UT Chem/Biochem Dept

The 2006-2007 Beckman Scholars: Nathan S. Froemming

Faculty Mentor: Professor Keith J. Stevenson Length of term: Summer 06, Fall 06, Spring 07, Summer 07 Honors & Awards:Dean's Scholar (Fall 04 +); University Honors (Fall 03, Spring 04, Fall 04, Spring 05); Undergraduate Research Fellowship (2005, 2006); Intel Foundation Award for Excellence in Physics Research (2006); Norman L. Hackerman Award for Excellence in Chemistry Research (2007) Publications:N. Froemming, G. Henkelman, J. Phys. Chem. (2009), 131 (23), 234103. Ojifinni, Rotimi A. ; Gong, Jinlong ; Froemming, Nathan S. ; Flaherty, David W. ; Pan, Ming; Henkelman, Graeme ; Mullins, C. Buddie JACS 130, no. 34 (2008): 11250-11251. Ojifinni, Rotimi A.; Froemming, Nathan S.; Gong, Jinlong; Pan, Ming; Kim, Tae S.;White, J. M.; Henkelman, Graeme; Mullins, C. Buddie JACS 130, no. 21 (2008): 6801-6812. Where is he now?Graduated with Bachelor of Science in Chemistry: Honors, with Special Departmental Honors in Chemistry, and with Bachelor of Science in Physics: Honors, with Special Departmental Honors in Physics, December 2008. Nathan worked at Sandia National Labs, Albuquerque, was a Citizen Schools Citizen teacher, a research scientist at UT Austin, and is currently in graduate school at the University of Washington, Seattle physics department. How can I contact him? natfro at gmail.com

Beckman research project in the Stevenson Group:

Download a copy of Nathan's Research report, entitled "Survival of the fittest: Using genetic evolutionary algorithms to design better fuel cell catalysts".

Platinum-based fuel cells offer an attractive alternative to internal combustion engines as an energy carrier for the future, however, severe shortcomings of such technologies must be resolved if they are to become practical and widespread. Some of these difficulties include a disparity between the reaction rates at the anode and cathode, energy loss due to a kinetic overpotential for oxygen reduction at the cathode, the short lifetime of electrodes in acidic environments, and the high material cost and limited supply of Pt itself. Better catalysts need to be developed; yet the task of discovering cheaper, more effective platinum alternatives is indeed nontrivial.

The objective of the proposed research is to investigate the catalytic properties of a large number of candidate bimetallic systems by incorporating genetic evolutionary algorithms into the present-day theoretical frameworks of density functional and transition state theories. In this way the increasing speed and accuracy with which such quantum chemistry calculations can be performed will help to reduce the time and cost associated with exploring such catalysts experimentally. Also, by applying genetic evolutionary algorithms within such a framework, the phenomenal combinatorial burden associated with simulating all possible amalgamations, proportions, and configurations of transition metals in such candidate bimetallic catalysts will be assuaged.

The first step toward implementing a genetic evolutionary algorithm within such a context is to develop a way of representing each candidate bimetallic system in terms of a chromosome unique to that particular system. For instance, a chromosome can be defined as a function of the number of atom layers away from the reaction surface as well as the metal composition of each layer. Once it is possible to specify the metal composition and configuration of any candidate bimetallic system in a chromosome, the application of a genetic evolutionary algorithm is relatively straightforward. In the case of the proposed research, I will first randomly generate an initial population of many different candidate bimetallic catalysts of arbitrary metal compositions and configurations and specify their respective chromosomes. Then, using density functional and transition state theories, I will evaluate the fitness of each system in the population based on its ability to catalyze the oxygen reduction reaction (ORR), the so-called "bottleneck" of current fuel cell technologies. Finally, I will select only the best members of the population based on their fitness and "breed" them by crossing their chromosomes, which will generate the next population to which the algorithm will be applied. The steps are repeated, and with time, subsequent generations continually evolve toward better fitness. The fittest catalysts will be able to catalyze the ORR at the fuel cell cathode most effectively. It is the hope that such a theoretically-based search-and-screen approach will not only help to reveal fuel cell catalysts of increased reactivity over platinum, but also that those features responsible for increased reactivity be communicated to and evaluated by experimentalists, and incorporated into the design of actual new, more efficient fuel cell catalysts in the future.