Chromatography Forum

I have an HP5890series II with dual FIDs, and I have been experiecing a large amount of drift between runs. The variation is not consistant but ranges from 2-15% or higher for duplicate runs of the same standard vial. I have tried two vials to avoid variation caused by solvent loss due to pierced septa and have seen no change. I have changed the needle in the autosamper, cleaned the injection port, changed the liner, and have experimented with different split flows, and temperatures at the injection port. I have swapped the FIDs and am relatively sure this is not a detector problem. This instrument had low drift, so I've been told, I just started a little over a month before. In my experience these instruments should be able to get below 2% drift between runs.
The one thing that I am not sure about is that we are using a splitless liner and running a split injection, but I have been assured that this is the way they have been doing it for years with good results. The FID was recently serviced (prior to me starting here) and the flows were reset to optimal; 10mL/min carrier, 20mL/min makeup, 30mL/min H2, 300mL/min AIR. I do not know what they were before, but the service tech is sure that they were not right. Apparently it has been since then that the repeatability problem has surfaced.
Any suggestions are welcome.

I would definitely use a splitless liner only for your split injections (we like SGE Focus liners). About a decade ago one of our contract manufacturers in another state bought a brand-new HP GC and couldn't get the reproducibility required for our cGMP assay. HP/Agilent had a copy of our test procedure, and was supposed to set up the GC and software for them, exactly as we had written it. To try to remedy this issue, more than one of their local engineers would replace the liner with a new one, but it didn't help. Finally from my location I faxed a page from a catalog to the HP service engineers there and they detailed that they had been replacing with split liners, as that's how it came from the factory (apparently to meet detection limit tests). I told them to install a split liner (which they did have) and the problem was solved forever. These days I would use a digital photo, but that wasn't available then. I'm getting dumber every day, so I have to ask about your carrier flow since you're obviously using capillary column, hence the split injection mode: isn't your carrier flow rate of 10 ml/minute carrier gas really high, if you mean flow through the column? If you mean total carrier flow with a portion through the column and some to the split vent, then I at least understand.

Apparently I do mean the flow through the column. I am new at FID, but I had questioned this, as my experience is with GC/MS and this flow through the column is much higher than anything I have ever worked with. The service tech set it up that way, so I have been struggling with this setting. What is a common column flow for FID using 30mX0.53 columns.

From top of my head: about 4 ml/min carrier for 0.53mm columns. We mostly use narrower-dimension columns here than 0.53mm.

If things were fine before the technician came in and reset the flows I would start by trying to find out what the flows were before and then set them to that value. I would also make sure if the jet was changed the proper jet was used. Just off the top of my head I would say your column flow is too high, the makeup flow is too low, and I suspect that your hydrogen flow is also too low. It has been years since I used an HP, but I seem to remember a hydrogen flow of 35 to 40 mL/min. If your flame is too hydrogen poor it may be unstable.

Was the service tech an Agilent employee or a third party contractor? Many people doing service are contractors and may not be as familiar with the older systems as they should be.

Thank you, the flows did seem to be the problem. I lowered the column flow back to where it was before the technitian changed it and raised the make-up and hydrogen flows. My drift is now consistantly below 1% from shot to shot.