Polymer-mixed Sb2Te3/Te nanocomposites exhibiting p-type to n-type conduction reversal and thermal conductivity reduction

Sb2Te3-based materials are potential room-temperature thermoelectric materials. In the present work, we choose polycrystalline Sb2Te3/Te nanocomposites and utilize Poly Methyl Methacrylate (PMMA) to reduce the thermal conductivity of Sb2Te3 samples. PMMA and polycrystalline Sb2Te3/Te were well mixed using ball milling. Pellets have been made by the cold press method. Thermoelectric transport properties of Sb2Te3/Te nanocomposites: composition, microstructure, and analysis are found to be influenced by PMMA. With increasing PMMA concentration a p-type to n-type transition has been observed because there are fewer charge carriers or the composites have a higher resistance. It is also observed that the thermal conductivity of Sb2Te3/Te nanocomposites decreases as the PMMA increases. This research paves the way for making the best thermoelectric materials by reducing thermal conductivity through the use of polymers.


Introduction
One of the things we frequently encounter and are aware of is energy, which can take several forms. Various forms of energy, including light, sound, electricity, heat, wind energy, etc can be converted into one another directly or indirectly. Electrical energy can be converted into thermal energy with the help of thermoelectrics. Thermoelectric (TE) energy conversion technology has recently attracted a lot of attention because of its advantages over other energy conversion methods. TE materials have a very low impact on the environment and lack moving components [1]. The dimensionless figure of merit, zT = S 2 σTκ −1 , which takes the Seebeck coefficient, electrical conductivity, absolute temperature, and thermal conductivity, respectively into account, can be used to describe the performance of TE materials [2]. Polymer-based thermoelectric materials are gaining much attention because of the ability of the polymer in altering the microstructure which can help in tuning the thermoelectric properties [3]. The TE properties of materials, Seebeck coefficient (S), electrical conductivity (σ), and thermal conductivity (κ) are dependent on the quenching medium, i.e., the cooling rate of melt solidification [4][5][6], annealing treatment [7,8], the synthesis process [9,10], and sintering process [11,12]. Knowledge of the phase diagram is important in choosing the appropriate composition and process temperature. The properties of the TE materials can be tuned by altering the microstructure. By adding a polymer with TE material, the microstructure can be modified which in turn can tune the physical properties of the material [3].
Also, polymer-based TE materials and their composites can be utilized to produce flexible TE material that can be utilized as a wearable thermoelectric generator for energy harvesting and for many other related applications like temperature-controlled body suits [13,14].
For room-temperature applications, Sb 2 Te 3 is one of the best p-type thermoelectric materials [12,15]. To improve the TE efficiency of Sb 2 Te 3 several efforts have been made like the formation of nanocomposite Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. [16][17][18][19][20][21][22][23], doping [15], orientation engineering [24], and nanostructuring [25,26]. The formation of nanocomposite has got much attention due to its advantage in reducing thermal conductivity while maintaining a high value of power factor (PF = S 2 σ) [27,28].
In recent years, polymer-based composites such as poly (3,4-ethylene dioxythiophene) (PEDOT) [3,4], polyaniline (PANI) [5,6], and polythiophene [7] have been applied as TE materials due to their versatile processability, low density, and low thermal conductivity, all of which are critical in improving TE properties. Recent research has concentrated on the synthesis of hybrid Sb 2 Te 3 composites by the blending of inorganic and organic ingredients to increase their application [29].
In recent years, inorganic-organic composites have attracted much attention as thermoelectric materials due to their versatile advantages such as low thermal conductivity, low density, and easy processability. By using inorganic-organic composites, bilateral benefits of good electrical conductivity with enhanced Seebeck coefficient by decoupling σ and S through an energy filtering effect and with low thermal conductivity can be achieved. Several groups have demonstrated the improved thermoelectric properties of using inorganic material and polymer composites [29][30][31][32][33][34][35][36][37].
Numerous thermoplastic polymers such as polyethylene (PE), polystyrene (PS), polyvinyl chloride (PVC), polyvinyl acetate (PVA), and polymethyl methacrylate (PMMA) were used as insulating polymer matrix to fabricate composite materials [29,[34][35][36][37]. Among them, poly (methyl methacrylate) (PMMA) has been one of the most widely studied polymer materials due to its excellent mechanical, chemical, and thermal stability properties. Conductive pathways can be generated with various inorganic fillers in the interfacial regions of the polymeric matrix during processing [35,36]. Because the arrangement of inorganic fillers with attractive electrostatic force depends heavily on the matrix shape, the composite could demonstrate high performance with a lined-up conductive pathway.
It is reported that composite formation with polymer (polyvinylpyrrolidone (PVP)) has reduced the κ of Bi 2 Te 3-based nanofiber [30]. However, there are not much work on the polymer mixed Sb 2 Te 3 nanocomposites.
Among several inorganic materials, the Bi 2 Te 3 /Sb 2 Te 3 -based thermoelectric materials were successfully utilized in room-temperature thermoelectric device applications. numerous studies were investigated for n-type Bi 2 Te 3 polymer composites and on the other hand few studies only reported for the p-type Sb 2 Te 3 . In both the n-type and p-type chalcogenides, reducing thermal conductivity and improving the Seebeck coefficient without affecting their electrical conductivity is the main motive to achieve high-performance thermoelectric materials. In the present study, we intend to study the p-type Sb 2 Te 3 and PMMA polymer composites due to the benefits of easy processability, large-scale processing of bulk compounds with excellent mechanical, chemical, and thermal stability properties of PMMA towards the improvement of Seebeck coefficient and reduction of thermal conductivity.
Among several inorganic materials, the Bi 2 Te 3 /Sb 2 Te 3 -based thermoelectric materials were successfully utilized for in room-temperature thermoelectric device applications. numerous studies were investigated for n-type Bi 2 Te 3 polymer composites and on the other hand few studies only reported for the p-type Sb 2 Te 3 . In both the n-type and p-type chalcogenides, reducing thermal conductivity and improving the Seebeck coefficient without affecting their electrical conductivity is the main motive to achieve high-performance thermoelectric materials.
None of the investigations have been published to investigate the thermoelectric characteristics of PMMAmixed Sb 2 Te 3 nanocomposites. The inorganic material in this study is Sb 2 Te 3 /Te and the polymer used is PMMA which is a non-conducting polymer. In the conducting pathways of Sb 2 Te 3 /Te, the polymer will act as an energy barrier that will restrict the transport of electrons as well as scattering centers to restrict the heat flow by scattering the phonons. It is well reflected in the temperature-dependent electrical conductivity and thermal conductivity of Sb 2 Te 3 and PMMA polymer composites. The thermal conductivity of Sb 2 Te 3 nanocomposites is found to be considerable. The variation in TE properties was understood based on the structural changes due to the addition of polymer.

Experiments and methods
Sb 2 Te 3 /5%Te nanocomposites were synthesized by melting the raw materials, Sb and Te (99.99% purity, Sigma Aldrich) by melt quenching method. Weighing was done on the Sb and Te elemental powders according to the Sb 2 Te 3 /5%Te composition. The weighed powders were transferred in a quartz ampoule and sealed under a vacuum of 10 −6 mbar. The ampoule was heated to 1073 K at a rate of 5 K min −1 in a rocking resistive furnace. The combination was subjected to a 10-hour rocking motion at 20 rpm and then the ampoule was quenched in an ice water medium. The ingots were ground to powder and then mixed with PMMA powder according to PMMA x Sb 2 Te 3 /5%Te (x = 0.00, 0.01, 0.03, 0.05) ratio. These mixed powders were ball milled for 6 h at 250 rpm in a planetary ball milling system and then cold pressed to produce pellets. The obtained pellets were sintered at 423 K for 6 h under a vacuum of 10 −3 mbar. This sintering temperature was kept low to avoid any decomposition or melting of the polymer. The density of the sintered pellets was measured using Archimedes' principle. For various structural and thermoelectric studies, the final samples were cut in the desired directions. Under identical conditions, all the samples were made twice and the structural and TE characterizations were independently carried out on these two sets of samples and the average of the two being used.
X-ray diffraction (XRD) was used to determine the crystal structure and purity of the samples using the Rigaku Smart Lab x-ray diffractometer with Cu K α radiation source (λ = 1.5418 Å). The sample's Raman spectra were recorded using a HORIBA JOBIN YVON HR 800 confocal Raman spectrometer (λ exc = 532 nm) and a 12 mW laser. For surface imaging, FEI Quanta 450 scanning electron microscopy (SEM) was used. Backscattered electron (BSE) images were taken with an FEI ESEM-Quanta 200. The powder samples were examined using a JEOL 2000 FX-II transmission electron microscope (TEM). Energy-dispersive x-ray spectroscopy (EDX) was used to determine the elemental compositions. The four probe and bridge methods, respectively, were used to measure the TE transport characteristics, σ and S in the temperature range of 300-390 K using a homemade setup under a vacuum of 10 −3 mbar. The uncertainty in electrical resistivity (σ) and Seebeck coefficient measurements (S) is ±5%. Hall effect measurement has been done for all the samples at room temperature using a homemade laboratory setup. The heat capacity was calculated using the formula κ = αC p d, where α is the thermal diffusivity, d is the density, and Cp is the heat capacity. In the temperature range 300-390 K, the thermal transport properties, thermal diffusivity (α) under argon atmosphere, and heat capacity (C p ) under nitrogen (N 2 ) atmosphere, were measured using LINSEIS LFA 1000 and differential scanning calorimetry apparatus TA SDT650, respectively. The Uncertainty in thermal conductivity (κ) measurements is ±5%.

Results and discussion
3.1. Powder XRD analysis Figure 1 shows the XRD pattern of polymer mixed Sb 2 Te 3 /Te nanocomposites. The indexing of all the XRD peaks has been done with p-type Sb 2 Te 3 (space group: R3m h, rhombohedral, PCD Card No. 1216385) and Te phase (space group: P3 1 21, trigonal, PCD Card No: 382599) [16]. Rietveld refinement analysis was carried out using the Full Proof software. The graphs of the XRD Rietveld refinement are displayed in figures 2(a)-(d). The simulated experimental x-ray diffraction patterns are shown by the dark blue line and observed experimental x-ray diffraction patterns are shown by the orange circles, respectively. The difference between the experimental and simulated curves can be seen by the pink continuous line at the bottom. The Bragg positions are shown by the thin, navy blue vertical lines. Table 1 lists the crystallite size, Lotgering factor, density as well as refined lattice parameters (a, b, c, and V). These values are found to be closely matching with the Sb 2 Te 3 lattice parameters that have been reported [16]. As vacancies tend to shorten the lattice parameters, the pure Sb 2 Te 3 lattice monotonically decreases with increasing polymer content. Two major phases are present in prepared samples as revealed by the Rietveld refinement. Sb 2 Te 3 phase has been found as the main phase and Te has been found as a secondary phase. A trace amount of the secondary phase of Te was observed in all the samples, due to the high amount of nonreactive Te present in the prepared samples.
The Rietveld refinement results show that the increase of polymer content in the Sb 2 Te 3 /Te composite tends to decrease the Sb 2 Te 3 phase and increases the Te phase. This is clearly evident in the observation of the presence of the Te phase (%) which increased from 8% to 18% and a simultaneous decrease in the Sb 2 Te 3 phase from 91% to 81% with the increase of polymer content from (0 to 0.05%). These observations show that the addition of polymer may segregate the Te phase from the Sb 2 Te 3 phase. In addition to this, the crystallite size, crystallite volume, and density of the polymer composite tend to decrease with the increase of PMMA content. There is a difference in the intensity of the (015) peak and (0015) peak is also observed, suggesting that the polymer mixed samples show a significant anisotropy of grain orientation without changing the crystal structure. The Lotgering  factor (F) is determined to confirm the anisotropic orientation from the XRD patterns of polymer mixed samples. From XRD peak intensity data, the degree of crystal grain orientation i.e. the Lotgering factor (F) is calculated in the case of layered structure materials since they are arranged mostly along the c-axis. The F is as follows ; Po can be calculated from the peak data of the JCPDS card. The F value is zero (F = 0) for a crystallographic isotropic (randomly oriented) sample i.e. the value of P is Po, and the F value is one (F = 1) for a completely oriented sample. F is increasing by increasing the polymer quantity which indicates the orientation of planes in a particular direction. Figure 1 shows that after adding the polymer the prepared samples have been oriented on the c-axis. From XRD data, the crystalline sizes (table 1)  q where D is the crystallite size, K is the shape factor (usually, the value of K is 0.94), λ is the x-ray wavelength (0.15418 nm), and β is the full-width half-maximum (FWHM) of the diffraction peak at θ, which is the diffraction angle. Crystalline size shows a monotonic decrease with increasing polymer content because of the shrinking of the vacancies. Anti-site defects are highly prevalent in the Sb 2 Te 3 system, and Sb atoms may occupy Te sites (Sb Te ) [38]. This may also be a reason behind the decreasing trend of the volume of prepared samples. Therefore, as described in the section on thermoelectric transport properties, the addition of polymer to the Sb 2 Te 3 system may significantly affect the TE properties of Sb 2 Te 3 /Te nanocomposites. The EDX spectra displayed in figure 3 show the presence of too much Te along with the Sb 2 Te 3 matrix. It is obvious that the Te distribution is not constant throughout the polymer mixed samples. Te-Te defects are produced by the secondary Te phase in composite materials, and these defects significantly affect the TE characteristics [39][40][41][42]. High-resolution TEM images have been taken to check the phase formation of the Sb 2 Te 3 matrix ( figure 4(a)). The selected area electron diffraction (SAED) pattern also confirms the phase formation of Sb 2 Te 3 . The ring pattern indicates the polycrystalline nature of the materials. Distance between the (015) planes has been calculated by ImageJ software and it has matched with the reported data of the Sb 2 Te 3 matrix. Figure 4(c) shows the Backscattered-Electron (BSE) image of the Sb 2 Te 3 /Te nanocomposite. Te has shown in dark black colour in figures 4(d)-(f). The addition of polymer increases the nonreactive Te in Sb 2 Te 3 which is clearly visible in figures 4(d)-(f). These images reveal that the mixing of polymer (PMMA) has a major impact on the microstructure. Te typically has nanograin sizes in the 30 to 50 nm range. The phonon means free path of Sb 2 Te 3 is in the range of nanoscale, which makes the Te nanograins effective in phonon scattering [43]. However, a clear difference has been observed between the pristine and polymer mixed samples. The following discussion discusses how microstructural modifications affect TE characteristics.

Raman and XPS analysis
To study more about the nanocomposites' molecular mode of vibration, Raman spectroscopy was used. Figures 5(a)-(d) displays the Raman spectra of the polymer mixed Sb 2 Te 3 /Te composites for x = 0, 0.01, 0.03, and 0.05 that were obtained using a 532 nm laser at room temperature. The silicon standard, which has a Raman peak of 520.5 cm −1 , was used to calibrate the Raman spectra. With the use of a Gaussian line profile, Raman peaks were fitted.
Raman scattering is used to identify the three Raman-active phonons (A , g 1 1 E , g 2 and A g 1 2 ) present in Sb 2 Te 3 -related compounds. These compounds have a rhombohedral structure made up of hexagonal closepacked atomic layers stacked along the c-axis [26]. The Raman spectra of Sb 2 Te 3 /Te nanocomposites (figures 5(a)-(d)) clearly display the Raman active vibrational modes at ∼65 cm −1 , ∼87 cm −1 , ∼116 cm −1 , ∼137 cm −1 , and ∼164 cm −1 belonging to A , were formed from Sb2Te3, whereas the other two modes E' and A 1 were produced by Te as shown in figure 5(e). This is again an agreement with the XRD results.
Out-of-plane and in-plane vibrations, respectively, are represented by the A 1g and E g modes. Out-of-plane (A 1g ) vibrations are found to be influenced by the addition of polymer to the Sb 2 Te 3 /Te composite. Raman peaks of A 1g plane vibrations are increasing with increasing polymer content indicating that the samples are oriented along the c-axis. It suggests that the mixing polymer with Sb 2 Te 3 /Te composite has a major impact on the thermal transfer, i.e., phonon transport.
The XPS of Sb 3d and Te 3d are displayed in figures 5(f) and (g), respectively. Using the Gaussian-Lorentzian (GL) function, the XPSPEAK41 program deconvoluted the generated XPS peaks. Sb 3d has three peaks at ∼530 eV, ∼532 eV, and ∼539 eV, which correspond to the ground state (3d5/2), O 1 s state, and excited state (3d3/2), respectively, while Te 3d exhibits two peaks at ∼576 eV and ∼586 eV, which correspond to the ground state (3d5/2), and the excited state (3d3/2), respectively. The oxidation states of Sb is 3+ and that of Te is 2-are confirmed by the energy gap between the excited state and ground state (E = 3d3/2-3d5/2 = ∼9.3 eV for Sb and ∼10.4 eV for Te, respectively) [44]. This XPS results again confirm the phase formation of the Sb 2 Te 3 /Te composite.

Thermoelectric transport properties
The electrical transport properties of polymer mixed Sb 2 Te 3 /Te nanocomposites measured in the temperature range 300-400 K were presented in figure 6. Figure 6(a) shows the temperature-dependent electrical conductivity of polymer-mixed Sb 2 Te 3 /Te nanocomposites. All the samples show metallic behavior with a decrease in electrical conductivity with increasing temperature. This might be caused by the interaction of the scattering mechanisms known as grain boundary/interfacial (metal-semiconductor) potential barrier scattering and acoustic phonon scattering (APS). [ [39][40][41][42]. The pristine Sb 2 Te 3 sample shows a higher electrical conductivity value of 70 × 10 3 S m −1 compared to other polymer mixed samples. The electrical conductivity value displays a decreasing trend with increasing the polymer content from 1% to 5%. From figure 6(a), the polymer mixed samples show a lower electrical conductivity due to hindrance in the transport of electrons between the metal and polymer interface. At a specific temperature, the observed fluctuation in the electrical conductivity value is directly related to the variation in carrier concentration (n c ) and carrier mobility (μ) and is represented as σ = n c eμ [41,42]. The Hall measurements were performed for all the samples at room temperature and the measured Hall mobility and carrier concentration values were presented in table 1. The carrier mobility shows a decreasing trend from the value of 67.21 cm 2 /V.s to 5.43, 0.18, and 0.065 cm 2 /V.s for 1% polymer, 3% polymer, and 5% polymer nanocomposites respectively. Pristine Sb 2 Te 3 /Te and polymer mixed nanocomposite show carrier concentration values in the range of 10 20 cm −3 . The decrease in both carrier concentration and carrier mobility values of polymer mixed nanocomposites resulted in low electrical conductivity values. Figure 6(b) shows the temperature-dependent Seebeck coefficient values for the pristine and polymer-mixed nanocomposites. The pristine sample shows p-type conductivity whereas the polymer mixed Sb 2 Te 3 /Te nanocomposites show n-type conductivity. As mentioned in table 1, the 'Te' phase is increased from 8% to 18% and the atomic percentage of 'Te' increased from 60.4 at% to 63.6 at%. This increase in 'Te' segregation can be directly correlated with the carrier concentration (table I) and Seebeck coefficient measurements ( figure 6(b)). It is clearly evident, that the carrier type has been changed from 'p-type' to 'n-type' in Seebeck coefficient measurement as well as the electron concentration increases from 0.89 × 10 20 to 25.19 × 10 20 with increasing PMMA content. In the case of increased 'Te' content in the Sb 2 Te 3 /Te polymer composites, the reduction in hole density of the composite might be due to the following reasons (i) The density of states (DOS) of 'Te' is smaller than the DOS of Sb 2 Te 3 and hence in the condition of increasing 'Te' contribution, the overall DOS and the releasing of holes to the polymer composites can be reduced. (ii) In the case of Sb vacancy, Sb atoms from Te sites will return back to Sb sites, and Te vacancy will become dominant and contributes excess electrons. Hence, we can conclude that with increasing PMMA, the 'Te' gets more segregation and the associated defects such as antisite defects, and vacancies may enhance the concentration of electrons and reverse the conductivity from p to n-type [26,45]. All the samples follow a similar trend, i.e., the magnitude of the S values increases with temperature, indicating a metallic behavior. This temperature-dependent trend is consistent with the reduction of σ with an increase in temperature. The linear temperature dependency of S may mean that the dominant mechanism of carrier transportation under a temperature gradient is diffusive in nature [42,46]. The enhancement in the Seebeck coefficient with temperature in polymer mixed Sb2Te3 composites may be attributed to the filtering of low energy charge carriers by the addition of PMMA embedded inside the Sb 2 Te 3 grain boundaries. The transition from p-type to n-type in the polymer matrix with the increase of PMMA might   be the cause of an increase in segregated Te phases which is clearly evident in XRD analysis. The observed results including carrier concentration, carrier mobility, and Seebeck coefficient supports the transition from p-type to n-type. Figure 6(e) shows how thermal conductivity varies with temperature (κ c ). In the case of narrow bandgap materials, i.e., Sb 2 Te 3 and Bi 2 Te 3 , the total thermal conductivity is the sum of carrier (κ c ), phonon (κ p ), and bipolar thermal conductivity (κ b ), i.e., it can be expressed as κ T = κ c + κ p + κ b [39][40][41]44]. At low temperature, below 423 K, κ b is negligible. The κ c value is estimated from Wiedemann-Franz relation κ c = LσT, L is the Lorenz number in this case. [39][40][41]44]. The single parabola band (SPB)-APS approximation is used to determine the L values from the S values. [40][41][42][45][46][47]. The κ c shows a similar trend of electrical conductivity, the decrease in carrier thermal conductivity with increasing temperature. Another reason for decreasing the electrical part of thermal conductivity is the density of the material. The decrease in density of Sb 2 Te 3 upon the addition of polymer affects the carrier transport. From figure 6, it can be seen that κ c contributes around ∼15%, whereas κ p dominates the κ value, contributing around ∼85% of total κ. In general, the κ drops with rising temperature as a result of phonon-phonon scattering, phonon-carrier scattering, and phonon interface scattering at Te nanograins [41,42]. The temperature-dependent phonon thermal conductivity (κ p ) and total thermal conductivity (κ T ) plots are depicted in figures 6(d) and (e) respectively. Both the plots show a decreasing trend with increasing temperature. The total thermal conductivity of the pristine Sb 2 Te 3 /Te sample shows a value of 2.75 W m −1 K −1 . However, with the increase of polymer content in Sb 2 Te 3 /Te, the thermal conductivity values are found to decrease up to 0.4 W/m/K for 5% polymer.
It should be mentioned that though the polymer addition brings interesting changes like reduction in thermal conductivity and conduction reversal from p to n-type, the figure of merit is found to decrease as shown in figure 6(f). In the present study, the higher zT value of 8 × 10 −2 was obtained for the pristine sample, and it almost decreased nearly to zero for the polymer added samples. In summary, the addition of polymer with Sb 2 Te 3 /Te nanocomposites resulted in a decrease in density, electrical conductivity, conduction reversal from p to n-type, and a reduction in thermal conductivity values.

Conclusions
The thermoelectric properties of poly methyl methacrylate (PMMA) polymer mixed Sb 2 Te 3 /Te nanocomposites made by melt-quenching followed by ball milling processes have been studied. The addition of polymer has a considerable impact on Sb 2 Te 3 /Te nanocomposites thermoelectric characteristics studied in the temperature range 300 to 390 K. A 30% reduction in thermal conductivity with the 5% PMMA composites has been observed. On the other hand, the polymer addition has significantly decreased the electrical conductivity. The addition of polymer also changes the conduction type from p to n-type. The positive impact of polymer addition in reducing thermal conductivity may help in designing high-performance thermoelectric materials based on Sb 2 Te 3 composites.