The idea is simple: take the measurements with your sample present in optical path and another one without sample. Then divide both spectra, point after point, and you're done. Of course, both spectra should be taken at the same wavelengths.
Such an approach correctly takes into account the simple facts: both the light source and the detector have their spectral characteristics, most often never examined individually, not even taken from manufacturer's data sheets, as they may differ from one slight source (detector) to the other.
I want to plot the graph but my transmittance data is upto 400...how can i convert it to %T @ sara sharifpur
Just use the same units for the signal with sample present and when there is no sample in the optical path, be it V, mV or even nV. Even if your detector is non-linear in amplitude (intensity of light), the result you obtain should never exceed 100%. That is when the noise is negligible. If it is not, then you simply have a great uncertainty. But for obvious reasons the result should lie within the interval [0,100]%, noise included.
how do you set the reference level? may be you set it incorrect. you must measure the %T than to the reference level and obviously it never exceed 100% !
Before calculating the transmittance of your sample you might have done the reference baseline correction. I doubt there may be some error in your baseline corrections. Make sure you have done the baseline correction for transmittance measurements and not for reflectance measurements.
I think he already have a data, you are asking to plot transmittance. Plot Wavelength Vs. Observed transmittance using Origin. Transmittance is always in % so it must be less than or equal to 100%. If you have absorption data, plot Wavelength Vs. absorption or convert the absorption data into transmittance using formula.
Ayan, how thick is your sample? Can you see interference effects? If yes, then the situation is a bit more complicated. Multiple internal reflections are not so trivial to account for correctly.
I see. This may be the cause of your problems. The idea I presented first will not apply here directly, it is only good for samples with thickness much higher than the wavelength of the light (inside investigated material). Now you will need to know the spectrum of reflection coefficient as part of your input data. This is not easy to find, if possible at all, when your sample is flat-parallel plate without well absorbing back side. Well, the back side may be curved (lens-shaped) for this purpose as well, but it is most likely not available for your samples, right?
Anyway, please read about light interference in thin layers or about "Newton rings". Or have your samples at least 10 times thicker, if possible ...
- Badges
- Science topic