accuracy of GC quantitaion

[ALT1]

What is the accuracy of GC analysis? I need to determine the amount of MeOH in H20 ( of course, with dilution) within the range of 48 to 52%, i.e the process can tolerate no more than 52% and no less than 48%. Is it possible or is it beyond the accuracy of GC analysis?
Thanks.

[Bruce Hamilton]

You should be able to easily get to about 0.2% Methanol by using an internal standard in your diluent. You will need an internal standard to compensate for injection variations, along with a set of standards ( equivalent to 47-53%? ).

You can probably obtain improved precision ( 0.05%? ), depending on your column, injection system ( autosamplers usually greatly improve precision ), type of column, number of replicates, consistency of dilution, and stability of your detector.

Bruce Hamilton

[Peter Apps]

If you do an uncertainty budget you will probably find that the main contributors to uncertainty are not the GC analysis step (as long as it is properly set up) but the sample prep (dilutions etc) and standards and calibration (purity of reference methanol, purity of internal standard, standard additions, dilutions etc).

Peter