GronOR: Massively parallel and GPU-accelerated non-orthogonal configuration interaction for large molecular systems.

GronOR is a program package for non-orthogonal configuration interaction calculations for an electronic wave function built in terms of anti-symmetrized products of multi-configuration molecular fragment wave functions. The two-electron integrals that have to be processed may be expressed in terms of atomic orbitals or in terms of an orbital basis determined from the molecular orbitals of the fragments. The code has been specifically designed for execution on distributed memory massively parallel and Graphics Processing Unit (GPU)-accelerated computer architectures, using an MPI+OpenACC/OpenMP programming approach. The task-based execution model used in the implementation allows for linear scaling with the number of nodes on the largest pre-exascale architectures available, provides hardware fault resiliency, and enables effective execution on systems with distinct central processing unit-only and GPU-accelerated partitions. The code interfaces with existing multi-configuration electronic structure codes that provide optimized molecular fragment orbitals, configuration interaction coefficients, and the required integrals. Algorithm and implementation details, parallel and accelerated performance benchmarks, and an analysis of the sensitivity of the accuracy of results and computational performance to thresholds used in the calculations are presented.

NMR shieldings from density functional perturbation theory: GIPAW versus all-electron calculations.

We present a benchmark of the density functional linear response calculation of NMR shieldings within the gauge-including projector-augmented-wave method against all-electron augmented-plane-wave+local-orbital and uncontracted Gaussian basis set results for NMR shieldings in molecular and solid state systems. In general, excellent agreement between the aforementioned methods is obtained. Scalar relativistic effects are shown to be quite large for nuclei in molecules in the deshielded limit. The small component makes up a substantial part of the relativistic corrections.

Capturing the elusive aromaticity of bicalicene.

The ring-current aromaticity of the bicalicene molecule arises, in spite of the 16 π carbon perimeter, from strong local diatropic circulations on the two pentagonal rings, as shown by current-density maps computed at the ipsocentric RHF/6-311G** and DFT/6-311G** levels of theory. Conjugated-circuit models cannot capture this pattern of circulation as it arises from 'ionic' contributions in a valence-bond picture. Canonical molecular-orbital analysis reveals a cancellation of paratropic and diatropic frontier-orbital contributions, which explains the difficulties that Hückel-based models have in producing qualitatively correct current-density maps for this molecule. Other measures of aromaticity reflect, to different extents, the dominance of the 'tetraionic' contribution to the aromaticity of this species.

Current density, chemical shifts and aromaticity.

The intimate link between chemical shifts and magnetic criteria for aromaticity prompts a search for detailed understanding of patterns of current density induced in pi systems by external magnetic fields. Conceptual and practical advantages of calculation of current densities with a specific method of distribution of origin of vector potential, the ipsocentric choice, where the induced current density at each point is calculated with that point as origin, are outlined. This choice leads uniquely to canonical molecular orbital contributions that are free of unphysical occupied-occupied mixing. Characteristic magnetic response of delocalized pi systems is then effectively restricted to the activity of a small number of frontier electrons, governed by simple symmetry and node-counting rules, and readily visualized in current density maps. Localized orbitals for sigma systems can also be used, again eliminating occupied-occupied mixing. For integrated properties (magnetizability and nuclear shieldings), the ipsocentric method gives, in a well-defined sense, the orbital contributions that are best for purposes of interpretation. The general theory is illustrated by maps for a set of annelated pentalenes; the known benzopentalene (1) and 1,2:4,5-dibenzopentalene (2), the still unknown isomer of 2, 1,2:5,6-dibenzopentalene (3), and cyclopent[b,c]acenaphthylene (4), an unknown isomer of pyracylene, all of which consist of fusions of formally aromatic and anti-aromatic pi-conjugated systems.

Survival and extinction of delocalized ring currents in clamped benzenes.

Direct visualisation of induced current density in clamped benzenes 1-4 distinguishes between saturated clamping groups, for which the central benzene ring retains a conventional diamagnetic ring current, and strongly interacting, unsaturated clamps, for which the central ring supports only the localised circulations expected of a 1,3,5-cyclohexatriene with fully fixed double bonds.

Ab initio calculations on iron-porphyrin model systems for intermediates in the oxidative cycle of cytochrome P450s.

Geometry optimizations for several spin states of the iron(III)-S-methyl- porphyrin complex, the iron (III)-oxo-S-methyl-porphyrin complex and the respective anions were performed in order to examine models for intermediates in the oxidative cycle of cytochrome P450. The aim of this study was to obtain insights into the ground states of the intermediates of this catalytic cycle and to use the ab initio calculated geometries and charge distributions to suggest better and more realistic parameters for forcefields which are generally used for modeling P450s. The results indicate that the ground states of both the iron(III)-S-methyl-porphyrin complex and the iron(III)-oxo-S-methyl-porphyrin complex are sextet spin states (high spin). The ground states of the anions of both complexes are probably quintet spin states. The fact that experimentally a shift from low spin to high spin is observed upon binding of the substrate suggests that the ab initio calculations for the iron(III)-S-methyl-porphyrin complex in vacuum give a correct representation of the (hydrophobic) substrate-bound state of the active site of P450. The ab initio geometries of the iron-porphyrin complexes are very similar to the experimentally observed geometries, except for the longer iron-sulfur bond in ab initio calculations, which is probably caused by the omission of polarization functions on the sulfur atom during the geometry optimization. The charge distribution in all ab initio calculated complexes can be described by a series of concentric rings of alternating charge, thus allowing a relatively large positive charge on the iron atom. The commonly used forcefields generally underestimate the charge differences between the iron atom and the different parts of the porphyrin moiety or ignore the charges completely. Although forcefield calculations can reproduce the experimental geometry of iron-porphyrin moieties, extension of the forcefields with charges obtained from ab initio calculations should give a better description of the heme moiety in protein modeling and docking experiments.