High-Throughput Experimentation Datasets for Palladium-Catalyzed Suzuki-Miyaura and Buchwald-Hartwig Cross-Coupling Reactions in SURF Format

This dataset collection contains High-Throughput Experimentation (HTE) data for palladium-catalyzed cross-coupling reactions formatted in the Standardized Uniform Reaction Format (SURF). The experimental procedure that generated the data is precisely described in the HTE OS paper. The datasets include experimental results from Suzuki-Miyaura (SM) and Buchwald-Hartwig (BH) coupling reactions. The "positive" datasets contain experiments where product formation exceeded 5% total area percent (product_1_area% > 0.05), while "all" datasets include both successful and unsuccessful reactions. The collection comprises four CSV files:• sm_all.csv – 3,426 Suzuki-Miyaura reactions (all outcomes including failures)• sm_positive.csv – 1,878 Suzuki-Miyaura reactions (significant product formation only)• bh_all.csv – 10,138 Buchwald-Hartwig reactions (all outcomes including failures)• bh_positive.csv – 3,441 Buchwald-Hartwig reactions (significant product formation only) Each reaction record includes:• Reaction conditions (temperature, time)• Starting materials with SMILES representations and molar equivalents• Catalyst and ligand information• Base/reagent details• Solvent system• HPLC area percentage of product as a yield proxy These datasets are suitable for machine learning model development for reaction outcome prediction, structure-activity relationship analysis, and benchmarking yield prediction algorithms. Record DOI: 10.5281/zenodo.18185850 This publication was created as part of NCCR Catalysis (grant number 225147), a National Centre of Competence in Research funded by the Swiss National Science Foundation