When making a biradical atom, both when creating a molecule from an RDKit Molecule (eg. reading from SMILES) and when incrementing a radical as part of a reaction recipe, we assume it will be a triplet. i.e. 2T rather than 2S.
For a few species, such as CH2 ([CH2]), CCH2 (C=[C]), HCOH ([CH]O) this is not the case, and we will look up the incorrect thermochemistry in our thermo libraries as a result.
Some models (which we may use as seed mechanisms or reaction libraries) have both the singlet and the triplet species in for some of these (eg. CH2), with different reactions. Others (eg. HCOH) they just include the singlet, but we look for the triplet.
We need to come up with some policy, then implement it.
Also, when we create 'augmented' InChI's, we append the mult# string to the inchi based on number of radicals, however this should not be done for singlet birads like CH2 (we need to distinguish between 2T and 2S).
When making a biradical atom, both when creating a molecule from an RDKit Molecule (eg. reading from SMILES) and when incrementing a radical as part of a reaction recipe, we assume it will be a triplet. i.e.
2T
rather than2S
. For a few species, such as CH2 ([CH2]
), CCH2 (C=[C]
), HCOH ([CH]O
) this is not the case, and we will look up the incorrect thermochemistry in our thermo libraries as a result. Some models (which we may use as seed mechanisms or reaction libraries) have both the singlet and the triplet species in for some of these (eg. CH2), with different reactions. Others (eg. HCOH) they just include the singlet, but we look for the triplet.We need to come up with some policy, then implement it.