Silica-based hybrid materials as biocompatible coatings for glucose sensors

Sensors and Actuators B 81 (2001) 68±75

Silica-based hybrid materials as biocompatible coatings for glucose sensors Alexander Krosa,1, Martijn Gerritsenb, Vera S.I. Sprakela, Nico A.J.M. Sommerdijkc,*, John A. Jansenb, Roeland J.M. Noltea a

Department of Organic Chemistry, University of Nijmegen, Nijmegen, The Netherlands b Department of Biomaterials, University of Nijmegen, Nijmegen, The Netherlands c Laboratory of Macromolecular and Organic Chemistry, Eindhoven University of Technology, P.O. Box 513, 5600 MB Eindhoven, The Netherlands Received 11 May 2001; accepted 20 August 2001

Abstract The preparation of sol±gel silica-based biocompatible coatings, which can be used for future implantable glucose sensors is described. Tetraethylorthosilicate (TEOS) was used as precursor for water-borne silicate gels of which the properties were varied by mixing the sol with polyethylene glycol (SG-PEG), heparin (SG-HEP), dextran sulfate (SG-DS), na®on (SG-NAF) or polystyrene sulfonate (SG-PSS). The toxicity of the coatings was examined in vitro using human dermal ®broblasts. All materials showed to be non-toxic and the cell proliferation rate of ®broblasts was found to be dependent on the additive. Glucose measurements using glucose oxidase-based sensors coated with the different hybrid ®lms were performed both in buffered solutions containing bovine serum albumin and in serum. Stable glucose responses were obtained for the coated sensors in both media. The SG-DS containing coating appeared to be most promising for future in vivo glucose measurements. # 2001 Elsevier Science B.V. All rights reserved. Keywords: Sol±gel; Silica; Biocompatibility; Amperometric biosensor; Glucose oxidase; Glucose

1. Introduction Since Clark and Lyons reported their enzyme electrode for measuring glucose in 1962, an enormous amount of literature has been published concerning enzyme-based biosensors [1,2]. Glucose is by far the most studied analyte in this ®eld of research, primarily due to its importance in human metabolic processes [3]. Continuous glucose sensing with a stable in vivo glucose sensor is expected to lead to improved regulation of the glucose concentration and thus to a reduction of the number of complications in diabetes mellitus patients. We have reported on a glucose sensor based upon polypyrrole microtubuli [4,5] in which glucose oxidase was immobilized. Levels of glucose could be measured in vitro for a period extending 3 months [6]. However, to obtain reliable glucose measurements in vivo, an outer

* Corresponding author. Tel.: ‡31-40-2475870; fax: ‡31-40-2451036. E-mail address: [email protected] (N.A.J.M. Sommerdijk). 1 Present address: Department of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, California, USA.

coating of the sensor is required in order to prevent contact between tissue and polypyrrole. The coating should minimize the encapsulation of the sensor after implantation to obtain a suf®ciently fast response to changes in the glucose level. Several coatings prepared for glucose sensors have been published in the literature, such as cellulose acetate [7], polyethyleneglycol [8,9], poly(vinyl chloride) [10], polyurethane [11] and na®on [12±14], but the in vivo tests in general were disappointing, making that there is still room for improvement [15]. Sol±gel derived silicates [16±24] are promising candidates for application as coatings, since they have been demonstrated to be highly compatible with enzymes, even stabilizing and protecting these and other biomolecules [25±32]. Drug delivery is a potential ®eld of application for these silica-based sol±gels. Some studies have already shown to some extent the biocompatibility of these materials [33±36]. Attempts are even made to implant sol±gel encapsulated pancreatic islets [37,38]. Further research, however, is required to develop this possible application as a potential treatment of diabetes mellitus. The chemical properties of the silica-based materials can be tuned by the addition of organic components as

0925-4005/01/$ ± see front matter # 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 2 5 - 4 0 0 5 ( 0 1 ) 0 0 9 3 3 - 9

A. Kros et al. / Sensors and Actuators B 81 (2001) 68±75

was exempli®ed by Tian et al. [39,40] who prepared a biodegradable hybrid material from silicates containing poly(e-caprolactam). In the ®eld of biosensors, Narang et al. [41] encapsulated glucose oxidase in a sol±gel and measured glucose for a period of over 2 months. No attempts, however, were made to test the biocompatibility of these hybrid materials in biomedia. In this paper we describe the cell and tissue response of tetraethylorthosilicate (TEOS)-based composite materials and hybrids containing heparin (SG-HEP), dextran sulfate (SG-DS), na®on SG-NAF, polyethylene glycol (SG-PEG) and polystyrene sulfonate (SG-PSS). In separate experiments we tested these hybrid materials as coatings for our glucose sensors (preliminary results of in vivo biocompatibility studies have also been reported [42]). 2. Experimental section 2.1. Preparation of the sol±gel coatings To TEOS, 4.5 ml was added 1.4 ml water and 0.1 ml aqueous 1.0 N HCl. The two-phase system was stirred vigorously for 3 h and mixed with an aqueous solution of the additive. Subsequently, 80 ml of the resulting sol was spin coated onto the surface of a Thermanox coverslip, using a spin speed of 3000 rpm. The coatings were aged at room temperature for at least 2 days. Before cell culture experiments, all substrates were sterilized by a radiofrequency glow discharge treatment (RFGD, 5 min, 100 mTorr, argon). 2.2. Contact angle measurements The contact angles of the sol±gel coatings were determined before and after RFGD treatment using the sessile drop method. A drop of milli-Q water (10 ml) was placed on the test substrates. The drops were photographed immediately after positioning. The contact angle (y) was calculated from the height (h) and base (b) of the drop according to a literature procedure [43]. 2.3. XPS measurements These experiments are done with a VG Escalab-200 spectrometer using an aluminum anode (Al Ka ˆ 1486:3 eV), with a background pressure of 2  10 9 mbar. Highresolution spectra of silicon (2p) and sulfur (2p) are recorded using the VG S500 data system. 2.4. Infra-red measurements For the infra-red measurements, the sols (80 ml) were spin coated onto the surface of a sodium chloride disc, using a spin speed of 3000 rpm. The coatings were aged at room temperature for at least 2 days. The measurements were performed on a Biorad FTS-25 spectrometer.

69

2.5. Cell proliferation For the cell proliferation assay, sol±gel coated polyethylene terephtalate coverslips (Thermanox) were positioned on the bottom of sterile 24 well plates (Greiner). Thermanox substrates without coating were considered as controls. Human dermal ®broblasts, obtained from a primary culture, were grown in MEM alpha medium with 10% heat inactivated fetal calf serum (Gibco). A suspension of human dermal ®broblasts (1 ml, 1  104 cells/ml) were spread on the sterilized sol±gel coated Thermanox coverslips and the cells were incubated for 1, 3, 6 and 8 days (378C, 5% CO2/ air), respectively. The growth medium was refreshed every 2 days. At the end of the incubation period, the substrates were removed from the wells and rinsed in phosphate buffered saline solution (PBS). Subsequently, the cells attached to the substrates were harvested by trypsin treatment (0.25% w/v) and counted using a Coulter counter. The assay was performed twice and each time in quadruplicate. Growth curves were plotted and compared. 2.6. Cell morphology For the cell morphology assay, a ®broblast suspension (1 ml, 1  104 cells/ml) were spread on the sterilized sol±gel coated Thermanox coverslips and the cells were incubated for 8 days (378C, 5% CO2/air). The non-attached cells were removed by rinsing with 0.1 M PBS (pH 7.4). The cells attached to the substrates were ®xed with 2% (v/v) glutaraldehyde in aqueous 0.1 M sodium cacodylate buffered solution for 30 min at 48C, rinsed twice in cacodylate buffer and dehydrated with a graded series of water±ethanol mixtures. Subsequently, the samples were dried with tetramethylsilane, sputtered with gold and investigated with a Philips SEM-500 scanning electron microscope (SEM) at an accelerating voltage of 12 kV and a working distance of 4 cm. The experiments were performed in triplicate. 2.7. Preparation of the glucose sensor Pyrrole was chemically polymerized within the pores of cyclopore membranes (pore size 1.0 mm, membrane thickness 10 mm) in a specially designed reaction vessel, whereby the membrane divides two compartments [4]. The ®rst chamber was ®lled with 3 ml of pyrrole (0.6 M) in distilled water and the second chamber was ®lled with an aqueous solution of ferric chloride (0.3 M). Polymerization of pyrrole occurred inside the pores of the membrane and was stopped after 1 min by rinsing the membrane with water. The excess of polymer was removed by carefully wiping the membrane surface with a tissue. A layer of platinum (thickness ca. 50 nm) was sputtered on the backside of the membrane with a Baltec SCD 005 sputter coater to ensure good electrical contact. Enzyme immobilization was achieved by agitating the membrane in 10 ml of an aqueous 0.1 M PBS (pH 7.4) containing glucose oxidase (5 mg/ml)

70

A. Kros et al. / Sensors and Actuators B 81 (2001) 68±75

and N,N0 -diethylamine-ethyl dextran (10 mg/ml) for 30 min. The membrane was subsequently dried over CaCl2 at 48C for 16 h. The immobilized track-etch membrane was then replaced in degassed aqueous PBS (pH 7.4) containing resorcinol (15 mM) and 1,3-diaminobenzene (15 mM), to prepare a coating capable of preventing electrochemical fouling of the sensor. Electropolymerization of resorcinol/ 1,3-diaminobenzene was achieved by repeatedly scanning (10 times) between 0.0 and 0.8 V versus Ag/AgCl with a scan speed of 2 mV/s. The sensor was placed on a spin coater and 200 ml of the sol was applied to one side of the membrane. After spinning for 30 s at 3000 rpm, the membrane was dried for 2 days at room temperature.

glucose analyzer. All glucose measurements described in this paper were performed in six-fold. 3. Results and discussion The organic±inorganic hybrid coatings were prepared from tetraethylorthosilicate and SG-PEG, SG-HEP, SGDS, SG-NAF or SG-PSS as described in Section 2. It was found that the additives had a positive effect on the physical behavior of the coatings. Whereas pure silicate gel (SG) showed to be brittle, the hybrid materials were more ¯exible, making them easier to handle. Addition of SG-HEP, SG-DS or SG-PEG to the sol±gel resulted in materials that showed signi®cantly lower contact angles with water (Table 1) compared to the pure silicate. In contrast, the addition of SG-NAF or SG-PSS did not have a signi®cant in¯uence on the contact angle. Prior to the cell culture experiments all coatings were sterilized using a RFGD, which led to a reduction of the contact angles for all coatings (Table 1). This modi®cation of the physicochemical characteristics of a materials surface is known for this kind of sterilization method [44]. Infra-red measurements showed that the RFGD treatment had no measurable effect on the composition of the sol±gels (data not shown). Cell proliferation analysis was used to assess the cytotoxicity of the coatings (Fig. 1A). Human dermal ®broblasts were grown on sol±gel coated Thermanox coverslips for several days and at the end of the incubation, the number of cells present was determined. Despite the fact that the hydrophilicities of the coatings were similar, distinct differences were found in their effect on cell proliferation. The SG, SG-HEP, SG-NAF and SG-PEG coatings give rise to a cell proliferation comparable to that of the reference surface. The rate of cell proliferation in the case of SG-PSS and SG-DS was found to be much lower. To further investigate the toxicity of the coatings, the morphology of the cells was investigated by SEM.

2.8. Amperometric measurements To perform amperometric measurements, the glucose sensor was placed as the working electrode in a threeelectrode ¯ow cell. The metal-coated rear of the membrane was attached to a glassy carbon disk (diameter ˆ 8 mm). To insulate the active surface of the membrane from the counter electrode, the sensor was covered with a Te¯on spacer of 1 mm thickness. In the spacer a duct of 0.15 cm2 was left, allowing the membrane to make contact with the solution. An Ag/AgCl electrode was used as a reference. The base of the ¯ow cell acted as the counter electrode. Buffer solution was driven through the ¯ow cell (1.75 ml/min) with a Watson Marlow peristaltic pump. The potential of the membrane was set on 300 mV and was applied continuously. When the background current had diminished (<100 nA), the buffer solution was replaced by a solution containing glucose and the current was monitored. For the measurements performed in bovine serum albumin solutions, 40 g/l of this protein was added to PBS solution (0.1 M, pH 7.4) and glucose was added to the solution after the background current stabilized. For the measurements in serum, glucose was added after the current stabilized. Exact glucose concentrations were determined externally by an APEC Table 1 Physical properties of sol±gel coatings Entry

SG SG-HEP SG-DS SG-NAF SG-PEG SG-PSS a

Additivea

Heparine (5)d Dextran sulfate (5)e Nafion (2.5)f Polyethylene glycol (10)g Polystyrene sulfonate (2.5)h

Contact angle (y)b Before RFGD

After RFGD

36.8 6.9 3.0 31.0 5.1 25.9

4.5 3.2 2.7 3.3 2.4 3.0

     

4.58 6.38 1.28 6.28 3.38 4.68

The weight percentage of additive in the sol±gel is given within brackets. Average and standard deviation of 10 measurements, RFGD ˆ radiofrequency glow discharge treatment. c Determined with XPS. d 20 mg of heparin was added to 320 ml of water and 80 ml of the sol. e 20 mg of dextran sulfate was added to 320 ml of water and 80 ml of the sol. f 500 ml of nafion was added to 500 ml of the sol. g 100 mg of polyethylene glycol was added to 500 ml of water and 500 ml of the sol. h 10 mg of polystyrene sulfonate was added to 320 ml of water and 80 ml of the sol. b

Ratio (S:Si)c

     

1.48 1.68 0.88 1.28 1.28 2.48

0.356 0.238 0.597

A. Kros et al. / Sensors and Actuators B 81 (2001) 68±75

71

Fig. 1. (A) Cell proliferation of human dermal fibroblasts on different sol±gel surfaces (mean  S:D:). (B) Scanning electron micrograph of a sol±gel heparin coated surface after 8 days of incubation. Bar represents 10 mm.

Evaluation showed that sol±gel coatings with SG-HEP, SGNAF and SG-PEG were covered with a uniform layer of human dermal ®broblasts after 8 days of cell growth. Fig. 1B shows a typical example. Electron micrographs of the assays containing SG-PSS or SG-DS con®rmed the low proliferation rate of the ®broblasts on these surfaces. Nevertheless, the morphology of the cells was normal. XPS was used to investigate the amount of sulfate groups present at the surface of the sol±gel coatings (Table 1).2 It was found that the S:Si ratio at the surface of SG-DS is smaller than that of SG-HEP. Furthermore, SG-PSS appears to have more sulfate 2

Measured for sg-hep, sg-pss, sg-ds and sg-naf.

groups at its surface than the other coatings. It is known that the cellular response of a coating is controlled by small reactive areas rather than by the non-reactive bulk material at the surface [45]. Based on these data, we tentatively propose that the differences in cell growth rates are related to the differences in the amount and the distribution of free sulfonate groups on the surfaces of the composite ®lms. Apparently, a large number of these groups decreases the amount of protein that is adsorbed, thereby lowering the rate of cell proliferation. The results presented above showed that all coatings studied are non-toxic for ®broblasts. In a subsequent series of experiments, these sol±gel ®lms were tested as coatings

72

A. Kros et al. / Sensors and Actuators B 81 (2001) 68±75

Fig. 2. (A) Scanning electron micrograph of a polypyrrole track-etch membrane, bar represents 10 mm. (B) Scanning electron micrograph of a polypyrrole track-etch membrane after coating with SG-HEP, bar represents 10 mm. (C) Scanning electron micrograph of a cross-section of a polypyrrole track-etch membrane, bar represents 1 mm. (D) Scanning electron micrograph of a cross-section of a polypyrrole track-etch membrane after coating with SG-HEP, bar represents 10 mm. The black lines in (C) and (D) mark the pore diameter, which are ca. 0.9 mm in (C) and ca. 0.5 mm in (D).

on an amperometric glucose sensor. This sensor contained glucose oxidase immobilized inside polypyrrole microtubuli [4,5]. The conducting polypyrrole microtubuli were prepared by diaphragmatic oxidative polymerization of pyrrole inside the pores of a track-etch membrane (Fig. 2A). The backside of this membrane was covered with a thin layer of platinum. Immobilization of glucose oxidase inside the polypyrrole-coated pores of this membrane was achieved by soaking the membrane in an aqueous solution of the enzyme followed by drying overnight. The glucose sensor thus obtained was covered with the sol±gels by spincoating. Energy dispersive spectrometry using a SEM (SEM±EDS) was applied to con®rm whether the sensors had been effec-

tively coated with a layer of the sol±gels. Not all pores at the surface were covered with the sol±gel (Fig. 2B). A crosssection of the membrane was made to investigate the coating inside the pores. The diameter of the untreated polypyrrole microtubuli was approximately 0.9 mm (Fig. 2C). After application of the sol±gel layer, this diameter decreased to 0.5 mm due to penetration of the sol into the interior of the microtubuli (Fig. 2D). No differences in coverage of the membranes were observed for any of the sol±gel coatings used and all sensors were completely coated with the gels used in this study. Amperometric measurements of glucose in PBS (pH 7.4) were subsequently performed in a ¯owcell at an applied potential of 300 mV versus Ag/AgCl (Fig. 3).

Fig. 3. (A) Schematic representation of the track-etch membrane biosensor. (B) Layout of the flow cell used in the biosensor activity measurements.

A. Kros et al. / Sensors and Actuators B 81 (2001) 68±75

73

Fig. 4. (A) Plot of the steady-state current of a sol±gel dextran sulfate coated glucose sensor measured at 0.300 V (vs. Ag/AgCl) as a function of the glucose concentration. Experiments were carried out in aqueous PBS of pH 7.3 under an ambient atmosphere. (B) Plots of the glucose responses of the uncoated and sol±gel coated sensors relative to glucose responses in phosphate buffered solutions (100%). Experiments carried out in albumin solutions (40 g/l, white bars) and in serum (black bars) (n ˆ 6, mean  S:D:).

Calibration of the coated sensors revealed responses for glucose of ca. 40 nA/mM. As an example, a typical calibration curve of a sensor coated with SG-DS is shown in Fig. 4A. The response time of the sol±gel coated sensors, determined as rise to 95% response after the addition of glucose, varied between 2 and 4 min. Continuous measurements showed that all sensors were stable over a period of at least 2 weeks without any signi®cant loss of sensitivity. Since 80% of the total amount of protein in serum or tissue ¯uid is albumin, ®rst the in¯uence of this protein on the glucose response was investigated. Measurements performed in bovine serum albumin solutions (40 g/l) showed stable responses for all sensors (Fig. 4B). No drift was observed during the period of glucose measurements,

typically 2 h. The SG-DS gave the highest response of all coatings, possibly because its free sulfate groups retard the adsorption of proteins most ef®ciently, see Fig. 4B. The results in this ®gure are in agreement with the cell proliferation experiments. To further investigate the properties of the sol±gel coatings, glucose was also measured in serum. In this medium the responses of the uncoated sensors dropped dramatically and the signal became unstable. A loss in glucose sensitivity of typically 20± 35% after 1 h of exposure was observed for the uncoated sensors. The SG-PSS coated sensor appeared to have the lowest sensitivity for glucose comparable to that of the uncoated sensor, although the signal remained stable, unlike the latter one. The reason for this is yet unknown.

74

A. Kros et al. / Sensors and Actuators B 81 (2001) 68±75

Again the sol±gel coating with SG-DS gave the highest response to glucose. 4. Conclusions In summary, we have shown that TEOS-based sol±gel materials can be prepared easily and the cell proliferation rate can be tuned by the incorporation of polymers. It is proposed that the amount and distribution of free sulfonate groups at the surface of these coatings controls the protein adsorption and thereby the rate of cell proliferation. Furthermore, the cell morphology experiment shows that none of the coatings are toxic. The present materials are easily applicable on a glucose sensor leading to a stable and higher glucose response in serum, when compared to an untreated glucose sensor. The sol±gel containing SG-DS appears to be the most promising coating for an implantable glucose sensor. This hybrid material showed to be non-toxic and also slows down the growth of ®broblasts on its surface, which is important since a thick capsule could inhibit the diffusion of glucose to an implanted glucose sensor. The glucose measurements also showed that the sensors coated with SG-DS containing sol±gel had the highest response and stability to glucose. In the future, glucose measurements in vivo will be carried out to evaluate the real merits of these types of hybrid materials as biocompatible coatings for glucose sensors. Acknowledgements This research was supported by the Technology Foundation STW, which is the Applied Science Division of The Netherlands Science Foundation NWO and the Technology Programme of the Ministry of Economic Affairs. We thank Prof. Dr. J.W. Niemantsverdriet and Dr. P.C. Thuene (Eindhoven University of Technology) for the XPS measurements. References [1] B. Eggins, Biosensors: An Introduction, Wiley, Chichester, 1996. [2] A.E.G. Cammann, U. Lemke, A. Rohen, J. Sander, H. Wilken, B Winter, Chemo- und biosensoren Ð Grundlagen und Anwendungen, Angew. Chem. 103 (1991) 519±541. [3] Diabetes control and complications trial research group, The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus, New Engl. J. Med. 329 (1993) 977±986. [4] C.G.J. Koopal, B. de Ruiter, R.J.M. Nolte, Amperometric biosensor based on direct communication between glucose oxidase and a conducting polymer inside the pores of a filtration membrane, J. Chem. Soc., Chem. Commun. (1991) 1691±1692. [5] C.G.J. Koopal, M.C. Feiters, B. de Ruiter, R.B.M. Schasfoort, R.J.M. Nolte, Glucose sensor utilizing polypyrrole incorporated in track-etch membranes as the mediator, Biosens. Bioelectron. 27 (1992) 461±471.

[6] A. Kros, PhD thesis, University of Wijegan (2000). [7] D.A. Gough, J.Y. Lucisano, P.H.S. Tse, Two-dimensional enzyme electrode sensor for glucose, Anal. Chem. 57 (1985) 2351±2357. [8] C.A.P. Quinn, R.E. Connor, A. Heller, Biocompatible, glucose permeable hydrogel for in situ coating of implantable biosensors, Biomaterials 18 (1997) 1665±1670. [9] T. Vladkova, N. Krasteva, A. Kostadinova, G.J. Altankov, Preparation of PEG-coated surfaces and a study for their interaction with living cells, J. Biomater. Sci. Polym. Edn. 10 (1999) 609±620. [10] S.M. Reddy, P.M. Vadgama, Surfactant modified poly(vinyl chloride) membranes as biocompatible interfaces for amperometric enzyme electrodes, Anal. Chim. Acta 350 (1997) 77±89. [11] K. Ihihara, H. Hanyuda, N. Nakabayashi, Synthesis of phospholipid polymers having a urethane bond in the side chain as coating material on segmented polyurethane and their platelet adhesion resistant properties, Biomaterials 16 (1995) 873±879. [12] R.F.B. Turner, D.J. Harrison, R.V. Rajotte, Preliminary in vivo biocompatibility studies on persulphonic acid polymer membranes for biosensor applications, Biomaterials 12 (1991) 361±368. [13] R.F.B. Turner, D.J. Harrison, R.V. Rajotte, H.P. Baltes, A biocompatible enzyme electrode for continuous in vivo glucose monitoring in whole blood, Sens. Actuators B 1 (1990) 561±564. [14] I. Galeska, D. Chattopadhyay, F. Moussy, F. Papadimitrakopoulos, Calcification-resistant nafion/Fe3‡ assemblies for implantable biosensors, Biomacromolecules 1 (2000) 202±207. [15] M. Gerritsen, A. Kros, J.A. Lutterman, R.J.M. Nolte, J.A. Jansen, Performance of subcutaneously implanted glucose sensors: a review , J. Invest. Surg. 11 (1998) 163±175. [16] J.D. Wright, N.A.J.M. Sommerdijk, Sol±gel Materials, Chemistry and Applications, Gordon and Breach, Amsterdam, London, 2001. [17] L.L. Hench, J.K. West, The sol±gel process, Chem. Rev. 90 (1990) 33±72. [18] D. Avnir, Organic chemistry within ceramic matrices: doped sol±gel materials, Acc. Chem. Rev. 28 (1995) 328±334. [19] B.M. Novak, Hybrid nanocomposite materials: between inorganic glasses and organic polymers, Adv. Mater. 5 (1993) 422±433. [20] J.Y. Wen, G.L. Wilkes, Organic/inorganic hybrid network materials by the sol±gel approach, Chem. Mater. 8 (1996) 1667±1681. [21] R.J.P. Corriu, D. Leqlercq, Recent developments of molecular chemistry for sol±gel processes, Angew. Chem. Int. Ed. Engl. 108 (1996) 1524±1540. [22] S. Mann, S.L. Burkett, S.A. Davis, C.E. Fowler, N.H. Mendelson, S.D. Sims, D. Walsh, N.T. Whilton, Sol±gel synthesis of organized matter, Chem. Mater. 9 (1997) 2300±2310. [23] J. Livage, Sol±gel processes, Curr. Opinion Solid State Mater. Sci. 2 (1997) 132±138. [24] L.L. Hench, Sol±gel materials for bioceramic applications, Curr. Opinion Solid State Mater. Sci. 2 (1997) 604±610. [25] L.M. Ellerby, C.R. Nishida, F. Nishida, S.A. Yamanaka, B. Dunn, J.S. Valentine, J.I. Zink, Encapsulation of proteins in transparent porous silicate glasses prepared by the sol±gel method, Science 255 (1992) 1113±1115. [26] J. Li, S.N. Tan, H. Ge, Silica sol±gel immobilized amperometric biosensor for hydrogen peroxide, Anal. Chim. Acta 335 (1996) 137±145. [27] J.D. Jordan, R.A. Dunbar, F.V. Bright, Aerosol-generated sol±gelderived thin films as biosensing platforms, Anal. Chim. Acta 332 (1996) 83±91. [28] S. Sampath, O. Lev, Inert metal-modified, composite ceramic-carbon, amperometric biosensor: renewable, controlled reactive layer, Anal. Chem. 68 (1996) 2015±2021. [29] B. Avnir, S. Braun, O. Lev, M. Ottolenghi, Enzymes and other proteins entrapped in sol±gel materials, Chem. Mater. 6 (1994) 1605±1614. [30] Q. Chen, L. Kenausis, A. Heller, Stability of oxidases immobilized in silica gels, J. Am. Chem. Soc. 120 (1998) 4582±4585.

A. Kros et al. / Sensors and Actuators B 81 (2001) 68±75 [31] E.H. Lan, B. Dunn, J.I. Zink, Sol±gel encapsulated anti-trinitrotoluene antibodies in immunoassays for TNT, Chem. Mater. 12 (2000) 1874±1878. [32] I. Gill, A. Ballesteros, Encapsulation of biologicals within silicate, siloxane and hybrid sol±gel polymers: an efficient and generic approach, J. Am. Chem. Soc. 120 (1998) 8587±8598. [33] P. Kortesuo, M. Ahola, S. Karlsson, I. Kangasniemi, A. Yli-Urpo, J. Kiesvaara, Silica xerogels as an implantable carrier for controlled drug delivery-evaluation of drug distribution and tissue effects after implantation, Biomaterials 21 (2000) 193±198. [34] P. Kortesuo, M. Ahola, M. Kangas, I. Kangasniemi, A. Yli-Urpo, J. Kiesvaara, In vitro evaluation of sol±gel processed spray dried silica gel microspheres as carrier in controlled drug delivery, Int. J. Pharm. 200 (2000) 223±229. [35] P. Kortesuo, M. Ahola, S. Karlsson, I. Kangasniemi, J. Kiesvaara, A. Yli-Urpo, Sol±gel-processed sintered silica xerogel as a carrier in controlled drug delivery, J. Biomed. Mater. Res. 44 (1999) 162±167. [36] E.M. Santos, S. Radin, P. Ducheyne, Sol±gel derived carrier for the controlled release of proteins, Biomaterials 20 (1999) 1695±1700. [37] K.P. Peterson, C.M. Peterson, E.J.A. Pope, Silica sol±gel encapsulation of pancreatic islets, Proc. Soc. Exp. Biol. Med. 218 (1998) 365± 369. [38] E.J.A. Pope, C.M.J. Peterson, K. Braun, Bioartificial organs. Part I. Silica gel encapsulated pancreatic islets for the treatment of diabetes mellitus, J. Sol±gel Sci. Technol. 8 (1997) 635. [39] D. Tian, P. Dubois, C. Grandfils, R. JeÂroÃme, P. Viville, R. Lazzaroni, J.-L. BreÂdas, P. Leprince, A novel biodegradable and biocomatible ceramer prepared by the sol±gel process, Chem. Mater. 9 (1997) 871±874. [40] B. Tian, P.H. Dubois, R. JeÂroÃme, Biodegradable and biocomatible inorganic±organic hybrid materials. Part I. Synthesis and characterization, Polymer 37 (1996) 3983±3987. [41] U. Narang, P.N. Prasad, F.V. Bright, K. Ramanthan, N.D. Kumar, B.D. Malhotra, M.N. Kamalasan, S. Chandra, Glucose sensor based on a sol±gel-derived platform, Anal. Chem. 66 (1994) 3139±3144. [42] M. Gerritsen, A. Kros, V.S.I. Sprakel, J.A. Lutterman, R.J.M. Nolte, J.A. Jansen, Biocompatibility evaluation of sol±gel coatings for subcutaneously implantable glucose sensors, Biomaterials 21 (2000) 71±78. [43] H.P. de Jong, A.W.J. van Pelt, J. Arends, Contact angle measurements on human enamel: an in vitro study of influence of pellicle and storage period, J. Dent. Res. 61 (1982) 11±13.

75

[44] R.M. Hesby, C.R. Haganman, C.M. Stanford, Effects of radiofrequency glow discharge on impression material surface wettability, J. Prosthet. Dent. 77 (1997) 414±422. [45] A.S.J. Hoffman, A general classification scheme for hydrophilic and hydrophobic biomaterial surfaces, J. Biomed. Mater. Res. 20 (1986) 9±11.

Biographies Alexander Kros studied chemistry at the University of Nijmegen and received his PhD in December 2000. Currently, he is a postdoctoral scholar in the group of professor Tirrell at Caltech, Pasadena, USA. His research interests are in the field of biosensors and conducting polymers. Martijn Gerritsen studied medicine at the University of Nijmegen and obtained his PhD in March 2000 in the group of professor J.A. Jansen. Currently, he is working in the TweeSteden Hospital in Tilburg from where he received his General Internal Medicine training. Vera S.I. Sprakel studied chemistry at the University of Nijmegen and is currently a PhD student in the group of professor R.J.M. Nolte. Her research interests are enzyme mimmicks and molecular receptors for catalysis. John A. Jansen studied dentistry at the University of Nijmegen and received his PhD in medical sciences in 1984. He currently is professor of biomaterials at the University of Nijmegen. His interests are in the field of wound healing phenomena around implants, in particular the preparation and biological behavior of percutaneous calcium phosphate ceramics. Roeland J.M. Nolte studied chemistry at the University of Utrecht and received his PhD in physical organic chemistry in 1973. He currently is professor of organic chemistry at the University of Nijmegen. His interests are in the field of supramolecular, macromolecular and physical organic chemistry, in particular the design and construction novel supramolecular nano-sized architectures. Nico A.J.M. Sommerdijk studied chemistry at the University of Nijmegen and received his PhD in physical organic and supramolecular chemistry in 1995. Currently, he is a university lecturer in biomimetic materials chemistry at the Eindhoven University of Technology. His research interests are the fabrication of organic±inorganic composite materials and the construction of polymer-enzyme hybrids using the assembly of molecular and macromolecular structures.