Properties of Mechanochemically Synthesized Famatinite Cu 3 SbS 4 Nanocrystals

: In this study, we report the optoelectric and thermoelectric properties of famatinite Cu 3 SbS 4 that was mechanochemically synthesized in a planetary mill from powder elements for 120 min in an inert atmosphere. The tetragonal famatinite Cu 3 SbS 4 was nanocrystalline with a crystallite size of 14 nm, as endorsed by Rietveld reﬁnement. High-resolution transmission electron microscopy showed several crystallites in the range of 20–50 nm. Raman spectroscopy proved the purity of the synthesized famatinite Cu 3 SbS 4 and chemical-state characterization performed by X-ray photoelectron spectroscopy conﬁrmed that the prepared sample was pure. The Cu 1+ , Sb 5+ , and S 2 − oxidation states in Cu 3 SbS 4 sample were approved. The morphology characterization showed homogeneity of the prepared sample. The photoresponse of Cu 3 SbS 4 was conﬁrmed from I–V measurements in the dark and under illumination. The photocurrent increase reached 20% compared to the current in the dark at a voltage of 5 V. The achieved results conﬁrm that synthesized famatinite Cu 3 SbS 4 can be applied as a suitable absorbent material in solar cells. The performed thermoelectric measurements revealed a ﬁgure of merit ZT of 0.05 at 600 K.


Introduction
Ternary semiconductor nanostructures have attracted considerable interest in recent years, owing to their potential applications in the field of optoelectronic devices, photosensitivity, and thermoelectrics [1,2]. Ternary copper-antimony-sulphide materials have gained favor in solar-cell investigations as alternative absorbers because of the fact that they are composed of low-toxicity and Earth-abundant elements [3,4]. Their thermoelectric properties are also interesting [5,6].
Famatinite Cu 3 SbS 4 belongs the group of semiconductors that have been extensively investigated recently. Cu 3 SbS 4 is a promising p-type semiconductor with a direct band gap of 0.9-1.8 eV, depending on the crystal structure, and a high absorption coefficient of 10 4 -10 5 cm −1 [7]. This material can be used in several applications, namely for solar photovoltaics [8] and as an advanced electrode for supercapacitor [9], as well as in thermoelectric devices [10].
There are many synthetic approaches for famatinite Cu 3 SbS 4 preparation: from compounds [11], from compounds and elements [12], or only from elements [13][14][15][16][17][18]. Flower-like Micro 2023, 3 459 Cu 3 SbS 4 particles were prepared easily and rapidly from mixtures of CuCl 2 , SbCl 3 , and thiourea in ethylene glycol by microwave radiation [11]. One-pot synthesis of Cu 3 SbS 4 nanocrystals by a surfactant-assisted solvothermal method was reported in [12]. Lee et al. [13] synthesized Cu 3 SbS 4 by mechanical alloying from pure elements copper, antimony, and sulfur, followed by consolidation using hot pressing. Ge-doped Cu 3 SbS 4 was synthesized from pure elements by mechanical alloying combined with spark plasma sintering (SPS) by Chen et al. [14]. High thermoelectric performance in Sn-doped Cu 3 SbS 4 prepared from pure elements by a combination of mechanical alloying and spark plasma sintering was also investigated by Chen et al. [15]. Cu 3 SbS 4 as thermoelectric material and the enhancement of its figure of merit (ZT) by chemical substitutions was reported in the paper by Suzumura et al. [16]. Chen et al. prepared Cu 3 Sb 1 − x Sn x S 4 samples from pure elements by mechanical alloying connected with spark plasma sintering in paper [17]. Thermoelectric properties and the thermal stability of Bi-and Sn-doped Cu 3 SbS 4 samples were investigated in the paper by Shen et al. [18].
In this connection, high-energy milling utilizing tools of mechanochemistry for the synthesis of nanomaterials also has a place [19]. Additionally, ball milling offers a timesaving alternative to conventional methods [20] and has been successfully applied in the synthesis of nanomaterials. This environmentally friendly route has been already applied for famatinite Cu 3 SbS 4 prepared by mechanochemical synthesis in the laboratory and industrial mill, as reported by our research group in a previous paper [21]. The present contribution is a study of the optoelectric properties of nanocrystalline famatinite synthesized via mechanochemistry in a planetary mill. Moreover, the Cu 3 SbS 4 powder was consolidated by spark plasma sintering and its suitability for thermoelectrics applications was examined.
Phase evolution was followed by X-ray powder diffraction (XRD). The XRD data were collected by D8 Advance diffractometer (Brucker, Bremen, Germany) with a CuKα radiation source. Data were collected in the range 10-65 • 2θ and 0.03 • step. The XRD data were further analyzed by Rietveld refinement (Fullprof software [22]), and fitted in the space group I-42m for Cu 3 SbS 4 . The XRD line broadening was refined by application of regular Thompson-Cox-Hastings pseudo-Voigt function parameters. The instrumental line broadening was verified by refinement of an LaB 6 standard specimen. The phase identification was performed by utilization of JCPDS PDF [23].
Scanning electron microscopy (SEM), transmission electron microscopy with high-resolution (TEM/HR-TEM) and electron diffraction (ED) techniques were used for the structural and morphological characterization of the sample. A few drops of an acetone dispersion from small amounts of powder sample were deposited on a carbon-coated nickel grid in a case of TEM measurements. The SEM microscope used to perform the morphological analysis was a Hitachi S-4800-FEG with field-emission filament. To carry out the microstructural characterization and analysis composition, a 200 kV JEOL-2100-PLUS microscope (JEOL, Akishima, Japan) with a LaB 6 filament (0.25 nm inter-dot resolution) and equipped with an energy-dispersive X-ray detector (EDX, Oxford Instrument, Nanolab Technologies Inc., Milpitas, CA, USA) was used. High-resolution image analysis and fast Fourier transform (FFT) were performed with a Gatan Digital Micrograph (Gatan Inc., Pleasanton, CA, USA) and JEM software (version number 3.3806U2009). The INCA program was used to quantify the chemical composition of the sample. The X-ray detector coupled to the microscope have internal standards that allow semi-quantitative analysis, showing an elemental ratio percentage.
Micro-Raman spectra were measured using a confocal Raman microscope (Spectroscopy & Imaging, Warstein, Germany) and a focused Ar laser beam with a wavelength of 514 nm.
Elemental compositions of the samples were evaluated by X-ray photoelectron spectroscopy (XPS; Kratos Axis Supra, Manchester, UK) with monochromatized AlKα radiation (hυ = 1486.69 eV) in the range of 0-1200 eV. The scanning was performed at an emission current of 4 mA using hybrid-lens mode. XPS survey and core-level spectra were recorded using scanning steps of 1.0 and 0.1 eV, respectively. Fitting and peak deconvolution were carried out using the Casa XPS software (Version 2.3.17, Casa Software Ltd., Teignmouth, UK). Prior to the data analysis, the C1s spectrum was calibrated at 284.8 eV. Low-resolution wide-survey spectrum and high-resolution spectra for Cu, Sb, and S were performed. The high-resolution spectra were deconvoluted using a Shirley-type background.
The measurement of zeta-potential (ZP) was carried out using a Zetasizer Nano ZS (Malvern Instruments Ltd., Malvern, UK). The ZP was obtained from the electrophoretic mobility by the Smoluchowski equation. The ZP was measured in the original dispersion medium and the measurements were repeated three times.
Micro PL spectra were obtained in the Vis-NIR region with the same confocal Raman microscope (Spectroscopy & Imaging, Warstein, Germany). To broaden the spectrum in the visible region, excitation at 488 nm was chosen by retuning the Ar laser.
I-V characteristics were measured on prepared devices using an Agilent semiconductor parameter analyzer (4155C). The devices were measured in the dark and under focused halogen lamp illumination. The illumination intensity of the focused beam reached 600 mW/cm 2 . The measured device was prepared by dropping a solution of nanocrystalline powder in isopropyl alcohol onto an interdigital structure with Au contacts. The connection of the interdigital structure to the socket was realized by means of wires glued with silver paste.
In order to allow thermoelectric characterization, the powder was placed into a graphite die (inner diameter 10 mm) and sintered into pellets about 5 mm thick at 723 K under 50 MPa with heating rate of 323 K/min and holding time of 3 min in a spark plasma sintering furnace model HP D10-SD (FCT Systeme GmbH, Frankenblick, Germany) in vacuum.
The pellets were cut for subsequent analyses by a diamond wire saw. The electrical resistivity and Seebeck coefficient were measured under nitrogen flow in a temperature range between 300 and 620 K via the four-probe method using a custom-made instrument. Thermal conductivity was calculated as a product of thermal diffusivity, heat capacity, and density. The first two properties were measured between 200 and 600 K using the Netzsch LFA 467 instrument utilizing Pyroceram 9606 as a standard for heat capacity. Density was determined from the dimensions and mass of the sample and considered temperature-independent. No diffraction peaks belonging to other phases in the Cu-Sb-S system were identified in the XRD pattern. The crystallite size determined by Rietveld refinement was~14 nm. The estimated microstrain was ε 0 19.8 × 10 −4 and the goodness parameter χ 2 was 3.85. Amorphization and the formation of small crystallites strongly influenced XRD pattern. TEM/ED/EDS results are presented in Figure 1b. The TEM image shows the nanocrystalline character of the sample. The crystallites were agglomerated into large particles, and the ED rings could be indexed in the famatinite tetragonal (T) structure in good agreement with the XRD results. Although the average crystalline domain shown by XRD was 14 nm, the HR-TEM images showed several crystallites in the range of 20-50 nm. The HR micrograph presents a crystal oriented along the [3 3 1] T zone axis and the corresponding Fast Fourier transform (FFT) is inset. A semiquantitative analysis of the sample was carried out by EDS analysis; small areas (only 2 or 3 crystallites) and larger zones (several microns) were measured, finding a homogenous result, presented in the EDX spectrum showing the average of the atomic ratio percentage: 35/14/51 for Cu/Sb/S, which is quite similar to the stoichiometric formula (37.5/12.5/50).

Structural and Microstructural Characterization
corresponding Fast Fourier transform (FFT) is inset. A semiquantitative analysis of the sample was carried out by EDS analysis; small areas (only 2 or 3 crystallites) and larger zones (several microns) were measured, finding a homogenous result, presented in the EDX spectrum showing the average of the atomic ratio percentage: 35/14/51 for Cu/Sb/S, which is quite similar to the stoichiometric formula (37.5/12.5/50).
Micro-Raman spectroscopy with laser excitation of 514 nm was used as an additional tool to analyze the crystallinity and structural phase, and also to distinguish the presence of intermediate phases or impurities. The corresponding micro-Raman spectrum of the synthesized Cu3SbS4 nanoparticles after 120 min of milling is shown in Figure 2. The measured spectrum has well distinguishable peaks at 130, 248, 271, 314, and 340 cm −1 , which are comparable to the published results described in [24][25][26] and the RRUFF database for the Cu3SbS4 phase. The RRUFF reference spectrum of famatinite Cu3SbS4 is included in the inset of Figure 2. The measured micro-Raman spectrum verified the synthesis of phasepure Cu3SbS4. These results are in good agreement with the XRD measurements and analysis. Micro-Raman spectroscopy with laser excitation of 514 nm was used as an additional tool to analyze the crystallinity and structural phase, and also to distinguish the presence of intermediate phases or impurities. The corresponding micro-Raman spectrum of the synthesized Cu 3 SbS 4 nanoparticles after 120 min of milling is shown in Figure 2. The measured spectrum has well distinguishable peaks at 130, 248, 271, 314, and 340 cm −1 , which are comparable to the published results described in [24][25][26] and the RRUFF database for the Cu 3 SbS 4 phase. The RRUFF reference spectrum of famatinite Cu 3 SbS 4 is included in the inset of Figure 2. The measured micro-Raman spectrum verified the synthesis of phase-pure Cu 3 SbS 4 . These results are in good agreement with the XRD measurements and analysis.

X-ray Photoelectron Spectroscopy (XPS)
X-ray photoelectron spectroscopy (XPS) was applied to elucidate the surface composition of the as-synthesized Cu 3 SbS 4 sample. XPS survey and core-level spectra of Cu 3 SbS 4 sample prepared after 120 min of milling are shown in Figure 3. The presence of Cu, Sb, and S as well as C from the reference and O impurity was proven by XPS survey spectrum of Cu 3 SbS 4 sample (Figure 3a). In order to assign the valency state for the Cu2p, Sb3d, and S2p the core-level spectra were also taken (Figure 3b-d). The Cu2p core-level spectrum is shown in Figure 3b. The core-level spectrum of Cu2p was well-defined and demonstrated the doublet of Cu2p 3/2 and Cu2p 1/2 at binding energies of 932.4 and 952.16 eV, respectively, with charge separation ∆E of 19.76 eV which has been adjudged to the Cu 1+ ion. Moreover, there was no Cu2p 3/2 satellite peak situated at about 942 eV which is characteristic of Cu 2+ . The values were consistent with the results reported in paper [11]. Figure 3c is the representative Sb3d core-level spectrum, showing that the Sb3d region had well-separated spin-orbit components observed at 530. 37 eV (Sb3d 5/2 ) and 539.69 eV (Sb3d 3/2 ). The Sb3d doublet had separation of 9.32 eV, which is in accordance with the value for Sb 5+ , as was reported in [11]. Additionally, the Sb3d spectrum of Cu 3 SbS 4 had a small signal at 531.70 eV that corresponded to Sb oxides [27]. The S2p core-level spectrum is shown in Figure 3d, where a well-defined doublet of S2p of the sample was observed at 162.20 eV (S2p 3/2 ) and 163.36 eV (S2p 1/2 ), resulting in a separation ∆E of 1.16 eV, which is consistent with sulphur in the sulphide chemical state S 2− in the Cu 3 SbS 4 . A further peak at a higher binding energy 164.45 eV reflected some oxidation in the Cu 3 SbS 4 sample. However, this observed peak had been previously ascribed to oxidized sulphur species in the form of SO x , which were located on the surface of the Cu 3 SbS 4 sample as was also confirmed by zeta potential value (see below Figure 4). Our results are in accordance with results published in [12]. The formation of Cu 3 SbS 4 was confirmed by higher intense peak values of S2p and Sb3d along with value of Cu2p peak in correlation with our XRD results. The confirmed Cu 1+ , Sb 5+ , and S 2− oxidation states were expected for the Cu 3 SbS 4 famatinite sample.
sentative Sb3d core-level spectrum, showing that the Sb3d region had well-separated spin-orbit components observed at 530. 37 eV (Sb3d5/2) and 539.69 eV (Sb3d3/2). The Sb3d doublet had separation of 9.32 eV, which is in accordance with the value for Sb 5+ , as was reported in [11]. Additionally, the Sb3d spectrum of Cu3SbS4 had a small signal at 531.70 eV that corresponded to Sb oxides [27]. The S2p core-level spectrum is shown in Figure  3d, where a well-defined doublet of S2p of the sample was observed at 162.20 eV (S2p3/2) and 163.36 eV (S2p1/2), resulting in a separation ΔE of 1.16 eV, which is consistent with sulphur in the sulphide chemical state S 2-in the Cu3SbS4. A further peak at a higher binding energy 164.45 eV reflected some oxidation in the Cu3SbS4 sample. However, this observed peak had been previously ascribed to oxidized sulphur species in the form of SOx, which were located on the surface of the Cu3SbS4 sample as was also confirmed by zeta potential value (see below Figure 4). Our results are in accordance with results published in [12]. The formation of Cu3SbS4 was confirmed by higher intense peak values of S2p and Sb3d along with value of Cu2p peak in correlation with our XRD results. The confirmed Cu 1+ , Sb 5+ , and S 2-oxidation states were expected for the Cu3SbS4 famatinite sample.

Zeta Potential
The zeta potential measurement provides appropriate information about the surface properties. The results for the Cu3SbS4 sample are displayed in Figure 4. The ZP values were measured in a pH range from 3 to 10. The balance between positive and negative charge on the surface expressed by the isoelectric point (IEP) was observed at pH 3.13. If

Zeta Potential
The zeta potential measurement provides appropriate information about the surface properties. The results for the Cu 3 SbS 4 sample are displayed in Figure 4. The ZP values were measured in a pH range from 3 to 10. The balance between positive and negative charge on the surface expressed by the isoelectric point (IEP) was observed at pH 3.13. If more alkali was added to the suspension, the particles tended to acquire more negative charge, down to −40.6 mV for pH 9. To the best of our knowledge, the IEP for famatinite has not been reported in the literature. The measured results indicate that the mechanochemically prepared famatinite can be oxidized very slightly on its surface [28,29], as confirmed by XPS analysis (Figure 3d). The slight increase in zeta potential observed above pH 9 may be caused by the compression of the double layer at high ionic strengths [30,31]. The achieved results are in agreement with ones observed for other mechanochemically synthesized sulfides reported in our previous papers [32,33] The chemical analysis was also studied by EDS, giving rise to an atomic percentage relationship for Cu/Sb/S (37/14/49) similar to that found by EDS in the TEM microscope. Elemental mapping was taken in several regions to verify the chemical homogeneity through the sample. A representative result is depicted in Figure 6 and shows that the three elements were homogeneously scattered throughout the sample. The chemical analysis was also studied by EDS, giving rise to an atomic percentage relationship for Cu/Sb/S (37/14/49) similar to that found by EDS in the TEM microscope. Elemental mapping was taken in several regions to verify the chemical homogeneity through the sample. A representative result is depicted in Figure 6 and shows that the three elements were homogeneously scattered throughout the sample.

Micro-Photoluminescence Spectroscopy
Nanocrystalline ternary chalcogenide compounds can emit radiation in a wide range of the spectrum, which is determined by the basic luminescence of the semiconductor and, in the case of nanoparticles, by their dimensions and the technology of preparation. In the literature for Cu 3 SbS 4 , there are ambiguous data reported about the direct 1.3 eV [7], 1.5 eV [12], and1.72 eV [34] and indirect 0.82 eV [12] and 1.2 eV [8] band gaps. Figure 7 shows the micro-PL spectrum of mechanochemically synthesized Cu 3 SbS 4 under excitation with a wavelength of 488 nm. The measured spectrum was broad with a maximum around 542 nm exhibiting various deflections which indicated the involvement of different luminescence centres in the radiative processes. The fitted peaks in the PL spectra presented in Figure 7 had a maximum determined at 542 nm (2.28 eV) and 688 nm (1.8 eV), indicating the quantum confinement effect of nanocrystalline Cu 3 SbS 4 . These results are comparable to absorption measurements in published results [8] where absorption maxima at 547 and 667 nm were measured for Cu 3 SbS 4 nanocrystals. The deflection of the PL spectra with a maximum at 886 nm (1.39 eV) may have been due to the bulk Cu 3 SbS 4 emission. The chemical analysis was also studied by EDS, giving rise to an atomic percentage relationship for Cu/Sb/S (37/14/49) similar to that found by EDS in the TEM microscope. Elemental mapping was taken in several regions to verify the chemical homogeneity through the sample. A representative result is depicted in Figure 6 and shows that the three elements were homogeneously scattered throughout the sample.

Micro-Photoluminescence Spectroscopy
Nanocrystalline ternary chalcogenide compounds can emit radiation in a wide range of the spectrum, which is determined by the basic luminescence of the semiconductor and, in the case of nanoparticles, by their dimensions and the technology of preparation. In the literature for Cu3SbS4, there are ambiguous data reported about the direct 1.3 eV [7], 1.5 eV [12], and1.72 eV [34] and indirect 0.82 eV [12] and 1.2 eV [8] band gaps. Figure 7 shows the micro-PL spectrum of mechanochemically synthesized Cu3SbS4 under excitation with a wavelength of 488 nm. The measured spectrum was broad with a maximum around 542 nm exhibiting various deflections which indicated the involvement of different luminescence centres in the radiative processes. The fitted peaks in the PL spectra presented in

Optoelectric Measurements
For the analysis of optoelectric properties the I-V characteristics of Cu3SbS4 nanoparticles mechanochemically synthesized after 120 min of milling were measured. As shown in Figure 8 the photoresponse of Cu3SbS4 was confirmed from I-V measurements in the dark and under illumination. The linear dependence of measured I-V characteristics corresponded to the creation of a sufficient ohmic contact for the measurements. As can be seen in Figure 8, the measured sample showed ~20% increase in photocurrent at 5 V when illuminated, compared to the current in the dark. The increase in the conductivity of the sample during illumination was due to the increased generation of charge carriers in nanocrystalline Cu3SbS4. The measured results are in good agreement with the I-V measurements given in other works [26,34,35].

Optoelectric Measurements
For the analysis of optoelectric properties the I-V characteristics of Cu 3 SbS 4 nanoparticles mechanochemically synthesized after 120 min of milling were measured. As shown in Figure 8 the photoresponse of Cu 3 SbS 4 was confirmed from I-V measurements in the dark and under illumination. The linear dependence of measured I-V characteristics corresponded to the creation of a sufficient ohmic contact for the measurements. As can be seen in Figure 8, the measured sample showed~20% increase in photocurrent at 5 V when illuminated, compared to the current in the dark. The increase in the conductivity of the sample during illumination was due to the increased generation of charge carriers in nanocrystalline Cu 3 SbS 4 . The measured results are in good agreement with the I-V measurements given in other works [26,34,35].

Thermoelectric Measurements
Cu3SbS4 is known to have moderate thermoelectric performance. Recent studies have shown that ideally doped Cu3SbS4 can attain values of the thermoelectric figure of merit ZT= S 2 T/(ρκ) between 0.6 and 0.7 at 623 K [14,18,36], where S is the Seebeck coefficient, ρ is the electrical resistivity, and κ is the thermal conductivity. Famatinite possesses fairly high lattice thermal conductivity for a thermoelectric material (about 1-4 W·m −1 ·K −1 at room temperature), unlike some related materials from the Cu-Sb-S system, such as skinnerite Cu3SbS3 or tetrahedrite Cu12Sb4S13 which profit from the presence of lone-pair electrons of the Sb atoms, which help reduce lattice thermal conductivity [37].
Electrical properties (Figure 9) show p-type semiconducting behaviour with electrical resistivity decreasing with temperature and a positive Seebeck coefficient showing a weak temperature dependence and fairly high values. Electrical resistivity (Figure 9a) dropped from 1800 mΩ·cm at room temperature to 400 mΩ·cm at 623 K. The Seebeck coefficient (Figure 9b) reached a value of 600 µV·K −1 at 600 K. Similar results were reported in the literature for pristine Cu3SbS4: 585 µV·K −1 at 623 K [13] and 580 µV·K −1 at 573 K [15], for samples with low carrier concentration (~10 17 cm −3 at 300 K), and ~500 µV·K −1 at 573 K [16,38], for samples with higher carrier concentration (~4.2 × 10 18 cm −3 at 300 K). Based on these reports we can assume the carrier concentration of our sample to be in the range 10 17 -10 18 cm −3 , which is some three orders of magnitude below the optimum concentration attainable by doping [15].
The measured thermal conductivity κ shown in Figure 9c solely reflected the lattice part of κ, as the electronic part was negligible given the high electrical resistivity. As stated above, the absolute value was rather high for a thermoelectric material: 2.1 W·m −1 ·K −1 at room temperature which decreased to the 1.0 W·m −1 ·K −1 at 600 K. This decrease was close to the T −1 dependence characteristic of the Umklapp phonon-scattering process. The absolute value was lower than the 1.7 W·m −1 ·K −1 at 573 K reported by Goto et al. [38] for a sample synthesized by solid-state reaction, which can be explained by increased scattering of phonons due to the stronger presence of grain boundaries and defects introduced by high-energy milling in the nanograined sample. Other studies, however, report even lower values mechanochemically synthesized samples, such as 0.74 W·m −1 ·K −1 at 623 K [13], or 0.7 W·m −1 ·K −1 at 623 K [15]. The principal difference seems to lie in the milling time, which was an order of magnitude longer than the 2 h in our case, thus further amplifying

Thermoelectric Measurements
Cu 3 SbS 4 is known to have moderate thermoelectric performance. Recent studies have shown that ideally doped Cu 3 SbS 4 can attain values of the thermoelectric figure of merit ZT = S 2 T/(ρκ) between 0.6 and 0.7 at 623 K [14,18,36], where S is the Seebeck coefficient, ρ is the electrical resistivity, and κ is the thermal conductivity. Famatinite possesses fairly high lattice thermal conductivity for a thermoelectric material (about 1-4 W·m −1 ·K −1 at room temperature), unlike some related materials from the Cu-Sb-S system, such as skinnerite Cu 3 SbS 3 or tetrahedrite Cu 12 Sb 4 S 13 which profit from the presence of lone-pair electrons of the Sb atoms, which help reduce lattice thermal conductivity [37].
Electrical properties (Figure 9) show p-type semiconducting behaviour with electrical resistivity decreasing with temperature and a positive Seebeck coefficient showing a weak temperature dependence and fairly high values. Electrical resistivity (Figure 9a) dropped from 1800 mΩ·cm at room temperature to 400 mΩ·cm at 623 K. The Seebeck coefficient (Figure 9b) reached a value of 600 µV·K −1 at 600 K. Similar results were reported in the literature for pristine Cu 3 SbS 4 : 585 µV·K −1 at 623 K [13] and 580 µV·K −1 at 573 K [15], for samples with low carrier concentration (~10 17 cm −3 at 300 K), and~500 µV·K −1 at 573 K [16,38], for samples with higher carrier concentration (~4.2 × 10 18 cm −3 at 300 K). Based on these reports we can assume the carrier concentration of our sample to be in the range 10 17 -10 18 cm −3 , which is some three orders of magnitude below the optimum concentration attainable by doping [15].
The measured thermal conductivity κ shown in Figure 9c solely reflected the lattice part of κ, as the electronic part was negligible given the high electrical resistivity. As stated above, the absolute value was rather high for a thermoelectric material: 2.1 W·m −1 ·K −1 at room temperature which decreased to the 1.0 W·m −1 ·K −1 at 600 K. This decrease was close to the T −1 dependence characteristic of the Umklapp phonon-scattering process. The absolute value was lower than the 1.7 W·m −1 ·K −1 at 573 K reported by Goto et al. [38] for a sample synthesized by solid-state reaction, which can be explained by increased scattering of phonons due to the stronger presence of grain boundaries and defects introduced by high-energy milling in the nanograined sample. Other studies, however, report even lower values mechanochemically synthesized samples, such as 0.74 W·m −1 ·K −1 at 623 K [13], or 0.7 W·m −1 ·K −1 at 623 K [15]. The principal difference seems to lie in the milling time, which was an order of magnitude longer than the 2 h in our case, thus further amplifying the nanosctructuring and defect concentration in the milled product. This suggests that even though 2 h of milling are sufficient for the synthesis of phase-pure powder, further milling has a beneficial effect on the lowering of thermal conductivity of the resulting material.
Since the sample was synthesized as pristine Cu 3 SbS 4 , we cannot expect good thermoelectric performance which has to be achieved by doping. Indeed, the maximum of figure of merit ZT reached only 0.05 at 600 K. This value was lower than the 0.14 at 623 K reported in other papers [13,15], where pristine famatinite was synthesized by mechanical alloying, but higher than reported in [16,38] for samples prepared by melting-sintering of elements. Moreover, the lower thermal conductivity of mechanochemically synthesized or mechanically activated samples has an impact on increasing of ZT in contrast to non-milled or non-activated samples [39,40]. Both treatments followed by SPS can be used to prepare ceramics with small grain size [41], where the sources of phonon scattering such as grain boundaries lead to reduced phonon mean free path and, as a consequence, lower lattice thermal conductivity. In the paper [42], it was presented that dry milling of natural tetrahedrite leads to the production of thermoelectrics with increased electrical resistivity, but decreased thermal conductivity, compared to the mineral. This can be attributed to grain size reduction, the role of defects in doping, and grain boundaries in phonon scattering.
Micro 2023, 3, FOR PEER REVIEW 11 the nanosctructuring and defect concentration in the milled product. This suggests that even though 2 h of milling are sufficient for the synthesis of phase-pure powder, further milling has a beneficial effect on the lowering of thermal conductivity of the resulting material.
Since the sample was synthesized as pristine Cu3SbS4, we cannot expect good thermoelectric performance which has to be achieved by doping. Indeed, the maximum of figure of merit ZT reached only 0.05 at 600 K. This value was lower than the 0.14 at 623 K reported in other papers [13,15], where pristine famatinite was synthesized by mechanical alloying, but higher than reported in [16,38] for samples prepared by melting-sintering of elements. Moreover, the lower thermal conductivity of mechanochemically synthesized or mechanically activated samples has an impact on increasing of ZT in contrast to nonmilled or non-activated samples [39,40]. Both treatments followed by SPS can be used to prepare ceramics with small grain size [41], where the sources of phonon scattering such as grain boundaries lead to reduced phonon mean free path and, as a consequence, lower lattice thermal conductivity. In the paper [42], it was presented that dry milling of natural tetrahedrite leads to the production of thermoelectrics with increased electrical resistivity, but decreased thermal conductivity, compared to the mineral. This can be attributed to grain size reduction, the role of defects in doping, and grain boundaries in phonon scattering. The XRD analysis of famatinite Cu3SbS4 after SPS was also performed. For comparison, the XRD patterns of both samples before and after SPS are shown in Figure 10. The very fine crystallite size resulting in the broad diffraction peaks (Figure 10a) was influenced by the SPS process (Figure 10b), as evidenced by the narrower peaks. An analysis Figure 9. Thermoelectric properties of mechanochemically synthesized Cu 3 SbS 4 in dependence of temperature; panels from (a-d) represent electrical resistivity, Seebeck coefficient, thermal conductivity, and figure-of-merit ZT, respectively.
The XRD analysis of famatinite Cu 3 SbS 4 after SPS was also performed. For comparison, the XRD patterns of both samples before and after SPS are shown in Figure 10. The very fine crystallite size resulting in the broad diffraction peaks (Figure 10a) was influenced by the SPS process (Figure 10b), as evidenced by the narrower peaks. An analysis of the XRD spectrum after SPS did not reveal any phases other than famatinite crystallizing in the space group I-42m with the lattice parameters a = b = 5.39153 Å and c = 10.76360 Å, α = β = γ = 90 • . The crystallite size determined by Rietveld refinement was~55.8 nm and the estimated microstrain was ε 0 = 4.9 × 10 −4 . These results could be ascribed to enhanced crystallinity and stress release due to sintering in contrast to as-milled famatinite sample before SPS. Micro 2023, 3, FOR PEER REVIEW 1 of the XRD spectrum after SPS did not reveal any phases other than famatinite crystalliz ing in the space group I-42m with the lattice parameters a = b = 5.39153 Å and c = 10.76360 Å, α = β = γ = 90°. The crystallite size determined by Rietveld refinement was ~ 55.8 nm and the estimated microstrain was ε0 = 4.9 × 10 −4 . These results could be ascribed to en hanced crystallinity and stress release due to sintering in contrast to as-milled famatinite sample before SPS.

Conclusions
In summary, we studied the properties of famatinite Cu3SbS4 nanocrystals mechano chemically synthesized by high-energy milling in a planetary mill in the argon atmos phere at room temperature for 120 min. The crystallite size of the tetragonal famatinite Cu3SbS4 determined by Rietveld refinement of XRD data was 14 nm. HR-TEM image showed several crystallites in the range of 20-50 nm, revealing agglomerated nanocrystal and also confirming their tetragonal structure. The zeta potential measurement indicated prepared famatinite Cu3SbS4 could be oxidized on its surface very slightly, which was also confirmed by the XPS method. The photoluminescence spectrum indicated the quantum confinement effect of nanocrystalline famatinite Cu3SbS4. The I-V characteristic of the sample with Cu3SbS4 nanoparticles showed ~20% increase in photocurrent at 5 V afte illumination, which proves that the synthesized tetragonal Cu3SbS4 famatinite exhibited the acceptable photoelectric properties and provides a prerequisite for use in photovolta ics. The thermoelectric measurements revealed a figure of merit ZT of 0.05 at 600 K for the undoped famatinite Cu3SbS4 sintered by SPS at 723 K. This value could be improved by adjusting the carrier concentration via doping. The prepared famatinite Cu3SbS4 repre sents a prospective material for future applications.

Conclusions
In summary, we studied the properties of famatinite Cu 3 SbS 4 nanocrystals mechanochemically synthesized by high-energy milling in a planetary mill in the argon atmosphere at room temperature for 120 min. The crystallite size of the tetragonal famatinite Cu 3 SbS 4 determined by Rietveld refinement of XRD data was 14 nm. HR-TEM images showed several crystallites in the range of 20-50 nm, revealing agglomerated nanocrystals and also confirming their tetragonal structure. The zeta potential measurement indicated prepared famatinite Cu 3 SbS 4 could be oxidized on its surface very slightly, which was also confirmed by the XPS method. The photoluminescence spectrum indicated the quantum confinement effect of nanocrystalline famatinite Cu 3 SbS 4 . The I-V characteristic of the sample with Cu 3 SbS 4 nanoparticles showed 20% increase in photocurrent at 5 V after illumination, which proves that the synthesized tetragonal Cu 3 SbS 4 famatinite exhibited the acceptable photoelectric properties and provides a prerequisite for use in photovoltaics. The thermoelectric measurements revealed a figure of merit ZT of 0.05 at 600 K for the undoped famatinite Cu 3 SbS 4 sintered by SPS at 723 K. This value could be improved by adjusting the carrier concentration via doping. The prepared famatinite Cu 3 SbS 4 represents a prospective material for future applications.  The support of COST Action CA18112 MechSustInd (www.mechsustind.eu), supported by the COST Association (European Cooperation in Science and Technology, www.cost.eu) is also acknowledged.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.