There is something called as "inner filtering efffect" which occurs at a high concentration. The reason why you need low absorbance for the band in the visible region is to avoid this effect in order to get emission from all the "chromophore molecules" in the path of the excitation beam. If the concentration is too high, not all the chromophores in the path are excited and won't get a true emission reading.

Thanks for your reply Dr. Peter. One of the reviewer of my paper has suggested that it should be below 0.2. Now in my case absorbance is below 0.2 in the region of excitation but the whole spectrum has some large absorbance regions (upto 1). Can i say that my studies are ok or i should perform studies again. 

There is almost no overlap between emission and excitation range. Is there any reference material for this?

Reference that mentions that 0.2 absorbance or below that at excitation wavelength is ok  for quantum yield calculation.  

thanks Dr. Peter Kapusta 

Best practice for dilute dye solutions is

Thank you Paul Miller 

There is something called as "inner filtering efffect" which occurs at a high concentration. The reason why you need low absorbance for the band in the visible region is to avoid this effect in order to get emission from all the "chromophore molecules" in the path of the excitation beam. If the concentration is too high, not all the chromophores in the path are excited and won't get a true emission reading.

usually it is good practic to keep OD below 0.1 at the abs. max. (or at times at the wavelebgth used to excite the samples) during QY measurements to avoid erronous results coming from factors like reabsorption/remission etc. however, depending on the difficulty (turbidity or other specific reasons) a relatively higher OD may be allowed but the experimenter must handle it carefully then to ensure that reabsorption does dont creep in to suppress the actual results. 

The limit of absorbance should be lower than 0.05, but really it depends on many factors. Besides inner filter effect it can be aggregation effect (dimerization of very hydrophobic dyes) and excimer formation.  

1.First you have to chose the red band wavelength for the excitation wavelength for the sample solution and absorbance should keep ~0.1 to avoid other interfering effects

2. Actually absorption spectrum contains many band (sometimes), but you have choose the S0->S1 band, which probably the red band in the absorption spectrum for QY calculations

3.Same solution you have to use for the PL measurement

4. For that particular excitation wavelength, the absorbance of reference must be also ~ 0.1

5. Not for all band, only the absorbance of  S0->S1 band must be ~0.1

Good luck

Suppose if your absorption and emission spectra have large stokes shift and they do not show any spectral overlap, is it still necessary to keep optical density below certain limit?

Even in case of a large Stokes shift it is usually customary to maintain a low OD value during QY measurement. A high conc. can lead to many other artifacts like reabsorption, conc. quenching etc. (specific mechanisms might be responsible for these depending on the specific case of study).

Dr. Bijan Kumar Paul are right, aggregation due to hydrophobic, specific interaction as well as excimer formation can distort the quantum yield but it will be difficult to predict at what concentration it will happends. It is need to investigate or find information about concentration effects.

thanks Dr.Bijan Kumar Paul and Prof. Sergei Nikolaevich Shtykov