Mysterious Gilson 321 pump failure

[griffith123]

We are experiencing a mysterious problem with our Gilson 321 pump (H2 pump heads). Since we are in Madison, Gilson has been out to work on the unit several times, and we have narrowed the problem down to one user, who needs to run 50 mM potassium phosphate/4 mM ammonium sulfate in pump head A and acetonitrile in pump head B, running a gradient from 0% to 50% B over 45 minutes (the other solvent system that is used on this system is 0.1% TFA/H2O in pump head A and acetonitrile in pump head B, and this always runs flawlessly).

Initially we observed that pump head B would suddenly drop in flow rate for no apparent reason. After trying all the obvious changes without eliminating the problem, we observed solid particles in the pump head B inlet tubing ! This observation, combined with the fact that the 1st run usually works fine and the unit fails in the 2nd run suggest that solvent A is leaking into pump head B and precipitating in the inlet line/check valve/pump head. I have been using HPLC for more than 10 years, including all major brands except for Gilson, and I have never heard of such a thing. What's more, the problem is not limited to our pump. Gilson loaned us a different 321 pump, which experienced the same problem! Now Gilson is claiming that we are simply asking too much from the pump.

I was wondering if anyone else has ever observed this in a Gilson pump, and if so, were you able to solve the problem? Is the Gilson really so limited? Isn't the whole idea of a check valve that liquid only flows in one direction? We have 3 Varian systems that run the same buffer system without any problems. Any ideas?

Byron in Madison

[Paulhurleyuk]

If the check valves are failing, then they would allow solvent in both directions.

Have you tried mixing solvent A and B together first, sot hat instead of going from 0 to 50% B you go from 0 to 100% A and B mixed together. This might produce less precipitation in the mixing valve as their is less of a sudden mixing.

Paul.

[griffith123]

Thanks for the input, Paul. Yes, we have used a mixture of A and B and it fixed the problem, but my post was directed more at the apparently faulty design of Gilson check valves. This problem occurs immediately after installing a new check valve, and we have installed several new ones, so it can't just be one faulty check valve. Again, my main concern is the apparently faulty design. It could be a bad manufacturing lot, but if others have observed the same thing on their Gilson's, then it would be general to the design and not a manufacturing problem.

Byron