Multi-dimensional Marvel

Following one of my recent presentations, an observer commented that while she enjoyed the talk and agreed that GC×GC is especially suited to analysis of complex mixtures, she felt generally overwhelmed by its complexity. She is not alone: while improved resolution sounds good to just about everybody, there is a widespread perception that multidimensional separations are more complex than “regular chromatography”. I want to take a few moments to debunk this complexity myth.

GC×GC is short-speak for comprehensive two-dimensional gas chromatography. Put simply, GC×GC uses two GC columns and harnesses the resolving power from both columns to provide improved separation. Recently, my team has been working on an approach that uses multiplexed detection and three separation columns. We configure the modulator in such a way that the instrument essentially comprises two parallel GC×GC column sets. What sets this apart is that both GC×GC column sets use the same detector.

A decade ago, this would have been complicated; in fact, any GC×GC was. Many users relied on homemade modulation systems and it could be a real battle to interpret results: there are stories of PhD students manually counting peaks and throwing in the towel after reaching a couple of thousand. Luckily, instrumentation and software advances mean that neither construction nor data interpretation are challenging today. Modern instruments are essentially walk-up: the most complex step is column installation.

Of course, a few lessons on GC×GC theory never go astray. I recall grappling with the concept of modulation for about a week in my early learning days, but using GC×GC to tackle new challenges has given me a great deal of satisfaction for more than a decade now and I don’t intend on stopping soon.

A non-exhaustive review of GC×GC papers published so far this year reveals a broad spectrum of applications that benefit from GC×GC analysis. Food/flavor/fragrance continues to be a field heavily using GC×GC, accounting for more than 30 percent of the published papers. Those associated with environmental contaminants account for 23 percent, with the combined areas of fuel and industrial chemistry making up a further 24 percent. Importantly, around 12 percent of the papers already published this year are associated with GC×GC theory or instrumentation and software advancement. This is a good indicator of a healthy field. The balance of papers can be classified as forensics, geochemistry and metabolomics/biomarker discovery, plus one fascinating investigation looking at the effect of the “shark necronome” on feeding behavior. As a resident of Australia, I will follow this study, which reports the potential discovery of a shark repellent, with keen interest!

Regarding our own recent investigations into multiplexed detection for parallel GC×GC, there is great benefit in obtaining three independent sets of retention data (or retention indices) in a single chemical analysis. GC×GC is good but it cannot separate everything, so the possibility of adding additional stationary-phase chemistry to further tease out resolution is inviting. Hyphenating GC×GC with fast mass spectrometry (MS) is very common, but having two MS instruments connected to a GC is not a viable option for most laboratories. Multiplexed detection makes our three-column system
MS-compatible.

It has been reported in the literature that comparing three complementary retention indices against tabulated data for an “unknown” analyte can provide strong enough evidence to assign identity – even without recourse to mass spectrometry. We are enthusiastically exploring the GC×GC / GC×GC-MS nexus using our new multiple-column approaches. Thus, while our colleagues in MS laboratories claim to be moving ever closer to eliminating chromatography, we shall continue our investigations, which we believe will alleviate our dependence on mass spectrometry.

It is not that I dislike mass spectrometry, but if we are aiming for a future filled with personalized measurement devices, multidimensional separation offers great potential for miniaturization. Chemical analyses performed by hyphenated techniques are structured so they alleviate the need for high resolution in any single analysis dimension. In a way, we are shrinking the chromatographer’s triangle of compromise, achieving speed without loss of separation.