Summary of How to calculate Bond Orders.

From: somsak (+ at +) atc.atccu.chula.ac.th (Somsak Tonmunphean)
Subject: Summary of How to calculate Bond Orders.
Date: Tue, 17 Dec 1996 10:40:31 +0000 (WET)
 Dear Netters,
 	A few days ago, I asked the question "How to calculate bond order
 of molecule ?", I have received some good anwering mails and here is my
 summary.
 	Thank you very much to the following persons who gave the great
 answers.
 	 Robert Ponec (RPONEC (+ at +) icpf.cas.cz)
 	 Dr.Stephan Irle (stephan.irle (+ at +) itc.uivie.ac.at)
 	 Dr. Jeff Gosper (Jeffrey.Gosper (+ at +) brunel.ac.uk)
 	 John R. Nash (nash (+ at +) chem.wisc.edu)
 	 Dr. Joel Polowin (polowin (+ at +) hyper.com)
 	 Elmar Gerwalin (gerwalin (+ at +) chemie.uni-kl.de)
 	 Alan Shuterman (Alan.Shusterman (+ at +) directory.Reed.EDU)
 	 Dr. Tapas Kar (TAPASKAR (+ at +) SIUCVMB.SIU.EDU)
 	 Dr. Damir Kovacek (dkovacek (+ at +) spider.irb.hr)
 ***************************************************************************
 From: Robert Ponec (RPONEC (+ at +) icpf.cas.cz)
 There is still some discussion about how to define bond order. The
 situation is clearer at SCF level where the generally accepted definition
 of bond order (sometimes called bon index) is that proposed some years
 ago by Wiberg. His bond indeces are sometimes referred to as Wiberg
 indices). What MOPAC calculates using BOND keyword is just this type of
 index. It has the advantage that his values for single bonds are close to
 1, dor double bonds close to 2 etc and for nonbonded atoms they are
 usually negligible and so they correspond well to the classical
 understanding of bond multiplicity. The Wiberg indices were originally
 defined for orthogonal basis sets ( i.e. they are straightforwardly
 applicabl;e at semiempirical level) but the extension to ab initio level
 was proposed by Mayer. If you are satisfied with SCF level only, it is
 best to use Wiberg or Mayer indices. A bit more complex situation is for
 correlated wave functions where the definition od bond index is still the
 subject of discussion. Since I have been working in this field quite
 recently, I can send you a reprint of my recent study on this subject.
  You can find all the references to Wiberg and Mayer works there. Let me
 please know whether you are interested in it.
 **************************************************************************
 From: stephan.irle (+ at +) itc.univie.ac.at (Stephan Irle)
 You should have asked me. I've implemented Mayer's bond orders
 into G92, and you can have it, of course with the references.
 **************************************************************************
 From: polowin (+ at +) hyper.hyper.com (Joel Polowin)
 It is not possible to calculate bond orders automatically with HyperChem,
 but they can be calculated manually from the density matrices, which can
 be printed in a log file.  The bond order between atoms A and B is the sum
 of the squares of the elements in the density matrix with rows for atom A's
 orbitals and columns for atom B's orbitals, or vice versa.  If you are
 doing UHF calculations, you must add the elements of the alpha and beta
 matrices together before squaring.  I tried this for C2H4 as model-built,
 with an AM1 single-point calculation, and got a C--C bond order of 2.0004.
 **************************************************************************
 From: "John R. Nash" <nash (+ at +) chem.wisc.edu>
 F. Weinhold's Natural Bond Order (NBO) calculation available with G92 (and
 G94) will calculate bond orders (and do decomposition of bonds into
 "ionic"
 and "covalent" parts if requested).  Check the manual for details, but
 at
 its simplest, you just put "POP=NBO" in the route card and $NBO $END
 at the
 end of the input file.  This gives the default NBO analysis, including bond
 orders.
 **************************************************************************
 From: Jeffrey J Gosper <Jeffrey.Gosper (+ at +) brunel.ac.uk>
 MOPAC can calculate bond orders (using the BONDS keyword). We are currently
 working on an Windows program that will take XYZ files, run MOPAC, and
 produce an 'extended' XYZ file that contain bond connectivities and orders
 which can be viewed in Re_View2.
 Until this is finished you can look at the degrees of bonding table in the
 MOPAC output.
 **************************************************************************
 From: Elmar Gerwalin <gerwalin (+ at +) iris1.chemie.uni-kl.de>
 In Gaussian you can calculate Bond orders very easily:
 Mulliken Overlap Populations are calculated by default.
 Better is the NBO-Program Package that mostly comes along with Gaussian (all
 Versions up to G94)
 Please refer to the Gaussian Homepage, your Software ADministrator or search the
 Web for NBO (Natural Bond Order Analysis).
 If I find a link to Gaussian/NBO in my own Bookmarks, I will tell you.
 **************************************************************************
 From: "Tapas Kar, Ph.D, Asst. Scientist" <TAPASKAR (+ at +)
 SIUCVMB.SIU.EDU>
 Here is some important references for bond order calculations:
 J. Chem.Education 65(1988)674, J. Mol.Structure(Theochem)209(1990)45,
 Chem.Phys.lett.173(1990),569 and 192(1992)14.
 You have to modify l601.F of gaussian92 to add bond order calculation
 because it is not available as standard calculation in G92. It is a
 simple program and you can add few lines in l601.F.
 **************************************************************************
 From Dr. Damir Kovacek <dkovacek (+ at +) spider.irb.hr>
 GAMESS (US version) can calculate total bond orders and I have programs
 that can extract pi-bond orders from GAMESS and MOPAC output. If you are
 interested just send e-mail.
 **************************************************************************