Computational Chemistry Software

News

(15.12.2013): Cygwin version of antechamber-1.27 binaries is now available

(14.12.2013): Cygwin version of GROMACS-4.6.5 binaries is now available
The compilation procedure is outlined here
Note: if you have no MPI installed, -nt 1 key may be useful!

(18.10.2012): An updated version of EGView is avalable: now ORCA output files (for geometries and energies) are suported

(25.08.2012): NwChem 6.1.1 binaries build with openmpi-1.4.5 for 32-bit LINUX are available (64-bit mpirun is included; 32-bit mpirun is here; please, mail to me and say, whether it works ok :) )

(08.05.2012): The NwChem Feature (but not a 'bug', IMHO):

* the Dunning basis sets were designed to be used with spherical basis functions.
  The NWChem default is cartesian, so the basis section input needs to iclude spherical keyword
  ("basis spherical" instead of "basis")
* the "freeze atomic" option might be needed to be added in MP2 calculations, since
   (1) this is the default in Gaussian and noobs will be confused when
       all-electron result does not agree with Gaussian, and
   (2) Dunning basis sets are supposed to be used with frozen-core for correlated calculations.

See also the discussion on the topic:

 * HF SCF energies: One can get consistent results from both programs by either specifying
   "basis cartesian" instead of "basis spherical" the NWChem input or specify 5D functions
   in the Gaussian03 input.
 * In Gaussian, specifying MP2 defaults to freezing the core orbitals, while in NWChem, MP2
   signifies full correlation. In order to get the same energies: specify "MP2(full)/6-31G*"
   in the Gaussian03 input or specify "freeze atomic" in the MP2 directive of an NWChem input.

(08.03.2012) New version of nwRun, a user-friendly GUI shell for running nwchem.exe is now available (see description here)

(08.03.2012) Be sure to use NWCHEM_BASIS_LIBRARY environment variable to specify the path to NwChem basis set description files

(08.03.2012) The nw-6.1-basis_library.zip arvhieve contains basis set collection from NwChem 6.1

(04.02.2012) NwChem 6.1 for Windows in now available (the simplest classical build - see note.txt in arcieve; 19 Mb)
More builds (including python, mpi, wfn generator etc) are coming soon...

(02.02.2012) NWChem developers announced that NWChem 6.1 has been released.

NWChem 6.0 for Windows

Download cygwin-based win32 binaries, built from original NwChem 6.0 source code with minor modifications to achieve successful compilation:

The user manual documentation is also avaiballe in single-file pdf format.

A GUI user-friendly shell for running nwchem.exe is also available (see description here)

Updated 01.10.2011
For technical support and/or any questions please, write to tim_mail (at) ukr (dot) net

EGView (version 09.07.2011)

BSSE-corrected Stabilization energy SE is obtained as
SE = E^A+B_DCBS - E^A_DCBS - E^B_DCBS = E^A+B_CP - E^A_MCBS - E^B_MCBS
where E^A+B_DCBS is BSSE-uncorrected electron energy, E^A_DCBS is the electron energy of fragment A in a complex obtained using DCBS (Dimer-Centered Basis Set), E^B_DCBS is the electron energy of fragment B in a complex obtained using DCBS (Dimer-Centered Basis Set), E^A+B_CP is BSSE-corrected electron energy using counterpoise method:
E^A+B_CP = E^A+B_DCBS + ΔE_BSSE with the BSSE-correction value ΔE_BSSE = E^A_MCBS - E^A_DCBS + E^B_MCBS - E^B_DCBS
Note that ΔE_BSSE > 0 sicnce E^X_DCBS < E^X_MCBS.

All the values from these equations appear in the Gaussian .out file if the counterpoise keyword is used.
Here are Gaussian keywords identifying different stages of the counterpoise procedure as they appear in Gaussian .out file:
E^A_DCBS

 Counterpoise: doing DCBS calculation for fragment   1 NewBq=T
 ...
 SCF Done:  E(RB+HF-LYP) =  -56.5595347922     A.U. after    5 cycles
             Convg  =    0.1804D-08             -V/T =  2.0095
             S**2   =   0.0000

Here NewBq is a synonym for NewGhost
and DCBS = Dimer centered basis set
E^B_DCBS

 Counterpoise: doing DCBS calculation for fragment   2 NewBq=T 
 ...
 SCF Done:  E(RB+HF-LYP) =  -114.503484307     A.U. after    6 cycles
             Convg  =    0.3794D-08             -V/T =  2.0090
             S**2   =   0.0000

E^A_MCBS

 Counterpoise: doing MCBS calculation for fragment   1 
 ...
 SCF Done:  E(RB+HF-LYP) =  -56.5577674045     A.U. after    5 cycles
             Convg  =    0.2728D-08             -V/T =  2.0092
             S**2   =   0.0000

Here MCBS = monomer-centered basis set
E^B_MCBS

 Counterpoise: doing MCBS calculation for fragment   2 
 ...
 SCF Done:  E(RB+HF-LYP) =  -114.503149487     A.U. after    5 cycles
             Convg  =    0.9897D-08             -V/T =  2.0090
             S**2   =   0.0000

For post-SCF methods keywords such as EUMP2 etc appears after "Counterpoise: doing XXXX calculation for fragment Y"
The following output appears regardless of the energy calculation method used:
E^A+B_CP

 Counterpoise: corrected energy =    -171.065201221288

and the BSSE-correction energy (ΔE_BSSE)

 Counterpoise: BSSE energy =       0.002102207582

Note that all these values are calculated for fragments geometries as they are in complex, so their deformation energy is not included. This is a good approximation (< 0.1 kcal/mole) only if interacting molecules are rigid enough.

Computational Chemistry Software

News

NWChem 6.0 for Windows

EGView (version 09.07.2011)

Some open-source QC code references

Leave a comment/question/suggestion