On the global hydration kinetics of tricalcium silicate cement

TL;DR

This study analyzes the hydration kinetics of tricalcium silicate cement, revealing two distinct power-law behaviors over time and suggesting a microstructure evolution consistent with experimental observations.

Contribution

It introduces a global second order rate model capturing the hydration process with a crossover from acceleration to deceleration, linked to hydrate microstructure development.

Findings

Hydration follows two power-law regimes with a crossover at ~16 hours.

Hydrates catalyze early hydration but inhibit it later.

Microstructure is likely plate-like, matching experimental data.

Abstract

We reconsider a number of measurements for the overall hydration kinetics of tricalcium silicate pastes having an initial water to cement weight ratio close to 0.5. We find that the time dependent ratio of hydrated and unhydrated silica mole numbers can be well characterized by two power-laws in time, $x/(1-x)\sim (t/t_x)^\psi$. For early times $t < t_x$ we find an `accelerated' hydration ($\psi = 5/2$) and for later times $t > t_x$ a `deaccelerated' behavior ($\psi = 1/2$). The crossover time is estimated as $t_x \approx 16 hours$. We interpret these results in terms of a global second order rate equation indicating that (a) hydrates catalyse the hydration process for $t<t_x$, (b) they inhibit further hydration for $t > t_x$ and (c) the value of the associated second order rate constant is of magnitude 6x10^{-7} - 7x10^{-6} liter mol^{-1} s^{-1}. We argue, by considering the hydration process actually being furnished as a diffusion limited precipitation that the exponents $\psi = 5/2$ and $\psi = 1/2$ directly indicate a preferentially `plate' like hydrate microstructure. This is essentially in agreement with experimental observations of cellular hydrate microstructures for this class of materials.