Minute-Made, High-Efficiency Nanostructured Bi2Te3 via High-Throughput Green Solution Chemical Synthesis

Scalable synthetic strategies for high-quality and reproducible thermoelectric (TE) materials is an essential step for advancing the TE technology. We present here very rapid and effective methods for the synthesis of nanostructured bismuth telluride materials with promising TE performance. The methodology is based on an effective volume heating using microwaves, leading to highly crystalline nanostructured powders, in a reaction duration of two minutes. As the solvents, we demonstrate that water with a high dielectric constant is as good a solvent as ethylene glycol (EG) for the synthetic process, providing a greener reaction media. Crystal structure, crystallinity, morphology, microstructure and surface chemistry of these materials were evaluated using XRD, SEM/TEM, XPS and zeta potential characterization techniques. Nanostructured particles with hexagonal platelet morphology were observed in both systems. Surfaces show various degrees of oxidation, and signatures of the precursors used. Thermoelectric transport properties were evaluated using electrical conductivity, Seebeck coefficient and thermal conductivity measurements to estimate the TE figure-of-merit, ZT. Low thermal conductivity values were obtained, mainly due to the increased density of boundaries via materials nanostructuring. The estimated ZT values of 0.8–0.9 was reached in the 300–375 K temperature range for the hydrothermally synthesized sample, while 0.9–1 was reached in the 425–525 K temperature range for the polyol (EG) sample. Considering the energy and time efficiency of the synthetic processes developed in this work, these are rather promising ZT values paving the way for a wider impact of these strategic materials with a minimum environmental impact.


Introduction
Our heavy dependence on non-renewable sources have led to a global energy crisis and as the consumption of these sources has increased, the prospect of their availability for a long term has reduced [1]. Certainly, to meet future energy challenges and climate change it is of paramount importance to develop renewable, and sustainable energy solutions. Thermoelectric (TE) materials have received significant attention because of their ability to convert waste heat to electrical power directly [2]. They can be used to harvest low-and high-grade waste-heat effectively, which requires the selection of proper material for the intended temperature range. As TE materials can be used for power generation, they can also be used for cooling or thermal management. TE devices have attracted significant interest because of their potential applications in energy harvesting, advantages of using MW heating are higher yields, and since the reaction is kinetically driven it ensures a narrow-size distribution of the particles [72]. A variety of nanomaterials have been synthesized with MWs, including 0D [63,[72][73][74][75][76][77][78], 1D [67,68,79,80] and 2D structures [81]. MW-assisted synthesis is also a sustainable, environmentally friendly, and economically viable heating method for bulk commercial scale production of strategic nanomaterials.
Most chemical syntheses use a large amount of conventional volatile or other organic chemicals, which create a pressing environmental concern. Therefore, the aim should be to minimize the use of toxic chemicals and follow the principles of green chemistry to synthesize nanomaterials [82]. The selection of the solvent for a reaction can dramatically affect the reaction outcome. Ethylene glycol (EG) is an abundant resource in the chemical industry, which can be produced from renewable biomass [83]. It is an odourless, non-volatile, low toxic solvent with a low dielectric constant (37 at room temperature), which has enormous applications in various industrial processes [84], including the polyol synthesis [64]. Water is the most environmentally friendly (i.e., green) solvent with a high dielectric constant (78 at room temperature), which is very effective for transferring electromagnetic energy into heat and thus driving the reduction reactions effectively [85]. Under MW heating, the superheating of water causes the hydrogen bonds to break easily, making it less polar and the self-ionization of water increases with temperature [62,86]. Due to superheating, the dielectric constant of water decreases at high temperatures (above its boiling point) and this influences the growth of the nanoparticles [86]. It has been reported that Bi 2 Te 3 particles grow more steadily in a solvent with low dielectric constant than in a solvent with a high dielectric constant [41]. Therefore, this makes the combination of MW heating and hydrothermal methods a suitable chemical synthesis route for the growth of nanocrystals. The pH of the reaction medium is key for the synthesis of the Bi 2 Te 3 nanocrystals. It has been reported that an alkaline medium (pH < [11][12] is required for the formation of Bi 2 Te 3 nanostructures [26,43,52,53,57,87]. The alkaline medium helps the tellurium to disperse in the reaction medium as telluride ions (Te 2− ), which is beneficial for the formation of the nanocrystals [46]. A growth promoting agent, like ethylenediaminetetraacetic acid (EDTA), is required to form the bismuth-antimony complexes which self-assemble to lamellar phases leading to the formation of Bi 2 Te 3 sheet nuclei, which facilitates the formation of nanosheets, nanoflakes and nanotubes [53,59]. The activity of EDTA is also dependent on the pH of the solution and by changing the pH, different morphologies of Bi 2 Te 3 have been obtained [51][52][53][54][55][56]87]. Control of the reaction time and temperature are essential in determining the structure and morphology of the nanocrystals [49,57]. It can be deduced that low temperature is suitable for the reaction of nanoparticles, nanoplates and nanotubes and as the reaction temperature increases, the dimensionality changes from 0D to 1D. The effect of reaction time and temperature on the morphology, structure, and crystallinity of Bi 2 Te 3 have been comprehensively summarized in Table S1.
We aim to provide facile and sustainable synthetic routes for large-scale production of Bi 2 Te 3 nanomaterials with promising TE performance. The design of the experiments has been done in-line with the green chemistry perspectives considering the solvents, precursors and the energy source. Taking advantage of the unique MW dielectric heating phenomenon, rapid heating and short reaction time, hexagonal nanoplatelets of Bi 2 Te 3 were obtained. These low dimensional structures have been effective in enhancing the charge transport properties and electrical conductivity, while reducing the thermal conductivity. Water and ethylene glycol are used as solvents making the reaction green, safe and nonhazardous, producing high purity TE nanopowders in a matter of minutes due to very effective volume heating-despite their significantly different dielectric constants. Besides yielding a large quantity of the investigated materials per batch, surface chemistry and the processing of these materials towards compacts are described, resulting in high ZT values of 1.03 and 0.88 for Bi 2 Te 3 materials synthesized through the polyol and the hydrothermal routes, respectively. Chemicals used for the synthesis were of analytical grade obtained from Sigma Aldrich (Stockholm, Sweden) and have been used as received without further purification: Bismuth chloride (BiCl 3 , 98% purity), sodium tellurite, (Na 2 TeO 3 , 99% purity), sodium hydroxide (NaOH, 97.0% purity) and sodium borohydride (NaBH 4 , 98% purity). BiCl 3 and Na 2 TeO 3 were used as the precursors for Bi and Te ions, respectively. Deionized (DI) water was used as the solvent, EDTA as the structure-directing agent, NaBH 4 as the reducing agent and NaOH to control the pH of the reaction media. The synthesis was performed in a Milestone flexiWAVE MW system (MILESTONE Srl, Sorisole, Italy) (900 watt magnetrons for a total of 1800 watt). One-pot synthesis was achieved by mixing all the precursors in the same media. For the synthesis of Bi 2 Te 3 , Bi:Te molar ratio of 2:3 was used. In a 100 mL Teflon vial, BiCl 3 , Na 2 TeO 3 , EDTA, and NaBH 4 were mixed where the pH was adjusted to 10.5-11 by the addition of NaOH. Two parallel reaction vessels were used in a single run. The reaction vials were kept under constant magnetic stirring for 30 min to ensure a homogeneous mixing. This was then subjected to MW heating to 220 • C with a ramp time of 4 min and dwell time of 2 min. The solution was then allowed to cool down to room temperature and a clear phase separation was observed with the product settled at the bottom, separated from water. The particles formed were easily separated from the reaction mixture by centrifugation and were then washed with isopropanol, acetone and water. Thereafter, the powder was dried in a vacuum oven at 60 • C. A schematic of the hydrothermal synthesis process is given in Figure 1. The sample is designated as hydro-Bi 2 Te 3 throughout the manuscript.
a Milestone flexiWAVE MW system (MILESTONE Srl, Sorisole, Italy) (900 watt magnetrons for a total of 1800 watt). One-pot synthesis was achieved by mixing all the precursors in the same media. For the synthesis of Bi2Te3, Bi:Te molar ratio of 2:3 was used. In a 100 mL Teflon vial, BiCl3, Na₂TeO₃, EDTA, and NaBH4 were mixed where the pH was adjusted to 10.5-11 by the addition of NaOH. Two parallel reaction vessels were used in a single run. The reaction vials were kept under constant magnetic stirring for 30 min to ensure a homogeneous mixing. This was then subjected to MW heating to 220 °C with a ramp time of 4 min and dwell time of 2 min. The solution was then allowed to cool down to room temperature and a clear phase separation was observed with the product settled at the bottom, separated from water. The particles formed were easily separated from the reaction mixture by centrifugation and were then washed with isopropanol, acetone and water. Thereafter, the powder was dried in a vacuum oven at 60 °C. A schematic of the hydrothermal synthesis process is given in Figure 1. The sample is designated as hydro-Bi2Te3 throughout the manuscript.

Synthesis by Polyol Method
All the chemicals used for the synthesis reaction were of analytical grade obtained from Sigma Aldrich and were used as received without further purification: BiCl3 (98% purity), Te powder (99.8% purity), ethylene glycol (EG, 99% purity thioglycolic acid (TGA, 98% purity), and trioctyl phosphine (TOP, 90% purity). The synthesis was performed in the same Milestone flexiWAVE MW system, as described in our earlier work [64]. Typically, BiCl3 was dissolved in EG as a source of Bi 3+ ions, by high power sonication for 2 min, then TGA was added under continuous stirring. Te precursor solution was prepared by complexing the Te powder with TOP using MW assisted heating for 90 s at 220 °C. The molar ratio of 2:3 was used for the concentration of Bi:Te respectively. Both precursors were mixed in a 100 mL Teflon vial, then loaded into the MW system heated to 220 °C with a ramp time of 4 min and dwell time of 2 min. Thereafter, the powders were washed several times with acetone and isopropanol and dried in a vacuum oven at 60 °C. A schematic of the polyol synthesis procedure is given in Figure 1. The sample is designated as polyol-Bi2Te3 throughout the manuscript.

Synthesis by Polyol Method
All the chemicals used for the synthesis reaction were of analytical grade obtained from Sigma Aldrich and were used as received without further purification: BiCl 3 (98% purity), Te powder (99.8% purity), ethylene glycol (EG, 99% purity thioglycolic acid (TGA, 98% purity), and trioctyl phosphine (TOP, 90% purity). The synthesis was performed in the same Milestone flexiWAVE MW system, as described in our earlier work [64]. Typically, BiCl 3 was dissolved in EG as a source of Bi 3+ ions, by high power sonication for 2 min, then TGA was added under continuous stirring. Te precursor solution was prepared by complexing the Te powder with TOP using MW assisted heating for 90 s at 220 • C. The molar ratio of 2:3 was used for the concentration of Bi:Te respectively. Both precursors were mixed in a 100 mL Teflon vial, then loaded into the MW system heated to 220 • C with a ramp time of 4 min and dwell time of 2 min. Thereafter, the powders were washed several times with acetone and isopropanol and dried in a vacuum oven at 60 • C. A schematic of the polyol synthesis procedure is given in Figure 1. The sample is designated as polyol-Bi 2 Te 3 throughout the manuscript.

Consolidation of the Powders via Spark Plasma Sintering
The powders were then compacted by spark plasma sintering (SPS, Dr Sinter 825, Fuji Electronic Industrial Co. Ltd., Tokyo, Japan), to obtain nanostructured bulk solid materials, in a 15 mm diameter die to form pellets for TE transport property measurements (see Figure S1). Bi 2 Te 3 powders from both the hydrothermal and the polyol syntheses methods were subjected to the same SPS process parameters for a fair comparison (see Table 1). The sintering temperature was set to 400 • C with the ramping heat rate of 30 • C/min, and 50 MPa pressure applied with the same ramping rate, as to reach the maximum pressure when the temperature reached a maximum 400 • C. The sintered pellets were cooled down to room temperature and polished for further structural analysis and TE transport measurements. The density (d pellet ) of the sintered specimens was measured with the Archimedes method and the packing density (ρ%) was calculated with reference to the bulk density (d bulk ) of 7.86 g/cm 3 .

Structural and Morphological Characterization
The phase structures are investigated by X-ray powder diffraction (XRPD) to identify the crystal structure, crystallinity, and the lattice parameters and compare the purity of the synthesized materials. XRPD analysis has been performed by using a Philips PANalytical X'Pert Pro Powder Diffractometer (Malvern Panalytical Ltd., Malvern, UK) with Cu-Kα 1 radiation (λ = 1.54059 Å). A scan speed of 0.04 • /s was used in continuous scan with a rotating sample holder. Morphology and microstructure analyses of the materials were performed using scanning electron microscopy (SEM; FEI Nova 200 (FEI Company, Hillsboro, OR, USA) and FE-SEM, TESCANBRNO-Mira3 LMU (TESCAN ORSAY HOLDING, Brno, Czech Republic) equipped with a 30 kV SEM FEG column. During the operation of the instrument, the working distance was maintained at 5 mm and the accelerating voltage was 10-15 kV. Transmission electron microscopy (TEM, JEM-2100F, JEOL Ltd., Tokyo, Japan) analysis was performed by pipetting a 100 µL aliquot of the powder samples dispersed in DI water onto a 200 mesh Cu-grid and drying prior to analysis. Zeta (ξ) potential of various samples were measured at room temperature in triplicates with Malvern Zetasizer Nano ZS90 at an incident angle of 90 • . All samples were dispersed in DI H 2 O, and titrated with 0.01 M HCl or 0.01 M NaOH to obtain pH dependent surface charge, and eventually to determine the isoelectric point (iep). The term iep represents a specific pH value where the net surface charge of the particle is zero, which is strongly linked to the surface chemistry and functionality on the nanoparticles.

Electronic and Thermal Transport Property Measurements
Electrical conductivity (σ) and Seebeck coefficient (S) measurements were performed simultaneously by using the commercial ZEM3 ULVAC-RIKO (ULVAC, Methuen, MA, USA) system. The measurements were conducted on SPS consolidated pellets in the system for both (hydrothermal and polyol) samples. The total thermal conductivity κ tot of the pellets were calculated according to the equation κ tot = C p ·α·ρ, where C p , α and ρ are specific heat capacity, thermal diffusivity, and density of mass respectively. Heat capacity measurements were performed with differential scanning calorimetry (DSC, PT1000 Linseis, Linseis Messgeraete GmbH, Selb, Germany). For measurement of the thermal diffusivity, α, 2 mm thick disk-shaped samples were used in the laser flash analysis (LFA 1000, Linseis) system.

X-ray Photoelectron Spectroscopy
X-ray photoelectron spectroscopy (XPS) was used for surface analysis. The XPS spectra were acquired using the PREVAC 426 system configuration, equipped with SCIENTA R3000 hemispherical photoelectron spectrometer (Scienta Omicron, Uppsala, Sweden) and monochromatic Al source. The base pressure during measurements was better than 3 × 10 −10 mbar. All acquired spectra were calibrated to adventitious carbon C1s at 285 eV. After subtraction of the Shirley-type background, the core-level spectra were decomposed into main components with mixed Gaussian-Lorentzian lines (70% G + 30% L for majority of photo-peaks) by a non-linear least squares curve-fitting procedure, using CasaXPS software (version 2.3.18, Casa Software Ltd., Cheshire, UK).

Structural Analysis
Structural analysis of as-synthesized materials through MW-assisted hydrothermal and polyol routes before and after SPS sintering are performed using XRPD. The diffraction patterns are presented in Figure 2, where the patterns are normalized for the most intense peak with Miller indices (015). As for the SPS sintered samples, XRPD was performed on the pellets' surface perpendicular to the sintering direction (pressing axis/direction is schematically shown in the inset of Figure 2a). The crystalline phases are indexed to Bi 2 Te 3 (ICDD: 01-089-2009) with rhombohedral crystal structure, and the corresponding indexing of the Bragg diffractions with relevant Miller indices are shown on the diffraction patterns in Figure 2. When as-made samples are compared, it can be clearly seen that polyol-Bi 2 Te 3 sample has more intense peaks for the Bragg diffractions with the same Miller indices (e.g., (1010), (110), (205)), which reveal a higher crystallinity as compared to the hydro-Bi 2 Te 3 sample. Upon sintering, the significant increase in the relative intensity of diffraction peaks with (001) index reveal a higher degree of texturing in the hydro-Bi 2 Te 3 sample. The intensity ratio of (006) peak to (015) peak could be used for the quantification of the texturing. The hydro-Bi 2 Te 3 sample has the peak intensity ratio (I 006/ I 015 ) of 0.38, while it is 0.19 for the polyol-Bi 2 Te 3 sample (see Figure S2). This indicates that after the SPS process the c-axis of the grains was preferentially oriented parallel to the pressing direction, resulting in a higher degree of texturing in the Hydro-Bi 2 Te 3 . Average crystallite size of the sintered samples was estimated from the XRD data using the Williamson-Hall model [88], as 231 nm for the polyol-Bi 2 Te 3 sample, and 295 nm for the hydro-Bi 2 Te 3 sample. A minor/secondary phase of Bi 2 TeO 5 observed in the hydro-Bi 2 Te 3 sample upon sintering. This phase was also observed in some earlier reports [89][90][91], and is estimated as 5.8% of the overall composition through the Rietveld refinement (see Figure S3). Possible reasons to observe Bi 2 TeO 5 (ICDD: 00-024-0154) impurity phase after the SPS process may be either due to crystallization of the pre-existing secondary phase upon sintering, or to the reaction of the powder with the trace amount of oxygen in the sintering medium.

Surface Analysis
XPS analysis was performed on as made powder samples synthesized through hydrothermal and polyol routes. The data are graphically presented in Figure 3 and are summarized in Table 2. It should be noted that in both the samples tellurium (Te) exists in two states, metallic and oxide. In the case of bismuth (Bi) content, the hydro-Bi2Te3 sample consists of Bi in two states, metallic and oxide, while in the case of the polyol-Bi2Te3 sample the element Bi exists only in the oxide state. The speciation of carbon (C) is slightly different in the materials, where it is a mixture of mainly C-O, and C-C for polyol-Bi2Te3, while it consists of C-O, C-C and higher O-C=O content for the hydro-Bi2Te3 sample. The chemical speciation of C on the surface can be seen from the XPS fitting details, where a high percentage of C-O was identified in the hydro-Bi2Te3 sample, followed by C=O groups. This may be due to a strong conjugation of the terminal carboxyl groups to the nanoparticles' surface through a bidentate or bridging configuration. In the polyol sample though, C is dominantly indexed to carboxyl, O-C=O, which is reasonable considering the conjugation of thioglycolic acid to the NP surface being through the sulfur (S) atoms. C content in the hydro-Bi2Te3 sample is higher than that of the polyol sample. This may originate from the precursors used in the synthetic process. Indeed, trace elements originating from the precursors are detected in the spectra of the respective samples. EDTA was used as a surface directing agent in the hydrothermal synthesis, which is known to be a hexadentate ligand at pH above 10. Naturally, it stays on the formed hexagonal platelets as an integral part, despite several washing-re-dispersing steps. The traces of EDTA can be seen through the nitrogen (N) signal (400.5 eV) in the XPS spectrum of the hydro-Bi2Te3 sample ( Figure  3a). In the polyol process, thioglycolic acid was used as a directing agent and its traces are also visible through the sulfur (S) signal (229.3 eV) in the XPS spectrum of this particular sample (Figure 3a).

Surface Analysis
XPS analysis was performed on as made powder samples synthesized through hydrothermal and polyol routes. The data are graphically presented in Figure 3 and are summarized in Table 2. It should be noted that in both the samples tellurium (Te) exists in two states, metallic and oxide. In the case of bismuth (Bi) content, the hydro-Bi 2 Te 3 sample consists of Bi in two states, metallic and oxide, while in the case of the polyol-Bi 2 Te 3 sample the element Bi exists only in the oxide state. The speciation of carbon (C) is slightly different in the materials, where it is a mixture of mainly C-O, and C-C for polyol-Bi 2 Te 3 , while it consists of C-O, C-C and higher O-C=O content for the hydro-Bi 2 Te 3 sample. The chemical speciation of C on the surface can be seen from the XPS fitting details, where a high percentage of C-O was identified in the hydro-Bi 2 Te 3 sample, followed by C=O groups. This may be due to a strong conjugation of the terminal carboxyl groups to the nanoparticles' surface through a bidentate or bridging configuration. In the polyol sample though, C is dominantly indexed to carboxyl, O-C=O, which is reasonable considering the conjugation of thioglycolic acid to the NP surface being through the sulfur (S) atoms. C content in the hydro-Bi 2 Te 3 sample is higher than that of the polyol sample. This may originate from the precursors used in the synthetic process. Indeed, trace elements originating from the precursors are detected in the spectra of the respective samples. EDTA was used as a surface directing agent in the hydrothermal synthesis, which is known to be a hexadentate ligand at pH above 10. Naturally, it stays on the formed hexagonal platelets as an integral part, despite several washing-re-dispersing steps. The traces of EDTA can be seen through the nitrogen (N) signal (400.5 eV) in the XPS spectrum of the hydro-Bi 2 Te 3 sample (Figure 3a). In the polyol process, thioglycolic acid was used as a directing agent and its traces are also visible through the sulfur (S) signal (229.3 eV) in the XPS spectrum of this particular sample (Figure 3a).  It is important to note that the relative ratio of metallic vs. oxide phases on the NPs' surface reveals that the polyol-Bi2Te3 sample is bearing more oxide phase than the hydro-Bi2Te3 sample. In the case of Bi in particular, the spectrum for polyol-Bi2Te3 (Figure 3c2) indicates Bi2O3 as the only identified phase. It is interesting to note this as water has been generally considered a stronger oxidizing solvent.  It is important to note that the relative ratio of metallic vs. oxide phases on the NPs' surface reveals that the polyol-Bi 2 Te 3 sample is bearing more oxide phase than the hydro-Bi 2 Te 3 sample. In the case of Bi in particular, the spectrum for polyol-Bi 2 Te 3 (Figure 3c 2 ) indicates Bi 2 O 3 as the only identified phase. It is interesting to note this as water has been generally considered a stronger oxidizing solvent.
Zeta (ξ) potential analysis was performed on both hydrothermal and polyol as-made powders dispersed in DI water. Isoelectric point (iep) is strongly dependent on the surface chemistry of the particles, and the dispersant. Determining the iep of as-synthesized Bi 2 Te 3 materials can reveal the information about the surface charge and chemistry of the as-made nanoparticles. For the polyol synthesized sample, the iep was reached around pH 6.3, while it was about 6.8 for the hydrothermally synthesized sample (Figure 4). These values are not significantly different, and both may derive from the high C content, carbonyl and carboxyl bearing surface groups. The small difference can be attributed to the TeO 2 , which has an iep of <3. The higher TeO 2 content of polyol-Bi 2 Te 3 sample can explain the slightly lower iep value. Zeta (ξ) potential analysis was performed on both hydrothermal and polyol as-made powders dispersed in DI water. Isoelectric point (iep) is strongly dependent on the surface chemistry of the particles, and the dispersant. Determining the iep of as-synthesized Bi2Te3 materials can reveal the information about the surface charge and chemistry of the asmade nanoparticles. For the polyol synthesized sample, the iep was reached around pH 6.3, while it was about 6.8 for the hydrothermally synthesized sample (Figure 4). These values are not significantly different, and both may derive from the high C content, carbonyl and carboxyl bearing surface groups. The small difference can be attributed to the TeO2, which has an iep of <3. The higher TeO2 content of polyol-Bi2Te3 sample can explain the slightly lower iep value.

Microstructure Analysis
Morphology and microstructure analyses of as-made nanoparticles and sintered samples were performed using SEM and few micrographs for powder samples at different

Microstructure Analysis
Morphology and microstructure analyses of as-made nanoparticles and sintered samples were performed using SEM and few micrographs for powder samples at different magnifications for hydro-Bi 2 Te 3 and polyol-Bi 2 Te 3 samples are given in Figure 5. Both the samples show dominantly hexagonal platelet morphology for the nanoparticles formed. The Bi 2 Te 3 crystals grow predominantly along the ab plane direction because the strong covalent bonds between Bi-Te can be easily extended along the plane. The platelets obtained in the polyol route were observed to be thinner than those from the hydrothermal route. In general, particles obtained from the hydrothermal route are smaller than those from the polyol route. The lateral dimensions show a great variation in the range 20-250 nm for the hydrothermal and 10-1000 nm for the polyol sample, respectively. These variations are attributed to the different solvents and the stabilizing ligands used in the two syntheses process, which have had a different impact on the nucleation and growth kinetics. Looking closer to the XRPD patterns in Figure 2, it can be assessed that the diffraction peak at 41 • , corresponding to the 110 Bragg diffraction, is much higher for the polyol sample, ascribed to larger lateral dimension (ab plane) of the platelets. A closer investigation was performed by TEM analysis and micrographs along with selected-area electron diffraction (SAED) patterns are shown in Figure 5c,d,g,h for hydro-Bi 2 Te 3 and polyol-Bi 2 Te 3 samples, respectively. Despite their large lateral dimensions, the platelets exhibit almost perfectly single crystalline character. There are some platelets aligned sideways, which reveal clearly the lamella-like structure along the c-axis, with a typical thickness of about 1 nm, as shown in Figure 5c,g. These layers are five atomic layers (Te-Bi-Te-Bi-Te) thick, which are known as the quintuple layer, held together by the van der Waals forces. SAED patterns, presented in Figure 5d,h, are indexed to typical Bragg diffractions of rhombohedral Bi 2 Te 3 (ICDD: 01-089-2009) as was also identified in the XRPD patterns ( Figure 2). syntheses process, which have had a different impact on the nucleation and growth kinetics. Looking closer to the XRPD patterns in Figure 2, it can be assessed that the diffraction peak at 41°, corresponding to the 110 Bragg diffraction, is much higher for the polyol sample, ascribed to larger lateral dimension (ab plane) of the platelets. A closer investigation was performed by TEM analysis and micrographs along with selected-area electron diffraction (SAED) patterns are shown in Figure 5c,d,g,h for hydro-Bi2Te3 and polyol-Bi2Te3 samples, respectively. Despite their large lateral dimensions, the platelets exhibit almost perfectly single crystalline character. There are some platelets aligned sideways, which reveal clearly the lamella-like structure along the c-axis, with a typical thickness of about 1 nm, as shown in Figure 5c,g. These layers are five atomic layers (Te-Bi-Te-Bi-Te) thick, which are known as the quintuple layer, held together by the van der Waals forces. SAED patterns, presented in Figure 5d,h, are indexed to typical Bragg diffractions of rhombohedral Bi2Te3 (ICDD: 01-089-2009) as was also identified in the XRPD patterns ( Figure 2). The as-synthesized nanostructured TE materials were consolidated using the SPS. The critical consolidation parameters (sintering temperature, holding time, pressure, and heating rates), and characteristics of the resultant pellets are presented in detail in Table  1. Despite both the samples being processed under identical conditions, the pellets reveal a big difference in their packing density; a packing density of about 80% for the polyol-Bi2Te3 and 90% for the hydro-Bi2Te3 sample. Densification under SPS for a given material as a function of particle size has been studied by Diouf et al. [93], where they reported a decrease in densification with increasing particle size of the specimen. In the case of the The as-synthesized nanostructured TE materials were consolidated using the SPS. The critical consolidation parameters (sintering temperature, holding time, pressure, and heating rates), and characteristics of the resultant pellets are presented in detail in Table 1. Despite both the samples being processed under identical conditions, the pellets reveal a big difference in their packing density; a packing density of about 80% for the polyol-Bi 2 Te 3 and 90% for the hydro-Bi 2 Te 3 sample. Densification under SPS for a given material as a function of particle size has been studied by Diouf et al. [93], where they reported a decrease in densification with increasing particle size of the specimen. In the case of the Bi 2 Te 3 samples, particle size of the polyol-Bi 2 Te 3 sample was assessed to be larger than the hydro-Bi 2 Te 3 sample, which may be the cause of the observed difference in the densification, in agreement with the earlier findings [93]. Microstructure of the pellets are displayed in Figure 6. The hydro-Bi 2 Te 3 sample shows a clear long-range stacking of hexagonal platelets, perpendicular to the c-axis, revealing a higher degree of texturing as compared to the polyol-Bi 2 Te 3 sample. This effect was observed in XRD (Figure 2), with a significant increase in the intensity ratio (I 006 /I 015 ) of the diffraction peaks for the hydro-Bi 2 Te 3 sample (see Figure S2).
hydro-Bi2Te3 sample, which may be the cause of the observed difference in the densification, in agreement with the earlier findings [93]. Microstructure of the pellets are displayed in Figure 6. The hydro-Bi2Te3 sample shows a clear long-range stacking of hexagonal platelets, perpendicular to the c-axis, revealing a higher degree of texturing as compared to the polyol-Bi2Te3 sample. This effect was observed in XRD (Figure 2), with a significant increase in the intensity ratio (I006/I015) of the diffraction peaks for the hydro-Bi2Te3 sample (see Figure S2).

Electronic Transport Property Evaluation
The electronic transport data in the temperature range of 300-523 K are presented in Figure 7. Table 3 summarizes the transport data at room temperature and at the temperature where the maximum ZT was obtained. No adverse effects in the transport properties were observed due to the presence of surface oxides in the as-made powders. The electrical conductivity (σ) is plotted in Figure 7a, where samples show reduction of σ by increasing temperature, which is typical of heavily doped semiconductor, or metallic character. The highest electrical conductivity value at room temperature was measured for hydro-Bi2Te3. The σ for hydro-Bi2Te3 changes from about 1300 S/cm at 300 K to about 600 S/cm at 523 K. This is likely due to the higher density of the compacted materials and larger layered structure of the hydro-Bi2Te3 in comparison to the polyol-Bi2Te3 sample. The secondary phase of Bi2TeO5 in the hydro-Bi2Te3 sample is an oxide with a larger band-gap than Bi2Te3 [90,94]. Due to its low content within the main phase, it is not expected to have a percolation path to influence the electronic transport properties significantly. This can also be seen from the initial high σ of hydro-Bi2Te3. However, this oxide phase may act as phonon scattering canters due to its different anisotropic crystal structure, influencing the thermal conductivity. The σ values change between 1100 S/cm and 700 S/cm for the polyol-Bi2Te3 sample, in the same temperature range. Reduction in the electrical conductivity over 300 K is generally a signature of increasing electron-phonon scattering. There is an interesting crossover at 373 K, where the σ for the hydro-Bi2Te3 sample goes below the polyol sample. This might be due to the electron-phonon scattering process being higher than the polyol-Bi2Te3 sample. The electrical conductivity of both samples are higher than those of other solution derived samples [95][96][97] comparable with that of p-type alloy ingot (on the level of 1 × 10 3 S/cm at 300 K) [98].
The Seebeck coefficient (S) is the key parameter to demonstrate the type of transport. Figure 7b shows the obtained S values for both samples, which have negative values, revealing n-type conduction. The polyol-Bi2Te3 sample starts the S from −120 μV/K at room temperature and increases in magnitude up to −165 μV/K at 525K, while the hydro-Bi2Te3 sample does not show a strong temperature dependence, displaying values between −140 μV/K and −150 μV/K. Electronic transport properties are strongly correlated with the microstructure, as well as charge carrier density, defects, and scattering events. This correlation is clearly visible for the hydro-Bi2Te3 sample, where a decreasing σ is accompanied

Electronic Transport Property Evaluation
The electronic transport data in the temperature range of 300-523 K are presented in Figure 7. Table 3 summarizes the transport data at room temperature and at the temperature where the maximum ZT was obtained. No adverse effects in the transport properties were observed due to the presence of surface oxides in the as-made powders. The electrical conductivity (σ) is plotted in Figure 7a, where samples show reduction of σ by increasing temperature, which is typical of heavily doped semiconductor, or metallic character. The highest electrical conductivity value at room temperature was measured for hydro-Bi 2 Te 3 . The σ for hydro-Bi 2 Te 3 changes from about 1300 S/cm at 300 K to about 600 S/cm at 523 K. This is likely due to the higher density of the compacted materials and larger layered structure of the hydro-Bi 2 Te 3 in comparison to the polyol-Bi 2 Te 3 sample. The secondary phase of Bi 2 TeO 5 in the hydro-Bi 2 Te 3 sample is an oxide with a larger band-gap than Bi 2 Te 3 [90,94]. Due to its low content within the main phase, it is not expected to have a percolation path to influence the electronic transport properties significantly. This can also be seen from the initial high σ of hydro-Bi 2 Te 3 . However, this oxide phase may act as phonon scattering canters due to its different anisotropic crystal structure, influencing the thermal conductivity. The σ values change between 1100 S/cm and 700 S/cm for the polyol-Bi 2 Te 3 sample, in the same temperature range. Reduction in the electrical conductivity over 300 K is generally a signature of increasing electron-phonon scattering. There is an interesting crossover at 373 K, where the σ for the hydro-Bi 2 Te 3 sample goes below the polyol sample. This might be due to the electron-phonon scattering process being higher than the polyol-Bi 2 Te 3 sample. The electrical conductivity of both samples are higher than those of other solution derived samples [95][96][97] comparable with that of p-type alloy ingot (on the level of 1 × 10 3 S/cm at 300 K) [98].

Thermal Conductivity
The total thermal conductivity values were calculated using the thermal diffusivity, heat capacity and mass density values of the samples. The total thermal conductivity of hydro-Bi2Te3 is slightly lower than polyol-Bi2Te3 at T = 293 K, as seen in Figure 8a. They  Table 3. A summary of the measured TE transport properties of Bi 2 Te 3 samples synthesized through MW-assisted hydrothermal and polyol routes: total thermal conductivity (κ tot ), Seebeck coefficient (S), electrical conductivity (σ), power factor (PF), and TE figure of merit (ZT) at the given temperature (T) values.

Sample
T κ tot S σ PF ZT (K) (W/(m·K)) (µV/K) (S/m) (µW/(K 2 ·cm)) The Seebeck coefficient (S) is the key parameter to demonstrate the type of transport. Figure 7b shows the obtained S values for both samples, which have negative values, revealing n-type conduction. The polyol-Bi 2 Te 3 sample starts the S from −120 µV/K at room temperature and increases in magnitude up to −165 µV/K at 525 K, while the hydro-Bi 2 Te 3 sample does not show a strong temperature dependence, displaying values between −140 µV/K and −150 µV/K. Electronic transport properties are strongly correlated with the microstructure, as well as charge carrier density, defects, and scattering events. This correlation is clearly visible for the hydro-Bi 2 Te 3 sample, where a decreasing σ is accompanied by an increase in the magnitude of the S. The band-gap (E g ) can be estimated from the S as it is sensitive to changes in the density of states near the Fermi level [99,100]. The E g for the materials were calculated using the electron charge (q), absolute value of the maximum Seebeck, |S max |, value and the corresponding temperature T max , using the following equation: The estimated E g from the S max for the hydro-Bi 2 Te 3 sample is 0.11 eV and that for the polyol-Bi 2 Te 3 sample is 0.17 eV. The estimated E g for the Hydro-Bi 2 Te 3 sample is in good agreement with the calculated E g of 0.11 eV by Mishra et al., using the density-functional theory with the spin-orbit interaction included [101]. However, the E g for the hydro-Bi 2 Te 3 sample is lower than the bulk Bi 2 Te 3 value of E g = 0.15 eV [102], while that of the polyol-Bi 2 Te 3 sample is slightly higher. The nanocrystal size effect on the shifts of valence and conduction band edge and the band-gap expansion of semiconductor nanocrystals has been modeled by Lang et al. [103]. This model, without any adjustable parameter, predicted that the band-gap expansion (along with shifts of valence and conduction band edges) is present with dropping size of nanocrystals because of the increase of surface/volume ratio. Crystallite size of the sintered pellets has been estimated as 295 nm for the hydro-Bi 2 Te 3 sample, and 231 nm for the polyol-Bi 2 Te 3 sample. Therefore, the estimated E g values agree with the trend expected from the relative size of the crystallites in the studied materials.
The power factor (PF) (S 2 σ) calculated for both hydrothermal and polyol samples, as presented in Figure 7c. The highest PF at room temperature was obtained for the hydro-Bi 2 Te 3 sample as 24 µW cm −1 K −2 , followed by the polyol-Bi 2 Te 3 sample as 16 µW cm −1 K −2 . The maximum power factors we achieved are 24 µW cm −1 K −2 for the hydrothermal sample at 300 K and 21 µW cm −1 K −2 at 423 K for the polyol sample. The magnitude of PF is comparable to or much higher than other solution-derived samples-with maximum power factor 1-9 µW cm −1 K −2 [66,96].

Thermal Conductivity
The total thermal conductivity values were calculated using the thermal diffusivity, heat capacity and mass density values of the samples. The total thermal conductivity of hydro-Bi 2 Te 3 is slightly lower than polyol-Bi 2 Te 3 at T = 293 K, as seen in Figure 8a. They both decrease first and then increase with increasing temperature, where it shows a dra-matic increase for the hydro-Bi 2 Te 3 sample after 373 K. Nanostructures at different length scales (diameter, roughness and point defects) showed efficient scattering throughout the phonon spectrum, reaching very low thermal conductivity values [23]. Looking at the microstructure of the pellets obtained, there could be a mixture of these nanostructures at different length scales, influencing the phonon propagation. The actual nanostructures include undefined structural features, such as interface defects, impurities, strain, etc., making it difficult to classify the contribution of each scattering process to phonon transport [104]. In order to make a clear assessment, we have examined lattice (κ lat ) and electronic thermal conductivity (κ el ) of both the samples by using the Wiedemann-Franz relationship, κ el = L o σT where L o , σ, and T are Lorenz number, the electrical conductivity and the absolute temperature, respectively. The κ lat values were estimated by subtracting the κ el from the total thermal conductivity. In order to calculate the Lorenz number (L o ), experimentally obtained Seebeck coefficient values were used and a single parabolic band model was assumed to solve the Fermi integral. Lorenz numbers for both the samples are graphically presented in Figure S4. The κ lat and κ e are shown in Figure 8b. Since the κ el is determined by the σ, it is expected to possess a similar trend to the electrical conductivity. In both the samples the κ el contribution is higher than the lattice contribution, κ lat , most likely due to the electrons' more dominant role on the thermal conductivity. The striking feature is the very low lattice contribution in these samples, strongly influenced by the increased density of grain boundaries that plays a crucial role in increased phonon scattering and thus a strong reduction in κ lat . The room temperature κ latt for the hydro-Bi 2 Te 3 and polyol-Bi 2 Te 3 were about 0.25 W/m·K and 0.37 W/m·K, respectively. The κ lat of the hydro-Bi 2 Te 3 sample increases with temperature after 373 K. This might be due to the highly crystalline texture (long layered structure) thus longer MFP of phonons. Crystallite size of the sintered pellets has been estimated as 295 nm for the hydro-Bi 2 Te 3 sample, and 231 nm for the polyol-Bi 2 Te 3 sample. The difference in the E g is the evidence of the correlation of the E g with the average crystallite size; a smaller crystallite size leading to a larger band-gap and vice versa. Due to the smaller E g of the hydro-Bi 2 Te 3 sample (0.11 eV), at elevated temperatures the minority carriers' diffusion (ambipolar effect characteristic for the electron-hole pair diffusion on the onset of the intrinsic conduction) contributes strongly to the total thermal conductivity κ tot [105,106]. Regardless of the synthetic route, the total thermal conductivity of the materials is lower than that of the current best-known n-type Bi2Te2.925Se0.075 (κtot = 1.65 W/m·K at 300 K in which the κlat was 1.27 W/m·K) [16], and are also comparable to those of ball milled and dc hot pressed n-type Bi2Te2.7Se0.3 bulk samples (κtot = 1.06 W/m·K at 300 K in which κlat was 0.7 W/m·K) [107]. This strongly suggests that the nanostructure obtained through the developed synthetic routes plays a very significant role in reducing the κlat through effective phonon scattering.
Temperature dependence of the estimated ZT values for the samples are shown in Figure 9. Room temperature ZT values of 0.8 and 0.52 were obtained for hydro-Bi2Te3 and polyol-Bi2Te3 samples, respectively. The maximum ZT values were obtained at slightly Regardless of the synthetic route, the total thermal conductivity of the materials is lower than that of the current best-known n-type Bi 2 Te 2.925 Se 0.075 (κ tot = 1.65 W/m·K at 300 K in which the κ lat was 1.27 W/m·K) [16], and are also comparable to those of ball milled and dc hot pressed n-type Bi 2 Te 2.7 Se 0.3 bulk samples (κ tot = 1.06 W/m·K at 300 K in which κ lat was 0.7 W/m·K) [107]. This strongly suggests that the nanostructure obtained through the developed synthetic routes plays a very significant role in reducing the κ lat through effective phonon scattering.
Temperature dependence of the estimated ZT values for the samples are shown in Figure 9. Room temperature ZT values of 0.8 and 0.52 were obtained for hydro-Bi 2 Te 3 and polyol-Bi 2 Te 3 samples, respectively. The maximum ZT values were obtained at slightly different temperatures, in the range 323-373 K for hydro-Bi 2 Te 3 , and 423-473 K for polyol-Bi 2 Te 3 . This is mainly because the samples display a maximum power factor and a minimum thermal conductivity at different temperatures. The maximum ZT value reaches about 0.9 at 373 K for hydro-Bi 2 Te 3 while for polyol-Bi 2 Te 3 it is about 1.03 at 473 K. These are among the highest ZT values reported for n-type Bi 2 Te 3 in the literature. This study shows two viable synthetic strategies, which yield materials with promising efficiency at slightly shifted temperatures that can be used for guiding the synthetic route for the intended temperature range of applications.
thermal conductivity (b) for SPS compacted pellets for hydrothermal and polyol synthesized Bi2Te3.
Regardless of the synthetic route, the total thermal conductivity of the materials is lower than that of the current best-known n-type Bi2Te2.925Se0.075 (κtot = 1.65 W/m·K at 300 K in which the κlat was 1.27 W/m·K) [16], and are also comparable to those of ball milled and dc hot pressed n-type Bi2Te2.7Se0.3 bulk samples (κtot = 1.06 W/m·K at 300 K in which κlat was 0.7 W/m·K) [107]. This strongly suggests that the nanostructure obtained through the developed synthetic routes plays a very significant role in reducing the κlat through effective phonon scattering.
Temperature dependence of the estimated ZT values for the samples are shown in Figure 9. Room temperature ZT values of 0.8 and 0.52 were obtained for hydro-Bi2Te3 and polyol-Bi2Te3 samples, respectively. The maximum ZT values were obtained at slightly different temperatures, in the range 323-373 K for hydro-Bi2Te3, and 423-473 K for polyol-Bi2Te3. This is mainly because the samples display a maximum power factor and a minimum thermal conductivity at different temperatures. The maximum ZT value reaches about 0.9 at 373 K for hydro-Bi2Te3 while for polyol-Bi2Te3 it is about 1.03 at 473 K. These are among the highest ZT values reported for n-type Bi2Te3 in the literature. This study shows two viable synthetic strategies, which yield materials with promising efficiency at slightly shifted temperatures that can be used for guiding the synthetic route for the intended temperature range of applications.

Conclusions
Rapid and scalable synthetic strategies for high-quality and reproducible thermoelectric (TE) materials are demonstrated in this work, exemplified by nanostructured n-

Conclusions
Rapid and scalable synthetic strategies for high-quality and reproducible thermoelectric (TE) materials are demonstrated in this work, exemplified by nanostructured n-type bismuth telluride with a promising TE performance. Use of microwaves enables an effective volume heating, leading to highly crystalline TE nanopowders in a few minutes. Polyol synthesis is shown to be a reliable and scalable route to materials synthesis. Hereby we show that not only EG, but also water is an excellent reaction media, which is certainly much more benign and abundant than any other solvent. Furthermore, hydrothermal synthesis is one-pot process, where all the precursors are mixed in the same reaction medium prior to heating, which makes this process facile and therefore more attractive. Despite the significant difference in their dielectric constant, both solvents (EG and water) showed impressive outcomes for the materials investigated. Crystal structure, crystallinity, morphology, microstructure and surface chemistry of the synthesized materials were evaluated using a library of analytical techniques, including XRD, SEM/TEM, XPS and zeta potential techniques. Particles with dominantly hexagonal platelet morphology were obtained, where the polyol-Bi 2 Te 3 generally exhibited a larger particle size than the hydro-Bi 2 Te 3 . Surface chemistry studied by XPS clearly revealed different levels of surface oxidation, based on the solvent used, and traces of the precursors used on the surface of as-synthesized materials. Surface directing agents are stably bound to the surface of the as-formed nanoparticles, thus assuring the maintenance of the hexagonal platelet morphology. The samples were then consolidated using SPS, prior to further transport measurements. The difference in the particle size of the specimens significantly influenced the densification behavior under SPS, where polyol-Bi 2 Te 3 with larger particle size reached about 10% lower densification compared to the hydro-Bi 2 Te 3 sample under identical conditions. Upon sintering, texturing is observed to be more predominant in the hydro-Bi 2 Te 3 sample, as assessed from XRD and SEM analyses. Thermoelectric transport properties were evaluated using electrical conductivity and Seebeck coefficient measurements up to 525 K, and the thermal diffusivity was assessed by using the LFA system. The final estimated TE figure-of-merit values about 0.8-0.9 was reached in the temperature range of 300-375 K for the hydro-Bi 2 Te 3 sample, while 0.9-1 was reached in the temperature range of 425-525 K for the polyol-Bi 2 Te 3 sample. Energy and time efficiency, scalability as well as the environmental impact of the synthetic processes presented here, along with the high ZT values achieved, pave the way for wider impact of these synthetic strategies for developing promising strategic materials in the field of thermal energy harvesting.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/nano11082053/s1, Figure S1: Schematic of spark plasma sintering (SPS) process, sample geometry for thermal diffusivity, using Laser Flash technique, and power factor measurements. Figure S2: Rietveld refinement for Hydro-Bi 2 Te 3 sample after SPS sintering, Figure S3: X-ray powder diffraction (XRPD) patterns showing a close up of the peaks with Miller indices of (006) and (015), Figure S4: Lorenz Number, Lo, estimated for Bi 2 Te 3 samples using parabolic band model, Table S1: Survey of the synthesis methods, reaction conditions, morphology and size of Bi 2 Te 3 .