[molpro-user] Regarding usage of effective core potentials for fluorine atom

Kirk Peterson kipeters at wsu.edu
Tue Jan 4 16:58:58 GMT 2011


Aleem,

When you use a ECP, core electrons are being replaced (that's why the total energies change so much, as Gerald has pointed out). Thus you have to drop these from your wf, occ, closed, con, etc directives. In your case you have accidentally calculated props of the doubly negative ions. 

Regards.

Kirk

Sent from my iPhone

On Jan 3, 2011, at 11:36 PM, syed aleem <aleem3176 at gmail.com> wrote:

> Dear Kirk,
>      
> Here am providing my complete input for both calculations.
> 
> Input with ECP
> 
> ***,ionization potential
> memory,100,m
> basis={h=aug-cc-pvtz,
>        c=aug-cc-pvtz,
>        f=aug-cc-pvqz
>        ECP,F,2,3,0;
>        1; 2,1.000000,0.000000;
>        1; 2,16.762800,96.241900;
>        1; 2,19.857600,-4.581900;
>        1; 2,0.031600,-0.001500;}
> geomtyp=xyz
> geometry={5
> equilibrium geometry
> c     0.000000    0.000000   -0.636134
> h     0.000000    1.029894   -0.984650
> h    -0.891914   -0.514947   -0.984650
> h     0.891914   -0.514947   -0.984650
> f     0.000000    0.000000    0.752306
> end}
>  rhf;occ,7,2;wf,18,1,0;orbital,2130.2;orbprint,5;
>  multi;occ,9,3;closed,4,0;start,2130.2;wf,17,1,1;state,2;                        
>  mrci;option,maxit=100;occ,9,3;closed,4,0;wf,17,1,1;state,2;select;con,2,2,1,0,0,2,2,0;con,2,1,2,0,0,2,2,0;
>  exit;
> 
> Input without ECP
> ***,ionization potential
> memory,100,m
> basis={h=aug-cc-pvtz,
>        c=aug-cc-pvtz,
>        f=aug-cc-pvqz;}
>       !ECP,F,2,3,0;
>       !1; 2,1.000000,0.000000;
>       !1; 2,16.762800,96.241900;
>       !1; 2,19.857600,-4.581900;
>       !1; 2,0.031600,-0.001500;}
> geomtyp=xyz
> geometry={5
> equilibrium geometry
> c     0.000000    0.000000   -0.636134
> h     0.000000    1.029894   -0.984650
> h    -0.891914   -0.514947   -0.984650
> h     0.891914   -0.514947   -0.984650
> f     0.000000    0.000000    0.752306
> end}
> rhf;occ,7,2;wf,18,1,0;orbital,2130.2;orbprint,5;
> multi;occ,9,3;closed,4,0;start,2130.2;wf,17,1,1;state,2;   
> mrci;option,maxit=100;occ,9,3;closed,4,0;wf,17,1,1;state,2;select;con,2,2,1,0,0,2,2,0;con,2,1,2,0,0,2,2,0;
> exit;
> 
> Thanks for your reply,
> 
> Regards,
> Aleem.
> 
> 
> On 4 January 2011 12:26, Kirk Peterson <kipeters at wsu.edu> wrote:
> Aleem,
> 
> It's difficult to see where the problem lies without the full inputs. Your ECP input looks ok. Note that your F basis should probably be used uncontracted since it's contraction coeffs are optimized for all-electron calculations. 
> 
> -Kirk
> 
> Sent from my iPhone
> 
> On Jan 3, 2011, at 9:54 PM, syed aleem <aleem3176 at gmail.com> wrote:
> 
>> Dear Users,
>> 
>> I am trying to use relativistic effective core potential for core electrons of Fluorine atom in ch3f molecule
>> 
>> My input according to manual is
>> 
>> ***,ionization potential
>> memory,100,m
>> basis={h=aug-cc-pvtz,
>>        c=aug-cc-pvtz,
>>        f=aug-cc-pvqz
>>        ECP,F,2,3,0;
>>        1; 2,1.000000,0.000000;
>>        1; 2,16.762800,96.241900;
>>        1; 2,19.857600,-4.581900;
>>        1; 2,0.031600,-0.001500;}
>> geomtyp=xyz
>> geometry={5
>> geometry parameters,wave function for multi and mrci calculation were followed
>> 
>> But the problem is when I submitted this job. state energies were drastically changed. I am providing both outputs (one with ecp and one without)
>> 
>> With ECP
>> 
>>      MRCI            MRCI            MULTI           MULTI           RHF-SCF
>>     -64.08157998    -64.10092216    -63.67925159    -63.69710084    -63.49847009
>> 
>> With out ECP
>> 
>>                MRCI            MRCI            MULTI           MULTI           RHF-SCF
>>    -138.93548516   -139.06809598   -138.57347287   -138.70164845   -139.10421155
>> 
>> Could you please check whether there was any mistake in my input. I would be grateful if you explain how to deal with  ECP (effective core potentials in molpro calculations)
>> 
>> Thanking you in advance for your help
>> 
>> Regards,
>> 
>> Aleem,
>> School of Chemistry,
>> University of Hyderabad,
>> INDIA.
>> 
>> _______________________________________________
>> Molpro-user mailing list
>> Molpro-user at molpro.net
>> http://www.molpro.net/mailman/listinfo/molpro-user
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20110104/b2015c8e/attachment.html>


More information about the Molpro-user mailing list