Page 1 of 1

Non-linearity When Working on the Slope of the UV Spectra

Posted: Wed May 16, 2018 11:20 pm
by adam
Hello

This seems like a basic question, but I cannot quite figure out what is causing it.

When doing HPLC with a UV/Vis detector, we often set the wavelength to a flat part of the curve (often the maxima). However, there are cases where we intentionally chose a less sensitive wavelength (for example, when we have multiple peaks and one has too much absorbance, and so would be off scale). But what I have found is that when using a detection wavelength that is on the slope of the spectra - and not on a flat part of the spectra - it sometimes causes a non-linearity in the response, or results in a significant intercept.

Now you may be thinking: There could be some slight drift in the position of the diffraction grating. And certainly that could cause it. But we are seeing this phenomenon with a Diode Array detector and, therefore, this cannot be the explanation.

Thanks in advance for any assistance. I would like to get a handle on why this is happening.

Sincerely
Adam

Re: Non-linearity When Working on the Slope of the UV Spectra

Posted: Fri May 18, 2018 9:18 am
by Khanom
Dear Adam,

What bandwidth you are using on your diode array detector.

Re: Non-linearity When Working on the Slope of the UV Spectra

Posted: Fri May 18, 2018 10:56 am
by HPLC chemist
Be sure you have 'Turned Off' the reference wavelength (minimum wavelength absorption). A slight drift in this area can cause non-linearity.

Re: Non-linearity When Working on the Slope of the UV Spectra

Posted: Fri May 18, 2018 12:32 pm
by lmh
could it also be that you're only doing this with things that give very high absorbances, to get them into a "linear range", but maybe the linear range isn't as high as it should be, and therefore all of these things are coming out non-linear, while the things that are present at low abundance (and therefore that you measure at a peak-apex in the spectrum) are actually a lower absorbance and therefore have better linearity?

Re: Non-linearity When Working on the Slope of the UV Spectra

Posted: Fri May 18, 2018 2:39 pm
by Consumer Products Guy
adam wrote:
When doing HPLC with a UV/Vis detector, we often set the wavelength to a flat part of the curve (often the maxima). However, there are cases where we intentionally chose a less sensitive wavelength (for example, when we have multiple peaks and one has too much absorbance, and so would be off scale).


Wouldn't it be easier to dilute before injecting or to inject smaller volume?

We never used a "slope" wavelength because the absorbance at the maxima was too large, but occasionally did when there was a matrix interference at maxima wavelength but none/considerably less on the slope. Just saying that we didn't do that, don't know if that part is relevant to you.

Re: Non-linearity When Working on the Slope of the UV Spectra

Posted: Tue May 22, 2018 7:39 am
by bunnahabhain
Maybe what you are seeing is related to the fact that bands in UV spectra are broadening when concentration increases.