Early reactivity of activated clays - impact on binder design and hydration

The readily available approach to cut down carbon footprint from the cement and concrete industry is the use of supplementary cementitious materials (SCMs). Clays and limestone are available globally and this has led to the development of calcined clay limestone cements (CCLC, also widely referred to as LC3). When the clays are activated using methods other than calcination (thermal activation), we refer to the binder as activated clay limestone cement (ACLC). This thesis focuses on kaolinitic clays, and smectitic clays. The impact of the physico-chemical properties of natural calcined clays (CCs) on rheology and early reactivity is the entry point of this thesis. It was shown that metakaolin content was not the main factor leading to differences in rheological properties and early reactivity. In addition, high early reactivity is also observed for some low specific surface area (SSA) CCs, showing the influence of factors other than SSA. High structural disorder of kaolinite was shown to be related to faster early reactivity. In a follow up study, the impact of physico-chemical properties and metakaolin content on the hydration of ACLC was followed up to 2 years. In order to better characterize the early reactivity of activated clays, a sulfate-limited model system (SLiM) test was developed. SLiM consists of an excess of CC and portlandite, and limited gypsum. It was demonstrated that the SLiM test can be used to probe physico-chemical properties, including the standard enthalpy of formation of metakaolin, and can predict mechanical performance and hydration of ACLC up to 3 days. Additionally, it lays the foundation for automated optimal sulfate determination, and it complements the R3 test in the assessment of activated clays or, in general, alumina-rich SCMs. Mechanical activation, and the combination of mechanical and thermal activation were explored. Results highlighted different mechanisms by which activated clays contribute to the hydration and mechanical performance of ACLC, and the influence of activation protocols on the properties of activated clays. Early reactivity was shown to be dependent on physical properties and structural defects, as highlighted by differences in portlandite consumption and hemicarboaluminate precipitation, while late-reactivity mainly depended on activation degree of clay minerals. Mechanical and mechano-thermal activation, as well as grinding aids, were shown to improve the early strength of smectite-based ACLC by 15 - 59 %. This was partly because, unlike calcination at 900 °C, mechanical and mechano-thermal activation does not result in a significant loss in SSA, and mechanical activation increases the initial dissolution of smectitic clays as shown using the SLiM test. Insights presented in this thesis improve understanding of the effect of physico-chemical properties of activated natural clays on workability and, early and late strength development. An early reactivity test (the SLiM test) which has widespread applications for alumina-rich SCMs was developed. In addition, the potential of mechanical activation and combination of mechanical and thermal activation as alternatives to thermal activation was highlighted.