Negative Peaks and Mobile Phase Degassing

[Eric Moore]

Hi -

We're running SEC with a KCl/KPhosphate mobile phase, pH 7. The mobile phase is filtered through a nylon 0.45um filter.

In the initial run, a negative peak at the void volume appears in our blank and sample injections. When the mobile phase (and samples) sat for a day, the peak went away.

We're thinking the mobile phase degassed overnight, and that's why it's different. Our plan is to sonicate the mobile phase under vacuum and see what happens. But I have a few questions:

What type of contaminant would go away with degassing?
Why didn't the filtration of the mobile phase provide enough degassing?
Should I be sonicating all my mobile phases?

Thank you.

Eric

[skunked_once]

Eric,

It is normal practice to degas the mobile phase during preparation and the mobile phase would probably reabsorb gas over time. Do you degas the mobile phase when you prepare it?

I suspect that the cause of the negative peak is due to the difference in absorbance (what wavelength are you using?) between the solvent in your blank and sample and the absorance of the mobile phase.

Maybe others can propose how the negative peak dissapears after time.

[Eric Moore]

Thank you, skunked_once.

The wavelength is 280. We typically have subscribed to the "filtration will provide sufficient degassing" theory, since we also have built-in degassers on our HPLCs.

Eric

[DR]

Degassing: Primary goal - to prevent formation of bubbles in the LC, particulary in the detector or in check valves
Secondary goal: To lower background absorbance of UV by O2 in the mobile phase - this only comes into play at wavelengths well below 280nm.

If all you have is a negative peak at the void, attribute it to a RI difference between your (presumably) degassed MP and your (presumably gas saturated) sample diluent and go on about your business, ignoring it, and anything else at the void voume.

If you have positive one day, negative the next slopes over the course of your gradient runs, and have trouble with integration because of it, you may want to look into it further. In this case, vigorous initial and a continuous trickle of He to displace the air from your MP may be helpful.

[tom jupille]

I agree with DR that it's probably an RI difference (I know, it's a UV detector, but they do respond somewhat to RI!).

The catch is that Eric is doing SEC, where you can't ignore t0 noise!

Ironically, the solution in this case might be to not degas, so that the dissolved air content in the mobile phase and sample are the same. Since it's an isocratic separation, formation of air bubbles should be much less of an issue than it would be in a gradient.

[danko]

Eric,
When you say void volume, do you mean exclusion volume or total permeation volume?
In other words; what is the retention in time units?

Best Regards

[Eric Moore]

Thanks, everyone.

Danko -

The negative peak is at about the RT of the smallest molecule on the column (or, to put it another way, at the last peak in the chromatogram.) This is the problem - it interferes (slightly) with the integration of the monomer (main) peak.

I've called that the void volume before, but maybe it is the exclusion volume?

Also - the negative peak is there when mobile phase is injected, (pipetted from the container into an hplc vial) so it doesn't seem to be due to a difference between sample and mobile phase gas content, as in this case both sample and mobile phase are degassed.

Eric

[danko]

I've called that the void volume before, but maybe it is the exclusion volume

The “exclusion volumeâ€

[Uwe Neue]

At 280 nm, you would not see oxygen. I suspect that your mobile phase is slightly "dirty", or that your detector is sensitive to RI changes (as pointed out by DR). To test for a dirty mobile phase, make an independent sample of the same composition as the mobile phase and inject it as a blank.

[Stryder08]

We have had this problem before. As Uwe stated, look for something in your mobile phase. Soap/surfactant in a mobile phase bottle will do it, even if it is a very tiny residue.

We've experienced this negative peak phenomenon with a "dirty" phosphate buffer and a contaminated organic solvent.

Order a brand new glass bottle. Make your solvents/mobile phases up fresh. Is the peak still there?

[danko]

What are you talking about guys? Eric told us that he took some mobile phase directly form the MP reservoir and injected it!

Uwe, of course you can “seeâ€

[Eric Moore]

Danko -

We'll try the experiments you suggested. However, I just re-read the thread, and I had forgotten about this part:

When the mobile phase (and samples) sat for a day, the peak went away.

And although the negative peak appears when injecting the mobile phase, it doesn’t mean that air is out of the picture. In fact it is not at all unusual that small air bubbles are introduced in the flow under the process of injection.

So, that would suggest it wouldn't be bubbles from the injector, as they would still be present after the mobile phase had sat. Same with dirty mobile phase or surfactants from residual soap (I think.) This is what makes me think it may be de-gassing related.

Thank you very much for all your responses, everyone. Keep them coming. I'll let you know what we find out.

Eric

[Eric Moore]

Well, the plot thickens. And forgive me for turning this into the "SEC Negative Peak Blog."

We noticed a leak in the system that had been running this method and switched to a new system. The negative peak was gone. This makes me think that the negative peak was somehow leak/air related.

Eric

[tom jupille]

And forgive me for turning this into the "SEC Negative Peak Blog."

No forgiveness needed. It's great to get feedback and find out how a problem worked out. What's really frustrating is to have the thread dropped and never know the outcome!