Composite of Au Nanoparticles and Molecularly Imprinted Polymer as

Anal. Chem. 2004, 76, 1310-1315

Composite of Au Nanoparticles and Molecularly Imprinted Polymer as a Sensing Material Jun Matsui,†,‡,§ Kensuke Akamatsu,†,‡,§ Shingo Nishiguchi,‡ Daisuke Miyoshi,‡ Hidemi Nawafune,†,‡,§ Katsuyuki Tamaki,‡ and Naoki Sugimoto*,†,‡,§

High Technology Research Center, Department of Chemistry, Faculty of Science and Engineering, and Frontier Institute for Biomolecular Engineering Research (FIBER), Konan University, 8-9-1 Okamoto, Higashinada-ku, Kobe 658-8501, Japan

A molecularly imprinted polymer with immobilized Au nanoparticles (Au-MIP) is reported as a novel type of sensing material. The sensing mechanism is based upon the variable proximity of the Au nanoparticles immobilized in the imprinted polymer, which exhibits selective binding of a given analyte accompanied by swelling that causes a blue-shift in the plasmon absorption band of the immobilized Au nanoparticles. Using adrenaline as the model analyte, it was shown that molecular imprinting effectively enhanced the sensitivity and selectivity, and accordingly, Au-MIP selectively detects the analyte at 5 µM. The combination of molecular imprinting and the Au nanoparticle-based sensing system was shown to be a general strategy for constructing sensing materials in a tailor-made fashion due to wide applicability of the imprinting technique and the independence of the sensing mechanism from the analyte recognition system. Recently, numerous studies have focused on the synthesis and application of nanostructured materials showing unique features due to the size quantization effects.1 Gold nanoparticles have drawn increasing attention as one such material, whose colloidal solutions exhibit particular colors due to collective oscillations of the surface electrons by visible light and show a colorimetric response for a change in the interparticle distance.2 Especially, Au nanoparticles combined with biomolecules are widely expected to develop new detection systems for specific reactions between biomolecules. Bangs demonstrated immunoassays based on the aggregation of antibody-modified gold nanoparticles through specific binding with an antigen.3 Mirkin et al. showed that the detection of polynucleotides with specific sequences can be performed by allowing the association of gold nanoparticles bearing the complementary nucleotide chains through hybridiza* To whom correspondence should be addressed. Telephone: +81-78-4314341. Fax: +81-78-435-2539. E-mail: [email protected]. † High Technology Research Center. ‡ Department of Chemistry. § Frontier Institute for Biomolecular Engineering Research (FIBER). (1) (a) Kreibig, U., Vollmer, M. Optical Properties of Metal Clusters; SpringerVerlag: Berlin, 1995. (b) Schmid, G., Ed. Clusters and Colloids, from Theory to Applications; VCH: New York, 1994. (c) Alivisatos, A. P. J. Phys. Chem. 1996, 100, 13226. (d) Fendler, J. H.; Meldrum, F. C. Adv. Mater. 1995, 7, 607. (2) (a) Takeuchi, Y.; Ida, T.; Kimura, K. Surf. Rev. Lett. 1996, 3, 1205-1208. (b) Kreibig, U.; Genzel, L. Surf. Sci. 1985, 156, 678-700. (3) Bangs, L. B. Pure Appl. Chem. 1996, 68, 1873-1879.

1310 Analytical Chemistry, Vol. 76, No. 5, March 1, 2004

tion with the target polynucleotides.4 Other detection systems for an antibody, lectin, and heavy metals have also been reported.5-7 While these systems utilize the aggregation of nanoparticles in solutions, gold nanoparticles immobilized on a solid support appear hopeful for developing new devices. For instance, monolayer-protected gold clusters have found practical uses in nanoelectronic applications8 and optical sensing applications.9 We envisage that gold nanoparticles immobilized in/on a solid support having a biomolecule-like activity would become a novel chemosensor that can promote a specific chemical reaction and convert the chemical event to useful signals such as electronic, spectroscopic, and mechanical outputs. In this study, we synthesized a molecularly imprinted polymer (MIP) as functional support for an Au nanoparticle, as shown in Figure 1, and demonstrated the selective colorimetric detection of a model target, adrenaline,10 using the composite material of MIP and Au nanoparticle (Au-MIP). MIPs are known to act as synthetic receptors that specifically capture given molecules that have been originally added as a template to the polymerization mixture.11 To date, some MIPs have been designed as chemosen(4) (a) Elghanian, R.; Storhoff, J. J.; Mucic, R. C.; Letsinger, R. L.; Mirkin, C. A. Science 1997, 277, 1078-1081. (b) Demers, L. M.; Mirkin, C. A.; Mucic, R. C.; Reynolds, R. A., III; Letsinger, R. L.; Elghanian, R.; Viswanadham, G. Anal. Chem. 2000, 72, 5535-5541. (c) Jin, R.; Wu, G.; Li, Z.; Mirkin, C. A.; Schatz, G. C. J. Am. Chem. Soc. 2003, 125, 1643-1654. (d) Taton, T. A.; Mucic, R. C.; Mirkin, C. A.; Letsinger, R. L. J. Am. Chem. Soc. 2000, 122, 6305-6306. (e) Storhoff, J. J.; Lazarides, A. A.; Mirkin, C. A.; Letsinger, R. L.; Mucic, R. C.; Schatz, G. C. J. Am. Chem. Soc. 2000, 122, 4640-4650. (f) Demers, L. M.; Mirkin, C. A.; Mucic, R. C.; Reynolds, R. A., III; Letsinger, R. L.; Elghanian, R.; Viswanadham, G. Anal. Chem. 2000, 72, 5535-5541. (5) Thanh, N. T. K.; Rosenzweig, Z. Anal. Chem. 2002, 74, 1624-1628. (6) Otsuka, H.; Akiyama, Y.; Nagasaki, Y.; Kataoka, K. J. Am. Chem. Soc. 2001, 123, 8226-8230. (7) Kim, Y.; Johnson, R. C.; Hupp, J. T. Nano Lett. 2001, 1, 165-167. (8) (a) Gittins, D. I.; Bethell, D.; Schiffrin, D. J.; Nichols, R. J. Nature 2000, 486, 67. (b) Chen, S.; Pei, R. J. Am. Chem. Soc. 2000, 72, 2190. (9) Nath, N.; Chilkoti, A. Anal. Chem. 2002, 74, 504-509. (10) Sensing of adrenaline using ISFET fucntionalized with Au nanoparticle: Kharitonov, A. B.; Shipway, A. N.; Willner, I. Anal. Chem. 1999, 71, 54415443. (11) (a) Komiyama, M., Ed. Molecular Imprinting-From Fundamentals to Applications; John Wiley & Sons: New York, 2002. (b) Sellergren, B. Molecularly Imprinted Polymers: Man-Made Mimics of Antibodies and their Application in Analytical Chemistry; Elsevier: Amsterdam, 2001. (c) Bartsch, R. A.; Maeda, M. Molecular and Ionic Recognition with Imprinted Polymers; ACS Symposium Series 703; The American Chemical Society: Washington, DC, 1998. (d) Haupt, K.; Mosbach, K. Chem. Rev. 2000, 100, 2495-2504. (e) Haupt, K.; Mosbach, K. Trends Biotechnol. 1998, 16, 468-475. (f) Wulff, G. Angew. Chem., Int. Ed. Engl. 1995, 34, 1812-1832. (g) Shea K. J. Trends Polym. Sci. 1994, 2, 166. 10.1021/ac034788q CCC: $27.50

© 2004 American Chemical Society Published on Web 01/24/2004

Figure 1. Schematic representation of the preparation of a molecularly imprinted polymer with immobilized Au nanoparticle (A) and the detection of an analyte upon selective swelling of the imprinted polymer (B). (A) Radical polymerization is conducted with a mixture of N,N′-methylenebis(acrylamide) (1), N-isopropylacrylamide (2), acrylic acid (3), adrenaline (4), and 11-undecanoic acid-protected Au nanoparticle (5). The resultant network polymer immobilizes 4 and 5 in the matrix, which is represented by hatching. Extraction of 4 results in leaving behind a complementary binding site with carboxylic moieties. The Au nanoparticle does not leak from the matrix due to its size. (B) The network polymer with a binding site, i.e., the imprinted polymer, is in the shrunken state without the analyte. Rebinding of 4 takes the imprinted polymer to the swollen state, which leads to an increase in the interparticle distance between the Au nanoparticles and the shift of the plasmon absorption band.

sors displaying spectroscopic changes in accordance with the uptake of a target molecule.12 In these cases, however, spectroscopically active monomers having interactions with template species were required to be designed and prepared for each imprinting system. We present a general strategy for constructing macromolecular sensors utilizing Au nanoparticles as a color indicator (Figure 1), where the colorimetry is based on the reversible swelling of a poly(N-isopropylacrylamide)-based MIP in response to the molecular recognition event that alters the proximity of the immobilized Au nanoparticles. EXPERIMENTAL SECTION Reagents. Acrylic acid, 2-aminoethanol hydrochloride, 2,2′azobis(isobutyronitrile) (AIBN), catechol, 3-(3,4-dihydroxyphenyl)DL-alanine (DL-DOPA), chloroform, dimethyl sulfoxide (DMSO), 2-phenylethylamine hydrochloride, 5-hydroxytryptamine hydrochloride (serotonin hydrochloride), tetraoctylammonium bromide (TOAB), toluene, hydrogen tetrachloroaurate(III) (HAuCl4‚4H2O), 11-mercaptoundecanoic acid (MUA) and sodium borohydride (12) (a) Matsui, J.; Higashi, M.; Takeuchi, T.; J. Am. Chem. Soc. 2000, 122, 5218-5219. (b) Turkewitsch, P.; Wandelt, B.; Darling, G. D.; Powell, W. S.; Anal. Chem. 1998, 70, 2025-2030.

were purchased from Wako Pure Chemicals (Osaka, Japan). 3-Hydroxytyramine hydrochloride (dopamine hydrochloride) was obtained from Sigma-Aldrich Japan (Tokyo, Japan), and N,N′methylenebis(acrylamide) (MBA) and N-isopropylacrylamide (NIPA) were purchased from Kanto Kagaku (Tokyo, Japan). DLAdrenaline hydrochloride was obtained from Tokyo Kasei Kogyo (Tokyo, Japan). Toluene, DMSO, and acrylic acid were purified by distillation, and MBA and AIBN were purified by recrystallization prior to use. All other reagents were used without further purification. Preparation of Size-Controlled Au Nanoparticles. This preparation was carried out according to a previously reported procedure.13 A solution of TOAB (3.3 g, 6.0 mmol) in toluene (400 mL) was added to an aqueous solution of HAuCl4‚4H2O (15 mM, 200 mL, 3.0 mmol). A solution of MUA (655 mg, 3.0 mmol) in toluene (100 mL) was then gradually added to the resulting mixture while vigorously stirring, followed by the dropwise addition of a freshly prepared aqueous solution of NaBH4 (0.30 M, 100 mL, 30 mmol). After the mixture was stirred for 1 h, the organic phase was separated and washed with distilled water. The solvent was then completely evaporated in a rotary evaporator and dried in vacuo for 1 day. The black solid thus obtained was heattreated at 155 °C at the heating rate of 2 °C min-1 for 30 min. The heat-treated product was dissolved in 50 mL of toluene, mixed with 800 mL of chloroform to remove the excess TOAB and MUA, and then filtered to give the Au nanoparticles. Preparation of Molecularly Imprinted Polymer. Typical procedure for the preparation of the adrenaline-imprinted polymer (IP) with embedded gold nanoparticles: In a glass tube with a screw cap, a mixture of N-isopropylacrylamide (3.65 mmol), acrylic acid (0.90 mmol), N,N′-methylenebis(acrylamide) (0.23 mmol), and DMSO (2.7 mL), including adrenaline hydrochloride (0.23 mmol) and the mercaptoundecanoic acid-modified Au nanoparticles (230 mg), was heated under a N2 atmosphere at 60 °C for 24 h. The obtained polymer was washed using methanol and water-acetic acid (9:1, v/v) until no desorption of the template was indicated by the UV-visible spectra and then thoroughly rinsed with water. The polymer was next dried in vacuo. A nonimprint polymer (NP) was prepared as a reference without the addition of the template and treated in the same fashion. Transmission Electron Microscope (TEM) Measurement. The dry IP was ground in a mortar and suspended in water. A portion of the suspension was dropped onto a carbon-coated copper grid that was placed on filter paper to drain the solvent. The sample was then dried under reduced pressure and examined using a TEM (JEOL, JEM-2000EX II, 200 kV). Spectroscopic Measurements. A photodiode-array spectrophotometer Multispec-1500 (Shimadzu, Kyoto, Japan) with a thermocell holder and an attached circulation bath (LCH-3000, Advantec, Tokyo, Japan) was used to obtain the UV-visible absorption spectra at 40 °C. Measurements were conducted with a piece of the polymer placed between the inner wall of a quartz cell (1 × 1 cm) and a cover glass that was held with siliconrubber stoppers (Figure 2). Determination of Binding Properties of the Polymers. Batch binding tests were carried out at 15-60 °C using a (13) Akamatsu, K.; Hasegawa, J.; Nawafune, H.; Katayama, H.; Ozawa, F. J. Mater. Chem. 2002, 12, 2862-2865.

Analytical Chemistry, Vol. 76, No. 5, March 1, 2004

1311

Figure 3. TEM image of the composite of the Au nanoparticle and the adrenaline-IP.

Figure 2. Assembling of the composite of the Au nanoparticle and the adrenaline-IP in a quartz cell for spectroscopic measurement.

thermostatic water bath. The polymers (50 mg), IP and NP, were immersed in various concentrations of aqueous adrenaline hydrochloride (5 mL) in screw-capped glass vials. After incubation for 5 h, the supernatants were analyzed by HPLC (Waters alliance HPLC system; column, Waters XTerra RP18; eluent, wateracetonitrile, 7:3) to estimate the amount of adrenaline bound to the polymers. Prior to the binding tests, it was confirmed that an equilibrium is established within 5 h under these conditions. RESULTS AND DISCUSSION Synthesis of Au Nanoparticle-Immobilized Polymers. The design of our sensing material, Au-MIP, is based on the assumption that the proximity of the immobilized Au nanoparticles could be altered by swelling of their support polymer in response to the binding of an analyte. The change in proximity could induce a shift in the plasmon absorption band of the immobilized Au nanoparticles. As the support material, a MIP was prepared using acrylic acid as the functional monomer that was expected to show interaction with adrenaline as a model template during the polymerization and to be immobilized in the resultant MIP as selective binding sites for the template. For composing the adrenaline-IP, NIPA was also employed as a comonomer, expecting the swelling property that depends on the presence of the target molecule.14 To observe the clear shift in surface plasmon resonance, the size and proximity of the Au nanoparticles immobilized in the polymer support would be critical factors. It is believed that the appropriate size of the Au nanoparticles for this purpose is ∼520 nm in diameter and that the interparticle distance that influences the plasmon absorption spectra is comparable or less than the diameter of the Au nanoparticles employed.1 Also, a narrow size distribution of the immobilized nanoparticles would (14) Watanabe, M.; Akahoshi, T.; Tabata, Y.; Nakayama, D. J. Am. Chem. Soc. 1998, 120, 5577-5578.

1312 Analytical Chemistry, Vol. 76, No. 5, March 1, 2004

be required for acquiring a sharp response. Therefore, we carefully determined the conditions for preparation of the Au nanoparticles15,16 and the concentration of Au nanoparticles in the prepolymerization mixture. The resultant IP was examined using a TEM to investigate the size and proximity of the nanoparticles. As shown in Figure 3, the TEM image confirmed the formation of Au nanoparticles with a mean size of 5.3 nm (CV 8.5%, n ) 445) in diameter and uniform dispersion of the nanoparticles in the dry polymer with the approximate interparticle distance of 1.7 nm. The surface of the Au nanoparticles was chemically protected with a polar and lipophilic group, 11-mercaptoundecanoic acid. This is because the nanoparticle exhibits good solubility in DMSO, which was used as the solvent in the polymer synthesis, that was appropriate for compromising the expected template-monomer interaction and the solubility of the cross-linker. Although the Au nanoparticles bear no polymerizable functional group, no leaking of the nanoparticles from the IP was observed during all the experiments, which was confirmed by no absorption around 520 nm in UV-visible spectra of the supernatant. This suggests that the polymer network is fine enough to immobilize Au nanoparticles and allows no localization or distribution of Au nanoparticles within the gel. Therefore, proximity of Au nanoparticles was expected to change only due to swelling/shrinking of the gel. Binding and Swelling Property of the Imprinted Polymer. To ensure the potency of the IP as a sensing material, the binding nature and swelling property of the polymers were studied by batch binding tests. The experiments were conducted at various temperatures ranging from 15 to 60 °C because the polymers, composed of a thermoresponsive monomer NIPA, were expected to show different volumes and resultant variations in their binding nature at different temperatures.17 Figure 4 shows the amount of adrenaline bound to the polymers, the adrenaline-IP, and the NP, when the polymers were incubated in 1 mM aqueous adrenaline hydrochloride. It appeared that the temperature influenced the binding ability of both polymers, and the most effective binding (15) Teranishi, T.; Hasegawa, S.; Shimizu, T.; Miyake, M. Adv. Mater. 2001, 13, 1699-1701. (16) Brust M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R. J. Chem. Soc., Chem. Commun. 1994, 801. (17) Heskins, M.; Guillent, J. E.; James, E. J. Macromol. Sci., Chem. 1968, A2, 1441-1455. (b) Schild, H. G. Prog. Polym. Sci. 1992, 17, 163.

Figure 4. Amount of adrenaline bound to the IP and the NP polymers versus the temperature in which IP and NP were soaked in an aqueous solution of adrenaline (1 mM).

of adrenaline took place at 25 °C. The decreasing amount of the bound adrenaline at the higher temperatures could be attributed to the lower stability of the electrostatic interaction that was expected to be formed between the adrenaline and carboxylic moieties in the polymers18 and to the tight structure of the highly shrunken gel that is not suitable for capturing the target molecule. The slightly lower binding capacity was also observed at 15 °C, at which the gel was highly swollen, suggesting that the too-loose structure is unfavorable for the binding. Although the effect of the swell on the binding is currently unclear, it would be speculated that the too-loose structure is unfavorable for the binding in terms of entropy and that the lower density of hydrophobic isopropyl groups and increased amount of water incorporated into the gel may create environments unfavorable for electrostatic interaction. Thus, the observed binding capacity would be a consequence of combined contributions of these temperature-dependent factors: strength of the electrostatic interaction, accessibility of the carboxylic moieties, and reproducibility of the structure established by the imprinting. Although both polymers exhibited similar profiles, the greater binding ability of IP was observed at all the temperatures tested; the amount of adrenaline bound to IP was 1.8-4.9 times greater than that to NP. The most enhanced binding of IP, compared with NP, was observed at 60 °C. This result is consistent with the formerly reported phenomenon that imprinted polymers exhibited their optimal binding ability at the temperature in which the polymer synthesis was conducted.19 The observed enhancement of binding capacity by imprinting could be moderate as compared to the most successful imprinting systems previously reported.11 This could be attributed to the solvents used for preparing and assessing IP. Although we aimed at detection of an analyte in aqueous media, DMSO was used for the gel preparation because of low solubility of the materials to water. This inconsistent solvent system could be a cause of the relatively moderate imprint effect. Saturation binding tests (data not shown) suggested that the number of binding sites formed in the IP was 75 µmol g-1, and ∼63% of the (18) Israelachvili, J. N. Intermolecular and Surface Forces; Academic Press: New York, 1992. (19) Spivak, D.; Gilmore, M. A.; Shea, K. J. J. Am. Chem. Soc. 1997, 119, 43884393.

Figure 5. Differences in relative volume (V/V0) of the IP and the NP polymers immersed in water and aqueous adrenaline hydrochloride (1.0 mM). V is volume of the polymer in water or aqueous adrenaline solution, and V0 is that at the preparation stage.

binding sites are occupied by adrenaline at 1 mM adrenaline under the conditions. Based on the results of the saturation test, however, estimation of a binding constant was not conducted because the affinity of the binding sites may vary by the swelling degree of the IP, which depends on the concentration of adrenaline as shown in the following experiments. The swelling property of IP and NP was also examined in the range of temperatures shown in Figure 5. The swelling degree (V/V0) appeared to depend not only on the temperature but also on the presence of adrenaline. Furthermore, it was found that the sensitivity to adrenaline had also been enhanced by the imprinting; NP exhibited a similar volume regardless of the presence of adrenaline at the identical temperature, whereas IP showed a significant swelling in response to adrenaline. The observed imprint effect on the swelling property could be attributed to the enhanced binding ability of IP. However, the response in the swelling degree to adrenaline was moderate at 25 °C, though the binding of adrenaline to IP occurred most effectively at that temperature. Also, no significant swelling was observed at 60 °C despite the result that IP exhibited the most enhanced binding capacity, as compared with that of NP. These results suggest that Analytical Chemistry, Vol. 76, No. 5, March 1, 2004

1313

the polymers have to be operated at a temperature that results in a moderately shrunken structure that compromises the capability for analyte uptake and the potential for swelling. Consequently, the most efficient swelling of IP was observed at 40 °C; IP swelled ∼170% at this temperature, while NP was only 110%. Therefore, the following spectroscopic experiments were carried out at 40 °C. It is a known phenomenon that imprinted polymer gels are swollen by a molecular recognition event;14 however, the mechanism of the response is currently uncertain. Because the carboxylic moieties strongly affect the swelling property of the gel, it would be reasonable to assume that the binding of adrenaline to the carboxylic moieties results in the drastic change in the gel volume. It could be speculated that binding of adrenaline causes conformational changes of the polymer gel that trigger incorporation of water by weakening hydrophobic interaction between isopropyl groups in the gel, which is the driving force of the temperature-dependent shrinkage. The observed swelling of IP could be converted to changes in the interparticle distance of the Au nanoparticles based on the assumption that the nanoparticles were in a face-centered cubic lattice. Electron microscope images showed that the interparticle distance was ∼1.7 nm in the dry polymer (IP) of which the volume was ∼44% as compared to that of IP at the preparation stage (V0). Accordingly, the interparticle distances were calculated to be 3.8 (V/V0 0.96) and 5.6 nm (V/V0 1.66) in water and adrenaline solution (1 mM), respectively, at 40 °C. Because it is known that coupling of the surface plasmon resonance influences the absorption spectra when the gold nanoparticles are present within a distance comparable to the particle diameter,1 it could be concluded that the density of the Au nanoparticles and the swelling property were suited to determine the presence of adrenaline by shifting the plasmon absorption band. Colorimetric Detection of Adrenaline. Finally, the IP was used for colorimetric detection of the template species, adrenaline. A slice of cylindrical IP was held between a cover glass and a quartz cell that was filled with an aqueous adrenaline sample and examined using a UV-visible spectrophotometer. Figure 6A shows a typical spectrum of the optical extinction measured in water and aqueous adrenaline hydrochloride (1 mM) at 40 °C. The plasmon absorption band originally appeared with a maximum absorbance at 533 nm when the IP was immersed in water; however, the maximum absorbance was observed at 511 nm with adrenaline, showing that the band was shifted by 22 nm due to the presence of 1 mM adrenaline. The spectroscopic measurements conducted in various concentrations of adrenaline elucidated the sensitivity and dynamic range of the optical response of IP, as shown in Figure 6B. The degree of the blue-shift depends on the concentration of adrenaline ranging from 5 × 10-6 to 2 × 10-3 M, demonstrating that the Au nanoparticles-MIP composite is potentially useful for quantifying the template. It is noteworthy that the sensitivity was comparable to that of the recently reported adrenaline-sensing system using Au nanoparticles and an ionsensitive field effect transistor (ISFET).10 For further assessing the ability of IP as a sensing material, the selectivity of the spectroscopic response was investigated using several kinds of chemicals having structures somewhat similar to adrenaline, as shown in Figure 7. It was found that imprinting 1314 Analytical Chemistry, Vol. 76, No. 5, March 1, 2004

Figure 6. Visible spectra of the imprinted polymer placed in water and adrenaline solution (1 mM) (A) and the wavelength of absorption maximum in various concentrations of adrenaline (B).

enhanced the selectivity for adrenaline; NP exhibited a similar response for most of the tested chemicals, whereas IP showed the largest shift when exposed to adrenaline. NP exhibited significantly larger response to the nontemplate chemicals than IP, as reported in other MIP-based sensor systems.20 In the case of NP, all acrylic acid molecules are potential nonspecific sites; therefore, the number of nonspecific sites in NP should be larger than that in IP. In addition, the volume of NP is larger than that of IP at 40 °C, which suggests that nonspecific sites in NP are sterically less hindered and could be more easily engaged in nonspecific binding, as compared to IP. The selectivity of IP implies a mechanism for the selective binding. A basic functionality would be essential for adsorption to the polymers because catechol induced no spectroscopic response, and the carboxylic moieties derived from acrylic acid would be effective for capturing such chemicals by an electrostatic interaction. Among the chemicals having a basic functionality, adrenaline exhibited a significant shift in the plasmon absorption of IP, while the others resulted in a slight shift, suggesting that the binding to IP is not a simple ion exchange phenomenon but molecular recognition based on multiple-point interactions with (20) Shoji, R.; Takeuchi, T.; Kubo, I. Anal. Chem. 2003, 75, 4882-4886.

Figure 7. Selectivity of the IP and the NP. Both polymers were incubated with 0.1 mM concentration of each chemical independently in water at 40 °C.

the carboxylic and other functionalities in IP.21 The fact that dopamine and DOPA induced a poor shift implies that the hydroxyl group of adrenaline could be a recognition site. Also, the result that serotonine and 2-phenylethylamine showed a slight shift implies that the 3,4-dihydroxylphenyl group could be supplementarily recognized. (21) (a) Annaka, M.; Tanaka, T. Nature 1992, 355, 430. (b) Pande, V. S.; Grosberg, A. Y.; Tanaka, T. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 12976. (c) Pande, S. V.; Grosberg, A. Y.; Tanaka, T. Macromolecules 1995, 28, 2218.

CONCLUSIONS The molecularly imprinted polymer with embedded gold nanoparticles (Au-MIP) was demonstrated as a potential sensing material that exhibits a blue-shift in its absorption spectrum in response to the target molecule. The sensitivity and selectivity of the polymer-Au nanoparticle composite was significantly enhanced by molecular imprinting. It was found that the detection limit was comparable to that for the previously reported MIPs prepared with an elaborately designed, spectroscopically active functional monomer.12a The Au-MIP presented in this study employed Au nanoparticles as an indicator that is independent of the analyte-recognizing mechanism, requiring no individual design or synthesis concerning the development of sensor functions. Also, the imprinting technology can be applied to a wide range of molecules simply by changing a template species in principle. Therefore, the presented strategy for composing the sensing material would be generally accepted and employed in various applications. Compared with the aggregation system, the degree of color changes would be moderate because the transfer of gold nanoparticles is limited within the polymer matrixes. However, considerable merits still lie with Au-MIP. The composite material will be extremely facile for repeated use and for composing highly sensitive sensors by installation in analytical instruments such as a surface plasmon resonance sensor. It should also be noted that the aggregation system requires an analyte to be quite large because it should work as a linker between Au particles, while Au-MIP is capable of capturing a small molecule by its threedimensional network as demonstrated in this study. ACKNOWLEDGMENT This work was partially supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture, Japan.

Received for review July 11, 2003. Accepted December 10, 2003. AC034788Q

Analytical Chemistry, Vol. 76, No. 5, March 1, 2004

1315