Preparation and evaluation of immobilized 4-methylbenzoylcellulose stationary phases for enantioselective separations

G Model

ARTICLE IN PRESS

CHROMA-357806; No. of Pages 7

Journal of Chromatography A, xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma

Preparation and evaluation of immobilized 4-methylbenzoylcellulose stationary phases for enantioselective separations Eric Francotte (Dr) ∗ , Tong Zhang 1 Novartis Pharma AG, Global Discovery Chemistry, 4002 Basel, Switzerland

a r t i c l e

i n f o

Article history: Received 8 April 2016 Received in revised form 30 July 2016 Accepted 3 August 2016 Available online xxx Keywords: Immobilized polysaccharides Chiral stationary phases Enantioselective chromatography

a b s t r a c t A photochemical method for immobilizing polysaccharide derivatives on silica gel has been developed and applied to 4-methylbenzoyl cellulose (PMBC). The photochemically immobilized materials have been used as chiral stationary phases (CSPs) for the chromatographic separation of the stereoisomers of chiral molecules. Through to the immobilization which makes the chromatographic material insoluble in almost all organic solvents, there no restriction regarding the kind of solvent used in the mobile phase. This feature permits to considerably extend the possibilities to improve the selectivity of the separations and or the solubility of the solute in the mobile phase. The influence of various parameters such as immobilization process, cross-linker type and amount on the chromatographic properties and chiral recognition ability of the resulting CSPs has been investigated using a set of chiral molecules. The impact of the amount of coated polysaccharide material on chiral recognition ability was also examined. © 2016 Published by Elsevier B.V.

1. Introduction Numerous chiral stationary phases are currently available for the analytical and preparative chromatographic separation of enantiomers [1,2]. With these tools, it is now possible to resolve almost all racemic compounds by chromatography [3]. While the technique has become the method of choice for the analytical determination of the optical purity of chiral compounds, it is also routinely applied as a means to obtain optically pure substance in preparative amounts up to several tonnes [4,5]. Among all available chiral stationary phases, those derived from polysaccharides [6–11] have been the most used ones, due to the easy modulation of their chiral recognition properties and their relatively high loading capacity [5]. However, one major drawback of the first generation of polysaccharide-based phases was their moderate to high solubility in many organic solvents such as tetrahydrofurane, dioxane, toluene, chlorinated solvents, or ethyl acetate. This feature considerably reduced the choice of mobile phase, thus limiting the possibility of increasing selectivity, of varying retention time, and of improving the solubility of the racemate. In order to

improve their properties with respect to this shortcoming, various approaches aiming to immobilize polysaccharide derivatives have been described. Okamoto and his group reported on first attempts to immobilize cellulose on silica gel through a dicarbamate linkage using diphenyl diisocyanate as a crosslinking agent in 1987 [12] but it appeared that the presence of the cross-linker negatively affects selectivity as the number of linkage increases. Improvement of this approach has permitted to prepare more stable and more efficient immobilized polysaccharide phases [13]. Other approaches have been developed, most of them includes the utilization of vinyl derivatives of polysaccharides. They have been reviewed a few years ago [11]. More recently another immobilization technique has been developed by Okamoto and his group and is based on the intermolecular polycondensation of polysaccharide derivatives and silica gel, both bearing triethoxysilyl groups [14,15]. We report here on the elaboration of a different immobilization concept for the preparation of insoluble polysaccharide stationary phases under very mild conditions, based on a photochemical approach. Preliminary and partial results arising from this concept have been described in an earlier patent [16].

∗ Corresponding author at: Novartis Pharma AG, WKL-122.P.25, Postfach, CH-4002 Basel, Switzerland. E-mail address: [email protected] (E. Francotte). 1 Present address: Chiral Technologies Europe, Parc d’Innovation, Bd. Gonthierd’Andernach, B.P. 80140, 67404 Illkirch, France. http://dx.doi.org/10.1016/j.chroma.2016.08.006 0021-9673/© 2016 Published by Elsevier B.V.

Please cite this article in press as: E. Francotte, T. Zhang, Preparation and evaluation of immobilized 4-methylbenzoylcellulose stationary phases for enantioselective separations, J. Chromatogr. A (2016), http://dx.doi.org/10.1016/j.chroma.2016.08.006

G Model CHROMA-357806; No. of Pages 7

ARTICLE IN PRESS E. Francotte, T. Zhang / J. Chromatogr. A xxx (2016) xxx–xxx

2

2. Experimental 2.1. Instrumentation The high-performance liquid chromatography (HPLC) systems used in this study consisted of 1) a Shimadzu SPD-6AV pump, a variable wavelength Shimadzu LC-6A UV–vis detector in series with a Perkin-Elmer (Model 241) polarimeter and a Reodyne injector fitted with a 20-microliter sample loop; the apparatus was connected to a PC and the data were managed with the GINA-NT (Bional AG, Dietikon, Switzerland) chromatographic software; 2) a Shimadzu LC-10AD pump, a variable wavelength Shimadzu SPD-10A UV–vis detector in series with a Jasco OR-990 chiral detector and a Reodyne injector fitted with a 20-microliter sample loop; the apparatus was connected to a PC loaded with the Shimadzu chromatographic software (version 1.62). 2.2. Synthesis of the stationary phases 2.2.1. Low molecular weight cellulose 150 g of microcrystalline cellulose (Avicel, Merck) are suspended in a mixture of 1.2 l toluene, 150 ml acetic acid, and 3 ml trifluoromethylsulfonic acid. 450 ml of acetic acid anhydride is added dropwise to this suspension. This mixture is then stirred at room temperature for 15 h. The resulting cellulose triacetate suspension is filtered, washed tree times with methanol, and dried under vacuum at 80 ◦ C for 6 h. Cellulose triacetate obtained as above are suspended in 1.3 l isopropanol. 320 ml of hydrazine monohydrate is added dropwise to this suspension. The mixture is stirred at 60 ◦ C for 60 h. The suspension is filtered, washed twice with isopropanol and once with methylene chloride, and dried under vacuum at 80 ◦ C for 6 h. Yield: 142 g low molecular weight cellulose. The degradation degree of the cellulose was characterized by a sharp decrease in viscosity of its para-methylbenzoate derivative from 0.75dl/g to 0.43dl/g, measured in chloroform at 25 ◦ C. 2.2.2. Preparation of (2,5-dihydro-3,4-dimethyl-2,5-dioxo-pyrrol-1-yl)-acetyl chloride (cross-linking spacer A) 183.2 g of (2,5-dihydro-3,4-dimethyl-2,5-dioxo-pyrrol-1-yl)acetic acid [17] are suspended in 360 ml toluene. The solution is heated under reflux for 16 h using a water separator. During that period, approximately 40 ml of toluene/water are distilled off azeotropically. The solution is then cooled to 70 ◦ C and 76.3 ml of thyonyl chloride are added dropwise in the course of 90 min. As soon the evolution of gas has ceased (approx. 2 h), the temperature is increased to 90 ◦ C for 2 h and then to 110 ◦ C for 30 min. After cooling, the solution is concentrated. The liquid residue is distilled and the fraction boiling at 182–184 ◦ C is collected. Yiel: 172.5 g (85.5%). Elemental analysis; Calc.: C 46.6; H 4.00; N 5.95; O 23.81; Cl 17.58. Found: C 49.42; H 4.21; N 6.71; O 22.93; Cl 16.83. 1 H NMR (CDCl3): 2.00 (s, CH3), 4.63 (s, CH2) 2.2.3. Preparation of 4-(2,5-dihydro-3,4-dimethyl-2,5-dioxo-pyrrol-1-yl)-benzoic acid chloride (cross-linking spacer B) 48 g (0.35 mol) of 4-aminobenzoic acid are dissolved in sodium hydroxide solution (14 g of NaOH in 300 ml water). To that mixture, a solution of 44.2 g of dimethylmaleic acid anhydride in 300 ml of dimethylacetamide is added dropwise with stirring. The solutionis heated at 90 ◦ C and, after 1.5 h, 175 ml of aqueous hydrochloric acid (2N) are added. The solution is cooled to room temperature and stirring is switched off. The crystalline product which has precipitated is filtered off, washed with water and dried in vacuo at 60 ◦ C. Yield: 73.6 g (85.7%). Melting point: 230–231 ◦ C. 73.6 g of this inter-

mediate are suspended in 700 ml of dry toluene. 32 ml of thionyl chloride are added dropwise to this suspension at 70 ◦ C. As soon as the evolution of gas has ceased (approximately 2 h), the temperature is increased to 80 ◦ C for 2 h. After cooling, the solution is concentrated using a rotary evaporator. The solid residue is recrystallized from toluene and then dried at 60 ◦ C. Yield 88%. Melting point: 199–200 ◦ C. 1 H NMR (CDCl3): 2.08 (s, CH3), 7.68 (d, phenyl), 8.20 (d, phenyl). 2.2.4. Preparation of 4-[(2,5-dihydro-3,4-dimethyl-2,5-dioxopyrrol-1-yl)methyl]-benzoic acid chloride (cross-linking spacer C) 100 g (0.66 mol) of 4-(aminomethyl)-benzoic acid are dissolved in sodium hydroxide solution (26.4 g of NaOH in 300 ml water). To that mixture, a solution of 83.3 g of dimethylmaleic acid anhydride in 500 ml of dimethylacetamide is added dropwise with stirring. The solution is heated at 90 ◦ C and, after 1.5 h, 330 ml of aqueous hydrochloric acid (2N) are added. The solution is cooled to room temperature and stirring is switched off. The crystalline product which has precipitated is filtered off, washed with water and dried in vacuo at 60 ◦ C. Yield: 155 g (90%). Melting point: 182–183 ◦ C. 120 g of this intermediate are suspended in 1000 ml of dry toluene. 50 ml of thionyl chloride are added dropwise to this suspension at 70 ◦ C. As soon as the evolution of gas has ceased (approximately 2 h), the temperature is increased to 80 ◦ C for 2 h. After cooling, the solution is concentrated using a rotary evaporator. The solid residue is recrystallized from toluene and then dried at 60 ◦ C. Yield: 155 g (80%). Melting point: 98–99 ◦ C. 1 H NMR (CDCl3): 1.98 (s, CH3), 4.72 (s, CH2), 7.45 (d, phenyl), 8.06 (d, phenyl). 2.2.5. Reaction of cellulose with acid chlorides (cross-linking spacers A, B or C) In a brown glass reactor, 3 g of low molecular weight cellulose prepared above are dried for 4 h at a bath temperature of 125 ◦ C with nitrogen flushing. The dried cellulose is suspended in 120 ml of dry pyridine and mixed with one of the above acid chloride derivative (cross linking spacers A, B or C) in the presence of catalytic amount of 4-(dimethylamino)-pyridine (0.2 ml). 151 ml triethylamine were then added dropwise to the suspension. The mixture is stirred for 24 h at 80 ◦ C. After cooling, the mixture is poured in 1 l methanol and the cellulose ester derivative is isolated by filtration and washed with methylene chloride. Various degrees of substitution have been obtained depending on the amount of acid chloride added. The products have been characterized by elemental analysis. 2.2.6. Esterification of the residual hydroxyl groups of the intermediates of cellulose obtained from cellulose and the cross-linking spacers 1–3 with 4-methylbenzoyl chloride 2.57 g of the intermediate derivative of cellulose are suspended 86 ml of dry pyridine and 22 ml of triethylamine in the presence of 10 mg of 4-(dimethylamino)-pyridine. 20 ml of 4-methylbenzoic acid chloride were added dropwise to this suspension and the mixture was stirred under nitrogen for 42 h at 60 ◦ C. After cooling, the mixture was poured into 200 ml methanol and the precipitate is filtered off. The crude product is dissolved in 200 ml of methylene chloride and the solution is filtered. The product is precipitated by addition of methanol to this solution. The precipitate is filtered off, washed with methanol and dried under high vacuum. The products have been characterized by elemental analysis. 2.4. Coating of the cellulose derivatives Two methods were used for coating, namely the evaporation and the precipitation techniques, which have been described in details in a previous paper [18].

Please cite this article in press as: E. Francotte, T. Zhang, Preparation and evaluation of immobilized 4-methylbenzoylcellulose stationary phases for enantioselective separations, J. Chromatogr. A (2016), http://dx.doi.org/10.1016/j.chroma.2016.08.006

G Model CHROMA-357806; No. of Pages 7

ARTICLE IN PRESS E. Francotte, T. Zhang / J. Chromatogr. A xxx (2016) xxx–xxx

Fig. 1. Structure of the cross-linking spacers.

Coating of the cellulose derivative on silica gel has been performed similarly for the three derivatives. Typically, 1 g of the cellulose derivative was dissolved in a mixture consisting of 2 ml of methanol and 20 ml of tetrahydrofuran. The obtained solution is divided into three equal portions. The first portion is mixed with 3.0 g of silica gel (Nucleosil Si 4000, 7 ␮m, from Macherey-Nagel), preliminarily modified with 3-aminopropyl-triethoxysilane [19]. The suspension is evaporated under vacuum in a rotavapor. The same procedure is repeated in the same manner with the two other portions. 2.5. Photochemical immobilization Immobilization of the silica-coated materials was achieved similarly for all cellulose derivatives. About 3–4 g of silica coated material are suspended in a mixture consisting of 80 ml of methanol and 240 ml of water in a 750 ml glass flask. The suspension is stirred with a mechanic stirrer and irradiated using a high pressure immersing mercury lamp (HPK 125, Philips) surrounded by a pyrex cooling jacket. After 16–20 h, the suspension is filtrated off and treated without stirring with dichloromethane for 16 h. The suspension is filtrated, washed with 100 ml of tetrahydrofuran. The solid residue is suspended again in 30 ml of dichloromethane. During moderate magnetic stirring, 300 ml of hexane were added dropwise to the above suspension (1.6 ml/min). After complete addition, the suspension is filtrated and dried at 120 ◦ C for 2 h. The amount of immobilized cellulose material is determined by elemental analysis (carbon content). 2.6. Column packing and testing Packing of the columns (stainless-steel column, 250 × 4 mm I.D., Macherey-Nagel) with the CSP materials obtained as above was performed using the slurry method in ethanol under constant pressure (150 bar). The mobile phase consisted of different mixtures as indicated in the Tables 1 and 2. Chromatographic runs were performed at ambient temperature at a flow rate of 0.7 ml/min. Tri-tert-butylbenzene peak was taken as a reference to determine the dead time of the system. 3. Results and discussion 3.1. Synthesis of the chiral stationary phases The designed immobilization strategy is based on the introduction of a photochemically reactive group, namely a dimethylmaleimide moiety (Fig. 1), which has proved to be a very efficient photochemically reactive crosslinker for various polymeric materials [20]. This moiety has the advantage to be stable under normal day light and over a large range of temperature, permitting the derivati-

3

sation of the other hydroxy groups without initiating spontaneous cross-linking during chemical or physical modifications of the materials. When exposed to UV treatment it rapidly reacts in a 2 + 2 cycloadditon reaction leading to the formation of a cyclobutane ring, causing a crosslinking of the polymer. Several cross-linker spacers have been synthesized and investigated in order to evaluate their influence on the chiral recognition properties (Fig. 1). All cross-linking spacers are constituted of the dimethylmaleimide moiety carrying an alkyl or aryl acid chloride group on the nitrogen atom. Preparation of the chiral stationary phases was achieved in five steps as shown in Fig. 2: i) derivatisation of cellulose by reaction with the acid chloride function of the cross-linking spacer, ii) reaction of the residual hydroxy functions on the polysaccharides with 4-methylbenzoyl chloride, iii) coating of silica gel with the obtained polysaccharide derivative, iv) photochemically induced cycloaddition of the dimethylmaleimide group, causing a crosslinking of the cellulose chains and v) conditioning of the material as a chromatographic stationary phase. Previous to derivatization, the microcrystalline cellulose was subjected to a degradation treatment in order to lower the molecular weight of the cellulose polymer (degree of polymerization) and to narrow its molecular weight distribution. It is indeed known that a too high degree of polymerization and a broad molecular weight distribution negatively affects the chiral recognition ability of the polysaccharide-based stationary phases [21]. The polymer degradation is achieved in two steps, starting by acetylation of the cellulose under acidic conditions and followed by cleavage of the ester with hydrazine monohydrate in isopropanol. The degradation degree of the cellulose was characterized by a sharp decrease in viscosity of its para-methylbenzoate derivative from 0.75 dl/g to 0.43 dl/g, measured in chloroform at 25 ◦ C. Table 1 summarizes the properties of the immobilized 4methylbenzoyl cellulose (PMBC) phases which were prepared according to the general synthetic scheme in Fig. 2 for the purpose of this study. 3.2. Structure of the tested racemates The chiral recognition ability of the individual CSPs was assessed by using a series of test racemates. The structures of the racemic compounds is shown on Fig. 3. For the purpose of comparing the chiral recognition ability of the differently immobilized materials, compounds which were known to be separated on the ‘classical’ non-immobilized cellulose 4-methylbenzoate CSP were selected. 3.3. Chromatographic properties of the CSPs 3.3.1. Influence of the type of cross-linking spacer The Influence of the type of cross-linking spacer was investigated by comparing the chromatographic properties of CSPs prepared from coated 4-methylbenzoyl cellulose containing the cross-linker structure A, B or C in a defined amount, before and after photochemical treatment (immobilization). The compositions of the coated CSPs are summarized in Table 1, A–C referring to the crosslinking agent (Fig. 1) and a or b denoting after or before irradiation respectively. For the photochemical treatment, a device consisting of a glassflask and equipped with a high pressure immersing mercury lamp surrounded by a pyrex cooling jacket was used. The silica gel coated chiral stationary phases were suspended in the inert solvent mixture methanol-water (80 ml/240 ml) and irradiated during 20 h at room temperature. The PMBC-0 column consists of the pure coated 4methylbenzoate of cellulose. It does not contain any cross-linker moiety and was not subjected to a photochemical treatment. It was used as a reference.

Please cite this article in press as: E. Francotte, T. Zhang, Preparation and evaluation of immobilized 4-methylbenzoylcellulose stationary phases for enantioselective separations, J. Chromatogr. A (2016), http://dx.doi.org/10.1016/j.chroma.2016.08.006

G Model

ARTICLE IN PRESS

CHROMA-357806; No. of Pages 7

E. Francotte, T. Zhang / J. Chromatogr. A xxx (2016) xxx–xxx

4

Table 1 Description of the prepared PMBC chiral stationary phases. CSP

Crosslinking spacer unit

Cross-linker content (average equivalent/glucose unit)

Coating method

Coating amount (weight%)

Immobilization yield (weight%)

Amount of coated/bonded cellulose derivative vs. silica (weight%)

PMBC-0 PMBC-A9a* PMBC-A9b* PMBC-B19a* PMBC-B19b* PMBC-C38a* PMBC-C38b* PMBC-C-P-25 PMBC-C-P-21 PMBC-C-P-16.7 PMBC-C-E-21

no A A B B C C C C C C

0.00 0.09 0.09 0.19 0.19 0.38 0.38 0.36 0.36 0.36 0.36

Evaporation Evaporation Evaporation Evaporation Evaporation Evaporation Evaporation Precipitation Precipitation Precipitation Evaporation

20 25 25 25 25 25 25 25 21 16.7 21

– 94.8 n.i 80.8 n.i 96.4 n.i 86.1 84.6 85.0 88.9

20 23.7 25 20.2 25 24.1 25 21.5 17.8 15 18.7

a*: after irradiation; b*: before irradiation; n.i. not immobilized. Table 2 Chromatographic resolution data of compounds 1–9 on the coated 4-methylbenzoylcellulose CSP (PMBC-0) and on the immobilized 4-methylbenzoylcellulose CSPs PMBC-A, −B, and −C; Mobile phase (MP): I) Hexane-Isopropanol (90/10 v/v); Flow rate: 0.70 ml/min, II) hexane/dichloromethane/2-propanol (85/15/2 v/v/v). Race-mate

1 2 3 4 5 6 7 8 9

PMBC-A9b MP I

PMBC-A9a MP I

PMBC-B19b MP I

PMBC-B19a MP I

PMBC-B19a MP II

PMBC-C38b MP I k1

␣

k1

␣

k1

␣

k1

␣

1.94 1.00 2.60 2.66 3.68 1.78 7.7 4.20 2.09

2.47 11.26 1.59 1.37 1.30 1.70 1.42 1.86 1.61

2.36 1.05 2.30 2.43 3.52 3.52 1.03 4.21 1.92

2.22 7.22 1.49 1.30 1.26 1.36 1.39 1.69 1.59

0.70 0.34 0.46 2.00 1.10 1.51 0.33 1.66 1.61

2.05 2.59 2.17 1.18 1.00 1.16 1.27 1.81 1.58

1.99 0.99 2.81 2.92 3.93 1.88 1.15 4.53 2.12

2.65 11.31 1.53 1.39 1.35 1.85 1.41 1.89 1.65

k1

␣

k1

␣

k1

␣

k1

␣

k1

2.17 0.89 2.21 2.27 3.43 1.65 1.01 3.95 1.79

2.00 5.69 1.31 1.26 1.24 1.30 1.40 1.65 1.51

2.84 1.30 1.96 2.40 3.35 2.69 1.03 4.62 1.84

1.67 2.71 1.21 1.20 1.16 1.00 1.35 1.49 1.43

1.79 0.81 2.17 2.27 3.30 1.57 0.97 3.63 1.73

2.31 9.58 1.49 1.34 1.30 1.56 1.39 1.80 1.59

2.26 1.01 2.06 2.29 3.44 1.94 0.98 4.11 1.75

2.04 5.55 1.40 1.27 1.24 1.35 1.35 1.60 1.56

0.64 0.36 0.42 2.10 1.16 1.52 0.31 1.82 1.47

1.89 2.22 1.95 1.18 1.07 1.13 1.26 1.69 1.63

PMBC-C38a MP I

PMBC-C38a MP II

PMBC-0 MP I

Table 3 Influence of coated amount and coating process on chromatographic results of the resolution of racemates 1–9. Mobile phase: Hexane/2-propanol (90/10 v/v); Column: 0.4 cm × 25 cm; Flow rate: 0.7 ml/min. Racemate

1 2 3 4 5 6 7 8 9

PMBC-P-25

PMBC-P-21

PMBC-P-16.7

PMBC-E-21

k1

␣

k1

␣

k1

␣

k1

␣

2.56 1.03 2.34 2.19 3.60 1.98 1.04 4.36 1.76

1.93 4.75 1.23 1.26 1.25 1.26 1.32 1.63 1.51

2.68 0.96 2.30 2.14 3.61 2.05 1.00 4.56 1.75

1.81 4.42 1.18 1.24 1.22 1.20 1.29 1.59 1.48

1.65 0.61 1.41 1.37 2.29 1.29 0.63 2.87 1.11

1.80 3.94 1.18 1.23 1.22 1.21 1.26 1.59 1.46

2.93 1.12 2.39 2.31 3.96 2.20 1.07 4.85 1.93

1.82 4.06 1.16 1.23 1.21 1.16 1.30 1.59 1.47

Table 2 shows the enantioselectivity (␣) and the retention factor k1 obtained for the test racemates 1–9 (Fig. 3) on the cellulosebased CSPs PMBC-A, −B, and −C before and after immobilization, using hexane/2-propanol (90/10 vol) as the mobile phase. Values obtained on PMBC-0 are also shown in Table 3 as reference. Data in Table 2 show that selectivities (␣) vary depending on the cross-linking spacer. In all instances, crosslinking agent A shows a poorer enantioselectivity while crosslinking agent C shows the highest enantioselectivity for all the tested compounds whatever before or after immobilization. Even though the CSPs made from crosslinking agent B is slightly inferior to the CSPs prepared from crosslinking agent C, it shows a similar chiral recognition ability, suggesting that the spacers containing an aromatic moiety have a positive effect on the chiral recognition mechanism. Moreover, selectivities obtained on the non-immobilized CSPs PMBC-Ab, −Bb, and −Cb are systematically higher than those observed on their immobilized counter parts. This indicates that the immobilization process probably causes slight changes in the supramolecular arrangement of the cross-linked macromolecular

assembly, affecting the chiral recognition power. Nevertheless, the immobilized phases retain a high chiral recognition ability, except for compound 6 on CSP PMBC-A9a. In terms of column efficiency (N), values ranging between 100 and 400 plates per column were measured for the immobilized phases, depending on the substances. These values are rather low and about 50% less compared to the coated phase, containing no crosslinking agent (PMBC) when packed under identical conditions. No particular efforts were done to optimize the packing conditions, the focus being on enantioselectivity which was the critical factor for assessing the value of the type and amount of incorporated cross-linking moieties. 3.3.2. Influence of the coated amount and coating technique Coating has been achieved using two different methods, namely the evaporation and the precipitation techniques, which have been described in details in a previous paper [17]. In order to determine the influence of the coated amount and coating technique on the chiral recognition performance of the immobilized PMBC phases,

Please cite this article in press as: E. Francotte, T. Zhang, Preparation and evaluation of immobilized 4-methylbenzoylcellulose stationary phases for enantioselective separations, J. Chromatogr. A (2016), http://dx.doi.org/10.1016/j.chroma.2016.08.006

G Model

ARTICLE IN PRESS

CHROMA-357806; No. of Pages 7

E. Francotte, T. Zhang / J. Chromatogr. A xxx (2016) xxx–xxx

5

Fig. 2. Synthesis and immobilization process of the CSPs.

O

Cl

N

O N

Et

O N

Cl

3

2

1

O

CH3

OCH3

O

OH

O

OH HO

5

4

O

7

HO

Et

O N H

N

6

O

8

9

Fig. 3. Structures of the tested racemates.

five different CSPs were prepared either by the precipitation or by the evaporation method, and applying amounts of cellulose derivative ranging between 16.7 and 25% on silica gel. For the preparation of all CSPs, 4-methylbenzoyl cellulose containing 0.36 equivalent of cross-linker C per unit of glucose was used (Table 1, PMBC-C-P and PMBC-C-E series).

Immobilization was achieved similarly to the other PMBCderived CSPs by suspending the coated materials in a methanolwater mixture and irradiation for 20 h with a mercury lamp. After extraction of the non-bonded material, the solid phase was filtered and then re-suspended in 30 ml of methylene chloride to which a 7-fold volume of hexane was added dropwise. The characteristics

Please cite this article in press as: E. Francotte, T. Zhang, Preparation and evaluation of immobilized 4-methylbenzoylcellulose stationary phases for enantioselective separations, J. Chromatogr. A (2016), http://dx.doi.org/10.1016/j.chroma.2016.08.006

G Model

ARTICLE IN PRESS

CHROMA-357806; No. of Pages 7

E. Francotte, T. Zhang / J. Chromatogr. A xxx (2016) xxx–xxx

6

(a)

0

5

10 15 20 25 30 35

(b)

0

2

4

6

Time (min.)

8

10 12 14

Time (min.)

Fig. 4. Separation of the enantiomers of racemic glutethimid 8 on the silica gel immobilized 4-methylbenzoyl cellulose PMBC-B19. Column 0.4 cm × 25 cm; Flow rate 0.7 ml/min. Mobile phase: a) hexane/2-propanol (90/10 v/v); b) hexane/dichloromethane/2-propanol (85/15/2 v/v/v).

of the prepared CSPs are summarized in Table 1. Interestingly, all CSPs show similar degree of immobilization. Racemate 1–9 were injected on the four columns under the same conditions and using a mixture of hexane/2-propanol (90/10 v/v) as the mobile phase. The obtained chromatographic data are shown in Table 3 and indicate that increasing the amount of coated material from 16.7% to 25% causes a slight increase of enantioselectivity, on average 4%, except for racemate 2 (troegers’ base) which is much more affected (+21%). Both coating techniques precipitation and evaporation (CSPs P-21 and E-21) shows similar ␣ values, indicating that this parameter does not substantially affect the chiral recognition ability, except again for racemate 2 which senses more the influence of the coating process. 3.3.3. Modulating enantioselectivity with the mobile phase type The expected advantage of the immobilized phases is obviously their ability to be applied with a broad variety of mobile phases, including those which are excluded with the non-immobilized phases such as chlorinated solvents, ethyl acetate, tetrahydrofuran, dichloromethane, dimethylformamide, due to their high solubility in these solvents. The possibility to utilize such ‘forbidden’ solvents in mobile phases is exemplified by the data reported in Table 2 which shows the chromatographic results for the separations of the enantiomers of racemic compound 1–9 with a mixture of hexane/dichloromethane/2-propanol 85/15/2 (v/v/v) as the mobile phase on the two CSPs PMBC-B19a (0.19 equivalent of crosslinking spacer B per unit of glucose) and PMBC-C38a (0.38 equivalent of crosslinking spacer C per unit of glucose). These results have been compared with those obtained on the same stationary phases (PMBC-B19a and PMBC-C38a), using the solvent mixture hexane/2-propanol 90/10 (MP I), which is typically applied with the non-immobilized phases (Table 2). Good enantioselectivities were generally observed for the tested racemates on both columns using the ‘classical’ mobile phase hexane/2-propanol (9/1, v/v), although CSP PMBC-C38 systematically exhibits slightly higher values. Applying the mixture hexane/dichloromethane/2-propanol 85/15/2 as the mobile phase,

good enantioselectivities were also observed and again, CSP PMBCC38 shows slightly higher enantioselectivity values except for compound 5 and to a much lower extent for compound 9. Interestingly, the mobile phase containing dichloromethane gives a better separation for compounds 3 and the hypnotic drug 8, demonstrating that the immobilized phases can become superior in terms of selectivity thanks to the possibility to apply other mobile phase types. This is not a general rule, but knowing that the mobile phase type and composition can have a dramatic influence on enantioselectivity [22], it considerably opens the space for improving chiral separations as well for accurate analytical purpose as for preparative separations for which the solubility of the racemic substrate is often a limitation for the productivity of the process. As an example, Fig. 4 shows the separation of the enantiomers of the chiral hypnotic drug glutethimide (compound 8) under two different mobile phase conditions.

4. Conclusion An effective method for immobilizing 4-methylbenzoyl cellulose (PMBC) on silica gel has been developed by incorporating photochemically reactive groups into the cellulose structure. Immobilization was successfully achieved by irradiation after conditioning the cellulose derivative as a suitable chromatographic support. The photochemically immobilized materials generally show a good chiral recognition ability for a variety of racemic compounds, even though there is a slight decrease of the selectivity after immobilization compared to the non-immobilized equivalent. However, the major advancement of these immobilized phases is their ability to tolerate almost all kind of organic solvents, permitting to considerably extend the possibilities to improve the selectivity of the separations by varying the mobile phase type and composition. This has been demonstrated for a few compounds, although the potential to improve selectivity has not been exhaustively exploited for the other tested racemic compounds.

Please cite this article in press as: E. Francotte, T. Zhang, Preparation and evaluation of immobilized 4-methylbenzoylcellulose stationary phases for enantioselective separations, J. Chromatogr. A (2016), http://dx.doi.org/10.1016/j.chroma.2016.08.006

G Model CHROMA-357806; No. of Pages 7

ARTICLE IN PRESS E. Francotte, T. Zhang / J. Chromatogr. A xxx (2016) xxx–xxx

References [1] E.M. Carreira, H. Yamamoto, Comprehensive chirality Separations and Analysis, vol. 8, Elsevier, Amsterdam, 2012. [2] E. Francotte, Chiral stationary phases for preparative enantioselective chromatography, in: G.B. Cox (Ed.), Preparative Enantioselective Chromatography, Blackwell Publishing, Oxford, 2005, pp. 48–77. [3] E. Yashima, C. Yamamoto, Y. Okamoto, Polysaccharide-based chiral LC columns, Synlett (1998) 344–360. [4] G.B. Cox, Preparative Enantioselective Chromatography, Blackwell Publishing, Oxford, 2005. [5] E. Francotte, Enantioselective chromatography as a powerful alternative for the ‘Preparation’ of drug enantiomers, J. Chromatogr. A 906 (2001) 379–397. [6] Y. Okamoto, Y. Kaida, Resolution by high-performance liquid chromatography using polysaccharide carbamates and benzoates as chiral stationary phases, J. Chromatogr. A 666 (1994) 403–419. [7] Y. Okamoto, E. Yashima, Polysaccharide derivatives for chromatographic separation of enantiomers, Angew. Chem. Int. Ed. 37 (1998) 1020–1043. [8] E. Yashima, Polysaccharide-based chiral stationary phases for high-performance liquid chromatographic enantioseparation, J. Chromatogr. A 906 (2001) 105–125. [9] C. Yamamoto, Y. Okamoto, Optically active polymers for chiral separation, Bull. Chem. Soc. Jpn. 77 (2004) 227–257. [10] J. Shen, Y. Okamoto, Efficient separation of enantiomers using stereoregular chiral polymers, Chem. Rev. 116 (2016) 1094–1138. [11] T. Ikai, C. Yamamoto, M. Kamigaito, Y. Okamoto, Immobilized polysaccharide-based chiral stationary phases for HPLC, Polym. J. 38 (2006) 91–108. [12] Y. Okamoto, R. Aburatani, S. Miura, K. Hatada, Chiral stationary phases for HPLC: cellulose tris(3,5-dimethylphenylcarbamate) and tris(3,5-dichlorophenylcarbamate) chemically bonded to silica gel, J. Liq. Chromatogr. 10 (1987) 1613–1628.

7

[13] Shouwan Tang, G. Liu, X. Li, Z. Jin, F. Wang, F. Pan, Y. Okamoto, Improved preparation of chiral stationary phases via immobilization of polysaccharide derivative-based selectors using diisocyanates, J. Sep. Sci. 34 (2011) 1763–1771. [14] H.-T. Qu, J.-G. Li, G.-S. Wu, J. Shen, X.-D. Shen, Y. Okamoto, Preparation and chiral recognition in HPLC of cellulose 3,5-dichlorophenylcarbamates immobilized onto silica gel, J. Sep. Sci. 34 (2011) 536–541. [15] S. Tang, Y. Okamoto, Immobilization of cellulose phenylcarbamate onto silica gel via intermolecular polycondensation of triethoxysilyl groups introduced with (3-glycidoxypropyl)triethoxysilane, J. Sep. Sci. 31 (2008) 3133–3138. [16] E. Francotte, Photochemically cross-linked polysaccharide derivatives as supports for the chromatographic separation of enantiomers, Patent (1996) WO 9627615 A1. [17] M. Baumann, V. Kvita, M. Roth, J.S. Waterhouse, Maleimide derivatives, Patent Ger. Offen. (1976) DE 2626795 A1 1976 1230. [18] E. Francotte, T. Zhang, Supramolecular effects in the chiral discrimination of meta-methylbenzoyl cellulose in high-performance liquid chromatography, J. Chromatog. A 718 (1995) 257–266. [19] Y. Okamoto, M. Kawashima, K. Hatada, Useful chiral packing materials for high-performance liquid chromatographic resolution of enantiomers: phenylcarbamates of polysaccharides coated on silica gel, J. Am. Chem. Soc. 106 (1984) 5357–5359. [20] J. Finter, Z. Haniotis, F. Lohse, K. Meier, H. Zweifel, A new class of photopolymers with pendent dimethylmaleimide groups III. Comparative study of different photopolymers with pendent olefinic structures including dimethylmaleimide groups, Angew. Makromol. Chem. 133 (1985) 147–170. [21] N. Enomoto, S. Furukawa, Y. Ogasawara, H. Akano, Y. Kawamura, E. Yashima, Y. Okamoto, Preparation of silica gel-bonded amylose through enzyme-catalyzed polymerization and chiral recognition ability of Its phenylcarbamate derivative in HPLC, Anal. Chem. 68 (1996) 2798–2804. [22] E. Francotte, D. Huynh, Immobilized halogeno-phenylcarbamate derivatives of cellulose as novel chiral stationary phases for enantioselective drug analysis, J. Pharm. Biomed. Anal. 27 (2002) 421–429.

Please cite this article in press as: E. Francotte, T. Zhang, Preparation and evaluation of immobilized 4-methylbenzoylcellulose stationary phases for enantioselective separations, J. Chromatogr. A (2016), http://dx.doi.org/10.1016/j.chroma.2016.08.006