valence check bug

Marvin thinks the C in the following compound has a valence error, but it does not.

Code:

<?xml version="1.0" ?> <MDocument>   <MChemicalStruct>     <molecule molID="m1">       <atomArray           atomID="a1 a2 a3"           elementType="Cl C Cl"           formalCharge="0 -1 0"           radical="0 trivalent 0"           x2="2.654325933058848 2.654325933058848 1.114325933058847"           y2="-4.697366437119687 -6.237366437119686 -6.237366437119686"           />       <bondArray>         <bond atomRefs2="a2 a1" order="1" />         <bond atomRefs2="a2 a3" order="1" />       </bondArray>     </molecule>   </MChemicalStruct> </MDocument>

Thanks, we will fix that.

This seems to be problematic in SMILES. As the molecule has the following SMILES output:


Cl[C-]Cl


If we import this, the carbon becomes monovalent radical.


The reason is the following:


C has 4 valence electrons from which 2 is used in sigma bond and one goes to the negative charge:


4 - 2 - 1 = 1


So the central carbon have 1 radical electron.





The other approach is that the plus electron (which gives the negative charge) may not located in the same orbital as a valence electron so alltogether 3 electron gives the radical number.





I think it is not possible to decide which approach to use if just the SMILES is given.


Moreover generally the first approcah is more plausible.

What does "trivalent radical" mean if not "three unshared electrons"?





And what does SMILES have to do with it? The molecule was given in MRV format. SMILES doesn't have a "trivalent radical" concept anyway.





It doesn't bother me that you don't give a valence error when you call it a monovalent radical. What bothers me is that you do give a valence error when you call it a trivalent radical.

OK


I just wanted to warn you that it is not possible to export and import it back the trivalent radical as SMILES.

Marvin 4.1.2 will contain the fix.