Valorization of diatomaceous earth as an alternative activator for pumice–based geopolymer concrete: Experimental and multiple linear regression modeling

Abstract

Geopolymer cement (GC) is gaining attention as a sustainable alternative to Portland cement due to its lower CO₂ emissions. However, GC formations rely on sodium silicate, which is expensive, and its production has a high carbon footprint. The development of alternative silica and aluminosilicate sources is therefore critical for improving the sustainability and economic viability of geopolymer concrete. This study focuses on the use of diatomaceous earth (DE) as an alternative silica source and examines the effect of the activator composition on the compressive strength of pumice-based geopolymer concrete, relative to commercial sodium silicate (CSS) as a control, using a one-step synthesis approach. The study evaluates the effects of curing temperature (30, 50, and 70 °C), sodium hydroxide concentration (6, 8, and 10 M), and silica mass in the activator (40, 80, and 100 g) on the compressive strength of the resulting geopolymer concrete. The compressive strengths for CSS and DE samples were comparable (<6%). The Multiple Linear Regression (MLR) model (y = 0.3319[NaOH]-0.0231[DE]+0.0304[Temp]+0.4508) indicated that NaOH has the highest positive effect on compressive strength, while excessive silica content has a negative effect on the compressive strength of pumice-based geopolymer concrete. Compressive strength increased (∼50%) from 2.8 to 4.41 MPa when NaOH concentration increased from 6 M to 10 M for DE samples, respectively. In contrast, an increase in silica mass from 40 g to 100 g reduced the strength by ∼28.9% and ∼38% for CSS and DE samples, respectively. The results indicate that at high NaOH concentration, low DE content and elevated temperatures, DE-based activator produced an activator with comparable results to CSS. In the absence of established geopolymer concrete standards in Kenya, a laboratory-scale mix design was adopted to enable consistent comparison between DE and CSS based systems. Further research should focus on mix optimization and strategies to enhance strength development for broader construction applications and standardization.