The Impact of Redox Atmosphere on the High-Temperature Melting Behavior of Basalt Materials

Abstract

This study systematically reveals the fundamental mechanisms controlling redox-induced phase transformations occurring in basalt melting processes via integrated high-temperature redox experiments combined with thermodynamic simulations. Our findings demonstrate that oxidizing conditions drive clinopyroxene dissolution and concurrent crystallization of refractory phases—hematite [(Fe,Ti,Al)₂O₃] and magnesioferrite [(Mg,Fe)(Fe,Al)₂O₄]—where distinct crystallization pathways govern magnesioferrite morphology evolution. Conversely, reducing environments suppress oxide mineral formation while promoting phase transformation from high-melting-point plagioclase to low-melting-point clinopyroxene solid solutions, thus lowering the system’s liquidus temperature to achieve full melting. This provides a theoretical basis for optimizing energy consumption in basalt fiber production and offers new insights into the effects of material melting temperature.

1. Introduction

Continuous basalt fiber (CBF), derived from basalt through high-temperature melting, exhibits superior performance compared with traditional fibers, making it applicable to various fields such as road construction, energy conservation and environmental protection, aerospace, military equipment, and marine ships [1,2,3,4,5,6,7,8]. The preparation of CBF primarily involves melting basalt, where the melting temperature and duration are essential technological parameters [9,10]. The melting behavior of basalt is largely determined by its chemical and mineral composition [11,12]. Additionally, the presence of a reducing atmosphere can significantly influence the melting temperature and the quality of the final fiber by altering the valence state and content of iron ions.

Research indicates that adding redox agents to basalt raw materials or heat-treating CBF in a reducing atmosphere significantly affects fiber quality and the behavior of iron ions. Specifically, the tensile strength of fibers increased approximately linearly with the rise in Fe²⁺ content [13,14,15]. The valence state and content of iron ions were critical factors influencing the crystallization behavior of silicate systems. Fe²⁺ was recognized for its capacity to lower the initial crystallization temperature [15], while Fe³⁺ functioned as nucleating agents, thereby enhancing crystallization ability [16,17]. Moreover, the crystalline phases present in silicate systems were dependent on the environmental conditions. In oxidizing conditions, feldspar was the dominant crystalline mineral, while pyroxene was more prevalent in reducing conditions [18,19,20,21].

Notably, the majority of studies primarily utilized supplementary redox agents in their experimental designs [13,15,18,22,23], and focusing mainly on basalt in its glassy state. This focus has resulted in neglect of a systematic investigation into the melting process of basalt. Furthermore, to our knowledge, little research has been conducted on the effect of atmospheric changes during the CBF preparation process. During the melting of basalt, variations in oxidizing and reducing atmospheres significantly affect multivalent ions (Fe), and the resulting changes in iron ions in turn exert a pronounced influence on the melting process. Importantly, implementing lower melting temperatures and shortening melting durations in industrial manufacturing processes contribute to a decrease in production costs. Producing CBF under reducing conditions not only has the potential to enhance physical properties but also to lower the melting temperature and duration. Therefore, it is crucial to study the influence of different atmospheric conditions on the melting behavior of basalt.

This study utilizes Cenozoic basalt from Hainan Province, China, as the raw material. Initially, the chemical and mineral compositions of the basalt were analyzed. Subsequently, melting experiments were conducted under varying atmospheric conditions and temperatures. The apparent Gibbs free energy for mineral melting and crystallization processes was calculated, and the results from thermodynamic simulations were compared with experimental findings. These experiments and analyses provide important insights into variations in basalt melting processes and mineral transformation under varying conditions. The findings offer a theoretical foundation for enhancing the production process of continuous basalt fibers.

2. Materials and Methods

2.1. Material

The experimental samples are basalt from the Duowen Formation (Qp²d) of the Pleistocene in northern Hainan Province, China (Figure 1). Two samples, B1 and B2, were selected, both of which are dense, massive basalt (Figure 2).

2.2. Preparation Method of Melting Samples

The melting experiment was conducted using a SAFTherm STM-20-17 high-temperature furnace (SAF Therm, Luoyang, China). Seven grams of 200-mesh basalt powder were placed in a corundum crucible. The samples were heated from 50 °C to the target temperature and held for 1 h. Subsequently, to preserve the molten state, the samples were promptly removed from the furnace and quenched in cold water. Previous research indicated that the primary minerals in basalt typically began to melt at approximately 1100 °C, with a significant liquid phase becoming apparent around 1200 °C [12]. Therefore, the experimental temperatures were set at 1100 °C, 1200 °C, and 1300 °C. To further investigate the influence of atmospheric variations on the high-temperature melting behavior of basalt, an additional temperature point of 1250 °C was included in the experimental design. Detailed information regarding the melting temperature, atmospheric conditions, and experimental number for each sample is provided in

Table 1. Experimental conditions and sample number.

Atmosphere	Sample	Melting Temperature
		1100 °C	1200 °C	1250 °C	1300 °C
Oxidizing	B1	B1-O-1100	B1-O-1200	B1-O-1250	B1-O-1300
	B2	B2-O-1100	B2-O-1200	B2-O-1250	B2-O-1300
Reducing	B1	B1-R-1100	B1-R-1200	B1-R-1250	-
	B2	B2-R-1100	B2-R-1200	B2-R-1250	-

.

In order to create a reducing atmosphere during the melting process, the basalt raw material was positioned within a small covered crucible, which was subsequently placed inside a larger covered crucible containing an excess of graphite powder (Figure 3). The graphite powder, weighing 15 g and with a particle size of 200 mesh, met the standards for analytical reagents. The atmospheric air is inherently oxidizing. Therefore, no additional conditions were required to maintain an oxidizing atmosphere; the same apparatus could be used without the inclusion of graphite powder. For details on the applicability and theoretical basis of this reduction device, please refer to Supplementary Material S1.

2.3. Analytical Methods

2.3.1. ICP-OES Element Analysis

The analysis of major elements in basalt samples was performed utilizing a 5300 DV ICP-OES (PerkinElmer, Waltham, MA, USA) at Chengdu Spectra Analysis Technology Co., Ltd. The findings are presented in

Table 2. Chemical compositions of basalt B1 and B2 (wt%).

Sample Number	SiO2	TiO2	Al2O3	Fe2O3	FeO	MnO	MgO	CaO	Na2O	K2O	P2O5	LOI	Total
B1	53.12	1.91	13.74	5.78	4.85	0.14	5.64	8.46	3.32	1.02	0.33	1.65	99.96
B2	53.44	1.65	13.43	5.19	5.5	0.18	6.60	9.00	3.04	0.62	0.28	0.28	99.21

. The Fe²⁺ content was determined through potassium dichromate oxidation titration, with detailed methodologies provided in article [25]. The Fe³⁺ content was calculated by subtracting the Fe²⁺ from the total iron content, with detailed methodologies provided in Supplementary Material S2.

2.3.2. X-Ray Diffraction

The mineralogical composition and crystalline structure of both the raw material and the melted samples were analyzed at the Chengdu University of Technology’s X-Ray Powder Diffraction Lab using a D8 ADVANCE XRD (BRUKER, Berlin, Germany). The testing conditions included a voltage of 40 kV and a current of 40 mA, with a diffraction slit of 0.6 mm, with a Cu target as the X-ray source. Various scanning rates were utilized depending on the sample type: 6°/min for raw materials, 0.8°/min for melting samples with significant diffraction peaks, and 0.2°/min for melting samples with a higher liquid phase, specifically B1-O-1300, B2-O-1300, B1-R-1200, B2-R-1200, B1-R-1250, and B2-R-1250. Quantitative phase analysis was conducted using TOPAS software (V6.0) (Bruker AXS, Germany) through Rietveld refinement methodology.

2.3.3. Scanning Electron Microscopy

The crystal morphology and composition of the melting samples were examined at the Scanning Electron Microscope Laboratory of Chengdu University of Technology using an FEI Nova NanoSEM 45 0 (FEI, Hillsboro, OR, USA) and EDAX AXF-650 APOLLO (EDAX, Berwyn, IL, USA). The instrument was operated at an accelerating voltage of 30 kV and a beam current of 140 μA. Due to the non-conductive nature of the basalt glass samples, a gilding process was required prior to SEM analysis.

3. Results

3.1. Raw Material Analysis

The chemical compositions of the two basalt samples are presented in Table 2. However, slight weathering in sample B1 caused the formation of a small amount of montmorillonite, leading to a minor Loss on Ignition (LOI). The test result error is less than 1%.

The primary minerals in both samples are plagioclase and clinopyroxene, with ilmenite as an accessory mineral (Figure 4). The quantitative analysis of the samples is presented in

Table 3. XRD analysis result of basalt B1 and B2 (%).

Sample Number	Plagioclase	Clinopyroxene		Ilmenite	Montmorillonite	Amorphous Phase
		Pigeonite	Diopside
B1	63.60	17.91	10.91	1.65	0.55	5.38
B2	55.64	21.19	12.04	2.46	-	8.68

.

3.2. Variation in Valence States of Iron

To evaluate the reducing capability of the experimental apparatus, this study compared the valence distribution and content differences of iron elements in molten samples between the experimental group (reducing atmosphere) and control group (oxidizing atmosphere) under 1250 °C conditions (Supplementary Material S2, Table 1). As shown in Figure 5 (including error analysis), the reducing group exhibited a significant increase in Fe²⁺ proportion, while the oxidizing group was predominantly composed of Fe³⁺. These results demonstrate that the experimental apparatus effectively regulates reaction atmospheres, promoting Fe³⁺ conversion to Fe²⁺ under reducing conditions, while facilitating Fe³⁺ formation under oxidizing conditions. The experimental data conclusively validate the system’s capacity to create reductive environments, providing empirical evidence for its application in high-temperature reduction processes.

3.3. Melting Samples Analysis

Figure 5 presents images of basalt samples B1 and B2, which were cooled after melting under varying atmospheric conditions and temperatures, along with their corresponding SEM images. In an oxidizing atmosphere, the samples exhibit notable changes as the temperature increases. At 1100 °C (Figure 6A), the sample appears brick red with a loose granular texture. As the temperature rises to 1200 °C (Figure 6B), the sample transitions to a glassy state, although fine-grained minerals remain visible on its surface. By 1250 °C (Figure 6C), there is a significant decline in crystalline minerals. Ultimately, at 1300 °C (Figure 6D), the sample is fully melted.

Conversely, under a reducing atmosphere, the melting characteristics differ. At 1100 °C (Figure 6E), the sample appears entirely black. As the temperature increases to 1200 °C (Figure 6F), the molten sample exhibits almost a glassy phase. Unfortunately, the results of the XRD analysis (Figure 7) indicate the presence of a minor quantity of minerals. However, a significant transformation occurs at 1250 °C (Figure 6G), where the sample is fully melted into a liquid phase, and the surface displays a glassy luster.

Figure 7 displays the XRD patterns of basalt samples B1 and B2, which were cooled after melting under different atmospheric conditions and temperatures. Under oxidizing conditions, when the basalt sample was heated to 1100 °C, plagioclase, clinopyroxene, pseudobrookite (formed through oxidation of ilmenite), and hematite were identified. Detailed characterization of the ilmenite oxidation process can be found in reference [26]. As the temperature increased to 1200 °C, clinopyroxene completely vanished, and magnesioferrite emerged. Further heating to 1250 °C led to the disappearance of plagioclase, leaving only minor amounts of hematite and magnesioferrite. At 1300 °C, the sharp peaks were no longer present, signifying that the basalt had entirely melted into disordered amorphous glass structures.

Under reducing conditions, the melting process of basalt was characterized by gradual reduction and the disappearance of mineral diffraction peaks with increasing temperature. At 1200 °C, ilmenite that had not undergone reduction disappeared (with no detectable diffraction peaks corresponding to reduction products such as rutile (TiO₂) or metallic Fe [27]). Complete melting of all mineral phases was achieved at 1250 °C.

To evaluate the reliability of the experimental results, repeated experiments were conducted using identical basalt under consistent conditions to establish a control group, with XRD experiments being replicated. The results of this control group are presented in Supplementary Material S3. Phase variations and content changes observed in the control group results were closely aligned with those from the initial experiment, demonstrating the reliability of the experimental findings.

Figure 8 shows the SEM results of the molten samples from B1 and B2 under different conditions. As the temperature rises, mineral content decreases. A sample at 1200 °C in a reducing atmosphere has lower mineral content than one in an oxidative atmosphere. The reducing sample becomes fully glassy at 1250 °C, while the oxidative sample requires 1300 °C.

4. Thermodynamic Calculation

With rising temperatures, the primary minerals in basalt—plagioclase and clinopyroxene—undergo melting reactions. These reactions are detailed in Equations (1) to (11). With increasing temperature, the primary minerals in basalt—plagioclase and clinopyroxene—undergo melting reactions, as described in Equations (1) to (4). Under oxidizing conditions, iron-titanium oxides and potentially involved spinel-group minerals crystallize from the melt, as delineated by chemical reaction Equations (5) and (10). Under reducing conditions, the primary reactions encompass the melting of plagioclase and clinopyroxene, alterations in the valence states of iron ions, and the formation of Ca-Tschermak (11). The study’s calculations are conducted under isobaric conditions with varying temperatures. The corresponding apparent Gibbs free energy calculations, which are crucial for understanding these reactions, are comprehensively detailed in Equations (12) to (26) [28,29,30].

NaAlSi₃O₈(s) = 2.5SiO₂(l) + 0.5Al₂O₃(l) + 0.5Na₂SiO₃(l)

(1)

CaAl₂Si₂O₈(s) = SiO₂(l) + Al₂O₃(l) + CaSiO₃(l)

(2)

CaFeSi₂O₆(s) = 0.5SiO₂(l) + 0.5Fe₂SiO₄(l) + CaSiO₃(l)

(3)

CaMgSi₂O₆(s) = 0.5SiO₂(l) + 0.5Mg₂SiO₄(l) + CaSiO₃(l)

(4)

Fe₂SiO₄(l) + 0.5O₂(g) = Fe₂O₃(s) + SiO₂(l)

(5)

Fe₂SiO₄(l) + TiO₂(g) = FeTiO₃(s) + SiO₂(l)

(6)

0.5Mg₂SiO₄(l) + Fe₂SiO₄(l) + 0.5O₂(g) = MgFe₂O₄(s) + 1.5SiO₂(l)

(7)

3Fe₂SiO₄(l) + O₂(g) = 2FeFe₂O₄(s) + 3SiO₂(l)

(8)

Fe₂SiO₄(l) + 2Al₂O₃(l) = 2FeAl₂O₄(s) + SiO₂(l)

(9)

Mg₂SiO₄(l) + 2Al₂O₃(l) = 2MgAl₂O₄(s) + SiO₂(l)

(10)

Al₂O₃(l) + CaSiO₃(l) = CaAl₂SiO₆(s)

(11)

At standard atmospheric pressure, the apparent free energy of the chemical reaction is defined by the following:

$${\Delta \mathrm{G}}_{\mathrm{T}}={\sum }_{\mathrm{i}=1}^n{{\mathrm{N}}_{\mathrm{i},\mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}}\Delta \mathrm{G}}_{\mathrm{i},\mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}}^{\mathrm{T}}-{\sum }_{\mathrm{i}=1}^n{{\mathrm{N}}_{\mathrm{i},\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{t}}\Delta \mathrm{G}}_{\mathrm{i},\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{a}\mathrm{n}\mathrm{t}}^{\mathrm{T}}$$

(12)

The calculation of apparent free energy for each substance is categorized into minerals (${\mathrm{G}}_{\mathrm{m}\mathrm{i}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{l}}^{\mathrm{T}}$), liquids (${\mathrm{G}}_{\mathrm{melt}}^{\mathrm{T}}$), and gases (${\mathrm{G}}_{\mathrm{g}\mathrm{a}\mathrm{s}}^{\mathrm{T}}$).

$$\begin{array}{cc}{\Delta \mathrm{G}}_{\mathrm{m}\mathrm{i}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{l}}^{\mathrm{T}}=& {\Delta \mathrm{H}}_{\mathrm{f}}+{\int }_{298}^T{\mathrm{C}}_{\mathrm{P},\mathrm{m}\mathrm{i}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{l}}\mathrm{d}T+{\int }_{{\mathrm{P}}_{\mathrm{r}}}^T\left[{\mathrm{V}}^{{\mathrm{P}}_{\mathrm{r}}}-T{\left(\partial \mathrm{V}/\partial \mathrm{T}\right)}_{\mathrm{P}}\right]\mathrm{d}P \\ & -T\left[{\mathrm{S}}_{298}^{\mathsf{\theta }}+{\int }_{298}^T\left({\mathrm{C}}_{\mathrm{P},\mathrm{m}\mathrm{i}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{l}}/T\right)\mathrm{d}T-{\int }_{298}^T{\left(\partial \mathrm{V}/\partial \mathrm{T}\right)}_{\mathrm{P}}\mathrm{d}P\right]\end{array}$$

(13)

$$\begin{array}{cc}{\Delta \mathrm{G}}_{\mathrm{m}\mathrm{e}\mathrm{l}\mathrm{t}}^{\mathrm{T}}=& ({\Delta \mathrm{H}}_{\mathrm{f}}+{\int }_{298}^{{\mathrm{T}}_{\mathrm{f}\mathrm{u}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}}}{\mathrm{C}}_{\mathrm{P},\mathrm{m}\mathrm{e}\mathrm{l}\mathrm{t}}\mathrm{d}T+{\mathrm{T}}_{\mathrm{f}\mathrm{u}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}}\Delta {\mathrm{S}}_{\mathrm{f}\mathrm{u}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}}+{\int }_{{\mathrm{T}}_{\mathrm{f}\mathrm{u}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}}}^T{\mathrm{C}}_{\mathrm{P}}^{\mathrm{l}\mathrm{i}\mathrm{q}}\mathrm{d}T) \\ & -T\left[{\mathrm{S}}_{298}^{\mathsf{\theta }}+{\int }_{298}^{{\mathrm{T}}_{\mathrm{f}\mathrm{u}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}}}\left({\mathrm{C}}_{\mathrm{P},\mathrm{m}\mathrm{e}\mathrm{l}\mathrm{t}}/T\right)\mathrm{d}T+\Delta {\mathrm{S}}_{\mathrm{f}\mathrm{u}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}}+{\int }_{{\mathrm{T}}_{\mathrm{f}\mathrm{u}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}}}^T\left({\mathrm{C}}_{\mathrm{P}}^{\mathrm{l}\mathrm{i}\mathrm{q}}/T\right)\mathrm{d}T\right]\end{array}$$

(14)

$${\mathrm{C}}_{\mathrm{P},\mathrm{m}\mathrm{i}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{l}}={\mathrm{C}}_{\mathrm{P},\mathrm{m}\mathrm{e}\mathrm{l}\mathrm{t}}={\mathrm{k}}_0+{{\mathrm{k}}_{1}T}^{-0.5}+{\mathrm{k}}_2{T}^{-2}+{\mathrm{k}}_3{T}^{-3}$$

(15)

$${\mathrm{G}}_{\mathrm{g}\mathrm{a}\mathrm{s}}^{\mathrm{T}}={\Delta \mathrm{H}}_{\mathrm{f}}-T{\mathrm{S}}_{298}+{\int }_{298}^T{\mathrm{C}}_{\mathrm{P},\mathrm{g}\mathrm{a}\mathrm{s}}\mathrm{d}T-T{\int }_{298}^T\left({\mathrm{C}}_{\mathrm{P},\mathrm{g}\mathrm{a}\mathrm{s}}/T\right)\mathrm{d}T+\left[{\mathrm{V}}^{{\mathrm{P}}_{\mathrm{r}}}+\mathsf{\alpha }\mathrm{V}\left(T-298\right)\right]P-\left(\mathsf{\beta }\mathrm{V}/2\right)P^2$$

(16)

$${\mathrm{C}}_{\mathrm{P},\mathrm{g}\mathrm{a}\mathrm{s}}=\mathrm{a}+\mathrm{b}T+\mathrm{c}{T}^{-2}+{\mathrm{d}T}^{-0.5}$$

(17)

In the process of mineral melting, hematite, magnesioferrite and magnetite involve polymorphic transitions. The apparent free energy of polymorphic transitions is added to the free energy function given by Equation (12). Note that if T is greater than the transition temperature, ${\mathrm{T}}_{\mathsf{\lambda }}$, the integrations in (19) and (20) are performed between ${\mathrm{T}}_{\mathrm{r}}$ and ${\mathrm{T}}_{\mathsf{\lambda }}$, under standard atmospheric pressure ${\mathrm{T}}_{\mathrm{d}}=0$.

$${\mathrm{C}}_{{\mathrm{P}}_{\mathsf{\lambda }}}=T{\left({\mathrm{l}}_1+{\mathrm{l}}_{2}T\right)}^2$$

(18)

$${{\Delta }_{\mathsf{\lambda }}\mathrm{H}}^{\mathrm{T}}={\int }_{{\mathrm{T}}_{\mathrm{r}}}^T{\mathrm{C}}_{{\mathrm{P}}_{\mathsf{\lambda }}}\mathrm{d}\mathrm{T}={\mathrm{x}}_1\left(T{-\mathrm{T}}_{\mathrm{r}}\right)+{\displaystyle \raisebox{1ex}{${\mathrm{x}}_2$}\!\left/ \!\raisebox{-1ex}{$2$}\right.}\left(T^2-{\mathrm{T}}_{\mathrm{r}}^2\right)+{\displaystyle \raisebox{1ex}{${\mathrm{x}}_3$}\!\left/ \!\raisebox{-1ex}{$3$}\right.}\left(T^3-{\mathrm{T}}_{\mathrm{r}}^3\right)+{\displaystyle \raisebox{1ex}{${\mathrm{x}}_4$}\!\left/ \!\raisebox{-1ex}{$4$}\right.}\left(T^4-{\mathrm{T}}_{\mathrm{r}}^4\right)$$

(19)

$${{\Delta }_{\mathsf{\lambda }}\mathrm{S}}^{\mathrm{T}}={\int }_{{\mathrm{T}}_{\mathrm{r}}}^T\left({\mathrm{C}}_{{\mathrm{P}}_{\mathsf{\lambda }}}/T\right)\mathrm{d}T={\mathrm{x}}_1\left(\ln T-\ln {\mathrm{T}}_{\mathrm{r}}\right)+{\mathrm{x}}_2\left(T{-\mathrm{T}}_{\mathrm{r}}\right)+{\displaystyle \raisebox{1ex}{${\mathrm{x}}_3$}\!\left/ \!\raisebox{-1ex}{$2$}\right.}\left(T^2-{\mathrm{T}}_{\mathrm{r}}^2\right)+{\displaystyle \raisebox{1ex}{${\mathrm{x}}_4$}\!\left/ \!\raisebox{-1ex}{$3$}\right.}\left(T^3-{\mathrm{T}}_{\mathrm{r}}^3\right)$$

(20)

$${{\Delta }_{\mathsf{\lambda }}\mathrm{G}}^{\mathrm{T}}={{\Delta }_{\mathsf{\lambda }}\mathrm{H}}^{\mathrm{T}}-T{{\Delta }_{\mathsf{\lambda }}\mathrm{S}}^{\mathrm{T}}$$

(21)

$${\mathrm{x}}_1={\mathrm{l}}_1^2{\mathrm{T}}_{\mathrm{d}}+2{\mathrm{l}}_1{\mathrm{l}}_2{\mathrm{T}}_{\mathrm{d}}^2+{\mathrm{l}}_2^2{\mathrm{T}}_{\mathrm{d}}^3$$

(22)

$${\mathrm{x}}_2={\mathrm{l}}_1^2+4{\mathrm{l}}_1{\mathrm{l}}_2{\mathrm{T}}_{\mathrm{d}}+3{\mathrm{l}}_2^2{\mathrm{T}}_{\mathrm{d}}^2$$

(23)

$${\mathrm{x}}_3=2{\mathrm{l}}_1{\mathrm{l}}_2+3{\mathrm{l}}_2^2{\mathrm{T}}_{\mathrm{d}}$$

(24)

$${\mathrm{x}}_4={\mathrm{l}}_2^2$$

(25)

The free energy function at temperatures above T_λ can be computed with the following:

$${\Delta \mathrm{G}}_{\mathrm{m}\mathrm{i}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{l},\mathsf{\lambda }}^{\mathrm{T}}={{\Delta }_{\mathsf{\lambda }}\mathrm{G}}^{\mathrm{T}}-\left(T-{\mathrm{T}}_{\mathsf{\lambda }}\right){{\Delta }_{\mathsf{\lambda }}\mathrm{S}}^{\mathrm{T}}$$

(26)

${\Delta \mathrm{H}}_{\mathrm{f}}$, ${\mathrm{S}}_{298}^{\mathsf{\theta }}$, and ${\mathrm{V}}^{{\mathrm{P}}_{\mathrm{r}}}$ are enthalpy, third law entropy, and molar volume, respectively, at the reference pressure (${\mathrm{P}}_{\mathrm{r}}=1 \mathrm{b}\mathrm{a}\mathrm{r}$) and temperature ($\mathrm{T}=298.15 \mathrm{K}$), ${\mathrm{T}}_{\mathrm{f}\mathrm{u}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}}$ is the liquid temperature, $\Delta {\mathrm{S}}_{\mathrm{f}\mathrm{u}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}}$ is the phase transition entropy, ${\mathrm{C}}_{\mathrm{P}}^{\mathrm{l}\mathrm{i}\mathrm{q}}$ is the liquid phase heat capacity, ${\mathrm{N}}_{\mathrm{i}}$ is the coefficient of substance i in the reaction formula, k₀, k₁, k₂, k₃, a, b, c and d are molar heat capacity polynomial coefficients, $\mathsf{\alpha }\mathrm{V}$ and $\mathsf{\beta }\mathrm{V}$ are the coefficients of thermal expansion and compressibility respectively multiplied by molar volume, which are the relevant thermodynamic calculation parameters to see

Table 4. Thermodynamic data for various mineral end-members [28,29].

Mineral	Formula	${\mathsf{\Delta }\mathbf{H}}_{\mathbf{f}}$ (kJ·mol−1)	${\mathbf{S}}_{298}^{\mathsf{\theta }}$ (J·K−1·mol−1)	k0	k1 × 10−2	k2 × 10−5	k3 × 10−7
Albite	NaAlSi3O8	−3921.618	224.412	393.64	−24.155	−78.928	107.064
Anorthite	CaAl2Si2O8	−4213.249	207.223	439.37	−37.341	0	-31.702
Diopside	CaMgSi2O6	−3200.583	142.5	305.41	−16.049	−71.66	92.184
Hedenbergite	CaFeSi2O6	−2845.389	171.431	307.89	−15.973	−69.925	93.522
Hematite	Fe2O3	−822	87.4	146.86	0	−55.768	52.563
Ilmenite	FeTiO3	−1231.947	108.628	150	−4.416	−33.237	34.815
Magnesioferrite	MgFe2O4	−1406.465	122.765	196.66	0	−74.922	81.007
Magnetite	FeFe2O4	−1117.403	146.114	207.93	0	−72.433	66.436
Hercynite	FeAl2O4	−1947.681	115.362	235.19	−14.37	−46.913	64.564
Spinel	MgAl2O4	−2300.313	84.535	235.9	−17.666	−17.104	4.062
Ca-Tschermak	CaAl2SiO6	−3298.767	140.751	310.7	−16.716	−74.553	94.878

,

Table 5. Thermodynamics data for melt components [28,29].

Melt	${\mathbf{T}}_{\mathbf{f}\mathbf{u}\mathbf{s}\mathbf{i}\mathbf{o}\mathbf{n}}$ (K)	$\mathsf{\Delta }{\mathbf{S}}_{\mathbf{f}\mathbf{u}\mathbf{s}\mathbf{i}\mathbf{o}\mathbf{n}}$ (J·K−1·mol−1)	${\mathbf{C}}_{\mathbf{P}}^{\mathbf{l}\mathbf{i}\mathbf{q}}$	${\mathsf{\Delta }\mathbf{H}}_{\mathbf{f}}$ (kJ·mol−1)	${\mathbf{S}}_{298}^{\mathsf{\theta }}$ (J·K−1·mol−1)	k0	k1 × 10−2	k2 × 10−5	k3 × 10−7
SiO2	1999	4.46	81.373	−901.554	48.475	127.3	−0.00010777	4.3127	−0.00004638
TiO2	1870	35.824	109.2	−944.75	50.46	77.84	0	−33.678	40.294
Al2O3	2320	48.61	170.3	−1675.7	50.82	155.02	−8.284	−38.614	40.908
Fe2SiO4	1490	59.9	240.2	−1479.36	150.93	248.93	−19.239	0	−13.91
Mg2SiO4	2163	57.2	271	−2174.42	94.01	238.64	−20.013	0	−11.624
CaSiO3	1817	31.5	172.4	−1627.427	85.279	141.16	−4.172	−58.576	94.074
Na2SiO3	1361	38.34	180.2	−1561.427	113.847	234.77	−22.189	0	13.53

The heat capacity formula of SiO₂ in the melt component is Cp = k₀ + k₁T + k₂T⁻² + k₃T^−0.5_.

and

Table 6. Thermodynamics data for gas [30].

Formula	${\mathsf{\Delta }\mathbf{H}}_{\mathbf{f}}$ (kJ·mol−1)	${\mathbf{S}}_{298}^{\mathsf{\theta }}$ (J·K−1·mol−1)	V	a	B × 10−2	C × 10−5	D × 10−7	$\mathit{\alpha }\mathit{V}$ (K−1)	$\mathit{\beta }\mathit{V}$ (kbar−1)
O2(g)	0	205.2	0	48.3	−0.00000691	4.992	−0.00004207	0	0
CO(g)	−110.53	197.67	0	45.7	−0.00000097	6.627	−0.00004147	0	0
CO2(g)	−393.51	213.7	0	87.8	−0.00002644	7.064	−0.00009989	0	0

. ${{\Delta }_{\mathsf{\lambda }}\mathrm{H}}^{\mathrm{T}}$, ${{\Delta }_{\mathsf{\lambda }}\mathrm{S}}^{\mathrm{T}}$ and ${\mathrm{C}}_{{\mathrm{P}}_{\mathsf{\lambda }}}$ represent the enthalpy, entropy, and heat capacity associated with polymorphic transitions at the reference pressure (${\mathrm{P}}_{\mathrm{r}}=1 \mathrm{b}\mathrm{a}\mathrm{r}$) and temperature ($\mathrm{T}=298.15 \mathrm{K}$). ${\mathrm{T}}_{\mathrm{r}}$ (usually 298.15 K) and ${\mathrm{T}}_{\mathsf{\lambda }}$ denote the temperatures of first-order and second-order (lambda) polymorphic transitions, respectively, ${\mathrm{l}}_1$ and ${\mathrm{l}}_2$ are the parameters for polymorphic transitions, as shown in

Table 7. Thermodynamics data for gas [29].

Mineral	Formular	${\mathbf{T}}_{\mathbf{r}}$ (K)	${\mathbf{T}}_{\mathbf{\lambda }}$ (K)	${\mathbf{l}}_1$× 102 (J/mol)−0.5 K−1	${\mathbf{l}}_2$ × 105 (J/mol)−0.5 K−1
Hematite	Fe2O3	298	955	−7.403	27.921
Magnesioferrite	MgFe2O4		665	15.236	−53.571
Magnetite	FeFe2O4		848	−19.502	61.037

.

Figure 9 presents the calculated results of the apparent Gibbs free energy for the aforementioned reactions. The melting reaction temperatures for NaAlSi₃O₈, CaAl₂Si₂O₈, CaMgSi₂O₆, and CaFeSi₂O₆ are determined to be 1035.15 K (797 °C), 1226.15 K (953 °C), 1070.15K (982 °C), and 1255.15 K (762 °C), respectively. Therefore, the primary melting temperatures of the principal minerals, plagioclase and clinopyroxene, are affected by their compositional variations, with ranges spanning from 1070.15 K (797 °C) to 1226.15 K (953 °C) for plagioclase and from 1035.15 K (762 °C) to 1255.15 K (982 °C) for clinopyroxene. However, melting remains challenging within the specified temperature range. To address this issue, it is advisable to elevate the temperature beyond the material’s melting point. In the case of silicate minerals, they generally do not melt rapidly until they are heated to approximately 100-150 K above their melting point [31,32]. This additional heating supplies the requisite energy to surmount the thermodynamic and kinetic barriers, thereby enhancing the melting process. It is recommended to maintain the temperature between 1373.15 K (1100 °C) and 1473.15 K (1200 °C) to promote melting [12]. Equations (5)–(10) are pertinent to the reactions occurring in the melt, where the value of ${\Delta \mathrm{G}}_{\mathrm{T}}$ is consistently less than zero, indicating a tendency for crystallization within the melt. For details on potential thermodynamic calculations of ilmenite-hematite solid solutions and spinel-group mineral solid solutions, refer to References [33,34,35,36,37].

5. Discussion

5.1. Influence of Atmosphere on Melting Temperature of Basalt

Based on the experimental and control group results, the phase and content changes of basalt during its melting process under different atmospheres were analyzed, with the results shown in Figure 10. Both basalt samples began to exhibit a significant presence of the liquid phase at 1100 °C under varying atmospheric conditions. However, there is a notable difference in temperature when the liquid content reaches 100%. In an oxidizing environment, the liquid phase composition attained 100% at 1300 °C, while in a reducing environment, this occurred at 1250 °C.

To further investigate the impact of atmospheric conditions on the melting process of basalt, thermodynamic simulations were conducted using the software FactSage 8.3 (GTT-Technologies, Herzogenrath, Germany and Thermfact/CRCT, Montreal, QC, Canada). The major element composition of the two samples, as detailed in Table 2, was selected for the simulations. The simulation conditions included a temperature range from 400 °C to 1400 °C, with equilibrium points established at every 100 °C interval. The pressure was maintained at 101.325 kPa, equivalent to standard atmospheric pressure. For the oxidation condition, the atmosphere was set to 21% O₂ and 79% N₂, while for the reduction condition, the atmosphere was set to 21% CO and 79% N₂. The FToxid oxide database was utilized for these simulations. The results of these simulations are presented in Figure 11.

Figure 11 shows the simulation results; the numerical simulation and experimental results are consistent in phase in terms of phase transformation. In an oxidizing atmosphere, iron-containing oxides form during melting, whereas in a reducing atmosphere, the melting temperature decreases.

This temperature decline results from the inhibition of Fe²⁺ oxidation to Fe³⁺, which limits the crystallization of high-melting-point minerals such as hematite and magnesioferrite. Additionally, experiments and simulations interestingly demonstrate that the melting behavior of clinopyroxene and plagioclase is significantly influenced by different atmospheric conditions. In experimental results, the melting trend of plagioclase becomes more pronounced under reducing conditions (Figure 12A). This trend is further supported by simulation results, which indicate a decrease in the disappearance temperature of plagioclase to 1200 °C (Figure 12B). Conversely, under oxidizing conditions, the melting of clinopyroxene becomes more significant (Figure 12C), while simulations show a decrease in its content (Figure 12D). It is essential to examine the alterations in the primary mineral contents of basalt, as plagioclase and clinopyroxene play a crucial role in influencing both the initial liquidus temperature and the melting temperature. Specifically, these temperatures tend to increase with a higher content of plagioclase and a lower content of clinopyroxene [12].

Therefore, melting basalt in a reducing atmosphere not only inhibits the crystallization of high-melting-point minerals (hematite and magnesioferrite) but also promotes the melting of plagioclase (a high-melting-point mineral). Based on XRD, SEM, and simulation results, compared to an oxidizing atmosphere, the melting temperature of basalt under reducing conditions can be effectively reduced by approximately 50 °C.

5.2. The Influence of Redox Atmosphere on the Crystallization Behavior of Basalt Materials During the Melting Process

The melting processes of basalt exhibit significant variations under different atmospheric conditions. Under oxidizing conditions, the primary rock-forming mineral clinopyroxene in basalt undergoes preferential melting, while hematite and magnesioferrite crystallize within the basaltic melt. In contrast, reducing conditions promote enhanced melting susceptibility of plagioclase, another principal rock-forming mineral in basaltic systems.

5.2.1. Influence of the Oxidizing Atmosphere on Basalt Melting Process

The discussion on melting mechanisms primarily encompasses two aspects: structural mechanisms and compositional mechanisms. The structural mechanism refers to crystal disintegration induced by crystal defects under external conditions [38,39,40], while the compositional mechanism involves elemental migration through dissolution/melting and diffusion processes [41,42,43,44]. Under the combined effects of these two mechanisms, clinopyroxene exhibits enhanced melting susceptibility under oxidizing conditions.

Recent studies propose that redox reactions can induce dislocation generation [45]. Significantly influenced by oxidation, Fe²⁺ in the clinopyroxene lattice oxidizes to Fe³⁺ during melting, as shown in Equation (27) [46]. This defect generation intensifies crystal disintegration and facilitates the breakage of bridging oxygen bonds, thereby promoting the melting of iron- and magnesium-rich clinopyroxene (pigeonite) (Figure 13①).

4Fe²⁺+O₂ → 4Fe³⁺+2O²⁻

(27)

The diffusion in silicate minerals can be categorized into volume diffusion and grain-boundary diffusion [41,42,43,44], where grain-boundary diffusion includes elemental migration at mineral-melt interfaces. During clinopyroxene melting, locally Fe-Mg-enriched melts form (Figure 13②). However, the iron and magnesium elements in the molten liquid phase are subsequently consumed to form magnesioferrite (as evidenced by octahedral granular magnesioferrite crystals observed on clinopyroxene surfaces with distinct cleavage planes in Figure 14A). The depletion of Fe-Mg elements promotes diffusion processes, thereby accelerating clinopyroxene melting.

Therefore, synthesizing the discussion of these two melting mechanisms, we conclude that the oxidation of Fe²⁺ to Fe³⁺ within the lattice under oxidizing conditions facilitates metal cations in overcoming lattice constraints. Simultaneously, the migration of Fe-Mg elements further accelerates the melting of Fe-Mg-rich clinopyroxene (pigeonite). As a result, at a temperature of 1100 °C the iron-rich pigeonite disappears, and the calcium-rich diopside remains, as illustrated in Figure 7 and Figure 10.

As confirmed by previous studies [47,48,49,50], iron in basaltic melts undergoes oxidative enrichment and subsequently crystallizes as hematite (Figure 13④). This process is quantitatively described by the chemical equilibria in Equations (5) and (6). Combined with Figure 14B,C, it is evident that hematite crystallized during the melting of basalt contains minor titanium and aluminum. Our experimental results demonstrate that sheet-shaped hematite microcrystals form through this crystallization mechanism at both 1200 °C (B) and 1250 °C (C), with their characteristic morphology documented in Figure 13. Magnesioferrite exhibits an octahedral shape (Figure 14A) or a triangular shape (Figure 14C), with their distinct morphologies being linked to their respective formation processes. In melts derived from clinopyroxene, Fe²⁺ is oxidized to Fe³⁺, accompanied by localized enrichment of Fe and Mg, leading to the formation of octahedral granular magnesioferrite (Figure 13③). The triangular sheet magnesioferrite is formed by the counter-diffusion process of Mg²⁺ within a hematite lattice (Figure 13⑤), and the chemical transformation is depicted in Equation (28) [51]. Additionally, combined with Figure 14B,C, it is evident that the magnesioferrite crystallized during the melting of basalt contains minor aluminum.

3Fe₂O₃ + 2Mg²⁺ → 2MgFe₂O₄ + 2Fe²⁺ + 0.5O₂

(28)

The iron-rich liquid phase present in the air is susceptible to oxidation, which facilitates the formation of hematite and magnesioferrite [15]. Similarly, basalt, characterized as a silicate system abundant in iron and aluminum, inherently faces challenges in avoiding crystallization. This crystallization tendency is evident both during cooling [52,53,54] and even during the melting process.

5.2.2. Influence of the Reducing Atmosphere on Basalt Melting Process

In a reducing atmosphere, plagioclase melts more easily because it transforms into clinopyroxene. Specifically, anorthite, a component of plagioclase, can decompose into Al- and Ca-rich clinopyroxene, known as the Ca-Tschermak molecule, as described in Equation (24) [52,53,54]. Ca-Tschermak molecule tends to form the stable solid solution of clinopyroxene [55], which can be synthesized at approximately 1200 °C under standard atmospheric pressure [56]. During our process, plagioclase incorporates into the diopside solid solution in the form of a Ca-Tschermak molecule, as illustrated in Figure 15 and represented by chemical Formulas (29) and (30). Therefore, at 1100 °C, clinopyroxene appears on the surface of plagioclase as short columnar microcrystals (Figure 16). During silicate depolymerization, the structural transition from framework silicates to chain silicates can occur [57], in which Mg²⁺ plays a critical role, as evidenced by the preferential incorporation of Mg²⁺ into chain-structured environments. Under high-temperature conditions, the four-membered rings within the framework structure of the plagioclase lattice undergo breakdown. This depolymerization exposes more non-bridging oxygen, which subsequently binds with metallic cations. The main cation in plagioclase are Al³⁺ and Ca²⁺, which contribute to the formation of a Ca-Tschermak molecule. However, plagioclase also contains smaller amounts of divalent cations, such as Mg²⁺ and Fe²⁺. These mental cations can integrate into crystal structures, with Mg²⁺ particularly favoring incorporation into chain silicates [57]. For details on potential thermodynamic calculations of Diopside-Ca-Tschermak solid solutions, refer to References [58,59]. Moreover, The oxidation of mental cations is difficult to occur under reducing atmospheric conditions. This reductive environment facilitates the development of chain silicates, such as clinopyroxene solid solutions. This observation corroborates the hypothesis proposed in Section 5.1: the enhanced basalt melting observed under reducing conditions is linked to plagioclase melting phenomena. This phenomenon occurs because the transformation of plagioclase (a high-melting-point mineral) into clinopyroxene (a low-melting-point mineral) renders basalt more susceptible to melting.

CaAl₂Si₂O₈(s) → CaAl₂SiO₆(s) + SiO₂

(29)

4(Ca,Mg,Fe)(Al,Si)₄O₈(s) + 2CaSiO₃(l)→
3CaAl₂SiO₆·(Ca,Mg,Fe)Al₂SiO₆·Ca(Mg,Fe)Si₂O₆(s) + 4SiO₂(l)

(30)

Under reducing conditions, the melting capacity of clinopyroxene is slightly constrained. This constraint is primarily due to the following two factors: the conversion of some plagioclase into clinopyroxene and the suppression of the transformation from Fe²⁺ to Fe³⁺ by the reducing environment. The CO atmosphere significantly influences the valence state of iron ions [60]. Notably, Fe²⁺ exhibits a greater tendency to integrate into the crystalline structure of clinopyroxene than Fe³⁺ [18]. The preference for Fe²⁺ incorporation restricts the release of iron ions from the clinopyroxene lattice, affecting the overall melting process of clinopyroxene.

6. Conclusions

• In an oxidizing atmosphere, the melting process is hindered due to the crystallization of high-melting-point minerals such as hematite and magnesioferrite. Conversely, a reducing atmosphere can lower the melting temperature by approximately 50. This decrease in melting temperature is attributed to the suppression of high-melting-point minerals (hematite and magnesioferrite), as well as the transformation of plagioclase into lower-melting-point clinopyroxene;
• Atmospheric conditions influence the melting of major minerals. Clinopyroxene melts more readily under oxidizing conditions, whereas plagioclase melts more easily under reducing conditions;
• Sheet-shaped hematite can crystallize from the liquid phase as Fe²⁺ oxidizes to Fe³⁺ at the liquid’s surface. Magnesioferrite crystallization can occur through two mechanisms: one involves the oxidation of Fe²⁺ and its combination with Mg²⁺ to form octahedral granular magnesioferrite, while the other involves the counter-diffusion of Mg²⁺ through the lattice of pre-crystallized hematite. Under reducing conditions, plagioclase partially transforms into the solid solution of diopside-Ca-Tschermak.

Our findings demonstrate that the reduction of melting temperature can effectively decrease energy consumption during CBF production. Furthermore, as evidenced by prior studies, the enhancement of Fe²⁺ content has been demonstrated to improve the physical properties of resultant fibers [13,14,15]. These mechanistic insights collectively validate that the controlled modulation of redox conditions during basalt melting processes constitutes a strategically advantageous approach for optimizing CBF manufacturing efficiency and product quality.