betaFeedback

WorksPhysics and Astronomy

Lattice dynamics study of (Bi1-xSbx)2Te3 compounds: Comparative compositional study of force constants and Grüneisen parameters

Abstract

We performed a comprehensive Raman spectroscopy study of (Bi1-xSbx)2Te3 (BST-x) compounds, with varying Sb concentration, under ambient and high-pressure conditions. At ambient conditions, the frequency of all observed Raman-active modes (E1g, A11g, E2g, A21g) is blue-shifting with increasing Sb concentration, albeit with varying concentration-dependent rate. Based on these results, we calculated the relative force constants as a function of Sb concentration using a harmonic approximation, which shows a mode-dependent variation. The force constants for the lower frequency Raman modes (E1 g and A11g) remain almost composition-independent, while for the higher-frequency ones a profound softening (E2g) or stiffening (A21g) was determined. From the Raman spectroscopy measurements under high-pressure conditions, the corresponding Grüneisen parameters for the three observed Raman-active modes (A11g, E2g, A21g) as a function of composition were determined by combining them with the previously reported Equation of States and the corresponding bulk modulus B0 of the ambient-conditions R-3m phase. The average Grüneisen parameter exhibits a clear maximum for BST-07. Based on Slack's equation, these results are directly related to the optimal ZT values for this composition. Our study provides a solid framework for connecting composition, lattice dynamics, and thermal-transport behaviour, towards establishing the microscopic origin of thermoelectric optimization in BST compounds.

Introduction

Thermoelectric (TE) materials, stand out as one of the most promising energy-conversion materials, aiming to meet pressing global energy needs. The thermoelectric performance of a TE material is governed by the coupled transport properties of the charge carriers and phonons, as described by the dimensionless figure of merit ZTZT (ZT=σS2T/(κe+κl)ZT=\sigma S^{2}T/(\kappa_{e}+\kappa_{l}), where SS is the Seebeck coefficient, σ\sigma is the electrical conductivity, TT is the absolute temperature and κe\kappa_{e} and κl\kappa_{l} are the electronic and lattice thermal conductivities, respectively). Thus, achieving a high ZTZT value requires not only favorable electronic transport properties but also effective suppression of the lattice-mediated thermal conductivity [1, 2]. Various approaches have been pursued to improve the performance (i.e., the ZTZT value) of TE materials, such as doping and alloying, novel synthesis methods etc. [3] Alloying provides a particularly powerful route to simultaneously and advantageously tune both the electronic transport properties and the lattice thermal conductivity. In addition to modifying the carrier concentration and the electronic structure, alloying alters the effective atomic masses, the interatomic force constants and lattice anharmonicity, all of which can influence phonon transport properties [4]. Therefore, understanding how alloying reshapes the lattice vibrational properties is essential for establishing the microscopic origin of thermoelectric optimization.

Among near-room-temperature TE materials, Bi2Te3\mathrm{Bi}_{2}\mathrm{Te}_{3} is considered an archetypal material and it is the most widely used commercially. Alloying Bi2Te3\mathrm{Bi}_{2}\mathrm{Te}_{3}, specifically with Sb, is one possible route to achieve higher ZTZT for the next-generation superior TE materials [5, 6]. Therefore, the Bi-Sb-Te (BST) system provides an ideal platform for examining such composition-dependent lattice effects. Both Bi2Te3\mathrm{Bi}_{2}\mathrm{Te}_{3} and Sb2Te3\mathrm{Sb}_{2}\mathrm{Te}_{3} crystallize in the same rhombohedral R3‾mR\overline{3}m structure under ambient conditions, consisting of Te1-Bi/Sb-Te2-Bi/Sb-Te1 quintuple layers separated by weak van der Waals gaps, and they form an extended substitutional Bi/Sb solid solution [7], that retains the R3‾mR\overline{3}m structure [8]. Optimized Bi/Sb ratios produce some of the most widely used p-type thermoelectric materials near room temperature, with compositions around (Bi0.3Sb0.7)2Te3(\mathrm{Bi}_{0.3}\mathrm{Sb}_{0.7})_{2}\mathrm{Te}_{3} and neighbouring BST alloys repeatedly serving as benchmark materials for high thermoelectric performance. The superiority of this composition range has commonly been discussed in terms of the carrier concentration, band structure, microstructure, and optimization of the electrical transport properties [9, 10, 11]. However, changing the Bi/Sb ratio unavoidably modifies the phononic dispersion relation as well. Although Bi and Sb are isovalent, they substantially differ in atomic mass and local bonding characteristics. Thus, alloying provides a continuous perturbation of both the inertial and the restoring-force terms governing the lattice vibrations. It is therefore natural to ask whether the favorable composition range is also associated with a distinctive lattice-dynamical response.

In this context, the phononic properties of BST compounds cannot be understood as a simple interpolation between the two end compounds based solely on changes in atomic mass. Element-resolved measurements and lattice-dynamical analyses of Bi2Te3\mathrm{Bi}_{2}\mathrm{Te}_{3} and Sb2Te3\mathrm{Sb}_{2}\mathrm{Te}_{3} have demonstrated substantial differences in their vibrational spectra. While the overall softening of the phonon density of states (PDOS) of Bi2Te3\mathrm{Bi}_{2}\mathrm{Te}_{3} relative to Sb2Te3\mathrm{Sb}_{2}\mathrm{Te}_{3} is partially consistent with their mass difference, the accurate reproductions of the optical phonon spectrum requires a significant change in the effective Bi-Te and Sb-Te force constants [12, 13]. This highlights a competition between mass and bonding effects during Bi/Sb substitution. Therefore, Raman spectroscopy appears advantageous for probing these changes, given that Raman shifts contain intertwined contributions from mass substitution, changes in bond stiffness, local strain, and atomic disorder. In the specific case of BST compounds, Raman spectroscopy is particularly sensitive to this competition, because the Raman-active modes (A1g1A_{1g}^{1}, A1g2A_{1g}^{2}, Eg1E_g^{1}, Eg2E_g^{2}) probe characteristic out-of-plane and in-plane atomic motions within the layered structure. Composition-dependent Raman investigations of BST at ambient conditions have indeed revealed systematic phonon-frequency shifts, mode-dependent evolution, and disorder-related broadening throughout the solid solution [14, 15, 16].

Going one step further, the isothermal Grüneisen γT\gamma_T parameter quantifies how a specific phonon mode responds to changes in crystal volume. Whereas Raman frequency at ambient conditions describes the curvature of the vibrational potential in the equilibrium structure, γT\gamma_T quantifies how this curvature evolves when the interatomic distances are modified. It therefore provides an important measure of mode-dependent anharmonic lattice response. High-pressure Raman measurements and first-principles calculations have previously yielded mode-specific and macroscopic Grüneisen parameters for Bi2Te3\mathrm{Bi_2Te_3}, confirming that its lattice dynamics is highly sensitive to compression [17, 18]. Likewise, studies of Sb2Te3\mathrm{Sb_2Te_3}-based thermoelectric systems have demonstrated that doping can alter phonon anharmonicity and its connection to thermal transport [19]. Although the Grüneisen parameters should not be interpreted as a direct measurement of lattice thermal conductivity they provide a key experimentally accessible evaluation of the anharmonic lattice response that participates in phonon-phonon interactions. This correlation is particularly intriguing for high-performance Bi-Sb-Te compositions.

Recent investigations on (Bi0.25Sb0.75)2Te3(\mathrm{Bi}_{0.25}\mathrm{Sb}_{0.75})_2\mathrm{Te_3} have revealed strong coupling between low-lying optical and acoustic phonons and shown that the low lattice thermal conductivity is associated with reduced phonon group velocities, along with enlarged scattering phase space, scattering rates, and Grüneisen parameters [20]. Such observations suggest that the favourable thermoelectric properties of BST may involve not only optimized electronic transport and extrinsic phonon scattering from defects or microstructures but also intrinsic features of its anharmonic phonon spectrum. However, results on an individual highperformance composition cannot establish whether this behavior is generic to the entire BST series or whether the anharmonic response itself exhibits a composition-selective enhancement. Despite extensive studies of composition-dependent Raman frequencies [15, 16], a systematic relationship between the Bi/Sb ratio, Raman-active phonons, force constants, mode-resolved Grüneisen parameters, and the thermoelectric-relevant composition range remains insufficiently established.

In this study, we systematically investigate the composition-dependent lattice dynamics of (Bi1−xSbx)2Te3(\mathrm{Bi}_{1-x}\mathrm{Sb}_{x})_{2}\mathrm{Te}_{3} (BST-x) compounds with varying Sb concentration using Raman spectroscopy under ambient and high-pressure conditions. The relative force constants of the Raman-active modes for the compositions x=0x=0 (Bi2Te3)(\mathrm{Bi}_{2}\mathrm{Te}_{3}), 0.20.2, 0.50.5, 0.70.7, 0.80.8, 0.90.9 and 11 (Sb2Te3)(\mathrm{Sb}_{2}\mathrm{Te}_{3}) are calculated based on a harmonic approximation. The results reveal that while the force constants of the lower-frequency Raman modes (Eg1(E_{g}^{1} and A1g1)A_{1g}^{1}) remain almost composition-independent, in agreement with previous Phonon Density of States (PDOS) studies [13], a diametric softening (Eg2)(E_{g}^{2}) and stiffening (A1g2)(A_{1g}^{2}) were determined for the higher-frequency ones. Moreover, the isothermal Grüneisen parameters for the A1g1A_{1g}^{1}, Eg2E_{g}^{2} and A1g2A_{1g}^{2} Raman modes were determined, by combining the Raman measurements under pressure in this study with the Equation of State (EOS) for the ambient-conditions R3ˉmR\bar{3}m phases reported in previous studies. Unlike a simple monotonic interpolation between the Bi2Te3\mathrm{Bi}_{2}\mathrm{Te}_{3} and Sb2Te3\mathrm{Sb}_{2}\mathrm{Te}_{3} end members, the Grüneisen parameters exhibit a pronounced composition dependence, with the strongest anharmonic response occurring near (Bi0.3Sb0.7)2Te3(\mathrm{Bi}_{0.3}\mathrm{Sb}_{0.7})_{2}\mathrm{Te}_{3}. Notably, this composition lies near the established high-performance p-type BST composition regime. This correspondence does not imply that the Grüneisen parameter alone determines the thermoelectric figure of merit; rather, it identifies an intrinsic lattice-dynamical feature that coincides with a favorable composition range. The results therefore suggest that enhanced phonon anharmonicity may constitute an additional, previously underemphasized phononic factor in optimizing BST thermoelectrics and provide a framework for connecting composition, bonding, lattice dynamics, and thermal-transport behavior.

EXPERIMENTAL METHODS

Raman spectroscopy

Raman studies were performed using a custom-made confocal micro-Raman system with the 660 nm line of a solid-state laser for excitation in the back-scattering geometry. The laser probing spot dimension was ≈4\approx4 μ\mum. Raman spectra were recorded with a spectral resolution of 2 cm−1^{-1} using a single-stage grating spectrograph equipped with a charge-coupled device (CCD) array detector. We kept the laser power on specimens below 0.2 mW to avoid any laser-induced decomposition. Ultra-low-frequency solid-state notch filters allowed us to measure Raman spectra down to 10 cm−1^{-1} [21]. For both the ambient and high-pressure studies, the BST-x powder specimens were from the same batch used in our recent high-pressure x-ray diffraction study [8].

High-pressure studies

Symmetric diamond anvil cells (DACs) with diamond culets of 300–400 μ\mum in diameter were used. Between the two diamonds, a high-pressure chamber was constructed using a pre-indented Rhenium gasket with a thickness of 30–40 μ\mum and a central hole diameter of ≈\approx 100 μ\mum. Ruby powder was also loaded inside the sample chamber and the shift of the ruby line [22] was used to determine the pressure in-situ. The BST-x powders with composition x = 0.2, 0.7 and 0.9 were individually loaded in the center of the chamber. The remaining volume was filled with neon (Ne), precompressed to ≈\approx 2 kbar (using a gas loader), acting as the pressure-transmitting medium (PTM). Neon (Ne) remains fairly hydrostatic up to >20>20 GPa [23], i.e., above the maximum pressure in this study.

RESULTS AND DISCUSSION

Raman spectroscopy of BST under ambient conditions

Figure 1(a) shows the Raman spectra of BST-x compounds (with x = 0 (Bi2_2Te3_3), 0.2, 0.5, 0.7, 0.8, 0.9 and 1 (Sb2_2Te3_3)) at ambient conditions. All BST compounds in this study adopt the rhombohedral 3‾m\overline{3}m structure (tetradymite-type) throughout the solid-solution series.

Raman spectra and frequency as a function of Sb concentration

Figure 1. (a) Raman spectra of BST-x for x=0x=0 (Bi2_2Te3_3), 0.20.2, 0.50.5, 0.70.7, 0.80.8, 0.90.9 and 11 (Sb2_2Te3_3)) at ambient conditions. The asterisks denote few impurity peaks from Te. (b) Frequency as a function of Sb concentration (xx) of the four Raman-active modes. The open symbols correspond to previously reported frequencies for Bi2_2Te3_3 (blue) [17] and Sb2_2Te3_3 (red) [26]. The schematic at the bottom shows the Raman-active modes eigenvectors, adapted from Ref. [25].

According to group theory, for the R3ˉmR\bar{3}m structure, 4 zone-center Raman-active modes are expected with symmetries Γ~R3ˉm=2A1g+2Eg\widetilde{\Gamma}_{R\bar{3}m}=2A_{1g}+2E_g. The EgE_g and the A1gA_{1g} Raman-active modes correspond predominantly to in-plane and out-of-plane atomic displacements, respectively, see Figure 1 bottom [15, 24, 25]. All four characteristic modes are resolved in the recorded spectra and can be assigned to the symmetries (in increasing frequency order) Eg1E_g^1, A1g1A_{1g}^1, Eg2E_g^2 and A1g2A_{1g}^2. The persistence of the same symmetry Raman-active modes across the investigated compositions is consistent with the preservation of the basic tetradymite framework upon Bi/Sb substitution.

The progressive blue shift of all four Raman modes indicates an overall increase of the optical phonons frequency as Sb concentration increases. This behavior can be understood primarily in terms of the reduced effective mass associated with substituting heavier Bi atoms with lighter Sb atoms, consistent with previous Raman and lattice-dynamical investigations of the Bi2_2Te3_3–Sb2_2Te3_3 solid-solution system [15, 27]. However, previous studies strongly indicate that mass substitution alone does not fully account for changes in the PDOS [13]. Specifically, the A1g2A_{1g}^2 mode shifts from approximately 134 in Bi2_2Te3_3 to 165 cm−1^{-1} in Sb2_2Te3_3, while the other three modes show a weaker composition dependence. This different composition-dependent behavior shall be related to the force constant of each mode.

Force constant evaluation

To further understand this mode-dependent behavior, we estimated the force constant for each mode using a harmonic approximation. All of the eigenvectors of the Raman active modes in the tetradymite-type structure involve vibrations of two pairs of Bi/Sb-Te1 atoms, either perpendicular (EgE_g) or parallel (A1gA_{1g}) to the c-axis, see Figure 1. We used a harmonic approximation ωi=kiMi\omega_i=\sqrt{\frac{k_i}{M_i}}, where kik_i and MiM_i denote the force constant and displacement-weighted effective mass associated with mode i, to estimate the force constant kik_i. The effective mass is then calculated according to the following equation: Mi=∑jmjuj,i2∑juj,i2M_i=\frac{\sum_j m_j u_{j,i}^2}{\sum_j u_{j,i}^2}, where the uj,iu_{j,i} defines the atomic displacement vector of atom j in mode i. The cation can be either Bi or Sb and the anion is always Te, thus we used the Bi/Sb ratio (e.g., 0.3*mαBim_{\alpha\mathrm{Bi}}+0.7*mαSbm_{\alpha\mathrm{Sb}} for BST=0.7) to calculate the mass of the cation. From the above, the force constant can be obtained by using the Raman frequency and the effective mass. We used equal displacement vector approximation for the case of the Eg1E_g^1 and Eg2E_g^2 modes [28]. For A1g1A_{1g}^1 and A1g2A_{1g}^2, the atomic displacement vectors for Sb2_2Te3_3 were used to approximate the whole concentration series [29]. The results of the normalized mode force constants, relative to Bi2_2Te3_3, are shown in Figure 2

Normalized force constants of four Raman active modes of BST-$x$ compounds as a function of Sb concentration

Figure 2. Normalized, to Bi2Te3\mathrm{Bi}_{2}\mathrm{Te}_{3}, force constants of the 4 Raman active modes of BST-xx compounds as a function of Sb concentration.

The force constants of the Eg1E^{1}_{g} and the A1g1A^{1}_{1g} modes, remain practically constant, taking into account the approximate nature of the calculations, throughout the compositional range. The force constant of the Eg2E^{2}_{g} mode strongly and monotonically decreases by ∼10%\sim10\% with increasing Sb concentration between x=0x=0 to 1. In contrast, the higher-frequency A1g2A^{2}_{1g} mode shows a distinctly different non-monotonic evolution, with an initial sharp increase, reaching a maximum near x=0.7x=0.7, followed by a slight decrease. Within the harmonic approximation model, this indicates that for the Eg1E^{1}_{g} and the A1g1A^{1}_{1g} modes, the frequency increase directly reflects the reduction in effective mass. On the other hand, for the Eg2E^{2}_{g} mode the reduction of the effective mass outweighs the decrease in the force constant. Surprisingly, the frequency increase of the A1g2A^{2}_{1g} mode is sufficiently pronounced, due to an enhancement of its force constant, in addition to the decrease of the effective mass, over much of the composition range. This is further supported by the superlinear trend of the mode frequency as a function of Sb concentration for the Eg1E^{1}_{g}, Eg2E^{2}_{g} and A1g1A^{1}_{1g} modes, as opposed to the sublinear for the A1g2A^{2}_{1g} mode. Notably, despite these contrasting force constant trends, all four modes undergo phonon blue-shifting with increasing Sb concentration

As discussed in Ref. [13], the PDOS for Sb2Te3\mathrm{Sb}_{2}\mathrm{Te}_{3} below ∼100 cm−1\sim100\ \mathrm{cm}^{-1} could be reproduced by the scaling of the Bi2Te3\mathrm{Bi}_{2}\mathrm{Te}_{3} PDOS, using a direct homology relation for the differences in effective masses. This agrees with our results throughout the compositional range. On the other hand, in the same study a ∼20%\sim20\% softening of the Sb force constant, compared to Bi, was needed to explain the PDOS shift between Bi2Te3\mathrm{Bi}_{2}\mathrm{Te}_{3} and Sb2Te3\mathrm{Sb}_{2}\mathrm{Te}_{3}, for phonon energies above ∼100 cm−1\sim100\ \mathrm{cm}^{-1}, even after accounting for the slight difference in lattice parameters. This well explains the progressive force constant softening behavior of the Eg2E_g^2 mode with increasing Sb concentration. However, the force constant of A1g2A_{1g}^2 shows exactly the opposite trend, namely a significant increase up to x=0.7x = 0.7. These contradicting force-constant trends may also reflect composition-dependent modifications of the local bonding environment and lattice structure that will result in changes of the phonon dispersion relations, especially at the center of the Brillouin zone, beyond the level of the collective PDOS.

Previous neutron diffraction investigations of BST compounds revealed a systematic contraction of the intralayer Bi/Sb-Te bond lengths upon Sb substitution, accompanied by changes in the inter-quintuple Te-Te separation [30]. Such structural modifications may influence the vibrational restoring forces by changing the intra- and inter-quintuple interactions. Importantly, these structural effects are expected to influence different vibrational modes unequally. This was also demonstrated in a recent study, which reported that the low-frequency A1g1A_{1g}^1 mode in Sb2Te3\mathrm{Sb}_{2}\mathrm{Te}_{3} is particularly sensitive to inter-quintuple decoupling, while the high-frequency A1g2A_{1g}^2 mode responds more strongly to intralayer bonding distortion [31]. Accordingly, the distinct non-monotonic behavior of the A1g2A_{1g}^2 force constant may reflect its different sensitivity to the local structural and bonding modifications induced by Sb substitution.

Nevertheless, since the present force constants were calculated using composition-independent atomic displacements, possible changes in the phonon displacement eigenvectors upon alloying are not explicitly considered. Therefore, the calculated trends should be interpreted as model-dependent variations in the estimated vibrational stiffness rather than direct evidence of individual bond strengthening or weakening. However, the fact that our simplified model accurately reproduces an almost concentration-independent force constant for the low frequency Eg1E_g^1 and A1g1A_{1g}^1 modes (both below 100 cm−1100\ \mathrm{cm}^{-1}) in full agreement with the comparative study of the PDOS in Ref. [13], further strengthens the qualitative validity of our approach.

Raman spectroscopy of BST under pressure

To further investigate the lattice vibrational response to compression, we preformed high-pressure Raman measurements on BST compounds with x=0.2x=0.2, 0.7, and 0.9. From the four expected Raman-active modes, three (A1g1A_{1g}^{1}, Eg2E_{g}^{2} and A1g2A_{1g}^{2}) were systematically observed under pressure and will be used in the following analysis and discussion. The pressure ranges up to ∼7\sim7-8 GPa, was constrained to the maximum pressure at which the ambient conditions R3m‾\overline{R3m} structure is retained without coexistence of the C2/mC2/m phase [8]. Figure 3 and Figure 4(a) show the corresponding Raman spectra collected at selected pressures. All observed Raman-active modes for all BST compounds examined in this study shift monotonically toward higher frequencies upon compression, indicating a pressure-induced blue-shift, consistent with previous Raman studies under pressure for Bi2Te3\mathrm{Bi}_{2}\mathrm{Te}_{3} and Sb2Te3\mathrm{Sb}_{2}\mathrm{Te}_{3}. This behavior is consistent with changes (increases) in the interatomic restoring forces resulting from reduced interatomic distances under pressure.

Raman spectra of BST-0.2 and BST-0.9 at selected pressures

Figure 3. Raman spectra of (a) BST-0.2 and (b) BST-0.9 at selected pressures upon pressure increase.

Raman spectra and mode frequencies of BST-0.7

Figure 4. (a) Raman spectra of BST-0.7 at selected pressures upon pressure increase. (b) Raman mode frequencies of BST-0.7 as a function of pressure. The solid curves are fits using a second-order polynomial; see text for details.

To quantitatively describe the pressure-induced frequency shifts, the pressure dependence of the frequency of each resolved Raman mode was fitted using a second-order polynomial: ω(P)=ω0+αP+βP2\omega(P) = \omega_{0} + \alpha P + \beta P^{2}, where ω0\omega_{0} is the ambient-pressure frequency and α\alpha and β\beta are the corresponding first- and second-order pressure coefficients [32]. Figure 4(b) shows the corresponding frequency as a function of pressure for BST-0.7 and Table I summarizes the results for all BST-x compounds studied under pressure, along with the results for Bi2Te3\mathrm{Bi}_{2}\mathrm{Te}_{3} and Sb2Te3\mathrm{Sb}_{2}\mathrm{Te}_{3} from previous studies [32, 33].

Grüneisen parameters

Using γT\gamma_{T} removes part of the trivial difference associated with the absolute phonon frequency and instead quantifies how strongly each vibrational mode responds to a fractional change in volume. Pressure-dependent Raman studies of Bi2Te3\mathrm{Bi}_{2}\mathrm{Te}_{3} have previously demonstrated substantial mode-dependent frequency shifts and yielded mode Grüneisen parameters for the rhombohedral phase [33]. First-principles lattice-dynamical calculations have likewise identified appreciable vibrational anharmonicity in Bi2Te3\mathrm{Bi}_{2}\mathrm{Te}_{3}, giving a macroscopic average Grüneisen parameter of approximately 1.56 [18, 27]. These studies establish that the tetradymite lattice is intrinsically sensitive to changes in interatomic spacing, but they do not address how this response evolves continuously across the (Bi1−xSbx)2Te3(\mathrm{Bi}_{1-x}\mathrm{Sb}_{x})_{2}\mathrm{Te}_{3} solid solution.

Based on the results in the previous section, the frequency response to pressure differs substantially among the Raman modes and, more importantly, varies with Bi/Sb composition. The corresponding Raman modes isothermal Grüneisen parameters for all observed Raman modes of the R‾3m\overline{R}3m phase were calculated according to γiT=−(∂ln⁡ωi∂ln⁡V)T=B0ωi,0(∂ωi∂P)T\gamma_{iT}=-\left(\frac{\partial\ln\omega_i}{\partial\ln V}\right)_{T}=\frac{B_{0}}{\omega_{i,0}}\left(\frac{\partial\omega_i}{\partial P}\right)_{T}, where B0B_{0} is the ambient-pressure bulk modulus. For the compositions x=0.2, 0.7 and 0.9 we determined the B0B_{0} from the Equation of State (EOS) results for the R‾3m\overline{R}3m phases reported in our previous high-pressure XRD study [8]. For the two end members (Bi2Te3\mathrm{Bi}_{2}\mathrm{Te}_{3} and Sb2Te3\mathrm{Sb}_{2}\mathrm{Te}_{3}) we have used the EOSs from previous studies [34, 35] for the R‾3m\overline{R}3m phases. We note that when a 3rd-order Birch–Murnaghan EOS [36] is used to fit the P-V data, substantial differences in B0B_{0} values might arise between various studies [33], due to the well-known interplay (opposite correlation) between B0B_{0} and B0′B_{0}^{\prime} values [37]. This variation, will have a profound effect on the determined γiT\gamma_{iT} values for different BST compounds and Raman modes. For this reason, a 2nd-order Birch–Murnaghan EOS (B′=4B^{\prime}=4, fixed) was fitted to all EOS data used in this study [8, 34, 35]. Although a 2nd-order EOS might not produce the optimum least-squares fit, it minimizes the effect of the variation in the B0B_0 values determined from different experimental studies. Nevertheless, our aim is a comparative study of γT\gamma_T values and not an absolute determination. Using this methodology, the determined γT\gamma_T values for all observed Raman modes under pressure, are listed in Table I and shown in Figure 5.

γT values as a function of Sb concentration

Figure 5. (a) γT\gamma_T values, as a function of Sb concentration, for the A1g1A^{1}_{1g}, Eg2E^{2}_{g} and A1g2A^{2}_{1g} Raman modes, see text for details, together with the average one from these modes.

From Figure 5, after taking into account relevant uncertainties and experimental errors, certain conclusions could be made. The γT\gamma_T of both the A1g1A_{1g}^{1} and Eg2E_g^{2} Raman modes shows a maximum at the composition x=0.7x=0.7, while the one of the A1g2A_{1g}^{2} appears practically composition independent. The maximum at x=0.7x=0.7 is also reflected in the average γT\gamma_T for the three modes. Although a definite explanation for the composition independence of γT\gamma_T for the A1g2A_{1g}^{2} mode is not possible; it is plausible to assume that this is related to the stiffening of the force constant upon increase of the Sb content. Based on Slack’s equation [38, 39], the thermal conductivity is inversely proportional to γT2\gamma_T^{2}. Although thermal conductivity is mainly correlated with acoustic phonons, our analysis provides a clear indication about the origin of the lower thermal conductivity for BST-07, which, as mentioned, is highlighted as the one with the maximum ZT values. Further studies are needed, mainly on the γT\gamma_T of acoustic phonons, to fully elucidate this correspondence.

CompositionBi2_2Te3_3BST-0.2BST-0.7BST-0.9Sb2_2Te3_3
B0B_045.045.943.642.840
ω0 A1g1\omega_0\ A^1_{1g}61.363.366.56870.6
ω0 Eg2\omega_0\ E^2_g101.1103.2107.8110.6114.2
ω0 A1g2\omega_0\ A^2_{1g}132.1144.1159.2162.7166.4
∂ω/∂P A1g1\partial\omega/\partial P\ A^1_{1g}3.713.574.874.294.42
∂ω/∂P Eg2\partial\omega/\partial P\ E^2_g3.493.344.523.802.86
∂ω/∂P A1g2\partial\omega/\partial P\ A^2_{1g}2.923.322.723.732.75
γT A1g1\gamma_T\ A^1_{1g}2.722.593.192.702.50
γT Eg2\gamma_T\ E^2_g1.551.481.831.471.00
γT A1g2\gamma_T\ A^2_{1g}0.991.060.740.980.66

Table 1. Bulk modulus B0B_0 (GPa), ambient-pressure Raman mode frequency ω0\omega_0 (cm−1^{-1}), frequency-pressure slope ∂ω/∂P\partial\omega/\partial P (cm−1^{-1}/GPa) and Grüneisen parameter γiT\gamma_{iT} for the BST-X compounds discussed in the text. The values for Bi2_2Te3_3 are from Refs. [34] (B0B_0) and [33] (Raman measurements) and for Sb2_2Te3_3 from Refs. [35] (B0B_0) and [32] (Raman measurements).

Conclusion

Based on the force-constant analysis using the Raman measurements at ambient conditions for the BST compounds in this study, a diverging response of the force constant as a function of the composition was concluded, beyond the PDOS scaling proposed in previous studies. This indicates that the force constants may reflect the distinct sensitivity of each Raman mode to local structural and bonding modifications induced by Sb substitution. The average Grüneisen parameter, calculated from the corresponding γT\gamma_T of the A1g1A^1_{1g}, Eg2E^2_g and A1g2A^2_{1g} Raman modes, exhibits a clear maximum for BST-07. Based on Slack’s equation, this is directly related to the optimal ZT values for this composition. Our study represents a critical step forward towards establishing the microscopic origin of thermoelectric optimization in BST compounds. Moreover, it provides the framework for connecting composition, bonding, lattice dynamics, and thermal-transport behavior.

Acknowledgements

We thank Azkar Saeed Ahmad for his help with specimens loadings and Muhamed Dawod and Yaron Amouyal for providing the BST specimens. The work was supported by funding from the Guangdong Technion Israel Institute of Technology, the Guangdong Provincial Key Laboratory of Materials and Technologies for Energy Conversion, MATEC (Grant No. 2022B1212010007, Guangdong Department of Science and Technology) and the Technion-GTIIT seed grant program.

Availability of data and materials

The data that support the findings of this study are available from the corresponding author upon reasonable request.

AI and AI-assisted tools Statement

AI-assisted tools were used solely for language and grammar editing.

Conflicts of interest

The authors have no conflicts to disclose.

References

  1. [1]J. He and T. M. Tritt, Advances in thermoelectric materials research: Looking back and moving forward, Science 357, eaak9997 (2017).DOI
  2. [2]X.-L. Shi, N.-H. Li, M. Li, and Z.-G. Chen, Toward efficient thermoelectric materials and devices: Advances, challenges, and opportunities, Chem. Rev. 125, 7525 (2025).DOI
  3. [3]T. Lu, L. Chen, X. Shi, W. Liu, M. Li, Y. Wang, X. Zeng, P. Song, J. Bell, G. J. Snyder, Z. Chen, and M. Hong, Advances in synthesis of high-performance thermoelectric materials, Small Struct. 6, 2400694 (2025).
  4. [4]Q. Xiong, G. Han, G. Wang, X. Lu, and X. Zhou, The doping strategies for modulation of transport properties in thermoelectric materials, Adv. Funct. Materials 34, 2411304 (2024).DOI
  5. [5]M. Zhou, H. Su, J. Pei, L. Wang, H. Zhuang, J.-F. Li, K. Song, H. Hu, J. Jiang, Q. Zhang, J. Li, and L. Li, Ultrahigh thermoelectricity obtained in classical BiSbTe alloy processed under super-gravity, Nat. Commun. 16, 7645 (2025).DOI
  6. [6]C. H. Champness, P. T. Chiang, and P. Parekh, Thermoelectric properties of Bi₂Te₃-Sb₂Te₃ alloys, Can. J. Phys. 43, 653 (1965).
  7. [7]A. N. Mansour, W. Wong-Ng, Q. Huang, W. Tang, A. Thompson, and J. Sharp, Structural characterization of Bi₂Te₃ and Sb₂Te₃ as a function of temperature using neutron powder diffraction and extended x-ray absorption fine structure techniques, J. Appl Phys. 116, 083513 (2014).DOI
  8. [8]C. Wei, M. Dawod, W. Lu, H. Zou, B. Sun, S. Feng, Q. Zhang, Z. Su, H. Kadobayashi, M. Kunz, B. Wang, A. S. Ahmad, Y. Amouyal, and E. Stavrou, Comparative high-pressure structural and electrical transport properties study of thermoelectric (Bi₁₋ₓSbₓ)₂Te₃ compounds, Discov. Mater. 5, 137 (2025).arxiv.org/abs/2503.06562
  9. [9]B. Poudel, Q. Hao, Y. Ma, Y. Lan, A. Minnich, B. Yu, X. Yan, D. Wang, A. Muto, D. Vashaee, X. Chen, J. Liu, M. S. Dresselhaus, G. Chen, and Z. Ren, High-thermoelectric performance of nanostructured bismuth antimony telluride bulk alloys, Science 320, 634 (2008).DOI
  10. [10]S. I. Kim, K. H. Lee, H. A. Mun, H. S. Kim, S. W. Hwang, J. W. Roh, D. J. Yang, W. H. Shin, X. S. Li, Y. H. Lee, G. J. Snyder, and S. W. Kim, Dense dislocation arrays embedded in grain boundaries for high-performance bulk thermoelectrics, Science 348, 109 (2015).DOI
  11. [11]Y. Yu, D.-S. He, S. Zhang, O. Cojocaru-Mirédin, T. Schwarz, A. Stoffers, X.-Y. Wang, S. Zheng, B. Zhu, C. Scheu, D. Wu, J.-Q. He, M. Wuttig, Z.-Y. Huang, and F.-Q. Zu, Simultaneous optimization of electrical and thermal transport properties of Bi₀.₅Sb₁.₅Te₃ thermoelectric alloy by twin boundary engineering, Nano Energy 37, 203 (2017).
  12. [12]T. Fang, X. Li, C. Hu, Q. Zhang, J. Yang, W. Zhang, X. Zhao, D. J. Singh, and T. Zhu, Complex band structures and lattice dynamics of Bi₂Te₃-based compounds and solid solutions, Adv. Funct. Materials 29, 1900677 (2019).
  13. [13]D. Bessa, I. Sergueev, H.-C. Wille, J. Perßon, D. Ebling, and R. P. Hermann, Lattice dynamics in bi₂ te₃ and sb₂te₃: Te and sb density of phonon states, Phys. Rev. B 86, 224301 (2012).
  14. [14]S. Elyamny, N. Imam, G. Aquilanti, H. Cabrera, and A. E.-H. B. Kashyout, Thermal transport properties for unveiling the mechanism of BiSbTe alloys in thermoelectric generation: A glance from synchrotron radiation Bi L₃-XAFS, J. Mater. Res. Technol. 18, 2261 (2022).
  15. [15]W. Richter and C. R. Becker, A Raman a and far-infrared investigation of phonons in the rhombohedral V₂−VI₃ compounds Bi₂Te₃, Bi₂Se₃, Sb₂Te₃, Bi₂(Te₁−ₓSeₓ)₃ (0<x<1) and (Bi₁−ᵧSbᵧ)₂Te₃ (0<y<1), Phys. Status Solidi B 84, 619 (1977).
  16. [16]C. Zhang, Z. Peng, Z. Li, L. Yu, and K. A. Khor, Controlled growth of bismuth antimony telluride BiₓSb₂−ₓTe₃ nanoplatelets and their bulk thermoelectric nanocomposites, Nano Energy (2015).DOI
  17. [17]W. Kullmann, J. Geurts, W. Richter, N. Lehner, H. Rauh, U. Steigenberger, G. Eichhorn, and R. Geick, Effect of hydrostatic and uniaxial pressure on structural properties and raman active lattice vibrations in bi₂ te₃, Phys. Status Solidi B 125, 131 (1984).
  18. [18]G. K. Pradhan, A. Bera, P. Kumar, D. Muthu, and A. Sood, Raman signatures of pressure induced electronic topological and structural transitions in Bi₂Te₃, Solid State Commun. 152, 284 (2012).DOI
  19. [19]D. Das, S. Das, P. Singha, K. Malik, A. K. Deb, A. Bhattacharyya, V. A. Kulbachinskii, R. Basu, S. Dhara, S. Bandyopadhyay, and A. Banerjee, Evolution of phonon anharmonicity in Se-doped Sb₂Te₃ thermoelectrics, Phys. Rev. B 96, 064116 (2017).DOI
  20. [20]F.-X. Bai, H. Yu, D. Peng, H.-J. Pang, L. Yu, L.-C. Chen, J. Mao, Q. Zhang, and X.-J. Chen, Strong acoustic-optical phonon coupling in high-performance thermoelectric material Bi₀.₅Sb₁.₅Te₃, Materials Today Physics 37, 101200 (2023).DOI
  21. [21]J. K. Hinton, S. M. Clarke, B. A. Steele, I.-F. W. Kuo, E. Greenberg, V. B. Prakapenka, M. Kunz, M. P. Kroonblawd, and E. Stavrou, Effects of pressure on the structure and lattice dynamics of α-glycine: a combined experimental and theoretical study, CrystEngComm 21, 4457 (2019).
  22. [22]K. Syassen, Ruby under pressure, High Pressure Res. 28, 75 (2008).
  23. [23]S. Klotz, J. Chervin, P. Munsch, and G. Le Marchand, Hydrostatic limits of 11 pressure transmitting media, J. Phys. D: Appl. Phys. 42, 075413 (2009).DOI
  24. [24]G. Keresztury, Raman spectroscopy: Theory, in Handbook of Vibrational Spectroscopy, edited by J. M. Chalmers and P. R. Griffiths (Wiley, 2001) 1st ed.
  25. [25]N. A. Abdullaev, I. R. Amiraslanov, Z. S. Aliev, Z. A. Jahangirli, I. Y. Sklyad­neva, E. G. Alizade, Y. N. Aliyeva, M. M. Otrokov, V. N. Zverev, N. T. Mamedov, and E. V. Chulkov, Lattice dynamics of Bi₂Te₃ and vibrational modes in Raman scattering of topological insulators MnBi₂Te₄n(Bi₂Te₃), JETP Lett. 115, 749 (2022).
  26. [26]K. Norimatsu, M. Hada, S. Yamamoto, T. Sasagawa, M. Kitajima, Y. Kayanuma, and K. G. Nakamura, Dynamics of all the raman-active coherent phonons in Sb₂Te₃ revealed via transient reflectivity, J. Appl. Phys. 117, 143102 (2015).
  27. [27]R. R. Urkude and U. A. Palikundwar, Phonon dynamics of (Bi₁−xSbx)₂Te₃ topological insulators by Raman spectroscopy, Phys. B: Condens. Matter. 655, 414754 (2023).DOI
  28. [28]G. C. Sosso, S. Caravati, and M. Bernasconi, Vibrational properties of crystalline Sb₂Te₃ from first principles, J. Phys.: Condens. Matter. 21, 095410 (2009).DOI
  29. [29]Y. Huang, J. D. Querales-Flores, S. W. Teitelbaum, J. Cao, T. Henighan, H. Liu, M. Jiang, G. De La Peña, V. Krapivin, J. Haber, T. Sato, M. Chollet, D. Zhu, T. Katayama, R. Power, M. Allen, C. R. Rotundu, T. P. Bailey, C. Uher, M. Trigo, P. S. Kirchmann, E. D. Murray, Z.-X. Shen, I. Savić, S. Fahy, J. A. Sobota, and D. A. Reis, Ultrafast measurements of mode-specific deformation potentials of Bi₂Te₃ and Bi₂Se₃, Phys. Rev. X 13, 041050 (2023).
  30. [30]F. Serrano-Sánchez, M. Gharsallah, N. M. Nemes, N. Biskup, M. Varela, J. L. Martínez, M. T. Fernández-Díaz, and J. A. Alonso, Enhanced figure of merit in nanostructured (Bi,Sb)₂Te₃ with optimized composition, prepared by a straightforward arc-melting procedure, Sci. Rep. 7, 6277 (2017).
  31. [31]J. Mertens, P. Kerres, Y. Xu, M. Raghuwanshi, D. Kim, C. Schön, J. Frank, F. Hoff, Y. Zhou, R. Mazzarello, A. R. Jalil, and M. Wuttig, Confinement-induced phonon softening and hardening in Sb₂Te₃ thin films, Adv Funct Materials 34, 2307681.
  32. [32]S. Souza, C. Poffo, D. Trichês, J. de Lima, T. Grandi, A. Polian, and M. Gauthier, High pressure monoclinic phases of Sb₂Te₃, Phys. B: Condens. Matter. 407, 3781 (2012).
  33. [33]R. Vilaplana, O. Gomis, F. J. Manjón, A. Segura, E. Pérez-González, P. Rodríguez-Hernández, A. Muñoz, J. González, V. Marín-Borrás, V. Muñoz-Sanjosé, C. Drasar, and V. Kucek, High-pressure vibrational and optical study of Bi₂Te₃, Phys. Rev. B 84, 104112 (2011).
  34. [34]M. Einaga, A. Ohmura, A. Nakayama, F. Ishikawa, Y. Yamada, and S. Nakano, Pressure-induced phase transition of Bi₂Te₃ to a bcc structure, Phys. Rev. B 83, 092102 (2011).DOI
  35. [35]N. Sakai, T. Kajiwara, K. Takemura, S. Minomura, and Y. Fujii, Pressure-induced phase transition in Sb₂Te₃, Solid State Commun. 40, 1045 (1981).DOI
  36. [36]F. Birch, Finite elastic strain of cubic crystals, Physical Review 71, 809 (1947).
  37. [37]H. Zou, A. Azulay, W. Lu, B. Sun, H. Kadobayashi, B. Wang, L. He, J. Hao, A. S. Ahmad, Y. Amouyal, and E. Stavrou, High pressure structural study of y- and la-doped camno3 perovskite, High Pressure Res. 46, 258 (2026).
  38. [38]G. Slack, Nonmetallic crystals with high thermal conductivity, J. Phys. Chem. Solids 34, 321 (1973).
  39. [39]E. J. Skoug, J. D. Cain, and D. T. Morelli, Structural effects on the lattice thermal conductivity of ternary antimony- and bismuth-containing chalcogenide semiconductors, Appl. Phys. Lett. 96, 181905 (2010).DOI

Paper details

vlnc:W_uu5zyefy
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

AI-assisted tools

View statement
Looking for related papers…
Contents