Bad Baseline

[bisnettrj2]

Has anyone seen a baseline do this before? I'm running an Agilent 1100 with an MWD detector, G1312 binary pump, G1322 degasser, G1329 autosampler with thermostat, thermostatted column compartment at 30 degrees C. Ran a calibration overnight two days ago, phenomenon did not manifest. Running samples and blanks (shown chromatogram is a sample, but blanks show same problems; peak at ~15 minutes is a surrogate), instrument is exhibiting very sudden rises and dips in the baseline. Mobile phase is 52:48 water and methanol, mixed by pump post-degasser, using a Dionex Acclaim E2 250x4.6 5 micron column, monitoring 214 nm and 254 nm, no reference wavelength.



I'm not sure what to point a finger at here, as before I ran a calibration on this instrument, it passed the MWD self-test, Cell Test, Filter Test, Pressure Test, Leak Test, column compartment temperature test, Wavelength calibration test, etc. This instrument, as far as I could tell, was tip-top. This morning I switched water sources (was running store-bought purified water, switched to water from our recently serviced Millipore unit, still have the baseline issue). Degassing shouldn't be an issue, as the degasser is a recently refurbished (by Agilent) unit. Even so, I switched the channel in which I have my water being degassed, to know avail. I also flushed the flow cell for 5 minutes at 4 mL/min with water and then 52:48 water:methanol in case it was an air bubble or other contamination. This instrument is a recent acquisition, so I don't know its particular history, but I'm running two other instruments with the same mobile phase, and one with the same type of column but a DAD detector, and I don't have this issue.

[danko]

It looks like a pumping problem - unless it’s some kind of adient of course. But it’s not a gradient method – or is it? Did you collect the pressure data? If so, do you see presser changes, coinciding with the baseline fluctuations?

Btw. it doesn’t look like an air bubble in the flow-cell issue. But if it was the case, it isn’t water one should use to remove the air, but rather some organic solvent i.e. methanol or acetonitrile for instance.

Best Regards

[grzesiek]

If you adjusted some older method, check if the gradient program is not hidden somewhere, it has happend to me and my collegues

[bisnettrj2]

Ok, this morning I pulled the flow cell, collected data without the flow cell in place - nice, flat baseline, so I don't think it's a lamp/detector issue.

I'm running isocratic, and I modified the method from my similar diode array instrument running the same type of column.

The pressure profile doesn't show anything to explain these sudden dips and rises.

When I flushed the flow cell yesterday, I used methanol, and then my mobile phase.

Also, I omitted in my first post that I am running 1.0 mL/min for a 62-minute runtime.

Here are a couple shots of a bad baseline and the associated pressure profile:

[HW Mueller]

Why would the pressure go up and down at the beginning and end, respectively?

[bisnettrj2]

Pressure goes down when the injection valve goes to bypass to fill the injection loop, goes back up to normal pressure once the injection is made. The pressure drops at the end because after my analytical run, I go from 52:48 H2O:MeOH to 100% MeOH to wash the column. My pump parameters look like:

Time - 0 - 52:48 H2O:MeOH
Time - 59 - 52:48 H2O:MeOH
Time - 63.6 - 0:100 H2O:MeOH
Time - 68.2 - 0:100 H2O:MeOH
Time - 72.8 - 52:48 H2O:MeOH
Total runtime - 82.8 minutes (for actual sample analysis - standards and blanks get run on an isocratic 62 minute run).

The first chromatogram I posted was from an isocratic run.

[grzesiek]

how about using "new" flow cell?
is your instrument close to any other electrical device that can cause a problem?
Anything changes when using reference wavelength?

[bisnettrj2]

The instrument is near my other HPLCs, but they don't show the same issues.

I'm running a no-injection run with the flow cell in the detector (after re-flushing the flow cell with acetonitrile at 5 mL/min for 10 minutes) to monitor the baseline.

I'll add a reference wavelength in the next no-injection blank run to compare the two.

Sadly, I do not have another MWD flow cell in-house, so I can't switch to a new one right now.

[danko]

It seems to me, the flow goes down to zero at about 60 min and then goes up again upon injection.
Btw. The pressure seems to be too high (when the flow’s up) with this column at 1 mL/min.

Best Regards

[bisnettrj2]

No, you're seeing a pressure profile that goes off-scale (drops to about 90 bar pumping Methanol at 1 mL/min). Flow rate stays constant throughout the anlysis.

Also, this isn't a C-18 column, it's a specialty column for explosives, and my typical backpressure is 190-210 bar at 1 mL/min for this 250x4.6 5 micron column.

I tried a reference wavelength, problem was still there. I decided to go for broke and pull the lamp and replace it with a new one. I'll be shooting a sequence overnight if my calibration check hits. If all is resolved, I'm blaming the lamp.

[Uwe Neue]

Did you do the no-injection runs already?

[bisnettrj2]

Yeah, that's what I meant about the reference wavelength not solving the issue. The variations in the baseline were smaller, though still present. The day is late, and since all problems happen to me on Friday afternoons, I've resigned myself to a Saturday morning visit to the lab to see if swapping lamps helped. If not, I'll resort to a sacrifice to the various chromatography gods. Thanks for the ideas!

[danko]

Hi Bisnettrj2,

Please let me guess; exchanging the lamp didn’t help. Right?

Best Regards

[bisnettrj2]

Yep. didn't help. Back in the lab Saturday at 4 AM. Re-injecting samples that showed the baseline problem, using a reference wavelength for these analytical runs. Will re-do no-injection runs this afternoon, both with and without reference. I sure do love a mystery....

[Uwe Neue]

What happens if you eliminate the 100% wash at the end of the run?