Discrimination of Saccharides with a Fluorescent Molecular Imprinting

Jun 9, 2009 - Thus, the inner filter effect exerted by the colored media on both the exciting light and emitted light is negligible. As the other main...
0 downloads 11 Views 876KB Size
Download PDF
Anal. Chem. 2009, 81, 5273–5280

Discrimination of Saccharides with a Fluorescent Molecular Imprinting Sensor Array Based on Phenylboronic Acid Functionalized Mesoporous Silica Jin Tan, He-Fang Wang, and Xiu-Ping Yan* Research Center for Analytical Sciences, College of Chemistry, Nankai University, 94 Weijin Road, Tianjin 300071, China A fluorescent indicator-displacement molecular imprinting sensor array based on phenylboronic acid functionalized mesoporous silica was developed for discriminating saccharides. D-Fructose imprinted material (FruIM), Dxylose imprinted material (XylIM) together with a control blank nonimprinted material (NIM) were synthesized as the elements of the imprinting sensor array. Spectrofluorimetric titrations of the three materials with eight selected saccharides were carried out, and Stern-Volmer quenching constants (KSV) of NIM, FruIM, and XylIM with the eight selected saccharides were obtained to investigate the interaction of the materials with saccharides. The present approach couples molecular imprinting technique to indicator-displacement strategy with the use of one conventional saccharide receptor (phenylboronic acid) and one commercially available fluorescent dye (Alizarin Red S., ARS) as the indicator, and allows identifying two template saccharides (D-fructose and D-xylose) plus eight nontemplate saccharides (D-arabinose, D-glucose, D-galactose, D-mannose, L-sorbose, D-ribose, L-rhamnose and sucrose). The principal component analysis (PCA) plot shows a clear discrimination of the 10 tested saccharides at 100 mM and the first principal component possesses 94.8% of the variation. Besides, the developed saccharide imprinted sensor array is successfully applied to discriminating three brands of orange juice beverage. As one of the primary biological materials, saccharides play fundamental roles in various biological phenomena.1,2 The design of sensors for saccharides is of considerable interest owing to their broad utility in food, cosmetic, and medicinal applications. The fact that most saccharides have only one kind of functional group (hydroxyl) with different stereochemistry makes the recognition of different saccharides extremely difficult. Unlike selective biosensors3 and chemical receptors4 for saccharides, which tend to suffer from cross-reactivity, chemical sensor * Corresponding author. Fax: (86)22-23506075. E-mail: [email protected]. (1) Davis, A. P.; Wareham, R. S. Angew. Chem., Int. Ed. 1999, 38, 2978–2996. (2) Hurtley, S.; Service, R.; Szuromi, P. Science 2001, 291, 2337. (3) Jelinek, R.; Kolusheva, S. Chem. Rev. 2004, 104, 5987–6016. (4) Shinkai, S.; Takeuchi, M. Biosens. Bioelectron. 2004, 20, 1250–1259. 10.1021/ac900484x CCC: $40.75  2009 American Chemical Society Published on Web 06/09/2009

arrays5-8 taking advantage of such cross-reactivity for differential sensing have been demonstrated to be highly effective tools for saccharide recognition.9-16 Indicator-displacement assay (IDA) is always adopted in saccharide sensor arrays.17,18 Two main approaches are employed in indicator-displacement differential sensors for saccharides. The first approach involves the ensemble of organic indicator dyes and phenylboronic acid/boric acid receptors as probe set.9,10 Although the compounds used in this approach are commercially available, lots of test work is required to select the best indicatorreceptor probe pairs. Synthetic boronic acid receptors are featured in the second approach.11-16 Compared with the first approach, more emphasis is placed on the organic synthesis of various boronic acid receptors with different binding behavior to a group of analyte saccharides. Both approaches show strong discrimination power toward saccharides, but each has its own intrinsic shortcoming. Thus, it is of great significance to develop an alternative approach for saccharide sensor arrays with easy and facile procedures both analytically and synthetically while maintaining the discrimination power. Molecular imprinting technique19-24 provides an alternative source for discrimination power with formalized synthesis procedure. Imprinted materials are highly cross-linked polymer matrices prepared in the presence of a template molecule. The (5) Jurs, P. C.; Bakken, G. A.; McClelland, H. E. Chem. Rev. 2000, 100, 2649– 2678. (6) Albert, K. J.; Lewis, N. S.; Schauer, C. L.; Sotzing, G. A.; Stitzel, S. E.; Vaid, T. P.; Walt, D. R. Chem. Rev. 2000, 100, 2595–2626. (7) Lavigne, J. J.; Anslyn, E. V. Angew. Chem., Int. Ed. 2001, 40, 3118–3130. (8) Wright, A. T.; Anslyn, E. V. Chem. Soc. Rev. 2006, 35, 14–28. (9) Lee, J. W.; Lee, J.-S.; Chang, Y.-T. Angew. Chem., Int. Ed. 2006, 45, 6485– 6487. (10) Lim, S. H.; Musto, C. J.; Park, E.; Zhong, W.; Suslick, K. S. Org. Lett. 2008, 10, 4405–4408. (11) Schiller, A.; Wessling, R. A.; Singaram, B. Angew. Chem., Int. Ed. 2007, 46, 6457–6459. (12) Schiller, A.; Vilozny, B.; Wessling, R. A.; Singaram, B. Anal. Chim. Acta 2008, 627, 203–211. (13) Zaubitzer, F.; Buryak, A.; Severin, K. Chem.sEur. J. 2006, 12, 3928–3934. (14) Koshi, Y.; Nakata, E.; Yamane, H.; Hamachi, I. J. Am. Chem. Soc. 2006, 128, 10413–10422. (15) Edwards, N. Y.; Sager, T. W.; McDevitt, J. T.; Anslyn, E. V. J. Am. Chem. Soc. 2007, 129, 13575–13583. (16) Duggan, P. J.; Offermann, D. A. Tetrahedron 2009, 65, 109–114. (17) Wiskur, S. L.; Ait-Haddou, H.; Lavigne, J. J.; Anslyn, E. V. Acc. Chem. Res. 2001, 34, 963–972. (18) Nguyen, B. T.; Anslyn, E. V. Coord. Chem. Rev. 2006, 250, 3118–3127.

Analytical Chemistry, Vol. 81, No. 13, July 1, 2009

5273

imprinting process involves the copolymerization of functional monomers and a cross-linker in the presence of the target analyte which acts as a molecular template (imprint molecule). The functional monomers initially form a complex with the template and then copolymerize with the cross-linker, their functional groups are thus held in position by the highly cross-linked polymeric structure. Subsequent removal of the template leaves binding sites that is complementary in size, shape and functionality to the analyte. In this way, the polymer keeps a “molecular memory” and exhibits specific binding characteristics for the template and structurally related compounds. Imprinted materials possess high chemical and thermal stability, and can be created to selectively bind a number of different analytes. They can also be rapidly and inexpensively generated, making them amenable to an array setting. The major drawbacks associated with imprinted materials, such as relatively low overall affinity and high level of cross-reactivity toward structurally highly similar species, can actually be advantageous for differential sensor arrays. As long as signals are generated for an analyte from one or more imprinted materials, a characteristic fingerprint for the analyte can be obtained. Pattern recognition can be completed by mathematical multivariate analyses such as principal component analysis (PCA) and linear discriminant analysis (LDA).8 To date, several encouraging works focusing on the design of imprinted materials based sensor arrays have been reported.25-28 Herein, we report a fluorescent molecular imprinting sensor array for discriminating saccharides based on phenylboronic acid functionalized mesoporous silica. It is a simple fluorescent indicator-displacement imprinting sensor array for saccharides taking advantage of molecular imprinting technique with formalized process of material synthesis. D-Fructose imprinted material (FruIM), D-xylose imprinted material (XylIM) together with a control blank nonimprinted material (NIM) are synthesized as the elements of the imprinting sensor array. The coupling of molecular imprinting technique and indicator-displacement strategy with the use of one conventional saccharide receptor (phenylboronic acid) and one commercially available fluorescent dye (Alizarin Red S., ARS) as the indicator provides a pattern-based sensing system for saccharides. As the two selected template saccharides (D-fructose and D-xylose) possess distinct structure and property, their imprinted materials would respond differently to several analyte saccharides, including both template and nontemplate saccharides. Consequently, the array composed of FruIM, XylIM, and NIM allows discriminating the two template saccharides plus eight nontemplate saccharides. The design, synthesis and characterization of such saccharide imprinted mesoporous materials and their applicability as sensor array to (19) (20) (21) (22) (23) (24)

(25) (26) (27) (28)

Dickey, F. H. Proc. Natl. Acad. Sci. U.S.A. 1949, 35, 227–229. Wulff, G.; Sarhan, A. Angew. Chem., Int. Ed. 1972, 11, 341. Wulff, G. Chem. Rev. 2002, 102, 1–28. Haupt, K.; Mosbach, K. Chem. Rev. 2000, 100, 2495–2504. Whitcombe, M. J.; Vulfson, E. N. Adv. Mater. 2001, 13, 467–478. Alexander, C.; Andersson, H. S.; Andersson, L. I.; Ansell, R. J.; Kirsch, N.; Nicholls, I. A.; O’Mahony, J.; Whitcombe, M. J. J. Mol. Recognit. 2006, 19, 106–180. Hirsch, T.; Kettenberger, H.; Wolfbeis, O. S.; Mirsky, V. M. Chem. Commun. 2003, 432–433. Greene, N. T.; Morgan, S. L.; Shimizu, K. D. Chem. Commun. 2004, 1172– 1173. Greene, N. T.; Shimizu, K. D. J. Am. Chem. Soc. 2005, 127, 5695–5700. Takeuchi, T.; Goto, D.; Shinmori, H. Analyst 2007, 132, 101–103.

5274

Analytical Chemistry, Vol. 81, No. 13, July 1, 2009

distinguish real-world samples are described and discussed in detail. EXPERIMENTAL SECTION Materials and Chemicals. All reagents used were of at least analytical grade. Ultrapure water (18.2 MΩ cm) obtained from a WaterPro water system (Labconco Corporation, Kansas City) was used throughout the experiments. D-Fructose, D-xylose, D-arabinose, D-glucose, sucrose, ARS, Sunset Yellow FCF, Tartrazine, cetyltrimethylammonium bromide (CTAB), and tetraethoxysilane (TEOS) were purchased from Guangfu Fine Chemical Research Institute (Tianjin, China). D-Galactose, D-mannose, L-sorbose, D-ribose, L-rhamnose and 3-isocyanatopropyltriethoxysilane were bought from Jingchun Reagent Co., Ltd. (Shanghai, China). The 3-aminophenylboronic acid monohydrate was obtained from Beijing Element Chem-Tech Co., Ltd. (Beijing, China). Three brands of commercial orange juice beverage including Coca Cola Minute Maid, President Orangeate, and Huiyuan orange juice beverage were purchased from local markets. Unless stated otherwise, commercial grade chemicals were used without further purification. Instrumentation. High performance liquid chromatography (HPLC) analysis was carried out with a Waters 600 system (Waters, U.S.). High resolution mass spectra (HRMS) were recorded on an IonSpec 7.0T FT-ICR mass spectrometer with an electronic spray ionization (ESI) source (IonSpec, U.S.). Elemental analysis was performed with a Vario EL analyzer (Elementar, Germany). IR spectra (4000-400 cm-1) in KBr were recorded on a Magna-560 spectrometer (Nicolet, U.S.). N2 adsorptiondesorption isotherms were recorded on a Tristar 3000 surface area and pore size analyzer (Micromeritics, U.S.) at 77 K. UV-vis absorption spectra were measured with a UV-3600 UV-vis spectrophotometer (Shimadzu, Japan). Fluorescence spectra were measured with an F-4500 fluorescence spectrophotometer (Hitachi, Japan). Synthesis of Phenylboronic Acid Appended Triethoxysilane. The 3-aminophenylboronic acid monohydrate (80 mg, 0.5 mmol) was dissolved in THF (3 mL), to which 3-isocyanatopropyltriethoxysilane (120 µL, 0.5 mmol) was added. The mixture was stirred at room temperature for 24 h, and was not isolated from the reaction solvent. The formation of phenylboronic acid appended triethoxysilane was confirmed by HPLC and HRMS. Preparation of Phenylboronic Acid Functionalized Mesoporous Silicas as Saccharide Imprinted Materials. Synthesis of FruIM. D-Fructose (1.8 g, 10 mmol) (template) and CTAB (0.44 g) were dissolved in water (22 mL), to which the phenylboronic acid appended triethoxysilane in reaction solvent (0.5 mmol, in 3 mL of THF) was added. After stirring the mixture for 1 h, TEOS (2.23 mL, 10 mmol, dissolved in 2 mL of MeOH) and NH3 · H2O (15 M, 2 mL) were added. The mixture was stirred for 48 h. The solid product was recovered by filtration and dried in vacuum at 65 °C for 24 h. The sol-gel material was crushed, ground into powder, suspended in acetone, and wet sieved through a 76 µm stainless steel sieve. The fractions passed through the sieve were collected. The resultant fine sol-gel powder was collected by filtration, air-dried, and washed with copious diluted HCl for several times to remove the surfactant and saccharide templates. To ensure the complete removal of the surfactant and saccharide templates, the material was finally

extracted with diluted HCl, washed with copious water and dried in vacuum. XylIM was synthesized in the same way except the addition of D-xylose (1.5 g, 10 mmol) as template. NIM was prepared in the same way without addition of any saccharide template and a control mesoporous material (MCM) was prepared in the same way as NIM but without addition of the phenylboronic acid appended triethoxysilane. Fluorescence Measurements. 0.25 M phosphate buffer solution (PBS) (pH 7.4), 0.5 M solution of saccharides and 10-4 M solution of ARS were all freshly prepared before use. All the fluorescence measurements were carried out with the excitation wavelength of 460 nm. The slit widths of excitation and emission were both 10 nm. In the spectrofluorimetric titration with ARS, to a set of 10 mL calibrated test tubes 5.0 mg of the materials, 2.0 mL of 0.25 M PBS (pH 7.4) and a given concentration of ARS solution were sequentially added. The mixture was then diluted to volume with ultrapure water and mixed thoroughly. In the spectrofluorimetric titration with saccharides, to a set of 10 mL calibrated test tubes 5.0 mg of the materials, 2.0 mL of 0.25 M PBS (pH 7.4), 1.0 mL of 10-4 M ARS and a given concentration of saccharide solution were sequentially added. The mixture was then diluted to volume with ultrapure water, and mixed thoroughly. In the array study for ten saccharides, to a set of 10 mL calibrated test tubes 5.0 mg of the materials, 2.0 mL of 0.25 M PBS (pH 7.4), 1.0 mL of 10-4 M ARS and 2.0 mL of 0.5 M saccharide solution were sequentially added. The mixture was then diluted to volume with ultrapure water and mixed thoroughly. The ten saccharides (D-fructose, D-xylose, Darabinose, D-glucose, D-galactose, D-mannose, L-sorbose, D-ribose, L-rhamnose, and sucrose) were tested against the array (NIM, FruIM, and XylIM) for five replicates, so that a 3 materials × 10 analytes × 5 replicates data matrix of fluorescence intensity was generated. The raw data matrices were processed using PCA in Matlab (The MathWorks Inc., U.S.). In the test of real-world samples, the three brands of orange juice beverage were unsealed and filtered before use, and then filtered by a 0.45 µm cellulose nitrate membrane. To a set of 10 mL calibrated test tubes 5.0 mg of the materials, 2.0 mL of 0.25 M PBS (pH 7.4), 1.0 mL of 10-4 M ARS and 1.0 mL of the filtrate were sequentially added. The mixture was then diluted to volume with ultrapure water and mixed thoroughly. The three samples were tested against the array (NIM, FruIM and XylIM) for five replicates to generate a 3 materials × 3 analytes × 5 replicates data matrix of fluorescence intensity. The raw data matrices were processed using PCA in Matlab. Prior to fluorescence test, the mixture solutions in 10 mL calibrated test tubes were ultrasonicated for 3 min to disperse the particles of the materials and to make the suspension homogeneous. The suspensions were then directly introduced into the cuvette and the fluorescence was measured without stirring (the emission of the suspensions remained constant for 2-3 min; after this time, the fluorescence decreased as a consequence of the particle settling).

RESULTS AND DISCUSSION Design and Synthesis of the Phenylboronic Acid Functionalized Mesoporous Silica as Saccharide Imprinted Material. With the well-known property of reversible covalent cyclic ester formation with diols, boronic acids are intensively used as the receptor elements for various kinds of saccharide sensors29-43 and saccharide imprinted materials.44-51 For the application of boronic acids in sensor arrays, the discrimination power mainly originates from different binding affinities of various boronic acid receptors to saccharides. Compared with synthetic route designed to obtain various boronic acid receptors, molecular imprinting technique provides an alternative source for discrimination power with formalized synthesis procedure. In this work, we chose phenylboronic acid as the receptor element in functional monomer and ARS as the fluorescent indicator to build an imprinting sensor array based on fluorescent indicator-displacement. Phenylboronic acid and ARS are inexpensive with easily commercial availability. ARS is widely used as a general reporter for studying saccharideboronic acid interactions due to a dramatic change in fluorescence intensity when ARS binds boronic acid.52,53 Although the design of more selective probes are most important for a single sensor to recognize a specific saccharide,54 for an imprinting sensor array the sensing element should be cross-responsive toward several analytes and a nonselective probe element is thus more welcome than a selective one. From this point of view, the system of phenylboronic acid and ARS is suitable for recognition and transduction in the design of a saccharide imprinted sensor array. (29) Tong, A.-J.; Yamauchi, A.; Hayashita, T.; Zhang, Z.-Y.; Smith, B. D.; Teramae, N. Anal. Chem. 2001, 73, 1530–1536. (30) Baker, G. A.; Desikan, R.; Thundat, T. Anal. Chem. 2008, 80, 4860–4865. (31) Sun, X.-Y.; Liu, B.; Jiang, Y.-B. Anal. Chim. Acta 2004, 515, 285–290. (32) Boduroglu, S.; El Khoury, J. M.; Reddy, D. V.; Rinaldi, P. L.; Hu, J. Bioorg. Med. Chem. Lett. 2005, 15, 3974–3977. (33) Kim, Y.; Hilderbrand, S. A.; Weissleder, R.; Tung, C.-H. Chem. Commun. 2007, 2299–2301. (34) Xue, C.; Cai, F.; Liu, H. Chem.sEur. J. 2008, 14, 1648–1653. (35) Fang, H.; Kaur, G.; Wang, B. J. Fluoresc. 2004, 14, 481–489. (36) Cao, H.; Diaz, D. I.; DiCesare, N.; Lakowicz, J. R.; Heagy, M. D. Org. Lett. 2002, 4, 1503–1505. (37) Cannizzo, C.; Amigoni-Gerbier, S.; Larpent, C. Polymer 2005, 46, 1269– 1276. (38) Elmas, B.; Senel, S.; Tuncel, A. React. Funct. Polym. 2007, 67, 87–96. (39) Wang, Q.; Li, G.; Xiao, W.; Qi, H.; Li, G. Sens. Actuators, B 2006, 119, 695–700. (40) Gamsey, S.; Baxter, N. A.; Sharrett, Z.; Cordes, D. B.; Olmstead, M. M.; Wessling, R. A.; Singaram, B. Tetrahedron 2006, 62, 6321–6331. (41) Kuzimenkova, M. V.; Ivanov, A. E.; Thammakhet, C.; Mikhalovska, L. I.; Galaev, I. Y.; Thavarungkul, P.; Kanatharana, P.; Mattiasson, B. Polymer 2008, 49, 1444–1454. (42) Ma, W. M. J.; Pereira Morais, M. P.; D’Hooge, F.; van den Elsen, J. M. H.; Cox, J. P. L.; James, T. D.; Fossey, J. S. Chem. Commun. 2009, 532–534. (43) Freeman, R.; Bahshi, L.; Finder, T.; Gill, R.; Willner, I. Chem. Commun. 2009, 764–766. (44) Friggeri, A.; Kobayashi, H.; Shinkai, S.; Reinhoudt, D. N. Angew. Chem., Int. Ed. 2001, 40, 4729–4731. (45) Granot, E.; Tel-Vered, R.; Lioubashevski, O.; Willner, I. Adv. Funct. Mater. 2008, 18, 478–484. (46) Deore, B.; Freund, M. S. Analyst 2003, 128, 803–806. (47) Miyahara, T.; Kurihara, K. Chem. Lett. 2000, 1356–1357. (48) Wang, W.; Gao, S.; Wang, B. Org. Lett. 1999, 1, 1209–1212. (49) Rajkumar, R.; Warsinke, A.; Mo¨hwald, H.; Scheller, F. W.; Katterle, M. Talanta 2008, 76, 1119–1123. (50) Fujita, N.; Shinkai, S.; James, T. D. Chem. Asian J. 2008, 3, 1076–1091. (51) Wulff, G. Pure Appl. Chem. 1982, 54, 2093–2102. (52) Springsteen, G.; Wang, B. Chem. Commun. 2001, 1608–1609. (53) Springsteen, G.; Wang, B. Tetrahedron 2002, 58, 5291–5300. (54) Mader, H. S.; Wolfbeis, O. S. Microchim. Acta 2008, 162, 1–34.

Analytical Chemistry, Vol. 81, No. 13, July 1, 2009

5275

Scheme 1. Schematic Representation of Molecularly Imprinted Mesoporous Silica as Indicator-Displacement Sensor for Saccharides

In the present work, phenylboronic acid moiety was introduced into the imprinting site through the reaction of 3-isocyanatopropyltriethoxysilane and 3-aminophenylboronic acid. A direct synthetic route involving the co-condensation of the obtained phenylboronic acid appended triethoxysilane with TEOS in the presence of the surfactant CTAB was utilized to covalently attach phenylboronic acid molecules into the mesoporous silica matrix.55-58 NIM was prepared without addition of any saccharide template, while for the two imprinted materials, a 20-fold excess of template saccharide with respect to the triethoxysilane monomer was used to ensure that a substantial proportion of the monomers was engaged in the interaction with the template saccharides during cross-linking. As illustrated in Scheme 1, after the removal of the saccharide template with diluted HCl, the binding sites for saccharide were left on the inner surface of the mesopores in the imprinted material. To demonstrate that the discrimination power in an imprinting sensor array mainly comes from the templating process, we chose D-fructose and D-xylose as the templates as they are quite different in both molecular size and binding affinity to boronic acid, and are also relatively common and inexpensive. D-Fructose is a sixcarbon ketose with relatively strong affinity toward boronic acid, whereas D-xylose, a five-carbon monosaccharide, forms moderate ester with boronic acid. Since the two saccharide templates leave distinct binding sites in their own templated materials, the two imprinted materials are expected to respond differently toward several saccharides with different molecular size, stereochemistry (55) Lim, M. H.; Blanford, C. F.; Stein, A. J. Am. Chem. Soc. 1997, 119, 4090– 4091. (56) Van Rhijn, W. M.; De Vos, D. E.; Sels, B. F.; Bossaert, W. D.; Jacobs, P. A. Chem. Commun. 1998, 317–318. (57) Fowler, C. E.; Lebeau, B.; Mann, S. Chem. Commun. 1998, 1825–1826. (58) Hoffmann, F.; Cornelius, M.; Morell, J.; Fro ¨ba, M. Angew. Chem., Int. Ed. 2006, 45, 3216–3251.

5276

Analytical Chemistry, Vol. 81, No. 13, July 1, 2009

and affinity to boronic acid. Furthermore, the templating process would make the two imprinted materials possess different response toward saccharides compared with NIM. Thus, the sensor array composed of the two imprinted materials and NIM should be able to discriminate several analyte saccharides. As shown in Scheme 1 and Supporting Information (SI) Figure S1, the access of ARS via the mesopores into the moieties of boronic acid on the inner surface of the mesopore of the synthesized materials increases the fluorescence due to the ester formation of catechols in ARS with boronic acids in the materials. The addition of analyte saccharides would change the equilibrium and drive some of the ARS molecules out of the binding site, which quenches the fluorescence intensity to some extent (see SI Figure S1). As long as the two imprinted materials and NIM show different binding affinities to analyte saccharides, fluorescence quenching would display unique diagnostic patterns for individual saccharides. Characterization of the Phenylboronic Acid Functionalized Mesoporous Silica. The mesoporosity of the materials is revealed by the N2 adsorption-desorption isotherms. N2 adsorption-desorption isotherms and the corresponding pore size distributions of MCM and FruIM are shown in Figure 1. Both the materials show MCM-41 characteristic type IV curves, with one well-defined step at intermediate partial pressures related to the capillary condensation of N2 inside the mesopores. The Brunauer-Emmett-Teller (BET) surface areas of MCM and FruIM were 576 m2 g-1 and 269 m2 g-1, respectively. Pore size distribution estimated by the Barrett-Joyner-Halenda (BJH) method shows a pore diameter of 2.10 nm and a pore volume of 0.264 cm3 g-1 for MCM while a pore diameter of 1.80 nm and a pore volume of 0.191 cm3 g-1 for FruIM. The reduction in surface area, pore diameter and pore volume is attributed to the functionalization of phenylboronic acid moiety onto the pore walls of the mesoporous silica.

Figure 1. N2 adsorption-desorption isotherms and corresponding pore size distributions of MCM and FruIM.

Figure 2. IR spectra of MCM and FruIM.

The IR spectra of FruIM with respect to MCM provide a solid evidence for the covalent attachment of phenylboronic acid moiety into the mesoporous silica (Figure 2). MCM has MCM-41 characteristic IR spectra, while FruIM, in addition of the MCM41 characteristic bands, shows a peak at 1655 cm-1 assigned to carbonyl in secondary amide and a peak at 1553 cm-1 attributed to secondary amine in amide. These two secondary amide characteristic peaks together with the ones at 2930 cm-1 and 2850 cm-1 which belong to -CH2CH2CH2- suggest that phenylboronic acid be covalently attached into the mesoporous silica matrix after the co-condensation of the phenylboronic acid appended triethoxysilane with TEOS. The correct incorporation of phenylboronic acid moiety into the silica matrix was also proved by the UV spectra of FruIM and 3-aminophenylboronic acid (see SI Figure S2). The 3-aminophenylboronic acid has a strong absorption band at 235 nm and a weak one at 294 nm. FruIM shows two similar bands, indicating the existence of the phenylboronic acid moiety in FruIM. The quantity of organic moieties grafted to the mesoporous silica surface was calculated from the percentage of nitrogen in the material, as estimated by elemental analyses. The result shows that there was (6.3 ± 0.4) × 10-4 mol g-1 of phenylboronic acid molecule in an individual material.

Spectrofluorimetric Titrations of the Materials with ARS. We first measured the response of the materials toward the indicator ARS. Upon addition of ARS to the suspension of FruIM in PBS at pH 7.4, the color of the mixture gradually turned from purple to orange within several minutes (see SI Figure S3). The fluorescence intensity reached 60% of the equilibrium value within 30 min. It required several hours to achieve complete equilibrium and then the fluorescence intensity kept relatively stable within 48 h. The indicator-displacement process is also slow, and it takes about 10 h to reach equilibrium, as indicated from the timedependent fluorescence response of the indicator-displacement system of NIM, FruIM and XylIM (5 mg/10 mL) with ARS (10-5 M) toward D-fructose (100 mM) at pH 7.4 (50 mM PBS) (See SI, Figure S4). However, after equilibrium the fluorescence kept relatively stable at least for 2 days. Considering that the relatively large dye molecule ARS needs time to penetrate into the gel matrix, the slow binding kinetics in this work can be understandable. Similar slow kinetics for other saccharide sensors using boronic acid functionalized gels and ARS were also reported.41,42 To guarantee the complete equilibrium of the fluorescence, we did all the following fluorescence test 12 h after the preparation of the test solution. To study the interaction of the materials with ARS, the spectrofluorimetric titrations of NIM, FruIM and XylIM with ARS were carried out (see SI Figures S5-S7). Upon titration of ARS, the fluorescence intensities all increased significantly but to different degrees for the three materials. Apparent binding constants (adsorption constants) were then determined by a Langmuir-type analysis of the fluorescence titration data.59,60 The binding constants of NIM, FruIM, and XylIM with ARS are 2.1 × 105, 3.7 × 105 and 3.2 × 105 M-1, respectively. The response of NIM, FruIM, and XylIM to ARS can be attributed to the access of ARS via the mesopores into the moieties of boronic acid on the inner surface of the mesopore of these materials due to the ester formation of catechols in ARS with boronic acids in the materials. Compared with NIM, FruIM and XylIM have stronger affinity toward ARS. In NIM, the functional groups and their orientation of boronic acid moieties are randomly distributed in the matrix, which may hinder the access of ARS to the binding sites of phenylboronic acid. However, in imprinted materials FruIM and XylIM, the functional groups are “frozen” in the position where they initially formed the complex with the templates, and the removal of the templates leaves space for the binding sites, facilitating the access of ARS to the binding sites and giving stronger affinity toward ARS. Spectrofluorimetric Titrations of the Materials with Saccharides. To investigate the interaction of the materials with saccharides, the spectrofluorimetric titrations of NIM, FruIM, and XylIM with eight selected saccharides were carried out. All the eight saccharides have moderate to relative large affinity toward phenylboronic acid. Typical spectrofluorimetric titration spectra of NIM and FruIM with D-fructose are presented in Figure 3. With the titration of D-fructose, both the fluorescence intensities of NIM and FruIM suspension mixtures decreased to some degrees. The (59) Descalzo, A. B.; Rurack, K.; Weisshoff, H.; Martı´nez-Ma´n ˜ez, R.; Marcos, M. D.; Amoro´s, P.; Hoffmann, K.; Soto, J. J. Am. Chem. Soc. 2005, 127, 184–200. (60) Descalzo, A. B.; Marcos, M. D.; Martı´nez-Ma´n ˜ez, R.; Soto, J.; Beltra´n, D.; Amoro´s, P. J. Mater. Chem. 2005, 15, 2721–2731.

Analytical Chemistry, Vol. 81, No. 13, July 1, 2009

5277

Figure 3. Changes in the spectrofluorimetric titration spectra of NIM (a) and FruIM (b) (5 mg/10 mL) with D-fructose in the presence of ARS (10-5 M) at pH 7.4 (50 mM PBS) (λex ) 460 nm). Inset: the corresponding Stern-Volmer plots.

extent of fluorescence quenching in the case of FruIM is larger than that in the case of NIM, indicating that FruIM shows positive imprinting effect to D-fructose than to ARS. The three materials responded differently toward the eight selected saccharides in spectrofluorimetric quenching titration. A Stern-Volmer quench equation was then used to analyze the spectrofluorimetric titration data (see SI Figures S8-S29). Stern-Volmer quenching constants (KSV) of NIM, FruIM, and XylIM with the eight saccharides are listed in Table 1. All the materials show relatively small, but distinct quenching constants with the eight saccharides. An untemplated saccharide can also give a fluorescence response due to the access of the saccharide via the mesopores to displace the ARS bound to the moieties of boronic acid on the inner surface of the mesopore of the materials NIM, FruIM, and XylIM. D-Ribose and L-sorbose give better signal as they have stronger affinity to phenylboronic acid in the three mesoporous materials (larger binding constants).53 Figure 4 was constructed to elucidate the imprinting effect of the two imprinted materials compared with NIM. We employed the ratio of the quenching constant of an imprinted material with a certain saccharide to that of NIM with the same saccharide to represent the imprinting effect. It is well-known that the imprinting effect is correlated to two major factors. One is the degree of similarity in size and shape between the target molecule and the template molecule. The other is the spatial arrangement of the functional groups of the binding sites for template molecule. As shown in Figure 4, FruIM has the strongest imprinting effect toward its template D-fructose (a six-carbon ketose). Moreover, the three five-carbon saccharides, D-xylose, D-arabinose, and D-ribose, have smaller molecular size with less steric hindrance 5278

Analytical Chemistry, Vol. 81, No. 13, July 1, 2009

effect and easier to access to the binding sites, so they receive more positive imprinting effect than the six-carbon saccharides, D-glucose, D-galactose, D-mannose, and L-sorbose. Similar phenomenon of such steric hindrance directed imprinting effect was also reported in another saccharide imprinted sensor.44 In the case of XylIM, the template D-xylose (five-carbon monosaccharide) owns the strongest imprinting effect, and another five-carbon saccharide, D-arabinose, receives a more positive imprinting effect than the five six-carbon saccharides due to the steric hindrance effect. However, another selected fivecarbon saccharide, D-ribose, shows the largest negative imprinting effect among the eight saccharides. It may result from the unique conformations of D-ribose in aqueous solutions. Unlike D-xylose and D-arabinose which are predominantly in pyranose form (more than 95%) in aqueous solutions, D-ribose shows a pyranose 80% and furanose 20% coexistence conformations.61,62 The spatial arrangement of the boronic acid groups in the mesopores makes XylIM prefer five-carbon saccharides with similar stereochemistry to its template D-xylose. As can be seen from Figure 4, the eight selected saccharides show quite different ranking patterns in terms of imprinting effect between the two imprinted materials, which offers the imprinting array composed of NIM, FruIM, and XylIM sufficient discrimination power to differentiate these saccharides. Fluorescent Imprinting Sensor Array for Saccharides and Real-World Samples. As each material behaved differently to a specific saccharide, a characteristic fingerprint for individual saccharides based on the patterns of KSV was obtained (Figure 5). The eight selected saccharides can be easily distinguished by their own unique fingerprints. Encouraged by this result, we added two more saccharides including L-rhamnose and sucrose into the analyte pool. We next measured the fluorescence response of an array composed of NIM, FruIM, and XylIM to the ten saccharides with five replicates to generate the 3 materials × 10 analytes × 5 replicates raw data matrix of fluorescence intensity (See SI Table S1). In order to reduce the dimensionality of the data set, to find underlying variables, and to expose the fingerprints more clearly, we applied PCA5 to the array. Figure 6 shows the obtained twodimensional PCA plot for D-fructose, D-xylose, D-arabinose, Dglucose, D-galactose, D-mannose, L-sorbose, D-ribose, L-rhamnose and sucrose to the array composed of NIM, FruIM, and XylIM. The PCA plot shows a clear discrimination of the 10 tested saccharides and the first principal component possesses 94.8% of the variation which is quite similar to that obtained in another saccharide sensor array (PC1 variance ) 95.4%; PC2 variance ) 3.4%) reported by Lee et al.9 The PCA plot suggests that the responses for each saccharide be clustered into tight distinct groupings, demonstrating the reproducibility of the response for each saccharide. These specific binding-based unique fingerprints generated by PCA allow the classification of these saccharides. Two similar works have been reported recently for the discrimination of several conventional carbohydrates (with only hydroxy groups). Lee et al.9 reported a colorimetric discrimination of 23 carbohydrates at 100 mM concentration based on a pH-indicator-pH-change-inducer array. Six probe pairs (boric acid, phenylboronic acid and four commercial organic dyes) were finally (61) Angyal, S. J. Adv. Carbohydr. Chem. Biochem. 1984, 42, 15–68. (62) Angyal, S. J. Adv. Carbohydr. Chem. Biochem. 1991, 49, 19–35.

Table 1. Stern-Volmer Quenching Constants (KSV) of NIM, FruIM and XylIM with Eight Selected Saccharidesa KSV (M-1)b

a

material

Fru

Xyl

Ara

Glu

Gal

Man

Sor

Rib

NIM FruIM XylIM

2.3 3.9 1.5

2.1 2.5 3.0

1.5 1.7 1.9

1.0 0.81 0.93

1.1 0.58 0.52

2.3 1.4 1.4

6.5 5.4 3.0

6.0 5.5 2.2

Saccharides: D-fructose (Fru), D-xylose (Xyl), D-arabinose (Ara), D-glucose (Glu), D-galactose (Gal), D-mannose (Man), L-sorbose (Sor), and (Rib). b All errors are