Chromatography Forum

I'm having a problem with the Agilent 1200 HPLC of my LCMS. Over the last couple of weeks, the retention times have gotten really long (almost doubled!) for some peptide standards I use for proteomics. I found a slow leak in the pump that I usually used for the organic solvent of my HPLC buffer system (the leaking solvent was kind of evaporating before it could trip the HPLC's leak detector), but taking care of that didn't really change anything. I made up fresh HPLC buffers, no improvement. I switched columns, no change (both columns were C18, 2.1X100 mm, 3 um). The pressures have actually been lower than usual. This morning I started measuring flow rates before and after the column: before the column (HPLC buffers going thru binary pump and autosampler), I got the expected flow rate (collected flow for 10 min. @ flow rate of 200 uL/min., measured volume of 2 mL), but after the column (same time and flow rate), I initially got 1.65 mL, then later in the day when I ran this again I got 1.75 mL. (Is this relatively small drop in flow rate enough to explain such a large increase in Rt?) The weird thing is I got the same drop in flow rate for BOTH pumps in the binary pumping system (I measure the flow rate for each pump separately). I was kind of expecting to see that the pump I usually use for the organic solvent (which is what elutes the peptide from the hydrophobic column) was underpreforming, but they both were! Is there something obvious I'm missing here?
Thanks for any suggestions!

Sean

The pressures have actually been lower than usual.

Anytime you see an unexpected decrease in pressure your first thought should be "what is leaking". Anytime you see an unexpected increase in pressure, your first thought should be "what is clogged?"

There are a lot more places in your system where a leak could have occurred that you didn't mention checking. Do you have a venting valve on the outlet of your pump? Check that when this is closed no fluid is leaking to the waste line of this valve. I had to fix one of mine recently after I realized there was a bypass occurring causing my RTs to jump around every injection.

Are you mixing mobile phase before hand or having the pump do it? If your proportioning valve isn't working properly the mobile phase composition you tell the machine won't be what it actually mixes.

You can also form leaks at the rotor seal on the autosampler valve, or on an Agilent the syringe is essentially just a pump piston and seal mounted in the autosampler which has to have the seal changed over time as well. You can get leaks in both of these places that would cause lower flow once the backpressure of the column is added. Tiny leaks in these spots don't show up until you take them apart for maintenance. Also your outlet check valve could be leaking and allowing mobile phase to leak back into the pump as the piston makes another draw. Just a few more places to look.

I have encountered a similar problem that was traced to a worn out rotor seal in the autosampler. The wear was slight so the leak only manifested itself at high backpressure. I think that is a key point here, that the leak is quite small so it only shows up at high pressure.

A leak test that I like to do is to plug the column outlet, build up some pressure (400 bar maybe), and then turn the pump off. Now measure how long it takes for the pressure to drop. On a leak-tight system it should take several minutes for the pressure to drop to 300 bar.

it may be a good idea to do the complete preventative maintenance, since you have at least two problem spots (the initial small leak and the new unknown leak).

Leaks are smart, can hide anywhere.

I've also seen leaks of Agilent purge valve where MP comes out the waste line, autosampler rotor valve, autosampler loop itself (broken weld from repetitive use), leak at needle seal.....so pretty much anywhere !!!

The other half of your question was whether small leaks can cause such drastic changes in retention time. In a gradient method run on a binary pump system, yes, smallish leaks can cause quite serious changes of retention time because they mess up the composition of the solvent.
You checked your pumps separately (with the same solvent, or different?) and found the same leak on each - so in theory, the composition shouldn't have changed, but it's obviously very unreliable to depend on a fault behaving consistently. Leaks are not only good at hiding, and finding new places to happen, but they have a talent for changing in ways that make life as confusing as possible.