Computational informatics: Guided discovery for combinatorial experiments

I. INTRODUCTION Combinatorial sciences offer an attractive paradigm to search for new materials. The challenge is to expand this paradigm from one of high-throughput screening to high-throughput knowledge. In order to accomplish this, there is a need to integrate the knowledge of the physics of the materials process into the statistically based strategies governing the combinatorial synthesis. In this chapter, we outline some of the challenges and approaches that exist to developing such integration. Materials discovery is still a process governed by empiricism and accidental discoveries (high-temperature ceramic superconductors and carbon fullerenes, to mention some recent examples). While incremental progress is made in specific technological areas of interest, we need to have a means of exploring vast combinations of structure-property relationships. If new, significant advances in materials science are to be made, we need to have search tools that can accelerate the discovery process.