The second RM pile at the Chalco (China Aluminum Corporation) Shandong Branch provided the RM utilized in this study. Table 1 displays the findings of a wet chemical analysis performed on a representative sample of the RM. According to the examination, the RM contains a considerable amount of Si, Al, and Fe, primarily in the form of stable oxides such as SiO₂, Al₂O₃, and Fe₂O₃. Heavy elements such as Mn, Zn, Cu, Cr, and Pb were also found in trace concentrations (Li et al., 2019). According to the RM sample’s wet sieving examination, 78% of the sample included particles smaller than 48 μm. The RM sample’s density and porosity were determined to be 2.0 g/cm³ and 45%, respectively. RM has diverse physical and chemical characteristics depending on the environment (Xu et al., 2013). Under normal circumstances, aragonite and calcite, which constitute the skeletal minerals of RM, serve as the source of the initial CaCO₃ in RM. Their total content, determined via XRD semi-quantitative analysis, is approximately 5.49 wt.%. These minerals exhibit limited hydrophilicity. Free SiO₂, Al₂O₃, and Fe₂O₃ are the compounds in RM that are actually hydrophilic. The amount of adsorbed water that forms increases with the concentration of these chemicals. Consequently, only a small quantity of gravitational water is released under vibration, if any, even though RM might have a moisture content of 40 to 70%. RM’s primary mineral constituents are Al₂Si₂O₅(OH)₄, Mg₂CO₃(OH)₂(H₂O)₃, and AlO(OH).

The desulfurization gypsum used in this experiment was obtained from Zouping County, Shandong Province. It appears white, with fine particles and a powdery texture. The measured pH was 5.68. The chemical composition of the desulfurization gypsum is shown in Table 2. The main chemical components of the desulfurization gypsum are CaO, SO₃, SiO₂, and MgO. The mineral composition of the desulfurization gypsum used in this experiment mainly consists of hydrated calcium sulfate.

The anhydrous calcium chloride used in this experiment was obtained from Tianjin Juhengda Chemical Co., Ltd. It appears as white, porous particles that are hygroscopic and easily soluble in water, ethanol, acetone, and acetic acid. The pH of the substance ranges from 8.0 to 10.0, with a relative molecular mass of 110.98. It complies with the standard HG/T 5439-2018 [Ministry of Industry and Information Technology (MIIT), 2018].

The high-purity CO₂ used in this experiment was supplied by Yantai Deyi Gas Co., Ltd., with a concentration of 99.9%, and complies with the standard GB/T 23938–2009 (General Administration of Quality Supervision, Inspection and Quarantine of the People’s Republic of China, Standardization Administration of the People’s Republic of China, 2009). The high-purity N₂ was supplied by Qingdao Deyi Gas Co., Ltd., with a concentration of 99.999%, and complies with the standard GB/T 8979–2008 (General Administration of Quality Supervision, Inspection and Quarantine of the People’s Republic of China, Standardization Administration of the People’s Republic of China, 2008).

The experiment utilized a Mettler Toledo bench-top pH meter, model FE28-Standard, equipped with a LE438 three-in-one glass pH composite electrode. The pH meter was calibrated before each test. After each measurement, the electrode was rinsed with deionized water and dried with filter paper to remove residual moisture, preventing interference with subsequent measurement accuracy.

The X-ray diffractometer used in this experiment is a Bruker X-ray diffractometer, model D8 ADVANCE. The samples were placed in a 2 mm deep sample holder. Mineral composition was determined between 5° and 80° with a resolution of 0.02° and a dwell time interval of 2 s (Ji, 2021). After the experiment, phase identification was performed using Jade 6.5 software, and semi-quantitative phase analysis was conducted using the reference intensity ratios (RIR) method. The mineral compositions of the RM and modified RM composite materials were analyzed and compared based on the PDF card library.

The thermogravimetric analyzer used in the test is a Mettler-Toledo TGA2 model. Place 5–10 mg of sample into a crucible, position the crucible on the crucible holder, then close the instrument. Use high-purity nitrogen as the protective gas. Set the test temperature range to 30–1,000 °C with a heating rate of 10 °C/min.

The scanning electron microscope used in this experiment was an Apreo SHiVac field emission scanning electron microscope equipped with a secondary electron detector. It operated at a voltage of 10 kV and employed area scanning. A carbon conductive tape was applied to the specimen holder, and the sample was firmly attached to the tape to ensure a secure fixation on the holder. After fixation, the sample was gold-coated for conductivity. The sample was subsequently magnified at an appropriate magnification for detailed observation, enabling further analysis of the surface morphology, crystal formation, and hydration state of the RM-based composite material at the microscopic level.

Under the condition of no external calcium source, the CO₂ and N₂ flow rate balance method was employed to introduce CO₂ at concentrations of 100%, 15%, and 1% into the RM suspension for testing its neutralizing effect on the RM (Li et al., 2013). CO₂ was introduced at a flow rate of 1 L/min for each concentration, while the suspension was continuously stirred with a magnetic stirrer at 200 rpm. After 24 h of CO₂ aeration, the gas supply was stopped. The pH and electrical conductivity (EC) of the RM suspension were continuously measured to assess the pH rebound of the RM when exposed to atmospheric conditions.

The gas absorption equilibrium curves for various CO₂ concentrations are shown in Figure 2. The pH of the RM suspension gradually decreased to equilibrium as the aeration time increased. When CO₂ at 100% concentration was introduced, the pH of the RM suspension rapidly decreased to equilibrium within 30 min. When CO₂ at 15% concentration was introduced, the pH gradually decreased to equilibrium within 3 h. With CO₂ at 1% concentration, the pH reached equilibrium after 15 h. The time to reach equilibrium decreased with increasing CO₂ concentration. The RM suspension aerated with 100% CO₂ reached equilibrium almost 15 h faster than the suspension aerated with 1% CO₂. The equilibrium pH values for the CO₂ concentrations of 100, 15, and 1% were 6.8, 7.6, and 8.3, respectively. The equilibrium pH varied with the CO₂ concentration, with the pH of the RM suspension aerated with 100% CO₂ being 1.5 units lower than that aerated with 1% CO₂. This is because CO₂ dissolves in the pore water of the RM suspension, reacting with water molecules to form carbonic acid (H₂CO₃). Carbonic acid is a weak acid that exists in dynamic chemical equilibrium in solution, transitioning between carbonate ions (CO₃²⁻) and bicarbonate ions (HCO₃⁻). The bicarbonate ions further react with more water molecules to form carbonate ions, which are precursors to carbonate precipitation. The chemical reaction principle is presented as Equations (1–6) (Ilahi et al., 2024):

Figure 3 shows the pH rebound of the RM suspension after exposure to the atmosphere following gas aeration equilibrium. Although the pH of the suspension rapidly decreased during CO₂ aeration, once the aeration was stopped and the reactor was opened to allow the RM suspension to fully interact with the atmosphere, the pH of the suspension quickly rebounded to between 9.3 and 9.4. Among them, the RM suspension aerated with 100% CO₂ showed a pH rebound of 2.6, the suspension aerated with 15% CO₂ showed a pH rebound of 1.77, and the suspension aerated with 1% CO₂ showed a pH rebound of 1.0. The RM suspension aerated with 100% CO₂ exhibited a greater pH rebound compared to the other two samples. This indicates that the pH decrease observed in Figure 2 is due to the dissolution of CO₂ gas in the aqueous phase in the form of H₂CO₃ or CO₂. In simulated experiments on RM systems, Smith et al. (2003) measured a dissolution rate of Ca₃Al₂O₆ at 1.2 × 10⁻⁸ mol/(m²·s) under conditions of pH 8.5 and 25 °C. The dissolution rate of Ca₃Al₂O₆ is significantly lower than the degassing rate of carbonates. Based on the pH rebound curve, the pH rapidly recovers within 4 h after gas cessation, consistent with the kinetics of gas degassing. Dissolved H₂CO₃ and HCO₃⁻ in the sealed system rapidly decompose into CO₂ upon exposure to the atmosphere, causing a decrease in H⁺ concentration and a rapid pH rebound. After 4 h of gas cessation, the pH of all samples increases slowly. This phenomenon stems from the gradual dissolution of Ca₃Al₂O₆. As the suspension is stirred under atmospheric conditions, the dissolved carbonates gradually degas. When considering only the influence of soluble alkalinity in the pore water, the pH equilibrium value of the RM is 10.0. The rapid pH decrease observed in the closed system is also the result of the depletion of alkalinity in the pore water. This finding highlights the importance of supplementing calcium sources, as the addition of calcium is more favorable for the formation of stable carbonate minerals, thereby achieving long-term CO₂ carbonation. Therefore, unless the original RM contains sufficient calcium, the addition of an external calcium source is considered a necessary condition for the carbonation of alkaline waste slag minerals.

Under atmospheric conditions at room temperature and pressure, a long-term batch experiment was conducted on three groups of RM suspensions, including those with various external calcium sources and those without any external calcium sources. The three groups of samples (RM, RM + CaSO₄, and RM + CaCl₂) were continuously stirred at 400 rpm using a magnetic stirrer for 56 days. Distilled water was added to the reactor every day to maintain a constant solution volume and total weight. After each volume adjustment, the pH and electrical conductivity (EC) of the suspension were measured 30 min later. Every 5 days, an aliquot was collected from each sample to measure the calcium ion concentration.

The pH and Ca²⁺ concentration changes over time for the three groups of samples during the experiment are shown in Figure 4. The addition of external calcium sources significantly enhanced the carbonation efficiency of the RM. The carbonation test of pure RM successfully eliminated most of the alkalinity in the pores of the RM. Under atmospheric exposure, the pH of all three groups of samples decreased to below 10. In contrast, the RM suspension with added external calcium sources exhibited a lower pH value. When considering only the effect of pore water alkalinity, the equilibrium pH of RM is 9.9 (Khaitan et al., 2009a), and the pH of the sample without an added calcium source is 10, which is in close agreement with the measured equilibrium pH of the pore water. The pH of the sample with added CaSO₄ reached 8.6, which is 1.1 lower than the pH of the sample without an added calcium source. The pH of the sample with added CaCl₂ reached 8.0, which is 1.7 units lower than that of the sample without added calcium source. This indicates that the addition of external calcium sources led to higher mineral carbonation. As the experiment progressed, the change in Ca²⁺ concentration verified the presence of the mineral carbonation reaction, as seen in Figure 4b. CaCl₂ is a highly soluble salt with an extremely rapid dissolution rate, dissociating instantaneously into Ca²⁺ and Cl⁻ ions within the experimental system. The initial Ca²⁺ concentration in the RM + CaCl₂ sample was significantly higher than in other groups, and its initial pH decline rate was slightly faster than that of the RM + CaSO₄ group. This indirectly demonstrates that its dissolution rapidly supplies Ca²⁺, preventing it from becoming a rate-limiting factor. CaSO₄ exists in dissolution equilibrium, enabling stable maintenance of Ca²⁺ concentration in aqueous solutions. In this study, the Ca²⁺ concentration in the RM + CaSO₄ sample remained consistently stable at 450–500 mg/L over the long term, consistent with the Ca²⁺ concentration range of saturated gypsum solutions. This indicates that its dissolution rate can continuously match the reaction demand driven by CO₂ dissolution. Since Ca²⁺ was present in the form of CaCO₃ precipitation or other calcium-containing phases, the pure RM sample had the lowest Ca²⁺ content.

Three sets of test samples (RM, RM + CaSO₄, and RM + CaCl₂) were placed in the atmosphere and mixed in the same manner to simulate the mineral carbonation reaction between RM, exposed to the atmosphere, and CO₂. The experiment lasted for 115 days. The RM layer in this study was laid at a thickness of 5 mm to ensure thorough contact between CO₂ and the RM. This experiment focused on the surface core region within the industrial pile where CO₂ can effectively diffuse, and reactions can fully occur, aiming to precisely elucidate the mechanism by which calcium sources promote the carbonation of RM. This experiment involved only one thorough mixing during sample preparation to ensure full contact between RM and the external calcium source. No additional mixing occurred during subsequent testing, fully simulating the static accumulation state of industrial piles and avoiding artificial mixing interference with the natural diffusion and reaction processes of CO₂. Samples were then taken for pH measurement and X-ray diffraction analysis to determine the main elemental composition of the RM.

The pH changes of the three test samples over time during the experiment are shown in Figure 5. On the 72nd day of the experiment, the pH of the pure RM sample decreased to 10.1. In the subsequent period of the experiment, the pure RM sample continued to be exposed to air, and its pH stabilized without further decline. This indicates that the alkalinity in the pore water of the RM was depleted due to the reaction with CO₂ in the atmosphere. Due to the rapid consumption of pore water alkalinity after the addition of external calcium sources, the two groups of samples with added CaSO₄ and CaCl₂ exhibited lower initial pH values. With the progression of the mineral carbonation reaction, the pH of these two groups of samples gradually decreased. The sample with added CaSO₄ showed a pH of 8.8 on the 56th day of the experiment, and then increased slightly to around 9.0 during the subsequent test. The pH of the CaCl₂-added specimens decreased to 8.4 on the 42nd day of the test and increased slightly during the subsequent tests, basically remaining at about 9.0. External calcium sources can influence the final pH equilibrium value. However, compared to the RM suspension test system, the rate of pH decrease in solid-state RM tests is less affected by the type of calcium source added.

The XRD patterns of the test samples before and after the experiment are shown in Figure 6. The main phases of Bayer-process RM are quartz, calcite, calcium zeolite, gibbsite, and boehmite. During the mineral carbonation process, carbonation reactions occur simultaneously, where Ca²⁺ in the sample reacts with CO₂ from the atmosphere to form CaCO₃ precipitation. To confirm the presence of CaCO₃, three sets of RM samples were measured by XRD tests after a period of exposure to the atmosphere. From the figure, after carbonation, the characteristic diffraction peak intensity of CaCO₃ (at 2θ = 29.4°) increased in all three sample groups. Among them, the sample containing added CaCl₂ exhibited the greatest increase in CaCO₃ diffraction peak intensity following carbonation. Using the RIR method for semi-quantitative analysis of CaCO₃ production, the uncertainty in the weight percentage of CaCO₃ is “±0.3 wt.%.” The weight percentage of CaCO₃ in pure RM samples increased by 2.8 wt.% after carbonation. In contrast, the weight percentage of CaCO₃ in samples with added CaSO₄ increased by 4.8 wt.% after carbonation, while that in samples with added CaCl₂ increased by 5.2 wt.%. The proportional increase in peak intensity directly reflects the trend of increased CaCO₃ phase content. Indicates that CO₂ reacts with alkaline substances in RM to form CaCO₃ with relatively complete crystallinity. It can be observed that compared to the samples with added CaSO₄ and CaCl₂, the pure RM sample shows the smallest change in the intensity of the diffraction peaks after the carbonation reaction, indicating that Bayer-process RM has a relatively low reactivity with CO₂. Specimens with added CaSO₄ and specimens with added CaCl₂ showed an increasing trend in CaCO₃ precipitation after 100 days of reaction in the atmosphere. This is a good indication that the addition of a calcium source facilitates the carbonation reaction of RM.

According to XRD test results, the primary change in RM after carbonation is the formation of CaCO₃ through the reaction of calcium-containing mineral phases. Therefore, for further investigation, thermogravimetric analysis (TGA) was conducted on the carbonated RM samples. Mass loss between 400 °C and 800 °C corresponds to the decomposition of various CaCO₃ phases (Wang et al., 2024). The mass loss at this temperature range was analyzed to quantify the proportion of CaCO₃ in the samples.

Calculate the actual CO₂ mineralization of the carbonized RM samples according to Equation 7 (Wang et al., 2024).

In the above equation:

The weight loss curves, CaCO₃ content, and actual CO₂ mineralization of the samples are shown in Figure 7. The carbon sequestration calculations for each sample are shown in Table 3. The CaCO₃ content in original red mud (RM₀) is only about 5.5 wt.%, primarily consisting of aragonite and calcite, which form the skeletal minerals of RM. After carbonation treatment, the CaCO₃ content in the RM group increased to 8.33 wt.%, representing a 2.83 wt.% increase compared to RM₀, while the CO₂ sequestration capacity simultaneously rose to 36.6 g/kg. The introduction of an external calcium source significantly enhanced the carbonation effect on RM. In samples treated with CaSO₄, the CaCO₃ content rose to 10.3 wt.%, representing a 4.8 wt.% increase over RM₀, with CO₂ sequestration capacity rising to 45.3 g/kg. Samples treated with CaCl₂ demonstrated optimal performance, achieving a CaCO₃ content of 10.72 wt.%—a 5.22 wt.% increase over RM₀—and corresponding CO₂ sequestration capacity of 47.2 g/kg. This result clearly indicates that exogenous calcium sources can effectively promote the formation of stable CaCO₃, thereby significantly enhancing its CO₂ sequestration potential. Notably, CaCl₂ exhibited a stronger enhancement effect than CaSO₄, consistent with findings from pH and XRD experimental studies. However, the DTG curve indicates that the samples containing CaSO₄ exhibit superior CaCO₃ crystallinity.

Figures 8a,b presents the SEM images of the sample before and after the carbonation test. The results show that during the CO₂ neutralization of the alkalinity in RM, a mineral carbonation reaction occurs. Initially, the alkalinity of the pore water solution in the RM is rapidly neutralized, resulting in the formation of soluble carbonate ions (CO₃²⁻). These ions then react with Ca²⁺ to form CaCO₃ precipitates. Subsequently, the alkalinity of the RM decreases gradually. Combining the results of mineral analysis tests with the principle of the RM carbonation reaction, it can be inferred that the composition formed after the RM carbonation reaction is CaCO₃. From the figure, it can be observed that after the carbonation reaction, the surface morphology of the RM particles has changed. Before carbonation, the surface of RM particles was relatively dense, exhibiting only a small number of primary pores. After carbonation, a porous CaCO₃ coating formed on the particle surface, accompanied by the emergence of new nanoscale pores within the particles. The encapsulation of the surface by fine particles is more evident. Spaces are present between the particles, and the structure of the sample is irregular. During the carbonation reaction, calcium ions leach out from the original RM particles, leaving behind some new nanoscale spaces. At the same time, the carbonates formed after carbonation fill the existing narrow spaces, leading to a reduction in the volume of large spaces and an increase in the number of nanoscale spaces (Yadav et al., 2010).

Surface scanning of the specimens before and after carbonation reaction was performed, with the mass percentage results for C/O/Ca elements and EDS images shown in Figures 8c,d. The mass percentages of all three elements in the post-carbonation sample increased to varying degrees compared to the pre-carbonation sample. Specifically, the mass percentage of C increased by 15.8%, while the Ca content rose from 0.88% to 3.17%, indicating that CO₂ was fixed in the form of carbonate. This result aligns with the enhanced characteristic peak of CaCO₃ observed in the XRD analysis, further confirming the occurrence of the carbonation reaction.