SCIENCE AT THE EDGE SEMINAR Friday, September 6, 2002 11:30 a.m., Room 1400 Biomedical and Physical Sciences Building Refreshments at 11:15 outside Room 1400 PREDICTIVE THEORY FROM MOLECULES TO MATERIALS* Rodney J. Bartlett Graduate Research Professor Quantum Theory Project Departments of Physics and Chemistry University of Florida ABSTRACT-- Today, a great deal of chemistry is done using quantum mechanics and mathematics instead of in experimental laboratories. This is made possible by describing the electrons in molecules and using very fast computers to approximately solve the equations. From these solutions to the Schroedinger equation, in principle, we can know essentially all there is to know about individual molecules including their structures, spectra, energetics, reaction paths, etc. For small molecules, the results obtained serve in the absence of experimental observation of the molecule. This introduces into chemistry the opportunity to use predictive quantum chemistry as a microscope to look for unusal bonding patterns and new ways to create very high energy molecules, prior to any experimental observation or synthesis. In this talk I'll discuss some polynitrogen molecules that don't exist but should, and how their eventaul synthesis will likely provide a new source of highly energetic rocket fuels, while also greatly expanidng upon our knowledge of the chemsitry of solely notrogen containing molecules. When we leave small molecules, reliable approximate quantum mechanical solutions are far more difficult to obtain, but if we are to reliably describe materials or large biomolecules and they phenomena they undergo, we require retaining the essence of predictve quantum chemistry in these studies for much more complicated systems. Only then can the results of computer simulations be believed. We will consider computer simulations of the fracture of amorphous silica and the role of water in accelerating its fracture to illusrate our use of a 'transfer Hamiltonian' as a way to introduce quantum chemical forces in large scale molecular dynamics, and their important consequences. Short movies will be shown. *This work is supported by the AFOSR and the NSF.