Chromatography Forum

I had just made a beautiful calibration a week previously. All my 8260 RSD's were well below 15%. Instrument is an Agilent 6890/5973 with an Archon 4551 autosampler and Tekmar 3000 P&T unit. About a week after doing the calibration I let a sample slip through that had 2000 ppb benzene. So, I had put the trap on bake and hold at 260C under UHP hydrogen flow. I meant to do it for 20 minutes or so. It was late and I forgot to check and left it on bake all night. The next morning I see my mistake as I am loading my first samples. Long story short, only a single analyte is adversely affected... Allyl Chloride has dropped from 40 ppb (1uL of 200 ppm std mix in 5mL of DI water) and is recovering only 28-30 ppb. The other 68 analytes are bang on and have been in every CCV and MS/MSD for over a week. I'm reluctant to do a new trap and calibration to take care of this. Particularly since most of my samples are just BTEX/TVH/MTBE so that allyl chloride is not even a target.

Any thoughts on what in a trap gets burnt out so that only allyl chloride is affected? I don't suppose there is something I can do to bring back that analyte?

The Supelco cat#24920-U Purge/Trap K Vocarb 3000 contains the following...
Carbopack B 10cm/Carboxen-1000, 6cm/Carboxen-1001, 1cm

Typical use is desorb at 250C (I use 235C) Bake at 260C and condition at 270C.

You may get by with just recalibrating and not replacing the trap. If you used hydrogen, it could have cleaned out some contaminates that were hanging around covering active sites on the tubing and now allyl chloride is being adsorbed by them. Run it for awhile and see if the response increases over time, if so then it is definitely active sites causing the problem.

Helium and Nitrogen flows tend to allow heavy compounds to collect and cap off active sites in the plumbing, but hydrogen will clean out the junk exposing those sites.

If there is still residual hydrogen stuck in your trap it can be reacting with the allyl chloride. I should go away with time.

Yes, I have my lab plumbed for Argon because I have an abundance from keeping a dewar for ICP-MS use. So, to save Helium, I use Argon for purging 11 minutes as per method then I dry purge for 1 minute using UHP Hydrogen and bake out is 8 minutes using hydrogen. I do the dry purge with hydrogen to reduce the amount of Argon that gets into my 5973 MS because Argon quenches MS response and being easier to ionize than Helium; it also messes with fragmentation mass ratios.

[Edit]: Essentially, the Archon is plumbed with Argon (or Nitrogen at the moment) from a dewar and the Tekmar 3000 is plumbed with Hydrogen. The GC-MS is plumbed with Helium.[/edit]

I've been operating sucessfully with these conditions for many years. I was actually using hydrogen as my carrier gas for a few years until I had a power failure that caused the 6890 EPC to back flow water and pump oil vapor (power failure makes the EPC fail open) which irreversibly contaminated the system. Even after a thorough cleaning and swapping in a different EPC and weldment. So no more hydrogen carrier for me and went back to helium.

With that background explained. It sounds like the Tekmar 3000 bake out path got thoroughly cleaned by the hydrogen and that pathway which is exposed during desorb is presenting active sites that are getting the vinyl chloride. And you think that is the best explanation and not that the trap contents were damaged by the bake out?

Its good to know I can probably eliminate the trap as a suspect. I think perhaps I can blame the variability of my dibromomethane surrogate recovery at the same door. Dibromomethane recovers within limits on CC and dirty samples but is recovering at the very low QC limits occasionally failing in method blanks and clean samples.

I agree with James_Ball, run a new calibration and move on.