Differences

This shows you the differences between two versions of the page.

--- doku:gaussion09 [2012/07/04 12:51] – external edit 127.0.0.1
+++ doku:gaussion09 [2021/06/14 09:28] (current) – removed jz
@@ Line 1: / Line 1: @@
-===== .bashrc entry for Gaussian Rev A.02  =====
-VSC-2:
-<code>
-# for Gaussian Rev A.02:
-export g09root=/opt/sw
-export GAUSS_SCRDIR="/scratch"
-. $g09root/g09/bsd/g09.profile
-</code>
-===== .bashrc entry for Gaussian Rev C.01  =====
-VSC-1:
-<code>
-# for Gaussian Rev C.01:
-export g09root=/opt/sw/g09_RevC.01
-export GAUSS_SCRDIR=/work/tmp
-. $g09root/g09/bsd/g09.profile
-</code>
-VSC-2:
-<code>
-# for Gaussian Rev C.01:
-export g09root=/opt/sw/g09-RevC01
-export GAUSS_SCRDIR=/scratch
-. $g09root/g09/bsd/g09.profile
-</code>
-===== Gaussian 09 Revision C.01 Release Notes =====
-==== Changes between Gaussian 09 Revisions B.01 and C.01: ====
-  - Changes to optimization algorithms and options:
-    * The selection of modes to include when stepping down from a region of wrong curvature during an optimization has been improved. This can also now be controlled by route options:<code>Opt=NoDownHill ... don't try to go downhill, just take RFO-like step.
-Opt=NGoDown=M ... Mix at most M eigenvectors in taking a downhill step.  The default is 3.</code>
-     * Linear bends are handled more reliably, and included in internal coordinate more frequently, than before. This avoids many optimization problems involving nearly linear angles becoming exactly linear.
-    c.  The connectivities of reactant and product are now merged in
-        generating the internal coordinates for the TS during QST2 and
-        QST3 optimizations.
-    d.  The maximum number of steps allowed ever in an optimization (i.e.,
-        include later restarts) can be reduced.  This is sometimes useful
-        for very large systems in order to reduce memory and disk usage.
-    e.  The program now checks if the standard orientation of a molecule
-        has flipped by 180 degrees during an optimization and avoids the
-        flip.  This avoids jumps when animating optimizations, IRCs, etc.
-        in GaussView and improves SCF convergence.
-    f.  The memory allocation for generation of internal coordinates is
-        now proportional to the amount of memory provided by %mem.  This
-        allows jobs with very large numbers of atoms or internal
-        coordinates which previously failed to run if enough memory is
-        provided.
-    g.  By default, internal coordinates for potential hydrogen bonds are
-        not generated automatically.  Bond coordinates are still added to
-        connect otherwise disjoint fragments, so coordinates for hydrogen
-        bonds which connect fragments will still be included.
-. Single-point BD calculations now default to frozen-core, with the
-    core orbitals uncorrelated but updated using the BD Fock matrix.
-    The previous default was to leave the core orbitals unchanged from
-    the HF values, or from the orbitals read in with BD=Read.  The new
-    default produces energies which are independent of the choice of
-    starting orbitals.  Gradients with BD still require and defaults to
-    full rather than frozen-core.  The OldFCBD keyword requests the
-    old-style frozen-core.
-  The memory required by very large ONIOM(MO:MM) and pure MM frequency
-    calculations has been reduced.
-. On some machines fully direct integral transformation and fully direct
-    MP2 are chosen if there is a large amount of memory, but the semidirect
-    algorithms are faster.  Tran=SemiDirect in the Default.Route file now
-    forces the SemiDirect algorithm for MP2 as well as the transformation
-    in higher level post-SCF calculations.  (All method keywords such as
-    MP2 are ignored in the Default.Route file, because otherwise they
-    would force that model in all calculations.)
-. Output=Wfn and Output=WfX with post-SCF methods now defaults to
-    Density=Current and Pop=NOAB, both of which are necessary for
-    the post-SCF density to be stored in the .wfn/.wfx file.  Problems
-    with the orientation of the forces in these files and in generating
-    them with ROHF wavefunctions and/or linearly dependent basis sets
-    have also been fixed.  Core densities are stored in the wfx file
-    for calculations which use ECPs, so that AIM and other analysis can
-    be done correctly for these cases.
-. Several customers have used the file generated for COSMORS as input
-    to their local simulation programs, so this capability has been put
-    back into G09.
-. Polar=Gamma has been added as more descriptive option for requesting
-    second hyperpolarizabilities.  It is synonym for Polar=(DCSHG,Cubic).
-. The definition of improper torsions in the Amber force field is dependent
-    on the ordering of atoms in the molecule.  Calculations in the Amber
-    program on typical proteins are consistent because of the standard
-    ordering of atoms withing residues and residues within a PDB file, but
-    for general molecules produced with GaussView the results depend on
-    the order of atoms in an arbitrary way.  G09 has been changed to average
-    over the 6 possible orders of atoms in an Amber improper torsion, making
-    the results slightly different than the standard Amber force field, but
-    making the energy independent of permuations of atoms in the molecule.
-. The RevTPSS exchange and correlation functionals have been added.
-. SDD now defaults to more recent basis sets for Actinides; OldSDD requests
-    the previous default.
-. Printing during Pop=MK with IOp33 increased has been restored to include
-    the data required for RESP charge fitting.  However, G09 can now generate
-    the data file for AnteChamber directly, by setting IOp(6/50=1) in the
-    Pop=MK job, and this is the recommended method for generating input for
-    RESP.
-. A bug in CIS frequencies with PCM solvation was fixed.
-. MaxDisk in a Default.Route file now applies to all steps of a compound
-    job; previously, only the first step was defaulted properly.
-. A bug which prevented reading AlpB parameters for AM1 was fixed.
-. Convergence during SCVS calculations is now checked more carefully.
-    Refer to the input files for tests 935-939 and 945 for examples of
-    using SCVS.
-. TB and TW can now be used to specify memory and disk allocations in
-    units of terabytes and terawords, respectively.
-. Pop=SaveBio in Stable=Opt jobs caused the stability calculation to
-    be wrong or fail.  This now works properly, saving the biorthogonal
-    orbitals only after the wavefunction has been made stable.
-. External point charges now work with symmetry turned on.
-. A bug in TD-DFT gradients with frozen core was fixed.
-. Print statements for NMR shielding were fixed to work with
-    more than 999 atoms.
-. A bug in DFTB using interpolated (not analytic) parameters with
-    d functions was fixed.
-. A rare problem with uncompleted write statements on slow file
-    systems was fixed.
-. Problems with some combinations of charge and multiplicity in
-    fragments during Guess=Fragment calculations were fixed.
-. Printing of Coriolis terms during Freq=VibRot was restored.
-. Some memory allocation problems for PBC calculations with large
-    unit cells were fixed.
-. Inconsistencies in how the geometry was modified in some cases by
-    Symm=Loose were fixed.
-. A bug in the ROMP4 triples energy when NoSymm was specified was fixed.
-. A warning "The extrapolated energy is higher than the direct energy"
-    is no long printed unnecessarily by the CBS extrapolation.
-. ONIOM(MO:MM) jobs which do microiterations and which fail to finish
-    are now restartable.
-. A bug in reading ECPs with ONIOM when the same ECP was placed on
-    multiple centers was fixed.
-. The combination of IRC and Freq, which did the frequency calculation
-    at the last point of the IRC rather than the TS, is now rejected.
-. Several unsupported combinations of Douglas-Kroll-Hess with properties
-    now generate an error message rather than incorrect answers.
-. A bug in generating the default (Harris) initial guess when using ECPs
-    on charged species was fixed.  The quality of the initial guess when
-    using ECPs has also been improved.
-. Several defaults for whether to use FMM and other integral options
-    have been updated for better performance on current models of CPU.
-. FormChk now writes -1 rather than ****** to the formatted checkpoint
-    file if the value exceeds 10^13-1.  This allows unfchk and other
-    utilities to process the resulting fchk file.
-. A bug affecting geom=check after numerical frequencies when using
-    ONIOM was fixed.
-. The Direct option is available for SAC-CI.  This requests an
-    integral-direct algorithm suitable for larger molecules.
-. A %oldchk link0 command has been added.  The contents of the
-    checkpoint file specified by %oldchk are copied to the checkpoint
-    file of the current job step at the start of the job step.
-    This allows data to be picked up from a previous calculation
-    without destroying anything on the chk file from it.
-. The combination of BD or W1BD with SCRF, which does not work
-    correctly, is now rejected by the route generator.
-. A new version of the ATLAS blas library is used on most platforms.
-    This fixes several problems when using very large amounts of memory.
-    In the event of such problems, IOp1=NoAssem can now be specified on
-    the route line to turn off use of the ATLAS matrix multiplication
-    routines.
-. A problem in reported transition moments between excited states
-    computed in SAC-CI jobs was fixed.
-. Empirical dispersion with DFT and ghost atoms now runs.  Empirical
-    dispersion and PBC now produces an error message, since it is not
-    implemented.
-. "Opt Freq" with ROHF/RODFT now works correctly, doing Freq=Numer
-    with the restricted open-shell wavefunction in the second job step.
-. Franck-Condon calculations now function correctly for forbidden
-    transitions.
-. The route generator rejects the combination of TD and double-hybrid
-    DFT methods, which never worked (previously, TD was done based on
-    only the SCF part of the double-hybrid).
-. IRC=(RCFC,GradientOnly) calculations now correctly use the Hessian
-    from the chk file.
-. Diffuse (aug-) functions were added for cc-pVDZ for the first
-    transition row.
-. A memory allocation bug for very large systems which could cause
-    a failure with the message "NIJ > Max2 in MMCore" was fixed.
-Changes specific to IBM AIX/Power machines:
-.  When building from source, the default is to build for the
-    current (Power7) processors.  To compile a version for the
-    Power5 or Power6 machines, instead of just "bldg09" you must
-    give the command "bldg09 all ibmp5".
-Changes specific to the Windows versions:
-.  A Windows64 version is now available.
-.  The external keyword functions correctly.  Look at
-    g09\tests\com\test726.gjf for an example of using it.
-.  A problem with the G09W front-end writing out multi-step
-    jobs when the --Link1-- lines were truncated was fixed.
-Corrections to deprecated features:
-.  Problems with setting non-integer nuclear charges in Massage input have
-    been corrected.  There is now a ZNuc function in Massage input which
-    changes the nuclear charge but not the atomic number.
-.  A problem with AIM analysis on Windows only was fixed.
-Changes between Gaussian 09 Revisions A.02 and B.01:
-.  A bug in MP2 frequencies with PCM was fixed.
-.  An updated version of the SAC-CI code is included.  This includes
-    a new integral-direct algorithm -- SACCI=(Direct,...) -- which is
-    much faster for large systems.
-.  The ExtraOverlay route keyword did not function in A.02; this has
-    been corrected.
-.  "Opt Freq" with TD now runs both job steps properly.
-.  NewZMat now writes out secondary structure information, if present,
-    with -opdb.
-.  NewZMat can now merge data from two input files.  Either two text
-    files or an input and a checkpoint file can be merged.  See the
-    website for examples.
-.  Problems with the dummy basis set used with molecular mechanics
-    when the system was highly charged or very high spin were fixed.
-.  Polar=(Cubic,DCSHG) can now be used to numerically differentiate
-    frequency-dependent hyperpolarizabilities (betas) to produce second
-    hyperpolarizabilities (gammas).  These polarizabilities are now
-    printed in the standard coordinate systems (i.e., with components
-    of beta along and perpendicular to the dipole moment).
-.  WfnX files, used by the newer versions of AIMPAC, can now be written
-    via Output=WfnX.
-. Performance for very large MM systems (>20K atoms) has been improved,
-    especially when range limits are applied to the Coulomb and
-    Van der Waals terms.  There is a new route option, Geom=Huge, which
-    turns off various actions, useful in QM calculations but
-    unnecessary and expensive with enormous MM runs.
-. MaxDisk can now be specified in the Default.Route file.
-. The free-format input routines have been generalized in order to make
-    data from newer DFTB parameter files acceptable.  These files still
-    require some modification to be used with G09; refer to the web site
-    for details.
-. The full tensors for ECD using TDDFT (including the quadrupole component)
-    are now printed.
-. The Hu, Lu, and Wang charge fitting model (JCTC 3, 1004-1013, 2007)
-    is now available via Pop=HLY.  The authors only parametrized the
-    atomic densities required for the model for the first 18 elements.
-    An alternative version, Pop=HLYGAt, uses the HLY fitting scheme but
-    with Gaussian's standard atomic densities, which are available for
-    the entire periodic table.  For systems which can be done either way,
-    the difference in atomic charges is usually between 1% and 5%.
-. The SCVS method of Todd Keith, which scales the molecule in order to
-    make the virial condition satisfied exactly, has been added.
-. The use of IOp's to specify user-selected ranges for integrals has
-    been updated in order to make it more general.
-. The default algorithm for optimizations when minimizing in a region
-    of incorrect curvature has been improved.
-. The initial guess for AM1 and PM6 has been improved.
-. More analysis of input ONIOM and MM parameters with respect to secondary
-    structure is now done (by default for systems with <10,000 atoms when
-    secondary structure information is available).  The net MM charges on
-    residues and average distances between residues are reported.
-. Various performance improvements, including ones for larger numbers
-    of SMP processors and for SCF frequency calculations.
-Usage Notes:
-.  If CIS frequencies are to be used with the Herzberg-Teller or
-    Franck-Condon-Herzberg-Teller analysis, the CIS frequencies must
-    be done numerically (Freq=Numer rather than Freq).  This is because
-    the transition dipole derivatives are not computed during the
-    analytic force constant evaluation.  The corresponding HF frequency
-    calculation on the ground state, which is also required, can be done
-    analytically as usual.
-.  CIS and CASSCF frequencies with PCM solvation must also be done numerically
-    using Freq=Numer.
-.  The linear scaling (FMM-based) algorithms are now Linda-parallel, so Linda
-    parallel jobs on large molecules do not need to specify NoFMM, and will run
-    faster with the default algorithms chosen by the program.
-.  Opt=GDIIS is still present but deprecated; the new default optimization
-    algorithm (Opt=GEDIIS) is better than GDIIS for the few cases where GDIIS was
-    better than the G03 default (Opt=RFO).
-.  Optimizations of large molecules which have many very low frequency vibrational
-    modes with DFT will often proceed more reliably when a larger DFT integration
-    grid is requested (Int=UltraFine).
-.  Density fitting can be made the default for jobs using pure DFT functionals
-    by adding the DenFit keyword to the route (-#-) line in the Default.Route file.
-    Fitting is faster than doing the Coulomb term exactly for systems up to several
-    hundred atoms (depending on basis set), but is slower than exact Coulomb using
-    linear scaling techniques (which are turned on automatically with exact Coulomb)
-    for very large systems.
-.  The default IRC algorithm has changed; refer to the User's Guide for details.
-    The default is to report only the energies and reaction coordinate at each
-    point on the path; if geometrical parameters along the path are desired, these
-    should be defined as redundant internal coordinates via Geom=ModRedundant or
-    as input to the IRC code via IRC=Report=Read.
-Changes in Usage and Defaults between Gaussian 03 and Gaussian 09:
-.  There are many changes in the PCM algorithms:
-    a.  The default surface integration is new and gives continuous
-        potential energy surfaces.  It is strongly recommended for all
-        new studies.  The route option SCRF=G03Defaults restores most
-        of the defaults to those in G03, but should be used only for
-        comparison with previous calculations done using G03.
-    b.  When using the default IEFPCM solvation method or SCRF=CPCM,
-        Gaussian 03 computed and reported non-electrostatic contributions
-        to the solvation energy but did not include these in the energies
-        and they were not included in the energies used for geometry
-        optimizations, frequencies, etc.  By default Gaussian 09 does not
-        compute these values at all.
-    c.  The new SMD solvation model is recommended for absolute solvation
-        energies and other properties for which the non-electrostatic
-        solvation terms are significant.  When SCRF=SMD is specified, the
-        SMD non-electrostatic terms are included in the basic energies
-        (the SCF energy reported in the "SCF Done" line, correlated energies,
-        etc.) and are included in the geometry optimization and frequency
-        calculations.  The non-electrostatic energy is also reported
-        separately.
-        Absolute solvation energies should be computed by doing a gas-phase
-        optimization and frequency calculation on the system, followed by
-        the same calculations with SCRF=SMD or SCRF=(SMD,Solvent=...).
-    d.  The SCFVac PCM input option has been removed.  If a preliminary
-        gas-phase energy is desired, do this in a separate job step before
-        the solvated calculation.
-    e.  If the quality of the PCM integration grid is to be changed, the
-        keyword PTSDens should be used in the PCM input section.  This specifies
-        the density of integration points in points per Angstrom^2.  The
-        old keyword TSAre specifies the area per tessera and hence implies use
-        of the old integration scheme involving tesserae, which is inferior.
-        The default is approximately 5 points per A^2, so PTSDens=10.0 will
-        cause about twice as many points to be used compared to the default.
-.  MP and CC calculations now default to a partial transformation (Tran=IABC).
-    This is faster on most systems, especially when several processors are used.
-    A full transformation can be requested using Tran=Full.
-.  The default SCF convergence is 10^-8 on the density for all calculations,
-    including single points.
-.  The physical constants used by default are those from the 2006 CODATA tables;
-    those used in Gaussian 03 can be requested via Constants=1998.
-.  AM1, PM3, and PM3MM by default use the new semi-empirical code, which has
-    proper analytic first and second derivatives but which gives slightly different
-    total energies because it computes the overlap integrals via 6-Gaussian expansions
-    rather than over Slater functions.  AM1=Old and Use=L402 both request use of the
-    old (MOPAC 6) code, through the regular links or through link 402, respectively.
-    The new code is strongly recommended except when comparison with results from
-    Gaussian 03 is required.
-.  Stable=Opt defaults to the usual (L502) SCF procedure for the initial SCF but
-    then uses SCF=QC for additional SCF calculations, if they are required.
-Changes between Gaussian 09 Revision A.01 and A.02:
-.  The logic for handling extra negative eigenvalues of the Hessian
-    during minimizations has been improved.
-.  The combination of DFT and General SCF, automatically turned on if
-    DFT is requested along with Int=DKHSO, does not work and is now
-    rejected by the route generator.
-.  ONIOM input is now checked for divalent link atoms.  The position
-    of these atoms is ill-defined unless the distance scale factors
-    are set to 1, and the model is usually poor if the scale factor
-    is forced to be 1.  Since this input is normally an error, it is
-    now rejected by link 101.  IOp 132 can be used to force acceptance
-    of this input, but this is strongly discouraged.
-.  Semidirect integral transformation is the the default.  This
-    code parallelizes better than the fully direct or in-core
-    algorithms and is similar in speed on a single processor.
-.  A bug which caused ONIOM=InputFiles to fail when PDB secondary
-    structure information was included in the input has been fixed.
-Building from source code:
-.  Building Gaussian 09 with Linda requires Linda version 8.2; the
-    executables will not build with previous version of Linda.
-.  Building on Intel Macs requires a case-sensitive file system.
-    In order to build the ia32 version you must specify:
-    bsd/bldg09 all mac32
-    as there is no way for the build script to determine that it is
-    running on a x86 rather than x86-64 machine and the default is to
-    build for x86-64.
-.  When building from source on IBM Power systems, the default is to
-    build for the current (Power7) processors.  To compile a version
-    for the Power5 or Power6 machines, instead of just "bldg09" you must
-    give the command "bldg09 all ibmp5".
-Missing Reference:
-The G09 User's Guide omitted the paper Clemente08 from the
-bibliography.  It is:
-F. Clemente, T. Vreven, and M. J. Frisch, in Quantum Biochemistry,
-Ed. C. Matta (Wiley VCH, 2008).