatiev, I.Ya. Gerlovin, I.E. Kozin, Y. Masumoto, Phys. Rev. B 71 (2005) 195323. [33] J.L. Blackburn, D.C. Selmarten, R.J. Ellingson, M. Jones, O. Micic, A.J. Nozik, J. Phys. Chem. B 109 (2005) 2625. [34] S.-H. Kim, R.H. Wolters, J.R. Heath, J. Chem. Phys. 105 (1996) 7957. [35] Y. Zhao, Z. Dong, S. Miao, A. Deng, J. Yang, B. Wang, J. Appl. Phys. 100 (2006) 123519. [36] L. Langof, L. Fradkin, E. Ehrenfreund, E. Lifshitz, O.I. Micic, A.J. Nozik, Chem. Phys. 297 (2004) 93. [37] K. Pemasiri, M. Montazeri, R. Gass, L.M. Smith, H.E. Jackson, J. Yarrison-Rice, S. Painam, Q. Gao, H.H. Tan, C. Jagadish, X. Zhang, J. Zou, Nano Lett. 9 (2009) 648.