Acetonitrile and HPLC gradients

[davida]

Over the last couple of years we have been using an increasing number of HPLC gradient analyses using acetonitrile as the organic component. Many of these gradients ramp up to 90% to 100% acetonitrile before returning to initial conditions. The problem we are having is the check valves on the low pressure gradient pumps are failing regularly and require thorough cleaning in propanol before becoming useable again. It has been stated by the HPLC manufacturer that using acetonitrile at concentrations above 90% leaves a sticky residue on the ball of the valves that renders them inoperable. To that end they supply a replacement type of valve with a ceramic ball to minimise the problem but even these require regular cleaning in propanol. We are using Merck Gradient grade acetonitrile as this has the lowest background absorbance of the several brands we have trialled. Does anyone know what the problem is and how to fix it?

[Jackus]

What pump (manufacturer) you use?

[AdrianF]

Are you using a buffer? Some buffer salts are insoluble in acetonitrile - phosphate is the worst. Precipitation of salts will cause check valve problems.

Consider not going above 85% acetonitrile and see if that improves matters.

[Celiagg]

I think you must prove the solubility of your buffer in % of ACN that you rrise at the end of the gradient. If salt precipitation occurs, you will see an increase of preassure in the system.

The polymeric pieces of the valves may be solved in organic solvents. You must buy seals of an addecuate material that resist acetonitryle.

Regards

[Jackus]

Are you using a buffer? Some buffer salts are insoluble in acetonitrile - phosphate is the worst. Precipitation of salts will cause check valve problems.

Consider not going above 85% acetonitrile and see if that improves matters.

According to my experience the worst is borate buffer. We are using phosphate quite a lot but problems with valves started when we used borate.

[DR]

If the problem is not ppt related, see if you can switch to ceramic ball and seat cartridge check valves. If your LC is a common one, Optimize Technologies and others probably make a kit for it.

[Mark Tracy]

If the problem is precipitated buffer, forcing DI water through the pump with a syringe usually is enough to restore it.

[davida]

Thanks for your quick replies. The problem is evident when no buffers are used in gradients going from say 30% ACN to 100% ACN. Several injection runs can be made before the problem arises causing the pump to lose pressure and stop the sytem. Tempoary cure is to sonicate valves in propanol for 10 minutes.
I came across this in another post:
[We´ve had problems with pumping 100% ACN; after some time the valves in the pumps didn´t work any more. The representive asked us to filter our ACN using 0,22 µm-Filters. This made no sense to us, because the solvent was already prefiltered in the same way by Merck (ACN gradient grade quality).
We now use a mixture of 10% water and 90 % ACN to clean the columns. Its works perfectly.
There´s also no problem using gradient elution. ]
We will try the ceramic valves and see how we get on.

[Bruce Hamilton]

Unlikely, but the problem may in your water, is it of suitable quality and residue-free?.

My problems with pure acetonitrile affecting valves usually happen only to my systems if the valves are stored in it overnight or longer. I've never seen failures during operation, and I would suspect contamination from somewhere.

I'd also be looking to see if the CH3CN is leaching something from the solvent system, as the Merck gradient grade has worked well for me, except on an Agilent 1100 preparative system, which seems to have check valves with sensitivity to long term storage in pure acetonitrile.
Adding 10% of water to acetonitrile fixed the problem for me.

The issues about check valves and acetonitrile are well documented, especially in US Patent 5002662, where the authors investigated the deposits, and is freely available via Goggle Patents ( a superior format ), or the USPTO.

Bruce Hamilton

[Mark Tracy]

I have seen 100% MeCN cause check valves to fail. These were ruby/sapphire with a PEEK housing. The MeCN was B&J's UV/HPLC grade. I observed two problems. One was a short-term loss of function and simply adding 2% H2O prevented it. The second was a failure after long exposure to 100% MeCN, and the cure was simply to set the check valve cartridge on the bench for a week and let it dry out. I offer no hypotheses to explain the observations.

[Patrik Appelblad]

Have you ever considered that it might be originating from the piston seals.
Even if you are using piston seals that is marked for organic solvents, I suspect that it is depletion and it enters into your check valves.
This effect is fast when using 100% ACN, but not so prominent at 90%.

Long time ago, we purchased a few "100% biocompatible pumps", as we had to feed very labile components issolved in dried ACN (100%) in to a flow system. After a few months, one pump collapsed, and some time later the other one. The PEEK housing and other plastic stuff were swollen and there were leakage into the housing etc. We had to throw them away.

[davida]

I have had a reply from Shimadzu the makers of one brand of HPLC we use and they supply ceramic check valves for use with 100% acetonitrile. So do Dionex who say that 100% Acetonitrile does not wet ruby ball valves and so they stick. We will try some ceramic valves and see what happens. I will post the results when I have them.

[XiZ]

I have had a reply from Shimadzu the makers of one brand of HPLC we use and they supply ceramic check valves for use with 100% acetonitrile. So do Dionex who say that 100% Acetonitrile does not wet ruby ball valves and so they stick. We will try some ceramic valves and see what happens. I will post the results when I have them.

Please DAVIDA, post your results with your new ceramic valves. I'm with the same problem and my HPLC's are from Shimadzu too, but I'm afraid to ask new valves to my boss and they doesn't resolve my problem.

[DR]

Optimize does make cartridge kits for Shimadzu pumps - in ceramic.

[Russ]

We have Shimadzu pumps and do not have problems with check valves with AN. Since we use multiple mobile phases, we let the system mix the proper phase for us so one of the pumps is always 100 % AN. We have the standard ruby ball check valves and can go over a year without any problems with the valves on that pump. We do filter our AN daily through a 0.22 um nylon membrane filter and even though it is supposed to be filtered into the bottle, we occasionally see "garbage" on the filter. We also discard (or use for sample preps) any AN that has been in the solvent reservoir after one week.