about single point calculation with larger basis set from smaller BS opt

[ardie]

Hi,
Does anybody know how to do a single point calculation using larger basis set by taking the initial guess and structure from the previous opimization results. I kept the alpha, beta, coord and control files from the opimizaition results with smaller basis set (B1) to do a larger basis set (B2) single point calculation. But error message was given and the job stopped. Can anyone give me some suggestion? Thank you !

The error:

              +--------------------------------------------------+
              |               basis set information              |
              +--------------------------------------------------+


              we will work with the  1s 3p 5d  7f  9g ... basis set
              ...i.e. with spherical basis functions...


   type   atoms  prim   cont   basis
   ---------------------------------------------------------------------------
    fe       1     47     22   Custom   [3s3p2d|7s5p5d]
    n        5     27     14   Custom   [3s2p1d|10s4p1d]
    c       28     27     14   Custom   [3s2p1d|10s4p1d]
    h       24      7      5   Custom   [2s1p|4s1p]
    o        1     27     14   Custom   [3s2p1d|10s4p1d]
    s        1     51     18   Custom   [4s3p1d|16s10p1d]
   ---------------------------------------------------------------------------
   total:   60   1184    636
   ---------------------------------------------------------------------------

   total number of primitive shells          :   94
   total number of contracted shells         :  292
   total number of cartesian basis functions :  673
   total number of SCF-basis functions       :  636


 integral neglect threshold       :  0.50E-11
 integral storage threshold THIZE :  0.10E-04
 integral storage threshold THIME :         5

 number of unique ecp types :           1

 symmetry group of the molecule :   c1

 the group has the following generators :
   c1(z)

    1 symmetry operations found

 there are 1 real representations :   a

  396 is greater than mxsao=(ndi4*ndi19)=  395 - use higher ndi19 !


 MODTRACE: no modules on stack

 <maksao> sao/ao dimension overflow
 dscf ended abnormally

[uwe]

Hi,

first of all: please avoid multiple posts.

To reuse old MOs, one has to call define and choose use instead of eht in the molecular orbital menu. See chapter 5.12 (page 23) of the tutorial how to do this.

The only thing one has to take care of is that define is not able to transform the orbitals which are currently loaded. So you have to create a second directory to either save your old run or where you prepare the new one. The probably easiest way is to:

• go to the directory where your optimized results are
• call
  
  cpc big
  
  this will create a sub-directory called big, it copies control and all the files that are referenced within control (the file=... lines) to this new directory. cpc stands for copy control and is a standard script which should be in your PATH.
  
• cd big
• call define and hit three times <Enter> until you are asked if you want to change the basis set:
  
  

  ATOMIC ATTRIBUTE DATA (BASES,CHARGES,MASSES,ECPS) HAVE BEEN
   TAKEN FROM THE DEFAULT INPUT FILE control.
   DO YOU WANT TO CHANGE THESE DATA ? DEFAULT=n  GOBACK=&
  

  
  Say 'y' here and change the basis set you would like to use.
• Hint: if you have used RI, and if you change your basis to a larger set, let define delete the old settings for the auxiliary basis set- define will ask you if they shall be deleted or not.
• proceed to the next menu and then enter:
  
  use ../control
  
  define will read in your old data and projects your old basis set to the new one. Works if symmetry and/or basis set has changed.
• Hint part 2: for RI calculations, just enter the ri or mp2/cc2 or rijk (for RI-DFT, RI-MP2/CC2, or RI-JK resp.) menu in the last section of define to re-assign new auxiliary basis sets.

.

• restart geometry optimization or single-point energy calculation.

Regards,

Uwe

[ardie]

Hi uwe,

Thanks. I tried to do everything according to your suggestion, and now I have one more question here.

After I change the basis set in define and then enter the "OCCUPATION NUMBER & MOLECULAR ORBITAL DEFINITION MENU", I type use ../control , it gives

    sao summary :
   irrep      number of sao's referring to
                old basis    new basis
    a             389          636

    mo occupation :
   irrep   mo's   occupied
    a      636      134

 number of basis functions   :          636
 number of occupied orbitals :          134


 reading orbital data $uhfmo_alpha  from file ../alpha .

 orbital characterization : scfconv=7

  mo         457  could not be found !
  mo         459  could not be found !

 MOLECULAR ORBITAL INPUT FAILED !
 COMMAND WILL BE IGNORED !


and it return to the "OCCUPATION NUMBER & MOLECULAR ORBITAL DEFINITION MENU". Does this error mean define failed to read the previous orbital generated from optimization job?

After that, I type command use ../control again, it gives


    sao summary :
   irrep      number of sao's referring to
                old basis    new basis
    a             389          636

    mo occupation :
   irrep   mo's   occupied
    a      636      134

 number of basis functions   :          636
 number of occupied orbitals :          134


 reading orbital data $uhfmo_alpha  from file ../alpha .

 orbital characterization : scfconv=7


 **********************************************************
 *                                                        *
 *    BBBBB   L        OOOO   W    W  U    U  PPPPP       *
 *    B    B  L       O    O  W    W  U    U  P    P      *
 *    BBBBB   L       O    O  W    W  U    U  P    P      *
 *    B    B  L       O    O  W WW W  U    U  PPPPP       *
 *    B    B  L       O    O  WW  WW  U    U  P           *
 *    BBBBB   LLLLLL   OOOO   W    W   UUUU   P           *
 *                                                        *
 *                                                        *
 *                  IS CARRIED OUT                        *
 *                                                        *
 *  MOLECULAR ORBITALS ARE BEING EXPRESSED IN NEW BASIS   *
 **********************************************************


 CALCULATING COMPLETE OVERLAP MATRIX


 EFFECTIVE NUMBER OF NON-VANISHING CARTESIAN
 OVERLAP MATRIX ELEMENTS   :  343289

 W A R N I N G : <phi(i)|phi(i)> is   1.00000000000153      for            i =
          47
 W A R N I N G : <phi(i)|phi(i)> is   1.00000000000160      for            i =
          54
 W A R N I N G : <phi(i)|phi(i)> is   1.00000000000251      for            i =
          61
 W A R N I N G : <phi(i)|phi(i)> is   1.00000000000269      for            i =
          78
 W A R N I N G : <phi(i)|phi(i)> is   1.00000000000403      for            i =
         102
 W A R N I N G : <phi(i)|phi(i)> is   1.00000000000236      for            i =
         124
 W A R N I N G : <phi(i)|phi(i)> is   1.00000000000236      for            i =
         131

 projection quality sqrt{sum[1-<i|i>]}/korb irrep a    :  0.71D-07

 ADJUSTING DATA GROUP  $drvopt  FOR CURRENT REQUIREMENTS

 FOUND KEYWORD $forceapprox IN control
 THIS MAKES ONLY SENSE IF YOU WANT TO CONTINUE WITH A
 PREVIOUS OPTIMIZATION RUN !
 DO YOU REALLY WANT TO USE $forceapprox (DEFAULT=y) ?

So I want to know whether this is fine?

Does define successfullly read the previous orbital and then use the new larger basis set ?


Thanks.

[uwe]

Hi,

it seems that in the first step define had a problem reading in the alpha MOs. This can be caused either by problems with the system (e.g. when running define on an NFS disk with temporary network problems) or by a problem with define. Could you please retry the whole procedure and check if it fails again at the first 'use' ? If the problem is reproducible, please send the files to the Turbomole Support - we will have a look at it and patch it if necessary.

Regards,

Uwe