Combinatorial materials science

Article By:

Ginley, David National Renewable Energy Laboratory, Golden, Colorado.

Teplin, Charles National Renewable Energy Laboratory, Golden, Colorado.

Taylor, Matthew National Renewable Energy Laboratory, Golden, Colorado.

van Hest, Maikel National Renewable Energy Laboratory, Golden, Colorado.

Perkins, John National Renewable Energy Laboratory, Golden, Colorado.

Last reviewed:2005

DOI:https://doi.org/10.1036/1097-8542.YB050670

Traditionally in materials science, researchers performed an experiment, analyzed the results, and then used these results to decide on the next experiment. This model of experimentation is a clear application of the scientific method and has been the foundation of many significant discoveries. However, the use of computers for both data acquisition and analysis has enabled high-throughput approaches to materials discovery experiments, that is, parallel rather than serial sample preparation, characterization, and analysis. Such experimentation is known as the combinatorial method, whereby many nearly simultaneous experiments are performed and analyzed at each step. This approach can greatly accelerate the rate at which science can be done and knowledge acquired. Combinatorial experimental methods are becoming common in a number of areas, including catalyst discovery, drug discovery, polymer optimization, phosphor development, and chemical synthesis. For example, complex catalysts have been discovered 10 to 30 times faster using the combinatorial method rather than conventional approaches. Perhaps the biggest challenge of combinatorial science is in analyzing rapidly the vast amount of data (hundreds, thousands, or millions of experiments) for important trends and results. This is typically accomplished by using data mining software.

The content above is only an excerpt.