Azobenzene undergoes trans → cis photoisomerization when irradiated with light with a wavelength of 300–400 nm.

By adding electron withdrawing groups, the wavelength can be red-shifted.

If I were to make the molecule shown below, would it be possible to selectively photoisomerize around each of the $\ce{N=N}$ at different wavelengths?

Proposed bis-diazo compound

  • 2
    $\begingroup$ Yes you can increase wavelength by substituents and by increasing conjugation length also, for example a simple polyene such as butadiene vs. carotene or a substituted polyene such as retinal. But no, you cannot just excite one bond in the whole conjugated part of the molecule. Using tailored femtosecond pulses there has been much effort to do 'bond selective' chemistry but what the results mean is still hotly disputed. $\endgroup$
    – porphyrin
    Oct 26, 2016 at 17:00
  • 1
    $\begingroup$ I don't remember the reference any more, but there was a way to have selectivity that was between cis- and trans- double bonds. I think as long as the energy difference is big enough relative to the spread in your input light, you should be able to have some selectivity. The only thing to watch out for is that as you increase substitution, you're going to start disfavoring the cis- form, which I gather is what you want to make. $\endgroup$
    – Zhe
    Oct 26, 2016 at 17:09
  • 1
    $\begingroup$ Related: chemistry.stackexchange.com/questions/31730/… $\endgroup$ Oct 26, 2016 at 17:20
  • 1
    $\begingroup$ Thanks Porphyrin and Zhe. I am trying to design a photoswitchable ion binder for a supramolecular chemistry class. I am trying to design a way to 'lock' the molecule in a particular conformation using a thiol-disulfide dynamic interchange. I would like to be able to switch between which of the rings is locked to the other by using different wavelengths. Pretty far fetched idea, needs lots of refining, but it would be very nice to be able to rotate around a bond selectively. Thanks again Klaus, I asked the azobenzene question the other time :) Still much to learn about photochemistry! :p $\endgroup$
    – Max
    Oct 26, 2016 at 17:27
  • 2
    $\begingroup$ cis and trans selectivity by wavelength, see for example hemithioindigo. Disclaimer: My former group. $\endgroup$
    – Jan
    Oct 26, 2016 at 19:50

1 Answer 1


In theory: yes

As soon as you have two distinct groups, you will have two different, discrete energy differences. If you manage to irradiate at exactly the correct wavelength, only one of the two will be excited.

In practice: possibly, but it depends

Two diazo groups are still rather similar. You will either need a beam that has a good wavelength focus or you will have to make their wavelengths sufficiently different.

By the way, note that both your average IR spectrum and your average UV/Vis spectrum show this property. At a specific wavelength, exactly one vibration (IR) or electron (UV/Vis) will be excited. With sufficient resolution, you should be able to excite the different nitro group vibrations separately. If this applies to one light wavelength band, it will work in others, too.

  • $\begingroup$ I think that you are not correct here, in the proposed molecule both cis and trans chromophores are in the same conjugation so are part of the same delocalised MO. If the were not connected electronically, say widely separated by a alkyl chain, then only if cis & trans absorb at different wavelengths would it be possible to change the one with the lower energy or longer wavelength. At shorter wavelengths where both absorb no discrimination would be possible. In solution levels are so broadened and are so overlapped that specific narrow band excitation has no particular effect. $\endgroup$
    – porphyrin
    Oct 27, 2016 at 7:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.