The LCGC Blog: My New Obsession: Gas Chromatography with Vacuum Ultraviolet Absorption Detection
In 2009, I was contacted by, VUV Analytics, Inc., a company out of Austin, Texas. The company was developing a new vacuum ultraviolet (VUV) absorption detector and had received Texas Emerging Technology Fund support to commercialize it. I had just won the LCGC Emerging Leader in Chromatography Award. They saw my face on the cover of the LCGC North America magazine and realized that I was located only a short, beautiful drive up I-35 from Austin to Arlington. Now, writing that sentence makes me snicker, because the drive between Austin and Arlington is not likely classified by many as “beautiful” — “monotonous” might be more accurate — and it has been a running joke associated with our collaboration for the past few years. Nevertheless, my colleagues at VUV Analytics made the drive many times to Arlington to talk about the potential development of a chromatography detector. They had worked previously in the semiconductor industry and had developed new technology to help measure ultrathin films. What they soon had realized is that the vacuum ultraviolet range of light (115–200 nm) has long been considered inaccessible for benchtop analytical instruments. Only bright-source synchrotron facilities have been able to effectively collect spectra in this wavelength range. You will only find one analytical textbook that pays it any credence as an analytically useful range for analysis, and that is the one I recently published with coauthors Dasgupta and Christian (1). Suffice it to say, I had some inside knowledge on the usefulness of this approach and thus advocated to include it, because I knew it was going to be the next best thing in gas chromatography (GC) detectors.
Soon after our meetings about the company’s innovative technology, we arrived at a focus on developing a GC detector. I will be the first to admit that the guys at VUV Analytics are the brains behind the detailed absorption measurement technology in this wavelength range, but I believe we quickly developed a very interesting relationship figuring out how to make a workable manifestation by connecting it to a GC system. On the one hand, the GC system made a great inlet for the detector. Analytes would be introduced into the absorption cell in the presence of a high background of helium. Every chemical compound absorbs in the 115–240 nm range captured by the detector, but helium has a relatively low absorptivity and can be easily subtracted from the background. Some engineering and upgrades were necessary to make sure that the full range of GC analytes that could be separated and eluted through modern column technology would also make it to the detector. Practitioners understand the importance of the heated transfer line between the GC and a mass spectrometer. We were facing the same issue — we did not want chemical compounds condensing in the transfer line or detector before the signal could be registered. We went through a variety of iterations and tests, both in Arlington and in Austin. In the end, the first commercially viable system was installed in my laboratory in 2013.
Very recently, we published the first article on the coupling between GC and vacuum ultraviolet absorption spectroscopy detection (GC–VUV) (2). If it is not clear from what I write above, this technology is characterized by universal detection and the fact that all chemical species have distinct gas-phase absorption spectra in the wavelength range monitored. Single-wavelength systems using VUV light have been investigated in the past as detectors for GC (3–5). But now for the first time, full-scan spectra can be collected with the use of a detuerium lamp, reflective optics, a flow cell appended by MgF windows, and a CCD photodetector, which captures the absorption of all wavelengths by the sample simultaneously. This is much like how a photodiode-array detector for high performance liquid chromatography (HPLC) works. Importantly, sample collection is fast enough (100 Hz) to easily keep up with modern fast GC and GCXGC applications.
We are investigating numerous applications, and we keep uncovering interesting ways to exploit the advantages of GC–VUV. Such is the excitement associated with new technology! We have spent significant time homing in on some of the applications where mass spectrometry (MS) has problems. If you think about it, GC detectors do not easily handle coeluted peaks. If two components are coeluted and enter an electron ionization source together, the resulting mass spectrum is a mix (and a mess) that will be difficult to search successfully within a library reference. MS provides qualitative information, but most other detection methods do not (flame ionization, electron capture, thermal conductivity, and so forth). Some might be selective, like nitrogen–phosphorus detection, but major limits still exist for deconvoluting overlapping peaks. The VUV detector potentially solves this issue. In GC–VUV, overlapping peaks will register spectra that are additive of their individual absorption properties. It is easy to tell that a peak is not pure if the spectrum for one of the coeluted components is known. If cross sections (that is, absorptivity across wavelength ranges) for both are known, it becomes a simple matter to deconvolute the relative contribution of each compound to the summed signal. Not only that — when a cross section is known for a molecule and an unknown amount of that compound is registered in the detector, it is literally possible to determine the number of molecules that caused the absorption event. Absent losses from sample preparation or the GC system (which can be characterized), this method lends possibilities for calibrationless detection. The VUV detector then becomes what can be called a pseudo-absolute detector (6). I do not want to go into more details on this here, as it will be the subject of a future blog. That possibility has the potential to greatly simplify analytical work flows.
In our recent article (2), you can see how GC–VUV is good for differentiating isomers of aromatic compounds and fatty acid methyl esters. It can be used to readily detect water, oxygen, carbon dioxide, and other permanent gases. The spectral features are simply beautiful, and for some small molecules, quite surprising. Take a look at the article and the associated supporting information document to see the large variety in spectral traces. If you cannot get access to the article, send me a note, and I will be happy to share. We show detection limits in the low- to mid-parts-per-billion regime (picograms on-column) for most compounds. Of course, different structures provide different chromophores and quantitative analysis is governed by the well-known Beer-Lambert law. Molecules possessing electrons that can undergo high probability transitions (for example, π à π*) are the most sensitively detected. We also take a look at the state of the art for computing absorption spectra and their comparisons with spectra we collect. I think this will be an interesting area of research to pursue further in the future. In the end, the availability of library spectra, coupled with the potential ability to target specific classes of compounds based on absorption features manifested by certain functional groups (in other words, the application of spectral filters to improve detector selectivity), will place GC–VUV in strong competition with GC–MS; in some cases, GC–VUV will be more favorable and much simpler.
I look forward to writing more about our advances (and those by others) using GC–VUV in the future. In a field that has not seen a new and truly transformative detector for many years, this is a very interesting opportunity. I have always said that I am easily distracted by shiny objects. It does not take much to gain my interest and consideration. However, here I think I have definitely found more of an obsession than just an interest.
Previous blog entries from Kevin Schug: