Enhancing crystallization and structural modification in SiO₂–CaO–Al₂O₃ glass-ceramics using mineral-based Mn source

Abstract

This work studies the effects of adding natural manganese ore in amounts between 0 and 15 wt% on the crystallization and microstructure of SiO₂–CaO–Al₂O₃ (SCA) glass-ceramics. The samples were prepared by melting and quenching with feldspar and calcium carbonate as base materials, followed by synthesizing at different temperatures. Characterization and physicochemical tests confirmed that the addition of manganese ore improves the crystalline phase formation of anorthite, rhodonite, and bustamite. Increasing the manganese content in the samples showed higher levels of non-bridging oxygen to total oxygen (NBO/T) ratio, which suggests more network depolymerization of the glass structure, which is known to enhance crystallization. X-ray diffraction and scanning electron microscopy clearly showed that rhodonite and bustamite grains grew larger and sharper at 1000 °C and that new Mn-containing phases formed, proving better long-range order. Physicochemical tests revealed manganese ore-rich samples were of higher hardness, lower water absorption, and higher corrosion resistance than their manganese ore-free counterpart. However, adding too much manganese ore (over 10 wt%) led to excessive grain growth and internal flaws, underscoring the need for careful compositional balance.

Introduction

The glass-ceramics systems, which result from controlled crystallization of parent glasses, have received major focus because of their hybrid properties, ranging from construction to biomedicine^1,2. Among the glass-ceramic systems, the SCA glass-ceramics is particularly pertinent owing to its superior thermal stability, mechanical strength, and chemical durability^3,4.

The tailoring of structural and functional attributes of glass-ceramic matrices has required the use of oxides such as manganese oxide (MnO) and other transition metal oxides. Acting as a glass network modifier, MnO changes the lattice by adding non-bridging oxygens which impacts the crystallization processes along with the resultant microstructure^5,6. Certain research indicates that MnO addition may contribute towards the formation of particular crystalline phases like bustamite and rhodonite, which enhance the mechanical strength of the glass ceramics^7,8.

The glass-ceramics crystallization process is determined by composition, heat treatments, and the actual presence of nucleating agents^9,10. Pace-controlled heating of SCA glasses with included MnO was reported to create favourable crystalline phases, resulting in improved hardness and thermal resistance^11,12.

Scope of widening the industrial by-products usage, especially electrolytic manganese residue (EMR), in the production of value-added products like MnO glass-ceramic is in tandem with sustainable development principles^13,14. This strategy improves the economic value by saving raw materials and addressing environmental issues.

The recent trends in this area are to achieve specific applications of glass-ceramics by optimizing the composition and processing parameters to develop glass-ceramics of higher-order features. The addition of more oxides like TiO₂ and ZrO₂ as nucleating agents has also been studied to modify the microstructure of MnO modified glass-ceramics and improve the performance^15,16. The development of low-temperature co-fired ceramics (LTCC) using SCA systems also created new prospects for electronic applications where MnO is known to significantly control dielectric properties^17,18.

There has been little research on the role of manganese ore in SCA glass-ceramics, especially in relation to crystallization and microstructural processes, despite advancements in the field. Although studies have been conducted on synthetic manganese oxides, the intricate composition of natural ores poses distinct possibilities and challenges that require additional research¹⁹.

This study attempts to fill the gaps by evaluating the impact of incorporating manganese ore (0–15 wt%) on the crystallization and microstructural features of SCA glass-ceramics with emphasis on phase change. The use of manganese ore in glass-ceramics as a nucleating agent has been attempted and its effect on the glass network structure using FTIR, XRD, and SEM-EDS, has been investigated. The results obtained will aid in the formulation of innovative sustainable high-performance glass-ceramics by demonstrating the use of natural feedstock in the synthesis of advanced materials. While prior works have examined synthetic Mn oxides, the innovation of this work lies in using natural manganese ore (a complex, low-cost mineral source containing MnO + Fe₂O₃ + minor impurities) as a sustainable feedstock. This approach addresses real-world challenges of ore complexity, reduces reliance on pure chemicals, and aligns with circular economy goals.

Experimental procedure

Preparation of glass-ceramic samples

In this study, SCA (SiO₂-CaO-Al₂O₃) glass-ceramics were synthesized using two raw materials: (1) feldspar (73 wt%) and (2) calcium carbonate (27 wt%). To simultaneously investigate the effects of manganese oxide and iron oxide, manganese ore was incorporated into the initial composition at 0, 5, 10, and 15 wt%, labelled as M0, M5, M10 and M15, respectively. The chemical compositions of the feldspar and manganese ore are provided in Table 1.

Table 1.

Chemical composition of the feldspar and Mn ore (wt%).

Item	SiO2	Al2O3	CaO	Fe2O3	Na2O	MgO	MnO	LOI
Feldspar	69.74	18.34	0.28	0.16	10.8	-	-	0.74
Mn ore	16.62	2.39	8.26	12.91	1.4	1.57	38	16.7

The raw materials were homogenized in a laboratory ball mill for 1 h. The mixture was then melted in an alumina crucible using an electric furnace at 1450 °C for 2 h, followed by rapid quenching in water to obtain glass frits. The frits were further pulverized in a ball mill to achieve a particle size finer than 200 mesh. Subsequently, 10 g of the milled powder was pressed into a cylindrical steel mold (3 cm diameter) under a uniaxial pressure of 160 bar. The pressed green bodies were synthesized in an electric furnace with a heating rate of 2 °C/min to target temperatures of 800, 900, 1000, and 1170 °C, held isothermally for 1 h at each temperature, to crystallize the samples. A schematic of the preparation process of the glass ceramic samples is illustrated in Fig. 1.

Fig. 1.

Preparation process of the glass ceramic samples.

Characterization techniques

Fourier-transform infrared spectroscopy (FTIR) was performed on the milled glass frits to analyze the glass network structure. X-ray diffraction (XRD, 10–90°, step size 0.02°) was conducted on glass-ceramic samples synthesized at 800, 900, and 1000 °C to identify crystalline phases. For scanning electron microscopy (SEM) coupled with energy-dispersive X-ray spectroscopy (EDS), the glass-ceramic surfaces (heat-treated at 1000 °C) were polished, etched with a 5 vol% hydrofluoric acid (HF) solution for 10 s, and sputter-coated with a thin gold layer to enhance conductivity (The 5 vol% HF etching for 10 s was optimized empirically to selectively reveal crystalline phases without excessive dissolution of the residual glass matrix or over-etching which could obscure morphology). Morphology and chemical composition of crystalline phases were evaluated using SEM-EDS.

The density and water absorption of the glass-ceramic samples were measured via the Archimedes technique. To measure the hardness of the samples, a Vickers hardness tester under conditions of 9.807 N load for 20 s, was used. Corrosion resistance was assessed by measuring weight loss after immersing samples in 20 wt% H₂SO₄ and NaOH solutions at 70 °C for 1 h. These tests were each repeated three times.

Result and discussion

Glass network structure

To investigate structural changes in glass frit networks and their relationship with crystallization behaviour in glass-ceramic materials, Fourier-transform infrared spectroscopy (FTIR) was utilized. Figure 2 presents the FTIR spectra of four glass frit samples (designated M0, M5, M10 and M15), revealing three primary absorption regions attributed to vibrational modes within the glass network. These regions span 400–580 cm⁻¹ (Band No. 1), 580–850 cm⁻¹ (Band No. 2), and 850–1400 cm⁻¹ (Band No. 3).

Fig. 2.

FTIR spectra of the frit samples.

Band No. 1, as described in previous studies²⁰, arises from bending vibrations of X-O-X bonds (X = Si or Al), a signature of the pyroxene phase frequently observed in glass-ceramic systems. Band No. 2 corresponds to bending modes of Si-O-Si and O-Si-O bonds, reflecting the silicate network’s structural framework. The broad and intense Band No. 3, linked to bending vibrations of [SiO₄] tetrahedra with differing numbers of bridging oxygens²¹, offers critical insights into the polymerization state of the glass network, which is a key factor influencing crystallization. For quantitative analysis, Band No. 3 was deconvolved via Gaussian fitting to identify Qⁿ species (Fig. 3), where Qⁿ denotes [SiO₄] tetrahedra with n bridging oxygens (n = 0–4)²². The relative contributions of these species, summarized in Fig. 4, were used to compute the non-bridging oxygen-to-total oxygen ratio (NBO/T).

Fig. 3.

Gaussian deconvolution results for Band No. 3: (a) M0, (b) M5, (c) M10, (d) M15.

Fig. 4.

Relative area fraction of Qⁿ in different samples.

Qⁿ species are assigned to distinct wavenumber ranges: Q⁰ (~ 850 cm⁻¹), Q¹ (~ 900 cm⁻¹), Q² (950–1000 cm⁻¹), Q³ (1050–1100 cm⁻¹), and Q⁴ (~ 1200 cm⁻¹)²³. Higher Q⁴ contributions signify fully polymerized networks (four bridging oxygens), while Q⁰ indicates depolymerized structures. The NBO/T ratio, calculated as (4Q⁰+3Q¹+2Q²+1Q³+0Q⁴)/ΣQⁿ²⁰, increased progressively from M0 (0.73), M5 (0.801), M10 (0.846) to M15 (0.912) which reveals significant alterations in the silicate network structure of the glass-ceramic system upon the incorporation of manganese ore. It seems that this trend is intrinsically linked to the role of transition metal oxides (MnO and Fe₂O₃) introduced by the manganese ore, which act as network modifiers and influence the polymerization state of the [SiO₄] tetrahedral framework. Mn²⁺ and Fe³⁺ ions act as network modifiers, disrupting the silicate network by breaking Si-O-Si bonds^24,25 and generating non-bridging oxygens (NBOs) as Q⁰, Q¹ and Q² increase constantly from M0 to M15. The increased NBO/T ratio directly correlates with enhanced crystallization propensity in the M15 sample. A depolymerized network (high NBO/T) reduces glass viscosity, facilitating ion mobility during heat treatment. This promotes the nucleation and growth of crystalline phases²⁶.

XRD patterns

Figure 5 displays the XRD patterns of the samples synthesized at 800 °C. As can be seen, at 800 °C, all samples (M0–M15) display a fully amorphous structure, characterized by the presence of broad humps in the pattern and the absence of sharp diffraction peaks. This suggests that no crystallization occurred at this temperature, and the glassy phase is dominant in this sample. The absence of crystalline phases even in the samples containing manganese ore indicates that 800 °C is below the crystallization temperature for these compositions⁹. This remains true in the presence of transition metal oxides such as MnO and Fe₂O₃, which, although known to act as network modifiers, are also potential nucleating agents and would suggest that they were not beyond the dissolution limit at this temperature.

Fig. 5.

XRD pattern of the glass ceramic samples at 800 °C.

In contrast, the XRD patterns of the samples synthesized at 900 °C (Fig. 6), the samples show the stages of crystallization. Distinct peaks of complex anorthite (CaAl₂Si₂O₈, ▲), rhodonite (MnSiO₃, ♦), bustamite (CaMnSi₂O₆, ■), and wollastonite (CaSiO₃, o) emerge, as reported by Li et al.¹³. It is noted that these phases are more marked for higher concentrations of MnO, which means manganese ore aids in the formation and growth of silicate crystals at high temperatures. It can also be deduced that anorthite appears as the dominant phase across all compositions at 900 °C due to the high Al₂O₃ and CaO content sourced from feldspar and calcium carbonate.

Fig. 6.

XRD pattern of the glass ceramic samples at (a) 900 °C and (b) 1000 °C.

Rhodonite and bustamite, both Mn-rich silicate phases, are prominently detected in M10 and M15 samples, confirming the successful incorporation of Mn into the silicate lattice.

Wollastonite, which may form at lower manganese ore concentrations, becomes less prevalent as manganese ore content increases, likely due to competitive crystallization with Mn-bearing phases.

In conclusion, the XRD results prove that manganese ore serves as a strong promoter of crystallization in the SCA glass-ceramic system. Crystallization is highly sensitive to temperature and concentration of manganese ore, with important phase changes occurring only over 800 °C.

Comparatively assessing XRD patterns at 900 °C and 1000 °C sheds greater light on the crystallization evolution of the SCA glass-ceramic system modified with manganese ore. Recall from earlier discussion that the synthesis at 900 °C sparks the beginning of the crystallization phenomena attested by the strong yet moderate diffraction peaks of anorthite, wollastonite, rhodonite, and bustamite exhibiting crystallized phases. It is noteworthy that the peaks are not only defined, but broad, pointing to the existence of small and imperfect crystallites in an amorphous matrix. Crystallization during this period is predominantly limited to nucleation, owing to growth being restrained by the greatly viscous parent glass.

The XRD patterns at 1000 °C show higher and sharper peaks for all samples, which confirms improved crystallinity and is a result of ionic mobility caused by thermal activation. In sample M0, the intensity of anorthite peaks increases noticeably while wollastonite can still be observed, but in lesser quantity. The increase in peak intensity reflects the advancement of crystallization in the matrix. Manganese ore additions (M5-M15) show more complex patterns upon addition. Bustamite and rhodonite peaks noticeably grew from M5 and M15, showing that silicates rich in manganese grew and stabilized. Anorthite remains a dominant phase across all samples but is accompanied by stronger bustamite and rhodonite signatures in manganese ore-rich compositions. In fact, XRD patterns clearly indicate the onset of crystallization above 800 °C, with progressive enhancement at 900–1000 °C.

Therefore, at 1000 °C, the thermal energy coupled with the depolymerization of glass networks (supported by FTIR) accelerates the growth of bustamite and rhodonite. Calcium oxide, manganese oxide, and silicon oxide together create these conditions, while transition metals (like Mn) modify the network, which encourages crystallization. It should be noted that the observed structural modifications and enhanced crystallization are primarily attributed to the combined action of transition metal oxides (MnO and Fe₂O₃) from the ore, with MnO playing the dominant role due to its higher concentration and known tendency to form specific Mn-rich phases (rhodonite and bustamite).

Microstructural observation

The SEM images of the samples synthesized at 1000 °C are presented in Fig. 7. Clear differences can be seen in the microstructure of the surface with respect to the increment of manganese ore content, which closely matches the crystallization behaviour as determined from the XRD and FTIR investigations.

Fig. 7.

SEM images of the glass ceramic samples, (a) M0, (b) M5, (c) M10, (d) M15.

In (M0), where manganese ore is absent, finer features of the microstructure consists of a glassy matrix that is more or less uniform, with no distinguishable crystalline features, if any at all. This observation corresponds to the results obtained from the XRD, since the only crystalline peaks that were detected were of anorthite, and those detected were weak. The absence of well-formed crystals suggests that the degree and rate of crystallization was limited due to, as suggested, the relatively high degree of polymerization in the glass network (lower NBO/T ratio) and lack of effective nucleating agents.

The addition of 5 wt% manganese ore (sample M5) appears to bring about the first changes to the microstructure. Crystalline structures begin to appear as small, dispersed angular grains that are distributed in a glassy matrix and are typically isometric in shape. This corresponds with the phase nucleation of bustamite and anorthite, as noted in the XRD data. However, the even more strikingly distributed and undeveloped crystals suggest that the amount of manganese ore at this concentration is not adequate for a wide-scale phase change.

Sample M10 has more prominent structural changes as it displays a higher density of interlocking crystalline structures, which are well defined and show signs of crystallization–nucleation activity and extensive growth of crystals, indicating strong nucleation density. The rod-shaped or slightly elongate crystal that is filled with bustamite and rhodonite phases dominates the microstructure. They were visibly present in the XRD patterns. The constriction of these Mn silicates further confirms that Mn²⁺ ions are not only acting as network modifiers, but also as crystal structure stabilizers. Their presence contributes to breaking the glassy microstructure, resulting in a multi-phase heterogeneous microstructure.

In sample M15, crystallization peaks both in its magnitude and definition. The microstructure is almost completely crystallized, consisting of large, well-defined grains with interconnected growth. These motifs indicate a high level of crystallization with coalescence of multiple silicate phases. The crystalline phases detected in XRD, especially anorthite, bustamite, and rhodonite, are now morphologically distinguishable from one another. Their interlocking nature with fewer residual glassy spaces strongly suggests that these phases were thermodynamically stabilized and grew at 1000 °C²⁴. This coincides with the peak NBO/T ratio retrieved from FTIR, considering the highly depolymerized silicate framework abundant in non-bridging oxygen preferred for crystal development.

Aligning the SEM results reveals significant consistency with the crystallinity developed within XRD and FTIR analyses. The incorporation of manganese ore weakly bonds into the glass network, which greatly modifies its structure, encourages phase segregation and aids the crystalline domains in nucleating and growing. Such structural modification results in a dense, three-dimensional crystalline microstructure. However, as can be seen in Fig. 7d, the addition of 15 wt% manganese ore led to the formation of defects such as cracks and pores. In sample M15, the high manganese ore content accelerates crystallization but results in internal stresses due to thermal expansion mismatch, volumetric shrinkage, and over-nucleation. These effects cause pore and crack formation, indicating that excessive manganese ore (> 10 wt%) degrades microstructure despite improved crystallinity, highlighting the need for an optimal amount of manganese ore.

The elemental analysis of point 1 (Fig. 8) highlighted in the SEM image of sample M5 shows that Si, Ca, Mn, O, and Al were present, with manganese levels much higher than the surrounding glassy matrix. This composition strongly suggests that the analysed region corresponds to a Mn-rich pyroxenoid phase, likely bustamite or rhodonite, which is consistent with the results of XRD. The contributions of both Ca and Mn, together with the high Si concentration, is the evidence for the crystallization of mixed Ca-Mn silicates modulated through the substitution mechanisms whereby Mn²⁺ ions partially substitute for Ca²⁺, in the crystal lattice.

Fig. 8.

EDS analysis of point 1 in Fig. 7b.

Also, the relatively low signals of Na and Fe, despite being present in the starting raw materials, suggest that these elements either remain in the residual glassy phase or have limited the amount of participation in the major crystalline phases under the current heat treatment conditions.

The presence of Al in very small quantities might suggest a substantial solid-solution addition of anorthite, or while being incorporated as Al³⁺ into the silicate framework, which is common in more complex multi-cation glass-ceramics²⁷.

The localized chemical characterization of microstructures confirms the hypothesis of the nucleation and growth of distinct manganese-bearing crystalline domains in the bulk microstructure. Combined with the XRD patterns and SEM imaging, EDS results confirm that the complex silicate phases have been formed through phase separation and selective crystallization processes owing to the incorporation of Mn. This also supports the earlier hypothesis that manganese ore acts as a modifier, considering the results from FTIR and NBO/T calculations, as well as a chemical constituent of stable crystalline silicates.

Hence, the EDS analysis offers convincing evidence concerning the microscale chemistry of the studied samples as well as direct evidence of the hypothesized crystallization mechanism and the thermodynamic stabilization of the studied M10 and M15 silicate compositions incorporating manganese.

The changes in the macroscopic appearance of the glass-ceramic samples upon varying synthesis temperatures qualitatively reflect the internal structural changes and densification processes taking place during the synthesis. As shown in Fig. 9, the samples notable changes in surface color, texture, and surface uniformity (appearance) with the increase in temperature, especially beyond 800 °C.

Fig. 9.

Appearance of the glass ceramic samples at different synthesis temperatures.

The samples appear to remain smooth and glassy at 800 °C, indicating minimal crystallization in an amorphous matrix. This also corresponds to the XRD results, where no crystalline peaks are observable at this temperature.

A change in surface appearance and color is apparent upon heating to 900 °C. The glass, which was previously transparent or semi-transparent, changes to opaque and becomes microscopically rough, indicating crystallization and phase separation within the bulk. This corresponds with the appearance of anorthite, bustamite, and rhodonite phases in the XRD analysis. The opacification is due to light scattering of early-stage crystals that formed in the glassy matrix, which means that it was possible to exceed the critical level for nuclei formation.

At 1000 °C, the surface of the samples is more deeply pitted and opaque; sometimes the color change is very pronounced, for instance, in samples with high Manganese content. These phenomena result from the more advanced crystallization and growth of large, more numerous crystalline domains, as seen in the SEM images. The interlocking crystals that emerge at this stage disrupt the smooth surface of those samples that were previously considered shiny, resulting in a dull, more mineral-like look. Moreover, indeterminate amounts of iron and manganese oxides could be responsible for some of the color through redox reactions when fired, yielding brown, gray, or even pinkish hues depending on the constituents of the sample.

For comparison, the viscous flow of the samples synthesized at 1170 °C can be found in Fig. 10. The different sample properties visually distinguish their degree of viscous flow and surface densification, and, with the addition of more manganese ore, become closely related to the increase in surface. Mn-ore free sample (M0) maintains the smoothest surface and most compact and deformed structure, which indicates the highest viscosity as well as slow ion mobility at higher ranges of temperatures. Sample M5 demonstrates some of the first changes with the addition of 5 wt% manganese ore that increases its roundness and results in a slight increase in diameter, presumably due to partial disruption of the silicate network. M10 is the most pronounced sample along with M0, as it exhibits flow behaviour, with a larger disc formation that corresponds to reduced viscosity and increased ion mobility of network modifiers (Mn²⁺ and Fe³⁺), leading to greater densification and structural reorganization. M15 takes the spot for the largest deformation and diameter expansion at almost 5 cm while demonstrating the roughest surface, which makes it more prone to excessive flow, explaining its high levels of manganese and iron oxides that result in lower viscosity. The increase in flow properties from M0 to M15 corresponds with the increase of NBO/T ratio as seen in FTIR, which confirms the importance of manganese ore in modifying the glass structure and increasing its flowability. The ruler (in cm) clearly demonstrates the increase in diameter, which directly showcases the increase in viscous flow with manganese ore, thus marking direct tangible evidence. It should be noted that at 1170 °C, samples exhibited significant viscous flow and deformation and the focus was on crystallization temperatures (up to 1000 °C) where well-defined microstructures form; higher temperature data (1170 °C) were included only for flow behaviour.

Fig. 10.

Pictures of the glass ceramic samples synthesized at 1170 °C.

Physicochemical properties

In Fig. 11, the density evolution as a function of synthesis temperature for the SCA glass-ceramic system with different additions of Mn ore (0 to 15 wt%) is shown. All the samples, without exception, demonstrated an increasing trend of density with temperature, while the increase was particularly pronounced for the manganese ore-rich compositions (M10 and M15). This behaviour is characteristic of a thermally induced densification process intimately associated with the crystallization kinetics and the level of structural polymerization within the glass matrix.

Fig. 11.

Plot of density vs. temperature for the glass-ceramic samples shown in Fig. 9 (the standard deviation for all measurements was at most ± 0.04 g/mLi).

For the manganese ore-free sample (M0), the density increases between 800 °C and 1000 °C is moderate, which, as indicated by XRD and SEM, is mostly due to anorthite crystallization, suggesting that crystallization is largely confined to anorthite. The NBO/T ratio (0.73) obtained from FTIR signifies that the relatively low non-bridging oxygen to total oxygen ratio indicates a silicate network that possesses NBO-T, suggesting a crystallization phase, indicating an abundant silicate network formation at the expense of ion mobility. Hence, in such a scenario the rate of crystalline phase formation and growth is obstructed thus retards crystallization. Therefore, the predominant controlling factor for densification tends to reside in the glassy phase undergoing sintering rather than in volumetric determining changes crystallization-dominated shrinkage.

The interlocking crystalline phases created in the glassy areas undergo remarkable compaction owing to microstructural lattice coherence and phase agglomeration. This rationalizes the sharp rise in densification of samples M10 and M15 at 1000 °C. This behaviour aligns with the densification mechanism exhibited in silicate glass-ceramics incorporating silicate-type crystal structures, where crystalline features greatly increase volume collapse and lower the remaining pore space²⁴. Despite localized defects of M15 (SEM images), the overall higher crystallinity and denser packing of Mn-rich phases result in net higher density at 1000 °C (crystallization compensates for minor porosity).

The reliability of our density measurements can be verified by their good agreement with the range of 2.6–3 g/cm³ reported by Zhao et al.²⁰ for similar silicate glass-ceramic systems. Also, Gaddam et al.²⁴ demonstrated that the addition of Mn to silicate glasses increases density due to network depolymerization (from 2.2 to 2.5 g/cm³) which is consistent with our observed trend.

Table 2 shows the Vickers hardness, water absorption, and corrosion resistance of SCA glass-ceramics fabricated at 1000 °C with different manganese ore contents. With an increase in the manganese ore content, hardness, porosity, and chemical durability all increased. The Vickers hardness increased from 461 HV in sample M0 to 498 HV in sample M15. As discussed in the XRD and SEM sections, this increase can be explained by the increase in the volume of rhodonite (MnSiO₃) and bustamite (CaMnSi₂O₆) crystals in the glass matrix. These dense Mn-bearing silicates with long-chain structures are widely known to increase hardness in glass-ceramics. Although excessive Mn ore (15 wt%) introduces some microstructural defects (visible in SEM images), the volume fraction and interlocking nature of hard Mn-rich crystalline phases (rhodonite and bustamite, which have inherently high hardness) still dominate, leading to the highest Vickers value (498 HV). Hardness is measured on polished surfaces and reflects local resistance to indentation, which benefits from increased crystallinity even if some porosity exists globally.

Table 2.

Physical and chemical properties of the glass ceramic samples synthesized at 1000 °C.

Sample ID	Vickers hardness (HV)	Water absorption (%)	Corrosion resistance (%)
			Acid resistance	Alkali resistance
M0	461 ± 3	0.45 ± 0. 02	92.5 ± 0.3	98.3 ± 0.2
M5	476 ± 2	0.38 ± 0. 01	94.1 ± 0.4	98.7 ± 0.2
M10	481 ± 3	0.32 ± 0. 02	95.8 ± 0.2	99.3 ± 0.1
M15	498 ± 5	0.32 ± 0. 02	96.3 ± 0.2	99.1 ± 0.1

Water absorption, which shows both porosity and densification, decreases from 0.45% in M0 to 0.32% in M10 and M15, implying a denser structure and more complete crystallization. In fact, higher amounts of manganese ore led to enhanced densification during synthesis.

The corrosion resistance in acidic and alkaline environments was enhanced with an increase in the manganese ore content, i.e., from 92.5% in M0 to 96.3% in M15 in and from 98.3% in M0 to 99.1% in M15 in acid and alkali environments, respectively. The reason can be explained by the formation of crystalline phases of Mn-containing silicates, which are chemically stable and inhibit ion diffusion. In addition, the pathways for chemical attacks are restricted by the increased crystallinity and decreased amount of the glass phase.

Conclusion

This research investigated the effects of adding natural manganese ore as an additive on the crystallization and performance of SiO₂–CaO–Al₂O₃ (SCA) glass-ceramics. The following are the main conclusions of the present study:

• The addition of natural manganese ore to SCA glass-ceramics alters the silicate network, i.e., the extra non-bridging oxygens raise the NBO/T ratio, encourage network depolymerization, and improve crystallization.

• The addition of manganese ore led to the facilitation of stable crystalline phases at 900 °C and especially at 1000 °C, which resulted in a more compact structure, whereas at 800 °C, an amorphous structure was obtained.

• The SEM results confirmed that with an increase in the manganese ore content, a mostly glass-containing matrix in the manganese ore-free sample changed to a microstructure with interlocking crystals. Nonetheless, excessive manganese ore (15 wt%) led to the formation of defects such as cracks and pores.

• The visual appearance of the samples showed that increasing the manganese ore amount and temperature led to improved viscous flow, meaning greater densification and spreading. Excessive manganese ore (15 wt%) caused surface roughness and over-deformation, implying decreased viscosity and possibly structural instability.

• Increasing the manganese ore content enhanced hardness and corrosion resistance and decreased water absorption, possibly due to the formation of Mn-rich crystalline phases with a denser microstructure.

Author contributions

Majid Bagheri: Conceptualization, Methodology, Investigation, Formal analysis, Writing - original draft, Writing - review & editing, Supervision, Project administration.Mostafa Karamouz: Investigation, Data curation, Validation.Mohsen Hajizamani: Methodology, Investigation, Resources, Writing - review & editing.All authors have read and approved the final version of the manuscript.

Data availability

All data generated or analysed during this study are included in this published article.

Declarations

Competing interests

The authors declare no competing interests.