TURBOMOLE Users Forum
TURBOMOLE Modules => Ricc2 => Topic started by: ztehrani on April 16, 2015, 10:31:46 AM
-
Dear Users:
Hi. I have a question. what should I do if I want to find TS @ ri-mp2? Is it available at this level or not? I found some discussion about how to find TS in TURBOMOLE but I have problem for writing script at this level. it is worth to mentioning that I have the initial structure of TS. As I found I should only insert this command in control file:
$statpt
hssfreq 0
itrvec 1
and then I think it may be something like this for calculating frequencies, read hessian, perform TS optimization and calculate frequencies.:
NumForce>numforce.out
jobex -ri -level cc2 -c 300 -energy 7 -gcart 4 -statpt -trans > ricc2-mpi.out
NumForce>numforce.out
in the submit script.
Any help will be appreciated.
Thanks in advance
Zahra
-
Hi!
Of course, you can run a TS search with MP2.
Your NumForce call is wrong. Just calling NumForce without options will give you the HF Hessian.
You would have to start with 'NumForce -level cc2' or 'NumForce -level cc2 -c'.'
'NumForce -h' will give you the possible options.
Also, you would have to remove the numforce subdirectory before running NumForce again in the same directory.
Cheers,
Arnim
-
Dear Arnim,
Hi. Thank you so much for your reply. Therefore, for finding transition state for acrolein with SVP at RI-MP2 level, the control file and the scripts should be as follow. It will be appertained if you let me know these files are correct or if not which kind of corrections should be applied to these file.
Best Wishes
Zahra
Control file:
$title
$operating system unix
$symmetry c1
$redundant file=coord
$statpt
hssfreq 0
itrvec 1
$coord file=coord
$user-defined bonds file=coord
$atoms
c 1-3 \
basis =c SVP \
cbas =c SVP
o 4 \
basis =o SVP \
cbas =o SVP
h 5-8 \
basis =h SVP \
cbas =h SVP
$basis file=basis
$rundimensions
dim(fock,dens)=3336
natoms=8
nshell=36
nbf(CAO)=80
nbf(AO)=76
dim(trafo[SAO<-->AO/CAO])=88
rhfshells=1
$scfmo file=mos
$closed shells
a 1-15 ( 2 )
$scfiterlimit 300
$scfconv 7
$thize 0.10000000E-04
$thime 5
$scfdamp start=0.300 step=0.050 min=0.100
$scfdump
$scfintunit
unit=30 size=0 file=twoint
$scfdiis
$scforbitalshift automatic=.1
$drvopt
cartesian on
basis off
global off
hessian on
dipole on
nuclear polarizability
$interconversion off
qconv=1.d-7
maxiter=25
$optimize
internal on
redundant on
cartesian off
global off
basis off logarithm
$coordinateupdate
dqmax=0.3
interpolate on
statistics 5
$forceupdate
ahlrichs numgeo=0 mingeo=3 maxgeo=4 modus=<g|dq> dynamic fail=0.3
threig=0.005 reseig=0.005 thrbig=3.0 scale=1.00 damping=0.0
$forceinit on
diag=default
$energy file=energy
$grad file=gradient
$forceapprox file=forceapprox
$lock off
$maxcor 20000
$denconv 0.10000000E-06
$cbas file=auxbasis
$freeze
implicit core= 4 virt= 0
$ricc2
mp2
geoopt model=cc2 state=(x)
$last step define
$end
Script.1 (subturb-rimp2-ts)
#!/bin/bash
`perl -pi -e "s/\r//" $1`
here=`pwd`
job=`basename $1 .dat`
proc=$2
sed 's/INPUT/'${job}'/g' /home/bin/turbo-rimp2-ts.sh > $here/${job}.cmd2
sed 's/NPROC/'${proc}'/g' $here/${job}.cmd2 > ${job}.cmd
qsub ${job}.cmd
########################################################################
echo ""
echo ""
Script 2. (turbo-rimp2-ts.sh)
#!/bin/bash
# pe request
#$ -pe mpi_NPROC NPROC
# our Job name
#$ -N INPUT
#$ -S /bin/bash
#$ -q all.q
#$ -cwd
WDIR=/work/$USER/INPUT.$$
mkdir -p $WDIR
DIR=`pwd`
cp * $WDIR
cd $WDIR
export PARA_ARCH=MPI
export TURBODIR=/opt/TURBOMOLE
export PATH=$PATH:$TURBODIR/bin/x86_64-unknown-linux-gnu_mpi:$TURBODIR/scripts:$TURBODIR/mpirun_scripts
export PARNODES=NPROC
NumForce -level cc2 > numforce.out
jobex -ri -level cc2 -c 300 -energy 7 -gcart 4 -statpt -trans > ricc2-mpi.out
NumForce>numforce.out
cp -rf * $DIR;
rm -rf $WDIR
-
Hi!
I can't really proofread all that in detail.
But this:
NumForce -level cc2 > numforce.out
jobex -ri -level cc2 -c 300 -energy 7 -gcart 4 -statpt -trans > ricc2-mpi.out
NumForce>numforce.out
Would have to look more like this:
NumForce -level cc2 > numforce.out
jobex -ri -level cc2 -c 300 -energy 7 -gcart 4 -statpt -trans > ricc2-mpi.out
NumForce -level cc2 >numforce2.out
However, I would not recommend to run these 3 steps in one submit. It is better submit only the NumForce step and check, if everthing worked and visualise the imaginary mode. Then, jobex and visualise the resulting geometry. And then NumForce again. So better, check the results after each step. That can save you time to analyse, if someting doesn't work.
Cheers,
Arnim
-
Dear Arnim,
Hi. Thanks a lot for your comment. I'll test your recommendation.
Best Wishes
Zahra
-
Dear Arnim,
Hi. I tested the job in that way you recommended but it does not work. I used the following steps:
Define
Basis set
EHT
Statp
SCF
MP2
FREEZE
CBAS
END
I am now confused because I also tested the sample test in the following way and it finished successfully. I just used:
Define
Basis set/cc-PVDZ level
EHT
Statp
itrvec 1
And end of define. I used the following command for its running (based on the manual):
jobex -trans
My question is that I did not define the method of calculation (i.e., DFT, RI-DFT, ri-mp2…) how this file was finished successfully. I mean it may calculated in some default method in TM. Another question is that it is necessary to define the method in control file or definition just in script file that’s ok. could you please help me what I should put in the control file @ RI-MP2 level.
Thanks in advance
Best Wishes
Zahra
-
Dear Zahra,
if you run just 'jobex -trans' with the settings in your control file, you run a Hartree-Fock calculation. When you look in the job.last file, you will see the outputs of dscf and grad telling you this. For an MP2 calculation the ricc2 module would be used.
'jobex -h' will give short explanation of the possible options.
Best wishes,
Arnim
-
Dear Arnim,
Hi. Thank you so much for you valuable comments. I tested them and they worked correctly on serial format. I have still some running problems by using scripts including these keywords. I read your comments in this case on previous topics of TM forum. As I found, I should use “-mfile” option or other possibility for running “numforce” in parallel mode at MPI machine (not SMP). As you mentioned before this job should be submitted in node’s cpu.
(http://www.turboforum.com/index.php/topic,936.msg2633.html#msg2633). I could not find a complete script for running job (in parallel or non-parallel mode) because I do not have enough knowledge for writing scripts. I tested a lot of script such as the following one but they are not worked. It will be appreciated if you help me which script I should use for solving this problem.
Best Wishes
Zahra
#!/bin/bash
# pe request
#$ -pe mpi_NPROC NPROC
# our Job name
#$ -N INPUT
#$ -S /bin/bash
#$ -q all.q
#$ -V
#$ -cwd
WDIR=/work/${USER}/INPUT.$$
DIR=`pwd`
mkdir -p $WDIR
cp -fr * $WDIR
cd $WDIR
export PARA_ARCH=MPI
export TURBODIR=/opt/TURBOMOLE
export PATH=$PATH:$TURBODIR/bin/x86_64-unknown-linux-gnu_mpi:$TURBODIR/scripts:$TURBODIR/mpirun_scripts
export PARNODES=NPROC
#NumForce -machinefile $TMPDIR/machines -n $NSLOTS > NumForce.out
#NumForce -level cc2 -machinefile $TMPDIR/machines > NumForce.out
NumForce -level cc2 -mfile $TMPDIR/machines > NumForce.out
cp -fr * $DIR
cd $DIR
rm -fr $WDIR
exit