Problem with CCSD(T) optimization followed by subsequent NBO and other calculations

Questions
#1

While trying to run an optimization followed by harmonic vibrational analysis I encountered the following error:
******************************
** OPTIMIZATION CONVERGED **
******************************
Requested basis set is non-standard
There are 35 shells and 89 basis functions

Q-Chem fatal error occurred in module libmdc/newfileman.C, line 376:

FileMan error: End of file reached prematurely reading (360) bytes in file FILE_DIPOLE_DERIVATIVES
Path: /home/scr/arikco/qchem3452141/133.0

Please submit a crash report at Q-Chem Crash Reporter

My input includes the following parameters/flags:

$rem
BASIS GEN
ECP GEN
ECP_FIT = TRUE
PURECART 1
method ccsd(t)
jobtype opt
geom_print TRUE
PRINT_DIST_MATRIX 999
GEOM_OPT_PRINT 5
USE_LIBPT true
NBO true
$end

I’ll be most thankful for any help in the matter.

Thanks a lot,

Arik

#2

Hi Arik,
Can you please post a minimal (but complete) input file that exhibits the problem? It looks like the geometry optimization completed successfully then something subsequent to that was a problem. That could be NBO=TRUE (as NBO was designed for SCF and not CCSD(T) calculations), or it could be something else. It may not be a good idea to combine all of these things together into a single job. Maybe optimize the geometry first and then run subsequent analysis calculations at that geometry, especially given that the optimization uses finite-difference gradients.

#3

Hi J ( :)),

Thank you you for the fast response and help above all.
Kindly see bellow the input file:

$comment
Optimize HXeNH2 then perform vibrational analysis on final structure.
$end

$molecule
0 1
N 0.000000 0.000000 0.000000
Xe 0.000000 0.000000 2.349089
H 0.080033 0.000000 4.191635
H -0.618583 0.777530 -0.228302
H -0.556748 -0.823225 -0.227285
$end

$rem
BASIS GEN
ECP GEN
ECP_FIT = TRUE
PURECART 1
method ccsd(t)
jobtype opt
geom_print TRUE
PRINT_DIST_MATRIX 999
GEOM_OPT_PRINT 5
USE_LIBPT true
NBO true
$end

$basis
N 0
S 6 1.00
6293.480 0.00196979
949.0440 0.0149613
218.7760 0.0735006
63.69160 0.2489370
18.82820 0.6024600
2.720230 0.2562020
SP 3 1.00
30.63310 0.111906 0.0383119
7.026140 0.921666 0.237403
2.112050 -0.00256919 0.817592
SP 1 1.00
0.684009 1.000000 1.000000
SP 1 1.00
0.200878 1.000000 1.000000
SP 1 1.00
0.0639000 1.0000000 1.0000000
D 1 1.00
1.8260000 1.0000000
D 1 1.00
0.4565000 1.0000000

Xe 0
S 1 1.00
0.7127000 1.0000000
S 1 1.00
0.5719000 1.0000000
S 1 1.00
0.1519000 1.0000000
P 1 1.00
1.2353000 1.0000000
P 1 1.00
0.3726000 1.0000000
P 1 1.00
0.1229000 1.0000000
D 1 1.00
4.5119000 1.0000000
D 1 1.00
2.4799000 1.0000000
D 1 1.00
1.2983000 1.0000000
D 1 1.00
0.5435000 1.0000000

H 0
S 3 1.00
33.86500 0.0254938
5.094790 0.190373
1.158790 0.852161
S 1 1.00
0.325840 1.000000
S 1 1.00
0.102741 1.000000
S 1 1.00
0.0360000 1.0000000
P 1 1.00
1.5000000 1.0000000
P 1 1.00
0.3750000 1.0000000

$end

$ecp
XE 0
XE-ECP 3 36
f potential
4
2 2.24699998 -2.98907089
2 6.12760019 -16.86190414
2 23.48080063 -80.85955811
1 72.73519897 -23.83045006
s-f potential
6
2 1.24269998 36.93946457
2 1.45309997 -117.99051666
2 1.92760003 217.58146667
2 2.56839991 -132.93127441
1 3.60890007 13.86918831
0 2.81739998 7.20968485
p-f potential
6
2 1.02950001 38.26582718
2 1.20910001 -120.96846008
2 1.59679997 217.59863281
2 2.10870004 -144.18939209
1 3.15689993 33.80591965
0 17.37170029 5.23532295
d-f potential
6
2 2.25279999 -17.56745148
2 2.73140001 46.21330261
2 3.76239991 -80.85101318
2 5.32089996 97.02409363
1 8.73659992 -0.88147599
0 18.91500092 8.33844185

$end

$nbo
nbo 2
$end

$geom_opt
hessian_verify = recomputed !Recompute Exact Hessian for verification
final_vibrational_analysis = true !Perform Vibrational Analysis on optimized Hessian
$end

I thank you again for your help!

Best Regards,

Arik

#4

Could you please tell us which Q-Chem version you are using? $nbo section in your input seems to be the issue. Please refer to the minimal input below that works fine. Also, refer to the NBO section in the Q-Chem User Manual (10.3.1 Interface to the NBO Package‣ 10.3 Orbital Analysis ‣ Chapter 10 Molecular Properties and Analysis ‣ Q-Chem 6.3 User’s Manual) for correct usage of keywords.

$molecule
0 1
O
H 1 1.1
H 1 1.1 2 110.
$end

$rem
BASIS sto-3g
PURECART 1
method ccsd(t)
jobtype opt
geom_print TRUE
PRINT_DIST_MATRIX 999
GEOM_OPT_PRINT 5
USE_LIBPT true
$end

$geom_opt
hessian_verify = recomputed !Recompute Exact Hessian for verification
final_vibrational_analysis = true !Perform Vibrational Analysis on optimized Hessian
$end

@@@

$molecule
read
$end

$rem
BASIS sto-3g
PURECART 1
method ccsd(t)
USE_LIBPT true
NBO 1
$end

$nbo
PLOT
$end
#5

This is what I was suggesting above - do this in separate jobs. First optimize with CCSD(T). Then run the NBO analysis with METHOD = HF (since it’s not going to use the correlated method anyway).