Usage¶

Three files are needed in order to run GIMIC:

The XDENS file containing the effective one-particle density, and the magnetically perturbed densities in AO basis;
The MOL file with information on molecular geometry and basis sets;
The GIMIC input file. The default name is gimic.inp, however, any other file name is acceptable.

The XDENS and MOL files are obtained externally using other quantum chemistry programs which support an interface to gimic. Details are presented below.

Before doing the actual calculation with GIMIC, it might be a good idea to check if the grid is correctly defined. The dryrun flag produces the grid.xyz file which can be examined in any molecular viewer.

$ gimic --dryrun

The calculation with GIMIC is started by:

$ gimic [gimic.inp] > gimic.out

The square brackets mean that it is not necessary to give the name of the input file if it is called gimic.inp since it is recognised by default.

Interfaces to GIMIC¶

The following sections explains how to obtain the XDENS and the MOL files.

Running TURBOMOLE¶

As of TURBOMOLE 5.10, the GIMIC interface is part of the official distribution. To produce the necessary files to run GIMIC, you first need to optimize the wavefunction/density of the molecule, before running the mpshift program which produces the perturbed density matrices. Before you run mpshift, you need to edit the control file and add the $gimic keyword. When the calculation has finished, TURBOMOLE writes two files called CAODENS (AO density data) and XCAODENS (perturbed density data). In case either of them is missing, the mpshift calculation has not converged. It is necessary to add the $csmaxiter N keyword to the control file, where N is larger than the default value of 30. Afterwards, run the turbo2gimic.py script (distributed with GIMIC) in the same directory to produce the MOL and XDENS files.

$ turbo2gimic.py > MOL

Running LSDalton¶

The scripts/input needed can be found in tools/lsdalton2gimic.

Written by C. Kumar, University of Oslo, chandan.kumar@kjemi.uio.no.

.GIMIC needs to be added in LSDALTON.INP.

Note that open-shell calculations are not supported.

Run in the same directory:

python lsdalton2gimic.py

Running GAUSSIAN¶

The nuclear shielding calculation on Gaussian needs to be performed by including the keyword IOp(10/33=2) in order to print the perturbed density matrices in the output file. Explicitly specifying NMR=GIAO is not necessary since using GIAO’s is the default in G09. The keyword Int=NoBasisTransform is needed in order to prevent Gaussian from transforming the generalized contraction basis sets. It ensures that the results will match with the ones obtained in Turbomole.

The general structure of the input file looks like this:

%chk=file.chk
#P <method>/<basis> nmr pop=regular Int=NoBasisTransform IOp(10/33=2)
<empty line>
title
<empty line>
<charge> <multiplicity>
<coordinates>
<empty line>

In the above, the lines with the < > symbols are supposed to be modified. These bracket symbols are not part of the actual input file.

If the basis set needs to be specified explicitly, the input file is structured as follows:

%chk=file.chk
#P <method> nmr pop=regular Int=NoBasisTransform IOp(10/33=2)
<empty line>
title
<empty line>
<charge> <multiplicity>
<coordinates>
<empty line>
<basis set specification>
<****>
<empty line>

If ECPs are needed, then the method specification line should like as in the example below.

#P TPSSTPSS/GenEcp nmr pop=regular Int=NoBasisTransform IOp(10/33=2)

It seems there is a small difference between the keywords nosymmetry and Symmetry=None. The latter should only be used if nosymmetry generates an error.

The scripts/input needed can be found in /tools/g092gimic.

This script has been provided by Vincent Liegeois from the University of Namur, vincent.liegeois@unamur.be.

The tool consists of two parts:

Gaussian2gimic.py which is the main script
BasisSet.py which is a module file containing the functions to read the basis set and to do the transformations from Spherical to Cartesian. This file just needs to be put in the same directory as Gaussian2gimic.py.

Gaussian2gimic.py uses optionparser to sets its different options. Therefore, Gaussian2gimic.py -h will give you the full description.

There are two options: -i and -t, however the latter is optional. The command which runs the script on a formatted checkpoint file from GAUSSIAN is the following:

Gaussian2gimic.py -i file.fchk

It will produce the XDENS and MOL files

The extra option -t accept the argument of the XDENS file, which matches the output of the interface to TURBOMOLE obtained with turbo2gimic.py.

Gaussian2gimic.py -i file.fchk -t XDENS_turbo

In addition to creating XDENS and MOL files, the script prints to the terminal a comparison with the values obtained from GAUSSIAN and from TURBOMOLE. This is intended as a way to make sure that both numbers are similar.

For example, for a calculation on paranitroaniline with HF/6-311G(2df,2pd), the maximum error on the density matrix is 5.8E-5 while the maximum errors on the perturbed density matrices are 2.1e-2, 2.6e-2, 5.0e-2.

For the same molecule but with HF/cc-pVTZ, the maximum errors are: 4.8e-5, 2.5e-2, 5.4e-2, 7.8e-2.

The MOL file produced by Gaussian2gimic is slightly different from the one obtained by turbo2gimic.`

Indeed, turbo2gimic.py gives the basis set exactly as obtained on https://bse.pnl.gov/bse/portal with “optimized general contractions” checked. However, the coefficients are not normalized with “optimized general contractions” checked. They are normalized when “optimized general contractions” is unchecked.

For example, the first atomic orbital in the cc-pVTZ basis set for carbon with “optimized general contractions” checked consist of a contraction of 8 GTOs, while there are 10 GTOs in the contraction if “optimized general contractions” is unchecked. The coefficients given in the website and in turbo2gimic.py are optimized for the contraction of the 10 GTOs rather than 8. Gaussian and therefore Gaussian2gimic.py give the coefficients that are normalized for the contraction of the 8 GTOs.

Example input for benzene:

%Chk=benzeneg09.chk
%mem=2000mb

#p B3LYP/Def2TZVP SCF=Tight NMR=GIAO Int=NoBasisTransform IOp(10/33=2)

Benzene Gaussian NMR example

0 1
C    1.2049777911    0.6956942520    0.0000000000
C    1.2049777911   -0.6956942520    0.0000000000
C    0.0000000000   -1.3913885041    0.0000000000
C   -1.2049777911   -0.6956942520    0.0000000000
C   -1.2049777911    0.6956942520    0.0000000000
C    0.0000000000    1.3913885041    0.0000000000
H    2.1430161769    1.2372709666    0.0000000000
H    2.1430161769   -1.2372709666    0.0000000000
H    0.0000000000   -2.4745419332    0.0000000000
H   -2.1430161769   -1.2372709666    0.0000000000
H   -2.1430161769    1.2372709666    0.0000000000
H    0.0000000000    2.4745419332    0.0000000000

Running Gaussian creates a file benzeneg09.chk You need to convert this *.chk file to a formatted *.fchk file.

$ formchk file.chk file.fchk

Then you can proceed as described above and generate the MOL and XDENS files with:

$ Gaussian2gimic.py --input=benzeneg09.fchk

Note that for open-shell cases you need to add “gfprint pop=regular iop(10/33=2)” and use the Gaussian *.log file instead of the *.fchk file.

Example input for the triplet dication of benzene:

%LindaWorkers=cib26-2
%NProcShared=20
%Chk=benzeneg09.chk
%mem=2000mb

#p POP=FULL GFPrint nosymmetry B3LYP/DEF2TZVP SCF=Tight NMR IOp(10/33=2)

Benzene Gaussian NMR example triplet dication

2 3
C    1.2049777911    0.6956942520    0.0000000000
C    1.2049777911   -0.6956942520    0.0000000000
C    0.0000000000   -1.3913885041    0.0000000000
C   -1.2049777911   -0.6956942520    0.0000000000
C   -1.2049777911    0.6956942520    0.0000000000
C    0.0000000000    1.3913885041    0.0000000000
H    2.1430161769    1.2372709666    0.0000000000
H    2.1430161769   -1.2372709666    0.0000000000
H    0.0000000000   -2.4745419332    0.0000000000
H   -2.1430161769   -1.2372709666    0.0000000000
H   -2.1430161769    1.2372709666    0.0000000000
H    0.0000000000    2.4745419332    0.0000000000

When the calculation completes, run in the terminal:

mv file.out > file.log
Gaussian2gimic.py --input=file.log

For the present example, a current strength susceptibility of 8.4 nA/T was calculated.

Running QChem and FERMION++¶

The scripts/input needed can be found in /tools/qchem.

Written by J. Kussmann, University of Munich, jkupc@cup.uni-muenchen.de.

Convert Output of Q-Chemor FermiONs++ for GIMIC (TURBOMOLE format)

For a list of options, type:

qc2tm -h

which prints

USAGE: qc2tm -t <qchem or fermions> -qcout <output-file> -scr
         <scratch-directory> -s2c (opt.) -openshell (opt.)

Running CFOUR¶

Do a normal NMR calculation and then run the xcpdens program distributed with GIMIC to make the XDENS file. Then run the MOL2mol.sh script to produce the MOL file.

Running ACES2¶

Using ACES2, the special driver script ’xgimic2.sh’ must be used to run the NMR shielding calculation. Modify the script to suit your needs (and set the paths correctly). If the NMR calculation is done with symmetry, the MOL file must be converted to C1 symmetry using the script MOL2mol.sh, prior to running GIMIC.

Example ZMAT:

CO2
O    2.14516685791074   0.00000000000000      0.00000000000000
C    0.00000000000622   0.00000000000000      0.00000000000000
O   -2.14516685791393   0.00000000000000      0.00000000000000

*ACES2(CALC=CCSD,BASIS=tzp,UNITS=BOHR
COORD=CARTESIAN
MEMORY=250000000
REFERENCE=RHF
SYMMETRY=ON
PROPERTY=NMR
MULTIPLICITY=1
CHARGE=0
SCF_MAXCYC=200,CC_MAXCYC=150,CC_EXPORDER=40
CC_CONV=10,SCF_CONV=10,LINEQ_CONV=10,CONV=10
LINEQ_EXPAN=30)

Run ACES2 via xgimic2.sh to produce the XDENS file:

$ xgimic2.sh --cc >aces2.out &

Convert the symmetry-adapted MOL file to C1 symmetry:

$ MOL2mol.sh

The new MOL file is now called mol.