(This question is forwarded by John Dolan for further comment by other participants)

I have an HP 1090 system that I have had for many years (~11). As you are aware they periodically leak oil into the mobile phase due to a micro crack in the gold faced diaphragm of the high pressure pump. This one has many thousands of hours of use. We use it 24-7. The question is how can I tell in advance when the diaphragm needs to be changed without waiting for oil in my mobile phase? I can change the diaphragm every 3-4 months but ....... Is there an easy way to check? I normally will collect 10 ml when I think
there may be a problem,and look for oil. But, I would like a more analytical approach. What about GC-MS analysis of the mobile phase every
week or two? Even then, I could have a leak just after the last check. What about putting a small guard column inline before the injector that
would bind the oil strongly? Any suggestions would be appreciated. We have been getting along with this for many years but the newer detectors
and methods we are developing do not allow for even micro amounts of oil.

Although I don't have direct experience with this problem, I seem to recall that the oil is spiked with a strongly fluorescent / UV-absorbing dye that will give dramatic baseline offsets with the slightest leak. If this is the case, a check of the background (before autozero) signal. Once the oil leaks, I would assume that your column will be toast and that the whole system will require flushing with organic (methylene chloride?) and perhaps repassivation with dilute nitric acid. I would check the service manual for advice on restoring the system to normal operation.

The 1090 booster pump diaphragm should last 1-2 years during typical operation if installed properly. In your case, I would suspect 9-12 months would be reasonable. I would change the diaphragm on a regular basis such as 6 months. Flushing the system with IPA, then hexane then IPA, for several hours should remove trace amounts of oil.

As for evidence of oil leaks, usually the background absorbance in the uv range increase little by little during the gradient. I have heard of some customers analyzing eluent in a cuvette via a spectrophotometer.

When replacing the diaphragm, remember to use new screws and use a torque wrench set to 9 NM (newton meters).

Hope this helps.