High-throughput Reaction Mechanism Exploration with Multi-reference Treatment

3,335 unique molecular structures and 2,227 elementary reaction steps from an automated high-throughput computational chemistry exploration. The data includes atomic coordinates, spin multiplicity, charge, and energies calculated at multiple quantum chemistry levels (DFT, HF, CCSD). It was produced by Jan P. Unsleber of ETH Zurich as supporting information for a journal article.

Use Cases

• Benchmarking quantum chemistry methods based on the provided DFT, HF, CCSD, and CCSD(T) energy calculations.
• Training machine learning models for energy prediction using the structure-energy pairs.
• Analyzing reaction networks based on the graph of elementary steps linking reactant, product, and transition state structures.
• Studying barrierless reaction dynamics based on the 1,260 identified barrierless steps.

Strengths

• Contains 3,335 de-duplicated molecular structures with atomic coordinates and electronic states.
• Includes 2,227 elementary reaction steps, with 1,260 explicitly identified as barrierless.
• Provides energies calculated at four levels of theory (PBE-D3BJ, HF, CCSD, CCSD(T)) for comparative analysis.

Limitations

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

Provenance

Source

ETH Zurich

Collection Method

Generated via automated high-throughput ab initio computational exploration in the cloud.