Magmas are generally composed of three main components, a melt, one or several types of minerals, and a vapor phase in the form of bubbles. The formation and growth of minerals (or crystals) and of the vapor phase (bubbles) exerts a major influence on the behavior of their host magmas because they modify the magma?s physical properties (e.g., viscosity/fluidity, volume) and influence its thermodynamic state (e.g. heat and mass distribution among phases), ultimately controlling what type of eruption (e.g. effusive vs. explosive) will occur. The formation of crystals and bubbles from the melt occurs via a process of spontaneous ?nucleation?, during which the crystal and vapor-forming components aggregate to a critical size. These nuclei then grow by further addition of components to the newly formed melt-crystal or melt-bubble interface. It is critical to understand the nucleation process in magmas because the initial number of nuclei controls the final distribution of crystals or bubbles. This proposal aims to better understand the formation of crystal and bubble nuclei via experimental and analytical approaches. The results could influence our understanding of nucleation theory, and experimental data could be implemented in predictive physical and numerical magma ascent and eruption models.
The energetics and kinetics of nucleation are increasingly well understood for glass ceramics, a class of engineered materials in which a glass hosts a high population density of nm-scale crystals. The development and characterization of industrial glass ceramics has stimulated technological developments in nanoanalysis and spectroscopic investigation of glasses, providing unprecedented opportunities for studying nucleation in naturally-occurring aluminosilicate melts. In its fundamental form, the Classical Nucleation Theory may be used to model experimental nucleation data for natural melts, but only if key assumptions concerning interfacial energy (for crystals) and surface tension (bubbles) are relaxed. Lack of firm understanding of these energy terms, operative at the nanometer scale and therefore elusive to conventional imaging techniques, is especially troubling because the theory stipulates that their influence over nucleation rate is profound. An additional complexity is that while nucleation in magma appears to occur homogeneously for crystals, thus involving just two phases, nucleation of bubbles may occur more typically by a heterogeneous process, involving energetic relationships among three phases. This project proposes to (1) experimentally investigate crystal and bubble nucleation in an intermediate (andesitic) and a silicic (rhyolite) melt, (2) revisit classical nucleation theory in the context of these natural magmas, (3) explore the spatial frontier (i.e., nanoscale) of incipient crystallization via cuttingedge analytical instruments presently available, and (4) perform a novel dynamic crystallization experiment to evaluate the influence of magma flow and shearing on crystal nucleation. The proposed effort to understand bubble and crystal nucleation kinetics hasfar-reaching implications for the field of physical volcanology, which is increasingly concerned with phase transformations occurring during magma ascent.
