Chromatography Forum

Hello!
I have used Agilent HPLC systems for five years. I am now using Waters systems for the first time and am unable to figure out how to use a reference wavelength as a was able to use with the Agilent systems. The only way I seem to be able to reduce my gradient baseline drift with Waters is to run a separate blank injection and subtract this from my sample chromatogram. I liked using a reference wavelength for only one injection. It saved time and was more accurate. Is the reference wavelength option an Agilent feature only?
Thanks!!!

Perhaps you could offer some more information regarding your mobile phase composition and gradient formation. Instrument models may also prove useful for specific responses. The idea is that others who observe a similar situation may be assisted should a resolution to your problem be found.

Here are a few initial links to follow.

http://www.sepsci.com/chromforum/viewtopic.php?t=7215
http://www.sepsci.com/chromforum/viewtopic.php?t=7081

Just thought I would give this thread a bump.

I have been looking for years for a reference that offers a good discussion of the topic of reference wavelength and the effect of slit width on sensitivity.

Has anyone come across such a thing.

Thanks

Reference wavelength issues (how to use) should be covered in the user's guides to the instruments. In terms of advantages/disadvantages, the short answer is "it depends".

The bad news is that a reference wavelength adds noise to the signal (essentially, the instrument must deal with two measurements instead of one). The good news is that it cancels certain types of noise (notably lamp fluctuations and, to a certain extent, RI effects). If those latter types of noise are large, then use of a reference wavelength reduces total noise. If those latter types of noise are small, then use of a reference wavelength hurts. As a minor issue, use of a reference wavelength can generate negative peaks if the sample has components which absorb at that wavelength.

Slit width issues are usually covered more in texts/articles on spectroscopy. Basically, the greater the intensity of the light getting to the photodetector, the less the relative noise (signal goes up as a linear function of intensity; noise increases roughly by the square root of the intensity). In a spectrophotometer, you can use a tall slit and then focus the light onto the photodetector to get an intense signal. That's not possible in a variable-wavelength HPLC detector, because you need to minimize cell volume. Using a wider slit (accepting a broader wavelength range) can compensate. Since UV spectra in solution have little structure, there is not much loss of information, but you do have to be careful. If you are detecting an an absorbance maximum, a wider slit takes in proportionally more signal of lower absorbance.

With a PDA, the situation is a bit more complex, because there is no physical slit width adjustment. It's done in software by combining measurements from adjacent photodiodes. The end result is similar, though because random electronic noise is unlikely to be correlated on all the diodes; hence averaging the signal tends to reduce the noise. As with the variable wavelength detector, if you are monitoring at an absorbance maximum, a wider wavelength range will take in more off-maximum light and lower the response, so there is a compromise to be made.

Hi Tammy,

I just realized that we all forgot to answer your initial question.
Waters detectors don’t have this feature (i.e. reference beam going through the flow cell)
These detectors utilize a reference beam of course, but it goes around the flow cell and at the same wavelength as the detection beam (the one that goes through the flow cell). The role of this beam is to serve as a reference i.e. “no light absorbanceâ€

There is one problem when you are working in GLP. Subtraction of blank baseline from sample baseline is not generally recomended and you may have problems with auditor.