Experimental study of CO₂ sequestration and H₂ generation potential through mineral carbonation in Saudi red mud

Abstract

This study examines the potential of Saudi red mud for CO₂ sequestration through mineral carbonation and its concurrent hydrogen generation. Two batch experiments were conducted under neutral (deionized water) and acidic (1 wt% HCl) conditions at elevated temperature and pressure. Under neutral conditions, the interaction between CO₂ and red mud led to measurable mineralization, with approximately 8.3% of the injected CO₂ converted into Minerals. In contrast, the acidic system induced extensive mineral dissolution and precipitation, releasing higher concentrations of reactive elements into solution. Increasing the amount of red mud in the reaction to twice that used in the DIW case resulted in approximately twice the amount of CO₂ being mineralized. This indicates a strong relationship between red mud mass and its CO₂ sequestration capacity, highlighting that providing more reactive solid material directly enhances the system’s ability to lock CO₂ into stable carbonate minerals. Gas analysis verified hydrogen generation, along with notable hydrogen sulfide formation, particularly under acidic conditions. Overall, the findings reveal that while red mud can facilitate partial CO₂ mineralization and hydrogen evolution, its reactivity strongly depends on pH, with acid-assisted reactions enhancing mineral mobility rather than stable carbonate formation.

Introduction

Rising atmospheric carbon dioxide (CO₂) levels have heightened the urgency of developing effective removal techniques to combat global climate change. Major industries, including power generation, steel production, and cement manufacturing, release vast amounts of CO₂ into the atmosphere daily. As a primary contributor to the greenhouse effect, CO₂ must be efficiently captured and mitigated, particularly from industrial flue gases, to reduce their environmental impact¹. Carbon capture and storage (CCS) technology is widely regarded as a promising solution for addressing climate change. This process entails capturing carbon dioxide emissions from stationary industrial sources, transporting the captured CO₂ via pipelines or ships, and securely storing it in underground geological formations. These storage sites can include depleted oil and gas reservoirs, saline aquifers, coal seams, and unconventional shale formations, effectively preventing the release of CO₂ into the atmosphere^2,3. CO₂ sequestration into subsurface geological formations has gained significant attention in recent years^4,5.

Mineral carbonation (MC) is considered one of the most secure methods for carbon dioxide removal (CDR). The process provides permanent storage and involves transforming CO₂ into carbonate minerals, preventing its release back into the atmosphere. MC is a chemical reaction in which CO₂ reacts with divalent cations like Mg²⁺ and Ca²⁺, which are naturally present in minerals such as serpentinite, olivine (magnesium silicates), and wollastonite (calcium silicate)^6,7,8,9,10. Additionally, they are found in alkaline industrial byproducts, including steelmaking slags, fly ash, cement waste, and red mud^{11,12,13,14,15}. Several million tons of alumina are produced each year globally through the Bayer process, a key method in the aluminum manufacturing industry¹. However, this process generates substantial quantities of red mud, also known as bauxite residue, as a byproduct¹⁶. On average, the production of one ton of alumina generates 1.0–1.5 tons of red mud, resulting in a global accumulation of over 66 million tons of red mud annually from aluminum manufacturing^17,18,19. The disposal of large amounts of such an alkaline sludge is both costly and environmentally challenging. Over the years, research on red mud repurposing has reported its potential as a low-cost and efficient adsorbent for removing heavy metals such as lead, chromium, arsenic, fluoride, cadmium, and zinc from aqueous solutions^20,21,22. Extensive global research has also been conducted to develop cost-effective and sustainable methods for repurposing red mud waste. Researchers have explored various practical applications for red mud, including agriculture, construction, water management, pipeline coating, and waste gas treatment. Red mud has been utilized as an acidic soil amendment and in the production of bricks, ceramics, tiles, glazes, and polymer composite panels as a wood substitute^{23,24,25,26,27}. Other applications include iron-rich cement, radiopaque materials, and CO₂ sequestration using liquid carbon dioxide^28,29,30. Red mud has also been utilized in the oil and gas upstream sector for oil well cement and drilling fluids applications^31,32. Recently, Al-Azani et al. ³³ proposed using red mud-derived nanoparticles as foam stabilizing agents, which proved their effectiveness in improving the stability and morphology of foam.

CO₂ sequestration through mineralization with red mud is another viable approach to repurposing red mud. It offers significant environmental benefits by simultaneously neutralizing the alkaline red mud waste and reducing atmospheric carbon levels. This dual-purpose approach not only addresses the challenges of industrial waste management but also contributes to climate change mitigation by converting CO₂ into stable mineral carbonates. Several studies have assessed CO₂ mineralization using red mud. Ilahi et al. ³⁴ presented a review on the principles and applications of CO₂ sequestration using red mud. Liu et al. ³⁵ also reviewed the application of red mud as a sink for CCUS and the factors influencing carbonation. Yadav et al. ²⁸ investigated CO₂ sequestration using red mud of varying particle sizes, analyzing the effects of reaction time and liquid-to-solid ratio under a constant pressure of 3.5 bars. Their characterization of red mud samples identified the presence of various carbonate minerals, including boehmite, cancrinite, chantalite, hematite, gibbsite, anatase, rutile, and quartz. Among these, chantalite and cancrinite were identified as the primary minerals responsible for the carbonation process in red mud. The highest carbonation was observed in the red mud sample with the highest cancrinite content, achieving a CO₂ uptake of 5.3 g CO₂ per 100 g red mud. Sahu et al. ³⁶ conducted neutralization experiments of red mud through CO₂ sequestration cycles at ambient conditions. The red mud samples were air-dried, ground in a mortar and pestle, sieved through a 1 mm mesh before analysis, and distilled water was used for the experiments. X-ray Diffraction (XRD), Scanning Electron Microscopy (SEM), Energy-Dispersive X-ray Spectroscopy (EDX), Fourier-transform infrared spectroscopy (FTIR) analysis, and auto titration revealed an increase in gibbsite intensity, ilmenite formation, and higher concentrations of Na, C, O, and Si in the carbonated filtrate. The amounts of sequestered CO₂ were 26, 58, 55, and 54 wt% per 10 g of RM in the neutralized red mud and successive filtrate cycles. Han et al.³⁷ investigated the simultaneous neutralization of red mud and atmospheric CO₂ mineralization through both laboratory and field pilot studies. Laboratory experiments were conducted over both short-term and long-term durations. To accelerate the reaction, additional Ca sources, such as flue gas desulfurization (FGD) gypsum or CaCl₂, were introduced as catalysts. The 55-day batch tests and longer-term field experiments showed consistent results, demonstrating that Ca addition significantly enhanced the neutralization rate. Without extra Ca, atmospheric CO₂ primarily reduced pore water alkalinity, whereas Ca addition facilitated further neutralization by promoting mineral carbonation, leading to CaCO₃ formation. Using the ball milling process at different times, Mucsi et al. ³⁸ mechanically activated red mud for enhancing its reactivity with CO₂ for CO₂ mineralization purposes. The reaction was carried out at 25 ℃ and 5 bars. Mechanical activation was found to enhance the CO₂ sequestration capacity of red mud by 1.7 wt%, as demonstrated in the investigations. Additionally, the reaction of red mud slurry with CO₂ led to a reduction in pH, indicating effective carbonation. During the measurements, the pH of the suspension decreased from 10 to 6.81, highlighting the neutralization effect of CO₂ on red mud. Recently, Suman and Tripathy³⁹ employed the high-energy ball milling method to mechanically activate red mud, aiming to enhance its CO₂ sequestration capacity. The study involved evaluating the kinetics of CO₂ sequestration, including adsorption and desorption, using various kinetic models. The results showed that the sequestration capacity of mechanically activated red mud was almost 6 times higher than the feed red mud (i.e., 18.28 mg/g compared to 3.75 mg/g). Avrami’s fractional order model was found to be the most suitable model, indicating that both physisorption and chemisorption play roles in improving the CO₂ sequestration capacity of red mud.

Further publications have also been reported to investigate the reaction conditions and the methods to improve the mineralization process with red mud. Xie et al. ⁴⁰ proposed an electrochemical CO₂ mineralization strategy for red mud treatment using hydrogen-cycled membrane electrolysis. This process efficiently sequesters CO₂ while recovering high-purity (99.4%) NaHCO₃ and sodium resources. The process enables direct mineralization of flue gas CO₂ (15% volume) without extra capture. Zhang et al. ⁴¹ studied the mineralization reaction of CO₂ in red mud under different pressures (i.e., 0–14 bars). The results of increasing the pressure correspond to a gradual increase in the carbonate content in the reaction product until it reaches a stable level. Their results showed that the carbonation reaction occurs as a liquid-phase reaction, with the carbonation products primarily deposited in the solution. Notably, CaCO₃ does not accumulate or form directly on the particle surface, preventing significant alterations in pore structure or excessive carbon deposition on the particle surface. Recently, Duraisamy and Chaunsali⁴² studied the mineralization of red mud at 25, 45, and 65 ℃ and a fixed pressure of 10 bars. They reported that experiments conducted at 45 ℃ resulted in the highest mineralization compared to experiments conducted at other temperatures. They also concluded that due to its low calcium and magnesium content, the red mud used in their study showed limited CO₂ uptake. Wang et al. ⁴³ proposed using carbide slag instead of NaOH to adjust the pH of the leaching solution for CO₂ mineralization in red mud, thereby facilitating the recovery of valuable elements while effectively sequestering CO₂. It was reported that replacing NaOH with carbide slag increased the Ca²⁺ concentration in the filtrate to 21,922 ppm after co-precipitation.

Although previous laboratory-scale studies have highlighted the potential of red mud for CO₂ mineralization, further research is essential to optimize process parameters, overcome current limitations, and achieve economic and operational feasibility for maximum CO₂ storage. Many previous studies have reported only partial carbonation, with CO₂ uptake limited to surface adsorption or the formation of transient carbonate species rather than stable, long-lasting mineral phases. Additionally, the comparative effects of neutral versus acidic conditions on mineral dissolution, precipitation pathways, and overall sequestration efficiency have received little attention. The potential co-production of valuable gases, such as hydrogen, during carbonation has also been largely overlooked, representing a significant research gap in exploring the broader benefits of this process. Hydrogen (H₂), as a clean energy carrier, plays a crucial role in the global energy transition⁴⁴. Its co-production during CO₂ sequestration not only facilitates greenhouse gas storage but also yields a valuable energy resource, thereby enhancing the contribution of CCS to sustainable energy systems. Investigating the generation of natural H₂ during CO₂ sequestration in red mud is well aligned with this global shift⁴⁵. Therefore, the objective of this work is to investigate the mineralization process of CO₂ in red mud and analyze the reaction products in all phases, namely, solid, liquid, and gaseous phases. The red mud waste used in this study was produced from a local alumina production facility in Saudi Arabia. Two reaction experiments were conducted in a confined reaction cell into which a predetermined solid/liquid ratio was introduced. The reaction slurry was prepared by adding the candidate red mud to either deionized water (DIW) (in Experiment 1) or a mixture of DIW and hydrochloric acid (1wt% HCl) (in Experiment 2) in the reaction cell, followed by CO₂ injection. An additional experiment (Experiment 3) was conducted to evaluate the effect of increasing the amount of red mud used in the mineralization process. The mineralization process was evaluated by analyzing the solid, liquid, and gaseous phases collected from the reaction cell after the reaction and comparing them with the original samples measured and characterized before the reaction. The solids were analyzed for their elemental and mineralogical composition (XRF and XRD) and for total carbon content (TC). The liquid was also analyzed for pH, electrical conductivity (EC), and solid and carbon content. Gas chromatography (GC) analyses were performed to monitor changes in gas composition during the reaction. The possibility of hydrogen production from the reaction was also investigated by analyzing the gas content at the end of the reaction tests.

Methodology

Material

The red mud used in this study is a byproduct of the alumina production industry, sourced from a local alumina plant in Saudi Arabia. Filtered deionized water (DIW) from an ELGA DV 25 Integral Water Purification System, with a conductivity of 4.88 µS/cm at room temperature, was utilized as the base liquid for the reaction slurry. Hydrochloric Acid (HCl, 37 wt%), purchased from Sigma-Aldrich, was used to prepare a 1 wt% solution for one of the reactions. High-purity CO₂ gas supplied by a local gas supplier was also used in this work.

Methodology

The Saudi red mud was characterized by elemental and mineralogical analyses using X-ray fluorescence (XRF) and X-ray diffraction (XRD), respectively. The XRD analysis was performed on an EMPYREAN Diffractometer (Malvern PANalytical). Measurements were conducted over a 2θ range of 4° to 70° with an angular step size of 0.01°. The instrument operated at 45 kV and 40 mA, and the collected diffraction data were processed and quantified using the Rietveld refinement method in HighScore Plus software. The system was equipped with a Pixcel1D detector, a reflection–transmission spinner sample stage, and a copper (Cu) X-ray source that generated Kα1 radiation at 1.54060 Å and Kα2 radiation at 1.54443 Å. The analysis revealed that the red mud is primarily composed of hematite, chantalite, and rutile, which will be further discussed in the Results section. The quality of the Rietveld refinements was assessed using the weighted profile R-factor (R_wp) and the goodness of fit (GoF). The R_wp quantifies the overall mismatch between the observed and calculated diffraction profiles, whereas the GoF reflects how well the refinement model accounts for the data relative to the expected statistical variation. Typically, smaller Rietveld error indices suggest a closer match between experimental and calculated diffraction patterns. Nonetheless, there are cases where an inaccurate model combined with poor-quality data can produce deceptively low error values, whereas a correct model with high-quality data may show higher indices. Consequently, these indices should not be relied upon as the only indicator of refinement quality. A visual examination of the fit is crucial to accurately evaluate the reliability of the refinement⁴⁶. However, for complex, multi-phase materials such as red mud, R_wp values below approximately 10 are generally considered acceptable, and GoF values exceeding 1 are common due to the inherent heterogeneity and peak overlap⁴⁷.

Three reaction experiments were conducted in a confined reaction cell into which a predetermined solid/liquid ratio was introduced. The reaction slurry was prepared by adding the candidate red mud with either DIW or hydrochloric acid (1wt% HCl) to the reaction cell, followed by CO₂ injection. The mineralization process was evaluated by analyzing the solid, liquid, and gaseous phases collected from the reaction cell after the reaction and comparing them with the original composition before the reaction. The solids were analyzed for their elemental and mineralogical composition (i.e., XRF and XRD) and for total carbon content. The liquid was also analyzed for its pH, electrical conductivity (EC), and its solid and carbon contents. Gas chromatography analyses were performed to investigate the changes in gas composition resulting from the reaction process. The possibility of hydrogen production from the reaction was also investigated by analyzing the gas content at the end of the reaction tests. Figure 1 illustrates the experimental procedure employed in this study. The following subsections provide a detailed description of the experimental work conducted.

Fig. 1

Flowchart of the experimental work.

Carbonation experiments

This work aimed to evaluate the potential of CO₂ mineralization utilizing Saudi red mud waste. The study began with a control experiment (Experiment 1), which provides a baseline for assessing the sequestration capacity of Saudi red mud. In this experiment, the reaction slurry was prepared by adding red mud and deionized water to a 1,000-ml titanium high-pressure, high-temperature (HPHT) reaction cell at a liquid-to-solid ratio of 14:1, followed by the introduction of CO₂. For this purpose, the reaction experiment was initiated by adding approximately 50 g of the as-received Saudi red mud into the cell, after which 700 mL of DIW was introduced. The cell was then capped and vacuumed to remove any air or gas impurities. CO₂ was then compressed and injected into the cell batch-wise with a gas booster to reach a pressure of 1,000–1,100 psig. Figure 2 shows the contents of the reaction cell during the reaction process. During pressurization, the cell was rocked for several minutes to thoroughly mix and agitate the samples, ensuring a sufficient quantity of CO₂ was dissolved in the liquid phase. The reaction cell was then left to equilibrate and then moved into a heating path (oven) operating at 50 ℃. The pressure of the cell was continuously monitored during heating, and upon thermal stabilization, the pressure reached approximately 2,300 psig. The reaction lasted for 14 days, with daily rocking performed to promote the reaction and mineral leaching. Meanwhile, pressure was continuously monitored to detect any leaks. At the end of the reaction period, the reaction cell was removed from the oven and allowed to cool to laboratory temperature. For consistency and reliability, this experiment was conducted twice to confirm repeatability and reproducibility. The results presented here correspond to the first experiment, denoted as Experiment 1 (or Exp 1 in chart or table notations). A second test was also performed to examine the leaching of elements from the red mud by introducing acid to the system. For this purpose, 1% HCl was added to the DIW, and the previously described procedure was followed. Hydrochloric acid was utilized in this study since previous studies have demonstrated its effectiveness in mineral leaching, including magnesium, aluminum, calcium, iron, scandium, and others. The leaching efficiency of HCl depends on many factors, including the concentration, the reaction conditions, and the number of stages utilized^{48,49,50,51,52}.

A third experiment (Experiment 3) was conducted to further explore the effect of red mud mass on CO₂ sequestration. The experimental setup and conditions were identical to those used in Experiment 1, including the same pressure, temperature, deionized water volume, and reaction duration. The only difference was that the mass of red mud was doubled from 50 g to 100 g. This adjustment was made to investigate whether increasing the amount of solid material could enhance the mineralization of CO₂. This is particularly relevant because, as mentioned earlier, large quantities of red mud are produced annually as an industrial byproduct, and finding effective ways to utilize it for carbon capture could have significant environmental benefits. By keeping all other parameters constant, the experiment isolates the effect of solid mass on CO₂ retention, allowing a clearer understanding of the relationship between red mud quantity and sequestration efficiency.

In addition to monitoring the mineral phases formed in the solid, the experiment also involved analyzing the liquid phase for pH, electrical conductivity, and total carbon content. Comparing these results with those from Experiment 1 can reveal how the increased mass of red mud influences ion release, carbon retention, and overall system behavior. Ultimately, this experiment aims to provide insight into the potential scalability of red-mud-based CO₂ sequestration and to inform strategies for maximizing CO₂ mineralization using a byproduct that is otherwise considered waste.

Fig. 2

Contents of the reaction cell during reaction experiments.

Post-reaction analysis

After the reaction cell was cooled down to laboratory temperature, gas sampling was conducted. The gas in the cell was carefully purged to ensure that no water vapor was present before connecting it to the gas chromatography for analyzing the gas composition. For this purpose, the Agilent 8860 GC system was utilized. Before performing the GC analysis, the lines of the gas chromatography were purged with the gas from the reaction vessel to ensure there were no trapped gases from previous users. The GC tests were performed multiple times by injecting gas at different cell pressures, and the average of the measurements is reported.

After gas sampling, the pressure of the cell was completely depleted, and the reaction test was terminated. The slurry was gently poured from the cell into a beaker, and the suspended solids were allowed to settle. After this, the liquids and solids were separated. Vacuum filtration was avoided to prevent potential organic contamination from the filter paper. Solid filtration was also not performed to retain suspended particles in the liquid for subsequent analysis (discussed later). The solid content was transferred to a porcelain evaporating dish and dried in a vacuum oven at 60 ℃ for two days to ensure complete removal of moisture. The solid powder was then stored for elemental and mineralogical characterization, including XRD, XRF, and TC analysis. The latter was performed using Elementar soli TOC^® CUBE analyzer.

Samples of the separated liquid were collected after the reaction and subjected to various analyses, including pH measurements, conductivity testing, and total carbon content analysis. The TC analysis for the liquid was performed using a Multi N/C 3100 from Analytik Jena. Since it is suspected that some of the produced carbonate minerals may be present as suspensions in the liquid phase during the reaction, the suspended solids were extracted from the liquid and analyzed for elemental, mineralogical, and carbon contents. This was done in two ways: the first involved vaporizing a specific amount of the liquid under vacuum at 60 °C, and the second involved precipitating the suspended carbonate minerals by gradually increasing the pH of a defined volume through the addition of NaOH while stirring. The addition of NaOH was stopped when the pH reached approximately 12.5, and a milky solution formed. The addition of NaOH results in neutralizing the carbonated water by driving a decarbonation reaction, which increases the concentration of CO₃²⁻ ions. This alteration in the fluid’s speciation creates conditions that thermodynamically favor the formation of carbonate minerals when divalent cations are available. As a result, raising the pH promotes the precipitation of carbonate minerals such as calcium carbonate (CaCO₃). The modified liquid was then transferred into capped centrifuge tubes to facilitate the separation of the precipitated solids. The samples were centrifuged for approximately 30 min at 3,000 rpm. The precipitated solids were then decanted, separated, transferred into a porcelain evaporating dish, and dried under vacuum at 60 °C.

CO₂ sequestration results

Sequestration potential of Saudi red mud

The XRD patterns of the raw red mud (RM) and the solids recovered at the end of Experiment 1 (DIW + CO₂ + RM) are presented in Fig. 3. In Experiment 1, where the red mud sample was reacted in DIW and CO₂, the diffraction peaks largely remained at the same 2θ positions as those of raw RM, indicating that the dominant crystalline phases were preserved. The slight changes in intensity suggest only limited dissolution or reorganization of minerals under neutral conditions⁵³. The XRD patterns indicate that the overall crystalline framework of the red mud remained largely intact, although variations in peak intensities suggest changes in phase abundances.

The quantified XRD results are also shown in Table 1. In the raw red mud, hematite (25.6 wt%), chantalite (24.2 wt%), and rutile (22.8 wt%) are the dominant crystalline phases, while calcite (7.6 wt%), magnesite (4.7 wt%), gibbsite (6.3 wt%), and boehmite (8.4 wt%) appear in lower proportions. Traces of cancrinite were also detected at approximately 0.3 wt%. These results are consistent with the initial description of red mud as an iron-rich residue with significant contributions from Al-, Ca-, and Mg-bearing phases that provide potential sources of cations for CO₂ mineralization. The refinement of the raw red mud yielded an R_wp of 9.14% and a GoF of 5.12. These values indicate an acceptable refinement quality, providing a consistent representation of the measured diffraction pattern. Although the fit is not highly precise, it aligns with expectations for a material characterized by numerous overlapping peaks and variable crystallinity.

The Figure 4 shows an SEM image of the red mud used in the reaction. The SEM image of raw red mud at 15,000× magnification shows a heterogeneous and compact microstructure composed of aggregated fine particles. The particles exhibit irregular, angular shapes and a rough, porous surface texture, indicative of incomplete crystallinity and the coexistence of multiple mineral phases, including iron oxides, aluminosilicates, and hydroxides. The clusters appear densely packed with smaller grains adhering to larger fragments, suggesting strong particle cohesion typical of unreacted bauxite residue. This morphology reflects the natural mineral complexity of red mud before CO₂ exposure, providing a baseline for assessing subsequent dissolution, surface modification, and carbonate formation during mineralization experiments.

The quantified results of the sample obtained after Experiment 1 support the diffractogram analysis, demonstrating that the relative abundance of most minerals increased compared to the raw sample, particularly hematite (27.4 wt%), rutile (27.9 wt%), gibbsite (16.4 wt%), and boehmite (16.8 wt%). The apparent enrichment of these phases suggests that only limited dissolution occurred under neutral aqueous CO₂ conditions, with the crystalline framework largely preserved. This implies that CO₂ uptake in this case is likely restricted to surface sorption or the formation of poorly crystalline carbonate phases that are not easily detected by XRD, resulting in relatively modest mineral trapping efficiency. Quantitative phase analysis also demonstrates that Ca–Al silicate minerals, particularly chantalite and cancrinite, were highly reactive, showing substantial dissolution after exposure to CO₂-enriched water. The disappearance of these phases was accompanied by a relative enrichment and reprecipitation of Al-bearing hydroxides, as reflected in the pronounced increases in gibbsite and boehmite contents. These results suggest that Al was mobilized from primary silicate phases and subsequently stabilized in secondary hydroxide minerals. Finally, the reduction in calcite (4.5 wt%) and magnesite (4.0 wt%) indicates the dissolution of these carbonate minerals from the red mud in the presence of carbonated water. Finally, the refinement of the solid residue obtained after Experiment 1 resulted in an R_wp of 7.44% and a GoF of 3.87, representing a good-quality fit for a multi-phase system. The deviation between the calculated and observed patterns is relatively limited, and the GoF, while above unity, falls within the typical range for heterogeneous industrial residues, reflecting satisfactory refinement robustness.

Fig. 3

XRD patterns of the as-received red mud and the solid red mud collected after Experiment 1.

Table 1 Summary of mineralogical composition of the as-received red mud and the red mud collected after experiment 1 based on XRD analysis.

Fig. 4

SEM image for the as-received Saudi red mud used in the reaction tests.

The XRF results (Fig. 5) further clarify the chemical transformations occurring during CO₂ interaction with red mud. In the raw sample, Fe (42.1 wt%), Al (17.3 wt%), Ca (17.3 wt%), and Si (11.0 wt%) are the dominant elements, consistent with the presence of hematite, alumina phases, calcite/magnesite, and silicates identified by XRD. The presence of calcium indicates that the red mud used has strong potential as a source material for the mineralization process. After treatment with DIW and CO₂ (Experiment 1), the concentrations of Fe, Al, and Si increased (49.4 wt%, 22.2 wt%, and 14.5 wt%, respectively), while Ca decreased substantially to 9.1 wt%. This trend indicates a preferential leaching of Ca-bearing phases under neutral CO₂ conditions. In contrast, Fe- and Al-bearing phases remained relatively stable, resulting in their apparent enrichment in the solid residue. The decline in Ca suggests some mobilization of this element into solution, which will be demonstrated later under the analysis of the reacted liquid. However, the limited changes in XRD imply that much of the Ca was not fully converted into crystalline carbonate phases. Additional time is likely required for the dissolved Ca²⁺ to precipitate as calcite or other stable carbonate phases in the presence of CO₂. The availability of Ca²⁺ ions in the solution plays a critical role in facilitating carbonate nucleation and growth, thereby enhancing CO₂ sequestration.

Fig. 5

XRF results of the as-received red mud and the solid red mud collected after Experiment 1.

The liquid separated from the reaction was analyzed for pH, electrical conductivity, and total carbon content (TC). The initial pH of the DIW was 7.6, while the DIW–RM slurry typically exhibits a pH of 9–10. After the reaction, the liquid pH decreased to 6.7, indicating a shift from an alkaline state to slightly acidic conditions due to the formation of carbonated water. This equilibrium pH was governed by bicarbonate (HCO₃⁻) ions, which are formed from dissolved CO₂. The TC of the reacted liquid reached 1.15 g/L, compared with only 2.01 × 10⁻³ g/L in the fresh DIW, confirming significant CO₂ uptake. The pH decrease is consistent with CO₂ dissolution and the formation of carbonic acid species; at this pH, dissolved inorganic carbon exists predominantly as HCO₃⁻, with a smaller fraction as dissolved CO₂/H₂CO₃. These results suggest that although the red mud provided a reactive alkaline environment, much of the captured CO₂ remained in the aqueous phase rather than being fully mineralized into stable carbonates. This could be due to either the limited availability of divalent cations (e.g., Ca²⁺) required for carbonate precipitation or kinetic barriers to the rapid formation of crystalline phases. Electrical conductivity also increased substantially, from 4.88 × 10⁻³ mS/cm in the fresh DIW to 12.75 mS/cm after the reaction, reflecting the release of soluble species (e.g., Na⁺, Ca²⁺, Al³⁺, and carbonate/bicarbonate ions) from the red mud into solution. This increase in ionic strength further supports the interpretation that CO₂ dissolution enhances both ions leaching and aqueous carbon storage. Nevertheless, only part of the dissolved carbon is converted to solid carbonate.

To demonstrate the presence of suspended solids in the liquid phase, a sample of the reacted liquid was vaporized at 60 °C under vacuum, allowing the suspended material to be recovered without additional treatment. The relatively low drying temperature was chosen to avoid potential crystal transformations that could occur at higher temperatures. Other samples of the reacted liquid were treated with NaOH to increase their pH, allowing the precipitation of any suspended carbonate particles. The treated liquid sample was then centrifuged at 3,000 RPM for 30 min to separate the precipitated solids, which were then dried at 60 ℃ under vacuum. The masses of the collected solids were measured, and the results showed that vaporizing 120 mL of the reacted liquid produced about 0.9485 g of solids, while treating 350 mL of the reacted liquid with NaOH resulted in the precipitation of approximately 0.1523 g of solids. The solids obtained from both methods were analyzed using XRD, XRF, and TC. It should be noted that NaOH treatment raised the pH to around 12.

Table 2 shows the XRD results for the solid extracted by vaporizing a sample of the reacted liquid. These results provide additional insight into the secondary mineral phases formed during the reactions. In Experiment 1 (DIW + CO₂), the suspended solids were dominated by sodium carbonates, primarily trona (64.3 wt%) and nahcolite (28.5 wt%), with low amounts of magnesian calcite (6.2 wt%). The dominance of trona and nahcolite phases indicates that CO₂ was effectively captured in aqueous form, though not permanently mineralized within the solid red mud matrix. This is consistent with the previous XRD and XRF observations that showed only modest dissolution of calcium- and magnesium-bearing minerals and preservation of the bulk mineralogical framework. The limited formation of magnesian calcite reflects some incorporation of Ca²⁺ and Mg²⁺ into carbonate phases, but the overall sequestration potential in this case is dominated by soluble, relatively unstable sodium carbonate species^54,55.

The XRD results for the solids extracted from samples after NaOH treatment of the recovered liquids are shown in Table 3. The solids that formed after raising the pH with NaOH tie the whole process together and show how solution chemistry and a simple pH adjustment determine which solid sinks end up storing CO₂-related species. When the pH of the reaction liquids from Experiment 1 (DIW + CO₂ + RM) was increased and the precipitate recovered, the solid was almost entirely magnesian calcite (≈ 98.5 wt%), with only trace thermonatrite and halite. This demonstrates that the sodium carbonate species (trona, nahcolite) observed in the vaporized liquid sample of Experiment 1 convert to a far more crystalline and Ca/Mg-rich carbonate when pH is increased. Practically, it means that the Ca²⁺/Mg²⁺ that were only partially visible as carbonates or in solution can be forced into stable carbonate minerals simply by pH adjustment with a pH buffer such as NaOH, which is considered a useful route to convert soluble/less-stable sodium carbonates into long-lived mineral carbonates. The high carbonate content, therefore, demonstrates that increasing the solution pH induced precipitation of insoluble carbonate minerals, converting the dissolved inorganic carbon species into stable solid forms. This result confirms that the CO₂ captured in solution as bicarbonate can be effectively mineralized through simple alkaline treatment.

The results of the XRF analysis of the solid samples are shown in Fig. 6. The high concentration of sodium in the NaOH-treated sample is due to the addition of a large amount of NaOH to adjust the pH.

Table 2 Summary of mineralogical composition of the solid particles collected by vaporizing the reacted liquid after experiment 1.

Table 3 Summary of mineralogical composition of the solid particles collected by NaOH treatment of the reacted liquid after experiment 1.

Fig. 6

XRF results of the solid obtained after vaporization and NaOH treatment of the reaction liquid of Experiment 1.

Table 4 shows the results of the analysis of the carbon content of all solid samples collected after Experiment 1. The total carbon (TC) analysis provides direct confirmation of the CO₂ storage pathways revealed by XRD, XRF, and the precipitation experiments. The raw red mud contained 0.947 wt% TC, representing its inherent carbonate and organic carbon fraction. After CO₂ mineralization in DI water, the solid residue from Experiment 1 showed a slight increase to 0.959 wt%, indicating minimal direct CO₂ uptake into the solid phase, consistent with the limited calcite and magnesite variation observed by XRD.

In contrast, the post-reaction liquids displayed substantial carbon enrichment upon secondary processing. The solid recovered by vaporizing the reacted liquid contained 5.07 wt% TC, while the solid obtained after NaOH pH adjustment contained 3.22 wt% TC. These values confirm that a significant fraction of CO₂ absorbed during the experiment remained in the aqueous phase as dissolved inorganic carbon and could subsequently be recovered as solid carbonate species. The higher TC in the vaporized sample reflects the precipitation of Na-carbonate and bicarbonate salts (trona, nahcolite), whereas the pH-adjusted sample predominantly contained Ca–Mg carbonates formed by induced precipitation at elevated alkalinity. Overall, the total carbon data indicate that under neutral DI-water conditions, CO₂ capture occurs primarily in the liquid phase rather than direct mineral carbonation of the red-mud solid. However, subsequent alkaline or evaporative treatment effectively converts this dissolved carbon into stable solid carbonates, improving the overall CO₂ fixation potential of the system.

Table 4 Total carbon content of the as-received red mud and the solids collected after experiment 1.

The percentage of CO₂ mineralized was determined using a carbon material balance approach, which involved measuring total carbon in both the solid and liquid phases before and after treatment. Initially, CO₂ was injected into a 300 mL accumulator at 1100 psi and 18 °C. The Peng-Robinson equation of state (PR-EOS) was used to determine the compressibility factor, Z = 0.19, under these conditions. Using this Z-factor, the mass of CO₂ initially injected into the accumulator was calculated from:

$$\:{\text{m}}_{\text{C}\text{O}2,\:\text{i}\text{n}\text{j}\text{e}\text{c}\text{t}\text{e}\text{d}}=\frac{\text{P}\text{V}\:\text{M}\text{w}}{\text{Z}\text{R}\text{T}}$$

(1)

where; P = injection pressure, V = accumulator volume available for CO₂, Mw = molar mass of CO₂, R = gas constant, T = injection temperature, Z = compressibility factor. Using these parameters, the initial injected CO₂ mass was determined to be 195.18 g.

When the system was subsequently increased to 2000 psi and 50 °C, and the accumulator was agitated to facilitate the dissolution of CO₂ into the brine, a portion of the injected CO₂ dissolved within the brine. The solubility of CO₂ in brine under these conditions was estimated using the correlation by Chang et al. ⁵⁶, which gives: mCO_2,dissolved≈37 g. Only this dissolved portion can undergo hydration to carbonic acid and subsequently participate in mineral-forming reactions with the solid phase. Thus, the maximum CO₂ available for mineralization is:

${\text{mC}}{{\text{O}}_{{\text{2}},{\text{available}}}}={\text{ mC}}{{\text{O}}_{{\text{2}},{\text{dissolved}}}}={\text{ 37 g}}$

The carbon retained in the system after the reaction was calculated from the sum of carbon measured in the treated solid and the TC in the remaining liquid (Eq. 1), as highlighted in the total organic carbon measurements in Table 4.

$$\:{C}_{retained}={C}_{Liquid\:after\:reaction}+{C}_{Solid\:after\:reaction}-{C}_{Solid\:before\:reaction}$$

(2)

where $\:{\text{C}}_{\text{r}\text{e}\text{t}\text{a}\text{i}\text{n}\text{e}\text{d}}$ denotes the total amount of retained carbon, $\:{\text{C}}_{\text{L}\text{i}\text{q}\text{u}\text{i}\text{d}\:\text{a}\text{f}\text{t}\text{e}\text{r}\:\text{r}\text{e}\text{a}\text{c}\text{t}\text{i}\text{o}\text{n}}$ denotes the amount of carbon in the liquid phase after the reaction, $\:{\text{C}}_{\text{S}\text{o}\text{l}\text{i}\text{d}\:\text{a}\text{f}\text{t}\text{e}\text{r}\:\text{r}\text{e}\text{a}\text{c}\text{t}\text{i}\text{o}\text{n}}$ denotes the amount of carbon in the solid phase after the reaction, and $\:{\text{C}}_{\text{S}\text{o}\text{l}\text{i}\text{d}\:\text{b}\text{e}\text{f}\text{o}\text{r}\text{e}\:\text{r}\text{e}\text{a}\text{c}\text{t}\text{i}\text{o}\text{n}}$ denotes the amount of carbon in the solid phase before the reaction. Based on the data presented in Table 4, the following are the amounts of retained carbon in each phase. The amount of carbon in the solid phase before reaction is:

$$\:{\text{C}}_{\text{S}\text{o}\text{l}\text{i}\text{d}\:\text{b}\text{e}\text{f}\text{o}\text{r}\text{e}\:\text{r}\text{e}\text{a}\text{c}\text{t}\text{i}\text{o}\text{n}}=\frac{{\text{T}\text{C}}_{\text{S}\text{o}\text{l}\text{i}\text{d}\:\text{b}\text{e}\text{f}\text{o}\text{r}\text{e}\:\text{r}\text{e}\text{a}\text{c}\text{t}\text{i}\text{o}\text{n} {\%}\:}\times\:\:\text{M}\text{a}\text{s}\text{s}\:\text{o}\text{f}\:\text{s}\text{o}\text{l}\text{i}\text{d}\:}{100}=\frac{0.947}{100}\times\:50=0.4735\:\text{g}$$

(3)

The amount of carbon in the solid phase after reaction is:

$$\:{\text{C}}_{\text{S}\text{o}\text{l}\text{i}\text{d}\:\text{a}\text{f}\text{t}\text{e}\text{r}\:\text{r}\text{e}\text{a}\text{c}\text{t}\text{i}\text{o}\text{n}}=\frac{{\text{T}\text{C}}_{\text{S}\text{o}\text{l}\text{i}\text{d}\:\text{a}\text{f}\text{t}\text{e}\text{r}\:\text{r}\text{e}\text{a}\text{c}\text{t}\text{i}\text{o}\text{n}{\%}}\:\times\:\:\text{M}\text{a}\text{s}\text{s}\:\text{o}\text{f}\:\text{s}\text{o}\text{l}\text{i}\text{d}\:\text{a}\text{f}\text{t}\text{e}\text{r}\:\text{r}\text{e}\text{a}\text{c}\text{t}\text{i}\text{o}\text{n}\:}{100}=\frac{0.959}{100}\times\:48.54=0.465\:\text{g}$$

(4)

The amount of carbon in the liquid phase after reaction is:

$$\:{\text{C}}_{\text{L}\text{i}\text{q}\text{u}\text{i}\text{d}\:\text{a}\text{f}\text{t}\text{e}\text{r}\:\text{r}\text{e}\text{a}\text{c}\text{t}\text{i}\text{o}\text{n}}={\text{T}\text{C}}_{\text{L}\text{i}\text{q}\text{u}\text{i}\text{d}\:\text{a}\text{f}\text{t}\text{e}\text{r}\:\text{r}\text{e}\text{a}\text{c}\text{t}\text{i}\text{o}\text{n}\:\text{g}/\text{l}}\times\:\text{L}\text{i}\text{q}\text{u}\text{i}\text{d}\:\text{v}\text{o}\text{l}\text{u}\text{m}\text{e}=1.19\times\:0.7=0.833\:\text{g}$$

(5)

The total amount of carbon retained:

$$\:{\text{C}}_{\text{r}\text{e}\text{t}\text{a}\text{i}\text{n}\text{e}\text{d}}=0.833+0.465-0.4735=0.824\text{g}$$

(6)

The amount of retained carbon is equivalent to 3.025 g of CO₂, and the corresponding CO₂ trapping as mineralization is = $\:\frac{3.025}{37}=\:8.31{\%}\:$.

Influence of HCl as reaction medium

In a parallel experiment to evaluate the effect of acidification on CO₂ mineralization, 50 g of red mud was suspended in 700 mL of deionized water, and hydrochloric acid (HCl) was added to lower the medium pH to 1%. After stabilization, 300 mL of CO₂ gas was injected into the system under a pressure of 1,000 psi and a temperature of 18 °C. The XRD patterns of the raw red mud and the solids recovered at the end of Experiment 2 (1% HCl + CO₂ + RM) are presented in Fig. 7. Experiment 2, which included 1 wt% HCl, shows more pronounced changes in the XRD profile in which peak intensities in the lower 2θ range increased, and sharper peaks emerged, indicating that the acidic environment enhanced crystallinity and facilitated mineral reorganization. The acidic medium promotes dissolution of Calcium- and magnesium-bearing minerals such as calcite, magnesite, and possibly chantalite, thereby releasing large amounts of Ca²⁺ and Mg²⁺ ions into solution, an essential step for stable carbonate formation⁵⁷. The sharper diffraction peaks also suggest that, while the more soluble phases dissolved, acid-resistant minerals such as hematite and rutile remained largely intact, and some recrystallization of residual or newly formed phases occurred⁵⁸.

The quantified XRD results from Experiment 2, shown in Table 5, reveal more substantial mineralogical changes, confirming the strong effect of acidic conditions. The contents of hematite and rutile decreased to 16.6 wt% and 15.3 wt%, respectively. The calcite (CaCO₃) content increased from 7.6 wt% in the raw red mud to 10.9 wt% in the reacted solid, while magnesite decreased slightly to 3.7 wt%. A new phase, gypsum (CaSO₄·2 H₂O), emerged at 26.1 wt%, indicating the combination of liberated Ca²⁺ with sulfate ions, which is derived from the red mud matrix. The appearance of gypsum in Experiment 2 is particularly striking, also indicating substantial dissolution of aluminosilicate and carbonate phases followed by sulfate incorporation and reprecipitation in the acidic medium. This newly formed gypsum phase provides strong evidence of active mineral transformation processes^59,60,61.

In addition to these changes, the carbonate-bearing cancrinite present in the raw red mud transformed into a sodium–aluminosilicate phase, identified as Na₃AlSiO₄·NaOH. This transformation suggests that the acid treatment destabilized carbonate-bearing cancrinite and favored recrystallization into a Na-rich, decarbonated aluminosilicate structure. The concentrations of aluminum hydroxides such as gibbsite (Al(OH)₃) and boehmite (Al₂O₃·H₂O) increased noticeably, reflecting secondary precipitation of Al phases from dissolved aluminosilicates. The hematite and titanium oxide (rutile and anatase) contents decreased slightly, implying that some iron- and titanium-bearing phases were partially dissolved or became less detectable due to the formation of abundant new minerals. The refinement of the solid residue obtained after Experiment 2 resulted in an R_wp of 8.86% and a GoF of 3.42. These values indicate a satisfactory refinement quality, reflecting a moderate level of misfit and a goodness of fit that falls within the expected range for complex industrial residues.

Fig. 7

XRD patterns of the as-received red mud and the solid red mud collected after Experiment 2.

Table 5 Summary of mineralogical composition of the as-received red mud and the red mud collected after experiment 2 based on XRD analysis.

The XRF results of the solid obtained after Experiment 2 (Fig. 8) show different geochemical fingerprint. Fe increased slightly to 44.7 wt% and Cl sharply rose to 15.8 wt%, while Al decreased to 15.8 wt%, Si to 7.6 wt%, Ca to 7.1 wt%, and S to 1.2 wt%, which shows consistency with strong acid leaching of aluminosilicate and carbonate-bearing phases. The changes in the concentration of Al, Si, and Ca suggest extensive dissolution of their host minerals. Although total Fe by XRF increased to 44.7 wt%, Rietveld analysis of XRD indicates a marked decline in identifiable hematite (to 16.6), suggesting redistribution of Fe into poorly crystalline or other Fe-bearing phases that are not indexed as hematite. The strong rise in Cl reflects incorporation of chloride from the HCl treatment into the reacted solid. In contrast, the slight decrease in S suggests leaching of sulfur-bearing phases into solution, with only a portion of the mobilized S reprecipitating as gypsum, as confirmed by XRD quantification. This strong association highlights the role of acid treatment in promoting dissolution-precipitation reactions that incorporate anions (Cl⁻ and SO₄²⁻) from the acidic medium into newly formed crystalline structures.

Fig. 8

XRF results of the as-received red mud and the solid red mud collected after Experiment 2.

When 1wt% HCl was used as the reaction medium instead of deionized water, the separated liquid exhibited markedly different behavior. The initial pH of the acid was 0.88, which increased to 3.77 after the reaction, indicating partial neutralization of the acid through dissolution of alkaline phases from the red mud, such as calcium and magnesium oxides and hydroxides. However, the solution remained within the acidic range. The TC of the reacted liquid increased to 43.52 mg/L compared with 18.02 mg/L in the initial 1wt% HCl, suggesting that some CO₂ dissolution occurred despite the strongly acidic condition, which is known to reduce CO₂ dissolution^62,63. This can also be attributed to the presence of other carbonate-bearing minerals, as indicated by the XRD results of the vaporized liquid sample shown in Table 6 (discussed later). However, since TC analysis of the liquid samples was performed after removal from the reaction cell, most of the dissolved CO₂ should have escaped, leaving the measured carbon content largely associated with suspended carbonate-bearing particles in the liquid phase.

In contrast to the DIW experiment, the electrical conductivity decreased substantially from 107.9 to 34.52 mS/cm, reflecting the consumption of ionic species during mineral dissolution and precipitation processes. This reduction in conductivity suggests that a portion of the dissolved ions, primarily Ca²⁺ and Mg²⁺, were removed from solution through the formation of solid carbonate and sulfate phases (i.e., calcite and gypsum) observed in the XRD results of the solid red mud^64,65.

Table 6 Summary of mineralogical composition of the solid particles collected by vaporizing the reacted liquid after experiment 2.

To investigate the solid content in the reacted liquid, 70 mL of the solution was vaporized, yielding 1.139 g of solids in Experiment 2. The XRD results of the solid collected after vaporizing a reacted liquid sample from Experiment 2 are shown in Table 6. The results show that the inclusion of 1wt% HCl yielded a different residue composition. The solids were mostly composed of gypsum (83.9 wt%), with minor amounts of trona (9.8 wt%), nahcolite (4.7 wt%), halite (1.0 wt%), and only traces of magnesian calcite (0.6 wt%). The dominance of gypsum aligns closely with the XRD quantification of the bulk solid, which showed gypsum as a major crystalline phase in the solid fraction of Experiment 2, and with the XRF data that revealed significant incorporation of S and Cl. This strongly suggests that acidic dissolution of aluminosilicate, carbonate, and hydroxide phases released Ca²⁺ and other cations, which subsequently reprecipitated with SO₄²⁻ from the acid medium to form gypsum as the principal stable phase. The sharp reduction in sodium carbonates compared to Experiment 1 also reflects the more aggressive dissolution environment, which redirected the chemistry from sodium-dominated carbonate precipitation toward calcium sulfate formation⁶⁶.

The solids extracted by the NaOH treatment of the recovered liquid from Experiment 2 (1wt% HCl + CO₂ + RM), as presented in Table 7, also show a different pathway. Unlike vaporization, pH adjustment causes mineral transformations, leading to the precipitation of slightly different chemical phases. However, the XRF results in Fig. 9 show that both post-processing steps result in nearly the same elements. The amount of solids separated and collected after treating 70 ml of the liquid was around 0.1607 g. The recovered solid is dominated by hydrocalumite (≈ 57.5 wt%) and a complex sodium, potassium, titanium, aluminum, silicon, and chlorine (Na–K–Ti–Al–Si–Cl) containing phase, which was found as Altisite (≈ 41.4 wt%), with small amounts of halite. This outcome is consistent with the observations that the acidic treatment resulted in the leaching of Al, Cl, and Ca, as demonstrated by the XRF of the solids presented in Fig. 9, which is also consistent with the XRD results of the bulk solid that revealed the formation of gypsum and cancrinite. Raising the pH by post-treatment of the recovered liquid with NaOH caused the dissolved Al and Ca to reprecipitate not as simple carbonates but as layered Ca-Al hydroxide phases (hydrocalumite) and aluminosilicate/halide phases that incorporate Na, K, Ti, and Cl. The absence of detectable calcite in the solid collected after NaOH treatment in Experiment 2 suggests that much of the calcium was already consumed in forming sulfate minerals (e.g., gypsum) or other sulfate phases during the acidic step, or that high concentrations of competing ions (such as Cl⁻, Al³⁺, and Ti-bearing phases) shifted the precipitation pathway away from simple carbonate formation. Similar behavior has been reported where gypsum addition during red mud treatment increases calcium reactivity toward sulfates and reduces carbonate phase formation^67,68.

Table 7 Summary of mineralogical composition of the solid particles collected by NaOH treatment of the reacted liquid after experiment 2.

Fig. 9

XRF results of the solid obtained by vaporization and NaOH treatment of the reacted liquid of Experiment 2.

The TC results provide further insight into the carbon distribution among the solid and liquid-derived products after CO₂ sequestration in the 1 wt% HCl system. As shown in Table 8, the raw red mud contained 0.947 wt% total carbon, which, as mentioned earlier, primarily reflects its inherent carbonate content before reaction. After exposure to CO₂ under acidic conditions, the TC in the recovered solid decreased to 0.582 wt%, indicating that the acidic environment caused partial dissolution of the original carbonates. This result aligns with the XRD findings, which showed a strong shift toward gypsum formation and reduced carbonate presence in the reacted solid.

Vaporization of the Experiment 2 solution produced almost no solid carbon (0.087 wt%), showing that dissolved inorganic carbon was not stabilized as sodium carbonate salts under acidic conditions. However, after pH adjustment, the carbon content increased to 1.37 wt%. Although the crystalline phases identified (i.e., hydrocalumite, altisite, and halite) do not contain structural carbonate, the elevated carbon offers the presence of poor crystalline or amorphous carbonate precipitates, or partial intercalation of carbonate ions into the layered structure of hydrocalumite, which is known to host various anions^69,70. This implies that NaOH treatment of the reacted liquid facilitates carbonate retention in non-ideal or amorphous phases, contributing to higher apparent carbon fixation that is weakly visible by XRD.

Table 8 Total carbon content of the as-received red mud and the solids collected after experiment 2.

Similar calculations of the amount of CO₂ mineralized were performed as in the previous experiment, and the total carbon mass after the experiment was less than its value before the experiment, reflecting that the red mud reaction with 1 wt% HCl led to mineral dissolution and transformation rather than the desired CO₂ mineral trapping.

Impact of doubling red mud quantity

In the third experiment, an attempt was made to investigate how increasing the amount of red mud used would influence the amount of CO₂ mineralized. This experiment was a replication of Experiment 1 except that the amount of red mud used was doubled (i.e., 100 g instead of 50 g) while fixing the other factors, including pressure, temperature, DIW volume, and duration. Figure 10 shows the XRD diffractogram of this experiment compared to that of the raw red mud. The pattern shows minimal mineral changes in the peaks, as in Experiment 1, indicating limited mineral transformations and no formation of new mineral phases. Table 9 also shows the quantified minerals present in the solid after removing it from the reaction test. The refinement of the solid residue obtained after Experiment 3 resulted in an R_wp of 6.00 and a GoF of 2.39, corresponding to a high-quality refinement. The relatively low R_wp and GoF values indicate minimal deviation between the calculated and observed patterns, confirming a reliable fit for this heterogeneous multi-phase system. Figure 11 shows the XRF results performed on the solid extracted after Experiment 3. The elemental composition closely resembles that observed in Experiment 1.

Fig. 10

XRD patterns of the as-received red mud and the solid red mud collected after Experiment 3.

Table 9 Summary of mineralogical composition of the as-received red mud and the red mud collected after experiment 3 based on XRD analysis.

Fig. 11

XRF results of the as-received red mud and the solid red mud collected after Experiment 3.

The pH of the liquid collected from Experiment 3 was about 6.45, essentially the same as in Experiment 1. The electrical conductivity, however, increased to 22.54 mS/cm, nearly double the 12.75 mS/cm measured in Experiment 1. This rise reflects the larger number of ions released into the solution during the reaction when more red mud is used for the reaction process. The total carbon measured in the liquid phase was 2.1 g/L, compared to 1.15 g/L in Experiment 1, reflecting the larger carbon content in the liquid when increasing the amount of red mud.

Carbon-balance calculations showed that around 16.5% of the introduced CO₂ was mineralized, compared to about 8.3% in Experiment 1. In other words, doubling the amount of red mud led to roughly double the amount of mineralized CO₂. This suggests that, under the tested conditions, the mineralization capacity of red mud scales with the amount of solid available for reaction, highlighting the importance of solid-to-liquid ratios in maximizing CO₂ sequestration efficiency.

Discussion and comparison of the results

The results from both experimental conditions (DIW vs. 1 wt% HCl) demonstrate that pH critically governs the geochemical pathways during the interaction of CO₂ with red mud. Under neutral conditions (Experiment 1), the system favors limited carbonation reactions, forming soluble sodium carbonates (trona, nahcolite) and minor Ca–Mg carbonates. These products represent a limited permanent CO₂ storage, as a significant fraction of the carbon remains in the aqueous phase. CO₂ trapping thus occurs mainly through transient mechanisms, such as sorption, bicarbonate formation, or amorphous carbonate precipitation, with the potential for enhanced solid carbonate formation upon post-treatment, such as the addition of NaOH.

In contrast, under acidic conditions (Experiment 2), the interaction between red mud and CO₂ does not result in stable carbonate formation. Instead, the lowered pH promotes intense dissolution of carbonate-, hydroxide-, and aluminosilicate-bearing minerals, releasing large quantities of Ca²⁺, Mg²⁺, and Al³⁺ into solution. However, rather than promoting carbonate precipitation, these reactions favor the formation of non-carbon-bearing secondary minerals such as gypsum, cancrinite-like phases, or amorphous aluminosilicates. It is suspected that there was a CO₂ release from previously bound carbonates, as the acidic environment destabilizes existing carbonate species and drives decarbonation. XRD and XRF analyses corroborate these observations, showing substantial mineral dissolution, loss of carbonate signals, and increased sulfate and silicate phase formation under acidic treatment. The persistence of hematite and rutile underscores the chemical robustness of these phases, which do not directly participate in CO₂ binding but provide a stable mineral framework^63,71. Thus, instead of enhancing CO₂ mineralization, acid-assisted treatment facilitates mineral reorganization and decarbonation, resulting in a net loss of stored carbon.

These contrasting outcomes emphasize the pivotal role of both the pH and the liquid used to make the red-mud slurry in controlling CO₂–mineral interactions. While neutral CO₂ treatment allows limited but genuine carbonate mineralization, acidic conditions lead to mineral dissolution, CO₂ release, and the formation of sulfate- and silicate-rich secondary phases. Therefore, effective CO₂ sequestration in red mud requires maintaining near-neutral conditions to stabilize carbonate phases and prevent acid-induced decarbonation and mineral transformation. They also emphasize the importance of choosing the right base liquid for preparing the slurry, or treating the red mud beforehand, to avoid conditions that trigger decarbonation and unwanted mineral changes.

Finally, doubling the amount of red mud in the system led to a significant increase in the concentration of cations, as reflected by the nearly doubled electrical conductivity of the liquid. These additional cations were available to react with CO₂, promoting the formation of carbonate minerals and enhancing CO₂ mineralization. This effect is further supported by the total carbon measurements, which also approximately doubled, indicating that more carbon was retained in the solid phase. In other words, increasing the mass of red mud provided more reactive sites for CO₂ binding, resulting in a higher fraction of CO₂ being stored in stable mineral forms.

Hydrogen production from red mud

A recent study by Al-Yaseri et al. ⁴⁵ demonstrated that hydrogen could be generated alongside carbon storage through CO₂ mineralization in basalt. This was attributed to the presence of iron-bearing minerals in the basalt. According to them, H₂ generation during CO₂ sequestration in basaltic rocks occurs through multiple geochemical pathways, primarily driven by redox reactions involving ferrous iron (Fe²⁺). The interaction of CO₂ and water with iron-bearing minerals promotes these reactions, accounting for the observed H₂ production. The possible reactions under the test conditions are^72,73:

$$\:2FeO+2{H}_{2}O\to\:2FeOOH+{H}_{2}$$

(7)

$$\:2FeO+{H}_{2}O\to\:2F{e}_{2}{O}_{3}+{H}_{2}$$

(8)

A similar concept could be applied to the red mud since it is rich in iron, as demonstrated earlier, it is important to examine its potential for hydrogen generation during CO₂ mineralization. For this purpose, the gas phase was analyzed at the end of the carbonation reaction, after the reaction cell had cooled to laboratory temperature. Table 10 summarizes the concentrations of the analyzed gases, while Fig. 12 displays the GC chromatograms, highlighting the peak positions of H₂, H₂S, and CO₂. A small but measurable quantity of hydrogen was detected, with a concentration of 0.001 wt%. Although this amount is minor, it indicates that under the applied experimental conditions, water did not solely act as a reactant for mineral carbonation but also participated in redox reactions that generated H₂. This observation aligns with the role of Fe-bearing minerals in red mud, where Fe²⁺ can act as an electron donor during oxidation, leading to the reduction of water and release of H₂, as described by:

$$\:2F{e}^{2+}+2{H}_{2}O\to\:2F{e}^{3+}+{H}_{2}+2O{H}^{-}$$

(9)

The GC results further revealed substantial H₂S production. In the experiment performed with 1 wt% HCl, H₂S reached 26.5 wt%, whereas in the experiment using DIW, only 5.3 wt% H₂S was detected. This large difference highlights the role of acidic leaching in mobilizing sulfur from the red mud. Under acidic conditions, sulfur present in phases such as sulfates or sulfide inclusions is released into solution, where it can react with the hydrogen generated in situ to form H₂S. A possible reaction is ⁷⁴:

$$\:FeS+2{H}^{+}\to\:F{e}^{2+}+{H}_{2}S$$

(10)

The repeated GC measurements across different experimental runs confirmed the reproducibility of these findings. Consistently higher H₂S levels were observed in the acidified system compared to the DIW system. This trend correlates with the previous XRF results (Figs. 6 and 9), which showed greater sulfur mobilization in the case of using 1 wt% HCl, providing direct evidence that sulfur-bearing phases were leached more effectively in the HCl experiment. Finally, increasing the amount of red mud used in the reaction resulted in an increased amount of H₂ production. The average amount of H₂ produced was around 0.005 wt%.

The production of H₂S during CO₂ sequestration is both a geochemical indicator and an operational concern. On one hand, it supports the concept of coupled redox–sulfur interactions, where hydrogen generated from Fe-driven reactions subsequently combines with mobilized sulfur to form H₂S. H₂S is toxic and corrosive, and its generation represents a potential environmental and safety risk in practical sequestration scenarios. Therefore, these findings emphasize the importance of designing pre-treatment or sulfur-removal strategies to minimize H₂S formation and improve the overall safety and efficiency of CO₂ mineralization in red mud.

Table 10 Composition of gases after the reaction experiments.

Fig. 12

GC chromatograms for Experiment 1 (a), Experiment 2 (b), and Experiment 3 (c).

Conclusions

This study demonstrates that Saudi red mud has significant potential as a feedstock for CO₂ sequestration through mineral carbonation, with the added benefit of generating minor quantities of hydrogen. The experimental results confirm that the system response is strongly dependent on pH, which dictates both the extent of mineral transformations and the efficiency of CO₂ trapping. Under neutral conditions, the interaction of CO₂ with red mud resulted in limited mineralogical changes, with the bulk mineral assemblage largely preserved and only weak carbonate stabilization observed. The total carbon content of the solid increased slightly from 0.947 wt% in the raw material to 0.959 wt% after treatment. At the same time, most of the injected CO₂ was retained in the aqueous phase as bicarbonates and transient sodium carbonates. Notably, a subsequent NaOH treatment step increased the carbon content in the solid to above 3.2 wt%, underscoring the potential for post-treatment processes to enhance sequestration efficiency substantially. Mass balance calculations revealed that approximately 8.3% of the injected CO₂ was mineralized under neutral conditions.

In contrast, the acidified system with 1 wt% HCl exhibited far greater mineral reactivity. Acid dissolution of Ca- and Al-bearing phases released substantial cations into solution, enabling extensive reprecipitation of new minerals. The acidic environment destabilized pre-existing carbonate and silicate phases, promoting the formation of secondary minerals such as gypsum and Na–Al–Si compounds. The process resulted in significant structural reorganization of the solid matrix and a more pronounced transformation of the red mud composition. Liquid analyses further supported these findings, showing a marked increase in total carbon in the reacted solution, coupled with a decrease in electrical conductivity due to ion consumption and mineral precipitation.

The results of Experiment 3 demonstrate that increasing the mass of red mud can substantially enhance CO₂ mineralization under the same reaction conditions. Doubling the amount of red mud from 50 g to 100 g led to higher ionic release into the liquid, higher total carbon in the system, and, most importantly, a doubling of the fraction of CO₂ retained in solid mineral phases (from about 8.3% to 16.5%). This finding indicates that the mineralization capacity of red mud scales with the available solid surface and suggests that optimizing the solid-to-liquid ratio is a key factor for improving CO₂ sequestration efficiency in red-mud systems. This demonstrates that the CO₂ sequestration capacity of red mud can be scaled by adjusting the amount of solid material in the system, highlighting the potential for utilizing large quantities of this industrial byproduct for effective carbon capture and storage.

Gas analysis also indicated hydrogen generation (0.001 wt%) and a pronounced increase in hydrogen sulfide concentration, reaching 26.46 wt% under acidic conditions, compared to 5.35% in the neutral system. Doubling the amount of red mud used in the reaction increases the H₂ yield to 0.005 wt%. In summary, these findings highlight the dual role of Saudi red mud as both a medium for CO₂ mineral sequestration and a potential source of hydrogen during carbonation reactions. While neutral treatment provides limited mineralization and relies heavily on post-reaction processes, acid-assisted treatment facilitates mineral reorganization and decarbonation, resulting in a net loss of stored carbon. This dual benefit not only advances the development of carbon removal technologies but also demonstrates a novel pathway for valorizing industrial waste materials in the context of circular economy practices, transforming red mud from an environmental burden into a functional resource for climate mitigation.

Future studies should systematically investigate the influence of reaction time, reaction strategy, and the use of different additives on the efficiency of CO₂ sequestration in red mud. Critical parameters such as solid-to-liquid ratio, acid concentration, applied pressure and temperature, and mixing methodology warrant detailed evaluation to optimize mineral carbonation pathways and maximize both CO₂ trapping and potential hydrogen generation. Under the current experimental conditions, the extent of stable carbonate mineral formation appears to be limited, suggesting that enhancing factors such as higher CO₂ partial pressure, longer reaction durations, or nucleation promoters may be necessary to achieve more permanent and efficient carbonate mineralization and H₂ generation. Finally, the generation of H₂S underscores the need for pretreatment or sulfur-removal steps to improve the safety of CO₂ mineralization in red mud.