sdarticle[2]-

7/27/2019 sdarticle[2]-

1/6

Available online at www.sciencedirect.com

Talanta 75 (2008) 301306

Differential amperometric determination of hydrogen peroxide inhoneys using flow-injection analysis with enzymatic reactor

Romulo Augusto de Abreu Franchini, Maria Auxiliadora Costa Matos,Rosana Colombara, Renato Camargo Matos

NUPIS (N ucleo de Pesquisa em Instrumentacao e Separacoes Analticas), Departamento de Qumica, Instituto d e Ci encias Exatas,

Universidade Federal de Juiz de Fora, 36036-330 Juiz de Fora, MG, Brazil

Received 30 August 2007; received in revised form 30 October 2007; accepted 7 November 2007

Available online 17 November 2007

Abstract

Hydrogen peroxide (H2O2) present in honey was rapidly determined by the differential amperometric method in association with flow-injection

analysis (FIA) and a tubular reactor containing immobilized enzymes. A gold electrode modified by electrochemical deposition of platinum was

employed as working electrode. Hydrogen peroxide was quantified in 14 samples of Brazilian commercial honeys using amperometric differential

measurements at +0.60 V vs. Ag/AgCl(sat). For the enzymatic consumption of H2O2, a tubular reactor containing immobilized peroxidase was con-

structed using an immobilization of enzymeson Amberlite IRA-743 resin. The linear dynamic range in H2O2 extends from1 to 100106 molL1,

atpH 7.0.At flow rateof 2.0 mLmin1 and injecting 150L sample volumes, the sampling frequency of the 90 determinations per hour is afforded.

This method is based on three steps involving the flow-injection of: (1) the sample spiked with a standard solution, (2) the pure sample and (3) the

enzymatically treated sample with peroxidase immobilized. The reproducibility of the current peaks for hydrogen peroxide in 105 molL1 range

concentration showed a relative standard deviation (R.S.D.) better than 1%. The detection limit of this method is 2.9 107 molL1. The honey

samples analyses were compared with the parallel spectrophotometric determination, and showed an excellent correlation between the methods.

2007 Elsevier B.V. All rights reserved.

Keywords: Amperometry; Honey; Hydrogen peroxide; Peroxidase immobilized

1. Introduction

Honey contains a complex matrix of components, which

presents a considerable analytical challenge. It is a liquid (or

semiliquid) product made up of about 80% solids. It is produced

by bees from thenectar of plants,as well as from honeydew. Bees

andplants arethe primary sources of components such as: carbo-

hydrates (fructose, glucose, maltose and sucrose with traces of

many other sugars depending on the floral origin), water, traces

of organic acids, enzymes, aminoacids, pigment, and other com-

ponents like pollenand waxwhich arise duringhoney maturation

[1]. The chemical analysis of honey has three main purposes:

(1) to determine the geographical and botanical origin, (2) veri-

fication of adulteration and (3) identification of pharmacological

active compounds. The first and second points assist with cer-

Corresponding author. Fax: +55 32 3229 3314.

E-mail address: [email protected] (R.C. Matos).

tification of quality of the product which is commonly used as

a food product; and the third area allows the examination of

content for the use of honey in medicinal purposes.

Hydrogen peroxide is a product of many biological reactions

catalyzed by several oxidase enzymes. All honeys contain per-

oxide, which imbues them with antibacterial properties. It has

been shown that the antibacterial activity of honey occurs due

to hydrogen peroxide generation [25]. Therefore, the determi-

nation of hydrogen peroxide is important in the characterization

and selection of honey samples for its use as an antimicrobial

agent. Hydrogen peroxide is generated by the enzyme glucose

oxidase when honey is diluted and maximum levels of hydro-

gen peroxide encountered in the diluted honeys are in the range

of 12mmol L1 [6]. Dilution is needed to decrease the acid-

ity of the medium and for adjusting the pH for proper action

of glucose oxidase. Weston [5] stated that the level of hydro-

gen peroxide in honey is essentially determined by the amount

of catalase, which originates from flower pollen, and glucose

oxidase, which originates from the hypopharyngeal glands of

0039-9140/$ see front matter 2007 Elsevier B.V. All rights reserved.

doi:10.1016/j.talanta.2007.11.011
mailto:[email protected]://localhost/var/www/apps/conversion/tmp/scratch_1/dx.doi.org/10.1016/j.talanta.2007.11.011http://localhost/var/www/apps/conversion/tmp/scratch_1/dx.doi.org/10.1016/j.talanta.2007.11.011mailto:[email protected]

7/27/2019 sdarticle[2]-

2/6

302 R.A.d.A. Franchini et al. / Talanta 75 (2008) 301306

bees. Manzoori et al. [7] have proposed the spectrofluorometric

determination of hydrogen peroxide in several honey samples

using crude extract of kohlrabi (Brassica oleracea gongylodes),

which is a rich source of peroxidase. Franchini et al. [8] have

described a versatile method for spectrophotometric determina-

tion of micromolar hydrogen peroxide in commercial Brazilian

honey samples using a peroxidase immobilized on resin and the

determination of trace metals using capillary zone electrophore-

sis without any treatment of honey samples.

Electrochemical determinations of hydrogen peroxide are

generally performed by oxidation on a platinum electrode [9].

Depending upon the pH of the solution, a very high positive

potential must be applied for the oxidation of hydrogen perox-

ide. The typical applied potentials are in the range of +0.7 to

+0.9 V vs. SCE [10]. As a result, many substances can interfere

with the measurements. The use of biosensors with immobilized

enzymes such as peroxidase and catalase has been extensively

investigated for hydrogen peroxide analysis, based on spec-

trophotometry [8,11], fluorometry [7,12], chemiluminescence

[13,14] and electrochemical [15] techniques.Strategies have been investigated to adapt quantification

methods to the range of sample concentrations with low cost. In

the analytical methods using enzymes, the reduction in cost of

the determination is generally associated with reduced enzyme

consumption. Recently, various ion-exchange resinshave gained

considerable attention not only for separation purposes but also

as carriers of catalytic active substances. Considerable thought

has been paid to their application for immobilization of enzymes

[11,14,1619]. The resins should meet several requirements.

Their porous structure must be strong enough to withstand the

enhanced pressure usually applied in forced flow bioreactors.

Furthermore, the membrane material must be chemically andphysically resistant. These requirements can be met by various

aromatic and aliphatic polyamides. Therefore, resin prepared

from these polymers is a suitable substrate for the immobiliza-

tion of enzymes [20]. The covalent binding of the enzyme to

the polymer matrix is one of the most prospective methods for

immobilization.

In the present work, we describe a versatile method for dif-

ferential amperometric determination of hydrogen peroxide in

honey, using a gold microelectrode modified by electrodepo-

sition of platinum, combined with an on-line tubular reactor

containing peroxidase immobilized on resin (Amberlite IRA-

743) without any treatment of samples. The concentration of

the hydrogen peroxide in each sample was calculated based onthe difference between the current measurement before and after

the enzymatic treatment.

2. Experimental

2.1. Enzymes immobilization

The procedure adopted to immobilize the peroxidase enzyme

was quick and very simple [15]. Amberlite IRA-173 resin was

selected as support, because it has active amine groups in its

chemical structure. The enzyme immobilization process begins

with the addition of 100L of glutaraldehyde 0.1% to 250 mg

of resin, and this mixture was stirred for 5 min. Subsequently,

200 units of enzymes were introduced into the mixture and

stirred for an additional 10 min. In the next step, the resin was

transferred to a length of tygon tubing (2.5 mm of i.d. and 25 mm

long) with one of its extremities closed with a thin layer of glass

wool to assemble the reactor. At this point, the other extrem-

ity of the tubing was then closed with glass wool. To adapt the

enzymatic reactor to a FIA (flow-injection analysis) system, the

tubing (0.8 mm of i.d) was attached at in each of its extremities

with the aid of a small piece of silicone tubing (1.3 mm i.d. and

5 mm long). Finally, the reactor was washed with 10 mmol L1

phosphate buffer solution (pH 7.0) to remove the excess of per-

oxidase.

2.2. Reagents and chemicals

All solutions used were of analytical grade. Hydrogen

peroxide, mono- and di-hydrogen phosphates were obtained

from Merck (Darmstadt, Germany). Solutions were prepared

by dissolving the solids in distilled water that was alsotreated with a nanopure system. Commercial peroxidase (EC

1.11.1.7115 U mg1) was obtained from Sigma (St. Louis,

MO, USA). The Amberlite IRA-743 ion-exchange resin and

glutaraldehyde were obtained from Aldrich (Milwaukee, WI,

USA). Diluted solutions of hydrogen peroxide were prepared

daily using deionized water.

2.3. Sample collection

This work was carried out on 14 samples in Brazil. The

samples were stored in the dark at room temperature prior to

analysis. For determination of hydrogen peroxide, 1 g of honeywas dissolved in 10 mL of purified water and injected in the

flow-injection system. Each sample was injected in triplicate.

2.4. Electrodes and instrumentation

The electrochemical cell comprised a platinum-modified

gold electrode (3.0 mm diameter). Modification was done by

electrochemical deposition of Pt (K2PtCl6 2103 molL1,

pH 4.8, at 1.00 V for 15 min). Microscopic observation of

the electrodes after electrodeposition showed uniform platinum

deposit, with a very rough surface. Electrodes so modified were

stable for at least 1 week under intense use. The reference elec-

trode was a miniaturized Ag/AgCl(sat) electrode constructed inour laboratory [21] and a stainless steel tube (1.2 mm i.d.) was

used as auxiliary electrode.

In this work, a double channel flow system was employed.

The solutions were propelled by pressurization, utilizing an

aquarium air pump to avoid the undesirable pulsation observed

when peristaltic pumps are employed [22]. Control of the flow

rate was done by adaptation of the aquarium valve outlet with

a pinched tygon tube inserted in the line. Teflon tubing of

0.5 mm i.d. was used throughout the flow system. The flow

system used during the development of this work consisted of

two lines, in first the sample was added in the detection sys-

tem, in the second the sample was inserted in the line that

7/27/2019 sdarticle[2]-

3/6

R.A.d.A. Franchini et al. / Talanta 75 (2008) 3 01306 303

contain the enzymatic reactor before of the detection system.

A potentiostat (-AUTOLAB) operating in the amperometric

mode was employed for FIA measurement. Temperature con-

trol was achieved using a THERMOMIX 18 BU B thermostatic

bath BraunBiotech. International. The systemcontains an aquar-

ium air pump, a pinch valve, sampling loop, a tubular reactor

( = 0.25 cm and 2.5 cm of length) with peroxidase chemically

immobilized in Amberlite IRA-743 resin, an electrochemical

cell and the potentiostat.

2.5. Procedure

For amperometric detection of hydrogen peroxide, +0.60 V

was found the most favorable potential to be applied to the gold

electrodemodified with platinum.The differential determination

of this analyte requires at last three measurements, one involv-

ing the sample containing a standard addition in the channel

without the reactor, a second containing just the sample in the

channel without the reactor, and the third measurement involves

a sample passes through the enzyme reactor. In the first case asignal, corresponding to hydrogen peroxide standard, hydrogen

peroxide of the sample and plus the interfering components is

registered. In the second case, the signal corresponds to the sam-

ple without hydrogen peroxide standard. In the third case, the

signal corresponds to the sample without H2O2 (i.e., only to the

interfering species). The calculated difference is compared with

a H2O2 standard.

3. Results and discussion

Preliminary tests employing platinum-modified electrodes

showed a very interesting behavior in the presence of hydrogenperoxide. The current enhancement was remarkable and in addi-

tion a decrease in the oxidation potential of hydrogen peroxide

occurs when the electrodes are modified. Part of the increase in

current can probably be attributed to the increase in the effective

area of the electrodes. Observations with a microscope showed

the formation of a very porous surface after platinum deposi-

tion.

3.1. Immobilized peroxidase and optimization of the flow

system

To examine the efficiency of the tubular reactor contain-

ing immobilized peroxidase in a resin, experiments involvingconsecutive injections of hydrogen peroxide solution were

performed. Responses of a gold electrode modified by elec-

trodeposition of platinum for injections of 150 L of hydrogen

peroxide 1 105 molL1 for a channel without enzyme and

with immobilized peroxidase were obtained. For a channel with-

out enzyme a current of 0.17A was measured, while for the

reactors with immobilized peroxidase currents of 0A was

found. The reactor with immobilized peroxidase was effective,

once it was able to eliminate completely the H2O2, a fundamen-

tal condition for applications in differential measurements.

The influence of parameters such as flow rate and sam-

ple volume was studied. Fig. 1A shows the amperometric

Fig. 1. Repetitive injections of hydrogen peroxide 1105 molL1 to find the

most suitable working conditions. (A) Flow rate from 0.5 to 5.0 mL min1 and

(B) samples volume injected from 50 to 250 L. Measurements made with a

gold electrode modified by electrodeposition of platinum. Applied potential,

+600 mV vs. Ag/AgCl(sat).

responses of a gold electrode modified with platinum for injec-

tions of 150L of hydrogen peroxide 1 105 molL1, as

a function of the flow rate, varied from 0.5 to 5.0 mL min1

with and without reactor. The signal remains virtually con-

stant when the flow rate is changed from 1.5 to 3.0 mL min1.For high flow rates, the peroxidase immobilized in the tubular

reactor was unable to eliminate completely the hydrogen per-

oxide. The elimination reaction rate has to decrease when theflow rate decreases. A flow rate of 2.0 mL min1 was chosen

as the most favorable, since it combines good reproducibil-

ity, high throughput (90 samples h1), lower consumption of

carrier solution and complete elimination of the hydrogen

peroxide.

Fig. 1B shows the influence of the sample volume on the

analytical signal which was also evaluated. Loops with internal

volumes varying from 50 to 250 L were tested. When the vol-

ume of the sample is increased, the amperometric signal grew

but the analysis time also increased, once the cell wash-out pro-

cess also requires a longer time. The volume of 150 L was

selected as the working volume in the following experiments.

7/27/2019 sdarticle[2]-

4/6


Fig. 2. FIA-amperometric measurements involving injections of 150L solu-

tion containing 1 10106 molL1 of hydrogen peroxide. The inset shows

the respective calibration plot. Conditions: sample volume, 150 L; flow rate,

2.0mLmin1; applied potential, +600 mV vs. Ag/AgCl(sat).

For all the volumes studied the peroxidase immobilized in the

tubular reactor completely eliminated the hydrogen peroxide in

the samples when used flow rate of 2.0 mL min

1

.An important characteristic observed for the immobilized

enzyme was a storage stability of at last 2 week under intenseuse

with hydrogen peroxide standard. After this period, a decrease

on the order of 3045% of the enzyme activity was observed.

When applied in the determination of hydrogen peroxide in

honey, the enzymatic reactor showed a loss in the enzyme activ-

ity after 50 injections, requiring construction of a new reactor.

When not in use, the reactors were stored in a freezer at 20 C

[8].

3.2. Calibration plot

Fig. 2 shows the amperometric response of the modi-

fied gold electrode for successive injections of 150L of

hydrogen peroxide from (a) 1molL1 to (e) 10molL1.

The proportionality between the amperometric current

and the hydrogen peroxide concentrations was con-

firmed from the calibration plot shown in the inset (i

(A)=3.77 103 + 1.67 102[H2O2](molL1); correla-

tion coefficient, 0.999). Notice the very favorablesignal-to-noise

ratio, demonstrated by the very stable base line obtained for

these low micromolar concentrations. The detection limit

for the conditions adopted in present study was found as

2.9 107 molL1 (3 times the standard deviation of the

blank) [23].

3.3. Determination of hydrogen peroxide in honey by

flow-injection analysis

The samples to be analyzed were mixed on-line with buffer

solution, used as the carrier solution. Fig. 3(AC) shows three

sequences of hydrogen peroxide analysis for three samples.

Each group of three peaks corresponds to 150L injec-

tions of the honey sample containing (I) a standard addition

(2 106 molL1 of the standard hydrogen peroxide), (II) the

sample without treatment, and (III) the sample after enzymatic

treatment with peroxidase immobilized. The standard addition

Fig. 3. Differential FIA-amperometric measurements of hydrogen peroxide in

three different samples of honey: (A) sample 3; (B) sample 4; and (C) sample 7.

The triplicate injection correspond to: (I) 150 L of sample +2106 molL1

of the standard hydrogen peroxide solution, (II) 150 L of the sample and (III)

150L of the sample after enzymatic treatment with peroxidase immobilized.

Other conditions as in Fig. 2.

7/27/2019 sdarticle[2]-

5/6

R.A.d.A. Franchini et al. / Talanta 75 (2008) 3 01306 305

Table 1

Results obtained in analysis of hydrogen peroxide (mg kg1) in honey samples

Number of sample Geographical origin H2O2 (mgkg1) (amperometry) H2O2 (mgkg

1) (spectrophotometry)

1 Vicosa 150 3 155 2

2 Teresopolis 139 3 135 4



5 Teresopolis 30 1 28 16 Teresopolis 8 0 9 0


8 Esprito Santo 157 3 158 5

9 Belo Horizonte 140 2 139 3

10 Tabuleiro 148 4 142 4

11 Coronel Pacheco 47 1 62 3

12 Coronel Pacheco 80 2 91 3

13 Volta Redonda 155 4 147 6

14 Juiz de Fora 37 0 47 1

Mean 103 109

Vmin 8 9

Vmax 205 214

method was carried out with the peaks (I) and (II) after subtract-

ing of the height of the peak (III).

Table 1 and Fig. 4 compare the results of the anal-

yses performed by amperometry developed in this work

and using the spectrophotometric detection [8] for 14 dif-

ferent samples (in triplicates). Comparing the amperometry

with gold/platinum electrode and spectrophotometry gives a

slope and intercept very close to unity and zero, respec-

tively. The confidence interval for the slope and intercept

are (0.95 0.03) and (8.14 4.04) mg kg1, respectively, for

a 95% confidence level. Taking into account these results,

no significant differences between the three methods wereobserved, which strongly indicates the absence of systematic

errors.

Fig. 4. Comparisonof the results obtained by differential amperometricanalysis

for 10 different samples of rainwater using a gold microelectrode modified by

the deposition of platinum and (A) differential amperometric analysis using a

mercury microelectrode and (B) spectrophotometric methods for the analysis of

hydrogen peroxide.

4. Conclusions

This work demonstrated the potentiality of the amperometric

method using gold electrodes modified with platinum coupled

with flow-injection analysis techniques, for the detection of

hydrogen peroxide in honey using peroxidase immobilized in a

tubular reactor. The very high sensitivity provided by amperom-

etry, combined with the low volume of the flow cell, allows us to

work with small sample volumes and at low concentrations. The

association of amperometric detection with flow-injection anal-

ysis and the possibility of avoiding cumbersome processes such

as separation, extraction and filtration substantially increase thespeed of analysis. These advantages offer a very favorable way

for the rapid analysis of hydrogen peroxide in honey samples

(throughput of 90-samples h1).

Acknowledgements

Authors would like to thank FAPEMIG (Fundacao de

Amparo a Pesquisa do Estado de Minas Gerais), CNPq

(Conselho Nacional de Desenvolvimento Cientfico e Tec-

nologico) and PROPESQ/UFJF (Pro-Reitoria de Pesquisa e

Pos-Graduacao da Universidade Federal de Juiz de Fora) for

financial support and grants.

References

[1] R.F. Torres, J.L.P. Bernal, M.A.B. Lopez, M.C. Mochon, J.C.J. Sanchez,

A.G. Perez, Talanta 65 (2005) 686.

[2] K.L. Allen, P.C. Molan, G.M. Reid, J. Pharm. Pharmacol. 43 (1991) 817.

[3] P.C. Molan, K.M. Russell, J. Apic. Res. 27 (1988) 62.

[4] M.J. Snow, M. Manley-Harris, Food Chem. 84 (2004) 145.

[5] R.J. Weston, Food Chem. 71 (2000) 235.

[6] L.M. Bang, C. Buntting, P. Molan, J. Altern. Complement. Med. 9 (2)

(2003) 267.

[7] J.L. Manzoori, M. Amjadi, M. Orooji, Anal. Sci. 22 (2006) 1201.

[8] R.A.A. Franchini, C.F. de Souza, R. Colombara, M.A.C. Matos, R.C.

Matos, J. Agric. Food Chem. 55 (17) (2007) 6885.

7/27/2019 sdarticle[2]-

6/6


[9] J.E. Harrar, Anal. Chem. 35 (1963) 893.

[10] G.G. Guibault, G.L. Lubrano, Anal. Chim. Acta 64 (1973) 439.

[11] R.C. Matos, E.O. Coelho, C.F. de Souza, F.A. Guedes, M.A.C. Matos,

Talanta 69 (2006) 1208.

[12] F. Schubert, F. Wang, H. Rinnerberg, Mikrochim. Acta 121 (1995) 237.

[13] R.W. Marshall, T.D. Gibson, Anal. Chim. Acta 266 (1992) 309.

[14] U. Spohn, F. Preuschoff, G. Blankenstein, D. Janasek, M.R. Kula, A.

Hacker, Anal. Chim. Acta 303 (1995) 109.

[15] R.C. Matos, J.J. Pedrotti, L. Angnes, Anal. Chim. Acta 441 (2001) 73.[16] A.C.A. Oliveira, V.C. Assis, M.A.C. Matos, R.C. Matos, Anal. Chim. Acta

535 (2005) 213.

[17] R.W. Marshall, T.D. Gibson, Anal. Chim. Acta 226 (1992) 309.

[18] P.D. Wentzell, S.J. Vanslyke, K.P. Bateman, Anal. Chim. Acta 246 (1991)

43.

[19] H. Hwang, P.K. Dasgupta, Mikrochim. Acta 3 (1985) 77.

[20] B.A. Watkins, D.D. Parrish, M. Trainer, R.B. Norton, J.E. Yee, J. Geophys.

Res. 100 (1995) 22831.

[21] J.J. Pedrotti, L. Angnes, I.G.R. Gutz, Electroanalysis 8 (1996) 673.

[22] R.C. Matos, I.G.R. Gutz, L. Angnes, R.S. Fontenele, J.J. Pedrotti, Quim.

Nova 24 (2001) 795.[23] J.C. Miller, J.N. Miller, Statistics for Analytical Chemistry, Harwood,

Chichester, 1992.

sdarticle[2]-

Documents