Impact of the supplementary cementitious materials on the kinetics and microstructural development of cement hydration

Supplementary cementitious materials (SCMs) are currently used to replace part of the clinker to reduce the environmental impact of cement and to favor the use of local materials. Two representative SCM have been used in this work: slag and fly ash. Each one corresponds to one of the SCM categories: blast furnace slag is a hydraulic material while siliceous fly ash is a pozzolanic material. A hydraulic material sets and hardens under water while pozzolans need calcium hydroxide (also referred as Portlandite) and water to produce hydrates. As a consequence when slag or fly ash is blended with clinker, its reaction occurs in addition to the cement reaction. This work aims to study the impact of the SCMs on cement hydration to get a generic knowledge on the changes in kinetics and microstructural development when SCMs replace a part of the Portland cement. These changes are important as they can modify the microstructure and change the long term properties. It is interesting to quantify and understand these effects to predict their performance and improve these sustainable cements. A large part of the reaction of clinker phases takes place within two days after the mixing with water. During the same period, slag and fly ash particles are not yet reacting, however they can enhance the hydration of the clinker phases. Using inert quartz powders with different fineness, our results show a clear relation between the inter-particle distance and the kinetics of the acceleration period. This can be explained by the shearing between particles which increases as the inter-particle distance decreases. The micrographs showed that reducing the distance or increasing the mixing speed produces more nucleation sites of C-S-H on the grains surfaces. The study of limestone indicates that limestone powder has an additional effect beyond the effect of increasing the shear produced by other SCMs. The nucleation density is much higher on limestone surface and the morphology of C-S-H is changed to individual needles. The arrangement of the atoms at the surface of limestone seems to explain the increased nucleation. In blended systems, the peak during the deceleration period has sometimes been considered as marking the beginning of the reaction of SCM. In this work, we showed that the peak only corresponds to the second dissolution of C3A although the SCM substitution level changes the kinetics of this reaction. It has been shown that this effect is caused by the faster depletion of sulfate ions, due to the faster reaction of clinker phases and faster formation of C-S-H. A part of the sulfate is adsorbed on the C-S-H structure. This leads to the faster depletion of sulfate in the solution and change the morphology of C-S-H. Our results showed that the adsorption of sulfate promotes the growth of C-S-H as we observed needles diverging from nucleation sites. When the sulfate desorbs the C-S-H needles grow more parallel to each other. This effect was confirmed with the observation of alite where sulfate is not present. The results from 1H NMR showed that the structure itself of C-S-H is affected: few gel pores are formed in presence of sulfate. Over the course of hydration, C-S-H fills the space available between the grains. Once the capillary pores reach a size of 6-8nm, the formation of C-S-H is slowed down and the capillary pores do not reduce further in size. The stabilization of the pore size was explained by the increase of the supersaturation index [...]