Computation screening for incorrectly determined cocrystal structures.

Area of Science:

• Solid-state chemistry
• Computational materials science
• Pharmaceutical formulation

Background:

• Pharmaceutical salts and cocrystals are critical for drug formulation, but regulatory distinctions are significant despite minor structural differences.
• Accurate differentiation is essential for drug development and regulatory approval.
• Previous DFT-based methods showed promise but had limitations with certain hydrogen bond lengths.

Purpose of the Study:

• To evaluate the benefits of the rSCAN functional over the PBE functional for distinguishing pharmaceutical salts and cocrystals.
• To expand a computational dataset and validate the method on a larger scale.
• To identify and investigate challenging cases at the salt-cocrystal continuum.

Main Methods:

• Density Functional Theory (DFT) calculations using various functionals (PBE, rSCAN, r2SCAN, PBE0, PBE50).
• Analysis of hydrogen bond lengths, particularly O-H...N interactions.
• Experimental validation through single-crystal X-ray diffraction for ambiguous structures.

Main Results:

• The expanded dataset of 404 models confirmed 301 cocrystals, identified 87 salt-cocrystal continuum forms, and flagged 16 cocrystals as potential salts.
• Experimental investigation of seven problematic structures revealed five to be true salts.
• The r2SCAN functional demonstrated reliability for O-H...N bonds longer than 2.554 Å, offering a balance of accuracy and efficiency.

Conclusions:

• The r2SCAN functional is recommended for accurate and efficient salt-cocrystal differentiation in pharmaceutical solid-state analysis.
• Computational methods, enhanced by advanced functionals, can significantly aid in classifying complex pharmaceutical solid forms.
• This approach helps resolve ambiguities in distinguishing salts from cocrystals, crucial for regulatory compliance and drug development.