Release Log for Q-Chem 5.4

image11

Q-Chem 5.4.2 Release

December 20, 2021

Changes to default behavior

  • Made default SCF convergence criterion for supersystem and fragment jobs in EDA and BSSE calculations consistent (Yuezhi Mao)

General features and improvements

  • Enabled mixed basis for AUTOSAD guess (Kevin Carter-Fenk, Yuezhi Mao, John Herbert)
  • Enabled compatibility with the NBO7 program (John Herbert)
  • Implementation of intrinsic bond orbital (IBO) analysis (Alexander Zech, Christopher Stein, Abdulrahman Aldossary, Martin Head-Gordon)
  • Resolved issues with:
    • frequency job failure when number of threads is thrice larger than number of atoms
    • frequency job failure when CPSCF segments are equal to number of atoms
    • incorrect alpha density generated when using new plots section format

Density functional theory and self-consistent field

  • Enabled analytic Hessian for TPPS/TM/SCAN TDDFT calculations
  • Added printing of information about memory requirements for TDDFT (John Herbert)
  • Added an experimental implementation of the X2C method for relativistic quantum chemistry (Diptarka Hait, Leonardo Cunha, Richard Kang, Martin Head-Gordon)
  • Improved CIS/TDA/RPA guess to avoid missing roots
  • Implementation of projection-based embedding with complex basis functions (Valentina Parravicini, Thomas Jagau)
  • Improved performance of the GOSTSHYP method through integral screening (Felix Zeller, Tim Neudecker, Eric Berquist)
  • Resolved issues with:
    • AIFDEM crash when a larger fragment is listed first
    • NAN in SCF energies using VV10 functionals
    • unrestricted RPA TDDFT analytic Hessian for singlet excited state
    • failure to compute non-adiabatic couplings (NACs) using pure TDDFT
    • incorrect TDDFT energies with FAST_XAS using multiple threads
    • incorrect results from projection-based embedding using LRC-DFT as the low-level theory (Yuezhi Mao)
    • incompletely converged energies in RPA calculations
    • failure to evaluate spin-orbit integrals in TDDFT SOC calculations
    • GPU acceleration of unrestricted pure DFT gradient when using BrianQC
    • incorrect ROHF gradient when using BrianQC

Correlated methods

  • Implementation of CCSD damped polarizability and first hyperpolarizability (Kaushik Nanda)
  • Resolved issues with:
    • wrongfully activated ECD properties with EOM-IP-CCSD (Josefine Andersen, Sonia Coriani)
    • convergence issues in EOM-DIP and EOM-DEA methods
    • symmetry check for v2RDM (Rain Li, Eugene DePrince)
    • failure to write ASCI energy to checkpoint files

Molecular dynamics

  • Enabled the use of new SCF drivers (GEN_SCFMAN=TRUE) in path integral MD
  • Resolved issues with:
    • missing energy-component file for AIMD when GEN_SCFMAN=TRUE

Fragment and energy decomposition analysis

  • When EDA2_MOM is used with EDA_BSSE, apply IMOM to BSSE calculations with ghost atoms to prevent collapsing to the lower-energy states (Yuezhi Mao)
  • Allowed SCFMI_MOM and EDA2_MOM to preserve the electronic configuration of the frozen state (Yuezhi Mao)
  • Multiple stability improvements in ALMO-EDA (Yuezhi Mao)
  • Implemented non-perturbative CT analysis for ALMO-EDA (Hengyuan Shen, Srimukh Prasad, Martin Head-Gordon)
  • Resolved issues with:
    • final print of the one-side CT energy in VFB CT analysis incorrectly contained the contribution from SMD's CDS (non-electrostatic) term, when using the SMD solvent model (Yuezhi Mao)
    • double-counting of environment frozen core orbitals with default N_frozen_core setting for projection-based embedding (Yuezhi Mao)
    • display of preparation energy for ALMO-EDA (Yuezhi Mao)
    • many-body expansion (MBE) geometry optimization (John Herbert)
    • convergence of linear solvers for orthogonal frozen decomposition (Yuezhi Mao)
    • the dispersion term in classic frozen decomposition in non-aufbau ALMO-EDA (Yuezhi Mao)

Miscellaneous

  • Disabled analytic force calculation with projection-based embedding (Yuezhi Mao)
  • Disabled complex SCF for fragment jobs (Yuezhi Mao)
  • Resolved issue with NAN printing efield file for in QM/MM calculations when external charges are set to zero
  • Added warning that CDFT does not support algorithms other than DIIS and RCA (Yuezhi Mao)
  • Added warning when 3c methods are used without recommended basis sets (John Herbert)
  • Added NBO version number in output (John Herbert)
  • Fixed minor spelling errors in the printing of TDDFT (Bushra Alam, John Herbert)

Q-Chem 5.4.1 Release

August 16, 2021

Changes to defaults and keywords

  • Renamed Onsanger SOLVENT_METHOD to Kirkwood (John Herbert)
  • Updated the SM8 solvation model to use Cartesian Gaussians (PURECART 2222) (John Herbert)
  • Renamed spin-specific keywords to EA_ALPHA, EA_BETA, IP_ALPHA and IP_BETA (Wojtek Skomorowski)

New features and improvements

  • Added Intrinsic Atomic Orbitals (IAO) and Intrinsic Bond Orbitals (IBO) (Abdulrahman Aldossary, Alexander Zech, Christopher Stein)
  • Added a new localization method, Oxidation State Localized Orbitals (OSLO) (Abdulrahman Aldossary, Alexander Zech, Christopher Stein)
  • Included installation of Romberg utilities

Density functional theory and self-consistent field

  • Implemented hybrid functionals for TAO-DFT (Shaozhi Li, Jeng-Da Chai)
  • Improved efficiency of range-separated DFT frequency calculations when run in parallel with shared memory and multithreading
  • Added option to turn off ground-state PCM calculations for TDDFT (John Herbert)
  • Added option to enforce level shifting in every SCF cycle for state-targeted energy projection (STEP) (Kevin Carter-Fenk)
  • Implemented projection-based embedding for unrestricted calculations (Yuezhi Mao)
  • Added printing of more digits for the TDDFT transition strength
  • Improved SCF guess for optimization jobs using BASIS2
  • Resolved issues with:
    • segmentation fault in CIS frequency calculations when using libqints
    • index out of bounds error with TDKS sample in manual (Hung-Yi Tsai, Jeng-Da Chai)
    • errors in unrestricted TDDFT Hessian calculations
    • small error in RPA excitation energies
    • incorrect SCF energy with libqints-based SRC-DFT
    • crash when computing numerical derivatives with BASIS=GEN
    • crash while running large frequency jobs due to insufficient memory in CPSCF
    • incorrect evaluation of iterative Hirshfeld charges (Abdulrahman Aldossary)

Correlated methods

  • Added options for custom scaling in complex basis function calculations (Florian Matz, Thomas Jagau)
  • Improved projected CAP-EOM-CC (James Gayvert)
  • Implemented EOM-DEA-CCSD two-photon absorption (Kaushik Nanda, Sahil Gulania, Anna Krylov)
  • Implemented complex-valued CC2 and RI-CCSD (Cansu Utku, Garrette Pauley Paran, Thomas Jagau)
  • Implemented effective nuclear charge approximation for SOCs using EOM (Saikiran Kotaru, Anna Krylov)
  • Resolved issues with:
    • freezing string method (FSM) reading SCF energy instead of correlated energy value
    • missing triples corrections for EOM calculations in ccman2 (Pavel Pokhilko)
    • using frozen core and virtual orbitals in projector-based embedding calculations (Yuezhi Mao)

Large systems, QM/MM, and solvation

  • Implemented user-defined permittivity grid for Poisson equation solver (PEqS) (Suranjan Kumar Paul)
  • Enabled SCRF for GEN_SCFMAN-based ROHF/ROKS calculations (Yuezhi Mao)
  • Implemented state-specific PCM/TDDFT (SS-PCM/TDDFT) method based on the constrained equilibrium theory (Haisheng Ren, Fan Wang, Xiangyuan Li, Yingli Su)
  • Improved GROMACS QM/MM interface (Vale Cofer-Shabica)
  • Improved gradient performance of the SM8 solvation model (John Herbert)
  • Improved memory usage of the SM8 solvation model (John Herbert)
  • Added TDDFT_PCM to control nth-order solvent correction (John Herbert)

Fragment and energy decomposition analysis

  • Enabled the linearized approximation in projection-based embedding (Yuezhi Mao)
  • Implemented POD2L and POD2GS for projection operator diabatization (POD) (Yuezhi Mao)
  • Enabled calculation of couplings between multiple pairs of diabatic orbitals for POD (Yuezhi Mao)
  • Added printing of separate energy components in the SAPT output (John Herbert)

Q-Chem 5.4 Release

May 17, 2021

Starting from this release, Q-Chem drops support for OpenMPI 1.10 on Linux and adds support for BrianQC on Windows.

Changes to default behavior

  • Use of automatically generated superposition of atomic densities SCF guess for custom basis sets (Yuezhi Mao, Kevin Carter-Fenk)
  • Use atomic size-corrected Becke weights for CDFT (Kevin Carter-Fenk)

New features and improvements

  • New methods to distort molecules using force and pressure: HCFF, X-HCFF, GOSTSHYP (Tim Stauch, Maximilian Scheurer)
  • Overhauled library of standard basis sets for consistency with Basis Set Exchange and extended support through element 118
  • Improved stability of ECP fitting and updated definitions of fitted ECPs (CRENBS, CRENBL, HWMB, LACVP, LANL2DZ, SBKJC)
  • Evaluation of electric field at nuclei (Yuezhi Mao)
  • Frequency calculations for rigid fixed-atom constraints (Saswata Dasgupta)
  • Save additional calculation output files to unique folder
  • Resolved issues with:
    • inconsistent application of quadrupole field to resolve orbital degeneracies
    • definition of jun-cc-pVDZ basis set (John Herbert)
    • some jobs crashing with the FILE_SET_SYM_REP read error
    • cleaning up in PES scan jobs on Windows
    • unnecessary gradient evaluation at every point of frozen PES scan

Density functional theory and self-consistent field

  • Vibronic and resonance Raman spectroscopy (Xunkun Huang, Huili Ma, WanZhen Liang)
  • Integrated DFT-D4 empirical dispersion model (Kuan-Yu Liu, Romit Chakraborty)
  • New implementation of direct propagation of the time-dependent Kohn-Sham equation (real-time TDKS) with support for unrestricted SCF and implicit solvation (Ying Zhu, John Herbert)
  • State-targeted energy projection method (Kevin Carter-Fenk, John Herbert)
  • Multiple improvements to frozen-density embedding methods (Cristina González-Espinoza, Alexander Zech, Tomasz A. Wesolowski)
  • Faster algorithm for wGDD tuning (John Herbert)
  • Improvements in the IP/EA omega tuning scripts for long range corrected functionals (John Herbert)
  • Support for high angular momentum in DFT frequency calculations
  • Superposition of atomic potentials (SAP) guess for SCF (Yu Zhang, Susi Lehtola)
  • Expand density functionals available for NMR chemical shift calculations (Jiashu Liang, Khadiza Begam, Barry Dunietz, Yihan Shao)
  • Nuclear gradient and analytical 2nd functional derivative of the VV10 functional (Jiashu Liang)
  • Performance improvements in the evaluation of DFT-D3 nuclear hessian contribution
  • Consistent constrained DFT and SCF convergence criteria (Kevin Carter-Fenk)
  • NVIDIA GPU computing improvements via interface with BrianQC:
    • Added support for GPU computing in Q-Chem for Windows
    • Accelerated force and vibrational frequency computations with range-separated functionals
    • Accelerated Fock derivative computation in DFT vibrational frequency jobs
  • Resolved issues with:
    • buffer overflow in a special case of very large DFT jobs
    • a special case of crashing unrestricted CIS derivative coupling calculations
    • evaluation of finite-difference nonlocal correlation orbital Hessian (Yuezhi Mao)
    • use of AO integrals in general response module
    • differences in DFT quadrature between Linux and macOS
    • using ghost atoms in MBD-vdW calculations (Kevin Carter-Fenk, Evgeny Epifanovsky)
    • using arbitrary density functionals for MBD-vdW and TS-vdW
    • crashing large CIS state following calculations
    • SOC constants with unrestricted TDDFT
    • RI-J/RI-K gradient
    • DFT hyperpolarizabilities

Correlated methods

  • Calculation of electronic circular dichroism (ECD) using EOM-CC (Josefine Andersen, Sonia Coriani)
  • Evaluation of spin-orbit couplings using CVS-EOM methods, L-edge XAS/XES spectroscopy calculations
  • Feshbach method with EOM-CC states and Coulomb wave expanded in terms of plane wave Gaussian type orbitals (Wojciech Skomorowski)
  • Improved performance in small to medium CC/EOM jobs via in-core computations
  • Improvements in projected CAP EOM-CC (James Gayvert)
  • IP/EA-ADC methods and intermediate state representation (ISR) properties (Adrian Dempwolff, Matthias Schneider, Alexander Paul)
  • Dramatic speedup of ADC(3) (Adrian Dempwolff)
  • Improved fourth-order static self-energy for all ADC variants (PP (EE), IP, EA) (Adrian Dempwolff)
  • Subspace-projected CAP-ADC for all ADC variants (PP (EE), IP, EA) (Adrian Dempwolff)
  • Evaluation of spin-orbit couplings using RAS-CI and RAS2-SF methods (Abel Carreras, Anna Krylov, David Casanova, Hanjie Jiang, Pavel Pokhilko, Paul M. Zimmerman)
  • Use of resolution-of-the-identity integrals in LibRASSF-based implementation of RAS-SF (Shannon Houck)
  • Implementation of the Bloch effective Hamiltonian approach within LibRASSF-based RAS-SF (Shannon Houck)
  • Experimental implementation of the CC2 and RI-CC2 methods (Garrette Paran, Thomas Jagau)
  • Implementation of the Brueckner CC2 method (Adam Rettig)
  • Implementation of direct RPA for the ground state correlation energy (Joonho Lee)
  • Cubic storage RI-MP3 and Laplace-transformed RI-MP2/RI-MP3 (Joonho Lee)
  • Added access to kappa-regularized orbital optimized MP2 via METHOD=koomp2
  • New implementation of v2RDM and v2RDM-CASSCF solvers (Rain Li, Wayne Mullinax, A. Eugene !DePrince III, Marcus Liebenthal)
  • Improved defaults in incremental FCI (Alan Rask)
  • Experimental implementation of tensor hypercontraction methods (Joonho Lee)
  • Resolved issues with:
    • CCSD electronic g-tensor initialization (Sven Kaehler)
    • 2 GB limit on temporary files in CC/EOM/ADC calculations on Windows
    • evaluation of analytic gradients of kappa-regularized OO-MP2
    • crashing in fragment excitation difference (FED) calculations due to insufficient memory (Aaditya Manjanath)
    • crashing in large RI-MP2 calculations
    • initial guess in EOM-DIP-CCSD calculations
    • crashing in large RI-CCSD calculations

Molecular dynamics

  • New AIMD variable (AIMD_INIT_VELOC_NANO_RANDOM) for better random seeds (Tarek Scheele)
  • Resolved issues with:
    • activating vibrational spectra computation in special cases

Large systems, QM/MM, and solvation

  • AIRBED: A simplified density functional theory model for physisorption on surfaces (Nick Besley, Stephen Mason)
  • Resolved issues with:
    • SM12 crashes with general basis set (Yuezhi Mao)
    • MM finite difference force calculations
    • printing of EFG principal components
    • SM12 gradient
    • implicit solvation in SCF and DFT response property calculations
    • requiring explicit derivative level to be set for IEF-PCM frequencies (John Herbert)
    • RI-MP2 + PCM jobs
    • out-of-memory error in large SMD jobs

Fragment and energy decomposition analysis

  • Enable geometry optimization on POL and VFB-CT surfaces in the presence of solvent (Yuezhi Mao)
  • Enable ALMO-EDA for systems with non-Aufbau electronic configurations (Yuezhi Mao)
  • Enable the separation of electrostatic and non-electrostatic terms in SMD solvation energy (Yuezhi Mao)
  • Improve error message when attempting ROHF-based SCFMI and EDA
  • Improve error message when attempting to use unsupported solvent models with SCFMI and EDA
  • Control number of subspace vectors and convergence threshold in SAPT CPSCF (Kevin Carter-Fenk)
  • Improved SAPT+aiDX and SAPT+MBD keywords (Kevin Carter-Fenk)
  • Resolved issues with:
    • crashing during large projection-based embedding calculations (Yuezhi Mao)
    • requiring explicit derivative level to be set for adiabatic EDA geometries and frequencies (Yuezhi Mao)
    • interoperability between SAPT features and various SAPT basis sets (Kevin Carter-Fenk)
    • crashing when using SAPT(KS) + cDFT with fragment-based Hirshfeld populations (Kevin Carter-Fenk)
    • memory usage in XSAPT (Kevin Carter-Fenk, John Herbert)

Other Release Logs: