[Big-Daddy]

How do we tell if a molecule will have dipole moments cancelling entirely to lead to 0 permanent dipole moment? I know the rule that "the molecule must be symmetric", or, in a way that is more straightforward, the central atom must have 0 lone pairs and all the surrounding substituents must be the same. But I am not sure I buy this later rule; XeF₂ and XeF₄ have lone pairs but if I remember right they are non-polar?

Also, what about something like propane-1,3-diol? Do the dipole moments on that cancel? How can I tell?

(Assume please that calculations aren't possible, I won't have bond dipole moment data with me when I have to answer questions like this!)

[Corribus]

First, you need a good structure.
Second, add up all bond dipoles as vectors.  If they cancel, overall dipole moment is zero.

That's the only full-proof way.

[Big-Daddy]

First, you need a good structure.
Second, add up all bond dipoles as vectors.  If they cancel, overall dipole moment is zero.

That's the only full-proof way.

I know that's the theoretical "definition" of bond dipole moments cancelling. But as I said I won't have the bond dipole moments on me to calculate with, when I'm asked this question!

[Corribus]

You don't really need to calculate anything.  If you have a structure, you can determine from inspection.  There will likely be times where some slight deviations from expectations can lead to small, unexpected permanent dipole moments.  But these occasions will be rare.

There really are no simple rules for this sort of thing.  It's really just something you have to learn to recognize.

[Big-Daddy]

OK, so for instance then propane-1,3-diol. Do the dipole moments cancel? I'm guessing instinctively that they do. In which case it really is the idea of complete molecular symmetry: if we can divide the molecule in half and the two halves are exactly identical - including stereoisomerism, which atoms there are, but probably the isotopes don't need to be the same - then the dipoles will cancel.

[Corribus]

OK, so for instance then propane-1,3-diol. Do the dipole moments cancel?

No, and here's why: molecular structures are not static.  A better example (and easier to visualize) than the one you mention:

1,2-dichloroethane.

Free rotation around the C-C bond creates several types of structures - gauche, staggered, eclipsed.  In some the dipoles cancel out.  In others they don't.  The "true" polarity will be an energy-averaged value of the polarities of all possible structures.  As you may imagine, it also depends on temperature.

[Big-Daddy]

Thanks that makes sense. What about the tetrathionate ion, is that non-polar? Or is it polar for the same reason as 1,2-dichloroethane, i.e. rotation around the S-S single bond.

[Corribus]

The dihedral angle appears to approximately 90 degrees.  This would definitely be polar in my view.  There might be structural intermediates, however, which are nonpolar.

Keep in mind that the accessibility of structural intermediates will depend a lot on sterics, temperature, crystallization state, and so for.

In 1,2-dichloroethane, imagine replace the chlorines with iodines, or something even more bulky.  You might get to the point where rotation is practically impossible at normal temperatures and one isomer becomes "frozen" as the only observable isomer (we speak in approximations here - there will always be SOME probability of rotation).  This needs to be taken into consideration. 

[Big-Daddy]

OK. What about the S₂O₆^2- ion. Is that polar?



If so then you will have to explain to me why C₂H₆ (which is exactly the same as this, given that resonance means each S-O bond is really 5/3) is non-polar.

[Corribus]

From the structure it would seem to have no permanent dipole moment, assuming all the S-O bond orders are equivalent.  I can't think of any reason they wouldn't be off the top of my head.

However in general you can't conclude that because sulfur molecules behave one way, then carbon molecules must behave the same way.  Sulfur bonds quite differently than carbon.

[Big-Daddy]

From the structure it would seem to have no permanent dipole moment, assuming all the S-O bond orders are equivalent.  I can't think of any reason they wouldn't be off the top of my head.

However in general you can't conclude that because sulfur molecules behave one way, then carbon molecules must behave the same way.  Sulfur bonds quite differently than carbon.

Well my idea is that, if a molecule has one or more central atoms (these central atoms must all be the same as one another) each of which has 0 lone electron pairs and otherwise identical substituents, it will be non-polar. Can you think of any examples to contradict this?

[Corribus]

The only examples that jump out of me right away are radicals.

Other than that, this is pretty much the basis of VSEPR theory.  That doesn't mean there aren't other examples that defy your "rule", but probably they will be rare.