MOF Geometric Screening Dataset v1: Pore-Limiting Diameter, Cavity Size, Density, and Target-Range Scores for 12,597 Structures

This dataset provides a geometry-based pre-screening and ranking of 12,597 lattice-verified metal–organic framework (MOF) structures, derived from an initial pool of ~71,500 candidates via geometric filtering (pore-limiting diameter ≥ 3.3 Å, crystal density ≥ 0.1 g/cm³, and related structural criteria). Rather than ranking structures by simple maximization of pore size or minimization of density — an approach known to favor extreme, often mechanically unstable frameworks — structures are scored against a trapezoidal target-range function that rewards geometric properties falling within ranges considered reasonable for stable, practically usable MOFs. The dataset is intended as a first-pass candidate shortlist to prioritize which structures are most worth carrying forward to grand canonical Monte Carlo (GCMC) adsorption screening, not as a substitute for it. This workbook contains the full geometric pre-screening and ranking of 12,597 lattice-verified metal–organic framework (MOF) structures, organized into six sheets: Documentation — methodology overview, data coverage statistics, scoring-function definitions, and ranking rules. Validation_Addendum — independent quality-control notes: verification that the scoring formulas reproduce exactly from raw values, and two flagged limitations (a large score plateau affecting 42.6% of structures, and duplicate entries arising from alternate relaxation protocols). Full_Dataset_Ranked — all 12,597 structures with raw geometric properties (PLD, LCD, density, void fraction where available), computed scores (LCD_score, density_score, core_score, VF_score, full_score), overall rank, and base-framework/duplicate-variant flags. Top_100_Shortlist — the 100 highest-ranked structures, live-linked to Full_Dataset_Ranked. Top_100_Unique_Shortlist — the 100 highest-ranked structures after deduplicating alternate relaxation variants of the same base framework, for diversity-aware candidate selection. Distribution_Charts — histograms of PLD, LCD, and density, and a scatter plot of the shortlist. All scores were computed using a trapezoidal target-range function (see accompanying reproduce_scores.py and PROVENANCE.md for the exact formulas, computational environment, and independent reproduction check). Scope and limitations: This is a purely geometric ranking. It does not capture chemical selectivity (e.g., CO₂ over N₂/H₂O), open-metal-site availability, adsorption enthalpy, framework flexibility, or hydrolytic stability under humid conditions, and should not be interpreted as a predictive performance ranking.