Combining crystal engineering and surface engineering to estimate the structure -functions relationship of Tafamidis solid state forms with the aid of machine learning.
Journal:
International journal of pharmaceutics
Published Date:
Dec 16, 2025
Abstract
The mutual benefits of surface engineering and crystal engineering led to the discovery of a pharmaceutical cocrystal with balanced biopharmaceutical properties. The surface engineering of a pharmaceutical API (Active Pharmaceutical Ingredient), namely, Tafamidis revealed the need of surface modification for optimized bio-performance characteristics. An unusual API cocrystal, namely, Tafamidis PEG cocrystal was synthesized by following a crystal engineering protocol. Tafamidis-PEG cocrystal contains a supermolecule built from two molecules of self-assembled Tafamidis (carboxylic acid dimer) and one molecule of PEG. The molecular hydrogen bonding pattern, surface chemistry, physicochemical properties and solubility of Tafamidis PEG cocrystal, were compared to that of the free API (Tafamidis) and API salt (Tafamidis meglumine salt). The molecular hydrogen bonding recognition of Tafamidis and Tafamidis-PEG cocrystal were established through various spectroscopic techniques, IR, Raman, NMR, and XPS. The intermolecular H-bond interactions between two heterosynthons brought significant changes in the molecular geometry and crystal properties of Tafamidis-PEG cocrystal. The H-bond interactions between donor and acceptor groups were reasoned considering Natural Bond Orbital (NBO) theory. The surface chemistry / properties were studied by various interfacial interactions techniques, such as Inverted Gas Chromatography Surface Energy Analyzer (IGCSEA), Contact Angle Measurement (CA), Surface Area Analyzer (SSA) and Dynamic Vapor Sorption Analyzer (DVS). The inter-intra relationships between various variables of structure-function domains were estimated with the aid of machine learning. The relationship between molecular structure, surface / crystal properties and solubility of the materials were found to be linear. The predictable structure / functions relationships helped to identify the most suitable material for pharmaceutical applications among free API, API salt, and API Cocrystal.
