OR WAIT null SECS
Kevin A. Schug is a Full Professor and Shimadzu Distinguished Professor of Analytical Chemistry in the Department of Chemistry & Biochemistry at The University of Texas (UT) at Arlington. He joined the faculty at UT Arlington in 2005 after completing a Ph.D. in Chemistry at Virginia Tech under the direction of Prof. Harold M. McNair and a post-doctoral fellowship at the University of Vienna under Prof. Wolfgang Lindner. Research in the Schug group spans fundamental and applied areas of separation science and mass spectrometry. Schug was named the LCGC Emerging Leader in Chromatography in 2009, and most recently has been named the 2012 American Chemical Society Division of Analytical Chemistry Young Investigator in Separation Science awardee.
It is much more efficient to screen column chemistries to achieve optimum separations for a set of target analytes, than it is to play with different mobile-phase conditions on a single column. Kevin Schug explains more.
I’ll admit that I might be a bit behind the curve on this one. I might be relaying a personal epiphany that some of you, especially those of you in some segments of industry, have known about for a long time. In fact, I am sure of it. I have recently spoken to established and highly successful researchers in the clinical analysis field, and they have affirmed what I now hold to be true: It is much more efficient to screen column chemistries to achieve optimum separations for a set of target analytes, than it is to play with different mobile-phase conditions on a single column. Why did it take me a while to come to this realization? I think it might have been a product of my scientific upbringing. Not to take anything away from the exceptional education I received, but I remember being fairly limited in the number of columns I had at my fingertips to test when I was in graduate school. I think that this is not an unlikely situation for an academic laboratory, but I also think it is a testament to the commercial development (and marketing, to be fair) of new phase chemistries over the past several years. Only recently have I really had the opportunity to work closely with a company (in this case, Restek Corporation) that has consciously prepared a set of column chemistries that they market to be complementary. In other words, they aim to provide a toolbox of reversed-phase–based columns that should solve a large majority of a lab’s separation problems. We have had the opportunity to test their offerings fairly comprehensively, and I can say that their collection does a nice job of offering what chromatographers really need to get their job done-selectivity. As we all know, the name of the game in chromatography is resolving chemical species. Resolution is a function of retention, efficiency, and selectivity, as described by the aptly named master resolution equation. I have written previously about the role of selectivity in liquid chromatography method development (1). It is the most influential factor in achieving separations-no selectivity, no separation. Selectivity is most profoundly dominated by the chemical nature of the stationary phase used in separations. If a stationary phase cannot tell the difference between two analytes, then there is little chance that a variation of mobile phase is going to have a profound effect, at least not beyond those things like pH, which can change the form of the analyte. Like any problem in chemistry, there are exceptions, but overall, this statement is true. Recently, we have been able to work with Shimadzu Scientific Instruments, Inc., to design a custom LC–MS instrument in our lab. The system is designed for trace quantitation, and in order to rapidly assess conditions for separations of various target compounds, we have incorporated automated method development hardware and software. Basically, the hardware consists of a selection valve, so that up to six different columns can be connected and checked for their performance using sequential injections. Two quaternary pumps are used to control a range of aqueous and polar organic mobile phase combinations, so that in all, up to 96 different stationary and mobile phase combinations could be theoretically evaluated. Software is provided to help design the mobile-phase compositions (both type of solvent and gradient conditions) to be tested, as well as to automatically build the batch-run table, which directs the instrument workflow. We did not start out with a defined target analyte set, in the sense that we were trying to develop and validate a method for a specific quantification application. Rather, as a means to study selectivity in the Restek column toolbox, we designed a set of analytes that exhibited diverse physicochemical characteristics. These ranged from very hydrophilic to very hydrophobic and included a broad range of chemical functionality. For LC–MS, there really is a limited set of volatile mobile-phase modifiers that are compatible, so choices to screen were straightforward. We chose to run a standard 95% aqueous to 98% polar organic gradient over 12 minutes, and to keep this constant for all mobile phase and stationary phase combinations. Thus, in a very short period of time, we could screen the performance of the stationary phases for separating the 28 different compounds we compiled in our test set. Based on multiple reaction monitoring traces in a triple quadrupole mass analyzer, we were able to take quite a comprehensive look at resolution of some critical pairs, peak shape, and detection sensitivity. The results of this study will soon be submitted for publication, but in general we found that even using just four column chemistries (a polar embedded C18, a high pH stable C18, a biphenyl, and a pentafluorophenyl propyl phase) we could find appropriate conditions for analysis of any of our target analytes. We could also quickly rule out combinations of mobile and stationary phases that did not work well, based on poor peak shape, poor retention, or poor separation. Certainly, if we were going to build a validated method for trace quantitation of some subset of these analytes, we would not have far to go to optimize the separation. Compared to some time I spent in front of an LC–MS instrument in graduate school, making incremental changes in mobile-phase compositions and gradient conditions to try to tease out just a little more resolution for a critical pair, the automated screening protocol is much more efficient. Rather than being bound to a particular column chemistry and trying to squeeze out a separation, a quick change of column often reveals major shifts in retention. Now that we have such automated method development capabilities, I firmly believe that this is most efficient route to method development. Recently, Dr. Russell Grant, Vice President for Research at Laboratory Corporation of America, one of the largest clinical research and diagnostics firms in the world, visited U.T. Arlington and gave a seminar. While Dr. Grant has certainly established himself in the clinical research and mass spectrometry fields, I was pleased to learn that he is a true chromatographer and separation scientist at heart. When we toured my lab and I showed him our system, he also remarked that column screening is a superior strategy for method development. He promptly shared with me some work that his group had been working on for quite a while. I quickly became more appreciative of the mind-set of the clinical lab and the automated throughput that could be brought to method development, to enable sensitive assays to be developed for target analytes from complex biological fluid samples. Method developers in a clinical lab clearly go many steps beyond what I have described above. First, they screen typical LC–MS mobile-phase compositions to ascertain which combination gives the best sensitivity. This is done simply by flow-injecting the analyte dissolved in different solvent conditions into the MS source with no column. From this result, they enter a screen of up to 32 different LC columns, investigating both reversed-phase and HILIC columns and conditions to find the best resolution for their target analytes. They then optimize extraction conditions to study analyte recovery and matrix effects. They often start with a simple liquid–liquid extraction protocol, but systematically investigate extraction in a 96-well format. Of course, matrix effects often dictate the need to alter separation conditions, so there is some iteration that may be necessary, but at this point, a huge method development space has been surveyed en route to designing and validating the final method. In academia, we are not so nearly driven by throughput as in the clinical laboratory, but the approach Dr. Grant described to me is just smart. Method development always takes time and can be a major bottleneck to progress in any analytical lab. It is clear to me that we need to take more of these efficiencies into our academic lab, not just for the sake of increasing productivity, but also to prepare our students better for their careers after they graduate. As I said before, I am not naïve. I know that there are many of you out there who already capitalize on well designed and even automated method development protocols. For many of us, when we were in our instrumental analysis laboratories in college, we had that one favorite column we used for LC, and we were instructed to see what happened to some peaks as we changed the mobile phase. If the mobile phase was too strong, then the peaks coeluted. If we got the composition of gradient just right, we got separation. It seems that was preordained based on the choice of parabens or phthalates as a simple analyte test set. Certainly there is some value to this exercise from the standpoint of seeing an LC separation for the first time. However, it is probably a disservice not to take this a step further. As I write this, I realize that I am quite ashamed that my students in laboratory courses do not get the opportunity to explore more column chemistries during their introduction to LC. I will resolve to change this, so that focus is placed on the more contemporary availability of a diverse range of column phases. Perhaps this will enable students to realize from the beginning that their best bet for achieving a challenging separation is to change the column, rather than to alter the mobile phase. Reference 1. K.A. Schug, “The Role of Selectivity in Liquid Chromatography Method Development,” Restek Advantage. January 2015, 22–23.
Kevin A. Schug is a Full Professor and Shimadzu Distinguished Professor of Analytical Chemistry in the Department of Chemistry & Biochemistry at The University of Texas (UT) at Arlington. He joined the faculty at UT Arlington in 2005 after completing a Ph.D. in Chemistry at Virginia Tech under the direction of Prof. Harold M. McNair and a post-doctoral fellowship at the University of Vienna under Prof. Wolfgang Lindner. Research in the Schug group spans fundamental and applied areas of separation science and mass spectrometry. Schug was named the LCGCEmerging Leader in Chromatography in 2009 and the 2012 American Chemical Society Division of Analytical Chemistry Young Investigator in Separation Science. He is a fellow of both the U.T. Arlington and U.T. System-Wide Academies of Distinguished Teachers.