-
- Posts: 19
- Joined: Mon Feb 06, 2012 9:53 pm
Long time lurker, great people here. I hope someone can drop me some knowledge! I've recently had to deal with an issue that I don't usually need to, as I am attempting a quantitation by LC/MS without the use of an isotopically labeled internal standard. I am using external calibration with the same, unlabeled analyte (despite informing my client that I'm not crazy about this approach).
However, this is becoming a problem, as I am unable to get reproducible peak areas over time. Specifically, this is not random but drifts in a general direction over the course of hours, but immediate effects can be seem in back to back injections. Sometimes, the drift goes toward increased peak area and some days the trend is reduced peak area. Still other days, the precision is very tight, ~ 1% RSD. Area of the final injection can be 200% or more of the initial injection, happening over the course of several hours. I have noticed drift in peak areas in the past in unrelated assays with entirely different chromatographic setup while using internal standards, but this was never an issue I had to deal with since the area ratios remained tight.
So, this issue seems independent of my current method, but the method causing me to beat my head against the wall utilizes HILIC chromatography with an ammonium formate (pH 3.0) buffer. The column was flushed with strong mobile phase for a long time and the ion source cleaned. I am having this issue with only 50 ng/mL purified standards before even trying samples, so matrix suppression seems unlikely to me (but who knows?).
I was thinking that maybe it has to do with an unstable electrospray, but as I can tell the capillary looks like it is good shape, and indeed the instrument was requalified by the vendor recently.
So, I guess I'm just wondering from your experience if this kind of signal drift is pretty typical with LC/MS, or if I might have a real problem here. Any advice on where to look to improve it or an explanation of the mechanism of this would be very helpful. I am confident that this is not the result of injector precision, sample instability, or carryover.
Thank you!
dja