[molpro-user] Bug in spin-orbit calculation
Alexander Mitrushchenkov
Alexander.Mitrushchenkov at u-pem.fr
Sat Nov 28 10:34:37 CET 2015
Hi,
The error seem to be numerical, probably due to small symmetry problem
in orbitals/integrals; but might be a bug as well. Before looking in
detail, could you please try the same calculation with much tighter
thresholds for integrals, orbitals, and CI coefficients?
Best,
Alexander
Le 26/11/2015 02:49, ZOUWL a écrit :
> Dear developers,
>
> There may be a bug in the following spin-orbit calculation. The
> low-lying 1D, 3P, 3F, and 5F states of Ta+ are calculated at the
> CASSCF(4,6) level, where the 5d6s orbitals are active. Group symmery
> used here is Ci, but the results and errors are the same if D2h is
> used. All the calculations are performed using the 2010 version. Input:
>
> ***,Ta+, SOC with ECP
> memory,100,m
>
> symmetry,xyz
>
> geometry={Ta}
>
> basis=vtz-pp
>
> {multi;natorb,ci;
> frozen,0;closed,1,3;occ,7,3;
> wf,12,1,0;state,5; ! 1D
> wf,12,1,2;state,10; ! 3P, 3F
> wf,12,1,4;state,7; ! 5F
> }
>
> {ci;noexc;wf,12,1,0;state,5;
> save,4201.3;}
> {ci;noexc;wf,12,1,2;state,10;
> save,4221.3;}
> {ci;noexc;wf,12,1,4;state,7;
> save,4241.3;}
>
> ! SOC: singlet and triplet states
> {ci;hlsmat,ecpls,4201.3,4221.3;}
> ! SOC: triplet and quintet states
> {ci;hlsmat,ecpls,4221.3,4241.3;}
> ! SOC: singlet and quintet states
> {ci;hlsmat,ecpls,4201.3,4241.3;}
> ---
>
> The CASSCF energies of L-S states are degenerate (normal):
> 1D: -56.47502225 (x 5)
> 3P: -56.49919112 (x 3)
> 3F: -56.48533379 (x 7)
> 5F: -56.52492675 (x 7)
>
> In the first (singlet + triplet) and the third (singlet + quintet) SOC
> calculations, the energies of atomic J terms are degenerate (normal):
>
> Nr Sym E E-E0 E-E0 E-E(1)
> E-E(1) E-E(1)
> (au) (au) (cm-1) (au)
> (cm-1) (eV)
> 1 1 -56.51000193 -0.01081081 -2372.70 0.00000000
> 0.00 0.0000
> 2 1 -56.50459652 -0.00540540 -1186.35 0.00540540
> 1186.35 0.1471
> 3 1 -56.50459652 -0.00540540 -1186.35 0.00540540
> 1186.35 0.1471
> 4 1 -56.50459652 -0.00540540 -1186.35 0.00540540
> 1186.35 0.1471
> 5 1 -56.50158234 -0.00239123 -524.81 0.00841958
> 1847.88 0.2291
> 6 1 -56.50158234 -0.00239123 -524.81 0.00841958
> 1847.88 0.2291
> 7 1 -56.50158234 -0.00239123 -524.81 0.00841958
> 1847.88 0.2291
> 8 1 -56.50158234 -0.00239123 -524.81 0.00841958
> 1847.88 0.2291
> 9 1 -56.50158234 -0.00239123 -524.81 0.00841958
> 1847.88 0.2291
> 10 1 -56.49671921 0.00247191 542.52 0.01328272
> 2915.22 0.3614
> 11 1 -56.49671921 0.00247191 542.52 0.01328272
> 2915.22 0.3614
> 12 1 -56.49671921 0.00247191 542.52 0.01328272
> 2915.22 0.3614
> 13 1 -56.49671921 0.00247191 542.52 0.01328272
> 2915.22 0.3614
> 14 1 -56.49671921 0.00247191 542.52 0.01328272
> 2915.22 0.3614
> ...
> Nr E E-E0 E-E0 E-E(1) E-E(1)
> E-E(1)
> (au) (au) (cm-1) (au) (cm-1)
> (eV)
> 1 -56.54141852 -0.01649177 -3619.53 0.00000000 0.00
> 0.0000
> 2 -56.54141852 -0.01649177 -3619.53 0.00000000 0.00
> 0.0000
> 3 -56.54141852 -0.01649177 -3619.53 0.00000000 0.00
> 0.0000
> 4 -56.53729558 -0.01236883 -2714.64 0.00412294 904.88
> 0.1122
> 5 -56.53729558 -0.01236883 -2714.64 0.00412294 904.88
> 0.1122
> 6 -56.53729558 -0.01236883 -2714.64 0.00412294 904.88
> 0.1122
> 7 -56.53729558 -0.01236883 -2714.64 0.00412294 904.88
> 0.1122
> 8 -56.53729558 -0.01236883 -2714.64 0.00412294 904.88
> 0.1122
> 9 -56.53111117 -0.00618441 -1357.32 0.01030736 2262.20
> 0.2805
> 10 -56.53111117 -0.00618441 -1357.32 0.01030736 2262.20
> 0.2805
> 11 -56.53111117 -0.00618441 -1357.32 0.01030736 2262.20
> 0.2805
> 12 -56.53111117 -0.00618441 -1357.32 0.01030736 2262.20
> 0.2805
> 13 -56.53111117 -0.00618441 -1357.32 0.01030736 2262.20
> 0.2805
> 14 -56.53111117 -0.00618441 -1357.32 0.01030736 2262.20
> 0.2805
> 15 -56.53111117 -0.00618441 -1357.32 0.01030736 2262.20
> 0.2805
> ...
>
> However, in the second SOC calculation (triplet + quintet), spherical
> degeneracy (degeneracy = J * 2 + 1) is broken because of unknown reasons:
>
> Nr E E-E0 E-E0 E-E(1) E-E(1)
> E-E(1)
> (au) (au) (cm-1) (au) (cm-1)
> (eV)
> 1 -56.54145102 -0.01652427 -3626.66 0.00000000 0.00
> 0.0000
> 2 -56.54145075 -0.01652400 -3626.60 0.00000027 0.06
> 0.0000
> 3 -56.54144331 -0.01651655 -3624.96 0.00000771 1.69
> 0.0002
> 4 -56.53734340 -0.01241665 -2725.14 0.00410762 901.52
> 0.1118
> 5 -56.53734093 -0.01241418 -2724.60 0.00411009 902.06
> 0.1118
> 6 -56.53731961 -0.01239286 -2719.92 0.00413141 906.74
> 0.1124
> 7 -56.53730864 -0.01238189 -2717.51 0.00414238 909.15
> 0.1127
> 8 -56.53730809 -0.01238134 -2717.39 0.00414293 909.27
> 0.1127
> 9 -56.53118445 -0.00625770 -1373.41 0.01026657 2253.25
> 0.2794
> 10 -56.53117231 -0.00624555 -1370.74 0.01027871 2255.92
> 0.2797
> 11 -56.53116127 -0.00623451 -1368.32 0.01028975 2258.34
> 0.2800
> 12 -56.53114109 -0.00621434 -1363.89 0.01030993 2262.77
> 0.2805
> 13 -56.53113027 -0.00620352 -1361.52 0.01032075 2265.14
> 0.2808
> 14 -56.53112879 -0.00620203 -1361.19 0.01032223 2265.47
> 0.2809
> 15 -56.53112337 -0.00619661 -1360.00 0.01032765 2266.66
> 0.2810
> ...
>
> There is no such a problem in the following all-electron calculation,
> so the error may be related to the ECPLS option.
>
> ***,Ta+, SOC with A.E.
> memory,100,m
>
> symmetry,xyz
>
> geometry={Ta}
>
> basis=vtz-dk
>
> set dkroll=1
>
> {hf;occ,21,19;wf,80,1,0}
>
> {multi;natorb,ci;
> ! 4f5s5p are inactive
> frozen,14,9;closed,15,19;occ,21,19;
> wf,72,1,0;state,5; ! 1D
> wf,72,1,2;state,10; ! 3P, 3F
> wf,72,1,4;state,7; ! 5F
> }
>
> {ci;noexc;wf,72,1,0;state,5;
> save,4201.3;}
> {ci;noexc;wf,72,1,2;state,10;
> save,4221.3;}
> {ci;noexc;wf,72,1,4;state,7;
> save,4241.3;}
>
> lsint
> {ci;hlsmat,ls,4201.3,4221.3;}
> {ci;hlsmat,ls,4221.3,4241.3;}
> {ci;hlsmat,ls,4201.3,4241.3;}
> ---
>
> Best wishes,
> Wenli
>
>
>
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user
--
Dr. Alexander Mitrushchenkov, IGR
Laboratoire de Modélisation et Simulation Multi Echelle
UMR 8208 CNRS
Université Paris-Est Marne-la-Vallée
5 Bd Descartes
77454 Marne la Vallée, Cedex 2, France
Phone: +33(0)160957316
Fax: +33(0)160957320
e-mail: Alexander.Mitrushchenkov at u-pem.fr
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20151128/c11ea666/attachment-0001.html>
More information about the Molpro-user
mailing list