Integrate individual components of a co-eluted peak?

[JMAC_1986]

Hello,

I am analizing cuticular hydrocarbons of insects by GC/MS-EI. I have identified all of them (methylated long-chain hydrocarbons) by their characteristic diagnostic fragment ions. My problem is that some of them co-eluted in a only peak.

Assuming that all these co-eluted methylated hydrocarbons ionize in an analogous way, my question: could their individual contribution (area) be estimated by performing: i) the summation (Xcalibur software) of all the ions of the co-eluted peak, ii) calculating the ratio between the different characteristic fragment ions of each component, and iii) using this ratio to estimate the individual area of each component with respect to the total area of the peak?

If this approximate option is possible, do you know any reference (books or scientific papers) where they have used it?

Thanks.

[Steve Reimer]

You may be able to find some deconvolution software for Xcaliber.
Caveat: I don't have any experience with that software.

[Multidimensional]

While software does exist to deconvolute peaks, the scientifically correct thing to do is to develop a method that corrects resolves ALL peaks apart. You can not properly integrate co-eluted peaks. Never assume that different samples will show the same response as others you have analyzed. Any actual experience running real samples will quickly convince you of the truth of this. This is one reason why we use standards and orthogonal methods for all samples, to avoid making huge mistakes in ID and also quant. Good chromatography follows the scientific method (and good fundamentals).

[lmh]

Interestingly, in a related field, Shimadzu's latest LC PDA software has a closely related feature. Given a partial coelution of two peaks (i.e. one misshapen peak with no hope of integrating, but the two underlying compounds not with exactly the same retention time) it will deconvolute the two, providing independent spectra for each component, and also independent chromatograms. Of course PDA data are safer than MS data, because it's reasonably rare for a mix of two chemicals to give anything other than the sum of the absorbances, and this general principle has been used since the birth of spectrophotometry in cuvettes. But it can be done in MS too.