0
$\begingroup$

In the absorption spectroscopy we can calculate transmittance $T$ of a given sample by comparing how the intensity of the incident beam $I_{0}$ is decreased with the distance (Lambert-Beer Law).

$A = -\ln{\frac{I}{I_{0}}}=\alpha L$

where $A$ is absorbence, $T = 1 -A$, $\alpha$ is an absorption coefficient and $L$ is the distance.

But I encountered something called Normalized transmittence I don't really understand how can I get it from measured data. Normalized how? It appears in scientific articles but I couldn't find any good source explaining it really.

I read that I should divide my measured $I$ by reference intensity (it can fluctuate so it is measured along the way too). So the normalized absorbance would be $A = -\ln{\frac{I}{I_{0}^2}}$ or just by definition $A = -\ln{\frac{I}{I_{0}}} $

Can anyone help me please?

The spectra then look like this: enter image description here

[https://www.researchgate.net/figure/Hydrogen-cyanide-H-13-C-14-N-2-3-rotational-vibrational-band-spectrum-obtained-by_fig4_242279131]

Improve this question
$\endgroup$
3
1
$\begingroup$

Recall the more familiar definition of light transmittance

$T = \frac{I}{I_o}$, many instruments quote percent transmittance as

$\%T = 100\frac{I}{I_o}$

Normalized transmittance is nothing but the maximum measured transmittance set to unity. This is easy to accomplish. Suppose you have 2000 data points (y-values) in your spectrum collected as a function of wavelength. Locate the maximum value of transmittance, say it was 95% or 0.95 in your data set (depending what was collected experimentally). Now divide the entire y-values with 95.65% or 0.9565. With that your maximum, transmittance will be exactly equal to 1 and the rest of the values will be less than 1. This process is called normalization.

It is just for convenience when you are comparing a lot of spectra.

Improve this answer
$\endgroup$
4
  • 1
    $\begingroup$ Upvoted, but it is not clear why the OP first posted at physics SE, accepted an answer and then came here. Modern double beam-in-time instruments are rather good, so double beam-in-space spectrometers are not necessarily preferred for source noise performance. $\endgroup$
    – Ed V
    Dec 24 '21 at 14:32
  • $\begingroup$ Oh thanks so much, it works. $\endgroup$
    – user76255
    Dec 24 '21 at 14:55
  • $\begingroup$ @EdV, I think the OP was confused how to normalize the experimental data. I did not know it was posted twice. $\endgroup$
    – M. Farooq
    Dec 24 '21 at 15:04
  • 1
    $\begingroup$ Yes, that is what I figured. But the HCN IR spectrum was apparently used to calibrate a laser system, based on what I find when searching. In the old days, a pulsed laser might have been used to pump a tunable dye laser and then pulse-to-pulse fluctuations in the pump laser’s output, typically a few percent, had to be normalized out. Missing and runt pulses were especially annoying and experimental data had to be corrected for those. I used to use a window comparator to ensure collection of data from non-defective laser pulses. $\endgroup$
    – Ed V
    Dec 24 '21 at 15:34