Nucleation and Medium Range Order in Silicate Liquids: Inferences from NMR Spectroscopy
- PDF / 775,396 Bytes
- 12 Pages / 414.72 x 648 pts Page_size
- 73 Downloads / 164 Views
amorphous materials. However, in organic molecules in solution, NMR has now become the method of choice for determining both bonding linkage topology and distances out to three, four, or even more bonds from a given carbon or hydrogen atom. The conformation of relatively large protein molecules (containing thousands of atoms) in solution can now be determined by NMR, using very high resolution, two-, three-, and even four-dimensional techniques. Organic liquids, with strong couplings to protons and rapid molecular motion to produce narrow NMR lines, are of course favorable cases. However, some of these same approaches (as well as techniques unique to the solid state) may begin to provide insights into the medium-range structure of inorganic materials as well. SILICATE GLASS AND LIQUID STRUCTURE Many studies of silicate glass structure have been done with NMR, beginning with now-classical work by Bray and colleagues on the distribution of borate species in borate and borosilicate glasses [4, 5]. More recently, most studies have used high resolution magic angle spinning (MAS) NMR of a variety of nuclides, especially 11B, 29Si, 27 A1, and 170. Several thorough reviews have been published recently [6, 7, 8]. Many of the most important findings of these studies have concerned short range structure, in particular the quantification of the distribution of coordination numbers for network forming cations B, Si, and Al. In situ, high temperature NMR studies are beginning to provide information on temperature effects on at least the average coordination numbers of Si, Al, Na, and Mg [9, 10, 11, 12, 13]. Such work is useful to constrain structure-based models of liquid and glass energetics and dynamics, which in turn are needed in any model of nucleation of liquid or crystalline phases. One type of model, for example, relates the thermodynamic activity of a crystallizing component, and thus the driving force for nucleation, to the abundance of a structurally similar species in the liquid. Although it is now clear for at least high-silica liquids that the lifetimes of large silicate "molecules" are probably too short to be of much consequence to structural relaxation and diffusion (see below), at the most local scale such "quasi-crystalline" approaches may have validity. For example, the mean coordination number of Al in high temperature aluminate, aluminosilicate, and fluoride melts has been estimated from in situ NMR studies [11, 12, 14], and is probably a key factor in controlling the crystallization behavior. In many theories of nucleation (or any other rate process modeled using simple energy barriers), activation energies are the most readily predictable and testable quantities. Understanding the structure of hypothesized activated complexes or transition states is thus important. For example, the addition of single Si0 4 tetrahedra to a growing silicate crystal may be the fundamental growth step [15]. For this to occur, one or more oxygen ions must diffuse away from the surface after attachment has occurred. A commonly
Data Loading...
