Contents lists available at ScienceDirect Cement and Concrete Research journal homepage: www.elsevier.com/locate/cemconres Influence of aluminates on the hydration kinetics of tricalcium silicate Elizaveta Pustovgar a , Ratan K. Mishra a , Marta Palacios a , Jean-Baptiste d'Espinose de Lacaillerie b , Thomas Matschei c,1 , Andrey S. Andreev b , Hendrik Heinz d , Rene Verel e , Robert J. Flatt a,⁎ a Institute for Building Materials, ETH Zürich, CH-8093 Zürich, Switzerland b Soft Matter Science and Engineering Laboratory, UMR CNRS 7615, ESPCI Paris, PSL Research University, 75005 Paris, France c Growth & Innovation, Holcim Technology Ltd, CH-5113 Holderbank, Switzerland d Department of Chemical and Biological Engineering, University of Colorado Boulder, Boulder, CO 80309-0596, USA e Department of Chemistry and Applied Biosciences, ETH Zürich, CH-8093, Switzerland ARTICLE INFO Keywords: Hydration Tricalcium silicate (C 3 S) Sodium aluminate Isothermal Calorimetry 27 Al NMR Molecular dynamics (MD) simulations ABSTRACT We used a combination of experimental and modelling techniques to study the effect of NaAlO 2 on C 3 S hy- dration. pH sensitive inhibition of C 3 S hydration occurred at an early age of reaction, but was followed by an increased amount of hydrates formed later. Most results suggest that aluminates hinder C 3 S dissolution. It is hypothesised that this takes place in active dissolution areas, present with a higher density on finer particles. Annealing reduces their number and increases retardation for a given dosage of aluminates. The view that aluminates act by hindering dissolution is supported by molecular dynamics (MD) simulations. They establish that aluminates can adsorb on the hydroxylated C 3 S mainly through strong ionic interactions between aluminate and calcium ions on the surface of silicate. Upon progress of hydration and at higher pH values, the binding strength of aluminates to the hydroxylated C 3 S decreases so that its passivating effect, and retardation, are reduced. 1. Introduction Hydration of silicates, and in particular, tricalcium silicate (C 3 S), the main clinker phase of commercial Portland cement, has been and continues to be studied extensively [1,2]. Nowadays, clinker replace- ment by supplementary cementitious materials (SCMs) is one of the most effective approaches to reduce carbon footprint and embodied energy of Portland cement production [3]. However, these materials are less reactive, in particular at early ages, so that various chemical routes are explored to enhance their reactivity [4] [5]. In non-sulfate optimized systems, many of these routes also lead to a larger release of aluminate ions since many SCMs are alumina-rich, and this, can delay cement hydration [6–11], thus counteracting efforts to increase the early reactivity of blended cements. An under-sulfated behavior can also be triggered by chemical admixtures [12], something that has been shown to result from increased ettringite formation at early ages [13]. More generally, this phenomenon may develop in the course of hy- dration even without chemical activators or chemical admixtures, compromising in these cases more the late than the early strength of cementitious materials. To overcome current challenges in producing highly blended cements, it therefore becomes necessary to better understand the molecular interactions between aluminate ions and calcium silicate surfaces with variable surface chemistry. Several mechanisms have been suggested to explain the inhibition of silicate reaction in presence of aluminium [6–9,14–16]. The first one is an action of aluminates on the dissolution of silicates. In this case, a surface poisoning is proposed to take place through the adsorption of aluminate ions on pre-existing etch pits or through covalent bonding between aluminate ions and the surface of silicates [7,14]. Another mechanism that has been proposed, is a negative effect on the nuclea- tion and growth of silicates hydrates. For instance, Begarin et al. [6] concluded that the calcium aluminate silicate hydrates (C-A-S-H) formed during alite hydration (containing 0.1% of aluminium) do not act as a nuclei for calcium silicate hydrates (C-S-H) and consequently extend the induction period. Finally, a mechanism proposed by Quennoz and Scrivener [8] involves a poisoning of alite by aluminium ions, but, according to the same authors, the presence of aluminate hydrates in the space available for the nucleation and growth of C-S-H may be also responsible for the passivation of the silicates hydration. The existence of these different interpretations, highlights the fact that the passivation mechanism of C 3 S hydration by Al remains poorly de- fined. http://dx.doi.org/10.1016/j.cemconres.2017.06.006 Received 3 August 2016; Received in revised form 3 April 2017; Accepted 29 June 2017 ⁎ Corresponding author. 1 Present address: Department of Civil Engineering, HTW Dresden (University of Applied Sciences), 01069 Dresden, Germany. E-mail address: flattr@ethz.ch (R.J. Flatt). Cement and Concrete Research 100 (2017) 245–262 0008-8846/ © 2017 Elsevier Ltd. All rights reserved. MARK