Failure of Conventional Density Functionals for the Prediction of Molecular Crystal Polymorphism: A Quantum Monte Carlo Study

Hongo, Kenta; Watson, Mark A.; Sánchez‐Carrera, Roel S.; Iitaka, Toshiaki; Aspuru‐Guzik, Alán

doi:10.1021/jz100418p

Cited by 65 publications

(90 citation statements)

References 51 publications

(93 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Our calculations are the largest which are currently feasible using the resources of the K computer. Our results show that our previous predictions in Ref 73,74 using a small…”

Section: Discussionsupporting

confidence: 89%

Diffusion Monte Carlo Study of Para-Diiodobenzene Polymorphism Revisited

Hongo

Watson

Iitaka

et al. 2015

J. Chem. Theory Comput.

Self Cite

View full text Add to dashboard Cite

We revisit our investigation of the diffusion Monte Carlo (DMC) simulation of p-DIB molecular crystal polymorphism. [J. Phys. Chem. Lett. 2010Lett. , 1, 1789Lett. -1794 We perform, for the first time, a rigorous study of finite-size effects and choice of nodal surface on the prediction of polymorph stability in molecular crystals using fixed-node DMC. Our calculations are the largest which are currently feasible using the resources of the K computer and provide insights into the formidable challenge of predicting such properties from first principles. In particular, we show that finite-size effects can influence the trial nodal surface of a small (1×1×1) simulation cell considerably. We therefore repeated our DMC simulations with a 1×3×3 simulation cell, which is the largest such calculation to date. We used a DFT nodal surface generated with the PBE functional and we accumulated statistical samples with ∼ 6.4 × 10 5 core-hours for each polymorph. Our final results predict a polymorph stability consistent with experiment, but indicate that results in our previous paper were somewhat fortuitous.We analyze the finite-size errors using model periodic Coulomb (MPC) interactions and kinetic energy corrections, according to the CCMH scheme of Chiesa, Ceperley, Martin, and Holzmann. We investigate the dependence of the finite-size errors on different aspect ratios of the simulation cell (k-mesh convergence) in order to understand how to choose an appropriate ratio for the DMC calculations. Even in the most expensive simulations currently possible, we show that the finite size errors in the DMC total energies are far larger than the energy difference between the two polymorphs, although error cancellation means that the polymorph prediction is accurate. Finally, we found that the T -move scheme is essential for these massive DMC simulations in order to circumvent population explosions and large time-step biases.

show abstract

“…Our calculations are the largest which are currently feasible using the resources of the K computer. Our results show that our previous predictions in Ref 73,74 using a small…”

Section: Discussionsupporting

confidence: 89%

Diffusion Monte Carlo Study of Para-Diiodobenzene Polymorphism Revisited

Hongo

Watson

Iitaka

et al. 2015

J. Chem. Theory Comput.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Since the QMC is not limited to single-reference and "gasphase" complexes and the reported protocol relies on the fixed-node error cancellation, it may also find application in cases that are intrinsically difficult for mainstream-correlated wave-function approaches. These include estimates of noncovalent interactions between open-shell organometallic and/or metal-organic systems (using simple multi-determinant trial functions instead of a single Slater determinant), studies of periodic models like physisorption on metal surfaces and 2D materials, or prediction of noncovalent crystal stability 31 . In the domain of noncovalent interactions, the quantum Monte Carlo FN-DMC method therefore appears to be very promising for its benchmark accuracy, low-order polynomial scaling with the system complexity, low memory requirements and nearly ideal scaling across thousands of proccessors 50 in parallel supercomputing environments.…”

Section: Discussionmentioning

confidence: 99%

Quantum Monte Carlo for noncovalent interactions: an efficient protocol attaining benchmark accuracy

Dubecký

Derian

Jurečka

et al. 2014

Phys. Chem. Chem. Phys.

View full text Add to dashboard Cite

Reliable theoretical predictions of noncovalent interaction energies, which are important e.g. in drug-design and hydrogenstorage applications, belong to longstanding challenges of contemporary quantum chemistry. In this respect, the fixed-node diffusion Monte Carlo (FN-DMC) is a promising alternative to the commonly used "gold standard" coupled-cluster CCSD(T)/CBS method for its benchmark accuracy and favourable scaling, in contrast to other correlated wave function approaches. This work is focused on the analysis of protocols and possible tradeoffs for FN-DMC estimations of noncovalent interaction energies and proposes an efficient yet accurate computational protocol using simplified explicit correlation terms with a favorable O(N 3 ) scaling. It achieves an excellent agreement (mean unsigned error ∼0.2 kcal/mol) with respect to the CCSD(T)/CBS data on a number of complexes, including benzene/hydrogen, T-shape benzene dimer, stacked adenine-thymine and a set of small noncovalent complexes A24. The high accuracy and reduced computational costs predestinate the reported protocol for practical interaction energy calculations of large noncovalent complexes, where the CCSD(T)/CBS is prohibitively expensive.

show abstract

“…The application of Density Functional Theory a) E-mail: jc.sancho@ua.es b) E-mail: aj.perez@ua.es (DFT) to the X23 dataset, necessarily with some correction for dispersion, was found to provide very accurate values showing mean absolute deviations of 1-2 kcal/mol compared to experimental reference values, 13 and thus very close to the uncertainties of the latter. Such accuracy might pave the way towards optimizing crystal structure parameters 14,15 or disclosing polymorphism issues, 16,17 although with due caution in very challenging cases, such as the quasi-degeneracy in crystalline aspirin, 18,19 oxalyl dihydrazide, 20 or para-diiodobenzene, 21 to mention some recently reported examples.…”

Section: Introductionmentioning

confidence: 99%

Determining the cohesive energy of coronene by dispersion-corrected DFT methods: Periodic boundary conditions vs. molecular pairs

Sancho‐García

Pérez-Jiménez

Olivier

2015

The Journal of Chemical Physics

View full text Add to dashboard Cite

We investigate the cohesive energy of crystalline coronene by the dispersion-corrected methods DFT-D2, DFT-D3, and DFT-NL. For that purpose, we first employ bulk periodic boundary conditions and carefully analyze next all the interacting pairs of molecules within the crystalline structure. Our calculations reveal the nature and importance of the binding forces in every molecular pair tackled and provide revised estimates of the effects of two-and three-body terms, leading to accurate results in close agreement with experimental (sublimation enthalpies) reference values. C 2015 AIP Publishing LLC. [http://dx

show abstract

Failure of Conventional Density Functionals for the Prediction of Molecular Crystal Polymorphism: A Quantum Monte Carlo Study

Cited by 65 publications

References 51 publications

Diffusion Monte Carlo Study of Para-Diiodobenzene Polymorphism Revisited

Diffusion Monte Carlo Study of Para-Diiodobenzene Polymorphism Revisited

Quantum Monte Carlo for noncovalent interactions: an efficient protocol attaining benchmark accuracy

Determining the cohesive energy of coronene by dispersion-corrected DFT methods: Periodic boundary conditions vs. molecular pairs

Contact Info

Product

Resources

About