[strewart]

I'm looking at dinitrotoluene and drawing its structure, the potential for hydrogen bonding and where in the structure that might occur is important. I know Nitrogen has 5 unpaired electrons in the outer shell, so would most readily accept 3 bonds. Now some places show the NO2 group on DNT as:
O=N-O

With a slightly + on the N, and slighty - on the single bond O. Others have two double bonds. Neither satisfies the N 3 bonds (remember it is attached to carbon from the hexane ring as well) which is the correct way to draw it out? It looks to me like the first case would readily allow a H-bond at the single bonded O, but I would have thought that O would be more likely to form another covalent bond with whatever it could.

[nj_bartel]

4 bonds to nitrogen with nitrogen positively charged and each oxygen with a half negative charge is the most important resonance structure.

[Yggdrasil]

The "driving force" for nitrogen to form three covalent bonds is the octet rule: most atoms would like to have eight valence electrons.  Combinations that fulfill the octet rule but involve a nitrogen with more or less than three bonds are still valid.  Also, when drawing the structures, remember to put lone pairs into the diagrams. 

The best representation of the nitro group is one bond between the carbon and nitrogen, one bond between the nitrogen and each oxygen, and one pair of electrons that is shared between the two oxygens and nitrogen.  This gives the oxygen atoms a partial negative charge and a the nitrogen a partial positive charge.  In general, electrons would like to spread out and cover a large area (this is energetically more favorable to confining them to a small area).  So, allowing the two electrons to spread out between three atoms instead of staying between two atoms gives a more stable molecule.  For this reason, the oxygen atoms in a nitro group are not very basic because binding a proton to one of the oxygen atoms would disrupt the ability of the pair of electrons to spread out along the entire nitro group.

[strewart]

Thanks for clearing that up. I am more of a surface scientist than organic, I do remember resonance structures now that you mention it.

So basically, this DNT molecule could possibly accept a hydrogen at one of the oxygens for hydrogen bonding, but there is no way it could have another molecule with strong electronegativity h-bonding to it anywhere? I know the CH3 group is a long way from anything, I was wondering if maybe the H attached to the C-N bond could because of the oxygens being even more electronegative. Like:

H⁺-C^--N⁺-O₂^-

(I know not the correct way to write it, I did that instead of trying to put delta in)

It looks pretty unlikely since H-bonds are usually a fair bit stronger than that, right?

[nj_bartel]

Hydrogen bonding doesn't occur with carbon.  Brief rundown on hydrogen bonding -

Electronegativity is the tendency of an atom to attract electrons toward itself.  When a highly electronegative atom, such as fluorine, nitrogen, or oxygen, bonds to hydrogen, you establish essentially a concentration gradient of electrons.  The electronegative atom draws more bonding electrons toward itself than does the much less electronegative hydrogen.  This creates an electric dipole.  You end up with a partial positive charge on hydrogen and a partial negative charge on the electronegative atom.  This molecule (positive charged on one end and negatively charged on the other) can interact with other similar molecules via electrostatic interactions (+ attracts -, and vice versa).

You could also view it by acid/base theory.  When the hydrogen binds to the electronegative atom, due to the uneven sharing of electrons, it becomes moderately acidic, and the electronegative atom becomes moderately basic.  You establish a network within the molecules of partial protonations/deprotonations.

This is why carbon isn't considered to take part in hydrogen bonding.

What Yggdrasil was saying can best be represented by the acid/base view.  Bases are compounds with highly localized electron concentration.  Because of resonance (delocalized charge), the oxygens are not as basic as they appear to be.  They can still take part in hydrogen bonding to an extent however, assuming another molecule has a protic hydrogen (one attached to an electronegative atom).

[Yggdrasil]

I know the CH3 group is a long way from anything, I was wondering if maybe the H attached to the C-N bond could because of the oxygens being even more electronegative. Like:

H⁺-C^--N⁺-O₂^-

(I know not the correct way to write it, I did that instead of trying to put delta in)

I'm not quite sure what you're trying to say here.  In dinitrotoluene, neither the C nor the N in the C-N bonds are attached to a hydrogen.  The nitrogen of the nitro group is attached to two oxygens and a carbon, while the carbon in the C-N bond is attached to two carbons and a nitrogen (with an unhybridized p-orbital participating in the aromatic system).

[strewart]

Yeah I know, I was clutching at straws a bit. I have read that there are a few rare cases where the H-bonding hydrogen doesn't need to be directly attached to the electronegative atom for it to H-bond. The oxygens draw electrons from the nitrogens, that makes the N slightly positive. Lol, I am trying to visualize the bonds like magnets which clearly isn't the case.

I had another molecule in a solid with DNT, and spectroscopically it seemed to suggest the carbonyl group on the other molecule was H-bonding to the DNT as the peak shifted slightly in areas where both molecules were in comparison to where just the other one was, but it doesn't seem possible since DNT doesn't have any H capable of H-bonding.

Thanks for the help.