$DEL-type optimisation in Gaussian

$DEL-type optimisation in Gaussian

Postby KTP » Fri Apr 15, 2016 7:24 am

Hello,

I'm wondering if anyone can help me here.

I'm performing a rather large $DEL-type optimisation in Gaussian using the "zero" keyword, as given in the example in this tutorial.


My system is much bigger than a water dimer however, and when I input the following into my input file:

$nbo print=0 nbosum $end
$del
zero 2 deloc from 1 to 2 from 2 to 1
$end


My output file gives me this:

Zero delocalization from unit 1 to unit 2
Zero delocalization from unit 2 to unit 1
Deletion of the NBO Fock matrix elements between orbitals:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 41
42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61
62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81
82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101
102 103 104
and orbitals:
182 183 184 1851007100810091010101110121013101410151016101710181019102010211022
10231024102510261027102810291030103110321033103410351036103710381039104010411042
10431044104510461047
Deletion of the NBO Fock matrix elements between orbitals:
40 105 106 107 108
and orbitals:


Before terminating the NBO programme. Basically it doesn't seem to be able to delete all the Fock matrix elements.

As can be seen, there is 1047 orbitals in total, and I noticed on the tutorial I linked it has the following line:

"Note that such $DEL-type optimizations are performed by non-analytic derivative methods that are quite time-consuming, hence restricted to a limited number of optimization variables in practical applications."

Are "optimization variables" the Fock matrix elements, and therefore, does my system exceed the limited number of variables for this calculation to be performed? If so, is there a way a round this or is my system simply too big for $DEL-type calculations.

I've used NBO6.0 on the exact same system for standard NBO calculations and it works fine. It's just this $DEL-type calculation that is giving me problems.

Thanks

KTP
KTP
 
Posts: 3
Joined: Fri Apr 15, 2016 7:17 am

Re: $DEL-type optimisation in Gaussian

Postby ericg » Thu May 05, 2016 3:29 pm

KTP,

The optimization variables referred to here are geometrical variables (bond length, bond angles, etc.). It's not clear to me why NBO6 would fail at this point in your calculation. If you could post a sample Gaussian input file for which NBO deletions fail in this way, I'll look into the issue more closely.

Eric
ericg
 
Posts: 265
Joined: Sat Dec 29, 2012 9:31 am

Re: $DEL-type optimisation in Gaussian

Postby KTP » Tue May 17, 2016 8:28 am

Eric,

Apologies for the late reply. Here's a sample of my input:

#p pbe1pbe/gen guess=noextra int=grid=ultrafine pseudo=read scf=maxcyc=512 pop=nbo6del

Title

1 1
C -3.139659 0.518549 -0.763243
C -4.128662 -0.021097 -1.534393
H -5.087981 0.430037 -1.745269
C -3.663998 -1.279281 -1.986599
H -4.191911 -1.975938 -2.622324
C -2.406781 -1.446351 -1.481903
C -3.168676 1.813185 -0.011995
C -1.778009 2.289660 0.254161
C -0.873998 2.400740 -0.817486
H -1.236080 2.226993 -1.826662
C 0.475442 2.662917 -0.607273
C 0.927642 2.853246 0.709623
H 1.983472 3.033119 0.884754
C 0.042024 2.792747 1.771814
H 0.395607 2.934617 2.785385
C -1.305331 2.488065 1.547522
H -1.978038 2.385445 2.391840
C 1.466005 2.585045 -1.721525
C 2.407221 1.446580 -1.481385
C 3.663941 1.278751 -1.987050
H 4.191713 1.975003 -2.623334
C 4.128347 0.020450 -1.534891
H 5.087289 -0.431199 -1.746384
C 3.139756 -0.518459 -0.762708
C 3.168824 -1.812679 -0.010803
C 1.778137 -2.289619 0.254547
C 0.874562 -2.400353 -0.817496
H 1.237046 -2.226252 -1.826472
C -0.474963 -2.662591 -0.607907
C -0.927699 -2.853301 0.708752
H -1.983618 -3.033133 0.883397
C -0.042498 -2.793197 1.771316
H -0.396503 -2.935365 2.784700
C 1.304955 -2.488497 1.547656
H 1.977327 -2.386080 2.392269
C -1.465118 -2.584304 -1.722534
N -2.043129 -0.342613 -0.706551
N 2.043658 0.343243 -0.705398
Th -0.000056 -0.000023 0.492679
H -2.035801 -3.518108 -1.781334
H -0.933367 -2.478505 -2.675895
H -3.707461 1.719499 0.938532
H -3.709059 2.560692 -0.605153
H 0.934626 2.480106 -2.675183
H 2.036976 3.518735 -1.779406
H 3.710070 -2.560194 -0.603149
H 3.706763 -1.718273 0.940143
B -0.000974 -0.000534 3.330130
H 0.998096 0.188699 2.575855
H -0.194378 0.995139 3.976551


And then after the PP and basis sets I this at the bottom of the file:

$nbo print=0 nbosum $end
$del
zero 2 deloc from 1 to 2 from 2 to 1
$end


Thanks

KTP
KTP
 
Posts: 3
Joined: Fri Apr 15, 2016 7:17 am

Re: $DEL-type optimisation in Gaussian

Postby ericg » Tue May 17, 2016 7:59 pm

Can you post the basis sets and PPs too? I need them too to determine why NBO is failing for you.

Eric
ericg
 
Posts: 265
Joined: Sat Dec 29, 2012 9:31 am

Re: $DEL-type optimisation in Gaussian

Postby KTP » Thu May 19, 2016 5:49 am

Sure. Also, I missed out 2 hydrogens when I was copying the geometry, so I've included them just above the basis sets:

H -0.999522 -0.189641 2.575243
H 0.192150 -0.996265 3.976563

C 0
cc-pVTZ
****
H 0
cc-pVDZ
****
N 0
cc-pVTZ
****
B 0
cc-pVTZ
****
Th 0
S 3 1.00
12059.3850000 0.000195
1804.8683000 0.001577
343.6386600 0.005956
S 1 1.00
93.6970580 1.0
S 3 1.00
54.6449460 0.040305
31.9081480 -0.215033
18.6141930 0.706484
S 1 1.00
9.3960940 1.0
S 1 1.00
2.5745050 1.0
S 1 1.00
1.3850100 1.0
S 1 1.00
0.5516260 1.0
S 1 1.00
0.2538890 1.0
S 1 1.00
0.0556970 1.0
S 1 1.00
0.0228960 1.0
P 5 1.00
2892.2961000 0.000039
641.4299000 0.000362
220.2668600 0.001338
80.4092960 0.005257
23.9598100 0.003058
P 1 1.00
13.9962800 1.0
P 1 1.00
8.1675990 1.0
P 1 1.00
3.5494310 1.0
P 1 1.00
1.7586970 1.0
P 1 1.00
0.6480260 1.0
P 1 1.00
0.2835080 1.0
P 1 1.00
0.0992710 1.0
P 1 1.00
0.0298450 1.0
D 6 1.00
448.4346700 -0.000007
150.8962600 0.000219
59.7091610 -0.000113
17.0286560 0.028374
9.9198280 -0.129626
3.2812650 0.453420
D 1 1.00
1.5927460 1.0
D 1 1.00
0.7041920 1.0
D 1 1.00
0.2547970 1.0
D 1 1.00
0.0827760 1.0
F 5 1.00
84.5709310 0.000157
24.4723760 0.001710
11.2617710 -0.001029
3.4474940 0.168891
1.5391530 0.342439
F 1 1.00
0.6468680 1.0
F 1 1.00
0.2484750 1.0
F 1 1.00
0.0812470 1.0
G 4 1.00
84.5709310 0.000233
24.4723760 0.000016
11.2617710 0.014242
3.4474940 0.034212
G 1 1.00
1.5391530 1.0
G 1 1.00
0.6468680 1.0
****

Th 0
ECP60MWB 5 60
H-Komponente
1
2 1.00000000 0.00000000
S-H
1
2 14.77642413 455.05596763
P-H
1
2 9.08318434 177.69179070
D-H
1
2 9.65385771 197.23270511
F-H
1
2 7.38076589 82.90872723
G-H
1
2 11.60902462 -55.12641906


Thanks

KTP
KTP
 
Posts: 3
Joined: Fri Apr 15, 2016 7:17 am

Re: $DEL-type optimisation in Gaussian

Postby ericg » Fri May 27, 2016 8:50 am

I encountered no problems with the calculation. You might try the most recent release of the NBO program (posted May 12), which I am using.

Eric
ericg
 
Posts: 265
Joined: Sat Dec 29, 2012 9:31 am


Return to General NBO Discussion

Who is online

Users browsing this forum: No registered users and 1 guest

cron