Diffusion-Free Mediator Based Miniature Biofuel Cell Anode

Sep 5, 2012 - Carbon-MEMS based electrodes offer precise manufacturing control on ... Rapid and Automated Measurement of Milk Adulteration Using a 3D ...
1 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/Langmuir

Diffusion-Free Mediator Based Miniature Biofuel Cell Anode Fabricated on a Carbon-MEMS Electrode Gobind S. Bisht,† Sunny Holmberg,‡ Lawrence Kulinsky,§ and Marc Madou*,†,§,∥ †

Department of Biomedical Engineering, ‡Department of Electrical Engineering & Computer Science, and §Department of Mechanical and Aerospace Engineering, University of California, Irvine, California 92617, United States ∥ Ulsan National Institute of Science and Technology (UNIST), Banyeon-ri 100, Ulsan 689-798, Korea S Supporting Information *

ABSTRACT: We report on the functionalization of a micropatterned carbon electrode fabricated using the carbonMEMS process for its use as a miniature diffusion-free glucose oxidase anode. Carbon-MEMS based electrodes offer precise manufacturing control on both the micro- and nanoscale and possess higher electron conductivity than redox hydrogels. However, the process involves pyrolysis in a reducing environment that renders the electrode surface less reactive and introduction of a high density of functional groups becomes challenging. Our functionalization strategy involves the electrochemical oxidation of amine linkers onto the electrode. This strategy works well with both aliphatic and aryl linkers and uses stable compounds. The anode is designed to operate through mediated electron transfer between 2,5-dihydroxybenzaldehyde (DHB) based redox mediator and glucose oxidase enzyme. The electrode was first functionalized with ethylene diamine (EDA) to serve as a linker for the redox mediator. The redox mediator was then grafted through reductive amination, and attachment was confirmed through cyclic voltammetry. The enzyme immobilization was carried out through either adsorption or attachment, and their efficiency was compared. For enzyme attachment, the DHB attached electrode was functionalized again through electro-oxidation of aminobenzoic acid (ABA) linker. The ABA functionalization resulted in reduction of the DHB redox current, perhaps due to increased steric hindrance on the electrode surface, but the mediator function was preserved. Enzyme attachment was then carried out through a coupling reaction between the free carboxyl group on the ABA linker and the amine side chains on the enzyme. The enzyme incubation for both adsorption and attachment was done either through a dry spotting method or wet spotting method. The dry spotting method calls for the evaporation of enzyme droplet to form a thin film before sealing the electrode environment, to increase the effective concentration of the enzyme on the electrode surface during incubation. The electrodes were finally protected with a gelatin based hydrogel film. The anode half-cell was tested using cyclic voltammetry in deoxygenated phosphate buffer saline solution pH 7.4 to minimize oxygen interference and to simulate the pH environment of the body. The electrodes that yielded the highest anodic current were prepared by enzyme attachment method with dry spotting incubation. A polarization response was generated for this anodic half-cell and exhibits operation close to maximum efficiency that is limited by the mass transport of glucose to the electrode.

■

INTRODUCTION

integrated into implantable biomedical microdevices, overcoming the stated limitations of conventional power sources. In this paper, we address an important challenge in the fabrication of implantable enzymatic biofuel cells by proposing a strategy that achieves high density of amine and carboxyl functional groups on pyrolytic carbon electrodes for immobilization of enzymes and mediators. This strategy is then applied to build an anodic glucose oxidase (GOx)-based biofuel halfcell operating through a mediated electron transfer mechanism. Several strategies for the immobilization of mediators and

Recent biomedical advances have led to a significant increase in the development of a number of implantable biomedical devices. Such implantable devices are proving to be extremely effective and in some cases are the only means to managing life threatening conditions such as epilepsy,1 neural/spinal injuries,2,3 and cardiac arrhythmia.4,5 Currently, chemical batteries available to operate such devices require an independent power supply that needs to be hermetically packaged within the device.6 This adds substantial bulk to the device. The operational lifetime of the device is limited by the battery capacity.7 Enzymatic miniature biofuel cells utilize the biochemical resources within the human body; they can be © 2012 American Chemical Society

Received: July 5, 2012 Revised: September 3, 2012 Published: September 5, 2012 14055

dx.doi.org/10.1021/la302708h | Langmuir 2012, 28, 14055−14064

Langmuir

Article

maximize functionalization yield. The electro-oxidation based grafting reaction proceeds through the oxidation of the amine group in an amine linker that is driven by an anodic potential applied at the electrode as shown in Figure 1. Our

enzymes have been considered to optimize the performance of the anodic half-cell. Similarly, immobilization of laccase from Trametes versicolor can be carried out on an anthracene functionalized surface to form a high efficiency cathodic halfcell,8,9 which can complement the proposed glucose oxidase based anodic half-cell and form a complete enzymatic biofuel cell. Pyrolytic carbon is chosen as an electrode material because it has a large window of electrochemical stability10 and good biocompatibility.11 It is also available in a variety of physical and chemical forms that can be engineered to offer desired conductivity and structural strength.12 Several manufacturing techniques have already been developed to produce pyrolytic carbon electrodes with precise dimensional control on both the micro- and nanoscale. Carbon-MEMS (C-MEMS) is one of the technologies that has been used to fabricate the electrodes discussed in this paper, and it allows direct two- and threedimensional (2D and 3D) micromanufacturing of carbon electrodes.13 This is achieved by formation of the initial micropatterned structure through photolithography of SU8 (a photosensitive epoxy) in single or multiple layers and by subsequent pyrolysis of the processed SU8 structure in an inert environment10 (Ar or N2) at 900 °C. The surface of such carbon electrodes can be further enhanced using additional micro/nanostructuring techniques (such as electrospinning), thus adding the advantage of a high reaction interface area per unit of footprint and increasing the net electron exchange density.14,15 Carbon nanostructures with fractal architectures can also be fabricated on these electrodes with sol−gel polycondensation of resorcinol/with formaldehyde gels (RF gels).16 Such fractal-like electrode architectures can minimize internal resistance while maximizing the surface-to-volume ratios.17 The ability to manufacture high electrochemical efficiency carbon electrodes combined with the higher electron conductivity of pyrolytic carbon as compared to competing redox hydrogels18,19 makes CMEMS electrodes a promising candidate for enzymatic biofuel cell applications. However, the electrode surface obtained in the CMEMS process is rendered less reactive during pyrolysis in a reducing environment, and conventional surface functionalization methods such as reactive plasma treatment20−23 and microwave activation24 do not produce a high enough density of functional groups. More vigorous functionalization methods such as nitric acid oxidation25,26 damage the electrodes, causing their disintegration. Electrochemical strategies, such as reduction of diazonium salts, have been used successfully in the past to generate specific chemical functionalities on pyrolytic carbon electrodes.27,28 However, the molecules that can be used with the diazonium functionalization technique are limited to mainly aryl compounds (nitrobenzene being a classic example) due to the unstable nature of most aliphatic diazonium salts.29 Aryl molecules are rigid and bulky and thus not highly preferred for the role of linkers in immobilization of redox molecules for mediated electron transfer. Aliphatic linkers on the other hand are flexible and can thus increase the redox interaction of the mediator with the electrode by increasing its mobility. Such a flexibility of linker has been found important in in previous studies.30 We therefore choose to use a more versatile and effective functionalization technique known as electro-oxidation for this study. This technique allows for direct grafting of both aliphatic and aryl linkers to carbon electrodes.31−33 Since, the amine compounds are stable at room temperature, the grafting reaction can be carried out slowly over a longer duration to

Figure 1. Mechanism for electro-oxidation of amines onto an electrode surface. Reprinted with permission from ref 31. Copyright 2004 American Chemical Society.

functionalization strategy in this study is based on the use of ethylene diamine (EDA) as a flexible linker for the mediator molecule31 and aminobenzoic acid (ABA)33 as a scaffold for the enzyme molecule. The reaction is carried out in an anhydrous medium to increase functionalization yield and prevent autooxidation of the carbon electrode surface. We use 2,5-dihydroxybenzaldehyde (DHB) as the mediator molecule due to its lower equilibrium redox potential compared to other mediator molecules such as ferrocene.34 This minimizes the potential loss during the shuttling of electrons by the mediator molecule between the enzyme and the electrode.30 The DHB mediator molecules are covalently immobilized onto the electrode surface using EDA linkers, minimizing problems of leaching during operation in a fluid environment. We then analyze the electron transfer characteristics of the attached mediator molecules. Two approaches for enzyme immobilization are tested: the first one involves physical adsorption followed by hydrogel entrapment of the enzyme, while the second approach utilizes covalent attachment using ABA linkers followed by hydrogel coating with entrapped enzyme, as shown in Figure 2. The hydrogel coating is used in both cases to protect the enzyme- and mediator-loaded electrode surface against denaturation, by providing a stable hydrated environment. The enzyme incubation was also carried out via two different strategies. In one strategy, called wet spotting incubation, the electrode was spotted with a droplet of the enzyme solution to cover the electrode area completely and then sealed in a humid environment, to prevent the droplet from drying. In the second strategy, called dry spotting incubation, the enzyme droplet was allowed to dry slowly over the course of 24 h (however, preventing complete drying that leads to flaking of the enzyme layer). The latter scheme is believed to allow for a gradual increase in the enzyme concentration as the droplet slowly dries on the functionalized electrode surface, leading to better surface reaction kinetics. The overall methodology of the experiments performed in this study is presented in Figure 3.

■

EXPERIMENTAL SECTION

Materials. 4-Aminobenzoic acid (ABA), ethylene diamine (EDA), 2,5-dihydroxybenzaldehyde (DHB), sodium cyanoborohydride (NaCNBH3), glucose oxidase from Aspergillus niger (GOx), glucose, gelatin, bovine serum albumin (BSA), glutaraldehyde (25%), and anhydrous ethanol were purchased from Sigma Aldrich Co. (St. Louis, MO). Nonaqueous reference electrode was made with the kit MF14056

dx.doi.org/10.1021/la302708h | Langmuir 2012, 28, 14055−14064

Langmuir

Article

Figure 2. Schematic of the electron transfer mechanism for the two anode designs: (a) Only mediator molecules are attached to the electrode surface through EDA linkers; enzyme immobilization is mostly due to surface adsorption and hydrogel entrapment. (b) Both mediator molecules and enzymes are attached to the electrode surface through EDA linkers and ABA linkers respectively; a hydrogel layer is applied for enzyme stabilization.

Figure 3. Stepwise block diagram for the functionalization of SU8 photo patterned carbon electrodes for fabrication of dihydroxybenzaldehyde mediated glucose oxidase anode. and 2.2 °C/min between 700 and 900 °C, and a nitrogen flow rate of 2200 sccm. The electrode geometry used is a square pad of 12.5 mm2 area which is connected to another square of same area through a 1 mm wide carbon strip that is used as contact pad. After contacting a wire to the electrode with indium as solder to minimize contact resistance, the whole electrode was coated with the Hysol epoxy, leaving only the reaction area exposed for functionalization. The electrodes’ reaction areas were cleaned before the functionalization process by wiping them with acetone, isopropanol, ethanol, and DI water to remove any surface contamination. The electrodes were stored in a closed container prior to the functionalization steps. Surface Functionalization of EDA Linkers for Mediator Attachment. The electrodes were conditioned in anhydrous ethanol for before starting the EDA electro-oxidation. The functionalization was carried out by cycling the voltage between 0.0 and 1.3 V (vs Ag/ Ag+ nonaqueous reference electrode) at a scan rate of 10 mV/s for 3 cycles, in anhydrous ethanol with 50 mM EDA and 100 mM LiClO4

2062 purchased from BASi Inc., wherein a silver wire was immersed in an acetonitrile solution containing 0.01 M AgNO3 and stored for at least 24 h to equilibrate the frit. A photosensitive epoxy SU8 2035 used for the fabrication of 2D carbon microelectrodes was purchased from Microchem Inc. p-Doped [100] silicon test wafers of 100 mm diameter were used as substrates for the carbon electrodes and were purchased from Addison Engineering Inc. (San Jose, CA). Hysol 9462 epoxy used for the protection of electrode contacts was purchased from McMaster Carr Inc. Preparation of 2D Carbon Microelectrodes. Carbon microelectrodes were prepared by depositing a 35 μm thick film of SU8 2035 on the silicon wafer using a Laurell spin coater model WS-650 SZ-6NPP/A1/BP1. The SU8 film was soft baked at 95 °C for 7 min and exposed for 13 s at 12 mW power with an AB-Minc Deep Ultra Violet Flood Exposure system, post baked at 95 °C for 5 min, and developed in Microchem SU8 developer to yield the SU8 pattern of the electrodes. The SU8 pattern was then pyrolyzed at 900 °C for 2 h with a temperature ramp rate of 7.5 °C/min between 25 and 750 °C 14057

dx.doi.org/10.1021/la302708h | Langmuir 2012, 28, 14055−14064

Langmuir

Article

Figure 4. (Left) Functionalization of carbon electrode with ethylene diamine. The irreversible peak at 0.8 V corresponds to the EDA aminooxidation. (Right) Functionalization of carbon electrode through ABA amino-oxidation after attachment of EDA-DHB. The irreversible peak around 1.0 V is the amino-oxidation peak. (as supporting electrolyte). The use of anhydrous ethanol ensures that the high oxidizing potential does not cause electrolysis of water that can interfere with the reaction. The cyclic voltammogram for the electro-oxidation is shown in Figure 4 (left). A single irreversible peak signifies a unique electron transfer reaction, attributed to the oxidation and attachment of a single amine group on the EDA molecule. The current density was found to drop substantially from cycle 1 to cycle 3, indicating saturation of the surface with the EDA molecules as the functionalization proceeds through subsequent cycles. Mediator Grafting to the EDA Linker. After the functionalization of the electrode surface with EDA, the electrodes were rinsed with ethanol and DI water for 30 s each to remove any unattached EDA adsorbed to the surface. The conjugation of DHB was carried out by using reductive amination chemistry where the amine groups generated on the carbon electrode surface are coupled to the CHO group on 2,5-dihydroxybenzaldehyde (DHB) at pH 7.4 with NaCNBH3 as the reducing catalyst. The reaction was carried out by dipping the electrodes in a phosphate buffer saline (PBS) solution of 50 mM DHB and 10 mM NaCNBH3 adjusted to pH 7.0. The solution was stirred throughout the incubation, which lasted for 24 h, to allow sufficient time for the surface reaction to happen. After the incubation, the electrodes were rinsed with DI water and then washed with PBS on a Labline Orbit Shaker 3520 for 15 min to remove any unattached DHB from the electrode. Surface Functionalization of ABA Linkers for Enzyme Attachment. After the attachment of DHB to EDA linkers, the electrode surface was functionalized with aminobenzoic acid (ABA) to generate carboxyl groups that can be conjugated to enzyme molecules through amide bond formation between the carboxyl groups in ABA and the amine groups on the amino acid side chains of the enzyme (discussed in more details below). ABA was electrochemically attached by cyclic voltammetry in a 10 mM ABA solution with 100 mM LiClO4 as supporting electrolyte and anhydrous ethanol as the solvent. The voltage was cycled between 0.0 and 1.3 V (vs nonaqueous Ag/Ag+ reference electrode) at a scan rate of 10 mV/s for three cycles as shown in Figure 4 (right). The oxidation peak of ABA appeared between 0.8 and 1.0 V on the cyclic voltammogram. The peak current progressively decreased with each subsequent cycle perhaps due to steric hindrance from already deposited molecules. Enzyme Immobilization via Enzyme Adsorption and Hydrogel Coating. Enzyme immobilization was carried out by a two-step process. First, the electrodes were incubated with GOx at a concentration of 8 kU/mL dissolved in PBS (pH 7.4) for 24 h. This allows the adsorption of the enzyme molecules onto the electrode surface. The incubation was performed using either wet spotting or dry spotting techniques as described earlier. The second step of enzyme immobilization was hydrogel coating. A thin layer of hydrogel with entrapped GOx and stabilized with BSA provides protection against shearing and denaturation of the adsorbed enzyme molecules. It also increases the enzyme loading, as illustrated earlier in Figure 2. A

solution of 4 kU/mL GOx and 20 mg/mL BSA in PBS was added to 20% gelatin solution in equal ratios and thoroughly mixed to form the base hydrogel coating that was then cross-linked with a 0.25% glutaraldehyde solution. This ratio of gelatin matrix to glutaraldehyde was found to yield an optimum cross-linking density that enables effective GOx entrapment while maximizing glucose mass transport to the electrode surface. The hydrogel was applied as a thin layer by dipping electrodes multiple times into a freshly prepared hydrogel solution (after adding the glutaraldehyde). The hydrogel coated electrodes were then allowed to cross-link for 30 min followed by a rinse with PBS. Prior to testing the electrodes were stored in a PBS solution. Enzyme Immobilization via Enzyme Attachment and Hydrogel Coating. The enzyme attachment to the electrode was done by first functionalizing the DHB attached functionalized electrode with ABA as described earlier. The free carboxyl groups available on the ABA molecules were then activated using EDC/NHS chemistry by incubating them for 24 h in a pH 7.0 PBS solution containing 50 mM EDC and 10 mM sulfo-NHS. The pH for EDC/ NHS activation was kept at 7.0 to minimize the hydrolysis of the NHS activated carboxyl groups, which otherwise becomes a competing reaction at a basic pH due to the slow reaction kinetics of surface conjugation. The carboxyl activated electrode surface is then incubated for 24 h in a PBS (pH 7.4) solution containing GOx (8 kU/mL). Several lysine side chains of the enzyme are available for forming amide bonds with the activated carboxyl groups on ABA and can directly attach to the electrode surface. The second step of enzyme immobilization is coating of the electrode surface with a thin film of hydrogel to stabilize and protect the attached enzyme layer as well as to increase the enzyme loading. The hydrogel was prepared according to the protocol described above. Electrochemical Characterization of the Electrodes. All the electrodes were characterized using cyclic voltammetry in the voltage range between −0.3 and +0.6 V, vs Ag/AgCl reference electrode at room temperature. The testing was done in PBS (pH 7.4). The glucose solution was prepared at a concentration of 0.1 M in PBS and allowed to mutarotate at room temperature for 24 h to reach anomeric equilibrium, before using it to test the anodic response of the GOx immobilized electrodes. All solutions were deoxygenated before electrochemical testing by purging ultrapure nitrogen gas through the solution for at least 20 min and maintaining the solution in a sealed environment during testing. A platinum wire was used as the counter electrode. A Gamry potentiostat model PC4 was used for data acquisition. The polarization curve for a representative anodic electrode was generated via a chronoamperometry routine by obtaining the steady state current at different cathodic potentials. 14058

dx.doi.org/10.1021/la302708h | Langmuir 2012, 28, 14055−14064

Langmuir

■

Article

RESULTS AND DISCUSSION DHB Attachment. The attachment of DHB molecules to EDA linkers was tested through cyclic voltammetry in deoxygenated PBS. The EDA-DHB functionalized electrodes were rinsed with PBS to remove any adsorbed DHB molecules. An adsorption control was also prepared by incubating untreated electrodes with same DHB solution as the other EDA functionalized electrodes. The graph in Figure 5 shows a

reaction for DHB molecule, the surface coverage density was found to be around 0.202 × 10‑7 mol/cm2, approximately twice that of previously reported values.30 The electrodes with EDA-DHB functionalization were further investigated using CV at several different scan rates as shown in Figure 6. The peak current density was then calculated and plotted against the scan rate as shown in the top inset for Figure 6. A linear trend reveals a nondiffusive redox behavior, confirming that the redox peaks in the CV are due to the immobilization of the DHB molecules on the electrode surface. The electron transfer kinetics rate for the DHB attached to the electrode surface was calculated using Laviron’s analysis.35 The plot between the difference of anodic peak potential and the formal equilibrium potential (ΔEp) and ln(scan rate) shows a linear trend (bottom inset of Figure 6), indicating a diffusionless redox system. Assuming a one electron transfer reaction between the DHB and the electrode,34 the electron transfer rate (k) was calculated and found to be 0.105 s−1. This value is lower than that for the electrodeposited DHB reported in an earlier work.34 The difference between two values is likely attributable to the reduced interaction of DHB molecules with the electrode surface due to the rigidity of the connecting intermediate linker and the steric crowding of the electrode surface by the EDA and ABA linkers that are redox inactive. However, the linker molecules are instrumental in providing the necessary mobility to the DHB molecules in order to facilitate their interaction with the large enzyme molecules. This mobility is necessary for the DHB molecule to access the enzyme redox center, which is buried deep in the bulk of the large enzyme molecule, far from the electrode surface.36,37 Direct access to this redox center is difficult and although carbon nanotubes have been used to accomplish this,38,39 the current density can be limited due to contact resistance between the carbon nanotubes and the supporting matrix.40−42 DHB Attachment with Enzyme Attachment. The enzyme attachment was accomplished by generating carboxyl groups on the electrode surface through electro-oxidation of ABA as described previously. The electrochemical activity of the attached DHB molecules was tested again after the ABA deposition, to confirm that there was no detrimental effect on the attached DHB redox properties due to the ABA deposition

Figure 5. CV scan of EDA functionalized electrode (dotted), EDADHB functionalized electrode (dark solid line), and electrode with no functionalization but incubated in DHB (light solid line) in deoxygenated PBS at 100 mV/s.

comparison between three electrodes: two EDA functionalized electrodes with (dark solid line) and without (light solid line) DHB attachment and the adsorption control (dotted line). The EDA functionalized electrode shows a larger redox current compared to the adsorption control. The larger redox current reflects the higher quantity of DHB that gets immobilized on the electrodes through the covalent attachment chemistry discussed in the Experimental Section under the heading mediator grafting to the EDA linker. The surface coverage of DHB was calculated through the integration of the voltammogram for EDA-DHB functionalized electrode and its subtraction from the voltammogram for EDA functionalized blank electrode both shown in Figure 5. Assuming a two electron

Figure 6. CV curves at scan rates between 10 and 60 mV/s (arrow shows direction of increasing scan rate), showing the mediator redox peaks after the attachment of DHB. (Top inset) Plot of peak current density vs scan rate from CV curves confirms the attachment of DHB to EDA linker, evident in the linear trend of the plot. (Bottom inset) Plot for peak separation (ΔEp) vs log of scan rate used for calculation of the electron transfer rate (k) using Laviron’s method for EDA-linked DHB redox mediator. 14059

dx.doi.org/10.1021/la302708h | Langmuir 2012, 28, 14055−14064

Langmuir

Article

Figure 7. CV curves at scan rates between 10 and 50 mV/s, showing the mediator redox peaks before (a) and after (b) the deposition of ABA on EDA-DHB functionalized electrode surface. (c) Plot for anodic peak current density vs scan rate. Decrease in DHB redox activity in seen after (triangles) vs before (diamonds) ABA deposition. (d) Plot for peak separation (ΔEp) vs log of scan rate used for calculation of the electron transfer rate (k) using Laviron’s method.

scan was first performed without any glucose to register the background current and then in the presence of glucose. These curves are shown in Figure 8 for electrodes prepared with enzyme immobilization using wet spotting (a,b) and dry spotting (c,d) incubation methods. An increase in the anodic current density (the current response in the positive potential region) was observed in the presence of glucose for all the enzyme immobilized electrodes. No such increase was seen in case of electrodes without any immobilized DHB (not shown). The increase in anodic current in presence of glucose is direct evidence for the activity of immobilized glucose oxidase and its interaction with the mediator molecule. Figure 8 demonstrates that the relative anodic current (the difference in the anodic current in the presence of glucose and the background, normalized over background) at 10 mV/s is higher for electrodes prepared with the dry spotting than with the wet spotting incubation method. This is believed to be caused by the higher amount of enzyme immobilization resulting from a higher effective concentration of enzyme available during incubation. The relative anodic current, however, drops at a higher scan rate of 100 mV/s, for both wet and dry spotting incubation methods. This effect is perhaps due to the slow glucose mass transfer through the hydrogel layer, which leads to a lower effective concentration of glucose at the electrode surface during higher scan rates. At higher scan rates, the voltage increases faster than the diffusion rate causing the steady state glucose concentration at the electrode to drop to a lower level than what is reached at lower scan rates. The drop in the relative anodic current is higher for dry spotting incubation than wet spotting and can be explained by the

process. The results are shown in Figure 7. The redox peaks in the CV were preserved, but a reduction in the magnitude of the peak current was observed. This could be caused by the attack of the ABA radical cations on the EDA-DHB groups during the ABA amino oxidation leading to the partial degradation of DHB groups, reactions that are well-known in electroorganic chemistry.43 Also the increased steric crowding of the electrode surface caused by the bulky benzene rings of the ABA linkers may act to reduce the mediator interaction with the electrode surface. The peak current density vs the scan rate continued to show a linear trend (Figure 7, bottom left), suggesting a diffusion-free electron exchange mechanism by the attached DHB molecules. The electron transfer kinetics for the remaining DHB molecules after ABA deposition does not seem to have been affected. This is reflected in the unaltered slope of the ΔEp vs ln(scan rate) plot before and after the ABA deposition (Figure 7, bottom right) as well as the similar electron transfer rates (kbefore = 0.17, kafter = 0.24) calculated from these plots using Laviron’s analysis. Hence, the remaining DHB molecules can continue to effectively participate in mediated electron transfer after the enzyme attachment to the ABA linkers. The carboxyl groups were then activated using EDC/NHS chemistry and incubated with glucose oxidase to allow attachment of the enzyme molecules to the ABA molecules. Anode Half-Cell: Enzyme Immobilization with Hydrogel Only. The anode half-cell was tested using cyclic voltammetry as discussed in previous sections. The potential on the electrode was scanned using cyclic voltammetry between −0.3 and +0.5 V at 10 and 100 mV/s in PBS at pH 7.4. The 14060

dx.doi.org/10.1021/la302708h | Langmuir 2012, 28, 14055−14064

Langmuir

Article

Figure 8. Cyclic voltammogram for the EDA linked DHB electrode after the immobilization of glucose oxidase with wet spotting incubation (a,b) and dry spotting incubation (c,d) in presence (dark curve) and absence of glucose (light curve). Scan rate: 10 mV/s (left) and 100 mV/s (right).

Figure 9. Cyclic voltammogram for the EDA linked DHB electrode after second functionalization with ABA and immobilization of glucose oxidase with wet spotting incubation (a,b) and dry spotting incubation (c,d), in the presence (dark curve) and absence of glucose (light curve). Scan rate: 10 mV/s (left) and 100 mV/s (right).

14061

dx.doi.org/10.1021/la302708h | Langmuir 2012, 28, 14055−14064

Langmuir

Article

increase in enzyme loading achieved through this method. Consequently, higher enzyme loading leads to an even faster glucose oxidation rate and therefore a stronger depletion effect on steady state glucose concentration on the electrode surface at higher scan rates. Anode Half-Cell: Enzyme Immobilization with Attachment and Hydrogel. The anode half-cell was tested using cyclic voltammetry as discussed in previous sections. The CV scans for both electrodes prepared with enzyme immobilization using wet spotting (a,b) and dry spotting (c,d) incubation methods are shown in Figure 9. The electrodes show an increase in the anodic current density in the presence of glucose, suggesting successful immobilization of redox active glucose oxidase and its favorable interaction with the immobilized mediator molecules. Similar to the previous set of electrodes discussed in the section on enzyme immobilization with hydrogel only, the relative anodic current in electrodes prepared with the dry spotting incubation method is observed to be higher here as well, compared to the electrodes prepared with the wet spotting. We also observe a higher relative anodic current for the electrodes prepared with enzyme attachment compared to the electrodes prepared with enzyme adsorption only. This could be a combined effect of (1) the increase in enzyme loading and (2) a more favorable mediator−enzyme interaction due to the increased proximity and ordering of the enzyme molecules in relation to the mediator molecules. Moreover, the relative anodic current drops at a higher scan rate (100 mV/s), as seen in Figure 9, and the effect is stronger in the electrodes prepared with dry spotting incubation method. This is consistent with the observation made in the previous electrodes discussed in the section on enzyme immobilization with hydrogel only. However, the overall magnitude of this effect is stronger in the electrodes prepared with enzyme attachment (utilizing the same incubation method). This is indeed a result of a higher enzyme loading achieved through enzyme attachment that causes an even faster depletion of glucose in the diffusion layer with a higher scan rate, compared to enzyme adsorption. Therefore, a more significant drop in the relative anodic current is observed when switching from a lower scan rate (10 mV/s) to a higher scan rate (100 mV/s). Among all the anodic electrodes discussed in this study, the electrodes prepared with carboxyl functionalization aided enzyme attachment combined with dry spotting incubation yields the highest increase in anodic current. These electrodes fabricated with immobilized enzyme-mediator interaction were further characterized via a polarization study. The polarization response was measured under steady state condition in a cathodic potential range of 0.0 to +0.5 V as shown in Figure 10. The curve shows an expected increasing trend with the cathodic potential that exhibits a slow saturation behavior at higher potentials similar to previous studies.38,44 This confirms the operation of the anodic cell in the mass transport limited regime also observed previously in the cyclic voltammograms of Figures 8 and 9. This suggests that the surface concentration of glucose oxidase and its interaction with the mediator molecules is high enough to catalyze mediated electron transfer at a rate comparable to the diffusive mass transfer rate of glucose to the electrode. The anode half-cell is therefore operational close to the maximum efficiency that is dictated by the rate limiting mass transport phenomenon of the glucose diffusion to the electrode.

Figure 10. Polarization curve for the glucose oxidase anode that contains EDA linked DHB redox mediator and ABA facilitated covalently immobilized glucose oxidase (incubated via dry spotting incubation method and coated with gelatin based hydrogel).

■

CONCLUSION We report on an efficient electrochemical functionalization strategy for micropatterned carbon electrodes fabricated via the CMEMS process. Sparing chemical functionalization strategies such as plasma treatment fail to deliver a high enough density of functional groups while harsh techniques such as acid oxidation damage the microstructured electrodes. Diazonium reduction approach on the other hand is only useful for use with aryl linkers that may not be highly flexible for effective mediator applications in biofuel cells. We have successfully achieved a high density of functionalization on CMEMS electrodes through electro-oxidation of ethylene diamine (EDA) and aminobenzoic acid (ABA). This functionalization strategy was employed to attach EDA as a linker molecule for the immobilization of DHB redox mediator to the electrode. The anodic half-cell was prepared by immobilization of GOx onto the DHB functionalized electrode, through either an adsorption or attachment strategy, wherein the enzyme incubation was carried out via a wet spotting or a dry spotting method. The dry spotting method is expected to allow for a gradual increase in the enzyme concentration during the incubation period leading to a better conjugation reaction kinetics and hence higher enzyme loading on the surface. All the electrodes were protected by a hydrogel coating, that not only stabilizes the enzyme against shearing and denaturation, but also increases the enzyme loading by providing a threedimensional scaffold for enzyme entrapment. The enzyme attachment was carried out through the utilization of ABA linkers as anchor molecules that were deposited onto the electrode after the DHB attachment step. The ABA deposition process was found to have no detrimental effects on the redox properties of the attached DHB mediator molecules as confirmed through cyclic voltammetry analysis. The carboxyl groups on the ABA linkers were then activated via EDC/NHS incubation at pH 6.0. The enzyme attachment was finally carried out through a coupling reaction between the amine group on the lysine side chains of the GOx molecule and the activated carboxyl groups on the ABA linkers at pH 7.4. The anode half-cells prepared through the above-stated methods were then tested through cyclic voltammetry, both in absence and in presence of glucose in a deoxygenated environment. Our results confirm the hypothesis that the dry spotting method allows for a higher efficiency of enzyme immobilization. The electrodes prepared through enzyme attachment combined with the dry spotting incubation method yielded the highest increase in anodic current. We believe that that this method allows for higher enzyme loading and a close proximity between enzyme and mediator molecules, thus 14062

dx.doi.org/10.1021/la302708h | Langmuir 2012, 28, 14055−14064

Langmuir

Article

(8) Blanford, C. F.; Heath, R. S.; Armstrong, F. A. A stable electrode for high-potential, electrocatalytic O2 reduction based on rational attachment of a blue copper oxidase to a graphite surface. Chem. Commun. 2007, 1710. (9) Thorum, M. S.; Anderson, C. A.; Hatch, J. J.; Campbell, A. S.; Marshall, N. M.; Zimmerman, S. C.; Lu, Y.; Gewirth, A. A. Direct, Electrocatalytic Oxygen Reduction by Laccase on Anthracene-2methanethiol-Modified Gold. J. Phys. Chem. Lett. 2010, 1, 2251. (10) Wang, C.; Madou, M. From MEMS to NEMS with carbon. Biosens. Bioelectron. 2005, 20, 2181. (11) Teixidor, G. T.; R. A. Gorkin, I.; Tripathi, P. P.; Bisht, G. S.; Kulkarni, M.; Maiti, T. K.; Battacharyya, T. K.; Subramaniam, J. R.; Ashutosh, S.; Park, B. Y.; Madou, M. Carbon microelectromechanical systems as a substratum for cell growth. Biomed. Mater. 2008, 3, 034116. (12) Singh, A.; Jayaram, J.; Madou, M.; Akbar, S. Pyrolysis of Negative Photoresists to Fabricate Carbon Structures for Microelectromechanical Systems and Electrochemical Applications. J. Electrochem. Soc. 2002, 149, E78. (13) Wang, C.; Jia, G.; Taherabadi, L.; Madou, M. A novel method for the fabrication of high-aspect ratio C-MEMS structures. J. Microelectromech. Syst. 2005, 14, 348. (14) Beidaghi, M.; Chen, W.; Wang, C. Electrochemically activated carbon micro-electrode arrays for electrochemical micro-capacitors. J. Power Sources 2011, 196, 2403. (15) Penmatsa, V.; Yang, J.-H.; Yu, Y.; Wang, C. Fabrication of porous carbon micropillars using a block copolymer as porogen. Carbon 2010, 48, 4109. (16) Sharma, C. S.; Kulkarni, M. M.; Sharma, A.; Madou, M. Synthesis of carbon xerogel particles and fractal-like structures. Chem. Eng. Sci. 2009, 64, 1536. (17) Park, B. Y.; Zaouk, R.; Wang, C.; Madou, M. J. A Case for Fractal Electrodes in Electrochemical Applications. J. Electrochem. Soc. 2007, 154, P1. (18) Calabrese Barton, S.; Gallaway, J.; Atanassov, P. Enzymatic Biofuel Cells for Implantable and Microscale Devices. Chem. Rev. 2004, 104, 4867. (19) Tamaki, T.; Yamaguchi, T. High-Surface-Area Three-Dimensional Biofuel Cell Electrode Using Redox-Polymer-Grafted Carbon. Ind. Eng. Chem. Res. 2006, 45, 3050. (20) Chen, C.; Liang, B.; Ogino, A.; Wang, X.; Nagatsu, M. Oxygen Functionalization of Multiwall Carbon Nanotubes by MicrowaveExcited Surface-Wave Plasma Treatment. J. Phys. Chem. C 2009, 113, 7659. (21) Pötschke, P.; Zschoerper, N. P.; Moller, B. P.; Vohrer, U. Plasma Functionalization of Multiwalled Carbon Nanotube Bucky Papers and the Effect on Properties of Melt-Mixed Composites with Polycarbonate. Macromol. Rapid Commun. 2009, 30, 1828. (22) Yook, J. Y.; Jun, J.; Kwak, S. Amino functionalization of carbon nanotube surfaces with NH3 plasma treatment. Appl. Surf. Sci. 2010, 256, 6941. (23) Coen, M. C.; Keller, B.; Groening, P.; Schlapbach, L. Functionalization of graphite, glassy carbon, and polymer surfaces with highly oxidized sulfur species by plasma treatments. J. Appl. Phys. 2002, 92, 5077. (24) Pandurangappa, M.; Ramakrishnappa, T. Microwave-assisted functionalization of glassy carbon spheres: electrochemical and mechanistic studies. J. Solid State Electrochem. 2010, 14, 687. (25) Lakshminarayanan, P. V.; Toghiani, H.; Pittman, C. U., Jr Nitric acid oxidation of vapor grown carbon nanofibers. Carbon 2004, 42, 2433. (26) Rosca, I. D.; Watari, F.; Uo, M.; Akasaka, T. Oxidation of multiwalled carbon nanotubes by nitric acid. Carbon 2005, 43, 3124. (27) Allongue, P.; Delamar, M.; Desbat, B.; Fagebaume, O.; Hitmi, R.; Pinson, J.; Savéant, J.-M. Covalent Modification of Carbon Surfaces by Aryl Radicals Generated from the Electrochemical Reduction of Diazonium Salts. J. Am. Chem. Soc. 1997, 119, 201. (28) Mahouche-Chergui, S.; Gam-Derouich, S.; Mangeney, C.; Chehimi, M. M. Aryl diazonium salts: a new class of coupling agents

enhancing the mediated electron transfer rate. Effects of mass transfer resistance to the glucose flow through the hydrogel layer were observed during the cyclic voltammetry study by the decrease in the relative anodic current as the scan rate was increased. The polarization curve for the anode half-cell reveals operation at a high catalytic efficiency that is limited by the mass transport of the glucose to the electrode.

■

ASSOCIATED CONTENT

S Supporting Information *

Schematic of the electrode, reductive amination reaction for DHB attachment and a molecular model of glucose oxidase revealing the amine terminated lysine groups. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Telephone: 949-824-6585. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors are grateful to Dr. Horacio Kido for useful discussions on surface chemistry and to Mr. Rahul Kamath for technical assistance and support; to Integrated Nanosystems Research Facility (INRF) at University of California, Irvine for the use of microfabrication equipment and resources. This research was supported by the National Science Foundation grant NIRT-0709085 and by UC Lab Fees Award 09-LR-09117362, Ignition Fund of the University of California Irvine and WCU (World Class University) program (R32-2008-00020054-0) through the National Research Foundation of Korea funded by the Ministry of Education, Science and Technology.

■

REFERENCES

(1) Halliday, A. J.; Campbell, T. E.; Razal, J. M.; McLean, K. J.; Nelson, T. S.; Cook, M. J.; Wallace, G. G. In vivo biocompatibility and in vitro characterization of poly-lactide-co-glycolide structures containing levetiracetam, for the treatment of epilepsy. J. Biomed. Mater. Res., Part A 2012, 100A, 424. (2) Kottink, A. I. R.; Tenniglo, M. J. B.; de Vries, W. H. K.; Hermens, H. J.; Buurke, J. H. Effects of an Implantable Two-Channel Peroneal Nerve Stimulator Versus Conventional Walking Device on Spatiotemporal Parameters and Kinematics of Hemiparetic Gait. J. Rehabil. Med. 2012, 44, 51. (3) Zhou, H.; Xu, Q.; He, J.; Ren, H.; Zhou, H.; Zheng, K. A fully implanted programmable stimulator based on wireless communication for epidural spinal cord stimulation in rats. J. Neurosci. Methods 2012, 204, 341. (4) Refaat, M. M.; Tanaka, T.; Kormos, R. L.; McNamara, D.; Teuteberg, J.; Winowich, S.; London, B.; Simon, M. A. Survival Benefit of Implantable Cardioverter-Defibrillators in Left Ventricular Assist Device−Supported Heart Failure Patients. J. Card. Failure 2012, 18, 6. (5) Otmani, A.; Trinquart, L.; Marijon, E.; Lavergne, T.; Waintraub, X.; Lepillier, A.; Chatellier, G.; Le Heuzey, J.-Y. Rates and predictors of appropriate implantable cardioverter-defibrillator therapy delivery: Results from the EVADEF cohort study. Am. Heart J. 2009, 158, 230. (6) Nagata, M.; Saraswat, A.; Nakahara, H.; Yumoto, H.; Skinlo, D. M.; Takeya, K.; Tsukamoto, H. Miniature pin-type lithium batteries for medical applications. J. Power Sources 2005, 146, 762. (7) Harvey, R. J.; Bankole, M. A.; Robinson, T. C.; Bernhard, W. F. Studies related to development of an implantable power source for circulatory assist devices. Am. J. Surg. 1967, 114, 61. 14063

dx.doi.org/10.1021/la302708h | Langmuir 2012, 28, 14055−14064

Langmuir

Article

for bonding polymers, biomacromolecules and nanoparticles to surfaces. Chem. Soc. Rev. 2011, 40, 4143. (29) Belanger, D.; Pinson, J. Electrografting: a powerful method for surface modification. Chem. Soc. Rev. 2011, 40, 3995. (30) Tamaki, T.; Ito, T.; Yamaguchi, T. Immobilization of Hydroquinone through a Spacer to Polymer Grafted on Carbon Black for a High-Surface-Area Biofuel Cell Electrode. J. Phys. Chem. B 2007, 111, 10312. (31) Adenier, A.; Chehimi, M. M.; Gallardo, I.; Pinson, J.; Vilà, N. Electrochemical Oxidation of Aliphatic Amines and Their Attachment to Carbon and Metal Surfaces. Langmuir 2004, 20, 8243. (32) Gallardo, I.; Pinson, J.; Vilà, N. Spontaneous Attachment of Amines to Carbon and Metallic Surfaces. J. Phys. Chem. B 2006, 110, 19521. (33) Liu, J.; Cheng, L.; Liu, B.; Dong, S. Covalent modification of a glassy carbon surface by 4-aminobenzoic acid and its application in fabrication of a polyoxometalates-consisting monolayer and multilayer films. Langmuir 2000, 16, 7471. (34) Pariente, F.; Tobalina, F.; Darder, M.; Lorenzo, E.; Abruña, H. D. Electrodeposition of Redox-Active Films of Dihydroxybenzaldehydes and Related Analogs and Their Electrocatalytic Activity toward NADH Oxidation. Anal. Chem. 1996, 68, 3135. (35) Laviron, E. General expression of the linear potential sweep voltammogram in the case of diffusionless electrochemical systems. J. Electroanal. Chem. 1979, 101, 19. (36) Wilson, R.; Turner, A. P. F. Glucose oxidase: an ideal enzyme. Biosens. Bioelectron. 1992, 7, 165. (37) Hecht, H. J.; Schomburg, D.; Kalisz, H.; Schmid, R. D. The 3D structure of glucose oxidase from Aspergillus niger. Implications for the use of GOD as a biosensor enzyme. Biosens. Bioelectron. 1993, 8, 197. (38) Ivnitski, D.; Branch, B.; Atanassov, P.; Apblett, C. Glucose oxidase anode for biofuel cell based on direct electron transfer. Electrochem. Commun. 2006, 8, 1204. (39) Cai, C.; Chen, J. Direct electron transfer of glucose oxidase promoted by carbon nanotubes. Anal. Biochem. 2004, 332, 75. (40) Tzeng, Y.; Chen, Y.; Liu, C. Electrical contacts between carbonnanotube coated electrodes. Diamond Relat. Mater. 2003, 12, 774. (41) An, L.; Friedrich, C. R. Measurement of contact resistance of multiwall carbon nanotubes by electrical contact using a focused ion beam. Nuclear Instrum. Methods Phys. Res., Sect. B 2012, 272, 169. (42) Wakaya, F.; Katayama, K.; Gamo, K. Contact resistance of multiwall carbon nanotubes. Microelectron. Eng. 2003, 67−68, 853. (43) Organic Electrochemistry; Lund, H., Hammerich, O., Eds.; Marcel Dekker, Inc.: New York, 2001. (44) Vaze, A.; Hussain, N.; Tang, C.; Leech, D.; Rusling, J. Biocatalytic anode for glucose oxidation utilizing carbon nanotubes for direct electron transfer with glucose oxidase. Electrochem. Commun. 2009, 11, 2004.

14064

dx.doi.org/10.1021/la302708h | Langmuir 2012, 28, 14055−14064