Inorganic polymerisation of Fe-rich slags: unravelling the complex reactions of a simple system

The abundance and accumulation of industrial residues are troublesome for future generations. Similarly, global warming is nowadays recognized as a problem and ordinary Portland cement (OPC) production plays an important role... [ view full abstract ]

The abundance and accumulation of industrial residues are troublesome for future generations. Similarly, global warming is nowadays recognized as a problem and ordinary Portland cement (OPC) production plays an important role in CO₂ emissions. The substitution of OPC by industrial residues provides thus an obvious synergy. A significant part of these residues is Fe-rich (30-60 wt% of FeO), which cannot be incorporated in blended cements due to their low reactivity. Examples are copper and lead slags. Novel material systems, like inorganic polymers, can incorporate these residues because of the higher alkalinity of the system. The reaction mechanism is largely determined by the behavior of Fe and is not comparable to activated blast furnace slags or metakaolin. The glassy nature, the multiplicity in oxidation states (Fe²⁺, Fe³⁺) and the possible coordination environments (4-6 fold coordination) renders the system complex. A combination of several techniques and experimental designs was necessary to shed light on the mechanism of the transition from slag to inorganic polymer at the nano-scale. The focus is placed on slags in the molar compositional range Ca_xFe_2-xSi_yO_2+2y (x = 0-1, y = 1-2) using in-situ ⁵⁷Fe Mössbauer spectroscopy and in-situ pair distribution function analysis derived from synchrotron X-ray scattering experiments. 

During dissolution of the slag, the short-range atomic arrangements are not affected. The species in solution are thus not broken up to the atomic level, but stay as ferrosialates in the pregnant activator. Potassium-based activators can influence the coordination environment of these species, observed as an increase in the coordination number of Fe. The reactions during setting result in 2 different products. Part of the Fe, which is in 2+ oxidation state in the slag, oxidizes to Fe³⁺ in tetrahedral configuration, taking residence in the silicate network and participating in the polymerization reactions. This oxidation is not affected by the oxygen level in the environment, shown by the lack of change in kinetics when preparing samples in a nitrogen glove box. Together with the Fe³⁺ containing network, a layered trioctahedral Fe²⁺ state is formed. This is a glassy equivalent of layered double hydroxide (LDH) minerals like Fe²⁺-bearing smectite clays or Fe²⁺ hydroxides. Whether it really constitutes a separate phase or clusters of a few atoms in layered configuration is not clear. The LDH oxidizes in a later stage and at elevated temperatures, because of drying and the consequent contact with air. A gradual reaction of the phase towards more tetrahedral Fe³⁺ is observed. The link of these transformations to the real-life behavior should be assessed to enhance control on the properties (e.g. durability), by the design of the resulting inorganic polymer phases.