Skip to content

Release 1.4.0

Choose a tag to compare

@nabbelbabbel nabbelbabbel released this 21 Oct 12:43
· 17 commits to master since this release

Release 1.4.0 (21.10.2021)

Functionalities

General/Other Features

  • SCF convergence thresholds were changed! The new defaults are
    • energy convergence threshold: 5e-8 (old: 1e-8)
    • density convergence threshold: 1e-8 (old: 1e-8)
    • max(FP-PF) threshold: 5e-7 (old: 1e-7)
  • Add Broken-Symmetry calculations via KS-DFT and sDFT (Anja Massolle).
  • Add a task that orthogonalizes orbitals between subsystems (Anja Massolle).
  • The EnergyTask can now evaluate the non-additive kinetic energy contribution
    from orthogonalized subsystem orbitals (Anja Massolle).
  • Add ECP gradients (Jan Unsleber).
  • Add multi-state FDE Electron Transfer (FDE-ET) and FDE-diab (Patrick Eschenbach).
  • Add a task that allows reading of orbitals from other programs.
    Currently, only the ASCII format from turbomole and Serenity's own format are
    supported (Moritz Bensberg).
  • Add calculation of quasi-restricted orbitals (Moritz Bensberg).
  • Makes Serenity compatible with the MoViPac program (Moritz Bensberg).
Local Correlation
  • Add occupied orbital partitioning into an arbitrary number of subsystems
    by the generalized direct orbital selection procedure (Moritz Bensberg).
  • Add input simplification tasks for local correlation calculations
    (LocalCorrelationTask) and DFT-embedded local correlation calculations
    (DFTEmbeddedLocalCorrelationTask) (Moritz Bensberg).
  • Add a task for coupled-cluster-in-coupled-cluster embedding by adjusting
    the DLPNO-thresholds for each region [see JCTC 13, 3198-3207 (2017)]
    (Moritz Bensberg).
  • Added a task that allows the fully automatized calculations of relative energies
    form multi-level DLPNO-CC (DOSCCTask) (Moritz Bensberg).
  • Core orbitals may be specified in the orbital localization task either by an
    energy cut-off, by tabulated, element-specific numbers, or by explicitly
    giving a number of core orbitals (Moritz Bensberg).

Polarizable Continuum Model

  • Add a task to calculate the PCM energy contributions for a given
    subsystem density (Jan Unsleber, Moritz Bensberg).
  • Add CPCM gradients (Moritz Bensberg).
  • Add cavity creation energy calculation from scaled particle
    theory (Moritz Bensberg).
  • Changed the default for "minDistance" in the PCM-input block from 0.1 to 0.2.

Response Calculations

  • Restricted/unrestricted CC2/CIS(Dinf)/ADC(2) excitation energies
    and transition moments from the ground state (Niklas Niemeyer).
  • Spin-component and spin-opposite scaled CC2/CIS(Dinf)/ADC(2) (Niklas Niemeyer).
  • Quasi-linear and DIIS nonlinear eigenvalue solver (Niklas Niemeyer).
  • Natural auxiliary functions (NAFs) for GW/BSE/CC2/CIS(Dinf)/ADC(2) (Niklas Niemeyer).
  • Non-orthonormal eigenvalue subspace solver (Niklas Niemeyer).
  • Restart system of non-converged eigenpairs in the iterative eigenvalue solvers (Niklas Niemeyer).
  • Gauge-origin invariant optical rotation in the length gauge (Niklas Niemeyer).
  • Virtual orbital space selection [tested for GW/BSE/TDDFT/TDA/CIS/TDHF/CC2/CIS(Dinf)/ADC(2)/MP2] (Johannes Tölle).
  • Diabitazation procedures (multistate FXD, FED, FCD) (Johannes Tölle).
  • GW and BSE (with and without environmental screening) (Johannes Tölle).
  • Partial response-matrix construction (TDA, TDDFT) (Johannes Tölle, Niklas Niemeyer).
  • LibXC support for TDDFT/TDA-Kernel evaluation (Johannes Tölle).
  • Mixed exact-approximate embedding schemes for ground and excited states (Johannes Tölle).
  • Reimplementation of natural transition orbitals and support for coupled TDDFT (Johannes Tölle).
  • Grimme's simplified TDA and TDDFT (Niklas Niemeyer).
  • Sigmavector for Exchange contribution using RI, support for long-range exchange and coupled sTDDFT support (Niklas Niemeyer, Johannes Tölle).
  • Löwdin transition, hole, and particle charges for response calculations (Anton Rikus, Niklas Niemeyer).
  • Transition densities, hole densities, and particle densities can be plotted with the PlotTask (Anton Rikus).
  • Natural Response Orbitals can now be plotted (Anton Rikus).

Cholesky Decomposition Techniques

  • Added Cholesky decomposition techniques (full Cholesky decomposition,
    atomic Cholesky decomposition, atomic-compact Cholesky decomposition) for the evaluation
    of Coulomb and exchange contributions (Lars Hellmann).
  • Added atomic and atomic-compact Cholesky basis sets to be used in place of the auxiliary
    basis sets used in the RI formalism (Lars Hellmann).
  • Added atomic and atomic-compact Cholesky basis sets to fit integrals in the range-separation
    approach (Lars Hellmann).

Electric Fields

  • Numerical external electric fields can now be included through point charges arranged in circular
    capacitor plates around a molecule (Niklas Niemeyer, Patrick Eschenbach).
  • Analytical external electric fields and corresponding geometry gradients can now be included through dipole integrals
    and their derivatives. (Niklas Niemeyer, Patrick Eschenbach).
  • Finite-Field Task for (FDE-embedded) numerical and semi-numerical
    calculation of (hyper) polarizabilities (Niklas Niemeyer, Patrick Eschenbach).

Technical Features

  • Update Libecpint to v1.0.4.
  • Rework of Libint precision handling.
  • Output modifications for simplified handling with MoViPac.
  • The MultipoleMomentTask now accepts multiple systems and is able to print
    their total multipole moments.
  • The GradientTask may now print the gradient for all atoms in all systems
    in one table.
  • Removed outdated keyword "dispersion" from GradientTask, GeometryOptimizationTask
    and HessianTask.
  • All basis-set files have been updated to the latest version available
    on www.basissetexchange.org.
  • Errors in the def2-series RI MP2 basis sets have been fixed. The old versions were
    actually the MP2 fitting-basis sets of the def-series.
  • Rework of DLPNO-MP2/CCSD/CCSD(T).
    Now significantly faster, linear scaling, and caches integrals on disk.
  • Fixed an error where the tabulated probe radii for the PCM cavity construction
    where given in Bohr instead of angstrom.
  • The Schwarz-prescreening threshold is now by default tied to the basis set size.
    It is calculated as 1e-8/(3M), where M is the number of Cartesian basis
    functions.
  • The settings of other tasks may now be forwarded with the block-input system.