Research: Lattice Energies of Molecular Crystals

Many organic molecules can crystallize into multiple possible crystal packing geometries, called polymorphs. These polymorphs can often have very similar energies. This is a challenge to drug development, because drugs on the market can, and sometimes do, recrystallize into an alternative polymorph that might not be effective as a drug (e.g., the famous case of the drug ritonavir). Thus, being able to compute whether a drug is likely to recrystallize into an alternative polymorph would be greatly desirable. Because of the very small energy differences between some polymorphs, a key prerequisite is the abilty to compute lattice energies of organic molecular crystals very accurately.

Unfortunately, the most popular method for computing crystal lattice energies, density functional theory, does not have the requisite accuracy to reliably discriminate between polymorphs. One must instead employ beyond-DFT quantum chemistry methods. Our group has been investigating the use of "gold-standard" coupled-cluster methods, which appear very promising for their high accuracy. The principal drawback is their very high computational cost. This cost can be greatly reduced by the use of a "many-body expansion," which adds up the crystal lattice energy as a sum of two-body, three-body, etc., contributions, obtained by computations on individual dimers, trimers, etc., of molecules extracted from the crystal. We have shown that such an approach is viable in systems where the intermolecular interactions are not extremely strong. Coupled-cluster computations are not nearly as time-consuming if they are performed on only two or three molecules at a time, so long as the molecules themselves are not too large. In addition, various approximate methods appear suitable for longer-range interactions for dimers and trimers where the constituent molecules are far apart from each other.

For larger molecules, we have developed our own domain-localized pair natural orbital (DLPNO) coupled-cluster code, which is much more efficient for large systems, and we are examining its accuracy for intermolecular interactions, particularly those in molecular crystals.

Concurrently with this high-accuracy work, we have also performed more qualitative studies on the physics of intermolecular interactions in molecular crystals, using symmetry-adapted perturbation theory (SAPT).

Representative Publications:

  • “Convergence of the Many-body Expansion with Respect to Distance Cutoffs in Crystals of Polar Molecules: Acetic Acid, Formamide, and Imidazole,” P. M. Nelson and C. D. Sherrill, J. Chem. Phys. 161, 214105 (2024) (doi: 10.1063/5.0234883)
  • “Benchmark Coupled-Cluster Lattice Energy of Crystalline Benzene, and Assessment of Multi-Level Approximations in the Many-Body Expansion,” C. H. Borca, Z. L. Glick, D. P. Metcalf, L. A. Burns, and C. D. Sherrill, J. Chem. Phys. 158, 234102 (2023) (doi: 10.1063/5.0159410)
  • “Assessment of Three-Body Dispersion Models against Coupled-Cluster Benchmarks for Crystalline Benzene, Carbon Dioxide, and Triazine,” Y. Xie, Z. L. Glick, and C. D. Sherrill, J. Chem. Phys. 158, 094110 (2023) (doi: 10.1063/5.0143712)
  • “Benchmarking Two-Body Contributions to Crystal Lattice Energies and a Range-Dependent Assessment of Approximate Methods,'' C. T. Sargent, D. P. Metcalf, Z. L. Glick, C. H. Borca, and C. D. Sherrill, J. Chem. Phys. 158, 054112 (2023) (doi: 10.1063/5.0141872)
  • “Range-Dependence of Two-Body Intermolecular Interactions and their Energy Components in Molecular Crystals,” D. P. Metcalf, A. Smith, Z. L. Glick, and C. D. Sherrill, J. Chem. Phys. 157, 084503 (2022) (doi: 10.1063/5.0103644)
  • “CrystaLattE: Automated Computation of Lattice Energies of Organic Crystals Exploiting the Many-Body Expansion to Achieve Dual-Level Parallelism,” C. H. Borca, B. W. Bakr, L. A. Burns, and C. D. Sherrill, J. Chem. Phys. 151, 144103 (2019). (doi: 10.1063/1.5120520)