Adding halogen nodes, AtomType for monovalent (non-terminal) atoms

[davidfarinajr]

This issue is related to halogens PR: https://github.com/ReactionMechanismGenerator/RMG-database/pull/392

When adding halogen atoms to the kinetics trees, I removed halogen atoms (F, Cl, and Br) from the R!H atomtype. Because I made general nodes with [H, Val7] atoms, I needed an atomtype that was not H and not Val7. My thinking was that, since we have more data for H than for Val7 atoms, creating nodes with [H, Val7] would provide better estimates of the kinetics of halogen reactions when data is laking for halogen species. General idea for tree looks like this (where R!H doesn't include Val7):

So that if there is no data for Val7, the [H,Val7] node would be used.

Alternatively, I could do this:

But then if there is no data for Val7, the R!H node would be used to estimate kinetics.

I don't think I should remove Val7 from R!H because it literally translates to R not H. I think it would be more appropriate to add a mono-valent atomtype (and a not mono-valent atomtype). The mono-valent atomtype could include H and Val7 (F1s,Cl1s,Br1s). Although this gets a little tricky because Cl and Br can have multiple bonds and exceed octet rule because it has access to d orbitals.

Any thoughts, comments, insight would be greatly appreciated!

[rwest]

So is the idea:

• A new atom type "monovalent" meaning can form 1 bond, i.e. include H and Val7
• A new atom type that is "R not monovalent" which in most cases replaces what we currently call "R not H"

?

[rwest]

Feedback from the RMG subgroup zoom just now (if I got this right, @alongd?): try to keep the changes backwards compatible, i.e. at first have "H" and "val7" be children of "monovalent", and "R not monovalent" be a child of "R not H" so that "R not H" still exists. Then gradually replace instances of "R not H" with "R not monovalent" as we confirm that's what they in fact mean.