# Modeling 3-center-2-electron bonds in Avogadro

Is it possible to model 3-center-2-electron bonds in Avogadro?

It seems to allow users to add however many bonds they would like to an atom, but I can't seem to find documentation on whether this is valid. When modeling $\ce{B2H6}$, for example, I can draw a $\ce{B-H-B}$ bond, but does Avogadro understand that this is a bridge bond?

For example, below is an $\ce{(AlH3)5}$ cluster I modeled in Avogadro. Two pairs of bridge bonds appear merged after UFF optimization, which doesn't seem right. This is what led me to wonder if modeling the bridge bonds the way I did was incorrect.

• somewhat related chemistry.stackexchange.com/questions/55646/… – orthocresol Aug 11 '16 at 0:24
• What do you mean by "merged?" Maybe this is better handled through the Avogadro mailing list or forum – Geoff Hutchison Aug 11 '16 at 21:06
• I'm really sorry for the confusion. I've added numbers on the bridge bonds to the picture for clarity. In the lower left corner, only 1 Al-H-Al bond is visible between each pair of Al atoms, while there are 2 everywhere else. I'll try to word it better on the forum. Thank you for the link! – concernedRock Aug 11 '16 at 23:04
• I believe that you are going to need some QM in order to optimise the geometry. Search the literature or ask on the forums to see what is the lowest level of theory you would need to obtain a reasonable geometry (if that is all you need). There is this, for instance: pubs.acs.org/doi/abs/10.1021/j100170a027?journalCode=jpchax A bit old, though. Once you have your optimised geometry you can check if Avogadro can "understand" it. You can also try Gabedit. If it does, you are done. Otherwise you can manually edit it and save it to a format that stores bond information (such as mol2). – Miro Aug 12 '16 at 20:36

First of all, I am not an Avogadro developer, just an occasional user. My view of this matter is the following:

• As 3D molecular drawing/editing software, Avogradro, does not need to "understand" B-H-B bonds. If it allows you to draw them in a sensible way, it is already fulfilling its function as a drawing/editing program.

• Avogadro is also an interface to a number of codes, either internal or external, that allow you to perform tasks such as geometry optimisations and conformational sampling. Now, there becomes relevant whether or not those methods can "understand" B-H-B bonding. Ab initio and other quantum mechanical methods should afford a reasonable treatment of such molecules, provided that an appropriate level of theory is used, whereas for semi-empirical and molecular mechanics methods it will depend on whether or not a given parametrisation model or a given the force field (version), respectively, have been parametrised to take such situations into account.

In summary, if you are able to draw and edit the structure in Avogadro, the real question is whether or not you are going to do something else with it within Avogadro and whether or not your method of choice is appropriate to treat your molecular systems.

• Thank you for your answer! I was indeed concerned about the validity of auto-optimization on a molecule with lots of bridge bonds. I'm working with (AlH$_3$)$_n$ clusters, and sometimes they don't look right after optimizing. I guess I'll continue for now. :) – concernedRock Aug 11 '16 at 19:42

I'm the developer of Avogadro.

Beyond the other answer, which is excellent, I want to emphasize two things:

• There is no rigorous definition of a "bond" in computational chemistry, particularly quantum chemistry. We like to visualize bonds as chemists, and Avogadro has a procedure for automagically interpreting bonding, and for allowing users to edit the bonds in the visualizer.
• The UFF force field method used by default in Avogadro does have a "bridging hydrogen" atom type for cases like diborane.

There are also some nice ways of visualizing 3-center 2-electron bonds, using QTAIM if your software generates a .wfn file. (Under Extensions -> QTAIM -> Molecular Graph). Using QTAIM, the dotted path indicates the bent 3-center 2-electron interaction.

e.g.: