Biosensors and Bioelectronics 21 (2005) 257–265

The potential use of hydrazine as an alternative to peroxidase in abiosensor: comparison between hydrazine and HRP-based glucose sensors

Md. Aminur Rahmana, Mi-Sook Wonb, Yoon-Bo Shima,∗a Department of Chemistry and Center for Innovative Bio-Physio Sensor Technology, Pusan National University, Busan 609-735, South Korea

b Korea Basic Science Institute, Keumjeong-ku, Busan 609-735, South Korea

Received 29 July 2004; received in revised form 15 September 2004; accepted 15 September 2004Available online 28 November 2004

Abstract

The potential use of hydrazine sulfate was examined for the catalytic reduction of enzymatically generated H2O2 in a biosensor system. Theperformance of the hydrazine-based sensor was compared with an HRP-based glucose sensor as a model of a biosensor. Hydrazine and HRPwere covalently immobilized onto a conducting polymer layer with glucose oxidase. The direct electron transfer reactions of the immobilizedhydrazine and HRP onto the poly-5,2′:5,2′′-terthiophene-3′-carboxylic acid (poly-TTCA) layer were investigated by using cyclic voltammetricm ntly reducedt wasc perometrict the appliedp eterminedt sb lood andu©

K

1

tmLbwatpp

dro-ffersnces,

hichermi-

lowlec-veryradishpe,

ardu-

i-i

0d

ethod and the electron transfer rate constants were determined. The glucose oxidase- and hydrazine-immobilized sensor efficiehe enzymatically generated H2O2 at −0.15 V versus Ag/AgCl. The surface of this GOx/hydrazine/poly-TTCA-based glucose sensorharacterized by QCM, SEM, and ESCA. Glucose-sensing properties were studied using cyclic voltammetric and chronoamechniques. Various experimental parameters were optimized according to the amount of hydrazine, pH, the temperature, andotential. A linear calibration plot was obtained in the concentration range between 0.1 and 15.0 mM, and the detection limit was d

o be 40.0± 7.0�M. Interferences from other biological compounds were studied. The long-term stability of the GOx/hydrazine sensor waetter than that of the one based on a GOx/HRP biosensor. The proposed glucose sensor was successfully applied to human whole brine samples for the detection of glucose.2004 Elsevier B.V. All rights reserved.

eywords:Amperometry; Conducting polymer; Glucose biosensor; Glucose oxidase; Horseradish peroxidase; Hydrazine

. Introduction

The development of amperometric biosensors has con-inued to receive much attention since the first ampero-etric glucose biosensor was invented in 1962 (Clark and

yons, 1962). To the present time, numerous amperometriciosensors for clinical (Wang, 1999), food (Warsinke, 1997),astewater, and environmental (Rogers, 1998; Lopez-Aviland Hill, 1997) applications have been reported. Many of

hese have relied on hydrogen peroxide detection, which is aroduct of an enzymatic reaction. The detection of hydrogeneroxide can be realized by its oxidation or reduction at a

∗ Corresponding author. Tel.: +82 51 510 2244; fax: +82 51 514 2430.E-mail address:ybshim@pusan.ac.kr (Y.-B. Shim).

solid electrode. However, the anodic determination of hygen peroxide requires a high oxidation potential and sufrom possible interferences by other oxidizable substasuch as ascorbic acid, uric acid, and acetaminophen, ware usually present in real samples. The cathodic detnation of hydrogen peroxide using a catalyst requires apotential. At this low potential, interfering species are etrochemically inactive and thus, this method provides aattractive biosensor system. The catalyst, such as horseperoxidase (HRP) (Ruzgas et al., 1996; Guo and Guadalu1997; Bartlett and Astier, 2000; Yang and Mu, 1997; Killet al., 1999; Tatsuma et al., 2001), mycobacterium tuberclosis KatG (Chouchane et al., 2000; Zhang et al., 2001), cy-tochromecperoxidase (Everse et al., 1991), soybean peroxdase and catalase (Zhang et al., 2002), myoglobin (Yonetan

956-5663/$ – see front matter © 2004 Elsevier B.V. All rights reserved.oi:10.1016/j.bios.2004.09.036

258 Md.A. Rahman et al. / Biosensors and Bioelectronics 21 (2005) 257–265

and Schleyer, 1967), and haemoglobin (Ortiz de Montellanoand Catalano, 1985) can catalyse the reduction of hydrogenperoxide at a low applied potential. Among these catalysts,HRP has been extensively used to catalyse hydrogen perox-ide. However, the enzymes including HRP could not give thesensor a long-term stability due to their inherent instability,which causes denaturation (Wang et al., 2004). Most notably,in a bi-enzyme based biosensor, the optimal operating con-ditions of two enzymes might be different, which do not givethe sensor long-term stability. Moreover, enzymes and pro-teins are expensive. Thus, the search for substitutes for HRPand proteins to improve performance with enhanced sensitiv-ity and stability, or to reduce the cost of a biosensor, wouldbe more attractive.

To determine a suitable substitute for HRP in a biosensor,the possible use of hydrazine has been explored. Hydrazine isa strong reducing agent. It is widely used in industry, agricul-ture, and medicine. It was historically used as a therapeuticagent in the treatment of tuberculosis, sickle cell anemia, andnon-specific chronic illnesses (Von-Burg and Stout, 1991).Hydrazine is found to be carcinogenic, and has cytotoxiceffects on hepatocyte cell cultures (Noda et al., 1987). How-ever, hydrazine sulfate is an anti-cachexia drug, which actsto reverse the metabolic processes of debilitation and weightloss in cancer, and secondarily acts to stabilize or regresstumors (Gold, 1981). Previously, hydrazine has been usedt tingt co-p oseod -e ty ofh

study( o-g inaryr lizedw oxi-d xan-t nicali n ofa ucoses sul-f r. Toc arriedo ta n-da ngp thecH paret e sen-s sen-s fromo ine-

based glucose sensor and finally applied to the human wholeblood and urine samples for the detection of glucose.

2. Experimental

2.1. Materials

Glucose oxidase, type VII-S (EC. 1.1.3.4, 196.0 units g−1)from Aspergillus nigerand horseradish peroxidase, type II(EC.1.11.1.7, 158 units mg−1), were purchased from SigmaCo. (USA). Hydrazine sulfate was obtained from Aldrich,USA. 1-Ethyl-3-(3-(dimethylamino)-propyl) carbodiimide(EDC) and dichloromethane (99.8%, anhydrous, sealedunder N2 gas), were received from Sigma and were usedas received. Tetrabutylammonium perchlorate (TBAP,electrochemical grade) was received from Fluka (USA),purified according to the general method (Shim and Park,1997) and then dried under vacuum at 1.33× 10−3 Pa. Aterthiophene monomer bearing a carboxylic acid group,5,2′:5,2′′-terthiophene-3′-carboxylic acid (TTCA), wassynthesized according to a previous report (Lee et al., 2002).H2O2 and�-glucose were obtained from Sigma Co. A phos-phate buffer solution was prepared by modifying 0.1 M ofdisodium hydrogen phosphate (Aldrich) with the admixtureof 0.1 M of sodium dihydrogen phosphate (Aldrich). Allo usedw erep roma

2

gp( ing,r yclicv usingp (S.K entw 17a A).Aq EMi 240.E tificE AlK

2

rode(m lledw lec-t M

o introduce reactive group for cross-linking by substituhe amide groups of the acrylamide and vinylimidazoleolymers into hydrazide groups in the fabrication of glucxidase, lactate oxidase, or HRP-based biosensors (Thierrye Lumley-Woodyear et al., 1995). However, to our knowldge, there has been no report on the catalytic activiydrazine in a biosensor system.

In the present study, we have extended our previousShankaran and Shim, 2002) on a hydrazine-based hydren peroxide sensor to the biosensor systems. Prelimesults showed that hydrazine sulfate could be immobiith glucose oxidase, glutamate oxidase, and xanthinease for the fabrications of the glucose, glutamate, and

hine sensors. Due to the fundamental, industrial, and climportance of glucose, the detail study for the fabricatio

glucose sensor was undertaken as a model. The glensor has been fabricated by immobilizing hydrazineate with glucose oxidase on a conducting polymer layeonstruct the sensor, we used a similar approach as cut in our previous work on choline biosensor (Rahman el., 2004), which involves, initially, the formation of a coucting polymer film by the potential cycling method (Leend Shim, 2001; Lee et al., 2002). We have used conductiolymer to make firm immobilization of enzyme throughovalent bond formation. We have also prepared a GOx andRP bi-enzyme-immobilized glucose biosensor to com

he sensing properties with the hydrazine-based glucosor. The experimental parameters, which can affect theor responses were optimized. The interference effectsther biological compounds were studied for the hydraz

ther chemicals were of extra pure analytical grade andithout further purification. All aqueous solutions wrepared in doubly distilled water, which was obtained fMilli-Q water purifying system (18 M� cm).

.2. Apparatus

Hydrazine sulfate and GOx-immobilized conductinolymer modified glassy carbon (area = 7 mm2), Ag/AgClin saturated KCl), and a Pt wire were used as workeference, and counter electrodes, respectively. Coltammograms and amperograms were recordedotentiostat/galvanostat, Kosentech model KST-P2orea). A quartz crystal microbalance (QCM) experimas performed using a SEIKO EG&G model QCA 9nd a PAR model 263A potentiostat/galvanostat (USn Au working electrode (area: 0.196 cm2; 9 MHz; AT-cutuartz crystal) was used for the QCM experiment. S

mages were obtained using a Cambridge StereoscanSCA experiments were performed using a VG ScienSCALAB 250 XPS spectrometer with monochromated� source with charge compensation at KBSI (Busan).

.3. Electropolymerization of the TTCA monomer

Prior to the polymerization, the glassy carbon electGCE) was polished to a mirror finish with 0.5�m of alu-ina slurry on a polishing cloth, and then rinsed with distiater. The poly-TTCA layer on the GCE was made by e

ropolymerization of 1.0 mM of TTCA monomer in a 0.1

Md.A. Rahman et al. / Biosensors and Bioelectronics 21 (2005) 257–265 259

TBAP/CH2Cl2 solution by cycling the potential between 0and 1.6 V five times at 100 mV s−1 (Lee and Shim, 2001;Lee et al., 2002). After the polymerization, the electrode waswashed with CH2Cl2 to remove the excess monomer.

2.4. Immobilizations of hydrazine and enzymes on thepolymer layer

Two GCEs coated with the poly-TTCA layer were used toimmobilize hydrazine/GOx and HRP/GOx, separately, usingthe procedure developed previously (Rahman et al., 2004). Atfirst, the poly-TTCA coated electrode (CPME) was immersedfor 6 h in a 0.1 M phosphate buffer solution (pH 7.0) contain-ing 10.0 mM of EDC to activate the carboxylic groups of thepoly-TTCA layer. Then, the CPME treated with EDC waswashed with a buffer solution and subsequently incubatedfor 6 h in a 0.1 M phosphate buffer solution (pH = 7.0) con-taining 0.5 mM of hydrazine sulfate at 4◦C. In this way, hy-drazine was immobilized onto the poly-TTCA layer throughthe formation of covalent bonds between the carboxylicacid groups of the polymer and the amine groups of hy-drazine. Glucose oxidase (GOx) was then successively im-mobilized onto the hydrazine/poly-TTCA layer by incubatingthe hydrazine/poly-TTCA electrode for 12 h in a 0.1 M phos-phate buffer solution (pH = 6.4) containing 0.5 mM of GOx

a Pa o-b Ef on-t lyi ofG s de-s e-a

2

O/+ teb etrice ntialoag ared9 cedi coses ithGs ring.C withc ntso erec .U

3. Results and discussion

3.1. Electrochemical behaviours of immobilizedhydrazine and enzymes on the poly-TTCA layer

The poly-TTCA layer was made by the electropoly-merization of the TTCA monomer on a GCE in a 0.1 MTBAP/CH2Cl2 solution. At the first anodic scan, the CVexhibited one oxidation peak at around 1.3 V, due to themonomer oxidation, to immediately form the polymer. Apolymer reduction peak appeared at around 1.1 V in the re-verse cathodic scan. The peak currents of the polymer filmat 1.1 and 1.3 V increased as the potential cycle numberincreased. SEM characterized the morphology of the poly-TTCA film formed on the electrode surface. The SEM pic-ture showed that the thickness of the polymer film was about300 nm (Lee et al., 2002).

Fig. 1A shows the CVs recorded for the hydrazine/poly-TTCA (dashed line) and poly-TTCA (solid line) modifiedelectrodes in a 0.1 M phosphate buffer solution of pH 7.0.A cyclic voltammogram also recorded for the poly-TTCAmodified electrode in a 0.1 mM of hydrazine solution (dottedline). A pair of redox peaks was observed at +0.05/+0.018 Vversus Ag/AgCl. The peak separation of the peaks was mea-sured to be 0.032 V, indicating the involvement of a two-electron transfer reaction. When the CV was recorded with

Fig. 1. (A) CVs recorded for the poly-TTCA (—) and hydrazine/poly-TTCA(- - - -) modified electrodes in a phosphate buffer solution and poly-TTCA(·····) electrode in a 0.1 mM hydrazine sulfate solution. (B) CVs recordedfor the poly-TTCA (—) and HRP/poly-TTCA (- - - -) modified electrodes ina phosphate buffer solution. The scan rate was 0.05 V/s.

t 4◦C (hereafter GOx/hydrazine/poly-TTCA). For the HRnd GOx bi-enzyme immobilization, HRP was first immilized onto the poly-TTCA layer by incubating the CPM

or 6 h in a 0.1 M phosphate buffer solution (pH = 7.0) caining 0.5 mM of HRP at 4◦C. GOx was then successivemmobilized onto the HRP/poly-TTCA layer from 0.5 mMOx in a 0.1 M phosphate buffer by the same process a

cribed for the GOx/hydrazine/poly-TTCA electrode (herfter GOx/HRP/poly-TTCA).

.5. Electrochemical measurement

Cyclic voltammograms were recorded for Gxhydrazine/poly-TTCA and GOx/HRP/poly-TTCA from0.25 to −0.35 V versus Ag/AgCl in a 0.1 M phosphauffer solution of pH 7.0, separately. Chronoamperomxperiments were carried out by applying the potef −0.15 V at GOx/hydrazine/poly-TTCA and−0.10 Vt GOx/HRP/poly-TTCA, separately to reduce the H2O2enerated from the enzymatic reaction. A freshly prep.0 ml of 0.1 M sodium phosphate buffer solution was pla

nto the cell and the steady-state current on the first gluample addition (1.0 ml of 0.5 mM) was monitored wOx/hydrazine/poly-TTCA and GOx/HRP/poly-TTCA,eparately, at the optimal pH and temperature, with stirhronoamperometric experiments were continuedonsecutive injections into the cell of varying amouf �-glucose. All electrochemical measurements warried out in a thermostat bath at 22◦C (PolyScience CoSA).

260 Md.A. Rahman et al. / Biosensors and Bioelectronics 21 (2005) 257–265

the hydrazine/poly-TTCA modified electrode in a phosphatebuffer solution, a pair of redox peaks at +0.084/+0.015 V ver-sus Ag/AgCl was also observed with a larger peak separation.

This redox peak was not observed when the CV wasrecorded for the poly-TTCA modified electrode in a blanksolution. This indicates that this redox peak was come solelyfrom the adsorbed hydrazine onto the poly-TTCA surface,which might have been related to the electron transfer reac-tion between the hydrazo and azo groups (Bard, 1974). Whenthe concentration of hydrazine in the immobilizing solutionincreased from 0.1 to 0.5 mM, the peak current of this redoxpeak also increased.Fig. 1B shows the CVs recorded for theHRP/poly-TTCA modified electrodes in a phosphate buffersolution of pH 7.0. A pair of redox peaks was observed at−0.014/+0.104 V versus Ag/AgCl, which was resulted fromthe redox process of immobilized HRP onto poly-TTCA.The peak separation of the peaks was measured to be about0.118 V. This large peak separation is probably ascribable tothe immobilized HRP molecules in various orientations (Liand Dong, 1997). For the both redox processes of immo-bilized hydrazine and HRP onto the poly-TTCA, the peakcurrents were directly proportional to the scan rate up to the0.5 V/s, indicating that the electrode reactions involved thesurface confined processes (Murray, 1984). The peak separa-tions of the both redox couples increased with the increase ofscan rate up to 0.5 V/s, indicating that the both redox reactionsw tants( de-t hem-i tob , att tt stert orh r re-p eta

3h

thea -TT e,t afte1 ont m-p ntopa re-p 3T de-t nto er,t mo-

bilization of GOx onto the hydrazine/poly-TTCA film, thefrequency decrease attained a steady-state within 1 h, indicat-ing that the immobilization of GOx onto the hydrazine/poly-TTCA layer completed within 1 h. The�f value of GOximmobilized onto the hydrazine/poly-TTCA layer after 1.0 hwas 0.60 kHz. The amount of GOx on the hydrazine/poly-TTCA layer was determined to be 0.66± 0.08�g, which wassimilar to the amount of hydrazine immobilized onto the poly-TTCA film. The frequency change during the immobilizationof HRP onto the poly-TTCA and during the immobilizationof GOx onto the HRP/poly-TTCA layers. From the frequencyshift, the amount of HRP after 1 h (�f= 0.52 kHz) was de-termined to be 0.57± 0.07�g, and that of GOx after 2 h(�f= 0.45 kHz) was determined to be 0.50± 0.09�g. Thenumber of molecules per area for hydrazine and HRP onto thepoly-TTCA layer, and that of GOx onto the hydrazine/poly-TTCA and HRP/poly-TTCA were determined. Thevalues were found to be 1.21± 0.3× 10−11 mol cm−2 and4.92± 0.41× 10−9 mol cm−2, respectively. The surfaceconcentration of hydrazine molecule was two orders magni-tude higher than that of HRP molecules. The number of GOx

molecule onto the hydrazine/poly-TTCA and HRP/poly-TTCA were found to be 4.12± 0.8× 10−12 mol cm−2 and3.12± 0.55× 10−12 mol cm−2, respectively. These values ofthe GOx surface concentration were comparable to the otherliterature values obtained from QCM studies (Blonder et al.,1

3G

-t sooT -T ei es-e pes,w thep atesm inem ofGa ob-t Al -T dG CAl

(fa singt ectraw hib-

ere quasi-reversible. The electron transfer rate consKs) for the redox reactions of hydrazine and HRP wereermined using the method for a surface bound electroccal system (Laviron, 1979). The values were calculatede 2.08 and 0.98 s−1 for hydrazine and HRP, respectively

he scan rate of 0.05 V/s. The higherKs value indicates thahe electron transfer rate for hydrazine is two times fahan that for HRP. TheKs value obtained in this study fydrazine is also very high when compared to the otheortedKs values for HRP (Ruzgas et al., 1995; Tatsumal., 1998).

.2. Determination of the amount of immobilizedydrazine and enzymes

QCM studies were carried out in order to determinemount of hydrazine and GOx on the GOx/hydrazine/polyTCA electrode and of HRP and GOx on the GOx/HRP/poly-TCA electrode. During the immobilization of hydrazin

he frequency decreased and attained a steady-state.5 h. This indicates that the immobilization of hydrazine

he poly-TTCA layer completed within 1.5 h at room teerature. After a 30 min immobilization of hydrazine ooly-TTCA layer, the frequency change (�f) was 0.58 kHznd mass change (�m) was determined using the equationorted previously (Lee and Shim, 2001; Rahman et al., 200).he amount of hydrazine immobilized after 30 min was

ermined to be 0.64± 0.08�g. Then, to determine the amouf GOx immobilized onto the hydrazine/poly-TTCA lay

he frequency change was again measured. During the im

r

998; Gooding et al., 1999).

.3. Surface characterization of theOx/hydrazine/poly-TTCA electrode by SEM and ESCA

The morphology of the GOx/hydrazine/poly-TTCA elecrode was characterized by SEM.Fig. 2shows SEM imagef the poly-TTCA layer on GCE (Fig. 2A), of hydrazinen the poly-TTCA layer (Fig. 2B), of GOx on the poly-TCA layer (Fig. 2C), and of GOx on the hydrazine/polyTCA layer (Fig. 2D). The SEM image of hydrazin

mmobilized onto the poly-TTCA layer shows the prnce of some small aggregates with different shahich were not observed in the SEM image foroly-TTCA layer. The presence of these small aggregight have been due to the immobilization of hydrazolecules on the poly-TTCA layer. The SEM imageOx immobilized onto the poly-TTCA layer (Fig. 2C)lso shows a very different morphology from that

ained for hydrazine immobilized onto the poly-TTCayer. From the SEM image of the GOx/hydrazine/polyTCA surface (Fig. 2D), it is clear that hydrazine anOx successfully attached to the surface of the poly-TT

ayer.Fig. 3 shows the ESCA spectra of C1s (Fig. 3A), of N1s

Fig. 3B), of O1s (Fig. 3C), and of S2p (Fig. 3D) peaksor the poly-TTCA, GOx/poly-TTCA, hydrazine/poly-TTCAnd GOx/hydrazine/poly-TTCA surfaces. Before asses

he peaks to the specific functional groups, all the spere deconvoluted. The C1s spectrum in poly-TTCA ex

Md.A. Rahman et al. / Biosensors and Bioelectronics 21 (2005) 257–265 261

Fig. 2. SEM images of (A) poly-TTCA, (B) hydrazine onto the poly-TTCA layer, (C) GOx on to poly-TTCA layer, and (D) hydrazine and GOx on to poly-TTCAlayer.

ited two peaks at 284.0 and 289.2 eV, which correspondedto the C H, C S, or C C bonds and COOH bond, respec-tively (Ng et al., 1998). An additional peak at about 286.0 eVwas observed after immobilization of hydrazine or GOx ontothe poly-TTCA layer. The peak at 284.0 eV slightly shiftedto a higher energy due to the formation of a covalent bondbetween the carboxalic acid groups in poly-TTCA and theamine groups in hydrazine or GOx. The peak observed at286.0 eV in hydrazine/poly-TTCA, GOx/poly-TTCA, andGOx/hydrazine/poly-TTCA corresponded to CN bond (Ng

F S2pp dG

et al., 1998). This indicates that hydrazine and GOx suc-cessfully immobilized onto the surface of the poly-TTCAlayer. The N1s spectrum in poly-TTCA exhibited no peakindicating the absence of theNH2 group. However, theN1s spectrum exhibited a clear peak at about 399.7 eV af-ter the immobilizations of GOx, hydrazine, and hydrazineand GOx. This peak at about 399.7 eV corresponded to the

N C bond (Li et al., 2004). The atomic percent of C1s peakwas found to be 68.97, 47.10, 47.14, and 47.32 and that ofthe N1s peak was found to be 0, 6.226, 4.914, and 9.378 inpoly-TTCA, hydrazine/poly-TTCA, GOx/poly-TTCA, andGOx/hydrazine poly-TTCA, respectively. To confirm the im-mobilisation of hydrazine and enzymes, the ratios of theatomic percent of N1s and C1s were determined. The val-ues of N/C atomic ratios in hydrazine/poly-TTCA (0.132),GOx/poly-TTCA (0.105), and GOx/hydrazine/poly-TTCA(0.20) were larger than that in poly-TTCA (0). This indi-cates that hydrazine and GOx successfully immobilized ontothe poly-TTCA layer. The intensity of the N1s peak in thecase of GOx/hydrazine/poly-TTCA was higher than that ofGOx/poly-TTCA and hydrazine/poly-TTCA. This was dueto the fact that in GOx/hydrazine/poly-TTCA, the amountof the N C group was higher than in GOx/poly-TTCA orhydrazine/poly-TTCA. The O1s spectrum in poly-TTCA ex-hibited two peaks at 531.5 and 532.5 eV, which correspondedto C O and C O, respectively (Ng et al., 1998). The peaka obi-l edoa

ig. 3. ESCA spectra of (A) C1s, (B) N1s, (C) O1s, and (D)eak for poly-TTCA, GOx/poly-TTCA, hydrazine /poly-TTCA, anOx/hydrazine/poly-TTCA.

t 532.5 slightly shifted to a higher energy after the immization of hydrazine and GOx. The S2p spectrum exhibitne peak at about 164.0 eV corresponding to CS bonds inll cases (Bernede et al., 1997).

262 Md.A. Rahman et al. / Biosensors and Bioelectronics 21 (2005) 257–265

3.4. Sensing properties of theGOx/hydrazine/poly-TTCA modified electrode

Fig. 4A shows the CVs recorded between +0.25 and−0.35 V versus Ag/AgCl for the GOx/hydrazine/poly-TTCA-modified electrode in a 0.1 M phosphate buffer so-lution without (solid line) or with a (dashed line) 0.5 mM�-glucose solution. No peak was observed when CVs wererecorded for poly-TTCA in the blank (dotted line) or glucosecontaining solution (figure not shown), due to the absence of�-glucose and GOx, respectively. When the CV was recordedfor GOx/hydrazine/poly-TTCA in a blank electrolyte solu-tion, no catalytic current was observed due to the absence ofglucose. A clear reduction peak was observed at−0.15 V ver-sus Ag/AgCl for GOx/hydrazine/poly-TTCA when the CVwas recorded in a 0.5 mM�-glucose solution. This means thatthe presence of GOx on the electrode surface generated H2O2through an enzymatic reaction, and that hydrazine on the elec-trode surface catalytically reduced H2O2. The cathodic peakat−0.15 V was proportional to the�-glucose concentration.To confirm whether this catalytic peak at−0.15 V came fromthe electroreduction of H2O2 and not from the O2 reductionin the air, the CV was recorded for the poly-TTCA electrodein the blank solution between +0.25 and−0.75 V. A cathodic

FbppGai7

peak appeared at−0.40 V, which gradually decreased whennitrogen gas was introduced into the electrochemical cell.This means that this peak at−0.40 V corresponded to the oxy-gen reduction. The catalytic reduction of H2O2 at −0.15 Vwas also confirmed with an authentic H2O2 sample on thehydrazine/poly-TTCA modified electrode. The catalytic re-duction of H2O2 generated through the enzymatic reaction in�-glucose at−0.15 V by GOx/hydrazine/poly-TTCA is veryeffective, because at this low potential, glucose sensing is lessprone to interferences from other biological species.

The sensing property of the GOx/hydrazine/poly-TTCAelectrode was compared with that of the GOx/HRP/poly-TTCA electrode.Fig. 4B shows the CVs recorded forGOx/HRP/poly-TTCA in a 0.1 M phosphate buffer without(solid line) and with (dashed line) 0.5 mM of�-glucose.The catalytic reduction of H2O2 by HRP was observedat −0.10 V, and the peak current was proportional to theglucose concentration. The cathodic peak current of theH2O2 reduction by GOx/hydrazine/poly-TTCA was higherthan that by GOx/HRP/poly-TTCA, indicating that theGOx/hydrazine/poly-TTCA electrode was more sensitivethan the GOx/HRP/poly-TTCA electrode.

3.5. Optimization of experimental parameters

The GO/poly-TTCA modified electrode with variousa diedi fort y-m neas ithG entra-

ig. 4. CVs recorded for (A) poly-TTCA coated GCE in a phosphateuffer solution (·····), GOx/hydrazine/poly-TTCA electrode in a phos-hate buffer (—), GOx/hydrazine/poly-TTCA electrode (- - - -) in a phos-hate buffer solution containing 0.5 mM�-glucose. (B) CVs recorded forOx/HRP/poly-TTCA electrode in a phosphate buffer solution (solid line)nd GOx/HRP/poly-TTCA electrode in a phosphate buffer solution contain-

ng 0.5 mM�-glucose. In both cases, pH of the solution was maintained at.0.

F hy-d ed forG

x

mounts of hydrazine in molar concentration was stun order to determine the optimum hydrazine quantityhe electroreduction of H2O2 generated through the enzatic reaction.Fig. 5A shows the effect of the hydrazimount on the cathodic peak current of 0.5 mM�-glucoseolution, which was deposited on the poly-TTCA layer wOx. The peak currents increased as the hydrazine conc

ig. 5. Effects of (A) hydrazine concentration, (B) deposition time ofrazine, (C) pH, and (D) temperature on the CV peak currents obtainOx/hydrazine/poly-TTCA electrode in 0.5 mM�-glucose.

Md.A. Rahman et al. / Biosensors and Bioelectronics 21 (2005) 257–265 263

tion increased from 0.1 to 0.6 mM. An increase of hydrazinequantity would improve the efficiency of the electrocatalyticreduction of H2O2. However, the further increase of the con-centration of hydrazine from 0.6 mM did not result in a notice-able change in the peak currents for the reduction of H2O2.Therefore, the concentration of hydrazine to be immobilizedonto the poly-TTCA layer was selected as 0.6 mM.

The effect of the deposition time on the electroreductionof H2O2 in �-glucose was studied between 1 and 15 h, andis shown inFig. 5B. The GOx/poly-TTCA electrode was im-mersed in a 0.6 mM of hydrazine solution at each depositiontime at 4◦C. The peak currents increased linearly as the de-position time increased from 1 to 6 h, indicating an enhance-ment of the uptake of hydrazine onto the GOx/poly-TTCAelectrode surface for a longer deposition time. The increasein the peak current continued until a constant current level wasattained due to the saturation of the active sites of hydrazinemolecules on the electrode surface. The peak currents werealmost constant for deposition times over 6 h. Therefore, theoptimum deposition time was chosen as 6 h.

The effect of the pH of the medium on the electrocatalyticreduction of H2O2 with GOx/hydrazine/poly-TTCA wasstudied over the pH range of 4.0–10.0.Fig. 5C shows the peakcurrent obtained at different pH. The peak height increasedfrom pH 4.0 to 7.0, and then decreased over pH 7.0. Although,hydrazine can be use for HO sensing in a basic medium,t beend ,t oo 0.

duc-t ew e oft Ho pera-t asea td from3 ost5o

tricc ion ofH sei oren ed at− ntialu us,t at− nts.

3

-t thep orbic

acid, dopamine, uric acid, acetaminophen, andl-cysteine. Noamperometric response was observed when 2.0 mM of otherbiological compounds were individually injected to the buffersolution. For confirmation, distilled water was also injectedbefore and after the individual injection of other biologicalcompounds. The oxidations of the above compounds did notoccur at the potential used (−0.15 V) in the amperometricexperiments with GOx/hydrazine/poly-TTCA electrode.

3.7. Calibration plots ofβ-glucose withGOx/hydrazine/poly-TTCA and GOx/HRP/poly-TTCA

Fig. 6A shows the typical current–time plots for theaddition of varying amounts of�-glucose in a 0.1 M phos-phate buffer solution. The applied potentials were separatelyset at−0.15 and−0.10 for the electroreduction of H2O2by GOx/hydrazine/poly-TTCA and GOx/HRP/poly-TTCAelectrodes, respectively. The catalytic reduction currents rosesteeply to a stable value as soon as the potential was applied.Both GOx/hydrazine/poly-TTCA and GOx/HRP/poly-TTCAelectrodes achieved 95% of steady state currents within 10 s.Fig. 6B shows the calibration plots for the glucose determi-nation, separately obtained from the amperometric responseswith GOx/hydrazine/poly-TTCA and GOx/HRP/poly-TTCAelectrodes. Under the optimized condition, the steady-statec ose

Fig. 6. (A) Current–time plots obtained with GOx/hydrazine/poly-TTCA(– – –) and GOx/HRP/poly-TTCA (- - - -) upon the addition of varyingamounts of �-glucose in the 0.1 M phosphate buffer solution at pH7.0. (B) Calibration plots for GOx/hydrazine/poly-TTCA (solid line) andGOx/HRP/poly-TTCA (dash line).

2 2he current decrease over pH 7.0 in this work might haveue to the poor enzyme activity of GOx at higher pH. Hence

he optimum pH for GOx/hydrazine/poly-TTCA in order tbtain the maximum response was determined to be 7.

The effect of the temperature on the electrocatalytic reion of H2O2 with the GOx/hydrazine/poly-TTCA electrodas studied.Fig. 5D shows the temperature dependenc

he current response in a 0.5 mM of�-glucose solution at a pf 7.0. The peak current gradually increased as the tem

ure increased from 15 to 30◦C, and then began to decres the temperature increased over 40◦C. The peak currenecreased about 25% when the temperature increased0 to 40◦C. Above 40◦C the peak current decreased alm0% because of the thermal inactivation of GOx. Thus, theptimum temperature was considered to be 30◦C.

The effect of the applied potential on the amperomeurrent response was also studied for the electroreduct2O2 on GOx/hydrazine/poly-TTCA. The current respon

ncreased as the applied potential went from 0.0 V to a megative potential. The maximum response was attain0.15 V and the application of the more negative potep to −0.30 V did not improve the current response. Th

he GOx/hydrazine/poly-TTCA electrode was polarized0.15 V versus Ag/AgCl in the amperometric experime

.6. Interference effects

The selectivity of the GOx/hydrazine/poly-TTCA elecrode for�-glucose was evaluated amperometrically inresence of other biological compounds, such as asc

urrents showed a linear relationship with the gluc

264 Md.A. Rahman et al. / Biosensors and Bioelectronics 21 (2005) 257–265

concentration in the range from 0.1 to 15.0 mM for bothsensors, although the GOx/hydrazine/poly-TTCA electroderesponded to 0.05 mM of�-glucose. The relative standarddeviations (RSD) at 0.1 mM of glucose were found tobe 5.4% and 6.1% with GOx/hydrazine/poly-TTCA andGOx/HRP/poly-TTCA electrodes, respectively. These lineardependences of the glucose concentration yielded the regres-sion equations ofip (�A) = (9.16± 1.72) + (6.86±0.22) [C](mM)) andip (�A) = (7.07± 1.37) + (4.76± 0.18) [C] (mM)for the GOx/hydrazine/poly-TTCA and GOx/HRP/poly-TTCA electrodes, respectively, with correlation coefficientsof 0.996 and 0.995. The detection limits of glucose weredetermined to be about 40.0± 7.0 and 90.0± 5.0�M byGOx/hydrazine/poly-TTCA and GOx/HRP/poly-TTCAelectrodes, respectively, based on three-times measurementfor the standard deviation of the blank noise (95% confidencelevel,k= 3,n= 5).

The analytical performance of the GOx/hydrazine/poly-TTCA sensor was compared with the earlier reported glucosebiosensors. The linear dynamic range for the detection ofglucose obtained with GOx/hydrazine/poly-TTCA was widerthan other reported works (Qian et al., 2004; Fang et al., 2003;Watanabe et al., 1988; Okawa et al., 1989). The detection limitof the GOx/hydrazine/poly-TTCA sensor was also lower thanthe other recently reported values (Qian et al., 2004; Fang etal., 2003).

3

nsesow howni shedw hateb ta-b boths itials asedf tot -T and

F e)a

it decreased by approximately 20% from its initial value after75 days. The above results clearly suggested that the sensorbased on GOx/hydrazine/poly-TTCA has a longer useabletime than the sensor based on GOx/HRP/poly-TTCA.

3.9. Real sample analyses

The proposed GOx/hydrazine/poly-TTCA sensor wastested to detect the glucose concentration in the healthy hu-man whole blood and in the human urine samples. The bloodand urine samples were carefully collected immediately be-fore the measurements. The blood samples were diluted 10times, whereas the urine samples were used as it after fil-tration. The standard addition method was followed and theglucose concentration in the samples was determined froma standard addition curve. The glucose concentrations in thewhole blood and urine samples were found to be 5.95± 0.5and 1.38± 0.02 mM, respectively, which are within the limitsfor normal physiological blood and urine glucose levels. Theglucose concentrations measured in the blood and urine sam-ples by our method were compared with that measured by abiochemical analyzer (Shimadzu 7200). The results obtainedby our method showed a good agreement with the resultsobtained by the biochemical analyzer.

4

usedw ses lm.T rizedu -Tig ngp wasc nd ite tionl tory( ivity.T ppliedt ands

A

J3-P area fullya

R

B , vol.

.8. Long-term storage stability

To examine the long-term storage stabilities, the respof GOx/hydrazine/poly-TTCA and GOx/HRP/poly-TTCAere examined with respect to the storage time, and are s

n Fig. 7. After each experiment, the electrodes were waith a phosphate buffer solution and stored in a phospuffer solution of pH 7.0 at 4◦C. The long-term storage silities of these electrodes were tested every 5 days. Forensors, up to 50 days, the sensitivity retained 98% of inensitivity. After 50 days, the response gradually decreor GOx/HRP/poly-TTCA, which might have been duehe loss of GOx and HRP activity. For GOx/hydrazine/polyTCA, the sensitivity retained almost 98% up to 75 days,

ig. 7. Effect of storage time of GOx/hydrazine/poly-TTCA electrode (cubnd GOx/HRP/poly-TTCA electrode (triangle).

. Conclusions

Hydrazine sulfate instead of HRP was successfullyith GOx in the fabrication of an amperometric glucoensor by immobilizing it onto a conducting polymer fihe surface of this new glucose sensor was charactesing QCM, ESCA, and SEM. The GOx/hydrazine/polyTCA sensor showed an electrocatalytic reduction of H2O2

n which H2O2 was enzymatically produced by GOx in alucose solution, at−0.15 V versus Ag/AgCl. The sensierformance of this sensor in the glucose detectionompared with an HRP-based glucose biosensor, axhibited a slight wider dynamic range and a lower detec

imit. The long-term stability of the sensor was satisfacabout 2 months) and it showed an excellent selecthe developed hydrazine-based glucose sensor was a

o the real sample (human blood and urine) analysesatisfactory results were obtained.

cknowledgments

The financial supports for this work (grants no. 02-PG6-EV05-0001) from the Ministry of Health and Welfnd Center for Integrated Molecular Systems are gratecknowledged.

eferences

ard, A.J., 1974. Encyclopedia of Electrochemistry of the Elements13. Marcel Dekker, New York, p. 174.

Md.A. Rahman et al. / Biosensors and Bioelectronics 21 (2005) 257–265 265

Bartlett, P.N., Astier, Y., 2000. Microelectrochemical enzyme transistors.Chem. Commun., 105–112.

Bernede, J.C., Tregouet, Y., Gourmelon, E., Martinez, F., Neculqueo, G.,1997. On the degradation of some thiophene oligomers after dopingby ion chloride. Polym. Degrad. Stab. 55, 55–64.

Blonder, R., Wilner, I., Buckmann, A.F., 1998. Reconstitution of apo-glucose oxidase on nitrospiropyran and FAD mixed monolayers ongold electrodes: photostimulation of bioelectrocatalytic features of thebiocatalyst. J. Am. Chem. Soc. 120, 9335–9341.

Chouchane, S., Lippai, I., Magliozzo, R.S., 2000. Catalase-peroxidase(Mycobacterium tuberculosisKatG) catalysis and isoniazid activation.Biochemistry 39, 9975–9983.

Clark Jr., L.C., Lyons, C., 1962. Electrode systems for continuous mon-itoring in cardiovascular surgery. Ann. N.Y. Acad. Sci. 102, 29–45.

Everse, J., Everse, K.E., Grisham, M.B., 1991. Peroxidases in Chemistryand Biology, vol. II. CRC Press, Boca Raton, FL.

Fang, A., Ng, H.T., Li, S.F.Y., 2003. A high-performance glucose biosen-sor based on monomolecular layer of glucose oxidase covalentlyimmobilised on indium-tin oxide surface. Biosens. Bioelectron. 19,43–49.

Gold, J., 1981. Anabolic profiles in late-stage cancer patients responsiveto hydrazine sulfate. Nutr. Cancer. 3, 13–19.

Gooding, J.J., Situmorang, M., Erokhin, P., Hibbert, D.B., 1999. An assayfor the determination of the amount of glucose oxidase immobilisedin an enzyme electrode. Anal. Commun. 36, 225–228.

Guo, Y., Guadalupe, A.R., 1997. Direct electrochemistry of horseradishperoxidase adsorbed on glassy carbon electrode from organic solu-tions. Chem. Commun., 1437–1438.

Killard, A.J., Zhang, S., Zhao, H., John, R., Iwuoha, E.I., Smith, M.R.,1999. Development of an electrochemical flow injection immunoassay(FIIA) for the real-time monitoring of biospecific interactions. Anal.

L weeptems.

L ofer.

L ionnal.

L of themic

L ction431,

L m.

M d.),pp.

N 998.acteri-

N andarm.

O ensornd to

O renes to

Qian, J.-M., Suo, A.-L., Yao, Y., Jin, Z.H., 2004. Polyelectrolyte-stabilized glucose biosensor based on woodceramics as electrode.Clin. Biochem. 37, 155–161.

Rahman, Md.A., Park, D.-S., Shim, Y.-B., 2004. A performance com-parison of choline biosensors: anodic or cathodic detections of H2O2

generated by enzyme immobilized on a conducting polymer. Biosens.Bioelectron. 19, 1565–1571.

Rahman, Md.A., Won, M.-S., Shim, Y.-B., 2003. Characterization of anEDTA bonded conducting polymer modified electrode: its applicationfor the simultaneous determination of heavy metal ions. Anal. Chem.75, 1123–1129.

Rogers, K.R., 1998. Biosensor technology for environment measurement.In: Meyers, R.A. (Ed.), Encyclopedia of Environmental Analysis andRemediation. Wiley, New York, pp. 755–768.

Ruzgas, T., Csoregi, E., Emneus, J., Gorton, L., Marko-Varga, G., 1996.Peroxidase modified electrodes: fundamentals and application. Anal.Chim. Acta 330, 123–138.

Ruzgas, T., Gorton, L., Emneus, J.G., Varga, M., 1995. Kinetic models ofhorseradish peroxidase action on a graphite electrode. J. Electroanal.Chem. 391, 41–49.

Shankaran, D.R., Shim, Y.-B., 2002. An amperometric sensor for hy-drogen peroxide based on a (3-mercaptopropyl) trimethoxysilaneself-assembled layer containing hydrazine. Electroanalysis 14, 704–707.

Shim, Y.-B., Park, S.-M., 1997. Electrochemistry of conductive poly-mers. XXII. Electrochemical and spectroelectrochemical studies ofpolyazulene growth and its characterization. J. Electrochem. Soc. 144,3027–3033.

Tatsuma, T., Ariyama, K., Oyama, N., 1998. Kinetic analysis of elec-tron transfer from a graphite coating to horseradish peroxidase. J.Electroanal. Chem. 446, 205–209.

T dase-for

0–

T ree-r for

. 67,

V 450.W . 71,

W ed onaste

W chu-tical

W idasehem.

Y niline71–78.

Y e re-n of

Z irectlase-. 74,

Z . Re-ercu-8.

Chim. Acta 400, 109–110.aviron, E., 1979. General expression of the linear potential s

voltammogram in the case of diffusionless electrochemical sysJ. Electroanal. Chem. 101, 19–28.

ee, T.Y., Shim, Y.-B., Shin, S.C., 2002. Simple preparationterthiophene-3′-carboxylic acid and characterization of its polymSynth. Met. 126, 105–110.

ee, T.Y., Shim, Y.-B., 2001. Direct DNA hybridization detectbased on the oligonucleotide-functionalized conductive polymer. AChem. 73, 5629–5632.

i, H., Zhang, F., Liu, S., 2004. Surface and interface state analysisTPD/Alq3 using X-ray photoelectron spectroscopy (XPS) and atoforce microscopy (AFM). Appl. Surf. Sci. 225, 162–169.

i, J., Dong, S., 1997. The electrochemical study of oxidation-reduproperties of horseradish peroxidase. J. Electroanal. Chem.19–22.

opez-Avila, V., Hill, H.H., 1997. Field analytical chemistry. Anal. Che69, 289R–305R.

urray, R.W., 1984. Chemically modified electrodes. In: Bard, A.J. (EElectroanalytical Chemistry, vol. 13. Marcel Dekker, New York,191–368.

g, S.C., Chan, H.S.O., Wong, P.M.L., Tan, K.L., Tan, B.T.G., 1Novel heteroarylene polyazomethines: their syntheses and charzations. Polymer 39, 4963–4968.

oda, A., Sendo, T., Ohno, K., Noda, H., Goto, H., 1987. Metabolismcytotoxicity of hydrazine in isolated rat hepatocytes. Chem. PhBull. (Tokyo) 35, 2538–2544.

kawa, Y., Tsuzuki, H., Yoshida, S., Watanabe, T., 1989. Glucose scarrying monomolecular layer of glucose oxidase covalently boutin (IV) oxide electrode. Anal. Sci. 5, 507–512.

rtiz de Montellano, P.R., Catalano, C.E., 1985. Epoxidation of styby hemoglobin and myoglobin: transfer of oxidizing equivalentthe protein surface. J. Biol. Chem. 260, 9265–9271.

atsuma, T., Ogawa, T., Sato, R., Oyama, N., 2001. Peroxiincorporated sulfonated polyaniline—polycation complexeselectrochemical sensing of H2O2. J. Electroanal. Chem. 501, 18185.

hierry de Lumley-Woodyear, T., Rocca, P., Lindsay, J., Dror, Y., Fman, A., Heller, A., 1995. Polyacrylamide-based redox polymeconnecting redox centers of enzymes to electrodes. Anal. Chem1332–1338.

on-Burg, R., Stout, T., 1991. Hydrazine. J. Appl. Toxicol. 11, 447–ang, J., 1999. Electroanalysis and biosensors. Anal. Chem

328R–332R.ang, Q., Lu, G., Yang, B., 2004. Hydrogen peroxide biosensor bas

direct electrochemistry of hemoglobin immobilized on carbon pelectrode by a silica sol–gel film. Sens. Actuators B 99, 50–57.

arsinke, A., 1997. Biosensors for food analysis. In: Scheller, F.W., Sbert, F., Fedrowitz, J. (Eds.), Frontiers in Biosensorics II. PracApplications. Birkhauser Verlag, Basel, pp. 121–140.

atanabe, T., Okawa, Y., Tsuzuki, H., Yoshida, S., 1988. Glucose-oxmodified SnO2 electrode as electrochemical glucose sensor. CLett., 1183–1186.

ang, Y., Mu, S., 1997. Bioelectrochemical responses of the polyahorseradish peroxidase electrodes. J. Electroanal. Chem. 432,

onetani, T., Schleyer, H., 1967. Cytochrome c peroxidase IX. Thaction of ferrimyoglobin with hydroperoxides and a comparisoperoxide-induced compounds of ferrimyoglobin and cytochromec per-oxidase. J. Biol. Chem. 242, 1974–1979.

hang, Z., Chouchane, S., Magliozzo, R.S., Rusling, J.F., 2002. Dvoltammetry and catalysis with mycobacterium tuberculosis cataperoxidase, peroxidases, and catalase in lipid films. Anal. Chem163–170.

hang, Z., Rusling, J.F., Chouchane, S., Magliozzo, R.S., 2001versible electrochemistry and catalysis with mycobacterium tublosis catalase-peroxidase in lipid films. Chem. Commun. 177–17