Hi all!

I'm experiencing a very strange issue and I was wondering if you could help troubleshoot.

My company produces ethanol for a variety of applications. I have been using a modified version of the USP method for analyzing ethanol quality using GC-FID, which includes quantifying the amount of acetaldehyde in our product. Recently, we installed an activated carbon filter system to further purify our product. The problem is, whenever I run one of these carbon filtered samples the peak response for acetaldehyde (and isovaleraldehyde, which is not present in the samples but is present in the standards) jumps up 2-10x. Any validation standards I run after the carbon filtered sample also show this increase in response for acetaldehyde. Blanks run do not show any signs of carryover. The issue continues after I recalibrate. Running unfiltered ethanol from the same process does not show this change in response. The other compounds in the validation standards do not show this increased response. If I run alternating carbon filtered samples and validation standards, the response in both increases over time. I do not experience this issue on my GCMS, and our other plants have been running carbon filtered material without much issue.

My instrument is an Agilent 6890N (refurbished) with a S/SL inlet and an FID detector. Column is a Restek 624, and I'm running helium as a carrier gas. I can provide method specifics if requested.

Please help!