Laser Heated Gradient NMR Studies of Ceramic Liquids

Arizona State University, Scottsdale AZ

Investigators

Abstract

The growing technological importance of ceramics has highlighted the need for new studies of the fundamental physical and chemical properties of ceramic melts, where temperatures are very high (up to and beyond 2,000 Celsius) and where much industrial processing is performed. We have built the apparatus required for both spectroscopic and time-domain NMR measurements of the properties of insulating materials at ultra-high temperatures, and have to date determined both chemical shifts and nuclear spin relaxation times of 27Al in several molten ceramics between 1700 and 2600 C. This apparatus will be used to extend these measurements in well-known refractory systems, e.g. silica-alumina, calcium-alumina and yttrium-alumina, as well as in new systems under investigation. In addition, we plan to further develop the technique of ultra high-temperature NMR measurements by incorporation of a magnetic field gradient into our NMR probe. This has the twin purposes of improving our understanding of sample temperature gradients in the apparatus, and of implementing the direct measurement of diffusivities. With the addition of these innovations to the well-known repertory of basic measurements, the application of NMR has the potential for significantly increasing our understanding of liquid phases over a broad range of ceramic systems. In molten ceramics, however, measurements are particularly challenging. The high temperatures and strong reactivities of a typical ceramic liquid's constituents preclude the use of sample containers. In our experiments a small sphere of material, about three mm in diameter, is levitated in an upward flow of argon gas and heated above ~2000 C by a CO2 laser. For NMR measurements the sample must also be located in a large magnetic field, at the center of a superconducting magnet. Sample temperature and composition may be varied, and the NMR data provide direct information about the effect of these variations upon the average chemical environments of different nuclei in the liquid. Diffusion rates can also be obtained. From this information a clearer understanding emerges, of the atomic scale structural and dynamic properties of molten ceramics. With this, we can better understand the macroscopic properties of a melt, for example its viscosity, and better control the properties of the resulting ceramic material for many applications. This project is co-funded with AFOSR (6.1).