Alkali-Activated Metakaolin as a Low-Carbon Stabilizer for Compressed Earth Blocks: Mechanical, Durability, Thermal and Microstructural Performance

Abstract

Compressed earth blocks (CEBs) stabilized with ordinary Portland cement (OPC) are extensively used in low-rise construction but exhibit a high embodied carbon footprint. In an effort to minimize the carbon footprint of CEBs, alkali-activated metakaolin (MK) was employed as a low-carbon alternative binder in CEBs prepared from lateritic red soil from Bengaluru, India. Five different systems of CEBs were prepared using 10% OPC as the control and four sodium hydroxide-activated MK systems at 5, 10, 15, and 20% MK to soil ratios. The raw materials were analyzed for particle size distribution, Atterberg limits, XRD analysis, and elemental composition. The Proctor compaction test was performed to determine the optimum moisture content and maximum dry density of each type of CEB. Samples were molded into 70.6 mm cubes and cured in ambient and mild thermal (60 °C) conditions. The samples were tested for compressive strength, dry density, water absorption, fire resistance, and microstructure using scanning electron microscopy (SEM) at 7 days, 28 days of curing, after 24 hours of water exposure, after exposure to 600 °C for 2 hours, and in their raw state. The 10% MK-geopolymer exhibited the highest strength within the geopolymer series (3.83 MPa after full curing, a ≈106% gain over 7-day strength) and retained ≈96% of its ambient strength after being exposed to fire, in comparison to the OPC mixture, which lost 39% of its strength after fire exposure. Furthermore, geopolymer blocks containing 5, 15 and 20% MK experienced disintegration upon immersion in water. SEM analysis of cross sections of geopolymer samples containing 10% MK revealed a more densely-packed and homogeneous structure relative to samples containing higher or lower percentages of MK. Thus, geopolymers containing 10% MK have exhibited various advantageous properties, indicating their potential as an efficient, fire-resilient, and low-carbon cement replacement in the construction of earthen buildings. Furthermore, a compact and exploratory analysis with limited data is also presented in this paper to shed light on the potential use of machine learning in the optimization of these cement-based materials.