-------------------------------
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.