Calcium Hexaboride Synthesis from Anhydrous Colemanite by Mechanochemical Method

Abstract

In this study, calcium hexaboride (CaB₆) was successfully synthesized from anhydrous colemanite (Ca₂B₆O₁₁) via a mechanochemical approach. The synthesis process was optimized in two stages: by adjusting the reaction time and varying the carbon-to-colemanite ratio. Structural and compositional analyses were performed using FT-IR, XRD, SEM, and EDX techniques. The optimal synthesis condition was found to be a carbon-to-colemanite ratio of 10:1 and a reaction time of 1020 min, yielding the highest Ca–B vibrational intensities in FT-IR spectra. The mechanochemical method enabled CaB₆ formation at ambient conditions without the need for high temperature or pressure, offering a significant advantage over traditional methods. The results suggest that this method can serve as a low-energy route for the synthesis of metal borides, which are promising candidates for applications in refractory materials, electronics, and hydrogen storage systems.

1. Introduction

Metal borides are a versatile family of compounds formed by boron and metallic elements, exhibiting exceptional hardness, high melting points, and chemical stability [1,2]. Recent studies highlight their potential in thermoelectric devices, magnetic storage, and metallurgical applications, where borides act as grain refiners, deoxidizers, and neutron absorbers [3,4]. Among them, calcium hexaboride (CaB₆) crystallizes in a cubic framework of boron octahedra with calcium atoms in interstitial positions, combining semiconducting behavior, low density, and high hardness [5,6,7]. These features make CaB₆ attractive for protective coatings, electrodes, and high-temperature materials processing [8,9,10]. In addition, boron-based compounds have also been investigated as catalysts for hydrogen generation from NaBH₄, further underlining their diverse energy-related applications [11]. Recent work has also reported enhanced density and functional performance in hexaboride ceramics [12].

Mechanochemical synthesis has recently gained renewed attention as a solvent-free, energy-efficient approach for advanced ceramics [12]. This renewed interest has been comprehensively reviewed in recent literature [13]. Previous reports describe CaB₆ production by mechanochemical activation of Ca/B₂O₃ blends [14], high-temperature sintering routes [15], and colemanite-based carbothermal methods involving transient B₄C phases [16]. These studies highlight the diversity of mechanochemical and thermal approaches; however, none of them address the direct, ambient-condition mechanochemical synthesis from anhydrous colemanite, which is the novelty of the present work. Figure 1 shows the schematic representation of the cubic CaB₆ crystal structure, where covalently bonded boron octahedra are stabilized by calcium atoms.

Traditionally, CaB₆ has been synthesized via high-temperature routes such as carbothermal, aluminothermic, or magnesiothermic reduction, often requiring temperatures above 1500 °C [17,18,19]. Electrochemical reduction in molten salts, combustion synthesis [20], and floating zone crystal growth [21]. have also been reported. Similar floating-zone growth studies have been carried out on SmB₆ single crystals [22]. Although these techniques yield CaB₆ with varying levels of purity and crystallinity, they generally involve high energy consumption, expensive equipment, or prolonged processing times.

Recently, mechanochemical synthesis has emerged as an alternative, energy-efficient method for producing non-oxide ceramics. High-energy ball milling enables solid-state reactions through repeated particle collisions, inducing phase transformations at much milder conditions [23]. In this context, the present study proposes, for the first time, the room-temperature synthesis of CaB₆ using anhydrous colemanite as the boron source, carried out under an argon atmosphere at ambient pressure. It is hypothesized that oxygen released from colemanite reacts with carbon during milling, generating gaseous by-products and triggering phase transformations that ultimately yield CaB₆ through direct atomic bonding. This approach not only avoids high-temperature processing but also demonstrates the potential of utilizing a low-cost, naturally abundant boron mineral for functional boride synthesis [24].

2. Material and Method

Calcium hexaboride (CaB₆) was synthesized at room temperature under ambient pressure via a mechanochemical route. To prevent oxygen contamination, all preparations were performed in a controlled-atmosphere workstation, Plas-Labs Simplicity 888, (Plas-Labs, Lansing, MI, USA) under an argon environment. Analytical grade activated charcoal (Merck) was used as the carbon source, while 99% pure anhydrous colemanite was supplied by ETİ Mining Operations (Ankara, Türkiye). In all experiments, colemanite acted as the limiting reagent and carbon was added in excess.

High-energy ball milling (Gazi University, Ankara, Türkiye) was carried out using a twelve-slot, three-dimensional impact mill operating at 1400 rpm. Milling jars were loaded with balls of various sizes to prevent dead-volume formation, and the ball-to-powder mass ratio was fixed at 4:1 [25]. Two sets of experiments were conducted: (i) optimization of the colemanite-to-carbon ratio, and (ii) determination of the required milling duration. Mixtures with different carbon/colemanite ratios were prepared under argon and analyzed by FT-IR to identify the optimal composition. Once the best ratio was established, samples collected at different milling times were characterized to determine the ideal reaction time. Throughout the experiments, milling speed, temperature (room temperature), pressure (ambient), and atmosphere (argon) were kept constant.

For FT-IR characterization, 0.01 g of ground sample was mixed with 1.99 g KBr in an agate mortar, pressed into pellets under 8 tons of pressure for 3 min, and analyzed using a Perkin Elmer Spectrum 65 FT-IR spectrometer (Perkin Elmer, Shelton, CT, USA) at Osmaniye Korkut Ata University Central Laboratory and a Jasco 480 Plus FT-IR spectrometer (Jasco, Cremella, Italy) at Gazi University, Department of Chemical Engineering). Each analysis was repeated at least three times to ensure reproducibility. Absorbance peak intensities were used to evaluate reaction progress and determine optimum synthesis conditions.

The as-milled product was purified to remove unreacted carbon and oxide by-products (CaO, B₂O₃). Approximately 8 g of the material was leached in 200 mL of 0.5 M HCl for 3 h with continuous stirring, followed by calcination in a Protherm muffle furnace (Protherm Furnaces, Ankara, Türkiye) at 900 °C for 3 h to burn off residual carbon. After purification, the samples were rinsed thoroughly with deionized water and dried.

For structural characterization, XRD analyses were conducted by Rigaku D 220 (Rigaku, Tokyo, Japan) at ETİ-MADEN R&D Center. Prior to analysis, the powders were sintered at 350 °C for 2 h under argon to promote crystallization. Diffraction patterns were obtained to confirm phase composition and lattice formation. SEM-EDX investigations were carried out at Atatürk University Central Laboratory using a Zeiss/FEI scanning electron microscope equipped with an EDX system (Carl Zeiss, Oberkochen, Germany); Atatürk University, Erzurum, Türkiye.

Semi-quantitative phase analysis was performed by comparing the observed XRD peak positions and relative intensity ratios (I/Imax) with reference CaB₆ patterns. FT-IR peak areas were baseline-corrected to estimate the relative contribution of Ca–B and B–O bands, and EDS was used to verify the Ca/B stoichiometry. No full Rietveld refinement was attempted. Although full Rietveld refinement was not attempted due to instrumental limitations, semi-quantitative evaluation of peak intensities was performed. Relative intensity ratios confirmed CaB₆ as the dominant phase, while CaO and B₂O₃ were identified as minor secondary products. A semi-quantitative phase composition based on integrated peak areas suggested that CaB₆ accounted for approximately 70–75 wt.%, whereas CaO and B₂O₃ together contributed about 25–30 wt.%. The Ca/B ratio estimated from EDX (~1:6) further supports the predominance of CaB₆ formation.

3. Results and Discussion

The optimal synthesis conditions were determined using FT-IR spectroscopy. Samp-les exhibiting the largest FT-IR peak areas were identified as the most favorable, as shown in

Table 1. Effect of the carbon-to-colemanite ratio (C: colemanite, CO₂ basis) on the FT-IR integrated peak area at a fixed reaction time of 840 min.

Experiment	Carbon/Colemanite Ratio	Time (min)	FT-IR Peak Area (a.u.)
Set A	5:1 (stoichiometric)	840	402
Set B	10:1 (excess, 2× sto.)	840	420
Set C	15:1 (excess, 3× sto.)	840	335
Set D	20:1 (excess, 4× sto.)	840	260

and

Table 2. Effect of reaction time on the FT-IR integrated peak area at the optimal carbon-to-colemanite ratio (10:1; 2× stoichiometric, CO₂ basis).

Experiment	Time (min)	10:1 (2× Stoichiometric, CO2 Basis)	FT-IR Peak Area (a.u.)
Set 1	600	10:1	2920
Set 2	800	10:1	3090
Set 3	960	10:1	3145
Set 4	1020	10:1	3149

. Peak intensities were calculated after baseline correction. Table 1. Effect of reaction time on FT-IR peak area at a fixed carbon/colemanite ratio 1.1.

At a constant carbon/colemanite ratio, the FT-IR peak area increased steadily with reaction time, reaching a maximum at 1020 min (Table 2). Beyond this duration, no further improvement was observed, indicating a reaction plateau. When the reaction time was fixed at 840 min and the carbon ratio was varied, the most favorable conditions were obtained with a 10:1 (2× stoichiometric, CO₂ basis) carbon-to-colemanite ratio (Table 1), which yielded the highest FT-IR peak area (3149 a.u.). This result indicates that excess carbon enhances molecular collisions, promoting the complete conversion of colemanite to CaB₆, although excessive carbon also reduced powder homogeneity and optical clarity, representing a limitation of the process. The optimized experimental conditions were therefore determined as a carbon-to-colemanite ratio of 10:1 and a reaction time of 1020 min. Nevertheless, due to the limited permeability caused by excess carbon, post-synthesis purification was essential. Calcination at 900 °C and subsequent acid washing successfully removed residual carbon and oxides.

3.1. XRD Analysis

XRD patterns of the sintered products (350 °C for 2 h under an argon atmosphere) are shown in Figure 2 and Figure 3. A relatively low sintering temperature was deliberately chosen to promote crystallization of the mechanochemically formed phases while preventing grain growth and the formation of unwanted carbide phases. Distinct diffraction peaks corresponding to CaB₆ (PDF# 73-1703) were observed, together with secondary reflections from CaO (PDF# 48-1467) and B₂O₃ (PDF# 06-0297). The CaB₆ peaks matched the cubic structure with space group Pm3̅m, consistent with reference data [18,24,25,26]. These PDF numbers were used to confirm phase identification, ensuring consistency with reference standards, and a semi-quantitative phase composition further suggested that CaB₆ accounted for approximately 70–75 wt.%, whereas CaO and B₂O₃ together contributed about 25–30 wt.%. The overall reaction for the mechanochemical synthesis of CaB₆ from anhydrous colemanite and carbon can be represented as

Ca₂B₆O₁₁ + 5C→CaB₆ + CaO + 5CO₂

(1)

In practice, minor residual CaO and B₂O₃ phases may remain due to incomplete conversion; depending on local conditions, a fraction of carbon may also be oxidized to CO₂.

The presence of CaO and B₂O₃ indicates that oxygen released from colemanite was only partially consumed by carbon, a finding that aligns with previous high-temperature carbothermal and magnesiothermic studies [19]. This confirms that while mechanochemical synthesis enables CaB₆ formation at room temperature, complete suppression of oxide by-products remains a general challenge in boride synthesis.

A sintering temperature of 350 °C was selected to enable XRD phase identification while minimizing grain growth and preventing the formation of secondary carbide phases. This temperature was considered adequate for structural analysis, as it allowed crystallization of CaB₆ without promoting additional reactions. The diffraction patterns were refined to suppress background noise, and characteristic peaks of both the desired product and the by-products were clearly resolved in Figure 3. The reflections of CaB₆ confirmed the formation of a cubic structure (Pm3̅m), while the accompanying peaks of CaO and B₂O₃ further indicated partial oxidation, consistent with earlier reports on boride syntheses from natural boron minerals [18,24,25,26].

The identification of CaO and B₂O₃ confirms that oxygen released from colemanite was only partially consumed during the reaction and became stabilized as oxide by-products. The diffraction peaks of CaB₆ correspond to a cubic structure (Pm3̅m), in agreement with reference data [18,24,25]. The coexistence of CaB₆ with CaO and B₂O₃ highlights a limitation of the method: although mechanochemical synthesis can proceed at room temperature, the complete elimination of oxide phases cannot be achieved. Similar by-products have been reported in conventional carbothermal and magnesiothermic syntheses, indicating that oxide formation is a general challenge in CaB₆ production rather than a drawback specific to the present method [26]. It is acknowledged that the optimized milling duration (1020 min, ~17 h) is relatively long compared to conventional furnace-based syntheses. However, the absence of high-temperature processing (≥1500 °C) and furnaces makes the overall energy consumption more competitive. For industrial applications, process intensification through higher-efficiency mills or optimized milling parameters could further reduce the required time, making the method more scalable.

In addition to XRD, FT-IR spectroscopy was employed to further characterize the reaction products. Comparison of the spectra with literature reports showed that peak areas varied depending on reactant concentration, reflecting changes in Ca–B bond intensities. The sample with the largest peak area was considered to represent the most favorable synthesis conditions, and its spectrum is presented in Figure 4.

3.2. FT-IR Spectroscopy

The FT-IR spectrum of the optimized sample is shown in Figure 4.

The FT-IR spectrum of the synthesized sample exhibited several absorption bands in the 500–4000 cm⁻¹ region. Distinct bands at 685 and 730 cm⁻¹ were attributed to B₂O₃, in agreement with previous reports [14,16,23,24]. Characteristic Ca–B vibrational modes were concentrated in the 500–1500 cm⁻¹ sub-region, and peaks at 1435 and 1160 cm⁻¹ confirmed Ca–B stretching vibrations [14,23,24]. These values are consistent with previously reported spectra of CaB₆ and boron oxides, confirming the successful formation of the target phase alongside minor by-products. Peak areas were integrated after baseline correction to enable semi-quantitative comparison of Ca–B bond formation, and the procedure is illustrated in Figure 5.

The main reflections at ~32.3°, ~45.0°, and ~65.7° (2θ, Cu Kα) correspond to CaB₆, while weak features at ~27–28° indicate minor boron oxides. Relative intensity ratios I(200)/I(110) are compatible with reference CaB₆ patterns within experimental variation.

The identification of CaO and B₂O₃ confirms that oxygen released from colemanite was only partially consumed during the reaction and became stabilized as oxide by-products. The diffraction peaks of CaB₆ correspond to a cubic structure (Pm3̅m), in agreement with reference data [18,24,25]. The coexistence of CaB₆ with CaO and B₂O₃ highlights a limitation of the method: although mechanochemical synthesis can proceed at room temperature, the complete elimination of oxide phases cannot be achieved. Similar by-products have been reported in conventional carbothermal and magnesiothermic syntheses, indicating that oxide formation is a general challenge in CaB₆ production rather than a drawback specific to the present method [26].

Peak-area ratios of Ca–B (1435, 1160 cm⁻¹) versus B–O (685–730 cm⁻¹) confirm that CaB₆ is the dominant phase, with only trace B₂O₃ by-products.

In addition to XRD, FT-IR spectroscopy was employed to further characterize the reaction products. Comparison of the spectra with literature reports showed that peak areas varied depending on reactant concentration, reflecting changes in Ca–B bond intensities. The sample with the largest peak area was considered to represent the most favorable synthesis conditions, and its spectrum is presented in Figure 5.

A two-stage optimization was employed. In the first stage, the carbon content was varied at a fixed reaction time of 840 min to identify the optimal carbon-to-colemanite ratio. In the second stage, keeping this optimal ratio (10:1; 2× stoichiometric on a CO₂ basis) constant, the reaction time was varied to determine the optimum duration.

EDS-derived Ca/B ratios are close to the theoretical 1:6 stoichiometry, further corroborating CaB₆ formation.

FT-IR peak areas served as semi-quantitative indicators of Ca–B bond formation. As shown in Figure 6, the carbon-to-colemanite ratio was varied at a fixed reaction time of 840 min. The highest peak area was obtained at 10:1 (2× stoichiometric, CO₂ basis), confirming that an excess of carbon is necessary to drive the reaction. Comparable requirements for surplus carbon have been reported in carbothermal syntheses of other metal borides [26], confirming that a stoichiometric excess of reductant is a general necessity in boride formation.

As shown in Figure 7, the reaction time was varied while keeping the optimal carbon-to-colemanite ratio (10:1; 2× stoichiometric, CO₂ basis) constant. The highest FT-IR peak area (3149 a.u.) was obtained at 1020 min, which was identified as the optimum duration. Beyond this point, no further increase was observed, indicating that the reaction had reached a plateau. The combination of a 10:1 (2× stoichiometric, CO₂ basis) and a reaction time of 1020 min resulted in the highest FT-IR peak area (3149 a.u.).

Because the presence of excess carbon reduced the permeability of the mixture and hindered further reaction progress, post-synthesis purification was applied. Residual carbon was removed by calcination at 900 °C in a muffle furnace, and remaining metal oxides were subsequently eliminated by washing the product with 0.5 M HCl solution.

XRD and FT-IR analyses consistently confirmed the formation of CaB₆ under these optimized conditions. The FT-IR spectra (Figure 8) revealed characteristic Ca–B absorption peaks at 1435 cm⁻¹ and 1160 cm⁻¹, which are in good agreement with the reported range of 1430–1125 cm⁻¹ for CaB₆ [14,23,24]. These results provide strong evidence that the mechanochemical method successfully produced crystalline CaB₆, while also corroborating earlier structural reports in the literature.

In addition to Ca–B vibrations, B₂O₃-related peaks were also detected in the FT-IR spectra. Previous studies have reported B₂O₃ absorptions in the 500–1000 cm⁻¹ region [14,16,23]. In the present analysis, distinct bands were observed at 685 cm⁻¹ and 730 cm⁻¹, which were attributed to B₂O₃. Although other studies reported slightly different positions, such as 685 cm⁻¹ and 738 cm⁻¹ [14,23], the observed values remain within acceptable alignment and confirm the presence of oxide by-products. This finding further supports the XRD results, which indicated residual CaO and B₂O₃ as unavoidable secondary phases.

Figure 9 presents the SEM images of the purified CaB₆ sample at three magnifications. At 10 µm scale (Image A), dispersed crystallites were visible, forming sprout-like clusters. At 1 µm scale (Image B), the particles appeared as aggregated grains with clearly defined boundaries. At 200 nm scale (Image C), individual crystallites displayed the characteristic cubic morphology of CaB₆. The morphology and particle size range (200 nm–1 µm) are consistent with previous reports on mechanochemical and carbothermal CaB₆ syntheses [18,19].

Figure 10 shows the EDX spectrum of the purified product.

The EDX spectrum (Figure 10) confirmed the elemental composition of the purified sample, indicating the presence of calcium, oxygen, and boron. Although the boron signal was clearly visible, its quantification is considered semi-quantitative due to the low atomic number of boron and its limited X-ray emission efficiency. Nevertheless, the simultaneous detection of Ca and B, together with residual O, is consistent with the formation of CaB₆ alongside minor traces of CaO. Similar limitations in accurately quantifying boron by EDX have been reported in previous studies [18,19]. Semi-quantitative EDX analysis indicated the presence of Ca, B, and O. After normalization, the Ca/B ratio was close to the theoretical 1:6 stoichiometry, whereas oxygen was attributed to residual CaO/B₂O₃ phases. Although boron quantification remains challenging due to its low atomic number, the combined XRD, FT-IR, and EDX data consistently verify CaB₆ as the main phase.

4. Conclusions

In this study, calcium hexaboride (CaB₆) was successfully synthesized from anhydrous colemanite via a mechanochemical activation method. The product was characterized using FT-IR, XRD, SEM, and EDX techniques. The optimum synthesis conditions were determined as a colemanite-to-carbon ratio of 1:10 and a reaction duration of 1020 min, which yielded the highest FT-IR peak area, consistent with the proposed reaction pathway.

The results demonstrated that the mechanochemical method enables CaB₆ formation at ambient temperature and pressure, eliminating the need for high-temperature furnaces or pressurization steps that are typically required in conventional carbothermal or magnesiothermic syntheses [14,15,16]. This one-step process significantly reduces energy consumption and production costs. The observed FT-IR vibrational bands of Ca–B bonds and the cubic CaB₆ reflections identified by XRD were in agreement with literature data [24,25,26,27], further validating the effectiveness of this route.

Nevertheless, certain limitations were also identified. Excess carbon, while necessary to ensure complete conversion of colemanite, reduced powder homogeneity and required post-synthesis purification. In addition, oxide by-products such as CaO and B₂O₃ were consistently detected, indicating incomplete oxygen removal—a challenge also frequently reported for high-temperature syntheses [26]. These findings highlight that while the proposed method provides energy efficiency, additional optimization in purification and reaction control is still required.

The study also suggests that this mechanochemical approach can be extended to the synthesis of other metal borides. Since borides are inherently oxygen-free, they can serve as efficient precursors for high-value boron-based materials such as boron nitride and boron hydrides, thereby minimizing the need for further purification steps. In addition, the excellent thermal and chemical stability of borides makes them suitable for refractory and high-temperature applications.

Overall, our findings complement recent studies highlighting the role of mechanochemistry and advanced crystal growth techniques in boride research [12,13,21].

In summary, the ambient-condition mechanochemical synthesis pathway presented here offers a practical and energy-efficient alternative to conventional high-temperature boride synthesis techniques while also providing insight into both the advantages and inherent limitations of this approach.