Cu2O–Au nanocomposites for enzyme-free glucose sensing with enhanced performances

Colloids and Surfaces B: Biointerfaces 95 (2012) 279–283

Contents lists available at SciVerse ScienceDirect

Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb

Short communication

Cu2 O–Au nanocomposites for enzyme-free glucose sensing with enhanced performances Qiyan Hu a,∗ , Fenyun Wang b , Zhen Fang b , Xiaowang Liu b,∗ a b

Department of Pharmacy, Wannan Medical College, Wuhu 241002, PR China The College of Chemistry and Materials Science, Anhui Key Laboratory of Functional Molecular Solids, Anhui Normal University, Wuhu 241000, PR China

a r t i c l e

i n f o

Article history: Received 2 January 2012 Received in revised form 17 February 2012 Accepted 17 February 2012 Available online 3 March 2012 Keywords: Nanocomposite Glucose Electrochemical sensor

a b s t r a c t A facile method for the synthesis of Cu2 O–Au nanocomposites has been reported by injecting Cu2 O nanocubes into Au precursor directly with the assistance of ultrasound radiation at room temperature. The ultrasound radiation is not a necessary requirement but can make the distribution of Au nanoparticles more homogenous. The formation of Cu2 O–Au nanocomposites is attributed to following two reasons. The first one is the difference in the reduction potential between Cu2+ /Cu2 O and AuCl4 − /Au, which can also be considered as the driving force for the redox reaction. The other one is the low lattice mismatch between (2 0 0) planes of Cu2 O and (2 0 0) facets of Au, which is favorable for the formation of heterostructure. The electrochemical investigation demonstrates that the performances of Cu2 O nanocubes in enzyme-free glucose sensing have been improved significantly after the decoration of Au nanoparticles which may be derived from the polarization effect provided by Au nanoparticles. As-prepared Cu2 O–Au nanocomposites have great potential in enzyme-free glucose sensing. © 2012 Elsevier B.V. All rights reserved.

1. Introduction As a great deal of success has been achieved in the economy development, people are now paying more attention to improve their living conditions. It has been reported that diabetes is one of the serious diseases which can cause severe complications such as lower limb amputations, blindness and cardiovascular disease. Sometimes, it can even lead people to death. There are about over 100 million people suffering from diabetes in the world. According to the data released by the American Diabetes Association, 6% of the general U.S. population over age 40 has been found to have diabetes and an equal amount has not been diagnosed yet. Although there is no efficient way to cure diabetes, diseaseassociated complications can be reduced through the tight control of blood glucose levels. As a result, exploration of fast and reliable methods for glucose concentration monitoring in the treatment of diabetes is of great interest [1]. In the past decades, tremendous effort has been made in this area [2–8]. Among those reported methods used to probe glucose, electrochemical technique is quite attractive due to its unique features, such as low detection limit, high selectivity, high sensitivity and low cost [9]. Generally speaking, electrochemical techniques can be broadly categorized into two types, enzyme-assistant strategy and enzyme-free route. The

∗ Corresponding authors. Fax: +86 553 3869303. E-mail addresses: [email protected] (Q. Hu), [email protected] (X. Liu). 0927-7765/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfb.2012.02.023

former one is based on the detection of oxidation signal of hydrogen peroxide or the reduction signal of dissolved oxygen, which is produced or consumed in the oxidation process of ␤-d-glucose to d-glucono-␦-lactone catalyzed by glucose oxidase [3,6,10]. The biggest challenge in enzyme-assistant amperometric biosensing is how to properly immobilize oxidase enzymes onto the surface of electrodes. Direct electrochemistry of most redox enzymes on bare solid electrodes is difficult to achieve because of the deeply embedded active site in the proteins and the instability of the biological matrix [11]. Thus, oxidase-based amperometric techniques can hardly avoid following drawbacks: (i) less stability in sensing process due to the intrinsic nature of the enzymes; (ii) lack of simplicity and reproducibility in the fabrication of electrode as a result of the difficulty in enzyme immobilization; (iii) a high overpotential needed for the oxidation of enzymatically generated H2 O2 . The high overpotential may initialize other side electrochemical reaction at the interface of electrode and definitely reduce the selectivity of as-prepared sensor [3]. In contrast, enzyme-free method is more applicable. Furthermore, the rapidly developing nanotechnology provides further opportunity to this kind of sensors with enhanced sensitivity and selectivity [6,12]. Tremendous effort has been made to design and fabricate enzyme-free glucose sensors exploiting different kinds of nanomaterials, such as noble metals [13,14], alloys [15–17], and metal oxides [9,18–21]. Given the features of metal oxides, such as low cost, good electronic conductivity, non- or low-toxicity and facile preparation, metal oxide nanostructure-based enzyme-free glucose sensors have great potentials. As high sensitivity is one of

2. Experimental 2.1. Materials CuCl2 ·2H2 O, glucose, l-ascorbic acid, NaOH and chloroauric acid were purchased from Shanghai Chemical Reagent Co. Ltd. and used as received. 2.2. Synthesis of Cu2 O nanocubes Cu2 O cubes with large scale were synthesized using l-ascorbic acid as reductive agent under alkaline condition [24]. In a typical synthesis, 5.0 mL of CuCl2 (0.1 M) and 15.0 mL of NaOH (0.2 M) were first added into 200 mL distilled water. The obtained mixture was kept at room temperature for 5 min under magnetic stirring, and then 10.0-mL l-ascorbic acid (0.1 M) was injected. The reaction was further kept for another 1 h at room temperature. Finally, Cu2 O cubes were dried under vacuum at 50 ◦ C for 6 h after being harvested and washed with distilled water for several times. 2.3. Synthesis of Cu2 O–Au nanocomposites Typically, dried Cu2 O nanocubes were re-dispersed into distilled water with a concentration of ca. 18 mg/mL for the synthesis of Cu2 O–Au nanocomposites. 0.04 mL of as-prepared Cu2 O nanocube colloid solution was injected into chloroauric acid (0.20 mL, 0.15 mM) at room temperature with the assistance of ultrasound radiation. The blackish precipitates were produced immediately once the Cu2 O nanocubes were injected into gold precursor. The products were obtained after the reaction which was kept at room temperature for 30 s. 2.4. Electrochemical measurement Electrochemical experiments were performed on CHI 660 electrochemical analyzer (CHI, USA). The modified electrodes were fabricated by a reported method [21]. Initially, the glassy carbon (GC) electrodes were polished and washed with double distilled water; then 10 ␮L of colloid solution containing Cu2 O crystals or Cu2 O–Au nanocomposites or pure Au NPs which was obtained by treating Cu2 O–Au nanocomposites under acidic condition, was cast on the surface of GC electrodes and dried in air. 2.5. Characterization X-ray powder diffraction (XRD) patterns of pure Cu2 O nanocubes and Cu2 O–Au nanocomposites were conducted on a Shimadzu XRD-6000 X-ray diffractometer with a 2 range from 20◦ to 80◦ . Transmission electron microscopy (TEM) or high-resolution transmission electron microscopy (HRTEM) analysis and energydispersive X-ray (EDX) spectrometry of the products were carried

(222)

Cu2O-Au Cu2O

a 20

(311)

(220)

(200)

b

(110)

the primary goals in the development of sensing devices, further improvement on nanostructures-based glucose sensors with high sensitivity is still necessary. Recent study has shown that the catalytic behaviors of Fe3 O4 nanoparticles toward the reduction of H2 O2 can be tuned by the decoration of Au nanoparticles on their surfaces [22]. Cu2 O is a kind of electrochemically active materials toward glucose [21], it is reasonable to expect that the activity of Cu2 O nanostructures may also be improved after the decoration of Au nanoparticles because of the polarization effect provided by Au nanoparticles [23]. In this work, a facile method has been reported for the synthesis of Cu2 O–Au nanocomposites and the electrochemical study demonstrates that the electrochemical activity of Cu2 O nanocubes toward the detection of glucose has been improved dramatically after the decoration of Au nanoparticles on their surfaces.

(111)

Q. Hu et al. / Colloids and Surfaces B: Biointerfaces 95 (2012) 279–283

Intensity

280

30

40

50

2 Theta

60

70

80

Fig. 1. XRD patterns of Cu2 O nanocubes and as-prepared Cu2 O–Au nanocomposites.

out on Tecnai G2 20 TEM with an acceleration voltage of 200 kV. Xray photoelectron spectroscopy (XPS) study of the nanocomposites was recorded on an ESCALab220-XL electron spectrometer from VG Scientific using a monochromated Al K␣ X-ray source (1486.6 eV). A drop of diluted solution containing pure Cu2 O or Cu2 O–Au sample was cast on a carbon-coated copper grid and dried before TEM or HRTEM characterization. 3. Results and discussion After the addition of l-ascorbic acid to the mixture containing CuCl2 and NaOH, the mixture turned from blue to green, and then yellow, and finally became orange. XRD pattern (Fig. 1, Curve a) of the product matches well with the diffraction pattern of cubic Cu2 O (JCPDS file no. 05-0667). No characteristic peaks from impurities, such as CuO and Cu(OH)2 , were detected, indicating that the product is pure Cu2 O. The produced Cu2 O sample (Fig. 2a) has regular cubic shape and smooth surfaces with an average edge width about 50 nm. The pale yellow HAuCl4 solution was turned into black immediately after the addition of produced Cu2 O nanocubes with the assistance of ultrasound radiation. TEM analysis (Fig. 2b) of the product illustrates that numerous smaller nanoparticles with homogenous distribution were formed on the surfaces of Cu2 O nanocubes. The average diameter of the smaller nanoparticles is about 4 nm. The smaller nanoparticles may be Au nanoparticles because the product contains Au, Cu and O elements as shown in EDX spectrum (Fig. 2c). Note that the existence of Au nanoparticles in the product cannot be confirmed by the XRD investigation as no additional diffraction peaks from Au (Fig. 1, Curve b) were detected as a result of the small size and the low content of Au nanoparticles [25]. To determine the composition of the smaller nanoparticles, XPS technique was employed and the results are shown in Fig. 2d. The XPS spectrum of the nanocomposites illustrates the peaks derived from C 1s, Cu 2p, O1s and Au 4f. It is worth noting that the peaks (inset in Fig. 2d) located at 84.0 and 87.7 eV are assigned to Au 4f7/2 and 4f5/2, respectively. These are typical values for Au (0) [26], indicating the formation of gold nanoparticles on surfaces of Cu2 O nanocubes. XPS measurement further implies that the surfaces states of Cu2 O nanocubes did not change obviously after the reaction with gold precursor as only fine spectra of Cu 2p3/2 and Cu 2p1/2, which are raised from Cu2 O in the surface region, were observed. Ultrasound radiation used in our method leads to a homogenous distribution of Au nanoparticles on the surfaces of Cu2 O nanocubes. However, it is not a necessary requirement for the formation of Cu2 O–Au heterostructures. When Cu2 O nanocubes were added

Q. Hu et al. / Colloids and Surfaces B: Biointerfaces 95 (2012) 279–283

281

Fig. 2. TEM images of Cu2 O nanocubes (a) and Cu2 O–Au nanocomposites (b); EDX (c) and XPS (d) spectra of Cu2 O–Au nanocomposites, inset in (d) is the XPS spectrum of Au 4f.

into dilute HAuCl4 solution in the absence of ultrasound radiation, the mixture also turned into black instantly that is same as that in the presence of ultrasound radiation. TEM measurement of the product further confirmed the formation of Au nanoparticles (Fig. 3a). However, from a panoramic view of the sample, we can see that Au nanoparticles formed on the surfaces of Cu2 O nanocubes were inhomogeneous, both in density and in diameter. To obtained Cu2 O–Au nanocomposites with Au nanoparticles in homogenous distribution, dispersion of Cu2 O nanocubes into gold precursor should be conducted as quickly as possible as the redox reaction between Cu2 O and HAuCl4 is very fast. Ultrasound radiation is considered as a versatile way to disperse and synthesize

nanomaterials in different solutions as a result of its high ultrasonic energy [27]. Once ultrasound radiation was exploited, Cu2 O nanocubes will be homogenously re-dispersed into Au precursor in a very short period of time, that is, each Cu2 O nanocube will react with gold precursor equally. As a result, Cu2 O–Au nanocomposites with homogeneous Au nanoparticle distribution were produced. The formation of Cu2 O–Au nanocomposites is attributed to the following two reasons. Firstly, a large difference in the reduction potential between Cu2+ /Cu2 O (0.203 V, vs SHE) and AuCl4 − /Au pairs (0.99 V) is the driving force [28]. Once Cu2 O nanocubes are in contact with gold precursor, AuCl4 − will be reduced immediately by Cu (I) on the surface of Cu2 O nanocubes and the oxidized portion

Fig. 3. (a) TEM image of Cu2 O–Au nanocomposites synthesized in the absence of ultrasound radiation and (b) HRTEM image of Cu2 O–Au nanocomposites prepared with the assistance of ultrasound radiation.

282

Q. Hu et al. / Colloids and Surfaces B: Biointerfaces 95 (2012) 279–283

100 50

100

(a)

0

0

Au Cu2O Cu2O-Au

-100 -150 -200

I /µA

I /µA

-50

-150 -200

-300

-250 0.6

0.4

0.2

0.0

-0.2

-0.4

-0.6

I II III IV V VI

-50 -100

-250

0.8

(b)

50

-0.8

0.8

0.6

0.4

0

2.0 µM

4.0 µM

-5

0.0

-0.2

-0.4

-0.6

-0.8

-10

0

-15

-10

(c) 6.0 µM 8.0 µM 10.0 µM

-5

14.0 µM

I /µ A

I /µA

0.2

E / V vs. Ag/AgCl

E / V vs. Ag/AgCl

-15

-20 100

-20 0

200

20

40

60

80 100

[glucose] / µ M

300

400

500

600

t/s

700

800

900 1000

Fig. 4. (a) CVs of Au nanoparticles, pure Cu2 O nanocubes and Cu2 O–Au nanocomposites modified glassy carbon electrode in 0.1 M NaOH solution at scan rate of 0.05 V/s. (b) CVs of Au nanoparticles (I and II), pure Cu2 O nanocubes (III and IV) and Cu2 O–Au nanocomposites (V and VI)-modified GC electrode in the presence of glucose with different concentrations (I, III, V, 20 ␮M; II, IV, VI, 30 ␮M) under the same conditions; (c) amperometric responses of Cu2 O–Au nanocomposites-modified GC electrode upon the successive addition of glucose into gently stirred 0.1 M NaOH at 0.50 V. Inset: the linear relationships between the catalytic current and glucose concentration.

of Cu (I) will release from the solids as Cu2+ with the assistance of H+ provided by HAuCl4 . The less difference in lattice mismatch between (2 0 0) planes of Cu2 O and (2 0 0) facets of Au also contributes to the formation of Cu2 O–Au nanocomposites. The biggest challenge in heterostructures synthesis is desired heterogeneous nucleation on the original materials surfaces often competes with homogeneous nucleation of separate nanocrystals of the secondary material; as a result, heterostructures as well as pure secondary nanostructures were produced. Similar lattice distances between their (2 0 0) planes favors the Au nucleation on the surface of Cu2 O nanocubes. HRTEM image of Cu2 O–Au nanocomposites shows that their (2 0 0) planes with d-spacing about 0.20 nm are parallel to each other at the interfaces, which means that Au were epitaxially grown on the surfaces of Cu2 O (Fig. 3b). This point is similar to a previous observation in Au–Cu2 O core–shell nanocomposites [29]. It has been confirmed that noble metal nanoparticles grown on semiconductor nanostructures’ surfaces can significantly influence their optical properties because of the interaction between electromagnetic waves and the charged particles in metal–semiconductor nanocomposites [30,31]. As the work function difference between noble metal nanoparticles and semiconductor portions, charges at the interfaces of the nanocomposites will re-distribute [32]. The redistribution of the surface charges may alter their electrochemically catalytic ability. Recently, growth of Au nanoparticles on metal oxides can even serve as an available strategy to tune the electrochemically properties of metal oxides [22]. Investigation on systematic effect between noble metal nanoparticles and metal oxides in electrochemical sensing process and exploration of suitable way to improve the sensitivity of enzymefree biosensors are very interesting issues in the domain of science.

Cyclic voltammogram (CV) and amperometric i-t technique have been used to study the systematic effect between Au nanoparticles and Cu2 O nanocubes in the enzyme-free sensing of glucose. There is no obvious redox peaks in CV (Fig. 4a) at Au nanoparticlemodified GC electrode in 0.1 M NaOH solution at scan rate of 0.05 V/s, but a weak peak at −0.068 V at both Cu2 O nanocubes and Cu2 O–Au nanocomposites-modified GC electrodes (Fig. 4a), ascribing to the oxidation of Cu (I) to Cu (II), which is consistent with a previous report [21]. It is worth noting that no obvious oxidation peaks in the voltage window between 0.2 and 0.8 V were observed at the three modified electrodes in 0.1 M NaOH solution. When glucose with various concentrations was injected into NaOH solution, no obvious change (Fig. 4b, I and II) in CVs was detected at Au nanoparticles-modified GC electrode. However, at both Cu2 O cubes and Cu2 O–Au nanocomposites modified electrodes, new oxidation peaks around 0.4–0.5 V were produced and the increases in the oxidation peaks were proportional to the concentrations of glucose introduced in the solution. Note that when the same amount of glucose was injected into NaOH solution, the increases in oxidation peak at both modified electrodes were varied significantly. Specifically, the peak (Curve IV in Fig. 4b) around 0.46 V increased to 64.0 ␮A at Cu2 O–Au nanocomposites-modified GC electrode when the concentration of glucose was increased to 20 ␮M; while the increase (Curve III in Fig. 4b) at Cu2 O nanocube modified electrode was only promoted to 44.9 ␮A. Similar phenomenon was observed when glucose concentration was enhanced to 30 ␮M. These results show that Cu2 O–Au nanocomposites have higher sensitivity than pure Cu2 O nanocubes in the sensing of glucose. The enhanced performances in sensing glucose at Cu2 O–Au nanocomposites may ascribe to the polarization provided by Au nanoparticles, which was similar to the results reported by Sun [22].

Q. Hu et al. / Colloids and Surfaces B: Biointerfaces 95 (2012) 279–283

I-t technique was employed to investigate the possibility of using Cu2 O–Au nanocomposites as an enzyme-free biosensor to determining the concentration of glucose in the solution. Amperometric responses for Cu2 O–Au nanocomposites-modified electrode to successive additions of glucose at 0.46 V is shown in Fig. 4c. Upon the injection of glucose into the NaOH solution, the responses of Cu2 O–Au nanocomposites-modified GC electrode to the changes of glucose concentration is about 1 s and steady-state signals were obtained within 2 s. The inset in Fig. 4c shows the linear relationship between the catalytic current and glucose concentrations injected into the NaOH solution, which illustrates that Cu2 O–Au nanocomposites-modified GC electrode has a linear response range from 2.0 to 100.0 ␮M. The fitting equation is I (␮A) = −3.028–0.185*C (␮M) with a correlation coefficient of 0.998. Cu2 O–Au nanocomposites-modified GC electrode also shows high stability and good reproducibility in detection of glucose, with a relative standard deviation (RSD, n = 12) of 2.8%. The detection limit was estimated as 0.50 ␮M at a signal/noise ratio of three. These results demonstrate that as-prepared Cu2 O–Au nanocomposites have shown excellent sensitivity; low detect limit and good linear relationship between the catalytic currents and glucose concentrations. In other words, Cu2 O–Au nanocomposites have great potential in electrochemical sensing glucose with enhanced performances. 4. Conclusions In conclusion, a facile method has been reported for the synthesis of Cu2 O–Au nanocomposites that have shown excellent performers in enzyme-free sensing of glucose as a result of the systematic effect between Au nanoparticles and Cu2 O nanocubes. The growth of Au nanoparticles on the surfaces of Cu2 O nanocubes is ascribed to the large difference in their reduction between Cu2+ /Cu2 O and AuCl4 − /Au couples and low lattice mismatch between their (2 0 0) planes. Due to the features of as-prepared Cu2 O–Au nanocomposites in the detection of glucose, such as excellent sensitivity; low detect limit and good linear relationship between the catalytic currents and the concentrations of glucose, they have great potential applications in enzyme-free glucose sensing.

283

Acknowledgments This work was supported by Grant for Young and Middleaged Teachers of Wannan Medical College (WK 200922) and National Natural Science Foundation of PR China (Nos. 21001006 and 21171007). References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32]

A. Heller, B. Feldman, Acc. Chem. Res. 43 (2010) 963. R.J. McNichols, G.L. Coté, J. Biomed. Opt. 5 (2005) 5. S. Park, H. Boo, T.D. Chung, Anal. Chim. Acta 556 (2006) 46. J.C. Pickup, F. Hussain, N.D. Evans, O.J. Rolinski, D.J.S. Birch, Biosens. Bioelectron. 20 (2005) 2555. J. Wang, M. Musameh, Anal. Chim. Acta 539 (2005) 209. K.J. Cash, H.A. Clark, Trends Mol. Med. 16 (2010) 584. M.-W. Shao, Y.-Y. Shan, N.-B. Wong, S.-T. Lee, Adv. Funct. Mater. 15 (2005) 1478. M.-W. Shao, M.-L. Zhang, Y.-Y. Shan, D.-D. Ma, N.-B. Wong, S.-T. Lee, Nanotechnology 17 (2006) 3574. X. Zhang, A. Gu, G. Wang, Y. Wei, W. Wang, H. Wu, B. Fang, CrystEngComm 12 (2010) 1120. X. Liu, Q. Hu, Q. Wu, W. Zhang, Z. Fang, Q. Xie, Colloids Surf. B 74 (2009) 154. A. Riklin, E. Katz, I. Willner, A. Stocker, A.F. Buchmann, Nature 376 (1995) 672. A.-J. Wang, J.-J. Feng, Z.-H. Li, Q.-C. Liao, Z.-Z. Wang, J.-R. Chen, CrystEngComm 14 (2012) 1289. F. Battaglini, P.N. Bartlett, J.H. Wang, Anal. Chem. 72 (2000) 502. X.H. Kang, Z.B. Mai, X.Y. Zou, P.X. Cai, J.Y. Mo, Talanta 74 (2008) 879. R. Qiu, X.L. Zhang, R. Qiao, Y. Li, Y.I. Kim, Y.S. Kang, Chem. Mater. 19 (2007) 4174. J.P. Wang, D.F. Thomas, A.C. Chen, Anal. Chem. 80 (2008) 997. X. Bo, J. Bai, L. Yang, L.P. Guo, Sens. Actuators B 157 (2011) 662. S. Li, Y. Zheng, G.W. Qin, Y. Ren, W. Pei, L. Zuo, Talanta 85 (2011) 1260. Y. Zhao, J. Zhao, Y. Li, D. Ma, S. Hou, L. Li, X. Hao, Z. Wang, Nanotechnology 22 (2011) 115604. Y. Ding, Y. Wang, L. Su, M. Bellagamba, H. Zhang, Y. Lei, Biosens. Bioelectron. 26 (2010) 542. L. Zhang, H. Li, Y. Ni, J. Li, K. Liao, G. Zhao, Electrochem. Commun. 11 (2009) 812. Y. Lee, M.A. Garcia, N.A.F. Huls, S. Sun, Angew. Chem., Int. Ed. 49 (2010) 1271. Q. Hua, F. Shi, K. Chen, S. Chang, Y. Ma, Z. Jiang, G. Pan, W. Huang, Nano Res. 4 (2011) 948. Z. Wang, H. Wang, L. Wang, L. Pan, J. Phys. Chem. Solids 70 (2009) 719. H. Zhang, Y. Xie, Z. Sun, R. Tao, C. Huang, Y. Zhao, Z. Liu, Langmuir 27 (2011) 1152. W. Cheng, S. Dong, E. Wang, Langmuir 19 (2003) 9434. A. Gedanken, Ultrason. Sonochem. 11 (2004) 47. G. Millazzo, S. Caroli, Tables of Standard Electrode Potentials, John Wiley & Sons Inc., New York, 1978. C.H. Kuo, T.-E. Hua, M.-H. Huang, J. Am. Chem. Soc. 131 (2009) 17871. Y.-P. Hsieh, C.-T. Liang, Y.-F. Chen, C.-W. Lai, P.-T. Chou, Nanotechnology 18 (2007) 415707. J. Lee, A.O. Govorov, J. Dulka, N.A. Kotov, Nano Lett. 4 (2004) 2323. X. Wang, C.J. Summers, Z.L. Wang, Appl. Phys. Lett. 86 (2005) 013111.