Data Infrastructure for High Throughput Materials Discovery

Published

March 7, 2016

Author(s)

Erik A. Pfeif, Kenneth G. Kroenlein

Abstract

Increases in computational capability enabled sophisticated materials design to evolve from trial-and-error approaches towards more informed methodologies that require large amounts of data. Expert designed tools and their underlying databases facilitate modern day High Throughput Computational Methods. Standard data formats and communication standards increase the impact of traditional data, and applying these technologies to high throughput experimental design provide dense, targeted materials data valuable for material discovery. Data curation has many components including selection, collection, organization, analysis, and communication. A multitude of factors dictate selection of experimental targets, with focuses on regions of sparse data or low-quality data. This selection process requires close communication between data curators and consumers. Improving this relationship is related to many of the goals of the Materials Genome Initiative. Integrated computational materials engineering requires both experimentally and computationally derived data. Harvesting this data comprehensively requires different methods of varying degrees of automation to accommodate variety and volume. Issues of data quality persist independent of type due to difficulties associated with the modern journal review process and data collection limitations.

Citation

APL Materials

Volume

4

Pub Type

Journals

Download Paper

DOI Link

Keywords

Materials Property Database, Database infrastructure, High Throughput Experimental Methods, High Throughput Computational Methods

Citation

Pfeif, E. and Kroenlein, K. (2016), Data Infrastructure for High Throughput Materials Discovery, APL Materials, [online], https://doi.org/10.1063/1.4942634 (Accessed May 8, 2026)

Additional citation formats