Mechanics and microstructure analysis of geopolymer utilizing ilmenite tailing and metakaolin powder as alkali-activated materials

Geopolymers derived from bulk solid waste are considered ideal substitutes for cement-based materials in developing environmentally friendly construction materials. The utilization of solid waste tailings powder as a precursor for alkali activation of geopolymer can enhance the mechanical characteristics of geopolymer. This study conducted extensive experimental studies on the compressive strength, flexural strength, SEM microstructure, thermogravimetric curve, and variable influence of alkali-activated ilmenite tailings (IT) geopolymer mortar, using the liquid-solid ratio and the number of tailings as variables. An analysis was conducted on the mortar's mechanical characteristics, microstructure, and machine-learning properties. Furthermore, the impact of different atomic ratios on the structure and characteristics of geopolymer mortars was also examined. The test results revealed that the metakaolin-iron titanium tailings geopolymer, when doped with 10 % ilmenite tailings, exhibited a compressive strength of 72.30 MPa at 14 days. Similarly, the flexural strength of the geopolymer reached 4.76 MPa at 14 days when doped with 30 % ilmenite tailings. XRD analysis showed that Magnesiohornblende particles Mg-F-A-S-H and C-A-S-H gel had positive and negative correlations, respectively, with the tailings doping level. TG results indicated that the specimens exhibit enhanced thermal stability following treatment with tailings. SEM analysis revealed that combining ilmenite tailings with metakaolin formed Me-(F)-A-S-H gels. The examination of surface porosity, pore shape, and unreacted material determined that adding ilmenite tailings increased the Si/Al and Fe/Si ratios, enhancing the transformation of the lamellar structure to the three-dimensional disordered gel mesh structure. Additionally, the tailings contributed a significant amount of Ca, elevating the Ca/Na ratio (0.1−0.6) and triggering calcium precipitation within a Ca/Si range of 0.03−0.2. This increase in calcium led to a decrease in the sample's strength as the curing time was extended. Based on prior research and the application of the Gray Wolf optimization algorithm in extreme learning, a clear negative correlation was identified between the liquid-solid ratio and strength. Furthermore, increasing the Si/Al ratio enhanced the geopolymer's strength, while an extended curing time negatively correlated with strength. The impact of tailings mixing was deemed insignificant when the tailings amount did not exceed 20 %. Geopolymers may derive significant environmental and industrial benefits from the improvement of their mechanical properties, thermal stability, and optimization using machine learning.