hi everyone!
I am using the turbochrome software of a perkin elmer HPLC, when I select the wavelenghts I want to use, I should also choose the reference wavelenghts...what is their function and what are the criteria to choose them?

thanks
annav

Choice of a proper reference is effective in reducing the baseline drift during gradient elutions. Without a reference, the chromatographic signal drifts upward
or downward due to changes in the mobile phase induced by the gradient. The drift is almost completely eliminated by use of a reference signal. wavelenght should be close to the sample wavelenght and try a bandwith for about 50nm for this wavelenght. but the ref-wavelenght may not overlab the samplewavelenght.

On a DAD, all of the light from the source lamp(s) is directed through the flowcell. By using a reference, changes in the lamp intensity as well as changes in MP absorbance can be negated from the sample signal.

In a variable wavelength detector, the lamp energy is split, so that there a 2 discreet photodiodes for the signal to be interpreted.

Sorry, do you mind telling me why the reference wavelength should be close to sample wavelength.

Thanks so much!!!

jin

You use the reference, to subtract the background at the sample wavelength doing the run, and therefore it should act as much as the sample wavelength.

how close to the sample wavelength should be the ref-wavelength? should it be higher or lower?

my sample has its UVmax at 230 nm, I am using acetonitrile as mobile phase, I choose 210 nm as ref-wavelenght but I can not see any peak.with another sample I had, chlorogenic acid, I used 325 nm and 360 nm as reference wavelength I did not see any peak, when I used 220 as ref-wavelength I had the peak(mobile phase MeOH)...
I am a little bit confused...

Hi
What's also important is the bandwidth for this ref wavelength. If it overlaps the sample wavelength you don't see any peak or maybe just a small peak. If you look at the iso-plot for your sample place you ref wavelength outside the peak, wher you have a low absorbans. I will suggest ref wavelength = 500 at 50 nm bandwidth. You can try to move it closer to the 230 nm, but don’t overlap, so minimum 300 nm.

thanks very much for the suggestion!
I will try in this way!

I don't get it. If you are using a reference wavelength close to the sample wavelength, you are suppressing signal. You said this yourself - if you select the reference wavelength too close to the sample wavelength, the signal goes away. Don't you need to select a reference wavelength way away from the measurement wavelength, preferentially at a wavelength where the signal of the sample is rather small? How small does the signal need to get to maximise signal/noise ratio?

Yes I will suggest that you start far away from the quantification wavelength, look at the S/N ratio, the peak height and the drift, then move it closer to the quantification wavelength and take another look. By doing that, you are optimising ref- wavelength. So you see it a question on how dose your solvent absorb at a high wavelength, I think not much compared to the quantification wavelength, therefore you use a high bandwidth to let in more energy for the photo multiplier. And remember this ref wavelength is the ref point for the measured signal, so every time you have a data point you also have a autozero, and therefore there is no drift from the gradient. On a simple UV-detector you do the autozero at the beginning of the run, with a ref energy point measured from a split light before entering the flowcell, and therefore it can not measure baseline movement caused by the solvent.
Hope you not more confused now!