Published: October 17, 2011

r 2011 American Chemical Society 2251 dx.doi.org/10.1021/la202945s | Langmuir 2012, 28, 2251–2261

ARTICLE

pubs.acs.org/Langmuir

Characterization of Nanoporous Gold Electrodes forBioelectrochemical ApplicationsMiche�al D. Scanlon,† Urszula Salaj-Kosla,† Serguei Belochapkine,† Domhnall MacAodha,§ D�onal Leech,§

Yi Ding,|| and Edmond Magner*,†,‡

†Department of Chemical and Environmental Sciences & Materials and Surface Science Institute, and ‡SFI-SRC in Solar EnergyConversion, University of Limerick, Limerick, Ireland§School of Chemistry, National University of Ireland, Galway, Ireland

)School of Chemistry and Chemical Engineering, Shandong University, Jinan 250100, P. R. China

bS Supporting Information

’ INTRODUCTION

Nanoporous gold (np-Au) is a material of considerable versa-tility and diverse application.1�3 np-Au is conductive, chemicallyand mechanically stable, free of surface contaminants (arisingfrom preparation in nitric or perchloric acid), biocompatible, andeasily functionalized.1�3 It has a high surface-to-volume ratio andlarge surface area with pore sizes that are tunable from thenanometer to micrometer scale. These properties make np-Auattractive for applications in the immobilization of enzymes orantibodies,4,5 energy storage,6 catalysis,7 and bio/fuel cells.8,9

The use of large surface area supports can potentially improve theperformance of devices, e.g., the sensitivity of an affinity-basedbiosensor can be increased by increasing the number of immo-bilized receptor molecules. Sensitive biosensors have beendeveloped for the detection of DNA,4 prostate specific antigen,5

and thrombin.10

The development of electrochemical devices based on directelectron transfer (DET) between an immobilized enzyme and anelectrode surface, has been hampered by the (a) limited numberof redox enzymes exhibiting DET at the surface of an electrode(approximately 5% of known redox enzymes),11 (b) limitedstability of enzyme electrodes, and (c) low, observed, current

densities.12 Achieving DET requires selective modification of theelectrode surface to enable optimal orientation of the enzymeand to minimize the distance between the electrochemicallyactive center and the electrode surface.13 Correct orientation iscrucial as the rate of electron transfer decreases exponentiallywith distance.14 The high surface-to-volume ratio and the abilityto functionalize the surface of np-Au with alkanethiolate self-assembled monolayers (SAMs) can result in high loadings ofoptimally orientated enzymes. Such high loadings ensure that themaximum signal is achieved and factors such as substrate diffu-sion or enzyme kinetics limit the response. Immobilization of anenzyme within the protective sheltered surroundings of a nano-pore, as opposed to on an exposed flat surface, may afford addi-tional protection for the enzyme from external environmentalconditions, potentially preventing desorption and stabilizing theprotein against reversible unfolding.15 Any desorption of enzyme

Special Issue: Bioinspired Assemblies and Interfaces

Received: July 28, 2011Revised: October 6, 2011

ABSTRACT: The high surface areas of nanostructured electrodes can provide forsignificantly enhanced surface loadings of electroactive materials. The fabrication andcharacterization of nanoporous gold (np-Au) substrates as electrodes for bioelectrochem-ical applications is described. Robust np-Au electrodes were prepared by sputtering agold�silver alloy onto a glass support and subsequent dealloying of the silver component.Alloy layers were prepared with either a uniform or nonuniform distribution of silver and,post dealloying, showed clear differences in morphology on characterization with scanningelectron microscopy. Redox reactions under kinetic control, in particular measurement ofthe charge required to strip a gold oxide layer, provided the most accurate measurements of the total electrochemically addressableelectrode surface area, Areal. Values of Areal up to 28 times that of the geometric electrode surface area, Ageo, were obtained. Fordiffusion-controlled reactions, overlapping diffusion zones between adjacent nanopores established limiting semi-infinite lineardiffusion fields where the maximum current density was dependent on Ageo. The importance of measuring the surface area availablefor the immobilization was determined using the redox protein, cyt c. The area accessible to modification by a biologicalmacromolecule, Amacro, such as cyt c was reduced by up to 40% compared to Areal, demonstrating that the confines of somenanopores were inaccessible to large macromolecules due to steric hindrances. Preliminary studies on the preparation of np-Auelectrodes modified with osmium redox polymer hydrogels andMyrothecium verrucaria bilirubin oxidase (MvBOD) as a biocathodewere performed; current densities of 500 μA cm�2 were obtained in unstirred solutions.

2252 dx.doi.org/10.1021/la202945s |Langmuir 2012, 28, 2251–2261

Langmuir ARTICLE

that does occur is more easily compensated due to the abundanceof enzyme initially adsorbed in comparison to that at a planarAu electrode.

Cytochrome c (cyt c) is a small (104 amino acid residues)redox protein capable of DET between its heme cofactor and asuitably modified electrode.16 The role of cyt c in the respiratorychains of mitochondria, coupled to its stability, wide availabilityand comprehensive physiochemical characterization ensures thatit is widely used as a model for larger, more complex electrontransfer (ET) systems.16 A stable electroactive monolayer of cyt cmay be physically adsorbed by electrostatic interaction of lysineresidues to a carboxyl-terminated alkane thiolate SAM on Au.17

Alternatively, cyt c can be covalently attached using carbodiimidelinkages.18 Bilirubin oxidases (BODs) are glycosylated enzymes,with molecular masses of 52�64 kDa, and belong to a groupof “blue” multicopper oxidases that include laccases, ascorbateoxidase, and ceruloplasmine.19 BOD from the fungiMyrotheciumverrucaria (MvBOD) and Trachyderma tsunodae (TtBOD) havebeen purified and used as cathode biocatalysts in biofuel cells.They catalyze the four-electron reduction of O2 to water withoutproducing H2O2 and do so at neutral pH and relatively lowoverpotentials.20,21 Ikeda et al. initially described the use of BODas a biocathode, reducing O2 to water in the presence of themediator 2,20-azinobis(3-ethylbenzothiazoline-6-sulfonate) atlow overpotentials.22 High current densities for O2 reductionwere achieved via a mediated electron transfer mechanism byincorporating MvBOD and TtBOD into osmium-based redoxhydrogel polymers.19,21,23,24

Here, we describe the fabrication and characterization of twonp-Au electrodes of different morphology and surface roughness.np-Au electrodes were fabricated by sputtering a Au�Ag alloy onto a glass support with subsequent dealloying of Ag. In contrast toelectrodes fabricated from 50%Ag:50%Au (white gold), theelectrodes were mechanically robust, which greatly facilitatedtheir ease of use. A series of electrochemical probes was used tocharacterize the electrochemically addressable surface area, Areal.The importance of using a probe appropriate for the intendedapplication is evident from the range of surface areas that wereobtained. For applications such as biofuel cells, which entailimmobilization of enzymes, cyt c was used as a model system toascertain the surface area accessible to modification by a redoxprotein, Amacro. Significantly increased (11-fold) loadings wereobserved. Preliminary studies on the preparation of np-Auelectrodes modified with osmium redox polymer hydrogels andMvBOD as a biocathode were performed.

’EXPERIMENTAL SECTION

Fabrication of Nanoporous and Planar Au Electrodes.Metal targets for substrate deposition, Au (AJA InternationalInc., USA), Ag and Ti (Kurt J. Lesker Company Ltd., UK), were>99.99% pure. Magnetron sputtering was carried out at roomtemperature in an ultrahigh vacuum chamber (ORION-5-UHVcustom sputtering system) onto plain precleaned glass micro-scope slides. Prior to metal deposition, the glass slides werecleaned in the vacuum chamber by Ar plasma. Sample substrateswere tilted at ∼70� from the surface normal and rotated at20�40 rpm to ensure uniform deposition. Sputter depositionrates were calibrated using a quartz crystal thickness monitor.The planar Au substrate was prepared by deposition of a∼10 nmthick Ti adhesion layer followed by a∼100 nm thick Au layer. Athinner Au substrate layer (35 nm) was used as a base for the

deposition of a composite layer of AgxAux�1 with a uniformdistribution of Ag and Au. The∼35 nm thick Au layer improvesadhesion and prevents delamination of the AgxAu1�x alloy layerduring dealloying.25 In addition, any potential redox activity ofthe underlying Ti layer is suppressed. Eachmetal was sputtered ata constant sputtering rate to produce a∼300 nm thick Ag67Au33alloy layer with a uniform distribution of Ag throughout the alloy.AgxAux�1 substrates with a nonuniform distribution of Ag wereprepared by simultaneously depositing Au at a constant sputter-ing rate while Ag was deposited at a gradually increasing sputter-ing rate. This procedure produced a ∼300 nm thick AgxAux�1

layer with a higher distribution of Ag near the surface of the alloylayer. Both “uniform” and “nonuniform” substrates were dealloyedin concentrated (70% w/v) nitric acid for 15 min at 38 �C. Onremoval from the nitric acid bath, the substrates were thoroughlyrinsed with deionized water and dried in a stream of nitrogen.The electrodes were electrically connected via a Ag wire attachedto the Au substrate using indium and an epoxy glue. A circularelectrode area (0.28 cm radius) was defined using an insulatingpaint (Gwent Electronic Materials Ltd., UK).The structure and chemical composition of np-Au was char-

acterized using a scanning electron microscope (SEM; HitachiSU-70) equipped with an energy dispersive X-ray spectrometer(EDS, Oxford Instruments). In order to minimize chargingeffects, beam voltages were adjusted between 3 and 5 kV, and acorner of the sample was electrically connected to the sampleholder with silver paint. The Au pore, ligament, and crack sizeswere determined manually by identifying a minimum of 10 ofeach feature and making measurements across the shortestdistance using ImageJ software.Preparation of Nanoporous Gold Leaf Electrodes. 12-carat

Au�Ag leaf (50 wt % Au) was obtained from Wilhelm WasnerBlattgold, Germany. After dealloying in (70% w/v) nitric acid for15 min at 30 �C, the free-standing np-Au films were washedseveral times with ultrapure water and floated on water. Thefloating np-Au films were attached to a polished glassy carbon(GC) electrode and allowed dry for 12 h at 4 �C to create a highsurface area np-Au/GC electrode, as described previously.26

Electrochemical Measurements and Reagents. All electro-chemical studies were performed using either a CHI832 bipo-tentiostat or CHI630A potentiostat (CH Instruments, Austin,Texas, USA) in a standard three-electrode electrochemical cellwith a 0.5 mm diameter Pt wire counter electrode (ALS Co. Ltd.,Tokyo, Japan) and Ag|AgCl|3 M KCl reference electrode(IJ Cambria Scientific Ltd., UK). Immediately prior to use, np-Auelectrodes were electrochemically cleaned by cycling the poten-tial between 0.0 and +1.5 V in 0.5 M sulfuric acid until a stable Auoxide cyclic voltammogram was obtained. Planar Au electrodeswere cleaned by dipping in piranha solution (1: 3 (v/v) mixtureof 30% H2O2: 98% H2SO4) for 30 s followed by rinsing inultrapure water in order to prevent electrochemical rougheningof the surface. (Warning: piranha solution reacts strongly withorganic compounds. Handle with extreme caution at all times. Do notstore the solution in a closed container.) All reagents were ofanalytical grade, purchased from Sigma-Aldrich Ireland, Ltd.,unless stated otherwise, and used as received. All solutions wereprepared in ultrapure water (resistivity of 18.2 MΩ cm) from anElgastat maxima-HPLC (Elga, UK) with the exception of thiolsolutions, which were prepared in absolute ethanol (Lennox Ltd.,Ireland). Cyt c from horse heart (Sigma C7752) was used asreceived. MvBOD was purchased from Amano Enzymes, Inc.(“Amano-3”, 2.63 unit mg�1) and used as received.

2253 dx.doi.org/10.1021/la202945s |Langmuir 2012, 28, 2251–2261

Langmuir ARTICLE

Electrochemical Characterization. Surface-confined or dif-fusion-controlled probes were used to determine the electro-chemically addressable surface areas (Areal) of each electrodedesign. With the exception of the Au oxide stripping technique(see below), Rf values were calculated by comparison of eithersurface coverage (Γ, pmol cm�2), double layer capacitance(Cdl, μF cm�2), or current density (I, mA cm�2), at a planar Auelectrode to that at a np-Au electrode. Γ, Cdl, and I values (TableS1) were normalized to the geometric electrode surface area(Ageo, 0.246 cm

2).Γ values for copper, 6-(ferrocenyl)hexanethiol(FcHxSH) and cyt c were obtained by integration of the anodiccurrent and converted to %monolayer coverage using theoreticalΓ values of 2000, 450, and 15 pmol cm�2 for complete mono-layer coverage of Cu,27 FcHxSH,28 and cyt c,29 respectively. Areal

values were determined by multiplying Ageo by Rf.Surface-Confined Electrode Reactions. (a) The charge re-

quired to strip a Au oxide layer in 0.5 M H2SO4 was determined,and a conversion factor of 390 μC cm�2 was applied to obtainAreal.

30 (b) Cdl values were calculated using31

Cdl ¼ Itotal=½ð2νAgeo � 10�6� ð1Þ

where Itotal is the sum of the anodic (ia) and cathodic (ic) currentsand ν is the scan rate. ia and ic were obtained at potentialsof �100 mV and �200 mV (where no faradaic processes wereoccurring) from cyclic voltammograms of planar and both np-Auelectrodes, respectively, in electrolyte containing 50 mM KH2PO4

and 29.1 mM NaOH at pH 7.32 (c) Underpotential deposition(UPD) of Cu was completed in aqueous solutions of either 1 or2 mM CuSO4 in 0.1 M H2SO4.

33�35 (d) Redox active thiolmonolayers were prepared by immersing electrodes in a solutionof 0.1 mM 6-(ferrocenyl)hexanethiol (FcHxSH) and 2 mM4-mercaptobutanol (HO(CH2)4SH) for 48 h. After washingwith ethanol, the response of the electrodes was examined in1 M HClO4.

28 (e) Covalent tethering of cyt c to thiol-modified Auelectrodes was accomplished, as described below. (f) A “blockingSAM”36 of long alkane chains (nine or more methylene units)was suitable to confirm the coating of all inner/outer goldsurfaces with thiol molecules. Au electrodes were modified byimmersion in 10 mM 1-hexadecanethiol (CH3(CH2)15SH)37

for 72 h.Diffusion-Controlled Electrode Reactions. (a) The electro-

chemical responses of the redox probes, potassium ferricyanide(5 mM Fe(CN)6

3� in 100 mM KCl) and ruthenium(III)hexamine chloride (5 mM Ru(NH3)6

3+ in 100 mM KCl) wereinvestigated at each electrode design. (b) Bulk deposition of Cuwas completed in aqueous solutions of either 1 or 2 mM CuSO4

in 0.1 M H2SO4.33,34 (c) Oxygen reduction was performed in

phosphate buffered saline (PBS, 50 mM K2HPO4�KH2PO4,150 mM NaCl, pH 7.4).Enzyme Immobilization and Biocathode Preparation.

Mixed monolayers of 1 mM 11-mercaptoundecanoic acid(HS(CH2)10COOH) and1mM6-mercapto-1-hexanol (HS(CH2)6-OH) were assembled at the Au surface. The surface carboxylgroups of the SAM were activated by immersion in 5 mMN-cyclohexyl-N0-(2-morpholinoethyl)carbodiimide metho-p-to-luenesulfonate (CMC), prepared in 100 mM K2HPO4�KH2-

PO4 buffer, pH 7, for 30 min at 4 �C. Covalent attachment ofthese activated carboxyl groups to lysine residues of cyt c wasachieved by exposure to 50 μM cyt c prepared in 4.4 mMphosphate (K2HPO4�KH2PO4) buffer, pH 7, for 60 min at4 �C. Cyclic voltammetry was performed in 4.4 mM phosphate

buffer, pH 7, at various scan rates. Laviron’s model38 was used tocalculate the electron transfer rate constant.A solution containing the osmium-based redox polymermediator

[Os(2,20-bipyridine)2(polyvinylimidazole)10Cl]+/2+ (Os(bpy)2 3PVI,

8 μL of a 6 mg/mL solution/suspension in water), poly(ethyleneglycol)diglycidyl ether (PEGDGE, 1.9 μL of a 15 mg/mL solu-tion in water) as a cross-linker, and the enyzme Myrotheciumverrucaria bilrubin oxidase (MvBOD, 4.8 μL of a 10 mg/mLsolution in water) as the biocatalyst was prepared in a 200 μLeppendorf tube and gently mixed. The solution was carefullydrop coated onto the Au electrode surface, ensuring all areas ofthe electrode were coated. The deposited film was allowed to dryfor precisely 24 h in darkness at room temperature. Themodifiedelectrodes were immersed in electrolyte solution (50 mM phos-phate buffer, 150 mMNaCl, pH 7.4) at 37 �C for at least 20 minprior to electrochemical measurements to allow for film swelling.The response in the absence of dissolved O2 was monitored inbuffer purged with nitrogen, and the bioelectocatalytic reductionof oxygen was monitored in buffer bubbled with O2. Currentswere measured at the peak potential, and all current densitieswere normalized to the geometric electrode area, Ageo.

’RESULTS AND DISCUSSION

SEM and EDS Characterization of np-Au Electrodes. Re-presentative top-down and cross-sectional SEM images of bothnp-Au electrode designs are shown in Figure 1. The “uniform”and “nonuniform”methods of sputtering Au�Ag alloys resultedin np-Au films with clear morphological differences. The electro-des consisted of three-dimensional networks of quasi-periodicnanoporous voids (or pores) and interconnected gold ligaments.Significant volume contraction (20�30%) of the films duringdealloying induces stresses that are manifested as cracks.39 np-Aufilms produced using the nonuniform sputtering method hadrelatively fine structures and an increased number of cracksbetween islands of ligaments. The pore, ligament, and crack sizeswere 12�20 nm, 15�18 nm, and 55�65 nm, respectively.Uniformly sputtered np-Au films were coarser with pore, liga-ment, and crack sizes of 17�23 nm, 18�25 nm, and 40�55 nm,

Figure 1. Top-down and cross-sectional SEM images of np-Au electro-des obtained after dealloying (A, C) a uniformly sputtered Au�Ag filmand (B, D) a nonuniformly sputtered Au�Ag film, in 70% (w/v) nitricacid at 38 �C for 15 min.

2254 dx.doi.org/10.1021/la202945s |Langmuir 2012, 28, 2251–2261

Langmuir ARTICLE

respectively. Cross-sectional SEM images (Figure 1C, D) demon-strated that the wormlike structure of interconnected ligamentswas regularly distributed throughout the entire volume of both

films after dealloying. The pure Au adhesion layer remainedintact after dealloying (Figure S1). Relative to uniform np-Au, thefiner pore structures and increased crack densities in nonuniform

Figure 2. Electrochemical characterization of nonuniform nanoporous (dashed lines), uniform nanoporous (dotted lines), and planar (solid lines) goldelectrodes using surface-confined probes: (A) gold oxide reduction; (B) gold hydroxide formation; (C) UPD of copper (1 mMCuSO4 in 0.1 MH2SO4

at a scan rate of 1 mV s�1); (D,E) bulk and UPD of copper (1 mM and 2 mMCuSO4, respectively, in 0.1 MH2SO4 at a scan rate of 1 mV s�1); and (F)6-(ferrocenyl)hexanethiol/4-mercaptobutanol SAM immobilization. Currents were normalized to the geometric surface areas of the electrodes. Unlessstated otherwise, a scan rate of 100 mV s�1 was used.

2255 dx.doi.org/10.1021/la202945s |Langmuir 2012, 28, 2251–2261

Langmuir ARTICLE

np-Au electrodes resulted in higher surface roughnesses, as outlinedbelow. Successive batches of both np-Au electrode designs wereprepared in a reproducible manner in terms of pore, ligament, andcrack sizes, as confirmed by SEM and Areal (calculated from thecharge required to strip a gold oxide layer).30 EDS data (notshown) confirmed the removal of Ag with only 0.6 and 2.8 wt %Agremaining on the uniform and nonuniform np-Au surfaces, respec-tively, post dealloying.Electrochemical Characterization. Cyclic voltammetry of

planar and np-Au electrodes in 0.5 M H2SO4 showed two broadoxidation peaks at 1.2 and 1.4 V, attributed to the formation of Ausurface oxides,30 and a sharp reduction peak at 0.9 V due tosubsequent removal of the oxides (Figure 2A). The absence ofany additional peaks confirmed that Ag was removed in thedealloying process, in agreement with EDS data. Cyclic voltam-mograms in a 50mMKH2PO4 and 29.1mMNaOH, pH 7, buffersolution confirmed the formation of AuOH at a potential of∼300 mV at each Au electrode design (Figure 2B). At potentialswhere faradaic processes were occurring, �100 mV for planarand �200 mV for both np-Au electrodes, large increases in Cdl

were observed at np-Au electrodes. The Rf values obtained byboth the Au oxide stripping and capacitance-based methods fornonuniform np-Au electrodes were in agreement with eachother. However, larger than expected Rf values were calculatedat uniform np-Au electrodes and greater electrode-to-electrodevariability was observed at each Au electrode design for thecapacitance-based method (Table 1).At both np-Au electrodes, the magnitudes of the observed

current signals were enhanced significantly in the potential rangeof Cu UPD, in comparison to those at planar Au electrodes(Figure 2C). No significant current enhancements were ob-served in the potential range for bulk Cu deposition (Figure 2D).Bulk Cu deposition is controlled by the diffusion of aqueous Cu2+

to the electrode surface, whereas Cu UPD is a surface-confinedelectrochemical process. Similar currents were observed for bulkCu deposition at planar and np-Au electrodes, indicating that, onthe time scale of the experiment, overlap of diffusion zones occurbetween individual nanopores at the np-Au surface establishing

a semi-infinite linear diffusion field. Thus, Cu2+ in solution isreduced only at the surface and, perhaps, the upper pore regionsof np-Au. The current densities of the diffusion-controlled andsurface-confined processes are dependent upon Ageo and Areal,respectively. Submonolayer coverage of Cu was observed at theplanar Au electrode. The UPD response was unchanged onincreasing the concentration of CuSO4 from 1 to 2 mM, indi-cating that a concentration of 1 mM CuSO4 was sufficient toachieve maximum coverage. However, the magnitude of the bulkdeposition peak current doubled, as expected for a diffusioncontrolled process (Figure 2E). The Rf value for a uniform np-Auelectrode calculated from the Cu UPD surface coverage was ingood agreement with the values obtained from the Au oxidestripping method. However, at nonuniform np-Au electrodes,the Rf value from the CuUPDwas less than that measured by theAu oxide stripping technique (Table 1). Nonuniform Au elec-trodes have smaller pore and ligament sizes, as confirmed bySEM, than uniform np-Au, indicating that some of the nanoporesmay be inaccessible to Cu2+.Mixed monolayers of FcHxSH/HO(CH2)4SH were coad-

sorbed on planar and np-Au electrodes in ethanol with a FcHxSHsolution mole fraction (Xsol) of 0.05. Cyclic voltammogramsobtained in 1 M HClO4 contained two peaks on both the anodicand cathodic sweeps for each electrode design (Figure 2F). Suchmultiple peaks are indicative of adsorption of FcHxSH moietiesin isolated (lower-potential) and clustered (higher-potential)states, respectively.40 The anodic and cathodic peak potentialsdecreased to less positive values at np-Au electrodes, indicative ofstabilization of the ferricenium form of the probe. This stabiliza-tion may arise from a more hydrophilic environment on thesurface of np-Au. Cdl was consistently larger at potentials positiveof the formal redox potential, E�0, for ferrocene oxidation. Thisobservation, in agreement with previous reports,28 may be due toincreased permeability of the monolayer to electrolyte ions and/or solvent on oxidation of ferrocene. Xsol did not correlatedirectly with surface mole fraction (Xsurf) as the adsorbate withthe longest alkyl chain, FcHxSH, is preferentially adsorbed.28 At aplanar Au electrode, Xsurf for FcHxSH was 0.36 (i.e., 36%

Table 1. Comparisons of Calculated Roughness Factors (Rf), Percent Monolayer Formation, and Electrochemically AddressableSurface Areas (Areal) as Determined Using Surface-Confined and Diffusion-Controlled Electrochemical Probes at Each AuElectrode Designa

roughness factor, Rf % monolayer formation

probe uniform np-Au nonuniform np-Au planar Au uniform np-Au nonuniform np-Au

UPD Cu2+/0b 15.50 ((0.02) 19.00 ((0.03) 66 ((1) 1031 ((15) 1258 ((30)

FcHxSH 7.40 ((0.05) 9.3 ((2.8) 36 ((1) 268 ((1) 334 ((12)

cyt c 9.3 ((1.2) 10.8 ((1.2) 100c 930 ((120) 1091 ((95)

electroactive area (Areal, cm2)

Au Ox/Red 16.1 ((1.3) 28.1 ((1.5) 0.251 ((0.013) 3.961 ((0.164) 6.931 ((0.238)

Cdl 21.2 ((2.3) 28.8 ((3.2) 0.246 5.214 ((0.556) 7.081 ((0.766)

bulk Cu2+/0b 1.05 ((0.14) 1.35 ((0.15) 0.246 0.259 ((0.030) 0.332 ((0.040)

Fe(CN)63‑/4‑ 1.16 ((0.07) 1.25 ((0.08) 0.246 0.286 ((0.018) 0.308 ((0.018)

Ru(NH3)63+/2+ 1.09 ((0.02) 1.02 ((0.01) 0.246 0.269 ((0.003) 0.251 ((0.002)

O2 2.07((0.12) 2.09 ((0.18) 0.246 0.509 ((0.029) 0.514 ((0.014)aAverages and standard deviation data from a minimum of three electrodes tested. bUPD and bulk Cu2+/0 data presented for experiments carried outwith 1 mM CuSO4 in solution. cTheoretical % monolayer formation of cyt c at a planar Au electrode.

2256 dx.doi.org/10.1021/la202945s |Langmuir 2012, 28, 2251–2261

Langmuir ARTICLE

monolayer coverage). Assuming that Xsurf remains constant at0.36 for both np-Au electrodes, this would yield Rf values of 7.4((0.05) and 9.3 ((2.8) for uniform and nonuniform np-Auelectrodes, respectively. These values were considerably lowerthan those calculated using all other surface confined techniques,even for the immobilization of more bulky cyt c molecules,discussed below. Control experiments were carried out toconfirm that all pore surfaces of np-Au could be coated withthiol molecules. A layer of 1-hexadecanethiol proved extremelyeffective in completely suppressing the redox responses ofpotassium ferricyanide (Figure 3C) and AuOH (data not shown)at planar and np-Au electrode surfaces, indicative of completemonolayer coverage. This observation eliminated the inability ofthiol molecules to access and modify all areas of the np-Aunanostructure as a factor in accounting for the low FcHxSHsurface coverages and, hence, the low Rf values observed.A possible explanation for the observed low surface coverages

is that different electrode surface morphologies produce none-quivalent adsorption sites and thus different Xsurf values forFcHxSH.41 The major morphological difference between planar

and np-Au electrodes is the presence of a higher density of defect-type sites (such as step-edges) in the latter.41 Defect sites areunlikely to be places of easily exchangeable molecules. Thus, anythiol adsorbed at a defect site will remain adsorbed and is lesslikely to exchange with free thiols in solution. Effectively, diluentmolecules stabilized at defect sites no longer exchange withFcHxSH. The adsorption of diluent subsequently becomes morefavored than at planar Au and Xsurf values for FcHxSH are reduced.Cyclic voltammograms for the diffusion-controlled redox

probes ruthenium(III) hexamine chloride (Figure 3A) andpotassium ferricyanide (Figure 3B) show no significant increasein current density on substituting planar with np-Au electrodes.As observed for bulk Cu deposition, semi-infinite linear diffusionfields are established on the time-scale of the experiment limitingthe observed current densities. Switching from planar to np-Audid, however, have a significant influence on the response observedfor oxygen reduction (Figure 3D). A positive shift and splitting ofthe oxygen reduction peak was observed, from�433 mV at planarAu to �210 (peak 1)/�346 (peak 2) mV, and �196 (peak 1)/�307 (peak 2) mV, at uniform and nonuniform np-Au,

Figure 3. Electrochemical characterization of nonuniform nanoporous (dashed lines), uniform nanoporous (dotted lines), and planar (solid lines) goldelectrodes using diffusion-confined probes: (A) Ru(NH3)6

3+/2+ and (B) Fe(CN)63�/4� in 0.1 M KCl. (C) Suppression of the Fe(CN)6

3�/4� redoxresponse in 0.1 M KCl by the formation of a SAMmonolayer on immersion in a 10 mM CH3(CH2)15SH solution for 72 h. (D) O2 reduction at a scanrate of 50mV s�1. Currents were normalized to the geometric surface areas of the electrodes. Unless stated otherwise, a scan rate of 100mV s�1 was used.

2257 dx.doi.org/10.1021/la202945s |Langmuir 2012, 28, 2251–2261

Langmuir ARTICLE

respectively. Also, a doubling of the current density was observedat the np-Au electrodes. As described previously,42 such observa-tions are a result of oxygen reduction proceeding at planar andnp-Au via different mechanisms due to the catalytic effect ofdefect sites present in np-Au. Oxygen reduction at np-Auelectrodes proceeds via a four electron process that takes placein two steps. Oxygen is first reduced to hydrogen peroxide andthen, due to the high reactivity of the defect sites in np-Au,undergoes reduction to water. At a planar Au electrode, oxygenreduction proceeds via a two electron reduction to hydrogenperoxide. Any subsequent reduction to water that may takeplace at planar Au occurs too slowly to be observed under theexperimental conditions employed.42

Direct comparisons between planar and np-Au systems for (a)adsorption of mixed monolayers of redox active thiols and (b)oxygen reduction are not valid due to the significant influencethe electrode morphologies, in terms of defect site densities, haveon the observed response. Comparisons of current densities atplanar and np-Au systems for diffusion-controlled redox reac-tions, as a reliable method to determine the electrochemicallyaddressable surface areas (Areal), are not valid as the overlappingdiffusion zones at np-Au limit the current response. Minimalincreases in the electroanalytical signal for redox probes with fastelectron transfer kinetics, despite substantial increases of Areal,have been reported for nanostructured electrodes based onmacroporous Au,43 Au nanopillars,44 nanoporous Au attached toGC,45 andmesoporous Pt.46 Redox active immobilized species orspecies with redox reactions under kinetic-control (i.e., slowkinetics) can provide a better measure of the electroactive sur-face area. Cu UPD, for example, is a suitable method to deter-mine Areal, although it may give slightly lower values for very finepore structures. The most commonly used, and most accurate,method to determine Areal is measuring the charge required tostrip a Au oxide layer. This method is widely used to characterizeAu based nanostructured electrodes.35,43,44,47 A less widely usedtechnique, determining the Cdl at potentials where no faradaicprocesses are occurring is a suitable complementary method toestimate Areal for np-Au electrodes. This capacitance-basedmethod should not, however, be used as a stand-alone techniqueto estimate Areal due to the relatively large electrode-to-electrodevariations, compared to the Au oxide stripping technique, andpossible overestimations of Areal observed.Voltammetric Characterization of Cyt c Immobilized on

Planar and np-Au Electrodes. The voltammetric responses ofcyt c covalently immobilized at planar and np-Au electrodesmodified with a HS(CH2)10COOH/HS(CH2)6OHmixed SAMare shown in Figure 4. In comparison to homogeneous SAMs,mixed SAMs result in faster and more reversible electron transferkinetics for immobilized cyt c .29 The Eo0 values of cyt c at both np-Au electrodes (Table 2) were constant over a range of scan rates(10�1000 mV s�1), in good agreement with similarly modifiedelectrodes.48 Plots of the anodic and cathodic peak currentsversus scan rate were linear (Figure 5A), indicating that cyt c wasadsorbed on the electrode surface. Voltammograms were nearlysymmetric, with some deviations in the peak separation (ΔEp)and full width at half-maximum (fwhm) from ideal behavior(0 and 90.6 mV, respectively; Table 2).49 Such deviations mayarise from a heterogeneous distribution of cyt c on themodified electrode surface and/or variations of the electrodesurface (pore opening, size etc). A distribution of formalpotentials may occur and, as described below, a distributionof heterogeneous electron transfer rate constants (ket) may

be observed.50�52 No changes in the voltammetric responsewith continuous potential cycling (30 scans) in aqueous buffer(not shown) were observed, demonstrating the stability of theadsorbed protein.Electrode topography, low Rf planar versus high Rf np-Au

surfaces, and heterogeneous mixed SAMs noticeably influencethe voltammetry of adsorbed cyt c. Rough electrode surfacesmodified with mixed SAMs to generate monolayers with higherdefect densities provide the optimal faradaic response.31,53 Theirregularly textured surfaces described here provide such aresponse (Figure 4), which is in contrast to the poor responseobtained at surfaces with minimal defects or smooth surfacesmodified with long chain homogeneous carboxylic acid termi-nated SAMs.53 The response (inset Figure 4) of cyt c at planar Auelectrode surfaces modified with a heterogeneous mixed SAMdisplayed more asymmetrical cyclic voltammograms (largerΔEpof 70 mV), with broader, less defined, anodic peaks in compar-ison to identical immobilization conditions on np-Au electrodes.The electron transfer coefficient (α) and heterogeneous rate

constant (ket) for cyt c at both np-Au electrodes were obtainedusing Laviron’s method38 (Table 2). An irreversible response(peak separation >200 mV/n, where n = 1) occurred for scanrates greater than 1.25 and 1.75 V s�1 for uniform and nonuni-form np-Au electrodes, respectively. Values of α were deter-mined from the slopes of the linear regions of plots of peak

Figure 4. Cyclic voltammograms of 50 μM cyt c covalently immobilizedon a 50/50 mol % HS(CH2)10COOH/HS(CH2)6OH mixed SAM atplanar (solid line and inset), uniform (dotted line), and nonuniform(dashed line) np-Au electrodes in 4.4 mM K2HPO4�KH2PO4 buffer,pH 7, at a scan rate of 100 mV s�1. Current densities were normalized tothe electrodes geometric surface area.

Table 2. Voltammetric Characterization of Cyt c/HS-(CH2)10COOH�HS(CH2)6OH SAM-Modified np-AuElectrodesa

electrode

design

E00

(mV) b

ΔEp(mV) c

fwhm

(mV) d α

ket(s‑1)

uniform np-Au �12 ((2) 18 ((1) 115 ((1) 0.53 2.4 ((0.9)

nonuniform np-Au �11 ((1) 18 ((1) 114 ((1) 0.50 3.3 ((0.9)aAverages and standard deviations represent data from four electrodes.bData obtained from scan rates in the range 10�1000 mV s�1. cΔEp =(Ep,a� Ep,c).

dΔEp and fwhm data obtained at a scan rate of 50 mV s�1.

2258 dx.doi.org/10.1021/la202945s |Langmuir 2012, 28, 2251–2261

Langmuir ARTICLE

potential versus scan rate (Figure 5B, C), while ket was deter-mined for scan rates in the range 1.25�4 V s�1 and 1.75�4 V s�1,for uniform and nonuniform np-Au electrodes, respectively.A range of rate constants, with average values of ket of 2.4( 0.9 s�1

for uniform np-Au (range of 1.3�3.5 s�1) and 3.3 ( 0.9 s�1 fornonuniform np-Au (range of 2.1�4.3 s�1) was obtained. Thisdistribution of ket values likely arises due to the variety of possibleorientations cyt c molecules adopt on the negatively chargedmixed SAM,50�52 as discussed earlier. The rate constants were ca.2 orders of magnitude less than those achieved at similarly modifiedplanar gold electrodes.54 The lower rate constants may arise fromchanges in the composition of the SAM, which have been observedwith FcHxSH. Such changes can significantly change the faradaicreponse of the protein.53

Surface coverages of electrochemically active cyt cwere∼9 and∼11 times that possible at a planar Au surface (assuming fullmonolayer coverage per geometric surface area) for uniform andnonuniform np-Au electrodes, respectively (Tables S1 and 1).Considering that the sensitivity of affinity-based biosensors islimited by the binding capacity of immobilized receptor mol-ecules, increasing the quantity of immobilized receptor mol-ecules on substrates such as np-Au has the potential tosignificantly increase the sensitivity of such sensors.Rf values, particularly at nonuniform np-Au electrodes, were

less than those obtained using other surface-confined techniques(Table 1). Whereas immobilization of cyt c can not provide anaccurate estimate of Areal, it does provide an indication of theaccessible area of a particular nanostructure to a protein or largemacromolecule, Amacro. Values of Amacro were 2.288 ((0.295)and 2.657 ((0.295) cm2, for uniform and nonuniform np-Auelectrodes, respectively, compared to Areal values of 3.961((0.164) and 6.931 ((0.238) cm2 (from the Au oxide strippingmethod, Table 1). The more pronounced decrease in Amacro fornonuniform np-Au electrodes reveals that a considerable portionof the finer nanostructure was inaccessible to xthe bulky cyt cmolecule (diameter of 6.6 Å).16 As discussed above, a controlexperiment with a “blocking” SAM, showed that the walls of thenanopores could be fully modified and that incomplete surfacemodification would not account for the reduced surface cov-erages of electrochemically active cyt c. In contrast to othersurface-confined methods, cyt c immobilization can be used toprovide a more accurate probe of the loading capacities of largemolecules on Au nanostructures.Comparative Study with np-Au Leaf Modified GC Electro-

des. The np-Au electrodes were mechanically robust and se-curely attached to the underlying glass support by a Ti/pure Auadhesion layer. This allowed ease of handling/manipulation and,as outlined, facile functionalization of the electrode surface withthiols and large biological molecules. An alternative methodof fabricating np-Au electrodes involved using free-standingAu�Ag leaf as a precursor film of np-Au. Attempts were made tomodify an np-Au/GC electrode with a mixed thiol SAM in anidentical manner as described for cyt c immobilization onsputtered np-Au electrodes. However, thiol modification of theAu was found to disrupt the weak attractive forces securing thenp-Au film to the underlying GC electrode, detaching the np-Aufilm from the GC. Attempts to modify the np-Au film with thiolmolecules prior to attachment on GC were not successful, eitherdue to the film breaking in solution or failure of the thiol modifiedfilms to attach to GC. Improved attachment of the np-Au film toGC has been reported by dropping a suspension of Nafion on adry np-Au/GC electrode.9,26 This, however, would preclude the

Figure 5. (A) Plots of the anodic and cathodic peak current densitiesfor immobilized cyt c as a function of scan rate at uniform (hollowcircles) and nonuniform (solid circles) np-Au electrodes; (B) depen-dence of the anodic and cathodic peak potentials for adsorbed cyt c onthe logarithm of the scan rate at uniform (hollow circles) andnonuniform (solid circles) np-Au electrodes; (C) linear regionsof the plots in (B) used to calculate α (peak separation >200 mV/n,with n = 1).

2259 dx.doi.org/10.1021/la202945s |Langmuir 2012, 28, 2251–2261

Langmuir ARTICLE

ability to modify the np-Au with thiols or any other potentiallinker molecules.

BiocathodeDevelopment at Planar andnp-Au Electrodes.The mediator Os(bpy)2 3 PVI has previously been screened foruse with biocathodes employing MvBOD. Fast electron transferkinetics and appreciable current densities were achieved.21 A drop-cast coating consisting of 38.6 wt % MvBOD, 38.6 wt %Os(bpy)2 3PVI, and 22.8 wt%PEGDGEwas utilized. This enyzme/redox polymer weight ratio of 1:1 was previously optimized for themediator [Os(4,40-dichloro-2,20bipyridine)2Cl]

+/2+ complexedwith polyacrylamide�poly(N-vinylimidazole), [(PAA�PVI)�Os(dcl-bpy)2]

+/2+, in combination with MvBOD.55 The cata-lytic current will be limited by the enzyme-catalyzed rate of O2

reduction if too little enzyme is present or by increasedresistance in the film if too much nonconductive enzyme ispresent.55 At 22.8 wt % of cross-linker, the films were stable,showing no variation in response after 100 scans in the absenceof O2, while enabling appreciable current densities to beachieved in the presence of O2.Near identical symmetric responses were observed for the

Os2+/3+ complex at each electrode design in terms of Em�0, ΔEp,half-peak widths, current density, and total charge transferred(Figure 6). Voltammograms differed only in the magnitudes ofthe double layer capacitances (Cdl) observed, with Cdl increas-ing with Rf, as expected. Em�0 values of 224 ((2) mV were inagreement with previous results.21 At 5 mV s�1ΔEp values of 35((5) mV and half-peak widths of 115 ((5) mV were measured,indicating minor deviations from the theoretical values for anideal Nernstian one-electron transfer reaction, as discussed for cytc. These deviations may be attributed to repulsive interactionsbetween redox centers or small variations in the local environ-ments of redox centers giving rise to a range of redox potentials.56

Scan rate (ν) studies of the polymer film in the absence of thebiocatalyst at nonuniform np-Au electrodes (not shown) re-vealed thin-film behavior for ν < 20mV s�1 (ip increasing linearlywith ν), followed by diffusion-controlled behavior at ν > 20 mVs�1 (ip increasing linearly with ν

1/2). Identical surface coveragesof the immobilized osmium complex, calculated from the chargeunder the anodic redox peaks following comprehensive electro-lysis at 5 mV s�1, of 44.2 ((2) nmol cm�2 were detected atplanar and both np- Au electrodes. On the time-scale of theexperiment, the diffusion layer thickness (δ) exceeded the filmthickness (ϕ), as indicated by the linear increase in ip for scanrates <20 mV s�1, at planar and both np-Au electrodes. Increas-ing the surface roughness of the electrodes did not increase thequantity of addressable Os2+/3+ centers in the hydrogel since, atslow scan rates, all of these highly mobile Os2+/3+ centers werealready fully oxidized and reduced during a potential cycle at aplanar Au electrode.At each electrode design, the onset of the O2 reduction current

began at a potential slightly positive of Em�0, and a limiting peakcurrent was reached at Em�0, which decayed gradually to apotential-independent plateau current at potentials ∼150 mVmore negative than Em�0. The catalytic current response at bothnp-Au electrodes revealed a splitting of the reduction peak, inparticular for uniform np-Au (Figure 6B). This may arise fromthe mediator/enzyme complex inhabiting different environ-ments within the pores; such peak splitting was not observedfor laccase and cellobiose dehydrogenase (data not shown). Therate of O2 reduction catalyzed by MvBOD and mediated byOs(bpy)2 3 PVI in the film was high when compared to the rate ofO2 diffusion in the electrolyte, resulting in depletion of O2 at theelectrode surface. Thus, concentration polarization, a conse-quence of not rotating the electrode, was manifested as a peak

Figure 6. Cyclic voltammograms in the absence (dashed line) andpresence (solid line) of O2 at (A) nonuniform nanoporous, (B) uniformnanoporous, and (C) planar gold electrodes modified with MvBOD(38.6 wt %), Os(bpy)2 3PVI (38.6 wt %), and PEGDGE (22.8 wt %) in50mMphosphate buffer saline, at 37 �C, and pH7.4. Scan rate of 5mV s�1.

2260 dx.doi.org/10.1021/la202945s |Langmuir 2012, 28, 2251–2261

Langmuir ARTICLE

current on the voltammogram. Near identical peak currentdensities of 476 ((74), 478 ((29), and 521 ((20) μA cm�2

were obtained at planar, uniform nanoporous, and nonuniformnanoporous Au electrodes, respectively. As outlined for diffusion-controlled electrode reactions, overlapping diffusion zones be-tween adjacent nanopores establish limiting semi-infinite lineardiffusion fields on the time-scale of the experiment and themaximum current density was dependent on Ageo.The observed current densities in Figure 6 are in agreement

with the results obtained for similar biocathodes at GC electrodeswhich were subject to moderate rotation speeds of 100 rpm.21

Heller described aMvBOD cathode “wired” using a [(PAA�PVI)�Os(dcl-bpy)2]+/2+ mediator, which was capable of operatingwith a maximum current density of 700 μA cm�2 under stagnantphysiological conditions at carbon fiber microelectrodes.57 Usingsimilarly modified carbon cloth electrodes under physiologicalconditions, maximum current densities of 5 mA cm�2 wereobtained at 1000 rpm.55 These higher current densities arise from(a) the use of microelectrodes and high rotation rates, whichincrease the rate of O2 transport to the electrode surface, and (b)the use of different osmium redox polymers, which significantlyinfluence the catalytic current.21

Detailed studies examining a range of redox polymers andenzymes for use as anodes and cathodes in biofuel cells areunderway. Improvements in stability expected within a shelterednanoporous environment, currently under investigation for theredox polymers used in this study, may be further augmented byanchoring and tethering redox polymers and enzymes ontosuitably pretreated surfaces. Improved stability by anchoringfunctionalized redox complexes and enzymes to suitably deriva-tized Au58 and carbon59 surfaces, using methods easily adaptablefor use at np-Au, was demonstrated. The anchoring surface canbe designed to contain amino, carboxyl, hydroxyl functionalgroups, and so forth, so that different preformed redox com-plexes, enzymes, or polymers can be tethered to the anchoringlayer using conventional coupling reactions. The response ofbiosensors and biofuel cells can be improved under flowingconditions by increasing the rate of mass transport of substrate tothe electrode surface. However, employing flowing conditionsmay increase the rate of desorption, and thus reduce the stability, ofsurface immobilized enzymes. Securely tethered enzymes withinthe protected conductive environment of np-Au can overcome thisbarrier and provide a stable, more sensitive response.

’CONCLUSIONS

np-Au electrodes of differing morphology and surface rough-ness were fabricated and characterized for use in biosensors andbiofuel cells. The suitability of a range of electrochemical probeswas examined to determine (a) the electrochemically addressablesurface area, Areal, and (b) the surface area accessible to mod-ification by a biological macromolecule, Amacro. Measuring thecharge required to strip a Au oxide layer and applying aconversion factor of 390 μC cm�2 was the most accurate andreliable method to determine Areal. Values for Areal ranged from0.251 ((0.013) cm2 [the geometric electrode area, Ageo, was0.246 cm2 giving a roughness factor, Rf, of 1.01 ((0.05)] atplanar Au, to 3.961 ((0.164) cm2 [Rf of 16.1 ((1.3)] and 6.931((0.238) cm2 [Rf of 28.1 ((1.5)] for uniform and nonuniformnp-Au electrodes, respectively. In addition, Cdl and UPD of Cu2+

were identified as suitable complementary methods to estimateAreal. The degree of steric hindrance a redox protein may

experience at a nanostructured electrode was demonstrated usingcyt c as a probe. Rf values of 9.3 ((1.2) and 10.8 ((1.2), whichequated toAmacro values of 2.288 ((0.295) and 2.657 ((0.295) cm2,for uniform and nonuniform np-Au electrodes, respectively, wereconsiderably less than Areal. The difference between Areal andAmacro was particularly large for nonuniform np-Au electrodesdue to the confines of the smaller pores becoming increasinglyinaccessible to cyt c. Surface coverages of electrochemically activecyt c were still, however, multiples of that possible at a planar Ausurface (assuming full monolayer coverage per geometric surfacearea). A comparative study with np-Au film modified GCelectrodes highlighted the importance of a robust electrodedesign as thiol modification of np-Au/GC electrodes, a crucialstep for DET-based biological devices, caused the np-Au film todetach. The presence of a Ti/pure Au adhesion layer for np-Auelectrodes produced via a sputtering method avoided suchcomplications. Initial studies on the use of np-Au as a biocathode,using an osmium redox polymer andMvBOD, demonstrated thatthe diffusion layer thickness exceeded the film thickness, anddiffusion zone overlap occurred between adjacent nanopores(at ν = 5 mV s�1) yielding equivalent charge transfer and currentdensities at planar and np-Au electrodes. Splitting of the ob-served reduction peak indicated that the mediator/enzymecomplex occupied different sites on the surface of the electrode.Future work will involve detailed studies of a range of redoxpolymers and enzymes for use as anodes and cathodes in biofuelcells. The combined advantages of a sheltered environmentwithin a nanopore and novel enzyme/redox polymer immobili-zation strategies at np-Au will be applied to the development ofbioloelectrochemical devices of optimal stability with improvedresponses under flowing conditions.

’ASSOCIATED CONTENT

bS Supporting Information. (Table S1) Summary of thesurface coverages (Γ, pmol cm�2), charges (Q, μC), double layercapacitances, (Cdl, μF cm

�2) and current densities (I, mA cm�2)determined using surface-confined and diffusion-controlled elec-trochemical techniques at each electrode design. (Figure S1)Cross-sectional image of a uniform np-Au electrode indicating thepresence of the pure gold adhesion layer. This material is availablefree of charge via the Internet at http://pubs.acs.org.

’AUTHOR INFORMATION

Corresponding Author*E-mail address: edmond.magner@ul.ie; tel.: +353 61 202629;fax: +353 61 213529.

’ACKNOWLEDGMENT

This work was supported by the European Union FP7 project3D-NanoBioDevice NMP4-SL-2009-229255 and the Programmefor Research in Third Level Institutions (INSPIRE). The assistanceof Dr. C. Dickinson and A. Singh in SEM analysis is acknowledged.

’REFERENCES

(1) Seker, E.; Reed, M.; Begley, M. Materials 2009, 2, 2188–2215.(2) Kertis, F.; Snyder, J.; Govada, L.; Khurshid, S.; Chayen, N.;

Erlebacher, J. JOM 2010, 62, 50–56.(3) Wittstock, A.; Biener, J.; B€aumer, M. Phys. Chem. Chem. Phys.

2010, 12, 12919–12930.

2261 dx.doi.org/10.1021/la202945s |Langmuir 2012, 28, 2251–2261

Langmuir ARTICLE

(4) Hu, K.; Lan, D.; Li, X.; Zhang, S. Anal. Chem. 2008, 80,9124–9130.(5) Shulga, O. V.; Zhou, D.; Demchenko, A. V.; Stine, K. J. Analyst

2008, 133, 319–322.(6) Stevens, G. B.; Reda, T.; Raguse, B. J. Electroanal. Chem. 2002,

526, 125–133.(7) Wittstock, A.; Zielasek, V.; Biener, J.; Friend, C. M.; B€aumer, M.

Science 2010, 327, 319–322.(8) Ding, Y.; Chen, M.; Erlebacher, J. J. Am. Chem. Soc. 2004,

126, 6876–6877.(9) Qiu, H.; Xu, C.; Huang, X.; Ding, Y.; Qu, Y.; Gao, P. J. Phys.

Chem. C 2008, 112, 14781–14785.(10) Qiu, H.; Sun, Y.; Huang, X.; Qu, Y.Colloids Surf., B: Biointerfaces

2010, 79, 304–308.(11) Christenson, A.; Dimcheva, N.; Ferapontova, E. E.; Gorton, L.;

Ruzgas, T.; Stoica, L.; Shleev, S.; Yaropolov, A. I.; Haltrich, D.;Thorneley, R. N. F.; Aust, S. D. Electroanalysis 2004, 16, 1074–1092.(12) Noll, T.; Noll, G. Chem. Soc. Rev. 2011, 40, 3564–3576.(13) L�eger, C.; Bertrand, P. Chem. Rev. 2008, 108, 2379–2438.(14) Marcus, R. A.; Sutin, N. Biochim. Biophys. Acta 1985, 811,

265–322.(15) Hudson, S.; Cooney, J.; Magner, E. Angew. Chem., Int. Ed. 2008,

47, 8582–8594.(16) Scott, R. A., Mauk, A. G., Eds. Cytochrome c, A Multidisciplinary

Approach; University Science Books: Sausalito, CA, 1996.(17) Song, S.; Clark, R. A.; Bowden, E. F.; Tarlov, M. J. J. Phys. Chem.

1993, 97, 6564–6572.(18) Collinson, M.; Bowden, E. F.; Tarlov, M. J. Langmuir 1992,

8, 1247–1250.(19) Christenson, A.; Shleev, S.; Mano, N.; Heller, A.; Gorton, L.

Biochim. Biophys. Acta 2006, 1757, 1634–1641.(20) Mano, N.; Kim, H.-H.; Heller, A. J. Phys. Chem. B 2002,

106, 8842–8848.(21) Jenkins, P. A.; Boland, S.; Kavanagh, P.; Leech, D. Bioelectro-

chemistry 2009, 76, 162–168.(22) Tsujimura, S.; Tatsumi, H.; Ogawa, J.; Shimizu, S.; Kano, K.;

Ikeda, T. J. Electroanal. Chem. 2001, 496, 69–75.(23) Mano, N.; Fernandez, J. L.; Kim, Y.; Shin,W.; Bard, A. J.; Heller,

A. J. Am. Chem. Soc. 2003, 125, 15290–15291.(24) Soukharev, V.; Mano, N.; Heller, A. J. Am. Chem. Soc. 2004,

126, 8368–8369.(25) Sun, Y.; Kucera, K. P.; Burger, S. A.; Balk, T. J. Scr. Mater. 2008,

58, 1018–1021.(26) Ge, X.; Wang, L.; Liu, Z.; Ding, Y. Electroanalysis 2011, 23,

381–386.(27) Kolb, D. M. In Advances in Electrochemical Engineering;

Gerischer, C., Tobias, C. W., Eds.; Wiley-Interscience: New York, 1978;Vol. 11, p 125.(28) Rowe, G. K.; Creager, S. E. Langmuir 1991, 7, 2307–2312.(29) El Kasmi, A.; Wallace, J. M.; Bowden, E. F.; Binet, S. M.;

Linderman, R. J. J. Am. Chem. Soc. 1998, 120, 225–226.(30) Trasatti, S.; Petrii, O. A. Pure Appl. Chem. 1991, 63, 711–734.(31) Leopold, M. C.; Black, J. A.; Bowden, E. F. Langmuir 2002,

18, 978–980.(32) Ng, Y.-F.; Strutwolf, J.; Garratt, P. J.; Williams, D. E. J. Electroanal.

Chem. 1999, 464, 263–267.(33) Hagenstr€om, H.; Schneeweiss, M. A.; Kolb, D. M. Langmuir

1999, 15, 7802–7809.(34) Herzog, G.; Arrigan, D. W. M. TrAC Trends Anal. Chem. 2005,

24, 208–217.(35) Huang, J.-F.; Lin, B.-T. Analyst 2009, 134, 2306–2313.(36) Finklea, H. O.; Avery, S.; Lynch, M.; Furtsch, T. Langmuir

1987, 3, 409–413.(37) Bain, C. D.; Troughton, E. B.; Tao, Y. T.; Evall, J.; Whitesides,

G. M.; Nuzzo, R. G. J. Am. Chem. Soc. 1989, 111, 321–335.(38) Laviron, E. J. Electroanal. Chem. 1979, 101, 19–28.(39) Parida, S.; Kramer, D.; Volkert, C. A.; R€osner, H.; Erlebacher, J.;

Weissm€uller, J. Phys. Rev. Lett. 2006, 97, 035504.

(40) Lee, L. Y. S.; Sutherland, T. C.; Rucareanu, S.; Lennox, R. B.Langmuir 2006, 22, 4438–4444.

(41) Voicu, R.; Ellis, T. H.; Ju, H.; Leech, D. Langmuir 1999, 15,8170–8177.

(42) Zeis, R.; Lei, T.; Sieradzki, K.; Snyder, J.; Erlebacher, J. J. Catal.2008, 253, 132–138.

(43) Szamocki, R.; Reculusa, S.; Ravaine, S.; Bartlett, P. N.; Kuhn, A.;Hempelmann, R. Angew. Chem., Int. Ed. 2006, 45, 1317–1321.

(44) Schr€oper, F.; Br€uggemann, D.; Mourzina, Y.; Wolfrum, B.;Offenh€ausser, A.; Mayer, D. Electrochim. Acta 2008, 53, 6265–6272.

(45) Qiu, H.-J.; Zhou, G.-P.; Ji, G.-L.; Zhang, Y.; Huang, X.-R.; Ding,Y. Colloids Surf., B: Biointerfaces 2009, 69, 105–108.

(46) Park, S.; Chung, T. D.; Kim, H. C. Anal. Chem. 2003, 75,3046–3049.

(47) Jia, F.; Yu, C.; Ai, Z.; Zhang, L.Chem.Mater. 2007, 19, 3648–3653.(48) Davis, K. L.; Drews, B. J.; Yue, H.; Waldeck, D. H.; Knorr, K.;

Clark, R. A. J. Phys. Chem. C 2008, 112, 6571–6576.(49) Bard, A. J.; Faulkner, L. R. Electrochemical Methods, Fundamentals

and Applications, 2nd ed.; John Wiley & Sons: New York, 2001.(50) Clark, R. A.; Bowden, E. F. Langmuir 1997, 13, 559–565.(51) Yue, H.; Waldeck, D. H.; Petrovi�c, J.; Clark, R. A. J. Phys. Chem.

B 2006, 110, 5062–5072.(52) Yue,H.;Waldeck,D.H.; Schrock, K.; Kirby,D.; Knorr, K.; Switzer,

S.; Rosmus, J.; Clark, R. A. J. Phys. Chem. C 2008, 112, 2514–2521.(53) Leopold, M. C.; Bowden, E. F. Langmuir 2002, 18, 2239–2245.(54) Dolidze, T. D.; Rondinini, S.; Vertova, A.; Waldeck, D. H.;

Khoshtariya, D. E. Biopolymers 2007, 87, 68–73.(55) Mano, N.; Kim, H.-H.; Zhang, Y.; Heller, A. J. Am. Chem. Soc.

2002, 124, 6480–6486.(56) Larsson, H.; Sharp, M. J. Electroanal. Chem. 1995, 381, 133–142.(57) Kim, H.-H.; Mano, N.; Zhang, Y.; Heller, A. J. Electrochem. Soc.

2003, 150, A209–A213.(58) Pita, M.; Gutierrez-Sanchez, C.; Olea, D.; Velez, M.; Garcia-

Diego, C.; Shleev, S.; Fernandez, V. M.; De Lacey, A. L. J. Phys. Chem. C2011, 115, 13420–13428.

(59) Boland, S.; Barri�ere, F.; Leech, D. Langmuir 2008, 24, 6351–6358.