Release Log for Q-Chem 6.2
Q-Chem 6.2.2 Release
November 21, 2024
Changes to keywords and default behavior:
- Renamed keyword TdNonEq to Td_CEq_NonEq to clarify that it is specifically for applying the constrained equilibrium principle in nonequilibrium solvation (Haisheng Ren)
- SYM_IGNORE keyword is now correctly set to TRUE if NO_REORIENT=TRUE was requested (John Herbert)
- Increased default printing threshold (CIS_AMPL_PRINT) to 15 for AIMD jobs with state analysis (Goran Giudetti, Anna I. Krylov)
- See a full list of printing and default changes that could potentially affect scripts and parsers here.
General features and improvements:
- ARM64 version available for Macs with Apple silicon chips
- Increased maximum number of atoms and basis functions for NBO to 300 atoms and 3000 basis functions (John Herbert)
- Improved documentation for NBO (John Herbert)
- Re-enabled increasing orbital printing precision via ORBITAL_ENERGY_PREC (John Herbert)
- Resolved issues with:
- Incorrect "current job invalid" failure when JOBTYPE = RAND and JOBTYPE = BH
Density functional theory and self-consistent field:
- New printing for properties in TAO-reference systems (Shaozhi Li, Jeng-Da Chai)
- Resolved issues with:
- Typo in printing of wGDD tuning parameter (John Herbert)
- Incorrect results for M06-2X analytic second derivative calculation for unrestricted systems
- Incorrect results for CIS/TDDFT frequency calculation when NSEG > 1
Correlated methods:
- Resolved issues with:
- Orbital decomposition in MO basis still proceeds for zero Dyson orbital norm in CCMAN2
- Failure of CASSCF geometry optimization with libopt3
- Failure when using a SAD guess for unpaired electron systems in RASCI1
- Incorrect results for EOM-EE-CCSD expectation-value polarizability code with right response vectors
Fragment and energy decomposition analysis improvements:
- SCF_CONVERGENCE for fragments in EDA calculations is now set to be the same as the parent job (Yuezhi Mao)
Q-Chem 6.2.1 Release
September 3, 2024
Changes to default behavior:
- Set correct default values of EDA_COVP_THRESH for COVP and NOCV thresholds, when not set by user (Yuezhi Mao)
- Corrected the list of optical dielectric constants defined for SolventName (John Herbert)
General features and improvements:
- Enable use of equals signs in several input sections (John Herbert)
- Enable higher initial angle to be called first in PES scan (Andrew Gilbert)
- Enable NTO analysis for Fast XAS TDDFT calculations
- Resolved issues with:
- Failure of some large CPSCF and Hessian jobs due to FileMan error
- Geometry optimizer reports failure to converge on the maximum cycle permitted by GEOM_OPT_MAX_CYCLES, even if that cycle has converged
Density functional theory and self-consistent field:
- Enable generation of CAP-HF electron density cube files (Charlotte Titeca, Thomas Jagau)
- TDDFT analytic frequency OpenMP parallel performance improvements
- Resolved issues with:
- Orbital parameterization bug in XES with DFT/CIS (Aniket Mandal, John Herbert)
- Real- and complex-valued molecular orbital cube files in CAP-HF calculations overwrote one another (Charlotte Titeca, Thomas Jagau)
- Bug in diabatic state energies for NEO-MSDFT for range-separated (electronic) functionals (Joseph Dickinson, Eno Paenurk, Sharon Hammes-Schiffer)
- SGM failure for quartet core ion ROHF optimization in 1C-NOCIS (Juanes Arias-Martinez, Martin Head-Gordon)
- Scaling bug in RI-JK nuclear gradient
- Integer overflow bug
- Thread-safety bug in libqints-based multipole and ECP
- Vibrational analysis in libopt3 ignores multipole field
- Vibrational analysis in libopt3 crashes for numerical gradient
- Failure of TDDFT derivative coupling jobs with external charges
- Incorrect D4 dispersion correction for charged species
Correlated methods:
- Enable setting EOM-CCSD(fT) method
- Added error-catching to ensure user-requested states in XOPT/MECP are consistent with requested CCSD/EOM-CCSD states
- Resolved issues with:
- Calling CCSD(dT) (unimplemented) would erroneously perform CCSD(fT)
- Overwriting of Molden files for natural Auger orbitals in resonant Auger decay calculations (Nayanthara Karippara Jayadev, Anna Krylov)
- State following code crash in CIS calculations (Justin Talbot, Juanes Arias-Martinez)
Molecular dynamics, non-adiabatic dynamics, embedding, and solvation improvements:
- Resolved issues with:
- Miscounting of frozen-core orbitals in post-SCF QM/MM calculations using default settings
- Update total energy print in gen_scfman jobs to include all solvent terms (Yuezhi Mao)
Fragment and energy decomposition analysis improvements:
- Enable the closed-shell fragments to be frozen in the polarization calculation for fragment spin alignment in EDA2 (Yuezhi Mao)
- Add print for combined alpha and beta result for unrestricted PolA and CTA (Yuezhi Mao)
- Simplify the print-out section of ALMO-EDA with implicit solvent models (Yuezhi Mao)
- Disable orthogonal frozen decomposition when external charges are present (Yuezhi Mao)
- Resolved issues with:
- Incorrect print about iterative ALMO-TDA memory requirement (Yuezhi Mao)
- Energy component analysis for SRC functionals (Felix Plasser)
Miscellaneous:
- Updated text of Q-Chem license documentation
Q-Chem 6.2.0 Release
May 6, 2024
Changes to default behavior:
- Set POINT_GROUP_SYMMETRY to FALSE for freq jobs reading guess from previous job to avoid reorientation
- Make NO_REORIENT = 1 and POINT_GROUP_SYMMETRY = 0 the defaults for harmonic confiner jobs
- Make improved initial guess Hessian the default for optimizations (INITIAL_HESSIAN = MODEL)
- Enable RI-K virtual orbital correction by default for occ-RI-K frequency jobs
General features and improvements:
- NEO Multistate DFT (NEO-MSDFT) (Joseph Dickinson, Qi Yu, Sharon Hammes-Schiffer)
- Added simulation of nuclear-electronic quantum dynamics based on real-time nuclear-electronic orbital (RT-NEO) (Tao E. Li, Mathew Chow, Sharon Hammes-Schiffer)
- Added support for D3 empirical dispersion correction (energies and gradients) and Effective Core Potential (ECP) (energies and gradients) for NEO approach and RT-NEO methods (Mathew Chow, Sharon Hammes-Schiffer)
- NEO Multistate DFT (NEO-MSDFT) (Joseph Dickinson, Qi Yu, Sharon Hammes-Schiffer)
- Constrained NEO (CNEO) for a single NEO center (Eno Paenurk, Sharon Hammes-Schiffer)
- Addition of SCS-RIMP2 and SOS-OOMP2 to NEO methods (Jonathan Fetherolf, Sharon Hammes-Schiffer)
- Faster NEO-SCF via simultaneous DIIS optimization (Mathew Chow, Sharon Hammes-Schiffer)
- Add atomic multipole moment calculation using IAOs (Alexandra McIsaac, Abdulrahman Aldossary, Martin Head-Gordon)
- Update the list of occasions not allowed for the AUTOSAD -> CORE fallback when atomic jobs fail (Yuezhi Mao)
- New syntax allowing external electric field to follow a bond (Tarek Scheele, Tim Neudecker)
- Real-time extension of TAO-DFT (RT-TAO) (Hung-Yi Tsai, Jeng-Da Chai)
- SCF energy prints now break one-electron terms into kinetic, nuclear attraction and remainder when SCF_FINAL_PRINT is set to TRUE
- Fixed issues with:
- Incorrect hydrogen isotope mass when using EPC functionals with NEO (Mathew Chow, Zhen (Coraline) Tao, Sharon Hammes-Schiffer)
- NEO Hessian implementation for range-separated functionals (Eno Paenurk and Sharon Hammes-Schiffer)
- Failure of the READ_REAL guess of complex RHF (Yuezhi Mao)
- Libopt3 misclassification of structure type for near-zero eigenvalues
- “SVD Failed" invalid step error in rare geometry optimization cases
Density functional theory and self-consistent field:
- Implemented DFT/CIS, including a new parameterization for X-ray spectroscopy (Aniket Mandal, John Herbert)
- Addition of new TZVP basis set for DFT/CIS to QCAUX (Aniket Mandal, John Herbert)
- Implemented suppression of open-shell mixing for ROKS calculations (Juanes Arias-Martinez, Martin Head-Gordon)
- Enable negative s8 values for Grimme’s D4 dispersion correction
- Dipole filtering for restricted and unrestricted TDKS (John Herbert, Avik Kumar Ojha)
- Print spatial overlaps of orbitals (intermediate overlaps for occ/virt pairs) (John Herbert)
- Make virtual orbital correction after SCF convergence optional for occ-RI-K
- Enable MakeMeSAD for GEN_SCFMAN
- Add SAD guesses for cc-pV6Z and aug-cc-pV6Z
- Disable AUTOSAD guess for CDFT-CI (John Herbert)
- Disable NTOs for RPA (John Herbert)
- Fixed issues with:
- Job control through FDIFF_DER for anharmonic frequency calculations (Yuezhi Mao)
- Anharmonic frequency calculation failure with DFT (Yuezhi Mao)
- Restart initialization for unrestricted TDKS (Hung-Yi Tsai, Jeng-Da Chai)
- Bug in RI related to basis sets that use multishells (note that this does not affect most RI calculations, unless an auxiliary multishell basis is used)
Correlated methods:
- Added hyperfine coupling code to calculate couplings between singlet/triplet excited CIS states (Samuel May, Joseph Subotnik)
- Generalization of 1C-NOCIS to two-electron open-shell singlets (Juanes Arias-Martinez, Hamlin Wu, Martin Head-Gordon)
- Implemented the ACP-EOMIP-CCSD method to compute partial Auger decay widths (Florian Matz, Thomas Jagau)
- Built-in implementation of calculations of x-secs and pads using Dyson orbitals (Madhubani Mukherjee, Sarai Folkestad, Anna Krylov)
- Natural Auger Orbitals for Auger decay, ICD, and related processes for CVS-EOM methods (Nayanthara K. Jayadev, Anna I. Krylov)
- Added BW-s2 method (Kevin Carter-Fenk, Martin Head-Gordon)
- Stochastic resolution of identity to CC2 (sRI-CC2) for excited state energy calculations (Chongxiao Zhao, Wenjie Dou, Joonho Lee)
- EOM-CCSDT for EE and SF states (Manisha, Prashant Uday Manohar)
- Fixed issues with:
- Legacy RI-MP2 code crashes when MEM_STATIC value is too large (Yuezhi Mao)
- Error in calculation of the separable part of the one-particle density matrix with CCSD (Yongbin Kim, Madhubani Mukherjee, Anna Krylov)
- Overwriting of Molden files for natural Auger orbitals in resonant Auger decay calculations (Nayanthara Karippara Jayadev, Anna Krylov)
Molecular dynamics, non-adiabatic dynamics, embedding, and solvation:
- External/embedded potential read (Huseyin Aksu, Barry Dunietz)
- Turn on PCM at the first SCF iteration when SCF_GUESS = READ, allowing faster convergence (Yuezhi Mao)
- Turn on PCM at the first SCF iteration in the 2nd geometry optimization cycle for FRAGMO and AUTOSAD (Yuezhi Mao)
- Clarified ambiguous output for COSMO (John Herbert)
- Added analytic X-HCFF Hessian (Rahel Weiss, Felix Zeller, Tim Neudecker)
- Fixed issues with:
- Out-of-bound access error for SMx with ghost atoms
- Undesirable behavior of SOLVENT_METHOD = TRUE invoking Kirkwood model (John Herbert)
- Add missing qarchive files for QM/MM calculations (Xiaoliang Pan, Yihan Shao)
- Incorrectly increasing van der Waals radii for I and At atoms in jobs with multiple SMD calculations (Zheng Pei)
Fragment and energy decomposition analysis:
- Add additional printout for libwfa regarding ionic/covalent states (Felix Plasser)
- Fixed issues with occasional crash in non-perturbative CTA (Yuezhi Mao)
Post-processing features:
- Calculations of g-tensors using ezMagnet (Sven Kähler, Antonio Cebreiro-Gallardo, Pavel Pokhilko, David Casanova, and Anna I. Krylov)
Previous Release Logs
- Release Log for Q-Chem 6.1
- Release Log for Q-Chem 6.0
- Release Log for Q-Chem 5.4
- Release Log for Q-Chem 5.3
- Release Log for Q-Chem 5.2
- Release Log for Q-Chem 5.1
- Release Log for Q-Chem 5.0