spin-orbit coupling

Find answers or ask questions regarding Dalton calculations.
Please upload an output file showing the problem, if applicable.
(It is not necessary to upload input files, they can be found in the output file.)

Post Reply
jtravers70
Posts: 10
Joined: 01 Dec 2013, 22:07
First name(s): John
Last name(s): Travers
Affiliation: Boston University
Country: United States

spin-orbit coupling

Post by jtravers70 » 22 Feb 2016, 00:48

Dear DALTON experts

I tried to compute the spin-orbit coupling constant between S0 and T1 using LR-TDDFT. The calculation works fine for some small test cases. But for the molecule I am working on (C11O3H6), the calculation always failed with an error message like:

READT: END reading file
/scr/john/test_08.1063000/DALTON_scratch_john/test_08_118923 UNIT 14
record length 17088

Any suggestions? I tried the calculation in both direct or non-direct fashion, but always failed. The output file is attached.

Thanks!

Best wishes

John
Attachments
test_08.out
(82.81 KiB) Downloaded 293 times

xiongyan21
Posts: 184
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: spin-orbit coupling

Post by xiongyan21 » 22 Feb 2016, 05:48

Dear Dr. Travers
There is no problem to finish your calculation with your original and unaltered input on OS X EI capitan 10.11.3 with Dalton2016.0, but it seems that the print level is too high for me because it pauses for a long time, maybe gets a stuck, in the output from HSOCTL at
IORBSH for ISHELL = 24
107
IORBSH for ISHELL = 25, meaning the Z SPNORB calculation print either encounters difficulties or is much more time-consuming.
The run with the default print level can render the following
...
**************************************************************************************
*** @ Excit. operator sym 1 & ref. state sym 1 => excited state symmetry 1 ( A ) ***
**************************************************************************************



@ Excited state no: 1 in symmetry 1 ( A )
----------------------------------------------

@ Excitation energy : 0.10204605 au
@ ... eV; ... cm-1; ... kJ / mol

@ Total energy : -648.03195 au

@ Operator type: X SPNORB
@ Spin-orbit coupling constant (X SPNORB) : ...
@ Operator type: Y SPNORB
@ Spin-orbit coupling constant (Y SPNORB) : ...
@ Operator type: Z SPNORB
@ Spin-orbit coupling constant (Z SPNORB) : ...


MO Overlap Diagnostic
---------------------



Delta_R Diagnostic (Aa): ...



PBHT MO Overlap Diagnostic
--------------------------

Reference: ...


The dominant contributions: ...


@ Overlap diagnostic LAMBDA : ...
Last edited by xiongyan21 on 27 Feb 2016, 12:53, edited 14 times in total.

taylor
Posts: 545
Joined: 15 Oct 2013, 05:37
First name(s): Peter
Middle name(s): Robert
Last name(s): Taylor
Affiliation: Tianjin University
Country: China

Re: spin-orbit coupling

Post by taylor » 22 Feb 2016, 09:30

Raising a print level should not, on the face of it, slow things down (unless there is some buffering/disk space problem writing to the output file), much less cause the code to "get stuck", so there may be a problem with the program in that respect. But otherwise it seems clear that the original poster's problem is with the build. The directory names suggest you (John) built not only the code but the OpenMPI etc. Further the version numbers of both compiler and OpenMPI are relatively "recent". If it is indeed a build issue I would suggest the following (especially because your system seems to be AMD-based, from the Linux version) and compatibility between the Intel compilers and AMD chips is not always absolutely guaranteed). Try building without OpenMPI. Your calculation is relatively modest in size and anyway you are running sequential in the calculation posted. This would eliminate any question as to whether the problem is with OpenMPI (just running with one process is not a definitive test of this). If the problem persists, rebuild the code with an earlier Intel compiler if available. You can also try gcc/gfortran (without OpenMPI because otherwise you will have OpenMPI compatibility problems unless it is rebuilt with gcc too). My experience is that it is nearly always best to avoid the latest OpenMPI.

Two quantum-chemical questions that are related to your calculation but not really to your posted problem. Your system has Cs symmetry but you explicitly disable symmetry. Any particular reason why? And although I am not an expert in this area I think using the full spin-orbit operator for a system comprising only C, O, and H is massive overkill. My own experiences even for heavier atoms (up to Kr) is that there is no difference between the results of AMFI calculations and those with the full operator, and the former are enormously cheaper.

Best regards
Pete

jtravers70
Posts: 10
Joined: 01 Dec 2013, 22:07
First name(s): John
Last name(s): Travers
Affiliation: Boston University
Country: United States

Re: spin-orbit coupling

Post by jtravers70 » 22 Feb 2016, 18:04

Dear Pete

Thank you for the suggestions! I will try with different build. Indeed, I recompiled a serial version of DALTON but got the same type of error, so I guess this might be related to the intel compiler (15.0.4). I will try with gfortran/gcc.

As for your questions on the symmetry and the choice of the spin-orbit operator, I am just testing the code because of the error.

Thanks again!

Best wishes

John

xiongyan21
Posts: 184
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: spin-orbit coupling

Post by xiongyan21 » 23 Feb 2016, 04:53

Dear Dr. Taylor and all
It only uses several gigabytes of a SSD, and I set 12 gigabyte memory, thus, I may not wait for that long to let it go further except that there
is something missing in the input, which is improper for debugging purpose.

Best Regards!

xiongyan21
Posts: 184
Joined: 24 Sep 2014, 08:36
First name(s): yan
Last name(s): xiong
Affiliation: CENTRAL CHINA NORMAL UNIVERSITY
Country: China

Re: spin-orbit coupling

Post by xiongyan21 » 25 Feb 2016, 08:22

Dear Dr.Taylor and all
I rerun the calculation using print 5 and the printing process is successful
...

... elements with absolute value ... not printed.

The numbers in parenthesis give the orbital symmetry.
Time used in HSOCTL is 12 minutes 50 seconds

@ Transition operator type: Z SPNORB
...

**************************************************************************************
*** @ Excit. operator sym 1 & ref. state sym 1 => excited state symmetry 1 ( A ) ***
**************************************************************************************

...

Best Regards!

Post Reply

Who is online

Users browsing this forum: No registered users and 7 guests