wavelength dependence of UV detector RI response

[tom jupille]

This carries on a discussion started in another thread (viewtopic.php?f=1&t=16138).

Refraction varies inversely with wavelength (blue light bends more than red light), but UV detector RI effects seem more pronounced at longer wavelength, as illustrated by this figure for the "mid-gradient-hump" in an acetonitrile/water gradient:



This is taken from an old Burdick & Jackson catalog. As far as I know, no column was connected.

Any suggestions as to why?

[HW Mueller]

Maybe the detector was not compensated for different light intensities at different wavelengths put out by the light source?

[danko]

Hi Tom,

First thought I get is: Too much variation. I wish these runs were repeated at least twice. Is it a linear gradient by the way? Not that it matters, but it illustrates the randomness of the baseline fluctuation.

Any way, if the shown data are reliable (would be if there are exact replicates of each wavelenght) there could be some issues with the light source, hence Hans' speculation. Besides, an Y-scale would've been nice since we're considering the signal intensity.

Best Regards

[Mike H.]

The first thing that I thought of is that the less energetic wavelengths are slowed down more when they pass into the medium. Is this close?

Edit: I just looked at a normalized pressure/abs overlay tested at 200nm and noticed that the two were close to mirror images. Does pressure play a role in the slope of gradients?

[danko]

Do I understand you correctly – i.e. the place where the baseline drops the pressure rises?
If so, then salt precipitation (salt staying in the column anyway) is more than likely.

Normally (if nothing wrong) increasing the acetonitrile will mean pressure decrease, because ACN is more compressible than water.

Pressure over the column does not influence the baseline, as the pressure in the flow-cell (where the absorbace is measured) is roughly constant.

Best Regards

[tom jupille]

Water/ACN and water/MeOH mixtures are non-ideal (in the thermodynamic sense that the properties of the mixture are not the average of the properties of the pure components). Viscosity, refractive index, boiling point, solubility of air, and even total volume are non-linear functions of composition. Pressure increases because the viscosity of water/ACN mixtures can be higher than that of either water or acetonitrile individually.

What I'm trying to get my head around is the wavelength dependence of the RI effect. The example I showed was typical of what I've experienced over the years, but at the moment I don't have access to a lot of examples where the same blank gradient was plotted out at different wavelengths, with a slightly out-of-algnment flow cell!