Growth of single-crystalline Bi2Te3 hexagonal nanoplates with and without single nanopores during temperature-controlled solvothermal synthesis

Bismuth telluride (Bi2Te3) is a promising thermoelectric material for applications near room temperature. To increase the thermoelectric performance of this material, its dimensions and thermal transport should be decreased. Two-dimensional nanoplates with nanopores are an ideal structure because thermal transport is disrupted by nanopores. We prepared Bi2Te3 nanoplates with single nanopores by a solvothermal synthesis and investigated their structural and crystallographic properties. The nanoplates synthesized at a lower reaction temperature (190 °C) developed single nanopores (approximately 20 nm in diameter), whereas the nanoplates synthesized at a higher reaction temperature (200 °C) did not have nanopores. A crystal growth mechanism is proposed based on the experimental observations.

www.nature.com/scientificreports www.nature.com/scientificreports/ thermoelectric performance of the nanoplates might be expected to further increase. However, to the best of our knowledge, there have been no reports on the crystal growth of Bi 2 Te 3 nanoplates containing nanopores.
In this study, we prepare Bi 2 Te 3 hexagonal nanoplates with single nanopores based on a solvothermal synthesis. For comparison, Bi 2 Te 3 hexagonal nanoplates with no nanopores were also prepared using solvothermal synthesis by changing the reaction temperature. The structural and crystallographic characteristics of the nanoplates were investigated. On the basis of our experimental observations and understanding of the crystal growth, the formation of the nanoplates with nanopores is discussed.

Methods
Single-crystalline Bi 2 Te 3 hexagonal nanoplates were formed in a solvothermal synthesis. The basic experimental method has been described in our previous work 28,29 . The nanoplates were prepared in a Teflon-container within a stainless-steel autoclave, involving a reaction under high pressure and high temperature. Stirring was applied with a hot plate and magnetic stirrer. Bi 2 O 3 (Fujifilm Wako Pure Chemical Co., >99.9%) and TeO 2 (Kojundo Chemical Laboratory Co., Ltd., >99.9%) were used as Bi 2 Te 3 sources, and C 2 H 6 O 2 (Fujifilm Wako Pure Chemical Co., >90.0%) was used as a ligand. NaOH (Fujifilm Wako Pure Chemical Co., >97.0%) and polyvinyl pyrrolidone (Fujifilm Wako Pure Chemical Co., K30, M s ~40,000) were contained in the solution. All chemicals were used as received without further purification. The Bi 2 Te 3 nanoplates were synthesized according to the following procedure: 0.4 g of PVP was dissolved in ethylene glycol (18 mL), followed by the addition of Bi 2 O 3 (0.02 mol/L), TeO 2 (0.07 mol/L), and 2 mL of a NaOH solution (5.0 mol/L). The resulting precursor solution was sealed in the autoclave and heated and maintained at either 190 °C or 200 °C for 4 h with magnetic stirring at 500 rpm. After the synthesis, the products were allowed to cool down below 50 °C naturally. The residue was washed several times with distilled water and absolute ethanol and the products were collected by centrifugation. Finally, the products were dried under vacuum at 60 °C for 24 h.
We analyzed the microstructure of Bi 2 Te 3 nanoplates with a field emission scanning electron microscope (FE-SEM, Hitachi S-4800). The precise structure of the nanoplates was analyzed with a high-resolution transmission electron microscope (TEM, JEOL JEM-ARM200F) and selected area electron diffraction (SAED) at an accelerating voltage of 200 kV. The chemical composition of the nanoplates was determined with an electron probe microanalyzer (EPMA, Shimadzu EPMA-1610). The compositions of the samples were calibrated using the ZAF4 program installed with the EPMA-1610. X-ray diffraction (XRD, Rigaku MiniFlex600) analyses were performed to determine the crystal lattice structure and the phase purity of the nanoplates. We used Cu-Kα radiation at a scanning rate of 0.02°/s over the 2θ range of 10° to 80°. www.nature.com/scientificreports www.nature.com/scientificreports/ Figure 2 shows SEM images of the Bi 2 Te 3 nanoplates formed at different reaction temperatures in the solvothermal synthesis. At a reaction temperature of 190 °C ( Fig. 2(a)), we obtained hexagonal nanoplates with edge lengths of approximately 1 μm. A single nanopore appeared in the center of the nanoplates. We also observed several samples manufactured under the same conditions as the sample shown in the SEM image and found a nanopore in the center of almost every nanoplate (The detailed information is provided in Fig. S1). Hence, the nanopores were naturally generated in the crystal growth process and not caused by accidental collisions between the nanoplates. At a reaction temperature of 200 °C ( Fig. 2(b)), the nanoplates had a similar edge length of approximately 1 μm but no nanopores were observed. We also fabricated Bi 2 Te 3 nanoplates at reaction temperatures of 180 °C and 230 °C, and observed their surface morphologies using FE-SEM (The detailed information is provided in Fig. S2). As a result, single nanopores were obtained in the nanoplates at 180 °C, but not in those formed at 230 °C. Therefore, we conclude that the synthesis at relatively low reaction temperatures yielded nanoplates with single nanopores and the nanopores disappeared as the reaction temperature was increased.

Results and Discussion
To further investigate the precise structure of the Bi 2 Te 3 nanoplates, we performed TEM observations, as shown in Fig. 3. The TEM images of the Bi 2 Te 3 nanoplates synthesized at a reaction temperature of 190 °C are shown in Fig. 3(a). The nanoplates had sharp edges with a distance of approximately 1 μm between the opposite edges, and a very smooth surface, which indicated excellent crystallinity. The mesh structure behind the nanoplate was visible indicating the low thickness of the plate. The size of the nanopores was approximately 20 nm, and each was located close to the center of each nanoplate. Figure 3(b) shows high-resolution TEM (HRTEM) images of the nanoplates near the nanopores. The SAED pattern inset in Fig. 3(b) shows a hexagonal symmetry diffraction spot pattern, which indicates its single-crystalline nature. Lattice fringes cannot be seen in the range of approximately 5 nm from the nanopore, suggesting an amorphous structure in this region. However, clear lattice fringes were observed in this region at distances greater than 5 nm from the edges of the nanopores. The lattice fringes were structurally uniform with a spacing of 0.21 nm, which is in good agreement with the d values of the (110) planes of rhombohedral Bi 2 Te 3 . From Fig. 3(a,b), we conclude that single-crystal nanoplates grew along the (00l) planes except for regions near the nanopores. Figure 3(c) shows a TEM image of a Bi 2 Te 3 nanoplate synthesized at a reaction temperature of 200 °C. The shape of the nanoplate was almost the same as that synthesized at 190 °C except for the presence of the nanopore. The HRTEM image in Fig. 3(d) clearly shows that the lattice fringes were also structurally uniform with a spacing of 0.21 nm, which is in good agreement with the d value of the (110) planes of rhombohedral Bi 2 Te 3 . The SAED pattern shown in the inset of Fig. 3(d) was indexed to the [00l] zone axis of rhombohedral Bi 2 Te 3 , indicating that this nanoplate was single-crystalline and had a preferential www.nature.com/scientificreports www.nature.com/scientificreports/ (00l) orientation. Notably, weak forbidden reflections were observed, owing to the broken translational symmetry of the Te/Bi antisite defects, which is consistent with previous reports 28,30,31 .
The atomic composition ratios [Te/(Bi + Te)] of the Bi 2 Te 3 nanoplates prepared at different reaction temperatures are listed in Table 1. The composition ratio of the nanoplates with single nanopores formed at 190 °C was 0.56, which is slightly lower than the stoichiometric proportion of 0.6. At a reaction temperature of 200 °C, the composition ratio of the nanoplates was 0.54. This result gives insight into the behavior of Te/Bi antisite defects, as shown in Fig. 3(d). The composition ratio of the nanoplates with no nanopores was lower than that of the nanoplates with nanopores; thus, a relatively large amount of Te atoms replaced Bi atoms in the nanoplates without nanopores.
The phase purity and crystal structure of the Bi 2 Te 3 nanoplates were examined by XRD analysis, as shown in Fig. 4. All peaks observed in the XRD patterns of the nanoplates formed at reaction temperatures of 190 °C and 200 °C were indexed to the standard diffraction pattern of Bi 2 Te 3 (JCPDS 15-0863) although both nanoplates exhibited slight deviations from the exact stoichiometry. The main peaks originated from the (006), (015), (10 10), and (00 15) planes. Therefore, the phase purity and the crystal structure depended only slightly on nanopores in the Bi 2 Te 3 nanoplates.
As mentioned above, the single nanopores appeared at the centers of the single-crystalline Bi 2 Te 3 hexagonal nanoplates at the lower reaction temperature. Here, we propose a growth mechanism and reaction process to account for the structures of these nanoplates. Figure 5 shows a schematic diagram of the crystal growth of the nanoplates with and without single nanopores. When Bi 2 Te 3 nuclei were generated in the solution, they aggregated. The atomic composition ratios of the nanoplates synthesized at different temperatures deviated slightly from the stoichiometric proportion; however, we consider that this deviation did not affect the aggregation process of the nuclei and Ostwald ripening process described below. When the radius of the aggregated nuclei  www.nature.com/scientificreports www.nature.com/scientificreports/ became larger than the critical nucleus, Bi 2 Te 3 nanoparticles were generated. In the Bi 2 Te 3 system, the formation of nanoplates is also attributed to the inherent crystal structure. Because of the high surface energy of the nuclei, the aggregated Bi 2 Te 3 particles were not in thermodynamic equilibrium and were metastable; Bi 2 Te 3 nanoplates with a thermodynamic preference for better crystallinity. After formation of Bi 2 Te 3 nanoplates, the Ostwald ripening process proceeded; however, the process differed at reaction temperatures of 190 °C and 200 °C 32 . At a reaction temperature of 190 °C, Ostwald ripening led to the formation of nanopore structures because of lower crystallinity or less dense particles in the colloidal aggregate, which gradually dissolved, whereas larger, better crystallized or denser particles in the same aggregate continued to grow at the lower reaction temperature. However, at 200 °C, nanopores were not formed in the nanoplates because the particles in the colloidal aggregate did not dissolve owing to the dense structures initially formed at the higher reaction temperature. Both types of Bi 2 Te 3 nanoplates grew gradually to form deep nanoplates with many crystalline planes from the top to bottom 33,34 . Rhombohedral Bi 2 Te 3 is built up from many layers extending along the top-to-bottom crystalline planes connected by van der Waals bonds, as shown in Fig. 1.

Conclusions
We used solvothermal synthesis at different reaction temperatures to prepare Bi 2 Te 3 hexagonal nanoplates. The structural and crystallographic characteristics of the nanoplates were investigated. The nanoplates synthesized at a lower reaction temperature (190 °C) developed single nanopores (approximately 20 nm in diameter), whereas the nanoplates synthesized at a higher reaction temperature (200 °C) did not have nanopores. Based on the experimental and analytical results, we propose a growth mechanism and reaction process for the nanoplates and the nanopore appearance at different reaction temperatures. We expect that the Bi 2 Te 3 nanoplates with single nanopores will feature improved thermoelectric performance.  www.nature.com/scientificreports www.nature.com/scientificreports/