Am I using the correct column to measure the ratio of water:ethanol? See details below.

Hi Everyone, I am looking to get guidance based on the following column (J&W DB-5 GC Column, 60 m, 0.25 mm, 0.25 µm, 7 inch cage; Part Number:122-5062).  I need to measure the ratio of water to ethanol in an organic solvent.  Would the column be good for this separation?  The organic solvent will be polar, but it can be a number of different ones. Any recommendations? Thanks :~)

  • Hi Steven,

    quantifying water with GC is not recommended. Especially on an apolar column like a DB-5ms you will have distorted peak shapes in irregular ways and water will slowly destroy the liquid phase.

    For really be sure about water it is better to use Karl-Fischer titration, you can still do ethanol by GC.

    If you really want to do water by GC (and as I said, not recommended) you need a detector that sees water (TCD, MS, HID). Water must never be the matrix if you want to quantify. The principle "same solves same" applies here too, polar component needs polar phase, i. e. at least a Wax phase, better a PLOT type (PoraPLOT Q, PoraPLOT U, PoraBOND U, HP-PLOT Q or U). The latter has the advantage that the phase will not be damaged, but the disadvantage of very high retention, so it depends on your matrix solvent. Polar columns will give better peak shapes for water. U polymers give you the best water peak shape of all phases.

    With our Scanview program (available here in the Community) you can search for an appropriate application note with component=water AND technique=GC and download them from the site, like A00582:

    With all liquid phase columns, be sure that your sample is fully evaporated in the injector (better use 270 °C injector temp as water has a high heat capacity and therefore cools down the liner during injection up to 80 °C). Water also has a very high evaporation volume that should fit into the liner's volume without backflash.



  • Thank you for the response! I forgot to state that the GC/MS is equipped with the MS detector. Would that change your answer in any way?  

  • Not really, except water being measured with MS is even worse than on a TCD, because of possible interactions in the ion source (and the low m/z of 18). GC is just not made for quantifying water.

Reply Children
No Data
Was this helpful?