Eyes on the Prize: Overcoming Uncertainty to Realize the Power of 2D-LC Separations
In this month’s column, I highlight some of the primary considerations we face in method development and point to resources that can help users overcome uncertainty and develop highly effective 2D-LC methods.
Two-dimensional liquid chromatography (2D-LC) is a technique that extends the separation capabilities of conventional liquid chromatography by adding a second separation step to resolve compounds that are coeluted from a first column. This approach holds tremendous potential to solve difficult separation challenges in fields ranging from pharmaceutical analysis to biofuel characterization. Currently, method development is a significant bottleneck impeding more widespread implementation of 2D-LC methods, particularly for new users who are uncertain about how to proceed. In this month’s column, I highlight some of the primary considerations we face in method development and point to resources that can help users overcome uncertainty and develop highly effective 2D-LC methods.
Two-dimensional liquid chromatography (2D-LC) holds tremendous potential to impact many areas of science that rely on liquid chromatography (LC) to move those fields forward. For example, 2D-LC can be used to rapidly resolve all the enantiomers of a molecule with multiple chiral centers, which can be very difficult to achieve with any single conventional 1D-LC method (1). Over the last 20 years, the development of 2D-LC as a technique itself has been impressive, including important advances in our theoretical understanding of the benefits and limitations of the technique, as well as the proliferation of commercially available instrument hardware and software that makes 2D-LC easier to implement in practice. In a recent review article, we noted that 2D-LC is slowly but steadily making its way out of academic research laboratories and into industrial ones, as indicated by the increasing number of peer-reviewed papers in the 2D-LC space with co-authors from industry (2). These developments are attracting new 2D-LC users that have no experience with the technique whatsoever, which is exciting in the sense that these new folks broaden the user community and enrich the range of compelling use cases for 2D separations. However, it can also be challenging, in the sense that new users need support in their development of new methods and applications, even through there are many research questions that need to be more fully resolved before we can provide advice that is as complete as we would like it to be.
In this installment of “LC Troubleshooting,” I briefly touch on several important aspects of the method development process for 2D-LC where I sense a high degree of uncertainty among new users about how to proceed. Overcoming these uncertainties is important for accelerating the development of effective 2D-LC methods and increasing the number of confident 2D-LC users in the separations community.
Selection of Stationary Phases and Optimization of Mobile-Phase Conditions
One of the essential principles of effective 2D-LC separations is that stationary phases and elution conditions must be chosen such that the selectivities of the first and second dimensions are complementary (3). Specifically, we aim for conditions that enable separation in the second dimension of compounds that are partially or fully coeluted from the first dimension (1D) column. If we are unable to realize this complementarity for any reason, then the overall prospects for the 2D-LC method will not be very good because we end up just repeating a separation with a given selectivity twice.
Resources that Can Guide the Selection of Stationary and Mobile-Phase Chemistries
In the interest of focusing this part of the discussion, I am largely restricting it to situations where reversed-phase (RP) separations are used in both dimensions. Other combinations of separation modes are very important in some application areas, such as protein characterization. Readers interested in a more detailed discussion of column selection in those contexts are referred to other resources (4).
The Hydrophobic Subtraction Model of reversed-phase selectivity (5) is the basis of a freely available database of characteristics for more than 750 commercial RP columns (www.hplccolumns.org). This data set has been a rich source of information supporting discussion about sets of RP columns for use in multidimensional LC separations for more than a decade (6–8). Theoretical studies that have considered the complementarity of thousands of different potential combinations of stationary phases drawn from this dataset invariably identify less frequently used phases (for example, graphite-like phases [8]) as having the most potential to yield the best 2D separations. However, these types of less frequently used phases are unlikely to be included in the pool of candidate stationary phases considered by large laboratories working with relatively short lists of preferred columns informed by decades of experience with them in the context of 1D-LC. Thus, it is instructive to look at the lists of columns that have been used in published experimental work that describe the use of arrays of 2D columns to systematically screen different selectivities in the context of impurity detection for small pharmaceutical molecules.
In 2013, Zhang, Chetwyn, and associates described a 2D-LC setup similar to that shown in Figure 1 involving an array of 2D columns mounted to a column selection valve (9). More recently, Wang, Regalado, and associates built on this concept through the addition of a column selection valve in the first dimension and mobile-phase selection valves in both dimensions (10). Such a system enables automated screening of hundreds of potential combinations of first and second dimension mobile- and stationary-phase chemistries. Most recently, in a paper published in this magazine last year (11), Lawler, Breitbach, and associates used an array of 2D columns (all from the same manufacturer) that included C8, C18, RP-amide, PFP, cyano, phenyl-hexyl, and biphenyl phases. They developed 2D-LC methods and used them to screen the different combinations of first (C18) and second dimension columns with an eye toward resolution of an API and its impurities. This group applied the screening approach to several different small molecule APIs and their associated impurities. Interestingly, although several of the 2D stationary phases yield good results, very often the best 2D method involved C18 phases in both the first and second dimensions. The authors concluded that using a different mobile-phase pH in the second dimension (compared to pH ~2 in the first dimension) was more important for resolving the APIs and their impurities than dramatically changing the stationary phase chemistry.
FIGURE 1: Illustration of a 2D-LC system hosting an array of different 2D column chemistries. These columns can be mounted on a column selection valve, such that the specific 2D column that is used for a particular analysis is controlled by a method setting. This enables automated, unattended screening of different stationary phases with one or more samples. Adapted from reference (9).
Over the past two years, Petersson, Euerby, and associates published a series of papers focused on identification of short lists of stationary and mobile-phase chemistries to include in screening 2D-LC method conditions for the specific purpose of detecting impurities related to therapeutic peptide APIs (12–14). After evaluating tens of different stationary- and mobile-phase chemistries by using performance metrics, including peak shape, general selectivity, and isomer selectivity, they were able to recommend short lists of stationary- and mobile-phase chemistries for use in 2D-LC screening systems that both leverage the complementarity of modern column chemistries (such as mixed-mode phases) and prioritize compatibility with mass spectrometric detection in the second dimension.
Resources that Can Guide the Selection or Refinement of Mobile-Phase Conditions
Once the stationary- and mobile-phase conditions have been chosen for the second dimension of a 2D-LC separation, one must decide how to go about choosing elution conditions—that is, isocratic or gradient, and if gradient, what program? Choosing a steep gradient that starts with minimal strong solvent (for example, mostly aqueous) and ends with mostly organic solvent provides the best chance at both getting some retention for the analytes of interest and ensuring that the compounds are eluted from the 2D column before the end of the analysis. However, such steep gradients are also the least likely to yield adequate resolution, particularly when attempting to separate closely related compounds as is very often the challenge in the 2D separation. Several studies have been published in recent years that describe the use of retention models in the context of non-comprehensive (that is, multiple heartcut, and selective comprehensive) 2D separations to help optimize 2D elution conditions. In my own research group, we have been focusing recently on the development and use of an iterative modeling approach that enables the use of any endpoint of the optimization workflow, including isocratic conditions. This is particularly challenging when working with molecules that are highly sensitive to mobile-phase composition, such as peptides and oligonucleotides, and this iterative modeling approach provides a means to develop such conditions in an efficient and systematic manner (12,15).
Choosing Conditions to Deal with Mobile-Phase Mismatch
One of the biggest challenges we encounter in 2D-LC is quite different from the situation in two-dimensional gas chromatography, is that in 2D-LC, the mobile phase from the 1D separation can seriously negatively impact the quality of the 2D separation. This is especially acute in cases where the 1D mobile phase contains a high concentration of a component that acts as a strong solvent in the second dimension. The most clear example of this is found in the use of hydrophilic-interaction chromatography (HILIC) in the first dimension, which relies on a high (typically greater than 80%) concentration of acetonitrile, and RP in the second dimension, where high concentrations of acetonitrile lead to very low retention. Unsurprisingly, the seriousness of this problem has attracted a great deal of attention from researchers over the years, and several potential solutions have been explored, such as evaporation of the strong solvent between the two dimensions, use of trapping cartridges between dimensions, and addition of a weak solvent diluent between dimensions. They are too numerous to describe here, but interested readers are referred to other resources that describe these approaches in detail (16). Recently, Pardon, Cabooter, and associates carried out a systematic study of the effects of parameters involved in the approach known as active solvent modulation (ASM) for managing the mobile-phase mismatch problem (17). A valuable product of this study was an easy-to-follow flow chart that guides the analyst through a series of decisions that can impact the effectiveness of the ASM approach. I strongly encourage newcomers to 2D-LC to consider using this flow chart when thinking about how to manage the mobile phase mismatch problem in 2D-LC generally, and when developing a method specifically involving ASM.
Summary
In this installment, I have addressed some of the most important sources of uncertainty for new users approaching the development of a 2D-LC method. As is the case with conventional 1D-LC, the fact that we have so many choices of stationary- and mobile-phase chemistries is both a blessing and a curse. Having a large number of choices provides tremendous opportunities to tailor the separation chemistry to the problem at hand; however, it can also be paralyzing if we don’t have a way to prioritize a small set of columns and mobile phases to try in actual experiments. Fortunately, several experimental studies in the past five years have demonstrated success in the development of short lists of conditions to use. Moreover, multiple groups have demonstrated the use of these conditions in 2D-LC systems set up with valves that enable automated selection of different mobile and stationary phases under software control. It seems likely at this stage this type of approach will be the cornerstone of 2D-LC method development for the foreseeable future. Finally, I briefly touched on recent work aimed at developing a systematic approach to managing the mobile phase mismatch problem that can be a major obstacle to implementation of effective 2D-LC methods. Taken together, guidance from these recent studies can help new 2D-LC users overcome uncertainty and realize the tremendous potential of 2D-LC for solving difficult separation challenges.
References
(1) Lin, J.; Tsang, C.; Lieu, R.; Zhang, K. Fast Chiral and Achiral Profiling of Compounds with Multiple Chiral Centers by a Versatile Two-Dimensional Multicolumn Liquid Chromatography (LC–mLC) Approach. J. Chromatogr. A 2020, 1620, 460987. DOI: 10.1016/j.chroma.2020.460987
(2) Van Den Hurk, R. S.; Pursch, M.; Stoll, D. R.; Pirok, B. W. J. Recent Trends in Two-Dimensional Liquid Chromatography. TrAC Trends Anal. Chem. 2023, 166, 117166. DOI: 10.1016/j.trac.2023.117166
(3) Giddings, J. C. Concepts and Comparisons in Multidimensional Separation. J. High Resol. Chromatogr. 1987, 10 (5), 319–323. DOI: 10.1002/jhrc.1240100517
(4) Stoll, D., R.; Pirok, B. W. J. Chapter 5: Selecting Modes and Selectivities for Mult-Dimensional LC. In Multi-Dimensional Liquid Chromatography: Principles, Practice, and applications; CRC Press,, 2022.
(5) Dolan, J. W.; Snyder, L. R. The Hydrophobic-Subtraction Model for Reversed-Phase Liquid Chromatography: A Reprise. LCGC N. Am. 2016, 34 (9), 730–741.
(6) Simpkins, S. W.; Bedard, J. W.; Groskreutz, S. R.; Swenson, M. M.; Liskutin, T. E.; Stoll, D. R. Targeted Three-Dimensional Liquid Chromatography: A Versatile Tool for Quantitative Trace Analysis in Complex Matrices. J. Chromatogr. A 2010, 1217 (49), 7648–7660. DOI: 10.1016/j.chroma.2010.09.023
(7) Lindsey, R. K.; Eggimann, B. L.; Stoll, D. R.; Carr, P. W.; Schure, M. R.; Siepmann, J. I. Column Selection for Comprehensive Two-Dimensional Liquid Chromatography Using the Hydrophobic Subtraction Model. J. Chromatogr. A2018, 1589, 47–55. DOI: 10.1016/j.chroma.2018.09.018
(8) Tirapelle, M.; Chia, D. N.; Duanmu, F.; Besenhard, M. O.; Mazzei, L.; Sorensen, E. In-Silico Method Development and Optimization of on-Line Comprehensive Two-Dimensional Liquid Chromatography via a Shortcut Model. J. Chromatogr. A 2024, 1721, 464818. DOI: 10.1016/j.chroma.2024.464818
(9) Zhang, K.; Li, Y.; Tsang, M.; Chetwyn, N. P. Analysis of Pharmaceutical Impurities Using Multi-Heartcutting 2D LC Coupled with UV-Charged Aerosol MS Detection: Liquid Chromatography. J. Sep. Sci. 2013, 36 (18), 2986–2992. DOI: 10.1002/jssc.201300493
(10) Wang, H.; Lhotka, H. R.; Bennett, R.; Potapenko, M.; Pickens, C. J.; Mann, B. F.; Ahmad, I. A. H.; Regalado, E. L. Introducing Online Multicolumn Two-Dimensional Liquid Chromatography Screening for Facile Selection of Stationary and Mobile Phase Conditions in Both Dimensions. J. Chromatogr. A 2020, 460895. DOI: 10.1016/j.chroma.2020.460895
(11) Lawler, J. T.; Lesslie, M. W.; Randstrom, C. E.; Zhao, Y.; Breitbach, Z. S. A Standardized 2D-LC Screening Platform for Peak Purity Determination in Pharmaceutical Analysis. LCGC N. Am. 2023, 220–224. DOI: 10.56530/lcgc.na.vr6271j2
(12) Stoll, D. R.; Sylvester, M.; Euerby, M. R.; Buckenmaier, S. M. C.; Petersson, P. A Strategy for Assessing Peak Purity of Pharmaceutical Peptides in Reversed-Phase Chromatography Methods Using Two-Dimensional Liquid Chromatography Coupled to Mass Spectrometry. Part II: Development of Second-Dimension Gradient Conditions. J. Chromatogr. A 2023, 1693, 463873. DOI: 10.1016/j.chroma.2023.463873
(13) Field, J. K.; Bruce, J.; Buckenmaier, S.; Cheung, M. Y.; Euerby, M. R.; Haselmann, K. F.; Lau, J. F.; Stoll, D.; Sylvester, M.; Thogersen, H.; Petersson, P. Method Development for Reversed-Phase Separations of Peptides: A Rational Screening Strategy for Column and Mobile Phase Combinations with Complementary Selectivity. LCGC Europe 2022, 440–449. DOI: 10.56530/lcgc.eu.qp3971p2
(14) Petersson, P.; Buckenmaier, S.; Euerby, M. R.; Stoll, D. R. A Strategy for Assessing Peak Purity of Pharmaceutical Peptides in Reversed-Phase Chromatography Methods Using Two-Dimensional Liquid Chromatography Coupled to Mass Spectrometry. Part I: Selection of Columns and Mobile Phases. J. Chromatogr. A2023, 1693, 463874. DOI: 10.1016/j.chroma.2023.463874
(15) Stoll, D.; Sylvester, M.; Meston, D.; Sorensen, M.; Maloney, T. D. Development of Multiple Heartcutting Two-Dimensional Liquid Chromatography with Ion-Pairing Reversed-Phase Separations in Both Dimensions for Analysis of Impurities in Therapeutic Oligonucleotides. J. Chromatogr. A2024, 1714, 464574. DOI: 10.1016/j.chroma.2023.464574
(16) Stoll, D., R.; Leme, G. M. Chapter 4: Instrumentation for Two-Dimensional Liquid Chromatography. In Multi-Dimensional Liquid Chromatography: Principles, Practice, and Applications; CRC Press, 2022.
(17) Pardon, M.; Chapel, S.; De Witte, P.; Cabooter, D. Optimizing Transfer and Dilution Processes When Using Active Solvent Modulation in On-Line Two-Dimensional Liquid Chromatography. Anal. Chim. Acta 2023, 1252, 341040. DOI: 10.1016/j.aca.2023.341040
About the Author
Dwight R. Stoll is the editor of “LC Troubleshooting.” Stoll is a professor and the co-chair of chemistry at Gustavus Adolphus College in St. Peter, Minnesota. His primary research focus is on the development of 2D–LC for both targeted and untargeted analyses. He has authored or coauthored more than 99 peer-reviewed publications and four book chapters in separation science and more than 150 conference presentations. He is also a member of LCGC’s editorial advisory board.
chroncich@mjhlifesciences.com
