Work

Catalyst Phylogenetic Tree: A Visualization Tool for Exploring Catalyst Research and Development from Tabular Data

Year: 2025

Type: article

Abstract: The rise of data-driven catalyst design has led to an increasing availability of curated catalyst datasets. These datasets likely contain historical trends that could guide research, but extracting such trends from high-dimensional, high-volume data remains challenging, limiting broader use beyond data scientists. This study proposes a catalyst phylogenetic tree, a pipelined method for visualizing vast catalyst data sets to provide an overview of their evolution. It groups catalysts by distinct elemental combinations, termed catalyst sets, and maps their physicochemical distances onto a phylogenetic tree. Applied to two publicly available datasets on oxidative coupling and dry reforming of methane, this method successfully identified catalyst lineages with similar designs emerging across different eras, as well as the standard catalyst designs for each lineage. This approach can be extended beyond catalyst data to various materials, maximizing the value of literature-based data curation and accelerating research and development across diverse fields.

Source: The Journal of Physical Chemistry Letters

DOI: 10.1021/acs.jpclett.5c01010

Authors: Sunao Nakanowatari, Toshiaki Taniike

Institution: Japan Advanced Institute of Science and Technology

Language: English

Topic: Machine Learning in Materials Science

Subfield: Materials Chemistry

Field: Materials Science

Domain: Physical Sciences

Open Access status: green

Funder: JST-Mirai Program

Award: G969364001