Trihalomethane analysis problems (GC/FID)

[TimJ]

Hi everyone
I am having problems measuring low concentrations of trihalomethanes (<100ug/L) by GC/FID. This is because I cannot get a stable baseline for some reason, and there are background peaks occuring near to where the THMs peak. I am using a headspace extraction method, so there could be contaminants in the air, but I have tried different types of air (lab, fume cupboard, outside, zero-grade air from cylinder) and they all give similar readings.
The use of a smaller syringe (100uL as opposed to 500uL) results in a much more stable baseline. Does this rule out column bleed as a reason for the peaks when using a 500uL syringe? Unfortunately, I need to use the larger syringe size for THM analysis.
If anyone has any suggestions what could be causing the peaks I would be very grateful.
I am using a capilliary column, 100C isothermal, 200C injector, 350C detector.
Many thanks
Timothy Jones

Univeristy of Wales, Bangor
United Kingdom

[xxx]

What type of column do you use?

You could follow this steps :

1-Try to condition it to its maximum isothermal temp for hours...
2-You could also cut the first 20-30 cm,to eliminate column contamination

Good luck

[TimJ]

Column is a Restek MXT-1 Crossbond 100% dimethyl polysiloxane, 30M x 0.53mm ID, 70 uM df.
I am trying your first suggestion at the moment. I have just put through an air sample at 300C isothermal instead of 100C and the peaks came out at the same place but much larger. Does this suggest to you a problem with the column?
Thanks
Tim

[MikeD]

What split ratio are you using?
For me, using your conditions and syringes, the FID limit of quantification would be 20-50 ug/L with a split ratio of 25:1. Have you tried seeing where a 1000 ug/L standard comes ?

Mike

[TimJ]

Hi Mike
I'm using a splitless injection, would you recommend using a split ratio?
I have eliminated the annoying peaks that we had by using a sample loop; the contaminents must have been coming from the septum or top of the injector port.
Standards are coming out nicely and we have the runtime down to 3 minutes
Do you know why the standard peaks get smaller the more brominated the THM is (same concentration)?
Many thanks
Tim

[MikeD]

Tim

Using an in-line sample loop was a good move. The reason I asked about split ratio was that the pressure pulse from splitless air injections by gas syringe can have undesirable effects even with the megabore capillary. The sample transfer from injector to column would take several seconds and in that time some or most of the sample could blowback through the septum. With the loop injection splitless transfer might still be slow but at least there is nowhere else the sample can go. If you get the job done without a split then that's OK.

Can't speculate on the brominated/chlorinated FID response difference. Clearly a difference in the propensity to form methane in the flame but don't know why.