Compensate for 13C contrebuting to other compound

[Aldijkhuizen]

After months of development of a method for three compounds with mass 248, 249 and 250 i finally have a method with good retention and peakshape for all compounds. Problem is that compound with mass 249 and 250 coelut.is it oké to substract 10% of the area of 249 from the area of the 250 peak? Anybody do this?I are really out of ideas to further tweak the seperation.

[James_Ball]

You would probably need to inject each analyte separately to see the exact contribution of one against the other to calculate the exact percentage for correction. Then document the results and the reason for the correction factor so that if any questions are asked in the future you can demonstrate why it is done.

[Peter Apps]

If you are doing EI (which sounds unlikely) then there may be another fragment mass you can pick for one of the co-eluters.

Peter

[lmh]

If you need to justify to anyone why the approach should work, you could draw an analogy with UV/Vis spectroscopy, where it is quite normal to estimate several chemicals by measuring the absorbance at several wavelengths and making a calculation based on the differing extinction coefficients of the chemicals at various wavelengths.

You could also point to Agilent's application note on LC-PDA which suggests dealing with coeluting A and B by taking a chromatogram at the absorbance maximum of "A" (X nm), and subtracting a background chromatogram at a wavelength Y chosen so that interfering chemical B has the same absorbance at Y as it does at X.

These are PDA examples of doing what you propose to do by MS.

I agree completely with the suggestion you should measure the contribution of 249 to 250 in a real, pure standard. You also need to bear in mind that this approach will affect your sensitivity in measuring 250 (i.e. if the compound that weighs 249 is absent, you may be able to measure 1 unit of 250 very precisely, but if 1000 units of 249 are present, the 1 unit of 250 is only 1% of the expected peak at 250 and becomes basically impossible to detect, let alone measure with precision).

You also need to keep in mind the underlying resolution of your mass spec; you're very dependent on the quality of the centroiding algorithm! Fortunately, from the little I've looked at, centroiding on all the systems I've tried is remarkably good.