Ee

YK

a

ARRAA

KAAMG(H

1

bssEttficitaitcl

bsa

0d

Electrochimica Acta 56 (2011) 6715– 6721

Contents lists available at ScienceDirect

Electrochimica Acta

jou rn al hom epa ge: www.elsev ier .com/ locate /e lec tac ta

lectrodeposition of gold–platinum alloy nanoparticles on carbon nanotubes aslectrochemical sensing interface for sensitive detection of tumor marker

a Li, Ruo Yuan ∗, Yaqin Chai, Zhongju Songey Laboratory on Luminescence and Real-Time Analysis, Ministry of Education, College of Chemistry and Chemical Engineering, Southwest University, Chongqing 400715, PR China

r t i c l e i n f o

rticle history:eceived 15 January 2011eceived in revised form 19 May 2011ccepted 19 May 2011vailable online 26 May 2011

eywords:mperometric immunosensorlpha-fetoprotein (AFP)

a b s t r a c t

A novel electrochemical sensing interface, electrodeposition of gold–platinum alloy nanoparticles(Au–PtNPs) on carbon nanotubes, was proposed and used to fabricate a label-free amperometricimmunosensor. On the one hand, the multiwalled carbon nanotubes (MWCNTs) could increase activearea of the electrode and enhance the electron transfer ability between the electrode and redox probe;on the other hand, the Au–PtNPs not only could be used to assemble biomolecules with bioactivity keptwell, but also could further facilitate the shuttle of electrons. In the meanwhile, horseradish peroxidase(HRP) instead of bovine serum albumin (BSA) was employed to block the possible remaining active sitesand avoid the nonspecific adsorption. With the synergetic catalysis effect of Au–PtNPs and HRP towards

ulti-walled carbon nanotubes (MWCNTs)old–platinum alloy nanoparticles

Au–PtNPs)orseradish peroxidase (HRP)

the reduction of hydrogen peroxide (H2O2), the signal could be amplified and the sensitivity could beenhanced. Using alpha-fetoprotein (AFP) as model analyte, the fabricated immunosensor exhibited twowide linear ranges in the concentration ranges of 0.5–20 ng mL−1 and 20–200 ng mL−1 with a detectionlimit of 0.17 ng mL−1 at a signal-to-noise of 3. Moreover, the immunosensor exhibited good selectivity,stability and reproducibility. The developed protocol could be easily extended to other protein detection

pote

and provided a promising

. Introduction

The elevated concentration of tumor marker in serum maye an early indication of certain cancer. Therefore, it is neces-ary and important to develop a method with high sensitivity andelectivity for the determination of tumor marker levels in serum.lectrochemical immunosensor, based on highly specific recogni-ion of antibody and antigen, is a very promising technique forhe assay of tumor marker and has been applied widely in manyelds such as environment analysis [1], food industry [2,3], andlinical applications [4], which was due to their potential util-ty as specific, simple, low cost, small size and short responseime [5]. Various types of electrochemical immunosensors suchs amperometric [6,7], potentiometric [8,9], capacitive [10,11] andmpedance immunosensors [12,13] have been reported. Amonghese immunosensors, the amperometric immunosensor is espe-ially promising for its relatively higher sensitivity, low detectionimit and wider linear range [14,15].

In the fabrication process of immunosensor, immobilization of

iomolecules with bioactivity kept well on the sensing electrodeurface has been considered to be one of the most important points,nd accordingly there are many literatures about immunosensor

∗ Corresponding author. Tel.: +86 23 68252277; fax: +86 23 68253172.E-mail address: yuanruo@swu.edu.cn (R. Yuan).

013-4686/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.oi:10.1016/j.electacta.2011.05.066

ntial in clinical diagnosis application.© 2011 Elsevier Ltd. All rights reserved.

based on different immobilization matrix and sensing interface[16]. With the rapid development of nanotechnology over the pastdecade, various nanomaterials have been synthesized and usedfor the construction of biosensors, especially metallic nanoparti-cles. Recently, bimetallic alloys have been of considerable interestin the field of catalysis and sensors because of the interactionbetween two components in bimetallic alloys. They often presentmany favorable properties in comparison with the correspondingmonometallic counterparts, which include high catalytic activ-ity, catalytic selectivity and better resistance to deactivation.Gold–platinum alloy nanoparticles (Au–PtNPs) are very attractiveamong various bimetallic alloys. Besides large surface-to-volumeratio, good biocompatibility and satisfied conductive capability[17,18], Au–PtNPs possessed excellent catalytic activities towardsH2O2, methanol and so on, due to the high synergistic actionbetween gold and platinum [19]. So it is significant to developAu–PtNPs based electrochemical sensors with appropriate char-acteristics such as high sensitivity, fast response time, wide linearrange, better selectivity and reproducibility, which was attribute tothe advantages of bimetallic nanoparticles. Based on the above rea-son, Luo et al. have reported Au–PtNPs for electrocatalytic methanoloxidation reaction and the result was satisfying [20].

On the other hand, carbon nanotubes (CNTs), which couldpossess unique advantages including enhanced electronic prop-erties, a large edge plane/basal plane ratio, and rapid electrodekinetics, have also been incorporated into electrochemical sensors.

6716 Y. Li et al. / Electrochimica Acta 56 (2011) 6715– 6721

F ) adso(

Coc

pmoctA(bnsttspaeopapc

2

2

(msCrNwgotr

ig. 1. Schematic illustration of the stepwise immunosensor fabrication process: (ac) anti-AFP loading, (d) HRP blocking, (e) AFP loading.

NT-based sensors generally have higher sensitivities, lower limitsf detection, and faster electron transfer kinetics than traditionalarbon electrodes [21].

Based on the above consideration, we utilized the uniqueroperties of MWCNTs and Au–PtNPs to fabricate an ampero-etric immunosensor. Initially, Au–PtNPs were electrodeposited

n the multiwalled carbon nanotubes (MWCNTs) modified glassyarbon electrode by constant potential stripping technique, andhen anti-AFP was adsorbed onto the sensing interface of theu–PtNPs/MWCNTs/GCE. Subsequently, horseradish peroxidase

HRP) instead of bovine serum albumin (BSA) was employed tolock possible remaining active sites of the Au–PtNPs to avoid theon-specific adsorption and was further used to amplify responseignal. The electron transfer between HRP and electrode surface ishe limiting factor in the operation of amperometric signal detec-ion. The hydroquinone (HQ) was used as an electron mediator tohuttle electrons between HRP and the electrode surface. Hydrogeneroxide (H2O2) was added into the working buffer as substrate,nd the response signal could be amplified effectively by the syn-rgistic catalysis action of Au–PtNPs and HRP towards the reductionf H2O2. Using alpha-fetoprotein (AFP) as model analyte, the pro-osed biosensor exhibited a wide linear response range for modelnalyte, good selectivity and sensitivity, which demonstrated theroposed immunosensor possessed potential applications in clini-al screening of cancer biomarkers.

. Experimental

.1. Reagents and material

Anti-AFP and AFP were purchased from Biocell CompanyZhengzhou, China), stored in the frozen state before used. The

ulti-walled carbon nanotubes (MWCNTs, >95% purity, synthe-ized by CVD method) were purchased from Chengdu Organichemicals Co. Ltd. of the Chinese Academy of Science. Chlorau-ic acid, chloroplatinic acid, bovine serum albumin (BSA, 96–99%),, N-dimethylformamide (DMF) and horseradish peroxidase (HRP)ere bought from Sigma Chemical Co. (St. Louis, MO, USA). Hydro-

en peroxide (H2O2, 30%, w/v solution) and hydroquinol (HQ) werebtained from Chemical Reagent Co. (Chongqing, China). All ofhe chemicals used were of analytical grade and were used aseceived without further purification. Phosphate-buffered solution

rption of MWCNTs film, (b) electrodeposition of gold–platinum alloy nanoparticles,

(PBS, 0.1 M) with various pH values was prepared with stock stan-dard solutions Na2HPO4 and NaH2PO4. The supporting electrolytewas 0.1 M KCl. Serum specimens provided by Southwest Hospital ofThird Military Medical University (Chongqing, China) were storedat 4 ◦C in a freezer. Double-distilled water was used throughout thisstudy.

2.2. Apparatus

Electrochemical measurements were carried out by CHI 660Delectrochemistry workstation (Shanghai CH Instruments Co.,China). The scanning electron micrograph was taken with scanningelectron microscope (SEM, S-4800, Hitachi). A conventional, three-electrode cell consisting of the modified glassy carbon electrode(GCE) as the working electrode, a platinum wire as the counterelectrode and a saturated calomel electrode (SCE) as the referenceelectrode was used.

2.3. Fabrication of the amperometric immunosensors

The bare GCE were respectively polished with 0.3 and 0.05 �malumina slurry to obtain mirror-like surface and ultrasonicallycleaned with ethanol and double distilled water for 5 min to removethe physically adsorbed substance. Then the electrodes wereallowed to dry at room temperature. Subsequently, 10 �L of blacksuspension of MWCNTs was dispersed in N, N-dimethylformamide(DMF) was cast on the pretreated bare GC electrode surface anddried in air. After that, the MWCNTs modified GCE was immersed in2 mL deposition solution (0.2 M Na2SO4 aqueous solution contain-ing 1 mM HAuCl4 and 1 mM H2PtCl6) [22] and applied a constantpotential for 200 s at −0.2 V to obtain Au–PtNPs/MWCNTs modi-fied electrode. Then, it was immersed in anti-AFP solution at 4 ◦Cfor 12 h. Finally, the obtained electrode was incubated in HRP solu-tion for 4 h at 4 ◦C to eliminate non-specific binding effect andblock possible remaining active sites. The schematic diagram of theimmunosensor was shown in Fig. 1.

2.4. Synthesis of Au–PtNPs

Au–PtNPs were synthesized according to the reference [22].1 mL 1 mM HAuCl4 and 1 mL 1 mM H2PtCl6 were mixed with 0.2 MNa2SO4 aqueous solution. The resulting solution was Au–PtNPs

Y. Li et al. / Electrochimica Act

s(A

2

ppiitfd2crptttr

3

3

eotdwpteocte(diewm

Fig. 2. SEM images of gold–platinum alloy nanoparticles.

olution and was stored at 4 ◦C. Scanning electron micrographsSEM) were performed to characterize the shape and size ofu–PtNPs (Fig. 2).

.5. Experimental measurements

The characteristic of the stepwise modified electrode wereerformed by cyclic voltammetry (CV) measurements, with theotential range from −0.3 to +0.4 V (vs SCE) at the rate of 50 mV s−1

n pH 7.4 PBS containing 2.0 mM hydroquinone (HQ). After beingncubated with various concentrations of AFP for 12 min at roomemperature, the immunosensors capturing AFP were washed care-ully with washing buffer and the detection was carried out byifferential pulse voltammetry (DPV) in working buffer containing.0 mM HQ and 8.0 mM H2O2. The immunoassay was based on thehange of reduction peak current response (�I) before and after theeaction of antibody and antigen. Before the immunoreaction takenlace, the current response (I0) was recorded. Owing to the forma-ion of bulky immunocomplexes restraining the electron-transfer,he current response of the immunosensor decreased after incuba-ion, which was recorded as I. Therefore, the immunosensor currentesponse was given by: �I = I0 − I.

. Results and discussion

.1. Electrochemical characterization on electrode surface

CV is an effective and convenient tool of showing the changes oflectrode behavior after each modification step. Fig. 3A shows CVsf differently modified electrodes in the potential range from −0.3o +0.4 V in work solution (pH 7.4 PBS containing 2.0 mM HQ). Well-efined CVs, characteristic of diffusion-limited redox processes,ere observed at the bare GCE electrode (Fig. 3A, curve a). Theeak currents increased after MWCNTs coating on the bare elec-rode (Fig. 3A, curve b), the reason was that MWCNTs facilitatelectron transfer. After Au–PtNPs monolayer was electrodepositedn modified electrode, the current further increased (Fig. 3A, curve), which was due to the promotion of Au–PtNPs towards the shut-le of electrons. When anti-AFP molecules were immobilized on thelectrode surface, a decreased current response signal was obtainedFig. 3A, curve d). After HRP was immobilized successfully, a furtherecrease of the peak currents was found with the fact that HRP

nsulates the conductive support and hinders the transmission oflectrons towards the electrode surface (Fig. 3A, curve e). Fig. 3Bas a contrast experiment and showed the CVs of the differentodified electrodes in the potential range from −0.2 to +0.6 V in pH

a 56 (2011) 6715– 6721 6717

7.4 PBS containing 5.0 mM K3[Fe(CN)6]/K4[Fe(CN)6] (1:1) and 0.1 MKCl. Fig. 3A and B displayed the CVs of the fabrication process ofthe immunosensor with different redox probes, respectively, whichboth proved the successful fabrication of the sensing interface.

3.2. The influence of MWCNTs on electrochemical sensinginterface

In order to investigate the electrochemical properties of MWC-NTs monolayer, a comparative study was carried out by calculatingelectro-active surface areas of different modified electrodes: (a)Au–PtNPs/MWCNTs/GCE, (b) Au–PtNPs/GCE. The CVs of two elec-trodes were performed in the presence of redox probe Fe(CN)6

3−/4−

at a series of scan rates, respectively. An enormous surface area-to-volume ratio, which is highly susceptible to heterogeneousredox chemistry with surrounding environment, is pivotal factor.The electro-active surface area of electrode can be estimated fora reversible and diffusion controllable process according to theRandles–Sevcik equation [23]:

Ip = 2.69 × 105A · D1/2n3/2v1/2c

where, Ip relates to the redox peak current; n represents the trans-ferring electron number. This constant can be used to estimate n foran electrode. The Fe(CN)6

3−/4− redox system used in this study isone of the most extensively studied redox couples in electrochem-istry and exhibits a heterogeneous one-electron transfer (n = 1). crepresents the concentration (mol cm−3) of the ferricyanide; A isthe area of the electrode (cm2), D is the diffusion coefficient of themolecule in solution (cm2 s−1), which is 6.70 ± 0.02 × 10−6 cm2 s−1

at 25 ◦C; and v is the scan rate of the potential perturbation (V s−1).D, C and n are constants. Ip (as well as the current at any otherpoint on the wave) is proportional to vl/2. According to the aboveequation, we can get the approximate value of A.

Fig. 4A and B were the CVs of the preparedAu–PtNPs/MWCNTs/GCE and Au–PtNPs/GCE in 5.0 mMFe(CN)6

3−/4− at different scan rates, respectively. The insetin Fig. 3A and B were respectively linear relations with theanodic and cathodic peak current of Au–PtNPs/MWCNTs/GCEand Au–PtNPs/GCE against the square root of scan rate in theranges of 10–300 mV s−1, suggesting a diffusion-controlled redoxprocess. The values of the electro-active surface area were respec-tively 25.44 mm2 for Au–PtNPs/MWCNTs/GCE and 14.75 mm2 forAu–PtNPs/GCE by the anodic peak current of electrode.

In Fig. 4C, it was the CVs of two detected electrodes in the same5.0 mM Fe(CN)6

3−/4− at scan rate of 90 mV s−1. Fig. 4C shows thatthe proposed electrode with MWCNTs (Fig. 4C, curve a) had higherconductivity than contrastive electrode without MWCNTs (Fig. 4C,curve b). This trend is consistent with the electro-active surfacearea and the peak current. The experimental results suggested thatthe proposed sensing interface possessed bigger electro-active sur-face areas and higher conductivity, which could prove the sensingcapabilities of the immunosensor.

3.3. The synergistic catalysis effect of HRP and Au–PtNPs

In order to make a comparison of catalysis activity of dif-ferent metal nanoparticles, the electrochemical performance ofthe HRP/anti-AFP/AuNPs/MWCNTs sensing interface (Fig. 5A): (a)without H2O2, (b) with H2O2; HRP/anti-AFP/PtNPs/MWCNTs sens-ing interface (Fig. 5B): (c) without H2O2, (d) with H2O2; andHRP/anti-AFP/Au-PtNPs/MWCNTs sensing interface (Fig. 5C): (e)

without H2O2, (f) with H2O2 were evaluated. Compared with theirmonometallic counterparts of AuNPs and PtNPs, Au–PtNPs exhib-ited more excellent catalytic property because of the synergisticeffect of the Au and Pt.

6718 Y. Li et al. / Electrochimica Acta 56 (2011) 6715– 6721

-0.2 0.0 0.2 0.4 0.6

-250

-125

0

125

250

a b c d e0

100

200

I / µ

A B

decb

a

I /µA

E / V-0.25 0.00 0.25 0.50

-200

-100

0

100

200

a b c d e0

75

150

I / µ

A

A

cbdea

I /µA

E / V

F Q in Pp /Au–P5

rFaeiA

3

3

ept

Fdap

ig. 3. The CVs of the modified electrodes in different redox probes: (A) 2.0 mM HH 7.4): (a) bare GCE, (b) MWCNTs/GCE, (c) Au–PtNPs/MWCNTs/GCE, (d) anti-AFP0 mV s−1 and all potentials are given versus SCE.

The electrocatalytic reactivity of immobilization matrix for theeduction of H2O2 was investigated by CVs in Fig. 5D. Curves g ofig. 5D show the CV of the sensing interface with BSA as blockinggent in the presence of H2O2. The HRP blocked sensing interfacexhibited better electrocatalytic property than BSA blocked sens-ng interface, owing to the synergistic catalysis effect of HRP andu–PtNPs.

.4. Optimization conditions for immunoassay

.4.1. Influence of pH on the response of immunosensor

The pH of the detection solution (0.1 M PBS) had a profound

ffect on the immunosensor. As shown in Fig. 6A, the influence ofH value on the current responses of the immunosensor was inves-igated in the range from 6.0 to 8.5. It was found that the current

ig. 4. The CVs of the different modified electrodes at different scan rates (from inner

ependence of redox peak currents on the square root of scan rates. (A) Au–PtNPs/MWCNt scan rate of 90 mV s−1: (a) with MWCNTs, (b) without MWCNTs. All experiments weotentials were given vs. SCE.

BS of pH 7.4; (B) 0.1 mM K3[Fe(CN)6]/K4[Fe(CN)6] (1:1) PBS (containing 0.1 M KCl,tNPs/MWCNTs/GCE, (e) HRP/anti-AFP/Au–PtNPs/MWCNTs/GCE. The scan rate was

responses increased with increasing pH value from 6.0 to 7.4 toreach the maximum and decreased thereafter. Thus, the optimal pH7.4 of working buffer was chosen throughout this study to obtain ahigh analytical sensitivity. On the other hand, the incubation solu-tion pH was selected as 7.4 (the physiological environment) in orderto keep the immunoreaction under optimal condition.

3.4.2. Influence of concentration of hydrogen peroxideThe influence of concentration of hydrogen peroxide in work-

ing solution for the catalytic activities of the electrode was studiedusing CV (Fig. 6B). With increasing the concentration of H2O2, the

reduction peak current increased gradually with the decrease ofthe oxidation peak current, which indicated that the immobilizedAu–PtNPs and HRP on the sensing surface had a typical electro-catalytic reduction process of H2O2. The maximum peak current

to outer): 10, 30, 50, 70, 90, 120, 150, 200 and 300 mV s−1. The inset showed theTs/GCE; (B) Au–PtNPs/GCE; (C) is the CVs of Au–PtNPs/MWCNTs modified electrodere in 5 mM K3Fe(CN)6/K4Fe(CN)6 (1:1) mixture under room temperature and all

Y. Li et al. / Electrochimica Acta 56 (2011) 6715– 6721 6719

Fig. 5. (A) The CVs of HRP/anti-AFP/AuNPs/MWCNTs modified electrode: (a) without H2O2, (b) with H2O2; (B) HRP/anti-AFP/PtNPs/MWCNTs modified electrode: (c) withoutH2O2, (d) with H2O2; (C) HRP/anti-AFP/Au–PtNPs/MWCNTs modified electrode: (e) without H2O2, (f) with H2O2; (D) (g) BSA/anti-AFP/Au–PtNPs/MWCNTs/GCE in workingsolution with H2O2. The scan rate was 50 mV s−1.

Fig. 6. Optimization of experimental parameters: (A) The immunosensor response in 2.0 mM HQ various pH values under room temperature: 6.0, 6.5, 7.0, 7.5, 8.0, 8.5; (B)The current response in the working solution with various concentration of H2O2 under room temperature; (C) influence of the incubation time on amperometric responseof immunosensor when immune-reacted with 100 ng mL−1 AFP. All potentials were given vs. SCE and the scan rate was 50 mV s−1.

6720 Y. Li et al. / Electrochimica Acta 56 (2011) 6715– 6721

0 50 100 150 200

20

40

60

80

-0.1 0.0 0.1 0.2

-80

-60

-40

-20

e

a

I / µ

AE / V

b

aΔ

I /µA

cAFP / (ng mL-1 )

Fig. 7. Calibration plots of the changes of DPV peak current response versus theconcentrations of AFP with the different immunosensors under optimal condi-tions: (a) HRP/anti-AFP/Au–PtNPs/MWCNTs/GCE with H2O2 in working solution;(b) BSA/anti-AFP/Au–PtNPs/MWCNTs/GCE without H2O2 in working solution. Theicu

ruH

3

twdisT

3

3

tciOatbahDctapwapplrie

3

s

Table 1Reproducibility assays using four immunosensors prepared in the same conditions.

Samplenumber

Standard value ofAFP (ng mL−1)

Intra-assaya

reproducibility (%)Inter-assayb

reproducibility (%)

1 0.5 4.4 5.82 1 2.3 4.73 10 5.7 6.14 20 4.0 3.9

a and b are RSD of three measurements.

Table 2The recovery of the proposed immunosensor in human serum.

Samplenumber

Standard value ofAFP (ng mL−1)

Found (ng mL−1)a Recovery (%)

1 1 1.05 ± 0.02 104.62 5 4.86 ± 0.08 97.23 10 9.90 ± 0.16 99.04 20 21.54 ± 0.33 107.6

nset was the DPV of the proposed immunosensor after incubation in different con-entrations of AFP standard solution (from a to e): 20, 50, 100, 150 and 200 ng mL−1

nder the optimal conditions.

esponse occurred in 8.0 mM H2O2, which corresponded to the sat-rated state. Consequently, the optimum concentration of 8.0 mM2O2 was employed for the test.

.4.3. Influence of incubation time on the immunoreactionsAs shown in Fig. 6C, the influence of immunochemical incuba-

ion time on current response was investigated. The immunosensoras incubated in a constant concentration of 100 ng mL−1 AFP forifferent time. The current response decreased with increasing

ncubation time and reached a platform at 12 min, indicating theaturated formation of immunocomplex on the modified electrode.herefore, the incubation time of 12 min was adopted in this work.

.5. Performance of the immunosensor

.5.1. DPV response and calibration curveIn this study, DPV technique was employed to investigate

he response performances of immunosensor under the optimalonditions. The inset of Fig. 7 shows the DPV response of themmunosensor when detecting different concentrations of AFP.bviously, the DPV peak currents of the immunosensor showed

decrease with the increase of AFP concentration in the incuba-ion solution. The reason was that more antigen molecules coulde bound to immobilized antibodies at higher concentrations ofntigens, and the immuocomplexes acted as an inert block layerindering the electron-transfer. As can be seen from Fig. 7a, thePV peak currents of the electrochemical immunosensor and theoncentrations of AFP showed a linear relationship in the concen-ration range from 0.5 to 20 ng mL−1 and 20 to 200 ng mL−1 with

detection limit of 0.17 ng mL−1 at a signal-to-noise of 3. As com-arative study, the immunosensor employed BSA as blocking agentas also prepared in Fig. 7b. The linear range was 5–10 ng mL−1

nd 10–100 ng mL−1 with a detection limit of 1.6 ng mL−1. Com-ared with the immunosensor employed BSA as blocking agent, theroposed immunosensor displayed higher sensitivity and a wider

inear range by the addition of H2O2 in the working solution. Theeason could be contributed to the facts that Au–PtNPs and block-ng agent HRP performed an effective amplification properties asxpected.

.5.2. Stability of the immunosensorThe stabilities of the immunoassay system were evaluated by

uccessive cycle scan. After 50 cycle successive CV measurements

5 50 51.25 ± 0.52 102.5

a Mean value ± SD of three measurements.

under the optimal conditions at 50 mV s−1, only 2.8% decrease ofthe initial current was observed. The long-time stability of theimmunosensor was investigated with 7 days. The immunosensorwas stored in the refrigerator at 4 ◦C, the relative standard deviationwas less than 6%. The excellent performance of the immunosen-sor owed to the good stability of the prepared Au–PtNPs/MWCNTssensing interface and the anti-AFP molecules were absorbed firmlyon the Au–PtNPs layer which provided a good biocompatiblemicroenvironment.

3.5.3. Reproducibility of the immunosensorThe reproducibility of the response of the immunosensor was

investigated by intra- and inter-assay coefficients of variation. Theintra-assay precision of the analytical method was evaluated byanalyzing four concentration levels using the equally preparedimmunosnesors. Similarly, the inter-assay precision of the ana-lytical method was also evaluated. The results of intra-assay andinter-assay were shown in Table 1, which suggested that the repro-ducibility of the proposed immunosensors was satisfying.

3.5.4. Preliminary analysis of real samplesTo investigate the feasibility of the prepared immunosensor for

clinical analysis, the immunosensor was applied to the determi-nation of different concentrations of AFP added into normal bloodserum samples. As shown in Table 2, the recovery was in the rangeof 97.2–107.6%, which indicated that the proposed immunosensormight be preliminarily applied for the direct determination of AFPin clinical diagnosis.

4. Conclusion

In this paper, a novel approach for fabrication of immunosensorwas described based on HRP/anti-AFP/Au–PtNPs/MWNTs modifiedGC electrode. The highlights of the prepared immunosensor canbe summarized as follows: the proposed electrode showed a largeelectro-active surface areas and high conductivity. With the syner-getic catalysis effect of Au–PtNPs and HRP towards the reductionof hydrogen peroxide (H2O2), the signal could be amplified and thesensitivity could be enhanced. Biomolecules could be immobilizedon the Au–PtNPs tightly with the bioactivity kept well. On the basis

of the above reasons, the proposed immunosensor showed widelinear range, high stability and reproducibility, superior sensitiv-ity and good catalytic activity. The simple fabrication procedures

ca Act

od

A

ttSi(S

R

[

[

[[

[[[

[[

[[

Y. Li et al. / Electrochimi

f the proposed method makes potentially attractive for the futureevelopment of practical devices for clinical diagnosis application.

cknowledgments

The authors are grateful for the financial supports provided byhe National Natural Science Foundation of China\ (21075100),he Ministry of Education of China (Project 708073), the Naturalcience Foundation of Chongqing (CSTC-2009BA1003), Special-zed Research Fund for the Doetoral Program of Higher Education20100182110015) and High Technology Project Foundation ofouthwest University (XSGX02), China.

eferences

[1] M. Farré, L. Kantiani, D. Barcelo, Trends Anal. Chem. 26 (2007) 1100.

[2] F. Ricci, G. Volpe, L. Micheli, G. Palleschi, Anal. Chim. Acta 605 (2007) 111.[3] Y. Wu, J.W. Zheng, Z. Li, Y.R. Zhao, Y. Zhang, Biosens. Bioelectron. 24 (2009)

1389.[4] J. Wu, Z.F. Fu, F. Yan, H.X. Ju, Trends Anal. Chem. 26 (2007) 679.[5] J. Zhang, J.P. Lei, C.L. Xu, L. Ding, H.X. Ju, Anal. Chem. 82 (2010) 1117.

[

[[

a 56 (2011) 6715– 6721 6721

[6] J.H. Lin, C.Y. He, L.J. Zhang, S.S. Zhang, Anal. Biochem. 384 (2009) 130.[7] Z.J. Song, R. Yuan, Y.C. Chai, B. Yin, P. Fu, J.f. Wang, Electrochim. Acta 55 (2010)

1778.[8] L. Qu, S.H. Xia, C. Biana, J.Z. Sun, J.H. Hana, Biosens. Bioelectron. 24 (2009) 3419.[9] R.P. Liang, H.Z. Peng, J.D. Qiu, J. Colloid Interface Sci. 320 (2008) 125.10] S. Loyprasertb, M. Hedströma, P. Thavarungkulb, P. Kanatharanab, B. Mattias-

sona, Biosens. Bioelectron. 25 (2010) 1977.11] S. Loypraserta, P. Thavarungkula, P. Asawatreratanakul, B. Wongkittisuksaa, C.

Limsakula, Biosens. Bioelectron. 24 (2008) 78.12] Y. Xiao, C.M. Li, Y.S. Liu, Biosens. Bioelectron. 22 (2007) 3161.13] X.J. Chen, Y.Y. Wang, J.J. Zhou, W. Yan, X.H. Li, J.J. Zhu, Anal. Chem. 80 (2008)

2133.14] J. Das, M.A. Aziz, H. Yang, J. Am. Chem. Soc. 128 (2006) 16022.15] W. Cheng, F. Yan, L. Ding, H.X. Ju, Y.B. Yin, Anal. Chem. 82 (2010) 3337.16] X. Che, R. Yuan, Y.Q. Chai, J.J. Li, Z.J. Song, J.F. Wang, J. Colloid Interface Sci. 345

(2010) 174.17] G. Vijayaraghavan, K.J. Stevenson, Langmuir 23 (2007) 5279.18] Y. Huang, Q. Wen, J.H. Jiang, G.L. Shen, R.Q. Yu, Biosens. Bioelectron. 24 (2008)

600.19] F. Xiao, F. Zhao, Y. Zhang, G. Guo, B. Zeng, J. Phys. Chem. C 113 (2009) 849.20] J. Luo, N. Njoki, Y. Lin, L. Wang, C. Zhong, Langmuir 22 (2006) 2892.

21] G.H. Lu, L.Y. Zhu, P.X. Wang, J.H. Chen, D.A. Dikin, R.S. Ruoff, Y. Yu, Z.F. Ren, J.

Phys. Chem. C 111 (2007) 17919.22] X.H. Kang, Z.B. Mai, X.Y. Zou, P.X. Cai, J.Y. Mo, Anal. Biochem. 369 (2007) 71.23] A.J. Bard, L.R. Faulkner, Electrochemical Methods: Fundamentals and Applica-

tions, 2nd ed., 2001, Wiley, New York, 1980, p. 231.