Green Mechanochemical Synthesis of Binary and Ternary Cadmium Chalcogenides with Tunable Band Gaps

1. Introduction

Cadmium chalcogenides, namely cadmium sulfide (CdS), cadmium selenide (CdSe), and cadmium telluride (CdTe), represent an important class of II–VI semiconductor materials with wide-ranging applications in optoelectronics, photovoltaics, and luminescent devices. Cadmium sulfide with a direct band gap of 2.42 eV was reported to possess immense potential applications in LED diodes and solar cells already two decades ago [1] and has recently been under intense investigation due to its outstanding catalytic, optical, and electronic properties, enabling its use in photocatalytic water splitting [2,3,4]. Cadmium selenide has served as the subject of many studies due to its good application prospects in optoelectronic devices [5]. The insertion of alkali metals and group 12 metals has led to the synthesis of a large group of chalcogenides with desirable electrical and optical properties [6,7].

Beyond these binary compounds, their ternary and quaternary solid solutions allow fine-tuning of band gaps and optical properties, thereby expanding their technological potential [8]. In particular, Cd-based ternary and quaternary (CdSeTe and CdSSeTe) compounds have shown superior properties that can be controlled by managing the composition, even without changing the particles’ size [9]. As an instance, the CdS : CdSe ternary system has been reported to exhibit a perfect choice for photodetectors due to its wide adjustable band gap range [10]. CdSSe/ZnS nanocomposites were reported for their applications in quantum-dot light-emission diodes (qLED) [11] while CdSeS ‘magic-size’ clusters have shown potential for applications for light scavenging and bioimaging [12].

Among the cadmium chalcogenides, cadmium telluride CdTe holds a prominent position as a II–VI semiconductor that has been widely investigated for its favorable optoelectronic properties [13]. With a direct band gap of ~1.5 eV, close to the Shockley–Queisser optimum for single-junction solar cells, and a high absorption coefficient, CdTe efficiently absorbs sunlight even in very thin layers [14,15]. These properties make it particularly attractive for thin-film photovoltaics, where CdTe-based devices have achieved competitive conversion efficiencies while maintaining lower production costs compared to crystalline silicon [16]. Today, CdTe solar cells represent one of the most commercially mature thin-film technologies and continue to play a key role in the development of scalable, high-efficiency, and cost-effective renewable energy systems [16,17,18].

While conventional methods for the synthesis of transition metal chalcogenides have long been known and established, they have key shortcomings that make them incompatible with the principles of green chemistry: they often require high temperatures, toxic precursors, or complex processing steps [19,20,21]. Among alternative preparation methods, which have been gaining popularity over the last decades, the single-source precursor method has emerged as an attractive approach for synthesizing metal chalcogenides, as both the metal and chalcogen are incorporated within one molecule, enabling better stoichiometric control [20,21,22]. While this method was also successfully applied during our previous research [23], it has a serious impediment due to the limited number of potential ligands containing sulfur, selenium and especially tellurium. Another method, which has emerged over the past two decades as a valuable technique for synthesizing nanostructured materials, is the sonochemical process. Ultrasound causes the phenomenon called acoustic cavitation, i.e., the formation, growth, and implosive collapse of gas bubbles in the sonicated solution or slurry. This process causes transitory high temperatures and pressures as well as rapid cooling rates, which can be exploited to create nanoparticles [20]. The sonochemical method was successfully applied during our previous research for the synthesis of binary CdS, CdSe, and CdTe nanoparticles in aqueous solutions [24,25] and was also successfully utilized for producing hexagonal CdS nanoplatelets in a continuous reactor [26]. Nevertheless, the aforementioned method obviously encounters its limitations when it comes to the synthesis of ternary compounds [27].

In contrast, the mechanochemical synthesis has emerged as a simple, energy-efficient, and environmentally benign alternative, enabling direct solid-state reactions at room temperature, in most cases without the need for solvents [19]. Mechanochemical reactions are the result of chemical transformations, occurring from grinding, milling, and alloying solid precursors. The method is carried out at relatively low temperatures (i.e., without external energy input), is reported to be reproducible, simple, and easy to operate, ensuring a high yield and control of the particle size by changing the milling conditions [20,28,29]. This approach offers advantages such as reduced reaction times, scalability, and the possibility of accessing metastable phases, making it an attractive and ‘green’ method for producing cadmium chalcogenides and their alloys [30,31]. While traditionally limited to niche applications, mechanochemistry has recently evolved from a laboratory curiosity to a powerful tool for the synthesis of nanosized materials [32], especially when combined with other energy sources like light irradiation, sound agitation, or electrical impulses [33]. Among other possible applications, the mechanochemical synthesis of nanoparticles for potential antimicrobial applications was reported recently [34].

Although mechanochemical syntheses of binary transition metal chalcogenides have been reported extensively in the literature, the first one being published more than two decades ago [35], this topic remains an active area of research. While the mechanochemical method is clearly suitable for the synthesis of binary chalcogenides [36,37], including the preparation of Cu, Cd, Al, Ga, and Ni chalcogenides reported in previous papers published by our group [38,39,40], the extension of this approach to ternary cadmium chalcogenides has, to the best of our knowledge, not yet been described and therefore represents the novel contribution of the present study.

2. Materials and Methods

2.1. Ball Milling

All reagents used were commercially available and employed without additional purification. Cadmium (powder, ~100 mesh, 99.5% trace metal basis) and tellurium (powder, 200 mesh, 99.8% trace metal basis) were obtained from Sigma-Aldrich, St. Louis, MO, USA. Selenium powder was acquired from Kemika d.d., Zagreb, Croatia, while sulfur (powder, 99.98%) was obtained from Merck KGaA, Darmstadt, Germany. The treatment of the mixtures was performed in a Fritsch premium line PULVERISETTE 7 Planetary Micro Mill (Fritsch, Idar-Oberstein, Germany) equipped with a 20 mL zirconia griding bowl and 10 mm diameter zirconia grinding balls. During all experiments, we used 10 milling balls (with a total mass of ≈ 30 g) and a 15 : 1 ball-to-powder mass ratio. As an example: for the synthesis of CdS_0.5Se_0.5, we weighed 1.3388 g (= 11.91 mmol) of Cd, 0.1909 g (= 5.95 mmol) of S, and 0.4703 g (= 5.96 mmol) of Se, to satisfy both the conditions: n(S) + n(Se) = n(Cd) and m(S) + m(Se) + m(Cd) ≈ 2 g. The masses of precursors during other experiments were calculated based on the same assumptions. To prevent overheating, the milling was conducted in 15 min intervals, followed by 15 min breaks. While initial experiments were performed in air atmosphere, we used argon atmosphere during some of the subsequent experiments to prevent oxidation, as explained further on.

2.2. Characterization

Powder X-ray diffraction (pXRD) data were collected on a PANalytical X’pert PRO diffractometer, (Malvern Panalytical Ltd., Malvern, UK) using CuKα1 radiation (λ = 1.5406 Å), scanning range from 10° to 70°, step size 0.02° and time/step = 1 s. Transmission electron microscopy (TEM) measurements were performed using a JEOL 2100 microscope (JEOL Inc., Peabody, MA, USA). The obtained powders were dispersed in isopropanol using an ultrasonic bath and placed on a carbon-coated copper grid. The elemental composition of the particles was analyzed using a Jeol JED 2300 EDXS system, (JEOL Inc., Peabody, MA, USA). The hydrodynamic particle size was measured by dynamic light scattering (DLS) with Zetasizer Nano-S equipment (Malvern Instruments, Malvern, UK) using disposable 1 cm polystyrene cuvettes. A UV-VIS spectrometer Varian Cary 50 Bio (Varian Inc., Palo Alto, CA, USA) was used to study the optical properties of the prepared cadmium chalcogenides. The samples were dispersed in ethanol in an ultrasonic bath prior to the measurement. The cadmium content was measured on a Varian SpectrAA-10 flame atomic spectrometer (AAS) (Varian Inc., Palo Alto, CA, USA) after digestion of the samples [41] with concentrated HNO₃ in PTFE vessels in a microwave oven (MDS-2000, CEM Corporation, Matthews, NC, USA). Thermal analysis of the samples was carried out by Mettler DSC 3 (Mettler Toledo, Switzerland) in 40 µL aluminum crucibles, using a heating rate of 10 K/min and a temperature range from 30 to 600 °C.

3. Results and Disussion

3.1. Binary Mixtures

Powder X-ray diffraction (pXRD) patterns of products obtained by ball milling equimolar Cd:S, Cd:Se, and Cd:Te mixtures for 15 min are shown in Figure 1, Figure 2 and Figure 3. In the case of the Cd : S system (Figure 1), all major reflections correspond to hexagonal CdS (ICDD PDF-4, file No. 00-006-0314). When milling was carried out in air, additional peaks attributable to CdO (ICDD file No. 00-005-0640) were observed. This oxidation was effectively suppressed when the milling was performed under an argon atmosphere, yielding phase-pure CdS. The Cd:Se system exhibited analogous behavior (Figure 2). The main product was cubic CdSe (ICDD file No. 00-019-0191), while minor CdO reflections appeared in samples prepared in air. As in the case of CdS, performing the reaction under argon completely prevented CdO formation. In contrast, the mechanochemical reaction of Cd and Te consistently yielded single-phase cubic CdTe (ICDD file No. 00-015-0770), even when milling was conducted in air (Figure 3), with no CdO impurities detected.

The pXRD results were further confirmed by DSC measurements for all precursors and ball-milled products, as presented in Figure 4 (a) – (c). For the Cd:S system (Figure 4a), the DSC curves of both elemental precursors display pronounced endothermic peaks corresponding to the melting points of Cd (T_m = 321 °C) and S (T_m = 119 °C). The minor endothermic peaks on the DSC curve of sulfur correspond to Sα → Sβ and Sβ → Sγ transitions [42] and the endotherm occurring at T ≈ 400 °C indicates the boiling point of sulfur; however, the changes beyond the melting point can be considered to be irrelevant to the discussion. After 15 min of milling, none of the precursor melting peaks remain, indicating complete conversion into CdS. This is consistent with the high melting point of CdS (≈ 1750 °C), which lies well beyond the accessible range of conventional DSC. The DSC curve of the product obtained after 60 min of milling is indistinguishable from that after 15 min, further confirming the rapid and complete formation of CdS. For comparison, a sample prepared by simple mortar-and-pestle grinding for ~ 5 min was also analyzed. Its DSC curve still shows both precursor melting peaks, albeit with significantly reduced intensity, suggesting that the reaction is initiated even under such mild mechanical treatment. A similar phenomenon was previously reported for the mechanochemical synthesis of CuS [38].

For the Cd:Se system (Figure 4b), the DSC curve of elemental selenium shows a sharp endothermic peak at ~221 °C, corresponding to its melting point. After 15 min of milling, the melting peaks of both precursors are absent, indicating complete conversion to CdSe. Extending the milling time to 60 min produces no further changes, confirming the rapidity of the reaction. Analogous behavior was observed for the Cd:Te system (Figure 4c). The DSC curve of tellurium exhibits a sharp endothermic peak at ~449 °C, corresponding to its melting point. After 15 min of milling, this peak disappears, consistent with complete reaction to CdTe. As CdTe has a melting point close to 1100 °C [43], no product-related endotherms can be detected within the accessible DSC temperature range.

The binary systems thus demonstrate that mechanochemical synthesis enables the rapid and complete formation of CdS, CdSe, and CdTe within 15 min of milling, with phase purity dependent only on the milling atmosphere in the cases of CdS and CdSe. Having established the efficiency and reliability of the method for binary cadmium chalcogenides, attention was next directed to ternary Cd–S–Se–Te systems. These compounds are of particular interest because their tunable compositions enable adjustment of structural and electronic properties, making them attractive for optoelectronic and photovoltaic applications.

3.2. Ternary Mixtures

Powder diffraction patterns of CdS_xSe_1-x samples, prepared using two different S : Se ratios, are presented in Figure 5 (a) and Figure 5 (b). The pattern of the sample, prepared by using a S : Se = 0.5 : 0.5 ratio, is shown together with those of both binary compounds, i.e., CdS and CdSe, in Figure 5 (a). Although no ICDD reference exists for CdS_0.5Se_0.5, the observed reflections fall almost exactly midway between the signals of CdS and CdSe, consistent with the formation of a solid solution. The sample prepared by using a S : Se = 0.8 : 0.2 ratio (Figure 5 (b)), matches perfectly with the pattern of ICDD file No. 00-040-0837, reported therein as ‘Cd₁₀S_8.13Se_1.87’. Notably, in contrast to the binary CdS and CdSe systems, no peaks of CdO were detected, even when milling was conducted in air.

A comparable outcome was obtained for the CdS_xTe_1−x system. The diffraction pattern of the sample with S:Te = 0.5:0.5 (Figure 6) displays reflections located midway between those of CdS and CdTe. No ICDD reference exists for CdS_0.5Te_0.5, however the absence of CdO peaks indicate that, as with CdS_xSe_1−x, no CdO could be detected even when milling was performed in air.

Finally, diffraction patterns for the CdSe_xTe_1−x system are depicted in Figure 7. Analogous to the previous two examples, the peaks of CdSe_0.5Te_0.5 are located between those of both binary chalcogenides. Although no ICDD reference exists for CdS_0.5Te_0.5, the measured peaks correspond closely to the reported pattern of hexagonal CdSe_0.6Te_0.4, (ICDD file No. 00-041-1325).

As with the binary systems, the pXRD findings were complemented by DSC measurements performed on all precursors and products, as presented in Figure 8 (a) – (c). In all three ternary systems, no precursor melting peaks were observed after 15 min of milling, confirming complete conversion. Prolonged milling up to 60 min produced no significant changes in the DSC curves, indicating that the reactions are already completed within the initial 15 min.

The crystallite sizes of all products were calculated using the Scherrer equation:

$$d_x={\displaystyle \frac{0,94\cdot \lambda \cdot 57,3}{\beta \cdot \cos \theta }}$$

Here, λ is the wavelength of the X – ray radiation (nm), β the full – width at half – maximum (FWHM) of the corresponding peak (^o) and θ is the diffraction angle (^o). The results, presented in Table 1, were obtained by calculating the average value of 2 – 3 most prominent peaks.

The morphology of the binary products obtained after 15 min of milling was examined by TEM (Figure 9 (a) – (c)). All samples consist of irregularly shaped aggregates with dimensions extending up to several hundred nanometers, a feature commonly observed in mechanically milled products due to the intense forces exerted during the process [40]. Closer inspection suggests the presence of smaller, nearly spherical primary nanoparticles with diameters in the 10–20 nm range. These dimensions are broadly consistent with crystallite sizes estimated from pXRD data using the Scherrer equation, although the clustered nature of the samples limits a more precise assessment. Results of the EDX analysis correspond well to the results obtained by pXRD and to Cd content, determined by AAS: 77.4 wt.% Cd for CdS (calc. 77.8%), 58.5 wt.% Cd for CdSe (calc. 58.7%), 46.3 wt.% Cd for CdTe (calc. 46.8%).

Similar conclusions can be drawn from TEM images of selected ternary products (CdS_0.5Se_0.5, CdS_0.5Te_0.5, and CdSe_0.5Te_0.5), as presented in Figure 10 (a)–(c). At higher magnifications, the presence of spherical primary nanoparticles is more clearly visible, as depicted in Figure 11 (a)–(c). AAS results for ternary products again showed good agreement with the data obtained from EDX and pXRD: 66.5 wt.% Cd for CdS_0.5Se_0.5 (calc. 66.9%), 58.5 wt.% Cd for CdS_0.5Te_0.5 (calc. 58.5%) and, 51.9 wt.% Cd for CdSe_0.5Te_0.5 (calc. 52.1%).

Particle size measurements and hydrodynamic size distribution were additionally corroborated by DLS measurements in water suspensions after applying an ultrasonic bath. For each of the samples, three independent measurements were performed and the average value calculated. The results are presented in Table 2. As expected, the hydrodynamic diameters obtained by DLS are significantly larger than crystallite sizes estimated from pXRD or the primary nanoparticle dimensions suggested by TEM. This discrepancy is well known and arises because DLS measures the apparent size of particles as they diffuse in suspension, which includes contributions from solvation layers and possible particle agglomeration in liquid media [40,44,45]. Within this general trend, the CdTe and CdSe_0.5Te_0.5 samples appear to exhibit comparatively smaller hydrodynamic sizes, although the origin of this behavior is not yet clear and would require more systematic investigation.

As already outlined in the Introduction, a key motivation for investigating ternary cadmium chalcogenides lies in the possibility of tailoring their band gaps by varying the composition [9,10]. Generally, the band gap width of bulk CsS is reported to be 2.42 eV, while that of CdSe is 1.74 eV, although variations exist in the literature due to size effects and sample morphology. As expected, there are variations in values reported throughout the literature, with the band gap increasing with decreasing nanoparticles size due to the quantum confinement effect [46]. On the other hand, CdS thin films have been reported with slightly reduced band gap values in the 2.27 – 2.33 eV range [47].

In ternary systems, the band gap typically falls between those of the two binary end members, but the relationship is often nonlinear due to the so-called bowing effect [48]. For the CdS_xSe_1−x system, the deviation from linearity is relatively small, since CdS and CdSe are miscible with almost zero enthalpy change due to small lattice mismatch [49]. By contrast, the CdS_xTe_1−x system exhibits a more complex dependence, with some of the ternary compositions reported to display either broader or narrower band gaps than any of the binaries [50,51]. Similar behavior was reported for the CdSe_xTe_1−x system, which has gained considerable attention in photovoltaic research, as partial incorporation of CdSe into CdTe thin film solar cells has been shown to improve cell performance [52,53]. At first glance, using a 1.74 eV band gap CdSe to increase the optical absorption of CdTe (E_g = 1.45 eV) seems counterintuitive, however alloying of CdSe into CdTe can effectively reduce the band gap below the values of either of the binary compounds due to the large bowing effect [54].

In the present work, the optical band gap energies of ternary cadmium chalcogenides were determined from UV–Vis absorption spectra using Tauc plots, according to the relation:

(αhν)ⁿ = A(hν - E_g)

Here, α is the absorption coefficient (cm⁻¹), hν the photon energy (eV), A is a constant, and n = 2 for a direct transition. By plotting (αhν)²vs. (hν) and extrapolating the linear portion of the curve to the x-axis, the band gap value E_g (eV) can be obtained as the intercept [40,46,47,49].

The optical band gap energies, calculated for CdS_0.5Se_0.5, CdS_0.5Te_0.5, and CdSe_0.5Te_0.5, were 2.01 eV, 1.72 eV, and 1.53 eV, respectively, as presented in Figure 12. The value obtained for CdS_0.5Se_0.5 fits perfectly between the values of the binaries, while being slightly higher when compared to the values of 1.8 – 1.95 eV reported in literature [48,49]. In contrast, the measured band gap value of CdS_0.5Te_0.5 is slightly lower than the values of 1.88–1.94 eV previously reported [50,55], though still in reasonable agreement. Finally, the result measured for CdSe_0.5Te_0.5 (= 1.53 eV) matches perfectly with ≈ 1.55 eV, reported in [56]. These findings support the expected tunability of band gaps in ternary cadmium chalcogenides and further confirm the successful synthesis of the targeted compositions.

4. Conclusions

Binary and ternary cadmium chalcogenides were successfully synthesized from elemental precursors using a mechanochemical approach. Powder XRD and DSC analyses confirmed the complete formation of CdS, CdSe, and CdTe within 15 min of milling, with phase purity for CdS and CdSe achievable under an inert atmosphere to prevent oxidation. Ternary alloys, namely, CdS_0.5Se_0.5, CdS_0.5Te_0.5, and CdSe_0.5Te_0.5, were successfully synthesized and their compositions confirmed by pXRD and AAS measurements, while their diffraction patterns indicate the formation of homogeneous solid solutions, free from detectable oxide impurities.

TEM imaging revealed that the products consist of irregular aggregates of several hundred nanometers, composed of smaller spherical primary nanoparticles in the 10–20 nm range, consistent with crystallite sizes estimated from XRD. DLS measurements in aqueous suspensions indicated larger hydrodynamic diameters, as expected due to particle solvation and aggregation. UV–Vis spectroscopy demonstrated tunable optical band gaps in the ternary systems, with measured values in good agreement with literature reports and consistent with expected bowing effects.

Overall, mechanochemical synthesis proves to be a simple, rapid, and environmentally friendly method for preparing binary and ternary cadmium chalcogenides, avoiding solvents, toxic precursors, and high-temperature treatments. The tendency of the products to form aggregates may be mitigated in future work through the use of surfactants or optimized milling protocols, enabling improved control over particle size and dispersion for potential applications in optoelectronic devices.