Thermoelectric materials are used to generate power and for solid-state cooling by exploiting their ability to convert heat into electricity (Seebeck effect) and heat pumping with electricity (Peltier effect). The energy conversion efficiency of thermoelectric devices is primarily determined by the dimensionless figure of merit, ZT = α²σκ^-1T, where Z is the temperature-dependent figure of merit, T is the absolute temperature, α is the Seebeck coefficient, σ is the electrical conductivity, and κ is the thermal conductivity [1-3]. Cu-based chalcogenides, which are composed of low-toxicity, low-cost, and earth-abundant elements, are considered promising p-type thermoelectric materials because of their excellent thermoelectric performance [3, 4]. The ternary Cu-Sb-S sulfide system comprises four major phases: CuSbS₂ (chalcostibite), Cu₁₂Sb₄S₁₃ (tetrahedrite), Cu₃SbS₃ (skinnerite), and Cu₃SbS₄ (famatinite). Famatinite crystallizes in an ordered zinc-blende superstructure with space group I4¯2m (no.121) [5, 6], and thus possesses a very simple crystal structure compared with other Cu-Sb-S ternary compounds [7-9]. Cu₃SbS₄ is a p-type thermoelectric material in which the major charge carriers are holes [5], and its bandgap energy is in the range 0.4–1.05 eV, which makes it a suitable thermoelectric material at intermediate temperatures [10-13]. Cu₃SbS₄ has a relatively low lattice thermal conductivity and a large Seebeck coefficient, but exhibits low electrical conductivity; thus, it is a likely candidate for improving electrical characteristics, to increase thermoelectric efficiency [8, 14, 15].

Because the thermoelectric performance of pure Cu₃SbS₄ is mainly limited by low electrical conductivity caused by the low carrier concentration, approaches have been studied to optimize the carrier concentration by partially substituting (doping) other elements for Cu or Sb. Shen et al. [15] reported a ZT of 0.76 at 623 K for Cu₃Sb_0.89Bi_0.06Sn_0.05S₄ by co-doping Bi and Sn at the Sb sites. Suzumura et al. [16] obtained a ZT of 0.095 at 323 K for Cu_2.91Ag_0.09Sb_0.98Ge_0.02S₄ by co-doping Ag at the Cu and Ge at the Sb sites. Chen et al. [17] reported ZT values of 0.37 at 523 K for Cu_2.25Ni_0.75SbS₄ and 0.22 at 523 K for Cu₃Ni_0.05Sb_0.95S₄ by substituting Ni into the Cu and Sb sites, respectively. Tanishita et al. [18] achieved a ZT of 0.67 at 623 K for Cu₃Sb_0.75Ge_0.1P_0.15S₄ by co-doping Ge and P at Sb sites. Pi et al. [19] reported a ZT of 0.67 at 623 K for Cu₃Sb_0.92Sn_0.08S₄ by substituting Sn at Sb sites. In our previous study [20], In-doped famatinite was successfully fabricated through mechanical alloying (MA) and hot pressing (HP), and Cu₃Sb_0.94In_0.06S₄ exhibited a ZT of 0.16 at 623 K. Although the improvement in thermoelectric performance by single doping with In was small, by double doping famatinite with Sn and In, the ZT value was expected to increase in this study due to additional lattice scattering and enhanced electrical properties (electrical conductivity and power factor). The Cu₃Sb_1-x-ySn_xIn_yS₄ compounds were prepared by partially introducing Sn and In at the Sb sites via the MA-HP process. A phase/microstructural analysis was conducted, and the thermoelectric properties were examined.

Sn and In double-doped famatinites, Cu₃Sb_1-x-ySn_xIn_yS₄ (x = 0.02, 0.04, 0.08, 0.10, 0.12, and y = 0.06, 0.08, 0.10), were prepared using MA and HP methods to prevent volatilization of the component elements and for homogeneous synthesis. Powders of Cu (purity 99.9%, <45 μm), Sb (purity 99.999%, <150 μm), Sn (purity 99.999%, <35 μm), In (purity 99.99%, <75 μm), and S (purity 99.99%, <75 μm) were used to prepare stoichiometric compositions. The starting powder was loaded into a stainless-steel jar with stainless-steel balls. The jar was evacuated and sealed with Ar gas, and then MA was performed for 12 h at a rotation speed of 350 rpm using a planetary ball mill (Pulverisette5, Fritsch). The synthesized powder was charged into a graphite mold and consolidated using HP at 70 MPa and 623 K for 2 h in vacuum.

The phases present in the MA-HP specimens were analyzed using an X-ray diffractometer (XRD; D8-Advance, Bruker) with Cu Kα radiation. The diffraction peaks were measured over a 2θ (diffraction angle) range of 10° to 90° in steps of 0.02°. The lattice constants were estimated using TOPAS, a Rietveld refinement program. The microstructures of the sintered specimens were observed using a scanning electron microscope (SEM; Quanta400, FEI) in the backscattering electron (BSE) mode. Composition analysis and elemental mapping were conducted using an energy-dispersive spectrometer (EDS; Quantax200, Bruker), where the energy levels of the elements employed were the Cu K series, Sb L series, Sn L series, In L series, and S K series. The charge transport parameters were measured using the van der Pauw method (Keithley 7065).

Rectangular bars of 3.5 × 3.5 × 10 mm³ were processed from sintered bodies for Seebeck coefficient and electrical conductivity measurements, which were performed using the DC four-probe method with the ZEM-3 system (Advance Riko) in a He atmosphere. To determine thermal conductivity, the thermal diffusivity, specific heat, and density were measured in a vacuum using the laser flash method with TC-9000H equipment (Advance Riko) on specimens cut into disks with a diameter of 10 mm and a thickness of 1 mm. Finally, the power factor and ZT values were calculated from the electrical conductivity, Seebeck coefficient, and thermal conductivity.

Figure 1 shows the XRD results for Cu₃Sb_1-x-ySn_xIn_yS₄ produced by the MA-HP process. All the samples contained the tetragonal famatinite phase (ICDD PDF# 00-035-0581); however, a secondary phase of CuInS₂ was identified in specimens with y = 0.10. The calculated lattice constants and relative densities are listed in Table 1. In previous studies the lattice constants of undoped Cu₃SbS₄ have been reported to be 0.5384 nm (a-axis) and 1.0749 nm (c-axis) [19]; however, in this study, Sn/In double doping increased the a-axis to 0.5387–0.5389 nm and changed the c-axis to 1.0744–1.0752 nm. Pi et al [19] reported that by doping Sn in Cu₃Sb_1-xSn_xS₄ (x = 0.02–0.10), the lattice constant of the a-axis slightly decreased from 0.5384 nm to 0.5377–0.5379 nm, but the c-axis increased significantly from 1.0749 nm to 1.0775–1.0778 nm. In our previous study [20], when In was doped into Cu₃Sb_1-yIn_yS₄ (y = 0.02–0.08), the a-axis was slightly reduced from 0.5384 nm to 0.5377–0.5380 nm, but the c-axis was greatly increased from 1.0749 nm to 1.0781–1.0791 nm.

However, in the present study, double doping with both Sn and In at the Sb sites in famatinite yielded changes in the lattice constants unlike those produced by single doping with Sn or In. The reason for this is still uncertain, but it is presumed that the tetragonality of famatinite was changed by the Sn/In double doping. The effect of double doping on the lattice constants of famatinite has not previously been reported.

Figure 2 shows the BSE-SEM images with elemental line scans and maps of Cu₃Sb_0.86Sn_0.08In_0.06S₄ and Cu₃Sb_0.78Sn_0.12In_0.10S₄. As confirmed by the XRD results in Fig 1, Cu₃Sb_0.86Sn_0.08In_0.06S₄ (Fig 2(a)) appeared as a single famatinite phase, and all elements were uniformly distributed. However, Cu₃Sb_0.78Sn_0.12In_0.10S₄ (Fig 2(b)) contained a secondary phase with a higher In content and lower Sb content than the matrix (famatinite), and from the XRD phase analysis shown in Fig 1, this was estimated to be the CuInS₂ phase. All specimens exhibited densely sintered bodies with relative densities of 96.1%–97.8% (Table 1) of the theoretical density (4.64 g cm^-3) of Cu₃SbS₄ [21]. Changes in the theoretical density arising from the doping contents of Sn and In were not considered.

The charge transport parameters of Cu₃Sb_1-x-ySn_xIn_yS₄ are presented in Table 1. All the samples had positive Hall coefficients, indicating p-type semiconductors whose major charge carriers were holes (not indicated in Table 1). As the Sn and In contents increased from Cu₃Sb_0.90Sn_0.04In_0.06S₄ to Cu₃Sb_0.80Sn_0.12In_0.08S₄, the carrier (hole) concentration increased from 1.12 × 10¹⁹ to 1.72 × 10¹⁹cm^-3, and the mobility also increased from 111 to 307 cm²V-1s-1. Chen et al. [9] reported that the carrier concentration of Cu₃SbS₄ was 3.2 × 10¹⁷cm^-3 and increased to 7.8 × 10¹⁹–1.0 × 10²¹cm^-3 for Cu₃Sb_1-xSn_xS₄ (x = 0.01–0.15). Goto et al. [22] reported that with increasing Sn content for Cu₃Sb_1-xSn_xS₄ (x = 0– 0.15), the carrier concentration increased from 4.2 × 10¹⁸ to 7.8 × 10²⁰ cm^-3. Shen et al. [15] reported that as the Sn content of Cu₃Sb_0.94-xBi_0.06Sn_xS₄ (x = 0.01–0.09) increased, the carrier concentration increased to 9.3 × 10¹⁹–7.4 × 10²⁰ cm^-3. Tanishita et al. [18] reported that the carrier concentration of Cu₃SbS₄ was 1.5 × 10¹⁷ cm^-3, and that of Cu₃Sb_0.85-xGe_xP_0.15S₄ (x = 0.05–0.4) increased to 0.5 × 10²¹–2.3 × 10²¹ cm^-3 as the Ge content increased.

Figure 3 shows the electrical conductivity of Cu₃Sb_1-x-ySn_xIn_yS₄. The electrical conductivity of a heavily doped semiconductor is defined as σ = enμ, where e is the electronic charge, n is the carrier concentration, and μ is the carrier mobility [23]. In this study, all the samples exhibited degenerate semiconductor behavior, where the electrical conductivity slightly decreased as the temperature increased. When the In content was constant at y = 0.06, as the Sn content increased from x = 0.04 to 0.08, the electrical conductivity increased from 1.2 × 10⁴–9.0 × 10³ S m^-1 to 2.8 × 10⁴–1.9 × 10⁴ S m^-1 at 323–623 K. These values are significantly higher than those (2.6 × 10²–1.1 × 10³ S m^-1 at 323–623 K) reported in our previous study [20]. As the Sn and In contents further increased, as can be seen in Table 1, the carrier concentration and mobility increased, and thus the electrical conductivity increased, with Cu₃Sb_0.80Sn_0.12In_0.08S₄ having the highest electrical conductivity of 5.2 × 10⁴–3.2 × 10⁴ S m^-1 at 323–623 K. Chen et al. [9] reported that the electrical conductivity of undoped Cu₃SbS₄ was 89 S m^-1 at 300 K, and that as the Sn content of Cu₃Sb_1−xSn_xS₄ (x = 0.01–0.15) increased, the electrical conductivity increased to 6.7 × 10³–1.3 × 105 S m^-1. Pi et al. [19] reported that the electrical conductivity of Cu₃SbS₄ was 57–490 S m^-1 at 323–623 K, but that as the Sn content of Cu₃Sb_1−xSn_xS₄ (x = 0.02– 0.10) increased, it increased to 4.0 × 10³–3.7 × 10⁴ S m^-1 at 323 K and to 4.2 × 10³–2.3 × 10⁴ S m^-1 at 623 K. Shen et al. [15] reported that as the Bi and Sn contents of Cu₃Sb_1-x-yBi_xSn_yS₄ (x = 0.02–0.14 and y = 0.01–0.09) increased, the electrical conductivity increased, achieving a maximum electrical conductivity of 9.4 × 10⁴ S m^-1 at 300 K for Cu₃Sb_0.85Bi_0.06Sn_0.09S₄. In the present study, the electrical conductivity had lower values compared to the reported values, in accordance with the lower carrier concentrations.

Figure 4 shows the Seebeck coefficients of Cu₃Sb_1-x-ySn_xIn_yS₄. All the samples exhibited positive Seebeck coefficients, similar to the signs of the Hall coefficient, confirming that they were p-type semiconductors. The Seebeck coefficient is simply expressed as α = Cm*Tn^-2/3 using the Pisarenko relationship [24], where C is a constant and m* is the effective carrier mass. The Seebeck coefficient increased as the temperature increased, in contrast to the temperature dependence of the electrical conductivity [25], and decreased as the Sn and In contents increased, because it has an inverse relationship with carrier concentration. The value of the Seebeck coefficient was 171–258 μV K^-1 at 323–623 K for Cu₃Sb_0.90Sn_0.04In_0.06S₄, but decreased to 124–201 μV K^-1 for Cu₃Sb_0.86Sn_0.08In_0.06S₄ and to 94–164 μV K^-1 for Cu₃Sb_0.80Sn_0.12In_0.08S₄.

Chen et al. [9] reported that the Seebeck coefficient of Cu₃SbS₄ was 670 μV K^-1 at 300 K, and for Cu₃Sb_1-xSn_xS₄ (x = 0.01–0.15), the carrier concentration increased with increasing Sn content, resulting in a decreased Seebeck coefficient of 264–72 μV K^-1. Pi et al. [19] reported that the Seebeck coefficient of Cu₃SbS₄ was 575–539 μV K^-1 at 323–623 K, and for Cu₃Sb_1-xSn_xS₄ (x = 0.02–0.10), it decreased from 260–338 μV K^-1 for Cu₃Sb_0.98Sn_0.02S₄ to 118–197 μV K^-1 for Cu₃Sb_0.90Sn_0.10S₄. In our previous study [20], the Seebeck coefficient of Cu₃Sb_1-yIn_yS₄ (y = 0.02–0.08) decreased as the In content increased: 499–488 μV K^-1 for Cu₃Sb_0.98In_0.02S₄ and 336–404 μV K^-1 for Cu₃Sb_0.92In_0.08S₄ at 323–623 K. Compared with the previous study, a significant decrease in the Seebeck coefficient was confirmed to have resulted from the double doping with Sn and In. Shen et al. [15] obtained lower Seebeck coefficients of 352–95 μV K^-1 at 300 K for Bi/Sn-double-doped Cu₃Sb_0.94-xBi_0.06Sn_xS₄ (x = 0.01–0.09).

Figure 5 shows the power factors of Cu₃Sb_1-x-ySn_xIn_yS₄. The power factor is expressed as PF = α²σ. Optimized carrier concentration is required to achieve a high power factor because it is affected by both the Seebeck coefficient, which is inversely proportional to the carrier concentration, and the electrical conductivity, which is proportional to the carrier concentration. The power factor of Cu₃SbS₄ reported by Pi et al. [19] was 0.02–0.14 mW m^-1 K^-2 at 323–623 K, and in our previous study [20] on Cu₃Sb_1-yIn_yS₄ (y = 0.02–0.08), Cu₃Sb_0.94In_0.06S₄ exhibited the highest power factor of 0.03–0.18 mW m^-1 K^-2 at 323–623 K, implying that the improvement in the power factor caused by In single doping was not significant.

However, in this study, the power factor was remarkably increased by the Sn and In double doping; it increased to 0.44–0.78 mW m^-1 K^-2 at 323–623 K for Cu₃Sb_0.86Sn_0.08In_0.06S₄, with Cu₃Sb_0.80Sn_0.12In_0.08S₄ exhibiting the highest power factor of 0.46–0.87 mW m^-1 K^-2. Pi et al. [19] reported that for Cu₃Sb_1-xSn_xS₄ (x = 0.02–0.10), Cu₃Sb_0.92Sn_0.08S₄ had the highest power factor, 0.94 mW m^-1 K^-2 at 623 K, as a result of the trade-off relationship between electrical conductivity and Seebeck coefficient. Shen et al. [15] reported that for Cu₃Sb_0.94-xBi_0.06Sn_xS₄ (x = 0.01–0.09), Cu₃Sb_0.85Bi_0.06Sn_0.09S₄ had the highest power factor of 1.40 mW m^-1 K^-2 at 623 K, a result of carrier concentration optimization caused by the Bi and Sn double doping. In the present study, the Seebeck coefficient of Cu₃Sb_0.80Sn_0.12In_0.08S₄ was similar to that of Cu₃Sb_0.85Bi_0.06Sn_0.09S₄ reported by Shen et al. [15], but its electrical conductivity was approximately half as great; thus, the power factor led to a lower value.

Figure 6 shows the thermal conductivity of Cu₃Sb_1-x-ySn_xIn_yS₄. In the temperature range 323–623 K, the thermal conductivity decreased as the temperature increased, and a bipolar effect did not occur. In our previous study [20], the thermal conductivity of Cu₃Sb_0.94In_0.06S₄ was 1.20–0.68 W m^-1 K^-1 at 323–623 K, but in the present study, the thermal conductivity increased with increasing Sn and In contents: 1.37–0.90 W m^-1 K^-1 for Cu₃Sb_0.86Sn_0.08In_0.06S₄ and 1.63–1.17 W m^-1 K^-1 for Cu₃Sb_0.80Sn_0.12In_0.08S₄ at 323–623 K. Pi et al. [19] reported that the thermal conductivity of Cu₃SbS₄ was 1.14–0.62 W m^-1 K^-1 in the temperature range 323–623 K and increased when Sn was doped, so that for Cu₃Sb_1-xSn_xS₄ (x = 0.02–0.10), the thermal conductivity was 1.26–0.72 W m^-1 K^-1 for Cu₃Sb_0.98Sn_0.02S₄ and 1.32–0.88 W m^-1 K^-1 for Cu₃Sb_0.90Sn_0.10S₄. Shen et al. [15] reported that the thermal conductivity of Cu₃Sb_0.94-xBi_0.06Sn_xS₄ (x = 0.01–0.09) increased to 1.74–2.28 W m^-1 K^-1 at room temperature as the Sn content increased from x = 0.01 to 0.09 as a result of the increase in electronic thermal conductivity, and for samples with Sn content lower than x = 0.05, the thermal conductivity was lower than that of Cu₃Sb_0.94Bi_0.06S₄. In this study, it was also confirmed that the electronic thermal conductivity contributed significantly to the increase in thermal conductivity due to Sn/In double doping.

Figure 7 shows (a) the electronic thermal conductivity (κ_E) of the carriers and (b) the lattice thermal conductivity (κ_L) of the phonons for Cu₃Sb_1-x-ySn_xIn_yS₄. The Wiedemann-Franz law (κ_E = LσT) was employed to estimate the electronic thermal conductivity, where L is the Lorenz number [26]. The Lorenz number [10^-8 V² K^-2] can be obtained using the relation L = 1.5 + exp(–|α|/116) [27,28], and the calculated values are listed in Table 1. As illustrated in Fig 7 (a), the electronic thermal conductivity was not significantly affected by temperature, but it increased as the electrical conductivity and Lorenz number increased with increasing Sn and In contents. Thus, over the temperature range 323–623 K, the electronic thermal conductivity of Cu₃Sb_0.90Sn_0.04In_0.06S₄ increased from 0.06 to 0.09 W m^-1 K^-1, that of Cu₃Sb_0.86Sn_0.08In_0.06S₄ from 0.17 to 0.20 W m^-1 K^-1, and that of Cu₃Sb_0.80Sn_0.12In_0.08S₄ from 0.33 to 0.34 W m^-1 K^-1. Pi et al. [19] reported that, over the same temperature range, the electronic thermal conductivity of Cu₃SbS₄ increased from 0.27 × 10^-3 to 0.40 × 10-2 W m^-1 K^-1, and that of Cu₃Sb_1-xSn_xS₄ (x = 0.02–0.10) increased from 0.27 × 10^-3–0.40 × 10-2 W m^-1 K^-1 at 323 K to 0.04–0.23 W m^-1 K^-1 at 623 K. In our previous study [20], the electronic thermal conductivity of Cu₃Sb_1-yIn_yS₄ (y = 0.02–0.08) was 0.42 × 10^-3–0.13 × 10-2 W m^-1 K^-1 at 323 K and 0.58 × 10-2–0.10 W m^-1 K^-1 at 623 K. The electronic thermal conductivity was higher than that of undoped Cu₃SbS₄ because of the increase in carrier concentration caused by the dopants.

Figure 7 (b) shows the lattice thermal conductivity of Cu₃Sb_1-x-ySn_xIn_yS₄, which is the value obtained by subtracting the electronic thermal conductivity from the total thermal conductivity. The lattice thermal conductivity decreased as the temperature increased, resulting from the Umklapp scattering caused by acoustic phonons (κ_L is in fact directly proportional to T-1) [29]. In our previous study [20], the lattice thermal conductivity of Cu₃Sb_0.94In_0.06S₄ was 1.20–0.67 W m^-1 K^-1 at 323–623 K, which was higher than the values 1.15–0.62 Wm^-1K^-1 for Cu₃SbS₄ reported by Pi et al. [19]. In the present study, Cu₃Sb_0.90Sn_0.04In_0.06S₄ showed the lowest lattice thermal conductivity of 1.19–0.70 W m^-1 K^-1 at 323–623 K. When the In content was y = 0.06, there was no significant change in the lattice thermal conductivity with increasing Sn content. However, when the Sn and In contents were further increased, the lattice thermal conductivity slightly increased, and for Cu₃Sb_0.80Sn_0.12In_0.08S₄, the lattice thermal conductivity was 1.30–0.83 W m^-1 K^-1 at 323–623 K. This is thought to be due to the formation of a secondary phase, such as CuInS₂, beyond the solubility limit of Sn and In at the Sb site of famatinite. Pi et al. [19] reported that the lattice thermal conductivity of Cu₃Sb_1-xSn_xS₄ (x = 0.02–0.10) was 1.24–1.10 W m^-1 K^-1 at 323 K and 0.68–0.64 W m^-1 K^-1 at 623 K. Shen et al. [15] reported that the lattice thermal conductivity of Cu₃Sb_0.94-xBi_0.06Sn_xS₄ (x = 0.01–0.09) was 1.69–1.83 W m^-1 K^-1 at 300 K and decreased to 0.75 W m^-1 K^-1 as the temperature increased to 623 K.

Values of the dimensionless figure of merit ZT of Cu₃Sb_1-x-ySn_xIn_yS₄ are presented in Fig 8. In the temperature range 323–623 K, the ZT values increased with increasing temperature. This was because intrinsic conduction and bipolar effects did not occur in the measurement temperature range [30]. Compared with the ZT values of 0.14 at 623 K for Cu₃SbS₄ [19] and 0.16 at 623 K for Cu₃Sb_0.94In_0.06S₄ [20], Cu₃Sb_0.86Sn_0.08In_0.06S₄ exhibited an improved ZT value of 0.53 at 623 K, resulting from the significantly increased power factor caused by double-doping with Sn and In. Although the highest power factor was achieved in Cu₃Sb_0.86Sn_0.12In_0.08S₄, this sample exhibited the relatively low ZT value of 0.45 at 623 K, because of its high thermal conductivity. Goto et al. [12] fabricated Cu₃SbS₄ and Cu₃Sb_0.85Sn_0.15S₄ via anvil pressing of powders synthesized by the melting-annealing method and obtained ZT values at 623 K of 0.03 and 0.1, respectively. Chen et al. [31] reported that Cu₃SbS₄ and Cu₃Sb_0.95Sn_0.05S₄ produced by the MA-SPS (spark plasma sintering) process exhibited ZT values at 623 K of 0.1 and 0.72, respectively. Shen et al. [15] prepared Cu₃Sb_0.89Bi_0.06Sn_0.05S₄ by the MA-SPS process and obtained a ZT of 0.76 at 623 K. Tanishita et al. [18] reported that Cu₃Sb_0.75Ge_0.1P_0.15S₄ produced by HP after synthesis using melting-annealing exhibited a ZT of 0.67 at 623 K, this value being due to an increase in carrier concentration from the Ge doping and a decrease in lattice thermal conductivity from the P doping.

Cu₃Sb_1-x-ySn_xIn_yS₄ (x = 0.02–0.12 and y = 0.06–0.10) famatinites double-doped with Sn and In were synthesized and sintered by MA-HP. According to the XRD and SEM analyses, a single famatinite phase was obtained in all samples, except for the sample with y = 0.10, in which a secondary phase of CuInS₂ was found. By double-doping with Sn and In at the Sb sites of famatinite, the lattice constant of the a-axis was slightly increased from 0.5384 nm to 0.5387–0.5389 nm, while the lattice constant of the c-axis changed from 1.0749 nm to 1.0744–1.0752 nm. Sn/In-double-doped famatinites were confirmed to be p-type semiconductors by their positive Hall and Seebeck coefficients. The temperature dependence of the electrical conductivity and Seebeck coefficient followed degenerate semiconductor behavior. The carrier concentration increased with increasing Sn and In contents, and thus, the electrical conductivity increased, and the Seebeck coefficient decreased. Cu₃Sb_0.80Sn_0.12In_0.08S₄ exhibited the highest power factor of 0.87 mW m^-1 K^-2 at 623 K. However, Cu₃Sb_0.86Sn_0.08In_0.06S₄ reached the highest ZT value of 0.53 at 623 K because of a high power factor of 0.78 mW m^-1 K^-2 and low thermal conductivity of 0.90 W m^-1 K^-1.