3
$\begingroup$

I have a water cluster(suppose) generated from MD run. Now, I want to optimize the geometry of only one water molecule of interest using QC methods (e.g. B3LYP/6-31+G(d,p)). I intend that the water molecule in my interest sees the other molecules, feels the forces from others, and changes its geometries. But the other molecules should not change their orientations. I initially tried to do a constrained geometry optimization, freezing the coordinates of all the other molecules and optimizing the desired molecule. I think my process is wrong because the central molecule always tries to get the optimized geometry in a vacuum. The bond parameters after constrained optimization are as follows,

Table-1 Optimized bond parameters of water molecule in a cluster

No of Runs O-H1 O-H2 <H-O-H
1 0.98 0.98 107.2
2 0.97 0.99 105.8
3 0.98 0.98 103.7
4 0.98 0.98 104.5

In each run, the input geometry were different. I think the water molecule do not see its surroundings. Is there any other procedure to perform this?

I am using gaussian and the part of input is provided below,

%nprocshared=12
# b3lyp/6-31+g(d,p) opt

Water_Clusters_2

0 1
O 0     -0.09390576      -0.45183448      -0.59180983
H 0     0.16750407      -0.14856647      -1.49420182
H 0     0.18256709      0.31842892      -0.02631596
O -1     1.00231196      1.52224084      1.1787738
H -1     1.25538061      2.24014492      0.65216815
H -1     1.87806003      1.29311014      1.6661256
O -1     1.08837153      -2.68452486      0.38127222
H -1     1.95970961      -2.80844083      0.7159187
H -1     0.81757222      -1.76132007      0.29217934
O -1     -2.67938906      -0.15180864      -1.36528883
H -1     -3.01086211      -0.19444878      -2.26724243
H -1     -1.75668721      -0.1300075      -1.40009288
O -1     1.23091274      0.89115464      -2.72058107
H -1     0.84775908      1.51321166      -3.39661546
H -1     2.1591499       1.12729822      -2.62721016
O -1     -3.36169761      -2.47394794      0.05216379
H -1     -2.90747991      -1.67693393      -0.34192527
H -1     -3.32528304      -2.49691151      1.00874685
O -1     -0.11898284      -1.37525577      -4.04574236
H -1     0.70166768      -0.85704547      -3.81234518
H -1     0.02456058      -2.25007808      -3.61809087
```
$\endgroup$
3
  • $\begingroup$ Did you try the ONIOM approach? $\endgroup$
    – Camps
    Nov 27, 2023 at 14:55
  • $\begingroup$ HI, Thank you for the suggestion. I did not look at ONIOM. I think the ONIOM=EmbedCharge will be useful. Great. But, I do want that other molecular orientations to be fixed. The problem will still persist with ONIOM model. What do you think? $\endgroup$
    – Pro
    Nov 28, 2023 at 1:16
  • $\begingroup$ I would go with the info (2d paragraph in the ONIOM page) about adding the freeze-code. Also, with the freezing/constraint info from here $\endgroup$
    – Camps
    Nov 28, 2023 at 12:45

0

You must log in to answer this question.