Species labeling in the output doesn't work properly

[bjkreitz]

Bug Description

When generating mechanisms for surface reactions, there is a bug in the naming of the species in the output file. RMG used to replace the SMILES string of the species for example C[Pt] has the label CH3X(8) in the final mechanism. This still works for some species. However, for a bunch of species RMG uses now the SMILES string in the final mechanism e.g. [Pt]CC[Pt] is [Pt]CC[Pt](26) instead of C2H4X2(26).

How To Reproduce

Run the catalysis/ch4_o2 example

Installation Information

Describe your installation method and system information.

• OS: macOS 14.3.1
• Installation method: from source
• RMG version information:
  • RMG-Py: 3.2.0-230-gbfaee1cad
  • RMG-database: 3.2.0-5-gcfb4910cf

[rwest]

I think what it does, or used to do, is

1. use a name from the input file (so the user has ultimate control over things they care about)
2. then try the thermo libraries (in order specified), so most people get a sensible name
3. then make up things with chemical formula or SMILES, depending on label length, etc. (I forget, off hand, the exact details).

I guess a couple of questions.

1. In what way has this changed recently?
2. What would you like it to do?

[rwest]

Oh, one thing that would probably be an improvement would be to replace "[Pt]" with "X" when it reverts to SMILES?
Or someone could try to code up your algorithm you describe at ReactionMechanismGenerator/RMG-database#640

X indicates a bond to the surface. It is always on the left hand site of an atom that is bonded to the surface e.g. XCCH2 it means that C is bonded to the surface.
If the X is on the right hand side and at the end of a label, it means that this species is physisorbed e.g. H2OX.

Although if the "use names from thermo libraries" works as intended, at least all your species will already be named well.