Solvothermal synthesis and growth mechanism of ultrathin Sb₂Te₃ hexagonal nanoplates with thermoelectric transport properties

Abstract

Uniform single-crystalline hexagonal-shaped Sb₂Te₃ nanoplates with a thickness of 30–40 nm have been successfully synthesized by a glucose-assisted solvothermal process in the mixed solvents of ethanediamine and water. It is found that the reaction time, concentration of glucose, reaction temperature and the volume ratio of ethanediamine to water play important roles in the formation of the uniform Sb₂Te₃ nanoplates. Based on the experimental results, the possible reaction process and formation mechanism of these hexagonal nanoplates is proposed. Different from previous reports, the growth process of such Sb₂Te₃ nanoplates can be reasonably explained by a self-assembly process and an Ostwald ripening mechanism. The thermoelectric transport properties are investigated by measuring the electrical conductivity and the Seebeck coefficient in the temperature range of 300–600 K. The samples show much more enhanced Seebeck coefficients than that of bulk Sb₂Te₃. Meanwhile, the size of the sample has much impact on both the electrical conductivity and the Seebeck coefficient.

---

Introduction

In the past few decades, nanostructured materials of specific morphologies have attracted much attention due to their morphology and size dependent properties.^1–3 Until now, numerous nanomaterials with different morphologies and sizes, such as nanowires, nanotubes, nanoplates, nanosheets etc., have been fabricated.^4–9 And versatile synthetic strategies have been developed to achieve such nanomaterials, including thermolysis of various metallic precursors, solvothermal/hydrothermal processes, electrochemical deposition technique, chemical vapor deposition and the hot-injection methods. Among these methods, the solvothermal/hydrothermal process is one of the most popular methods for controllable synthesis of nanomaterials with different sizes and morphologies.

Antimony telluride (Sb₂Te₃) based compounds are very promising materials for room-temperature thermoelectric applications. Recent studies suggest that materials with nanostructures have higher figures-of-merit (ZT) due to both a high density of states and an increased phonon scattering or reduced lattice thermal conductivity in the nanosystems.^10,11 For instance, the p-type nanostructured thin film superlattices of Bi₂Te₃ and Sb₂Te₃, grown from chemical vapor deposition, have ZT values of about 2.4 at room temperature.¹² This result greatly stimulated interest in the synthesis of Sb₂Te₃ nanostructures with well-controlled morphologies and sizes. Therefore, Sb₂Te₃ based nanomaterials with different morphologies and sizes have been synthesized by various methods. For instance, previous work reported that Sb₂Te₃ hexagonal nanoplates had been fabricated through rapid microwave-assisted wet chemical¹³ and hydrothermal/solvothermal routes.^14,15 A vapor–liquid–solid growth method^16,17 and electrochemical deposition route have been employed to prepare arrays of Sb₂Te₃ nanowires.¹⁸ Sb₂Te₃ nanorings¹⁹ and nanobelts²⁰ have been synthesized by a simple hydrothermal process. Although many methods have been applied for the synthesis of Sb₂Te₃ hexagonal nanoplates, limited reports have been involved with the synthesis of hexagonal nanoplates with thicknesses less than 50 nm and the detailed formation mechanism.

Herein, we report the fabrication of Sb₂Te₃ hexagonal nanoplates through a facile solvothermal method with the assistance of glucose. Some factors affecting the morphologies of Sb₂Te₃ crystals, such as reaction time, concentration of glucose, reaction temperature, and the volume ratio of ethanediamine to water are systematically investigated. On the basis of time-dependent SEM observations, a formation mechanism different from those previously reported for such hexagonal nanoplate morphology is proposed, which is related to self-assembly and an Ostwald ripening process. In addition, the diameter and thickness of the nanoplates can be tuned by adjusting the concentration of glucose. Moreover, the size-dependant thermoelectric transport properties are clearly demonstrated, which present much enhanced Seebeck coefficients as compared to that of bulk Sb₂Te₃.

Experimental section

In this work, all the solvents and chemicals were purchased from Shanghai Chemical Reagent Co. and used without further purification unless otherwise stated.

Preparation of Sb₂Te₃ nanoplates

In a typical procedure, 0.135 g of antimony potassium tartrate (K(SbO)C₄H₄O₆·0.5H₂O), 0.135 g of sodium tellurite (Na₂TeO₃·0.5H₂O) and 0–0.9 g of glucose were dissolved in 10 mL water, then 20 mL of ethanediamine was added under constant stirring. Subsequently, the resulting solution was transferred into a 40 mL Teflon-lined stainless-steel autoclave. Then the autoclave was sealed and maintained at 180 °C for 12 h. After that, the autoclave was cooled naturally to ambient temperature. The prepared silvery grey products were collected, centrifuged and washed with distilled water and absolute ethanol several times, then dried under vacuum at 80 °C for 12 h.

Characterization

XRD measurements were performed on a Rigaku-D/MAX-2550PC diffractometer, using Cu-Kα radiation at 40 kV, 50 mA, employing a scanning rate of 4° min⁻¹ in the 2θ range from 10° to 90°. The morphologies of the products were characterized by an FEI Quanta 200F field emission scanning electron microscope (FESEM). The fine crystal structures of the particles were detected using transmission electron microscopy (TEM, FEI Tecnai G2 S-Twin operating at 300 kV). Infrared analysis was detected on an FTIR spectrometer (Shimadzu) in the range of 400–4000 cm⁻¹ with the samples mulled in KBr wafers.

Thermoelectric transport measurement

To measure the electrical conductivity and Seebeck coefficient of the as-prepared samples, a pressed bar like sample (10 × 4.6 × 1.2 mm) was prepared under 20 MPa pressure at room temperature. Silver pastes were used as electrical contacts for the electrodes to the specimen. The four-probe method was adopted for the electrical conductivity measurement. For the measurement of Seebeck coefficient, a microheater was applied to create a temperature difference of about 3–15 K between the cool and hot ends of the specimen. A temperature gradient was established in the samples when the electrical power was applied by a thermoelectric pile. The two thermocouples were contacted with the surface of the samples to detect the temperature drop (ΔT), while the resulting thermally induced voltage ΔV was tested by the voltage probes. Then the Seebeck coefficient can be obtained using the formula, S = −ΔV/ΔT. The data of electrical conductivity and Seebeck coefficient were collected by a computer-controlled multifunctional measuring system with a flow of argon of 1 mL min⁻¹ (Keithley 2400 source meter, Keithley 2700 multimeter, Keithley Instruments Inc., America).

Results and discussion

Using the Rietveld refinement method, the crystal structures of Sb₂Te₃ synthesized at 180 °C for 12 h is refined with the room temperature, high-resolution powder XRD data. Rietveld refinement fitting results are presented in Fig. 1. The experimental intensity data are depicted by a dotted line, while the solid line is the calculated pattern of Sb₂Te₃. The bottom line represents the difference between the observed and calculated patterns. The refined result is converged to lattice parameters with reliability factors of Rwp = 0.0942 and Rp = 0.0725, thus confirming the presence of Sb₂Te₃ as a single-crystalline phase in the powder. The morphology and size of the as-prepared samples are studied by FESEM. Fig. 2a shows the overall morphology of a typical sample composed of numerous uniform platelike structures. The magnified SEM images shown in Fig. 2b and 2c clearly display that the obtained products are predominantly hexagonal-based plates, and the edge length of the plates is in the range of 300–500 nm with a thickness of about 30–40 nm. To further investigate the microstructural details of the Sb₂Te₃ platelike nanostructures, TEM and HRTEM analysis were carried out. Fig. 2d shows the TEM image of typical Sb₂Te₃ hexagonal nanoplates. Observed carefully, the surface of the nanoplate is rough. The corresponding HRTEM lattice and spot pattern of SAED, shown in Fig. 2e and 2f, demonstrate the single crystalline nature of the nanoplate. The spacing of 0.209 nm is consistent with the lattice spacing of (110) planes of Sb₂Te₃.

Fig. 1 Observed and calculated XRD powder profiles for Sb₂Te₃ with the difference curve shown below.

Fig. 2 The morphology and structure of typical Sb₂Te₃ nanoplates: (a) low magnification SEM image; (b) middle magnification SEM image; (c) high magnification SEM image; (d) TEM image; (e) HRTEM image recorded from the individual nanoplate; (f) SAED pattern of the nanoplate.

Growth mechanism

To demonstrate the formation mechanism of the Sb₂Te₃ hexagonal nanoplates, time-dependent experiments were carried out, and the products prepared at different growth stages were examined by FESEM and TEM observations. As illustrated in Fig. 3a, irregular nanoplates and some nanoparticles are obtained in the early stages (2 h). A higher magnification SEM image shown in Fig. 3b reveals that the obtained nanoplates have a uniform thickness of less than 30 nm. To further study the structure of the nanoplates, TEM and SAED were performed. As indicated in Fig. 3c, the obtained products have an irregular plate-like morphology, which is in accordance with SEM observations. The SAED taken from the edge of any nanoplate in Fig. 3d indicates the polycrystalline nature of the Sb₂Te₃ nanoplates. It is worth pointing out that the obtained nanoplates composed of numerous nanocrystals are comparatively sensitive to electron beam irradiation, thus, we can not get HRTEM information for the single Sb₂Te₃ nanocrystal. When the reaction time is prolonged to 4 h, more uniform nanoplates can be obtained accompanying some hexagonal-based nanoplates (Fig. 3e). Further increasing the reaction time to 6 h, the products are dominated by hexagonal-based nanoplates, however, some irregular nanoplates still exist (Fig. 3f). As the reaction proceeded, the irregular nanoplates gradually became uniform and changed into hexagonal nanoplates, which can be seen in Fig. 2a–c. At the same time, the edge length of the nanoplates decreases while the thickness increases gradually. Fig. 3g shows the XRD patterns of the above three samples. When the solvothermal time is 2 h, the diffraction peaks are indexed to elemental Te, hexagonal Sb₂Te₃ and (Sb,Te) (JCPDS no. 15–0041). However, the diffraction intensity is lower, which further proves the comparatively poor crystallinity of the sample. Increasing the aging time to 4 h and 6 h, the peaks became higher and sharper because of the better crystallinity of the samples. Simultaneously, the elemental Te gradually disappeared with the increasing reaction time.

Fig. 3 The morphology and structure of the Sb₂Te₃ crystals synthesized in the presence of 0.6 g glucose at 180 °C for different time intervals: (a) low magnification, (b) high magnification SEM images, (c) TEM image, and (d) SAED pattern of the products obtained at 2 h; (e) 4 h; (f) 6 h; (g) XRD patterns.

Based on the above experimental results and analysis, the whole morphology evolution process of the Sb₂Te₃ hexagonal nanoplates is proposed, as illustrated in Fig. 4. In this formation process, time is the key controlling factor. Such a process involves a fast nucleation of amorphous primary particles accompanied by an aggregation and ripening process. In the initial stage, the TeO₃²⁻ ions are first reduced into element Te, and simultaneously a disproportionation reaction occurs in the alkaline solution, resulting in the formation of Te²⁻ and TeO₃²⁻ ions. Then the Te²⁻ ions and Sb³⁺ ions react to form Sb₂Te₃. When no TeO₃²⁻ ions are added into the solution, no solid product could be obtained. This result suggests that the Sb(III)(tart) complex can not be reduced to elemental Sb in this reaction system. As depicted above, the reaction process can be formulated as:

TeO₃²⁻ + H₂NCH₂CH₂NH₂ → Te + H₂O + OH⁻ (1)

3Te + 6OH⁻ → 2Te²⁻ + TeO₃²⁻ + 3H₂O (2)

3Te²⁻ + 2Sb³⁺ → Sb₂Te₃ (3)

As the Sb₂Te₃ nuclei are generated, the glucose molecules will be adsorbed on the surface due to the higher surface energy. Because of the hydrogen bond and electrostatic effects of glucose, the capped nuclei are quickly assembled into large plate-like structures. Such capped nuclei assembled into larger entities is similar to those proposed for the formation of crystalline CaCO₃.²¹ Zhang's research groups²² reported that the layered anisotropic lattice structure of Sb₂Te₃ tended to grow along the top–bottom crystalline plane rather than that along the c-axis, which led to the formation of nanoplates. In our system, the formation of the nanoplate is also attributed to the inherent crystal structure. Because of the high surface energy of the nuclei, the aggregated Sb₂Te₃ particles are not in thermodynamic equilibrium and are metastable, thus, the better crystalline Sb₂Te₃ nanoplates become thermodynamically preferred with the elimination of the glucose molecules from the nuclei. So the transformation from the poor crystallites to better crystalline Sb₂Te₃ through Ostwald ripening can be observed. In addition, the packing structure of rhombohedral Sb₂Te₃ viewed along the c-axis can be described as an infinite hexagon with the composition of alternating Sb and Te ions, resulting in the formation of hexagonal platelike crystals. This intrinsic crystal property that dominates the shape of the particles can also be found in the formation of Bi₂Te₃ hexagonal nanoplates.²³ As the Ostwald ripening process proceeds, the smaller nanoparticles adsorbed around the plate are consumed and grew gradually to make the nanoplate bigger.

Fig. 4 Schematic illustration of the proposed formation mechanism of the Sb₂Te₃ nanoplates.

On the basis of the above experiments results, glucose molecules may promote the aggregation of the Sb₂Te₃ nanoparticles and affect the crystal growth. To further verify the effects of glucose, the products obtained at different time intervals are characterized by FT-IR spectra, as shown in Fig. S1.† In the first step (2 h), two absorption peaks located at about 3462 and 1650 cm⁻¹ correspond to O–H stretching vibration and C=O stretching vibration, respectively, which demonstrates that the glucose molecules can not be washed off by water and ethanol, indicating the strong interaction between glucose and the nanoparticles. As the reaction time is prolonged, the interaction strength starts to decrease with the crystal growth, that is, the effect of glucose decreases gradually when the Sb₂Te₃ crystals become more complete.

Influencing factors

For the sake of clarifying the role glucose molecules play in controlling the morphology of the product, we have synthesized the Sb₂Te₃ micro-platelike structures without the addition of glucose. Fig. 5a and 5b present a SEM image of the prepared sample. From the image, it can be observed that the main product is microplates with an edge length of 4–10 μm and an average thickness of 600 nm, but most of the microplates have irregular shapes. When the dosage of glucose is tuned to 0.3 g, the Sb₂Te₃ samples remain the same platelike structures except that the length and thickness of the Sb₂Te₃ samples decrease to 0.5–4 μm and 50–80 nm, respectively, whereas, some hexagonally shaped nanoplates form (Fig. 5c and 5d). With increasing the amount of glucose to 0.6 g, the hexagonally shaped platelike structures become more uniform and the size continues to decrease, as can be seen in Fig. 2a–c. Further increasing the amount of glucose to 0.9 g, the morphology and the size of the samples have little changes. As discussed above, the obtained nuclei can be protected due to the existence of glucose molecules, which slows down the reaction speed. As the mass transport is impeded, the products become smaller and more regular as they grow gradually. These results indicate that more glucose molecules will reduce the products' size and make them more regular.

Fig. 5 SEM images of Sb₂Te₃ crystals formed using various amounts of glucose: (a, b) 0 g; (c, d) 0.3 g; (e, f) 0.9 g.

It is found that the reaction temperature plays an important role in the formation of Sb₂Te₃ hexagonal nanoplates. Except for the variation of temperature, all the samples are synthesized under the same reaction conditions. When the temperature is decreased to 140 °C, the morphology of the product is obviously different from the product mentioned above in Fig. 2. In Fig. 6a, the product is still composed of numerous nanoplates together with some rod-like morphology. The corresponding XRD pattern shows the obtained sample is composed of Te and Sb₂Te₃ (Fig. 6c). As reported in the literature, the lower temperature leads to a lower reaction rate in the reaction system.^14,24 In this work, a similar phenomenon can be observed, the lower reaction temperature results in a lower formation rate of Te nanorods and transformation rate of Te to Te²⁻. When the temperature is increased to 180 °C and 220 °C, the reaction rate accelerates, simultaneously, the balance between the Te and Te²⁻ is destroyed. Therefore, the uniform hexagonal nanoplates can be obtained (Fig. 6b). These results further prove the reaction process discussed above.

Fig. 6 SEM images and XRD pattern of the products obtained with different reaction temperatures: (a) 140 °C; (b) 220 °C; (c) 140 °C.

Solvent is also a very important factor for the synthesis of Sb₂Te₃ nanoplates. When less ethanediamine (2 mL) is involved in the reaction system, Te nanorods combined with a few Sb₂Te₃ nanoplates are obtained which can be seen in Fig. 7b. The XRD pattern also shows the existence of Te and Sb₂Te₃ (Fig. 7a). The reason why a few Sb₂Te₃ nanoplates can be synthesized under these conditions is the lack of OH⁻ anions. In other words, in the weak alkali medium, the disproportionation from Te to Te²⁻ and TeO₃²⁻ ions is restrained. This phenomenon further proves the disproportionation reaction mechanism of the Te element in the solution. To determined the effect of pH value in the solution, the pH values before and after solvothermal treatment were measured, as shown in Table S1.† With a pH value of 10.8 (2 mL of ethanediamine), the disproportionation reaction can not proceed completely. Thus, Te and Sb₂Te₃ coexist. When 5 mL of ethanediamine is employed, the pH value of the solution increases (pH = 11.7), which accelerates the reaction greatly. As a result, the amount of Te is sharply reduced, meanwhile, the Sb₂Te₃ nanoplates increase (Fig. 7c). When the volume of ethanediamine reaches 10 mL (pH = 12.1), the disproportionation rate from Te to Te²⁻ and TeO₃²⁻ ions will be further increased. And the Te nanorods absolutely transform into irregular Sb₂Te₃ nanoplates, as can be seen in Fig. 7d. With a further increase in the volume of ethanediamine, some hexagonally shaped nanoplates with a wider size distribution are obtained at 15 mL of ethanediamine (Fig. 7e). If the volume ratio of ethanediamine to water increases to 20 : 10, the reaction rate will be further accelerated, and uniform hexagonal nanoplates produced. As the volume ratio of ethanediamine to water continues to increase, the increasing reaction rate will lead to the uniform morphology and size distribution (Fig. 7f).

Fig. 7 The SEM images and XRD pattern of the prepared Sb₂Te₃ products after aging at 180 °C for 12 h in media with different volume ratios of ethanediamine and water: (a) 2 : 28; (b) 2 : 28; (c) 5 : 25; (d) 10 : 20; (e) 15 : 15; (f) 25 : 5.

Thermoelectrical transport properties

The four-probe electrical conductivity measurements in the temperature range of 300–600 K reveal that the three types of samples with different particle size have different electrical conductivities. The sample with the larger particle size has the highest electrical conductivity. The particle size influence on the electrical conductivity has been reported in the literature; it was demonstrated that electrical conductivity is lower for samples with smaller particle size and increases with increasing particle size because of the scattering of charge carriers by the increased density of grain boundaries for smaller grains.^3,25 For all three types of samples, the electrical conductivity decreases with temperature, which are 59.1 S cm⁻¹, 45.5 S cm⁻¹, and 41.1 S cm⁻¹ at room temperature and 30.1 S cm⁻¹, 22.3 S cm⁻¹, 21.4 S cm⁻¹ at the temperature of 600 K for 500 nm, 60 nm and 40 nm particle samples, respectively (Fig. 8a). The measured electrical conductivity for all the three Sb₂Te₃ samples is two orders of magnitude lower than that of the corresponding bulk sample²⁶ and exhibits strong size dependence. Moreover, the electrical conductivities of the three samples are lower than those of Sb₂Te₃ single crystalline nanowires²⁶ and spark plasma sintering Sb₂Te₃ nanosheets,²⁷ while higher than that of porous particle films with similar sizes.²⁸ The possible explanation for the observed different electrical conductivity originates in the density of the pellets. The denser sample leads to the higher electrical conductivity. In order to determine the effect of the density, Archimedes principle is employed to measure the density of the sample, and the results are shown in Table S2.† The densities of the three samples have little difference, which further proves that the size of the products has much effect on the electrical conductivity rather than the density of the pellets. However, the density of the sample pressed in our work is possibly higher than that of porous films and lower than that prepared by spark plasma sintering process, hot pressed method or isostatic cool pressing technique. Thus, a higher value of electrical conductivity can be obtained in denser samples, e.g. hot pressed and spark plasma sintering pellets.

Fig. 8 (a) The electrical conductivities and (b) Seebeck coefficients of the obtained samples in the temperature range of 300–600 K.

Fig. 8b shows the Seebeck coefficients measured as a function of temperature for the three different sized Sb₂Te₃ samples. As can be seen, the Seebeck coefficients increase with increasing temperature for all the three samples, and they are affected by the variation in the particle size. Fig. 8b also demonstrates that all the three samples are p-type semiconductors with high maximum Seebeck coefficients about 119.9 μV K⁻¹, 142.4 μV K⁻¹, and 161.7 μV K⁻¹ at 600 K for 500 nm, 60 nm and 40 nm particle samples. These values are higher than that of bulk Sb₂Te₃ crystals (79 μV K⁻¹);^28,29 but are lower than that of Sb₂Te₃ nanoparticles²⁸ with the size of 20 nm and that of spark plasma sintering Sb₂Te₃ nanosheets.²⁷ The enhanced Seebeck coefficient in smaller particles is thought to be possibly related to the energy filtering effect due to the charge carrier trapped in the grain boundaries regions,^30,31 or to be connected with the increasing difference between the Fermi level and the average mobile carrier energy resulting from the quantum confinement effect.^28,32

Conclusions

In summary, well-defined Sb₂Te₃ hexagonal nanoplates with a thickness of 40 nm were successfully prepared by a solvothermal treatment in the presence of glucose. Based on a series of contrast experimental results, a possible reaction process and formation mechanism of these hexagonal nanoplates have been proposed, which was related to self-assembly and an Ostwald ripening process. The glucose plays a crucial role in the formation process. The size and morphology of Sb₂Te₃ can be tuned by adjusting the concentration of the glucose. Moreover, the electrical conductivity and Seebeck coefficient were investigated as a function of temperature and thickness of the Sb₂Te₃ plate. The electrical conductivities are much smaller than the bulk value, while the Seebeck coefficients show much enhancement in smaller particle size.

Acknowledgements

This work is financially supported by the National Natural Science Foundation of China (Project no. 20871036, 21071036), Province Natural Science Foundation of Heilongjiang Province (ZD201011) and Scientific Research Innovation Foundation of Harbin Institute of Technology (Project no. GFCQ98332122).

References

1. Y. N. Xia, P. D. Yang, Y. G. Sun, Y. Y. Wu, B. Mayers, B. Gates, Y. D. Yin, F. Kim and Y. Q. Yan, Adv. Mater., 2003, 15, 353–389.
2. H. Q. Liu, D. B. Kuang, A. W. Xu, Y. P. Fang, C. Frommen and D. Fenske, Adv. Mater., 2003, 15, 1747–1750.
3. M. S. Toprak, C. Stiewe, D. Platzek, S. Williams, L. Bertini, E. C. Muller, C. Gatti, Y. Zhang, M. Rowe and M. Muhammed, Adv. Funct. Mater., 2004, 14, 1189–1196.
4. J. S. Lee and U. A. Joshi, Small, 2005, 1, 1172–1176.
5. Z. H. Wang, L. Pan, H. B. Hu and S. P. Zhao, CrystEngComm, 2010, 12, 1899–1904.
6. M. Sun, D. Z. Li, W. J. Zhang, X. Z. Fu, Y. Shao, W. J. Li, G. C. Xiao and Y. H. He, Nanotechnology, 2010, 21, 355601.
7. Y. Z. Jin, Y. F. Yang, H. P. He and Z. Z. Ye, CrystEngComm, 2010, 12, 2663–2665.
8. P. O'Brien, S. S. Garje, D. J. Eisler, J. S. Ritch, M. Afzaal and T. Chivers, J. Am. Chem. Soc., 2006, 128, 3120–3121.
9. W. L. Suchanek, J. M. Garces, P. F. Fulvio and M. Jaroniec, Chem. Mater., 2010, 22, 6564–6574.
10. P. D. Yang, W. J. Liang, A. I. Hochbaum, M. Fardy, O. Rabin and M. J. Zhang, Nano Lett., 2009, 9, 1689–1693.
11. A. Majumdar, A. I. Hochbaum, R. K. Chen, R. D. Delgado, W. J. Liang, E. C. Garnett, M. Najarian and P. D. Yang, Nature, 2008, 451, 163-U165.
12. R. Venkatasubramanian, E. Siivola, T. Colpitts and B. O'Quinn, Nature, 2001, 413, 597–602.
13. J. H. Zhu, B. Zhou, Y. Ji, Y. F. Yang and X. H. Li, Cryst. Growth Des., 2008, 8, 4394–4397.
14. H. J. Zhang, W. D. Shi, L. Zhou, S. Y. Song and J. H. Yang, Adv. Mater., 2008, 20, 1892–1897.
15. W. Z. Wang, B. Poudel, J. Yang, D. Z. Wang and Z. F. Ren, J. Am. Chem. Soc., 2005, 127, 13792–13793.
16. Y. Cui, S. Meister, H. L. Peng, K. McIlwrath, K. Jarausch and X. F. Zhang, Nano Lett., 2006, 6, 1514–1517.
17. H. Park, J. S. Lee, S. Brittman and D. Yu, J. Am. Chem. Soc., 2008, 130, 6252–6258.
18. R. V. Devireddy, D. Pinisetty, M. Gupta, A. B. Karki and D. P. Young, J. Mater. Chem., 2011, 21, 4098–4107.
19. D. L. W. Z. Wang, Y. J. Liang, G. L. Zhang, B. Q. Zeng and Q. Y. He, Langmuir, 2011, 27, 815–819.
20. H. J. Zhang, W. D. Shi, J. B. Yu and H. S. Wang, J. Am. Chem. Soc., 2006, 128, 16490–16491.
21. A. W. Xu, M. Antonietti, H. Colfen and Y. P. Fang, Adv. Funct. Mater., 2006, 16, 903–908.
22. W. D. Shi, L. Zhou, S. Y. Song, J. H. Yang and H. J. Zhang, Adv. Mater., 2008, 20, 1892–1897.
23. Z. L. Wang, W. G. Lu, Y. Ding, Y. X. Chen and J. Y. Fang, J. Am. Chem. Soc., 2005, 127, 10112–10116.
24. Z. Fang, X. Y. Wang, J. M. Shen, X. Lin, Y. H. Ni and X. W. Wei, Cryst. Growth Des., 2010, 10, 469–474.
25. M. Takashiri, K. Miyazaki, S. Tanaka, J. Kurosaki, D. Nagai and H. Tsukamoto, J. Appl. Phys., 2008, 104, 084302.
26. P. Kim, Y. M. Zuev, J. S. Lee, C. Galloy and H. Park, Nano Lett., 2010, 10, 3037–3040.
27. Y. J. Zhu, G. H. Dong and L. D. Chen, J. Mater. Chem., 2010, 20, 1976–1981.
28. Q. Y. Yan, J. Chen, T. Sun, D. Sim, H. Y. Peng, H. T. Wang, S. F. Fan, H. H. Hng, J. Ma, F. Y. C. Boey, S. Li, M. K. Samani, G. C. K. Chen, X. D. Chen and T. Wu, Chem. Mater., 2010, 22, 3086–3092.
29. C. Drasar, M. Steinhart, P. Lost'ak, H. K. Shin, J. S. Dyck and C. Uher, J. Solid State Chem., 2005, 178, 1301–1307.
30. J. Martin, L. Wang, L. Chen and G. S. Nolas, Phys. Rev. B, 2009, 79, 115311.
31. T. C. Harman, D. L. Spears and M. J. Manfra, J. Electron. Mater., 1996, 25, 1121–1127.
32. D. V. Talapin, R. Y. Wang, J. P. Feser, J. S. Lee, R. Segalman and A. Majumdar, Nano Lett., 2008, 8, 2283–2288.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c1ra00642h

---