Using A Wavelength that is on a Slope (of UV Spectra)

[Jimi]

Hello

This is a classic question, and has probably been discussed here in the forum at some point over the years. But I would like to discuss the concern about working at a wavelength (in a UV detector) that is not on a fairly flat part of the spectra (i.e. working on a slope).

Now the obvious answer is that there may be somewhat more variability in response.

However, what I am getting hung up on are other issues. Specifically, the fact that the linearity is usually not as good and the intercept is generally larger, in these cases. But it is not clear to me why this should be.

Thanks in advance for any insight.

[DR]

Think about the spectrum... Picture it.

The spectrum is a measure of absorbance over a range of wavelengths. Think about taking a measurement in a flat spot at or near a max. If you move left or right a nanometer, the absorbance doesn't change much, if at all.
If, however, you do the same thing where there is a steep slope, your absorbance can change a lot by moving a nanometer in either direction.

As a function of having moving parts, UV/Vis detectors do not stop at nor do they remain perfectly at whatever wavelength they're set to. Wavelength accuracy specifications are typically ±1 or 2nm. As a result, one should expect that precision with respect to area counts is a function of the steepness of the UV spectrum at a given wavelength.

[Jimi]

Thanks DR

Any thoughts on the second part of the question. I have re-stated it directly below for convenience.

Why (when working on the slope) does the linearity tend to be not as good, and the intercept is generally larger. It is not clear to me why this should be.

Thanks in advance

[DR]

Uncertainty increases because any change in wavelength read on a sloped portion of the spectrum will result in a change in the resulting value more than if reading in a flatter part.

To the extent variability increases, there may be a change in the slope and intercept, especially if errors in wavelength readings are not uniformly distributed around the true wavelength.

All you can tell for certain without assessing the magnitude and bias of errors is that precision will be worse when you're reading from a sloped portion of the spectrum.

I can't tell you why the slope and intercept would be higher or lower as a function of the slope of the spectrum where you are taking readings, but the measure of linearity as expressed by r or r² is tied to variability.