Towards the development of an integrated capillary electrophoresis optical biosensor

Alessandra Bossi1

Laura Castelletti1

Sergey A. Piletsky2

Anthony P. F. Turner2

Pier Giorgio Righetti1

1Department of Science andTechnology,University of Verona,Verona, Italy

2Institute of BioScience andTechnology,Cranfield University at Silsoe,Silsoe, Bedfordshire, UK

Towards the development of an integrated capillaryelectrophoresis optical biosensor

Extending the previous preliminary study on the construction of a capillary electro-phoresis (CE)/sensor for the detection of reducing analytes, we focus the interest onthe simultaneous detection of redox active species, which are important indicatorsof the oxidative damage in tissues, of food preservation, and of pollution. The CE/sensor was built by modifying the detector-portion of the capillary with the redox-sensitive polymer polyaniline (PANI). The analyte is detected by monitoring thechanges in optical absorption of the PANI film. The CE/sensor was tested, withgood results, with ascorbic acid, glutathione (GSH), as well as with compounds withvery close similarity (ascorbic and isoascorbic acid). The kinetics of oxidation andreduction of PANI were evaluated. Further a PANI/CE-biological sensor was devel-oped by coupling an enzyme, glucose oxidase (GOD), to the PANI-modified portionof the capillary. The stability of the immobilized GOD and the sensitivity of the CE/biosensor were studied, by using glucose as test analyte in concentrations within thephysiological range. The results indicate that the CE/biosensor had good stability(more than 75% of original activity retained after 30 operational days), manufacturingreproducibility and a sensing range convenient for monitoring physiological glucose(1–24 mM).

Keywords: Biosensor / Capillary electrophoresis / Chemical sensor / Glucose oxidase / Redoxanalytes DOI 10.1002/elps.200305588

1 Introduction

Several successful attempts to achieve an integration be-tween capillary electrophoresis (CE) and sensors havebeen reported [1–3]. The integrated CE/sensors wereinspired by the desire to provide a second dimension tothe separation process thus, analytes benefit, in additionto the separation based on the electrophoretic mobility, ofa selection and discrimination achieved by the sensorcomponent. Mostly electrochemical (EC) detection wascoupled with CE, gaining significance in analyticalscience, especially for inorganic ions and analytes lackingof absorbance/fluorescence properties, offering a widerange of sensitivities [4]. Technical improvements allowedto minimize the interferences between the separation andthe sensing voltages. The fixed voltage or current appliedto the EC detector is the factor discriminating for the mostthe class of analytes sensed, but as the analyte pos-sesses a high oxidation potential, the sensor resultsresponsive to a number of interfering substances. Fewerexamples exist of optical sensors coupled to CE, despite

some results indicate high sensitivities (nM range) asshown for fiber-optic sensors developed for environmen-tal measurements [2, 5].

Exploring new optical-sensors assets, we exploited theoptical properties of the semiconductive polymer polyani-line (PANI) [6, 7] for the development of a CE-integratedoptical sensor for the detection and quantification ofredox species such as L-ascorbic acid (L-AA) [8]. The opti-cal sensor was obtained by modification of the detectorportion of the capillary column with a thin film of theredox-sensitive conjugated polymer PANI. The injectionof AA resulted in a reduction reaction that changed theoptical absorbance of PANI generating a detectable sig-nal. The CE/sensor showed short analysis time (ca. 2 min)and provided the possibility to work in different ranges ofanalyte concentration (low ppm or hundreds of ppm).

As redox analytes are implicated in oxidative stress, i.e.,environmental pollution [9] and health (cancer and agingprocesses) [10–13], we expand our work on CE-polyani-line redox sensor (CE/PANI sensor), for the simultaneousdetection and quantification of redox-active species.Chosen analytes were ascorbic acid (AA), glutathione(GSH), D-isoascorbic acid (IAA) and hydrogen peroxide(H2O2). The CE/PANI sensor was further modified withan enzyme to overcome the limits of application imposed

Correspondence: Dr. Alessandra Bossi, Department of Scienceand Technology, University of Verona, Strada le Grazie 15,I-37134 Verona, ItalyE-mail: [email protected]: +39-045-8027929

3356 Electrophoresis 2003, 24, 3356–3363

2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Electrophoresis 2003, 24, 3356–3363 Integrated CE optical sensor and biosensor 3357

by chemical sensing to the sole redox-active analytesand generalizing the use of this CE-sensing device tonumerous classes of compounds which are substratesof enzymes such as oxidases and dehydrogenases [14,15].

A CE/biosensor for the determination of glucose wasdeveloped by coupling the enzyme glucose oxidase(GOD) to the PANI layer. The method for immobilizingGOD on PANI was covalent, via the bifunctional reagentglutaraldehyde, as described earlier [16, 17]. PANI worksas an electron acceptor for the enzyme reaction (seeFig. 1) [18]. The CE/biosensor takes advantage of thepossibility to achieve a direct electron transfer betweenGOD and the PANI layer, as previously demonstrated inexperiments by measuring the oxidation currents in glu-cose solutions at a fixed potential [19]. The stability ofthe immobilized GOD and the sensitivity of the CE/bio-sensor were studied, using glucose as test analyte in thephysiological range concentrations. Results indicate thatthe CE/biosensor had good stability, manufacturing re-producibility, and a sensing range convenient for monitor-ing physiological glucose. The CE-coupled optical bio-sensor offers a new class of analytical devices, whosecharacteristics meet the demand for integrated analysisof complex samples.

Figure 1. Scheme of the optical biosensor made by GODas biological element and a PANI layer as sensing ele-ment. Glucose is oxidized by GOD, the produced elec-trons are directly transferred to PANI, giving rise to achange in the optical properties of the polymer (changein the ratio phenylene/quinidine units), which is detectableat 650 nm.

2 Materials and methods

2.1 Reagents

Aniline hydrochloride, ammonium persulfate, AA, asparticacid, glucose, GOD from Aspergillus niger type VII,sodium phosphate, potassium phosphate, and sodiumtetraborate were from Sigma (St. Louis, MO, USA). Hydro-chloric acid was from Fluka (Buchs, Switzerland). Allreagents, purchased from commercial sources, were ofanalytical grade. The trisubstituted piperazine [N(methyl-N-�-iodobutyl)-N’-methylpiperazine] (Q-Pip) was kindly

provided by Prof. A. Citterio, Politecnico of Milan, Italy.Fused-silica capillaries (50 �m ID�375 �m OD) werepurchased from Polymicro Technologies (Phoenix, AZ,USA).

2.2 Modification of capillaries with PANI

Capillaries were modified as described previously [8].After treatment with 0.1 M NaOH for 2 h and water for 3 h,capillaries of 24 cm total length and 19.4 cm effectivelength were coated with the PANI film in the portion be-tween the detector window and the outlet. Chemical po-lymerization of aniline/HCl with ammonium persulfate300:25 (mM:mM) was carried out for 30 min. Coated capil-laries were washed thoroughly with 1 M HCl solution, Milli-Q water, and stored at room temperature.

2.3 CE/sensor running conditions

CE analysis were performed on a Bio-Rad (Hercules, CA,USA) BioFocus 3000 instrument, at 25�C, with PANI-mod-ified capillaries (50 m ID). Prior to start, capillaries weretreated with 50 mM ammonium persulfate (400 psi�s) tooxidize the PANI film, washed with the quaternary pipera-zine Q-Pip (500 psi�s) to confer an uniform positivecharge to the capillary walls and conditioned with runningbuffer (400 psi�s) [20]. The running buffer was 250 mM

borate, pH 8.0. Reducing analytes (L-AA and GSH) wereinjected by pressure (2 psi�s). The applied voltage was4 kV. The detection wavelength was 650 nm. Redox activ-ity of the analytes was detected as marked irreversibledecrease in absorbance at 650 nm. Riboflavin-5P wasadded to the samples as internal standard and detectedat 444 nm. For oxidizing compounds the capillary wastreated with 300 mg/L AA (400 psi�s) to reduce the PANIfilm, washed with Q-Pip (500 psi�s) and conditioned with50 mM aspartic acid at pH 2.7 (400 psi�s). Hydrogen per-oxide with a concentration range varying from 3 to 30%was injected (25 psi�s) and PANI oxidation was revealedby a sharp increase in absorbance at 650 nm. Every tenruns the capillary was treated with ammonium persulfateor AA, respectively to recover the initial oxidation orreduction state of the PANI film.

2.4 Simultaneous determination of L-AA andD-IAA in the CE/sensor

Running conditions for the simultaneous determination ofL-AA and D-IAA were 250 mM borate buffer at pH 8.0, 4 kVapplied voltage, and detection wavelength at 650 nm.L-AA and D-IAA in the concentration range of 100–1 mg/Lwere injected by pressure (2 psi�s).


CE

and

CE

C

3358 A. Bossi et al. Electrophoresis 2003, 24, 3356–3363

2.5 Kinetics of reduction and oxidation ofPANI films

Methacrylate cuvettes were coated with a PANI filmgrafted using the same protocol as for capillaries (see Sec-tion 2.2). Modified cuvettes were conditioned in 50 mM

aspartic acid, pH 2.7. Samples of AA, GSH, hydrogen per-oxide, and potassium dichromate with different concen-trations were added to the modified cuvette. The kineticsof oxidation and reduction was observed by monitoring thespectrophotometric changes of absorbance at 650 nmover time. Experimental curves were then analyzed withequations for the fitting of sigmoidal curves, e.g., Boltz-mann equation and dose-response curves.

y � bottom � top � bottom

1 � eV50�xSlope

2.6 The CE/biosensor

Methacrylate cuvettes were coated with a PANI film, asdescribed in Section 2.5, then treated with 0.1 M ammo-nium hydroxide to convert the PANI salt to base form andwashed thoroughly with deionized water. PANI was con-ditioned with 50 mM phosphate buffer at pH 5.8 for 30 min,then treated for 1 h with a solution 2.5% glutaraldehyde in50 mM phosphate buffer, pH 5.8. Afterwards, cuvetteswere washed with fresh phosphate buffer and enzymeimmobilization was performed for 1 h with 1 mg/mL GOD(E.C. 1.1.3.4. from Aspergillus niger type VII-S) at pH 5.8.GOD-cuvettes were washed several times with 1 M NaCl,then with 2 mM glycine and finally with phosphate buffer.GOD-cuvettes were stored in potassium phosphate buf-fer, pH 7, at 5�C when not in use. Response to glucose inGOD-cuvettes was measured spectrophotometrically be-tween 400 and 800 nm, at room temperature and at pH5.8. The CE/biosensor was obtained by immobilizingGOD onto a PANI-modified capillary (Section 2.2) washedwith water, activated with 1 M NaOH (500 psi�s) andconditioned at pH 5.8. in phosphate buffer (500 psi�s).For GOD immobilization a 2.5% glutaraldehyde solutionat pH 5.8 was injected into the capillary portion betweenthe detector and the outlet end of the capillary and letreact with PANI for 1 h. Capillaries were washed with50 mM phosphate buffer, pH 5.8 (500 psi�s). A 1 mg/mLenzyme solution (254 U) in pH 5.8 buffer was injectedin the column and let react for 1 h. The CE/biosensorwas then washed with 1 M NaCl (500 psi�s), 2 mM gly-cine (500 psi�s) and 50 mM phosphate buffer (500 psi�s).The electroosmotic flow was inverted by treating thecapillary with 2 mM Q-Pip. Runs were carried out in aBio-Rad (Hercules, CA, USA) BioFocus 3000 instrument,injecting glucose concentration in the range of 1–40 mM

(10 psi�s). The running buffer was 50 mM phosphate,pH 5.8, voltage was 2 kV, and detection wavelengthwas 580 nm.

3 Results and discussion

3.1 Simultaneous determination of reducingagents by CE/sensor

In the perspective of developing a system for the analysisof samples containing different redox-active species, weextended the previous work on the CE/PANI sensor [8],exploring the possibility of simultaneous quantification ofreducing analytes. Experiments were performed to detecttwo reducing agents, AA and GSH. The electrophero-grams (Fig. 2) clearly show that as a reducing analytepasses through the PANI film it produces an irreversibledecrease of the optical absorbance of the polymer layer.AA concentration was fixed at 200 mg/L (1 mM), while theconcentration of GSH was progressively increased (from6 to 18 mM). The order of elution is GSH first, then AA. Thequantification of the reducing agents was made by meas-uring the magnitude of the decrease of optical absorb-ance (in mAU units). Such decrease was constant for AA,the concentration of which was kept constant, butincreased with the increase in GSH concentration,accounting for proportionality between the signal and theredox activity of the ‘sensed’ analyte.

Experiments for the simultaneous determination of the re-ducing agents L-AA and D-IAA were also performed. Asshown in Fig. 3, AA and IAA acid were separated on theCE/sensor. The CZE separation of the diastereoisomerswas achieved only in buffer borate at pH 9.0, under con-ditions where the borate esters formed by complexationof AA and IAA are different (the first is a 6-member ringcomplex while the second is a 5-member ring) [21]; theresolution for AA and IAA is close to Rs = 1 and the kineticsof response of the polyaniline polymer became crucial forthe discrimination of the two analytes.

3.2 Determination of oxidizing agents in theCE/PANI sensor

In order to explore the performance of the CE/PANI sen-sor in detection and quantification of oxidizing com-pounds, experiments were conducted with hydrogen per-oxide as model analyte. Different concentrations ofhydrogen peroxide were injected into the CE/PANI sensor(Fig. 4A). The sensor response to different concentrationof hydrogen peroxide was then plotted in the calibrationscurve (Fig. 4B). The range of response of the CE/PANIsensor seems limited to very high concentrations ofhydrogen peroxide, i.e., between 1 and 30% v/v.



Figure 2. Profile of three electropherograms obtainedby injecting (a) 18 mM GSH and 1 mM AA, (b) 12 mM GSHand 1 mM AA, and (c) 6 mM GSH and 1 mM AA. PANIreduction was revealed by a sharp and permanentdecrease in absorbance at 650 nm caused by the changein optical absorbance occurred at the polymer surface.

Figure 3. Typical determination of AA and IAA on the CE/PANI sensor. Conditions: 250 mM borate buffer, pH 9.0;applied voltage, 4 kV; detection wavelength, 650 nm.Samples were injected by pressure (2 psi�s). 2 mM D-IAA,2 mM L-AA.

Figure 4. CE/PANI sensor and oxidizing compounds.(A) Two concentrations of hydrogen peroxide (3% and24%, respectively). Conditions: 50 mM aspartic acidbuffer, pH 2.7; 20 kV; T = 25�C; � = 650 nm. (B): CE/sensorresponse to different concentrations of hydrogen perox-ide plotted as a calibration curve.



3.3 Redox kinetics on PANI

The results with reducing and oxidizing analytes showsthat the response of the CE/sensor to the oxidation andto the reduction is nonsymmetrical. The kinetics of oxida-tion and reduction was further studied on cuvettes modi-fied with PANI. Solutions with increasing concentration ofanalyte (L-AA, D-IAA, GSH, potassium dichromate, hydro-gen peroxide) were added to the modified cuvettes andthe PANI response was monitored spectrophotometri-cally over time (Fig. 5). The redox changes d(Abs650nm)/dtvaried between 0.62 mAU/s in the case of potassiumdichromate and 0.067 mAU/s for GSH. The changes inoptical absorbance were compared with the standardredox potentials of the analytes. Despite hydrogen perox-ide has a standard potential E� = �1.77 V and was used ata concentration of 1.89 M, the optical changes were lowand also the reaction kinetics was extremely slow,d(Abs650nm)/dt = 0.1 mAU/s. It confirmed the nonsignifi-cant sensitivity of the polymer to H2O2, while oxidationwith potassium dichromate (E� = �1.33 V) gave a clearresponse. The best sensitivity was obtained with AA(E�=�0.127 V).

3.4 CE/biosensor for the determinationof glucose

In the second part of our work, an enzyme was linked to thePANI-modified portion of the capillary, in order to create aCE/PANI biosensor. Model enzyme for testing the biosen-sor was the flavoprotein GOD [22], which has been widelyused in the field of biosensors [23]. GOD is a stable enzymeand has optimum of activity at pH 5.5. It is characterizedby high specificity for D-glucose while 2-deoxy-D-glucose,D-mannose, and D-galactose exhibit low activities as sub-strate [24]. The CE/PANI biosensor was built by treating thecapillary with the quaternary piperazine agent, whichcovalently modified the capillary inverting the electroos-motic flow [20]; then, the detector portion of the capillarywas modified with PANI (16:1 mol/mol aniline/ammoniumpersulfate) and GOD was covalently linked to PANI via theglutaraldehyde reaction. In parallel, control experimentswere performed modifying PANI cuvettes with GOD.

Cuvettes were an ideal system to test the sensitivity of theGOD-biosensor, without dealing with potential interfer-ences given by the application of voltage, as during elec-

Figure 5. Graphs of the changesin absorbance versus time forPANI-modified cuvettes. Ana-lytes, dissolved in 50 mM aspar-tic buffer, pH 2.7, were addedat the following concentra-tions: 13 �M L-AA, 3 mM GSH,10.4 �M potassium dichromate,and 3.6% (1.89 M) hydrogenperoxide. The initial reduction/oxidation rate was evaluated asV = d(Abs650nm)/dt.



trophoresis. Thus, the sensitivity range of the GOD-cuv-ettes was measured by adding increasing concentrationof glucose and monitoring the changes in optical absorb-ance, after ca. 2 min incubation. The operational range ofGOD-cuvettes was assessed between 0.2–36 mM glu-cose (Fig. 6). The GOD-cuvette biosensor showed a non-linear response, that might be described by a saturationequation, model of the Michaelis-Menten: y = �Absmax*X/(Kd�X), with �Abs max = �10.32 and Kd = 12.The value of Kd seems to be related to the KM of GOD.

Next, the sensitivity of the CE/PANI biosensor was tested.Increasing concentrations of glucose, ranging from 1 to50 mM, were injected in the capillary at pH 5.8. Glucosemigrates along the capillary transported by the inverseEOF and reached the detector in 10 min. Transit time forglucose varies as the numbers of runs performed on thecapillary increased (tR 1st run = 10 min, tR 6th run = 14 min).The phenomenon is due to the depletion of the Qpip layer,responsible for the inversion of the EOF in the capillary.Perfect reproducibility of transit time is possible by condi-tioning the capillary with Qpip prior to each run. To pre-vent possible damages to GOD and loss in the catalyticactivity, Qpip conditioning was limited to once per day.

Upon glucose addition a signal was recorded in the CE/PANI biosensor as a drop in the observed optical absorb-ance (see Fig. 7). A calibration curve was built for the CE/PANI biosensor, indicating an operational range within1–24 mM. Also the CE/PANI biosensor response might bedescribed by a saturation equation: y = �Abs max*X/(Kd�X), with �Abs max = �1.16 and Kd = 10, indicatingthat the changes in the PANI optical absorbance are cor-related with the enzymatic activity and the Kd calculatedis related to the KM of GOD. Moreover, results on the CE/PANI biosensor show that the GOD activity and the elec-tron transfer between GOD and PANI are not significantlyaffected by the electric field.

A comparison between the response of the GOD-cuvetteand the PANI/CE biosensor shows that the time of con-tact between the analyte and the enzyme is determinantfor the signal development: the GOD-cuvettes were incu-bated with glucose for ca. 2 min (100 s) before reading theabsorbance, while in the PANI/CE biosensor the glucoseresidence-time on the polyaniline-enzyme layer is esti-mated as ca. 10 s; the response of the GOD-cuvette is10 times higher than the response of the PANI/CE biosen-sor. The sensitivity range of the CE/PANI biosensor iscomparable with other GOD-biosensors [25–29] and fitswith the range of concentrations required for clinical anal-ysis of glucose (e.g., normal 4–6 mM, diabetes �8 mM), asshown in the curve of Fig. 8. Thus, the PANI/CE-biosensorcould be adapted for clinical analysis.

Figure 6. Range of sensitivity of the PANI/GOD-modifiedcuvettes. Response to glucose concentrations (1–36 mM)were monitored at 580 nm, in potassium phosphatebuffer, pH 5.8.

Figure 7. Typical electropherograms of glucose samplesobtained with the CE/PANI biosensor. Conditions: 50 mM

phosphate buffer, pH 5.8; applied voltage, 2 kV; T = 25C�,detection, � = 580 nm. Samples were injected by pressure(10 psi�s). (A) 4 mM glucose; (B) 16 mM glucose.



Figure 8. Calibration of the CE/PANI biosensor. The dropin absorbance given by the passage of each glucosesample was measured as �abs at 580 nm and plotted asfunction of the glucose concentration. The sensitivityrange of the CE/PANI biosensor was 1–24 mM glucose.

The regeneration of the initial oxidation state of the poly-mer is achieved by treating the PANI polymer with an oxi-dant, ammonium persulfate, as demonstrated previously[7, 8, 30]. The storage stability of the CE/PANI biosensorwas investigated by checking the optical response to8 mM glucose over different days. When not in use, theCE/PANI biosensor was stored in 50 mM potassium phos-phate buffer, pH 7, at 5�C or at 20�C. Fig. 9 shows the var-iation in the sensor response with time for the CE/PANIbiosensor. The response was well stable for more than30 days from the day of fabrication. Results indicate thatafter 80 days the CE/PANI biosensor retains still the40%of its original sensitivity, accounting for a fairly good

Figure 9. Stability of the CE/PANI biosensor, measuredas sensor response. The first day response was consid-ered as 100%. Further measurements were referred asresidual sensor response. The CE/PANI biosensor wastested both for storage at 20�C and at 5�C.

robustness of the CE/PANI biosensor. The CE/PANI bio-sensor stored at 20�C does loose significantly responseability after the first 10 days.

4 Concluding remarks

The introduction of CE methodologies into sensing pro-vides a unique and powerful element of selectivity forremote analyses and should offer unparalleled versatilityand reusability. In the present work, the simple exploita-tion of the optical properties of a redox sensing polymer(PANI), layered onto the detector window of a CE capillarycolumn, allowed to fabricate a sensor detector for thequantification of redox analytes, instead of a UV-detector.The modification of the capillary column is easy to per-form, relatively inexpensive, regenerable, and resusablefor several days.

The present work extends the previous study on the con-struction of a CE/PANI sensor demonstrating the simulta-neous analysis of different redox active species. The CE/PANI sensor resolved mixtures of reducing agents, e.g.,AA and GSH, as well as compounds with very close simi-larity like the diastereoisomers of AA IAA. The responsekinetics of the polymer to reduction and oxidation is cru-cial for the detection and should match the high resolutionability of CE. Studies should be addressed to the ameli-oration of the kinetics of PANI response.

A further progress was achieved in building a CE/PANIbiosensor, immobilizing the enzyme GOD, chosen as amodel biocatalyst, onto the PANI layer. Coupling an en-zyme broadens the applicability of the CE/PANI sensor,as the biological element determines ultimately the selec-tivity of the sensor, allowing a highly specific detectionand quantification of components which does not haveredox activity. The characteristics of the CE/PANI biosen-sor were evaluated. Results indicate a fairly high stabilityof the biocomponent in the CE/biosensor, a sensitivityrange comparable with other GOD biosensors describedand an operational range compatible with measurementsof physiological glucose. The response kinetics in the CE/biosensor might be described by a Michaelis-Mentenequation, thus seems directly correlated to the enzymeactivity, accounting for direct electron transfer and mini-mal interferences of the electric field applied in CE for theseparation. The entity of the response depends on thecontact-time between enzyme and substrate.

The model of PANI/CE-biosensor here considered is afirst example of PANI-CE optical biosensor. It shows inter-esting performance, possibility to be tuned towards thedesired selectivity by coupling the most appropriateenzymes and it offers analysis based on the bidimension-



ality of the separation/sensing, which could be highlybeneficial for the discrimination of interferents, inhibitorymolecules, and substrate-related compounds.

A.B. is grateful to COFIN 2001 (prot. 2001033797_003)and to Agenzia2000 CNR for supporting the present re-search.

Received May 19, 2003

5 References

[1] Haber, C., Jones, W. R., Soglia, J., Surve, M. A., McGlynn,M., Caplan, A., Reineck, J. R., Krstanovic, C., J Capill. Elec-trophor. 1996, 3, 1–11.

[2] Stokes, D. L., Sepaniak, M. J., Vo-Dinh, T., Biomed Chroma-togr. 1997, 11, 187–192.

[3] Whelan, R. J., Zare, R. N., Anal Chem. 2003, 75, 1542–1547.[4] El Rassi, Z. (Ed.), Electrophoresis 2002, 22–23, 3789–4052.[5] Dickens, J. E., Sepaniak, M. J., J. Environ. Monit. 2000, 2,

11–16.[6] MacDiarmid, A. G., Epstein, E. P., Faraday Discuss. 1989, 88,

317.[7] Kang, E. T., Neoh, K. G., Tan, K. L., Progr. Polym. Sci. 1998,

23, 277–324.[8] Castelletti, L., Piletsky, S. A., Turner, A. P. F., Righetti, P. G.,

Bossi, A., Electrophoresis 2002, 23, 209–214.[9] Foyer, C., in: Alscher, R. G., Hess, J. L. (Eds.), Antioxidants in

Higher Plants, CRC Press, Boca Raton, FL 1993, pp. 31–58.[10] Kohen, R., Biomed. Pharmacother. 1999, 53, 181–192.[11] Beckman, K. B., Ames, B. N., Physiol. Rev. 1998, 78, 547–

581.[12] Heber, U., Miyake, C., Mano, J., Ohno, C., Asada, K., Plant

Cell. Physiol. 1996, 37, 1066–1072.

[13] Rice-Evans, C. A., Burdon, R. H. (Eds.), Free Radical Dam-age and its Control, Elsevier, Amsterdam 1994.

[14] Laschi, S., Mascini, M., Turner, A. P. F. (Eds.), Biosensors,Kirk-Othmer Encyclopedia of Chemical Technology (on-lineedition), Wiley, New York 2002.

[15] Bossi, A., Piletsky, S. A., Righetti, P. G., Turner, A. P. F., J.Chromatogr. A 2000, 892, 143–153.

[16] Parente, A. H., Marques, E. T. Jr, Azevedo, W. M., Diniz, F. B.,Melo, E. H., Lima Filho, J. L., Appl. Biochem. Biotechnol.1992, 37, 267–273.

[17] de Melo, J. V., Bello, M. E., de Azeveado, W. M., de Souza,J. M., Diniz, F. B., Electrochim. Acta 1999, 44, 2405–2412.

[18] Gerard, M., Chaubey, A., Malhotra, B. D., Biosens. Bioelec-tron. 2002, 17, 345–359.

[19] Lu, S. Y., Li, C. F., J. Electroanal. Chem. 1994, 364, 31–36.

[20] Sebastiano, R., Gelfi, C., Righetti, P. G., Citterio, A., J. Chro-matogr. A 2000, 894, 53–61.

[21] Obi, N., Katayama, M., Sano, J., Kojima, Y., Shigemitsu, Y.,Takada, K., New J. Chem. 1998, 22, 933–934.

[22] Bentley, R., in: Colowick, S., Kaplan, N. (Eds.), Methods inEnzymology, Vol. IX, Academic Press, N.Y. 1996, p. 86.

[23] Clarke, L. C., Lyons, C., Ann. N. Y. Acad. Sci. 1962, 102, 29–45.

[24] O’Malley, J., Weaver, J., Biochemistry 1972, 11, 3527.

[25] Zhou, X., Arnold, M. A., Anal. Chim. Acta 1995, 304, 147–156.

[26] Ohashi, E., Karube, I., J. Biotechnol. 1995, 40, 13–19.

[27] Sotiropoulou, S., Chaniotakis, N. A., Anal. Bioanal. Chem.2003, 375, 103–105.

[28] Hu, S., Luo, J., Cui, D., Anal. Sci. 1999, 15, 585–588.

[29] Brahim, S., Narinesingh, D., Guiseppi-Elie, A., Biosens.Bioelectron. 2002, 17, 53–59.

[30] Bossi, A., Piletsky, S. A., Piletska, E. V., Righetti, P. G.,Turner, A. P. F., Anal. Chem. 2000, 72, 4296–4300.


Documents

Towards the development of an integrated capillary electrophoresis optical biosensor