Keywords

Introduction

With increasing energy consumption and intensifying environmental problems, the environment-friendly thermoelectric technology attracts wide attention as one of the alternative ways to tackle the above mentioned problems. Thermoelectric conversion efficiency depends on the Carnot efficiency and dimensionless figure of merit, the ZT value, defined as ZT = α2T/ρκ, where α, T, ρ and κ are the Seebeck coefficient, absolute temperature, electrical resistivity and thermal conductivity, respectively. The performance of thermoelectric materials has made great progress as the advancement of the theories of thermoelectric materials and synthetic preparation [1, 2]. However, the complicated and time- and energy-consuming fabrication process, high cost, toxic elements and the use of rare elements limit large-scale applications of thermoelectric materials.

Tetrahedrite (Cu12Sb4S13) thermoelectric materials consist mainly of earth-abundant and environment-friendly elements copper and sulfur, moreover, natural mineral tetrahedrites can be used as direct source of the thermoelectric materials [3]. Researches on the low-cost, environment-friendly, high performance tetrahedrite thermoelectric materials are expected to create new opportunities of thermoelectric energy generation applications. Till now researches have been mainly focused on the enhancement of the thermoelectric performance of Cu12Sb4S13 tetrahedrite by substitution Cu with Mn, Fe, Co, Ni, Zn and other elements. The highest ZT values reported for the Zn, Ni, Ni and Zn, Fe, Mn and Te doped tetrahedrites are 0.9 [3], 0.7 [4], 1.03 [5], 0.8 [3], 1.13 [6] and 0.92 [7], respectively. However, the preparation process of tetrahedrites is time and energy consuming. The melting time is about three days and the annealing time is two weeks [3]. In our previous work [8], a process was developed to prepare undoped tetrahedrite Cu12Sb4S13 with good thermoelectric performance in a considerably reduced melting and annealing time. In this work, the effect of annealing time on the microstructure and thermoelectric properties of Mn doped tetrahedrite Cu11.5Mn0.5Sb4S13 was studied, hoping to shorten the fabrication time of bulk doped tetrahedrites. Results show that good thermoelectric performance doped tetrahedrite can also be obtained by a fast preparation process without annealing.

Experimental Procedures

Ten gram Cu11.5Mn0.5Sb4S13tetrahedrite ingots were prepared by a direct melting method. The starting materials, Cu shots (99.999%, Alfa-Aesar), Mn pieces (99.95%, Alfa-Aesar), Sb shot (99.9999%, Alfa-Aesar) and S pieces (99.999%, Alfa-Aesar), were weighed out in stoichiometric ratios and loaded into silica ampoules. The ampoules were evacuated to a pressure of 0.1 Pa and then sealed. The sealed ampoules were suspended in a vertical tube furnace, and heated to 923 K at a rate of 15 Kmin−1 and held for four hours. Then the furnace was switched off, and the ampoules cooled down to room temperature.

To study the effect of annealing time on the microstructure and thermoelectric properties of Cu11.5Mn0.5Sb4S13 tetrahedrite, the ingots were crushed and ground into powders, and then cold pressed and placed into silica ampoules and a box furnace to anneal at 723 K for one day, 3, 7 and 14 days, respectively. The annealed pellets were ground into powders, loaded into graphite dies with a diameter of 15 mm, and then hot pressed at 743 K and 62 MPa for 40 min. For comparison, a disc was also hot pressed under the same condition with the un-annealed powders. The sintered disks were cut for characterization of thermoelectric properties.

The samples were examined by powder X-ray diffraction (XRD, Rigaku Rint 2000, Cu Kα). The microstructure and phase composition of the ingot and sintered samples were also examined by SEM and EDS. The densities of the sintered samples were measured using the Archimedean method. The Seebeck coefficient and electrical resistivity were measured using the standard 4-probe method (LINSEIS LSR-3) in helium atmosphere. The thermal conductivity was calculated from the measured thermal diffusivity D, specific heat C p, and density d according to the relationship κ = DC p d. Thermal diffusivity of the samples were measured by a laser flash method (LINSEIS LFA 1000) and the specific heat C p of the samples was calculated using the Dulong-Petit law. The measurement errors for the electrical resistivity, the thermal conductivity and the Seebeck coefficient are ±5–7, ±5–7 and ±5%, respectively.

Results and Discussion

X-ray Diffraction Patterns. For the Cu11.5Mn0.5Sb4S13tetrahedrite powder samples are shown Fig. 1 The major peaks of the un-annealed sample match the Cu12Sb4S13 phase, and a few peaks match Cu3SbS4, implying that the tetrahedrite phase forms in the melt during cooling, and the ingots consist of principal Cu11.5Mn0.5Sb4S13 tetrahedrite and secondary Cu3SbS4. The annealed samples also contain secondary phase Cu3SbS4. Wang [8, 9] and Chetty [10] and Suekuni [11] also observed the secondary phase Cu3SbS4 in the synthesized tetrahedrite materials. It seems long time annealing cannot eliminate the second phase Cu3SbS4 in tetrahedrites.

Fig. 1
figure 1

X-ray diffraction patterns for the Cu11.5Mn0.5Sb4S13 tetrahedrite powder samples

Full size image

  • SEM Back Scattering Images. of the Cu11.5Mn0.5Sb4S13 tetrahedrite ingot and sintered samples with powders before and after annealing for 1 day and 14 days are shown Fig. 2. It can be seen that the ingot consists of matrix and secondary phases in light grey, dark grey and white. EDS show they are Cu11.5Mn0.5Sb4S13, Cu3SbS4, Cu2S and CuSbS2, respectively. The sintered samples consist of matrix Cu11.5Mn0.5Sb4S13 and secondary Cu3SbS4 and CuSbS2. The impurity phase CuSbS2 is not found in the X-ray diffraction patterns (Fig. 1). The reason is likely that the content of impurity phase CuSbS2 is too less to be detected, whose diffraction peak is obscured in comparison to the peaks of other phases. The impurity phase Cu2S in the ingot disappears in the sintered samples, possibly due to the Cu2S phase reacts further into other phase under high temperature and high pressure during the sintering process. In addition, the micro-porous area fraction of the sample sintered with the one day annealed powders is significantly less than that of the samples sintered with the powders before and after annealed for 14 days. As is known, the longer the annealing time the better the crystallization quality is for the annealed powders. And the better crystallization quality the poorer sintering property is for the annealed powders. This maybe the reason why the sample sintered with the one day annealed powders has less micro-porous in comparison to other samples.

    Fig. 2
    figure 2

    Back-scattering images of the Cu11.5Mn0.5Sb4S13 Tetrahedrite ingot (a) and samples sintered with powder before (b) and after annealing for 1 day (c) and 14 days (d)

    Full size image

  • Electrical Conductivity. As a function of temperature for the Cu11.5Mn0.5Sb4S13 samples sintered with powders before and after annealing is shown Fig. 3. The electrical resistivity of all the samples of Cu11.5Mn0.5Sb4S13 increases with temperature. The tendency of the electrical resistivity change with temperature agrees with that reported by Wang [9] and Chetty [10]. The electrical resistivity of the samples decreases with increase of annealing time, and then increases with increasing annealing time when the annealing time exceeds one day. The electrical resistivity of the un-annealed sample is about 16.335–20.932 × 10−6 Ω m in whole temperature range. The electrical resistivity of the one day annealed sample drops to 14.145–18.608 × 10−6 Ω m, which is the lowest among the samples. And the electrical resistivity goes up to 15.542–19.548 × 10−6, 16.597–21.228 × 10−6 and 16.928–21.603 × 10−6 Ω m for the 3, 7 and 14 days annealed samples, respectively. The trend of electrical resistivity change with annealing time is inconsistent with that of Cu12Sb4S13 tetrahedrite [8], which goes down with increasing annealing time. The crystallization quality, homogeneity of the doped Mn (the real doping amount) and density combine to affect the electrical resistivity of the samples. As the annealing time increases, the crystallization quality and the homogeneity of the doped Mn are improved, and the density drops slightly under the same sintering condition. The electrical resistivity of the samples decreases with increasing of crystallization quality, and increases with increasing of real Mn doping amount and dropping of density. That’s why the electrical resistivity of the samples drops at first, and then increase with increasing annealing time.

    Fig. 3
    figure 3

    Temperature dependence of the electrical resistivity for theCu11.5Mn0.5Sb4S13 tetrahedrite samples sintered with powders before and after annealing

    Full size image

  • Seebeck Coefficient. As a function of temperature for the Cu11.5Mn0.5Sb4S13 tetrahedrite samples sintered with powders before and after annealing is shown in Fig. 4. The Seebeck cofficient of all the Cu11.5Mn0.5Sb4S13 tetrahedrite samples is positive and rises with temperature, which indicates the majority carriers are hole and the tetrahedrites exhibit p-type conduction. Although the seebeck coefficients of the sample is different, and the difference is small within the measurement error, which means that annealing time has a negligible effect on the seebeck coefficient. In the temperature range of 300–700 K, the seebeck coefficient of Cu11.5Mn0.5Sb4S13 tetrahedrite samples is between 86 and 150 μV/K, close to that of Cu11.5Mn0.5Sb4S13 tetrahedrites reported by Wang [9] and Chetty [10].

    Fig. 4
    figure 4

    Temperature dependence of the Seebeck coefficient for the Cu11.5Mn0.5Sb4S13 tetrahedrite samples sintered with powders before and after annealing

    Full size image

  • Thermal Conductivity. As function of temperature for the Cu11.5Mn0.5Sb4S13 tetrahedrite samples sintered with powders before and after annealing is shown in Fig. 5. The thermal conductivity of the samples rises with increasing temperature, and then drops slightly over the temperature range of 650–700 K, similar to those reported before [6, 9, 10]. It was attributed to a low lattice thermal conductivity resulting from complex crystal structure and strong lattice anharmonicity and dominant electronic thermal conductivity which increases with temperature according to Wiedemann-Franz law [7]. Among all the samples, the thermal conductivity of the one day annealed sample is the highest, 1.12–1.22 Wm−1 K−1, over the temperature range of 400–700 K, while that of the un-annealed sample is the lowest, 0.96–1.00 Wm−1 K−1. The thermal conductivity of the more than one annealed samples is very close, the difference of the thermal conductivity is within the measurement errors, indicating that prolonged annealing time has negligible effect on the thermal conductivity of Cu11.5Mn0.5Sb4S13 tetrahedrite.

    Fig. 5
    figure 5

    Temperature dependence of the thermal conductivity for the Cu11.5Mn0.5Sb4S13 tetrahedrite samples sintered with powders before and after annealing

    Full size image

  • Dimensionless Figure of Merit (ZT). As function of temperature for all the Cu11.5Mn0.5Sb4S13 tetrahedrite samples is calculated based on the above data and is presented in Fig. 6. It can be seen that the samples with different annealing time have similar ZT value in the temperature range of ~700 K, and the ZT value of the samples exceeds 0.6 at above 650 K. Theun-annealed sample has the largest ZT value of 0.74, which is very close to that of Cu11.6Mn0.4Sb4S13 tetrahedrite reported by Chetty [10]. This means it is unnecessary to prolong annealing to improve thermoelectric properties of Cu11.5Mn0.5Sb4S13tetrahedrite. Cobalt doped tetrahedrite Cu11.5Co0.5Sb4S13 was also prepared by the process of melting and hot press, without annealing. The ZT value of the tetrahedrite is ~0.7. It seems that preparation time of high performance doped tetrahedrite thermoelectric materials can be cut considerably.

    Fig. 6
    figure 6

    Temperature dependence of the figure of merit for the Cu11.5Mn0.5Sb4S13 tetrahedrite and Cu11.5Co0.5Sb4S13 tetrahedrite samples

    Full size image

Summary

Manganese doped tetrahedrite Cu11.5Mn0.5Sb4S13 was fabricated by fast melting and annealing method. The Cu11.5Mn0.5Sb4S13tetrahedrite phase forms in the melt during cooling. The ingot consists of principal phase of Cu11.5Mn0.5Sb4S13 and secondary phases of Cu3SbS4, Cu2S and CuSbS2. Annealing has negligible effect on eliminating the Cu3SbS4 and CuSbS2 phases in Cu11.5Mn0.5Sb4S13 tetrahedrite. Sintering can eliminate Cu2S phase. Long time annealing has slight effect on electrical resistivity, and negligible effect on Seebeck coefficient and thermal conductivity of Cu11.5Mn0.5Sb4S13 tetrahedrite. All the Cu11.5Mn0.5Sb4S13 samples have ZT in excess of 0.6 at above 650 K and the maximum ZT value obtained in this study is 0.74 for the un-annealed sample. Cobalt doped tetrahedrite Cu11.5Co0.5Sb4S13 fabricated by the un-annealed process also obtained ZT value of ~0.7. Based on our experimental results, the time for preparing doped tetrahedrites can be cut considerably.