I think it has something to do with the energy levels. if you have big discrete energy states you will have a sharp peak but if you have overlapping energy states (degenerated states ) you will have a broader peaks

Thanks Tomer, for sharing this great point, however how isomerism can result in such a difference?

Does the Sharp vs. broad tells any thing about the nature of transitions? such as π → π* vs. n → π* ? 

Well we all know the were we have a maximum this correspond to an allowed transition (epsilon at the wavelength is significant) so a sharp peak will correspond as I said to a discrete allowed transition. now a broad peak might suggest that you dont have a clear discrete energy levels (as I said by degeneration) or the transition is not pure allowed (by the rules) so epsilon is not significant enough, some people might also suggest to look at a radical mechanism   

OK, Hammad Cheema since you asked I'll give it a try. What you are seeing is inhomogeneous broadening, which for the I form is smaller and my guess is that the difference in dipole moment between the excited state and the ground state is (considerably) bigger in THDM-T-DHDM. What is the solvent you used?  If you have an indication that I am correct, I'll explain why.

Did you look at, or do you have literature on both forms of phthalic acid, which is the chromophore? I could look myself, but maybe you already have, or maybe (as many people do nowadays) you did quantum chemical calculations. 

The spectra of both the iso- and tere- form of phthalic acid already show narrower lines for the iso form (I found them on NIST chemistry webbook). There must be more literature on these compounds. 

 Dear Dr. Zwan, thanks for the feedback. These measurements were done in acetonitrile, can you please explain how dipole moment is related with peak broadening?

This is a quite recent question for me to answer, where i have been able to find data on Ter-phthalic acid but nothing on Iso-phthalic acid, some of the useful references are below, how you see the low energy transition, as either π → π* vs. n → π* ? as there is no agreement in literature, specially the first reference below.

http://www.nature.com/pj/journal/v15/n3/pdf/pj198329a.pdf

http://polymerphysics.net/pdf/Polymer_31_2023_90.pdf

The broadening you see is inhomogeneous broadening, which is caused by the slightly different environments the molecule finds itself in, in solution, hence the word inhomogeneous.  There is also homogeneous, or life time broadening, but that is usually much smaller than inhomogeneous broadening. The difference can be observed directly in low temperature experiments where the environment is fixed but nevertheless random, such as hole-burning or fluorescence line narrowing spectroscopy of chromophores in glasses.

In polar solvents, such as acetonitril, the fluctuating environment causes electric field fluctuations at the position of your molecule. In fact, every volume element in the solvent has a fluctuating dipole moment, the more polar (higher dielectric constant) the bigger. These dipoles cause electric fields, and these are the fields that act on the states of the molecule, in a number of ways.  I always think of a molecule as a collection of states and dipole moments. In its simplest form there are two states: the ground state |0> and the excited state |1>, and three dipole moments.  

Each state, ground and excited, has a dipole moment. This can be found as the expectation  value of the dipole operator d (a vector) in that state.  Thus, the ground state dipole moment is

dg =

and the excited state dipole moment is

de =

These dipole moments are usually not the same. For terephthalate the ground state dipole moment will be zero, due to symmetry. For highly symmetric molecules (such as benzene) all dipole moments of all the states will be zero.

These are the dipole moments the fluctuating electric fields act on. Basically the interaction energy is -d.E which you can view as a perturbation on the electronic states. Thus you have to calculate the new transition energies, and if you do that correctly, you will see that only the difference dipole moment de-dg enters the expression. Finally you have to average over the polarization fluctuations to get the broadened spectrum. I attached a paper (rather theoretical though) where this idea was carried much further to time resolved fluorescence, and I'll upload some lectures of mine to RG later where I explain more about these ideas. In this way polarity of the medium and difference dipole moments conspire to give you the broadened spectrum.

Therefore, if there is no difference dipole moment you will see very narrow lines, for instance for benzene or pyridine, where the spectrum clearly shows vibrational structure. In most broadened spectra the vibrational bands overlap and are no longer separately visible. This was the reason for my remark about the difference dipole moment. I can only infer from the structure that the T form will not have a ground state dipole moment.

Do not confuse the difference dipole moment with the transition dipole moment, the third dipole of the two level system. That quantity is , which is not an observable. Its square, , is, and this is proportional to the oscillator strength of the transition. The two dipole moments (difference and transition) are unrelated.

I think you are dealing with pi-pi* transitions. These type of molecules always have two of those, perpendicular to each other and close in energy. The references you give, I only looked at them quickly, are indeed not very clear, but I'll have a closer look at them. It is surprising that so little information can be found on UV spectroscopy of these compounds. Do they fluoresce? I found a paper (APPLIED SPECTROSCOPY

Volume: 58 Issue: 7 Pages: 823-830) which I do not have access to, which suggests they do. 

Hammad,

I am almost certain these are n → π* transitions. A way to prove this is by checking their molar absorptivity which should be in the log(ε) ~ 2-3 (analogous to acetophenone), and by checking a blue solvatochromism happening with increasing solvent polarity: the more polar the solvent is, the larger the absorption energy of the S0 → S1 transition -- i.e. if you use toluene, ethyl acetate and acetonitrile as solvents and take the absorption spectra in that order, you should progressively see a blue shift in the low energy absorption band as the n-state relaxes in the more polar environment and the π* remains fairly constant.

Maybe it'd be helpful to understand the conditions at which you took these spectra first, so as to avoid effects from aggregates. Also, as Gert suggested, you should plot these in Energies/Wavenumbers instead of Wavelengths in the X-axis to better understand the vibrational spacing, plus I suggest you use molar absorptivity in the Y-axis as opposed to just absorbance. This will give more information about what's going on.

Dear Dr. Zwan and Estrada, thanks for the feedback.

Dr. Zwan please see attached the paper you referenced. The compounds shown above have reasonably weak fluorescence (1-2% in ACN compared to upto 8% of ter-phthalic acid under same conditions) in polar solvent such as acetonitrile (ACN), however these strongly fluoresce in toluene. Also trans based compounds (trimers actually) show two and half fold more singlet oxygen in ACN, compared to Iso based trimer. No singlet oxygen for both in toluene. The triplet-triplet signal in my other question is for Trans trimer, where as iso based trimer do no show any triplet-triplet signal in ACN. I also do not have valid reason to understand if there is no nanosecond transient signal for triplet-triplet excited state why i see singlet oxygen from iso-based trimer? 

@ Estrada, thanks for suggestion on confirming the nature of transition, i will look into that.

Dear Hammad Cheema,

Thanks for the paper, I looked through it, but it does not appear to give information on the spectroscopy of your compounds that would be helpful in explaining your spectra. Looking a bit further in the literature, I am surprised that there is so little information in general about the spectroscopy of these compounds. I found a paper with DFT calculations (Spectrochim. Acta A, 57, (2001), 993) for terephthalic acid. These calculations do seem to confirm Leandro's suspicion that the lowest transitions are  n → π*, but also has the strange result of giving a rather large dipole moment in the ground state. I would have expected a dipole moment of 0 in view of the symmetry. Also, in the past the compound was apparently used to determine the presence of OH radicals and metals. I did not dig deep enough to see what the mechanism for that was. 

I do not know what the goal of your research is, but it seems to me that it would be advantageous to figure out some of the more fundamental spectroscopic properties of the i and t phthalic acids, (or simpler derivatives like the methylated compounds) before you can infer what exactly is going on in the excited state. In this I also agree with Leandro that you should investigate the properties in solvents of different polarity, maybe looking at the fluorescence as well (and using Lippert-Mataga's expression to get an estimate of the difference dipole moment).