Chromatography Forum

Uwe, I turned it back on...

HW, I did have the system running with the degasser on and it did work fine for quite awhile. However, last Monday the problem with the baseline started. When I contacted the group responsible for developing the assay I was told to turn off the degasser, so I did. I was told that they were surprised it worked so long with the degasser on. But, it's a vacuum degasser, so it should not have mattered. However, I did what I was told and shut it off. Of course, it didn't help any.

I even went so far as to change out the flow cell altogether to try and correct the problem, it didn't. If it were a stuck bubble, wouldn't the problem go away with a new flow cell?

The flow cell that I'm using for the FLD is rated at 10bars, so it's quite fragile. I've asked Waters if there was a more robust flow cell and was told there wasn't. So, I guess I'm stuck with it.

These experiments I mentioned show that a stuck bupple (apparently not visible) has a very high frequency of occurrence (when starting out, or changing the plumbing), but is usually not noticed as the baseline can be smooth and one doesn´t usually have a reference point of baseline level. A better indicator than replacing the cell is to jolt the flow as Mark or I mentioned, but that only works permanently with degassed mobile phase. If repeated jolts don´t lower noise AND the baseline, chances are that your problem is dirt.
So: I would first run the system as you did your last time, then remove the column and run the stuff directly through the detector without changing parameters, then keep going with vacuum degassed mobile phase (any foam will immediatly disappear on going to atm pressure, or if its unusually stubborn, withdraw from the bottom...), then one should cut in the degasser again.....

Thank you, HW, I'll do what you suggest.

And tell us about results?

Well, I went through the process describe to clear out the system. Everything looks better, but I still have the characteristic baseline; it's just much lower than it was before clearing out the system.

There is obviously something else that is causing this problem. There was quite a bit of dirt in the system before I started, but I don't see any now, and I've been running the system for quite awhile now.

OK, one problem out of the way. Does your cleaning solvent look better when pumped directly into the detector as compared to via the column?

The cleaning solvent looks the same pumped either way.

With your "characteristic baseline" or smooth?

Both ways show the "characteristic baseline".

Forgot, did you state somewhere what the cleaning solvent is? Anyway, you would have to have your whole (noncolumn) system horribly gunked or use a totally wrong flushing solvent to get these "mountains" from dirt (do you have spectra of different points on the baseline?), so my guess is that you still have an air problem. If your cleaning solvent is pure organic (MeOH, ACN, or especially higher alcohols) there should be no air problem unless you have a leak which syphons in air. (With air saturated ethanol I have never seen "air noise", only a higher, smooth baseline, as a matter of fact, I have removed air problems from sytems, especially those including the column, by running degassed MeOH through).

The cleaning solvent is 100% MeOH, degassed. I've checked for leaks and have none. In fact, I just had a tech from Waters do a preventive maintenance check on the system.

What happens if you push MeOH through your detector with a syringe directly attached to the detector (careful, though, its difficult not to inject air this way, one has to "fiddle" around a bit)? What I am saying is that such problems are solved by elimination ......
How do you know that there is no leak? Some leaks "suck" in air, rather than loosing mobile phase.

Ok, I pushed MeOH manually through the FL detector, and I got a nice straight baseline with a very sharp sharp peak. I did this 2 or 3 times and got the same results. I, then, pushed MeOH manually through both the UV and FL detectors, as it would be running in the system, and I got the same thing. Again, doing this 2 or 3 times and got the same results.

As far as there being any leaks, how would I check for ones that are sucking in air? I checked every connection I could find and didn't see or hear anything unusual. The Waters tech was here and didn't see any leaks either. Is there anything I could do that would indicate that the HPLC was sucking in air?

Just a thought - but we always used IPA when trying
to get rid of air bubbles from the detector (but I'm not
sure if it will do a better job than MeOH).

Theoretically, iso-propyl should be a bit better than MeOH as air should dissolve better in the former, MeOH is much cheaper.
steve2, you know that you found your leak when the baseline is smooth, keep hunting (the biggest chance for pulling in air is, of course, the suction side of the pump, and especially if "last drop frit" parts protrude above the mobile phase).