Analytica Chimica Acta 516 (2004) 19–27

Electrochemical impedance detection of DNA hybridization based on theformation of M-DNA on polypyrrole/carbon nanotube modified electrode

Ying Xu, Ying Jiang, Hong Cai, Pin-Gang He, Yu-Zhi Fang∗

Department of Chemistry, East China Normal University, Shanghai 200062, China

Received 21 November 2003; received in revised form 15 March 2004; accepted 7 April 2004

Available online 25 May 2004

Abstract

A new selective and sensitive biosensing strategy for electrochemical impedance spectroscopy (EIS) measurement of DNA hybridizationis described. The detection approach relied on the doping of nucleic acid probes within electropolymerized polypyrrole (PPy) film onto acarboxylic group-functionalized multi-walled carbon nanotubes (MWNTs-COOH) modified electrode and monitoring the impedance changesprovoked by the metallation of helix DNA after hybridization. Oligonucleotide probes served as the solo counter anions during the growth ofconducting PPy film on the carbon nanotube modified electrode. As a consequence of hybridization and the formation of metallized double-stranded DNA (M-DNA), significant changes in electrochemical impedance values (both in real componentZre and imaginary componentZim)coming from the change of electronic transport resistance of the modified electrode were observed, especially a visible decrease in theZre.Based on the unique response�Zre after hybridization and metallation at 5469 Hz, only the complementary DNA sequence had an obvioussignal of the impedance decrease when compared with 1-base, 3-base mismatched and non-complementary sequences; hybridization amountsof 1-base, 3-base mismatched sequences were obtained only 35.7 and−5.8% responses. The protocol also offered high sensitivity with thedetection limitation was 5× 10−11 M using 3 S.D.,n = 11. Results showed that Zn2+-DNA had the best ability to transport electrons inM-DNA double-stranded chains when compared with Co2+-DNA and Ni2+-DNA on the same condition.© 2004 Elsevier B.V. All rights reserved.

Keywords:Impedance detection; Polypyrrole; Carbon nanotubes; DNA hybridization; Metallized double-stranded DNA

1. Introduction

Interests in the DNA hybridization biosensors have beenincreased due to the major developments in human genetics.An effective DNA-sensing system to detect specific nucleicacid sequences is playing an important role in many areas,such as clinical diagnosis, medicine, epidemic prevention,environmental protection and bioengineering. Various tech-niques including radiochemical[1], electrochemical[2], col-orimetric [3] and chemiluminescent[4] methods have beendeveloped for DNA analysis. Among them, electrochemicaltechnique is a novel and developing technique which com-bines biochemical, electrochemical, medical and electronictechniques with the advantages of being simple, reliable,cheap, sensitive and selective for genetic detection, as wellas can be compatible with DNA biochip. Such devices rely

∗ Corresponding author. Fax:+86-21-62451921.E-mail address:yuzhi@online.sh.cn (Y.-Z. Fang).

on conversion of the DNA base-pair recognition event intoa useful electrical signal. It can detect the increased cur-rent signal of the redox indicator[5,6], including cationicmetal complexes[7], organic intercalating compounds[8,9];metal nanoparticles labels[10,11] or enzyme labels[12,13](the latter are accomplished by an electrochemical measure-ment of the product of the enzymatic reaction[14]); it alsocan monitor the changes in the instinctive redox activityof DNA occurred from the hybridization event[15,16], ex-ploit different rates of electron transfer[17,18] and detectchanges in the conductivity[19,20], capacitance[21] or theFaradic impedance[22–24]accompanying the hybridizationevent. For instance, Barton and coworkers[17,18]monitoredchanges between an electroactive methylene blue intercala-tor and a ferricyanide redox species using a gold electrodemodified with thiolated DNA. Korri-Youssoufi et al.[19]found that 13-mer oligonucleotide-substituted PPy film dis-played a decrease current response during the duplex forma-tion due to the higher rigidity formation along the polymer

0003-2670/$ – see front matter © 2004 Elsevier B.V. All rights reserved.doi:10.1016/j.aca.2004.04.013

20 Y. Xu et al. / Analytica Chimica Acta 516 (2004) 19–27

backbone. Berggren et al.[21] monitored the capacitancechange of a thiolate-oligonucleotide modified gold electrodein high sensitivity and speed hybridization event.

Changes in the properties of conducting polymers accom-panying DNA hybridization hold a particular promise for de-tection DNA interaction[20,25]. One of our previous papersreported an indicator-free DNA hybridization detection ap-proach based on the DNA-doped polypyrrole film[26]. DNAprobe can serve as the solo dopant if no other electrolyteexisting in the polymerization solution when electropoly-merization occurs. The entrapped DNA probes maintaintheir hybridization activity within the host polymer network.In this work we applied electrochemical impedance spec-troscopy (EIS) to detect the polymeric impedance changeas the DNA hybridization signal. We found that the forma-tion of metallized double-stranded DNA after hybridizationcould further amplify the impedance signal to detect DNAhybridization.

M-DNA is usually formed from double-stranded DNAat pH above 8 in the presence of Zn2+, Co2+ or Ni2+ butnot Mg2+ or Ca2+ [27]; metal ions will be in the center ofthe helix coordinating to the N3 of T and N1 of G in everybase-pair (shown inFig. 1). Formation of M-DNA woulddecrease the electronic transport resistance of the electrodeinterfacial properties. This has already been proved by threemethods: First, fluorescence quenched when duplexes wereconstructed with a donor fluorophore at one end and accep-tor at the other on the formation of M-DNA[28]. Second,direct measurement of conductivity of M-DNA by detect-ing the current as a function of voltage has been formed byplacing the DNA between two gold electrodes[29]. Third,electrochemical methods such as cyclic voltammetry couldalso be used to measure electronic transfer rates throughM-DNA [30]. All the results showed that M-DNA has ametallic-like conductivity property and is able to increasethe DNA’s electronic transfer rate. In this field, Lee andco-workers have done a lot of work[27–29,31,32]andproposed the use of M-DNA as the molecular wire fornanoelectronics and biosensing.

Herein, we applied M-DNA to DNA hybridization detec-tion based on EIS choosing the impedance spectroscopy asthe analytical signal. A decrease of electronic transport re-sistance and an increase of double-layer capacitance in M-DNA/PPy compared with ssDNA/PPy modified electrodecould be obtained. The result showed that it was a low-costsingle-frequency analyzer, discarded the complex labeling

Fig. 1. The preferred model for M-DNA in which the imino protons ofT and G are replaced by the metal ion.

chemical manipulation and was free to use quite expensivereagents and analytical equipment.

2. Experimental

2.1. Reagents

Carboxylic group-functionalized multi-walled carbonnanotubes (MWNTs-COOH, the inner diameter was 10–20 nm, the outer diameter was 30–50 nm, the length was1–10�m) were purchased from Shenzhen Nanotech PortCo., Ltd. Company (Shenzhen, China). The various oligonu-cleotides were purchased from Shenggong Bioengineer-ing Ltd. Company (Shanghai, China) with the followingsequences—probe: 24-mer 5′-GAGCGGCGCAACATTTC-AGGTCGA-3′; the complementary target oligonucleotides:5′-TCGACCTGAAATGTTGCGCCGCTC-3′; 1-base mis-matched oligonucleotides: 5′-TCGACCTGAAACGTTG-CGCCGCTC-3′; 3-base mismatched oligonucleotides: 5′-TCGTCCTGAAACGTTGCGCCTCTC-3′.

A 0.05 M Tris–HCl buffer solution was used. Otherreagents were commercially available and were all of an-alytical reagent grade. All of the solutions were preparedwith ultrapure water from an Aquapro system.

2.2. Apparatus

A.C. impedance (ACI) measurements were all performedwith a CH Instruments Model 660 electrochemical analyzer(CH Instruments Inc., US). The three-electrode system con-sisted of a glassy carbon working electrode (ca. 9 mm2 sens-ing area), a Ag/AgCl reference electrode with KCl saturated,and a platinum wire counter electrode. All electrochemicalmeasurements were carried out in a 10 mL cell.

2.3. Construction of a MWNTs-COOH modified electrodetransducer

Prior to its modification, the glassy carbon electrode waspolished with a 0.3�m alumina slurry and was thoroughlywashed with sterilized water. Ten milligrams of MWNTs-COOH was dispersed into 10 mLN,N-dimethylformamide(DMF) with sonication for 10 min. Four microliters ofMWNTs-COOH/DMF was dropped onto the fresh elec-trode. After the DMF volatilized, a stable MWNTs-COOHmodified electrode was obtained.

2.4. Immobilization of single stranded DNA probes ontothe transducer

The polypyrrole electropolymerization proceeded witha repetitive cyclic voltammetric scanning (between 0.0and +0.75 V; 50 mV/s; 6 cycles) with the MWNTs-COOH/electrode dipping in the electropolymerization

Y. Xu et al. / Analytica Chimica Acta 516 (2004) 19–27 21

solution containing 0.05 M pyrrole and 10−5 M 24-merssDNA probe sequences. Pyrrole was distilled before used.

2.5. Hybridization of target DNA sequences with probeoligonucleotides entrapped in PPy network and metallationthe double-stranded DNA

Hybridization reaction was carried out using a stirring0.75 M NaCl+ 0.05 M Tris–HCl buffer blank solution (pH7.2) containing the target DNA sequence for 30 min at38◦C. After that, the dsDNA/PPy modified working elec-trode was incubated in a stirring 0.2 mM Zn2+ + 0.3 MNaCl+ 0.05 M Tris–HCl buffer solution (pH 8.6) for 3 h atroom temperature.

2.6. Electrochemical detection

Impedance measurements were performed in a 0.3 MNaCl + 0.05 M Tris–HCl buffer solution (pH 8.6) in thefrequency range from 105 down to 10 Hz with a samplingrate of 12 points per decade (A.C. amplitude: 5 mV, biaspotential: 0 V). Before every measurement, the workingelectrode was washed with 0.3 M NaCl+ 0.05 M Tris–HClbuffer solution (pH 8.6) for 15 min.

3. Results and discussion

3.1. Electrochemical synthesis of PPy/ssDNA film oncarbon nanotube modified electrode

3.1.1. The characteristics of the carbon nanotube modifiedelectrode

The morphology of carbon nanotube modified electrodewas investigated by AFM and shown inFig. 2 insert. It

Fig. 2. Differential pulse voltammetry of 1 mM K3[Fe(CN)6]:K4[Fe(CN)6] (1:1) in 0.1 M KCl from−0.2 to+0.6 V (vs. Ag/AgCl) with (a) MWNTs-COOHmodified electrode and (b) glassy carbon electrode as working electrode. The current curves were both subtracted the background current in the 0.1 MKCl without K3[Fe(CN)6]:K4[Fe(CN)6] with the same working electrode. Insert: AFM image of the carbon nanotube modified electrode.

illustrates that there were carbon nanotube clusters lyingrandomly on the electrode with the average diameter was180 nm and length was irregular. Additionally, we inves-tigated the carbon nanotube modified electrode by differ-ential pulse voltammetry (DPV) with the redox species,[Fe(CN)6]3−/4−, in the 0.1 M KCl solution when com-pared with the bare glassy carbon electrodes (GCE) asworking electrode. As shown inFig. 2, using the MWNTs-COOH modified electrode fabricated inSection 2.3, theoxidation current integral of [Fe(CN)6]3−/4− with volt-age was 4.152E−6 V A after subtracting the backgroundcurrent integral in 0.1 M KCl solution without containing[Fe(CN)6]3−/4−. However, using the bare GCE, the currentintegral after subtracting the background current integralwas 1.524E−6 V A. The result illustrates that the MWNTs-COOH modified electrode had higher active real surfaceand the detection surface has increased to about three timesin our protocol after the MWNTs-COOH modification. Itwas due to that each carbon nanotube had high surface areaand high electrical conductivity, therefore each nanotubeacted as an electrode and ion did not have to travel a longdistance to exchange the electrons.

3.1.2. Electrochemical synthesis of PPy/ssDNA film on thecarbon nanotube modified electrode

The results of electrochemical synthesis of PPy/ssDNAfilm onto the MWNTs-COOH modified electrode and bareelectrode were shown inFig. 3. The use of oligonucleotideas counter anion showed a normal polymer growth withincreasing current on repetitive scanning, which revealedan efficient film deposition and ssDNA incorporation intothe PPy network for maintaining the polymer’s electricalneutrality in the same manner as other anionic macro-molecules adopt when doping PPy[33]. The growth of

22 Y. Xu et al. / Analytica Chimica Acta 516 (2004) 19–27

-0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

0.00

-1.00

-2.00

-3.00

-4.00

b

a

Cur

rent

/ 1e

- 4A

Potential /V (vs. Ag/AgCl)

Fig. 3. Repetitive cyclic voltammograms for the PPy/ssDNA polymeriza-tion reaction on MWNTs-COOH modified glassy carbon electrode (a),glassy carbon electrode (b) in 0.05 M pyrrole+ 10−5 M ssDNA probesequences with six continuous cyclic scanning at 50 mV/s.

PPy/ssDNA film began at the potential of about+0.4 V(versus Ag/AgCl) on the MWNTs-COOH modified elec-trode, while on the bare electrode the growth occurred atabout+0.6 V. It shows that the surface of carbon nanotubeelectrode was more active than that of the flat glassy car-bon electrode due to the active characteristics of nanotubesand therefore lower nucleation energy was required at thebeginning of PPy/ssDNA polymerization. One the otherhand, the growth current of PPy/ssDNA on MWNTs-COOHmodified electrode was much greater compared with on abare electrode, and the growth current was 9.3:1. It wasdue to the very high real surface area from the nanometersize of nanotube and the MWNTs-COOH modified elec-trode decreasing the nucleation energy at the beginning ofPPy/ssDNA polymerization, therefore a large volume-freeradicals of pyrrole would concentrate on the MWNTs mod-ified electrode and greater volume of PPy/ssDNA couldbe deposited on the electrode surface without increasingthe thickness PPy/ssDNA film. The similar results havebeen obtained in literature[34]. The resulted PPy/ssDNAcoated electrode having an open, porous structure with athin surface that could offer an easy way for the targetDNA to contract the probe DNA, which would contributeto the sensitive detection of DNA hybridization. In addi-tion, due to the high strength and good adhesion betweencarbon nanotubes and PPy, the PPy/ssDNA film could notbe peeled off easily. Additionally, after the electropoly-merization in the pyrrole and DNA probe solution, thePPy/ssDNA has coated outsides the carbon nanotubes uni-formly; the similar results have been gained when using acarbon nanotube electrodes grown on flat titanium substrate[34]. Fan et al.[35] demonstrated that the nanotubes onlywere considered as composite of carbon nanotubes withPPy, there was no significant interaction between carbonnanotubes with PPy. Such well-uniform PPy/ssDNA film

would be important to the quantitative reproducibility of thedetection.

3.2. Electrochemical impedance spectroscopy detection forDNA sequence

In our protocol, the electrode interfacial properties’change were detected by the frequency dependent methodEIS, for it has proved to be a powerful and convenienttool for investigating electrode process by monitoringimpedance[36] and it can offer a lot of information aboutthe electrode interfacial properties such as electronic trans-port resistance and double-layer capacitance. Herein, for-mation of M-DNA would further decrease the impedanceafter hybridization, especially decrease the real compo-nent of the impedance which chosen as the hybridizationsignal.

3.2.1. DNA hybridization detection based on theimpedance change in the real part, Zre

In our experiment, the impedance measurementswere performed with a CH Instrument model 660electrochemical analyzer, recording the values of theimpedanceZ, the impedance’s real partZre (real impedance)and the impedance’s imaginary partZim (imaginaryimpedance) of the electrochemical system at each fre-quency from 105 down to 10 Hz. The Bode diagram oflogZ versus frequency from 105 to 10 Hz for the ss-DNA/PPy/electrode (a), dsDNA/PPy/MWNT/electrode (b)and M-DNA/PPy/MWNT/electrode (c) was recorded inFig. 4a. It shows that a visible decrease in the logarith-mic impedance would occur caused by the hybridizationbetween the ssDNA probe on the PPy film and the DNAtarget (10−6 M) in the hybridization solution. Between 104

and 102 Hz, the decrease was much clearer and stable, es-pecially around 103.5 Hz. After the metal, Zn2+, dopinginto the double-stranded DNA, the logarithmic impedancewould further decrease about to three times. The obviousdecrease of the impedance values may be due to the muchhigher conductivity properties of M-DNA which is able toincrease the DNA’s electronic transfer rate. Based on thesedata the CHI recorded, we compared the decrease inZ, ZimandZre after the hybridization and metallation, respectively(shown inFig. 4b–d). As we could notice that the visibledecrease in theZ, caused by the hybridization and metalla-tion, came mostly from the decrease in theZre. From 105

to 10 Hz, theZre decreased with the similar trend that washybridization would decrease the value ofZre, metallationwould further decreaseZre to about three times (shown inTable 1). However, the trend in theZim change was not sim-ilar with theZ or Zre. Oppositely, theZim began to increaseafter the hybridization and metallation from about 103 downto 10 Hz. As a result, herein we investigated the change inZre between ssDNA and M-DNA as the hybridization signal(frequency= 5469 Hz) to detect the DNA sequences for itwas stable and obvious for observation.

Y. Xu et al. / Analytica Chimica Acta 516 (2004) 19–27 23

Fig. 4. (a) Impedance curves in Bode diagram of the modified glassy carbon electrode. Curves: a, ssDNA/PPy/MWNT/electrode; b, ds-DNA/PPy/MWNT/electrode; c, M-DNA/PPy/MWNT/electrode detected in 0.3 M NaCl+0.05 M Tris–HCl buffer (pH 8.6) at an open circuit voltage from105 down to 10 Hz. Six cycles CV for DNA probe entrapment in PPy was same as inFig. 3. Hybridization condition: with 10−6 M target oligonucleotideat 38◦C for 30 min in 0.75 M NaCl+ 0.05 M Tris–HCl buffer (pH 7.2). Metallation condition: with 0.2 mM Zn2+ for 3 h in 0.3 M NaCl+ 0.05 MTris–HCl buffer (pH 8.6). (b) The plots of−�Z vs. frequency from 105 down to 10 Hz.Z was the impedance of the modified electrode. Other conditionsas in (a). (c) The plots of−�Zre vs. frequency from 105 down to 10 Hz.Zre was the impedance’s real component of the modified electrode. Otherconditions as in (a). (d) The plots of−�Zim vs. frequency from 105 down to 10 Hz. Zim was the impedance’s imaginary component of the modifiedelectrode. The other conditions were same as in (a).

Table 1Hybridization amount of various 24-mer target oligonucleotide detected byZre change at frequency= 5469 Hz between ssDNA/PPy and M-DNA/PPy filmmodified electrode. The target DNA was 10−7 M; impedance was measured in 0.3 M NaCl+0.05 M Tris–HCl buffer (pH 8.6) with an open circuit voltage

Working electrode withMWNT modification

Metallation Oligonuleotides �Zre (�) Hybridization (%)

Yes Yes Complementary −12.34 1001-Base mismatch −4.41 35.73-Base mismatch 0.715 −5.8

No Complementary −3.94 31.9

No Yes Complementary −2.57 20.8

24 Y. Xu et al. / Analytica Chimica Acta 516 (2004) 19–27

Fig. 5. Nyquist diagram of the ssDNA/PPy (a), dsDNA/PPy (b), M-DNA/PPy (c) and dis-metallized dsDNA/PPy (d) film on the MWNTs-COOH modifiedglassy carbon electrode. Dis-metallation condition: with EDTA for 40 min in 0.3 M NaCl+ 0.05 M Tris–HCl buffer (pH 8.6). Other conditions as inFig. 4. Insert: standard Randles circuit and its typical Nyquist diagram.Rs: the solution resistance,Ret: the electronic transport resistance,Zre: the realcomponent of impedance,Zim: the imaginary component of impedance,ω0: the ω whereZim in the semicircle is maximum and equalsRet/2.

3.2.2. The Nyquist diagram of the several DNA modifiedelectrodes in our protocol

Additionally, we investigated the Nyquist diagram(shown in Fig. 5) based on theZim and Zre from highfrequency to low frequency for ssDNA/PPy/MWNT, ds-DNA/PPy/MWNT and M-DNA/PPy/MWNT modifiedelectrode, respectively. The plots were similar to a typicalNyquist diagram for a standard Randles equivalent circuit;standard Randles equivalent circuit and its typical Nyquistdiagram were shown inFig. 5 insert. As illustrated inliterature [37], the shape of a typical Faradic impedancespectrum for a standard Randles circuit is a semicircle por-tion observed at the higher frequencies lying on theZre-axisand a straight line following the semicircle characteristic ofthe lower frequencies; at the higher frequencies, the spec-troscopy is characteristic of the electron transfer limitedprocess and at the lower frequencies is the diffusion-limitedelectron-transfer process. The electronic transfer resistance(Ret) which represents one interfacial property of the elec-trode is equal to the semicircle diameter. The impedancevalues deduced for a standard Randles circuit are

Cdl = (ω0Ret)−1 (1)

whenω → ∞,

Zre = Rs + Ret

1 + ω2C2dlR

2et

= Rs + ω20Ret

ω2 + ω20

(2)

Zim = ωCdlR2et

1 + ω2C2dlR

2et

= ωω0Ret

ω2 + ω20

(3)

whenω → 0,

Zre = Rs + Ret + σω−1/2 (4)

Zim = σω−1/2 + 2σ2Cdl = σω−1/2 + 2σ2

ω0Ret(5)

where σ is a constant, andω0 is found where Zimin the semicircle part is maximum and equalsRet/2[38].

If we take the typical Nyquist diagram into consideration,theRet in our protocol continued decreasing after hybridiza-tion and metallation with the diameter of the semicircle de-creasing. It was due to that M-DNA has better conductivitythan the double-stranded DNA and dsDNA is much con-ductively active when compared with single stranded DNA.And in our experiment we observed that theω0 of the sev-eral modified electrodes was corresponding to a frequencyof 54 690 Hz. Therefore, based on the deduced equation(1) with the decreasing ofRet, the double-layer capaci-tance (Cdl) for the modified electrodes was increasing fromssDNA/PPy/MWNT/GCE to M-DNA/PPy/MWNT/GCE.This was also found in literature[39]: if electronics dou-ble layers are formed on the electrode,Cdl will increase.In our protocol that the M-DNA was formed from ssDNAwould result in double layers forming on the PPy modi-fied electrode surface and as a resultCdl would increase.Additionally, based on the deduced equations (2)–(5),with Ret decreasedZre should decrease both during higherfrequencies and lower frequencies;Zim should decreasein high frequency range and increase in low frequency

Y. Xu et al. / Analytica Chimica Acta 516 (2004) 19–27 25

range. Our results (shown inFig. 4c and d) showed thechange.

However, after the M-DNA/PPy/MWNT modified elec-trode reacted with EDTA (disodium ethylenediaminetetraacetate) for 40 min, the Nyquist plot returned near tothe plot of the dsDNA/PPy/MWNT modified electrode.It was due to that metallized M-DNA could be easily re-versed to the normal double-stranded DNA in the presenceof EDTA by the complexation between Zn2+ living in theM-DNA helix strands and EDTA; when the complexationreaction occurred, the Zn2+ in the helix chains would bereadily replaced by H+ and which resulted in the normaldsDNA formation[27].

Fig. 6. (a) The plots of−�Zre vs. frequency from 105 down to 10 Hz with different complementary oligonucleotide concentrations and 10−7 Mnon-complementary sequence. Other conditions as inFig. 4. �Zre was theZre change from ssDNA/PPy/MWNT/electrode to M-DNA/PPy/MWNT/electrode.(b) The plots of−�Zre (at frequency= 5469 Hz) vs. complementary oligonucleotide concentration. Other conditions as inFig. 4.

3.3. Other experimental relevant conditions

In order to establish optimal conditions for impedancespectroscopy as a platform for the reagentless DNA-sensingassay based on MWNTs-COOH/PPy and metallation, sev-eral relevant experimental parameters, such as the MWNTs-COOH modification, the metallation time, the deferent metalions metalling dsDNA/PPy film and the hybridization tem-perature were investigated in our protocol. In our experi-ment, the real impedance (Zre) would decrease along withthe metallation time for 3 h when the impedance decreasedto a platform. Therefore, we chose 3 h as the metallationtime to make the metallation reach its equilibrium state. As

26 Y. Xu et al. / Analytica Chimica Acta 516 (2004) 19–27

we could see inTable 1, that on the non-MWNTs-COOHmodified electrode, the signal of DNA hybridization wasonly 20.8% compared with the one on the MWNTs-COOHmodified electrode. Which could be due to that with nan-otubes film offering very high real active surface area, moreprobe DNA would immobilized onto the electrode surfacewhen polymerization occurred.

We also investigated Co2+-DNA and Ni2+-DNA basedon the same protocol with 10−7 M target DNA. Aftermodifying electrode with carbon nanotubes, hybridizingDNA target with probe and metalling the dsDNA on thesame condition, there was a 82.1 and 80.3% impedancedecrease from ssDNA/PPy/MWNT/electrode to Co2+-DNA/PPy/MWNT/electrode and Ni2+-DNA/PPy/MWNT/electrode when compared with Zn2+-DNA/PPy/MWNT/electrode at 5469 Hz detected by�Zre, respectively. Theresult was consisted with the metal resistance sequence(when t = 20◦C, ρZn < ρCo ≈ ρNi). It also reveals thatcomparing the three M-DNAs, Zn2+-DNA had the low-est resistance for electron to transport in the helix chains,therefore we chose Zn2+ as the doping ion to decrease theimpedance property of DNA/PPy network and amplify thehybridization signal in this protocol.

The temperature for hybridization also has an obvious in-fluence on the detection sensitivity and selectivity of DNAhybridization. Herein, for the melting temperature relativeto this 24-mer-base ssDNA target based on the PPy film wasabout 55◦C, the hybridization amount would increase withthe hybridization temperature increasing till to about 40◦C;and to 50◦C the duplex DNA would gradually denature.Therefore, the temperature was chose at 38◦C where thehybridization for complementary sequence reached a maxi-mum however the 3-base mismatched sequence had a neg-ligible signal.

3.4. Sensitivity and selectivity of this DNA hybridizationdetection protocol

By varying the target oligonucleotide concentration from10−10 to 10−6 M in hybridization solution, we investigatedthe sensitivity of this DNA hybridization assay with�Zreas the detecting signal to quantify DNA hybridization. Itillustrates in Fig. 6a that the hybridization signal wouldbe increased with the concentration of complementary tar-get DNA increasing; and the analytical signal,�Zre (atthe frequency of 5469 Hz), could be described as a linearfunction of the logarithmic value of the target DNA con-centration (shown inFig. 6b). The regression equation wasY = 25.856 + 1.834X (X: log[target DNA], Y: −�Zre,frequency= 5469 Hz) with the correlation coefficient (R)0.98153. Detection limitation was 5× 10−11 M using 3S.D. of blank solution,n = 11. With the metallation am-plification step for the impedance detection, the detectionlimitation was expected to enhance as the result ofZredecreasing compared with the non-metallation detection.However, with 10−7 M non-complementary sequence in the

same hybridization and metallation conditions, the changein Zre could hardly be noticed.

The selectivity of this EIS hybridization assay was in-vestigated by detecting theZre change between ssDNA/PPyand M-DNA/PPy which formed with the same two steps:firstly, the ssDNA probes immobilized in PPy film was hy-bridized with 10−7 M complete matched 1-base mismatchedor 3-base mismatched sequences, respectively, then the DNAmodified electrode was metallized in the presence of 0.2 mMZn2+ (0.3 M NaCl+ 0.05 M Tris–HCl buffer, pH 8.6) for3 h. The results show that only the completely matched se-quence could give a significant response (listed inTable 1).Therefore, this protocol based on impedance detection andmetallation technology could differentiate complementarysequence and mismatched sequence. Besides, much clearerdifferentiation result could be obtained with the metalla-tion step: before metallation 1-mis and 3-mis matched se-quences had 51 and 8.2% hybridization signals comparedwith complementary sequence; after metallation the signalswere down to 35.7 and−5.8%. Thus and so, the metallationstep could enhance not only the detection limitation but alsothe ability of our protocol to recognize DNA sequences.

4. Conclusion

In conclusion, we have demonstrated oligonucleotideprobes could be doped into PPy film and immobilized ontothe carbon nanotube modified electrode. Such entrapmentallows ssDNA probes to hybridize target DNA sequencesmore easily. After the hybridization and metallation, ssDNAwas converted into M-DNA and concomitantly changedin the impedance values, especially decreased in the realimpedance (Zre). And both a decrease in electronics trans-port resistance (Ret) and an increase in double-layer ca-pacitance (Cdl) were obtained. The metallation step couldamplify the hybridization signal by 3.2 times comparedwith the non-metallation protocol. The fact that the sensorcan give difference between completely matched sequencesand mismatched sequences coupled with the simple andsensitive operation makes this protocol a hopeful method todetect DNA sequences.

Acknowledgements

We thank the National Nature Science Foundation ofChina (NSFC), which financially supported this work (No.29875008).

References

[1] K. Jacobs, S.F. Wolf, L. Haines, J. Fisch, J.N. Kremsky, J.P.Dougherty, Nucl. Acids Res. 15 (1987) 2911.

[2] D.J. Caruana, A. Heller, J. Am. Chem. Soc. 121 (1999) 769.

Y. Xu et al. / Analytica Chimica Acta 516 (2004) 19–27 27

[3] W. Kan, F.F. Chehab, Proc. Natl. Acad. Sci. U.S.A. 86 (1989) 9178.[4] A.P. Abel, M.G. Weller, G.L. Duveneck, M. Ehrat, H.M. Widmer,

Anal. Chem. 68 (1996) 2905.[5] S. Takenaka, K. Yamashita, M. Takagi, Y. Uto, H. Konodo, Anal.

Chem. 72 (2000) 1334.[6] J. Wang, X. Cai, G. Rivas, H. Shiraishi, Anal. Chim. Acta 326 (1996)

141.[7] J. Wang, J.R. Fernandes, L.T. Kubota, Anal. Chem. 70 (1998) 3699.[8] G. Marrazza, Z. Chianella, M. Mascini, Biosens. Bioelectron. 14

(1999) 43.[9] G. Marrazza, G. Chiti, M. Mascini, M. Anichini, Clin. Chem. 46

(2000) 31.[10] H. Cai, Y.Q. Wang, P.G. He, Y.Z. Fang, Anal. Chim. Acta 469 (2002)

165.[11] H. Cai, Y. Xu, N.N. Zhu, P.G. He, Y.Z. Fang, Analyst 127 (2002)

803.[12] T. de Lumley, C.N. Campbell, A. Heller, J. Am. Chem. Soc. 118

(1996) 5504.[13] O. Bagel, C. Degrand, B. Limoges, M. Joannes, F. Azek, P. Brossier,

Electroanalysis 12 (2000) 1447.[14] J. Wang, D. Xu, A. Erdem, R. Polsky, M.A. Salazar, Talanta 56

(2002) 931.[15] D.H. Johnston, K. Glasgow, H.H. Thorp, J. Am. Chem. Soc. 117

(1995) 8933.[16] E. Palecek, S. Biiova, L. Havran, R. Kizek, A. Miculkova, F. Jelen,

Talanta 56 (2002) 919.[17] S.O. Kelley, E.M. Boon, J.K. Barton, N.M. Jackson, M.G. Hill, Nucl.

Acids Res. 27 (1999) 4830.[18] E.M. Boon, D.M. Ceres, T.G. Drummond, M.G. Hill, J.K. Barton,

Nat. Biotechnol. 18 (2001) 1096.[19] H. Korri-Youssoufi, F. Garnier, P. Srivtava, P. Godillot, A. Yassar, J.

Am. Chem. Soc. 119 (1997) 7388.[20] J. Wang, M. Jiang, A. Fortes, B. Mukherjee, Anal. Chim. Acta 402

(1999) 7.

[21] C. Berggren, P. Stalhandske, J. Brundell, G. Johanson, Electroanalysis11 (1999) 156.

[22] I. Willner, F. Patolsky, A. Lichtenstein, Anal. Sci. 17 (2001) 1351.[23] L.Q. Luo, J.Y. Liu, Z.X. Wang, X.R. Yang, S.J. Dong, E.K. Wang,

Biophys. Chem. 94 (2001) 11.[24] G. Lillie, P. Payne, P. Vadgame, Sens. Actuators B 78 (2001)

249.[25] G. Farace, G. Lillie, T. Hianik, P. Payne, P. Vagama, Bioelectro-

chemistry 55 (2002) 1.[26] H. Cai, Y. Xu, P.G. He, Y.Z. Fang, Electroanalysis 15 (2003) 1864.[27] J.S. Lee, L.J.P. Latimer, R.S. Reid, Biochem. Cell Biol. 71 (1993)

162.[28] P. Aich, S.L. Labiuk, L.W. Tari, L.J.T. Delbaere, W.J. Roesler, K.J.

Falk, R.P. Steer, J.S. Lee, J. Mol. Biol. 294 (1999) 477.[29] A. Rakitin, P. Aich, C. Papadopoulos, Y. Kobzar, A.S. Vadeneev,

J.S. Lee, J.M. Xu, Phys. Rev. Lett. 86 (2001) 3670.[30] R.P. Ianek, W.R. Fawcett, A. Ullman, Langmuir 14 (1998) 3011.[31] S.D. Wettig, D.O. Wood, J.S. Lee, J. Inorg. Biochem. 94 (2003)

94.[32] S.D. Wettig, C.Z. Li, Y.T. Long, H.B. Kraatz, J.S. Lee, Anal. Sci.

19 (2003) 23.[33] J.M. Pernaut, R.C. Peres, V. Juliano, M.A. De Paoli, J. Electroanal.

Chem. 274 (1989) 225.[34] J.H. Chen, Z.P. Huang, D.Z. Wang, S.X. Yang, W.Z. Li, J.G. Wen,

Z.F. Ren, Synth. Met. 125 (2002) 289.[35] J.H. Fan, M.X. Wan, D.B. Zhu, B.H. Chang, Z.W. Pan, S.S. Xie,

Synth. Met. 102 (1999) 1266.[36] M. Sluyters-Rehbach, J.H. Sluyters, in: A.J. Bard (Ed.), Electroana-

lytical Chemistry, vol. 4, Marcel Dekker, New York, 1970.[37] A. Bardea, E. Katz, I. Willner, Electroanalysis 12 (2000) 1097.[38] M.S. Lun, AC Impedance Spectroscopy Principles and Applications,

National Defence and Industry Press, Beijing, 2001.[39] C. Xu, H. Cai, Q. Xu, P.G. He, Y.Z. Fang, Fresen. J. Anal. Chem.

369 (2001) 428.