[wildfyr]

Is an azide more or less electron withdrawing than a chloride?

I ask because I converted a chloro dye to an azido dye, and got a blue shift. I always thought azide was a stronger withdrawing group.

[Corribus]

what is the chromophore?

[pgk]

Azide is an electron donating group.

[wildfyr]

Corribus,
The chromaphore is a cyanine dye. The group is on one of central methines, not the benzene rings on the termini.

pgk, some Hammett plot data I've found says EWG. However I did find this paper which says the story is.... complex https://www.chempap.org/file_access.php?file=282a243.pdf

From their observations it appears that the azide group is electron-donating
in nature. However, if the azide group is in the interaction with a strong electron-donating
substituent, then the electron-withdrawing nature of this group is dominant

later in paper
From integrated absorption intensities of the va
s(NNN) bands (Table 1) it follows
that with azides having electron-releasing substituents the azide group exhibits an
electron-withdrawing ability and that having electron-withdrawing substituents does
the reverse. These observations are in accord with conclusions described in [3].
 

so.... 


Someone also pointed out to me that the electron withdrawing strength of azide can diminish if it is twisted out of plane with respect to the conjugated system, especially since I am looking at the IR absorbing band near 800 nm, which I believe is the π-->π* transition. I'm not sure if I have access to molecular modeling software that I can get a reliable answer from on this.

[pgk]

This is exactly, what I mean (p. 244 in the cited article).
Try, the conjugation of p-nitro-phenyl azide, as an example.
Also, try the conjugation of your azido cyanine dye, if the azide group is in conjugation with the π-system.

[pgk]

Besides, the Hammett equation plot refers to thermodynamic/kinetic data of reactions that involve substituted aromatic rings that may not always coincide with their bathochromic shift (e.g. halide aromatic substitution), neither with their vibrational behavior.
As a comparison example:
Assume the hypothetical nitration of phenyl azide, p-hydroxy-phenyl azide and p-nitro-phenyl azide. The rate and the direction of nitration will be different in each case, depending on the relative electron donating/attraction strength of azide group, compared with the neighbor one; but their UV bathochromic shift will be cumulative (if not conjugative), and with a strong IR band at ≈ 2150 cm-1 that corresponds to the azide group.   
Furthermore, the strength of conjugation is also a question of bond angles, which can explain the decrease any electron withdrawing character of azide, when being out of plane.

[wildfyr]

I used the free molecular modeling program Avogadro today and found that the azide is indeed twisted out of the plane, almost 90°. Part of the molecule is cut out to hide its full ID, but the central carbon chain is conjugated and the azide is in blue.

[pgk]

As mentioned above, spectroscopic behavior does not always follow the chemical behavior a (conjugated) compound.
In your case:
1). Structurally, azide group contains of triple bond that conjugates to a system with two adjacent double bonds:

-N(-)-N(+)ΞΝ  ↔  -Ν=Ν(+)=Ν(-)

The left form is electron donating and the right form is electron withdrawing. Thus, the hypothetical hydrobromination a double bond in conjugation with it, will lead to a mixture of azidobromide isomers at a ratio that depends on the contribution of each conjugative form in the entire conjugation.
2). But bathochromically, azide behaves as an electron donating group that conjugates with the cyanine quaternary nitrogen, via the double bonds sequence, regardless if being out of plane (branched) or not (at the linear end of the conjugated double bonds chain). And this has nothing (or better: a little) to do with its chemical behavior.

[wildfyr]

Thank you pgk. Between your explanation and the undeniable shift seen when adding azide to two different dyes, I am convinced that azides are primarily electron donating.


The Avogadro MM results should be taken with a grain of salt, I did some work with another molecule modeling setup (ChemDraw's) and it show much less out of plane twisting. Not sure which to trust.

[pgk]

Unfortunately, these softwares have a static imaging and fail to correctly represent conjugation, which is a dynamic property.