Dear Santu Mazumder ,

Hope you are well!

If the graph in arbitrary units then you can not convert it to any absolute unit as for example w/m^2.

If it is in counts/s m^2 then you can convert it into watts/m^2

From the value of the wavelength you can get the the photon energy as

Eph= hc/lambda

where

h is the Planck's consatnt,

c is the speed of light

lambda is the wave length of the emitted radiation.

Then you can convert the no. of counts n /s/m^2 into w/m^2

using the the simple formula

The spectral intensity = n Eph / m^2

So, you just need to multiply counts n by the photon energy at the intended wavelength.

Best wishes

Dear Abdelhalim abdelnaby Zekry Sir,

Thank you so much for your kind attention and reply, and hope you are well too. The articles I have gone through has given the unit of intensity as counts only. They did not mention whether it is per meter^2 per sec or not. Previously I have done the method you mentioned here, but as the unit was not mentioned properly I was a bit more worried about the reliability of the intensity value I got. If I get to know about the method of taking PL intensity(unit) data by any particular PL spectrometer in general, that would be very helpful.

But Abdelhalim abdelnaby Zekry ,

what area is involved?

I think depending on the type of monochromator it is a) the net detector area behind the exit slit of the monochromator which picks up the light or b) the net entrance slit area of the monochromator.

And what is about the spectral resolution (Res e.g. in nm) and the step width (SW also e.g. in nm)?

I think in addition to your calculation one has to apply a factor SW/Res onto your formula. So the counts per step are now normalized to the step width with the resolution appropriately considered. For the luminescence band one now has only to add up the counts at that band (or peak) to get the photon number n, which is mentioned above.

What is about the monochromator efficiency? It is less than 1. What is about the detetcor efficiency? (also a bit less less than 1).

So Santu Mazumder , you see, that there are a lot of uncertainties when calculating a real trustable W/m² value at the point of the detector.

If you do not want to figure out a luminescence yield factor, there is no necessity to have a W/m² value.

So people mostly use only the counts they have measured (from this values you can estimate the noise in the spectrum) but others only use a.u. when they do not operate in a count mode of a detector ( e.g. in current mode of a PMT).

Dear Gerhard Martens sir,

Thank you very much for your prompt reply sir. I need to extract the luminescence intensity to generate a spectrum file. Thank you all again for your kind suggestions. I will look into it, sirs.

But Santu Mazumder ,

a 'spectrum file' also works with a.u.- , count- or counts/s-units of the y-axis.

What do you intend to do with that spectrum?

Real energy- or power-flux values are rarely needed.

Dear Gerhard Martens sir,

The simulator I am using needs a file with extension .spe with the intensity value in W/m^2; it does not have options for feeding the data in counts or counts/s units, sir.

Dear Gerhard Martens , and Santu Mazumder

Welcome!

I agree with you completely Gerhard Martens .

If the area is not known and the counts are not of the input photons at the specific wave length range then one has to has to know completely the detector parameters you have listed above.

What is the solution of this problem?

I think the most important property is the profile of the spectral emissivity that is how the emitted spectrum intensity depends on the wavelength. The absolute value of photo luminesce may be of less interest since it could be changed by the input excitation.

It remains that you make yourself the PL spectral measurements. Or you can modify your software to accept the existing units of the PL spectrum.

Best wishes

A.U (Arbitrary Unit)