Re: Please specify a default basis or define basis sets for all atoms!

[qia...@theochem.uni-stuttgart.de]

There is a known "MINAO-PP" issue in the IBO localization program of Molpro for atoms Si--Ni when ECPs is used. As far as I know correlation consistent AO basis sets to be used with ECPs are not available for these atoms, and the corresponding minimal basis sets (MINAO-PP) required for IBO orbital localization were not developed.

You may try using a different orbital localization method in PNO calculations (e.g., Pipek-Mezey with "loc_method=pm"). Do check the localized orbital composition in the output to see whether the MOs are reasonably localized though: if that is not the case, the PNO-LCCSD calculations are not expected to be accurate.

On Friday, July 16, 2021 at 8:35:43 AM UTC+2 benjam...@gmail.com wrote:

Dear Molpro forum,

In a calculation, I have defined the basis set for all atoms, as seen in the input file below for PNO-LCCSD(T) calculations but when I submit the calculation, it says that it 'cannot find default basis MINAO-PP for atom Ti' and 'Please specify a default basis or define basis sets for all atoms!' What do you think I have done wrong in the input file?

Best wishes and many thanks for the help,

Benjamin

***,molpro

!memory,700,m

file,2,molpro-dft.wfu,new

orient,noorient

!gdirect

symmetry,nosym

geometry=input.geometry

set,charge=606

{lattice,input.lattice}

basis={

Ti1=def2-tzvpp;

O1=def2-tzvpp;

O2=def2-tzvpp;

ECP,Ti2,18,3;

1; 2,1.000000,0.000000;

2; 2,0.860000,9.191690; 2,0.172000,0.008301;

2; 2,0.860000,17.997720; 2,0.172000,-0.032600;

2; 2,1.600000,-9.504310; 2,0.320000,-0.151370;

s, Ti2, 654125.31, 97995.933, 22302.012, 6315.5811, 2059.3549, 742.55919, 288.60525, 118.65832, 50.372387, 19.281107, 9.0193318, 4.1177245, 1.3063361, 0.53009179, 0.08139871, 0.03171529, 0.01268612

c, 1.17,-0.0000015,-0.0000114,-0.0000591,-0.0002519,-0.0008839,-0.0028046,-0.0071352,-0.0155078,-0.0175351,0.00922554,0.09836177,0.08328293,-0.1980669,-0.3316394,0.59524717,0.44878795,0.12478378

set,jkfit,context=jkfit;

Ti1=def2-tzvpp/jkfit;

O1=def2-tzvpp/jkfit;

O2=def2-tzvpp/jkfit;

s, Ti2, 654125.31, 97995.933, 22302.012, 6315.5811, 2059.3549, 742.55919, 288.60525, 118.65832, 50.372387, 19.281107, 9.0193318, 4.1177245, 1.3063361, 0.53009179, 0.08139871, 0.03171529, 0.01268612

c, 1.17,-0.0000015,-0.0000114,-0.0000591,-0.0002519,-0.0008839,-0.0028046,-0.0071352,-0.0155078,-0.0175351,0.00922554,0.09836177,0.08328293,-0.1980669,-0.3316394,0.59524717,0.44878795,0.12478378

set,mp2fit,context=mp2fit

Ti1=def2-tzvpp/mp2fit;

O1=def2-tzvpp/mp2fit;

O2=def2-tzvpp/mp2fit;

s, Ti2, 654125.31, 97995.933, 22302.012, 6315.5811, 2059.3549, 742.55919, 288.60525, 118.65832, 50.372387, 19.281107, 9.0193318, 4.1177245, 1.3063361, 0.53009179, 0.08139871, 0.03171529, 0.01268612

c, 1.17,-0.0000015,-0.0000114,-0.0000591,-0.0002519,-0.0008839,-0.0028046,-0.0071352,-0.0155078,-0.0175351,0.00922554,0.09836177,0.08328293,-0.1980669,-0.3316394,0.59524717,0.44878795,0.12478378

}

{df-rhf,df_basis=jkfit,maxit=1000;shift,-1.0,-0.5}

{pno-lccsd(t),basis_mp2=mp2fit

CORE,30}

[Hans-Joachim Werner]

There are MINAO basis sets for all atoms, but here you are looking for a MINAO-PP set (minimal basis for pseudo-potential) for Ti which is not available since there are no default PP basis sets for Ti. The total charge of 606 in your input indicates that you seem to do something special, perhaps use ECPs as environment? The MINAO sets are needed for IBO localization, but I don’t know if this could ever work for such a case. You could try to use Pipek-Mezey or Boys localization, these do not require MINAO. MINAO is also needed in the orbital guess program. Perhaps you can get start orbitals without these ECPs. Or you have to define a MINAO-PP basis for Ti; this should simply be a generally contracted minimal basis set, containing only the valence s,p,d functions.
Best wishes
Joachim

> --
> You received this message because you are subscribed to the Google Groups "molpro-user" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to molpro-user...@googlegroups.com.
> To view this discussion on the web, visit https://groups.google.com/d/msgid/molpro-user/52891241-946b-41d5-a1ef-3a2a93bad354n%40googlegroups.com.

[Benjamin Shi]

Hi Joachim,

Thanks for the response! And apologies for the late response - only just saw it now as I had deleted this message since I found a solution (with thanks to Daniel Kats) that worked before anyone responded. These calculations were for embedded cluster calculations which I was trying to set up, hence the large total charge to use the ECPs as an embedding environment. The solution that worked most easily for me was to define an AO basis with ao=  when calling pno-lccsd(t), such as below.

Best wishes,

Benjamin

***,molpro

!memory,700,m

file,2,molpro-dft.wfu,new

orient,noorient

set,aomin

Ti1=MINAO;

O1=MINAO;

O2=MINAO;

s, Ti2, 654125.31, 97995.933, 22302.012, 6315.5811, 2059.3549, 742.55919, 288.60525, 118.65832, 50.372387, 19.281107, 9.0193318, 4.1177245, 1.3063361, 0.53009179, 0.08139871, 0.03171529, 0.01268612

c, 1.17,-0.0000015,-0.0000114,-0.0000591,-0.0002519,-0.0008839,-0.0028046,-0.0071352,-0.0155078,-0.0175351,0.00922554,0.09836177,0.08328293,-0.1980669,-0.3316394,0.59524717,0.44878795,0.12478378

}

{df-rhf,df_basis=jkfit,maxit=1000;shift,-1.0,-0.5}

{pno-lccsd(t),ao=aomin,basis_mp2=mp2fit

CORE,30}