by
HbJ » Thu Mar 29, 2007 6:05 pm
Please note: I'm a GC guy and I have no experience with LC but as your request seems to be very urgent I think it's okay to just share my thoughts.
Whenever I have to do quantification I take a calibration curve of my analyte to determine the linearity the detector has for my compound. That way I found e.g. that my TCD gives a too high response for very small amounts of water in solvents.
Generally speaking, quantification is a difficult task:
If your detector gives you an (about) equal response for every compound of your sample (including the solvent) and the detector was not saturated (i.e. the main peak(s) are cut off) then just go for the 100 percent method:
The percentage your peak of interest has from the total area of your chromatogram (i.e. the added area of all of your peaks, baseline-corrected) is the percentage of your analyte in the sample.
If your detector was saturated for the other components, height is the way to go assuming your peak of interest is not cut off. That is an older technique coming from the days when integrators were rare (and unreliable) devices so the determination of the area was difficult (cutting out the 'peaks' from the paper trail and weighting them on the analytical balance was one of the more arcane area determination techniques) while measuring the height of a peak just takes a ruler.
The most simple approach (which I found to be quite robust) is the addition method: After you've analyzed your sample you spike it with an additional amount of the compound of interest. Assuming your detector response is linear, your detector doesn't get to full scale etc. you can (somewhat) safely determine the content of your component.
No matter which way you prefer a calibration curve is the way to go for a sensible quantification IMHO.
If you don't have time (or access to the instrument) check your previous runs if you've chromatograms of samples which contained a known amount of lutein.
