Review of Chiral Stationary Phase Development and Chiral Applications

The development of chiral column technology since the review in 2008 is discussed by Thomas Beesley, a pioneer in the chromatographic separations of enantiomeric compounds. He discusses the use of generic chiral stationary phase (CSP) screening to find the optimum column, the use of smaller diameter columns for high performance liquid chromatography (HPLC) and supercritical fluid chromatography (SFC), clones of off-patent phases and new concepts in CSP development.

The chiral business continues to grow, albeit at different rates in different parts of the globe. A handful of successful chiral stationary phases (CSPs) recently developed have prompted a reliance on a generic approach to screening for selectivity. The primary goals of speed and efficiency have been demonstrated through the application of supercritical fluid chromatography (SFC) and columns of smaller dimensions with high solvent throughput. Standard HPLC still dominates the methodologies because of its perceived robustness, transferability and wide instrument availability. A number of new clones of the popular cellulose and amylose derivatives have entered the field with a special emphasis on a 3-chloro-5-methylphenyl carbamate derivative of both cellulose and amylose. Two new concepts in chiral stationary phase development have been introduced: one based on bonded derivatized cylcofructans and the second on a derivatized spiral chiral polymer. Several instruments introduced to support chiral screening and detection have also furthered the goals of achieving speed and efficiency.

The success of chiral separations is highly dependent on several factors, including the availability and performance of the appropriate CSPs, the knowledge base of the operator and the methods used to obtain results, including the instrumentation. In this review we would like to cover each of these aspects to reflect on what has occurred in the industry since the last chiral review in April 2008.¹

Continued discoveries of significant differences in effectiveness and toxicity of individual enantiomers in biological systems has maintained the importance of chiral separations. In addition, global governmental rules for dealing with compounds that have stereogenic centres has put a mechanism in place for the rapid advancement of chiral technology. As a result of these factors, the chiral knowledge base has grown exponentially. Developments in broadly applicable CSPs and continued expansion of the applications data base have created a more routine analytical methodology, which has also been aided by appropriate developments in instrumentation.

The current available classes of CSPs offer more generic approaches, although there are still certain structural types that require new concepts in CSP development. For example, it has been found to be difficult to separate compounds with stereogenic centres that are nonpolar — offering no hydrogen bonding or interactive groups — by HPLC, as well as a significant number of chiral primary amines. In addition, multiple chiral centres offer unique challenges — both analytically and preparatively. A call to expand the menu of chiral selectors was heard from many experts in the field at Pacifichem 2010.² On the plus side, advances in screening technologies — such as the use of SFC combined with MS, multiple column switching and parallel column screening systems — has offered better and faster outcomes: a major focus and goal of the pharmaceutical industry. The Express LC systems of Eksigent (Dublin, California, USA) offer an eight microcolumn screening system that claims to reduce method development time by factors of 10 with 1/10 the solvent consumption. Sepiatec GmbH (Berlin, Germany) also offers an eight parallel column HPLC screening system trademarked Sepmatix, as well as a similar eight column system for SFC.

Numerous publications on a range of chiral screening strategies and their evaluation,^3,4 new insights into mechanisms⁵ and molecular modeling⁶ have all helped to accelerate the advances in the chromatographic separation of stereoisomers. The cited papers represent only a small sample of the many publications available. The first book dedicated solely to the mechanisms of various CSPs, edited by Alain Berthod, was published by Springer in May 2010 called Chiral Recognition in Separation Methods, Mechanisms and Applications. While the emphases on single enantiomer production has remained in tack as a strategy, the pharmaceutical industry in the United States has now placed a greater emphasis on proteomics and biologics, while the small molecule effort has seen a major shift to Europe and Asia. Europe has seen a significant increase in chiral analysis. China is making a major effort to take its long history of effective natural products and bring their production into the 21st century, the development of which typically requiring chiral evaluations. Chiral publications in China from 2000–2009 exceeded the US by 177 to 142.² China and India are only now in the early stages of developing the necessary expertise, with India showing a somewhat more advanced progress in chiral analysis. Even though the apparent numbers of chiral compounds in the US pipeline has not increased significantly, the stress on analytical chemists to come up with chiral methods quickly and efficiently has not diminished. Little time or effort is given to optimizing a separation or searching for the optimum CSP. Speed and efficiency remain the targets. This seems like a flawed strategy when method redevelopment has to be done when the needs of the method changes, such as requiring a change in mobile phase type from normal phase (NP) to reversed phase (RP) in manufacturing processes or polar organic to SFC for preparative chromatography needs. Aspects of drug development require changes to tolerable mobile phase components.

The State of Chiral Screening Strategies

The success of chiral methods development is often complimented by the introduction of appropriate instrumentation. This includes everything from column heater-coolers — a very important operating parameter in chiral chromatography — to screening and detection systems, such as optical rotation (LC–OR), circular dichroism (LC–CD) and mass spectrometry (LC–MS).

In the last coverage of this subject, a number of approaches were being investigated, including sequential screening of four to as many as twelve CSPs under an established protocol, and the introduction of a number of parallel screening systems as well as packed capillary screening systems described in the previous review. Statistics are not available, but from a brief survey of users in the field, these options appear to have now settled down in general to a screen of four or eight chiral stationary phases with backup schemes when the first pass through fails. A number of recent publications have also introduced the concept of using an SFC tandem column screening system.^7,8 This seems to be a growing trend. In the one case,⁷ a two by four column system was operated in parallel, claiming that for five sequential samples it took 29 min on average for the method development of each sample. Screening of each four channel columns typically took 5 min and each optimization took ~15 min. Of the more than 100 samples processed by this method, the authors claimed remarkable success. The system was comprised of two banks of four columns (150 mm × 2.1 mm) that included ChiralPak AD-H, AS-H, ChiralCel OD-H, OJ-H (All available from Chiral Technologies, West Chester, Pennsylvania, USA) and the Chirobiotics V, T, R and TAG (from Sigma-Aldrich/Supelco, St Louis, Missouri, USA).

To further expand this generic screening activity, PDR-Chiral (Lake Park, Florida, USA) has developed a 12- or 24-column automated method screening system with a 10- or 20-bottle solvent mixer. The system can be operated as an HPLC or SFC system purchased with, or without, heating and cooling and is equipped with both UV and laser polarimeter detectors to identify chiral components and elution order. In a recent paper, four currently available HPLC chiral detectors; three polarimeters and a circular dichroism detector, where evaluated for their performance in chiral screening systems.⁹ The CD detector was considered the most useful as its response is both linear and sensitive except for non-UV analytes where the polarimeter offers a distinct advantage.

The battle for fast chiral analysis has taken many paths to date looking for the optimum solution. In the absence of sub-2 μm CSPs availability, a recent paper demonstrated the effectiveness of UHPLC for fast chiral separation of drugs by first derivatizing the analytes; rac amphetamine, rac methamphetamine and several β-blockers with AITC and Marfey's reagent respectively. The separation of the resulting diastereoisomers was run on the sub-2 μm Acquity C18 (Waters Corp. Milford, Massachusetts, USA) resulting in the separation of up to ten enantiomers in less than 3 min.¹⁰ While this may be a useful methodology for pure substances it presents its own analytical problems, especially for the clinical analysis of amines.

The State of Current CSPs

Of the more than 110 CSPs that have been introduced in the last 10 years, there are currently a handful of CSPs that have appeared to satisfy the majority of chiral separation needs of the industry. The derivatized cellulose and amylose phases still dominate the chiral market, while the macrocyclic glycopeptides applications have significantly expanded and taken on a complimentary role. Cyclodextrin derivatives, like in the Cyclobond (Sigma-Aldrich/Supelco) line, and π-complex CSPs like the Whelco-01 (Regis Technologies, Morton Grove, Illinois, USA) have largely fulfilled the balance of the needs, the choice depending on compound structure and the particular mobile phase requirements of the application. In some instances, as in the area of clinical applications, the protein phases can be added to the preceding list. From a review of the literature, many of these latter applications outperformed the cellulose and amylose phases based on sample solubility due to mobile phase components, selectivity and/or separation of multiple chiral centres or degradation products within the sample.

Polysaccharide Chiral Stationary Phases

At the time of the last review, Chiral Technologies (Daicel), West Chester, Pennsylvania, USA had just introduced the Chiralpak IC, a bonded 3,5-dichlorophenylcarbamate derivative of cellulose. This phase had never been made available as a coated version due to its high solubility in organic solvents. As a bonded column it has demonstrated good selectivity for those analytes previously not resolved or poorly resolved by the other immobilized cellulose and amylose derivatives. This column is now considered to be complementary to the other two bonded phases. More complex method development protocols and the expanded use of high levels of methyl-tert-butyl ether MTBE with this phase, however, limits its applicability in certain applications. Chiral Technologies has also added two new 5 coated derivatives to its line, the methylchlorophenyl carbamate of cellulose (Chiralpak AY-H) and of amylose (Chiracel OZ-H). These phases were originally available coated on 20 μm silica particles for preparative applications only. Figure 1 and 2 demonstrate a comparison to the original more standard dimethylphenyl carbamate derivatives of cellulose (OD) and amylose (AD) demonstrating enhanced selectivity. In addition, they have extended the Chiracel OD and Chiralpak AD to 3 μm material for LC–MS applications. A 50 mm × 4.6 mm column producing 7000 plates offers high-speed analysis with manageable pressures. Stability testing up to 250 bar (3625 psi) showed no deterioration in aqueous mobile phase systems.

Figure 1: Separation of methyl 1-benzyl-5-oxo-3-pyrrolidinecarboxylate enantiomers. Red trace = Chiralcel OZ-H. Blue trace = Chiralcel OD-H. Column dimensions = 250 mm × 4.6 mm; mobile phase = 80:20 hexane–ethanol; flow rate = 1.5 mL/min; temperature = 25 °C. (Courtesy of Chiral Technologies).

Macrocyclic Glycopeptide Phases

An HPLC method development strategy and applications update had been published in 2009¹¹ for the macrocyclic glycopeptides (Chirobiotic phases) from Supelco/Sigma-Aldrich. The article cites their unique breath and versatility in a variety of mobile phase conditions suitable for clinical applications using LC–MS methods, drug discovery screening and preparative methods. The article identifies the potential mechanisms at work and the mobile phase that drives those specific interactions. For these and other phases in their line, Sigma-Aldrich/Supelco has developed a new search engine for obtaining chromatographic information on a particular chiral analyte of interest, compound class or applications on a particular chiral column. This information can be found on the Sigma-Aldrich website under chiral e-Times. They also now offer ideas, tips and tricks to chiral chromatographers on Twitter. For more information on this latest innovation in chiral communication go to www.twitter.com/chiralchrom.

Figure 2: Separation of cyclandelate diastereomers. Red trace = Chiralpak AY-H. Blue trace = Chiralpak AD-H. Column dimensions = 250 mm × 4.6 mm; mobile phase = 90:10 hexane–ethanol; flow-rate = 1.5 mL/min; temperature = 25 °C. (Courtesy of Chiral Technologies).

Cyclodextrin-based Phases

Cyclodextrin technology has a long history of development dating back to the 1970s. It is ubiquitous in its breadth of applications both for chiral and achiral analytes. It continues to find its mark in current chiral processing. A surprise has been the unique opportunities it has afforded in SFC¹² for the π-acidic derivative Cyclobond DNP (Supelco). Another unexpected discovery was the ability of the Cyclobond DM, a fully methylated version with no hydrogen bonding capabilities, demonstrating good selectivity for single ring structures with bulky side arms and fused ring structures with a dominance of steric interactions as the driving force.^13–15

π-complex-based Phases

With the increasing interest in SFC, both for analytical and preparative applications, the π-complex phases have seen an increase in activity due primarily to their success in normal phase conditions. The Whelk-01 and -02 have been the work horses in this area. A recent study by Regis Technologies demonstrated the effect of varying co-solvents in SFC on a Whelk-O1 CSP. The study showed that and increase in polarity of the modifier correlated with a decrease in selectivity and a decrease in retention (Figure 3).

Figure 3: Effect of varying co-solvents in SFC method development on a Whelk-O 1 CSP. Group 1 = Troger’s base; Group 2 = chlormezanone; Group 3 = indapamide; Group 4 = devrinol. (Courtesy of Regis Technologies).

Preparative Applications

While SFC has taken a very substantial role in the chiral business it has been mostly used so far, though not exclusively, in preparative applications for its green benefits, as well as for its speed. Comments from analysts in the field indicated that the major reason that HPLC still seems to be the method of choice is the robustness, transferability and wide instrument availability of standard HPLC equipment. The rule seems to be that methods based on typical NP solvents or, to a lesser degree from the polar organic mode, can be transferred to SFC, while those developed with a RP system were less successful. In addition to speed and the reduction in solvent consumption it has substantially reduced the solvent waste stream. Development of a fraction collector for the inline concentration of purified enantiomers by Modular SFC, (North Attleboro, Massachusetts, USA) will certainly further its use and cost effectiveness.

There are still a significant number of applications requiring conventional HPLC and the recent introduction of a versatile bench-top simulated moving bed (SMB) system: the Semba Octave by Semba Biosciences (Madison, Wisconsin, USA) should help bridge the gap in cost effectiveness. Supelco (Bellefonte, Pennsylvania, USA) was a test site for this device and has now introduced a set of eight, 15 μm particle size, 5 mm × 10 mm columns designed to operate with this system or any bench-top HPLC system (Figure 4).

Figure 4: Columns = Astec Chirobiotic V2, 5 cm × 10 mm, 15 μm (set of eight); sample feed = 0.15 mL/min; mobile phase = methanol; sample = 5-methyl 5-phenylhydantoin (10 mg/mL); productivity = 20 mg/h for each enantiomer under above conditions. Increased throughput up to 60 mg/h was achieved by increasing concentration and focusing on either raffinate or extract. (Courtesy of Supelco/Sigma-Aldrich).

The Year of the Clones

With expiration of the Chiral Technologies (Daicel) patents on their highly successful coated versions of the derivatized cellulose and amylose phases, an increasing number of companies have attempted to produce generic versions with varying degrees of success.

The successful Kromasil AmyCoat, a 3,5-dimethylphenyl carbamate derivative of amylose, and CelluCoat, the 3,5-dimethylphenylcarbamate derivative of cellulose from Akzo Nobel (Bohus, Sweden), who have announced an exclusive agreement with SigmaAldrich/Supelco for exclusive distribution in the United States, Canada and Puerto Rico. In a recent poster, Akzo Nobel demonstrated the optimization of small-scale chiral HPLC preparative separation of tropic acid on a Knauer bench top SMB system (Berlin, Germany) using eight 4.6 mm × 150 mm (25 μm) AmyCoat columns. Total development time was said to be two working days with the Knauer SMB-Guide software.

Phenomenex (Torrance, California, USA) continues to expand the Lux chiral line that includes a 3-chloro-4-methylphenyl carbamate derivative of cellulose (Lux Cellulose-2) and 5-chloro-2-methylphenylcarbamate derivative of amylose (Lux Amylose-2). There is also, as part of this line, a replacement claimed to match the Chiracel OD-H from Chiral Technologies referred to as Lux Cellulose-1. At Pittcon 2011 they introduced Lux Cellulose-3, the tris(4-methylbenzoate) derivative and Lux Cellulose-4, the tris(4-chloro-3-methylphenylcarbamate). All these phases show complimentary selectivity.

ES Industries (West Berlin, New Jersey, USA) has also introduced a 3-chloro-5-methylphenyl carbamate derivative of cellulose referred to as Chromega Chiral CC2 coated on a 5 μm spherical 1000 Å silica that they claim has improved efficiency. This line has now been expanded to include a Chromega Chiral CC4, a 4-chloro-3-methylphenylcarbamate derivative.

Table 1: Examples of racemic separations on Epitomize CSPs.

Orochem Technologies (Lombard, Illinois, USA) has now introduced four products in this area, trade name Epitomize, that include the typical 3,5-dimethylpheny carbamate derivatives of both cellulose (CSP-1C) and amylose (CSP-1A) as well as an interesting 3-chloro-4-methylphenyl carbamate derivative, again of both cellulose (CSP-1Z) and amylose (CSP-1K). For a comparative enantiomeric response see Table 1. In addition to the usual variety of particle sizes and column lengths Orochem has added a UHPLC version for LC–MS using 1.7 μm particles (Figure 5). This format is available for all four coated versions with a pressure limit of ~3500 psi.

Figure 5: Separations achieved using Epitomize CSP-1C 1.7 and 5 μm. Left: Column = 250 mm × 4.6 mm, 5 μm; mobile phase = 10% isopropanol–heptane; flow-rate = 0.5 mL/min; back pressure = 330 psi; mobile phase used = 4.6 mL. Right: Column = 50 mm × 2.1 mm, 1.7 μm; mobile phase = 10% isopropanol–heptane; flow-rate = 0.5 mL/min; back pressure = 3620 psi; mobile phase used = 0.4 mL. (Courtesy of OraChem Technologies).

New Chiral Stationary Phases

A number of the earlier attempts to introduce new concepts in CSPs have all but disappeared, including chiral molecular imprinting. Chiral ionic liquids have been developed for a wide variety of achiral applications including their recent successful application in the chiral area for GC¹⁶ and HPLC.¹⁷ One of the consistent failures in the chiral separation business has been the separation of a significant number of primary amines. The crown ethers do a very good job in separating many of these types of compounds, but not all, and they fail to deliver on preparative, clinical and some manufacturing processes because of the lack of mobile phase volatility.

Dr Daniel Armstrong of the University of Texas at Arlington has now published on a new series of bonded CSPs based on derivatized cyclofructans and appears to have more than solved the problem of difficult chiral amines.^18,19 His new company AZYP (Arlington, Texas, USA) has produced four products in this series that have been launched this year at Pittcon. The first is based on an isopropyl-derivatized C6 cyclofructan, designated Larihc CF6-P, that efficiently separates chiral primary amines. The second is a highly aromatic derivatized C6 cyclofructan that loses its ability to separate these chiral amines but has excellent selectivity towards a broad spectrum of other types of racemates and is called Larihc CF6-RN.²⁰ The interesting fact is that the dominant mode of separation for both these phases is either NP or polar organic mode, both of which allow for direct transfer to efficient SFC methodologies. (Figure 6) With the isopropyl derivative on the CF6 over 20 chiral primary amines were separated to baseline in one day. Stability was addressed, indicating no change in performance in all common organic solvents even after over 1000 injections. These stationary phases also seem to have good sample capacity as demonstrated by the 4200 μg in 100 μL resolved to baseline on an analytical column, in this case limited only by the sample solubility.

Figure 6: SFC separation on isopropyl-CF6. Column = Isopropyl-CF6; mobile phase = 80:20 carbon dioxide–methanol with 0.1% diethylamine; flow-rate = 3 mL/min. (Courtesy of Dan Armstrong, University of Texas, Arlington, USA).

Toyko Kasei (TCI) from Tokyo, Japan has introduced a new concept in chiral stationary phase development with a spiral chiral polymer that offers steric recognition as the primary mechanism. Three different products result when chiral side groups are attached to the backbone adding additional mechanisms. In addition to steric recognition from the side chain, the three offer electrostatic interactions between an analyte containing a carbonyl group and a nitrogen side chain and π–π interaction between an analyte and the side chain. The polarity of the three products ranges from low for the TCI Chiral MB-S, to moderate for the TCI Chiral BP-S, to high for the TCI Chiral CH-S covering a broad range of chiral applications in either normal or reversed phase conditions. The products come in 3 μm and 5 μm formats.

A new approach for biomedical analysis of chiral compounds was demonstrated with the direct separation and quantification of lactic acid enantiomers in human urine. This was reported using for the first time a highperformance immunoaffinity LC–MS system. The antibody was immobilized onto a Poros-OH 20 μm support material from Perceptive Biosystems (Cambridge, Massachusetts, USA) activated and packed into a 4.6 × 150 mm stainless steel column. Pure alpha hydroxyl acids were injected and eluted with a buffer demonstrating the selective binding of the D-enantiomer but not the L. It was determined that the limit of detection when binding the D-hydroxy acid antibody for lactic acid enantiomers using MS as a detector was 63 μM.²¹

Conclusions

Chiral analysis continues to expand and to gain in speed and efficiency driven by the use of SFC and small column design at high linear velocity. Generic screening for selectivity remains the method of choice with four to eight columns operated either sequentially or in parallel with the availability of new screening instrumentation. New generic versions of the popular coated cellulose and amylose derivatives continue to enter the market. Two new concepts in CSP development have been introduced based on a bonded and derivatized six carbon cyclofructan and a derivatized spiral chiral polymer.

Thomas E. Beesley is founder and former CEO of Quantum Industries, which designed pre-absorbent and channeled TLC, which was sold to Whatman in 1978. He was founder and CEO of Advanced Separation Technologies which manufactured chiral, HPLC, Cyclobond and Chirobiotic HPLC columns and Chiraldex capillary GC columns. The company was sold to Sigma-Aldrich in 2006. He is currently a lecturer, expert witness and contract problem solver with a speciality in the separation of chiral molecules.

Column Watch editor, Ronald E. Majors, is a senior scientist at the Columns and Supplies Division, Agilent Technologies, Wilmington, Delaware, USA and is a member of LCGC Europe's editorial advisory board. Direct correspondence about this column should be addressed to "Column Watch", LCGC Europe 4A Bridgegate Pavilion, Chester Business Park, Wrexham Road, Chester CH4 9QH, UK or e-mail the editor, Alasdair Matheson, at amatheson@advanstar.com.

References

1. LCGC North America Supplemental Issue, April (2008).

2. M. Rouhi. C&EN, 89(1), 25–26 (2011).

3. H.A. Wetli and E. Francotte., J. Sep. Sci., 30, 1255–1261 (2007).

4. B.L. He and Y. Shi, Am. Pharm. Res., 11, 47–52 (2008).

5. C.R. Mitchell, N.J. Benz and S. Zhang, J. Chrom. B, 875, 65–71 (2008).

6. A. Berthod, S.C. Chang and D.W. Armstrong, Anal. Chem., 64, 395–404 (1992).

7. L. Zeng et al., J. Chrom. A, 1169, 193–204 (2007).

8. C.J. Welch et al., Chirality, 19, 184–189 (2007).

9. L. Kott et al., J. Pharma & Biomed Analysis, 43, 57–65 (2007).

10. D. Guillarme et al., Chirality, 22, 320–330 (2010).

11. T.E. Beesley and J.T. Lee, J. of Liq. Chromatogr & Related Technologies, 32, 1733–1767 (2009).

12. K.H. Gahm, Poster 78 of the 18th International Conference on Chirality, ISCD-18, Busan, S.Korea, June 24–28 (2006).

13. X. Han et al., Sep. Sci. Tech, 40(13), 2745–2759 (2005).

14. X Han et al., J. Chrom. A., 1063, 111–120 (2005).

15. D.D. Schumacher et al., J. Liq. Chromatogr. & Rel. Tech, 28(2), 169–186 (2005).

16. J. Ding, T. Welton and D.W. Armstrong, Anal. Chem., 76, 6819–6822 (2004).

17. S. Yu, S. Lindeman and C.D. Tran, J. Org. Chem., 73, 2576–2591 (2008).

18. P. Sun et al., Anal. Chem., 81, 10215–10226 (2009).

19. P. Sun et al., Analyst, 136, 787–800 (2011).

20. K. Kalikova et al., J. Chromatog. A, 1218, 1393–1398 (2011).

21. E.J. Franco, H. Hofstetter and O. Hofstetter, J. Pharma & Biomed. Analysis, 49, 1088–1091 (2009).