Low-pH cements based on blends of MgO and hydromagnesite

Most of the alternative low CO2-cements that have been proposed are based on calcium silicates and aluminates. Their main calcium source is limestone, even though the amounts required may be somewhat lower than for Portland cement clinker. MgO can be used to make a variety of hydraulic binders as well as binders that harden by carbonation. Potential raw materials are Magnesium silicates or Mg-containing brines, which could be carbonated and partially calcined afterwards to obtain mixture of MgO and basic magnesium carbonates. If the final composition of the hardened binder were sufficiently rich in carbonate the CO2 sequestered by it could fully compensate for the CO2 emitted by the production of the energy needed to drive the manufacturing process, we could talk of a truly carbon-neutral binder. We could even envisage carbon-negative binders for which there would be net consumption of CO2 during manufacture and use. Besides their ability to bind CO2, such binders have a low pH (around 10-11), which could be of potential interest regarding waste encapsulation. The current study aimed at increasing the initial carbonate content of such binders in order to enhance their effectiveness as a means of reducing CO2 emissions. The hydration of reactive periclase (MgO) in the presence of hydromagnesite (Mg5(CO3)4(OH)2·4H2O) was investigated by a variety of physical and chemical techniques. Isothermal calorimetry, X-ray powder diffraction, thermal analysis, vibrational spectroscopy (infrared and Raman spectroscopy) and thermodynamic modelling were applied to assess the hydration mechanism. Isothermal calorimetry showed that partial replacement of the MgO with hydromagnesite accelerated the hydration of the pure MgO at early ages (i.e. for about the first half day). The MgO-hydromagnesite blends always produced a significant amount of an unknown hydrate with a thermal decomposition peak centered on about 100°C as determined by thermogravimetry. This hydrate, which was not observed in a previous study, formed quickly – it was already present by one day of hydration and did not increase significantly after that. We hypothesise that this peak represents an amorphous or very poorly-crystalline phase that leads to cohesive binding in these blends. The other main difference between the hydrated MgO-hydromagnesite pastes and the pure MgO paste is the observation of very broad X-ray diffraction peaks for brucite whenever hydromagnesite is present. Furthermore, reflections of a dypingite-like phase (Mg5(CO3)4(OH)2·5H2O) were identified by X-ray diffraction. The formation of artinite, which was calculated to be the thermodynamic stable phase, could not be confirmed, even if a few % of pure artinite were added as seeds to the mix.