Surface Modification of Bi₂Te₃ Nanoplates Deposited with Tin, Palladium, and Tin/Palladium Using Electroless Deposition

Abstract

Surface-modified nanoplate-shaped thermoelectric materials can achieve good thermoelectric performance. Herein, single-crystalline Bi₂Te₃ nanoplates with regular hexagonal shapes were prepared via solvothermal techniques. Surface modification was performed to deposit different metals onto the nanoplates using electroless deposition. Nanoparticle-shaped tin (Sn) and layer-shaped palladium (Pd) formed on the Bi₂Te₃ nanoplates via electroless deposition. For the sequential deposition of Sn and Pd, the surface morphology was mostly the same as that of the Sn-Bi₂Te₃ nanoplates. To assess the thermoelectric properties of the nanoplates as closely as possible, they were compressed into thin bulk shapes at 300 K. The Sn-Bi₂Te₃ and Sn/Pd-Bi₂Te₃ nanoplates exhibited the lowest lattice thermal conductivity of 1.1 W/(m·K), indicating that nanoparticle-shaped Sn facilitated the scattering of phonons. By contrast, the Pd-Bi₂Te₃ nanoplates exhibited the highest electrical conductivity. Thus, the highest power factor (15 μW/(m∙K²)) and dimensionless ZT (32 × 10⁻³) were obtained for the Pd-Bi₂Te₃ nanoplates. These thermoelectric properties were not as high as those of the sintered Bi₂Te₃ samples; however, this study revealed the effect of different metal depositions on Bi₂Te₃ nanoplates for improving thermoelectric performance. These findings offer venues for improving thermoelectric performance by sintering nanoplates deposited with appropriate metals.

1. Introduction

With the development of wireless sensor networks for the Internet of Things (IoT), securing a power supply for sensors has become a major issue [1]. Therefore, energy harvesting has received considerable attention [2]. These methods convert environmental energy, including light [3,4], vibration [5,6], heat [7,8], and electromagnetic waves [9,10], into electrical energy. Among the different types of environmental energy, thermal energy is the most accessible, owing to its ubiquity; electrical energy can be obtained from it using a thermoelectric generation system. The performance of thermoelectric generation depends directly on the properties of the thermoelectric materials. The main thermoelectric properties are the power factor, PF, and the dimensionless figure of merit, ZT, which are expressed as PF = σS² and ZT = σS²T/κ, respectively, where σ, S, T, and κ are the electrical conductivity, Seebeck coefficient, absolute temperature, and thermal conductivity, respectively.

Theoretical and experimental studies of thermoelectricity began in earnest in the 1950s [11], and various structures of thermoelectric generators have since been proposed [12,13,14,15,16]. Various thermoelectric materials have been considered, including inorganic and organic materials [17,18,19,20,21,22,23]. However, despite these efforts, the best-performing materials at temperatures of approximately 300 K have been bismuth-telluride-based alloys developed in the 1950s [24,25,26,27,28,29,30]. These alloys have hexagonal unit cells, and the lattice constant along the c-axis is almost seven times greater than that along the a- or b-axis, contributing to the significant anisotropic thermoelectric properties of the material. These alloys exhibit properties of an indirect-transition-type semiconductor with a band gap ranging from 0.1 to 0.2 eV. The typical single-crystalline Bi₂Te₃ bulk exhibited a ZT of approximately 0.8 [31]. Therefore, when thermoelectric generators are used to power wireless sensors, bismuth-telluride-based alloys are the most practical. However, their material performance is not sufficient to power wireless sensors, requiring further performance improvements. One approach to improving material performance is nanostructuring, which can lower thermal conductivity owing to phonon scattering and high Seebeck coefficient, which are due to lower dimensionality [32]. To this end, various nanostructured bismuth-telluride-based alloys have been developed, including superlattices [33,34], nanocrystals [35,36,37,38], nanoparticles [39,40,41], and nanoplates [42,43,44]. In superlattices, Venkatasubramanian et al. fabricated the p-type Bi₂Te₃/Sb₂Te₃ superlattices using metal–organic chemical vapor deposition and achieved a ZT of 2.4 at 300 K [33]. In nanocrystals, Poudel et al. fabricated bismuth antimony telluride bulk alloys made by hot-pressing nanoparticles and achieved a ZT of 1.4 at 373 K [35]. Wu et al. fabricated nanoporous (00l)-oriented Bi₂Te₃ nanoplates using a facile process combining chemical and physical methods, including annealing and a reaction-engineered solution method, and the estimated ZT was larger than 1.2 at 300 K [44]. In particular, nanoplates have excellent geometries that can be both nanostructured and low-dimensional. Because nanoplates are extremely thin (tens of nanometers) compared with their diameter, they can be considered two-dimensional structures. Therefore, the synthesis of bismuth-telluride-based alloy nanoplates and the evaluation of their thermoelectric properties have been actively investigated [45,46,47,48,49]. Nanocomposites of nanoplates with other materials have also been developed [50,51,52,53,54].

Recently, studies have been conducted to further improve the performance of nanoplates. One approach is the surface modification of nanoplates with different materials [55,56]. Zheng et al. fabricated Au-Sb₂Te₃ nanoplates with a beneficial interface barrier and high phonon scattering, resulting in synergistically optimal electrical and thermal properties [57]. In our previous study, tin nanoparticles were deposited on the surface of a Bi₂Te₃ nanoplate using electroless deposition, which reduced the lattice thermal conductivity [58]. When different materials were deposited on the surfaces of nanoplates, the crystal growth modes differed depending on the lattice matching and interfacial free energy [59]. Therefore, the morphology of the nanoplate surface varies depending on the type of material deposited on its surface, affecting its thermoelectric performance.

In this study, we prepared Bi₂Te₃ nanoplates via solvothermal techniques and deposited different metals (tin (Sn), palladium (Pd), and Sn/Pd) on the nanoplate surface using electroless deposition. Sn and Pd have been typically used in electroless deposition, exhibiting different deposition modes [60]. In addition, the magnitude of diffusion in Bi₂Te₃ is different for Sn and Pd. Sn tends to diffuse into Bi₂Te₃, while Pd tends to remain on the Bi₂Te₃ surface [61]. Therefore, different Bi₂Te₃ surface structures can be fabricated by using Sn and Pd, and the difference can be reflected in the thermoelectric performance. Their structural and thermoelectric properties were examined using various analytical techniques. The surface morphology of the nanoplates depended on the deposited metal. To measure the thermoelectric properties of the nanoplates, the nanoplate assemblies were compressed into thin bulk shapes. The thermoelectric properties of the bulk samples varied depending on the metal type and surface morphology of the nanoplates. Finally, we discuss the different modes of crystal growth on the nanoplates depending on the metal type and their effect on the thermoelectric properties.

2. Materials and Methods

The studied Bi₂Te₃ nanoplates were fabricated via a solvothermal technique. The detailed fabrication procedure is described in previous studies [62,63,64]. In brief, Bi₂Te₃ nanoplates were fabricated as follows: 0.4 g of polyvinylpyrrolidone (PVP) (99.9% purity, K30, Ms ~40,000, Fujifilm Wako Co., Osaka, Japan) was dissolved in 18 mL of C₂H₆O₂ (99.5% purity, Fujifilm Wako Co., Osaka, Japan). Subsequently, 20 mM of Bi₂O₃ (99.9% purity, Fujifilm Wako Co., Osaka, Japan), 70 mM of TeO₂ (99.9% purity, Kojundo Chemical Lab., Sakado, Japan), and 2 mL of NaOH solution (0.5 M) (purity > 97.0%, Fujifilm Wako Co., Osaka, Japan) were added. The resultant solution was placed in an autoclave and heated to 200 °C for 4 h under continuous stirring. The precipitated products were collected using centrifugation and washed several times with distilled water and ethanol. The collected products were dried under vacuum at 60 °C for 24 h. The total weight of dried Bi₂Te₃ nanoplates was 0.71 g. The production efficiency, which is defined as the total weight of the dried Bi₂Te₃ nanoplates in the precursor solution divided by the weight of Bi and Te in the initial Bi₂O₃ and TeO₂ powders, was 54%.

The process of metal deposition on the studied Bi₂Te₃ nanoplates is shown in Figure 1. A similar electroless deposition process has been described in our previous study [58]. To deposit the metals on the Bi₂Te₃ nanoplates, a displacement deposition method was used for electroless deposition [65]. Three types of metals were deposited on the Bi₂Te₃ nanoplates: Sn, Pd, and the lamination of Sn and Pd (Sn/Pd). For the Sn deposition, a precursor solution comprising 60 mM of SnCl₂ (purity > 97.0%, Fujifilm Wako Co., Osaka, Japan) and 0.66 M of HCl (concentration of 35–37%, Fujifilm Wako Co., Osaka, Japan) was used. For the Pd deposition, a precursor solution comprising 1.1 mM of PdCl₂ (purity > 99.0%, Fujifilm Wako Co., Osaka, Japan) and 0.66 M of HCl (concentration of 35–37%, Fujifilm Wako Co., Osaka, Japan) was used. The Bi₂Te₃ nanoplates (0.65 g) were immersed in each solution for 10 min. For the Sn/Pd deposition, the same conditions were used for the two solutions, as mentioned earlier. First, Bi₂Te₃ nanoplates (0.65 g) were immersed in the Sn-containing solution for 10 min. After rinsing with deionized water, the Sn-deposited Bi₂Te₃ nanoplates were immersed in the Pd-containing solution for 10 min. Thereafter, the three types of metal-deposited Bi₂Te₃ nanoplates were corrected using filter paper and rinsed with deionized water. In the end, the nanoplates were dried under vacuum at 60 °C for 24 h. In order to evaluate the thermoelectric properties of the prepared nanoplates by representing their properties as close as possible to those of pristine nanoplates, but not the melting structure, 0.38 g of nanoplates were pressed at approximately 300 K under uniaxial pressure (0.45 GPa). The samples were 22 mm long, 5 mm wide, and 0.6 mm thick.

The surface morphologies of the prepared nanoplates were examined using field-emission scanning electron microscopy (FE-SEM; S-4800, Hitachi, Tokyo, Japan), transmission electron microscopy (TEM; JEM-ARM200F, JEOL, Tokyo, Japan), and selected-area electron diffraction (SAED). The atomic composition of the nanoplates was determined using an electron probe microanalyzer (EPMA; EPMA-8050G, Shimadzu, Kyoto, Japan). The phase purity and crystal structure of the nanoplates were characterized using X-ray diffraction (XRD; MiniFlex 600, Rigaku, Tokyo, Japan) with Cu-Kα radiation (λ = 1.54 Å with 2θ ranging from 10° to 80°).

Thermal conductivity κ was calculated using the equation κ = αρC, where α, ρ, and C are thermal diffusivity, density, and specific heat, respectively. The thermal diffusivity α was measured at 300 K using a thermophysical-property-measuring instrument (Thermowave Analyzer, Bethel, Ishioka, Japan) based on spot-periodic heating radiation thermometry with an accuracy of ±5% [66]. This is a non-contact method that independently measures both in-plane and cross-plane thermal conductivity. Density was determined by measuring the size and mass. Literature values were used for the specific heat [14,67]. The electrical conductivity and Seebeck coefficient were measured at 300 K using the ZEM-3 method (Advance Riko, Yokohama, Japan) in a helium atmosphere with an accuracy of ±5%. Finally, the power factor and the dimensionless figure of merit were calculated.

3. Results and Discussion

3.1. Structural Characteristics of Bi₂Te₃ Nanoplates Deposited with Different Metals

Figure 2 shows the structural characteristics of pristine Bi₂Te₃ nanoplates. The SEM image in Figure 2a shows that the Bi₂Te₃ nanoplates exhibited a significantly flat surface and a regular hexagonal shape with an approximate lateral size of 1 μm. The atomic composition of the plates (measured using EPMA) was 37 at% Bi and 63 at% Te, differing by 3% from the stoichiometric ratio (Bi:Te = 40:60). A TEM image of a typical Bi₂Te₃ nanoplate is shown in Figure 2b. The nanoplates were considerably thin (less than 50 nm) because of the overlap between the nanoplates and the mesh behind them. The SAED pattern shown in the inset of Figure 2b was indexed to the (00l) zone axis of rhombohedral Bi₂Te₃, which indicated that this nanoplate was single-crystalline. In brief, the mechanism of crystal growth of Bi₂Te₃ nanoplates is as follows [68]. After the chemical reactions, the Bi₂Te₃ nuclei were formed and subsequently aggregated. Bi₂Te₃ nanoparticles were generated when the radius of the aggregated nuclei exceeded their critical nucleus size. The nanoplate’s formation was attributed to its inherent crystal structure. Due to the high surface energy of the nuclei, the aggregated particles of Bi₂Te₃ were not in thermodynamic equilibrium and were metastable. As a result, Bi₂Te₃ nanoplates with better crystallinity became thermodynamically preferred. After Bi₂Te₃ nanoplates were formed, the Ostwald ripening process took place. The smaller nanoparticles adsorbed around the plates were consumed and grew gradually, leading to larger nanoplates along the top–bottom crystalline plane. This phenomenon happened because the rhombohedral Bi₂Te₃ can be described as a stack of infinite layers extending along the top–bottom crystalline plane connected by van der Waals bonds. From a thermodynamic perspective, the free energy of a broken covalent bond is higher than that of a dangling van der Waals bond. This suggests that Bi₂Te₃ crystal growth is faster along the top–bottom crystalline plane than along the c-axis, which is perpendicular to the top–bottom planes. This is because crystalline facets tend to develop on the low-index plane.

Figure 3 shows the SEM images of the Bi₂Te₃ nanoplates after electroless deposition of different metals. Notably, the SEM image of the Sn-Bi₂Te₃ nanoplates is referred to from our previous report [58]. As shown in Figure 3a, when Sn was deposited on the Bi₂Te₃ nanoplates, nanoparticles with a size of several tens of nanometers were dispersed on their surface. The concentration of Sn, as measured using EPMA, was 1 at%, suggesting that nanoparticle-shaped Sn was deposited on the prepared Bi₂Te₃ nanoplates. The surface morphology of the Pd-deposited Bi₂Te₃ nanoplates was different from that of the Sn-deposited nanoplates, as shown in Figure 3b. The nanoplate surface was covered with a thin, speckled Pd layer with a Pd concentration of 3 at%. Therefore, the shape and amount of deposition on the prepared Bi₂Te₃ nanoplates varied depending on the type of metal. This difference was attributed to the different liquid–solid interfacial free energies of Sn-Bi₂Te₃ and Pd-Bi₂Te₃ [59]. When the interfacial free energy was higher (as in the case of Sn-Bi₂Te₃), the crystals grew into island shapes. By contrast, when the interfacial free energy was lower (as in the case of Pd-Bi₂Te₃), the crystals grew in a layered shape. As shown in Figure 3c, when Sn and Pd were sequentially deposited on the nanoplate surface, the surface morphology was similar to that of the Sn-deposited nanoplates. The concentration of Sn was 1 at%, which was the same as for the Sn-deposited nanoplates; however, the concentration of Pd was less than 1 at%, which was lower than for the Pd-deposited nanoplates. These results suggest that Pd readily adhered to the Bi₂Te₃ surface but struggled to adhere to the Sn surface. The phase purity and crystal structure of the Bi₂Te₃ nanoplates with the electroless deposition of different metals were examined via XRD analysis, as shown in Figure 4. Bi₂Te₃ showed rhombohedral tetradymite-type crystal structures with the space group $D_{3d}^5(R\stackrel{-}{3}m)$ and hexagonal unit cells. The lattice constant along the c-axis was 30.3 Å, while that along the a- or b-axis was 4.35 Å. The unit cell consists of five covalently bonded monatomic sheets arranged along the c-axis in the sequence –Te⁽¹⁾–Bi–Te⁽²⁾–Bi–Te⁽¹⁾–. The superscripts (1) and (2) indicate two distinct chemical states of the anions. The bonds between Te(1) and Bi include both covalent and ionic bonds, while Te(2) and Bi are purely covalently bonded. There is a very weak van der Waals attraction between adjacent Te(1) layers. Most of the peaks observed in the XRD patterns of the nanoplates without metal deposition were indexed to the standard diffraction pattern of Bi₂Te₃ (JCPDS 15-0863). The main peaks were observed in the c-axis-oriented (00l), (015), and (1010) planes, and several smaller peaks were also observed. Although electroless deposition was performed to deposit different metals on the nanoplate surface, the XRD patterns did not exhibit any peaks corresponding to those attributed to the deposited metals. This was due to the low crystallinity, i.e., extremely small crystallite size or high defect density, of the electrodeposited metals. Similar phenomena have been reported for the electrodeposited layers of bismuth-telluride-based alloys [58,69,70].

3.2. Thermal Conductivity of Bi₂Te₃ Nanoplates Deposited with Different Metals

Figure 5 shows the in-plane thermal conductivities of the bulk Bi₂Te₃ nanoplates with different metal depositions. The total thermal conductivity, κ_total, was estimated based on the measured thermal diffusivities, densities, and specific heats. The electronic thermal conductivity, κ_e, described in the inset was calculated from the experimentally measured electrical conductivities and using the Wiedemann–Franz law, given by κ_e = LσT. Here, the Lorenz number, L, was 2.44 × 10⁻⁸ WΩ/K² for the classical free-electron model. The lattice thermal conductivity, κ_lattice, was obtained by subtracting the electronic thermal conductivity. The lattice thermal conductivities of all nanoplates were mostly the same as the total thermal conductivities, owing to their low electronic thermal conductivities, as shown in the inset. Notably, the thermal conductivities of the pristine Bi₂Te₃ nanoplates and the Sn-Bi₂Te₃ nanoplates are referred to from our previous report [58]. When electroless deposition was not performed on the Bi₂Te₃ nanoplates, the lattice thermal conductivity was 1.9 W/(m·K), corresponding to the lattice thermal conductivity of single-crystalline Bi₂Te₃ in the a- or b-axis directions (1.3–2.3 W/(m·K)) [71]. Following the metal deposition on the Bi₂Te₃ nanoplates, the lattice thermal conductivities decreased. In particular, the Sn-Bi₂Te₃ and Sn/Pd-Bi₂Te₃ nanoplates exhibited the lowest value of 1.1 W/(m·K), followed by the Sn/Pd-Bi₂Te₃ nanoplates with the value of 1.4 W/(m·K). This phenomenon indicates that the dispersion of metal nanoparticles on the nanoplate surface caused more phonons to scatter compared with a structure with a metal layer on the nanoplate surface, reducing the lattice thermal conductivity. This phenomenon was consistent with a previous report [72].

3.3. Thermoelectric Properties of Bi₂Te₃ Nanoplates Deposited with Different Metals

Figure 6 shows the in-plane thermoelectric properties of bulk-formed Bi₂Te₃ nanoplates deposited with different metals. In Figure 6a, the most negative Seebeck coefficient of −141 μV/K was observed in the Bi₂Te₃ nanoplates without metal deposition. The deposition of Sn and Pd on the Bi₂Te₃ nanoplates, respectively, decreased the negative Seebeck coefficient, which was strongly related to the carrier concentration. The relationship between the Seebeck coefficient S and carrier concentration n is given by Equation (1) [73,74], as follows:

$$S=\frac{8{\pi }^2{k}_{\mathrm{B}}^{2}T}{3{eh}^2}{m}^{*}{\left(\frac{\pi }{3n}\right)}^{\frac{2}{3}},$$

(1)

where m* denotes effective mass. According to the above equation, the Seebeck coefficient scales inversely with the carrier concentration. Therefore, the negative Seebeck coefficient decreased because the carrier concentration corresponding to the metal concentration increased, owing to the metal deposition. In particular, the Pd-Bi₂Te₃ nanoplates exhibited the lowest negative Seebeck coefficient of −98 μV/K, owing to the high carrier concentration as determined by EPMA. By contrast, the Seebeck coefficient of the Sn/Pd-Bi₂Te₃ nanoplates negatively increased to −127 μV/K, corresponding to the value located between that for the Bi₂Te₃ nanoplates without metal deposition and that for the plates with Sn deposition. This was because the Sn/Pd-Bi₂Te₃ nanoplates had higher carrier concentrations compared to that of the Bi₂Te₃ nanoplates without metal deposition and lower carrier concentrations compared to that of Sn-Bi₂Te₃ and Pd-Bi₂Te₃ nanoplates. The mechanism for the lower carrier concentration when Sn and Pd were deposited simultaneously compared to when each was deposited alone is unclear. However, a possible explanation is that the carriers generated by the deposition of Sn and Pd could be offset by the coexistence of these two metals [75].

In Figure 6b, the electrical conductivity of the Bi₂Te₃ nanoplates without metal deposition is shown to be 86 S/m. The highest electrical conductivity of 1572 S/m was observed for the Pd-Bi₂Te₃ nanoplates, followed by 447 S/m for the Sn-Bi₂Te₃ nanoplates, owing to the increase in the carrier concentration compared to the Bi₂Te₃ nanoplates without metal deposition. By contrast, the Sn/Pd-Bi₂Te₃ nanoplates exhibited the lowest electrical conductivity of 60 S/m, which was lower than that of the pristine Bi₂Te₃ nanoplates. This phenomenon cannot be explained by the magnitude of the carrier concentration alone. Electrical conductivity σ is strongly related to the carrier concentration n as well as mobility μ, as expressed by Equation (2):

$$\sigma =ne\mu .$$

(2)

Therefore, the Sn/Pd-Bi₂Te₃ nanoplates had a higher carrier concentration but lower mobility than the pristine Bi₂Te₃ nanoplates. The low mobility of Sn/Pd-Bi₂Te₃ nanoplates is due to electron scattering at the interfaces between the Sn and Pd layers.

The power factor of the Bi₂Te₃ nanoplates with different metal depositions is shown in Figure 6c. The power factor showed a trend similar to that of electrical conductivity because the inter-sample variation of the electrical conductivity was greater than the inter-sample variation of the Seebeck coefficient. The power factors of the Sn-Bi₂Te₃ and Pd-Bi₂Te₃ nanoplates were 6.4 and 15.2 μW/(m∙K²), respectively, which are higher than that of the pristine Bi₂Te₃ nanoplates at 1.7 μW/(m∙K²). Therefore, the power factor was improved by single-metal deposition on the nanoplatelets, and the power factor depended on the type of metals. However, the Sn/Pd-Bi₂Te₃ nanoplates exhibited the lowest power factor of 1.0 μW/(m∙K²), which was approximately 60% lower than that of pristine Bi₂Te₃ nanoplates, owing to the lowest electrical conductivity. Therefore, in this study, the simultaneous metal deposition did not enhance the power factor of the nanoplates.

The dimensionless ZT of the Bi₂Te₃ nanoplates with different metal depositions is shown in Figure 6d. The Pd-Bi₂Te₃ nanoplates exhibited the highest ZT of 32 × 10⁻³. This is because they had extremely high electrical conductivity among the four samples even though they did not have the lowest thermal conductivity or the highest negative Seebeck coefficient. The second highest ZT of 18 × 10⁻³ was exhibited by the Sn-Bi₂Te₃ nanoplates. Due to the lowest thermal conductivity of the Sn-Bi₂Te₃ nanoplates, the difference in ZT between the Pb-Bi₂Te₃ and Sn-Bi₂Te₃ nanoplates decreased compared to the difference in power factor between them. The ZT of the Sn/Pd-Bi₂Te₃ nanoplates was 2.7 × 10⁻³, which was the same value as the pristine Bi₂Te₃ nanoplates owing to the low thermal conductivity of the Sn/Pd-Bi₂Te₃ nanoplates. Therefore, all cases of the metal deposition on the Bi₂Te₃ nanoplates did not decrease the ZT. However, the highest ZT in this study was lower than that for the sintered Bi₂Te₃ bulk samples (ZT = 0.8 at 300 K) [76]. This was because the samples in this study underwent no sintering (e.g., spark plasma sintering) to assess the thermoelectric properties of the nanoplates by representing their properties as close as possible to those of the pristine nanoplates without melting. This study reveals the effects of different metal depositions on Bi₂Te₃ nanoplates, with implications regarding their thermoelectric performance. In addition, we consider there is room to improve thermoelectric performance by optimizing the type of metal and the amount of metal deposited. In the future, thermoelectric performance will be further improved by sintering nanoplates deposited with appropriate metals.

4. Conclusions

The studied Bi₂Te₃ nanoplates were deposited with Sn, Pd, and Sn/Pd, to modify their surface morphologies and improve their thermoelectric properties. The studied Bi₂Te₃ nanoplates were prepared using solvothermal techniques and metal deposition was performed using electroless deposition. Sn nanoparticles at a concentration of approximately 1 at% were deposited on the prepared Bi₂Te₃ nanoplates. For Pd deposition, a Pd layer with a concentration of approximately 3 at% was deposited on the Bi₂Te₃ nanoplates. When Sn and Pd were sequentially deposited on the nanoplate surface, the surface morphology was similar to that of the Sn-deposited nanoplates where the Sn and Pd concentrations were 1 at% and less than 1 at%, respectively. In order to assess the thermoelectric properties of the nanoplates by representing their properties as close as possible to those of the inherent nanoplate but not the melting structures, the nanoplate assemblies were compressed into thin bulk shapes at 300 K. The lattice thermal conductivity was affected by the surface morphology of the nanoplates. The Sn-Bi₂Te₃ and Sn/Pd-Bi₂Te₃ nanoplates exhibited the lowest value of 1.1 W/(m·K), indicating that the dispersion of metal nanoparticles on the nanoplate surface caused more phonon scattering. However, the Pd-Bi₂Te₃ nanoplates exhibited the highest power factor and dimensionless ZT owing to their significantly high electrical conductivities.