Development of Practical MO Techniques for Prediction of the Properties and Behavior of Materials
Abstract
Our previous semiempirical treatments have been based on the ZDO approximations introduced by Pople, MNDO and AMI using the best of these (NDDO). Although there were good reasons to believe that the ZDO approximation is wholly unacceptable, we were forced to use it in order to obtain a treatment of practical value, i.e. one that could be applied directly to the molecules of interest to organic chemists rather than to meaninglessly simplified 'models', using generally available computers. Although the resulting treatments proved remarkably successful and have led to major developments in chemical theory, in particular in studies of reaction mechanisms. The situation has, however, changed dramatically in recent years with almost incredible increase in computing power of minicomputers due to the invention of the integrated circuit. The primary purpose of the research supported by this grant was to make use of the greatly improved computing facilities now available to us to develop a semiempirical treatment based on a full LCAO SCF MO approximation, i.e. one in which overlap is retained.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 31, 1994
- Accession Number
- ADA285079
Entities
People
- Michael J. Dewar
Organizations
- University of Florida