Tracking materials science data lineage to manage millions of materials experiments and analyses

Abstract

In an era of rapid advancement of algorithms that extract knowledge from data, data and metadata management are increasingly critical to research success. In materials science, there are few examples of experimental databases that contain many different types of information, and compared with other disciplines, the database sizes are relatively small. Underlying these issues are the challenges in managing and linking data across disparate synthesis and characterization experiments, which we address with the development of a lightweight data management framework that is generally applicable for experimental science and beyond. Five years of managing experiments with this system has yielded the Materials Experiment and Analysis Database (MEAD) that contains raw data and metadata from millions of materials synthesis and characterization experiments, as well as the analysis and distillation of that data into property and performance metrics via software in an accompanying open source repository. The unprecedented quantity and diversity of experimental data are searchable by experiment and analysis attributes generated by both researchers and data processing software. The search web interface allows users to visualize their search results and download zipped packages of data with full annotations of their lineage. The enormity of the data provides substantial challenges and opportunities for incorporating data science in the physical sciences, and MEAD's data and algorithm management framework will foster increased incorporation of automation and autonomous discovery in materials and chemistry research.

Benchmarking the acceleration of materials discovery by sequential learning

“…The datasets for simulated SL were constructed from high throughput experiments described previously. 24,28 Parallel synthesis and processing of a composition library proceeded with inkjet printing 29 of elemental precursors to produce a discrete library with 2121 unique compositions comprising all possible unary, binary, ternary and quaternary compositions from a 6 element set with 10 at% intervals. Following conversion to metal oxide samples via calcination at 400 C for 10 hours, accelerated aging of the catalysts is performed via parallel operation for 2 hours.…”

Combinatorial screening yields discovery of 29 metal oxide photoanodes for solar fuel generation

“…The 1-dimensional composition gradient on a 2-dimensional substrate also provides nominally duplicate samples that can be evaluated for reproducibility and/or under different electrochemical conditions, such as different electrolyte pH. Mining the photoanode experiments in the materials experiments and analysis database (MEAD) 18 resulted in enumeration of the 29 metal oxide phases in Table 1 that to the best of our knowledge have not been reported as photoanodes.…”

Benchmarking the Acceleration of Materials Discovery by Sequential Learning

“…The datasets for simulated SL were constructed from high throughput experiments described previously. 24,28 Parallel synthesis and processing of a composition library proceeded with inkjet printing 29 of elemental precursors to produce a discrete library with 2121 unique compositions comprising all possible unary, binary, ternary and quaternary compositions from a 6 element set with 10 at.% intervals. Following conversion to metal oxide samples via calcination at 400 • C for 10 hours, accelerated aging of the catalysts is performed via parallel operation for 2 hours.…”

“…Each collection of 2121 FOMs is treated as an independent dataset for sequential learning simulation, and each such dataset contains considerable catalyst composition diversity with 6, 15, 20, and 15 unary, binary, ternary and quaternary composition spaces, respectively. The different plates and their respective plate ID in our database 28 and composition system are shown in Table 1, and the distribution of catalyst overpotentials in each dataset is shown in Fig. 1.…”