(386a) A Cyber-Infrastructure for Catalysis Science

Authors: 
Delgass, W. N., Purdue University
McLennan, M. J., Purdue University
Krishnamurthy, B. B., Purdue University
Nandi, S., Purdue University
Delgass, L. L., Purdue University
Wang, H., Purdue University
Arns, L. L., Purdue University
Dunlop, S. R., Purdue University
Hsu, S., Purdue University
Lasinski, M. E., Purdue University
Orcun, S., Purdue University
Blau, G., Purdue University
Goyal, A., Purdue University
Cao, J., Purdue University
Midkiff, S. P., Purdue University


Progress in the development of a cyber infrastructure for catalysis science will be reported. The cyber infrastructure includes a number of components. An e-Lab Notebook has been developed that enables ingress of XML annotated data into an ontologically-enabled database, including digital signature and electronic notarization for IP certification. The database supports standard SQL searches. In addition, the database is currently being extended to support ontologically based searches so that the user can employ higher level logic in querying the database. An interactive visualization environment has been developed that includes multiple spreadsheets, 2D graphics and 3D graphics all of which are linked and connected to the database and archived copies of the original e-Lab Notebook data. For example, a user can click on an outlier on one 2D plot and the software will immediately highlight that data point on all other 2D and 3D plots as well as connected cells in the spreadsheets, and if desired directly access the original data in the e-Lab Notebook. The visualization environment is scaleable from a single desk top monitor, to a multi-head display, to a large tiled wall. The visualization environment is connected to MatLab or user-written C or Fortran code for data analysis. Special purpose nonlinear statistical analysis tools have been developed for model building and will be integrated into the analysis environment. Special purpose chemical compilers have been written for automatically generating and solving large sets of ODEs to describe complex chemical kinetics, where the equations are generated from near-English language chemistry rules. Finally, for the computationally intense parts of this project, we have developed code suitable for massively parallel computers on the TeraGrid. The critical feature of this cyber infrastructure for catalysis science is not the individual components of the infrastructure, but rather that the components are integrated in to an organic whole that can be used by catalysis researchers that want to focus on catalysis science rather than the cyber details.