SFRIS: Optimized Molecular Structures for 17 Benchmark Systems

All optimized molecular structures generated by the SFRIS (Systematic Fragmentation and Recombination Isomer Search) algorithm, as described in the accompanying manuscript. The dataset contains structures for 17 benchmark systems (7–17 atoms; C, H, O, N, and S), organized into directories for main runs and robustness analyses. Structures were optimized at the B3LYP-D3(BJ)/def2-TZVP level of theory using ORCA 6.1.1 and are provided in XYZ format with energies.

Use Cases

• Benchmarking isomer search algorithms based on the provided 17 molecular systems.
• Analyzing molecular geometry and energy relationships using the DFT-optimized structures and included energies.
• Testing the robustness of computational chemistry pipelines using the multiple independent runs provided for some molecules.
• Developing machine learning models for molecular property prediction using the optimized 3D structures.

Strengths

• Structures are optimized at a consistent, high level of theory (B3LYP-D3(BJ)/def2-TZVP).
• Includes robustness analysis data with up to five independent runs per molecule for some subsets.
• Covers 17 benchmark molecular systems containing 7–17 atoms and elements C, H, O, N, and S.
• Energies (in Hartree) are included in the comment line of each XYZ file.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The dataset's 9.7 MB size suggests a relatively small number of molecular structures.

Provenance

Source

DAPENG ZHANG via figshare

Collection Method

Generated by the SFRIS algorithm, with structures optimized using ORCA 6.1.1 and initial screening with xTB 6.7.1.

Freshness

Last updated 2026-04-14 03:45:25; freshness should be verified.