Separation of m- and p- Xylenes by Gas Chromatography on Dimethyl Polysiloxane – Is it Possible?

One of the first gas chromatography (GC) projects I tackled when I worked for the US Environmental Protection Agency a few decades ago was the separation of m- and p- xylenes for a microbial degradation experiment.  I did it by purge-and-trap GC with whole column cryotrapping, a technique invented by Jim Pankow at the Oregon Graduate Center.  Purging an aqueous sample onto an adsorbent trap, followed by desorption of that trap to a capillary GC column maintained at -80°C, allowed for sufficient trapping of volatile compounds.  (Believe it or not, you can get 30 mL/min of flow through 0.25mm and 0.32mm column at moderate head pressures when the column is at -80°C; unlike a liquid, gas viscosity decreases with temperature.)  After trap desorption, the GC column was programmed to effect a separation.  In the xylenes application, I used a short piece of 5% diphenyl 95% dimethyl column connected to a wax column.  Cryofocusing was done on the 5 (which still has some diffusivity at -80°C) and rapid heating to 40°C restored the wax to a “liquid” to allow the efficient separation of m- and p-xylenes.  A wax is one of the only commercially-available capillary GC stationary phases that will separate m- and p- xylenes.  I remember calling Jim to discuss the results with him and he said that sounded like an “elegant” solution to the problem.  This was high praise from the master for a teenage chromatographer!

I’ve retained (a little GC humor there...) an interest in the separation of m- and p- xylenes since the EPA days, so when I saw I’d done it on an Rtx-DHA-150 in a GCxGC experiment (see the figures), I knew I had a blog, since the DHA is dimethyl polysiloxane, a stationary phase that would not be expected to separate those xylenes, at least in a typical 30 or 60m column.  GCxGC aficionados know already that this separation isn’t trivial to achieve on any column, since we often sacrifice first dimension separations in pursuit of that ultra sexy GCxGC contour plot.  However, in this case, I practiced “true peak capacity increase” GCxGC according to Leon Blumberg’s suggestions and maintained my first column’s peak capacity.  Thanks again, Leon!

Exxon Mobil Gasoline by GCxGC on a 150m x 0.25mm x 1.0µm Rtx-DHA-150 and 1m x 0.18mm x 0.18µm Stabilwax column combination.

Separation of p- and m- xylenes on a 150m x 0.25mm x 1.0µm Rtx-DHA-150 in the first dimension of a GCxGC run.