A

cbSTwod©

K

1

ianlstgmg[wei

0d

Talanta 71 (2007) 1842–1848

Voltammetric determination of glucose based on reduction ofcopper(II)–glucose complex at lanthanum hydroxide

nanowire modified carbon paste electrodes

Li Liu a, Jun-feng Song a,∗, Peng-fei Yu b, Bin Cui b

a Institute of Analytical Science, Northwest University, Xi’an 710069, Chinab Department of Chemistry, Northwest University, Xi’an 710069, China

Received 13 June 2006; received in revised form 2 August 2006; accepted 5 August 2006Available online 28 November 2006

bstract

A novel voltammetric method for the determination of �-d-glucose (GO) is proposed based on the reduction of Cu(II) ion in Cu(II)(NH3)42+–GO

omplex at lanthanum(III) hydroxide nanowires (LNWs) modified carbon paste electrode (LNWs/CPE). In 0.1 mol L−1 NH3·H2O–NH4Cl (pH 9.8)uffer containing 5.0 × 10−5 mol L−1 Cu(II) ion, the sensitive reduction peak of Cu(II)(NH3)4

2+–GO complex was observed at −0.17 V (versus,CE), which was mainly ascribed to both the increase of efficient electrode surface and the selective coordination of La(III) in LNW to GO.he increment of peak current obtained by deducting the reduction peak current of the Cu(II) ion from that of the Cu(II)(NH3)4

2+–GO complex

as rectilinear with GO concentration in the range of 8.0 × 10−7 to 2.0 × 10−5 mol L−1, with a detection limit of 3.5 × 10−7 mol L−1. A 500-foldf sucrose and amylam, 100-fold of ascorbic acid, 120-fold of uric acid as well as gluconic acid did not interfere with 1.0 × 10−5 mol L−1 GOetermination.

2006 Elsevier B.V. All rights reserved.

Gluc

mtBibsmstc

pa

eywords: Glucose; Lanthanum hydroxide nanowires; Carbon paste electrode;

. Introduction

Quantitative determination of �-d-glucose (GO) is of greatmportance in clinical, biological and pharmacy samples as wells food processing and fermentation. The electrochemical tech-ique for GO detection was introduced to achieve a low detectionimit and low-cost equipment. Various electrochemical detectiontrategies have been designed, which continue to use essen-ially two ways. One way is based on the enzymatic reactionenerally using glucose oxidase (GOD). Recently, metal andetal oxide [1,2], metal hexacyanoferrates [3–5], methyl violo-

en [6], metal–organic complexes [7–12], ferrocene derivatives13–17], nanoparticles [18–23] and others are co-immobilized

ith GOD on electrode surface. These substances were used

ither as the catalyst to reduce the operation potential for mon-toring the product of the enzymatic reaction, as the redox

∗ Corresponding author. Tel.: +86 29 8830 3448; fax: +86 29 8830 3448.E-mail address: songjunf@nwu.edu.cn (J.-f. Song).

tnrpw

e

039-9140/$ – see front matter © 2006 Elsevier B.V. All rights reserved.oi:10.1016/j.talanta.2006.08.042

ose–copper(II) complex

ediators to accelerate the enzymatic reaction or as electronransfer mediators to realize the direct electron transfer of GOD.ut it still suffers from the problem of stability and biocompat-

bility due to the intrinsic nature of enzyme. The other way isased on the direct electro-oxidation of GO on different sub-trates such as metal [24–26], various alloys electrodes [27–31],etal–complex modified electrodes [32,33] and others [34] in

trongly alkaline media. However, the practical application ofhese methods based on the direct electro-oxidation of GO isonfined by low sensitivity and poor selectivity.

It has been proved that nanomaterials with many attractiveroperties such as carbon nanotubes have brought a series ofdvantages in developing sensing systems for biomolecule. Upo now, more attention has been paid to the functionalization ofanomaterials for further improving the function of nanomate-ials. Prospectively, when nanomaterials with some additional

roperties were used as electrode materials, these propertiesould endow electrochemical methods with novel character.Lanthanum(III) hydroxide nanowires (LNWs) as new rare

arth nanaomaterial have some unique properties such as

ta 71 (2007) 1842–1848 1843

oiGmdohcicmtbhCra

2

2

taUuaear

2

i1ibwNiwmmot1

2

1miEi

waCdm0wsbssm

2

itTrtpw

bitrm

L. Liu et al. / Talan

ptical, catalytic and magnetic properties [35,36]. In addition,t is known that La(III) and Cu(II) ions can coordinate withO through deprotonated and free hydroxyl groups in alkalineedia, respectively, and the coordination sites in GO are

ifferent, for La(III) ion the hydroxyl group at C-2 and C-3r C-3 and C-4 [37,38], for Cu(II) ion most probably theydroxyl group at C-1 and oxygen atom at C-5 [39–41]. Byonsidering these, in this work, LNWs were immobilizedn carbon paste electrode (CPE) to prepare LNWs modifiedarbon paste electrode (LNWs/CPE), Cu(II) ion in ammoniumedium was used to transfer electroinactive GO into elec-

roactive Cu(II)(NH3)42+–GO complex and the voltammetric

ehavior of the Cu(II)(NH3)42+–GO complex at LNWs/CPE

as been investigated. The reduction peak current of theu(II)(NH3)4

2+–GO complex was greatly enhanced, and theelated mechanism was discussed. Based on this, a sensitivend selective method for GO determination was proposed.

. Experimental

.1. Apparatus

A JP-303 polarographic analyzer (Chendu Instrument Fac-ory, Chendu, China) is used for single sweep voltammetry and

CHI 660 electrochemical workstation (CH Instrument Inc.,SA) is used for cyclic voltammetry. A three-electrode config-ration includes a homemade LNWs/CPE working electrode,platinum wire counter electrode and a saturated calomel ref-

rence electrode (SCE). All the potentials quoted in this workre referred to the SCE. All the measurements are carried out atoom temperature.

.2. Materials

GO was of analytical reagent grade (purity > 99%, Yix-ng Chemical Reagent Factory No. 3, Yixing, China). A.0 × 10−2 mol L−1 GO stock standard solution was preparedn aqueous solution. Working standard solutions were preparedy appropriate dilution of the stock standard solution withater. A 1.0 × 10−2 mol L−1 Cu(II) solution and 1.0 mol L−1

H3·H2O–NH4Cl buffer (pH 9.8) were prepared. All chem-cals were of analytical reagent grade. Twice distilled wateras used throughout. LNWs were prepared by hydro-thermalethod [35,36], and were directly used without any further treat-ent. The transmission electron microscopy (TEM) micrograph

f LNWs is showed in Fig. 1. As can be seen, that any impuri-ies were hardly found, the LNWs had an average diameter of5–20 nm.

.3. Electrode preparations

Unmodified carbon paste electrode was prepared by mixing.0 g graphite powers and 0.2 mL paraffin oil adequately in agate

ortar. A portion of the resulting paste was then packed firmly

nto the electrode cavity (2.6 mm diameter) of a PTFE sleeve.lectrical contact was established via a copper wire. Other mod-

fied CPEs were prepared in the similar manner. The LNWs/CPE

sNb

Fig. 1. TEM image of La(OH)3 nanowires.

as prepared by mixing 0.98 g graphite powders, 0.02 g LNWsnd 0.2 mL paraffin oil. The carbon nanotubes (CNT) modifiedPE (CNT/CPE) was prepared by mixing 0.98 g graphite pow-ers, 0.02 g CNT and 0.2 mL paraffin oil. And the La(NO3)3odified CPE (La(NO3)3/CPE) was prepared as following:

.02 g La(NO3)3 was dissolved in 2 mL water. Then the solutionas mixed with 0.98 g graphite powders. After evaporating the

olvent water at room temperature, La(NO3)3/CPE was preparedy mixing the obtained dry powers with 0.2 mL paraffin oil. Theurface of all the modified and unmodified CPEs were carefullymoothed on weighing paper and rinsed with water prior to eacheasurement.

.4. General procedure

After mixing appropriate GO and 5.0 × 10−5 mol L−1 Cu(II)on in 0.1 mol L−1 NH3·H2O–NH4Cl buffer (pH 9.8) for 35 min,he obtained solution was transferred into a voltammetric cell.he second-order derivative single sweep voltammogram was

ecorded by applying a cathode-going potential scan from 0.5o −0.5 V at 100 mV s−1. The second-order derivative reductioneak current i′′p of the Cu(II)(NH3)4

2+–GO complex at −0.17 Vas measured. The increment �i′′p of peak current was obtained

y deducting the reduction peak current of the Cu(II)(NH3)42+

on from that of the Cu(II)(NH3)42+–GO complex, the calibra-

ion curve was obtained by plotting the increment �i′′p of theeduction peak current versus GO concentration. Cyclic voltam-etry was conducted under the same experimental condition.

After each measurement, the LNWs/CPE was regenerated by

uccessive potential cycling in the range of 0.5 to −0.5 V in theH3·H2O–NH4Cl buffer (pH 9.8) several times until the stableackground current was obtained. If necessary, the electrode

1 ta 71

sow

2

LTpTt

3

3

0

Fwo(twtbnwtlt

m

Fc

844 L. Liu et al. / Talan

urface was renewed by pushing a folium (2–3 mm in thickness)f the carbon paste of the tube out, then polishing and washingith water.

.5. Samples analysis

A 1.00 mL of GO injection (Xi’an Lijun Pharmaceutical Co.td., China) was diluted in 250 mL volumetric flask with water.he determination of GO in the obtained sample solution waserformed according to the general procedure mentioned above.he content of GO in the sample was calculated according to

he calibration curve.

. Results and discussion

.1. Voltammetric behavior of GO at LNWs/CPE

Typical cyclic voltammetric curves at LNWs/CPE in.1 mol L−1 NH3·H2O–NH4Cl buffers (pH 9.8) are shown in

aaii

ig. 2. Cyclic voltammograms at the LNWs/CPE (A), CNT/CPE (B), bare CPE (ontaining (a) 1.0 × 10−4 mol L−1 GO, (b) 5.0 × 10−4 mol L−1 Cu (II) and (c) 5.0 ×

(2007) 1842–1848

ig. 2A, when GO was present alone, no voltammetric responseas observed (Fig. 2A, curve a). When Cu(II) ion was presentnly, a pair of the well-defined redox peaks were observedFig. 2A, curve b). A reduction peak at −0.12 V was the reduc-ion of Cu(II)(NH3)4

2+ ion, and an oxidation peak at −0.02 Vas the oxidation of Cu(0). However, up adding GO, the reduc-

ion peak current of Cu(II)(NH3)42+ ion was greatly enhanced

y about 620% accompanying with a potential shift of 50 mV inegative direction, whilst the oxidation peaks current of Cu(0)as also increased by about 180% accompanying with a poten-

ial shift of 30 mV in positive direction (Fig. 2A, curve c). Theittle shift of peak potential was attributed to the re-formation ofhe Cu(II)(NH3)4

2+–GO complex in ammonium buffer [32].In order to check the reasons on great increase of the voltam-

etric responses, the voltammetric responses of K3Fe(CN)6

nd Cu(II)(NH3)4

2+–GO complex at bare CPE, LNWs/CPEnd CNT/CPE with the same geometrical surface were exam-ned, respectively. As shown in Fig. 3, K3Fe(CN)6 exhib-ted a pair of redox peaks at bare CPE in pH 6.7 phosphate

C) and La(NO3)/CPE (D) in 0.1 mol L−1 NH3·H2O–NH4Cl buffer (pH 9.8)10−4 mol L−1 Cu (II) and 1.0 × 10−4 mol L−1 GO. Scan rate 0.1 V s−1.

L. Liu et al. / Talanta 71 (2007) 1842–1848 1845

FCb

bactca(tCCCtgittfr(rflt

iCwTow

Nac

Fc1

pwbiCtrur(darLoCt

cobL

cfdmc

3

ig. 3. Cyclic voltammograms (CV) of 1.0 × 10−3 mol L−1 K3Fe(CN)6 at barePE (a), CNT/CPE (b) and LNWs/CPE (c). Supporting electrolyte: phosphateuffer (pH 6.7), scan rate 0.1 V s−1.

uffer. The difference �Ep of peak potentials was about 90 mVnd the ratio iPa/iPc of peak current was about 1 (Fig. 3,urve a). When LNWs/CPE was instead, the �Ep reducedo about 80 mV and the iPa/iPc still was about 1 (Fig. 3,urve c). Moreover, the voltammetric behavior of K3Fe(CN)6t CNT/CPE was almost the same as that at LNWs/CPEFig. 3, curve b). The results demonstrated that the elec-rode process of K3Fe(CN)6 was reversible at bare CPE,NT/CPE and LNWs/CPE. However, the peak currents at bothNT/CPE and LNWs/CPE were much higher than that at barePE. When scan rate � increased from 50 to 500 m Vs−1,

he peak current of the oxidation peak of K3Fe(CN)6rown gradually, according to the Randles–Sevcik equation:

pa = 2.69 × 105n3/2AC0DR1/2 v1/2 and presuming the parame-

ers such as n, C0, DR were constant, the linear regression equa-ions were ipa (�A) = 0.0298 + 0.1188 v1/2 (mV s−1)1/2, r = 0.999or the bare CPE, ipa (�A)= 0.3082 + 1.010 v1/2 (mV s−1)1/2,= 0.999 for the LNWs/CPE, and ipa (�A) = 0.306 + 1.012 v1/2

mV s−1)1/2, r = 0.998 for the CNT/CPE, respectively. From theatio of slopes of the three equations, the efficient electrode sur-aces of the LNWs/CPE and CNT/CPE were about 10 times asarge as that of bare CPE, and the efficient electrode surface ofhe LNWs/CPE was almost the same as that of CNT/CPE.

In 0.1 mol L−1 NH3·H2O–NH4Cl buffers (pH 9.8) contain-ng Cu(II) ion and GO, the reduction peak currents of theu(II)(NH3)4

2+–GO complex at LNWs/CPE (Fig. 2A, curve c)as higher about six times than that at CPE (Fig. 2C, curve c).he increase of the peak current was just related with the increasef the efficient electrode surface of LNWs/CPE in quatity, whichas owing to the introduction and surface effect of LNW.

Additionally, the voltammetric responses in 0.1 mol L−1

H3·H2O–NH4Cl buffers (pH 9.8) containing Cu(II) ion withnd without GO at CNT/CPE and LNWs/CPE with the same effi-ient electrode surface were compared. Without GO, the redox

Rtrs

ig. 4. Effect of pH value on reduction peak current of Cu(II)(NH3)42+–GO

omplex in NH3·H2O–NH4Cl buffer containing 5.0 × 10−5 mol L−1 Cu(II) and.0 × 10−5 mol L−1 GO. Scan rate: 0.1 V s−1.

eak currents of Cu(II)(NH3)42+/Cu(0) couple at CNT/CPE

ere almost the same as that at LNWs/CPE (Fig. 2B, curve). On adding the GO, the redox peak currents of Cu(II) ionsncreased (Fig. 2B, curve c). However, the peak currents of theu(II)(NH3)4

2+–GO complex at LNWs/CPE were higher aboutwo times than that at CNT/CPE. Additionally, the voltammet-ic responses at La(NO3)/CPE and CPE were also comparednder the same condition. Up adding GO, the redox peak cur-ents of the Cu(II)(NH3)4

2+–GO complex increased at both CPEFig. 2C, curve c) and La(NO3)3/CPE (Fig. 2D, curve c) toifferent extent. However, the increment of the peak currentt La(NO3)3/CPE was higher than that at CPE. The increaseesulted from the chemical coordination reaction of La(III) inNWs with GO [30,31] because different coordination sitesf La(III) and Cu(II) ions in GO made it possible for theu(II)(NH3)4

2+–GO complex to re-coordinate with La(III) ando be accumulated on the LNWs/CPE surface.

In summary, from these results mentioned above, it can beoncluded that the great increase of the voltammetric responsesf Cu(II)(NH3)4

2+–GO complex at LNWs/CPE resulted fromoth the surface and the chemical coordination function ofNWs.

With the reduction peak current of the Cu(II)(NH3)42+–GO

omplex at LNWs/CPE, a highly sensitive and selective methodor GO determination was proposed. Single sweep second-ordererivative voltammetry was employed used to record voltam-ogram as it can fast run and had good resolving power. The

onditions were optimized.

.2. Effect of pH values

Several alkaline supporting electrolytes such as Britton–

obinson, NH3·H2O–NH4Cl buffers and NaOH solution were

ested. The results showed that the second-order derivativeeduction peaks of the Cu(II)(NH3)4

2+–GO complex were moretable and sensitive in NH3·H2O–NH4Cl buffer than that in

1 ta 71 (2007) 1842–1848

otcTwbracpcoCfTnp

3

CtCCbctttuw

3

L

FC5

F 2+

cC

ptadd

3

Nognt

846 L. Liu et al. / Talan

ther media. The effect of pH values on the second-order deriva-ive reduction peak currents of the Cu(II)(NH3)4

2+–GO mixedomplex was examined in the pH range of 8.0–11.0 (Fig. 4).he peak currents increased with the increase of pH valueshen pH value was less than 9.7, and kept almost unchangedetween pH 9.7 and 9.8. When pH value was above 8.3, theeduction peak of the Cu(II)(NH3)4

2+–GO complex broadenednd the peak current reduced, which might resulted from thehange of the composition of the Cu(II)(NH3)4

2+–GO com-lex. Thus, 0.1 mol L−1 NH3·H2O–NH4Cl buffer (pH 9.8) washosen as supporting electrolyte. Meanwhile, with the increasef pH values, the peak potentials of reduction peak of theu(II)(NH3)4

2+–GO complex shifted positively, and obeyed theollowing equations: Epc (V) = −0.545 + 0.051pH (r = 0.997).he shift of peak potential might depend on the different coordi-ation states of the Cu(II)(NH3)4

2+–GO complex under differentH conditions.

.3. Effect of Cu(II) ion concentration

As the measurement depended on the formation of theu(II)(NH3)4

2+–GO complex and the reduction of Cu(II) ion inhe Cu(II)(NH3)4

2+–GO complex at LNWs/CPE, the effect ofu(II) ion concentration was examined. The results showed thatu(II) ion concentration was closely related to the peak currentut hardly to the peak potential. As shown in Fig. 5, the peakurrent increased with Cu(II) ion concentration increasing inhe range of (0–4.0) × 10−5 mol L−1. When Cu(II) ion concen-ration was in the range of 4.0 × 10−5 to 5.0 × 10−5 mol L−1,he peak current achieved a maximum value and was nearlynchanged. Accordingly, the final concentration of Cu(II) ionas 5.0 × 10−5 mol L−1.

.4. Effect of the mass radio of LNW in LNWs/CPE

Experimental results showed that the mass radio of LNW inNWs/CPE had a significant influence on peak current but not on

ig. 5. Effect of Cu (II) concentration on reduction peak current ofu(II)(NH3)4

2+–GO complex in NH3·H2O–NH4Cl buffer (pH 9.8) containing.0 × 10−5 mol L−1 Cu (II) and 1.0 × 10−5 mol L−1 GO. Scan rate: 0.1 V s−1.

mgcb

4

4

itcsoa�

or

(woo

ig. 6. Effect of reaction time on reduction peak current of Cu(II)(NH3)4 –GOomplex in NH3·H2O–NH4Cl buffer (pH 9.8) containing 5.0 × 10−5 mol L−1

u (II) and 1.0 × 10−5 mol L−1 GO. Scan rate: 0.1 Vs−1.

eak potential. The maximum peak current was obtained whenhe mass ratio of LNW and graphite power was 1:49. However,s the mass radio of LNW increased further, the peak currentecreased. Presumably, more LNWs reduced the electronic con-uctance of the electrode.

.5. Effect of reaction time

When GO was introduced into 0.1 mol L−1 NH3·H2O–H4Cl buffer (pH 9.8) containing Cu(II) ion, the peak currentf the reduction peak of the Cu(II)(NH3)4

2+–GO complex wasreater, and decreased gradually in half an hour and then it wasearly unchanged (Fig. 6). The reason of the decrease might behat GO first combined with more Cu(II)(NH3)4

2+ ions to formultinuclear oligomers, and the multinuclear oligomers then

radually transformed to be a more stable Cu(II)(NH3)42+–GO

omplex through a homogeneous ligand-exchange reactionetween Cu(II)(NH3)4

2+ ion and GO.

. Analytical performance

.1. Calibration curve for GO

In 0.1 mol L−1 NH3·H2O–NH4Cl buffer (pH 9.8) contain-ng 5.0 × 10−5 mol L−1 Cu(II) ion, the relationship betweenhe second-order derivative reduction peak current and GOoncentration was examined by second-order derivative linearweep voltammetry at scan rate of 0.1 V s−1. The peak sharpf the second-order derivative reduction peak recorded wascute and easy to be measured (Fig. 7), and the incrementsi′′p of the peak current were proportional to GO concentration

ver the range of 8.0 × 10−7 to 2.0 × 10−5 mol L−1. The linear

egression equation was �i′′p (�A s−2) = 0.013 + 0.014 × 106C

mol L−1), r = 0.9996. A detection limit of 3.5 × 10−7 mol L−1

as obtained (based on 3s1/m, where s1 is the standard deviationf the intercept and m is the slope), which was lower than thatf other electrochemical methods reported.

L. Liu et al. / Talanta 71 (2007) 1842–1848 1847

Table 1Determination and recovery test of GO in GO injections

Samples (injection) Found (g/250 mL) R.S.D. (%) Added (10−6 mol L−1) Found (10−6 mol L−1) Recovery (%) R.S.D. (%)

G

4

ttido

4

sww1sft

4

iTtd1m

F(G

5

pGbcofo

A

e

R

O 9.97 2.1 1.510.01 1.9 3.0

.2. Reproducibility

The proposed method showed good reproducibility, the rela-ive standard deviation (R.S.D.) for eight successive determina-ions of 5.0 × 10−6 mol L−1 GO at a LNWs/CPE was 2.1%, andndividual determination of 5.0 × 10−6 mol L−1 GO using fiveifferent LNWs/CPEs with the same surface area gave a R.S.D.f 2.8%.

.3. Interferences

In order to further investigate the selectivity of LNWs/CPE,everal potential interferences were tested. The tolerable limitas defined as the concentrations of foreign substances,hich gave an error less than ±5.0% in the determination of.0 × 10−5 mol L−1 GO. The results showed that 500-fold ofucrose and amylam, 100-fold of ascorbic acid, as well as 120-old of uric acid did not interfere with the determination. Fromhe results, the selectivity of the method is good.

.4. Determination of GO in clinic injections

The proposed method was applied to the determination of GOn clinic GO injections (10 g per 250 mL). The results (shown inable 1) were in good agreement with the content marked in

he label. The recovery experiments were carried out by stan-ard addition method and the recovery ranged from 97.6 to01.2%, which suggested acceptable accurate and precise of theethod.

ig. 7. Second-order derivative voltammetric curves in NH3·H2O–NH4Cl bufferpH 9.8) containing (a) 5.0 × 10−5 mol L−1 Cu (II); (b) a + 1.0 × 10−5 mol L−1

O. Scan rate: 0.1 Vs−1.

[

[[[[[

[[[

[

[

[[[

[

[

1.47 98.2 2.23.07 100.2 2.9

. Conclusion

In this work, a novel strategy for GO determination was pro-osed without either enzymatic reaction or direct oxidation ofO. The LNWs as electrode material were provided with dou-le function, not only increasing the electrode surface but alsohemically coordinating with GO. The sensitive reduction peakf the Cu(II)(NH3)4

2+–GO complex at LNWs/CPE can be usedor GO determination. Additionally, with the cathodic polarizedperation, more easily oxidizable compounds were tolerated.

cknowledgements

Thanks for the financial support of the National Natural Sci-nce Foundation of PR China (Grant No. 20475043).

eferences

[1] D.R. Shankaran, N. Uehara, T. Kato, Biosens. Bioelectron. 18 (2003)721–728.

[2] G.L. Luque, M.C. Rodrıguez, G.A. Rivas, Talanta 66 (2005) 467–471.[3] F. Ricci, D. Mosconea, C.S. Tuta, G. Palleschi, A. Amine, A. Posciab, F.

Valgimiglib, D. Messeri, Biosens. Bioelectron. 20 (2005) 1993–2000.[4] D.W. Pan, J.H. Chen, L.H. Nie, W.Y. Tao, S.Z. Yao, Anal. Biochem. 324

(2004) 115–122.[5] P.A. Fiorito, C.M.A. Brett, S.I.C.D. Torresi, Talanta 69 (2006) 403–408.[6] M.E. Ghica, C.M.A. Brett, Anal. Chim. Acta 532 (2005) 145–151.[7] S. Reiter, K. Habermuller, W. Schunmann, Sens. Actuators. B 79 (2001)

150–156.[8] N.J. Forrow, S.J. Walters, Biosens. Bioelectron. 19 (2004) 763–770.[9] Y. Nakabayashi, A. Omayu, S. Morii, S. Yagi, Sens. Actuators. B 66 (2000)

128–130.10] K. Yamamoto, H.S. Zenga, Y. Shen, M.M. Ahmed, T. Kato, Talanta 66

(2005) 1175–1180.11] T. Tatsuma, K. Saito, N. Oyama, Anal. Chem. 66 (1994) 1002–1006.12] J.J. Fei, K.B. Wu, F. Wang, S.S. Hu, Talanta 65 (2005) 918–924.13] H. Patel, X. Li, H.I. Karan, Biosens. Bioelectron. 18 (2003) 1073–1076.14] A. Gulce, H. Gulce, J. Biochem. Biophys. Methods 62 (2005) 81–92.15] Y.X. Huang, W.J. Zhang, H. Xiao, G.X. Li, Biosens. Bioelectron. 21 (2005)

817–821.16] I.L. Mattos, M.C.C. Areias, Talanta 66 (2005) 1281–1286.17] J. Wang, J.W. Mo, J. Porter, Anal. Chim. Acta 441 (2001) 183–189.18] H. Tang, J.H. Chen, S.Z. Yao, L.H. Nie, G.H. Deng, Y.F. Kuang, Anal.

Biochem. 331 (2004) 89–97.19] M. Yemini, M. Reches, E. Gazit, J. Rishpon, Anal. Chem. 77 (2005)

5155–5159.20] Y. Liu, M.K. Wang, F. Zhao, Z.A. Xu, S.J. Dong, Biosens. Bioelectron. 21

(2005) 984–988.21] S.Q. Liu, H.X. Ju, Biosens. Bioelectron. 19 (2003) 177–183.22] J. Wang, M. Musameh, Anal. Chim. Acta 539 (2005) 209–213.23] J. Wu, Y.H. Zou, N. Gao, J.H. Jiang, G.L. Shen, R.Q. Yu, Talanta 68 (2005)

12–18.24] I.G. Casella, M. Gatta, M.R. Guascito, T.R.I. Cataldi, Anal. Chim. Acta

357 (1997) 63–71.25] S.B. Aoun, Z. Dursun, T. Koga, G.S. Bang, T. Sotomura, I. Taniguchi, J.

Electroanal. Chem. 567 (2004) 175–183.

1 ta 71

[[[[[[

[

[

[

[[

[

[

848 L. Liu et al. / Talan

26] P.F. Luo, S.V. Prabhu, R.P. Baldwin, Anal. Chem. 62 (1990) 752–755.27] S. Mho, D.C. Johnson, J. Electroanal. Chem. 500 (2001) 524–532.28] Y. Sun, H. Buck, T.E. Mallouk, Anal. Chem. 73 (2001) 1599–1604.29] D.J. Watson, G.A. Attard, Electrochim. Acta 46 (2001) 3157–3161.30] I.H. Yeo, D.C. Johnson, J. Electroanal. Chem. 484 (2000) 157–163.31] E. Scavetta, M. Berrettoni, R. Seeber, D. Tonelli, Electrochim. Acta 46

(2001) 2681–2692.32] M.D. Socorro, M. Quintino, H. Winnischofer, M. Nakamura, K.

Araki, H.E. Toma, L. Angnes, Anal. Chim. Acta 539 (2005) 215–222.

33] E. Garnier, C. Bachmann, K. Servat, K.B. Kokoh, K. Bamba, J.M. Leger,Electrochim. Acta 50 (2005) 3341–3346.

34] J. Wang, M. Musameh, Anal. Chim. Acta 539 (2005) 209–213.

[

[

[

(2007) 1842–1848

35] X. Wang, Y.D. Li, Chem. Eur. J. 9 (2003) 5627–5635.36] X. Wang, X.M. Sun, Y.D. Li, Adv. Mater. 15 (2003) 1442–1444;

H. Tsukube, S. Shinoda, H. Tamiaki, Coord. Chem. Rev. 226 (2002)227–234.

37] Y.M. Yang, S.F. Weng, J.G. Wu, G.X. Wu, Chem. J. Chin. Univ. 15 (1994)646–650.

38] J.H. Feng, X.J. Li, F.K. Pei, X. Chen, S. Li, Y.X. Nie, Anal. Biochem. 301(2002) 1–7.

39] G. Cerchiaro, A.C.S. Ana, M.L.A. Temperini, A.M.D.C. Ferreira, Carbo-hydr. Res. 340 (2005) 2352–2359.

40] A. Diaz, I. Garcia, R. Cao, H. Beraldo, M.M. Salberg, D.X. West, L.G.E.Ochoa, Polyhedron 16 (1997) 3549–3555.

41] J. Fan, Talanta 42 (1995) 317–321.