Could not find a valid valance for Atom - Cu

[mit-eremin]

Hi!

I'm already stressed Duarte group gonna hate me and @t-young31 in particular, lol, but:

I'm working on acetylene behavior in water for CuAAC reaction and conformer search works great as it allows to get more good structures rather than me guessing where to put water molecules. But I am now moving to the next step when I want to use Cu(I) coordinated to triple bond and find lowest complex with several water molecules around. I use an xyz input for alkyne-Cu(I) complex. Unfortunately, autodE gives me the following error in the INFO file:

2022-04-08 21:40:11 autode.log.log: INFO     Getting atoms from PhC2H_CuCl.xyz
2022-04-08 21:40:11 autode.log.log: INFO     Generating a Molecule object for molecule
2022-04-08 21:40:11 autode.log.log: WARNING  Not requested any solvent - returning None
2022-04-08 21:40:11 autode.log.log: INFO     Generating molecular graph with NetworkX
2022-04-08 21:40:11 autode.log.log: INFO     Setting the π bonds in a species
2022-04-08 21:40:11 autode.log.log: WARNING  Cu not found in π valency dictionary - assuming not a π-atom
2022-04-08 21:40:11 autode.log.log: WARNING  Cu not found in π valency dictionary - assuming not a π-atom
2022-04-08 21:40:11 autode.log.log: WARNING  Cu not found in π valency dictionary - assuming not a π-atom
2022-04-08 21:40:11 autode.log.log: INFO     Setting the stereocentres in a species
2022-04-08 21:40:11 autode.log.log: WARNING  Could not find a valid valance for Atom(Cu, 0.7681, -0.3143, -2.5625). Guessing at 6

The ade input I use is basically similar to water trimer example and it worked great for a number of alkynes I used so far.

Am I missing something? Any suggestions would be awesome, thank you!

[t-young31]

Hi @mit-eremin

Don't worry about it – I'm always happy to answer questions.

So that statement is just a warning, and can be safely ignored (probably!). Specifically, the maximum valance hasn't been defined for all elements in this function. This property is used when constructing a molecular graph from just atom positions, which is in turn used to discard any conformers with different connectivity. For your system it's definitely worth turning off checks on connectivity, in which case the warning you get definitely doesn't matter! There's an example here that just boils down to setting allow_connectivity_changes=True. Hope that helps

what is a conformer?!?

[mit-eremin]

@t-young31, thank you for suggestions. Somethings began to work, some did not yet. Let me respond after I finish a few calculations to better describe what and how is going on.

Also, sorry for the confusion. I used the word conformer, because after generating possible complexes I use function xxx.find_lowest_energy_conformer.