I'm not really expecting an answer here, but heres an interesting and baffling problem we've been having.........

We are measuring low concentrations of drug in biological samples on a phenyl column. There is lots of interferance, and so the run is a slow gradient to separate out all the substances in the sample.

Over about a year of development, we managed to separate all interference by using a formate buffer mobile phase (weak acid pH) with mecn, P-2-ol and meoh . The meoh% is changed to provide the gradient.Detection is by fluorescence at UV wavelength Excitation and emission.

The assay was working fine till a few weeks ago when all of a sudden a huge baseline drift started occuring after about 10mins into the run (increasing as MeOH% increases) and this occurs even when injecting a mobile phase blank.....

We have changed every possible component in the system to try and get rid of it, but no change... a 50:50 MeOH:H2O wash didnt help either...

I chatted to a chemist and he suspected some impurity in the MeCN.... and suggested we buy some really expensive stuff to test it (£100/L)
(we've not tried this yet)

We also have the same problem using the same conditions on a different HPLC with a different drug, but the drug elutes before the drift starts, so theres no problem there...

Anyone have any ideas, We're totally stumped!

Do you get the drift without injecting anything?

Maybe you are now (not before) inadvertently injecting air? Normally you would get a broad peak, but sometimes air also produces a drift (usually downward, but with a gradient....?).

I am curious about using a gradient to improve sundry interferences. Did your interferences all elute early?

yep , we get the drift with a 0 ul injection. hmm, perhaps if the needle wash was dry and not re-primed.... interesting thought :)

Our interferences eluted both sides of the drug peak, as we had no idea what these interferences were (and too small to analyse on our puny single quad MS!) we had to use trial and error until we shifted them. Changing the pH and proportion of the various solvents in our mobile phase - coupled with good quality HPLC bottled water seemed to do the trick.

Good luck getting rid of the interferences!

Drifting baselines are common while performing a gradient analysis due to the
changing composition of the mobile phase. In other situations, drifting baselines
indicate that a column is still equilibrating or the detector is warming up.
Contamination problems may also be a factor.