[molpro-user] optg and symmetry

Hans-Joachim Werner werner at theochem.uni-stuttgart.de
Wed Nov 4 11:46:20 GMT 2009


The problem is that you try to optimize with a symmetric structure,  
while
the minimum is less symmetric. With other words, you are looking for a
saddle point. Numerical deviations from the symmetric structure can then
lead to the symmetry breaking error you have encountered. If the  
symmetric
structure is really what you want, then can can use

{optg;inactive,d1,d2}

and it will converge. The result is
  E(MULTI000) / Hartree    -721.44795033  -721.44795072     0.00000000
  FS / ANGSTROM               1.55737684     1.55730460      
1.55728459     0.00003211     3.16565802
  SO1 / ANGSTROM              1.43718585     1.43697248      
1.43697882    -0.00001987     2.76653608
  OSO1 / DEGREE             104.44542861   104.46248545    
104.46257370     0.00000048     0.00020110
  SO2 / ANGSTROM              1.47416992     1.47392292      
1.47384721     0.00015022     2.84645040
  OSO2 / DEGREE             104.20339288   104.18121886    
104.17590854     0.00000114     0.00018690
  SO3 / ANGSTROM              1.47416992     1.47392292      
1.47384721     0.00015022     2.84645040
  OSO3 / DEGREE             104.20339288   104.18121886    
104.17590854     0.00000114     0.00018690
  Convergence:                0.00000000  (line search)      
0.00024431     0.00014935  (total)

If you are looking for the minimum, you should start with a slightly  
displaced geometry.
This yields after optimization (in C1 symmetry)
  Variable                      Last           Current         
Next           Gradient       Hessian

  E(MULTI000) / Hartree    -721.47056745  -721.47056820     0.00000000
  FS / ANGSTROM               1.56554151     1.56560614      
1.56561237     0.00001544     3.24796822
  SO1 / ANGSTROM              1.59686592     1.59709395      
1.59713240    -0.00003870     2.17830248
  OSO1 / DEGREE              98.23051197    98.22898085     
98.21213578     0.00000330     0.00019792
  SO2 / ANGSTROM              1.42700826     1.42714076      
1.42721765    -0.00019504     3.93132875
  OSO2 / DEGREE             107.78577103   107.85573425    
107.87763529    -0.00000190     0.00016594
  D1 / DEGREE               111.71389954   111.69366244    
111.69255063     0.00000119     0.00010771
  SO3 / ANGSTROM              1.41826777     1.41838093      
1.41849831    -0.00036016     3.81880552
  OSO3 / DEGREE             106.47258790   106.43860815    
106.43073999    -0.00000018     0.00021831
  D2 / DEGREE               137.60538845   137.58281179    
137.56488041     0.00000331     0.00017498
  Convergence:                0.00000000  (line search)      
0.00065227     0.00036855  (total)

Joachim Werner
Am 04.11.2009 um 11:47 schrieb Luis Poveda:

> Dear Tereza,
> May be you have a symmetry-breaking error. OPTG move the starting  
> geommetry along the energy gradient. If you specify symmetry to  
> start and the atoms begin to move, the new set of integrals have to  
> be calculated at a lower symmetry and the program can't figure out  
> what to do. use "nosym" in geometry card, if the optimization  
> procedure work well you will get a nearly symmetric structure...
>
>
> On Mon, Nov 2, 2009 at 2:39 PM, tereza sedivcova <tereza.sedivcova at gmail.com 
> > wrote:
> Dear Tim Lillestolen and others MOLPRO users,
> did you solve your problem?
> Because I have the same now, and I can not see a bug (bugs) in my Z- 
> matrix.
>
> I am interested in FSO3 radical in its C3v symmery, however even if  
> I tryed  it
> for Cs or C1 symmetry I obtain the same error.
> The inputs for all three geometries are below.
>
> Thank you very very much for your help,
> Tereza
> **************************************************************
> Laboratory of High Resolution Molecular Spectroscopy
> Department of Analytical Chemistry
> Institute of Chemical Technology, Prague
> **************************************************************
>
> input:
> ***C3v symmetry point group
>
>          !
>  memory,500,m
>  gprint,orbital,civector
>  geometry={
>  Angstrom;
>  f;
>  s,f,fs;
>  o1,s,so,f,oso;
>  o2,s,so,f,oso,o1,D;
>  o3,s,so,f,oso,o2,D;
>  }
>
>  fs   =        1.542;
>  so   =        1.443;
>  oso  =      105.404;
>  D    =      120.000;
>
>  basis=vdz;
>  int;
>  {uhf; occ, 17,8; wf,49,2,1;}
>
>  {multi; occ, 19,9; closed,15,6; wf,49,2,1;state,1;
>  maxiter,100;}
>
>  {optg;}
>  forces;
>
>  put, molden, 1.molden;
>
>  &&&&&&&&&&&&&&&&&&&&&&&&&&&&
> ***Cs
> memory,500,m
>  gprint,orbital,civector
>  geometry={
>  Angstrom;
>  s;
>  f,s,fs;
>  o1,s,so1,f,oso1;
>  o2,s,so1,f,oso1,o1,D1;
>  o3,s,so2,f,oso2,o2,D2;
>  }
>
>  fs   =        1.542;
>  so1   =        1.443;
>  so2   =        1.443;
>  oso1  =      105.404;
>  oso2  =      105.404;
>  D1    =      120.000;
>  D2    =      120.000;
>
>  basis=vdz;
>  int;
>  {uhf; occ, 17,8; wf,49,2,1;}
>
>  {multi; occ, 19,9; closed,15,6; wf,49,2,1;state,1;
>  maxiter,100;}
>
>  {optg;}
>  forces;
>
>  put, molden, 1.molden;
>
> &&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&
> ***,C1
>
> memory,500,m
> gprint,orbital,civector
> geometry={
> Angstrom;
>  s;
>  f,s,fs;
>  o1,s,so1,f,oso1;
>  o2,s,so2,f,oso2,o1,D1;
>  o3,s,so3,f,oso3,o2,D2;
> }
>
> fs   =        1.542;
> so1   =        1.443;
> so2   =        1.443;
> so3   =        1.443;
> oso1  =      105.404;
> oso2  =      105.404;
> oso3  =      105.404;
> D1    =      120.000;
> D2    =      120.000;
>
> basis=vdz;
> int;
> {uhf; occ, 17,8; wf,49,2,1;}
>
> {multi; occ, 19,9; closed,15,6; wf,49,2,1;state,1;
> maxiter,100;}
>
> {optg;}
> forces;
>
> put, molden, 1.molden;
>
>
>
>
>
> On Wed, Nov 5, 2008 at 12:24 PM, lillestolen
> <timothy.lillestolen at nottingham.ac.uk> wrote:
> > Dear Molpro users,
> >
> > I feel like I'm probably missing something simple here, but I still
> > can't figure out what I am doing wrong.  I am trying to optimise the
> > geometry of carbonic acid (H2CO3) with C2V symmetry at the
> > B3LYP/aug-cc-pVDZ level of theory.  I am including the  input file  
> at
> > the end of the message.  The first energy evaluation works just  
> fine,
> > and the program detects the correct C2V symmetry.  However, after  
> the
> > first displacements are calculated I get the following error:
> >
> >
> > Optimization point 1
> >
> >  Variable                      Last           Current
> > Next           Gradient       Hessian
> >
> >  E(KS-SCF000) / Hartree      0.00000000  -264.87588720      
> 0.00000000
> >  OC1 / ANGSTROM              0.00000000     1.22200000      
> 1.28411786
> > -0.25660683     4.00134221
> >  OC2 / ANGSTROM              0.00000000     1.22200000      
> 1.25951398
> > -0.32956964     8.18713291
> >  OCO / DEGREE                0.00000000   120.00000000
> > 119.07590136     0.00134251     0.00164275
> >  DIH4 / DEGREE               0.00000000   180.00000000
> > 180.00000000     0.00000000     0.00000000
> >  HO / ANGSTROM               0.00000000     0.88000000      
> 0.94600782
> > -0.41770827     6.28691438
> >  HOC / DEGREE                0.00000000   120.00000000
> > 117.35334871     0.00102677     0.00046113
> >  DIH5 / DEGREE               0.00000000     0.00000000
> > 0.00000000     0.00000000     0.00000000
> >  DIH6 / DEGREE               0.00000000   180.00000000
> > 180.00000000     0.00000000     0.00000000
> >  Convergence:                0.00000000  (line search)
> > 0.19172405     0.32724871  (total)
> >  The request for symmetry elements X,Y, cannot be honoured
> >  ? Error
> >  ? Symmetry problem
> >  ? The problem occurs in zmatrix.f:zmat_orient
> >
> >  ERROR EXIT
> >  CURRENT STACK:      MAIN
> >
> >
> >   
> **********************************************************************************************************************************
> >  The request for symmetry elements X,Y, cannot be honoured
> >
> >  Prototype coordinates
> >                   1             2             3
> > 4             5             6
> >         1     0.0000000     0.0000000     0.0000000     0.0000000
> > 0.0000000     0.0000000
> >         2    -0.0240505     1.2534949     1.1355597    -2.4015615
> > 0.1491318     0.3359890
> >         3    -0.0252998     2.0378084    -2.1038457     0.0864578
> > -3.5947566     3.5720959
> >  ? Error
> >  ? Symmetry problem
> >  ? The problem occurs in zmatrix.f:zmat_orient
> >
> >  RECURSIVE CALL OF FEHLER. EMERGENCY STOP.
> >  FILE INFORMATION MAY BE LOST
> >
> >  RECURSIVE CALL OF ERROR HANDLER, FILE INFORMATION MAY BE LOST
> >
> >  EMERGENCY STOP
> >
> >
> > I get the same message if I try and do the optimisation with hartree
> > fock (replacing ks,b3lyp with hf).  Am I doing something incorrectly
> > with my z-matrix?  Or is there another option that I am missing  
> out on?
> > I really need to get the final geometry with the correct symmetry  
> as I
> > am calculating some atomic properties using these geometries and I  
> need
> > symmetric atoms.  Much thanks in advance for any advice/help,
> >
> > best regards,
> >
> > Tim Lillestolen
> > Wheatley Group
> > University of Nottingham
> >
> >
> > input file:
> >
> > ***,h2co3
> >
> > basis=avdz
> >
> > geometry={
> > angstrom;
> >  c;
> >  o1, c,oc1;
> >  o2, c,oc2,o1,oco;
> >  o3, c,oc2,o1,oco,o2,dih4;
> >  h1,o2,ho,c ,hoc,o3,dih5;
> >  h2,o1,ho,c ,hoc,o2,dih6;
> > }
> >
> > oc1  =      1.222;
> > oc2  =      1.222;
> > ho =       0.88;
> >
> >
> > oco=       120.000;
> > hoc=       120.000;
> >
> > dih4=      180.000;
> > dih5=         0.000;
> > dih6=       180.000;
> >
> > ks,b3lyp;
> > optg,symmetry=auto,coord=zmat;
> >
> >
> > This message has been checked for viruses but the contents of an  
> attachment
> > may still contain software viruses, which could damage your  
> computer system:
> > you are advised to perform your own checks. Email communications  
> with the
> > University of Nottingham may be monitored as permitted by UK  
> legislation.
> >
> > _______________________________________________
> > Molpro-user mailing list
> > Molpro-user at molpro.net
> > http://www.molpro.net/mailman/listinfo/molpro-user
> >
> _______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user
>
> _______________________________________________
> 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/20091104/81a0ee0d/attachment.html>


More information about the Molpro-user mailing list