College of Design – School of Building Construction – PhD Proposal Defense – Taofiq Mohammed

Date: Tuesday, September 1, 2026

Time: 2:00 – 4:00 PM EST

Location: Caddell Building, Conference Room 212 & Microsoft Teams

Microsoft Meeting Link: 

https://teams.microsoft.com/meet/275012286837302?p=tYjHzReQ3Ao5blaC1O

Meeting ID: 275 012 286 837 302

Passcode: fg7aW3gx

Committee:

• Dr. Ebenezer Fanijo – School of Building Construction, Georgia Tech. (Advisor)
• Dr. Baabak Ashuri – School of Building Construction and School of Civil and Environmental Engineering, Georgia Tech.
• Dr. Georg Reichard – School of Building Construction, Georgia Tech.
• Dr. Valerie Thomas – School of Industrial and Systems Engineering and School of Public Policy, Georgia Tech.
• Dr. Moe Sharbaf – EP Power Minerals.

Title: Performance Evaluation of Geopolymer Mixes with Alternative Precursors: Mechanical, Microstructural, Durability, and Sustainability Insights

Abstract

Geopolymers, synthesized through the alkali activation of aluminosilicate precursors, have emerged as promising alternatives to ordinary Portland cement (OPC) because of their potential for lower carbon emissions, competitive mechanical and durability performance, and enhanced resistance to elevated temperatures. Geopolymer are a family of alumino-silicate binders, which is the result of an inorganic polycondensation reaction; a so-called geopolymerisation. Such reaction yields an amorphous to semicrystalline three-dimensional aluminosilicate frameworks with the general empirical formula of Mn{−(SiO2)z–AlO2−}n. wH20 - where M is a charge-balancing cation (e.g., Na⁺, K⁺, or Ca²⁺), n is the degree of polycondensation, and z is the Si/Al molar ratio governing geopolymer concrete properties – similar to conventional cement. Although established geopolymer cement commonly rely on aluminosilicate precursors such as fly ash (FA), ground-granulated blast-furnace slag (GGBFS), and metakaolin (MK), each precursor presents distinct limitations related to reactivity, material availability, compositional variability, cost, as well as environmental sustainability and optimization. For instance, FA shows source-dependent compositional variability - reporting up to 25% variation in strength across FA sources, and slow ambient-cured strength development, while GGBFS, despite high reactivity, can exhibit drying shrinkage 2–3 times that of OPC. MK provides more consistent reactivity but has high liquid demand and requires energy-intensive calcination (600–900 °C). Supply is also increasingly constrained: U.S. FA production declined by ~68% from 2002 to 2024, while the shift toward electric-arc-furnace steelmaking reduces availability of GGBFS. These limitations motivate hybrid geopolymer systems combining abundant natural and industrial precursors to exploit complementary physicochemical properties while mitigating individual precursor deficiencies. Analogous to the multi-component strategy of limestone calcined clay cement (LC3), optimized precursor combinations could achieve competitive engineering performance with reduced environmental and economic burdens. Therefore, this study establishes a pathway for translating alternative geopolymer precursors into construction-ready concrete, beginning with precursor reactivity and reaction mechanisms, followed by systematic mixture design, performance and durability validation, and environmental and economic evaluation. Ultimately, these efforts aim to identify a technically robust, low-impact, and cost-competitive binder with demonstrated potential for field implementation.

Toward this goal, a state-of-the-art review was first conducted to synthesize recent advances in the selection and application of commonly used geopolymer precursors, including FA, GGBFS, and MK, with particular emphasis on their chemical composition and reactivity, geopolymerization mechanisms, microstructural evolution, engineering performance, and system-level sustainability, as well as the key physicochemical factors governing their early-age behavior and long-term durability. The findings showed that performance is particularly governed by the combined effects of precursor chemistry, activator composition, and curing conditions, with FA- and GGBFS-based systems exhibiting distinct reaction pathways dominated by N–A–S–H and C–A–S–H generations, respectively. While both systems demonstrated mechanical and durability performance comparable to or better than OPC, GGBFS-based geopolymer generally exhibited lower energy demand and emissions. Importantly, the review highlighted the role of calcium-rich chemistry in GGBFS-based geopolymer promote C–A–S–H formation and strength development. Together with the need to reduce dependence on conventional precursors, this finding motivated the exploration of alternative calcium-rich industrial materials such as calcium carbide residue (CCR).

Informed by this review, the first research work investigates CCR, a high-calcium by-product of acetylene production, as a supplementary precursor to partially replace GGBFS and MK in binary and ternary geopolymer binder systems. The goal is to develop a blended geopolymer using an abundant Ca-rich precursor to reduce GGBFS demand while enhancing precursor reactivity and engineering performance while minimizing reliance on FA and MK. Findings showed that selected CCR replacement levels, particularly at lower dosage, achieved strengths comparable to the GGBFS control, while ternary CCR–MK blends retained 96–99% of the control 28-day strength. Microstructural analysis revealed that CCR promoted the formation of hybrid reaction products, while its combination with MK facilitated matrix densification and the development of a more refined geopolymer microstructure. At 20% CCR replacement, embodied carbon decreased from 574.6 to 560.6 kg CO₂-e/m³, while material cost decreased by approximately 6%.  These findings demonstrated that CCR can be valorized as a Ca-rich supplementary material to reduce GGBFS consumption, embodied carbon, and material cost while maintaining performance. However, CCR availability is inherently linked to calcium-carbide-based acetylene production, resulting in regional supply constraints and potentially higher transportation costs and motivating the investigation of more widely available natural precursors. More fundamentally, the changes in reaction products and microstructural development observed with alternative precursor substitution demonstrate that replacing conventional precursors can significantly alter geopolymerization pathways. This highlights a critical knowledge gap in how precursor chemistry controls dissolution kinetics and, consequently, gel formation and long-term network development.

Furthermore, a second mechanistic study, guided by computational modeling, was conducted to establish the relationship between precursor dissolution rate and kinetics and the subsequent evolution and reorganization of the geopolymer network. Real-time measurements using inductively coupled plasma mass spectrometry (ICP-MS) showed rapid early dissolution of Al and Si, with initial rates of approximately 0.605 and 4.697 μmol L⁻¹ s⁻¹, respectively, followed by a transition toward polycondensation. Multiscale characterization further revealed progressive gel growth and surface densification, with surface roughness decreasing by approximately 39% by 7 days and the Si/Al ratio stabilizing at about 1.7 – 1.8 as a dense and chemically stable N–A–S–H network developed. These findings showed that early dissolution behavior directly influences subsequent gel chemistry and structural development, providing reaction-level insight for selecting and combining precursors in the hybrid system.

This understanding motivated a hybrid precursor system combining highly reactive MK with lower-energy pumice and limestone powder (LP) as the third research objective. MK provides readily available Si and Al, while pumice supplies additional aluminosilicates for continued polycondensation and LP contributes Ca alongside filler and nucleation effects. Their complementary chemistry promotes interactions between N–A–S–H and Ca-containing gels, providing a basis for optimizing reaction kinetics, microstructural development, engineering performance, and overall binder sustainability. To systematically evaluate precursor interactions across the ternary composition space, a full-range augmented simplex lattice design was deployed to establish composition–property relationships for flowability, compressive strength, sorptivity, and drying shrinkage. Multi-response optimization identified a balanced mixture containing approximately 7.4% MK, 75.6% pumice, and 17.0% LP. Experimental validation yielded 106.25% flowability, 3.62 ksi 28-day compressive strength, 0.0447 mm/√s sorptivity, and 0.056% drying shrinkage. Life-cycle assessment showed a GWP of ~374 kg CO₂-e/m³, compared with 543 and 689 kg CO₂-e/m³ for MK-based GPC and OPC, respectively, demonstrating the environmental benefit of increased pumice and LP substitution. However, transport and dimensional-stability responses were highly composition-dependent, with sorptivity varying by >13-fold (0.0062–0.0823 mm/√s) and drying shrinkage by >3-fold (0.042–0.134%), contributing to larger prediction errors during model validation.

To reduce variability and prediction errors in the initial optimization, the composition space around the initial optimized mixture composition was refined to 5–20% MK, 70–85% pumice, and 10–25% LP, leading to the fourth research goal focused on refining and validating the optimized ternary system. This constrained optimization targets <10% deviation between predicted and experimental responses while identifying a mixture with improved overall performance, including the flowability, compressive strength, sorptivity, and drying shrinkage. The optimized system subsequently undergo integrated life-cycle and techno-economic assessments (LCA- TEA), benchmarked against conventional geopolymer systems to evaluate its environmental impact, production cost, and circular economy. The optimized mixture (9.8% MK, 77.1% pumice, 13.1% LP) achieved 105.48% flowability, 3.83 ksi 28-day strength, 0.042 mm/√s sorptivity, and 0.057% drying shrinkage, with prediction errors of 0.30–5.80%. XRD, FTIR, and SEM-EDS attributed this balanced performance to a dense, silica-rich hybrid C-(N)-A-S-H matrix with moderate Ca incorporation. Compared with conventional geopolymer systems, the mixture demonstrated competitive engineering performance while reducing cost to $286.30/m³ (58.8% below MK-based GPC) and GWP to 378.3 kg CO₂-eq/m³ (30.4% and 10.9% below MK- and slag-based systems, respectively). Strength-normalized analysis further demonstrated improved cost and carbon efficiency relative to MK- and FA-based systems.

Finally, future work will advance the final optimized binder toward field and application-scale validation as a repair or overlay /concrete pavement. Bond performance with existing concrete will be evaluated through pull-off testing in accordance with ASTM C1583 and slant-shear testing based on ASTM C882, together with other durability testing, to determine its suitability for practical repair and overlay applications. Overall, this research progresses from understanding conventional and alternative precursor behavior to mechanism-informed, sustainability-guided mixture development, optimization, and application-scale validation, with the goal of advancing technically viable, lower-carbon geopolymer binders for practical construction applications.