The LCGC Blog: The Dimensionality of Separations: Mass Spectrometry Is Separation Science
On May 1, my group and I were attending a semiannual North Texas Mass Spectrometry Working Group (NTMSWG) meeting in Grapevine, Texas. We were sharing our plans for the upcoming American Society for Mass Spectrometry (ASMS) conference, including abstracts of planned presentations at the conference by different groups (and instrument vendors) in the area. I have mentioned before to some that I am proud about how our group can sit on the fence between chromatography and mass spectrometry (MS). I firmly believe that each needs the other to tackle some of the most challenging analytical problems. Even so, I am not out to make waves; I would happily wear a reversible t-shirt that says “I am a chromatographer” on one side and “I am a mass spectrometrist” on the other. I could then switch it around depending on the company I was keeping at the time. But, at our NTMSWG meeting, I found that I was being given a little grief (all in fun) for my strong ties to the separation science community. It was not at all bothersome, but I couldn’t help thinking “wait a minute, mass spectrometry is separation science, too!”
A great deal of research has been performed on multidimensional separations, including two-dimensional liquid chromatography (2D-LC) (1) or gas chromatography (GC x GC) (2). These techniques use orthogonal retention modes in tandem to expand the separation space and increase the peak capacity (the limit on the number of peaks that can be theoretically resolved in a given run) (3,4). Thus, highly complex mixtures of sample components can be resolved in a single run if sufficient peak capacity can be attained. Although single-dimension high performance liquid chromatography (HPLC) separations might give rise to peak capacities around 50, by adding a second dimension, 2D-LC separations might be expected to reasonably provide peak capacities closer to 500. For GC and GC x GC, these numbers are about two times higher (theoretical peak capacity of 100 in one dimension, and 1000 with two dimensions). It is a synergistic effect.
In many cases, multidimensional chromatographic separations are interfaced with MS detection to obtain high sensitivity and specificity. Ions are generated for the various peaks eluted from the chromatography column, and these ions are then manipulated using electric or magnetic fields (or both) to separate the ions based on their mass-to-charge ratio (m/z). Most scientists know this. Then why is MS typically not considered as just another separation dimension? In fact, MS in different instruments can inherently provide multiple dimensions of separation. An ion trap mass analyzer can provide higher order tandem MS fragmentation to differentiate isomers that exhibit only minor structural differences. We recently published the discrimination of some steroid hormone metabolites, which differed only by the α- or β- orientation of a hydroxyl group, by subjecting them to MS5 (meaning that the fragments of a fragment (MS4) of a fragment (MS3) of a fragment (MS-MS) of a precursor ion distinguished the isomers) (5). Yet, a big difference is that these separations are not really orthogonal. The same mechanism is used to differentiate the ions in each dimension. The orthogonal separation modes available in chromatographic separations are not available in MS (unless you distinguish between electric and magnetic fields, which is reasonable). Even so, no one doubts the power of MS. Instruments that boast high resolving power can theoretically resolve an extraordinarily large number of ions, differing only by parts-per-million in their m/z values from one another. I think that this is one reason that some mass spectrometrists would not be willing to wear the reversible shirt. They would rather try to eliminate chromatography all together. This might be a worthy effort, but there will always be limitations based on ion yields, especially where complex mixtures are desired to be ionized. Combining chromatography with MS clearly has its advantages. Perhaps one of the most seminal contributions for combining 2D-LC with MS was in the development of multidimensional protein identification technology (MudPIT) by John Yates (6). Digests of whole cell lysates certainly fall into the category of highly complex mixtures. The ability of the mass spectrometer to resolve species that chromatographically have been coeluted is clearly a necessity in the MudPIT workflow. Yet, you rarely hear about the concept of peak capacity in stand-alone MS.
As a separation scientist, one area of “gas phase ion manipulation” techniques that I really like is ion mobility spectrometry (IMS). Although I have never had the pleasure of trying it myself, being raised as a chromatographer, it makes sense to me. It also makes technological sense, in terms of increasing the dimensionality of separations. Ion mobility separates gas phase ions based on their differential mobility (controlled by their shape and charge) in the presence of a drift gas. The drift gas retards the mobility of ions as they traverse a flight tube under the influence of a potential gradient. Typically chromatographic separations take seconds to minutes. IMS takes milliseconds, and MS takes microseconds (for time-of-flight MS, at least). Thus, IMS-MS detectors for chromatography are optimal, because you would like successive separation dimensions to take less time than the preceding ones. Studies on peak capacity for IMS-MS have been performed and there is a strong dependence on the type of drift gas used (7,8). Using a sample of tryptic peptides, researchers found peak capacities to range from approximately 2500 (for He) to 7500 (for N2). Because we know that the combination of different peak capacities for different dimensions are synergistic, you can see how increasing the dimensionality of separation from the initial sample preparation, through the traditional (or multidimensional) chromatographic steps, and into the detection regime can provide the capacity to resolve an enormous number of compounds in a sample. At that point, dynamic range of the method versus abundance of components of the mixture probably becomes a limiting factor, but that is a topic for future discussion.
(1) F. Bedani, P.J. Schoenmakers, and H.-G. Janssen, J. Sep. Sci. 35, 1697–1711 (2012).
(2) B. Mitrevski, P. Wynne, P.J. Marriott, Anal. Bioanal. Chem. 401, 2361–2371 (2011).
(3) J.C. Giddings, Anal. Chem. 39, 1027–1028 (1967).
(4) C.G. Horvath and S.R. Lipsky, Anal. Chem. 39, 1893–1893 (1967).
(5) L. Tedmon, J.S. Barnes, H.P. Nguyen, and K.A. Schug, J. Am. Soc. Mass Spectrom. 24, 399–409 (2013).
(6) M.P. Washburn, D. Wolters, and J.R. Yates, Nat. Biotechnol. 19, 242–247 (2001).
(7) B.T. Ruotolo, K.J. Gilliq, E.G. Stone, and D.H. Russell, J. Chromatogr. B 782, 385–392 (2002).
(8) B. Ruotolo, J.A. McLean, K.J. Gilliq, and D.H. Russell, J. Mass Spectrom. 39, 361–367 (2004).
Previous blog entries from Kevin Schug: