Data-Driven Classification of Solubility Space in Deep Eutectic Solvents: Deciphering Driving Forces Using PCA and...

Area of Science:

• Pharmaceutical Science
• Physical Chemistry
• Computational Chemistry

Background:

• Traditional drug formulation relies on empirical screening, which is inefficient.
• Deep eutectic solvents (DESs) offer tunable properties for drug delivery but require rational design for optimal solubility.
• Predicting drug solubility in DESs is crucial for effective pharmaceutical development.

Purpose of the Study:

• To develop a data-driven framework for classifying and predicting drug solubility in DESs.
• To enable rational formulation design by moving beyond empirical approaches.
• To identify key factors influencing drug solubility in various DESs.

Main Methods:

• Analysis of 2010 solubility measurements for 21 pharmaceutical compounds across multiple DESs.
• Principal Component Analysis (PCA) to reduce 16 COSMO-RS descriptors into four interpretable dimensions.
• K-means clustering to identify distinct drug-DES solubility regimes.

Main Results:

• PCA identified two key factors: global solvation propensity (PC1) and specific interaction complementarity (PC2).
• Four distinct solubility regimes were identified, categorizing DES-drug combinations.
• Choline chloride showed broad utility, while menthol and betaine exhibited specialized roles.
• Case studies demonstrated how solubility regimes guide formulation strategies.

Conclusions:

• The integrated PCA-clustering framework transforms DES development from screening to targeted design.
• The taxonomy provides formulation scientists with a rational approach for DES selection.
• This method accelerates sustainable pharmaceutical formulation by offering fundamental insights into solubility mechanisms.