Abstract
Sb2Te3 alloys are promising thermoelectric materials because of their outstanding electrical transport properties in the midtemperature range of 500–700 K, while codoping with multiple elements has been successful to improve their thermoelectric performance. In this study, enhanced thermoelectric properties with a maximum thermoelectric figure of merit of 0.97 are reported for singly and lightly Pb-doped Sb2Te3 polycrystalline alloys (Sb2–PbTe3). Very light Pb doping in the range in the Sb2–PbTe3 alloys yielded significantly improved carrier transport properties and increased electrical conductivity while the Seebeck coefficient is decreased moderately, since the density-of-state effective mass is improved much. As a result, power factor for the Pb-doped Sb2Te3 is largely increased up to 3.7 mW/mK2 at 300 K. The lattice thermal conductivity decreased considerably owing to the additional point defect phonon scattering by the Pb despite slight doping. Consequently, a maximum thermoelectric figure of merit of 0.97 was obtained for Sb1.9875Pb0.0125Te3 () at 600 K, which is the highest reported value for singly doped Sb2Te3-based alloys. A maximum energy conversion efficiency was calculated to be 9.0% for a temperature difference of 350 K, which is higher than that for other singly or codoped Sb2Te3 alloys.
1. Introduction
Thermoelectric materials have been attracting attention for solid-state cooling and energy harvesting because they can convert heat directly into electricity. Therefore, there exist many potential applications, including cryoprobes, on-chip cooling, and power generators based on waste heat from automobiles and industries [1–4]. However, the widespread applications of thermoelectric materials depend on their conversion efficiency, which is typically characterized by the thermoelectric figure of merit. The thermoelectric figure of merit is a dimensionless quantity given by , where , , , and are the Seebeck coefficient, electrical conductivity, absolute temperature, and total thermal conductivity, respectively. Note that the average figure of merit () is an important measure of the power generation efficiency of a thermoelectric material over a temperature range. The maximum energy conversion efficiency () of a thermoelectric material can be estimated using the following equation: where is the temperature of the cold side of the thermoelectric material, is the temperature of the hot side of the thermoelectric material, is , and is .
(Bi,Sb)2Te3 solid solution alloys are the most widely used -type thermoelectric materials at room temperature. In particular, Sb-rich alloys such as Bi0.5Sb1.5Te3 and Bi0.4Sb1.6Te3 are known to exhibit high -type thermoelectric performance because of their optimized carrier transport properties [5–7]. Pristine Bi2Te3 typically exhibits -type electrical transport behavior; interestingly, when Bi2Te3 is alloyed with Sb2Te3, which exhibits -type electrical transport properties, a high of >2,000 S/cm can be obtained [8, 9]. Consequently, (Bi,Sb)2Te3 alloys can exhibit a high thermoelectric performance in the low-temperature range of 300–400 K. However, their performance degrades rapidly at higher temperatures owing to the occurrence of bipolar conduction, which is unfavorable for electric power generation [10]. In contrast, Sb2Te3 alloys demonstrate a high thermoelectric performance in the midtemperature range of 500–700 K, with no significant occurrence of bipolar conduction. Nevertheless, the maximum of Sb2Te3 alloys is lower than that of (Bi,Sb)2Te3 alloys.
Several studies have been conducted to enhance the of Sb2Te3 alloys in the midtemperature range. In particular, codoping Sb2Te3 with multiple elements has been shown to be successful [11–14]. For instance, Qin et al. reported a of 1.0 at 673 K for Sb2Te3 codoped with Mn and In [12]; in addition, Hu et al. reported a maximum of 0.92 at 710 K for Sb2Te3 codoped with In and Ag [13]. Moreover, a of 0.96 at 680 K was reported for Sb2Te3 codoped with Mg and In [14]. However, Sb2Te3 has been typically shown to exhibit a high thermoelectric performance when codoped with two or more elements [15–18]. The maximum achievable value for singly doped Sb2Te3 alloys is 0.7, which is significantly lower than that obtained via codoping [13, 19].
In this study, Sb2Te3 was doped with a single dopant, Pb, to enhance its thermoelectric performance. The performance of (Bi,Sb)2Te3 alloys has been shown to improve when Pb was used as an acceptor dopant [20–22]. To this end, a series of Sb2–PbTe3 alloys with , 0.0025, 0.005, 0.0075, 0.01, and 0.0125 was synthesized, where is the doping content. Pb doping significantly enhanced the power factor of the alloys, and a maximum of 0.97 at 600 K was achieved for (i.e., Sb1.9875Pb0.0125Te3). The thermoelectric properties of the Sb2–PbTe3 alloys were analyzed using various phenomenological parameters, including the density of state (DOS) effective mass () and weighted mobility (). was found to increase linearly with the doping content, which in turn positively affected the thermoelectric performance of the Sb2–PbTe3 alloys. In addition, the further doping of Pb in Sb2–PbTe3 was investigated by introducing large amounts of Pb corresponding to , 0.1, 0.15, and 0.2 (Supplementary Information (available here)).
2. Experimental Methods
Polycrystalline Sb2–PbTe3 (, 0.0025, 0.005, 0.0075, 0.01, and 0.0125) samples were prepared using the traditional solid-state reaction method. Stoichiometric quantities of high purity Sb (99.999%; 5N Plus), Pb (99.999%; Alfa Aesar), and Te (99.999%; 5N Plus) were weighed according to the nominal composition of the desired alloy, placed in a quartz tube, and sealed under vacuum (10–5 Torr). The sealed quartz ampoules were heated at 1,000 K for 4 h, and this temperature was further maintained for 6 h. Next, the ampoules were quenched in water at room temperature, following which the respective metal ingots were extracted from the ampoules. Ball milling (SPEX 8000D, Costa Mesa, USA) was performed in an Ar atmosphere to convert the ingots into powders. Subsequently, the powders were sintered at 573 K for 5 min at a pressure of 70 MPa under vacuum (10–6 Torr) to form cylindrical bulk pellets using spark plasma sintering (SPS; SPS-1030, Sumitomo Coal Mining Co. Ltd., Tokyo, Japan). Note that the sintered bulk samples had a relative density of more than 99%.
The crystal structure of the samples was analyzed using X-ray diffraction (XRD; D8 Discover, Bruker, USA) with Cu–Kα1 radiation. Atomic percentage of Pb dopant in the samples is measured using energy-dispersive spectroscopy (EDS, Quantax Xflash 6-60, Bruker, USA) by scanning electron microscope (SEM, SU8010, HITACHI, Japan). The and values of the samples were measured using a thermoelectric property measuring system (ZEM-3M8, Advance Riko, Japan) in a He atmosphere between 300 and 650 K. The Hall measurements were conducted in the van der Pauw configuration at 300 K using a Hall measurement system (HMS-5300, Ecopia, Korea). A laser flash analyzer (LFA457, Netzsch, Germany) was used to obtain the thermal diffusivity () of the samples, which was used to calculate , where and are the density and specific heat capacity of the sample, respectively. was measured (relative density of ≥97%), while J/g·K was used based on previously reported data [23]. The results corresponding to high Pb doping (i.e., ) are presented in the Supplementary Information.
3. Results and Discussion
Figures 1(a) and 1(b) show the XRD patterns of the polycrystalline bulk Sb2-PbTe3 (, 0.0025, 0.005, 0.0075, 0.01, and 0.0125) samples obtained along the directions parallel (PA) and perpendicular (PE) to the sintering direction, respectively. The XRD patterns confirm the synthesis of different Pb-doped Sb2Te3 phases in accordance with the JCPDS#01-071-0393 card. The lattice parameters ( and ) of the samples were calculated using the (015) and (1010) diffraction peaks. Figure 1(c) shows the lattice parameters as functions of . The error bars indicate the measurement error of the diffraction data of Figures 1(a) and 1(b). The lattice parameter did not vary significantly with , whereas the lattice parameter increased monotonically with . This increase in with can be explained by the difference between the ionic radii of Sb3+ (90 pm) and Pb2+ (133 pm), which enables the successful doping of Pb at the Sb sites. In addition, Table 1 shows atomic ratio in Sb2-PbTe3 measured by EDS, and it was confirmed systematic increase in the amount of Pb (see Supplementary Information Figure S1 for EDS mapping). However, PbTe appeared as an impurity in the highly doped samples (i.e., ; see Supplementary Information Figure S2(a)). This was further confirmed by the behavior of the corresponding lattice parameters with (see Supplementary Information Figure S2(b)).

(a)

(b)

(c)
Figures 2(a) and 2(e) show as a function of temperature for the Pb-doped Sb2Te3 samples along the PE and PA directions, respectively. of the Sb2–PbTe3 samples increased with along both the PE and PA directions. The values at 300 K along the PE direction were 2,240, 2,440, 2,620, 3,080, 3,570, and 3,770 S/cm for , 0.0025, 0.005, 0.0075, 0.01, and 0.0125, respectively, whereas the values at 300 K along the PA direction were 1,640, 1,870, 2,020, 2,370, 2,430, and 2,730 S/cm for , 0.0025, 0.005, 0.0075, 0.01, and 0.0125, respectively. For the highly doped samples (), there was a sharp increase in , which had values in the range 5000–6000 S/cm; however, decreased with a further increase in owing to the formation of secondary PbTe phases in these samples (see Supplementary Information; Figure S3(a)). This decrease in with an increase in the doping content is because of the relatively low of the PbTe phase of approximately 120–200 S/cm [24, 25].

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)
Figures 2(b) and 2(f) show as a function of temperature for the Sb2–PbTe3 samples along the PE and PA directions, respectively. increased with temperature for all the samples. The value for was slightly higher than that for ; however, for , decreased with along both the measured directions. Compared to the relatively large increase in with (Figures 2(a) and 2(e)), the decrease in with was less significant. However, with further Pb doping (i.e., ), decreased significantly, reaching a minimum value of 30 μV/K (see Supplementary Information Figure S3(b)). Figures 2(c) and 2(g) show the calculated power factor of the Sb2–PbTe3 samples as a function of temperature along the PE and PA directions, respectively. The power factor of the Pb-doped samples was significantly higher than that of pristine Sb2Te3 across the entire temperature range. A maximum power factor of 3.66 mW/mK2 was obtained for at 300 K along the PE direction, which is 39% higher than that for . A maximum power factor of 2.75 mW/mK2 was obtained for along the PA direction, which is 40% higher than that for (1.96 mW/mK2). A gradual decrease in the power factor was observed with further doping (see Supplementary Information Figure S3(c)).
Figures 2(d) and 2(h) show as a function of temperature for the Sb2–PbTe3 samples along the PE and PA directions, respectively. is related to the maximum theoretical efficiency of a thermoelectric material and can be defined using the following analytical approximation of the Drude-Sommerfeld free-electron model for [26]: where and denote Planck’s constant and the electron mass, respectively. Note that is proportional to the maximum power factor of the sample, which can be obtained by optimizing the Hall carrier concentration . In general, increased with the doping content along both the PE and PA directions. This confirms that the Pb-doped samples have enhanced electrical transport properties and hence can exhibit enhanced thermoelectric performance.
Figures 3(a) and 3(b) show and (along both the PE and PA directions) of the Sb2–PbTe3 samples as functions at 300 K. steadily increased with increasing Pb content and reached a maximum value of for . By contrast, decreased linearly with along both the measured directions. Moreover, along the PA direction was approximately 27% lower than that along the PE direction. The values along the PE direction were 281, 258, 254, 239, and 218 cm2/Vs for , 0.0025, 0.005, 0.0075, 0.01, and 0.0125, respectively, whereas the values along the PA direction were 221, 216, 214, 202, and 172 cm2/Vs for , 0.0025, 0.005, 0.0075, 0.01, and 0.0125, respectively. The increase in with can be attributed to the increase in with , which in turn is owing to the substitution of the Sb3+ ions by the Pb2+ ions. Supplementary Information Figure S4 presents the corresponding results of the highly doped samples.

(a)

(b)

(c)

(d)
The electrical transport properties of semiconductors are determined by the DOS at the Fermi level. The magnitude of the DOS is directly related to . Thus, can be defined in terms of changes in the electron structure using the single parabolic band model [27], such that
Figures 3(c) and 3(d) show of the Sb2–PbTe3 samples as a function of along the PE and PA directions, respectively. The insets in Figures 3(c) and 3(d) show as a function of for different samples, where the curves correspond to fixed values. increased with the Pb content along both the PE and PA directions, similar to . The reason for this increase in with is that the rate of increase of with is faster than the rate of decrease of with . The values at 300 K along the PE direction were 0.960, 1.05, 1.09, 1.16, 1.17, and 1.26 ( being the rest mass of an electron) for , 0.0025, 0.005, 0.0075, 0.01, and 0.0125, respectively. along the PA direction also exhibited a similar trend. The maximum value of along the PA direction was 1.34 for . However, for , the value at 300 K decreased linearly with increasing doping content (Supplementary Information Figures S4(c) and (d)). For , of the highly doped samples was lower than that of pristine Sb2Te3.
Figure 4 shows and the lattice thermal conductivity of the Sb2–PbTe3 samples as functions of temperature along the PE and PA directions. of the Pb-doped samples varied nonmonotonically with along both the measured directions, as shown in Figures 4(a) and 4(c). The insets in Figures 4(a) and 4(c) show the electrical thermal conductivity of the Sb2–PbTe3 samples as a function of temperature, which was calculated using the Wiedemann-Franz law [28], namely, (where is the Lorenz number). can be calculated using the following equation [29]:

(a)

(b)

(c)

(d)
Figures 4(b) and 4(d) show as a function of temperature along the PE and PA directions, respectively, which was calculated by subtracting from . decreased with owing to an increase in the point defect phonon scattering by the Pb2+ ions. The for high-doped samples for and 0.0012 for the PE direction was calculated very low, but this seems to be due the error in the calculation, which can be expected from the error in for very high conductive alloys ( S/cm) [25]. (See Figure S5 in Supplementary Materials for thermal transport properties of the highly doped Sb2–PbTe3 samples along the PE direction.)
Figures 5(a) and 5(b) show the of the Sb2–PbTe3 samples as a function of temperature along the PE and PA directions, respectively. A maximum or of 0.97 was obtained for at 600 K along the PE direction, which is approximately 45% higher than that of pristine Sb2Te3 ( at 550 K). A of 0.84 was obtained for at 550 K along the PA direction, which is 33% higher than that of pristine Sb2Te3 ( at 550 K). (See Figure S6 in Supplementary Materials for of the highly doped Sb2–PbTe3 samples.)

(a)

(b)

(c)

(d)

(e)

(f)
Figure 5(c) compares the values of Sb1.9875Pb0.0125Te3 (i.e., ) and several high-performance thermoelectric alloys in the midtemperature (namely, CeFe4Sb12, Yb0.3Co4Sb12, and Sn0.098Cu0.02Se) and near-room temperature (namely, -type Bi0.5Sb1.5Te3 and Bi0.4Sb1.6Te3) range [30–34]. The values of Sb1.9875Pb0.0125Te3 in this temperature range were comparable to those of filled skutterudites such as XFe3Sb12 and XCo3Sb12 (where , Ce, In, etc.), which require a complex and lengthy synthesis process [32–34]. In addition, the values of Sb1.9875Pb0.0125Te3 were comparable to those of Sb2Te3 alloys codoped with In (Figure 5(d)). Furthermore, the values of Sb1.9875Pb0.0125Te3 were much higher than those of codoped Sb2Te3 alloys in the temperature range 300–600 K. Consequently, the of Sb1.9875Pb0.0125Te3 was 0.71, which is higher than that of the codoped Sb2Te3 alloys (Figure 5(e)).
Figure 5(f) compares the values of Sb1.9875Pb0.0125Te3 and other alloys. A of 9.0% was obtained for Sb1.9875Pb0.0125Te3 corresponding to a temperature difference of 350 K, which was calculated using Eq. (1). This is 32% higher than that of singly doped Sb1.85In0.15Te3 () [13]. Moreover, although Mn0.02Sb1.83In0.15Te3 () exhibited the highest [12], of Sb1.9875Pb0.0125Te3 exceeded that of Mn0.02Sb1.83In0.15Te3. The singly and lightly Pb-doped Sb2Te3 could outperform other singly or codoped Sb2Te3 with respect to power generation in midtemperature range.
4. Conclusion
In this study, we synthesized a series of polycrystalline bulk Sb2–PbTe3 (with , 0.0025, 0.005, 0.0075, 0.01, and 0.0125) samples using the traditional solid-state reaction method. The electrical conductivity of the samples increased from 2,240 S/cm for pristine Sb2Te3 to 3,570 S/cm with Pb doping. However, the Seebeck coefficient of the samples was not affected significantly by the improved effective mass. An optimal power factor of 3.7 mW/mK2 was obtained at 300 K for , which is 39% higher than that of the pristine sample (2.6 mW/mK2). The lattice thermal conductivity of the Pb-doped samples decreased significantly with the doping content owing to an increase in point defect phonon scattering. Consequently, a maximum value of 0.97 was obtained for , which is 45% higher than that of pristine Sb2Te3. We also compared the performance of the Sb2–PbTe3 alloys developed in this work with that of other high-efficiency thermoelectric alloys in the midtemperature range, including filled skutterudites and codoped Sb2Te3 alloys.
Data Availability
Data are available upon reasonable request.
Conflicts of Interest
The authors declare that they have no competing interests.
Authors’ Contributions
Okmin Park and Kyu Hyoung Lee contributed equally to this work.
Acknowledgments
This study was supported by the Nano·Material Technology Development Program under the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (2022M3H4A1A04076667).
Supplementary Materials
The energy-dispersive spectroscopy (Quantax Xflash 6-60, Bruker, USA) by scanning electron microscope (SU8010, HITACHI, Japan) is measured to confirm the doping for the samples. Table S1: atomic percentage in Sb2-PbTe3 samples obtained from energy-dispersive spectroscopy (EDS). Figure S1: EDS mapping images of Sb2-PbTe3 samples (, 0.0025, 0.005, 0.075, 0.01, and 0.0125). The higher doping of Pb in Sb2–PbTe3 beyond was investigated by introducing large amounts of Pb corresponding to , 0.1, 0.15, and 0.2, and the results are shown in Supplementary Materials. Figure S2: (a) XRD patterns and (b) lattice parameters of the highly doped Sb2–PbTe3 samples. Figure S3: thermoelectric transport properties of the highly doped Sb2–PbTe3 samples along the PE direction: (a) electrical conductivity, (b) Seebeck coefficient, and (c) power factor as functions of temperature. Figure S4: carrier transport properties of the highly doped Sb2–PbTe3 samples along the PE direction: (a) Hall carrier concentration, (b) Hall mobility, and (c) DOS effective mass, as functions of the doping content. (d) Logarithmic carrier concentration as a function of the absolute Seebeck coefficient. Figure S5: thermal transport properties of the highly doped Sb2–PbTe3 samples along the PE direction. (a) Total and (b) lattice thermal conductivity, as functions of temperature. The inset in (a) shows the electron thermal conductivity as a function of temperature. Figure S6: (a) thermoelectric figure of merit and (b) weighted mobility of the highly doped Sb2–PbTe3 samples, as functions of temperature. (Supplementary Materials)