Statistical Mechanical Approximations to More Efficiently Determine Polymorph Free Energy Differences for Small...

Area of Science:

• Crystallography
• Computational Chemistry
• Materials Science

Background:

• Accurate prediction of crystal polymorph stability is essential for crystal structure predictions (CSPs).
• Current methods like quasi-harmonic approximations (QHA) and replica exchange molecular dynamics (REMD) have limitations in capturing dynamic structural changes.
• REMD simulations reveal that many organic crystal structures can rearrange to lower energy states upon heating, a phenomenon not captured by QHA.

Purpose of the Study:

• To develop and evaluate intermediate computational methods for determining the relative stability of organic crystal polymorphs.
• To bridge the gap in accuracy and computational cost between QHA and full thermodynamic cycle calculations using REMD.
• To identify a cost-effective yet accurate approach for screening crystal polymorph stability.

Main Methods:

• Development of intermediate computational methods combining aspects of QHA and REMD.
• Utilizing REMD simulations to explore the energy landscape of organic crystals.
• Boltzmann weighting of harmonic free energies from REMD trajectories.

Main Results:

• The proposed intermediate methods achieve significant computational savings (42-80%) compared to full REMD.
• These methods introduce a small error (0.16 ± 0.04 kcal/mol) relative to the full thermodynamic cycle.
• A method involving Boltzmann-weighted harmonic free energies from REMD trajectories demonstrated particular promise.

Conclusions:

• Intermediate computational methods provide a viable and efficient alternative for assessing crystal polymorph stability.
• These methods offer a practical balance between computational expense and predictive accuracy.
• The Boltzmann-weighted harmonic free energy approach is recommended for screening crystal polymorph stability in CSPs.