Multiobjective Design of Electrolyte Solvents via Physics-Based Modeling and Reinforcement

A dataset of molecular candidates for electrochemical devices, generated using a reinforcement learning framework. The molecules were designed to optimize multiple properties, including viscosity, boiling point, and HOMO energy, with validation from physics-based simulations. The dataset was authored by Mohammad Atif Faiz Afzal and last updated in April 2026.

Use Cases

• Benchmarking molecular generation algorithms based on the described reinforcement learning framework.
• Analyzing trade-offs between viscosity and boiling point for electrochemical device materials.
• Training surrogate machine learning models for molecular properties like HOMO energy.
• Validating physics-based simulation methods for boiling point and viscosity estimation.

Strengths

• Includes 200 top-performing candidate molecules validated with physics-based simulations.
• Properties were computed using established methods: molecular dynamics for viscosity/boiling point and density functional theory for HOMO energy.
• Dataset size is 63.4 MB, indicating a non-trivial collection of molecular data.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

figshare

Collection Method

Generated via a reinforcement learning-based framework using SMILES strings, with property predictions from trained machine learning models and validation via physics-based simulations.

Time Range

null

Freshness

Last updated 2026-04-23 19:43:48; freshness should be verified.

Geography

null