Does ammonium carbonate pH 10 destroy pumps?

[bunnahabhain]

Dear all

We are struggling with dying pumps here at the moment. It all started with an Agilent 1100, mixing 75 % ACN with 25 % ammonium carbonate (50 mM pH 10). This was fine for one hour or two, then we saw big pressure ripples (> 10 bar) soon followed by complete failure: The pistons moved, but no liquid was pumped at all.
I changed the inlet valve because I thought the liquid would flow back instead of forward due to a leak in there => Did not help.
I exchanged the pump seals (they were 2 months old, but anyway) => Pumping worked again, but again only for some hours.
Then we had a service technician here who exchanged the whole pump head => was fine for some hours, same problem again
Then we had the same technician, exchanging the whole pump to a binary model => was fine for some hours, then same problem again
I then transferred the method to an older model (Gynkotek M480). Same thing there. Pressure is stable for about two hours, then failure. At least this pump works fine with methanol/water after purging. The Agilents are all dead now.

So my question of the year: What is it in ammonium carbonate that kills the Agilent pumps? And how can I revive them?

Kind regards
Jörg

[tkubowicz]

Hello

I'm using ACN with Ammonium Carbonate (10-20mM, pH 9) with different ratio and never had any problems with Agilent pumps (binary/quat - 1100/1200/1260/1290)
Try to mix both in glass test probe to see if there is any precipitation (50mM can be too high)

I always use "seal wash" option with buffers (10% IPA in water)

Also check other lines, perhaps there are some traces of previos solvent in system that can cause troubles.

I'd try to flush pump with water (hot can do the trick) and IPA.

Regards

Tomasz Kubowicz

[Gerhard Kratz]

I agree with Tomasz, precipitation might be the problem.
Do you see the same problem when running with MeOH?
Good luck Jörg.

[bunnahabhain]

Dears

I think precipitation is not the problem: There is no turbidity when mixing buffer and acetonitrile or after letting stand over night.
This eluent (pre-mixed) now runs fine on our third HPLC, also Agilent 1100 (I was sweating blodd and tears during the runs).

This is really strange...
Jörg

[James_Ball]

Have you examined the pump pistons from the failed pumps under a microscope?

Look to see if there are micro scratches on them, if so then some type of precipitate has formed at some time and caused the problem. It could be prior damage and now that combined with a little precipitation from current mobile phase is causing premature failure of the seals.

[bunnahabhain]

Thanks James for the suggestion. I will ask the service technician to do that (I am in the at least partially lucky situation that this is a warranty issue).
However, I guess this cannot be the reason for the failing binary pump, because mixing is done after contact with the pistons. Channel A with contact with the buffer died, channel B (acetonitrile) is still OK.
I'll keep you updated.

Jörg