ABSTRACT

Shortly after the discovery of NMR, the potential of the technique in studies of cementing minerals was demonstrated by Pake (Pake 1948) who observed the 'H-^H dipolar coupling in the water molecules of gypsum, and showed that information about the inter-proton distance may be derived from either *H single-crystal or powder NMR spectra. In the following three decades, *H NMR was used in a small number of studies of the hydration reactions and kinetics of cements by detecting the variations in aH spin-lattice and spin-spin relaxation times (Elaine 1960; Blinc et al. 1978; Kawachi et al. 1955; Seligmann 1968). When the use of high-field superconducting magnets and the magic-angle spinning (MAS) technique (Andrew et al. 1958, Lowe 1959) became routine in NMR in the early 1980s, the possibility of high-resolution studies of a number of important spin nuclides such as 29Si and 27A1 in solid powders immediately became apparent. Pioneering work on 29Si MAS NMR by Lippmaa et al. showed that the 29Si isotropic chemical shift (that is, the variation in the resonance frequency) reveals the degree of condensation of SiO4 tetrahedra in silicate minerals (Lippmaa et al. 1980) including calcium silicates (Magi et al. 1984) and in the amorphous calcium silicate hydrate (C-S-H) produced by Portland cement hydration (Lippmaa et al. 1982). 29Si MAS NMR has subsequently been employed in a large number of detailed studies of the hydration of the calcium silicates alite and belite and of these phases in Portland cement-based materials. From 27A1 MAS NMR experiments Miiller et al. showed that aluminium in tetrahedral and octahedral coordination can be clearly distinguished by their different 27A1 isotropic chemical shifts (Miiller et al 1977, 1981). This result was subsequently utilized in studies of the hydration of calcium aluminate (CA), high aluminate cements, and aluminium minerals in Portland cements. Furthermore, magic-angle spinning has been combined with multi-pulse decoupling sequences to reduce the line broadening which arises from 1H-1H dipolar couplings in *H NMR spectra. Grimmer and Rosenberger (Grimmer and

Rosenberger 1978, Rosenberger and Grimmer 1979) used this technique in their measurements of *H isotropic chemical shifts in hydrous silicates. These results have since been used to distinguish different JH species in the C-S-H produced in Portland cement hydration. The hydration reactions of calcium silicate minerals have also been studied by 17O MAS NMR (Cong and Kirkpatrick 1993<z) which requires the use of 17O (7=5/2, 0.04 per cent natural abundance) isotopically enriched starting materials. Furthermore, specially designed NMR equipment has been developed for a number of applications to cement materials. Notable are hightemperature NMR probes for investigating the clinkering reactions of model raw-mix materials and the structure and dynamics of molten silicates and calcium aluminates at elevated temperatures (Bonafous et al. 1995; Massiot et al. 1995).