In contrast to combinatorial chemistry efforts in drug discovery where hundreds of compounds (libraries of products) are desired, in process development only one molecule is of interest and synthetic ways of producing this molecule are the main focus (''libraries'' of reaction conditions). Thus, the evaluation of many possible combinations of reaction parameters has to be achieved by parallelization, automation, and miniaturization. Furthermore, split-and-mix approaches cannot be applied in this setting.
Proper identification of reaction conditions requires very carefully controlled reaction parameters and the acquisition of many data sets including, for example, the control of nonuniform temperatures and stirring rates. Furthermore, standardized and repetitive runs have to be included to allow proper statistical validation of the reaction yields and conditions.
Within the area of process development for basic chemicals, parallelization is mostly used for the development of heterogeneous catalysts and, here, predominately for gas-phase reactions. Catalyst screenings are classified by their degree of parallelization, the complexity, and type of information gained from each experiment. A typical primary screening classifies a large number of solvents or catalysts as ''good'' or ''bad'' according to a set of standard conditions, whereas a quaternary screening gives detailed information for a single catalyst and an optimized set of process parameters. These expressions do not correspond to the nomenclature used in the context of process development for fine chemicals. The discussion of combinatorial chemistry in heterogeneous catalysis would go beyond the scope of this article and the reader is referred to the corresponding literature. A current list of examples, where combinatorial methods have been used, can be found in an article by Senkan . Recently, a special issue was devoted to topical developments in combinatorial heterogeneous catalysis .
Was this article helpful?