Formate families

[bjkreitz]

I added new families for formate and other similar species that are bidentate species with a vdW_bond. This PR needs to be reviewed in tandem with RMG-Py PR#2706.

I added the following families:

• Surface_Monodentate_to_vdW_Bidentate, example: [X]OC(H)O + [X] -> [X]~OC(H)O[X]
• Surface_Dissociation_vdW_Bidentate, example: [X]~OC(H)O[X] -> [X]O + O=C[X]H
• Surface_Dissociation_vdW_Bidentate_Beta, example: [X]~OC(H)O[X] -> OCO.[X] + H[X]
• Surface_Abstraction_vdW_Bidentate_Beta, example: [X]~OC(H)O[X] + O=[X]-> OCO.[X] + O[X] + [X]

This is by no means complete and formate can undergo further reactions. However, it should cover some of the most relevant reaction types.
I still have to clean the families up and provide reasonable kinetic parameters, but this allows testing of the RMG-Py PR.
Here is an input file that can be used input.py

I had to get a little creative with the reaction templates since RMG cannot form vdW_bonds. For example, take a look at the template for Surface_Dissociation_vdW_Bidentate:

 *1-*2=*3              *1     *2=*3
  |     :     ---->    ||      |
~*4~  ~*5~~           ~*4~ + ~*5~~

RMG cannot break a bond between *3 and *5 since there is technically no bond. The workaround is to first increment the bond order to form a single bond and then break it. Here is the recipe:

recipe(actions=[
    ['CHANGE_BOND', '*1', 1, '*4'],
    ['BREAK_BOND', '*1', 1, '*2'],
    ['CHANGE_BOND', '*3', 1, '*5'],
    ['BREAK_BOND', '*3', 1, '*5'],
    ['FORM_BOND', '*2', 1, '*5'],
])