Tuesday, November 6, 2007 - 5:15 PM

The Predictive Power Of Atomically-Detailed Simulations

Karl Johnson1, Sudhakar V. Alapati2, Bing Dai3, and David S. Sholl2. (1) University of Pittsburgh, 1249 Benedum Hall, Department of Chemical Engineering, Pittsburgh, PA 15261, (2) Carnegie Mellon University, Department of Chemical Engineering, Carnegie Mellon University, Pittsburgh, PA 15213, (3) Chemical Engineering, University of Pittsburgh, 1249 Benedum Hall, Pittsburgh, PA 15261

Professor Gubbins has been one of the foremost advocates of the use of molecular simulations as a tool for predicting the behavior (thermodynamics, phase transitions, structure, etc.) of macroscopic systems. This talk focuses on the use of quantum mechanical methods for predicting the behavior of macroscopic systems, specifically considering the thermodynamics and kinetics of materials related to hydrogen storage in the solid state.

The on-board storage of hydrogen is one of the most vexing problems associated with the development of viable fuel cell vehicles. The challenge is to safely and efficiently store hydrogen at high volumetric and gravimetric densities. Hydrides of period 2 or 3 metals can store hydrogen at high gravimetric and volumetric densities. However, existing hydrides either have unacceptable thermodynamics or kinetics. New materials for hydrogen storage are therefore needed. We demonstrate how first principles density functional theory (DFT) can be used to screen potential candidate materials for hydrogen storage. We have used DFT calculations to screen over 300 reactions that have not been investigated experimentally. We have identified several interesting destabilized metal hydrides having favorable thermodynamics. We have predicted the van't Hoff plots for many of the promising candidate reactions identified though our modeling and as part of this work we have predicted van't Hoff plots for known reactions and compared our results with experimental data where available. We have also examined the effect of substitutional alloying on the thermodynamics of different metal hydrides.

We have studied the kinetics of selected surface reactions on several hydride-related materials using DFT. Our calculations help explain the difficulty of hydrogenating Mg2Si, an otherwise promising destabilized metal hydride. Our calculations indicate that the initial dissociation of H2 on Mg2Si is facile at room temperature and above. However, our calculations also show that Mg2Si is very susceptible to oxidation, which completely passivates the surface toward hydrogen dissociation.