Dear Yifan,

technically I canot see a reason why you should not be able to use it if you can export your data to a software where you can perform it (e.g. Excel). We have been developing the methods some years back for mixture analysis to obtain reasonable normalization in case that one/few compounds have a high degree of variation in concentration. This works for NMR because for this method all signals are proportional to the concentration of the signals which induce the signal. I do not know how this is in IR?

I hope this helps

Alfred

Dear Alfred,

Thanks for your attention! In consideration of your answer, I would like to make further description of my problem. Firstly, I am working on the problem that physical variation (that is, light scattering) blends with high degree concentration variation. Secondly, light scattering in NIR is a multiplicative phenomenon, like xi=ai*zi, where xi is the measure spectrum of sample i, zi is the true spectrum, ai is the coefficient of light scattering. In this case, I am not sure if PNQ can be used for estimating ai?

Yifan

Dear Yifan

I'd guess that "ai" is characterisitic for the substance you investigate. If this is case the only thing that helps is to have reverence data measured with the substance you investigate with a known concentration. If everything works fine the concentration of the same substance at different concentration is simply given by the "ratio of the spectra" which is expected to be identical for all wavenumbers. If the response of the IR is however "nonlinear" (e.g. very high amplitudes "look" smaller than expected) the only thing that helps is to measure spectra of a dilution series. I'm afraid in non of both cases our PQN is of any advantage.

Hope this helps

Kind regards

Alfred

Dear Alfred

It is yes that "ai" is characterisitic for individual sample. For now, we are studying linear system. From your answer, I consider that PNQ may not be advantageous for estimating "ai". But for linear system, there is method that can be used for estimating "ai" without reference spectra, which can be found in the following website.

Article Quantitative Spectroscopic Analysis of Heterogeneous Mixture...

Thank you for your sincere advice!

Yifan