Pressure becomes unstable during long runs

[dkreller]

Hi,

I have an older Waters 600 solvent delivery system. I He sparge my aqueous mobile phase throughout the experiments. My experimental mobile phase flow rate is 1.0 mL/min.

Everytime I start a new experiment and/or change out the mobile phase reservoir I am careful to ensure that there are no gas bubbles in the lines or pump. I draw several mL through the system, and then I run the pump, w/ the column bypassed, for a few minutes at 10.0 mL/min until I am sure there are no bubbles in the pump.

I start my mobile phase flowing through the system w/ the column in place. We pack our own columns (for geochemical research purposes) w/ quartz sand w/ relatively large particle sizes and the backpressures are not especially high, around 100-300 psi.

Everything is looking good, the pressure is stable, and so I start the experiment, which has 90 injections total. However the next morning when my student and I look at the data, it is apparent that the pressure and the flow became unstable during the night. My hopethesis is that bubbles got into the pump system.

I have heard from many people that this type of Waters system really needs higher pressures, to be sure that the check valves are working properly. My next move may involve doing something to make the backpressures higher. I have a 75 psi backpressure regulator from Upchurch but I haven't yet installed it. An instrumental technician also suggest that we could try to put a long section of line inbetween the pump and the autosampler. The idea is that the long piece of line would contribute to backpressure but not add to axial broadening of sample.

Any input, respected Chromatography forum community?

Thanks,

David

[tom jupille]

Actually, my recollection of those old Waters systems is that they had spring-loaded outlet check valves, and so didn't need a lot of back pressure to operate. My gut feeling is that a couple of hundred psi should be sufficient back pressure to keep the pump happy. On the other hand, a meter or so of 0.010" id tubing can't hurt, and since it would be upstream of your injector, have no effect on performance.

My recollection is also that once an air bubble got stuck, they were a bear to get re-primed.

One other possibile culprit: teflon is somewhat permeable to air, so you may be getting some air diffusing back into your solvent between the (degassed) reservoir and the pump. Look for evidence of "phantom bubbles" in the line just before it enters the pump (you shouldn't see any). If you do, there are a number of "in line degassers", aka "membrane degassers" available from the major chromatography suppliers. These go in line immediately before the pump and so minimize this problem.

[dkreller]

Thanks for your input. I will share this reply with my instrumental support person.

David

[Noser222]

The other option is you just make your student stay all night.

[dkreller]

That wouldn't solve anything though. The problem is that this is an undergraduate student. She wouldn't know what to do if the system pressure started to go all over the place. She might call her supervisor in the middle of the night and then it would be apparent that he doesn't know what to do either.

David

[HW Mueller]

You already have a temporary (until you find the cause) solution: Start over again by running high flow through the pump, etc. Your undergraduate should be able to handle that.
It should not be forgotten that He does not degas, but rather exchanges air with He which has a lower solubility than the gases in air, but it is present. If sparged carefully it probably does not outgas, but if you draw in some micro-He-bubbles which might be plentiful in the reservoir if sparging is vigorous, I can immagine pump outgassing.
On the other hand, sparging might stop over night?

[dkreller]

TJ and HWM,

Tom, you said that vacuum (membrane) degassers are commercially available. I remember using a Waters Alliance system a few years ago while post-doc'ing and if memory serves me right all that system had was a vacuum degassing unit. A couple of questions. i) In your experience, are the vacuum degassing units more effective than the He sparging systems? and ii) can you imagine retaining the He sparging and adding a vacuum degassing unit to the system and running both at the same time?

This has been such a major and ongoing issue for me - causing around 40% of my experiments to go down the drain - that I might approach this by doing both i) adding a vacuum degassing unit and ii) adding some length of small ID peek tubing ahead between the pump and autosampler to boost backpressures.

Also, wrt using small diameter tubing to boost backpressures, do you see any advantages of putting the length of line b/w the pump and autosampler, vs putting the new length of tubing after the detectors?

Thanks,

Dave

[tom jupille]

As HW Mueller pointed out, He sparge works because helium is only sparingly soluble (much less that nitrogen or oxygen). As you pointed out, most modern systems use in-line degassing and dispense with the helium sparge entirely; with an in-line degasser, helium sparge really doesn't contribute anything. So:

i) In your experience, are the vacuum degassing units more effective than the He sparging systems?

Yes

and ii) can you imagine retaining the He sparging and adding a vacuum degassing unit to the system and running both at the same time?

I can *imagine* it, but I wouldn't do it

do you see any advantages of putting the length of line b/w the pump and autosampler, vs putting the new length of tubing after the detectors?

You have to be very careful to not have too much pressure on the detector cell or you can start having problems with leakage at the cell itself. Adding a capillary restrictor after the detector is a fairly common trick for preventing outgassing at the column outlet. Since the purpose in your case is to minimize pump problems, I'd put the restrictor between the pump and the injector just to be on the safe side.

[dkreller]

Thanks Tom,

Okay, I think you have helped me determine the steps I am going to take. I am going to put a length of small ID tubing b/w the pump and injector, and see if the experimental problems have been alleviated. If that fixes it then that will support the hypothesis that the pressure instability problems were indeed associated with lowish backpressures.

However, if the problems persist even after this change, then I will think seriously about changing over from the He sparge system to a membrane degassing unit.

David

[khw]

I have seen pressure fluctuations in my system when I was not careful enough about my aqueous phase and had algae growing which spoiled my check valves (Agilent 1100 hardware). I assume you checked this, but I just wanted to mention it.

[dkreller]

Dear KHW,

No I haven't checked for this. It has been many months since I have flushed the system. Good idea. Disassemble the pump and check for biofilm on the check valves, I assume.

[Bruce Hamilton]

Assuming the above added restrictor and cleaning solutions fix your problem and all will be well, the following is unnecessary..

However, if you still have problems, be aware that the the solvents also contain significant amounts of dissolved gases, and if you can premix ( if isocratic ), or partially mix components ( if gradient eg A =80:20, B =20:80 ) you may also minimise bubble formation problems.

Be aware that temperature variations can affect gas solubility, if the solvents warm whilst passing through the column the gas solubility reduces.

Another solution may be to split your long run into two shorter runs, with a suitable flushing method in between.

Please keep having fun,

Bruce Hamilton

[HW Mueller]

This is a typical case of where a careful analysis of result and cause is necessary. Your info, that flushing in the beginning, etc., results in good pumping speaks against a microbe film as cause (how do you explain that the same mobile phase removes and deposits microbes?). "Remote" analysis of the problem has led several people here to suspect air. Thus a systematic check on air seems to be most promising. For instance, I would make sure that the plumbing ahead of the pump is tight, then I would vacuum degas the mobile phase, even risking a change of mobile phase composition (unimportant here, since one is checking pump performance). If pump instability set in I would vacuum degas again . . .maybe even using pure solvents, closely observing behavior continuously.

[juddc]

Have you checked / changed / removed the filter stones on the ends of your solvent lines?

What you've described is classic bahavior for a 600 with either clogged filter stones or insufficient degassing. My advice would be the following: Put NEW filter stones on the solvent and sparge lines then draw degassed methanol through the lines and pump that for a few minutes at an elevated rate, rinse with degassed water, then put your filtered and degassed mobnile phase on. There should be no need to have both vacuum degassing and He sparging on the same system. 100-300psi should de fine for a 600, but if you want more, why not put a guard column between your pump and injector?

An outside possiblility would be your gradient proportionaing valve, but that would only give problems if you're using it to mix solvents on-line. I have seen those fail, but it's rare - I've replaced perhaps 2-3 in the last 16 years.

Good luck!

[dkreller]

I replaced the filter stone on the solvent draw line for the reservoir/line in question and indeed, the bubbles that were forming in that line stopped appearing.

That is a good suggestion, i.e. placing a guard column b/w the pump and injector. I've never bought one. Any recommdations on guard column brand, length, inner diameter, packing material composition, packing material particle size distribution?

Thanks,

David