openbabel-devel

[OpenBabel-Devel] Multiplicity

[Noel O'B. <bao...@gm...>]

According to the OB API, OBMol.SetTotalMultiplicity() works as follows:
"Set the total spin multiplicity of this molecule to spin (i.e., 0 =
singlet (default), 1 = doublet, 2 = triplet, etc.)"

But shouldn't a singlet have a multiplicity of 1, triplet have
multiplicity of 3. See for example:
http://qcl.theochem.tu-muenchen.de/qcl/help/multiplicity_e.html

Is this just an error in the docs?

Noel

Re: [OpenBabel-Devel] Multiplicity

[Andrew D. <da...@da...>]

> According to the OB API, OBMol.SetTotalMultiplicity() works as  
> follows:
> "Set the total spin multiplicity of this molecule to spin (i.e., 0 =
> singlet (default), 1 = doublet, 2 = triplet, etc.)"

The code says

         //Jan 05 Implicit valency now left alone; use spin  
multiplicity for impl
icit Hs
         int mult = atom->GetSpinMultiplicity();
         if(mult==2) //radical
           hcount-=1;
         else if(mult==1 || mult==3) //carbene
           hcount-=2;
         else if(mult>=4) // as in CH, C etc
           hcount -= mult-1;

         if (hcount < 0)
           hcount = 0;
         if (hcount)
           {
             vhadd.push_back(pair<OBAtom*,int>(atom,hcount));
             count += hcount;
           }
       }



and


     //if there are any explicit Hs on an atom, except if their  
_isotope!=0,
     //then they consitute all the Hs.
     //Any discrepancy with the expected atom valency is because it  
is a radical
of some sort.
     //So SMILES CC[2H] is interpreted as CH3CH2D; CC[H] is methyl  
carbene.
     for (atom = BeginAtom(k);atom;atom = NextAtom(k))
       {
         if ((!atom->IsHydrogen() && atom->ExplicitHydrogenCount 
(true)!=0)//exclu
de D,T
             || atom->HasNoHForced())
           {
             diff=atom->GetImplicitValence() - (atom->GetHvyValence()  
+ atom->Exp
licitHydrogenCount());
             if (diff)
               atom->SetSpinMultiplicity(diff+1);//radicals =2; all  
carbenes =3
           }
       }



				Andrew
				da...@da...

Re: [OpenBabel-Devel] Multiplicity

[Chris M. <c.m...@ds...>]

Noel O'Boyle wrote:
> According to the OB API, OBMol.SetTotalMultiplicity() works as follows:
> "Set the total spin multiplicity of this molecule to spin (i.e., 0 =
> singlet (default), 1 = doublet, 2 = triplet, etc.)"
 > But shouldn't a singlet have a multiplicity of 1, triplet have
 > multiplicity of 3.

This is for OBMol and is actually the number of unpaired electrons.
Even more confusing, for OBAtom:

    //! \return the atomic spin, e.g., 0 (default) for singlet,
    //!   2 for radical  1 or 3 for carbene
    int GetSpinMultiplicity() const { return(_spinmultiplicity); }
which is the same as the quantity "Radical" in MDL mol file Property block.

Both of them are mis-named and should really be regarded as conventions. 
The atom version reflects hydrogen deficiency as well as spin, and 
really should have been called something like "radical" (my fault).

Perhaps for OBMol we should have a GetHighSpinMultiplicity() and 
SetHighSpinMultiplicity() with default=1, doublet=2, triplet=3, and 
deprecate GetTotalMultiplicity() and SetTotalMultiplicity().

Chris

Re: [OpenBabel-Devel] Multiplicity

[Noel O'B. <bao...@gm...>]

On 30/08/2007, Chris Morley <c.m...@ds...> wrote:
> Noel O'Boyle wrote:
> > According to the OB API, OBMol.SetTotalMultiplicity() works as follows:
> > "Set the total spin multiplicity of this molecule to spin (i.e., 0 =
> > singlet (default), 1 = doublet, 2 = triplet, etc.)"
>  > But shouldn't a singlet have a multiplicity of 1, triplet have
>  > multiplicity of 3.
>
> This is for OBMol and is actually the number of unpaired electrons.

If this is true, then there must be a bug in the Gaussian output file
writer. If you use "SetTotalSpinMultiplicity" to set the value to X,
then this value of X is used by the Gaussian output file writer as the
multiplicity.

>>> import pybel
>>> a = pybel.Molecule()
>>> a.OBMol.SetTotalSpinMultiplicity(1)
>>> a.write("gau")
'#Put Keywords Here, check Charge and Multiplicity.\n\n \n\n0  1\n\n'
>>> a.OBMol.SetTotalSpinMultiplicity(2)
>>> a.write("gau")
'#Put Keywords Here, check Charge and Multiplicity.\n\n \n\n0  2\n\n'

Here "0  2" means charge of 0 and multiplicity of 2. For your regular
molecules, a multiplicity of 1 is generally used (singlet).

The multiplicity boils down to the number of unpaired electrons + 1. I
would rather see the documentation (and internal code) changed to
reflect this, rather than changing the Gaussian code (and perhaps
others) to conform to a misleading convention. (Then again, I probably
won't be the person doing the changing, so ...)

> Even more confusing, for OBAtom:
>
>     //! \return the atomic spin, e.g., 0 (default) for singlet,
>     //!   2 for radical  1 or 3 for carbene
>     int GetSpinMultiplicity() const { return(_spinmultiplicity); }
> which is the same as the quantity "Radical" in MDL mol file Property block.
>
> Both of them are mis-named and should really be regarded as conventions.
> The atom version reflects hydrogen deficiency as well as spin, and
> really should have been called something like "radical" (my fault).
>
> Perhaps for OBMol we should have a GetHighSpinMultiplicity() and
> SetHighSpinMultiplicity() with default=1, doublet=2, triplet=3, and
> deprecate GetTotalMultiplicity() and SetTotalMultiplicity().

I don't understand why you would need to introduce a high/low spin
method. The molecule either has a particular multiplicity or it
doesn't. As I said, I'm all for redefining the current method to avoid
needless confusion.

So as I see it, there are three reasonable options:
(1) do nothing, but need to fix various parsers that didn't read the
documentation for SetTotalSpinMultiplicity()
(2) change the name of SetTotalSpinMultiplicity() to
SetTotalUnpairedElectrons(), and fix the parsers as for (1)
(3) keep the name, change the docs to adhere to convention, and fix
internal code to use multiplicity instead of number of unpaired
electrons

Noel

> Chris
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a browser.
> Download your FREE copy of Splunk now >>  http://get.splunk.com/
> _______________________________________________
> OpenBabel-Devel mailing list
> Ope...@li...
> https://lists.sourceforge.net/lists/listinfo/openbabel-devel
>