Nucleation and growth studies by small angle neutron scattering and results for a glass ceramic

Abstract

THE nucleation of phase separation in bulk solids remains one of the more elusive phenomena in both experimental and theoretical terms¹. Experimentally, the difficulties centre on the problem of gaining detailed structural information about the small composition fluctuations from which the precipitate develops. Theoretically, progress has been made in calculating stable microstructures in simple materials such as the rare gases² and amorphous metals³ as well as identifying transient clusters in liquids which are precursors to long range order and thus may act as nucleation embryos⁴. However, it remains a formidable problem to relate such calculations to the microstructure of a multicomponent solid or liquid. The most readily observed physical property of nuclei is their ability to facilitate growth of a new phase in appropriate conditions. Consequently, the study of the growth process immediately following nucleation is the most accessible though indirect method of inquiry into nucleation. It is, however, often difficult to follow the very early stages of growth either because it occurs too rapidly, or because the size or total mass of the precipitate falls outside of the conventional range of structural techniques. We report here the striking effects of nucleation as revealed by a neutron small angle scattering study (SANS) of a glass ceramic (2MgO, 2Al₂O₃, 5SiO₂+10% TiO₂ nucleating agent) during the early stages of growth of the precipitate. As a result of the continuous viscosity–temperature relationship, the rate of phase separation in glasses can be put on a favourable time-scale to follow the kinetics of growth simply by working at the appropriate temperature. Also as will be evident below, neutron small angle scattering provides important advantages in following the growth of precipitates in bulk. Three particular advantages have been exploited. (1) The range in scattering vector Q(Q = (4π sin θ)/λ) which it is possible to study includes the region around Q = 0.005 Å⁻¹ which generally falls between the scope of X-ray small angle scattering and light scattering measurements; (2) the low absorption of neutrons even at long wavelengths (∼10 Å) allows the easy investigation in transmission of a bulk glass which is heated directly in the neutron beam; and (3) the precipitating phase (an aluminium titanate solid solution) is rich in titanium atoms, and the negative scattering length of titanium ensures that there is good scattering contrast between the precipitate and the host silicate glass. All the experiments were carried out at the Institut Laue–Langevin on the D11 spectrometer⁵.