One-Step Preparation of Strongly Luminescent and ... - ACS Publications

2 downloads 85146 Views 3MB Size Report
Aug 9, 2010 - Namgu, Pohang, 790-784, South Korea, and Samsung AdVanced Institute ... annealing.13-18,20,21,23-25 Recently, Seo and co-workers have.
14362

J. Phys. Chem. C 2010, 114, 14362–14367

One-Step Preparation of Strongly Luminescent and Highly Loaded CdSe Quantum Dot-Silica Films Sanghwa Jeong,† JinSik Lee,† Jutaek Nam,† Kyuhyun Im,‡ Jaehyun Hur,‡ Jong-Jin Park,‡ Jong-Min Kim,‡ Bonghwan Chon,† Taiha Joo,† and Sungjee Kim*,† Department of Chemistry, Pohang UniVersity of Science and Technology, San 31, Hyojadong, Namgu, Pohang, 790-784, South Korea, and Samsung AdVanced Institute of Technology, Mt. 14-1, Nongseo-Dong, Giheung-Gu, Yongin-Si, Gyeonggi-Do 446-712, South Korea ReceiVed: April 22, 2010; ReVised Manuscript ReceiVed: July 12, 2010

Strongly luminescent and highly loaded CdSe quantum dot (QD)-silica composite films are prepared by a simple one-step method. QDs are grown in situ as the silica matrix forms from spin-on-glass organosiloxane polymers. The nucleation and growth of QDs in the silica matrix is directly observed using TEM. The QD composites are comprehensively characterized, which includes photoluminescence(PL) decay, PL efficiency, and QD volume fraction. The QD emission can span the entire visible range and can be tuned by the preparation condition. The PL of QDs is very strong, reaching the quantum efficiency up to 35%. Very high QD loading is achieved with the volume fraction as high as 13%, which makes our composite the highest loaded QD-silica composite material. We further showcase the versatile applicability of our method by a QD-silica composite capped down-conversion pseudo white LED. 1. Introduction Colloidal semiconductor nanocrystal quantum dots (QDs) are being extensively studied to exploit their unique properties, including the zero-dimensional discrete electronic energy feature, high photostability, tunable and bright emission, and solution processability. They can promise many potential applications, including biological fluorescent probes,1,2 light-emitting diodes (LEDs),3 optical modulators,4 lasers,5 and solar cells.6,7 Before QDs can be actively applied to many device applications, they often need to be stably embedded in an appropriate matrix. Polymers8-12 or inorganic matrixes13-25 can be used to fabricate QD-embedded composites. QD-silica composites through various synthetic routes have been reported because silica is attractive as a transparent and nonscattering matrix and is wellsuited for many optical components of waveguides, lasers, optical filters, and modulators.13-18,20,21,23-27 QD-silica composites can be typically prepared by compatibilizing QDs with silica sol precursors, which is followed by gelation and annealing.13-18,20,21,23-25 Recently, Seo and co-workers have reported bright QD-silica monoliths by using their simultaneous ligand-exchange and shell formation method.23 Brock and coworkers have also demonstrated highly luminescent QD monoliths using an arrested precipitation strategy.24 In general, stabilizing QDs into solid matrixes typically requires multistep processes. The compatibilization needs surface modification of QDs and/or addition of surfactants to avoid aggregation in the silica sol precursor medium. This step can risk losing the high photoluminescense (PL) quantum efficiency (QE) of as-prepared QDs. In addition, surface molecules on QDs or encapsulating surfactants limit obtaining high QD loading fractions as significantly occupying volumes in the QD composite medium. The QD volume fraction can be a critical factor in many * To whom correspondence should be addressed. E-mail: sungjee@ postech.ac.kr. † Pohang University of Science and Technology. ‡ Samsung Advanced Institute of Technology.

applications, including inventory control and lasing devices. For example, multicolor QD beads can only guarantee their full multiplexing power when they can reach a high QD loading fraction.28 The QD volume fraction is pivotal for lasing applications due to the fast Auger process.29 On the other hand, QD-silica composites can be also prepared by melting semiconductor precursors in a glass host.15,16 However, this meltingpot process has an inherent limitation to reach a high volume fraction of QDs because of the low solubility of semiconductors in glass. In addition, the melting-pot process is energy-inefficient as typically demanding a very high temperature of ∼1600 °C. We report herein a simple and rapid method to prepare QD-silica composites by using QD precursors and spin-onglass (SOG) organosiloxanes. Highly fluorescent QD-silica films can be obtained by mixing QD precursor solutions with SOG, spin-casting, and baking at a relatively low temperature of ∼240 °C. QDs are in situ synthesized concurrently with the silica matrix formation. The QD-silica composites are strongly luminescent with the QE as high as reaching up to 35%. The QD loading fraction can be also very high, reaching up to 13%, which makes our composite the highest loaded QD matrix, as far as the authors know. This QD loading fraction is extremely high, at least an order of magnitude larger than conventional QD composites. 2. Experimental Section 2.1. Procedures for QD Precursors. All chemicals were used without further purifications. Precursor preparations were carried out under an inert atmosphere. The Cd precursor was prepared by dissolving cadmium acetate dihydrate (98%, SigmaAldrich) in methanol (Mallinckrodt). The Cd concentration was adjusted to be 2 M. The Se precursor was prepared by complexing selenium (pellets, 99.999+%, Sigma-Aldrich) with excess tris(3-hydroxypropyl)phosphine (THPP) (HISHICOLIN P-540, Nippon Chemical Industrial) in ethanol (anhydrous, Sigma-Aldrich). The Se precursor solution contained 2 M selenium in 1:1 (vol/vol) THPP and ethanol.

10.1021/jp103617m  2010 American Chemical Society Published on Web 08/09/2010

Preparation of CdSe QD-Silica Films

J. Phys. Chem. C, Vol. 114, No. 34, 2010 14363

Figure 1. (a) Representative QD-silica composite films on glass substrates under UV illumination. (b) Size series of normalized emission spectra.

2.2. Fabrication of QD-Silica Composite Films. Glass slides (Microslide 2947-3X1, Corning) were cut to 2.5 cm by 2.5 cm for the substrates. The glass substrate was washed with Alconox and sonicated in acetone for 30 min and in ethanol for another 5 min. It was rinsed with distilled water and dried under a nitrogen blow. For a typical fabrication, the Cd and Se precursor solutions prepared above were added and mixed well in SOG solution (Accuglass T-512B from Honeywell, Morristown, NJ). The volume ratio of Cd precursor solution/Se precursor solution/SOG solution was made to be 1:1:3. The mixture was spin-cast onto the glass substrate with 2000 rpm for 30 s. The substrate was baked at ∼240 °C in a tube furnace under a continuous nitrogen flow. The baking time was adjusted for a few minutes up to hours depending on the desired emission color. 2.3. Optical Measurements. Absorption spectra were obtained using an Agilent UV-vis spectrophotometer 8453. PL spectra were taken by an Ocean Optics USB4000-UV-vis spectrophotometer using a 385 nm LED (SU50-RUP501AT385, Seoul Semiconductor Co., LTD) as the excitation source. PL QEs were obtained by an absolute measurement method using an integrating sphere (4P-GPS-033-SL, Labsphere) connected by a Jobin Yvon Horiba Fluorolog-2 system with a Symphony Si CCD. Mello’s method was used for the QE measurements.30 The QE measurement apparatus setting was built by following the publication by Beeby and co-workers.30 All the components inside of the integrating sphere were made of white Teflon to minimize the light absorption. 2.4. TEM, EDS, and AFM Characterization. A JEOL JEM-1011 was used for transmission electron microscopy (TEM) images and energy-dispersive spectroscopy (EDS), and 400 mesh copper grids (Ted Pella 01822, Inc., Redding, CA) were used. For TEM sample preparation, the TEM grid was placed on top of a glass slide center. The QD precursor SOG solution was spin-cast on the glass slide, and the substrate was baked in the oven. The TEM grid with a very thin QD-silica composite layer on top was retrieved from the substrate using a tweezer. For AFM characterization, the film thickness of the QD-silica composites on the glass slides was determined using a Dimension 3100 (VEECO) in tapping mode. To reveal the film thickness by AFM, QD-silica composite films were scratched by a razor. 2.5. Time-Resolved Photoluminescence Experiment. For time-resolved PL experiments, the light source was by homebuilt cavity-dumped mode-locked Ti:sapphire laser pulses. The excitation wavelength was 400 nm, which was doubled by the fundamental light (800 nm, 380 kHz, 20 fs) with BBO 100 µm. The 101.6 mm parabolic mirror was used for collecting the emission in the “backscattering” geometry. The detecting system consisted of a CCD (Andor, DU401A-BV) for time-integrated

PL (TIPL) and an MCP-PMT (Hamamatsu, R3809-U51) for PL. Time-resolved photoluminescence (TRPL) was monitored via time-correlated single-photon counting, which provides the sub-10 ps time resolution with deconvolution. The abovementioned setup was previously reported.31 2.6. Fabrication of a Down-Conversion Pseudo White LED. A 385 nm emitting LED (SU50-RUP501AT385, Seoul Semiconductor Co., LTD) was dip-coated into the QD precursors and SOG mixture solution for 10 s. After the dip-coating, the LED was dried in air for a minute and baked at ∼200 °C in a tube furnace under a continuous nitrogen flow. 3. Results and Discussion QD-silica composites can be prepared by a simple and mild one-step mix-and-bake procedure. Figure 1a shows representative CdSe QD-silica composite films on glass substrates. The films are flat and smooth and look glossy with strong fluorescence under a UV lamp. Figure 1b shows a size series of emission spectra from typical CdSe QD-silica composite films. The emissions cover almost the entire visible range, with the PL bandwidth as narrow as 60 nm. For a typical preparation of the QD-silica composite films, QD precursors are mixed with SOG, spin-cast, and baked in an oven. For the Cd precursor, Cd salts, such as cadmium acetate, are dissolved in methanol. For the Se precusor, a Se-THPP complex was made and dissolved in ethanol. The QD precursors are dissolved in alcohols because the SOG solution is well miscible with them. THPP has a phosphine functional group that can easily form a Se-P complex bond. It also has three hydroxy functional groups that can ensure alcohol solubility and the potential ability to incorporate into SOG silica networks by forming cross-linked Si-O-C bonds. QD precursor solutions and SOG solution mixture can be spin-cast or molded by necessity and baked to obtain the QD-silica composites. QDs are being synthesized concurrently as the SOG matrix forms. The QD precursors are readily soluble in the SOG solution. The Cd acetate and Se-THPP complex can decompose at relatively low temperatures. As the pyrolysis of the QD precursors proceeds, the chemical changes of the QD precursors result in a rapid reduction in solubility. The abrupt solubility drops trigger nucleations of QDs as the precursor concentration overshoots the minimum concentration for nucleation. At the initial growth stage, the QD growth occurs concurrently with nucleation until the effective precursor concentration drops below the concentration for nucleation. QDs continue to grow presumably mostly by Ostwald ripening. The growth is usually quenched by cooling as the substrate is removed from the oven. By using the pyrolysis method, our process can be performed in remarkably lower temperatures when compared with the glass melting process.

14364

J. Phys. Chem. C, Vol. 114, No. 34, 2010

Jeong et al.

Figure 2. (a) Absorption and normalized emission spectra of QD-silica composites: samples prepared at different baking temperatures of 210 °C (solid lines), 240 °C (dashed lines), and 270 °C (dotted lines) using the baking time of 15 min and a 1:1 Cd-to-Se precursor ratio. (b) Absorption and normalized emission spectra of QD-silica composites: samples prepared for different baking times of 10 min (solid lines), 15 min (dashed lines), and 30 min (dotted lines) using the baking temperature of 240 °C and a 1:1 Cd-to-Se precursor ratio. (c) Absorption and normalized emission spectra of QD-silica composites: samples prepared using different Cd-to-Se precursor ratios of 1:1 (solid lines), 3:2 (dotted lines), and 2:3 (dashed lines) using the baking temperature of 240 °C and the baking time of 15 min. (d) PL decays of QD-silica composites emitting at 514 nm (black), 535 nm (green), 556 nm (blue), and 608 nm (red).

TABLE 1: PL Decay Life Times τ and the Corresponding Amplitude Percentages A for Samples in Figure 2d.a χ2 is the Reduced Chi-Square Value for Each τavg sample no.

emission peak [nm]

A1

τ1 [ns]

A2

τ2 [ns]

A3

τ3 [ns]

A4

τ4 [ns]

τavg [ns]

χ2

1 2 3 4

514 535 556 608

64.2 76.8 60.4 35.3

0.075 0.051 0.096 0.19

21.9 9.0 19.7 20.3

1.32 1.80 2.49 4.09

10.9 7.9 10.7 21.6

10.4 17.3 20.1 32.0

3.0 6.4 9.2 22.8

83.4 93.1 117 141

3.98 7.49 13.4 40.0

1.29 1.19 1.18 1.18

a

Each decay was fitted using a quadro-exponential function.

As shown in Figure 1, the QD-silica composite emission can span the entire visible range and can be tailored by the preparation conditions. Figure 2 shows absorption and normalized emission spectra of QD-silica composites with changing the preparation conditions, such as baking temperature, baking time, and precursor ratio between Cd and Se. Figure 2a shows the dependency upon the baking temperature. Other parameters, such as baking time and precursor ratio, are maintained identically. As the baking temperature increases, we find the absorption and emission profiles shift to longer wavelengths. The emission peak shifted from ∼500 to ∼700 nm as the baking temperature increased from 210 to 270 °C. Higher temperatures causes enhanced mobility of the QD precursors and SOG siloxane polymers. This can promote faster growth of QDs, which results in the red shift in absorption and emission. However, the absorption peak becomes significantly broader as it red shifts. It is believed that emission from trap sites increases at higher baking temperatures. The absorption shift is mostly due to the increased QD size. On the other hand, the emission shift is contributed by the increased QD size and increased trap emissions. The broad PL profile results from the convolution of trap emissions and the inhomogeneity from the QD size distribution. Insufficient surface passivating organic ligands may have contributed to the trap emissions. The low mobility of the QD precursors in the SOG matrix may have limited the QDs to grow more uniformly in size. Figure 2b shows samples that are prepared using different baking times. As the baking extends, absorption and emission shift to longer wavelengths. The emission peak shifted from ∼500 to ∼600 nm when the baking time was extended from 10 to 30 min. The precursor ratio between Cd and Se was investigated in Figure 2c. The absolute concentration of the limiting reagent was kept constant. We found that a 3:2 Cd-to-Se precursor ratio produced larger QDs than the cases using a 1:1 or 2:3 ratio. As the ratio of Cd to Se increases, absorption appears at longer wavelengths, which indicates faster particle growth (Figure 2c). A cadmium excess condition seems to promote rapid QD growth. This has been

reported for solution-based colloidal CdSe QD syntheses as supposedly Cd precursors have a lower activation energy for QD surface binding than Se precursors.32 When the preparation conditions are modified, different colored QD-silica composites can be obtained. To elucidate the origin of the PL, time-resolved PL measurements were performed for different colored QD-silica composites (Figure 2d). Intensity-weighted average lifetimes were obtained using quadro-exponential fittings.31 Lifetimes and amplitudes for each element are listed in Table 1. Four samples with their emission peaks at 514, 535, 556, and 608 nm were used. They showed mean decay lifetimes of 4.0, 7.5, 13.4, and 40.0 ns, respectively (Table 1). The decay lifetimes are comparable to those of typical colloidal QDs synthesized in solution or in glass.21,33 In addition, our samples have the PL QEs that are comparable to those of colloidally synthesized QDs. This corroborates the hypothesis that our sample PL mostly originates from the band-edge excitonic emissions. We find that the decays become slower for the longer emission wavelength samples. The prolonged decay also accords with colloidally synthesized QDs, which is due to the well-known QD size effect.34,35 Our QD-silica composites glow brightly when excited under a UV lamp. Their PL QEs were measured by an absolute method of Mello’s equation using an integrating sphere.30,36 They show strong fluorescence with a typical PL QE of ∼10%. The brightest sample reaches the PL QE up to 35%. Yellow emitting samples typically showed the highest PL QEs. PL QEs from different sized QDs fall in the same range. Small or large QDs tend to show lower PL QEs, presumably due to the lack of enough annealing for QD crystallinity or increased trap sites on the larger QD surface area. The PL QEs were moderately reproducible for samples prepared by an identical recipe, showing the typical deviation of (5% point. In-situ formed QDs typically show low PL QEs, presumably because of the lack of proper surface passivation. The strong fluorescence from our QD-silica composites can be ascribed to THPP acting as the QD surface ligand. Phosphine is a well-known QD passivating

Preparation of CdSe QD-Silica Films

J. Phys. Chem. C, Vol. 114, No. 34, 2010 14365

Figure 3. TEM images of quantum dot-silica composite films baked at 240 °C for 7 (a), 10 (b), and 15 min (c). Scale bars represent 50 nm. (d) TEM image of a QD-silica composite film baked at 240 °C for 10 min by using a 1:1 Cd-to-Se precursor ratio. Point A and point B are chosen for QD-rich and QD-sparce areas, respectively. (e) For points A and B, elemental analysis is performed using EDS. (f) For points A and B, the relative atomic compositions for major elements are tabled.

surface ligand, efficiently keeping QDs with high PL QEs.37 THPP has three terminal hydroxyl groups that may have participated in the silica gelation. It can be postulated that THPPs form stable networks between QD surfaces and the host silica matrix. The brightest QD-silica composites prepared from solution-synthesized QDs typically show the PL QEs of 3-10%.21,23-25 It is noted that our QD-silica composites are comparably bright when compared to the previously reported best samples. The lifetime measurements indicate that the PL of our QD-silica composites originates from QD excitonic emission. To obtain direct evidence of QD growth in our composite, the direct observation of QD growth in the composite was made by TEM. We have observed very thin QD-silica composite films that were baked for different times using TEM. For the TEM sample, a grid was placed on the center of the glass substrate. Spin-casting was made using a 1:1 Cd-to-Se ratio QD precursor-SOG mixture solution, and the baking temperature was set for 240 °C. The grid on top of the glass slide acts as an ascended substrate, which accommodates very thin film. To follow the time evolution of the QD growth, TEM images were taken for the samples using different baking times. Figure 3a shows the TEM image after 5 min of baking, where small and blurred particles of approximately 2 nm in size begin to appear (Figure 3a). At 7 min, QDs are visualized in a more distinctive fashion. They have grown to an average diameter of 3.4 nm in (Figure 3b). After 15 min, QDs have grown to an average diameter of 5.5 nm with the size distribution of 27% (Figure

3c). The size distribution is rather large when compared with colloidally synthesized QDs. However, the one-step process of QD composites (concurrently with the matrix solidification) inherently has to trade some degree of uniformity in QD size for the simple and rapid preparation. It is noted that the growth of QDs accompanies a decrease in the number density of the particles. This indicates that larger QDs are growing upon sacrifices of smaller ones. We further confirmed the existence of CdSe QDs by using EDS measurements. Figure 3d shows the TEM image of a QD silica composite that was synthesized using the condition of a 1:1 Cd-to-Se precursor ratio, baking temperature of 240 °C, and the baking time of 10 min. In the figure, points A and B were assigned for a QD-rich area and a QD-sparse area, respectively. EDS spectra can be found in Figure 3e with the relative atomic compositions for the major elements in Figure 3f. It is noted that Cd and Se are colocalized in the QD-rich area A. This indicates that the dark particles under TEM are indeed CdSe and not elemental Cd or Se. The QD-sparse area B has much less relative atomic compositions for Cd and Se than area A. The TEM images and EDS data show direct evidence for in situ formations of QDs in the silica matrix. Higher-magnification QD TEM images were not accessible due to the limitations from the silica matrix electron scattering. For QD composite applications, the loading fraction is a critical factor. The QD loading fraction plays an important role, especially in inventory control and lasing applications. QD composites often have low loading fractions with the volume

14366

J. Phys. Chem. C, Vol. 114, No. 34, 2010

Jeong et al.

Figure 4. (a) Absorption spectrum of a QD-silica composite film. (b) AFM field scan across a scratched region for film thickness measurement. (c) The depth profile of QD-silica film along the white line in (b) by AFM.

fraction less than 1%. When presynthesized QDs are added in a matrix resin for composite fabrication, a high loading fraction is limited by the phase separations or segregations. Typically, significantly volume occupying additives are necessary to overcome the difficulties, which inherently limit achieving a high QD fraction.9 QDs can be also embedded in a premade host matrix. For example, a polymer matrix can be swollen and invite QDs in.28 In this case, QDs usually reside near the surface of the matrix and fail to reach deep in the matrix, which makes the QD loading fraction very low. In our case, QDs are in situ synthesized with the matrix formation. QD precursors are provided in high concentrations because they are very miscible with the host SOG. By using in situ QD preparations, no additives are required to prevent phase separations. Therefore, our QD-silica composite preparation is suited for high QD loading. Loading fractions of our QD composite films were investigated using the QD absorption and the film thickness. Figure 4a shows absorption spectrum of a QD-silica composite film, where a high precursor concentration was used for maximal QD loading. The first excitonic absorption peak can be found at 416 nm, which corresponds to the average QD diameter of 1.1 nm. The QD concentration in the composite was determined using a CdSe QD molar extinction coefficient of 24 000 M-1 cm-1 at 350 nm. The extinction coefficient was borrowed from a previous report.38 For CdSe QDs, 350 nm is relatively high in energy when compared with the band gap. The CdSe electronic states converge close to the bulk continuum at 350 nm. As a result, the oscillator strength at 350 nm is quite insensitive to the QD size. In addition, QD precursors and SOG do not have a notable absorption at 350 nm. The film was measured to have the absorbance of 0.39 at 350 nm and the QD molar concentration of 0.30 mol L-1. The average thickness of the composite film was measured to be 550 nm using tapping mode AFM as swiping across a razor scratched line. Figure 4b shows the AFM image with the scratched area on the left and the film on the right. The depth profile across the white in Figure 4b is shown in Figure 4c. On the basis of the QD concentration, QD average size, and film thickness, the QD volume fraction was determined as 13%. This is a massive amount of QDs loaded in the silica matrix. In the case of close-packed QDs without any protective host matrix, the volume fraction was reported as ∼20%.10 A very high QD loading fraction that is comparable to closely-packed QD films can be achieved by our QD-silica composite preparation. The volumes occupied by organic molecules are minimized because no excess surface molecules or surfactants are used. Unlike the glass melting process, our method does not suffer from the limited solubility

Figure 5. (a) A QD-silica composite capped down-conversion white LED. (b) Emission spectra from a 385 nm LED (black) and from a QD-silica composite capped down-conversion pseudo white light LED (red).

of QD precursors. Large amounts of precursors can be dissolved in SOG solution, which results in the high QD packing density. However, when QDs occupy more than 5% volume fractions, PL QEs begin to significantly decrease because of the enhanced self-absorption and energy transfers between QDs. The low temperature of our process can enable QD-silica composite fabrication onto preprocessed devices. To demonstrate the versatility of our process, a down-conversion pseudo white LED was made by simply coating a QD-silica composite on a 385 nm GaN LED (see the Experimental Section for details). The epoxy packaged blue LED was dip-coated in the QD precursor SOG mixture solution and baked at 200 °C. A pseudo white color emission was clearly observed from the QD-silica capped LED (Figure 5a). A broad QD emission from ∼500 to ∼700 nm renders the white light illumination. Figure 5b is the pseudo white LED emission spectrum. This demonstrates that our QD-silica composite preparation can be easily applied to nonflat substrates. 4. Conclusions In conclusion, CdSe QD-silica composites were made by a simple one-step mixing and baking preparation. The QD emission can span the entire visible range and can be tuned by the preparation condition. The PL of QDs was very strong, reaching the QE of up to 35%. The QD volume fraction can reach as high as 13%. Growth of QDs in the silica matrix was confirmed by PL decay lifetime and TEM measurements. We further showcased the excellent applicability of our method by the QD-silica composite capped down-conversion pseudo white LED. Acknowledgment. This work was supported by the Korea Science and Engineering Foundation (KOSEF) grant funded by

Preparation of CdSe QD-Silica Films the Korea Government (MOST) (M10703001036-08M030003610, R0A-2008-000-20114-0(2008), and KRF-2008-331C00140) and the Priority Research Centers Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (20090094037 and 20090090897). References and Notes (1) Bruchez, M.; Moronne, M.; Gin, P.; Weiss, S.; Alivisatos, A. P. Science 1998, 281, 2013. (2) Bhang, S. H.; Won, N.; Lee, T.-J.; Jin, H.; Nam, J.; Park, J.; Chung, H.; Park, H.-S.; Sung, Y.-E.; Hahn, S. K.; Kim, B.-S.; Kim, S. ACS Nano 2009, 3, 1389. (3) Coe, S.; Woo, W. K.; Bawendi, M.; Bulovic, V. Nature 2002, 420, 800. (4) Bang, J.; Chon, B.; Won, N.; Nam, J.; Joo, T.; Kim, S. J. Phys. Chem. C 2009, 113, 6320. (5) Klimov, V. I.; Mikhailovsky, A. A.; Xu, S.; Malko, A.; Hollingsworth, J. A.; Leatherdale, C. A.; Eisler, H. J.; Bawendi, M. G. Science 2000, 290, 314. (6) Bang, J.; Park, J.; Lee, J. H.; Won, N.; Nam, J.; Lim, J.; Chang, B. Y.; Lee, H. J.; Chon, B.; Shin, J.; Park, J. B.; Choi, J. H.; Cho, K.; Park, S. M.; Joo, T.; Kim, S. Chem. Mater. 2009, 22, 233. (7) Huynh, W. U.; Dittmer, J. J.; Alivisatos, A. P. Science 2002, 295, 2425. (8) Zhang, H.; Cui, Z. C.; Wang, Y.; Zhang, K.; Ji, X. L.; Lu, C. L.; Yang, B.; Gao, M. Y. AdV. Mater. 2003, 15, 777. (9) Lee, J.; Sundar, V. C.; Heine, J. R.; Bawendi, M. G.; Jensen, K. F. AdV. Mater. 2000, 12, 1102. (10) Li, S.; Toprak, M. S.; Jo, Y. S.; Dobson, J.; Kim, D. K.; Muhammed, M. AdV. Mater. 2007, 19, 4347. (11) Zhang, H.; Wang, C. L.; Li, M. J.; Ji, X. L.; Zhang, J. H.; Yang, B. Chem. Mater. 2005, 17, 4783. (12) Skaff, H.; Sill, K.; Emrick, T. J. Am. Chem. Soc. 2004, 126, 11322. (13) Thoma, S. G. W. J. P.; Abrams, B. L. United States of America, 2008. (14) Chan, Y.; Steckel, J. S.; Snee, P. T.; Caruge, J. M.; Hodgkiss, J. M.; Nocera, D. G.; Bawendi, M. G. Appl. Phys. Lett. 2005, 86, 073102.

J. Phys. Chem. C, Vol. 114, No. 34, 2010 14367 (15) Guerreiro, P. T.; Ten, S.; Borrelli, N. F.; Butty, J.; Jabbour, G. E.; Peyghambarian, N. Appl. Phys. Lett. 1997, 71, 1595. (16) Lipovskii, A.; Kolobkova, E.; Petrikov, V.; Kang, I.; Olkhovets, A.; Krauss, T.; Thomas, M.; Silcox, J.; Wise, F.; Shen, Q.; Kycia, S. Appl. Phys. Lett. 1997, 71, 3406. (17) Kolobkova, E. V. Glass Phys. Chem. 2006, 32, 404. (18) Bodnar, I. V.; Gurin, V. S.; Molochko, A. P.; Solovei, N. P. Inorg. Mater. 2004, 40, 797. (19) Jasieniak, J.; Pacifico, J.; Signorini, R.; Chiasera, A.; Ferrari, M.; Martucci, A.; Mulvaney, P. AdV. Funct. Mater. 2007, 17, 1654. (20) Jasieniak, J.; Sada, C.; Chiasera, A.; Ferrari, M.; Martucci, A.; Mulvaney, P. AdV. Funct. Mater. 2008, 18, 3772. (21) Selvan, S. T.; Bullen, C.; Ashokkumar, M.; Mulvaney, P. AdV. Mater. 2001, 13, 985. (22) Tetsuka, H.; Ebina, T.; Mizukami, F. AdV. Mater. 2008, 20, 3039. (23) Wang, Q. B.; Iancu, N.; Seo, D. K. Chem. Mater. 2005, 17, 4762. (24) Arachchige, I. U.; Brock, S. L. J. Am. Chem. Soc. 2007, 129, 1840. (25) Yang, P.; Li, C. L.; Murase, N. Langmuir 2005, 21, 8913. (26) Iancu, N.; Sharma, R.; Seo, D. K. Chem. Commun. 2004, 2298. (27) Wang, Q. B.; Seo, D. K. Chem. Mater. 2006, 18, 5764. (28) Han, M. Y.; Gao, X. H.; Su, J. Z.; Nie, S. Nat. Biotechnol. 2001, 19, 631. (29) Klimov, V. I.; Mikhailovsky, A. A.; McBranch, D. W.; Leatherdale, C. A.; Bawendi, M. G. Science 2000, 287, 1011. (30) Porre`s, L.; Holland, A.; Pa˚lsson, L.-O.; Monkman, A.; Kemp, C.; Beeby, A. J. Fluoresc. 2006, 16, 267. (31) Chon, B.; Ryu, S. R.; Hong, Y.-J.; Yoo, J.; Yi, G.-C.; Joo, T.; Jung, Y. M. J. Mol. Struct. 2008, 883-884, 209. (32) Qu, L.; Peng, Z. A.; Peng, X. Nano Lett. 2001, 1, 333. (33) Wuister, S. F.; Donega, C. D.; Meijerink, A. J. Phys. Chem. B 2004, 108, 17393. (34) Donega, C. D.; Hickey, S. G.; Wuister, S. F.; Vanmaekelbergh, D.; Meijerink, A. J. Phys. Chem. B 2003, 107, 489. (35) Garrett, M. D.; Dukes, A. D.; McBride, J. R.; Smith, N. J.; Pennycook, S. J.; Rosenthal, S. J. J. Phys. Chem. C 2008, 112, 12736. (36) de Mello, J. C.; Wittmann, H. F.; Friend, R. H. AdV. Mater. 1997, 9, 230. (37) Kim, S.; Bawendi, M. G. J. Am. Chem. Soc. 2003, 125, 14652. (38) Leatherdale, C. A.; Woo, W. K.; Mikulec, F. V.; Bawendi, M. G. J. Phys. Chem. B 2002, 106, 7619.

JP103617M