3426 Faiz Ali Won Jo Cheong Department of Chemistry, Inha University, Namku, Incheon, South Korea Received July 30, 2014 Revised September 9, 2014 Accepted September 10, 2014
J. Sep. Sci. 2014, 37, 3426–3434
Research Article
C18 -bound porous silica monolith particles as a low-cost high-performance liquid chromatography stationary phase with an excellent chromatographic performance Ground porous silica monolith particles with an average particle size of 2.34 m and large pores (363 Å) exhibiting excellent chromatographic performance have been synthesized on a relatively large scale by a sophisticated sol–gel procedure. The particle size distribution was rather broad, and the d(0.1)/d(0.9) ratio was 0.14. The resultant silica monolith particles were chemically modified with chlorodimethyloctadecylsilane and end-capped with a mixture of hexamethyldisilazane and chlorotrimethylsilane. Very good separation efficiency (185 000/m) and chromatographic resolution were achieved when the C18 -bound phase was evaluated for a test mixture of five benzene derivatives after packing in a stainless-steel column (1.0 mm × 150 mm). The optimized elution conditions were found to be 70:30 v/v acetonitrile/water with 0.1% trifluoroacetic acid at a flow rate of 25 L/min. The column was also evaluated for fast analysis at a flow rate of 100 L/min, and all the five analytes were eluted within 3.5 min with reasonable efficiency (ca. 60 000/m) and resolution. The strategy of using particles with reduced particle size and large pores (363 Å) combined with C18 modification in addition to partial-monolithic architecture has resulted in a useful stationary phase (C18 -bound silica monolith particles) of low production cost showing excellent chromatographic performance. Keywords: C18 modification / Fast analysis / Monolith particles / Stationary phases DOI 10.1002/jssc.201400811
1 Introduction Surface modification on metal oxide surfaces (either particle surface or flat surface) is a useful strategy for expanding the application of such materials in a variety of fields. Development of separation materials via controlled living radical polymerization was reviewed by Wang et al. [1]. Materials of various functional coatings were introduced [2]. A review on optoelectronic applications such as photovoltaics, light emitting iodides, thin-film transistors, sensors, etc., was presented [3]. In another review, breakthrough applications were proposed in the fields of sustainable energy, environmental remediation, biomaterials, pharmaceutical industry, and catalysis [4]. Materials with great versatility and control in drug delivery could be prepared by the functionalCorrespondence: Professor W. J. Cheong, Department of Chemistry, Inha University, 253 Yonghyundong, Namku, Incheon 402751, South Korea E-mail: [email protected] Fax: +82-328675604
Abbreviations: BET, Brunauer–Emmett–Teller; BJH, Barrett– Joyner–Halenda; FE-SEM, field-emission scanning electron microscopy; FE-TEM, field-emission transmission electron microscopy; HETP, height equivalent to a theoretical plate; RAFT, reversible addition-fragmentation chain transfer; TMOS, trimethylorthosilicate C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ization of mesoporous silica nanoparticles [5]. A wide range of hydrophilic polymer brushes were grown for potential applications in surface lubricity and tribology, tunable surface wetting, and cell detachment [6]. Especially, porous/nonporous silica materials are subject to various surface modifications that result in diverse chromatographic and SPE media [7–11]. Reduction of the average size of silica particles has been the goal in developing chromatographic separation media throughout chromatographic history because of the theoretically expected improvements in chromatographic efficiency. The general trend of reducing the particle size from 100 m to sub-2 m with enhancement of chromatographic performance has been reported during the continuous developing stages [7,12–19]. Stationary phases comprising ultimately small particles with high separation efficiency have been compiled in some review articles [20–22] where core–shell particles and very fine porous particles are the typical phases of interest. Studies of monolith particles prepared by crushing bulk monoliths have been rare although molecularly imprinted polymer particles were prepared in such a way in the early years [23–25]. Ground organic monolith particles were employed as new chromatographic separation media very recently [26]. In our laboratory, silica monolith particles prepared from bulk monoliths have been studied as chromatographic www.jss-journal.com
J. Sep. Sci. 2014, 37, 3426–3434
separation media. We also have reduced the size of silica monolith particles from 10 m to sub-2 m in a series of studies [12,13,27,28] followed by their modification to various stationary phases where polystyrene-bound phases [12, 13] have been reported to be very efficient stationary phases. Somewhat lower separation resolution and lower yield of the final stationary phase due to the many preparation steps are the disadvantages associated with the polystyrene bound phases. In this study, 2.34 m (average) silica particles with large pore size (363 Å) have been prepared from bulk monoliths. The sol–gel process of silica monolith formation has been further elaborated to achieve the goal. The optimized formulation was found after preliminary preparation and examination of a series of batches of silica monolith particles. The scale of silica monolith production of this study was increased by three times in comparison with that of Ref. [12]. The resultant silica monolith particles were then chemically modified by reaction with a C18 reagent followed by an end-capping reagent, and packed in a column (1.0 mm × 150 mm) for examination of chromatographic performance.
Liquid Chromatography
3427
2.3 C18 modification and end capping of silica monolith particles Ground silica monolith particles 600 mg (calcined at 550⬚C for 48 h) were dried at 120⬚C overnight in a vacuum oven and suspended in 17 mL anhydrous toluene followed by stirring for 30 min in a 50 mL round bottom flask. The reaction mixture was subjected to reflux at 115⬚C. Then 350 mg chlorodimethyloctadecylsilane (dissolved in 5 mL anhydrous toluene) was added dropwise with a dropping funnel. The reflux was maintained for 40 h. The temperature of the reaction mixture was adjusted to 60⬚C. Hexamethyldisilazane 120 L and chlorotrimethylsilane 120 L dissolved in 3 mL anhydrous toluene were added to the reaction mixture dropwise. The reaction at 60⬚C was continued overnight. The temperature of the flask was increased to 110⬚C and refluxed for 5–6 h. The product was washed with toluene at room temperature (30 min) and reflux temperature (overnight), and rinsed with acetone (HPLC grade). The solution was filtered and C18 -bound silica monolith particles were dried in a vacuum oven at 70⬚C overnight.
2 Materials and methods 2.1 Reagents and materials All chemical reagents and solvents were of analytical or HPLC grade. Acetic acid, urea, PEG 10 000, trimethylorthosilicate (TMOS), chlorodimethyloctadecylsilane, hexamethyldisilazane, chlorotrimethylsilane, anhydrous toluene, and styrene were purchased from Sigma–Aldrich (St. Louis, MO, USA). HPLC-grade methanol, acetonitrile, acetone, and water were obtained from Mallinckrodt Baker (Phillipsburg, NJ, USA). All the reagents were used as received. Screen frits (3.2 mm diameter, 0.08 mm thickness, and 0.5 m pore size) were obtained from Valco (Houston, TX, USA). Glass-lined stainless-steel tubing (15 cm, 1.0 mm id, 1/8 inch od) and silica capillaries (50 m id, 365 m od) were purchased from Grace (Deerfield, IL, USA).
2.2 Synthesis of ground silica monolith particles The synthesis procedure for soft silica monoliths was further modified to get smaller (2.34 m) silica monolith particles with larger pores in comparison to that of Ref. [12]. The PEG and urea contents were further increased with respect to TMOS, and the relative amount of catalytic solvent (dilute acetic acid) was also increased. Thus, 4000 mg PEG and 4125 mg urea were dissolved in 36 mL 0.01 M acetic acid. A 10 mL aliquot of TMOS was added to the solution in an ice/water condition and stirred vigorously for 60 min. The solution was gelled at 40⬚C in an oven for 48 h, and aged at 120⬚C in an autoclave for 48 h. The residual water was decanted off and the silica monolith was dried at 70⬚C for 20 h, ground with a mortar and pestle for 10 min, and calcined at 550⬚C for 50 h. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2.4 Characterization of the bare silica and C18 modified silica monolith particles A BEL-Japan (Osaka, Japan) BELSORP-Max was used to measure the Brunauer–Emmett–Teller (BET)/Barrett–Joyner– Halenda (BJH) N2 adsorption–desorption isotherms at 77 K, BJH pore sizes, and BET-specific surface areas for the bare and C18 -bound silica monolith particles. The samples were out-gassed at 373 K for 10 h to obtain a residual pressure of less than 10–3 Torr. The amount of N2 adsorbed at a relative pressure of P/P0 = 0.99 was used to determine the total pore volume. The SEM images of C18 -bound silica monolith particles were taken by Hitachi (Tokyo, Japan) S-4200 field-emission scanning electron microscopy (FE-SEM) while the TEM images were captured by JEOL (Tokyo, Japan) JEM2100F field-emission transmission electron microscopy (FE-TEM). A Thermo Electron (Waltham, MA, USA) Flash EA1112 elemental analyzer was used to obtain the carbon load data. A Malvern (Worcestershire, UK) Mastersizer 2000 particle size analyzer was used to get particle size distribution.
2.5 Column packing and HPLC An HPLC system similar to that of Ref. [13] was used where a homemade 150 mm glass-lined micro-column, a 10AD pump (Shimadzu, Tokyo, Japan), a Valco (Houston, TX, USA) C14W.05 injector with 50 nL injection loop, a membrane degasser (Shimadzu DGU-14A), and a UV-VIS capillary window detector (Jasco UV-2075) were assembled to construct the LC system. The software Multichro 2000 from Youlingisul (Seoul, Korea) was used for acquisition and processing chromatographic data. www.jss-journal.com
3428
F. Ali and W. J. Cheong
In our previous work, we prepared the polystyrenebound partially sub-2 m particles by reversible additionfragmentation chain transfer (RAFT) polymerization and packed the phase in columns of 1 mm id and 300 mm length, and the columns showed excellent separation efficiency partially owing to the pseudo-core–shell character of the final phase [12, 13]. In current study, we prepared C18 -modified partially sub-1 m particles having no core–shell character at all. Some difficulties were encountered in packing this phase, thus the column dimension was adjusted to 150 mm length. The presence of high population of fine particles below sub-1 m imposes a serious negative effect upon achieving a good packing quality. Thus a major portion of fine particles below sub-1 m were removed by repeated sedimentation before packing. Removal of fine particles was carried out by suspending the final phase in methanol by stirring for 5 min, sonicating for 10 min and keeping the vial for 30 min undisturbed. The C18 -bound particles settled down to the bottom while the fine particles were still suspended in the supernatant. The supernatant was carefully removed and the precipitate was again suspended in methanol and the whole process of sedimentation was repeated three times. Both phases before and after sedimentation treatment were subjected to measurements of particle size distribution. A loss of 22.4% was observed as the fine particles by the sedimentation treatments (three times). These recovered fine particles may be used for other applications. The term “partially sub1 m” was adopted to address the product of reduced particle sizes having some sub-1 m particles although the final product was obtained after removal of a major part of sub1 m particles (the smallest particles). The column was packed as follows. A 0.5 m commercial screen frit was placed in the 1/8 inch port of a reducing union (1/8–1/16), and a glass-lined stainless-steel column (1.0 mm id, 15 cm length) was fitted to the port, and the tubing was connected to the column packer. The slurry was prepared by suspending 170 mg stationary phase in 4.5 mL methanol, and was fed into the reservoir. A pressure sequence was applied. After packing, the inlet union was installed in the same way as the outlet union (with a frit). The column inlet was connected to the injector, and the column outlet, to the capillary window detector. The stock sample solution was prepared by dissolving phenol (0.88 L), acetophenone (0.14 L), 4-methyl-2-nitroaniline (0.32 mg), benzene (2.93 L), and toluene (1.46 L) in 1 mL mobile phase and was stored at 4⬚C. The sample was further diluted for injection.
3 Results and discussion 3.1 Stationary phase characterization 3.1.1 Architecture and morphology The FE-SEM images (A, B, C), FE-TEM images (X, Y), and a microscopic view (Z) of the C18 -bound silica monolith C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Sep. Sci. 2014, 37, 3426–3434
particles are shown in Fig. 1. The microscopic view (Z) and the wide SEM view (A) show that the average size of C18 -modified silica monolith particles is clearly less than 3 m. The close SEM views (B, C) indicate that the surface roughness of silica monolith particles was not altered by C18 modification. On the other hand, surface modification of silica monolith particles with thin polystyrene layers via RAFT polymerization resulted in the particles of much smoother surface [12, 13]. FE-TEM images of C18 -modified silica monolith particles (X, Y) show dim but visible pore structures as projected in dark irregular spots. The average pore size seems to be quite large (Fig. 1X), and this was confirmed by the N2 adsorption–desorption analysis below.
3.1.2 Particle size distribution The particle size distribution data (both number and volume based) of the bare and C18 -bound silica monolith particles are assembled in Table 1 while the volume-based plot is given in Fig. 2A. The effect of sedimentation treatment for the C18 bound particles was also included in Table 1. Visual comparison of volume-based particle size distribution between the bare and C18 -bound silica monolith particles and that between the untreated stationary phase and the sedimentation-treated stationary phase are given in Fig. 2A and 2B, respectively. It is interesting to note that the particle size distribution of the sedimentation-treated C18 -bound phase was narrowed down both in the smaller and larger particle size areas (Fig. 2B and Table 1). The disappearance of the region of smaller size is natural (due to removal of fine particles below sub1 m), but the decay of the region of larger size is against the common expectation. As shown in Table 1, d(0.9) before sedimentation treatment was 7.12 m while d(0.9) after treatment was 5.15 m. We assume that some of the C18 -bound silica monolith particles were physically aggregated during the prior drying stage, and were separated into smaller particles by vigorous stirring and sonication during the sedimentation process. A loss of 22.4% was observed as the fine particles by the sedimentation treatments (three times). Special care was taken to ensure a high quality packing in this study. It is known to be difficult to get the expected separation efficiency when columns of diameter below the standard (4.6 mm) are packed or the particle size is reduced [29–31]. Column dimensions critically affect separation efficiency as reported by Mazzeo et al. [30]. Packing procedures regarding size of packing materials and diverse column dimensions have been reviewed by Unger et al. [32]. High pressure up to 2000 bar can be applied for packing silica particles [33]. We have been making improvements in column packing such as modifying the slurry packer for use under higher pressure up to 1250 bar, adopting home-made tapered connection tubing from the slurry reservoir to the column for smooth flow of the suspended particles, incorporation of vigorous mechanical vibration during the packing procedure, and sufficient conditioning under a lower pressure at the end of packing, etc. Thus column packing www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2014, 37, 3426–3434
3429
Figure 1. FE-SEM images (A, B, C), FE-TEM images (X, Y), and microscopic view (Z) of C18 -bound silica monolith particles. The scale bars of A, B, C, X, Y, and Z are 10 m, 1 m, 1 m, 50 nm, 10 nm, and 2 m, respectively. Table 1. Comparative analysis of particle size distribution (number-based and volume-based) between the bare and C18 -bound silica monolith particles and between the C18 -bound particles before and after removal of fine particles. The length unit is m.
Silica monolith particles (m)
C18 -bound silica monolith particles Including fine particles
Number based Volume based
Without fine particles
d(0.1)a)
d(0.5)b)
d(0.9)c)
Rd d)
d(0.1)a)
d(0.5)b)
d(0.9)c)
Rd d)
d(0.1)a)
d(0.5)b)
d(0.9)c)
Rd d)
0.59 0.82
0.85 2.34
1.86 5.76
0.32 0.14
0.68 1.01
1.03 3.23
2.15 7.12
0.32 0.14
0.73 1.22
1.11 2.63
2.05 5.15
0.36 0.24
a) Particle diameter corresponding to the integrated area ratio of 0.1 when integrated in the range of 0–d diameter. b) Particle diameter corresponding to the integrated area ratio of 0.5 when integrated in the range of 0–d diameter. c) Particle diameter corresponding to the integrated area ratio of 0.9 when integrated in the range of 0–d diameter. d) d(0.1)/d(0.9)
was carried out according to the procedure published in previous studies [12, 13], [34–36] with a little modification as described in Section 2.5.
3.1.3 BET/BJH analysis (pore size distribution) The BET/BJH adsorption–desorption analysis results of this study are summarized in Table 2 and compared with those of C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
previous studies. The BJH adsorption pore size distribution of this study is plotted in Fig. 3 for the bare and C18 -bound silica monolith particles. The data in Table 2 shows that a sequential increase of pore size (and the corresponding pore volume) of silica monolith as 133 [36] < 212 [35] < 343 [13] < 363 Å (current study) results in a sequential decrease of surface area as 343 [36] > 283 [35] > 136 [13] >110 m2 /g (current study). As shown in Fig. 3, the pore size distribution was slightly shifted to the left by C18 modification. This was
www.jss-journal.com
3430
J. Sep. Sci. 2014, 37, 3426–3434
F. Ali and W. J. Cheong
Figure 2. Comparison of volume-based particle size distribution between bare (䊊) and C18 -bound (䊉) silica monolith particles (A) and comparison of volume-based particle size distribution of C18 -bound silica monolith particles before (䊉) and after () removal of fine particles (B). Table 2. Comparison of BET/BJH analysis results of the bare and ligand modified silica monolith particles for the current and previous studies
Silica particles
Ligand bound silica particles Polystyrene bound phases previous studies
˚ a) Pore size (A) Pore volume (cm3 /g)b) Surface area (m2 /g)c)
Current study C18 -bound phase
[36]
[35]
[13]
Current study
[36]
[35]
[13]
133 0.64 343
212 0.83 283
343 1.06 136
363 1.09 110
131 0.32 168
146 0.53 161
252 0.84 131
303 0.95 83
a) BJH adsorption average pore diameter. b) Total pore volume at P/P0 = 0.99. c) BET-specific surface area. Table 3. Elemental analysis results of the phase of this study in comparison with those of previous studies
Stationary phase
Carbon % Hydrogen %
˚ This study 363 A, partially sub-1 m monolith, C18
6.92 0.80
Previous studies
˚ 5 m [36] 133 A, Lichrosorb, polystyrene
˚ 3–5 m [35] 212 A, monolith, C18
[13] 343 A˚ partially sub 2-m, polystyrene
19.96 1.98
7.12 1.61
10.0 1.0
also confirmed in Table 3 showing a minimal decrease in pore volume from 1.09 to 0.95 cm3 /g for C18 binding of this study in contrast to a much greater decrease observed for polystyrene binding in previous studies. In this study, two basic strategies were adopted in preparation of silica monolith particles, that is, decreasing particle size and increasing pore size, which have been successfully achieved as demonstrated above. The main purpose of such strategies is to increase the separation efficiency of the resultant stationary phase. Reduction of particle size surely contributes to improved mass transfer kinetics, but it makes column packing more difficult, thus we had to adjust the column dimension to a shorter length (15 cm) instead of
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
30 cm of the previous study [13]. Increasing the pore size is certainly helpful for better mass transfer kinetics, but the consequently reduced surface area (and number of silanol groups) resulted in low carbon load after C18 modification. We should also note that the pseudo-core–shell character of polystyrene-bound monolith particles [12, 13] is not realized with the C18 -modified monolith particles of this study. We concluded that such conflicting effects were roughly selfcompensating based on the performances of the columns packed with the phase of this study. The common advantage of stationary phases based on silica monolith particles is the partial monolithic structure to enable high flux solvent flow during packing [12, 13].
www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2014, 37, 3426–3434
Figure 3. BJH-adsorption pore size distribution of the bare (o) and C18 -bound (䊉) silica monolith particles.
3.1.4 Elemental analysis The carbon load of the C18 -modified monolith particles of current study (6.92%) seems better than expected if it is compared to that (7.12%) of the previous study [35] (Table 4). The pore size of current silica monolith (363 Å) is much greater than that (212 Å) of Ref. [35]. Therefore, the surface area for current silica monolith (110 m2 /g) is smaller than that (283 m2 /g) of Ref. [35]. The lower surface area should cause a lower number of surface silanol groups and hence a lower carbon load. It seems that the silanol groups in the larger pores were subject to higher chance of silanization reaction with the C18 reagent resulting in a better reaction yield. 3.2 Chromatographic performance of the phase of current study in comparison with those of a commercial phase and the phases of previous studies The chromatographic performance of the phase of this study is demonstrated in Fig. 4 in comparison with a commercial
3431
C18 phase (Alltima C18 , 3 m, 80 Å) and the C18 -bound phase based on silica monolith particles reported in the previous study [35]. The separation efficiency and plate height data of Fig. 4 are assembled in Table 4. The column-to-column and day-to-day reproducibility data in retention times and separation efficiency for the phase of this study (Fig. 4A) are given in Table 5. The separation efficiency data of fast analysis are given in Table 6. The chromatographic resolution data of the phase of this study are given in Table 7 in comparison with those of other phases. The van Deemter plot based on the chromatographic data obtained with the C18 -modified silica monolith particles is given in Fig. 5. The mobile phase was 70:30 v/v acetonitrile/water with 0.1% TFA. The optimized flow rate was 25 L/min. The column backpressure at the optimum flow rate was 98 bar. The average number of theoretical plates (N, Table 4) of the column packed with the C18 -bound silica monolith particles of this study was 27 600. This value corresponds to 184 000/m, which is close to or even better than those of the previous studies (165 000/m [13] or 188 000/m [12]) obtained with the polystyrene bound silica monolith particles bearing some core-shell character. In this study, the monolith formation procedure was modified to result in reduction of particle size, and C18 modification was adopted instead of polystyrene modification of the previous studies [12, 13] to produce more feasible stationary phase for realization of disposable microcolumns in the future. Thus the number-based average particle size of the phase of this study was decreased to 1.11 m while that of Ref. [12] was 1.61 m. On the other hand, the average pore size (363 Å) was increased in comparison to that (252 Å) of Ref. [12]. The scale of silica monolith preparation has also been increased by three times compared to that of Ref. [12,13]. The above-mentioned changes induced some effects. Particle size reduction was beneficial in mass transfer kinetics although its packing was challenging. Increase of pore size contributed somewhat to improvement in mass transfer kinetics. Adoption of C18 modification instead of polystyrene
Table 4. Comparison of number of theoretical plates and height equivalent to a theoretical plate (HETP) (m) measured at the optimum flow rate of the van Deemter curve among the columns packed with various phasesa)
Analytes
Phenol Acetophenone 4-Methyl-2-nitro-aniline Benzene Toluene Average
Column type C18 -bound silica monolith particles of this studyb)
Commercial Alltima C18 (3 m)b)
C18 -bound silica monolith particles (3–5 m) [35]c)
N-value
HETP
N-value
HETP
N-value
HETP
22 000 ± 400 29 000 ± 940 29 200 ± 1200 28 900 ± 1260 28 800 ± 1020 27 600
6.8 5.2 5.1 5.2 5.2 5.4
11 700 ± 250 16 000 ± 220 16 000 ± 200 13 700 ± 380 13 200 ± 410 14 100
12.7 9.38 9.4 10.9 11.3 10.6
27 000 26 900 29 700 29 700 24 600 26 400
11.1 11.2 10.1 10.1 12.2 11.4
a) The mobile phase was 70/30 v/v acetonitrile/water with 0.1% TFA and the detection wavelength was 214 nm. b) The column dimension was 1.0 mm × 150 mm. The mobile phase flow rate was 25 L/min. c) The column dimension was 0.5 mm × 300 mm. The mobile phase flow rate was 7L/min.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
3432
J. Sep. Sci. 2014, 37, 3426–3434
F. Ali and W. J. Cheong
Figure 4. Comparison of chromatograms obtained at 214 nm with different stationary phases under optimized conditions. Chromatograms (A), (B), and (D) were from current study, and (C), from ref. [31]. (A) Our phase of this study; (B) commercial Alltima C18 (3 m); (C) C18 -modified silica monolith particles (3–5 m) of previous study [31]; (D) our phase of current study showing fast analysis. The column dimension of (A), (B), and (D) was 1.0 mm × 150 mm, and that of (C) was 0.5 mm × 300 mm. The mobile phase was 70:30 v/v acetonitrile/water with 0.1% TFA. The flow rate of (A) and (B) was 25 L/min, while those of (C) and (D) were 7 and 100 L/min, respectively. The elution order was phenol, acetophenone, 4-methyl2-nitroanailine, benzene, and toluene. Table 5. Column-to-column and day-to-day reproducibility in N and retention time for the columns (1.0 mm × 150 mm) packed with the phase of this studya)
Analytes
Phenol Acetophenone 4-Methyl- 2-nitroaniline Benzene Toluene
Column to column reproducibility
Day to day reproducibility
N-value
Retention times (min)
N-value
Retention times (min)
Average
% RSD
Average
% RSD
Average
% RSD
Average
% RSD
22 400 28 100 28 100 27 600 27 800
1.8% 3.2% 4.1% 4.4% 3.5%
6.94 8.31 8.85 10.50 12.51
0.6% 1.6% 2.0% 2.7% 3.6%
22 200 28 900 29 200 28 600 28 600
0.9% 0.2% 0.13% 0.74% 0.8%
6.92 8.24 8.74 10.29 12.21
0.3% 0.7% 0.7% 0.8% 1.1%
a) For column-to-column reproducibility data three columns were packed and the averages for both N-values and retention times were calculated while for day-to-day reproducibility only one column was evaluated on three consecutive days. Table 6. Numbers of theoretical plates obtained with our column of current study using a high flow rate (fast analysis)a)
Analytes
Phenol
Acetophenone
4-Methyl-2-nitroaniline
Benzene
Toluene
Average
N-value
10 100 ±150
9150 ± 120
9800 ± 40
8100 ± 80
7800 ± 110
9000
a) Flow rate was 100 L/min with the column dimension of 1.0 mm id × 150 mm length, and the mobile phase was 70/30 v/v acetonitrile/water with 0.1% TFA and the detection wavelength was 214 nm.
modification was beneficial in view of cost reduction, simplicity, and yield of the final phase at the price of loss of the core-shell character [12] of the polystyrene-bound phase. Such conflicting effects were found roughly self-compensating, resulting in similar separation efficiencies per unit length for the columns packed with the phase of this study and that of Ref. [12]. It should be noted that the column length of this study was reduced to 150 mm to ensure good packing instead C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
of 300 mm of the previous study [12]. Thus the separation efficiency per column was only one half, but separation efficiency per meter was similar to that of the previous study [12]. Nevertheless, the chromatographic resolution (average: 6.05) of this study was even better than that of Ref. [12] (average: 5.88) as shown in Table 7. The chromatographic resolution of the C18 -modified monolith particles of this study (Fig. 4A) was close to that (average: 6.02, Fig. 4B) of the Alltima C18 www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2014, 37, 3426–3434
3433
Table 7. Comparison of resolution (R) between current and previous studies
Stationary phase
C18 -bound silica monolith particles of this study b) Fast analysis with the phase of this study c) Polystyrene-bound silica monolith particles [12] d) Polystyrene-bound silica monolith particles [13] d) C18 -bound silica monolith particles (3–5m) [35] e) Alltima commercial C18 (3m) b)
Solute pair a) 1–2
2–3
3–4
4–5
Average
7.15
2.70
7.05
7.30
6.05
3.93
1.52
3.98
3.71
3.28
8.15
3.82
5.25
6.31
5.88
5.15
1.43
3.76
3.35
3.42
5.21
1.29
2.63
2.60
2.91
7.37
2.73
6.93
7.08
6.02
a) Solutes: 1, phenol; 2, acetophenone; 3, 4-methyl-2-nitronaniline; 4, benzene; 5, toluene. b) The column dimension was 1.0 mm × 150 mm. The mobile phase was 70:30 v/v acetonitrile/water 25 L/min. c) The column dimension was 1.0 mm × 150 mm. The mobile phase was 70:30 v/v acetonitrile/water 100 L/min. d) The column dimension was 1.0 mm × 300 mm. The mobile phase was 60:40 v/v acetonitrile/water 15 L/min. e) The column dimension was 0.5 mm × 300 mm. The mobile phase was 70:30 v/v acetonitrile/water 7 L/min.
Figure 5. The van Deemter plot based on chromatographic data obtained with the C18 -modified silica monolithic particles of this study using a mobile phase of 70:30 v/v acetonitrile/water with 0.1% TFA.
(spherical, 3 m, 80 Å) phase (Table 7). Note that the retention times for our phase (Fig. 4A) were much shorter than those of the Alltima C18 phase (Fig. 4B) owing to the lower carbon load (6.9% against 15.4%) induced by the larger pore size. Reduction of carbon load is usually accompanied by reduction in resolution if the other conditions are the same. The reason for the comparison of our phase with Alltima C18 , is that the volume-based average particle size distribution (2.63 m) of our phase is comparable to that of Alltima C18 (spherical, C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
with 0.1% TFA at a flow rate of with 0.1% TFA at a flow rate of with 0.1% TFA at a flow rate of with 0.1% TFA at a flow rate of
3 m, 80 Å). The advantageous character of our phase over spherical Alltima C18 particles is that pacing our phase gives rise to partial monolithic character in the packed structure enabling good packing quality and enhanced mass transfer kinetics. The chromatogram shown in Fig. 4C has been taken from Ref. [35] for comparison between the previous C18 -modified silica monolith particles (3–5 m) [35] and the phase of this study. We can see that the chromatographic performance of the material made in this study is far better than that of Ref. [35] as demonstrated in Fig. 4, Tables 4 and 7. For example, compare the separation efficiency: 184 000/m versus 88 000/m. We also found the possibility of our phase for fast analysis owing to the reduced retention based on the low carbon load. When a flow rate of 100 L/min was used, all the analytes were eluted within 3.5 min with good resolution (3.28) and reasonable separation efficiency (60 000/m) as demonstrated in Fig. 4D, Tables 6 and 7. The column backpressure at this flow rate was 371 bar. The column packed with our phase is not a typical monolithic column but it has a partial monolithic character in the packed structure. The plot of height equivalent to theoretical plate (H) as a function of mobile phase velocity (van Deemter plot) is given in Fig. 5. As shown in Fig. 5, the lowest H was ca. 5 m for the test solutes except for phenol. The optimized flow rate was found 25 L/min. The reason why the separation efficiency of phenol is poorer and apart from other solutes is not clear at present. We guess it may be related to the shortest retention time of phenol. Our phase is somewhat hybrid of spherical and typical monolithic phase
www.jss-journal.com
3434
F. Ali and W. J. Cheong
in view of permeability and trend of van Deemter plot [13]. The strategy of packing ground silica monolith particles seems to be more powerful when incorporated with C18 modification in view of production cost of the stationary phase. Modification of silica monolith particles with polystyrene via RAFT polymerization is a lengthy process including attachment of a halogen-terminated spacer, coupling of a RAFT initiator moiety, RAFT polymerization, and various washing steps leading to a rather low overall yield of the final phase [12, 13]. The ground silica monolith particles were claimed to play an important role in future realization of disposable microcolumns owing to the inexpensive production procedure of silica monolith particles [12, 13]. Now it seems to be more feasible to realize disposable microcolumns on account of successful synthesis of C18 -modified silica monolith particles of excellent chromatographic performance. A relevant study is underway.
4 Conclusion Porous silica monolith particles of reduced particle size (average 2.34 m) with an increased pore size (363 Å) have been synthesized, C18 -modified and end-capped. The C18 -modified silica monolith particles were packed in a glass-lined column (1.0 mm × 150 mm) to give the similar separation efficiency per unit length to that of polystyrene bound silica monolith particles of the previous study. Nevertheless, its chromatographic resolution was better, which was comparable to that of a column of the same length packed with a commercial 3 m spherical C18 phase. Moreover, the retention times were much smaller than those of the 3 m spherical C18 phase. Our phase also showed some potential for fast analysis. Thus a new stationary phase of good chromatographic performance and low cost has been developed, and this phase may be useful for realization of disposable microcolumns. This work was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and Future Planning (2012-R1A1A2006066). The authors have declared no conflict of interest.
J. Sep. Sci. 2014, 37, 3426–3434
[6] Edmondson, S., Armes, S. P., Polym. Int. 2009, 58, 307– 316. [7] Qiu, H., Liang, X., Sun, M., Jiang, S., Anal. Bioanal. Chem. 2011, 399, 3307–3322. [8] Namera, A., Miyazaki, S., Saito, T., Nakamoto, A., Anal. Methods. 2011, 3, 2189–2200. [9] Ali, I., Aboul-Enein, H. Y., J. Sep. Sci. 2006, 29, 762–769. [10] Haginaka, J., J. Chromatogr. B 2008, 875, 12–19. [11] Kara, D., Fisher, A., Sep. Purif. Rev. 2012, 41, 267–317. [12] Ali, F., Kim, Y. S., Lee, J. W., Cheong, W. J., J. Chromatogr. A 2014, 1324, 115–120. [13] Ali, F., Cheong, W. J., Al Othman, Z. A., Al Majid, A. M., J. Chromatogr. A 2013, 1303, 9–17. [14] Wang, Y., Ai, F., Ng, S., Tan, T. T. Y., J. Chromatogr. A 2012, 1228, 99–109. [15] Brady, R., Woonton, B., Gee, M. L., O’Connor, A. J., Innov. Food Sci. Emerg. Technol. 2008, 9, 243–248. ´ Boissiere, ` [16] Prouzet, E., C., C. R. Chim. 2005, 8, 579. ¨ M., Buchel, ¨ ¨ [17] Unger, K. K., Kumar, D., Grun, G., Ludtke, S., Adam, T., Schumacher, K., Renker, S., J. Chromatogr. A 2000, 892, 47–55. [18] Mutton, I. M., Chromatographia 1998, 4, 291. [19] Halasz, I., Naefe, M., Anal. Chem. 1972, 44, 76–84. [20] Fekete, S., Olah, E., Fekete, J., J. Chromatogr. A 2012, 1228, 57–71. [21] Guiochon, F., Gritti, J. Chromatogr. A 2011, 1218, 1915– 1938. [22] Guiochon, G., J. Chromatogr. A 2007, 1168, 101. [23] Cheong, W. J., Yang, S. H., Ali, F., J. Sep. Sci. 2013, 36, 609–628. [24] Wei, S., Mizaikoff, B., J. Sep. Sci. 2007, 30, 1794–1805. [25] Owens, P. K., Karlsson, L., Lutz, E. S. M., Andersson, L. I., TrAC, Trends Anal. Chem. 1999, 18, 146–154. [26] Lee, J. W., Ali, F., Kim, Y. S., Cheong, W. J., Bull. Korean Chem. Soc. 2014, 35, 2033. [27] Ko, J. H., Baik, Y. S., Park, S. T., Cheong, W. J., J. Chromatogr. A 2007, 1144, 269–274. [28] Han, K. M., Cheong, W. J., Bull. Korean Chem. Soc. 2008, 29, 2281–2283. [29] de Villiers, A., Lestremau, F., Szucs, R., Gelebart, S., David, F., Sandra, P., J. Chromatogr. A 2006, 1127, 60– 69. [30] Mazzeo, J. R., Neue, U. D., Kele, M., Plumb, R. S., Anal. Chem. 2005, 77, 460A–467A.
5 References
[31] Olah, E., Fekete, S., Fekete, J., Ganzler, K., J. Chromatogr. A 2010, 1217, 3642.
[1] Wang, H., Dong, X., Yang, M., TrAC, Trends Anal. Chem. 2012, 31, 96.
[32] Unger, K. K., Skudas, R., Schulte, M. M., J. Chromatogr. A 2008, 1184, 393–415.
[2] Olivier, A., Meyer, F., Raquez, J., Damman, P., Dubois, P., Prog. Polym. Sci. 2012, 37, 157–181.
[33] Roumeliotis, P., Chatziathanassiou, M., Unger, K. K., Chromatographia 1984, 19, 145–150.
[3] Moad, G., Chen, M., Haussler, M., Postma, A., Rizzardo, E., Thang, S. H., Polym. Chem. 2011, 2, 492–519.
[34] Ali, F., Kim, Y. S., Cheong, W. J., Bull. Korean Chem. Soc. 2014, 35, 542.
[4] Soler-Illia, G. J. A. A., Azzaroni, O., Chem. Soc. Rev. 2011, 40, 1107–1150.
[35] Lee, S. M., Zaidi, S. A., Cheong, W. J., Bull. Korean Chem. Soc. 2010, 31, 2943–2948.
[5] Li, Z., Barnes, J. C., Bosoy, A., Stoddart, J. F., Zink, J. I., Chem. Soc. Rev. 2012, 41, 2590–2605.
[36] Hwang, D. G., Zaidi, S. A., Cheong, W. J., Bull. Korean Chem. Soc. 2009, 30, 3127–3130.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
J. Sep. Sci. 2014, 37, 3426–3434
Research Article
C18 -bound porous silica monolith particles as a low-cost high-performance liquid chromatography stationary phase with an excellent chromatographic performance Ground porous silica monolith particles with an average particle size of 2.34 m and large pores (363 Å) exhibiting excellent chromatographic performance have been synthesized on a relatively large scale by a sophisticated sol–gel procedure. The particle size distribution was rather broad, and the d(0.1)/d(0.9) ratio was 0.14. The resultant silica monolith particles were chemically modified with chlorodimethyloctadecylsilane and end-capped with a mixture of hexamethyldisilazane and chlorotrimethylsilane. Very good separation efficiency (185 000/m) and chromatographic resolution were achieved when the C18 -bound phase was evaluated for a test mixture of five benzene derivatives after packing in a stainless-steel column (1.0 mm × 150 mm). The optimized elution conditions were found to be 70:30 v/v acetonitrile/water with 0.1% trifluoroacetic acid at a flow rate of 25 L/min. The column was also evaluated for fast analysis at a flow rate of 100 L/min, and all the five analytes were eluted within 3.5 min with reasonable efficiency (ca. 60 000/m) and resolution. The strategy of using particles with reduced particle size and large pores (363 Å) combined with C18 modification in addition to partial-monolithic architecture has resulted in a useful stationary phase (C18 -bound silica monolith particles) of low production cost showing excellent chromatographic performance. Keywords: C18 modification / Fast analysis / Monolith particles / Stationary phases DOI 10.1002/jssc.201400811
1 Introduction Surface modification on metal oxide surfaces (either particle surface or flat surface) is a useful strategy for expanding the application of such materials in a variety of fields. Development of separation materials via controlled living radical polymerization was reviewed by Wang et al. [1]. Materials of various functional coatings were introduced [2]. A review on optoelectronic applications such as photovoltaics, light emitting iodides, thin-film transistors, sensors, etc., was presented [3]. In another review, breakthrough applications were proposed in the fields of sustainable energy, environmental remediation, biomaterials, pharmaceutical industry, and catalysis [4]. Materials with great versatility and control in drug delivery could be prepared by the functionalCorrespondence: Professor W. J. Cheong, Department of Chemistry, Inha University, 253 Yonghyundong, Namku, Incheon 402751, South Korea E-mail: [email protected] Fax: +82-328675604
Abbreviations: BET, Brunauer–Emmett–Teller; BJH, Barrett– Joyner–Halenda; FE-SEM, field-emission scanning electron microscopy; FE-TEM, field-emission transmission electron microscopy; HETP, height equivalent to a theoretical plate; RAFT, reversible addition-fragmentation chain transfer; TMOS, trimethylorthosilicate C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
ization of mesoporous silica nanoparticles [5]. A wide range of hydrophilic polymer brushes were grown for potential applications in surface lubricity and tribology, tunable surface wetting, and cell detachment [6]. Especially, porous/nonporous silica materials are subject to various surface modifications that result in diverse chromatographic and SPE media [7–11]. Reduction of the average size of silica particles has been the goal in developing chromatographic separation media throughout chromatographic history because of the theoretically expected improvements in chromatographic efficiency. The general trend of reducing the particle size from 100 m to sub-2 m with enhancement of chromatographic performance has been reported during the continuous developing stages [7,12–19]. Stationary phases comprising ultimately small particles with high separation efficiency have been compiled in some review articles [20–22] where core–shell particles and very fine porous particles are the typical phases of interest. Studies of monolith particles prepared by crushing bulk monoliths have been rare although molecularly imprinted polymer particles were prepared in such a way in the early years [23–25]. Ground organic monolith particles were employed as new chromatographic separation media very recently [26]. In our laboratory, silica monolith particles prepared from bulk monoliths have been studied as chromatographic www.jss-journal.com
J. Sep. Sci. 2014, 37, 3426–3434
separation media. We also have reduced the size of silica monolith particles from 10 m to sub-2 m in a series of studies [12,13,27,28] followed by their modification to various stationary phases where polystyrene-bound phases [12, 13] have been reported to be very efficient stationary phases. Somewhat lower separation resolution and lower yield of the final stationary phase due to the many preparation steps are the disadvantages associated with the polystyrene bound phases. In this study, 2.34 m (average) silica particles with large pore size (363 Å) have been prepared from bulk monoliths. The sol–gel process of silica monolith formation has been further elaborated to achieve the goal. The optimized formulation was found after preliminary preparation and examination of a series of batches of silica monolith particles. The scale of silica monolith production of this study was increased by three times in comparison with that of Ref. [12]. The resultant silica monolith particles were then chemically modified by reaction with a C18 reagent followed by an end-capping reagent, and packed in a column (1.0 mm × 150 mm) for examination of chromatographic performance.
Liquid Chromatography
3427
2.3 C18 modification and end capping of silica monolith particles Ground silica monolith particles 600 mg (calcined at 550⬚C for 48 h) were dried at 120⬚C overnight in a vacuum oven and suspended in 17 mL anhydrous toluene followed by stirring for 30 min in a 50 mL round bottom flask. The reaction mixture was subjected to reflux at 115⬚C. Then 350 mg chlorodimethyloctadecylsilane (dissolved in 5 mL anhydrous toluene) was added dropwise with a dropping funnel. The reflux was maintained for 40 h. The temperature of the reaction mixture was adjusted to 60⬚C. Hexamethyldisilazane 120 L and chlorotrimethylsilane 120 L dissolved in 3 mL anhydrous toluene were added to the reaction mixture dropwise. The reaction at 60⬚C was continued overnight. The temperature of the flask was increased to 110⬚C and refluxed for 5–6 h. The product was washed with toluene at room temperature (30 min) and reflux temperature (overnight), and rinsed with acetone (HPLC grade). The solution was filtered and C18 -bound silica monolith particles were dried in a vacuum oven at 70⬚C overnight.
2 Materials and methods 2.1 Reagents and materials All chemical reagents and solvents were of analytical or HPLC grade. Acetic acid, urea, PEG 10 000, trimethylorthosilicate (TMOS), chlorodimethyloctadecylsilane, hexamethyldisilazane, chlorotrimethylsilane, anhydrous toluene, and styrene were purchased from Sigma–Aldrich (St. Louis, MO, USA). HPLC-grade methanol, acetonitrile, acetone, and water were obtained from Mallinckrodt Baker (Phillipsburg, NJ, USA). All the reagents were used as received. Screen frits (3.2 mm diameter, 0.08 mm thickness, and 0.5 m pore size) were obtained from Valco (Houston, TX, USA). Glass-lined stainless-steel tubing (15 cm, 1.0 mm id, 1/8 inch od) and silica capillaries (50 m id, 365 m od) were purchased from Grace (Deerfield, IL, USA).
2.2 Synthesis of ground silica monolith particles The synthesis procedure for soft silica monoliths was further modified to get smaller (2.34 m) silica monolith particles with larger pores in comparison to that of Ref. [12]. The PEG and urea contents were further increased with respect to TMOS, and the relative amount of catalytic solvent (dilute acetic acid) was also increased. Thus, 4000 mg PEG and 4125 mg urea were dissolved in 36 mL 0.01 M acetic acid. A 10 mL aliquot of TMOS was added to the solution in an ice/water condition and stirred vigorously for 60 min. The solution was gelled at 40⬚C in an oven for 48 h, and aged at 120⬚C in an autoclave for 48 h. The residual water was decanted off and the silica monolith was dried at 70⬚C for 20 h, ground with a mortar and pestle for 10 min, and calcined at 550⬚C for 50 h. C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
2.4 Characterization of the bare silica and C18 modified silica monolith particles A BEL-Japan (Osaka, Japan) BELSORP-Max was used to measure the Brunauer–Emmett–Teller (BET)/Barrett–Joyner– Halenda (BJH) N2 adsorption–desorption isotherms at 77 K, BJH pore sizes, and BET-specific surface areas for the bare and C18 -bound silica monolith particles. The samples were out-gassed at 373 K for 10 h to obtain a residual pressure of less than 10–3 Torr. The amount of N2 adsorbed at a relative pressure of P/P0 = 0.99 was used to determine the total pore volume. The SEM images of C18 -bound silica monolith particles were taken by Hitachi (Tokyo, Japan) S-4200 field-emission scanning electron microscopy (FE-SEM) while the TEM images were captured by JEOL (Tokyo, Japan) JEM2100F field-emission transmission electron microscopy (FE-TEM). A Thermo Electron (Waltham, MA, USA) Flash EA1112 elemental analyzer was used to obtain the carbon load data. A Malvern (Worcestershire, UK) Mastersizer 2000 particle size analyzer was used to get particle size distribution.
2.5 Column packing and HPLC An HPLC system similar to that of Ref. [13] was used where a homemade 150 mm glass-lined micro-column, a 10AD pump (Shimadzu, Tokyo, Japan), a Valco (Houston, TX, USA) C14W.05 injector with 50 nL injection loop, a membrane degasser (Shimadzu DGU-14A), and a UV-VIS capillary window detector (Jasco UV-2075) were assembled to construct the LC system. The software Multichro 2000 from Youlingisul (Seoul, Korea) was used for acquisition and processing chromatographic data. www.jss-journal.com
3428
F. Ali and W. J. Cheong
In our previous work, we prepared the polystyrenebound partially sub-2 m particles by reversible additionfragmentation chain transfer (RAFT) polymerization and packed the phase in columns of 1 mm id and 300 mm length, and the columns showed excellent separation efficiency partially owing to the pseudo-core–shell character of the final phase [12, 13]. In current study, we prepared C18 -modified partially sub-1 m particles having no core–shell character at all. Some difficulties were encountered in packing this phase, thus the column dimension was adjusted to 150 mm length. The presence of high population of fine particles below sub-1 m imposes a serious negative effect upon achieving a good packing quality. Thus a major portion of fine particles below sub-1 m were removed by repeated sedimentation before packing. Removal of fine particles was carried out by suspending the final phase in methanol by stirring for 5 min, sonicating for 10 min and keeping the vial for 30 min undisturbed. The C18 -bound particles settled down to the bottom while the fine particles were still suspended in the supernatant. The supernatant was carefully removed and the precipitate was again suspended in methanol and the whole process of sedimentation was repeated three times. Both phases before and after sedimentation treatment were subjected to measurements of particle size distribution. A loss of 22.4% was observed as the fine particles by the sedimentation treatments (three times). These recovered fine particles may be used for other applications. The term “partially sub1 m” was adopted to address the product of reduced particle sizes having some sub-1 m particles although the final product was obtained after removal of a major part of sub1 m particles (the smallest particles). The column was packed as follows. A 0.5 m commercial screen frit was placed in the 1/8 inch port of a reducing union (1/8–1/16), and a glass-lined stainless-steel column (1.0 mm id, 15 cm length) was fitted to the port, and the tubing was connected to the column packer. The slurry was prepared by suspending 170 mg stationary phase in 4.5 mL methanol, and was fed into the reservoir. A pressure sequence was applied. After packing, the inlet union was installed in the same way as the outlet union (with a frit). The column inlet was connected to the injector, and the column outlet, to the capillary window detector. The stock sample solution was prepared by dissolving phenol (0.88 L), acetophenone (0.14 L), 4-methyl-2-nitroaniline (0.32 mg), benzene (2.93 L), and toluene (1.46 L) in 1 mL mobile phase and was stored at 4⬚C. The sample was further diluted for injection.
3 Results and discussion 3.1 Stationary phase characterization 3.1.1 Architecture and morphology The FE-SEM images (A, B, C), FE-TEM images (X, Y), and a microscopic view (Z) of the C18 -bound silica monolith C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
J. Sep. Sci. 2014, 37, 3426–3434
particles are shown in Fig. 1. The microscopic view (Z) and the wide SEM view (A) show that the average size of C18 -modified silica monolith particles is clearly less than 3 m. The close SEM views (B, C) indicate that the surface roughness of silica monolith particles was not altered by C18 modification. On the other hand, surface modification of silica monolith particles with thin polystyrene layers via RAFT polymerization resulted in the particles of much smoother surface [12, 13]. FE-TEM images of C18 -modified silica monolith particles (X, Y) show dim but visible pore structures as projected in dark irregular spots. The average pore size seems to be quite large (Fig. 1X), and this was confirmed by the N2 adsorption–desorption analysis below.
3.1.2 Particle size distribution The particle size distribution data (both number and volume based) of the bare and C18 -bound silica monolith particles are assembled in Table 1 while the volume-based plot is given in Fig. 2A. The effect of sedimentation treatment for the C18 bound particles was also included in Table 1. Visual comparison of volume-based particle size distribution between the bare and C18 -bound silica monolith particles and that between the untreated stationary phase and the sedimentation-treated stationary phase are given in Fig. 2A and 2B, respectively. It is interesting to note that the particle size distribution of the sedimentation-treated C18 -bound phase was narrowed down both in the smaller and larger particle size areas (Fig. 2B and Table 1). The disappearance of the region of smaller size is natural (due to removal of fine particles below sub1 m), but the decay of the region of larger size is against the common expectation. As shown in Table 1, d(0.9) before sedimentation treatment was 7.12 m while d(0.9) after treatment was 5.15 m. We assume that some of the C18 -bound silica monolith particles were physically aggregated during the prior drying stage, and were separated into smaller particles by vigorous stirring and sonication during the sedimentation process. A loss of 22.4% was observed as the fine particles by the sedimentation treatments (three times). Special care was taken to ensure a high quality packing in this study. It is known to be difficult to get the expected separation efficiency when columns of diameter below the standard (4.6 mm) are packed or the particle size is reduced [29–31]. Column dimensions critically affect separation efficiency as reported by Mazzeo et al. [30]. Packing procedures regarding size of packing materials and diverse column dimensions have been reviewed by Unger et al. [32]. High pressure up to 2000 bar can be applied for packing silica particles [33]. We have been making improvements in column packing such as modifying the slurry packer for use under higher pressure up to 1250 bar, adopting home-made tapered connection tubing from the slurry reservoir to the column for smooth flow of the suspended particles, incorporation of vigorous mechanical vibration during the packing procedure, and sufficient conditioning under a lower pressure at the end of packing, etc. Thus column packing www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2014, 37, 3426–3434
3429
Figure 1. FE-SEM images (A, B, C), FE-TEM images (X, Y), and microscopic view (Z) of C18 -bound silica monolith particles. The scale bars of A, B, C, X, Y, and Z are 10 m, 1 m, 1 m, 50 nm, 10 nm, and 2 m, respectively. Table 1. Comparative analysis of particle size distribution (number-based and volume-based) between the bare and C18 -bound silica monolith particles and between the C18 -bound particles before and after removal of fine particles. The length unit is m.
Silica monolith particles (m)
C18 -bound silica monolith particles Including fine particles
Number based Volume based
Without fine particles
d(0.1)a)
d(0.5)b)
d(0.9)c)
Rd d)
d(0.1)a)
d(0.5)b)
d(0.9)c)
Rd d)
d(0.1)a)
d(0.5)b)
d(0.9)c)
Rd d)
0.59 0.82
0.85 2.34
1.86 5.76
0.32 0.14
0.68 1.01
1.03 3.23
2.15 7.12
0.32 0.14
0.73 1.22
1.11 2.63
2.05 5.15
0.36 0.24
a) Particle diameter corresponding to the integrated area ratio of 0.1 when integrated in the range of 0–d diameter. b) Particle diameter corresponding to the integrated area ratio of 0.5 when integrated in the range of 0–d diameter. c) Particle diameter corresponding to the integrated area ratio of 0.9 when integrated in the range of 0–d diameter. d) d(0.1)/d(0.9)
was carried out according to the procedure published in previous studies [12, 13], [34–36] with a little modification as described in Section 2.5.
3.1.3 BET/BJH analysis (pore size distribution) The BET/BJH adsorption–desorption analysis results of this study are summarized in Table 2 and compared with those of C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
previous studies. The BJH adsorption pore size distribution of this study is plotted in Fig. 3 for the bare and C18 -bound silica monolith particles. The data in Table 2 shows that a sequential increase of pore size (and the corresponding pore volume) of silica monolith as 133 [36] < 212 [35] < 343 [13] < 363 Å (current study) results in a sequential decrease of surface area as 343 [36] > 283 [35] > 136 [13] >110 m2 /g (current study). As shown in Fig. 3, the pore size distribution was slightly shifted to the left by C18 modification. This was
www.jss-journal.com
3430
J. Sep. Sci. 2014, 37, 3426–3434
F. Ali and W. J. Cheong
Figure 2. Comparison of volume-based particle size distribution between bare (䊊) and C18 -bound (䊉) silica monolith particles (A) and comparison of volume-based particle size distribution of C18 -bound silica monolith particles before (䊉) and after () removal of fine particles (B). Table 2. Comparison of BET/BJH analysis results of the bare and ligand modified silica monolith particles for the current and previous studies
Silica particles
Ligand bound silica particles Polystyrene bound phases previous studies
˚ a) Pore size (A) Pore volume (cm3 /g)b) Surface area (m2 /g)c)
Current study C18 -bound phase
[36]
[35]
[13]
Current study
[36]
[35]
[13]
133 0.64 343
212 0.83 283
343 1.06 136
363 1.09 110
131 0.32 168
146 0.53 161
252 0.84 131
303 0.95 83
a) BJH adsorption average pore diameter. b) Total pore volume at P/P0 = 0.99. c) BET-specific surface area. Table 3. Elemental analysis results of the phase of this study in comparison with those of previous studies
Stationary phase
Carbon % Hydrogen %
˚ This study 363 A, partially sub-1 m monolith, C18
6.92 0.80
Previous studies
˚ 5 m [36] 133 A, Lichrosorb, polystyrene
˚ 3–5 m [35] 212 A, monolith, C18
[13] 343 A˚ partially sub 2-m, polystyrene
19.96 1.98
7.12 1.61
10.0 1.0
also confirmed in Table 3 showing a minimal decrease in pore volume from 1.09 to 0.95 cm3 /g for C18 binding of this study in contrast to a much greater decrease observed for polystyrene binding in previous studies. In this study, two basic strategies were adopted in preparation of silica monolith particles, that is, decreasing particle size and increasing pore size, which have been successfully achieved as demonstrated above. The main purpose of such strategies is to increase the separation efficiency of the resultant stationary phase. Reduction of particle size surely contributes to improved mass transfer kinetics, but it makes column packing more difficult, thus we had to adjust the column dimension to a shorter length (15 cm) instead of
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
30 cm of the previous study [13]. Increasing the pore size is certainly helpful for better mass transfer kinetics, but the consequently reduced surface area (and number of silanol groups) resulted in low carbon load after C18 modification. We should also note that the pseudo-core–shell character of polystyrene-bound monolith particles [12, 13] is not realized with the C18 -modified monolith particles of this study. We concluded that such conflicting effects were roughly selfcompensating based on the performances of the columns packed with the phase of this study. The common advantage of stationary phases based on silica monolith particles is the partial monolithic structure to enable high flux solvent flow during packing [12, 13].
www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2014, 37, 3426–3434
Figure 3. BJH-adsorption pore size distribution of the bare (o) and C18 -bound (䊉) silica monolith particles.
3.1.4 Elemental analysis The carbon load of the C18 -modified monolith particles of current study (6.92%) seems better than expected if it is compared to that (7.12%) of the previous study [35] (Table 4). The pore size of current silica monolith (363 Å) is much greater than that (212 Å) of Ref. [35]. Therefore, the surface area for current silica monolith (110 m2 /g) is smaller than that (283 m2 /g) of Ref. [35]. The lower surface area should cause a lower number of surface silanol groups and hence a lower carbon load. It seems that the silanol groups in the larger pores were subject to higher chance of silanization reaction with the C18 reagent resulting in a better reaction yield. 3.2 Chromatographic performance of the phase of current study in comparison with those of a commercial phase and the phases of previous studies The chromatographic performance of the phase of this study is demonstrated in Fig. 4 in comparison with a commercial
3431
C18 phase (Alltima C18 , 3 m, 80 Å) and the C18 -bound phase based on silica monolith particles reported in the previous study [35]. The separation efficiency and plate height data of Fig. 4 are assembled in Table 4. The column-to-column and day-to-day reproducibility data in retention times and separation efficiency for the phase of this study (Fig. 4A) are given in Table 5. The separation efficiency data of fast analysis are given in Table 6. The chromatographic resolution data of the phase of this study are given in Table 7 in comparison with those of other phases. The van Deemter plot based on the chromatographic data obtained with the C18 -modified silica monolith particles is given in Fig. 5. The mobile phase was 70:30 v/v acetonitrile/water with 0.1% TFA. The optimized flow rate was 25 L/min. The column backpressure at the optimum flow rate was 98 bar. The average number of theoretical plates (N, Table 4) of the column packed with the C18 -bound silica monolith particles of this study was 27 600. This value corresponds to 184 000/m, which is close to or even better than those of the previous studies (165 000/m [13] or 188 000/m [12]) obtained with the polystyrene bound silica monolith particles bearing some core-shell character. In this study, the monolith formation procedure was modified to result in reduction of particle size, and C18 modification was adopted instead of polystyrene modification of the previous studies [12, 13] to produce more feasible stationary phase for realization of disposable microcolumns in the future. Thus the number-based average particle size of the phase of this study was decreased to 1.11 m while that of Ref. [12] was 1.61 m. On the other hand, the average pore size (363 Å) was increased in comparison to that (252 Å) of Ref. [12]. The scale of silica monolith preparation has also been increased by three times compared to that of Ref. [12,13]. The above-mentioned changes induced some effects. Particle size reduction was beneficial in mass transfer kinetics although its packing was challenging. Increase of pore size contributed somewhat to improvement in mass transfer kinetics. Adoption of C18 modification instead of polystyrene
Table 4. Comparison of number of theoretical plates and height equivalent to a theoretical plate (HETP) (m) measured at the optimum flow rate of the van Deemter curve among the columns packed with various phasesa)
Analytes
Phenol Acetophenone 4-Methyl-2-nitro-aniline Benzene Toluene Average
Column type C18 -bound silica monolith particles of this studyb)
Commercial Alltima C18 (3 m)b)
C18 -bound silica monolith particles (3–5 m) [35]c)
N-value
HETP
N-value
HETP
N-value
HETP
22 000 ± 400 29 000 ± 940 29 200 ± 1200 28 900 ± 1260 28 800 ± 1020 27 600
6.8 5.2 5.1 5.2 5.2 5.4
11 700 ± 250 16 000 ± 220 16 000 ± 200 13 700 ± 380 13 200 ± 410 14 100
12.7 9.38 9.4 10.9 11.3 10.6
27 000 26 900 29 700 29 700 24 600 26 400
11.1 11.2 10.1 10.1 12.2 11.4
a) The mobile phase was 70/30 v/v acetonitrile/water with 0.1% TFA and the detection wavelength was 214 nm. b) The column dimension was 1.0 mm × 150 mm. The mobile phase flow rate was 25 L/min. c) The column dimension was 0.5 mm × 300 mm. The mobile phase flow rate was 7L/min.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
3432
J. Sep. Sci. 2014, 37, 3426–3434
F. Ali and W. J. Cheong
Figure 4. Comparison of chromatograms obtained at 214 nm with different stationary phases under optimized conditions. Chromatograms (A), (B), and (D) were from current study, and (C), from ref. [31]. (A) Our phase of this study; (B) commercial Alltima C18 (3 m); (C) C18 -modified silica monolith particles (3–5 m) of previous study [31]; (D) our phase of current study showing fast analysis. The column dimension of (A), (B), and (D) was 1.0 mm × 150 mm, and that of (C) was 0.5 mm × 300 mm. The mobile phase was 70:30 v/v acetonitrile/water with 0.1% TFA. The flow rate of (A) and (B) was 25 L/min, while those of (C) and (D) were 7 and 100 L/min, respectively. The elution order was phenol, acetophenone, 4-methyl2-nitroanailine, benzene, and toluene. Table 5. Column-to-column and day-to-day reproducibility in N and retention time for the columns (1.0 mm × 150 mm) packed with the phase of this studya)
Analytes
Phenol Acetophenone 4-Methyl- 2-nitroaniline Benzene Toluene
Column to column reproducibility
Day to day reproducibility
N-value
Retention times (min)
N-value
Retention times (min)
Average
% RSD
Average
% RSD
Average
% RSD
Average
% RSD
22 400 28 100 28 100 27 600 27 800
1.8% 3.2% 4.1% 4.4% 3.5%
6.94 8.31 8.85 10.50 12.51
0.6% 1.6% 2.0% 2.7% 3.6%
22 200 28 900 29 200 28 600 28 600
0.9% 0.2% 0.13% 0.74% 0.8%
6.92 8.24 8.74 10.29 12.21
0.3% 0.7% 0.7% 0.8% 1.1%
a) For column-to-column reproducibility data three columns were packed and the averages for both N-values and retention times were calculated while for day-to-day reproducibility only one column was evaluated on three consecutive days. Table 6. Numbers of theoretical plates obtained with our column of current study using a high flow rate (fast analysis)a)
Analytes
Phenol
Acetophenone
4-Methyl-2-nitroaniline
Benzene
Toluene
Average
N-value
10 100 ±150
9150 ± 120
9800 ± 40
8100 ± 80
7800 ± 110
9000
a) Flow rate was 100 L/min with the column dimension of 1.0 mm id × 150 mm length, and the mobile phase was 70/30 v/v acetonitrile/water with 0.1% TFA and the detection wavelength was 214 nm.
modification was beneficial in view of cost reduction, simplicity, and yield of the final phase at the price of loss of the core-shell character [12] of the polystyrene-bound phase. Such conflicting effects were found roughly self-compensating, resulting in similar separation efficiencies per unit length for the columns packed with the phase of this study and that of Ref. [12]. It should be noted that the column length of this study was reduced to 150 mm to ensure good packing instead C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
of 300 mm of the previous study [12]. Thus the separation efficiency per column was only one half, but separation efficiency per meter was similar to that of the previous study [12]. Nevertheless, the chromatographic resolution (average: 6.05) of this study was even better than that of Ref. [12] (average: 5.88) as shown in Table 7. The chromatographic resolution of the C18 -modified monolith particles of this study (Fig. 4A) was close to that (average: 6.02, Fig. 4B) of the Alltima C18 www.jss-journal.com
Liquid Chromatography
J. Sep. Sci. 2014, 37, 3426–3434
3433
Table 7. Comparison of resolution (R) between current and previous studies
Stationary phase
C18 -bound silica monolith particles of this study b) Fast analysis with the phase of this study c) Polystyrene-bound silica monolith particles [12] d) Polystyrene-bound silica monolith particles [13] d) C18 -bound silica monolith particles (3–5m) [35] e) Alltima commercial C18 (3m) b)
Solute pair a) 1–2
2–3
3–4
4–5
Average
7.15
2.70
7.05
7.30
6.05
3.93
1.52
3.98
3.71
3.28
8.15
3.82
5.25
6.31
5.88
5.15
1.43
3.76
3.35
3.42
5.21
1.29
2.63
2.60
2.91
7.37
2.73
6.93
7.08
6.02
a) Solutes: 1, phenol; 2, acetophenone; 3, 4-methyl-2-nitronaniline; 4, benzene; 5, toluene. b) The column dimension was 1.0 mm × 150 mm. The mobile phase was 70:30 v/v acetonitrile/water 25 L/min. c) The column dimension was 1.0 mm × 150 mm. The mobile phase was 70:30 v/v acetonitrile/water 100 L/min. d) The column dimension was 1.0 mm × 300 mm. The mobile phase was 60:40 v/v acetonitrile/water 15 L/min. e) The column dimension was 0.5 mm × 300 mm. The mobile phase was 70:30 v/v acetonitrile/water 7 L/min.
Figure 5. The van Deemter plot based on chromatographic data obtained with the C18 -modified silica monolithic particles of this study using a mobile phase of 70:30 v/v acetonitrile/water with 0.1% TFA.
(spherical, 3 m, 80 Å) phase (Table 7). Note that the retention times for our phase (Fig. 4A) were much shorter than those of the Alltima C18 phase (Fig. 4B) owing to the lower carbon load (6.9% against 15.4%) induced by the larger pore size. Reduction of carbon load is usually accompanied by reduction in resolution if the other conditions are the same. The reason for the comparison of our phase with Alltima C18 , is that the volume-based average particle size distribution (2.63 m) of our phase is comparable to that of Alltima C18 (spherical, C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
with 0.1% TFA at a flow rate of with 0.1% TFA at a flow rate of with 0.1% TFA at a flow rate of with 0.1% TFA at a flow rate of
3 m, 80 Å). The advantageous character of our phase over spherical Alltima C18 particles is that pacing our phase gives rise to partial monolithic character in the packed structure enabling good packing quality and enhanced mass transfer kinetics. The chromatogram shown in Fig. 4C has been taken from Ref. [35] for comparison between the previous C18 -modified silica monolith particles (3–5 m) [35] and the phase of this study. We can see that the chromatographic performance of the material made in this study is far better than that of Ref. [35] as demonstrated in Fig. 4, Tables 4 and 7. For example, compare the separation efficiency: 184 000/m versus 88 000/m. We also found the possibility of our phase for fast analysis owing to the reduced retention based on the low carbon load. When a flow rate of 100 L/min was used, all the analytes were eluted within 3.5 min with good resolution (3.28) and reasonable separation efficiency (60 000/m) as demonstrated in Fig. 4D, Tables 6 and 7. The column backpressure at this flow rate was 371 bar. The column packed with our phase is not a typical monolithic column but it has a partial monolithic character in the packed structure. The plot of height equivalent to theoretical plate (H) as a function of mobile phase velocity (van Deemter plot) is given in Fig. 5. As shown in Fig. 5, the lowest H was ca. 5 m for the test solutes except for phenol. The optimized flow rate was found 25 L/min. The reason why the separation efficiency of phenol is poorer and apart from other solutes is not clear at present. We guess it may be related to the shortest retention time of phenol. Our phase is somewhat hybrid of spherical and typical monolithic phase
www.jss-journal.com
3434
F. Ali and W. J. Cheong
in view of permeability and trend of van Deemter plot [13]. The strategy of packing ground silica monolith particles seems to be more powerful when incorporated with C18 modification in view of production cost of the stationary phase. Modification of silica monolith particles with polystyrene via RAFT polymerization is a lengthy process including attachment of a halogen-terminated spacer, coupling of a RAFT initiator moiety, RAFT polymerization, and various washing steps leading to a rather low overall yield of the final phase [12, 13]. The ground silica monolith particles were claimed to play an important role in future realization of disposable microcolumns owing to the inexpensive production procedure of silica monolith particles [12, 13]. Now it seems to be more feasible to realize disposable microcolumns on account of successful synthesis of C18 -modified silica monolith particles of excellent chromatographic performance. A relevant study is underway.
4 Conclusion Porous silica monolith particles of reduced particle size (average 2.34 m) with an increased pore size (363 Å) have been synthesized, C18 -modified and end-capped. The C18 -modified silica monolith particles were packed in a glass-lined column (1.0 mm × 150 mm) to give the similar separation efficiency per unit length to that of polystyrene bound silica monolith particles of the previous study. Nevertheless, its chromatographic resolution was better, which was comparable to that of a column of the same length packed with a commercial 3 m spherical C18 phase. Moreover, the retention times were much smaller than those of the 3 m spherical C18 phase. Our phase also showed some potential for fast analysis. Thus a new stationary phase of good chromatographic performance and low cost has been developed, and this phase may be useful for realization of disposable microcolumns. This work was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and Future Planning (2012-R1A1A2006066). The authors have declared no conflict of interest.
J. Sep. Sci. 2014, 37, 3426–3434
[6] Edmondson, S., Armes, S. P., Polym. Int. 2009, 58, 307– 316. [7] Qiu, H., Liang, X., Sun, M., Jiang, S., Anal. Bioanal. Chem. 2011, 399, 3307–3322. [8] Namera, A., Miyazaki, S., Saito, T., Nakamoto, A., Anal. Methods. 2011, 3, 2189–2200. [9] Ali, I., Aboul-Enein, H. Y., J. Sep. Sci. 2006, 29, 762–769. [10] Haginaka, J., J. Chromatogr. B 2008, 875, 12–19. [11] Kara, D., Fisher, A., Sep. Purif. Rev. 2012, 41, 267–317. [12] Ali, F., Kim, Y. S., Lee, J. W., Cheong, W. J., J. Chromatogr. A 2014, 1324, 115–120. [13] Ali, F., Cheong, W. J., Al Othman, Z. A., Al Majid, A. M., J. Chromatogr. A 2013, 1303, 9–17. [14] Wang, Y., Ai, F., Ng, S., Tan, T. T. Y., J. Chromatogr. A 2012, 1228, 99–109. [15] Brady, R., Woonton, B., Gee, M. L., O’Connor, A. J., Innov. Food Sci. Emerg. Technol. 2008, 9, 243–248. ´ Boissiere, ` [16] Prouzet, E., C., C. R. Chim. 2005, 8, 579. ¨ M., Buchel, ¨ ¨ [17] Unger, K. K., Kumar, D., Grun, G., Ludtke, S., Adam, T., Schumacher, K., Renker, S., J. Chromatogr. A 2000, 892, 47–55. [18] Mutton, I. M., Chromatographia 1998, 4, 291. [19] Halasz, I., Naefe, M., Anal. Chem. 1972, 44, 76–84. [20] Fekete, S., Olah, E., Fekete, J., J. Chromatogr. A 2012, 1228, 57–71. [21] Guiochon, F., Gritti, J. Chromatogr. A 2011, 1218, 1915– 1938. [22] Guiochon, G., J. Chromatogr. A 2007, 1168, 101. [23] Cheong, W. J., Yang, S. H., Ali, F., J. Sep. Sci. 2013, 36, 609–628. [24] Wei, S., Mizaikoff, B., J. Sep. Sci. 2007, 30, 1794–1805. [25] Owens, P. K., Karlsson, L., Lutz, E. S. M., Andersson, L. I., TrAC, Trends Anal. Chem. 1999, 18, 146–154. [26] Lee, J. W., Ali, F., Kim, Y. S., Cheong, W. J., Bull. Korean Chem. Soc. 2014, 35, 2033. [27] Ko, J. H., Baik, Y. S., Park, S. T., Cheong, W. J., J. Chromatogr. A 2007, 1144, 269–274. [28] Han, K. M., Cheong, W. J., Bull. Korean Chem. Soc. 2008, 29, 2281–2283. [29] de Villiers, A., Lestremau, F., Szucs, R., Gelebart, S., David, F., Sandra, P., J. Chromatogr. A 2006, 1127, 60– 69. [30] Mazzeo, J. R., Neue, U. D., Kele, M., Plumb, R. S., Anal. Chem. 2005, 77, 460A–467A.
5 References
[31] Olah, E., Fekete, S., Fekete, J., Ganzler, K., J. Chromatogr. A 2010, 1217, 3642.
[1] Wang, H., Dong, X., Yang, M., TrAC, Trends Anal. Chem. 2012, 31, 96.
[32] Unger, K. K., Skudas, R., Schulte, M. M., J. Chromatogr. A 2008, 1184, 393–415.
[2] Olivier, A., Meyer, F., Raquez, J., Damman, P., Dubois, P., Prog. Polym. Sci. 2012, 37, 157–181.
[33] Roumeliotis, P., Chatziathanassiou, M., Unger, K. K., Chromatographia 1984, 19, 145–150.
[3] Moad, G., Chen, M., Haussler, M., Postma, A., Rizzardo, E., Thang, S. H., Polym. Chem. 2011, 2, 492–519.
[34] Ali, F., Kim, Y. S., Cheong, W. J., Bull. Korean Chem. Soc. 2014, 35, 542.
[4] Soler-Illia, G. J. A. A., Azzaroni, O., Chem. Soc. Rev. 2011, 40, 1107–1150.
[35] Lee, S. M., Zaidi, S. A., Cheong, W. J., Bull. Korean Chem. Soc. 2010, 31, 2943–2948.
[5] Li, Z., Barnes, J. C., Bosoy, A., Stoddart, J. F., Zink, J. I., Chem. Soc. Rev. 2012, 41, 2590–2605.
[36] Hwang, D. G., Zaidi, S. A., Cheong, W. J., Bull. Korean Chem. Soc. 2009, 30, 3127–3130.
C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.jss-journal.com
