# How to access implicit bond angles that don't appear in Gaussian's z-matrix input?

Suppose I have a Gaussian script to calculate the energy of a relatively large molecule like hexamethyldisiloxane:

# method ...

Title
0 1
O
Si      1        1.63776
Si      1        1.63776     2      148.48211
C       2        1.87584     1      107.42571     3      198.44765
H       4        1.08758     2      111.25460     1      180.71860
H       4        1.08686     2      110.86273     1      300.80346
H       4        1.08691     2      110.96626     1       60.56110
C       2        1.88065     1      110.11571     3      318.27904
H       8        1.08696     2      111.18628     1      301.17666
H       8        1.08749     2      111.53142     1       61.67217
H       8        1.08768     2      110.77221     1      181.40186
C       2        1.87887     1      109.39896     3       78.96340
H      12        1.08704     2      110.94415     1       57.63532
H      12        1.08764     2      111.00260     1      177.47274
H      12        1.08699     2      111.36071     1      297.48941
C       3        1.87887     1      109.39896     2       78.96340
H      16        1.08707     3      111.36057     1      297.48596
H      16        1.08696     3      110.94715     1       57.63340
H      16        1.08766     3      110.99845     1      177.47927
C       3        1.88071     1      110.11501     2      318.27662
H      20        1.08768     3      110.76810     1      181.39759
H      20        1.08698     3      111.18376     1      301.17588
H      20        1.08741     3      111.52706     1       61.67191
C       3        1.87581     1      107.42600     2      198.44461
H      24        1.08749     3      111.25560     1      180.71260
H      24        1.08679     3      110.86218     1      300.80333
H      24        1.08694     3      110.96749     1       60.55900


The idea I have is to modify the values (bond length, angles, dihedral) of the script to make a rigid scan and plot energy curves.

However, there are angles in the molecule that are not explicit in Gaussian script, such as HCH. This also occurs with some dihedrals.

Would I be able to add a line to the script to characterize the HCH angle?

What I would like was to have an input for the HCH angle in the Gaussian script and to vary it.

Should I use dummy / ghost atoms?

Any suggestions are welcome.

• For such large number of atoms, you might be interested in supporting the creation of a Stack Exchange site for large-scale molecular modeling: area51.stackexchange.com/proposals/122958/…. You could ask here, but that site is going to be exclusively modeling questions, and a higher concentration of experts. Make sure to click "sign-up" before you click commit. – user1271772 Mar 30 '20 at 23:11