Data-Driven Synthesis of Covalent Organic Frameworks with Integrated Success-Failure Data

A dataset of 4,425 synthesis records for covalent organic frameworks (COFs), integrating 1,822 in-house failed attempts and 2,603 successful cases from the literature. The data was used to train a random forest model achieving 91% accuracy in solvent prediction and high R² scores for reaction temperature and time. The dataset was created by Bing Ma and last updated on 2026-05-21.

Use Cases

• Train ML models for solvent prediction in COF synthesis based on integrated success and failure records.
• Predict reaction temperature and time for novel COF frameworks based on the described random forest model.
• Validate synthesis routes for structurally complex COFs, such as M(salen)-COFs and M(Rabson)-COFs, using the provided experimental data.

Strengths

• Integrates 1,822 failed synthesis records with 2,603 successful cases, providing a balanced training set.
• The derived ML model achieved 91% accuracy in solvent prediction and R² scores of 0.9129 for temperature and 0.9596 for time.
• Experimental validation successfully synthesized eight structurally complex and synthetically challenging COFs under predicted conditions.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data size is 598.7 KB, indicating a limited scope.

Provenance

Source

figshare

Collection Method

Curated from in-house failed synthesis records and successful cases extracted from the literature.

Freshness

Last updated 2026-05-21 09:21:10; freshness should be verified.