Molcas Forum

emilioerc

Member

Registered: 2021-12-27

Posts: 9

Surface hopping does only 1 MD step

Dear all,

With the input described below I would like to run a SH calculation from a Wigner distribution, but it stops at the first step without displaying any error. Hope you can help me!

&GATEWAY
 COORD=test.xyz
 BASIS= cc-pvtz
 GROUP= nosym

>> EXPORT MOLCAS_MAXITER=10
>> DOWHILE

&SEWARD

>> IF ( ITER = 1 )

&RASSCF
Symmetry=1
Spin=2
Nact=7 0 0
Inac=1
RAS2=11
CIROot= 3 3 1
MDRLXR = 1
>> COPY $Project.JobIph $Project.JobOld

>> ENDIF

&RASSCF
 JOBIPH; CIRESTART
Symmetry=1
Spin=2
Nact=7 0 0
Inac=1
RAS2=11
CIROot= 3 3 1
 MDRLXR= 1
>> COPY $Project.JobIph $Project.JobOld

&surfacehop
 TULLY
 SUBSTEP = 200
 DECOHERENCE = 0.1
 PSUB

&ALASKA

&Dynamix
 VELVer
 DT= 10.0
 VELO= 1
 THER= 0
 TEMP=298

>> END DO

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,012

Re: Surface hopping does only 1 MD step

Without test.xyz, no one can try it.

emilioerc

Member

Registered: 2021-12-27

Posts: 9

Re: Surface hopping does only 1 MD step

Sorry, I did not notice. These two files come from the Dynamixtool script for generating the initial conditions from a Wigner distribution.
test.xyz is:

5

C -0.0007127864755757133 -0.0012606427444886107 0.0009820003233501086
H -0.001297436341570797 0.004206893452240651 1.079686998090965
H 1.0256222458811735 0.009189848844910047 -0.36701044317196496
H -0.512758782848106 -0.8722667062395215 -0.36254591804989894
H -0.5030788264909528 0.8738802270842687 -0.36180397044572793

and test.velocity.xyz is:

-0.000694 -0.001001 -0.000506
0.057508 -0.019487 -0.001932
-0.012544 0.047056 0.055068
-0.031462 0.038231 0.008139
0.015036 -0.024657 -0.040479

Ignacio

Administrator

From: Uppsala

Registered: 2015-11-03

Posts: 1,012

Re: Surface hopping does only 1 MD step

It works fine for me if I copy the velocity file:

>> COPY $CurrDir/test.velocity.xyz $Project.velocity.xyz

     Molecular Dynamics specifications (time =     90.0 a.u.)
     ==========================================

     Kinetic energy         0.1252D+02 a.u.
     Total linear momentum  0.1181D+03 a.u.
     RMS deviation          0.5215D+00 Bohr

Otherwise it fails, but it also gives an informative error:

     Dynamics in full dimensionality.
EOF reached for file=/scratch/test.3886306/test.velocity.xyz
 ###############################################################################
 ###############################################################################
 ###                                                                         ###
 ###                                                                         ###
 ###    ERROR: Error in Get_Ln                                               ###
 ###                                                                         ###
 ###                                                                         ###
 ###############################################################################
 ###############################################################################

(i.e., the file is not there)

emilioerc

Member

Registered: 2021-12-27

Posts: 9

Re: Surface hopping does only 1 MD step

It works fine for me if I copy the velocity file:

>> COPY $CurrDir/test.velocity.xyz $Project.velocity.xyz

     Molecular Dynamics specifications (time =     90.0 a.u.)
     ==========================================

     Kinetic energy         0.1252D+02 a.u.
     Total linear momentum  0.1181D+03 a.u.
     RMS deviation          0.5215D+00 Bohr

Otherwise it fails, but it also gives an informative error:

     Dynamics in full dimensionality.
EOF reached for file=/scratch/test.3886306/test.velocity.xyz
 ###############################################################################
 ###############################################################################
 ###                                                                         ###
 ###                                                                         ###
 ###    ERROR: Error in Get_Ln                                               ###
 ###                                                                         ###
 ###                                                                         ###
 ###############################################################################
 ###############################################################################

(i.e., the file is not there)

Thank you very much, Ignacio. Now the issue is that it seems not working properly, due to the results make no sense. I was wondering how to perform MD from a Wigner distribution from Dynamixtools utility.