Potential Energy Surface of the Chromium Dimer Re-re-revisited with Multiconfigurational Perturbation Theory.

The chromium dimer has long been a benchmark molecule to evaluate the performance of different computational methods ranging from density functional theory to wave function methods. Among the latter, multiconfigurational perturbation theory was shown to be able to reproduce the potential energy surface of the chromium dimer accurately. However, for modest active space sizes, it was later shown that different definitions of the zeroth-order Hamiltonian have a large impact on the results. In this work, we revisit the system for the third time with multiconfigurational perturbation theory, now in order to increase the active space of the reference wave function. This reduces the impact of the choice of zeroth-order Hamiltonian and improves the shape of the potential energy surface significantly. We conclude by comparing our results of the dissocation energy and vibrational spectrum to those obtained from several highly accurate multiconfigurational methods and experiment. For a meaningful comparison, we used the extrapolation to the complete basis set for all methods involved.

Molcas 8: New capabilities for multiconfigurational quantum chemical calculations across the periodic table.

In this report, we summarize and describe the recent unique updates and additions to the Molcas quantum chemistry program suite as contained in release version 8. These updates include natural and spin orbitals for studies of magnetic properties, local and linear scaling methods for the Douglas-Kroll-Hess transformation, the generalized active space concept in MCSCF methods, a combination of multiconfigurational wave functions with density functional theory in the MC-PDFT method, additional methods for computation of magnetic properties, methods for diabatization, analytical gradients of state average complete active space SCF in association with density fitting, methods for constrained fragment optimization, large-scale parallel multireference configuration interaction including analytic gradients via the interface to the Columbus package, and approximations of the CASPT2 method to be used for computations of large systems. In addition, the report includes the description of a computational machinery for nonlinear optical spectroscopy through an interface to the QM/MM package Cobramm. Further, a module to run molecular dynamics simulations is added, two surface hopping algorithms are included to enable nonadiabatic calculations, and the DQ method for diabatization is added. Finally, we report on the subject of improvements with respects to alternative file options and parallelization.

Accurate calculations of geometries and singlet-triplet energy differences for active-site models of [NiFe] hydrogenase.

We have studied the geometry and singlet-triplet energy difference of two mono-nuclear Ni(2+) models related to the active site in [NiFe] hydrogenase. Multiconfigurational second-order perturbation theory based on a complete active-space wavefunction with an active space of 12 electrons in 12 orbitals, CASPT2(12,12), reproduces experimental bond lengths to within 1 pm. Calculated singlet-triplet energy differences agree with those obtained from coupled-cluster calculations with single, double and (perturbatively treated) triple excitations (CCSD(T)) to within 12 kJ mol(-1). For a bimetallic model of the active site of [NiFe] hydrogenase, the CASPT2(12,12) results were compared with the results obtained with an extended active space of 22 electrons in 22 orbitals. This is so large that we need to use restricted active-space theory (RASPT2). The calculations predict that the singlet state is 48-57 kJ mol(-1) more stable than the triplet state for this model of the Ni-SIa state. However, in the [NiFe] hydrogenase protein, the structure around the Ni ion is far from the square-planar structure preferred by the singlet state. This destabilises the singlet state so that it is only ∼24 kJ mol(-1) more stable than the triplet state. Finally, we have studied how various density functional theory methods compare to the experimental, CCSD(T), CASPT2, and RASPT2 results. Semi-local functionals predict the best singlet-triplet energy differences, with BP86, TPSS, and PBE giving mean unsigned errors of 12-13 kJ mol(-1) (maximum errors of 25-31 kJ mol(-1)) compared to CCSD(T). For bond lengths, several methods give good results, e.g. TPSS, BP86, and M06, with mean unsigned errors of 2 pm for the bond lengths if relativistic effects are considered.

Theoretical Study of the Dissociation Energy of First-Row Metallocenium Ions.

The bond dissociation energy of a series of metallocenium ions, i.e., the energy difference of the reaction MCp2(+) → MCp(+) + Cp· (with M = Ti, V, Cr, Mn, Fe, Co, and Ni), was studied by means of multiconfigurational perturbation theory (CASPT2, RASPT2, NEVPT2) and restricted coupled cluster theory (CCSD(T)). From a comparison between the results obtained from these different methods, and a detailed analysis of their treatment of electron correlation effects, a set of MCp(+)-Cp binding energies are proposed with an accuracy of 5 kcal/mol. The computed results are in good agreement with the experimental data measured by threshold photoelectron photoion coincidence (TPEPICO) spectroscopy but disagree with the more recent threshold collision-induced dissociation (TCID) experiments.

Theoretical modelling of photoswitching of hyperpolarisabilities in ruthenium complexes.

Static excited-state polarisabilities and hyperpolarisabilities of three Ru(II) ammine complexes are computed at the density functional theory (DFT) and several correlated ab initio levels. Most accurate modelling of the low energy electronic absorption spectrum is obtained with the hybrid functionals B3LYP, B3P86 or M06 for the complex [Ru(II)(NH3)5(MeQ(+))](3+) (MeQ(+)=N-methyl-4,4'-bipyridinium, 3) in acetonitrile. The match with experimental data is less good for [Ru(II)(NH3)5L](3+) (L=N-methylpyrazinium, 2; N-methyl-4-{E,E-4-(4-pyridyl)buta-1,3-dienyl}pyridinium, 4). These calculations confirm that the first dipole- allowed excited state (FDAES) has metal-to-ligand charge-transfer (MLCT) character. Both the solution and gas-phase results obtained for 3 by using B3LYP, B3P86 or M06 are very similar to those from restricted active-space SCF second-order perturbation theory (RASPT2) with a very large basis set and large active space. However, the time-dependent DFT λ(max) predictions from the long-range corrected functionals CAM-B3LYP, LC-ωPBE and wB97XB and also the fully ab initio resolution of identity approximate coupled-cluster method (gas-phase only) are less accurate for all three complexes. The ground state (GS) two-state approximation first hyperpolarisability ß(2SA) for 3 from RASPT2 is very close to that derived experimentally via hyper-Rayleigh scattering, whereas the corresponding DFT-based values are considerably larger. The ß responses calculated by using B3LYP, B3P86 or M06 increase markedly as the π-conjugation extends on moving along the series 2→4, for both the GS and FDAES species. All three functionals predict substantial FDAES ß enhancements for each complex, increasing with the π-conjugation, up to about sevenfold for 4. Also, the computed second hyperpolarisabilities γ generally increase in the FDAES, but the results vary between the different functionals.

Electronic spectra of N-heterocyclic pentacyanoferrate(II) complexes in different solvents, studied by multiconfigurational perturbation theory.

Ligand-field and charge-transfer spectra of N-heterocyclic pentacyanoferrate(II) complexes [Fe(CN)5L]n− were investigated using multiconfigurational perturbation theory. The spectrum of [Fe(CN)5(py)]3­ was studied in detail under vacuum and in the following polarizable continuum model (PCM) simulated solvents: acetone, acetonitrile, dimethylsulfoxide (DMSO), ethanol, methanol, and water. The ligand-field states proved to be rather insensitive to the solvent environment, whereas much stronger solvent effects were observed for the charge-transfer (CT) transitions. The nature of the intense band was confirmed as a metal-to-ligand charge transfer originating from a 3d(xz) → π(b1)* (L)-orbital transition. The difference between the calculated and experimental transition energy of this CT transition is minimal for aprotic solvents, but increases strongly with the solvent proton donor ability in the protic solvents. In an attempt to improve the description of this CT state, up to 14 solvent molecules were explicitly included in the quantum model. In DMSO, the spectra of complexes with ligands L (where L is pyridine, 4-picoline, 4-acetylpyridine, 4-cyanopyridine, pyrazine, and N-methylpyrazinium) correlate very well with the experiment.

Parallelization of a multiconfigurational perturbation theory.

In this work, we present a parallel approach to complete and restricted active space second-order perturbation theory, (CASPT2/RASPT2). We also make an assessment of the performance characteristics of its particular implementation in the Molcas quantum chemistry programming package. Parallel scaling is limited by memory and I/O bandwidth instead of available cores. Significant time savings for calculations on large and complex systems can be achieved by increasing the number of processes on a single machine, as long as memory bandwidth allows, or by using multiple nodes with a fast, low-latency interconnect. We found that parallel efficiency drops below 50% when using 8-16 cores on the shared-memory architecture, or 16-32 nodes on the distributed-memory architecture, depending on the calculation. This limits the scalability of the implementation to a moderate amount of processes. Nonetheless, calculations that took more than 3 days on a serial machine could be performed in less than 5 h on an InfiniBand cluster, where the individual nodes were not even capable of running the calculation because of memory and I/O requirements. This ensures the continuing study of larger molecular systems by means of CASPT2/RASPT2 through the use of the aggregated computational resources offered by distributed computing systems.

Fourteen-electron ring model and the anomalous magnetic circular dichroism of meso-triarylsubporphyrins.

The MCD spectra of meso-triarylsubporphyrins show a sign anomaly which is correlated with the acceptor properties of the aryl substituent. From the spectra, magnetic moments of the excited states are determined. In the context of a simplified orbital model, the sign change is attributed to the quenching of the magnetic moment of the LUMO by acceptor orbitals of the substituent. The actual calculation of this moment presents a major challenge to computational methods. It is shown that wave function techniques based on CASSCF underestimate the covalency effects that are responsible for the quenching. In contrast, a CI method based on DFT orbitals yields excellent results, which fully support the orbital model.

A Multiconfigurational Perturbation Theory and Density Functional Theory Study on the Heterolytic Dissociation Enthalpy of First-Row Metallocenes.

The heterolytic dissociation enthalpy of a series of first-row metallocenes M(C5H5)2, M = V, Mn, Fe, and Ni, was studied by (restricted) multiconfigurational perturbation theory and density functional theory. The results were compared directly to the experimental values, taking into account all necessary contributions to the relative energy. Of the tested functionals, B3LYP performs best in reproducing the binding energy, while the PBE0 functional gives the best structures. High quality multiconfigurational perturbation calculations were also carried out, demonstrating the superior performance of a larger, restricted active space. The spin crossover behavior of manganocene is correctly predicted by multiconfigurational perturbation theory as opposed to the three functionals B3LYP, PBE0, and M06, which (severely) overstabilize the high-spin with respect to the low-spin state.

Multiconfigurational Second-Order Perturbation Theory Restricted Active Space (RASPT2) Studies on Mononuclear First-Row Transition-Metal Systems.

A series of model transition-metal complexes, CrF6, ferrocene, Cr(CO)6, ferrous porphin, cobalt corrole, and FeO/FeO(-), have been studied using second-order perturbation theory based on a restricted active space self-consistent field reference wave function (RASPT2). Several important properties (structures, relative energies of different structural minima, binding energies, spin state energetics, and electronic excitation energies) were investigated. A systematic investigation was performed on the effect of: (a) the size and composition of the global RAS space, (b) different (RAS1/RAS2/RAS3) subpartitions of the global RAS space, and (c) different excitation levels (out of RAS1/into RAS3) within the RAS space. Calculations with active spaces, including up to 35 orbitals, are presented. The results obtained with smaller active spaces (up to 16 orbitals) were compared to previous and current results obtained with a complete active space self-consistent field reference wave function (CASPT2). Higly accurate RASPT2 results were obtained for the heterolytic binding energy of ferrocene and for the electronic spectrum of Cr(CO)6, with errors within chemical accuracy. For ferrous porphyrin the intermediate spin (3)A2g ground state is (for the first time with a wave function-based method) correctly predicted, while its high magnetic moment (4.4 µB) is attributed to spin-orbit coupling with very close-lying (5)A1g and (3)Eg states. The toughest case met in this work is cobalt corrole, for which we studied the relative energy of several low-lying Co(II)-corrole π radical states with respect to the Co(III) ground state. Very large RAS spaces (25-33 orbitals) are required for this system, making compromises on the size of RAS2 and/or the excitation level unavoidable, thus increasing the uncertainty of the RASPT2 results by 0.1-0.2 eV. Still, also for this system, the RASPT2 method is shown to provide distinct improvements over CASPT2, by overcoming the strict limitations in the size of the active space inherent to the latter method.

Multiconfigurational Second-Order Perturbation Theory Restricted Active Space (RASPT2) Method for Electronic Excited States: A Benchmark Study.

The recently developed second-order perturbation theory restricted active space (RASPT2) method has been benchmarked versus the well-established complete active space (CASPT2) approach. Vertical excitation energies for valence and Rydberg excited states of different groups of organic (polyenes, acenes, heterocycles, azabenzenes, nucleobases, and free base porphin) and inorganic (nickel atom and copper tetrachloride dianion) molecules have been computed at the RASPT2 and multistate (MS) RASPT2 levels using different reference spaces and compared with CASPT2, CCSD, and experimental data in order to set the accuracy of the approach, which extends the applicability of multiconfigurational perturbation theory to much larger and complex systems than previously. Relevant aspects in multiconfigurational excited state quantum chemistry such as the valence-Rydberg mixing problem in organic molecules or the double d-shell effect for first-row transition metals have also been addressed.

Copper corroles: the question of noninnocence.

In this paper, the results are presented from a comparative study of the electronic and geometric structure of copper corroles by means of either density functional theory (DFT, using both pure and hybrid functionals) and multiconfigurational ab initio methods, starting from either a complete active space (CASSCF) or restricted active space (RASSCF) reference wave function and including dynamic correlation by means of second-order perturbation theory (CASPT2/RASPT2). DFT geometry optimizations were performed for the lowest singlet and triplet states of copper corrole, both unsubstituted and meso-substituted with three phenyl groups. The effect of saddling on the electronic structure was investigated by comparing the results obtained for planar (C(2v)) and saddled (C(2)) structures. With DFT, the origin of the saddling distortion is found to be dependent on the applied functional: covalent Cu 3d-corrole π interactions with pure functionals (BP86, OLYP), antiferromagnetic exchange coupling between an electron in the corrolate (C(2)) b type π orbital, and an unpaired Cu(II) 3d electron with hybrid functionals (B3LYP, PBE0). The CASPT2 results essentially confirm the suggestion from the hybrid functionals that copper corroles are noninnocent, although the contribution of diradical character to the copper-corrole bond is found to be limited to 50% or less. The lowest triplet state is calculated at 0-10 kcal/mol and conform with the experimental observation (variable temperature NMR) that this state should be thermally accessible.

Multireference ab initio calculations of g tensors for trinuclear copper clusters in multicopper oxidases.

EPR spectroscopy has proven to be an indispensable tool in elucidating the structure of metal sites in proteins. In recent years, experimental EPR data have been complemented by theoretical calculations, which have become a standard tool of many quantum chemical packages. However, there have only been a few attempts to calculate EPR g tensors for exchange-coupled systems with more than two spins. In this work, we present a quantum chemical study of structural, electronic, and magnetic properties of intermediates in the reaction cycle of multicopper oxidases and of their inorganic models. All these systems contain three copper(II) ions bridged by hydroxide or O(2-) anions and their ground states are antiferromagnetically coupled doublets. We demonstrate that only multireference methods, such as CASSCF/CASPT2 or MRCI can yield qualitatively correct results (compared to the experimental values) and consider the accuracy of the calculated EPR g tensors as the current benchmark of quantum chemical methods. By decomposing the calculated g tensors into terms arising from interactions of the ground state with the various excited states, the origin of the zero-field splitting is explained. The results of the study demonstrate that a truly quantitative prediction of the g tensors of exchange-coupled systems is a great challenge to contemporary theory. The predictions strongly depend on small energy differences that are difficult to predict with sufficient accuracy by any quantum chemical method that is applicable to systems of the size of our target systems.

Performance of CASPT2 and DFT for Relative Spin-State Energetics of Heme Models.

The accuracy of the relative spin-state energetics of three small Fe(II) or Fe(III) heme models from multiconfigurational perturbation theory (CASPT2) and density functional theory with selected functionals (including the recently developed M06 and M06-L functionals) was assessed by comparing with recently available coupled cluster results. While the CASPT2 calculations of spin-state energetics were found to be very accurate for the studied Fe(III) complexes (including FeP(SH), a model of the active site of cytochrome P450 in its resting state), there is a strong indication of a systematic error (around 5 kcal/mol) in favor of the high-spin state for the studied Fe(II) complexes (including FeP(Im), a model of the active site of myoglobin). A larger overstabilization of the high-spin states was observed for the M06 and M06-L functionals, up to 22 and 11 kcal/mol, respectively. None of the tested density functionals consistently provides a better accuracy than CASPT2 for all model complexes.

Theoretical description of the structure and magnetic properties of nitroxide-Cu(II)-nitroxide spin triads by means of multiconfigurational ab initio calculations.

The structural, electronic and magnetic properties of two different models of the heterospin polymer chain complexes of Cu2+ hexafluoroacetylacetonate with two pyrazole-substituted nitronyl nitroxides Cu(hfac)2L(R) have been studied by means of multiconfigurational perturbation theory based on a CASSCF (complete active space self-consistent field) wave function, i.e. the CASPT2 method. Our calculations reveal the presence of two minima in the electronic energy curve along the Cu-O(L) bond, separated by only 6 kcal/mol, and corresponding to the X-ray structures of the CuO6 centers in Cu(hfac)2L(Pr) at 115 and 293 K, respectively. The two energetic minima are characterized by a different electronic structure, thus giving rise to a different three-spin exchange coupling and explaining the thermally induced spin transitions in this family of compounds. The concomitant variations in the magnetic properties, i.e. g factors and magnetic moments mu(eff)(T) were calculated and compared with the experimental data of Cu(hfac)2L(Pr). Even if the correspondence is only qualitative, our calculations provide a convincing explanation of the observed magnetic peculiarities. In particular, at low temperatures, the predicted ground-state is 2A(u), well separated from the 2A(g), 4A(u) states and therefore exclusively populated. Its calculated g factors, g(parallel) = 1.848, g(perpendicular) = 1.965, 1.974, qualitatively correspond to the observed g < 2 signals in the low-temperature EPR spectra. The previously assumed formal spin assignment >N-O*-Cu-*O-N < for these linear spin triads is challenged by our calculations, pointing instead to a more important role of the end-standing NO in the exchange interactions with Cu(II).

Multiconfigurational g tensor calculations as a probe for the covalency of the copper-ligand bonds in copper(II) complexes: [CuCl4]2-, [Cu(NH3)4]2+, and plastocyanin.

Calculations of the g tensor of three copper(II) complexes [Cu(NH3)4]2+, [CuCl4]2-, and plastocyanin are presented. Two different sum-over-states-based approaches are considered, making use of the multistate CASPT2 method for excitation energies and PMCAS (perturbation modified CAS) wave functions for the computation of the angular momentum and spin-orbit coupling matrix elements. Test calculations on [Cu(NH3)4]2+ and [CuCl4]2- point to the need of including in the MS-CASPT2 treatment the specific charge-transfer state with an electron excited out of the bonding counterpart of the ground-state SOMO. The computed g shifts for these two molecules present a considerable improvement with respect to the results obtained from our previous g tensor calculations based instead on CASSCF/CASPT2. This is shown to be related to an improved description of the covalency of the Cu-L bonds. For the calculations on plastocyanin, different models are used, taken from a recent (QM/MM) DFT study by Sinnecker and Neese. The effect of the surrounding protein is taken into account by surrounding the central cluster either with a dielectric continuum (epsilon = 4) or with a set of point charges. The second approach is found to be indispensable for an accurate description of environmental effects. With this approach, the calculated g values compare to within 30 ppt with the experimental data of plastocyanin.

Relative energy of the high-(5T2g) and low-(1A1g) spin states of the ferrous complexes [Fe(L)(NHS4)]: CASPT2 versus density functional theory.

High-level ab initio calculations using multiconfigurational perturbation theory [complete active space with second-order perturbation theory (CASPT2)] were performed on the transition energy between the lowest high-spin (corresponding to (5T2g) in Oh) and low-spin (corresponding to 1A1g in Oh) states in the series of six-coordinated Fe(II) molecules [Fe(L)(NHS4)], where NHS4 is 2,2'-bis(2-mercaptophenylthio)diethylamine dianion and L=NH3, N2H4, PMe3, CO, and NO+. The results are compared to (previous and presently obtained) results from density functional theory (DFT) calculations with four functionals, which were already shown previously by Casida and co-workers [Fouqueau et al., J. Chem. Phys. 120, 9473 (2004); Ganzenmuller et al., ibid. 122, 234321 (2005); Fouqueau et al., ibid. 122, 044110 (2005); Lawson Daku et al., ChemPhysChem 6, 1393 (2005)] to perform well for the spin-pairing problem in these and other Fe(II) complexes, i.e., OLYP, PBE0, B3LYP, and B3LYP*. Very extended basis sets were used both for the DFT and CASPT2 calculations and were shown to be necessary to obtain quantitative results with both types of method. This work presents a sequel to a previous DFT/CASPT2 study of the same property in the complexes [Fe(H2O)6]2+, [Fe(NH3)6]2+, and [Fe(bpy)3]2+ [Pierloot et al., J. Chem. Phys. 125, 124303 (2006)]. The latter work was extended with new results obtained with larger basis sets and including the OLYP functional. For all considered complexes, the CASPT2 method predicts the correct ground state spin multiplicity. Since experimental data for the actual quintet-singlet (free) energy differences are not available, the performance of the different DFT functionals was judged based on the comparison between the DFT and CASPT2 results. From this, it was concluded that the generalized gradient OLYP functional performs remarkably well for the present series of ferrous compounds, whereas the success of the three hybrid functionals varies from case to case.

Calculation of EPR g tensors for transition-metal complexes based on multiconfigurational perturbation theory (CASPT2).

The computation of the electronic g tensor by two multireference methods is presented and applied to a selection of molecules including CN, BO, AlO, GaO, InO, ZnH, ZnF, O(2), H(2)O(+), O(3) (-), and H(2)CO(+) (group A) as well as TiF(3), CuCl(4) (2-), Cu(NH(3))(4) (2+), and a series of d(1)-MOX(4) (n-) compounds, with M=V, Cr, Mo, Tc, W, Re and X=F, Cl, Br (group B). Two approaches are considered, namely, one in which spin-orbit coupling and the Zeeman effect are included using second-order perturbation theory and another one in which the Zeeman effect is added through first-order degenerate perturbation theory within the ground-state Kramers doublet. The two methods have been implemented into the MOLCAS quantum chemistry software package. The results obtained for the molecules in group A are in good agreement with experiment and with previously reported calculated g values. The results for the molecules in group B vary. While the g values for the d(1) systems are superior to previous theoretical results, those obtained for the d(9) systems are too large compared to the experimental values.

Relative energy of the high-(5T2g) and low-(1A1g) spin states of [Fe(H2O)6]2+, [Fe(NH3)6]2+, and [Fe(bpy)3]2+: CASPT2 versus density functional theory.

High-level ab initio calculations using the CASPT2 method and extensive basis sets were performed on the energy differences of the high-[(5)T(2g):t(2g) (4)e(g) (2)] and low-[(1)A(1g):t(2g) (6)] spin states of the pseudo-octahedral Fe(II) complexes [Fe(H(2)O)(6)](2+), [Fe(NH(3))(6)](2+), and [Fe(bpy)(3)](2+). The results are compared to the results obtained from density functional theory calculations with the generalized gradient approximation functional BP86 and two hybrid functionals B3LYP and PBE0, and serve as a calibration for the latter methods. We find that large basis set CASPT2 calculations may provide results for the high-spin/low-spin splitting DeltaE(HL) that are accurate to within 1000 cm(-1), provided they are based on an adequately large CAS[10,12] reference wave function. The latter condition was found to be much more stringent for [Fe(bpy)(3)](2+) than for the other two complexes. Our "best" results for DeltaE(HL) (including a zero-point energy correction) are -17 690 cm(-1) for [Fe(H(2)O)(6)](2+), -8389 cm(-1) for [Fe(NH(3))(6)](2+), and 3820 cm(-1) for [Fe(bpy)(3)](2+).