Kernel learning assisted synthesis condition exploration for ternary spinel
Yutong Liu † \dagger, Mehrad Ansari † \dagger , ∗ \ast, Robert Black, Jason Hattrick-Simpers ∗ \ast · 2025
Read Kernel learning assisted synthesis condition exploration for ternary spinel by Yutong Liu † \dagger, Mehrad Ansari † \dagger , ∗ \ast, Robert Black, Jason Hattrick-Simpers ∗ \ast (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-19637.txt.
Description
Yutong Liu and three co-authors used an interpretable machine learning model to find which recipe choices make a mixed-metal oxide form as a single clean phase. The target was an iron-zinc-cobalt spinel, made by high-throughput co-precipitation, a fast method that still leaves many failed or mixed products. Their kernel classifier learned to separate single-phase successes from the rest.
To explain the model, they applied a global form of SHAP analysis, which prices each experimental feature's contribution to the verdict. The ranking pointed at the precursors and the precipitating agent as the decisive ingredients. Those answers matched established crystal growth theory, a reassuring sign that the model learned chemistry rather than noise. The authors frame the approach as a template: pair fast automated synthesis with explanations a chemist can argue with, and experimental design gets smarter rather than merely faster.
Catalog record
| CREATOR | Yutong Liu † \dagger, Mehrad Ansari † \dagger , ∗ \ast, Robert Black, Jason Hattrick-Simpers ∗ \ast |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 43 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-19637