-
- Posts: 7
- Joined: Fri Feb 24, 2017 12:46 am
- Location: University of Washington
I've got an intractable issue with loss (apparently) of sterol acetates (cholesterol acetate, stigmasterol acetate, etc) when trying to quantify them via GC-FID. I routinely get 60-70% of what I should be observing. I can observe this both when acetylating standard sterols or when measuring a directly made-up cholesterol acetate standard. My measurement of my standards isn't a problem, as I see the same 30% deficit (for cholesterol) when comparing an acetylated sterol standard with a silylated one (i.e. the TMS derivative gives me what I think I should see, the acetate derivative doesn't). This isn't happening to n-alkyl alcohols, and it's not dependent on retention time (later-eluting, non-sterols like archaeol acetate show up just fine), suggesting it's not an injection discrimination problem.
I see a baseline drop after my sterol acetates peak, which does not occur with underivatized sterols or TMS-ethers, or any other class of OH-acetate derivative. The shape of the baseline around the peak would seem to indicate decomposition of the compound on the GC column. However, the problem persists when a brand new column is used, or when a different phase is used (we have used both VF-17 and DB-5 columns, mid-polarity and non-polar, respectively), and there's no obvious reason why these compounds, which are generally quite stable, would be decomposing mid-flight. We've tried reducing the max oven temperature, changing flow rates, varying inlet temperatures, two different GCs, etc. Still keep getting ~70% response. We've tried this with both a PTV inlet and a SSL inlet. We're injecting in toluene. I added a retention gap to reduce interaction with the phase while the compound was 'waiting' to volatilize in the temp program, but this didn't change anything.
Any thoughts? Has anyone ever run into something like this? Or know why an unsaturated sterol ester would decompose on a GC column? I assumed there could be reactions with silanol groups, but the consistency of that that 70% regardless of column age or run parameters is extremely vexing.
I know there are corrections that can be made for the differential FID response of different types of molecule (in this case, a sterol acetate vs my n-alkane response factor standard) for the MOST precise quantification, and I've used simple mass balance corrections or carbon number corrections when comparing acetate and TMS derivatives, but I've never had a problem comparing alkanes and esters to a 1st order approximation before, certainly not to the tune of a 30% difference in response. Thoughts?
I see a baseline drop after my sterol acetates peak, which does not occur with underivatized sterols or TMS-ethers, or any other class of OH-acetate derivative. The shape of the baseline around the peak would seem to indicate decomposition of the compound on the GC column. However, the problem persists when a brand new column is used, or when a different phase is used (we have used both VF-17 and DB-5 columns, mid-polarity and non-polar, respectively), and there's no obvious reason why these compounds, which are generally quite stable, would be decomposing mid-flight. We've tried reducing the max oven temperature, changing flow rates, varying inlet temperatures, two different GCs, etc. Still keep getting ~70% response. We've tried this with both a PTV inlet and a SSL inlet. We're injecting in toluene. I added a retention gap to reduce interaction with the phase while the compound was 'waiting' to volatilize in the temp program, but this didn't change anything.
Any thoughts? Has anyone ever run into something like this? Or know why an unsaturated sterol ester would decompose on a GC column? I assumed there could be reactions with silanol groups, but the consistency of that that 70% regardless of column age or run parameters is extremely vexing.
I know there are corrections that can be made for the differential FID response of different types of molecule (in this case, a sterol acetate vs my n-alkane response factor standard) for the MOST precise quantification, and I've used simple mass balance corrections or carbon number corrections when comparing acetate and TMS derivatives, but I've never had a problem comparing alkanes and esters to a 1st order approximation before, certainly not to the tune of a 30% difference in response. Thoughts?
