Study on Phosphorus Compound/Catechol-Catalyzed Dehydrative Amidation and Its Database Development for Machine...

Area of Science:

• Organic Chemistry
• Catalysis
• Computational Chemistry

Background:

• Dehydrative amidation is crucial for synthesizing amides, essential in pharmaceuticals.
• Phosphorus compounds with P─H bonds are known catalysts, but their efficiency can be limited.
• Anilines are often unreactive substrates in amidation reactions.

Purpose of the Study:

• To enhance the catalytic activity of phosphorus compounds in dehydrative amidation reactions.
• To identify effective catechol derivatives as additives for these reactions.
• To develop a machine-learning model for predicting catalytic performance.

Main Methods:

• Systematic screening of four phosphorus compounds and 17 catechol derivatives.
• Utilizing dimethyl phosphite and 2,3-dihydroxynaphthalene as an effective catalytic combination.
• Developing multilayer perceptron (MLP) machine-learning models with quantum chemical and RDKit descriptors.

Main Results:

• Substantial enhancement of catalytic activity using catechol derivatives as additives.
• Successful amidation of carboxylic acids with amines, including challenging aniline substrates.
• MLP models accurately predicted catalytic effectiveness and provided chemical interpretability via SHAP analysis.

Conclusions:

• Catechol derivatives are effective additives for phosphorus-catalyzed dehydrative amidation.
• The developed catalytic system offers a new route for pharmaceutical synthesis, particularly for anilines.
• Machine learning models can successfully predict and interpret catalytic performance in organic reactions.