Development of Integrated Actinide Chemistry Application, AACE, for Acceleration of Actinide Chemistry Experiments

¹ Tokyo Institute of Technology, Institute of Innovative Research, laboratory for Zero-Carbon Energy, 152-8550, 2-12-1 Ookayama, Meguro-ku, Tokyo, Japan
² Nagoya University Graduate School of Medicine, Department of Biostatistics, 466-8550, 65 Tsurumai-cho, Showa-ku, Nagoya city, Japan

^* Corresponding author

Published online: 31 January 2025

Abstract

Efficient separation of Minor Actinide (MA) from High-Level Liquid Waste (HLLW) is essential in next generation reprocessing systems. To enable this, finding an efficient separation system is inevitable. We employ a chemoinformatic approach to find suitable diluents, particularly fluorinated diluents, due to some advantages. To efficiently find the candidate molecules we developed Acceleration of Actinide Chemistry Experiment (AACE), which can deploy transfer learning (TL) and human-in-the-loop machine learning (HITL-ML). Our approach utilizes Hansen’s solubility parameters derived from molecular structures to predict solubility and extractability, create extraction models for MA surrogate Lanthanide (Ln) and MA. This manuscript outlines the methodology for diluent exploration and the function of AACE.