Except for high concentrations of BTEX (high ppmv values) the static dilution is not accurate essentially due to adsorption on walls. May be large gas cylinder (as used for gas sair sampling) may be better (compressed calibration compressed gases are sold in the ppm range).

Adsorption on the walls at lower concentrations is a problem. It can be minimized by using isotope dilution internal standard GC/MS technique. There are multiple label variations for each of these compounds. The viability of this would be dependent on being able to add the internal standards to your actual samples in a consistent way.

For calibrating our SPME needles, we used dynamic (continuous) dilution of 10 ppm BTEX cylinders, using flowmeters, down to ppb values.

Sadegh - Probably be better off using a tedlar or Teflon bag.  I'm no air expert but you might find more information here: http://epa.gov/ttnamti1/airtox.html  I'd look at method TO-11A and TO-3 for reference information.  I know our field chemist spikes known concentration (prepared in methanol) (10 ug/L) into full 40 mL VOA vial with water and no head space.  When ready for analysis they draw off 10 mL which leaves head space above water with very stable concentrations when kept on ice (imperitive).  It works great for doing passive air sampling for a number of volatile compounds including your BTEX compounds.  I should think it would work well for calibrating your spme strips by hanging them in the head space.  Reply to this if you would like more information on the specific field test method including the individual concentrations of the compounds found in the head space (used for calibration). 

The tedlar bags idea is good and certainly cheaper than stainless steel vessels. (Anywway, I don't understand how to know the composition of the water headspace in the Peter's method). I have no experience with such bags, the only problem I foresee is the determination of the volume of air introduced in the bag. The TO-3 method describes a static calibration using these bags, which seems useful. The heating step for VOCs less volatile than benzene is also a problem although I am not sure that it is necessary. Note that the TO-3 method will produce high concentrations of VOCs, and dilution steps may be necessary for obtaining a range compatible with usual air concentrations.  

As Jean noted, the big problem with the method is determination of VOC concentration in head space of the vial with 10 ml of aliquots. I think this should be good for calibration of SPME for vocs in water.

Hi guys - We have a number of systems to determine concentrations of contaminants in air.  We have 2 labs devoted to this analysis and typically use summa canisters in the air toxics analysis.  However our field chemist uses the above mentioned method in mobile laboratory when doing soil gas surveys on numerous sites throughout the year.  Our mobile laboratory is not large enough to be equipped with Entech preconcentrator, lig nitrogen cryo, and GC/MS instrumentation.  Using headspace is a standard practice for field screening and is only used to calibrate the GC instrument, the subsequent air samples (soil gas) are collected in tedlar bags.  Through extensive experimentation we have found the 10mL headspace above liq standard at 10 ug/L at 0-1 degree C has the following BTEX concentrations.

Benzene =  151 ppb/v

Toluene = 159 ppb/v

Ethylbenzene = 145 ppb/v

m/p-xylenes = 136 ppb/v

o-xylene = 112 ppb/v

I was thinking if you wanted to try a very quick, easy, cheap, and simple experiment you could try this with your spme.  Prepare 2 standards in 40 mL VOA vials. One for calibrating instrument, one for spme.   add spme to 1.  Draw 10 mL of liq standard out of each vial and discard. Equilibrate 4 min then Inject 200 uL (depending on instrument sensitivity adjust volume if necessary) into GC - check response of air standard to that of spme.  Now you have a spme air calibration standard at the above listed concentrations.   This would take about an hour to do....  might be worth a try

Regards

I seems worth trying. I know that water vapor has a minor impact on SPME response depending. There are certainly data in the literature on that matter.

Peter's method is a nice quick&dirt way to have a reference atmosphere in the field, but in my opinion it''s not a good calibration for SPME. Carboxen containing fibres (i.e. black, light blue, pink ones) are adsorpion based and will deplete the headspace very rapidly, making the "calibration" useless for ambient samplings.

The simplest option seems to me the use of a "big" (>15L) Tedlar or Nalophan (cheap!!) bag. The bag has to be big to avoid depletion (i.e. concentration change): its volume is linked to the expected exposure time.

Good point Salvatore. I though that the vial contained enough BTEX to avoid depletion, but I don't know how to evaluate that. Have you an idea of the the capacity of a SPME coating, carboxen or any others?

@Salvatore I am doubtful that 4 min exposure to spme would deplete or drive equilibrium between vapor/aqueous especially at 00 C. , if timed properly the concentration in vapor (headspace) will "rebound" after taking spme out.  First injection would answer that question though, if spme has > response than the calibration then your theory would be proven. I agree bags are probably best bet -  As I recall one of the problems with the bags, especially big ones is it's tough to keep them at a constant temperature.  Anyways, Sadegh keep us posted - would like to hear outcome.

@Peter, @Jean-François,

my point is that the SPME fiber depletes the headspace, not the entire solution. The low temperature (high viscosity) and lack of agitation (slow mass transfer rate) are key for this. The SPME does not sample a constant concentration atmosphere, which is what is needed by Sadegh for an ambient sampling - therefore my suggestion: bigger volumes or flow conditions (both for calibration and sampling!!)

Anyway, SPME is not ideal for quantitative enironmental evaluations