- Email: [email protected]
Silica-based monolithic capillary columns—Effect of preparation temperature on separation efficiency
Journal of Chromatography A, 1217 (2010) 5737–5740
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Short communication
Silica-based monolithic capillary columns—Effect of preparation temperature on separation efficiency Josef Planeta ∗ , Dana Moravcová, Michal Roth, Pavel Karásek, Vladislav Kahle Institute of Analytical Chemistry of the ASCR, v. v. i., Veveˇrí 97, 60200 Brno, Czech Republic
a r t i c l e
i n f o
Article history: Received 4 February 2010 Received in revised form 25 June 2010 Accepted 7 July 2010 Available online 15 July 2010 Keywords: Monolithic column Sol–gel process Bypass channel Temperature effect Silica capillary column
a b s t r a c t The temperature effects during the sol–gel process and ageing of the silica-based monolith on the structure and separation efficiency of the capillary columns (100 m i.d., 150 mm) for HPLC separations were studied. The tested columns were synthesized from a mixture of tetramethoxysilane, polyethylene glycol and urea under the acidic conditions. The temperature was varied from 40 ◦ C to 44 ◦ C and formation of bypass channels between the silica mold and the capillary wall was examined. The temperature of 43 ◦ C was estimated as optimal for preparation of efficient silica capillary columns which were subsequently modified by octadecyldimethyl-N,N-diethylaminosilane or covered by poly(octadecyl methacrylate) and tested using standard mixture of alkylbenzenes under the isocratic conditions. © 2010 Elsevier B.V. All rights reserved.
1. Introduction At present, monolithic continuous porous rods have become frequented in separation science and proved to be sufficiently flexible and reliable as stationary phases for liquid chromatography. The massive development and applications of monolithic capillary columns started approximately 10–15 years ago with both silica[1–4] and organic polymer-based [5–7] monoliths. Silica monoliths used for capillary chromatography are usually prepared by sol–gel process involving sequential acid hydrolysis and polycondensation of alkoxy silicon derivatives, mainly tetramethoxysilane (TMOS) in the presence of porogen (polyethylene glycol, PEG) at elevated temperature. Comprehensive studies of preparation and modification methods for monolithic silica columns in analytical as well as capillary formats outlined Tanaka et al. [4,8–10]. Generally, the composition of the starting mixture, i.e., the type and concentration of precursor (usually TMOS) and the concentration of PEG, can control the size of silica skeletons and through-pores, independently to some extent, and of course it has an effect on performance of the prepared capillaries [10,11]. Surprisingly, only spare information has been available on the impact of preparation temperature on the character of resultant silica-based monoliths. However, we have found that the temperature during the gelation step is an important factor affecting the final performances of monolithic capillary columns because
∗ Corresponding author. Tel.: +420 532290176; fax: +420 541212113. E-mail address: [email protected] (J. Planeta). 0021-9673/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2010.07.010
of formation of void spaces along the capillary wall. This paper presents the results of our study concerning the effect of gelation temperature on the separation efficiency of prepared monolithic capillary columns. 2. Experimental 2.1. Instrumentation HPLC equipment consisted of a modified piston micropump MHPP20 (Laboratory Instruments, Prague, Czech Republic) connected to an electrically actuated Valco E90-220 sampling valve with a 60 nL internal sample loop and a lab-made splitter with fused silica capillary (30 m i.d./100 mm). Capillary monolithic column was inserted into the body of splitter, as near as possible to the rotor with the loop. Because of geometric limitations of the set-up configuration, the column end had to be connected with a PTFE sleeve to a 75 m i.d., 170 mm long fused silica capillary with optical cell window in the polyimide coating. UV detection was performed by a Spectra 100 detector (Thermo Separation Products, Waltham, MA, USA) at 254 nm for acetone and at 220 nm for alkylbenzenes. The detector signal was evaluated by chromatography station software Clarity (DataApex, Prague, Czech Republic). 2.2. Materials All fused silica capillaries of 75 and 100 m i.d./375 m o.d. were purchased from Agilent Technologies, Germany. TMOS (purity 99%+), PEG 10,000, urea, n-octadecyl methacrylate (ODM, technical grade), azobisisobutyronitrile (AIBN, 98%)
5738
J. Planeta et al. / J. Chromatogr. A 1217 (2010) 5737–5740
Table 1 Separation efficiencies of unmodified silica-based monolithic columns for different preparation temperatures. The RSDs were calculated for three batches. A ChromolithTM CapRodTM C18 column was employed as a reference column. Mobile phase: 80% (v/v) acetonitrile in water, UV detection at 254 nm. Column dimensions: 100 m i.d., 150 mm length. Linear velocity: 1 mm/s for all experiments. Temperature [◦ C]
Efficiency [plates/m, RSD%]
Total porosity, εT
Permeability, K [×10−14 m2 ]
40 42 43 44 Chromolith
5000 [25%] 74,000 [12%] 95,000 [8%] 86,000 [8%] 98,000
0.970 0.957 0.936 0.929 0.924
10.55 3.24 3.40 3.25 8.66
and octadecyldimethylchlorosilane (95%) were purchased from Sigma–Aldrich (Prague, Czech Republic). Acetic acid (100%, extra pure glacial) and acetonitrile (CHROMASOLV, HPLC gradient grade) were obtained from Riedel-de Haën (Prague, Czech Republic). 3Trimethoxysilylpropyl methacrylate (␥-MAPS, 97%) was purchased from ABCR (Karlsruhe, Germany). Diethylamine (>99.5%) and alkylbenzenes (number of methylene groups n = 0–6) (all >99%) were obtained from Fluka, Switzerland. Ethanol (99.8%, UVAPUR) and acetone (purity p.a.) were purchased from Chromservis (Prague, Czech Republic). Distilled water was produced using own distilling equipment. ChromolithTM CapRodTM C18 column was provided by Merck (Darmstadt, Germany). Column heating was performed in the oven of a HRGC 5300 gas chromatograph (Carlo Erba, Italy). The accuracy of the oven temperature control was checked with a Pt-100 sensor and thermometer (JUMO dTRON 316, Brno, Czech Republic) and was found reproducible with a maximum difference of ±0.1 ◦ C. 2.3. Preparation of monolithic silica capillary columns Monolithic silica capillary columns were prepared following previously reported procedures [8–10] with some modifications in the preparation conditions. A solution of PEG (0.9 g) and urea (0.9 g) in 10 mL 0.01 M acetic acid was prepared and degassed in an ultrasonic bath. TMOS (1 mL) was added to 2.5 mL of the solution in a 10 mL vial and the mixture was magnetically stirred for 10 min. It was observed that the two-phase mixture gradually became homogeneous and the temperature of reaction mixture slightly increased. The prepared solution was injected by syringe into fused silica capillaries 100 m i.d., 25 cm in length. The ends of capillaries were carefully sealed by methane–oxygen micro-flame burner, the capillaries were placed into the GC oven, and gelation of the sol followed at a pre-set temperature for 20 h. Then, the temperature was raised with gradient of 0.5 ◦ C/min up to 120 ◦ C and kept constant for 2 h. After cooling, the column ends were cut off and the prepared silica gels were rinsed with water until neutral reaction of the eluent was obtained. The water in silica rods was replaced by absolute ethanol in amount of about 15 column volumes and the capillary was then allowed to dry at 25 ◦ C for 24 h. Calcination of the prepared silica monoliths was performed by heating from 45 ◦ C with a gradient of 0.5 ◦ C/min up to 120 ◦ C and a gradient of 1.0 ◦ C/min up to 320 ◦ C followed with a final isothermal step for 1 h. After cooling to ambient temperature, the pure silica monolithic column was cut down to 150 mm and used directly for measurement or submitted to chemical modification. Two kinds of stationary phases were prepared: ODS by chemical modification of monolith with octadecyldimethyl-N,Ndiethylaminosilane (ODS-DEA) [12] and ODM by coating with poly(octadecyl methacrylate) [13].
Separation impedance 379,000 5640 3260 4160 1200
on the structure and separation efficiency of the final monolithic capillary columns. The first part of experiments was focused on the study of monolith moulding in the capillary at a temperature from 40 ◦ C to 44 ◦ C. Acetone as an unretained analyte and 80% (v/v) acetonitrile in water as a mobile phase were utilized to investigate the dispersion of solute in the unmodified silica-based monolithic columns. A commercial monolithic capillary column ChromolithTM CapRodTM C18 of the same diameter was tested under the same conditions as the prepared capillary columns and used as a reference column. The test results are shown in Table 1. Monolithic capillary prepared at 40 ◦ C provides an extremely low separation efficiency for unretained analyte (only about 5000 plates/m) and a pronounced peak broadening which can be explained by the presence of bypass zones formed between the monolith and the capillary wall (see Fig. 1). Temperature of 40 ◦ C results in slower reaction rates of hydrolysis and polycondensation of TMOS and in the development of void spaces. An elevated temperature causes a faster kinetics of the condensation reaction and a shorter time available for the gel network to shrink. Remarkable improvement in separation efficiency was observed when the temperature of sol–gel preparation and ageing was increased to 43 ◦ C which was estimated as an optimum temperature where the maximum number of theoretical plates and the minimum separation impedance were obtained (see Table 1). Separation impedances were calculated from E = H2 /K, where H is the height of the theoretical plate of a column for the tested solute and K is the column permeability, K = F × × L/(p × × r2 ), where F is the mobile phase flow rate, is the mobile phase viscosity, p is the pressure drop across the column, L is the column length, and r is the inner radius of the column [14]. The effect of bypass channels on the number of theoretical plates, permeability and separation impedance of the prepared columns is essential. When comparing the columns prepared at 43 ◦ C with ChromolithTM CapRodTM C18, the number of plates is
3. Results and discussion The primary objective of this study was to investigate the effect of temperature during sol–gel forming and ageing of the monolith
Fig. 1. Micrograph of monolithic silica-based capillary columns (100 m i.d., 360 m o.d.) prepared at 40 ◦ C and 43 ◦ C. The sections with disrupted contact between monolith and capillary wall (at 40 ◦ C) are marked by arrows.
J. Planeta et al. / J. Chromatogr. A 1217 (2010) 5737–5740
5739
Fig. 2. The van Deemter plots obtained for ODS () and ODM (䊉) modified silicabased monolithic columns using hexylbenzene as testing solute in mobile phase 80% (v/v) acetonitrile/water. The curves were fitted to equation h = A + B/u + Cu.
similar while the separation impedance is relatively high and, consequently, another optimization (composition, tailoring process) is necessary. A rise in the gelation temperature also resulted in a decrease of the total porosity εT of the column as presented in Table 1. Fig. 1 shows a photograph of the columns prepared at 40 ◦ C and 43 ◦ C where bypass channels are present only at the lower temperature. According to our observations, the drying step was not a critical part of the preparation process and all void spaces, if present, were developed in the wet gel after the tailoring step. Silica monoliths prepared at the optimized temperature of 43 ◦ C were modified to ODS and ODM stationary phases. Fig. 2 presents the resultant van Deemter curves for the ODS and ODM columns. The minimum plate height is equal to 7.5 m (for hexylbenzene) corresponding to 133,000 theoretical plates/m. Both ODS and ODM
Fig. 3. Isocratic separation of a mixture of alkylbenzenes (benzene–hexylbenzene, n = 0–6). Test mixture: 50 L of each in 20 mL of 80% (v/v) acetonitrile, injection 60 nL loop, splitter, UV detection 220 nm. Retention factors for hexylbenzene: k (ODS) = 0.54, k (ODM) = 1.02.
columns use identical monolithic silica support and therefore they show similar separation efficiencies. When comparing the HPLC system used here (see Section 2.1) with direct on-column detection [8–10], it is obvious that the transfer capillary adds extracolumn volumes and partly reduces the separation efficiency. The transfer line contribution to the observed peak variance can be estimated from the Taylor–Aris dispersion
Fig. 4. Scanning electron micrographs of monolithic silica-based columns: (a) pure silica, (b) ODS modification, (c) ODM modification, and (d) clogged monolithic column after ODM modification. The monolithic silica was prepared at optimal temperature (43 ◦ C).
5740
J. Planeta et al. / J. Chromatogr. A 1217 (2010) 5737–5740
equation [15,16] provided that there is no retention of the analyte in the transfer line and the flow in the transfer line is laminar. Employing the diffusion coefficient of hexylbenzene at 25 ◦ C estimated from the literature data [17], the extracolumn broadening in the transfer line was found to be approximately 21% of the total observed variance. Fig. 3 shows the chromatograms of isocratic separations of a mixture of alkylbenzenes (benzene–hexylbenzene, n = 0–6) on ODS and ODM columns. The alkylbenzenes retain more on ODM column which provides a better separation of toluene and ethylbenzene [12,13,18] at the same time. These phenomena can be explained by a higher amount of stationary phase and a higher capacity of ODM column which depend on the radical reaction resulting in formation of a network structure of octadecyl methacrylate polymer. On the contrary, a stationary phase formed from monochlorosilane is represented only by a single layer of alkyls on the silica gel surface for the ODS column. These facts are in agreement with the values of total porosities for the tested columns determined using 80% (v/v) acetonitrile in water as a mobile phase and acetone as an unretained analyte. The total porosity of the monolithic silica-based capillary columns prepared in this study is somewhat reduced after modification with alkyl chains in the order: pure silica-based monolithic column (εT = 0.94) > ODS modified monolithic capillary column (εT = 0.90) > ODM modified monolithic capillary column (εT = 0.79). Permeabilities of the prepared columns were found to be similar for both modifications (ODS 3.20 × 10−14 m2 , ODM 3.15 × 10−14 m2 ) and the prepared columns possessed separation impedances less than 2000 (EODS = 1860, EODM = 1760) calculated for hexylbenzene which is close to well packed particle-filled capillary columns as well as close to commercial ChromolithTM CapRodTM C18 columnwith E of 1320. Three ODS and three ODM monolithic columns were prepared for the determination of preparation reproducibility using hexylbenzene as a test analyte. For both column types, the efficiencies fell within 15% of the respective mean value. Fig. 4a–d shows the morphologies of the monolithic columns prepared at optimal temperature (43 ◦ C) and described above. Fig. 4a corresponds to a pure silica monolithic column, Fig. 4b shows silica-monolithic column modified to ODS stationary phase and a monolithic capillary column modified to ODM stationary phase is in Fig. 4c. We observed a typical “sponge-like” structure of silica with flow-through pores of about 1.5 m size. Fig. 4d shows overloading with the ODM polymer layer where the resulting column is
too dense and clogged and such a column is useless for practical applications. 4. Conclusions Silica monolithic capillary columns achieving good separation efficiency and low pressure drop were prepared and tested in our laboratory. A clear correlation was observed between the sol–gel process temperature and formation of the bypass channels in capillaries. The capillaries prepared at elevated temperature of 43 ◦ C provided the best separation results. This is presumably due to the rate of condensation reaction when wet silica gel network is formed. Acknowledgements This work has been supported by the Czech Science Foundation (Projects GA203/08/1465 and GA203/08/1536), by the Ministry of Education, Youth and Sports of the Czech Republic (Grant No. LC06023), and by the Academy of Sciences of the Czech Republic through Institutional Research Plan No. AV0Z40310501. References [1] [2] [3] [4] [5] [6] [7] [8] [9]
[10] [11] [12] [13] [14] [15] [16] [17] [18]
A.-M. Siouffi, J. Chromatogr. A 1000 (2003) 801. G. Guiochon, J. Chromatogr. A 1168 (2007) 101. ˜ O. Núnez, K. Nakanishi, N. Tanaka, J. Chromatogr. A 1191 (2008) 231. N. Tanaka, H. Kobayashi, N. Ishizuka, H. Minakuchi, K. Nakanishi, K. Hosoya, T. Ikegami, J. Chromatogr. A 965 (2002) 35. E.F. Hilder, F. Svec, J.M.J. Fréchet, J. Chromatogr. A 1044 (2004) 3. J. Urban, P. Jandera, J. Sep. Sci. 31 (2008) 2521. F. Svec, J. Chromatogr. A 1217 (2010) 902. N. Ishizuka, H. Minakuchi, K. Nakanishi, N. Soga, K. Hosoya, N. Tanaka, J. High Resolut. Chromatogr. 21 (1998) 477. N. Tanaka, H. Nagayama, H. Kobayashi, T. Ikegami, K. Hosoya, N. Ishizuka, H. Minakuchi, K. Nakanishi, K. Cabrera, D. Lubda, J. High Resolut. Chromatogr. 23 (2000) 111. T. Hara, H. Kobayashi, T. Ikegami, K. Nakanishi, N. Tanaka, Anal. Chem. 78 (2006) 7632. T. Hara, S. Makino, Y. Watanabe, T. Ikegami, K. Cabrera, B. Smarsly, N. Tanaka, J. Chromatogr. A 1217 (2010) 89. C. Yang, T. Ikegami, T. Hara, N. Tanaka, J. Chromatogr. A 1130 (2006) 175. ˜ O. Núnez, T. Ikegami, W. Kajiwara, K. Miyamoto, K. Horie, N. Tanaka, J. Chromatogr. A 1156 (2007) 35. P.A. Bristow, J.H. Knox, Chromatographia 10 (1977) 279. A. Alizadeh, C.A. Nieto de Castro, W.A. Wakeham, Int. J. Thermophys. 1 (1980) 243. J.G. Atwood, J. Goldstein, J. Phys. Chem. 88 (1984) 1875. J. Li, P.W. Carr, Anal. Chem. 69 (1997) 2530. ˜ O. Núnez, T. Ikegami, K. Miyamoto, N. Tanaka, J. Chromatogr. A 1175 (2007) 7.
Contents lists available at ScienceDirect
Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma
Short communication
Silica-based monolithic capillary columns—Effect of preparation temperature on separation efficiency Josef Planeta ∗ , Dana Moravcová, Michal Roth, Pavel Karásek, Vladislav Kahle Institute of Analytical Chemistry of the ASCR, v. v. i., Veveˇrí 97, 60200 Brno, Czech Republic
a r t i c l e
i n f o
Article history: Received 4 February 2010 Received in revised form 25 June 2010 Accepted 7 July 2010 Available online 15 July 2010 Keywords: Monolithic column Sol–gel process Bypass channel Temperature effect Silica capillary column
a b s t r a c t The temperature effects during the sol–gel process and ageing of the silica-based monolith on the structure and separation efficiency of the capillary columns (100 m i.d., 150 mm) for HPLC separations were studied. The tested columns were synthesized from a mixture of tetramethoxysilane, polyethylene glycol and urea under the acidic conditions. The temperature was varied from 40 ◦ C to 44 ◦ C and formation of bypass channels between the silica mold and the capillary wall was examined. The temperature of 43 ◦ C was estimated as optimal for preparation of efficient silica capillary columns which were subsequently modified by octadecyldimethyl-N,N-diethylaminosilane or covered by poly(octadecyl methacrylate) and tested using standard mixture of alkylbenzenes under the isocratic conditions. © 2010 Elsevier B.V. All rights reserved.
1. Introduction At present, monolithic continuous porous rods have become frequented in separation science and proved to be sufficiently flexible and reliable as stationary phases for liquid chromatography. The massive development and applications of monolithic capillary columns started approximately 10–15 years ago with both silica[1–4] and organic polymer-based [5–7] monoliths. Silica monoliths used for capillary chromatography are usually prepared by sol–gel process involving sequential acid hydrolysis and polycondensation of alkoxy silicon derivatives, mainly tetramethoxysilane (TMOS) in the presence of porogen (polyethylene glycol, PEG) at elevated temperature. Comprehensive studies of preparation and modification methods for monolithic silica columns in analytical as well as capillary formats outlined Tanaka et al. [4,8–10]. Generally, the composition of the starting mixture, i.e., the type and concentration of precursor (usually TMOS) and the concentration of PEG, can control the size of silica skeletons and through-pores, independently to some extent, and of course it has an effect on performance of the prepared capillaries [10,11]. Surprisingly, only spare information has been available on the impact of preparation temperature on the character of resultant silica-based monoliths. However, we have found that the temperature during the gelation step is an important factor affecting the final performances of monolithic capillary columns because
∗ Corresponding author. Tel.: +420 532290176; fax: +420 541212113. E-mail address: [email protected] (J. Planeta). 0021-9673/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.chroma.2010.07.010
of formation of void spaces along the capillary wall. This paper presents the results of our study concerning the effect of gelation temperature on the separation efficiency of prepared monolithic capillary columns. 2. Experimental 2.1. Instrumentation HPLC equipment consisted of a modified piston micropump MHPP20 (Laboratory Instruments, Prague, Czech Republic) connected to an electrically actuated Valco E90-220 sampling valve with a 60 nL internal sample loop and a lab-made splitter with fused silica capillary (30 m i.d./100 mm). Capillary monolithic column was inserted into the body of splitter, as near as possible to the rotor with the loop. Because of geometric limitations of the set-up configuration, the column end had to be connected with a PTFE sleeve to a 75 m i.d., 170 mm long fused silica capillary with optical cell window in the polyimide coating. UV detection was performed by a Spectra 100 detector (Thermo Separation Products, Waltham, MA, USA) at 254 nm for acetone and at 220 nm for alkylbenzenes. The detector signal was evaluated by chromatography station software Clarity (DataApex, Prague, Czech Republic). 2.2. Materials All fused silica capillaries of 75 and 100 m i.d./375 m o.d. were purchased from Agilent Technologies, Germany. TMOS (purity 99%+), PEG 10,000, urea, n-octadecyl methacrylate (ODM, technical grade), azobisisobutyronitrile (AIBN, 98%)
5738
J. Planeta et al. / J. Chromatogr. A 1217 (2010) 5737–5740
Table 1 Separation efficiencies of unmodified silica-based monolithic columns for different preparation temperatures. The RSDs were calculated for three batches. A ChromolithTM CapRodTM C18 column was employed as a reference column. Mobile phase: 80% (v/v) acetonitrile in water, UV detection at 254 nm. Column dimensions: 100 m i.d., 150 mm length. Linear velocity: 1 mm/s for all experiments. Temperature [◦ C]
Efficiency [plates/m, RSD%]
Total porosity, εT
Permeability, K [×10−14 m2 ]
40 42 43 44 Chromolith
5000 [25%] 74,000 [12%] 95,000 [8%] 86,000 [8%] 98,000
0.970 0.957 0.936 0.929 0.924
10.55 3.24 3.40 3.25 8.66
and octadecyldimethylchlorosilane (95%) were purchased from Sigma–Aldrich (Prague, Czech Republic). Acetic acid (100%, extra pure glacial) and acetonitrile (CHROMASOLV, HPLC gradient grade) were obtained from Riedel-de Haën (Prague, Czech Republic). 3Trimethoxysilylpropyl methacrylate (␥-MAPS, 97%) was purchased from ABCR (Karlsruhe, Germany). Diethylamine (>99.5%) and alkylbenzenes (number of methylene groups n = 0–6) (all >99%) were obtained from Fluka, Switzerland. Ethanol (99.8%, UVAPUR) and acetone (purity p.a.) were purchased from Chromservis (Prague, Czech Republic). Distilled water was produced using own distilling equipment. ChromolithTM CapRodTM C18 column was provided by Merck (Darmstadt, Germany). Column heating was performed in the oven of a HRGC 5300 gas chromatograph (Carlo Erba, Italy). The accuracy of the oven temperature control was checked with a Pt-100 sensor and thermometer (JUMO dTRON 316, Brno, Czech Republic) and was found reproducible with a maximum difference of ±0.1 ◦ C. 2.3. Preparation of monolithic silica capillary columns Monolithic silica capillary columns were prepared following previously reported procedures [8–10] with some modifications in the preparation conditions. A solution of PEG (0.9 g) and urea (0.9 g) in 10 mL 0.01 M acetic acid was prepared and degassed in an ultrasonic bath. TMOS (1 mL) was added to 2.5 mL of the solution in a 10 mL vial and the mixture was magnetically stirred for 10 min. It was observed that the two-phase mixture gradually became homogeneous and the temperature of reaction mixture slightly increased. The prepared solution was injected by syringe into fused silica capillaries 100 m i.d., 25 cm in length. The ends of capillaries were carefully sealed by methane–oxygen micro-flame burner, the capillaries were placed into the GC oven, and gelation of the sol followed at a pre-set temperature for 20 h. Then, the temperature was raised with gradient of 0.5 ◦ C/min up to 120 ◦ C and kept constant for 2 h. After cooling, the column ends were cut off and the prepared silica gels were rinsed with water until neutral reaction of the eluent was obtained. The water in silica rods was replaced by absolute ethanol in amount of about 15 column volumes and the capillary was then allowed to dry at 25 ◦ C for 24 h. Calcination of the prepared silica monoliths was performed by heating from 45 ◦ C with a gradient of 0.5 ◦ C/min up to 120 ◦ C and a gradient of 1.0 ◦ C/min up to 320 ◦ C followed with a final isothermal step for 1 h. After cooling to ambient temperature, the pure silica monolithic column was cut down to 150 mm and used directly for measurement or submitted to chemical modification. Two kinds of stationary phases were prepared: ODS by chemical modification of monolith with octadecyldimethyl-N,Ndiethylaminosilane (ODS-DEA) [12] and ODM by coating with poly(octadecyl methacrylate) [13].
Separation impedance 379,000 5640 3260 4160 1200
on the structure and separation efficiency of the final monolithic capillary columns. The first part of experiments was focused on the study of monolith moulding in the capillary at a temperature from 40 ◦ C to 44 ◦ C. Acetone as an unretained analyte and 80% (v/v) acetonitrile in water as a mobile phase were utilized to investigate the dispersion of solute in the unmodified silica-based monolithic columns. A commercial monolithic capillary column ChromolithTM CapRodTM C18 of the same diameter was tested under the same conditions as the prepared capillary columns and used as a reference column. The test results are shown in Table 1. Monolithic capillary prepared at 40 ◦ C provides an extremely low separation efficiency for unretained analyte (only about 5000 plates/m) and a pronounced peak broadening which can be explained by the presence of bypass zones formed between the monolith and the capillary wall (see Fig. 1). Temperature of 40 ◦ C results in slower reaction rates of hydrolysis and polycondensation of TMOS and in the development of void spaces. An elevated temperature causes a faster kinetics of the condensation reaction and a shorter time available for the gel network to shrink. Remarkable improvement in separation efficiency was observed when the temperature of sol–gel preparation and ageing was increased to 43 ◦ C which was estimated as an optimum temperature where the maximum number of theoretical plates and the minimum separation impedance were obtained (see Table 1). Separation impedances were calculated from E = H2 /K, where H is the height of the theoretical plate of a column for the tested solute and K is the column permeability, K = F × × L/(p × × r2 ), where F is the mobile phase flow rate, is the mobile phase viscosity, p is the pressure drop across the column, L is the column length, and r is the inner radius of the column [14]. The effect of bypass channels on the number of theoretical plates, permeability and separation impedance of the prepared columns is essential. When comparing the columns prepared at 43 ◦ C with ChromolithTM CapRodTM C18, the number of plates is
3. Results and discussion The primary objective of this study was to investigate the effect of temperature during sol–gel forming and ageing of the monolith
Fig. 1. Micrograph of monolithic silica-based capillary columns (100 m i.d., 360 m o.d.) prepared at 40 ◦ C and 43 ◦ C. The sections with disrupted contact between monolith and capillary wall (at 40 ◦ C) are marked by arrows.
J. Planeta et al. / J. Chromatogr. A 1217 (2010) 5737–5740
5739
Fig. 2. The van Deemter plots obtained for ODS () and ODM (䊉) modified silicabased monolithic columns using hexylbenzene as testing solute in mobile phase 80% (v/v) acetonitrile/water. The curves were fitted to equation h = A + B/u + Cu.
similar while the separation impedance is relatively high and, consequently, another optimization (composition, tailoring process) is necessary. A rise in the gelation temperature also resulted in a decrease of the total porosity εT of the column as presented in Table 1. Fig. 1 shows a photograph of the columns prepared at 40 ◦ C and 43 ◦ C where bypass channels are present only at the lower temperature. According to our observations, the drying step was not a critical part of the preparation process and all void spaces, if present, were developed in the wet gel after the tailoring step. Silica monoliths prepared at the optimized temperature of 43 ◦ C were modified to ODS and ODM stationary phases. Fig. 2 presents the resultant van Deemter curves for the ODS and ODM columns. The minimum plate height is equal to 7.5 m (for hexylbenzene) corresponding to 133,000 theoretical plates/m. Both ODS and ODM
Fig. 3. Isocratic separation of a mixture of alkylbenzenes (benzene–hexylbenzene, n = 0–6). Test mixture: 50 L of each in 20 mL of 80% (v/v) acetonitrile, injection 60 nL loop, splitter, UV detection 220 nm. Retention factors for hexylbenzene: k (ODS) = 0.54, k (ODM) = 1.02.
columns use identical monolithic silica support and therefore they show similar separation efficiencies. When comparing the HPLC system used here (see Section 2.1) with direct on-column detection [8–10], it is obvious that the transfer capillary adds extracolumn volumes and partly reduces the separation efficiency. The transfer line contribution to the observed peak variance can be estimated from the Taylor–Aris dispersion
Fig. 4. Scanning electron micrographs of monolithic silica-based columns: (a) pure silica, (b) ODS modification, (c) ODM modification, and (d) clogged monolithic column after ODM modification. The monolithic silica was prepared at optimal temperature (43 ◦ C).
5740
J. Planeta et al. / J. Chromatogr. A 1217 (2010) 5737–5740
equation [15,16] provided that there is no retention of the analyte in the transfer line and the flow in the transfer line is laminar. Employing the diffusion coefficient of hexylbenzene at 25 ◦ C estimated from the literature data [17], the extracolumn broadening in the transfer line was found to be approximately 21% of the total observed variance. Fig. 3 shows the chromatograms of isocratic separations of a mixture of alkylbenzenes (benzene–hexylbenzene, n = 0–6) on ODS and ODM columns. The alkylbenzenes retain more on ODM column which provides a better separation of toluene and ethylbenzene [12,13,18] at the same time. These phenomena can be explained by a higher amount of stationary phase and a higher capacity of ODM column which depend on the radical reaction resulting in formation of a network structure of octadecyl methacrylate polymer. On the contrary, a stationary phase formed from monochlorosilane is represented only by a single layer of alkyls on the silica gel surface for the ODS column. These facts are in agreement with the values of total porosities for the tested columns determined using 80% (v/v) acetonitrile in water as a mobile phase and acetone as an unretained analyte. The total porosity of the monolithic silica-based capillary columns prepared in this study is somewhat reduced after modification with alkyl chains in the order: pure silica-based monolithic column (εT = 0.94) > ODS modified monolithic capillary column (εT = 0.90) > ODM modified monolithic capillary column (εT = 0.79). Permeabilities of the prepared columns were found to be similar for both modifications (ODS 3.20 × 10−14 m2 , ODM 3.15 × 10−14 m2 ) and the prepared columns possessed separation impedances less than 2000 (EODS = 1860, EODM = 1760) calculated for hexylbenzene which is close to well packed particle-filled capillary columns as well as close to commercial ChromolithTM CapRodTM C18 columnwith E of 1320. Three ODS and three ODM monolithic columns were prepared for the determination of preparation reproducibility using hexylbenzene as a test analyte. For both column types, the efficiencies fell within 15% of the respective mean value. Fig. 4a–d shows the morphologies of the monolithic columns prepared at optimal temperature (43 ◦ C) and described above. Fig. 4a corresponds to a pure silica monolithic column, Fig. 4b shows silica-monolithic column modified to ODS stationary phase and a monolithic capillary column modified to ODM stationary phase is in Fig. 4c. We observed a typical “sponge-like” structure of silica with flow-through pores of about 1.5 m size. Fig. 4d shows overloading with the ODM polymer layer where the resulting column is
too dense and clogged and such a column is useless for practical applications. 4. Conclusions Silica monolithic capillary columns achieving good separation efficiency and low pressure drop were prepared and tested in our laboratory. A clear correlation was observed between the sol–gel process temperature and formation of the bypass channels in capillaries. The capillaries prepared at elevated temperature of 43 ◦ C provided the best separation results. This is presumably due to the rate of condensation reaction when wet silica gel network is formed. Acknowledgements This work has been supported by the Czech Science Foundation (Projects GA203/08/1465 and GA203/08/1536), by the Ministry of Education, Youth and Sports of the Czech Republic (Grant No. LC06023), and by the Academy of Sciences of the Czech Republic through Institutional Research Plan No. AV0Z40310501. References [1] [2] [3] [4] [5] [6] [7] [8] [9]
[10] [11] [12] [13] [14] [15] [16] [17] [18]
A.-M. Siouffi, J. Chromatogr. A 1000 (2003) 801. G. Guiochon, J. Chromatogr. A 1168 (2007) 101. ˜ O. Núnez, K. Nakanishi, N. Tanaka, J. Chromatogr. A 1191 (2008) 231. N. Tanaka, H. Kobayashi, N. Ishizuka, H. Minakuchi, K. Nakanishi, K. Hosoya, T. Ikegami, J. Chromatogr. A 965 (2002) 35. E.F. Hilder, F. Svec, J.M.J. Fréchet, J. Chromatogr. A 1044 (2004) 3. J. Urban, P. Jandera, J. Sep. Sci. 31 (2008) 2521. F. Svec, J. Chromatogr. A 1217 (2010) 902. N. Ishizuka, H. Minakuchi, K. Nakanishi, N. Soga, K. Hosoya, N. Tanaka, J. High Resolut. Chromatogr. 21 (1998) 477. N. Tanaka, H. Nagayama, H. Kobayashi, T. Ikegami, K. Hosoya, N. Ishizuka, H. Minakuchi, K. Nakanishi, K. Cabrera, D. Lubda, J. High Resolut. Chromatogr. 23 (2000) 111. T. Hara, H. Kobayashi, T. Ikegami, K. Nakanishi, N. Tanaka, Anal. Chem. 78 (2006) 7632. T. Hara, S. Makino, Y. Watanabe, T. Ikegami, K. Cabrera, B. Smarsly, N. Tanaka, J. Chromatogr. A 1217 (2010) 89. C. Yang, T. Ikegami, T. Hara, N. Tanaka, J. Chromatogr. A 1130 (2006) 175. ˜ O. Núnez, T. Ikegami, W. Kajiwara, K. Miyamoto, K. Horie, N. Tanaka, J. Chromatogr. A 1156 (2007) 35. P.A. Bristow, J.H. Knox, Chromatographia 10 (1977) 279. A. Alizadeh, C.A. Nieto de Castro, W.A. Wakeham, Int. J. Thermophys. 1 (1980) 243. J.G. Atwood, J. Goldstein, J. Phys. Chem. 88 (1984) 1875. J. Li, P.W. Carr, Anal. Chem. 69 (1997) 2530. ˜ O. Núnez, T. Ikegami, K. Miyamoto, N. Tanaka, J. Chromatogr. A 1175 (2007) 7.