Pump loses prime !!

[kate_m]

Hey everybody

I'm using a quat pump and at the same point in my method the pump suddenly loses prime. All 4 lines are full although only using A&B and i purge every morning. I've never had this problem before and no alterations have been made to the method.

It happens with ~90% ACN (B). (A) is a very weak buffer (10 mM). The pressure suddenly drops. it has happened twice now and always after a couple of injections in the sequence.

Any ideas?

thanks!

[Consumer Products Guy]

Are you purging all four channels even though you use only two??? We did.

[uzman]

Sonicate the checkvalves in acetone , then water , it may be a checkvalve problem.

[James_Ball]

I have also seen the proportioning solenoids go bad and not open all the way and cause a cavitation air bubble in the solvent line which can cause drop in pump pressure. Maybe try and switch to C & D lines and see if the same thing happens. If it goes away then one of the solenoids is bad or the needle is sticking.

[kate_m]

Hi everyone thanks for the replies.

i always purge all 4 lines before starting. again over the weekend the LC stopped at the same point. always high ACN %.

I'm pretty sure it is the check inlet valve. if i slightly loosen and allow liquid to run through just by static pressure, the bubble escapes and the pump works fine again.

then a few injections and the same problem again.

i am going to try sonicating my check inlet to see if this solves anything.

kate

[DR]

::Assuming we're talking about Waters 2695::

Another option is to switch from the default ruby/sapphire to ceramic cartridges. They are far less prone to "floating" due to surface imperfections that arise when the ruby/sapphire version are left sitting in ACN.

[kate_m]

its an agilent 1260 series.

i'm using the G7111B quat pump. anyone have any idea what type of check valve this is? (active/passive?) in ref to the inlet.

it's still happening. now on 2 different methods, but at a similar point i.e high acetonitrile concentration.

each time, the only way to remove the bubble (that is stuck in the check valve) is to slightly unscrew and allow mobile phase to trickle through, bringing the bubble with it. i'm concerned if i continue to do this i will damage the thread on the pump head/check valve itself....

[tom jupille]

That's an Agilent pump, and it should have an active check valve (to confirm, look to see if there's a honkin' big solenoid attached). Active valves are much less prone to bubble problems compared to passive check valves, which suggests that the valve is defective and needs to be replaced. Before doing that, there are a few other things to check:
- go over all connections in the inlet line to make sure they are tight and leak-free. Teflon tubing can "cold-flow" over time and result in a leaky connection. On the intake side of the pump, a leak will result in air being aspirated.
- while you'e checking the inlet line, look carefully for any signs of air / vapor bubbles
- does your system have an in-line degasser? if it does, check to make sure that it is functioning OK. If it doesn't, you might consider getting one (probably cheaper to replace the active inlet valve, however)
- If your gradient starts with a non-zero strong solvent concentration (e.g., if the range is something like 20 - 60%), then you might try pre-mixing the weak mobile phase; in the example I just gave, the "A" solvent would be 20% ACN and the gradient would run from 0 - 40%

[James_Ball]

That's an Agilent pump, and it should have an active check valve (to confirm, look to see if there's a honkin' big solenoid attached). Active valves are much less prone to bubble problems compared to passive check valves, which suggests that the valve is defective and needs to be replaced. Before doing that, there are a few other things to check:
- go over all connections in the inlet line to make sure they are tight and leak-free. Teflon tubing can "cold-flow" over time and result in a leaky connection. On the intake side of the pump, a leak will result in air being aspirated.
- while you'e checking the inlet line, look carefully for any signs of air / vapor bubbles
- does your system have an in-line degasser? if it does, check to make sure that it is functioning OK. If it doesn't, you might consider getting one (probably cheaper to replace the active inlet valve, however)
- If your gradient starts with a non-zero strong solvent concentration (e.g., if the range is something like 20 - 60%), then you might try pre-mixing the weak mobile phase; in the example I just gave, the "A" solvent would be 20% ACN and the gradient would run from 0 - 40%

To add, the active inlet on the Agilent has a small cartridge inside the solenoid that you will see if you unscrew it completely. That is a consumable and does need to be replaced from time to time. They can become clogged or wear enough that it doesn't open correctly or doesn't close and seal or I have also had them fail and begin to leak, but it is an easy replacement and not too expensive.

If you observe the instrument when it is near the high ACN point in the gradient, do you ever see any bubbles in the tubing from the 4 way proportioning solenoids to the active inlet at the pump? If you ever see bubbles there then it could be the ACN solenoid on the proportioning valve is clogged or not opening as it should.

[Multidimensional]

Some info that may help u. If you have an Agilent G7111B Quat pump (1260 Series II), then most likely, your pump Does NOT have any ACTIVE INLET valve ([b]AIV)[/b], just a conventional gravity valve (prone to sticking and failure). Awhile back, Agilent's team made a very bad marketing decision to start selling their near bullet-proof reliable HPLC pumps without the Active Inlet Valve as standard, on some models, and have customers pay even more money to buy it as an option. Since the AIV feature is what makes these pumps (since 1992) super reliable and rarely ever have priming or check valve problems, their decision to do this was profound (and companies like Waters and Shimadzu are still thanking them for their decision). Why anyone would delete this valve from the basic pump configuration to save a few hundred dollars and reduce the reliability (?) can only be answered by Agilent. It is well know in the industry by insiders as 'THE' monumental marketing blunder of the last ten years.

There is good news though. Agilent still sells versions which INCLUDE the AIV as standard in the pump (the only way you should buy one). They also offer a "KIT" to add the AIV to systems that are sold without it (everything is there to plug it back in). Agilent P.N. G5699A. You can buy their kit or just buy the AIV, cartridge and new connection line to the MCGV for much less too. Please purchase and install the AIV in your pump and provided you take care of your HPLC system, you will never have these issues again!

Additionally, to troubleshoot any type of HPLC or FPLC loss of prime issue, here is a link to a free article with troubleshooting help:

"Tips and Advice for Priming your HPLC PUMP (or similar pumps, FPLC or UHPLC Pump)", https://hplctips.blogspot.com/2020/09/tips-and-advice-for-priming-your-hplc.html

[Hollow]

Is the "sticky air bbubble" proven or just an assumption?
If the air is real, then your degasser may be insufficient/malfunctioning or there is some other source of air along your connections. As Tom suggested.
Maybe you can try to degas the mobile phases manually before you put them on the system and see if it gets better. If it does, check the degasser.

If it's "just" a sticky check valve, have a look at the various threads abou this topic and the article from J.W. Dolan from 2008:
https://www.lcresources.com/lcr/2008/06 ... tonitrile/

Maybe you can try to sonicate the cv-cartridge for some minutes in methanol or even in nitric acid (ca. 6 mol/l) (! Be careful and don't mix nitric acid with any organic solvent! Flush everrthing with plenty of water before and after the treatment! Also check the chemical compatibilityof your cv.)
If nothing helps, it may be the time (and cheaper; take into account the time spent/wasted...) to replace the cv.

[kate_m]

thank you all so much for your replies!!!

its definitely a bubble. the pressure profile reflects it and its difficult to re-prime the pump - the liquid in the waste line moves to and fro but in the same place. either pumping IPA or slightly unscrewing the check valve removes it.

I haven't observe the bubbles coming from the prportioing valve that is something i am going to test...

However - another related question. i thought that if the case is a problem with the proportioning valve (and to eliminate a mixing issue between high % organic and aq.) i am running my gradient with both channel A and B filled with ACN and a restriction after the injector (column disconnected). in theory - the pressure profile should run smooth across the entire run right? as the solvent is the same in both channels... Or should there be some small pressure differences due to the delay of the proportioning valve?

i ask this because my profile isn't smooth across the entire run. even some steps in places... always lining nicely to the B channel (which is usually my organic channel in a normal run). so maybe my theory is right!

as for premixing my mobile phase - this is not a problem i have ever had before,. it suddenly started with no changes to method or eluents... so why suddenly now? I'm not doing method development or anything, kinda just routine analysis...


thanks a bunch!! kate

[James_Ball]

thank you all so much for your replies!!!

its definitely a bubble. the pressure profile reflects it and its difficult to re-prime the pump - the liquid in the waste line moves to and fro but in the same place. either pumping IPA or slightly unscrewing the check valve removes it.

I haven't observe the bubbles coming from the prportioing valve that is something i am going to test...

However - another related question. i thought that if the case is a problem with the proportioning valve (and to eliminate a mixing issue between high % organic and aq.) i am running my gradient with both channel A and B filled with ACN and a restriction after the injector (column disconnected). in theory - the pressure profile should run smooth across the entire run right? as the solvent is the same in both channels... Or should there be some small pressure differences due to the delay of the proportioning valve?

i ask this because my profile isn't smooth across the entire run. even some steps in places... always lining nicely to the B channel (which is usually my organic channel in a normal run). so maybe my theory is right!

as for premixing my mobile phase - this is not a problem i have ever had before,. it suddenly started with no changes to method or eluents... so why suddenly now? I'm not doing method development or anything, kinda just routine analysis...


thanks a bunch!! kate

This makes it seem like the proportioning valves are the problem. They should simply switch between the two channels when doing solvent mixing with them and if it is ACN in both channels then the pump shouldn't see but one solvent coming in always at the same percentage (100% in this case).

If a solenoid is not opening as it should it can cause cavitation in the solvent, which draws out trapped gasses that form a bubble(or possibly vaporizes some of the solvent to make the bubble). You can also have this happen if the filter frit is clogged but the degasser should remove any bubbles that form because of that. I have seen frits on the aqueous channel become clogged with algae to the point that you have a series of bubbles in the line between the bottle and the degasser, those are signs someone has not checked that frit in a long time.

Can you try the same thing you just did but using channel A and C or A and D or even C and D and see if it goes away? That would tell you if B is the problem.

[Multidimensional]

This should be really easy to troubleshoot. If you think that one of the four channels is the problem, then simply start by running a step gradient (vs some organic with 0.1 or 1% max Acetone in solution). Set the UV to monitor the change. Run each channel, A, B, C against channel D to start with (Usually channel 'D' is the least used so in best condition). Look for any anomalies in the steps (Use the restriction cap, no column). If one of the channels appears to be not working correctly, flush out all channels with fresh IPA (100%, one-at-a-time). Run at 0.5 mL/min, 100% A for 20 minutes; Same with B, C and D (or just the channel that is suspect if you want a quick answer. Then re-run the step gradient tests. Filling and flushing IPA through the MCGV sometimes re-hydrates the solenoid seals allowing them to function properly. They can get fouled up from dirty mobile phase.

If you have not done so already, make sure you replace your PFTE frit to (EVERY month as PM. They are $5 each!). A dirty PTFE frit may cause baseline problems, but not normally failure to prime a channel (still, it should be checked). Loss of prime is covered in the article.

Has the pump been professionally serviced ? Was a leak test done? What was the pressure drop in 2 minutes? Bad piston seals can also contribute to loss of prime.

A damaged or contaminated degasser can also cause all kinds of problems with flow and composition (similar to a bad check valve). Clogged solvent pickup frits can also do the same. They require cleaning/replacement per PM too. Many of these items are covered in the article I previously referenced.

If you had the AIV in the pump, then you would probably never have any issues with liquid alternating in the outlet check valve. That is what happens when a 'regular' check valve gets stuck OR the other items that I noted occur.