Bubbles ...

[HPLC_Help]

Hi, all -

We are running a Waters IC-Pak Ion-Exclusion column for fermentation products. Our mobile phase is 0.01N H2SO4 (degassed), and our detection is RI. Our HPLCs are Shimadzu.

We run at 0.9 mL/min, and after each run (typically in the middle of the night) we reduce the flow to 0.2 mL/min to conserve mobile phase.

Lately we've been seeing lots and lots of bubbles coming out of the HPLC in the morning (and the usual baseline choppiness associated with bubbles) ... BUT only when the flow has been reduced to 0.2. If the flow remains at 0.9, no bubbles are seen. Our analysts increase the flow to 0.9, and the bubbles go away, and the next run proceeds with no problems.

We have degassers on our HPLCs, and both the 0.9 and 0.2 mL/min methods make use of them.

There seems to be a decrease in column longevity associated with this new bubble phenomenon, which makes me think the bubbles may be forming before the column and affecting the packing material (however this is not confirmed - it may just be a bad column, as we've gotten a few over the years). But when we placed the line from the pump in a beaker of water and let it pump at 0.2, we didn't see any bubbles coming out of the line.

When we put the line OUT OF THE COLUMN into a beaker of water, we observed tiny bubbles coming out at 0.2 ml/min, but NOTHING coming out at 0.9 ml/min. This seems to indicate the bubbles are forming in the column? But why wouldn't they come out at 0.9 mL/min? And how would bubbles form in a closed system, where no leaks are observed?

We've seen no leaks.

So far it seems like we haven't been able to conclude anything about the cause of what we're seeing. If anyone has had any similar experiences, or ideas of what to try next, I'd appreciate it!

H_H

[HPLC_Help]

An update ...

In front of our column we have a Shodex SH-G guard column. We found NO bubbles going into the guard column, but MANY bubbles coming OUT of the guard column! So, somehow bubbles are forming inside the guard column (though apparently only at 0.2 mL/min) and then being pumped through the column and the rest of the instrument.

So my question is ... how do bubbles form inside a guard column when no leak is visible? (The column is at 70C, so I guess any leaks would evaporate pretty quickly ...) And why (apparently) only at low flow rates?

H_H

[HPLCCONSULT]

A quick thought. If you have not correctly swaged your fittings to the tubing in your HPLC system, then all of the problems you describe can occur. Poor quality connections, internal mixing chambers and incorrect tubing ID selection can result in bubble formation which is more noticeable at low flow rates.

I routinely see HPLC systems at BIG pharma companies which have improper connections and tubing. These often cause problems with the chromatography which are often overlooked until the system is "pushed" a bit where the problems become more obvious.

*An article on HPLC tubing and fittings can be found here: http://www.lc-ms.com/hplc_tubing_and_fittings.htm

[HW Mueller]

If you have a leak in the sense of liquid coming out you will not get gas into your system while it leaks liquid. Sometimes a Venturi system is created if the fitting is inproperly set (as mentioned by HPLCCONSULT), that draws in air only if you have a flow. In this case I would rather think that your mobile phase is not properly degassed and bubble formation needs a "seed" like a frit, so that bubbles don´t form when the pump is directly pumping against atmospheric pressure.
This is a typical example for having to do systematic checks, without skipping involved parts by making wrong assumptions. (An example of what I mean with the last statement: You assumed maybe wrongly that you don´t have a leak, so you may have wrongly skipped checking the fittings).

[Gerhard Kratz]

I agree with both HPLCCONSULT and HW Mueller. In "High Pressure" LC a leak means solvent is going out, but it is impossible that air/gas gets in. Your tubing ID I guess is 0,5mm. Check if your degasser is working properly first. Use a dead volume free column to column connector between your guard column and the analytical column. Replace the tubing from the column outlet to your detector by a tubing with 0,1mmID, and length should be as short as possible. Check your back pressure, it will increase a little bit. Control the temperature and make sure that 70°C are on the guard and the analytical column, and isolate the tubing to the detector. If you see bubbles in the detector flow cell you can connect a 0,1mmID tubing, 30cm length to the detector outlet as a restrictor. But be careful with the back pressure inside the flow cell. If air is still in your solvent, due to the pressure drop and the large ID of the tubing, at 70°C air is no longer dissolved in your solvent and you see air bubbles. Good luck

[HW Mueller]

Gerhard,
you misunderstood me. I tried to say that one can not have a leak of liquid out and air in at the same time in one fitting. Certainly, I have had a Venturi effect drawing in air at fittings which were not tight enough. In other words it is definately possible to have a leak which transports air into the plumbing at full pressure, as long as there is a flow.