BSSE error


  • Guest -
7:14:29 AM PDT - Thu, Apr 21st 2011
.

  • Guest -
7:30:37 AM PDT - Thu, Apr 21st 2011
Hi...

I try to test a simple calculation of Na+ and H2O interaction energy with c-dft and bsse correction.
But it keeps giving an error message in BSSE stage like "gridrtunc: no basis on ghost atom? 2".
I posted my input and corresponding output part.
1) What shoudl I chage or miss in my calculation?
2) What is the role of ghost atoms in BSSE?
thanks in advance.

Contents

>Input<

echo

start

scratch_dir ./scratch
permanent_dir ./scratch

    1. molecular system ##
charge 1
geometry noautoz
symmetry
Na 0. 0. 0.5
O 0. 0. 2.
H 0. 1. 3.
H 0. -1. 3.
end

    1. basis sets ##
basis SPHERICAL
Na library 3-21g
O library 3-21g
H library 3-21g
BqNa library Na 3-21g
BqO library O 3-21g
BqH library H 3-21g
end

    1. dft options ##
dft
xc b3lyp
convergence nolevelshifting
odft
mult 2
iterations 500
tolerances tight
cdft 1 1 charge 1.0 pop
decomp
end

driver
maxiter 200
end

task dft optimize

bsse
mon metal 1
charge 1
mon water 2 3 4
end
task dft optimize

>End of Input<


>Error in output<

                                NWChem DFT Module

                                -----------------




 Caching 1-el integrals 

------------------------------------------------------------------------

gridrtunc: no basis on ghost atom?        2

------------------------------------------------------------------------

------------------------------------------------------------------------

 current input line : 

   54: task dft optimize

------------------------------------------------------------------------

------------------------------------------------------------------------

There is an error in the specified basis set

------------------------------------------------------------------------

For more information see the NWChem manual at http://nwchemgit.github.io/index.php/NWChem_Documentation




For further details see manual section:
Last System Error Message from Task 0:: Inappropriate ioctl for device
0:0:gridrtunc: no basis on ghost atom?:: 2
(rank:0 hostname:rcc-pok-idg-2088.cloud9.ibm.com pid:6983):ARMCI DASSERT fail. armci.c:ARMCI_Error():260 cond:0
 0: ARMCI aborting 0 (0).

 0: ARMCI aborting 0 (0).
system error message: Invalid argument
 atom                       2

 tag    bqO      

coord      0.000002   -0.004400    1.988395

 charge    0.00000000000000     

 ic1 ic2                     0                   -1

=End of output


  • Guest -
9:39:25 AM PDT - Fri, Apr 22nd 2011
I think there are some small bug in ghostname specification.
Try it with "bqX" using small character instead of "Bq" of chapital letters.

  • Guest -
12:11:50 PM PDT - Wed, May 4th 2011
function of Bq
You asked about the function of Bq atoms in a BSSE calculation. Basically, in order to do a Counterpoise correction you need to define two "pieces" of the molecule. NWChem chooses to do this by defining "normal" atoms and "ghost (bq)" atoms. Gaussian, for example, defines "layer 1" and "layer 2".

Forum Vet
3:08:57 PM PDT - Fri, May 6th 2011
Yes, there is a small bug that forces the BSSE module to only recognize lower case bq and not Bq. This will be fixed in the next release.

Thanks,

Bert


Quote: Apr 22nd 4:39 pm
I think there are some small bug in ghostname specification.
Try it with "bqX" using small character instead of "Bq" of chapital letters.

Forum Vet
3:27:12 PM PDT - Fri, May 6th 2011
Also, BSSE will not work with constrained DFT as it is not set up to handle constrained charges on a ghost atom (i.e. a bqNa with a forced +1 charge does not work).

Bert


Quote:Bert May 6th 10:08 pm
Yes, there is a small bug that forces the BSSE module to only recognize lower case bq and not Bq. This will be fixed in the next release.

Thanks,

Bert


Quote: Apr 22nd 4:39 pm
I think there are some small bug in ghostname specification.
Try it with "bqX" using small character instead of "Bq" of chapital letters.

  • Guest -
1:49:03 PM PDT - Thu, Oct 6th 2011
error
dft optimize failed 0 ------------------------------------------------------------------------ ------------------------------------------------------------------------
 current input line :

   37:  task dft optimize

------------------------------------------------------------------------

------------------------------------------------------------------------

This type of error is most commonly associatated with calculations not reaching convergence criteria

------------------------------------------------------------------------

For more information see the NWChem manual at http://www.emsl.pnl.gov/docs/nwchem/nwchem.html




For further details see manual section:            [google][yahoo]google[/yahoo][/google]                                                                              

                                           [u][s]underline[/s][quote=username]quote[/quote][/u]

Forum Vet
4:54:06 PM PDT - Thu, Oct 6th 2011
Look up a little earlier in the output file. Many possibilities here:

1. You reached the maximum number of geometry optimization steps. Solution, increase number of steps.

2. You did not converge the DFT energy calculation. Solution, increase number of steps.

3. You reached the time limit of your job and the code noticed and stopped.

Which case applies to you can be found in the output file, but not in this last line.

Bert


Quote: Oct 6th 8:49 pm
dft optimize failed 0 ------------------------------------------------------------------------ ------------------------------------------------------------------------
 current input line :

   37:  task dft optimize

------------------------------------------------------------------------

------------------------------------------------------------------------

This type of error is most commonly associatated with calculations not reaching convergence criteria

------------------------------------------------------------------------

For more information see the NWChem manual at http://www.emsl.pnl.gov/docs/nwchem/nwchem.html




For further details see manual section:            [google][yahoo]google[/yahoo][/google]                                                                              

                                           [u][s]underline[/s][quote=username]quote[/quote][/u][/quote]

Gets Around
9:43:50 AM PST - Mon, Nov 21st 2011
Hi,

could someone also implement the automatic assignment of basis set to ghost atoms?
I mean i would like to be able to write:

geometry noautosym
O 8.0 8.0 9.245956
bqO 8.0 8.0 8.0
end
basis spherical
* library aug-cc-pvdz
end

Currently this fails with:

gridrtunc: no basis on ghost atom? 2

so i have to specify the basis for bqO explicitly.

Marcin


Forum >> NWChem's corner >> NWChem functionality