Label free electrochemical immunoaffinity sensor based on impedimetric method for pesticide detection

Phama tel: + 33-1-57277224 b Institute for Tropical Technology ITT, Vietnamese Academy of Science and Technology, 18 Hoang Quoc Viet, Hanoi, Vietnam *e-mail: piro@univ-paris-diderot.fr R

Label-Free Electrochemical Immunoaffinity Sensor Based on

Impedimetric Method for Pesticide Detection

H V Tran,a

S Reisberg,a

B Piro,*a

T D Nguyen,b

M C Phama

tel: + 33-1-57277224

b Institute for Tropical Technology (ITT), Vietnamese Academy of Science and Technology, 18 Hoang Quoc Viet, Hanoi, Vietnam

*e-mail: piro@univ-paris-diderot.fr

Received: July 4, 2012

Accepted: October 10, 2012

Published online: January 24, 2013

Abstract

We present a new approach using a conducting polymer to combine immunoaffinity and electrochemical impedance

spectroscopy for atrazine detection The system is based on a competitive complexation between atrazine

anti-body and either atrazine present in the analyzed sample or hydroxyatrazine immobilized on the sensor surface The

process allows to detect atrazine at a very low detection limit (0.2 ng L1) in a true label-free format (no redox

probe added in solution) by following changes in the electrochemical impedance of the sensor.

Keywords: Conducting polymers, Electrochemical impedance spectroscopy (EIS), Enzyme-linked immunosorbent

assays (ELISA), Immunosensors, Label-free detection, Cross-reactivity, Atrazine, Pesticides

DOI: 10.1002/elan.201200331

1 Introduction

The increased use of pesticides and herbicides has led to

serious problems of contamination of soil and water The

maximum value for atrazine in drinkable water set by the

World Health Organization is 2 mg L1, and this limit is

even lower in the European Union (0.1 mg L1) Given

this, atrazine has been often chosen as a model to develop

analytical devices

Immunoassay technology with enzyme-linked

immuno-sorbent assays (ELISA) is now seen as a gold-standard of

immunoassays for pesticide analysis The immunoassays

kits are inexpensive, simple, adaptable to field use and

constitute a rapid way to determine contaminants in

envi-ronmental samples [1, 2] ELISA test is based on

anti-body technology and involves the immobilization of a

re-actant (an antibody, an antigen or a part of an antigen,

called hapten) onto a solid surface with enzymes being

used as markers for the presence of a specific

antibody-antigen (Ab/Ag) or antibody-hapten (Ab/Hp) complex

However, ELISA tests are not efficient for simultaneous

multiple analysis (multiplexing), continuous detection and

transduction into an electronically processable signal

Conversely, label-free electrochemical immunosensors are

efficient for these tasks [3, 4] Among them, impedimetric

immunosensors are the subject of special attention [5–7]

Electrochemical Impedance Spectroscopy (EIS)

consti-tutes an efficient way to follow the Ab/Ag or Ab/Hp

in-teractions at electrode surfaces, by probing changes in ion

diffusion and electrical capacitance [8] In the last five

years, the applications of EIS to the detection of small an-tigenic organic pollutants, as bisphenol A [9], organic toxins [10, 11] or pesticides, mainly atrazine [7, 12, 13] have shown growing interest Hleli et al reported a detec-tion limit of 20 ng mL1 atrazine using an impedimetric immunosensor based on mixed biotinylated self-assem-bled monolayer [14] Later, Cosnier et al reported the detection of extremely low atrazine concentration (10 pg mL1) using a label-free impedimetric immunosen-sor based on a conducting polymer cleverly modified to bind the antibody probe using affinity interactions [7] They showed that the immunoreaction of atrazine on the attached anti-atrazine antibody induces an increase in the charge transfer resistance proportional to the atrazine concentration

Among the key steps in the design of an immunosen-sor, the immobilization of the bioreceptor is decisive The introduction of appropriate functionalities through chemi-cal modification of a monomer can provide polymer films with specific characteristics Conducting polymers such as polypyrrole or polyaniline have been extensively studied for their great functionalization potentialities and advan-tageous electrochemical properties [15–18] Polyquinone derivatives are less investigated but, nevertheless, present great and remarkable electrochemical properties as well

as good biocompatibility, easy bio-functionalization and

a very stable electroactivity in neutral aqueous medium [19] These properties can be used to probe biomolecular interactions [20–24] due to the high sensitivity of the qui-none group to its local physico-chemical environment

In the present study, we describe an innovative strategy

based on an original electrogenerated polyquinone film

functionalized by a hydroxyatrazine moiety (HATZ) for

sensitive and direct detection of atrazine (ATZ) The

syn-thesized monomer

[N-(6-(4-hydroxy-6-isopropylamino-

1,3,5-triazin-2-ylamino)hexyl)-5-hydroxy-1,4-naphthoqui-none-3-propionamide] (JUG-HATZ) contains three

func-tional groups: the hydroxyl group for

electropolymeriza-tion, the quinone group to be used as transducer, and

hy-droxyatrazine (a structural analogue of ATZ) as

biore-ceptor element (Scheme 1) Electropolymerization of

JUG-HATZ leads to

poly[N-(6-(4-hydroxy-6-isopropyla-

mino-1,3,5-triazin-2-ylamino)hexyl)5-hydroxy-1,4-naph-thoquinone-3-propionamide], poly(JUG-HATZ) [25] By

this method, the quinone and the hydroxyatrazine

func-tions are preserved We have shown that

poly(JUG-HATZ) is able to specifically bind a-ATZ, the antibody

directed to atrazine, due to cross-reactivity of a-ATZ for

HATZ a-ATZ is eventually removed from the electrode

surface if ATZ (the natural ligand of a-ATZ) is present

in solution (see Figure 1) This architecture differs from

other devices used in label-free impedimetric

imunosen-sor, for which the antibody is immobilized on the surface

In the present work, electrochemical impedance

spectros-copy (EIS) is used to identify the physico-chemical

prop-erties of the electrode/electrolyte interface (such as

ca-pacitances and charge transport resistances), which are

dramatically influenced by the antibody

binding/unbind-ing, Changes in these parameters are significant and can

be used individually to determine ATZ in solution, for

concentration as low as 0.2 pg mL1

2 Experimental

2.1 Chemicals

All reagents and solvents were PA type Phosphate buffer

saline (PBS, 137 mM NaCl; 2.7 mM KCl; 8.1 mM

Na2HPO4; 1.47 mM KH2PO4, pH 7.4) was provided by

Sigma Aqueous solutions were made with ultrapure

(18 MW cm) water Glassy carbon electrodes (GC, 3 mm

in diameter, S = 0.07 cm2

) were purchased from BASInc

Anti-atrazine (a-ATZ) antibody (monoclonal) was

pur-chased from Thermo Scientific, USA Atrazine (ATZ)

and atrazine-desethyl-2-hydroxy (ATD) were purchased

from Supelco (USA) 5-hydroxy-1,4-naphthoquinone

(JUG), 1-naphthol (1-NAP), lithium perchlorate

(LiClO4), acetonitrile (ACN) and ethanol (EtOH),

practi-cal grade, were from Sigma-Aldrich Dry Argon (Ar) was

purchased from Air Liquide (France)

N-(6-(4-hydroxy-6-

isopropylamino-1,3,5-triazin-2-ylamino)hexyl)-5-hydroxy-1,4-naphthoquinone-2(3)-propionamide (JUG-HATZ)

and 5-hydroxy-3-thioacetic acid-1,4-naphthoquinone

(JUGA) were synthesized in the lab [21]

2.2 Preparation of the Poly(JUG-HATZ)-Modified Electrodes

GC electrodes were polished by 1 mm alumina slurry then sequentially washed in an ultrasonic bath with distilled water, EtOH and ACN for 5 minutes The electrochemi-cal synthesis of the polymer films was carried out by elec-trooxidation of 5  103M JUG-HATZ + 103M 1-NAP + 0.1 M LiClO4 in ACN on GC electrodes, under Ar, at

a constant potential of 1.25 V (vs SCE) during 600 s After that, electrodes were repeatedly washed with aceto-nitrile to remove residual monomers then put into an electrochemical cell containing PBS and scanned between

0.9 V and 0 V at a scan rate of 50 mV s1

under Ar at-mosphere until complete stabilization of the voltammo-grams After this treatment, electrodes were characterized

by electrochemical impedance spectroscopy (EIS)

after complexation and decomplexation.

Electroanalysis 2013, 25, No 3, 664 – 670  2013 Wiley-VCH Verlag GmbH & Co KGaA, Weinheim www.electroanalysis.wiley-vch.de 665

2.3 Complexation of a-ATZ on

Poly(JUG-HATZ)-Modified Electrodes

Poly(JUG-HATZ)-modified electrodes were immersed

into PBS containing a-ATZ (the concentration of which

depends on the experiment) overnight at 37 8C to achieve

complexation between HATZ and a-ATZ After that, the

electrodes were rinsed thrice with ultrapure water and

immersed into PBS for 1 h at 37 8C to remove

nonspecifi-cally adsorbed a-ATZ After this step, electrodes were

characterized by EIS Such electrodes are referred as

poly(JUG-HATZ)/a-ATZ The maximum surface

concen-tration of a-ATZ is ca 0.2 pmol cm2, i.e 109

antibodies

on an electrode area of 0.07 cm2

2.4 Detection of ATZ

For ATZ detection, poly(JUG-HATZ)/a-ATZ electrodes

were dipped 2 h at 37 8C into a 100 mL sample of water

containing ATZ at concentrations between 0.1 pM and

0.1 mM, then washed with PBS at 37 8C for 30 min After

that, electrodes were characterized by EIS

2.5 Methods

EIS was performed using an Autolab PGSTAT30

equipped with the FRA module Impedance spectra were

recorded in PBS buffer at room temperature at a given

potential within a frequency range from 10 kHz to

100 mHz with a perturbation amplitude of 10 mV As

a pretreatment before each experiment, a constant

poten-tial corresponding to the one used for EIS was imposed,

for 120 s Solutions were systematically deaerated with Ar

before and during experiments

3 Results and Discussion

3.1 JUG-HATZ Electroactivity

JUG, JUGA and JUG-HATZ were dissolved into PBS

buffer at a concentration of 0.1 mM Their electroactivity

was investigated using cyclic voltammetry from0.7 V to

0.1 V (vs SCE) at a scan rate of 50 mV s1using GC

elec-trodes The results are showed on Figure 2 The redox

peaks (Epa/Epc) of JUG, JUGA and JUG-HATZ are

situ-ated at 0.27/0.31 V; 0.28/0.33 V and 0.29/0.36 V

vs SCE, respectively Averaged peak potentials (E1/2) are

0.29 V; 0.30 V and 0.33 V (vs SCE), respectively

Changes in potentials for JUGA and JUG-HATZ

com-pared to JUG reflect the presence of lateral chains of the

substituents, which are slightly electrodonors (so that the

reduction of the quinone moiety is slightly more difficult

and occurs at lower potentials) Changes in currents, for

the same concentration, reflect a decrease of the diffusion

coefficient for JUGA and JUG-HATZ compared to

JUG

3.2 EIS on Poly(JUG-HATZ)-Modified Electrodes

Electroactivity of poly(JUG-HATZ)-modified electrodes

in PBS buffer is shown in Figure 3 Two main redox cou-ples are present, centered at around 0.44/0.53 V and

0.72/0.76 V vs SCE A third couple appears as should-ers at 0.27/0.30 V vs SCE This redox system is com-parable to that of the monomer (Figure 2), with widened peaks due to electronic delocalisation within the polymer structure In aqueous solution, quinones transfer two elec-trons and two protons in a concerted process As previ-ously reported, we assume that the presence of three cou-ples of peaks in PBS is due to three different types of

JUG-HATZ Concentration: 104M Scan rate 50 mV s1 Potential range: 0.7 to 0.1 V (vs SCE) Medium: PBS GC electrode, S = 0.07 cm 2

.

argon-satu-rated solution of PBS (pH 7.4), scan rate 50 mV s 1 E = 0.9 to

0 V (vs SCE).

nones on the electrode surface: at the electrode/polymer

interface, in the bulk of the polymer and at the polymer/

electrolyte interface [20–25] The electrochemical stability

of this modified electrode is excellent It takes ca 50

vol-tammetric cycles to stabilize the system in PBS [25] then

it becomes perfectly stable (no measurable changes) for

more than 250 cycles, or several days of storage in PBS

Electrochemical Impedance Spectroscopy (EIS) is an

efficient method to probe the interfacial properties of

sur-face-modified electrodes [26] We performed EIS on

poly(JUG-HATZ)-modified electrodes in the same

po-tential domain than for cyclic voltammetry (Figure 4)

To fit the experimental spectra, the equivalent circuit

presented on Figure 5 was used It includes Rf, Refand Rs

which are the resistances of the film, of the

electrolyte-film interface and of the electrolyte, respectively The

constant phase elements represent the

pseudocapacitan-ces of the electrolyte-film interface (CPEef) and of the film (CPEf), respectively In addition, we introduced an-other constant phase element (CPEd) to characterize dif-fusion of ions from the electrolyte bulk to the electrode, which better fits experimental data than the classical War-burg element [27]

The impedance of a CPE is expressed as

ZCPE¼ Q1

0 wn

with n a corrective term between 1 (perfect capacitor) and 0, and Q0=1/j Z j for n = 1 and w = 1 rad s1 For our fitting, values of n were kept between 0.8 and 1

Values extracted from fittings (given on Figure 6) are

in good agreement with cyclic voltammetry (Figure 3): diffusion (CPEd) appears to be maximal at around

0.4 V, where the slope of the film capacitance (CPEf) is the highest The electrolyte-film capacitance (CPEef) is the lowest at 0.5 V, as well as the film resistance (Rf) These last three parameters are the most informative about changes which can occur on the electrode surface and can be used to characterize, at a given potential, anti-body complexation or decomplexation

3.3 ATZ Detection

It is noteworthy to recall the ability of the combining site

of an antibody to react with more than one antigenic de-terminant This is known as cross-reactivity and arises be-cause the cross-reacting antigen shares a structure similar

to that of the immunizing antigen (here ATZ) used to generate antibodies The phenomenon of cross-reactivity

is an intrinsic characteristic of all antibodies but depends also on the relative concentrations of cross-reactant In our case, the antibody a-ATZ generated from ATZ (chlorinated s-triazine) can also bind to HATZ (the hy-droxylated s-triazine), but with lower affinity [28–30] We took advantage of this cross-reactivity to design our sens-ing strategy As explained, poly(JUG-HATZ) is able to bind to a-ATZ, which covers the electrolyte-film inter-face However, a-ATZ recognizes HATZ with less

affini-ty than for ATZ (which is the natural ligand of a-ATZ)

so that, when put into contact with ATZ, a-ATZ is re-moved from the electrode surface (this is illustrated on Figure 1)

We have demonstrated in previous studies that the electroactivity of polyquinone-modified electrodes is sen-sitive to changes in cation diffusion, sodium in particular [19] This can be generated by heavy molecules at the vi-cinity of the film surface, influencing the diffusion layer

by steric hindrance This is what occurs when a-ATZ is present at the electrolyte-film interface a-ATZ has a mo-lecular weight of ca 150 kDa and an hydrodynamic volume of about 25 nm3

, which makes a projected surface

of 252/3=8.5 nm2

, so that a-ATZ can strongly inhibits ion exchange Upon ATZ detection, the antibody is released and the electrolyte-film interface is cleared, so that the

electrode at different formal DC potentials vs SCE in 100 mM

PBS buffer solution, pH 7.4 Spectra are presented between

100 kHz to 1.265 Hz Symbols correspond to experimental data

and lines correspond to fittings.

Fig 5 Equivalent circuit used to fit experimental spectra R s is

the solution resistance, R ef and CPE ef are the electrolyte-film

in-terface resistance and pseudocapacitance (CPE is a constant

phase element), R f and CPE ef are the film resistance and

pseudo-capacitance For the CPE ef , values of n were kept between 0.9

and 1.

Electroanalysis 2013, 25, No 3, 664 – 670  2013 Wiley-VCH Verlag GmbH & Co KGaA, Weinheim www.electroanalysis.wiley-vch.de 667

film resistance Rf associated to ion diffusion is expected

to decrease

For the same reasons, a-ATZ decomplexation should

influence the film pseudocapacitance, CPEf, considering

that this pseudo-capacitance comes from ion diffusion

The electrolyte-film interface pseudocapacitance CPEef

should also change when a-ATZ is removed, for a

differ-ent reason

CPEefis given by:

CPEef ¼ ee0=d

with e the relative permittivity of the medium, e0the

per-mittivity of vacuum and d the double layer thickness The

antibody at the electrolyte-film interface does not change

the permittivity of the medium but due to steric

hin-drance, it significantly increases the double layer

thick-ness d On the contrary, when the antibody is removed,

this double layer thickness decreases, so that CPEef

should increase Rf, CPEfand CPEef are given as a

func-tion of ATZ concentrafunc-tion, on Figure 7 Values for Rf,

CPEf and CPEef before addition of ATZ are 1306 W, 1.05 mF and 0.12 mF, respectively (these values appear as dashed lines on Figure 7) As shown, at an ATZ concen-tration of 1013M, changes for Rf, CPEf and CPEef are not significant, due to the standard deviation However, for 1012M, ca 0.2 ng L1, variations become significant This value can be considered as the limit of detection These variations are relatively linear (in a log scale) up to

107M (ca 20 mg L1), which covers practical concentra-tions Considering the size of a-ATZ, the maximum sur-face concentration of a-ATZ is ca 109

antibodies on an electrode area of 0.07 cm2

This value is in accordance with the lowest detectable quantity of ATZ which is 6 

108 molecules (1012M in 100 mL)

3.4 Selectivity The transduction mechanism is based on the cross-reac-tivity of a-ATZ towards HATZ That is why it is impor-tant to evaluate the selectivity of this sensor towards an-other structural analogue of atrazine, for example

Fig 6 Effect of the offset potential on the characteristics of poly(JUG-HATZ) modified electrode: (a) diffusion peudo-capacitance, (b) film pseudo-capacitance, (c) electrolyte-film interfacial pseudo-capacitance, and (d) film resistance Results obtained for offset po-tentials between 0.8 V and 0 V vs SCE, in PBS, fitted from the equivalent circuit of Figure 5 in the range of 100 kHz–1.265 Hz For the CPEs, values of n were kept between 0.8 and 1.

amino-4-chloro-6-isopropylamino-1,3,5-triazine (desethy-latrazine ATD, see Scheme 1)

Experiments were performed as described in Section 3.3, with ATD then with ATZ, at two different concentra-tions: 1 nmol L1and 10 nmol L1 Relative changes in re-sistance (Rf), film capacitance (CPEf) and electrolyte-film capacitance (CPEef) are summarized and compared in Table 1

It appears that changes related to ATD are systemati-cally and significantly lower than those related to ATZ For CPEef, changes to ATD represents only 34 % of the one obtained for ATZ, at 10 nM ATD As expected, the selectivity is lower for high concentrations than for low concentrations

Scheme 1 Chemical structures of atrazine (ATZ), its 2-hydroxy analog (HATZ), its dealkylated derivative (ATD) and JUG-HATZ.

Table 1 Relative changes in resistance (DR f ), film capacitance (DCPE f ) and electrolyte-film capacitance (DCPE ef ) as a function

of the target DR f (%) = 100  (Ra-ATZR a-ATZ/antigen )/Ra-ATZ; DCPE f

(%) = 100  (CPE f a-ATZ CPE f a-ATZ/antigen )/CPE f a-ATZ ; DCPE ef

(%) = 100  (CPE ef a-ATZ CPE ef a-ATZ/antigen )/CPE ef a-ATZ

Fig 7 Changes in films resistance (R f ); capacitance of

conduc-tive film (CPE f ) and capacitance of film-electrolyte interface

(CPE ef ) as function of the concentration of atrazine in solution

in the decomplexation step Results obtained for an offset

poten-tial of 0.5 V vs.SCE, fitted from the the equivalent circuit of

Figure 5, in the range of 100 kHz–1.265 Hz For the CPEs, values

of n were kept between 0.8 and 1 Values for R f , CPE f and CPE ef

before addition of ATZ appear as dashed lines.

Electroanalysis 2013, 25, No 3, 664 – 670  2013 Wiley-VCH Verlag GmbH & Co KGaA, Weinheim www.electroanalysis.wiley-vch.de 669

4 Conclusions

We have shown that atrazine can be directly detected by

electrochemical impedance spectroscopy using a novel

multifunctional conducting polymer, poly(JUG-HATZ)

It acts both as the complexation and the transduction

ele-ment for a label-free, reagentless electrochemical

immu-nosensor Experiments show that atrazine can be detected

down to ca 1 pM (~ 0.2 pg mL1), which is an excellent

result compared to similar immunosensors based on EIS

[7, 10–13] This was obtained when using an original

transduction scheme which takes profit from the

cross-re-activity of the antibody The sensing sequence consists in:

1) immobilization of anti-ATZ antibodies on an

electro-active surface carrying an analogue of ATZ (HATZ); 2)

decomplexation of anti-ATZ by ATZ present in the

ana-lyzed solution, which triggers changes in several

electro-chemical impedance parameters such as the film

resist-ance, the film capacitance and the interfacial

electrolyte-film capacitance This approach is versatile: it can be

ex-tended to other targets (pesticides, organic pollutants or

toxins) providing that they are antigenic Works are in

progress with such molecules, e.g bisphenol A,

ochratox-in and isoproturon

References

[1] C Mouvet, S Broussard, R Jeannot, C Maciag, R

Abu-knesha, G Ismail, Anal Chim Acta 1995, 311, 311.

[2] G Xiong, J Liang, S Zou, Z Zhang, Anal Chim Acta

1998, 371, 97.

[3] A P F Turner, I Karube, G S Wilson, Biosensors:

Funda-mentals and Applications Oxford University Press 1989.

[4] J L Marty, B Leca, T Noguer, Analusis 1998, 26, M144.

[5] E Katz, I Willner, Electroanalysis 2003, 15, 913.

[6] A Ramanavicius, A Finkelsteinas, H Cesiulis, A

Ramana-viciene, Bioelectrochemistry 2010, 79, 11.

[7] R E Ionescu, C Gondran, L Bouffier, N

Jaffrezic-Re-nault, C Martelet, S Cosnier, Electrochim Acta 2010, 55,

6228.

[8] A Sargent, O A Sadik, Electrochim Acta 1999, 44, 4667.

[9] M A Rahman, M J.A Shiddiky, J S Park, Y B Shim, Bio-sens Bioelectron 2007, 22, 2464.

[10] A E Radia, X Munoz-Berbel, V Lates, J L Marty, Bio-sens Bioelectron 2009, 24, 1888.

[11] A Viga, A Radoi, X Munoz-Berbel, G Gyemantc, J L Marty, Sens Actuators B 2009, 138, 214.

[12] E Valera, J Ramon-Azcon, A Rodrıguez, L M Castaner,

F J Sanchez, M P Marco, Sens Actuators B 2008, 129, 921 [13] J Ramon-Azcon, E Valera, A Rodrıguez, A Barranco, B Alfaro, F Sanchez-Baeza, M P Marco, Biosens Bioelectron.

2008, 23, 1367.

[14] S Hleli, C Martelet, A Abdelghani, N Burais, N Jaffrezic-Renault, Sens Actuators B 2006, 113, 711.

[15] S Cosnier, Biosens Bioelectron 1999, 14, 443.

[16] M Gerard, A Chaubey, B D Malhotra, Biosens Bioelec-trontron 2002, 17, 345.

[17] S Cosnier, Anal Bioanal Chem 2003, 377, 507.

[18] S Cosnier, M Holzinger, Chem Soc Rev 2011 40, 2146 [19] B Piro, J Haccoun, M C Pham, L D Tran, A Rubin, H Perrot, C Gabrieli, J Electronal Chem 2005, 577, 155 [20] S Reisberg, D F Acevedo, A Korovitch, B Piro, V Noel,

M C Pham, Talanta 2010, 80, 1318.

[21] B Piro, A Kapella, V H Le, G Anquetin, Q D Zhang, S Reisberg, V Noel, L D Tran, H T Duc, M C Pham, Elec-trochim Acta 2011, 56, 10688.

[22] Q D Zhang, G March, V Noel, B Piro, S Reisberg, L D Tran, L V Hai, E Abadia, P E Nielsen, C Sola, M C Pham, Biosens Bioelectron 2012, 32, 163.

[23] M C Pham, B Piro, L D Tran, T Ledoan, L H Dao, Anal Chem 2003, 75, 6748.

[24] B Piro, Q D Zhang, S Reisberg, V Noel, L A Dang,

H T Duc, M C Pham, Talanta 2010, 82, 608.

[25] H V Tran, S Reisberg, B Piro, N Serradji, T D Nguyen,

L D Tran, C Z Dong, M C Pham, Biosens Bioelectron.

2012, 31, 62.

[26] C Gabrielli, O Haas, H Takenouti, J Appl Electrochem.

1987, 17, 82.

[27] J Bisquert, G Garcia-Belmonte, P Bueno, E Longo,

L O S Bulhes, J Electroanal Chem 1998 452, 229 [28] J M Schlaeppi, W Foery, K Ramsteiner, J Agric Food Chem 1989, 37, 1532.

[29] M Wortberg, G B Jonesb, S Kreissig, D M Rocke, S J Gee, B D Hammock, Anal Chim Acta 1996, 319, 291 [30] K Charlton, W J Harris, A J Porter, Biosens Bioelectron.

2001, 16, 639.