Solutions for a interfering (co-eluting) contaminant.

[LabRat010101]

We are using a dual-column GC-ECD to detect halogenated contaminants in drinking water.

We suspect that a contaminant in the sample matrix is co-eluting with an analyte due to a consistent discrepancy between the results of the primary and confirmation columns, where the primary column result is always higher. The contamination is not observed in standards.

Is it possible to use some technique, such as standard addition, to quantify or confirm the presence of a matrix contaminant co-eluting with an analyte in the primary column? I think this isn't possible due to the fact that we can't obtain sample without analyte or the contaminant.

An alternative would be to attempt to resolve the contaminant from the analyte, but short of selecting a new column, I'm not sure how to go about doing that.

Perhaps GC-MS could be used to distinguish the contaminant from the analyte? Can the MS distinguish spectra of two different molecules that co-elute?

[James_Ball]

If you have GC/MS and the analyte/contaminate have a high enough concentration you should be able to separate them unless they are isomers. If you can inject a standard and see the analyte, then inject the sample and see if the spectra are different.

[LabRat010101]

If you have GC/MS and the analyte/contaminate have a high enough concentration you should be able to separate them unless they are isomers. If you can inject a standard and see the analyte, then inject the sample and see if the spectra are different.

I'll have to develop a method for the GCMS, which I am new at. Any considerations for method development or protecting the GCMS column from junk? I imagine that I will start with a split at around 50:1, a standard oven program, then inject a sample and try to figure out which peaks are which.

The GCMS is equipped with a DB-624 column (20 x 0.18). The samples are extracts in MTBE that I would consider "dirty".