Purge and trap problems

[marie0213]

Hello,

I have so much problems with the purge and trap these past months and I can't get help from anyone so I'm taking a chance here. I'm analyzing volatiles in water on a purge and trap (trap #9) connected to a GCMS. I'm using the P&T to add my IS and surrogates and to do my calibration curve. Everything was working fine for 3 years. Now, I can't do a curve from 1 to 20 ppb. I tried changing the trap, manually doing the curve myself, adding the IS myself, injecting using the water method and the soil method because they have a different flow path... Nothing works, I get a parabola as a calibration curve for the less volatile compounds. The other day, I was rushed and only injected a calibration curve without adding IS nor surrogates. Everything was perfect. I did multiple tries and it seems that the IS/surrogates affect my analytes responses... I injected a curve without IS and one with IS one after the other and both were fine.. I'm thinking maybe I have an active site but can't determine where and why it only affects my IS/surrogates.?? Any clues?? Thank you very much for your time. Marie

[James_Ball]

When you do a calibration with the IS/Surrogates, are the peak areas consistent in all standards or do they increase with increasing standard concentration?

What concentration are the Internal Standards?

When was the last time you changed the column and cleaned the source or changed the vacuum pump oil?

Also, what model is the purge and trap and the GC/MS.

[marie0213]

Hi,

I noticed, when everything was working properly, that the area of my IS and surrogate increased as the concentration of my analytes increased. Now, it seems that the area is stable for std 1 ppb, 2 ppb and 5 ppb and sometimes 10 ppb. The 20 ppb is generally lower but excluding this point doesnt give me a proper calibration curve. The concentration of the IS and surrogates is 10 ppb. First thing I did was to clean the ion source, cut the column and even change to another one without any change... The pump we use has no oil, not sure how it is called. The purge and trap is an Atomx from Teledyne Tekmar, the GC is a 7890B and the MS a 5977A.
Thanks, Marie

[James_Ball]

What is the split ratio at the inlet of the GC? With that instrument and those concentrations you should be 60:1 up to 120:1 and still be able to have acceptable sensitivity. If too much water is transferred at lower splits it can cause similar problems.

If any samples have foamed during purge, they can contaminate the trap, plumbing and even the transfer line to the GC. You may need a new Silcosteel transfer line if that has happened.

I don't know if the 5977 behaves the same as the 7000, but if you go into tune/manual tune does the Emission Current read zero with the filament turned off? I had a problem with a 7000 where the filament was reading 25uA when it was turned off, and when turned on the filament would barely even glow because it was applying current to make 35uA but with the offset the actual current was 10uA and caused a very weak ionization to occur. Changing filaments helped the problem.

Also, if you haven't already done it, use a flow meter to verify that the purge gas flow rate is correct, if it is too fast or too slow it could be causing a discrimination in purge efficiency between analytes.

One last thing you can try is to increase the concentration of the internal standards 5x and run another curve, it might possibly load up any active sites enough to minimize the effects on the calibration.

[Steve Reimer]

The symptoms you describe are ones I see after running some of my nasty samples. Usual fix is to clean out everything along the way, especially the 6 port valve in the concentrator and transfer line along with the source and head of the column.

[marie0213]

Hi Steve,

How do you clean it?? You inject methanol or a special solution or you change everything to new pieces??

Thank you

[Steve Reimer]

Cleaning depends upon the piece and the contaminant. The transfer line from the concentrator to the GC is flushed with water while heated to 100 degrees. If the contaminant is creosote, I will do a methanol flush first, at room temp.(no fires please!) The 6 port in my OI 4660 can be included in that flush, switch the valve back and forth a couple of times part way through the steam clean.
The tubing from the sparge vessel to the 6 port valve can either be flushed with methanol in place or removed and soaked depending on how hard it is to move. Inlet and column maintenance should be obvious and of course change the trap.
Heavy accumulations will make it to your source, I find that usually the repeller cruds up first and just needs a rinse rather than a scrubbing. If you retune and your repeller voltage and ion focus voltage keep going up your source is getting dirty.
On my Archon the water sampling needle and tubing needs to be cleaned but the rest of it is usually OK.

[LALman]

I'll just add that if you do decide to flush transfer lines with Methanol, be sure to follow that flush with deionized water before you restore heat to the transfer lines. Leaving Methanol in the lines can create active sites as the line heats up.

[James_Ball]

I actually had an adapter made up with 1/16" swagelok on one end and a luer fitting on the other so I could connect it to the fitting at the inlet and push solvent back through the system and out through a blank trap into a beaker. I also flushed at temp with methanol then water, it will vaporize the methanol at first, but once it cools the surface all the way through you get hot methanol in the beaker, then repeat with DI Water and a couple empty syringes of air. After reconnect the trap, or use the blank trap and set it to desorb to dry, before reconnecting to the GC inlet.

Worst sample I had to clean after was an iodine powder that foamed and pushed all the way through to the column, the rinses were red for several rinses and took forever to scrub it out of the inlet. Another guy in the lab actually overfilled a sparge tube enough that water went into the column when it desorbed, you could actually see water droplets in the megabore 105m Rtx-505.2 column. I set the oven to 120C and let it bake overnight and was able to recalibrate the next day and keep going. I was surprised that the column and the MS(5971) survived.

[LALman]

I have just such an adapter also! I purge all my sample vials with nitrogen or argon after adding the stir bars. So, I have that gas line right there and I set it for 20 psi and swap the luer lock end for a 1/8" swaglok adapter and let it blow through for 15-20 minutes.