headspace sample preparation and ISTD

[foren]

Dear members of this forum,

I am trying to develop a HS-GC-FID method (on an agilent ?890 and a7694HS sampler) for quantifying toluene in urine of auto painters. I am out of town and carry out the task with help of friends, by phone.
Looking for low ppb range, the calibration curve was not linear due to sample introduction. 1 ml of toluene spiked urine plus 50ul %.5 IPA as ISTD, and 1 ml of citrate buffer was my initial sample preparation. I changed to add 1ml saturated NaCl solution with an injector after crimping the vial in order to avoid toluene loss.

Also I tried to tune instrument conditions a bit. I don't like the design of Agilent since it introduces extra parameters such as loop and transfer line temp and loop fill and eq. time but it is the only instrument available and literature shows that it works for many applications. Current settings are as follows:

Vial temp: 60C, Loop Temp: 120C, Trans. Line Temp:140 C, Loop Fill time:0.2', Inject time:1', Loop eq. time:0.05', Eq. Time: 13'
GC Oven:60C initial for 6min, then ramp to 150C, N2 as carrier, splitless inj for 1 min.

I had requested friends to prepare 10 samples(10ppb), with extra care to prepare each and every one with the same procedure. Results are interesting. RSD was high but there is a readily detectable systematic error: ISTD areas are increasing, toluene areas are decreasing significantly.

I do not have much experience with agilent HS design so I suspect I am doing something wrong. I (to be more precise: we) made trial runs for different parameters but they did not satisfy me.

Do you have some suggestions for me, to reduce RSD, due to sample preparation and/or sample introduction?

My next step is to change the ISTD, with a similar bp and henry constant to toluene. Benzene is a candidate for trial work but since it is also one of the chemicals people are exposed to, will not work for the method.

I thank you for your patience to read this long post.

Best regards,
Foren

[chromatographer1]

Foren,

I would add the 1mL urine to 100mg NaCl. Perhaps you should add 10µL of a 1µg/mL Cumene in methanol as an IStd although you should not need one, but go ahead and do it anyway.

You can create std addition samples by adding toluene in methanol as you did the internal std.

Your temperatures and injection parameters seem ok.

Your problem concerning the systemic error is a matter of sample preparation and I am aware of the issues and of the corrective actions.

Please call or email me and I will be glad to give you a workable solution.

I validated a similar analysis using 1µL of urine measuring toluene at 1ppm with linearity of 0.9995 from 1ppm to 100ppm with RSDs of 5% at 50 ppm.

Lower concentrations are easier than the higher concentrations for toluene.

Rodney George
Senior Research and Development Scientist
Gas Separations Research
Supelco
595 North Harrison Road
Bellefonte, PA 16823

814-359-5737 voice
814-359-5459 fax
rgeorge@sial.com