Journal of Chromatography A, 1358 (2014) 208–216

Contents lists available at ScienceDirect

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

Single-step approach for fabrication of vancomycin-bonded silica monolith as chiral stationary phase Ming-Lung Hsieh a , Lai-Kwan Chau a,b,∗ , Yung-Son Hon a,1 a

Department of Chemistry and Biochemistry, National Chung Cheng University, 168 University Road, Min-Hsiung, Chia-Yi 62102, Taiwan, ROC Center for Nano Bio-Detection and Advanced Institute of Manufacturing with High-tech Innovations, National Chung Cheng University, 168 University Road, Min-Hsiung, Chia-Yi 62102, Taiwan, ROC b

a r t i c l e

i n f o

Article history: Received 17 March 2014 Received in revised form 28 June 2014 Accepted 2 July 2014 Available online 9 July 2014 Keywords: Capillary electrochromatography Vancomycin Monolithic column Sol–gel process Chiral separation

a b s t r a c t A vancomycin-bonded silica monolithic column for capillary electrochromatography (CEC) was prepared by a single-step in situ sol–gel approach. This sol–gel process incorporates a synthetic sol–gel precursor which contains a macrocyclic antibiotic, vancomycin, to form a porous silica network inside a fused-silica capillary. To avoid degradation of vancomycin during the column fabrication, a mild step was adopted into the sol–gel process. The performance of the vancomycin chiral stationary phase was investigated by CEC in both the reversed-phase mode and the normal-phase mode. The vancomycin chiral stationary phase was optimized with respect to vancomycin loading in the reversed-phase mode for chiral separation of thalidomide enantiomers. The best efficiency and resolution values of 94 600 plates/m and 5.79, respectively, were achieved. The optimized column was further applied to chiral separation of alprenolol enantiomers. A plate height of less than 7 ␮m for the first eluted enantiomer of alprenolol was obtained in an aqueous mobile phase at a flow rate of 0.74 mm/s. Using enantiomers of seven ␤-blockers and some other basic enantiomers as test analytes, separation efficiencies of up to 148 100 plates/m in the reversed-phase mode and up to 138 100 plates/m in the normal-phase mode were achieved. © 2014 Elsevier B.V. All rights reserved.

1. Introduction The separation of racemic compounds is a very interesting topic of research in analytical chemistry, especially in the pharmaceutical field. Therefore, development of analytical chiral separation methods offering high efficiency and high resolution in a short time is highly desirable. In this perspective, capillary electrochromatography (CEC) is attractive as it combines the high efficiency of capillary electrophoresis (CE) and high selectivity of high-performance liquid chromatography (HPLC) [1–3]. Enantioselectivity in CEC can be achieved by the interactions with an appropriate chiral selector either bound or adsorbed to the capillary wall (open-tubular capillary) [4,5] or to a stationary phase (packed capillary or monolithic column) [6–10].

∗ Corresponding author at: Center for Nano Bio-Detection and Advanced Institute of Manufacturing with High-tech Innovations, National Chung Cheng University, 168 University Road, Min-Hsiung, Chia-Yi 62102, Taiwan, ROC. Tel.: +886 5 2720411x66411; fax: +886 5 2721040. E-mail address: [email protected] (L.-K. Chau). 1 Dedicated to the memory of Professor Yung-Son Hon (1955–2011). http://dx.doi.org/10.1016/j.chroma.2014.07.003 0021-9673/© 2014 Elsevier B.V. All rights reserved.

Monolithic stationary phases are increasingly considered as a viable alternative for columns packed with particles in HPLC and CEC because of their easy preparation, excellent properties and high performance [11–15]. Depending on the nature of the monolithic material, two major classes of monolithic columns can be identified: (1) organic polymer-based monoliths and (2) silicabased monoliths [11–15]. Organic polymer-based monoliths are created by a one-step polymerization of an organic monomer in the presence of a cross-linker, an initiator, and a porogen. A critical drawback associated with some organic polymer-based monolithic columns is its tendency to swell/shrink during exposure to the organic solvent in the running mobile phase. Such a structural change may reduce the mechanical stability and the permeability of the monolith. Silica-based monoliths, consisting of a silica skeleton and interconnecting macropores, generally are prepared using sol–gel technology. Inside the skeleton a large number of mesopores is present. The macropores can provide fast flow while the mesopores provide a large surface area which is necessary for a high sample loading. Additionally, the porous sol–gel network can offer high permeability, high efficiency, high mechanical strength, and good solvent resistance. The popularity of silica monoliths can be linked to the availability of different chemistries that can be used for surface modification and ligand attachment. As such, sol–gel

M.-L. Hsieh et al. / J. Chromatogr. A 1358 (2014) 208–216

processing presents an attractive inorganic alternative for preparation of silica-based monoliths [16]. Silica monoliths fabricated by sol–gel processing often incorporate the use of porogen, say, poly(ethylene glycol) (PEG), to manipulate phase separation and thus control macropore size and volume [11]. Monoliths thus formed generally possess high through-pore-to-skeleton size ratios to provide good column permeability and separation efficiency. The chiral selectors utilized for the preparation of silica-based monoliths ranged from cyclodextrin and its derivatives, macrocyclic antibiotics, chiral ion-exchangers, proteins, ligand exchange-chiral stationary phases, cellulose derivatives to molecular chiral imprinted polymer [17–25]. Up to now, the immobilization of a chiral selector in most of these CEC columns was based on in situ encapsulation or physical adsorption or post-modification of silica monoliths. In 2000, Hayes and Malik used a C18-silicon alkoxide to prepare monolithic octadecylated silica columns in a single step for CEC [26]. Our previous study used a sol–gel precursor with a quaternary ammonium functionality to prepare monolithic columns in a single step for anion-exchange CEC [27]. Recently, a single-step approach to prepare monolithic columns by sol–gel processing of an organofunctional silicon alkoxide precursor that contains a macrocyclic moiety, cyclodextrin, was also described by our group [28]. The powerful enantioseparation capability of glycopeptide antibiotics was first introduced by Armstrong and coworkers as chiral selectors in HPLC and later on widely applied to CE system for separation of enantiomers [29–32]. Macrocyclic glycopeptides are efficient chiral selectors for several reasons: (1) they contain ionizable functional groups, which can be either acidic or basic depending on pH; (2) they have multiple stereogenic centers; (3) they process numerous functional groups contributive to stereoselectivity; and (4) they contain both hydrophobic and hydrophilic group. Hence, the chiral separation mechanism to form transient noncovalent diastereomeric complexes with glycopeptides antibiotic is based on electrostatic interactions as well as secondary interactions such as hydrophobic, hydrogen bonds, dipole-dipole, ␲–␲ interactions, and steric repulsion. In which, vancomycin and teicoplanin are the most frequently used and have successfully applied to HPLC and also to CEC for chiral separation of compounds of pharmaceutical interest using packed stationary phases in the reversed-phase and the normal-phase modes [8,33–37]. So far, many vancomycin stationary phases (such as commercial Chirobiotic VTM and LiChrospher® diol silica modified with vancomycin) have been applied to CEC in the form of packed columns. Nevertheless, a comprehensive survey of enantioseparations by CEC using vancomycin or norvancomycin monolithic column shows only few publications [9,20,38–40]. Maruska and coworkers used an organic polymeric continuous-bed and postmodified with vancomycin as a chiral stationary phase (CSP) [9,38]. Dong and coworkers used a silica monolith and post-modified with vancomycin [20]. Ding and coworkers used a silica monolith and post-modified with norvancomycin [39,40]. Enantioseparations by HPLC using vancomycin monolithic column has also been reported by Pittler and Schmid using dynamic coating with a vancomycin derivative [41]. Although post-modified monolithic column has found applications in various fields, the elimination of such complicated tasks of chemical treatment is desirable. Here, we describe a new and easy-to-prepared method to fabricate a well-controlled functional 3D skeletal silica monolith based on a single-step in situ sol–gel process. The vancomycin CSP precursor for sol–gel processing was synthesized via addition reaction by covalently attaching triethoxysilyl group to the vancomycin moiety. Copolymerization of the vancomycin CSP precursor with tetramethoxysilane (TMOS) under acid-catalyzed sol–gel reaction results in a functionalized

209

monolithic column. During the sol–gel process, proper addition of PEG will cause phase separation [11]. Upon removal of PEG, a porous monolithic stationary phase for CEC separation is ready. Thus, the time and labor associated with column fabrication are reduced. To the best of our knowledge, we are the first to prepare vancomycin-bonded silica monolithic columns by a single-step in situ sol–gel process. The capillaries thus formed were tested for the CEC chiral separation of a number of racemates of pharmaceutical interest. In particular, chiral separation of the enantiomers of ␤-blockers has been successfully achieved with good efficiency and resolution in both the reversed-phase mode and the normal-phase mode.

2. Materials and methods 2.1. Chemicals and reagents All chemicals were of highest available purity and used directly without any further pretreatment. Tetramethoxysilane (TMOS) was obtained from Acros Organics (USA). ␤-Blockers (acebutolol, alprenolol, atenolol, labetalol, metoprolol, pindolol, propranolol), bromacil, bupivacaine, devrinol, thalidomide, and warfarin were obtained from Sigma–Aldrich (USA). Poly(ethylene glycol) (PEG) with an average molecular mass of 10 000, glacial acetic acid (HOAc), and 3-isocyanatopropyltriethoxysilane (ICNPTES) were obtained from Fluka Chemika (USA). N,N-Dimethylformamide (DMF), acetone, acetonitrile (ACN), ethanol, diethyl ether, and methanol (MeOH) were purchased from TEDIA Company (USA). Triethylamine (TEA) and Coumachlor were obtained from Riedel-De Haën (German). Vancomycin was purchased from Duchefa Biochemie (Nederlands). Triethylammonium acetate (TEAA) buffer was prepared by adjusting 0.1% or 1% solution of TEA with HOAc to the appropriate pH. Ammonium acetate solution was prepared by dissolving 10 mmol of acetic acid in water, adding ammonium hydroxide up to the desired pH value, and diluting to the final volume of 100 mL with water. The pH values of the running electrolytes were measured by an Orion 420A pH meter (USA). All aqueous solutions were prepared with water that had been purified by a Milli-Q water purification system (Millipore) with a specific resistance of 18.2 M cm. Aqueous mobile phases were prepared by adding the desired volume of the HPLC grade organic solvents to the pH controlled buffer solutions. Before use, all solutions were filtered through a 0.22 ␮m membrane (Corning) and degassed by vacuum and sonication. Polar organic mobile phases were prepared by adding a desired ratio of TEA and HOAc to a mixture of MeOH and ACN. Analyte stock solutions (4.0 mg/mL) were dissolved into the HPLC grade MeOH or ACN and stored at 4 ◦ C. The racemic samples for CEC experiments were daily diluted to the desired concentrations with an aqueous buffer or MeOH.

2.2. Synthesis of vancomycin CSP precursor To a stirred solution of vancomycin (60 mg, 0.04 mmol) in 0.5 mL dry DMF, ICNPTES (30 ␮L, 0.12 mmol) and TEA (12 ␮L, 0.07 mmol) were added. The resulting solution was stirred at room temperature under nitrogen for 24 h. After adding diethyl ether (6 mL) and stirring for another 5 min, the precipitate collected by centrifugation was rinsed by diethyl ether. The purified precipitate was dried under vacuum to afford the title compound as a light yellow solid and identified by a LTQ ion trap mass spectrometer (Thermo Electron Corp., CA). The product was then used as the CSP precursor for further sol–gel processing.

210

M.-L. Hsieh et al. / J. Chromatogr. A 1358 (2014) 208–216

2.3. Fabrication of vancomycin monolithic capillary column The approach used in this study is a modified version of our previous study [27,28] and the preparation procedure of monolithic columns was similar to those reported previously [42–45]. To prepare a vancomycin monolithic capillary column, a segment of a fused-silica capillary (75 ␮m i.d. × 375 ␮m o.d., Polymicro Technologies, USA) of a desired length was pretreated with 1 N NaOH rinse, followed by a water rinse, a 0.1 N HCl rinse, a water rinse, an acetone rinse and then dried with nitrogen. Then 500 ␮L TMOS was added to a solution of PEG (60 mg) in 0.01 M acetic acid (0.9 mL) and stirred at 0 ◦ C for 1 h. After storing at 0 ◦ C for another 7 h, a solution of vancomycin CSP precursor in 200 ␮L water was added to the TMOS sol–gel solution and the whole mixture was then vigorously stirred at 0 ◦ C for 5 min. Subsequently, the resulting homogeneous mixture was injected into the pretreated fused-silica capillary by a syringe pump (KD Scientific, USA) and both ends were sealed. The assembly was placed in an oven at 40 ◦ C for 1 h and then allowed to age at room temperature for 24 h. The functionalized silica monolith thus formed was then treated by a wash of water and methanol with a HPLC pump (Model 203, EverSeiko, Japan). The capillary was then stored at 4 ◦ C until use.

separations. The performance of the vancomycin column was evaluated in the reversed-phase mode, and an electroosmotic flow (EOF) study using a unretained marker, acetone, and chiral separation of a ␤-blocking drug, alprenolol, was also included. By a method similar to previous reports [8,34,36], linear velocities at various voltages and van Deemter curves were also investigated on the vancomycin monolithic column. Separations of racemic analytes were then investigated in both the reversed-phase mode and the normal-phase mode at the optimal mobile phase conditions. The enantiomeric resolution (Rs ) and the number of theoretical plates (N) were calculated according to Rs = 1.18 ×

N = 5.54 ×

t − t  R2 R1 w2 + w1

 t 2 R

w

where tR is the retention time and w is the peak width at half height. 3. Results and discussion 3.1. Synthesis of vancomycin CSP precursor

2.4. Characterization of vancomycin silica monolith Short lengths of the vancomycin column (5 mm) were cut off and dry at 90 ◦ C oven for 30 min. The structural morphology of the silica monolith inside the capillary was obtained at a fractured surface by scanning electron microscopy (SEM) (Hitachi S-4800, Japan) at 15 keV after vacuum sputtering (Hitachi E-1030, Japan) with gold/palladium for 20 s. The specific surface area, total pore volume, and average pore diameter of the mesopores of all the materials were determined by measurement of adsorption–desorption isotherms at 77 ◦ K with a Micrometrics ASAP 2020 micropore analyzer (USA) according to the BET method. Prior to nitrogen sorption analysis, the monoliths were crushed to a fine powder, lyophilized by a freeze dryer (Kingmech FD-2-12P, Taiwan) and degassed at 110 ◦ C for 4 h. The permeability of the column was examined by the pressure drop, according to Darcy’s law which leads to the definition of the specific permeability of the column [46]. When an aqueous mobile phase was forced through the capillary at a flow rate of 10 ␮L/min using a HPLC pump, the pressure drop of the vancomycin column was determined. 2.5. Electrochromatographic experiments All electrophoretic experiments were performed on a HP3D CE capillary electrophoresis instrument (Hewlett-Packard, USA) equipped with a diode array detector and an external pressurization system (N2 ). The columns for all electrophoretic experiments were prepared by jointing a monolithic capillary (25 cm) and a bare fused-silica capillary (8.5 cm) with teflon tubing. A detection window of about 2 mm in length was made by burning off the polyimide coating on the fused-silica capillary by a lighter. The detection wavelength was kept at 214 nm. In general, CEC separations were performed at 15 ◦ C at an applied voltage up to 25 kV and an equal pressure of 2 bar was applied at both ends of a capillary column. Prior to use, each new monolithic column was preconditioned with water or a running electrolyte degassed by sonication by pressurizing the column inlet with a HPLC pump, typically at a flow rate of 10 ␮L/min. Then the column was further conditioned electrokinetically in the CE instrument by driving a mobile phase intended for separation through the capillary at a desired separation voltage until a stable baseline was achieved, typically required 1–2 h. Sample solutions were injected under electrokinetic mode for 3–5 s at the electric field strength difference desired for the

Nucleophilic additions of amine or hydroxyl groups are known reactions and have been investigated for the derivatization of vancomycin [29,47]. Since vancomycin possesses nine hydroxyl groups (including three phenolic moieties) and two amine functions (one primary and one secondary), the ICNPTES-vancomycin precursor was synthesized by allowing the reaction to run with appropriate amounts of ICNPTES and TEA. The mass spectra of the CSP precursor exhibited fragment ions at 1635.75 and 1798.75, which correspond to the desired hydrolyzed products (ICNPTESvancomycin, [M+Na]+ and (ICNPTES)2 -vancomycin, [M+Na]+ ). The dominant 1798.75 peak suggests that the prevalent vancomycin CSP precursor form is (ICNPTES)2 -vancomycin. This vancomycin CSP precursor has two important features: (1) the multi-chirality sites and functional groups are capable of providing chromatographic interactions with racemic analytes and (2) the three ethoxy groups attached to the silicon atom can undergo hydrolysis and the following condensation, thereby facilitates the in situ creation of a vancomycin-bonded monolithic matrix throughout the entire solution-filled inner capillary volume. 3.2. Preparation and characterization of vancomycin monolithic column In the sol–gel process, the macropore volume can be manipulated by reducing the water fraction of the solvent phase to possess the appropriate through-pore-to-skeleton size ratio for chromatography. Although large through-pores provide high column permeability, they can also cause slow mobile-phase mass transfer and large eddy diffusion (increase of A-term) [2,48]. Finer domains may lead to better column efficiency (larger N or smaller H) due to smaller A-term and shorter path of diffusion (decrease of C-term) for a solute into and out of the stationary phase to minimize peak broadening at higher mobile phase flow rate [49,50]. The macroporous structure is formed shortly after gelation while dissolution-reprecipitation in a basic solution results in pore tailoring [51,52]. In an earlier attempt to prepare vancomycin monolithic column, we first used our previous approach with procedures similar to that of the ␤-cyclodextrin functionalized monolith [28]. Unfortunately, the column efficiency was merely adequate, though thalidomide enantiomers were resolved with acceptable enantioselectivity in CEC in an aqueous mobile phase, 0.1% TEAA, pH 4.5/ACN (80:20).

M.-L. Hsieh et al. / J. Chromatogr. A 1358 (2014) 208–216

211

Fig. 1. Scanning electron micrographs (cross-sectional view) of a vancomycin monolithic column (at magnification of 4000×).

The insufficient column efficiency by the first attempt is perhaps due to partially degraded macrocyclic moiety on the silica network under base-catalyzed pore tailoring reactions at high temperature. To improve this, the vancomycin monolithic column was fabricated with the adoption of a mild step into the preparation procedures which consist of: (1) decreasing the water to silica ratio and (2) applying a wash of water without ammonia during the heat treatment step [11,42–45]. By this improved method, the specific surface area and porosity of the vancomycin monolith were altered. According to the BET measurement, the improved monolith possesses a specific surface area of ∼280 m2 /g, which is smaller than that by our previous process (about 500 m2 /g) [28] and similar to that of columns packed with 5 ␮m silica particles (∼300 m2 /g) [11]. The amount of vancomycin CSP precursor (15–60 mg) used in the column fabrication was also investigated on chiral separation. In the first round of optimization, capillaries based on various loadings of vancomycin CSP precursor were employed in CEC in the reversed-phase mode, 0.1% TEAA, pH 4.5/ACN (80:20, v/v), for chiral separation of thalidomide enantiomers. Those with a loading of at least 30 mg result in enhanced resolution and efficiency. In the second round of optimization, one ␤-blocking drug, alprenolol, was baseline-separated in a mobile phase of 1% TEAA, pH 5.0/MeOH (10:90, v/v), when the vancomycin monolithic columns were fabricated with loadings of 30, 45 and 60 mg vancomycin CSP precursor, and the latter showed the best resolution. Thus, the optimum loading of 60 mg was used for subsequent vancomycin monolithic column preparations. Fig. 1 shows the typical SEM cross-sectional views of a vancomycin monolithic column. It can be seen that the monolith has continuous skeletons and large through-pores. The through-pore size is around 2.0 ␮m and the average size of the mesopores is determined to be 2.2 nm by N2 adsorption-desorption method. In addition, the permeability of the column was examined by the pressure drop [46]. When an aqueous mobile phase was used at a flow rate of 10 ␮L/min, the pressure drop of the vancomycin monolithic column was about 30 MPa. Such a pressure drop corresponds to a K value of 3.7 × 10−13 m2 , which is lower than that by our previous process (6.3 × 10−13 m2 ) [28] and higher than that of columns packed with 5 ␮m silica particles (typically, ∼4 × 10−14 m2 ) [11]. 3.3. Column evaluation Each new column was evaluated with a system suitability test prior to further exploratory experiments [34]. The test included chiral separation of thalidomide enantiomers and measurement

Fig. 2. Plot of (a) linear velocity versus applied electric field and (b) van Deemter curve for alprenolol with a vancomycin monolithic column. Conditions: capillary, 25 cm (overall length 33.5 cm) × 75 ␮m ID; mobile phase, 1% TEAA, pH 5.0/MeOH (10:90, v/v); sample injection, 5 kV, 5 s; detection wavelength, 214 nm; 15 ◦ C; 2 bar pressure support.

of linear velocities at various voltages for a non-retained neutral marker, acetone. Determination of retention time reproducibility and resolution was also calculated for each column. The efficiency and resolution of this new CSP were first investigated by examination of thalidomide enantiomers using an aqueous mobile phase, 0.1% TEAA, pH 4.5/ACN (80:20, v/v). Based on the reversed-phase mode conditions for chiral separation, MeOH or ACN as an organic additive with a buffer system of TEAA was subsequently explored for all studied racemic analytes. It is interesting to note that during the screening of analytes and running buffers, baseline separation of the enantiomers of ␤-blockers were achieved with reasonable retention times and satisfactory efficiency values in the 1% TEAA buffer with MeOH modifier. Since this is an uncommon mobile phase used in CEC, the effects of applied voltage on EOF, linear velocity, and plate height of the selected racemic analyte were examined on this vancomycin monolithic column. In this study, the electroosmotic mobility of a typical vancomycin monolithic column was determined to be 4.5 × 10−5 cm2 V−1 s−1 , when a buffer solution of 1% TEAA in pH 5.0/MeOH (10:90, v/v) was used as the mobile phase. Additionally, a higher EOF (6.4 × 10−5 cm2 V−1 s−1 ) was also obtained in a polar organic mobile phase of MeOH/ACN/TEA/HOAc (80:20:0.1:0.1 by volume). The observed velocity of alprenolol in the running buffer of 1% TEAA in pH 5.0/MeOH (10:90, v/v) with increasing applied field strength is seen to be linear (correlation coefficient, r = 0.9992) as shown in Fig. 2a, indicating the absence of both thermal and double-layer overlap effects [53]. A van Deemter curve obtained with the first eluting enantiomer of alprenolol is

212

M.-L. Hsieh et al. / J. Chromatogr. A 1358 (2014) 208–216

shown in Fig. 2b. The plate height (H) minimum was determined as 6.8 ␮m corresponding to 1.5 × 105 theoretical plates per meter at a linear velocity of approximately 0.74 mm/s. The result indicates that reduction in the skeleton size in this column, as such, should improve the column efficiency. The column was also tested for run-to-run reproducibility (n = 5). Results show that the relative standard deviation (RSD) values for retention time and resolution of thalidomide are 1.1% and 7.5%, respectively. For column-to-column (n = 4) reproducibility under the same experimental conditions, the RSD values for retention time and resolution are 9.4% and 11.8%, respectively, which are lower than the run-to-run reproducibility but acceptable.

Table 1 Racemic compounds examined on the vancomycin monolithic column in CEC using the reversed-phase mode. Racemate

Mobile phase

tR1 (min)

N1 (plates/m)

N2 (plates/m)

Rs

Acebutolol Alprenolol Atenolol Metoprolol Pindolol Propranolol Bupivacaine Coumachlor Thalidomide

a a a a a a a c a c d b d

24.4 18.7 31.6 19.0 17.6 21.9 17.4 35.4 21.4 14.4 27.2 27.6 42.2

104 900 102 000 76 300 83 100 93 700 74 500 84 600 112 300 93 100 94 600 114 200 136 700 147 200

101 100 101 500 55 000 82 200 50 200 67 000 65 600 88 900 55 200 64 100 103 700 127 200 148 100

2.41 2.61 2.47 1.86 1.84 1.68 6.15 7.59 7.20 5.79 12.69 5.32 9.75

Warfarin

3.4. Reversed-phase mode In order to compare the enantioselectivity of this new vancomycin monolithic column in CEC to that reported earlier [8,9,34,35], the evaluation was based on the reversed-phase mode. The influence of the mobile phase composition (pH, buffer concentration, and organic modifier content) on retention time, efficiency,

a 1% TEAA, pH 5.5/MeOH (10:90, v/v), 15 kV. b 0.1% TEAA, pH 4.5/ACN (80:20, v/v), 15 kV. c 0.2% TEAA, pH 4.5/ACN (80:20, v/v), 20 kV. d 100 mM ammonium acetate, pH 6.0/water/ACN (5:5:90, v/v/v), 15 kV.

Fig. 3. CEC enantioseparation of (a) thalidomide, (b) warfarin, (c) bupivacaine, (d) alprenolol, (e) metoprolol, and (f) pindolol on a vancomycin monolithic column. Experimental conditions: (a) 0.2% TEAA, pH 4.5/ACN (80:20, v/v), 20 kV, (b) 0.1% TEAA, pH 4.5/ACN (80:20, v/v), 15 kV, and (c)–(f) 1% TEAA, pH 5.5/MeOH (10:90, v/v), 15 kV; other conditions are the same as described in Section 2.5.

M.-L. Hsieh et al. / J. Chromatogr. A 1358 (2014) 208–216

213

Fig. 4. Effect of TEAA buffer pH on efficiency and resolution for separation of alprenolol enantiomers on a vancomycin monolithic column. Mobile phase, 1% TEAA at the desired pH/MeOH (10:90); applied voltage, 15 kV; all other conditions are held constant as described in Fig. 3.

Fig. 5. Evaluation of ACN content in the polar organic mobile phase for its effect on the resolution of racemic alprenolol and efficiency of the alprenolol firsteluting enantiomer on a vancomycin monolithic column by keeping the content of TEA/HOAc at 0.1/0.1, v/v. Experimental conditions are the same as described in Section 2.5.

and resolution were optimized. A number of basic, neutral, and acidic molecules were separated by utilizing a buffer system of TEAA/ACN mobile phase. Of the racemic analytes tested in the traditional reversed-phase mode conditions, best efficiency and resolution was achieved for thalidomide (94 600 plates/m and Rs = 5.79), coumachlor (112 300 plates/m and Rs = 7.59) and warfarin (136 700 plates/m and Rs = 5.32), as shown in Table 1 and Fig. 3. Other analytes were just partially resolved. The efficiency and resolution values are comparable to those obtained by the packed CSP using the commercial Chirobiotic VTM or organic polymeric continuous-bed modified with vancomycin [9,34]. In addition, using a mobile phase with a composition of 100 mM ammonium acetate solution, pH 6.0/water/ACN (5:5:90, v/v/v) [35,36], further

improved efficiency and resolution are achieved for thalidomide and warfarin enantiomers, as shown in Table 1. As expected, none of the basic compounds were resolved by this vancomycin CSP in the reversed-phase mode which consists of ACN predominantly. In previous reports, when the CSP was tested for chiral separation of several basic compounds using MeOH instead of ACN, a general increase in efficiency and resolution were observed [34–36]. To study the effect of the organic modifier added to the mobile phase on the retention time, efficiency, and resolution, ACN or MeOH was added to the TEAA buffer (pH 5.0) such that its concentration is in the range of 0.1% to 1.5% (initial concentration), and preliminary experiments on the chiral separation of thalidomide, bupivacaine, and alprenolol enantiomers were carried out.

Fig. 6. Chiral separation of four ␤-adrenergic blockers (a) acebutolol, (b) alprenolol, (c) pindolol, and (d) propranolol using the normal-phase mode on a vancomycin monolithic column. MeOH/ACN/TEA/HOAc (v/v/v/v): (a) and (c) 80:20:0.1:0.1, (b) 90:10:0.1:0.1, and (d) 95:5:0.1:0.1; applied voltage, 15 kV. For other conditions see Section 2.5.

214

M.-L. Hsieh et al. / J. Chromatogr. A 1358 (2014) 208–216

Table 2 Racemic compounds examined on the vancomycin monolithic column in CEC using the normal-phase mode. Racemate

Mobile phase

tR1 (min)

N1 (plates/m)

Acebutolol Alprenolol Atenolol Labetalol Metoprolol Pindolol Propranolol Bupivacaine Thalidomide

c b d c d c a c d

14.7 9.0 8.4 22.8 9.9 7.4 8.7 10.8 11.8

123 600 75 500 87 400 64 800 138 300 73 000 100 000 40 100 138 000

a

N2 (plates/m) 105 300 71 300 58 500 66 700 85 500 50 300 79 600 32 700 95 800

MeOH/ACN/TEA/HOAc (95:5:0.1:0.1, v/v/v/v), 15 kV. b MeOH/ACN/TEA/HOAc (90:10:0.1:0.1, v/v/v/v), 15 kV. MeOH/ACN/TEA/HOAc (90:10:0.05:0.05, v/v/v/v), 15 kV.

A general increase in efficiency and resolution by raising the content of organic modifier in the mobile phase was obtained, probably due to the stronger interaction of the analytes with the vancomycin CSP. Traditionally, MeOH in combination with the TEA/HOAc aqueous mobile phase will result in greater enantioselectivity and less retention due to its hydrogen-bonding capability [21]. In order to achieve chiral separation of the studied compounds, various concentrations of the TEAA buffer were examined. Results show that the EOF decreases while efficiency increases at higher ionic strength, leading to relatively long retention time. In summary, it was found that the maximum efficiency and resolution were reached at approximately 1% TEAA buffer with MeOH modifier. The increase of MeOH content causes a general increase in resolution and a maximum resolution for thalidomide was achieved at 80% of organic modifier, while the best enantioselectivity for bupivacaine and alprenolol was achieved at 90% MeOH. By using ACN, retention times were strongly reduced and no enantioseparation was observed. It is known that the vancomycin molecule has six pKa values of 2.9, 7.2, 8.6, 9.6, 10.4, and 11.7, the first three of which could possibly contribute to ionization changes on the molecule between the pH values commonly used [34]. In the synthetic route of the ICNPTES-vancomycin precursor, the isocyanate group can react with the two amino groups or aliphatic hydroxyl residues through the urea group or the carbamate group, respectively. Although this may only have a slight influence on the pKa values of all proteolytic groups, such an approach in preparing the vancomycin monolithic column may result in the net ionization of the vancomycin molecule to become more anionic and therefore contributes to positive EOF. The increase of the pH from 4.5 to 6.5 caused an increase of the EOF and a slight increase of the retention time. As can be observed in Fig. 4, the highest number of theoretical plates and resolution for alprenolol were achieved at pH 5.5. The results indicate that the pH of the TEAA solution is a very important parameter because it influences the charge of both the chiral selector and the racemic analytes, probably playing an important role in the electrostatic interactions involved in the chiral separation mechanism. Once the vancomycin CSP had been evaluated in this aqueous mobile phase, a number of racemic compounds, including acebutolol, atenolol, pindolol, and propranolol, were also examined. Fig. 3 shows the baseline separation of several enantiomers of ␤-blockers by using the vancomycin monolithic column in a mobile phase of 1% TEAA, pH 5.5/MeOH (10:90, v/v). The optimum experimental conditions found for the above racemic compounds were also tested for other enantiomers, good efficiency and resolution values were achieved as shown in Table 1. 3.5. Normal-phase mode Since Armstrong and Owens et al. demonstrated the potential of the polar organic mobile phase for the macrocyclic antibiotic CSP in CEC, there have been numerous reports in the literature

Rs (This study) 2.73 2.06 2.28 1.5, 2.4, 4.4 2.37 1.68 2.40 2.14 5.53

c

Rs (Other study)

1.4, 3.1, 2.2 [34] 2.74 [34] 2.17 [8]

MeOH/ACN/TEA/HOAc (80:20:0.1:0.1, v/v/v/v), 15 kV.

d

about the application of non-aqueous mobile phase for chiral separations [8,34]. In this study, a number of racemic pharmaceutical molecules were screened in the polar organic mode on this new vancomycin monolithic column for chiral separations. The majority of these were basic molecules, but neutral and acidic molecules were also examined. The polar organic mobile phase consisted of a mixture of two organic solvents (MeOH and ACN) and an acid/base pair (HOAc and TEA). In enantioselective CEC, the mobile phase composition strongly affects the enantioselectivity due to the strong influence of the enantiorecognition mechanism (charge–charge interaction, hydrogen bonding, ␲–␲ interactions, etc.). Hence, the effect of the concentration ratio of MeOH/ACN on retention time, efficiency, and resolution of several basic compounds was investigated by keeping the content of TEA/HOAc constant. The mobile phase consisted of a mixture of two organic solvents, MeOH and ACN, and the ratio of these was examined. As shown in Fig. 5, alprenolol was selected as the model racemic analyte for evaluating the mobile phase composition since it was well separated (Rs = 2.06) in approximately 9 min with a high efficiency value (75 500 plates/m) under standard conditions. By changing the organic modifier concentration from 100% MeOH to 90%, the resolution for alprenolol enantiomers increases. On the other hand, by increasing ACN content, both the efficiency and resolution for alprenolol enantiomers decrease. These results are in agreement with previous reports on macrocyclic antibiotic-based CSPs, where better enantioselectivity generally is obtained using a high MeOH concentration in combination with small amounts of acid/base additives, probably due to less nonselective hydrogen interactions [34]. The observations described above for alprenolol enantiomers are applicable to other racemic analytes. The effect of concentration ratio of TEA/HOAc in a constant solvent mix on enantioselectivity was also examined. It can be observed that for most basic racemic analytes tested on this vancomycin monolithic column, the effect of concentration ratio of TEA/HOAc has a lower influence than the modifier ratio. The change of the MeOH/ACN concentration ratio affects the efficiency and resolution values more than the change of the TEA/HOAc concentration ratio. The resolution was not changed significantly over the range examined, but significantly decreased at higher concentrations of acid and base additives. From the above results, it can be remarked that the selection of an appropriate concentration ratio of the two organic solvents is of paramount importance in controlling retention time, efficiency, and resolution of a chiral separation. With optimum combination of the above factors, high efficiency and resolution values for chiral separation of acebutolol, alprenolol, pindolol, and propranolol enantiomers were achieved as illustrated in Fig. 6a–d, respectively, and in Table 2. Well-resolved separations with good efficiency and resolution values were also achieved for other racemic compounds using this vancomycin monolithic column, as shown in Table 2. Also shown in Table 2, the enantiomeric resolutions of this vancomycin monolithic column for

M.-L. Hsieh et al. / J. Chromatogr. A 1358 (2014) 208–216

chiral separation of labetalol, pindolol, and bupivacaine are comparable to those by known vancomycin bonded chiral phases when a mobile phase of very similar composition was used [8,34].

[14]

4. Concluding remarks

[15]

This study has demonstrated the feasibility of preparing a vancomycin-bonded silica monolithic column via a single-step in situ sol–gel approach. In contrast to the post-modification approach which is time-consuming and probably has limited CSP loading since the size of the mesopores does not favor a high molecular weight chiral selector to be functionalized inside the mesopores, our in situ approach to copolymerize the skeleton precursor and CSP precursor into a monolithic network has fewer steps and provides an opportunity to increase the CSP loading. In an aqueous mobile phase, the enantioselectivity is often lower than that in a polar organic mobile phase [21]. Nevertheless, our results show that a number of neutral, basic and acidic analytes were successfully separated with good enantioselectivity on this CSP in the aqueous mobile phase. It should be mentioned that while the chiral separation of ␤-blockers was not realized using the post-modified vancomycin-bonded silica-based monolithic column [21], enantiomers of six ␤-blockers were separated using our column in the reversed-phase mode. Moreover, all seven ␤-blockers in our list, including labetalol, were enantioseparated in the normal-phase mode. We believe this single-step approach to prepare the chiral silica-based monolithic capillary columns are extendable to other macrocylic antibiotics.

[16]

Acknowledgements This research was supported by National Science Council of ROC (NSC 96-2113-M-194-012-MY3) and National Chung Cheng University. The authors are grateful to Dr. Guor-Tzo Wei for his valuable suggestions in the chiral separation.

[17]

[18]

[19]

[20]

[21]

[22]

[23] [24]

[25]

[26] [27] [28]

References [29] [1] L.A. Colon, Y. Guo, A. Fermier, Capillary electrochromatography, Anal. Chem. 69 (1997) 461A. [2] N. Ishizuka, H. Minakuchi, K. Nakanishi, N. Soga, H. Nagayama, K. Hosya, N. Tanaka, Performance of a monolithic silica column in a capillary under pressure-driven and electrodriven conditions, Anal. Chem. 72 (2000) 1275. [3] M.M. Dittmann, G.P. Rozing, Capillary electrochromatography – a highefficiency micro-separation technique, J. Chromatogr. A 744 (1996) 63. [4] D. Wistuba, V. Schurig, Enantiomer separation by pressure-supported electrochromatography using capillaries packed with Chirasil-Dex polymer-coated silica, Electrophoresis 20 (1999) 2779. [5] Z. Liu, K. Otsuka, S. Terabe, Chiral separation by open tubular capillary electrochromatography with adsorbed avidin as a stationary phase, J. Sep. Sci. 24 (2001) 17. [6] S. Li, D.K. Lloyd, Packed-capillary electrochromatographic separation of the enantiomers of neutral and anionic compounds using ␤-cyclodextrin as a chiral selector: effect of operating parameters and comparison with free-solution capillary electrophoresis, J. Chromatogr. A 666 (1994) 321. [7] A.S. Carter-Finch, N.W. Smith, Enantiomeric separations by capillary electrochromatography using a macrocyclic antibiotic chiral stationary phase, J. Chromatogr. A 848 (1999) 375. [8] H. Wikström, L.A. Svensson, A. Torstensson, P.K. Owens, Immobilization and evaluation of a vancomycin chiral stationary phase for capillary electrochromatography, J. Chromatogr. A 869 (2000) 395. [9] O. Kornyˇsova, P.K. Owens, A. Maruska, Continuous beds with vancomycin as chiral stationary phase for capillary electrochromatography, Electrophoresis 22 (2001) 3335. [10] S. Fanali, P. Catarcini, G. Blaschke, B. Chankvetadze, Enantioseparations by capillary electrochromatography, Electrophoresis 22 (2001) 3131. [11] N. Tanaka, H. Kobayashi, K. Nakanishi, H. Minakuchi, N. Ishizuka, Monolithic LC columns: a new type of chromatographic support could lead to a higher separation efficiencies, Anal. Chem. 73 (2001) 420A. [12] H. Zou, X.D. Huang, M.L. Ye, Q. Luo, Monolithic stationary phases for liquid chromatography and capillary electrochromatography, J. Chromatogr. A 954 (2002) 5. [13] C. Legido-Quigley, N.D. Marlin, V. Melin, A. Manz, N.W. Smith, Advances in capillary electrochromatography and micro-high performance liquid

[30]

[31] [32] [33] [34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

215

chromatography monolithic columns for separation science, Electrophoresis 24 (2003) 917. S. Eeltink, W.M. Decrop, G.P. Rozing, P.J. Schoenmakers, W.T. Kok, Comparison of the efficiency of microparticulate and monolithic capillary columns, J. Sep. Sci. 27 (2004) 1431. I. Tanret, D. Mangelings, Y.V. Heyden, Monolithic stationary phases in classic and chiral pharmaceutical analysis with CEC and pCEC, J. Chromatogr. Sci. 47 (2009) 407. D. Allen, Z. El Rassi, Silica-based monoliths for capillary electrochromatography: methods of fabrication and their applications in analytical separations, Electrophoresis 24 (2003) 3962. Z. Liu, K. Otsuka, S. Terabe, M. Motokawa, N. Tanaka, Physically adsorbed chiral stationary phase of avidin on monolithic silica column for capillary electrochromatography and capillary liquid chromatography, Electrophoresis 23 (2002) 2973. Z.L. Chen, H. Ozawa, K. Uchiyama, T. Hobo, Cyclodextrin-modified monolithic columns for resolving dansyl amino acid enantiomers and positional isomers by capillary electrochromatography, Electrophoresis 24 (2003) 2550. Z.L. Chen, T. Hobo, Chemically l-phenylalaninamide-modified monolithic silica column prepared by a sol–gel process for enantioseparation of dansyl amino acids by ligand exchange-capillary electrochromatography, Anal. Chem. 73 (2001) 3348. B. Preinerstorfer, D. Lubda, W. Lindner, M. Lämmerhofer, Monolithic silicabased capillary column with strong chiral cation-exchange type surface modification for enantioselective non-aqueous capillary electrochromatography, J. Chromatogr. A 1106 (2006) 94. X. Dong, J. Dong, J. Ou, Y. Zhu, H. Zou, Preparation and evaluation of a vancomycin immobilized silica monolith as chiral stationary phase for CEC, Electrophoresis 28 (2007) 2606. M. Lammerhofer, A. Gargano, Monoliths with chiral surface functionalization for enantioselective capillary electrochromatography, J. Pharm. Biomed. Anal. 32 (2010) 1091. D. Wistuba, Chiral silica-based monoliths in chromatography and capillary electrochromatography, J. Chromatogr. A 1217 (2010) 941. J. Ou, X. Li, S. Feng, J. Dong, X. Dong, L. Kong, M. Ye, H. Zou, Preparation and evaluation of a molecularly imprinted polymer derivatized silica monolithic column for capillary electrochromatography and capillary liquid chromatography, Anal. Chem. 79 (2007) 639. X. Dong, R. Wu, J. Dong, M. Wu, Y. Zhu, H. Zou, The covalently bonded cellulose tris(3,5-dimethylphenylcarbamate) on a silica monolithic capillary column for enantioseparation in capillary electrochromatography, J. Chromatogr. B 875 (2008) 317. J.D. Hayes, A. Malik, Sol–gel monolithic columns with reversed electroosmotic flow for capillary electrochromatography, Anal. Chem. 72 (2000) 4090. T.-A. Lin, G.-Y. Li, L.-K. Chau, Sol–gel monolithic anion-exchange column for capillary electrochromatography, Anal. Chim. Acta 576 (2006) 117. M.-L. Hsieh, G.-Y. Li, L.-K. Chau, Y.-S. Hon, Single-step approach to ␤cyclodextrin-bonded silica as monolithic stationary phases for CEC, J. Sep. Sci. 31 (2008) 1819. D.W. Armstrong, Y. Tang, S. Chen, Y. Zhou, C. Bagwill, J.-R. Chen, Macrocyclic antibiotics as a new class of chiral selectors for liquid chromatography, Anal. Chem. 66 (1994) 1473. D.W. Armstrong, K.L. Rundlett, J.-R. Chem, Evaluation of the macrocyclic antibiotic vancomycin as a chiral selector for capillary electrophoresis, Chirality 6 (1994) 496. C. Desiderio, S. Fanali, Chiral analysis by capillary electrophoresis using antibiotics as chiral selector, J. Chromatogr. A 807 (1998) 37. D.W. Armstrong, U.B. Nair, Capillary electrophoretic enantioseparations using macrocyclic antibiotics as chiral selectors, Electrophoresis 18 (1997) 2331. A. Dermaux, F. Lynen, P. Sandra, Chiral capillary electrochromatography on a vancomycin stationary phase, J. High Resol. Chromatogr. 21 (1998) 575. C. Karlsson, L. Karlsson, D.W. Armstrong, P.K. Owens, Evaluation of a vancomycin chiral stationary phase in capillary electrochromatography using polar organic and reversed-phase modes, Anal. Chem. 72 (2000) 4394. C. Desiderio, Z. Aturki, S. Fanali, Use of vancomycin silica stationary phase in packed capillary electrochromatography: I. Enantiomer separation of basic compounds, Electrophoresis 22 (2001) 535. S. Fanali, P. Catarcini, M.G. Quaglia, Use of vancomycin silica stationary phase in packed capillary electrochromatography: III. Enantiomeric separation of basic compounds with the polar organic mobile phase, Electrophoresis 23 (2002) 477. C. Karlsson, H. Wikström, D.W. Armstrong, P.K. Owens, Enantioselective reversed-phase and non-aqueous capillary electrochromatography using a teicoplanin chiral stationary phase, J. Chromatogr. A 897 (2000) 349. O. Kornyˇsova, R. Jarmalaviciene, A. Maruska, A simplified synthesis of polymeric nanoparticulate stationary phases with macrocyclic antibiotic as chiral selector for capillary electrochromatography, Electrophoresis 25 (2004) 2825. G.-S. Ding, A.-N. Tang, Capillary electrochromatographic separation of racemates on a norvancomycin-bonded monolithic column under reversed-phase conditions, J. Chromatogr. A 1208 (2008) 232. A.-N. Tang, X.-N. Wang, G.-S. Ding, X.-P. Yan, On-line preconcentration and enantioseparation of thalidomide racemates by CEC with the hyphenation of octyl and norvancomycin monoliths, Electrophoresis 30 (2009) 682. E. Pittler, M.G. Schmid, Enantioseparation of dansyl amino acids by HPLC on a monolithic column dynamically coated with a vancomycin derivative, Biomed. Chromatogr. 24 (2010) 1213.

216

M.-L. Hsieh et al. / J. Chromatogr. A 1358 (2014) 208–216

[42] H. Minakuchi, K. Nakanish, N. Soga, N. Ishizuka, N. Tanaka, Octadecylsilylated porous silica rods as separation media for reversed-phase liquid chromatography, Anal. Chem. 68 (1996) 3498. [43] N. Ishizuka, H. Kobayashi, H. Minakuchi, K. Nakanishi, K. Hirao, K. Hosoya, T. Ikegami, N. Tanaka, Monolithic silica columns for high-efficiency separations by high-performance liquid chromatography, J. Chromatogr. A 960 (2002) 85. [44] M.C. Breadmore, S. Shrinivasan, K.A. Wolfe, M.E. Power, J.P. Ferrance, B. Hosticka, P.M. Norris, J.P. Landers, Towards a microchip-based chromatographic platform: Part 1. Evaluation of sol–gel phases for capillary electrochromatography, Electrophoresis 23 (2002) 3487. [45] C.H. Xie, J. Hu, H. Xiao, X.Y. Su, J. Dong, R.J. Tian, Z.K. He, H.F. Zou, Electrochromatographic evaluation of a silica monolith capillary column for separation of basic pharmaceuticals, Electrophoresis 26 (2005) 790. [46] I. Gusev, X. Huang, C. Horvath, Capillary columns with in situ formed porous monolithic packing for micro high-performance liquid chromatography and capillary electrochromatography, J. Chromatogr. A 855 (1999) 273. [47] A. Berthod, U.B. Nair, C. Bagwill, D.W. Armstrong, Derivatized vancomycin stationary phases for LC chiral separations, Talanta 43 (1996) 1767.

[48] M. Motokawa, H. Kobayashi, N. Ishizuka, H. Minakuchi, K. Nakanishi, H. Jinnai, K. Hosoya, T. Ikegma, N. Tanaka, Monolithic silica columns with various skeleton sizes and through-pore sizes for capillary liquid chromatography, J. Chromatogr. A 961 (2002) 53. [49] H. Minakuchi, K. Nakanishi, N. Soga, N. Ishizuka, N. Tanaka, Effect of skeleton size on the performance of octadecylsilylated continuous porous silica columns in reversed-phase liquid chromatography, J. Chromatogr. A 762 (1997) 135. [50] T. Hara, H. Kobayashi, T. Ikegami, K. Nakanishi, N. Tanaka, Performance of monolithic silica capillary columns with increased phase rations and small-sized domains, Anal. Chem. 78 (2006) 7632. [51] K. Nakanishi, N. Soga, Phase separation in gelling silica-organic polymer solution: systems containing poly(sodium styrenesulfonate), J. Am. Ceram. Soc. 74 (1991) 2518. [52] K. Nakanishi, H. Shikata, N. Ishizuka, N. Koheiya, N. Soga, Tailoring mesopores in monolithic macroporous silica for HPLC, J. High Resol. Chromatogr. 23 (2000) 106. [53] Q.-H. Wan, Capillary electrochromatography: effect of electrolyte concentration on electroosmotic flow and column efficiency, J. Chromatogr. A 782 (1997) 181.

Single-step approach for fabrication of vancomycin-bonded silica monolith as chiral stationary phase.

A vancomycin-bonded silica monolithic column for capillary electrochromatography (CEC) was prepared by a single-step in situ sol-gel approach. This so...
1MB Sizes 1 Downloads 3 Views