Modeling the Formation of Geopolymers

Geopolymers are a class of X-ray amorphous alkali aluminosilicate gel binder materials with potential applications in a wide range of areas. In particular, geopolymers can provide significant improvements over traditional Portland cement technology in applications requiring resistance to acid or salt attack, or thermal stability at temperatures up to 1000°C. The quasi-zeolitic nature of some of the phases formed during geopolymerization is also of significant interest in immobilization of cationic waste streams. However, it is only recently that the structures and synthesis mechanisms of geopolymers have begun to be modeled. Microto nanostructural information has been obtained by MAS-NMR, microscopy and synchrotron pair distribution function analysis, which together have provided for the first time the ability to analyze both framework and non-framework cation sites and ordering within the geopolymer gel binder phase in detail. Comparison between the results of an empirical reaction kinetic model and data obtained by in situ energy dispersive synchrotron X-ray diffractometry is presented, and insight into the geopolymerization process and its influence on the microstructure of geopolymers is undertaken. The results presented will have significance in determining the performance of geopolymers in applications requiring controlled setting rates and rheology, or where long-term chemical stability is important. Introduction Geopolymeric materials show significant potential for utilization in a wide range of applications, including as a replacement for traditional Portland cements, as a possible encapsulant for toxic and/or radioactive wastes, and also as a relatively inexpensive yet heatresistant ceramic material [1]. However, due to their primarily X-ray amorphous nature and the high levels of impurities introduced by the use of waste materials as a solid aluminosilicate source for geopolymerization, detailed analysis of the structure and reactivity of geopolymers has historically been somewhat elusive [2]. The development of such an understanding is central to the future widespread utilization of geopolymers, particularly in waste immobilization applications where extreme durability is required, but also (and no less importantly) in the construction industry, where the ability to predict whether or not a material will retain its structural integrity over a 50 year service life under loaded conditions is critical. The experience of 200 years’ usage of Portland cements cannot be replicated in the short term in the laboratory, so the only way to persuade industry that geopolymer technology is sufficiently mature for use in construction applications is to develop a more complete, theoretically sound understanding of geopolymer properties and performance. An intensive recent research effort has provided some very significant advances in the development of such an understanding by the use of simplified model systems and the development of appropriate experimental techniques [3, 4]. Some of the results of these investigations, and their consequences for the understanding of geopolymer structure and synthesis, are discussed here. Raw Material Sources for Geopolymerization Geopolymers are formed by reaction of an alkaline solution (usually containing very high levels of dissolved hydroxide and/or silicate) with a solid aluminosilicate powder, forming an alkali-aluminosilicate gel phase with inclusions comprising unreacted solid precursor particles and/or any added fillers, for example aggregates [5, 6] or fibers [7-10]. Metakaolin, coal fly ash and blast furnace slag are the three aluminosilicate sources most commonly investigated – metakaolin primarily for higher-value ceramic-type applications due to its cost, and fly ash and slag for larger-scale concrete replacement applications. However, the calcium present in some fly ashes and in slags can greatly complicate the analysis of these systems. Synthetic aluminosilicate precursors are also used when a very high-purity raw material is necessary for analytical purposes or specific applications. Some of the analytical work presented here utilizes an aluminosilicate powder synthesized by the PVA-steric entrapment method [11], however the bulk of the results presented are for metakaolin-based systems. Extension of the results presented here to the analysis of fly ash systems is ongoing [1, 12], although modifications to some of the experimental techniques used are necessary to account for the different rheology and high impurity levels of fly ash geopolymers. The Geopolymerization Reaction Process Geopolymerization takes place via a complex multistep mechanism. The initial dissolution of the solid aluminosilicate source releases small silicate and aluminate species into the surrounding solution. These species are highly labile, and so undergo a series of rapid exchange and oligomerization reactions, also involving any silicate species that are initially present in the activating solution. As larger and larger oligomers form due to the very low water content and therefore the strong driving force for polymerization present in the system, the solution phase undergoes a gelation process. This greatly hinders the diffusive transport of dissolved species from the solid particle surfaces to the bulk of the geopolymer, meaning that in most cases unreacted aluminosilicate source particles will be present as inclusions in the binder. The structure of the gel continues to evolve and harden, eventually becoming a predominantly fully coordinated (Q) aluminosilicate network [13, 14], which is what is described as the ‘geopolymeric binder’ phase. This is clearly visible in Figure 1, which is a SEM micrograph of a polished metakaolin-based geopolymer specimen, showing the smooth binder phase, with voids where the very soft unreacted metakaolin particles have been removed during polishing. Figure 1. SEM micrograph of a geopolymer with overall (superficial) SiO2/Al2O3 = 3.90, synthesized by mixing metakaolin with sodium silicate solution. From [3]. Figure 2 presents a simplified conceptual model of some of the chemical processes occurring during the initial setting and later structural evolution of geopolymers. Figure 2. Processes occurring during geopolymerization. To develop a detailed description of the process of geopolymerization, an understanding of each of these individual steps is highly desirable, however separating the effects of a single step from the others that are happening simultaneously, in a highly constrained and rapidly-solidifying system, is quite challenging. Initial work in this field has focused on the use of model systems, in particular aluminosilicate hydrogels [15] and zeolite synthesis systems [4], to describe certain aspects of the chemistry and rheology of reacting geopolymer slurries. However, to ensure that the full range of competitive and synergistic effects between the different processes is able to be analyzed, a means of examining the process of geopolymerization as a whole – from both experimental and computational viewpoints – is necessary. Energy Dispersive X-ray Diffractometry Energy dispersive X-ray diffractometry has been carried out in situ using white-beam synchrotron radiation on a laboratory-sized geopolymer sample, characterizing the rate of geopolymerization during the first 3 hours of the reaction process. By carrying out the reactions at a temperature (~40°C) where the geopolymer is just completing the solidification step shown in Figure 2, the rate of formation of this initial geopolymeric gel phase is able to be described. It must be noted that this phase will differ structurally from the final geopolymer gel observed after extended curing, as the presence of moisture and warmth allows the gel to continue rearranging itself into a more thermodynamically favorable form, involving very high degrees of crosslinking and also the formation of nanosized crystallites. These two stages of gel evolution, Dissolution of solid aluminosilicate source Rearrangement and exchange between dissolved units Silicate species in activating solution