Page 2 of 2

Posted: Thu Oct 22, 2009 1:14 pm
by Peter Apps
Hi Francesc

How good was the performance, especially repeatability ?, and how does its cost compare to fully automated systems ?

Peter

Posted: Fri Oct 23, 2009 9:09 am
by fsistere
Hi Peter

Now i worked in headspace with a combipal and obiously it's better. With the manual headspace you need the internal standard.
I don't know how does it cost.
I use it to analize benzene and i have a repeatibility about 10-15% (with ISTD).
I write about it when i read that can't buy an automated system.
I think for "manual headspace" is a good option (but better the automated)

Posted: Fri Oct 23, 2009 3:43 pm
by JGK
Well, I was able to get quite a decent results by using combination of home-made sand bath (i.e. beaker with sand, hot plate with fine adjustment and a thermometer), 2 mL crimp-cap vials and classical autoinjector (not even a gas-tight syringe!). Used it for determination of residual hexane in oils and pressed seeds. Each sample was equilibrated in sand bath for 1 h, and afterwards I had to run with a hot vial to another room, but it did the job. The most problematic step was actually preparation of spiked samples (it's a bit hard to spike 1 mL/0.5 g of sample with e.g. 0.1 or 0.3 uL of hexane), but after some practice I was able to get excellent linearity and very good precision (sorry, I don't remember the exact RStDev).
I did a method for analysis of an anasthetic from blood on a 5890 and a non-HS autosampler.

We rigged the sample trays up to a waterbath with a pumping thermostirrer unit (we only needed 40C temp), replaced the syringe with a large bore 200µL syringe (non-gas tight), lubricated the plunger with a viscous silicon oil. For the samples we placed filter a paper cylinder into the vial and injected the sample into the sealed vial. Capilliary action took the liquid into the filter paper and increased surface area for evaporation into the headspace. this also minimised the possibility of liquid uptake into the injection syringe.