AIFDEM triplet state

Questions
#1

I am new to QCHEM (AIFDEM) and want to calculate the multiexciton states for water dimer. The input file is shown below. When I run the calculations I always get the following error message
" Missing file /scratch/ms1771-241935/qchem28267/SFTripFrg1/400.0 in ReadFrgmFile

Q-Chem fatal error occurred in module exciton/././aifdem_util.h, line 30:"

The input file

$rem
BASIS aug-cc-pvdz
EXCHANGE HF
CIS_N_ROOTS 3
CIS_SINGLETS TRUE
CIS_TRIPLETS TRUE
XPOL TRUE
AIFDEM TRUE
AIFDEM_EMBED_RANGE 0
AIFDEM_NTOTHRESH 90
NTO_PAIRS 1
AIFDEM_SINGFIS TRUE
AIFDEM_CTSTATES TRUE
$end

$molecule
0 1
–H2O 0
0 1
O 1.74078 1.59716 -1.49814
H 2.22908 2.18316 -2.08914
H 0.88038 2.04726 -1.32684
–H2O 1
0 1
O 1.31998 -1.18934 -1.91734
H 1.49988 -0.22974 -1.89044
H 1.69058 -1.52594 -1.07704
$end

#2

This looks like it might be a bug; suggest you send it to Q-Chem support (or someone may check this page). Not sure what version of Q-Chem you are running but with the latest trunk, this job crashes for me in a different way that suggests that XPol with CM5 charges is incompatible with AIFDEM. (The XPol default was changed to CM5 at some point fairly recently, would not have been the default in earlier versions.) However, if I turn off XPol altogether, then I do generate your error. I notice that there is no sample in the manual or in the distribution that uses AIFDEM_SINGFIS so someone at Q-Chem should probably create one.

#3

Thank you John for your response. It worked smoothly after switching XPOL off. You are right, the QCHEM team should implement an input sample for the multiexciton stuff into the manual.

#4

That’s good to know. I wonder, with XPol turned on, whether it works if you add the following input section:

$xpol
embed charges
charges chelpg
$end

That will use CHELPG charges for the embedding rather than CM5. The former used to the be default and we switched it for other reasons (for use with XPol + SAPT where CM5 works better), and apparently did not test this with AIFDEM.

#5

Hello mfshibl,

Thank you for the quesiton. You need to add AIFDEM_SEGSTART and AIFDEM_SEGEND (12.17 Ab InitioFrenkel Davydov Exciton Model (AIFDEM)‣ Chapter 12 Fragment-Based Methods ‣ Q-Chem 5.4 User’s Manual) to include multiexciton states (AIFDEM_SINGFIS) as follows. We will update the manual to emphasize the correlation. Please let me know if you have additional questions.

$rem
BASIS aug-cc-pvdz
EXCHANGE HF
CIS_N_ROOTS 3
CIS_SINGLETS TRUE
CIS_TRIPLETS TRUE
XPOL TRUE
AIFDEM TRUE
AIFDEM_EMBED_RANGE 0
AIFDEM_NTOTHRESH 90
NTO_PAIRS 1
AIFDEM_SINGFIS TRUE
AIFDEM_CTSTATES TRUE
SCF_PRINT_FRGM true
AIFDEM_SEGSTART 1
AIFDEM_SEGEND 2
$end

$molecule
0 1
-- H2O 0
0 1
O 1.74078 1.59716 -1.49814
H 2.22908 2.18316 -2.08914
H 0.88038 2.04726 -1.32684
-- H2O 1
0 1
O 1.31998 -1.18934 -1.91734
H 1.49988 -0.22974 -1.89044
H 1.69058 -1.52594 -1.07704
$end
$xpol
embed charges
charges chelpg
$end
#6

I have updated the manual to include a sample. This should appear in 5.4.2 or possibly in 6.0.

#7

Thank you Kyliu. appreciated.