GCxGC-FID of a Petroleum Diesel and Biodiesel Blend: Rtx-200 x Rxi-17Sil MS and Rxi-5ms x Rxi-17Sil MS

One of the benefits of going to the 8^th GCxGC Symposium and 35^th International Symposium on Capillary Chromatography is that you not only get to hear presentations by experts like Professor John Seeley of Oakland University, but you also get to interact with those same experts outside of the talks, which is a great learning experience.  John gave the excellent talk, “Rationalizing and Predicting GCxGC Separations with a Simple Solvation Parameter Model”.  One of the most interesting GCxGC separations he showed was for a diesel:biodiesel blend on a DB-210 (100% trifluoropropylmethyl-siloxane) primary column x BPX50 (50% phenyl polysilphenylene-siloxane) secondary column.  The fatty acid methyl esters (FAMEs) that “are” the biodiesel in this case were positioned slightly below and to the right of the bulk petroleum hydrocarbons, unlike any GCxGC separation I had previously seen for diesel:biodiesel, including that in one of John’s publications, Analysis of Biodiesel/Petroleum Diesel Blends with Comprehensive Two-Dimensional Gas Chromatography, where he used a 5% phenyl x wax setup.  In that report, and others, the FAMEs elute above (later in the second dimension) n-alkanes in a region where interferences could occur with the cyclic alkanes and monoaromatics in petroleum diesel. 

I couldn’t wait to get back in the lab and try this myself with Restek’s Rtx-200 (trifluoropropylmethyl-siloxane, slightly less than DB-210) x Rxi-17Sil MS (50% phenyl polysilarylene-siloxane) in the GCxGC-FID.  Using the same second dimension Rxi-17Sil MS, I also tried an Rxi-5ms (5% diphenyl 95% dimethyl polysiloxane, same length and ID as the Rtx-200, and the same GCxGC-FID conditions) as an alternative to the Rtx-200 in the first dimension.

Let’s get to the data.  You can already see in the first contour plots (GCxGC chromatograms) below that I’m somewhat achieving what John showed, with the FAMEs in a dilution of the Diesel/Biodiesel 80:20 Blend Standard (soy biodiesel) showing up either at or below the n-alkanes “line” along the 17Sil MS dimension when the Rtx-200 is the primary column.  Contrast this against the Rxi-5ms primary column, where the FAMEs are now eluting above the n-alkanes, which complicates their unbiased determination.

How can this “retention cross-over phenomenon” of alkanes and FAMEs happen?  We’re using EXACTLY the same second dimension column, the 17Sil MS.  Well, it’s all about elution temperature, in the first dimension.  The hydrocarbons in the petroleum diesel are less retained by the Rtx-200 versus the Rxi-5ms, which means that many (all?) of the petroleum diesel compounds experience more retention on the Rxi-17Sil MS since they’re “injected” onto that column by the GCxGC modulator while that secondary column is at a lower temperature.  You can see this clearly by comparing the second set of contour plots of Rtx-200 x Rxi-17Sil MS and Rxi-5ms x Rxi-17Sil MS, which are on exactly the same time scale.  Note the greater “spread” of compounds along the 17Sil MS axis for the Rtx-200 setup.

So...not only do the petroleum hydrocarbons elute earlier on the Rtx-200 and show increased retention on the 17Sil MS as a consequence, but the FAMEs elute proportionately later versus the Rxi-5ms.  In the case of the 5ms, the alkanes (and other hydrocarbons) elute so late they have less retention on the 17Sil MS and they cross over with the FAMEs.  A bit confusing, I know, but perhaps the table shown below with GCxGC retention times in sec will make it more clear (like mud, right?).

I’ll follow up soon with another post on why this separation is significant, and it has something to do with not wanting FAMEs to show up at all in certain fuels, meaning you need some way to check for that.

Solvation parameter model of comprehensive two-dimensional gas chromatography separations (Seeley et al., JOCA 2009)