Illuminating Spectra Part 8 - Is the cup empty?
Extending on the example used in the previous post (Part 7 - watch the baseline) you may be tempted to guess what will happen to the NIR spectra if the sample cup is empty!
Well as there is no sample in the cup to interact with the NIR light, all the available light goes through the sample cup and almost none will reflect back to the detector. In theory, absorbance levels should go through the roof and approach infinity (as the reflection approaches 0%) because the detector struggles to see in the dark, but in practice absorbance is never infinity and the detector will sense some voltage due to electronic noise and/or stray light from optics. Therefore, there will always be an absorbance ceiling (also known as stray light ceiling).
In Figure 1 below, the red lines are the spectra of an empty cup. As you can see, the noise level in the spectra is increased dramatically when the cup is empty. It is always the case that when the detector is deprived of light (due to a highly absorptive sample or no sample at all) the level of noise in the spectra increases. This can have a negative impact on the accuracy and precision of NIR predictions. As the level of sample in the cup increases, more light gets reflected back and subsequently the baseline is pushed down, the noise level is reduced, and reproducibility of the NIR spectra improves in parallel.
Figure 1 - Duplicate NIR scans of a whole wheat sample in a cup with varying fill heights
So... how full should the cup be for an optimum NIR measurement?
Well, it depends on the type of sample and the depth of your cup! But you can determine this for yourself by scanning cups of varying filling heights and see when the baseline lowers and noise is minimised. But to be on the safe side, fill up the cup or, if you don't have enough sample, switch to a smaller cup with the same depth.
Please bear in mind that the above description does not apply to transmission NIR (or NIT)... we will discuss that another day.