Effects of carbonation methods on mechanical behavior and carbonation mechanisms of coal gangue based geopolymer stabilized saline soil Authors: Xiaoyun Yang, Yongxiang Ren, Rongjie Cui, Wenjing Li, Jiale Liu Published: 2026 ABSTRACT AbstractThis study systematically investigated the effects of different carbonation methods on the failure patterns, unconfined compressive strength (UCS), and strength growth rate of geopolymer stabilized saline soil. Carbonation mechanisms were examined using X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), thermogravimetric analysis (TG), scanning electron microscopy (SEM), and energy dispersive spectroscopy (EDS). The results show that the unconfined compressive strength (UCS) of carbonated stabilized soil is significantly improved, exceeding 4.1 MPa at a geopolymer content above 25% under mixed carbonation. For conventional carbonation curing and superimposed carbonation, the strength exceeds 3.8 MPa and 6.2 MPa, respectively, at geopolymer contents over 20%. Among the carbonation methods studied, the strength of 20–35% geopolymer stabilized soil decreases in the order: 2 h superimposed carbonation, 1 h superimposed carbonation, 2 h carbonation curing, 1 h carbonation curing, and mixed carbonation. With increasing geopolymer content, the failure mode changes from cracking to disintegration and peeling, with failure behavior shifting from ductile to brittle. All carbonation methods show obvious strengthening effects when geopolymer content exceeds 25%. Microstructural analysis indicates that carbonation promotes the formation of CaCO3 and C–A–S–H, N–A–S–H, C–N–A–S–H gels, resulting in a denser microstructure. Thermogravimetric analysis shows that carbonated geopolymer-stabilized soil decomposes in three stages: removal of free and bound water, decomposition of low‑crystallinity CaCO3, and decomposition of high‑crystallinity CaCO3. These results provide a theoretical basis and practical reference for the engineering application of carbonated geopolymer stabilized saline soil.Subject terms: Energy science and technology, Engineering, Environmental sciences, Solid Earth sciences Introduction Saline soils are extensively distributed worldwide, mainly in arid and semi-arid inland basins and coastal lowlands, covering a global area of 1.381 billion hectares, approximately 10.7% of the total global land area1. About 70% of the global saline soil is distributed across 10 core countries, with occurrences recorded on all continents. Representative regions include the Central Asian area surrounding the Aral Sea, the arid western part of Australia, and the regions adjacent to the Sahara Desert in Africa. In China, the total area of saline soil is about 100 million hectares, accounting for roughly 10% of the country’s total land area, which is mainly located in Northwest China, the North China Plain, and the eastern coastal areas2–4. Its high content of soluble salts, such as sodium chloride and sodium sulfate, leads to pronounced physicochemical instability5. Under wetting conditions, salt dissolution disrupts interparticle cementation, causing rapid disintegration and significant subsidence. During drying, salt crystallization induces volume expansion, surface hardening, and the formation of dense crust layers that hinder construction activities. Repeated dry–wet cycles therefore result in severe volume changes and structural degradation, leading to uneven settlement of road subgrades, cracking of building foundations, and damage to anti-seepage structures in hydraulic engineering projects6. These characteristics pose serious challenges to the construction quality and long-term serviceability of infrastructure built on saline soils. Conventional stabilizing agents, such as Portland cement and lime, have been widely used to improve the mechanical properties of saline soil7,8. However, their production is energy intensive and associated with substantial CO₂ emissions, which conflicts with current demands for green, low carbon, and sustainable development9,10. Therefore, developing new types of curing agents that are environmentally friendly, low energy consuming and suitable for the special engineering environment of saline soil has become a key issue that urgently needs to be addressed in the field of geotechnical engineering. As a new type of alkali activated cementitious material, geopolymer has shown broad application prospects in the field of soil stabilization due to its excellent mechanical properties, good resistance to chemical erosion and significant low carbon and environmental protection characteristics11–14. Through dissolution, migration, and polycondensation reactions under alkaline activation, geopolymers form a three-dimensional inorganic polymer gel that effectively binds soil particles and enhances long-term stability15,16. Compared with Portland cement, geopolymer production avoids high temperature calcination and can reduce carbon emissions by more than 70%17. Moreover, their dense microstructure and superior resistance to salt corrosion make geopolymers particularly suitable for stabilizing saline soils in aggressive environments. Previous studies have confirmed that geopolymer stabilization can substantially improve the strength of saline soil, with reported increases in unconfined compressive strength ranging from several-fold to nearly an order of magnitude. Razeghi et al.18 conducted multiple tests to compare the effects of alkali activated cementitious materials such as volcanic ash and slag in improving saline soil with Portland cement under the same conditions, confirming the technical feasibility of alkali activated materials to fully replace traditional cement as a stabilizer for saline soil. Li et al.19 used geopolymer to stabilize saline soil and found that when the ratio of sodium silicate to fly ash remained unchanged and the quicklime content was 3%, the unconfined compressive strength of the stabilized soil could be increased by 10 times compared with that of natural saline soil. Hamid et al.20 used geopolymer-stabilized saline sabkha soil and concluded that the unconfined compressive strength of the stabilized soil increased by 340–952% compared with that of the untreated soil. Beyond soil stabilization, geopolymers have recently attracted attention for their potential in CO₂ sequestration21. The presence of alkaline pore solutions and unreacted reactive silica–alumina phases enable geopolymers to react with CO2 and form stable carbonate minerals, allowing long-term carbon storage while simultaneously enhancing material strength. Experimental studies have demonstrated that geopolymers exhibit considerable CO₂ adsorption capacity, which can be further improved through material modification or optimized curing conditions. Freire et al.22 evaluated the advantages of geopolymer materials in the separation or purification of CO2, and explored their potential application as alternatives to Portland cement in CO2 storage wells in the oil and gas industry. Minelli et al.23 analyzed the carbon dioxide adsorption capacity of geopolymer monomer columns and found that geopolymer exhibits excellent CO2 adsorption capability, with a maximum adsorption rate reaching 0.6 mmol/g. Chen et al.24 discovered that amine grafted geopolymers have excellent CO2 capture capabilities and the CO2 adsorption capacity of the solid adsorbent at 1 atm and 60 ℃ was 1.82 mmol/g. Schneider25 investigated the influence of curing temperature on the ability of synthetic phosphate-waste residue geopolymers to adsorb CO2. The experimental data indicated that the adsorption capacity could reach 2.24 mmol/g at 80 °C. These findings suggest that geopolymers can serve a dual function as both soil stabilizers and carbon utilization media. The combined application of industrial solid waste–based geopolymers and CO2 therefore provides a synergistic approach for saline soil treatment, achieving mechanical reinforcement while promoting resource utilization and carbon reduction. Despite these advances, existing studies have largely investigated geopolymer stabilization and CO2 sequestration as separate processes. Systematic research on the coupled application of geopolymer and CO2 for saline soil stabilization remains limited, particularly regarding how different carbonation methods influence strength development and microscopic mechanism. To address this gap, the present study investigated the effects of three carbonation approaches (mixed carbonation, carbonation curing, and superimposed carbonation) on the mechanical performance of coal gangue based geopolymer stabilized saline soil. The strengthening mechanisms were elucidated through comprehensive microstructural characterization using X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), thermogravimetric analysis (TG), scanning electron microscopy (SEM), and energy dispersive spectroscopy (EDS). Materials and methods Materials Coal gangue, ground granulated blast furnace slag (GGBFS), and saline soil were sourced from Yuncheng, Shanxi Province, China. The coal gangue was used as an activated micro powder. The GGBFS was S95 grade. Saline soil was collected from the vicinity of Huanchi Village, Salt Lake District, Yuncheng. A modified sodium silicate activator was prepared by blending industrial-grade liquid sodium silicate with analytical grade NaOH to adjust the silicate modulus to 0.6. The liquid sodium silicate contained 29.99% SiO₂ and 13.75% Na2O, with a Baumé degree of 50 °Bé and an initial modulus of 2.25. The chemical compositions of the raw materials are presented in Table 1. The soluble salt composition, physical properties and state indices of the soil sample are listed in Tables 2 and 3. The soil is classified as sodium sulfate-chloride saline soil, with a total salt content of 1.62%, which falls into the category of moderately saline soil26. Table 1Materials composition /%.MaterialsSiO2Al2O3CaOFe2O3SO3K2OMgOTiO2OthersCoal gangue53.844.10.30.50.10.4–0.60.2GGBFS28.213.346.00.42.60.47.31.20.6Saline soil59.713.612.07.00.13.62.41.10.5 Materials composition /%. Table 2Soluble salt content of soil sample /%.Ca2+Mg2+Na++K+CO32−HCO3−Cl−SO42−0.420.270.1200.030.300.48 Soluble salt content of soil sample /%. Table 3Physical properties and state indexes of soil sample.Natural density ρ/g·cm− 3Moisture content ω/%Optimal moisture content ωop/%Maximum dry density ρ/g·cm− 3Plastic limitωP/%Liquid limitωL/%Plasticity index IPLiquidity index IL1.4820.514.51.7818.230.512.30.19 Physical properties and state indexes of soil sample. Carbonation methods Three carbonation approaches were investigated: mixed carbonation, carbonation curing, and superimposed carbonation. (1) Mixing carbonation method: CO2 was injected during the mixing stage of the geopolymer slurry and saline soil. The resulting mixture was molded into cylindrical specimens and cured in a standard curing chamber until the specified age, followed by strength testing. (2) Carbonation curing method: Specimens were first prepared without CO₂ injection. The molded specimens were then placed in a carbonation chamber for a prescribed carbonation duration, after which they were transferred to a standard curing chamber to complete subsequent curing before strength testing. (3) Superimposed carbonation method: This approach combined mixed carbonation and carbonation curing, i.e., CO2 was injected during mixing, and the molded specimens subsequently underwent carbonation chamber curing before standard curing. Preparation of specimens of carbonated geopolymer stabilized soil Preliminary trials indicated that adjusting the soil water content to 16% prevented the saline soil from extracting excessive water from the geopolymer slurry during mixing, thereby facilitating geopolymerization and improving strength development. Accordingly, saline soil was sieved through a 2 mm mesh, adjusted to 16% moisture content, and sealed for 24 h prior to specimen preparation. The geopolymer precursor consisted of coal gangue and GGBFS mixed at a 1:1 mass ratio. The modified sodium silicate activator (modulus 0.6) was dosed at 18% of the total mass of the solid precursor (coal gangue and slag)27. Geopolymer slurry was incorporated into the soil at five dosage levels: 15%, 20%, 25%, 30%, and 35% (by mass of the soil sample). All specimens were compacted into cylindrical molds (39.1 mm diameter × 80 mm height). Each specimen was compacted in three layers, then demolded to form a soil column. Specimens were prepared using the three carbonation methods described below. Mixing carbonation method: Mixing carbonation was conducted in a self-made mixing carbonation system. The system consists of a CO2 cylinder, a CO2 pressure gauge, a flow meter, an injection tube, and a covered mixer. The solid precursor (coal gangue and slag) and alkaline activator were mixed to form a geopolymer slurry. The preconditioned saline soil was then added. During the mixing of slurry and soil, CO2 was injected for 2 min (The output pressure of CO2 was 0.1 MPa, the concentration was 99.9%, and the injection flow rate was 2 L/min) to produce the mixed carbonated blend. The mixture was immediately molded, compacted, demolded, wrapped with cling film, and cured in a standard curing chamber. Mixing carbonation method: Mixing carbonation was conducted in a self-made mixing carbonation system. The system consists of a CO2 cylinder, a CO2 pressure gauge, a flow meter, an injection tube, and a covered mixer. The solid precursor (coal gangue and slag) and alkaline activator were mixed to form a geopolymer slurry. The preconditioned saline soil was then added. During the mixing of slurry and soil, CO2 was injected for 2 min (The output pressure of CO2 was 0.1 MPa, the concentration was 99.9%, and the injection flow rate was 2 L/min) to produce the mixed carbonated blend. The mixture was immediately molded, compacted, demolded, wrapped with cling film, and cured in a standard curing chamber. Carbonation curing method: The soil–geopolymer mixture was prepared without CO₂ injection and molded into cylindrical specimens. Specimens were then placed in a carbonation chamber (20 ± 2 °C, RH ≥ 75%, CO2 concentration = 20%) for 1–2 h28,29. After carbonation, specimens were wrapped with cling film and transferred to a standard curing chamber for subsequent curing. Carbonation curing method: The soil–geopolymer mixture was prepared without CO₂ injection and molded into cylindrical specimens. Specimens were then placed in a carbonation chamber (20 ± 2 °C, RH ≥ 75%, CO2 concentration = 20%) for 1–2 h28,29. After carbonation, specimens were wrapped with cling film and transferred to a standard curing chamber for subsequent curing. Superimposed carbonation method: CO2 was injected during mixing for 2 min (as in mixed carbonation). The molded specimens were then subjected to carbonation chamber curing for 1–2 h, followed by cling film wrapping and standard curing. A schematic of the carbonation procedures is provided in Fig. 1. Superimposed carbonation method: CO2 was injected during mixing for 2 min (as in mixed carbonation). The molded specimens were then subjected to carbonation chamber curing for 1–2 h, followed by cling film wrapping and standard curing. A schematic of the carbonation procedures is provided in Fig. 1. Schematic diagram of carbonation test. (a) Mixing carbonation, (b) Carbonation curing, (c) Superimposed carbonation. In accordance with relevant specifications30,31, the 7 day unconfined compressive strength (UCS) is adopted as the quality control index for stabilized soils. Therefore, unconfined compressive strength tests were conducted at curing ages of 7 d and 14 d for all specimen groups. Triplicate specimens were prepared for each mix proportion and curing condition in accordance with ASTM D5102/D5102M-2232. The coefficient of variation of test results was controlled within 10%. To quantify the effect of carbonation relative to the corresponding uncarbonated geopolymer stabilized soil, the strength change rate (SCR) was calculated as Eq. (1). It should be noted that when calculating the SCR for carbonation-cured and superimposed carbonation specimens, the uncarbonated control specimens were treated identically to the carbonated ones (20 °C, 75% RH, 1–2 h) except without CO₂, thereby ensuring the same moisture evaporation history before curing. Where, SCR is the strength change rate of carbonated geopolymer stabilized soil relative to uncarbonated geopolymer stabilized soil; UCScarbonation is the unconfined compressive strength of carbonated geopolymer stabilized soil, MPa; UCSuncarbonated is the unconfined compressive strength of uncarbonated geopolymer stabilized soil, MPa. Material characterization Specimens exhibiting relatively higher strength were selected for microstructural and chemical characterization to clarify the carbonation induced strengthening mechanisms. Mineralogical phases were identified using an X-ray diffractometer (XRD, Bruker D8 Advance, Cu target, the scanning angle ranged from 5° to 90°, with a scanning rate of 8° per min). Functional groups and changes in chemical bonding were analyzed using an FTIR spectrometer (Thermo Fisher Scientific iS5, wavenumber range was 400–4000 cm−1). Micro-morphology was examined using an SEM (Hitachi Regulus 8100) at a magnification of 2 × 104. Elemental distribution and composition were obtained using an EDS system (Bruker Quantax XFlash 7). Thermal stability and decomposition behavior were evaluated using a thermogravimetric analyzer (TA SDT-Q600). Samples were heated from 25 °C to 1000 °C at 10 °C/min under a nitrogen atmosphere. Strength analysis of geopolymer stabilized saline soil Strength analysis of uncarbonated geopolymer stabilized saline soil Figure 2 depicts the unconfined compressive strength (UCS) of uncarbonated geopolymer stabilized soil. The increase in geopolymer content and the extension of the curing period can effectively improve the strength of the stabilized soil. The UCS of the stabilized soil with a 15% geopolymer content is 1.4 MPa and 1.9 MPa at 7 d and 14 d, respectively. When the geopolymer content is increased to 35%, the UCS at 7 d and 14 d is 2.2 MPa and 5.0 MPa, which is 1.6 times and 2.6 times higher than that of the 15% geopolymer stabilized soil. The UCS of uncarbonated geopolymer stabilized soil. (a) 7d, (b) 14d. The failure morphology of uncarbonated geopolymer-stabilized soil with different geopolymer contents is presented in Fig. 3, based on representative specimens selected from each group. For the 15% geopolymer stabilized soil, distinct diagonal cracks are observed on the specimen surface, which is consistent with the shear failure characteristics commonly reported for low cementitious stabilized soils33. The 20% geopolymer stabilized soil exhibits an overall sliding along a longitudinal plane, forming a relatively rough shear surface, with geopolymer particles visibly adhered to the soil matrix. The 25% geopolymer stabilized soil shows a vertical penetrating crack, indicating a transition in the failure plane orientation. As the geopolymer content increased to 30%, multiple diagonal penetrating cracks and local spalling are observed, accompanied by limited overall deformation, which aligns with the brittle shear failure behavior of cementitious-stabilized materials reported in the literature34,35. At 35% geopolymer content, the specimen fractured into irregular fragments with protruding residues, and blocky geopolymers are clearly visible inside, indicating severe structural instability after reaching peak strength. Based on the above observations, the failure mode of geopolymer stabilized soil shows a trend of transition from shear-dominated behavior to brittle failure characteristics as the geopolymer content increases. The failure morphology of uncarbonated geopolymer stabilized soil. (a) 15% geopolymer, (b) 20% geopolymer, (c) 25% geopolymer, (d) 30% geopolymer, (e) 35% geopolymer. Strength analysis of mixing carbonation geopolymer stabilized saline soil Figure 4 describes the UCS of the mixed carbonated coal gangue based geopolymer stabilized saline soil. The UCS of the stabilized soil increases monotonically with geopolymer content, and a significant strength enhancement is achieved at contents exceeding 25%. The UCS of stabilized soil with 25% to 35% geopolymer range from 1.2 MPa to 5.1 MPa at 7 days (8.0–33 times that of untreated soil), and between 4.1 MPa and 5.8 MPa at 14 days (27.2–38 times that of untreated soil). The UCS of mixing carbonation geopolymer stabilized saline soil. (a) 7 d, (b) 14 d. Figure 5 presents the failure mode of mixing carbonation stabilized soil. 15% of geopolymer stabilized soil exhibits a shear failure mode, with multiple intersecting diagonal cracks visible on the surface. The cracks extend downward from the top of the sample and intersect in the middle, forming a clear shear slip surface, indicating that it mainly bear shear stress during the stress process, leading to shear failure along the weak surface, and exhibiting certain plastic deformation characteristics after failure. 20% geopolymer stabilized soil showed no obvious penetrating cracks, only slight micro cracks and surface peeling phenomena in local areas, indicating that the increase in geopolymer content has improved soil strength. The stabilized soil samples with a geopolymer content of over 25% exhibit typical brittle failure modes such as oblique section shear failure, collapse failure, and vertical section failure. Moreover, with the increase of geopolymer content, obvious blocky, clustered, and flaky distribution of geopolymer can be seen inside the soil, indicating that its internal bonding strength is high and after the external force reaches the ultimate strength, the soil exhibits sudden fracture phenomenon. The reason is that under high content, the soil sample has strong deformation resistance but poor ductility, resulting in brittle failure. The failure morphology of mixing carbonation geopolymer stabilized soil. (a) 15% geopolymer, (b) 20% geopolymer, (c) 25% geopolymer, (d) 30% geopolymer, (e) 35% geopolymer. To summarize, the increase of geopolymer leads to the transformation of the failure mode of mixing carbonation stabilized soil from cross cracks to collapse and fracture, and the ductile failure dominated by shear gradually evolves into sudden failure dominated by brittle fracture, reflecting the transition of materials from soil behavior to cementitious behavior. Compared with uncarbonated geopolymer stabilized soil, mixing carbonation geopolymer stabilized soil shows brittle failure when the content is above 25%, which corresponds to its higher strength. Strength analysis of carbonation curing geopolymer stabilized saline soil Figure 6 indicates the UCS of geopolymer stabilized saline soil after 1 h and 2 h of carbonation curing. For carbonated curing 1 h geopolymer stabilized soil, during the 7 d curing period, the UCS of stabilized soil with a geopolymer content of more than 20% can reach over 2.1 MPa. During the 14 d curing period, the UCS is all above 2.1 MPa. The overall UCS of geopolymer stabilized soil after carbonation curing for 2 h is higher than that carbonation curing for 1 h. At geopolymer contents exceeding 20%, the UCS is significantly enhanced, exceeding 3.8 MPa (more than 25.2 times that of untreated soil). The UCS of carbonation curing geopolymer stabilized saline soil. (a) Carbonation curing 1 h at 7 d, (b) Carbonation curing 1 h at 14 d, (c) Carbonation curing 2 h at 7 d, (d) Carbonation curing 2 h at 14 d. Figure 7 shows the failure morphology of carbonation curing soil. The failure morphology of 1 h and 2 h carbonation curing geopolymer stabilized soils is similar. The 15% geopolymer stabilized soil exhibits a distinct longitudinal through-crack, which is a typical manifestation of shear failure. Moreover, only agglomerated, spherical geopolymer particles were formed within the specimen. The failure mode of 20% to 25% geopolymer stabilized soil shows a longitudinal main crack along the vertical direction, with uneven crack width and significant opening in some local areas. The soil on both sides undergoes displacement and separation, indicating that it has endured considerable shear stress during the loading process and exhibits a shear splitting composite failure mode. For geopolymer-stabilized soils with a content of 30–35%, significant damage was observed in the top region, characterized by surface cracking, extensive irregular fissures, and spalling. The cracks were predominantly longitudinal and did not form through-going planes. The overall failure mode reflects the sudden failure of geopolymer stabilized soil after reaching the ultimate strength, showing brittle failure. The reason is the mechanical property that the stiffness of the soil increases but the ductility decreases under the condition of high geopolymer content. The failure morphology of carbonation curing geopolymer stabilized soil. In summary, the failure mode of carbonation curing geopolymer stabilized soil also reflects a process from plasticity to brittleness, and the failure mode does not vary with the increase of carbonization time. Moreover, the higher the content of geopolymer, the more it reflects the brittle failure of the cementitious material. Strength analysis of superimposed carbonation geopolymer stabilized saline soil Figure 8 shows the UCS of stabilized saline soil with superimposed carbonated geopolymer. When the geopolymer content reaches 20%, the UCS of the stabilized soil can be significantly improved. The UCS of stabilized soil with 1 h superimposed carbonation reaches 1.2 MPa and 5.2 MPa at 7 d and 14 d, which are 7.7 and 34.4 times that of untreated soil, respectively. The UCS of stabilized soil with superimposed carbonation for 2 h is higher than that of stabilized soil with superimposed carbonation for 1 h, and the UCS of 20% geopolymer stabilized soil can reach 1.7 MPa and 6.2 MPa (11.3 and 41.1 times of untreated soil) at 7 d and 14 d, respectively. The UCS of superimposed carbonated geopolymer stabilized saline soil. (a) Superimposed carbonation 1 h at 7 d, (b) Superimposed carbonation 1 h at 14 d, (c) Superimposed carbonation 2 h at 7 d, (d) Superimposed carbonation 2 h at 14 d. The failure mode of geopolymer stabilized soil after 1 h and 2 h superimposed carbonation is similar, as shown in Fig. 9. 15% to 25% of geopolymer stabilized soil exhibits longitudinal cracks, and the number of cracks increases with the increase of geopolymer. 30% geopolymer stabilized soil shows obvious peeling in some areas, except for cracks, and internal geopolymer can be seen. 35% geopolymer stabilized soil severely disintegrates, fractures, and peels off, exposing large areas of internal geopolymer and exhibiting brittle failure. The failure morphology of superimposed carbonation geopolymer stabilized soil. Therefore, the failure mode of superimposed carbonation geopolymer stabilized soil with low polymer content (15–25%) is the presence of longitudinal cracks, while high geopolymer content (30–35%) shows disintegration and fragmentation, reflecting the brittle failure of cementitious materials. Strength analysis of geopolymer stabilized soil with different carbonation methods The strength of carbonated geopolymer is influenced by carbonation method, geopolymer content, and curing time. To examine the individual and interactive effects of these three factors, a three-way analysis of variance (ANOVA) was conducted on the strength data of all carbonated specimens. The results are presented in Table 4. All three main effects were highly significant on strength (P < 0.001). The relative importance of the factors affecting strength decreases in the following order: geopolymer content, curing time and carbonation method. Table 4Results of Three-Way ANOVA.Source of variationSum of squaresDegrees of freedomMean squareF valueP valueSignificanceCarbonation method32.80548.2016.80< 0.001***Geopolymer content222.091455.52346.03< 0.001***Curing time49.934149.93441.40< 0.001***Error48.246401.206---***Significant at P < 0.001. Results of Three-Way ANOVA. ***Significant at P < 0.001. To compare the effects of different carbonation methods on the strength of geopolymer stabilized soil, the strength change rate (SCR) of carbonated geopolymer stabilized soil relative to that of uncarbonated geopolymer stabilized soil was calculated, as shown in Fig. 10. For the SCR of mixing carbonation stabilized soil, during the 7 d curing period, the SCR of stabilized soil with geopolymer content less than 25% is negative, indicating that mixing carbonation inhibited strength development. However, a concern is whether this negative SCR simply indicates poor compaction from rapid hardening. To address this, all specimens were prepared using the same compaction effort, and their dry density was measured after molding. The measured dry density difference between carbonated and uncarbonated specimens was less than 0.02 g/cm³ (e.g., uncarbonated: 1.81 ± 0.02 g/cm³, carbonated: 1.83 ± 0.02 g/cm³), ruling out poor compaction as the cause of the negative SCR. Instead, the negative SCR at low geopolymer contents is mainly due to CO₂ induced alkalinity reduction that suppresses geopolymerization, outweighing any cementing effect from CaCO₃ precipitation. While for the geopolymer content was greater than 25%, it showed an enhanced carbonation phenomenon. The SCR of the sample with a content of 35% reaches 127.7%. After 14 d, the SCR of samples with a content of over 25% is positive, and the SCR is the highest when the content is 30%, reaching 49.6%. It is worth noting that although accelerated carbonation improves the 14-day compressive strength, the carbonation process permanently reduces the pH of pore solution. In geopolymer systems, a low-pH environment may impair the long-term stability of geopolymer gels, thereby weakening the resistance of solidified soil to coupled environmental effects such as salt erosion and dry-wet cycles. The strength change rate of geopolymer stabilized soil with different carbonation methods. (a) 7 d, (b) 14 d. Therefore, the strength of soil stabilized by geopolymer through mixing carbonation, carbonation curing, and superimposed carbonation all increase with the increase in geopolymer content. When the geopolymer content reaches 25% in soil stabilized by mixing carbonation, the strength can be significantly improved. For soil stabilized by carbonation curing and superimposed carbonation, the strength can be effectively improved when the geopolymer content reaches 20%. The SCR of soil strength through mixing carbonation and carbonation curing basically shows a higher value in the early stage and a relatively smaller value in the later stage, proving that carbonation can effectively improve early strength. However, the SCR of superimposed carbonation stabilized soil shows a larger value in the later stage. The main reason for this is likely that mixing carbonation and carbonation curing facilitate the rapid formation of low crystallinity calcium carbonate and the generation of a small amount of geopolymer gel on the surface and near surface of stabilized soil specimens. These early-formed products quickly fill the large pores on the specimen surface, forming a preliminary dense protective layer that effectively inhibits the penetration of external water and air, thereby resulting in a higher early-stage SCR36. In contrast, superimposed carbonation involves multi-round carbonation treatment, which promotes the formation of more stable calcium carbonate and thus leads to a higher late-stage SCR. Regarding the soil strength under different carbonation methods, when the geopolymer content ranges from 20% to 35%, the relative performance decreases in the following order: 2 h superimposed carbonation, 1 h superimposed carbonation, 2 h carbonation curing, 1 h carbonation curing, and mixing carbonation. Microscopic mechanism Mineral composition analysis Microscopic test analysis was conducted on 30% geopolymer stabilized saline soils with higher strength during the 14 d curing period (uncarbonated geopolymer stabilized soil, mixing carbonation geopolymer stabilized soil, carbonation curing for 2 h geopolymer stabilized soil, and superimposed carbonation curing for 2 h geopolymer stabilized soil). Figure 11 shows the XRD patterns of geopolymer-stabilized saline soil. The main mineral phases of coal gangue-based geopolymer stabilized soil include quartz, calcite, gismondine, analcime and thomsonite. Gismondine is a typical crystalline product derived from the C–A–S–H gel system, analcime corresponds to the characteristic crystalline phase of the N–A–S–H gel, and thomsonite represents the crystalline phase of the mixed C–N–A–S–H gel system. After carbonation, no additional new characteristic diffraction peaks appear in the XRD patterns, only the intensity of existing peaks changes. This demonstrates that carbonation does not produce new types of mineral phases, but mainly alters the relative content and crystallinity of original minerals. Moreover, the characteristic diffraction peak intensity of calcite is markedly strengthened after carbonation, revealing that Ca2+ released from the geopolymer system reacts with CO₂ to form additional calcite crystals. The XRD pattern of geopolymer stabilized saline soil. Functional group analysis The changes in functional groups and chemical bonds of the sample can be used to analyze the chemical changes and structural evolution during the carbonization process. The main peaks in the FTIR images of the four samples are basically the same, as shown in Fig. 12. The spectral band of 770–790 cm− 1 is formed by Si–O–Si bending vibration37. Vibration and contraction of Si–O–T (where T is Si or Al) at 860–870 cm− 138. Near 1020 cm− 1 is the main peak of C–(N)–A–S–H39. O–C–O corresponds to 1460 cm− 1, indicating the formation of carbonate40,41. At 1680 cm− 1 and 3480 cm− 1, there were bending vibrations of H–O–H and stretching vibrations of –OH, respectively, indicating that bound water is generated during the geopolymerization process42,43. After carbonation, all the main peaks of the sample are enhanced, reflecting the intervention of CO2, which promotes the formation of geopolymers and carbonates. Moreover, the superimposed carbonation peak is the strongest, which is consistent with its relatively high strength. The primary cause of this phenomenon is that once CO₂ enters the system, it rapidly reacts with free Ca2+ to form CaCO₃, which fills the pores of the soil matrix and creates a denser spatial network structure. Meanwhile, the original C–A–S–H gel in the system undergoes decalcification under carbonation, converting into CaCO₃ and releasing reactive silica–alumina fractions. In the alkaline activation environment, these liberated reactive components further participate in polycondensation reactions to regenerate secondary N–A–S–H and C–N–A–S–H gel phases44,45. In addition, moderate carbonation promotes the leaching of residual alkali within the system, which further reacts with unreacted geopolymer precursors to generate new C–A–S–H, N–A–S–H and C–N–A–S–H gels. Collectively, these carbonate and gel products serve as supplementary cementitious phases to strengthen the matrix structure, ultimately leading to an enhancement in mechanical strength. The FTIR of geopolymer stabilized saline soil. Thermogravimetric analysis Figure 13 shows the TG-DTG diagram of geopolymer-stabilized saline soil. It can be seen from the TG curve that the sample is divided into three steps during the heating process. The first step is 50–150 °C, where the free water and bound water in the geopolymer evaporate. The second step is 500–750 °C, where the low-crystallinity CaCO3 decomposes. The third step is 750–900 °C, where the high crystallinity CaCO3 decomposes46,47. From the DTG curves, the peaks corresponding to free water and bound water of carbonated geopolymer stabilized soil are relatively prominent, indicating the formation of more geopolymer gels, which is consistent with the FTIR analysis results. The peak intensity of CaCO3 corresponding to the carbonated sample is higher than that of the uncarbonated sample, indicating that the carbonation process provided sufficient CO2, thereby forming a larger amount of CaCO3. Meanwhile, based on the thermogravimetric analysis and calculations using Eqs. (2) and (3)48, the CO₂ uptake of mixing carbonation, carbonation curing and superimposed carbonation were determined to be 1.52%, 2.94%, and 3.41%, respectively. The TG-DTG diagram of geopolymer stabilized saline soil. Where, ΔmCO2 is the mass loss of during decarbonization over the temperature range of 550 to 950 °C, and m105°C is the mass of the dried sample. CO2carbonated is the mass ratio of decarbonated in carbonated samples, CO2uncarbonated is the mass ratio of decarbonized in uncarbonated samples. Microscopic morphology analysis Figure 14 represents the SEM images of geopolymer stabilized saline soil. It can be seen from the figure that a large number of clump like, granular, and rod-like C–(N)–A–S–H gels have formed in the geopolymer stabilized soil, which adhere to the soil surface, fill between soil particles, and encapsulate soil particles, thereby enhancing the strength of the stabilized soil. No salt-like substances are observed in the stabilized soil. The reason is that during the geopolymeration reaction, free chloride ions and sulfate ions in the soluble salts of saline soil react with calcium ions and aluminum ions in the system to form Friedel’s salt and ettringite, respectively, while sodium ions are embedded in the aluminosilicate network to form N-A-S-H gels49,50. In the mixing carbonation sample, distinct rod like zeolite like cementitious materials can be observed. Abundant spherical CaCO3 particles are present in both carbonation curing and superimposed carbonation. The SEM of geopolymer stabilized saline soil. (a) Uncarbonated, (b) Mixing carbonation, (c) Carbonation curing, (d) Superimposed carbonation. Elemental analysis To more clearly understand the changes in elements during the carbonation process of geopolymer stabilized soil, energy spectrum line scanning analysis was conducted on the samples, as shown in Fig. 15. The main elements of geopolymer stabilized soil are C, O, Si, Al, Ca, Na, Mg, etc., and the corresponding geopolymers are C–A–S–H, N–A–S–H, and C–N–A–S–H respectively. At the line scan in Fig. 15a, it can be seen that within the range of 0–4.0 μm, elements such as Ca, Na, Mg, Al, and Si are uniformly distributed on the surface. This substance is mainly C–N–A–S–H gel. Si elements are enriched above 4.0 μm, while other elements are relatively scarce, indicating that this area is soil particles. In Fig. 15b, at 0–3.0 μm and 4.0–6.0 μm, N–A–S–H gel is mainly present, while at 3.0–4.0 μm, C–A–S–H is mainly present. In Fig. 15c, N–A–S–H and C–A–S–H coexist and are distributed in the line scan area. In Fig. 15d, at 0–1.8 μm and 3.7ཞ4.25 μm, it is mainly C–A–S–H, and N–A–S–H is at 1.8–3.7 μm and 4.25–6.0 μm. The EDS of geopolymer stabilized saline soil Table 5 shows the content of each element in the EDS line scan. The carbonated sample has relatively less Si, indicating that carbonation promotes geopolymerization reactions, thereby consuming Si elements. The abundance of Ca elements is related to the fact that carbonation promotes the formation of carbonates. The values of Ca/Si, Ca/(Si + Al), and (Ca + Na)/(Si + Al) of the carbonated samples ranged from 0.40–0.53, 0.22–0.39, and 0.27–0.54 respectively, all of which are greater than the corresponding values of the uncarbonated samples. This indicates that the generated C–A–S–H and C–N–A–S–H are relatively more, thereby enhancing the overall strength of the stabilized soil. Table 5The Content of each element in geopolymer stabilized saline soil /wt%.SamplesCOSiAlCaNaMgUncarbonated13.3360.4518.722.743.270.281.20Mixing carbonation11.3353.4914.1912.375.741.541.34Carbonation curing11.5655.7812.816.847.63.062.35Superimposed carbonation9.6859.4118.162.877.311.710.85 The Content of each element in geopolymer stabilized saline soil /wt%. Conclusion The effects of mixing carbonation, curing carbonation and superimposed carbonation on the failure mode, unconfined compressive strength (UCS) and strength development rate of coal gangue based geopolymer stabilized saline soil are examined, and the corresponding mechanisms are analyzed. The main conclusions are summarized as follows. The strength development of geopolymer stabilized saline soil is strongly dependent on geopolymer content and carbonation method. For mixed carbonation, a noticeable strength improvement occurs when the geopolymer content exceeds 25% (above 4.1 MPa), whereas carbonation curing and superimposed carbonation provide effective strengthening at geopolymer contents above 20% (above 3.8 MPa and 6.2 MPa). The strength ranking of stabilized soils with 20–35% of geopolymer content under different carbonation methods from highest to lowest is: superimposed carbonation for 2 h, superimposed carbonation for 1 h, carbonation curing for 2 h, carbonation curing for 1 h, and mixing carbonation. The reason for this is that more CaCO3 and geopolymer are generated in the superimposed carbonation geopolymer soil. With increasing geopolymer content, the failure mode of carbonated specimens gradually shifts from cracking to disintegration and spalling, indicating a transition from predominantly plastic to more brittle failure behavior. The strength development of geopolymer stabilized saline soil is strongly dependent on geopolymer content and carbonation method. For mixed carbonation, a noticeable strength improvement occurs when the geopolymer content exceeds 25% (above 4.1 MPa), whereas carbonation curing and superimposed carbonation provide effective strengthening at geopolymer contents above 20% (above 3.8 MPa and 6.2 MPa). The strength ranking of stabilized soils with 20–35% of geopolymer content under different carbonation methods from highest to lowest is: superimposed carbonation for 2 h, superimposed carbonation for 1 h, carbonation curing for 2 h, carbonation curing for 1 h, and mixing carbonation. The reason for this is that more CaCO3 and geopolymer are generated in the superimposed carbonation geopolymer soil. With increasing geopolymer content, the failure mode of carbonated specimens gradually shifts from cracking to disintegration and spalling, indicating a transition from predominantly plastic to more brittle failure behavior. When the geopolymer content exceeds 25%, the strength change rate (SCR) of carbonated geopolymer stabilized soil remains consistently positive, indicating an overall strengthening effect associated with carbonation. Relatively higher SCR values are observed at early curing ages for mixed carbonation and carbonation curing, suggesting that carbonation may play an important role in enhancing early-stage strength development. When the geopolymer content exceeds 25%, the strength change rate (SCR) of carbonated geopolymer stabilized soil remains consistently positive, indicating an overall strengthening effect associated with carbonation. Relatively higher SCR values are observed at early curing ages for mixed carbonation and carbonation curing, suggesting that carbonation may play an important role in enhancing early-stage strength development. Microstructural and compositional analyses indicate that the principal phases in carbonated coal gangue based geopolymer stabilized saline soil include SiO2, CaCO3, C–A–S–H, N–A–S–H, and C–N–A–S–H, with major constituent elements such as C, O, Si, Al, Ca, Na, and Mg. Compared with uncarbonated specimens, carbonated samples exhibit higher CaCO3 content as well as increased Ca/Si, Ca/(Si + Al), and (Ca + Na)/(Si + Al) ratios. These results suggest that the carbonation process is likely associated with enhanced formation of carbonate phases and calcium rich aluminosilicate gels, which may contribute to the observed strength improvement. Microstructural and compositional analyses indicate that the principal phases in carbonated coal gangue based geopolymer stabilized saline soil include SiO2, CaCO3, C–A–S–H, N–A–S–H, and C–N–A–S–H, with major constituent elements such as C, O, Si, Al, Ca, Na, and Mg. Compared with uncarbonated specimens, carbonated samples exhibit higher CaCO3 content as well as increased Ca/Si, Ca/(Si + Al), and (Ca + Na)/(Si + Al) ratios. These results suggest that the carbonation process is likely associated with enhanced formation of carbonate phases and calcium rich aluminosilicate gels, which may contribute to the observed strength improvement. Thermogravimetric analysis indicates that the thermal decomposition behavior of carbonated coal gangue based geopolymer stabilized saline soil can be divided into three main stages: evaporation of free and bound water, decomposition of low-crystallinity CaCO3, and decomposition of high-crystallinity CaCO3. The relatively stronger mass-loss peaks associated with bound water in carbonated specimens imply a higher degree of geopolymer gel formation, while the increased decomposition peaks of CaCO3 compared with uncarbonated specimens suggest that sufficient CO2 exposure promotes the crystallization of carbonate phases. Thermogravimetric analysis indicates that the thermal decomposition behavior of carbonated coal gangue based geopolymer stabilized saline soil can be divided into three main stages: evaporation of free and bound water, decomposition of low-crystallinity CaCO3, and decomposition of high-crystallinity CaCO3. The relatively stronger mass-loss peaks associated with bound water in carbonated specimens imply a higher degree of geopolymer gel formation, while the increased decomposition peaks of CaCO3 compared with uncarbonated specimens suggest that sufficient CO2 exposure promotes the crystallization of carbonate phases. Author contributions X.Y. Yang: Conceptualization, Data curation, Formal analysis, Funding acquisition, Methodology, Project administration, Resources, Supervision, Visualization, Writing - original draft, Writing - review & editing. Y.X. Ren: Writing - review & editing, Validation. R.J. Cui, W.J. Li and J.L. Liu: Data curation, Investigation, Software, Validation, Writing - review & editing. Funding This research was funded by the Basic Research Program Project of Shanxi Province, China (No. 202403021212151) and the Science and Technology Innovation Project of Higher Education Institutions in Shanxi Province, China (No. 2023L508). Data availability Data will be made available on request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement Data will be made available on request.