TURBOMOLE Users Forum

TURBOMOLE Modules => Aoforce and Numforce => Topic started by: chandra on July 17, 2017, 06:15:10 PM

Title: vibrational frequency analysis for symmetric molecule
Post by: chandra on July 17, 2017, 06:15:10 PM
Hi,

I am looking at the optimization of -NH2 functionalized adamantane  molecule (cs symmetry) using DFT (PBE/def2-tzvp) method. I am able to optimize the structure. To confirm that the the optimized geometry that I have obtained is local minima I should perform vibrational frequency analysis. I tried  but unable to calculate the vibrational frequencies (please see the attached file). I am always getting an error. Please let me know what is the issue. In the manual it is mentioned that for high symmetric systems turbomole fails to calculate the hessain matrix/forces.  I also want to know if there is a way to confirm that the optimized structure that I obtained is minimum energy geometry without  calculating the vibrational frequeny analysis.
Title: Re: vibrational frequency analysis for symmetric molecule
Post by: antti_karttunen on July 17, 2017, 06:58:51 PM
Hi,

It seems that the actual error message is not printed in the output. So, you would find it in a batch system log file or in your terminal. My guess is that this is a memory-related issue. Please see the following FAQ: http://www.turbo-forum.com/index.php/topic,23.0.html

Quote
In the manual it is mentioned that for high symmetric systems turbomole fails to calculate the hessain matrix/forces
I'm pretty sure that Turbomole has no problems for highly symmetric systems. Even some really high-symmetry point groups that do not exist in most programs work fine with aoforce (e.g. D24h). The only case where I've encountered problems with aoforce is point groups with complex irreducible (E) representations. For those, aoforce prints a warning and you probably should use NumForce. However, Cs is certainly no problem.

Quote
I also want to know if there is a way to confirm that the optimized structure that I obtained is minimum energy geometry without  calculating the vibrational frequeny analysis.
To confirm the structure as true local minimum at the potential energy surface, you need to run a vibrational frequency analysis. Making sure that you have the lowest energy conformer/isomer is another type of problem and for this you need to systematically check the energies of other reasonable conformations/isomers.

Best,
Antti
Title: Re: vibrational frequency analysis for symmetric molecule
Post by: chandra on July 18, 2017, 10:20:18 AM
Hi Antti,

Thank you for the reply. I increased the memory but it didn't help me. I got an error 137 and following lines are written in error.output.

/pfs/data4/software_fh2/bw/chem/turbomole/7.1_tmolex42/TURBOMOLE/bin/em64t-unknown-linux-gnu_smp/aoforce: line 40: 25104 Killed   $TURBODIR/bin/`sysname`/aoforce_omp $STROPT

Please find the attached input and output files. Let me know if i am missing anything in my input.
Title: Re: vibrational frequency analysis for symmetric molecule
Post by: chandra on July 18, 2017, 10:28:56 AM
Following is my control file.

control file:

$lastdiag
$rundimensions
dim(fock,dens)=125226
   natoms=28
   nshell=189
   nbf(CAO)=498
   dim(trafo[SAO<-->AO/CAO])=968
   rhfshells=1
$energy    file=energy
$scfdump
$drvopt
cartesian  on
   basis      off
   global     off
   hessian    on
   dipole     on
   nuclear polarizability
$maxcor    5000
$dft
 functional pbe
 gridsize m3
$subtitle
SP,DFT/no-RI,PBE,charge 0,cs,def2-TZVP,FREQ
$dipole from dscf
  x     0.40232309308333    y     0.22808244082158    z    -0.00000000000001    a.u.
   | dipole | =    1.1755095280  debye
$scfiterlimit    30
$coord    file=coord
$jbas    file=auxbasis
$thime    5
$scfconv    6
$charge from dscf
          0.000 (not to be modified here)
$scforbitalshift    automatic=.1
$grad    file=gradient
$optimize
 internal off
 redundant off
 cartesian on
 global off
$symmetry    cs
$tmole
$basis    file=basis
$cosmo_out    file=job_notRun_65.cosmo
$scfmo   file=mos
$title    tmolecal
$atoms
h 11-17,19-28 \
   basis = h def2-TZVP
n 18 \
   basis = n def2-TZVP
c 1-10 \
   basis = c def2-TZVP
$scfdamp     start=0.700 step=0.050 min=0.100
$operating system unix
$thize    0.10000000E-04
$scfintunit
unit=30       size=0        file=twoint
$scfdiis
$last SCF energy change = -445.62776
$closed shells
 a'      1-27                                   ( 2 )
 a"      1-15                                   ( 2 )
$actual step      force
$orbital_max_rnorm 0.17612274894689E-04
$restart force
$end

Title: Re: vibrational frequency analysis for symmetric molecule
Post by: antti_karttunen on July 18, 2017, 07:12:25 PM
Hi,

did you just increase $maxcor or also the user limits as explained in the FAQ (for example, ulimit -s unlimited)?

Antti