6-311G basis set for transition metals

Clicked A Few Times

2:09:08 AM PDT - Tue, Jun 16th 2015

Hi All,

I want to use the 6-311G basis set as this was suggested by another group to give good results for Cu(I) complexes.

However, this other group uses Gaussian and there the 6-311G basis set seems to contain first row transition metals, especially Cu, which is not the case in the EMSL /NWChem basis sets. Is there a version of the 6-311G basis set that also contains 1st row transition metals for NWChem?

This information might help: When I tell Gaussian to print the Cu 6-311G basis set I obtain:

     1 0

S   6 1.00       0.000000000000

     0.3372000000D+06  0.5986907519D-03

     0.5007290000D+05  0.4639853327D-02

     0.1137340000D+05  0.2399039370D-01

     0.3239820000D+04  0.9540564768D-01

     0.1071970000D+04  0.2958387587D+00

     0.3950990000D+03  0.6682885518D+00

S   1 1.00       0.000000000000

     0.1583990000D+03  0.1000000000D+01

S   1 1.00       0.000000000000

     0.6735910000D+02  0.1000000000D+01

S   1 1.00       0.000000000000

     0.2229830000D+02  0.1000000000D+01

S   1 1.00       0.000000000000

     0.9393570000D+01  0.1000000000D+01

S   1 1.00       0.000000000000

     0.2578480000D+01  0.1000000000D+01

S   1 1.00       0.000000000000

     0.9640800000D+00  0.1000000000D+01

S   1 1.00       0.000000000000

     0.1133030000D+00  0.1000000000D+01

S   1 1.00       0.000000000000

     0.4079100000D-01  0.1000000000D+01

P   5 1.00       0.000000000000

     0.2245290000D+04  0.3179198104D-02

     0.5321060000D+03  0.2589780402D-01

     0.1721950000D+03  0.1216254283D+00

     0.6532390000D+02  0.3580114679D+00

     0.2705510000D+02  0.6107858516D+00

P   1 1.00       0.000000000000

     0.1174350000D+02  0.1000000000D+01

P   1 1.00       0.000000000000

     0.4693820000D+01  0.1000000000D+01

P   1 1.00       0.000000000000

     0.1906670000D+01  0.1000000000D+01

P   1 1.00       0.000000000000

     0.7114450000D+00  0.1000000000D+01

D   3 1.00       0.000000000000

     0.5364780000D+02  0.5683578684D-01

     0.1507470000D+02  0.3080428785D+00

     0.5103920000D+01  0.7640422112D+00

D   1 1.00       0.000000000000

     0.1727430000D+01  0.1000000000D+01

D   1 1.00       0.000000000000

     0.5283220000D+00  0.1000000000D+01

F   1 1.00       0.000000000000

     0.1440000000D+01  0.1000000000D+01

Can this output be somehow converted to match the NWChem basis set format or is there another way to obtain the 6-311G basis set for Cu?

Gets Around

11:21:59 AM PDT - Tue, Jun 16th 2015

As it happens I have recently worked on code for basis set conversion to NWChem from Gaussian 94 style basis data as used by Psi4: https://github.com/mattbernst/ebsel

Here's the converted output for NWChem/copper using your data above as input:

#BASIS SET reformatted: [9s,5p,3d,1f]
#origin: manually supplied data
basis "Cu_6-311G" spherical
Cu     S
  337200.0000000               0.0005987
   50072.9000000               0.0046399
   11373.4000000               0.0239904
    3239.8200000               0.0954056
    1071.9700000               0.2958388
     395.0990000               0.6682886
Cu     S
     158.3990000               1.0000000
Cu     S
      67.3591000               1.0000000
Cu     S
      22.2983000               1.0000000
Cu     S
       9.3935700               1.0000000
Cu     S
       2.5784800               1.0000000
Cu     S
       0.9640800               1.0000000
Cu     S
       0.1133030               1.0000000
Cu     S
       0.0407910               1.0000000
Cu     P
    2245.2900000               0.0031792
     532.1060000               0.0258978
     172.1950000               0.1216254
      65.3239000               0.3580115
      27.0551000               0.6107859
Cu     P
      11.7435000               1.0000000
Cu     P
       4.6938200               1.0000000
Cu     P
       1.9066700               1.0000000
Cu     P
       0.7114450               1.0000000
Cu     D
      53.6478000               0.0568358
      15.0747000               0.3080429
       5.1039200               0.7640422
Cu     D
       1.7274300               1.0000000
Cu     D
       0.5283220               1.0000000
Cu     F
       1.4400000               1.0000000

Edit: manually added basis .. spherical header.

Clicked A Few Times

1:09:01 AM PDT - Wed, Jun 17th 2015
Hi Mernst, many thanks for your fast response and the conversion, that really helps a lot!!! One more question, do you have any idea why Gaussian, in contrast to EMSL basis set exchange, has included 1st row transition metals (e.g. Cu) to the 6-311G basis sets.

Gets Around

12:20:21 PM PDT - Wed, Jun 17th 2015
The Gaussian documentation gives 3 citations for the first row transition metal parameters of 6-311G: http://www.gaussian.com/g_tech/g_ur/m_basis_sets.htm I would guess that nobody has integrated the additional parameters into 6-311G for submission to the BSE. Or perhaps there has been a combined submission that was rejected on the grounds that the parameters for first row transition metals came from work that predated the 6-311G basis set. The earliest citation regarding 6-311G in the BSE notes is from 1980 while the Wachters parameters for first row transition metals, ultimately incorporated into Gaussian's version of 6-311G, were published in 1970.

Forum >> BSE: Basis Set Exchange >> General BSE Topics