1 Introduction to Mechanochemistry

Mechanochemistry is defined as a branch of chemistry that is concerned with chemical and physicochemical changes of substances of all states of aggregation due to the influence of mechanical energy [1]. According to the International Union of Pure and Applied Chemistry (IUPAC), which also recognizes it, a mechanochemical reaction is a chemical reaction that is induced by the direct absorption of mechanical energy [2]. In simple words, it is an alternative method to the traditional solution-based chemistry due to the availability to perform chemical reactions completely in solid state in the absence of toxic solvents and external elevation of temperature and pressure. It was defined as a separate research field at the beginning of twentieth century [3] and has been proven to be applicable in distinct research areas since then, as it is demonstrated in Fig. 1. Its applicability is mostly in chemistry, materials science, and environmental sciences.

Fig. 1
A chart has a photo of white balls inside a pot at the center engulfed by an arrow labeled mechanochemistry. It is surrounded by catalysis, main group chemistry, supramolecular chemistry, metal-organic framework, organometallic compounds, waste management, polymer technology, cosmetics, extraction of biological active compounds from natural sources, and others.

Diverse applications of mechanochemistry [4]

1.1 High-Energy Milling

High-energy milling is a tool to perform mechanochemical reactions. Apart from conventional ball milling that delivers only limited amount of energy mainly to reduce the particle size of the treated material, additional effects like mechanical activation and defect formation are observed by applying high-energy mills. Conventional ball milling also most often does not lead to the occurrence of chemical reaction. Today, there is a number of high-energy ball mills that efficiently deliver the energy to the treated powders (examples are shown in Fig. 2). For less energy-demanding reactions like organic ones, mixer or shaker mills are usually employed (Fig. 2b). The effect of milling is achieved by an intensive shaking from one side to the other, and the milling chambers usually have the cylindrical shape with a volume of up to 20 mL. However, to perform more energy-demanding inorganic synthesis, the planetary ball mills where the milling chamber is located on the disc and is being intensively rotated to the one side, while the disc is being rotated to the other one, are being used (Fig. 2a). The volumes of the chambers used in the planetary ball mills are usually up to 500 mL. Shaker/mixer and planetary mills are exhaustively applied for the lab-scale experiments, but they are not suitable for implementation on a larger scale. In order to scale up mechanochemical reactions, eccentric vibratory (Fig. 2c) or attrition mills (Fig. 2d) can be applied.

Fig. 2
4 photographs of a planetary ball mill, an electric shaker mill, a vibratory mill, and an attrition mill. The planetary ball mill is shaped as a weighing scale. The electric shaker mill is cylindrical. The vibratory mill is cuboidal and has a glass encasing on the front. The attrition mill has a cylindrical bottom and top. A whip from the top is linked to a beater shaft.

Examples of high-energy mills used for mechanochemical processes: (a) planetary ball mill, (b) electric shaker mill, (c) eccentric vibratory mill, (d) attrition mill. (Adapted from Springer Nature Customer Service Centre GmbH: Springer Nature [5] [COPYRIGHT] (2021))

There are plenty of experimental conditions that can be changed when performing the high-energy milling treatment. The most common ones are presented in Fig.3. Depending on the treated material, different conditions might be suitable. One of the greatest challenges in mechanochemistry is the abrasion of milling media, which contaminates the product. This problem can be overcome, or at least significantly reduced, by changing the conditions like the material of milling media, revolution speed, or using the grinding aids.

Fig. 3
A chart describes how mill type, milling time, milling chamber, breaks, chamber opening, milling balls, mass of sample, milling speed, surfactant, grinding aid, and the presence of liquid or solvent affect milling conditions.

Most commonly changed experimental conditions during high-energy milling. (Reprinted by permission from Springer Nature Customer Service Centre GmbH: Springer Nature [5] [COPYRIGHT] (2021))

The high-energy milling treatment of the powders can lead to two scenarios. Either just physicochemical characteristics of the treated solids (such as particle size or specific surface area) are changed, however, no clear chemical change is observed. In this case, the process can be defined as mechanical activation. On the other hand, when a clear chemical change occurs, we are talking about mechanochemical reaction. To illustrate this difference, Fig. 4 is shown. The two processes can take place in one system, and mechanical activation of the reactants is often necessary before they undergo the mechanochemical reaction.

Fig. 4
A schematic illustration of before milling and after milling. Before milling, there are reactants A and B. These on mechanical activation result in a product that has alternating combination of A and B in 4 layers. The reactants on mechanochemical reaction result in a new product, C.

Schematic illustration showing the difference between mechanical activation and mechanochemical reaction (A, B- reactants, C- product of mechanochemical reaction). (Reprinted by permission from Springer Nature Customer Service Centre GmbH: Springer Nature [5] [COPYRIGHT] (2021))

In most research groups working with mechanochemistry around the world, it is being tested as an alternative to the procedures that are already established there (as an example, performing organic reactions via the traditional solution-based methods versus mechanochemical one can be mentioned). However, the authors of this chapter are part of the department, where mechanochemistry is the main research focus and the high-energy milling is being applied for the processing of a wide variety of materials (Fig. 5).

Fig. 5
A chart has a map of Slovakia on which Kosice and Bratislava are marked. It points to the text, mechanochemistry. This further points to a photo of a vibratory mill, which is connected to a zoomed in photo of milling balls inside a coupling. A photo of a device below the text points to chalcogenides synthesis, oxide synthesis, environmental issues, organic mechanochemistry, and other applications.

Variety of research activities being performed at the Department of Mechanochemistry, Institute of Geotechnics, Slovak Academy of Sciences in Košice, Slovakia. (Reprinted from [6], Copyright (2020), with permission from Elsevier)

As it can be seen, one of the main research directions is the mechanochemical synthesis of chalcogenides. This has been a permanent agenda at the Department as manifested in the reviews [7, 8] and monographs from the years 2000–2021 published by the authors of this chapter [5, 9, 10]. In the present work, the overview of recent works of our research group from the year 2018 is provided. This is complemented by providing overview tables on the mechanochemical syntheses of chalcogenides by other research groups around the world. As the mechanochemical synthesis can also be performed in the scale-up fashion, the text is subdivided into a lab-scale and scalable synthesis. Further structuring is based on the chemical composition of chalcogenides, namely to binary, ternary, and quaternary ones.

2 Laboratory-Scale Mechanochemical Synthesis of Chalcogenides

The simplicity of mechanochemical synthesis of chalcogenides in comparison with other methods can be well demonstrated by the fact that just high-energy milling of the solid precursors is satisfactory to obtain nanocrystalline products [8]. Most often, two methodologies are applied; either the elements or compounds are used as precursors. The second option became preferred with time, as the obtained nanoparticles are usually smaller and thus more suitable for advanced applications, for example, as quantum dots. The principle of the two methodologies is demonstrated in the example of sulfides synthesis below (Fig. 6). The concept of using metal acetate and sodium sulfide as the precursor of metal and sulfur, respectively, has been introduced by the research group of the authors of this chapter already in [11]. This synthetic pathway is sometimes referred to as a wet one because sodium sulfide is used as nonahydrate and the resulting reaction mixture often resembles a slurry. As a side-product, sodium acetate, which can be easily washed out, is formed.

Fig. 6
2 flow diagrams present mechanochemical synthesis of sulfides using a, element precursors and b, compounds. In a, metal and sulphur on high-energy milling yield semiconductor nanocrystals. In b, metal acetates and sodium sulphide on high-energy milling, washing, and decantation yield sodium acetate. The remaining product on drying yields semiconductor nanocrystals.

Mechanochemical synthesis of sulfides using: (a) elemental precursors, (b) compounds (metal acetate and sodium sulfide). (Reprinted by permission from Springer Nature Customer Service Centre GmbH: Springer Nature [8] [COPYRIGHT] (2017))

3 Binary Systems

3.1 Cobalt Pentlandite, Co9S8

Co9S8 nanoparticles from cobalt and sulfur in stoichiometric ratio in a planetary ball mill in an inert atmosphere were successfully prepared via mechanochemical synthesis in [12]. The cubic Co9S8 nanoparticles were nanocrystalline with crystallite size of about 16 nm, as confirmed by both Rietveld analysis and transmission electron microscopy. They consist of nanocrystals closely aggregated into spherical-like objects. The kinetics reaction studied by X-ray diffraction (XRD) is depicted in Fig. 7. The patterns were recorded after 60–180 min of milling. After the first 60 min, only small amounts of desired sulfide are visible next to the non-reacted cobalt and sulfur precursors. Another 60 min of milling was needed to obtain final product cobalt pentlandite Co9S8 (01-073-6395), which crystallized in the cubic face-centered structure. Further treatment until 180 min did not lead to any significant change in the phase composition.

Fig. 7
A graph of intensity counts versus 2 theta in degrees. In it, X R D patterns for mechanochemically synthesized C o 9 S 8 after milling times of 60, 90, 120, and 180 minutes are plotted from bottom to top. The patterns are noisy and go through several prominent spikes. The patterns for 90, 120, and 180 minutes have 4 prominent peaks.

XRD patterns of mechanochemically synthesized pentlandite Co9S8 after different milling time: (a) 60 min, (b) 90 min, (c) 120 min, and (d) 180 min. (Reprinted from [12], Copyright (2018), with permission from Elsevier)

The transition from paramagnetic to weak ferromagnetic or ferrimagnetic behavior by changing the magnetic field was proven by magnetic measurements (Fig. 8).

Fig. 8
A 4 quadrant graph of magnetization in electromagnetic units per gram and magnetic field in kilooersted for C o 9 S 8 after 120 minutes of milling. In it, S shaped hysteresis loops are plotted at 7 different temperatures. There is an inset graph of magnetization versus temperature. In it, a right skewed bell curve is plotted.

Hysteresis loops measured at various temperatures and temperature dependence of magnetization (inset) of mechanochemically synthesized pentlandite Co9S8 after 120 min of milling. (Reprinted from [12], Copyright (2018), with permission from Elsevier)

3.2 Copper Sulfides-Chalcocite Cu2S and Covellite CuS

Copper sulfide is a semiconductor with multidisciplinary applications, the ones in biomedicine being of particular interest. In our older studies, we have shown the possibility to prepare this chalcogenide within a second range [13] and that despite the precursors are consumed, the phase transformations still go on for a couple of minutes [14]. The equilibrium was reached after 15 min and 30 min in the case of CuS and Cu2S, respectively. The obtained products were properly characterized using a rich pallette of experimental techniques (SEM/EDS, FTIR, XPS, photoluminescence, nitrogen adsorption, thermal analysis) [15]. In addition, the optical properties by UV-vis spectroscopy were determined (Fig. 9). The optical bandgap of CuS (blue) and Cu2S was calculated to be 1.92 eV and 3 eV, respectively. In the former case, a red shift in comparison with its bulk analogue was evidenced. The ability of both CuS and Cu2S to photocatalytically degrade the methyl orange (MO) dye was tested, showing both materials being good photocatalysts. Better results were achieved using Cu2S (Fig. 9b), as complete degradation of MO dye was observed within 150 min. Covellite (CuS) showed 80% MO degradation after the same duration. Interestingly, the results for the products were better than for the commercial TiO2 (P25) or ZnO. It has to be noted that the adsorption stage was not considered in the experiments and all the decrease in the dye concentration was ascribed to the photocatalytic effect [15].

Fig. 9
a is a line graph of absorbance versus wavelength in which lines for C u S and C u 2 S are plotted. The line for C u 2 S descends to a trough first and then ascends. The line for C u S ascends and goes through a peak and trough. b is a line graph of degradation ratio versus time. The lines for C u S and C u 2 S decay exponentially.

(a) UV-vis absorption spectra (inset Tauc’s plots) of CuS and Cu2S products. (b) The irradiation time-dependent degradation efficiencies of MO solution in the presence of CuS and Cu2S photocatalysts. (Reprinted from [15], Copyright (2018), with permission from Elsevier)

Mechanochemical synthesis of copper sulfide was also the main topic of the report [16], which was a common work with our Kazakh colleagues. The initial idea was to obtain copper sulfide/sulfur composite by combining the already explained acetate route (milling copper acetate and sodium sulfide) and simultaneously synthesize elemental sulfur (by using sodium thiosulphate as its source and citric acid as a catalyst). However, instead of obtaining CuS/S composite, the needle-like nanocrystals of CuS were obtained. This is of particular interest to mechanochemists, as the shape control is a large issue due to a constant supply of energy and stress to the milled solids, which hampers them from creating a well-crystallized structures. Simultaneously, the problem with synthesis of CuS using acetate route with the formation of oxidized products reported in the past, was also overcome by the proposed methodology. The whole paper is based on the comparison between CuS obtained using the traditional acetate route and the sample prepared by using also the additional sulfur source. The as-received CuS had a needle-like shape with a uniform thickness of around 6 nm length of up to 60 nm (aspect ratio 1:10), whereas the sample obtained using just copper acetate and sodium sulfide exhibited an isometric morphology with particle size around 20 nm. The formation of the specific morphology in the first case is most probably a result of the presence of elemental sulfur, which may in situ react with CuS by enhancing the S − S bonding and thus causing a preferential growth of CuS along the basal planes. There was also a significant difference in the grain size (as very often when using the mechanochemical synthesis), the individual nanoparticles are assembled into larger grains, being significantly finer in the case of needle-like CuS, namely, the average grain size was around 340 nm for CuS prepared without using the additional sulfur source and 170 nm when using it. The obtained products were tested as antibacterial agents using the agar well diffusion method, and it was shown that the elongated CuS exhibited activity against both gram-negative E. coli and gram-positive S. aureus, whereas the other sample was active just against the gram-negative bacteria. The whole story of the paper is summarized in Fig. 10 below.

Fig. 10
A chart illustrates how clusters of N a 2 S 2 O 3 dot 5 H 2 O and citric acid yield sulfur, clusters of C u C H 2 C O O twice dot H 2 O and N a 2 S dot 5 H 2 O yield C u S and clusters of sulfur and C u S yield n C u S. Below is an illustration of a rotating ball mill jar filled with milled materials and grinding balls. On its right is a double bar graph of R I Z D % for S, C u S and n C u S. Bars are plotted for E coli and S aureus. At the bottom are T E M images of C u S and n C u S with cloudy and needle like structures, respectively.

Graphical abstract of the report [16] showing the possibility to prepare needle-like CuS nanocrystals using mechanochemical synthesis. (Reprinted with permission from [16]. Copyright 2019 American Chemical Society)

The mechanochemical synthesis of CuS was also the topic of the study [17]; however, the selected reaction was used just as a model to investigate the amount of iron wear by means of magnetometry. The CuS product was formed in 15 min (in this case, the reaction was much slower than in [13], because the atomized powder Cu with spherical morphology, neither electrolytically prepared Cu powder with needle-like one was used). The possibility to use the magnetization value to calculate the actual amount of iron (even if it is present in trace amounts) in powders treated in steel milling jars was proposed. The amount of iron has rapidly increased between 15 and 30 min of milling, when its content was 0.13 and 0.76%, respectively. This increase is ascribed to the fact that after 15 min, the reaction is finished and further energy input is used inefficiently.

3.3 Co, Cu-Transition-Metal Selenides

In our recent works, nanostructured CoSe, CuSe, and Cu2Se semiconductors were successfully synthesized by simple, direct, and relatively inexpensive short-time solid-phase mechanochemical reactions in a high-energy planetary ball mill [18,19,20]. The synthetic Co9Se8 with the pentlandite structure, (Fe,Ni)9S8, was also prepared; however, in this case, also a short subsequent thermal annealing was necessary [21]. The X-ray diffraction patterns of the mechanochemically synthesized cobalt and copper selenides with the corresponding time of mechanochemical synthesis and subsequent thermal treatment, respectively, are shown in Fig. 11.

Fig. 11
A graph of intensity versus 2 theta in degrees. In it, X ray diffraction patterns for C o S e, C o 9 S e 8, C u S e and C u 2 S e at different annealing and milling times are plotted. The patterns have several prominent spikes.

X-ray diffraction patterns of synthesized cobalt and copper selenides; tM: milling time, tA: annealing time; F: freboldite-CoSe

The diffraction peaks were identified based on the ICDD-PDF cards and correspond to CoSe (freboldite 01-077-7572), Co9Se8 (01-089-4180), CuSe (00-34-0171), and Cu2Se (00-47-1448). Rietveld analyses confirmed that mechanochemical synthesis of CoSe is completed after 120 min of milling with an average crystallite size of 26 nm, and mechanochemical/thermal synthesis produced Co9Se8 nanoparticles with an average crystallite size of 67 nm. After 5 min of milling, CuSe and Cu2Se nanoparticles with an average crystallite size of 24 and 25 nm, respectively, were obtained.

TEM/HRTEM images of the agglomerated nanoparticles of prepared Co and Cu-containing transition metal selenides are shown in Fig. 12 and Fig. 13. The larger clusters composed of nanoparticles with tens of nanometer in size were observed in both cases. Within the single agglomerate of Co9Se8, the crystallites are in close contact, separated by distinct grain boundaries, almost like in sintered ceramics (see Fig. 12b)).

Fig. 12
a is a T E M image of an agglomerated cluster of cobalt selenide nanoparticles. It has a cloudy structure. b is a H E T E M image of the agglomerate that presents C o 9 S e 8 structures as grainy clouds. E D S spectrum graph is in the inset.

(a) TEM image of an agglomerated cluster of CoSe nanoparticles with a typical EDS spectrum (inset). The Cu peaks stem from the copper TEM-grid. Reprinted by permission from Springer Nature Customer Service Centre GmbH: Springer Nature [20] [COPYRIGHT] (2017); (b) HRTEM image from the thin part of the agglomerate shows larger Co9Se8 crystallites in close contact (GBs = grain boundaries). EDS analysis from this region yielded Co/Se ratio of 1.16, which is close to 9:8 as expected. Reprinted by permission from Springer Nature Customer Service Centre GmbH: Springer Nature [21] [COPYRIGHT] (2019)

Fig. 13
a and b are H R T E M and T E M images, respectively, of agglomerated crystals of C u S e and C u 2 S e, respectively. The crystals have a cloudy structure with a darker region in the center. In the inset are S A E D diffraction patterns of the crystals. The patterns have a bright spot in the middle surrounded by several concentric circles.

(a) HRTEM image of agglomerated crystals of CuSe with selected area electron diffraction (SAED) pattern (inset); (b) TEM image of agglomerated Cu2Se crystals; (c) SAED pattern of the same crystals. (Reprinted by permission from Springer Nature Customer Service Centre GmbH: Springer Nature [19] [COPYRIGHT] (2021))

By comparing our mechanochemical synthesis results with the literature, only Campos and co-authors managed to prepare CoSe in more than 72 h by milling in a SPEX shaker mill [22], Co9Se8 has not been mechanochemically synthesized yet. With regard to copper selenides, Ohtani and co-authors were the only researchers who synthesized γ-CuSe by mechanical alloying for 60 min using a high-energy ball mill [23]. In 2017, Bulat with co-workers produced a single-phase material containing the α-Cu2Se phase after 2 h of milling in a high-energy ball mill [24] and in 2018, Li and co-workers synthesized almost single-phase β-Cu2-xSe after 20 min of milling in a vibratory mill [25].

The overview table showing the precursors, necessary milling time, achieved particle size, and the application of the binary chalcogenides prepared recently by other research groups is shown below in Table 1.

Table 1 Overview of binary chalcogenides including selected mechanochemical conditions, obtained crystallite size, and application prepared by other research groups since 2018

4 Ternary Systems

4.1 Chalcopyrite, CuFeS2

One of the most interesting chalcogenides in ternary systems is chalcopyrite CuFeS2. The rapid mechanochemical synthesis of CuFeS2 particles with good magnetic, optical, and electro-optical properties prepared by high-energy milling in a planetary mill from copper, iron, and sulfur precursors was firstly investigated by our research group [54]. The progress of the mechanochemical synthesis was studied by XRD (Fig. 14). Nanocrystalline CuFeS2 with a tetragonal body-centered crystal structure and crystallite size around 38 nm could be prepared within 60 min of milling. No significant change, either in the lattice parameters or in the microstrain, was observed after prolonged milling until 120 min.

Fig. 14
A graph of intensity counts versus 2 theta in degrees. a to f are X R D patterns for C u F e S powders before milling and after milling times of 5, 15, 30, 60 and 120 minutes, respectively, plotted from bottom to top. The patterns are horizontal with several prominent spikes. a has spikes labeled S, F e and C u. b and c has spikes labeled C u S and F e. d has spikes labeled C u F e S 2 and F e. e and f have spikes labeled 112, 220, and 312.

XRD patterns of Cu-Fe-S powders before milling (a) and after 5 min (b), 15 min (c), 30 min (d), 60 min (e), and 120 min (f) of milling (the lines labelled by their Miller indexes alone belong to CuFeS2 nanoparticles). (Reprinted with permission from Ref. [54]. Copyright 2018, Elsevier)

This mineral has been considered as a potential thermoelectric material due to the abundance of the constituting elements, as well as its environmental friendliness. Nanophase and defect formation via milling is a possible route to decrease the fairly high thermal conductivity by increasing phonon scattering. In this respect, mechanochemical pretreatment by high-energy milling in CuFeS2 synthesis is an innovative procedure. Recently, a couple of papers to elucidate the thermoelectric properties of mechanochemically treated chalcopyrite have been published [55,56,57]. The influence of high-energy milling on structure as well as on the semiconducting properties was studied. While dependence of specific surface area SBET and crystallinity X is already known and tested for several sulfides [10] (Fig. 15 left), a possibility to influence of bandgap is a totally new agenda. In Fig. 15 (right), the dependence of bandgap value on milling time is shown, which illustrates a possibility to manipulate with semiconducting properties (bandgap engineering) of chalcopyrite.

Fig. 15
On the left is a graph of specific surface area and crystallinity versus milling time in minutes. The trend for surface area ascends linearly until (175, 4.8) and then descends to (250, 4.5). The trend for crystallinity descends first with a higher slope and then with a smaller slope. On the right is a graph of E subscript g versus milling time. The plotted line descends linearly.

The influence of milling time, tM on specific surface area, SBET (1), crystallinity, X (2) and bandgap, Eg (3) for chalcopyrite CuFeS2. (Reprinted with permission from Ref. [57]. Copyright 2018, Elsevier)

The progress of chalcopyrite synthesis from elements can be also investigated via observing the magnetization changes. This has been accomplished in the study [58]. Namely, at the beginning, the magnetism is very high due to the presence of highly magnetic elemental iron. As it becomes incorporated into chalcopyrite structure, the magnetization value decreases rapidly (from around 65 Am2/kg to around 2 Am2/kg upon 120 min of milling).

4.2 Copper Antimony Sulfides-:Chalcostibite, CuSbS2 and Famatinite, Cu3SbS4

CuSbS2 nanocrystals were mechanochemically prepared in an argon atmosphere after only 30 min of milling in a laboratory planetary mill [59]. The progress of mechanochemical synthesis is illustrated in Fig. 16 by means of XRD. The analysis validated the formation of orthorhombic structure of CuSbS2 with crystallite size around 25 nm determined by LeBail refinement. UV-vis spectrum revealed absorption of CuSbS2 nanocrystals in the entire visible range with the determined optical bandgap 1.54 eV.

Fig. 16
A graph of intensity versus 2 theta in degrees. In it, X R D patterns for C u S b S 2 for milling times of 5, 15, and 30 minutes are plotted. In the one for 5 minutes, the spikes are labeled S b 2 S 5, C u S and S b. In the one for 15 minutes, spikes labeled C u S b S 2 and S b are marked.

XRD patterns of CuSbS2 synthesized for different milling times (the Miller indices in the upper-most pattern belong to CuSbS2 nanoparticles). (Reprinted by permission from Springer Nature Customer Service Centre GmbH: Springer Nature [59] [COPYRIGHT] (2021))

The formation of nanocrystalline chalcostibite CuSbS2 was also validated by transmission electron microscopy (Fig. 17). The sample consists of small nanocrystalline domains, which aggregate giving rise to larger particles as can be observed in the TEM image (Fig. 17a). All the SAED rings (Fig. 17a (inset) could be indexed in the orthorhombic symmetry of the CuSbS2, thus being in accordance with the XRD results (Fig. 16). HRTEM images (Fig. 17c, d) indicate nanocrystals directed along two distinct zone axes.

Fig. 17
a is a T E M image of mechanochemically synthesized C u S b S 2. It has diffused clusters. In its inset, there is a T E M image that has a square placed over a bright spot surrounded by concentric circles. b is an E D S spectrum graph with peaks for C u, N i, S, and S b. c and d are H R T E M images that have a grainy structure filling the entire panel.

TEM analysis of mechanochemically synthesized CuSbS2: TEM image at low magnification with indexed SAED pattern (inset) (a), EDS spectrum (b), HRTEM images and the corresponding FFTs and simulated ED patterns (inset) (c, d). (Reprinted by permission from Springer Nature Customer Service Centre GmbH: Springer Nature [59] [COPYRIGHT] (2021))

In addition to chalcostibite, CuSbS2, also famatinite Cu3SbS4 was mechanochemically synthesized using copper, antimony, and sulfur in a stoichiometric ratio 3:1:4 in a planetary ball mill with the utilization of protective atmosphere without the next post-heating [60]. The progress of the mechanochemical synthesis of Cu3SbS4 for milling times 60–120 min is illustrated in Fig. 18.

Fig. 18
A graph of intensity versus 2 theta in degrees. In it, X R D patterns for C u 3 S b S 4 for milling times of 60, 90, and 120 minutes are plotted. The patterns are horizontal with several prominent spikes in between. The spikes are labeled C H, F t and F c. Tetragonal famatinite has the highest intensity.

XRD patterns of Cu3SbS4 synthesized in laboratory mill for different milling time. The identified phases are marked as follows: CuSbS2 – chalcostibite (CH), Cu3SbS4 – tetragonal famatinite (Ft), Cu3SbS4 – cubic famatinite (Fc). (Reprinted with permission from Ref. [60]. Copyright 2021, Elsevier)

The mechanochemical reaction was completed after 120 min of milling, when no other phases could be detected in the XRD pattern, indicating the high purity of the product. The crystallite size 14 nm of the prepared tetragonal famatinite was calculated by LeBail refinement.

The optical properties of Cu3SbS4 were studied using UV-vis spectroscopy (Fig. 19). The Tauc plot to deduce the direct bandgap from derived UV-vis spectrum is shown as an inset of Fig. 19. The determined direct optical bandgap 1.31 eV is blue-shifted relatively to the bulk Cu3SbS4. The observed blue shift could be attributed to the existence of very small nanocrystallites.

Fig. 19
A graph of absorbance and wavelength. The famatinite line ascends with minor fluctuations. In the inset is a graph of alpha times h times nu whole squared versus h times nu. In it, a line that ascends in a concave up manner is plotted.

UV-vis spectrum of mechanochemically synthesized Cu3SbS4 and Tauc plot (inset). (Reprinted with permission from Ref. [60]. Copyright 2021, Elsevier)

4.3 Copper Tin Sulfide Mohite, Cu2SnS3

By means of mechanochemical synthesis, we managed to prepare Cu2SnS3 (mohite) phase in 15 min using planetary ball milling starting from elements [61]. The XRD patterns measured after 15 s, 5, 10, 15, and 30 min are shown in Fig. 20a. Already after a few seconds, the elemental Cu and S rapidly reacted to form CuS (already after 15 s, the diffractions of CuS are visible). The exothermic reaction of CuS formation is in line with the phenomena described in [13] and was corroborated by a significant increase of pressure in the milling chamber after about 13 s. Until 15 min, tin, together with additional sulfur, was subsequently incorporated into the crystal structure. Further milling did not bring about any changes in the phase composition, only the occurrence of agglomeration phenomenon. The product was in the form of nanocrystals embedded in large micron-sized agglomerates. The presence of small nanocrystals was confirmed by TEM (HRTEM), showing that the size of all the nanocrystals was below 20 nm (Fig. 20b). The obtained crystals were defective, and such planar defects are marked by an arrow in the figure. The electrical resistivity showing a clear negative correlation with the grain size was also discovered therein.

Fig. 20
A is a 3 D graph of milling time versus 2 theta degrees versus intensity. The different crystallographic phases of covellite, tin, and mohite are marked in the patterns. Mohite is pointed highest in the mixture, followed by copper tin. b is a grainy H R T E M image of C u 2 S n S 3 product with 4 arrows pointing toward the planar defects.

(a) XRD patterns of the copper-tin-sulfur reaction mixtures milled for different time. The identified crystallographic phases are marked in the figure as follows: C—covellite (JCPDS 00–006-0464), T—tin (JCPDS 03–065-0296), M—mohite (JCPDS00–027-0198), (b) HRTEM image showing the Cu2SnS3 product obtained after 15 min of milling of the stoichiometric elemental mixture. The arrows point to the planar defects. (Reprinted by permission from Springer Nature Customer Service Centre GmbH: Springer Nature [61] [COPYRIGHT] (2018))

Very recently, an update to this study concentrating mainly on the Raman spectroscopy characterization of these samples was published. Due to the high sensitivity of Raman spectroscopy, an additional phase of copper sulfate was found, and its presence was postively correlated with the amount of CuS phase, which most probably serves as a precursor of its formation. Thus, the sample after 15 min was the most stable [62].

In our research, we also managed to obtain pure Cu2SnS3 phase after 60 min of milling in a laboratory planetary ball mill using covellite CuS and SnS (previously obtained by a mechanochemical synthesis) as precursors. The obtained nanocrystals exhibited size of approximately 10–15 nm. Reaction progress calculated from the Rietveld refinement of the XRD patterns is illustrated in Fig. 21a, where the dependence of conversion degree on milling time is shown. Since CTS phase exists in several polymorphs not detectable by XRD, the Raman spectroscopy was used for deeper inspection (Fig. 21b). There was a peak broadening due to the phonon confinement effect. The peak with a maximum at 335 cm−1 is undoubtedly assigned to the tetragonal CTS phase. The Raman spectroscopy revealed also the presence of impurities like CuS, SnS, and SnS2, which are formed as intermediates during the mechanochemical synthesis.

Fig. 21
a is a scatter plot of conversion degree versus milling time. Exponentially decaying curves fit the datasets for C u S and S n S. A curve that ascends in a concave downward manner fits the dataset for C u 2 S n S 3. Alpha inf is the intersection of S n S and C u 2 S n S 3 trends. b is a graph of conversion degree versus Raman shift. In it, 4 intersecting bell curves with ascending peaks are plotted. The peaks are labeled S n S, S n 2 S 3, and C u 2 S n S 3 from left to right. An average curve is also plotted.

(a) Reaction kinetics of Cu2SnS3 synthesis in a laboratory mill calculated from the XRD data using the Rietveld refinement, (b) fitted Raman spectra of the product synthesized after 60 min. (Reprinted with permission from Ref. [63]. Copyright 2018, Elsevier)

The overview table showing the precursors, necessary milling time, achieved particle size, and the application of the ternary chalcogenides prepared recently by other research groups is shown below (Table 2).

Table 2 Overview of ternary chalcogenides prepared by other research groups since 2018

5 Quaternary Systems

From the quaternary systems, most of our work was devoted to Cu-Fe-Sn-S and Cu-Zn-Sn-S systems. Initially, we planned to apply the obtained materials as adsorber layer in the solar cells; however, the manufacturing of actual cell was necessary to test the performance, and this was not feasible, so therefore, we turned to testing these compounds as potential thermoelectric materials.

5.1 Stannite, Cu2FeSnS4

With regard to Cu-Fe-Sn-S (CFTS) system, we have intensively investigated the mechanochemical synthesis of stannite Cu2FeSnS4 phase in the past [75,76,77]. The synthesis included a rapid formation of CuS as the first step, then incorporation of tin and finally iron into the structure. However, there was always an admixture of rhodostannite Cu2FeSn3S8 phase. More recently, two studies describing the spectroscopic properties of the CFTS nanocrystals were published. In the first one [78], the detailed far-infrared study on the prepared CFTS nanocrystals was conducted, and as demonstrated in the spectra obtained for the sample milled for different times, the signals corresponding to binary sulfides (reaction intermediates) disappear with time to reach only CFTS after 2 h of the treatment (Fig. 22). The same facts have been confirmed in another study utilizing Raman measurements [79].

Fig. 22
A scatterplot of reflectivity versus wave number. In it, spectra of stannite C u 2 F e S n S 4 at 4 different milling times are plotted. Descending lines fit the experimental spectra presented by circles. Arrows labeled S n S and F e S point to the spectra.

Far-infrared reflection spectra of stannite Cu2FeSnS4 obtained after different milling times (given in the figure). Experimental spectra are presented by circles. The solid lines are calculated spectra obtained by a fitting procedure based on the Maxwell-Garnet mixing rule. (Reprinted with permission from Ref. [78]. Copyright 2018, Elsevier)

5.2 Kesterite, Cu2ZnSnS4

We also succeeded in mechanochemically synthesizing the analogue with zinc (instead of iron), which is called kesterite, Cu2ZnSnS4 (CZTS) [80]. We have investigated an option to prepare it by using elemental precursors and also a mixture of compounds and elements (see equations below):

$$ 2\ \textrm{Cu}+\textrm{Zn}+\textrm{Sn}+4\ {\textrm{S}\rightarrow\textrm{Cu}}_2{\textrm{ZnSnS}}_4\vspace*{-30pt} $$
(1)
$$ 2\ \textrm{CuS}+\textrm{SnS}+\textrm{Zn}+{\textrm{S}\rightarrow\textrm{Cu}}_2{\textrm{ZnSnS}}_4\vspace*{-20pt} $$
(2)

The reaction pathway is of course different depending on the option selected. Whereas when the combination of compounds and elements is used, CuS and SnS react to form Cu2SnS3, and the elements react to form ZnS (Fig. 23, left), the immediate formation of CuS and subsequent incorporation of Sn and Zn takes place in the other case (Fig. 23, right). In both cases, the pure kesterite phase was obtained after 60 min of milling, although in the last 30 min, just the incorporation of last traces of zinc takes place.

Fig. 23
a and b are graphs of counts versus 2 theta. In both, X ray diffractometry patterns for mixtures at 5 different times are plotted using different colors. The patterns go through 4 prominent spikes and are labeled as C Z T S. Markers are used to plot counts for C u S, S n S, and Z n at intervals of 2 theta in a. In b, markers are used to plot counts for C u, S n, Z n and S.

X-ray diffraction patterns of stoichiometric mixtures milled according to Eq. 2 (from compounds/elements mixture) (a) and Eq. 1 (from elements) (b) for various times. (Reprinted with permission from Ref. [80]. Copyright 2019, Elsevier)

TEM analysis has shown that in both cases, the CZTS nanocrystals have the size ranging from 5 to 20 nm and that they contain defects. The elemental distribution seems to be slightly better in the case of the sample prepared by a combination of compounds and elements.

The overview table showing the precursors, necessary milling time, achieved particle size, and the application of quaternary chalcogenides prepared recently by other research groups is shown below (Table 3). Similarly to ternary chalcogenides, only the reports on sulfides were found.

Table 3 Overview of quaternary chalcogenides prepared by other research groups since 2018

6 Scale-Up of Mechanochemical Synthesis of Chalcogenides

Although the lab-scale processing opens up new research pathways and leaves enough space for scientists to investigate interesting phenomena, the large-scale production is always questioned by the companies and the applied research. Mechanochemistry is well-scalable; however, only selected methods are suitable for it. In our research, we have shown that the synthesis of chalcogenides can be well-performed in a scalable manner by using eccentric vibratory milling. In this section, we provide a brief overview to show the feasibility of scalable mechanochemical synthesis of chalcogenides via this method.

The concept of eccentric vibratory mill (ESM) has been developed in Germany by the group of professor Gock [92, 93]. The schematic view of the ESM and its photograph is provided in Fig. 24a and b, respectively. The key difference in comparison with the traditional vibratory mills is that this mill produces elliptic, circular, and linear vibrations, which lead to unequalled amplitudes of vibration up to 20 mm and a high degree of disaggregation of the grinding media. This results in more intensive milling, and the specific energy consumption is reduced by 50%.

Fig. 24
a is a front view schematic of eccentric vibratory mill. It has a large circle at the center labeled grinding pipe filled with balls labeled grinding balls, connected on the left to unbalanced drive and on the right to balanced mass. b is its side view photo.

Eccentric vibratory mill: (a) schematic view, (b) photograph. (Part A is reprinted with permission from Ref. [93]. Copyright 1999 Elsevier)

As it can be seen in Fig. 24b, there is a gray vessel in the middle of the front wall of the mill. This is called a “satellite” chamber and mirrors the exact events going on inside of the big mill (capable of the treatment of tens of kilograms) in a smaller volume. In this configuration, milling of the overall mass up to hundreds of grams is possible, and in our research, we have mostly used this satellite chamber.

The overview table showing various chalcogenides synthesized by our research group using an eccentric vibratory milling (in a satellite chamber) together with the information about the batch size and milling time is provided in Table 4. As it can be seen, mainly copper-based chalcogenides were synthesized. The duration of milling was usually longer than in the case of lab-scale synthesis; however, the scalability of the process is of utmost importance. Finally, we managed to prepare covellite in 7.5 kg batch upon operating the ESM device in the full mode. In the majority of works, we have evaluated the thermoelectric performance of the prepared chalcogenides.

Table 4 Overview of recent studies of our research group reporting the scalable mechanochemical synthesis using eccentric vibratory mill

7 Conclusion

In this chapter, mechanochemical synthesis has been shown as a very perspective method to prepare nanocrystalline chalcogenides. The high-energy milling process and the events of mechanical activation and mechanochemical reaction play a decisive role in achieving the products with desired properties. The overview tables have shown that a rich plethora of binary, ternary, and quaternary chalcogenides have been recently prepared by research groups all around the world. The works of our research group focusing on the lab-scale preparation of mainly copper-based sulfides and selenides have been reviewed. The scalability of mechanochemical synthesis upon using eccentric vibratory milling was outlined in the last part of the chapter.