Resolution of Enantiomers of (RS)-Baclofen by Ligand-Exchange Thin-Layer Chromatography

Abstract

A new chromatographic method has been developed for direct enantioresolution of (RS)-baclofen by ligand-exchange thin-layer chromatography (TLC) adopting two different approaches; (A) TLC plates were prepared by mixing the ligand exchange reagents (LER) with silica gel slurry and the chromatograms were developed with different achiral solvents or solvents having no chiral additive, and (B) the LER consisting of Cu(II)–l-amino acid complex was used as chiral mobile phase additive and the plain plates of silica gel having no chiral selector were used. Cu(II) acetate and four l-amino acids (namely, l-tryptophan, l-histidine, l-proline and l-phenylalanine) were used for the preparation of LERs. Spots were located by the use of iodine vapor. Effect of temperature and the mole ratio of Cu(II)-to-amino acid on enantioresolution were also studied. The results for the two methods have been compared, and the issue of involvement of the Cu(II) cation for the best performance of the two methods has been discussed with respect to the same mobile phase. l-Trp proved to be a good ligand using a common mobile phase in each case.

Introduction

There exists a general awareness among the scientific community, at least, about the consequences of the existing or potential differences between pharmacodynamics and pharmacokinetics of the enantiomers of racemic drugs. Nowadays, the regulatory agencies in many countries, involved in the registration of new active ingredients, insist on registration of single enantiomers of a new drug and ask the pharmacologists to present full information on the stereochemistry and stereoselectivity of both enantiomers including the necessary stereoselective analytical methods. Nevertheless, synthetic racemic drugs are being approved for marketing and pharmaceutical applications, of course, with the appropriate demonstration and justification that the racemates and/or fixed amounts of enantiomers are not toxic or do not have any other harmful effects.

For the same reasons, the investigations are focused on the development of new processes for production and separation of chiral substances. Depending upon commercial viability, two alternative approaches can be considered to develop one of the single enantiomers, (i) enantioselective synthesis of the desired enantiomer or (ii) separation of both isomers from an enantiomeric mixture. The economic viability of a separation/resolution method would remain an important point of consideration. Nevertheless, efficient methods of enantioseparation are always required to control the enantiomeric purity, or to separate the target molecule or one of its chemical precursors (obtained from conventional synthetic procedures), or for monitoring the completion of enantioselective reaction process (since the production of single enantiomers is a real difficult task).

Both direct and indirect approaches of enantioseparation (using LC) have their own advantages and limitations depending upon the situation, particularly, in terms of sources and amount of samples available, chemical structure of the analyte and the ease of availability of laboratory facilities. Experience suggests that both high-performance liquid chromatography (HPLC) and thin–layer chromatography (TLC) techniques should be considered complementary rather than competing, with TLC being used widely in analytical research and in routine quality control of pharmaceuticals. Little attention has been paid to TLC despite its advantages. Advantages of TLC comparing to HPLC in pharmaceutical and drug analyses (1) and in enantiomeric analysis in particular (2) have been described in the literature. Some of the advantages of TLC include semi-preparative isolation of the separated compounds and photography of the chromatogram as a clearly visible evidence of separation. Preparative HPLC generally requires high capital investment.

At present, chiral stationary phases (CSPs) and bonded phases are by no means cheap because the development costs are formidable. The CSPs, which are already on the market, have relatively limited applications and restricted performance. Besides, high amounts of HPLC grade solvents are required for elution of the desired compound from such columns.

Use of an enantiomerically pure compound as chiral impregnating reagent constitutes one of the bases of separation of enantiomers in TLC; it is a less exploited area that has great potential. The adsorption characteristics are changed (while the chiral environment is created) without covalently affecting the inert character of the adsorbent. Different approaches of impregnation of thin silica gel plates with various l-amino acids and certain other enantiomerically pure compounds for direct enantioresolution of a variety of racemic compounds by TLC have been reviewed (2).

Based on the work of Davankov et al. (3, 4) who modified commercial HPLC columns by grafting alkyl derivatives of α-amino acids such as n-decyl-l-histidine or n-hexa decyl-l-proline, onto the resin for enantiomeric resolution (explained as ligand-exchange chromatography, LEC), a new chiral selector (2S,4R,2′RS)-N-(2′-hydroxy dodecyl)-4-hydroxy proline was developed by Degussa AG, Germany (5) and was used for ligand-exchange TLC resolution of enantiomeric amino acids (6–9).

Applications of l-amino acids as chiral impregnating reagents (2) in TLC and complexes of l-amino acids with a metal ion, particularly Cu(II), as chiral ligand exchange reagents (LER) or as chiral mobile phase additives both in TLC and HPLC have been reviewed for direct enantioresolution of a variety of racemic compounds (10). The presence of certain functional groups (such as amino, hydroxy or carboxy) on both the selector and the analyte is considered to be the precondition for successful enantioseparation by LEC via the formation of diastereomeric ternary mixed metal complexes (11).

Baclofen

It is a γ-aminobutyric acid analog and is chemically known as 4-amino-3-(4-chlorophenyl) butyric acid (Figure 1). It is extensively used as a stereoselective agonist for GABA_B receptor (12, 13). It has been used as a muscle relaxant, and also used to treat spasticity due to multiple sclerosis, cerebral and spinal cord injury, cerebral palsy and complex region pain syndrome (14, 15). Bac is marketed and used as a racemic mixture. However, it is claimed that only the (R)-enantiomer is stereoselectively active on GABA_B-receptors and is more active than the (S)-enantiomer (16, 17).

Figure 1.

The structure of (RS)-Baclofen.

Literature reports on HPLC separation of enantiomers of (RS)-Bac by a direct approach include use of chiral stationary phase consisting of Crownpak^® CR (18), macrocyclic antibiotic teicoplanin (19) and use of chiral mobile phase consisting of aqueous copper(II) acetate and N,N-di-n-propyl-l-alanine (20) and d-penicillamine-based ligand exchange chiral columns (21). Enantioseparation of (RS)-Bac by an indirect approach has been reported using CDRs such as (S)-naproxen chloride (22), o-phthaldialdehyde combined with N-acetyl-l-cysteine (23), 1-fluoro-2,4-dinitrophenyl-5-l-alaninamide (24) and N-(4-chloro-6-piperidinyl-[1,3,5]-triazine-2-yl)-l-phenylalanine (25). Besides, dichloro-s-triazine and monochloro-s-triazine (26) reagents have also been used as CDRs for enantioseparation of (RS)-Bac by using reversed-phase HPLC.

The literature dealing with TLC analysis/separation of certain important enantiomeric drugs (27), along with the literature cited above and the references cited therein, reveals a scope for developing new sensitive, simple TLC methods of enantioseparation of (RS)-Bac though it has a high prescription rate and is easily available. Moreover, to the best of authors' knowledge, there are no reports available on the enantioseparation of (RS)-Bac by ligand-exchange TLC in the literature. Because of the lack of literature reports (particularly by ligand exchange TLC) on sensitive enantioseparation of (RS)-Bac, it was considered to establish validated experimental protocols for analytical enantioseparation of (RS)-Bac. Considering these aspects and the structure of Bac, the following experiments were planned and carried out.

Present work

l-tryptophan (l-Trp), l-histidine (l-His), l-proline (l-Pro) and l-phenylalanine (l-Phe) were chosen for preparing chiral LERs (considering their success in enantioresolution of certain β-blockers (28) by ligand-exchange TLC). TLC was performed by the following two different approaches: (A) TLC plates were prepared by mixing the LER with silica gel slurry and the chromatograms were developed with different solvents, and (B) the LER consisting of Cu(II)–l-amino acid complexes was used as the chiral mobile phase additive and the plain plates of silica gel having no chiral selector were used. To the best of authors' knowledge, this is the first report for direct enantioresolution of (RS)-Bac adopting these two approaches (impregnation by LER and use of LER in the mobile phase as an additive). By choosing one mobile phase, the performance of the two methods was compared to discuss the issue of involvement of Cu(II) for the best resolution.

Experimental section

Chemicals and instrumentation

(RS)-Bac; l-Trp, l-His, l-Pro and l-Phe were obtained from Sigma–Aldrich (St Louis, MO, USA). pH was measured with a Cyberscan 510 pH meter (Singapore), UV–visible spectrometry was performed with a Hitachi U2001 spectrometer (Tokyo, Japan) and optical rotations were determined by using a polarimeter (model Kruss P3001RS; Germany). Silica gel G with 13% calcium sulfate as binder, having chloride, iron and lead impurities up to 0.02% and with pH 7.0 in a 10% aqueous suspension, was obtained from Merck (Mumbai, India). The other chemicals and reagents used were of analytical-reagent grade and were obtained from SISCO Research Laboratory (Mumbai, India), Merck (Mumbai, India) and BDH (Mumbai, India). Stock solutions were prepared in double-distilled water and purified (18.2 MΩ cm) with a Milli Q system of Millipore (Bedford, MA, USA).

Thin-layer chromatography

Preparation of Cu(II)–l-amino acid complex

l-Trp, l-His, l-Pro and l-Phe were chosen as chiral selectors. Solutions of l-amino acid (4 mM) and copper(II) acetate (2 mM) were prepared in purified water–methanol (95:5); the solutions of Cu(II) and each of the chiral selectors were mixed in a ratio of 1:2 and the pH of the solution was maintained by adding a few drops of ammonia. In total, four LERs were prepared. The UV spectra for the complexes were different from the UV spectrum of the Cu(II) solution.

Preparation of TLC plates

The TLC plates were prepared as described below. TLC plates (10 cm height × 5 cm width, and 0.5 mm thickness) were prepared in the laboratory by spreading the slurry of silica gel G (25 g) with a Stahl-type applicator. The plates were allowed to set at room temperature and then were activated for 8–10 h at 60 ± 2°C. These were considered to be plain plates.

Use of LER in TLC

LER were used in the following manner.

1. Impregnation of plain plates by mixing the LER with slurry. Slurry of silica gel G (25 g) was prepared in the solution of each of the four LERs (50 mL), and the plates were prepared as described previously. The chromatograms were developed with different solvents.

2. LER as mobile phase additive. The LER consisting of Cu(II)–l-amino acid complexes was used as a mobile phase additive for developing the chromatograms and the plain plates of silica gel having no chiral selector were used.

Development of chromatograms

Solution of (RS)-Bac (10 µL, 10⁻² M, in 0.1 M NaHCO₃) was spotted on the TLC plates with the help of a 25-µL Hamilton syringe. The clean, dry and paper-lined rectangular glass chamber was pre-equilibrated for 15 min at each temperature fixed for chromatographic development inside an incubator and allowed to reach the specific temperature before development, for each experiment under approaches (A) and (B). The chromatograms were developed for 25–30 min.

Binary and ternary mixtures of methanol, acetonitrile, acetone, chloroform, water and dichloromethane were tried as mobile phase to achieve enantiomeric resolution. Chromatograms were dried in an oven at 40°C and then cooled to room temperature; spots were located in an iodine chamber. For separation of enantiomers of (RS)-Bac, optimization of conditions with respect to temperature and concentration of the chiral selectors was carried out.

To determine repeatability, solutions of known concentration of (RS)-Bac (10⁻² mM) were applied on TLC plates six times.

Mole ratio of Cu(II) to amino acid

Experiments were performed to optimize the ratio of l-amino acid to Cu(II) by using each of the amino acids in four different concentrations. The plates were prepared by using 1, 2, 4 and 6 mM concentrations of each of the amino acids with 2 mM of Cu(II) acetate; this provided the l-amino acid–Cu(II) ratio of 1:2, 1:1, 2:1 and 3:1. Further experiments were carried out by using a fixed concentration of amino acid (4 mM) and varying the concentrations of Cu(II) acetate (2, 4 and 6 mM); in this case, the ratio of l-amino acid–Cu(II) was 2:1, 1:1 and 2:3, respectively. While keeping the ratio of l-amino acid–Cu(II) as 2:1, the following combinations were also tried, 2 mM l-amino acid–1 mM Cu(II), 4 mM l-amino acid–2 mM Cu(II), 6 mM l-amino acid–3 mM Cu(II) and 8 mM l-amino acid–4 mM Cu(II).

Effect of temperature

The chamber was pre-equilibrated for 20–30 min with mobile phases at different temperatures, 16, 20, 24, 28 and 32°C (when the room temperature was ∼28°C). Each temperature was maintained/controlled using an incubator. The chromatograms so developed were processed as described previously.

Results

In this study, resolution of enantiomers of (RS)-Bac was achieved by two different LEC approaches [(A) and (B)], and different mobile phases were tried for each case. The two approaches were found successful for enantioseparation of (RS)-Bac via a ligand-exchange mechanism.

Mobile phases and hR_F (R_F ×100) values

The successful mobile phases and hR_F (R_F ×100) values are listed in Table I. The R_F values are the average of at least five experiments performed on the same day and on different days. The (S)-isomer was found to have a higher R_F value than the (R)-isomer. The resolution (R_S) of two adjacent spots was calculated by dividing the distance between the two spot centers with the sum of their radii; the two spots were considered to have reasonably separated when R_S > 1.2 (29). Table I shows that approach (A) provides better resolution.

Table I.

The Successful Mobile Phases and hR_F (R_F× 100) Values for Enatioresolution of (RS)-Bac Under Approaches (A) and (B)

Chiral selector	Approach (A)					Approach (B)
	Solvent ratio	hRF				Solvent ratio	hRF
		(R)	(S)	Pure (S)	Rs		(R)	(S)	Pure (S)	Rs
l-Trp	6:2:2	54	82	82	3.2	6:3:1	51	68	68	2.1
l-His	5:1:4	48	61	61	2.9	5:2:3	47	53	53	1.8
l-Pro	6:1:3	58	79	79	3.0	7:2:1	43	56	56	1.9
l-Phe	4:2:4	41	54	54	2.7	4:2:4	34	27	27	1.2

Solvent system, MeCN–CHCl₃–MeOH; solvent front, 8.5 cm; development time, 25–30 min; R_s = resolution; hR_F = R_F × 100; detection, by iodine vapors.

The performance of the two methods was compared for the best resolution by choosing one mobile phase. It was found that one particular ratio of the three solvents (MeCN–CHCl₃–MeOH) was hardly successful in all the situations; a variation in the ratio of the three solvents was to be made to achieve resolution. The representative photographs of chromatograms showing resolution of (RS)-Bac by use of Cu(II)–l-Trp complexes are shown in Figure 2.

Figure 2.

Photographs of chromatograms showing resolution of (RS)-Baclofen by use of Cu(II)–l-Trp complex by approach (A) and approach (B). Experimental details are given in Table I. From left to right: Spot 1: lower spot for (R)-enantiomer and the upper spot for (S)-enantiomer resolved from the racemate; Spot 2: pure (S)-isomer. This figure is available in black and white in print and in color at JCS online.

Effect of temperature

To study the effect of temperature on resolution of enantiomers, it was varied systematically and the effect was noted by observing eight-shaped spots, tailing or clear resolution. Experiments carried out at 16, 20, 24, 28 and 32°C using the successful solvent systems (MeCN–CHCl₃–MeOH) showed that the best enantioresolution of (RS)-Bac was obtained at 28°C under two approaches with any of the chiral selectors used in these studies. The decrease in temperature to 16°C showed eight shaped structures or no resolution and the increase in temperature to 32°C resulted in tailing of spots.

Effect of mole ratio of Cu(II) to amino acid

The ratio of l-amino acid to Cu(II) was optimized for (RS)-Bac. It was found that the best resolution (R_S) was achieved when (4 mM) l-amino acid and (2 mM) Cu(II) were used, i.e. a ratio of 2:1.

Precision and limit of detection

The relative standard deviation (RSD) was between 1.35 and 1.75%. Different concentrations of (+)-enantiomer were spiked into a fixed concentration of (−)-enantiomer in the range of 0.1–2% using standard solutions of the two isomers to establish detection limits. Using approach (A), the chromatograms were developed followed by visualization with iodine vapors. l-Trp was used as a chiral selector and MeCN–CHCl₃–MeOH (6:2:2, v/v) was used as the mobile phase. Detection was successful up to 0.5%.

Discussion

Ligand exchange TLC resolution of (RS)-Bac

Davankov and Rogozhin (30) introduced chiral ligand-exchange chromatography (CLEC). Separations by means of CLEC are based on the formation of labile ternary metallic complexes in the mobile phase and/or in the stationary phase. TLC enantiomeric separations based on ligand exchange were published by Günther et al. (6, 7) and Weinstein (31) in 1984 for the first time. Although the procedures differed in their choice of chiral selectors and range of applicability, they had a very similar methodology. Separation models developed for ligand exchange HPLC (32, 33) are also valid for TLC. The chosen chiral selector (the l-amino acid, in this study too) has carboxylic and amino functional groups, which are capable of interacting with the metal ion, Cu (II). In approaches (A) and (B), two chiral selector amino acid molecules act as chelating ligands for the Cu(II) ion. In the course of the enantioseparation, one chelating amino acid molecule is replaced by the competing enantiomer molecule of the separated enantiomeric mixture. This results in the formation of diastereomeric complexes of the two enantioseparated compounds having different stabilities and thus the enantioresolution is observed. This explanation has been described in the literature for the separation of enantiomers of α-amino acids and α-hydroxy acids and is well satisfactory for the separation of enantiomers of the (RS)-Bac, under study, as suggested by its structure. On the other hand, the mechanism of separation involves a series of complexation equilibria in the mobile phase and in the stationary phase (34) when the LER is used as a mobile phase additive; it has been reported that the enantioseparation takes place because of different affinities of the diastereomeric complexes for the stationary phase rather than the stereoselectivity observed in solution (35).

Thus, it can be contended that the mechanism of enantioselectivity in the LEC depends on whether the chiral ligand is linked on the stationary phase or it is added in the mobile phase. The stability of the diastereomeric complexes formed in LEC is higher than the stability of the diastereomeric adducts formed by other chiral selectors (36). Schmid et al. (37) proposed formation of ternary mixed metal complexes between the selector and the analyte in LEC.

Many interdependent factors are considered to affect the formation of complexes. The better results with approach (A) are in agreement with the previous explanations for enantioresolution of certain β-blockers by ligand-exchange TLC (28). In approach (A), the formation of ternary complex is probably taking place during the chromatographic development, as the Cu(II) is available at the same time to the l-amino acid (present in the impregnated form) and to the analyte molecule spotted on the plate. The two approaches are successful for enantioseparation via the same ligand exchange mechanism for which the method of impregnation and successive formation of ternary complexes are different.

Conclusion

The results indicate that the TLC method using the ligand-exchange approach via impregnation or by adding LER to the mobile phase can be applied for the detection of each enantiomer in low amounts up to 0.5%. The use of home-made plates is quite common for successful and less expensive routine works. The method presented herein provides a rapid and effective approach in planar mode for the control of enantiomeric purity of (RS)-Bac (and other structurally similar pharmaceuticals), which can be realized even in a small laboratory. The method is less expensive and simple in comparison to existing methods requiring either chiral HPLC columns or derivatization reactions for indirect mode of enantioseparation. Both the approaches presented herein have additional advantages that many other different chiral selectors (e.g., l-, or d-amino acids) can be easily tested to resolve the analyte and only a small amount of ligands is required which results in increased rate of ligand exchange and enhanced separation efficiency. Chromatographic separation and detection take place separately in TLC, which enables us to carry out the analysis at different times and to make full use of detection techniques for the analysis of constituents. An advantage of TLC is the possibility of localization of all constituents on chromatograms contrary to HPLC where lack of detection may be observed in the case of the retention of a constituent on the column.

Conflict of Interest statement. The authors certify that they have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript; this includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript.