Postg

Overview

Postg is a program that implements the calculation of the dispersion energy and related quantities using the exchange-hole dipole moment(XDM) model. This program reads the output of a Gaussian (http://www.gaussian.com/) calculation (a wfn, wfx, fchk or molden file) and computes the XDM atomic moments and volumes, dispersion coefficients ($C_6$, $C_8$, and $C_{10}$), Hirshfeld charges, the dispersion energy and its first and second derivatives, and the total energy. The code can be used to calculate the dispersion energy at a single point or to perform geometry optimizations and frequency calculations using Gaussian's "external" keyword.

To use postg, you will need the parameters for the damping function, which depend on the exchange-correlation functional you are using and also on the basis set (if it is far from complete). A list of parameters is shipped with the postg distribution (a file called xdm.param). This data can also be found online at https://erin-r-johnson.github.io/software/ .

Compilation

To build postg, first uncompress the package or clone the repository:

tar xjvf postg.tar.bz2
cd postg/

and then edit the Makefile and uncomment the block corresponding to your compiler (or write your own if it's not in the list). I normally use the Intel fortran compiler for postg, so my compilation flags look like:

FC = ifort
FCFLAGS = -O2 -openmp
LDFLAGS = -O2 -openmp

The "-openmp" option activates the shared-memory parallelizaton. After this, you should be able to generate the "postg" binary by running:

make

No environment variables need to be set to run postg nor any special setup is required. Shared-memory parallelization is implemented using the OpenMP library, so the OMP_NUM_THREADS environment variable can be used to control the number of processors that is used in the calculation. For instance, use:

export OMP_NUM_THREADS=8

in your .bashrc (or the equivalent setenv in your .cshrc if you use csh) to get postg to use 8 threads in your calculations. Usually postg takes mere seconds to complete the calculation, but in large systems this may be important.

Use

In order to use postg, you need a gaussian-style wfn, wfx, fchk (formatted checkpoint) or molden file. The first two are generated by passing the "output=wfn" or "output=wfx" keywords to the gaussian input, followed by the name of the file at the end. The fchk can be obtained from the checkpoint file (use "%chk" before the route section) by applying the formchk utility program to it. Many programs (including psi4, gaussian, and gamess) generate molden-style wavefunction files (extension .molden).

If, for instance, your wavefunction file is calc.wfn, then either

postg 0.754 1.102 calc.wfn lcwpbe
postg BJ 0.754 1.102 calc.wfn lcwpbe

will calculate the XDM(BJ) dispersion energy using BJ-damping parameters $a1=0.754$, $a2=1.102$ Å and the atomic volumes corresponding to the LC-wPBE functional.

To use XDM(Z), which uses the new (2025) one-parameter damping function based on atomic numbers, then either

postg 138857 calc.wfn lcwpbe
postg Z 138857 calc.wfn lcwpbe

will calculate the XDM(Z) dispersion energy using the Z-damping parameter $z_damp=138857$ and the atomic volumes corresponding to the LC-wPBE functional.

To use XCDM, the new (2025) exchange-correlation-hole dipole moment model, with either BJ or Z damping, then append "xcdm" to the end of the input arguments:

postg BJ 0.754 1.102 calc.wfn lcwpbe xcdm
postg Z 138857 calc.wfn lcwpbe xcdm

Damping function parameters for common functionals and basis sets can be found in https://erin-r-johnson.github.io/software/ . or in the included xdm.param file. The functional keyword can be one of:

• blyp
• b3lyp
• bhandhlyp, bhandh, bhah, bhahlyp
• camb3lyp, cam-b3lyp
• pbe
• pbe0
• lcwpbe, lc-wpbe
• pw86, pw86pbe
• b971, b97-1
• A number between 0 and 1 specifying the amount of exact exchange in the functional. In this case, the atomic volumes of the lighter elements are found by interpolation.

Using wfx instead of wfn is strongly recommended. We never got the format of the wfn nailed down so using wfns may fail for large systems with thousands of primitives.

Optimizations and frequency calculations

The "external" keyword in gaussian can be used together with the "gxdmopt.sh" script, provided in the postg package, to perform geometry optimizations ("opt") and frequency ("freq") calculations. The typical input is:

#p external='./gxdmopt.sh' opt=(nomicro) freq

title

0 1
... geometry ...

The "gxdmopt.sh" script has to be modified to specify the functional, basis, ECPs, etc. Inside this script, you can find a "modify this" section at the beginning, that encapsulates all the information that needs to be tweaked:

## modify this
chf="lcwpbe"          # the functional keyword for postg. Can be a number
c1=1.0875             # a1 parameter
c2=0.4850             # a2 parameter
basis="aug-cc-pvtz"   # basis set file or (if not found) basis set keyword
ecp=""                # ecp file
G09="g09"             # gaussian executable
POSTG="postg"         # postg executable
verbose=""            # verbose flag: write more info about the
                      # intermediate gaussian calculation. verbose=1
                      # gives huge output files.
guesstrick=""         # if guesstrick=bleh, use bleh.chk as checkpoint file
                      # for the underlying calculation and re-use it from
		  # one step to the next. Speeds up geometry optimizations.
cat > $2.route <<EOF  # this is the route section of the calculation
%mem=2GB              # it is passed as-is to the underlying gaussian process
%nprocs=4             # basis set and ecp's pseudo=read are inserted automatically
#p lc-wpbe int(grid=ultrafine)
EOF
#########

The directory "examples" contains some input files showing how to use this script with gaussian:

• ecps: optimization of the Zn-He dimer using ECPs for Zn.
• opt: optimization of the CO2 dimer.
• optfreq: optimization and frequency calculation in the Ne dimer.
• opt_open: methane/methyl radical optimization (open shell).
• sp: fixed-geometry calculation in the methane dimer. This can be done using the gxdmopt.sh script (ch4_ch4_script) or manually by requesting the wfx file to be written with 'output=wfx' and then running postg (ch4_ch4_manual and runme.sh).

References and citation

The basic references for postg are:

• A. Otero-de-la-Roza and E. R. Johnson, J. Chem. Phys. 138, 204109 (2013). (http://dx.doi.org/10.1063/1.4807330)
• F. O. Kannemann and A. D. Becke, J. Chem. Theory Comput. 6, 1081 (2010). (http://dx.doi.org/10.1021/ct900699r)

Please, consider citing us if you find this program useful. The basic XDM references are:

• E. R. Johnson and A. Becke, J. Chem. Phys. 123, 024101 (2005). (http://dx.doi.org/10.1063/1.1949201)
• A. D. Becke and E. R. Johnson, J. Chem. Phys. 122, 154104 (2005). (http://dx.doi.org/10.1063/1.1884601)
• A. D. Becke and E. R. Johnson, J. Chem. Phys. 123, 154101 (2005). (http://dx.doi.org/10.1063/1.2065267)
• A. D. Becke and E. R. Johnson, J. Chem. Phys. 124, 014104 (2006). (http://dx.doi.org/10.1063/1.2139668)
• E. R. Johnson and A. D. Becke, J. Chem. Phys. 124, 174104 (2006). (http://dx.doi.org/10.1063/1.2190220)
• E. R. Johnson and A. D. Becke, Chem. Phys. Lett. 432, 600 (2006). (http://dx.doi.org/10.1016/j.cplett.2006.10.094)
• A. D. Becke and E. R. Johnson, J. Chem. Phys. 127, 124108 (2007). (http://dx.doi.org/10.1063/1.2768530)
• A. D. Becke and E. R. Johnson, J. Chem. Phys. 127, 154108 (2007). (http://dx.doi.org/10.1063/1.2795701)
• F. O. Kannemann and A. D. Becke, J. Chem. Theory Comput. 5, 719 (2009). (http://dx.doi.org/10.1021/ct800522r)
• F. O. Kannemann and A. D. Becke, J. Chem. Theory Comput. 6, 1081 (2010). (http://dx.doi.org/10.1021/ct900699r)
• A. D. Becke, A. Arabi, and F. O. Kannemann, Can. J. Chem. 88, 1057 (2010). (http://dx.doi.org/10.1139/V10-073)
• E. R. Johnson, J. Chem. Phys. 135, 234109 (2011). (http://dx.doi.org/10.1063/1.3670015)
• A. Arabi and A. Becke, J. Chem. Phys. 137, 014104 (2012). (http://dx.doi.org/10.1063/1.4731342)
• A. Otero-de-la Roza and E. R. Johnson, J. Chem. Phys. 138, 054103 (2013). (http://dx.doi.org/10.1063/1.4789421)

And XDM applied to solid-state calculations:

• A. Otero-de-la Roza and E. R. Johnson, J. Chem. Phys. 136, 174109 (2012). (http://dx.doi.org/10.1063/1.4705760)
• A. Otero-de-la Roza and E. R. Johnson, J. Chem. Phys. 137, 054103 (2012). (http://dx.doi.org/10.1063/1.4738961)
• E. R. Johnson and A. Otero-de-la Roza, J. Chem. Theory Comput. 8, 5124 (2012). (http://dx.doi.org/10.1021/ct3006375)

And XCDM and Z-damping:

• K. R. Bryenton and E. R. Johnson, (Manuscript in preparation 2025)
• A. Becke, J. Chem. Phys. 160, 20 (2024). (http://dx.doi.org/10.1063/5.0207682)

Copyright notice

Copyright (c) 2013-2025 Alberto Otero de la Roza aoterodelaroza@gmail.com, Kyle R. Bryenton kyle.bryenton@gmail.com, Felix Kannemann felix.kannemann@dal.ca, Erin R. Johnson erin.johnson@dal.ca, Ross M. Dickson ross.dickson@dal.ca, Hartmut Schmider hs7@post.queensu.ca, and Axel D. Becke axel.becke@dal.ca

postg is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.