blown degasser? dead lamp? air in flow cell? air in system?

[VibekeC]

Dear forum,

In my last HPLC run with a TSK Gel CM-STAT cation exchange column which ran Thursday afternoon and overnight until Friday I have run in to instrument problems.

My first injection is just mobile phase A (10mM sodium phosphate) + 2uL carboxy peptidase.
The first chromatogram looked harmless like previous ones obtained last week and this week (where I am trying to deal with a retention issue, but that is another problem, and not covered in this question).

It is a 55 minute gradient with mobile phase B (10mM sodium phosphate + 100mM sodium chloride)
The method is described in more detail on page 6 here:
https://mc.usp.org/sites/default/files/ ... %200.2.pdf
"Procedure 3: LIMIT OF CHARGE VARIANTS (AFTER CARBOXYPEPTIDASE TREATMENT): CATION EXCHANGE
CHROMATOGRAPHY"

Then in the second chromatogram after approx. 16 minutes (early on the gradient) something goes ballistic…
In a split second it goes from around 0 mAU to 300 mAU and criss-crosses between 250 and 350 for the rest of the analysis.
Is it possible to post a picture of the chromatogram here? Wonder if this will work?


The rest of the sequence is similar intense baseline noise in all analyses.

I have tried the obvious during Friday, took out the column after it was washed through.
Put in a bit of peak tubing and flushed the system with MeOH20:80% first and then 100% MeOH. up to 5mL/min.
No change in baseline. I tried to pull air from the system lines and especially the A line seemed somewhat prone to air.
B also. D line (100% MeOH) and C line (20:80 MeOH:H2O) were both okay. Minimum air.

The lamp does not show any significant change in intensity. I have turned it off and back on again. No change.
I have tried to put back pressure on the flow cell to release possible air bubbles and the 100% MeOH should have helped as well. But there is no change.

The pressure is unchanged. No indication of leaks or blockage. No salt deposits on the syringe or the injection seal.

Any suggestions to explain how such a quick and massive deterioration of the output from the system is possible are very welcome...
Because it happens so suddenly, I was convinced it was “just” an air bubble trapped in the flow cell.
Now I am not so sure any more. Am I missing something? Could it be large amounts of air in the pumping system?
Could the degasser have blown a fuse?
My system is an old but normally quite reliable system.

Help...

[uzman]

Check the detector cable going to your data system ( at both ends ), the cable may be broken.

In order to find the problem , you may also short circuit the detector output signal to see if the signal goes to zero.

[kubowicz.tomasz]

Hello

I'd recommend procedure:

1.Create new method with 0 ml/min (run time 60min) and run blank to see if problem still exists. Don't create it using problematic method as a template - create new method from scratch
2.Make sure you don't have any timetable for detector (wavelength changes during run or lamps on/off)

Regards

Tomasz Kubowicz

[badger]

Sounds like you might have gas bubbles in your detector flow cell (indicative of degasser problems) try adding a piece of narrow diameter stainless steel tubing after your flow cell to keep the back pressure high enough to prevent outgassing.

[VibekeC]

Short update. Thanks for all the creative suggestions for solutions.

It is not my method or the specific assay run which is causing the spikes.
The spikes are still there even if I just turn on the pump and run manual data acquisition.
Same size, but now auto-zeroed to between -50 and 50 mAU.

I have checked all connections for leaks, checked syringe and injection part for bubbles and leaks.
I have degassed mobile phases extra, and the solvent degasser is running as normal, it was just hard to hear because the 20%MeOH:80%water solution I am running through now is very well degassed (plus the radio was on...).

I have double checked all electrical connections and I have a C18 column sitting tight now in the column compartment.
Back pressure is nice and high.
I have a small pressure drop on injection. at a 1 mL /min flow my back pressure is 164 bar and on injection it drops to approx. 152 but quickly climbs back up to a steady 164 again in 3-4 seconds.

I have checked the flow and the purge valve delivers 1 mL/min when running at 1 mL/min, and the same goes for the outlet after the detector.
No bubbles to be seen in the outlet tubing (the only bit which is not peek or stainless steel).

I have changed the lamp, checked the flow cell etc. and I can't see indications of leaks or bubbles anywhere.
I have turned the detector on and off (lamp as well) and it comes back nice and smooth, does its internal spectrum calibration without errors.

Could it be as simple as a blown connection to my detector array? or would that be registered by the instrument during the auto calibration?

I still see the crazy criss cross pattern even when I do data acquisition at a flow rate of 0 mL/min...
Just a bit more tight in the spikes.
If I have gas bubbles in the flow cell, should the 200 mL 100% MeOH I ran through not have gotten rid of it...?
Is Isopropanol more effecient?

I will disconnect the detector today and run manual injection of water and MeOH or Isopropanol (or both) straight into the detector.

Any ideas to solutions are very much appreciated!

[kubowicz.tomasz]

Hello

Could you please paste chromatogram? It'd be handy.
Regards

Tomasz Kubowicz

[VibekeC]

Hope this works,

Otherwise press this link and it takes you to ImageShack where I uploaded the chromatogram to.

[VibekeC]

https://imageshack.com/i/ip6L9DIfj

You have to right click on the picture icon and open in new tap.

I cant seem to get the full picture shown here in line with the txt.

[kubowicz.tomasz]

Hello

Try to run it without cell to see if it is the same. Is it always at the same time?
If you replaced lamp it could be optical unit or mainboard.

Regards

Tomasz Kubowicz

[VibekeC]

No since that specific injection I have only had crazy noise on my baseline.
It got auto zeroed on the next injection, but stayed between -50 and +50 mAU.
I normally have a noise level at 0.05 mAU...

I changed mobile phase today to 80% MeOH and have injected a 0.5% toluene test mix to see if I could get a signal.
The noise is slightly lower now. Between -10 and +10 mAU.
Still catastrophically high though.
Will post a few more chromatograms later today.

If it is optical unit or main board... my detector is to old to fix I fear...

I have a couple of other possible fixes to try. Will keep you posted.
Thanks for all the help!

[VibekeC]

Chromatograms:

This is what it looked like last week when all was normal...
It is a gradient run, so ignore the baseline drift.


This is the injection which contains the sudden abnormal behaviour.


This is a toluene run from today at 214 nm, which shows the continued elevated noise level
but also a "normal" response to a compound toluene


Noise in toluene run at 230 nm:


The toluene peak seen at 260nm:

[ScottHorn]

So you say that this noise is present even with no flow? Is the amplitude of the noise consistent across the entire wavelength range of the detector? Also, what model of detector is it?

[VibekeC]

Yes, the noise is also present at zero flow. more compact, but yet present.
When the flow is turned on the noise spreads out a bit. But with same intensity.

I have tried three different lamps and problem remains.
It is across all wavelengths from 200 - 595. And yes, at same amplitude.
The detector is a Gynkotek 340S.

I inspected the flow cell visually and couldn't see bubbles or dirt, but I am contemplating a manual injection to clean it properly if needed.

Thanks!

[ScottHorn]

Have you tried back-flushing the flow cell? Connect a syringe to the outlet tube, disconnect the inlet tube from the column and push some solvent backwards through the flowcell. You may want to try a few different solvents, but if there's some small particle stuck in there then just about any solvent should do the trick.

[VibekeC]

Thx will try that tomorrow!
Was planning to do it the other way around with a manual injection through the connection after the column, but that is all peek tubing and doing it from the outlet tubing is much easier. And as you say, if something is stuck, it would be better to try to get it back out where it came from...