The interplay of intra- and intermolecular errors in modeling conformational polymorphs

Abstract

Conformational polymorphs of organic molecular crystals represent a challenging test for quantum chemistry because they require careful balancing of the intra- and intermolecular interactions. This study examines 54 molecular conformations from twenty sets of conformational polymorphs, along with the relative lattice energies and 173 dimer interactions taken from six of the polymorph sets. These systems are studied with a variety of van der Waals-inclusive density functionals theory (DFT) models, dispersion-corrected spin-component-scaled second-order Møller-Plesset perturbation theory (SCS-MP2D), and domain local pair natural orbital coupled cluster singles, doubles, and perturbative triples (DLPNO-CCSD(T)). We investigate how delocalization error in conventional density functionals impacts monomer conformational energies, systematic errors in the intermolecular interactions, and the nature of error cancellation that occurs in the overall crystal. Density functionals such as B86bPBE-XDM, PBE-D4, PBE-MBD, PBE0-D4, and PBE0-MBD are found to exhibit sizable 1-body and 2-body errors versus DLPNO-CCSD(T) benchmarks, and the level of success in predicting the relative polymorph energies relies heavily on error cancellation between different types of intermolecular interactions or between intra- and intermolecular interactions. Models like SCS-MP2D, and, to a lesser extent, ωB97M-V exhibit smaller errors and rely less on error cancellation. Implications for crystal structure prediction of flexible compounds are discussed. Finally, the 1-body and 2-body DLPNO-CCSD(T) energies taken from these conformational polymorphs establish the CP1b and CP2b benchmark data sets which could be useful for testing quantum chemistry models in challenging real-world systems with complex interplay between intra- and intermolecular interactions, a number of which are significantly impacted by delocalization error.

Coupled Cluster Benchmarking of Large Noncovalent Complexes in L7 and S12L as Well as the C60 Dimer, DNA–Ellipticine, and HIV–Indinavir

“…The ΔCCSD­(T) correlation correction term is where the TightPNO threshold (TCutPNO = 1 × 10 –7 , TCutDO = 5 × 10 –3 , TCutPairs = 1 × 10 –5 , and TCutMKN = 1 × 10 –3 ) has been used in DLPNO-CCSD­(T) with noniterative semicanonical triples [sometimes referred to as DLPNO-CCSD­(T 0 )] and CBS′ means using smaller basis sets to do the basis-set extrapolation. An even tighter TCutMKN parameter (1 × 10 –4 ), which determines the size of the initial domain in which the PNOs are expanded, on top of TightPNO has been employed in enthalpies of formation, reaction energies and barriers, and conformational and dimer interaction energies. , The typical coronene dimer in L7, which is a stepping stone between the benzene dimer and graphene bilayer but shows a binding discrepancy over the so-called “chemical accuracy” of 1 kcal/mol between local CCSD­(T) and FN-DMC, ,, is used here to test the influence of TCutMKN on bindings in large noncovalent complexes. Further tightening TCutMKN to 1 × 10 –4 changes the C2C2PD binding at the level of DLPNO-CCSD­(T 0 )/CBS from −20.93 to −20.91 kcal/mol, which is well within its estimated 0.44 kcal/mol binding uncertainty.…”

Effect of Fluorination on the Polymorphism and Photomechanical Properties of Cinnamalmalononitrile Crystals

“…Because the CM and 4FCM molecules are not strictly planar and generalized gradient approximation density functionals like B86bPBE-XDM are known to exhibit biases toward certain conformations in systems with extended π-conjugation, final single point energies of the crystal structures were computed by employing an intramolecular energy correction which has previously been demonstrated to be important in a number of other polymorphic crystals. − This correction adjusts the B86bPBE-XDM lattice energy based on the intramolecular energy difference computed with SCS-MP2D and B86bPBE-XDM (computed in the gas-phase using the molecular geometry directly extracted from each crystal): E crystal / Z = E crystal ( DFT ) / Z − E molec ( DFT ) + E molec ( SCS−MP2D ) See ref for details. This correction shifted the relative lattice energies in CM and 4FCM by 1–2 kJ/mol, on average.…”

XDM-Corrected Hybrid DFT with Numerical Atomic Orbitals Predicts Molecular Crystal Lattice Energies with Unprecedented Accuracy

“…43,[65][66][67] Lastly, GGA functionals give a poor description of conformational energies, which are important when comparing molecular crystals composed of flexible molecules. 44,[68][69][70] In this work, we address these shortcomings by combining XDM functionals with the numerical atomic orbital (NAO) basis sets in the Fritz Haber Institut ab initio materials simulation (FHI-aims) package. [71][72][73][74] Being finite-support functions, NAOs have the double advantage of making the calculation scale linearly with size as well as allowing the relatively inexpensive use of hybrid functionals compared to plane-wave approaches.…”

“…This is important because hybrid functionals can be used to mitigate delocalization error, 64,66,[75][76][77][78][79] and, generally, they are also more accurate than GGAs for conformational energies. 70 One drawback of NAOs is the possible appearance of basis-set incompleteness error (BSIE), which is known to have a deleterious effect on the description of non-covalent interactions. [80][81][82] Dispersion-corrected DFT methods with NAOs have been used before for molecular crystals, particularly in combination with the Tkachenko-Scheffler (TS) 83 and many-body dispersion (MBD) 84,85 family of corrections.…”