DESIGN OF ALCOHOL SYNTHESIS CATALYSTS ASSISTED BY A KNOWLEDGE-BASED EXPERT SYSTEM

A new knowledge-based, expert system approach is presented for the design of a family of heterogeneous catalysts for alcohol synthesis. The knowledge base for solving this problem is a hybrid consisting of numerical data, declarative facts, and heuristic rules. Design rules were derived from an integration of the information about the performance of the catalyst, reaction mechanism, and the nature of the active sites. An example for a specific case is constructed on the basis of compositional and kinetic data for mixed metal oxide catalysts used to reduce carbon monoxide with hydrogen to alcohols. An inference engine was constructed by simulating the kine of thinking that an experienced catalyst designer employs. By this method, an expert system can guide the search for an optimum catalyst formulation, its preparation procedure, and the best reaction conditions to achieve the specified selectivity of the desired product. In this way, the expert system can provide a path to the design of the catalyst through a systematic and logically structured database that is accessible and interactive.