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 and two surface hopping algorithms are included to enable nonadiabatic calculations. Finally, we report on the subject of improvements with respects to alternative file options and parallelization.

Density functional theory (DFT) calculations of molecular hyperfine tensors were implemented as a second derivative property within the two-component relativistic zeroth-order regular approximation (ZORA). Hyperfine coupling constants were computed for systems ranging from light atomic radicals to molecules with heavy d and f block elements. For comparison, computations were also performed with a ZORA first-order derivative approach. In each set of computations, Slatertype basis sets have been used. The implementation allows for nonhybrid and hybrid DFT calculations and incorporates a Gaussian finite nucleus model. A comparison of results calculated with the PBE nonhybrid and the PBE0 hybrid functional is provided. Comparisons with differing basis sets and incorporation of finite-nucleus corrections are discussed. The second derivative method is applied to calculations of paramagnetic NMR ligand chemical shifts of three Ru(III) complexes. The results are consistent with those calculated using a first-order derivative method, and the results are consistent for different functionals used. A comparison of two different methods of calculating pseudo-contact shifts, one using the full hyperfine tensor and one assuming a point-charge paramagnetic center, is made for the Ru(III) complexes.

The electronic structure and magnetic properties of neptunyl(VI), NpO2(2+), and two neptunyl complexes, [NpO2(NO3)3](-) and [NpO2Cl4](2-), were studied with a combination of theoretical methods: ab initio relativistic wavefunction methods and density functional theory (DFT), as well as crystal-field (CF) models with parameters extracted from the ab initio calculations. Natural orbitals for electron density and spin magnetization from wavefunctions including spin-orbit coupling were employed to analyze the connection between the electronic structure and magnetic properties, and to link the results from CF models to the ab initio data. Free complex ions and systems embedded in a crystal environment were studied. Of prime interest were the electron paramagnetic resonance g-factors and their relation to the complex geometry, ligand coordination, and nature of the nonbonding 5f orbitals. The g-factors were calculated for the ground and excited states. For [NpO2Cl4](2-), a strong influence of the environment of the complex on its magnetic behavior was demonstrated. Kohn-Sham DFT with standard functionals can produce reasonable g-factors as long as the calculation converges to a solution resembling the electronic state of interest. However, this is not always straightforward.

A method is reported by which calculated hyperfine coupling constants (HFCCs) and paramagnetic NMR (pNMR) chemical shifts can be analyzed in a chemically intuitive way by decomposition into contributions from localized molecular orbitals (LMOs). A new module for density functional calculations with nonhybrid functionals, global hybrids, and range-separated hybrids, utilizing the two-component relativistic zeroth-order regular approximation (ZORA), has been implemented in the parallel open-source NWChem quantum chemistry package. Benchmark results are reported for a test set of few-atom molecules with light and heavy elements. Finite nucleus effects on (199)Hg HFCCs are shown to be on the order of -11 to -15%. A proof of concept for the LMO analysis is provided for the metal and fluorine HFCCs of TiF3 and NpF6. Calculated pNMR chemical shifts are reported for the 2-methylphenyl-t-butylnitroxide radical and for five cyclopentadienyl (Cp) sandwich complexes with 3d metals. Nickelocene and vanadocene carbon pNMR shifts are analyzed in detail, demonstrating that the large carbon pNMR shifts calculated as +1540 for Ni (exptl.: +1514) and -443 for V (exptl.: -510) are caused by different spin-polarization mechanisms. For Ni, Cp to Ni π back-donation dominates the result, whereas for vanadocene, V to Cp σ donation with relaxation of the carbon 1s shells can be identified as the dominant mechanism.

Electronic structures and magnetic properties of actinyl ions AnO2(n+) (An = U, Np, and Pu) and the equatorially coordinated carbonate complexes [UO2(CO3)3](5–), [NpO2(CO3)3](4–), and [PuO2(CO3)3](4–) are investigated by ab initio quantum chemical calculations. The complex [PuO2(NO3)3](−) is also included because of experimentally available g-factors and for comparison with a previous study of [NpO2(NO3)3](−) (Chem.—Eur. J. 2014, 20, 7994-8011). The results are rationalized with the help of crystal-field (CF)-type models with parameters extracted from the ab initio calculations, and with the help of natural orbitals and natural spin orbitals contributing to the magnetic properties and the unpaired spin distribution, generated from the spin–orbit wave functions. These orbitals resemble textbooklike representations of the actinide 5f orbitals. Calculated paramagnetic susceptibilities are used to estimate dipolar 13C chemical shifts for the carbonate ligands. Their signs and order of magnitude are compared to paramagnetic effects observed experimentally in NMR spectra. The results indicate that the experimental spectra are also influenced by contact shifts.