Chromatography Forum

Does anyone have experience using dual wavelength detection for assay and impurities determination? Have you ever compared the area of an impurity at one wavelength with the area of the parent compound at another wavelength (using a relative response factor)? Obviously, one impurity has poor absorbance at the lambda max of the main analyte. Any commentary would be helpful.

just cause you use dual wavelength, you don't need dual asking...

(viewtopic.php?t=13509)

Hi Lanon,

As the energy intensity of detector lamps varys with the wavelength, I’m not completely convinced that a Relative Response Factor would be universally valid i.e. independent of system, time etc.
Maybe, if you include a daily check for (and potential correction of) the RRF (could be using the SST sample) you’ll be able to justify such a procedure. Validation of the procedure would/should be mandatory.

Best Regards

Dual wavelength is o.k. We use it a lot. RRFs have to be determined for each impurity (clean impurity standards are available?). As RRFs are often cut down to two decimal places (pharmacopoiea says: no RRFs applied between 0.8 and 1.2), I wouldn't care about different lamp intensities, however check recovery on different instruments (PDA, VWD, maybe old+new lamp).

Alex

What´s this business about the lamp intensity? HPLC is calibration dependent, and recalibration or checks needs to be done ever so often. So if you use two different wavelength for two different substances one calibrates as always.
(Was "dual wavelengths" used to describe the method in which two chromatograms at two diff. wavelength were recorded before?).

Hi Hans,

It is a case of area percentage, relative to the main peak. So, no need for genuine calibration here. Never the less, there will be needed some kind of kontrol of the the ratio, as mentioned earlier. And reworded by Alex Thanks for that.

Best Regards

Call it check or calibration, if you properly set up the analysis the lamp intensity differences at different lambda is compensated like everything else (provided differences are constant).

Exactly