A novel electrochemical SPR biosensor

A novel electrochemical SPR biosensor

Electrochemistry Communications 3 (2001) 489±493 www.elsevier.nl/locate/elecom A novel electrochemical SPR biosensor Xiaofeng Kang, Guangjin Cheng, ...

116KB Sizes 0 Downloads 15 Views

Electrochemistry Communications 3 (2001) 489±493

www.elsevier.nl/locate/elecom

A novel electrochemical SPR biosensor Xiaofeng Kang, Guangjin Cheng, Shaojun Dong

*

Laboratory of Electroanalytical Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun, 130022, People's Republic of China Received 22 February 2001; received in revised form 23 April 2001; accepted 25 April 2001

Abstract In this paper, we demonstrate for the ®rst time that upon electrochemical oxidation/reduction, the transition in the conductivity of polyaniline (PAn) ®lm on gold electrode surface leads to a large change of surface plasmon resonance (SPR) response due to a change in the imaginary part of dielectric constant of PAn ®lm. Based on the amplifying response of SPR to the redox transformation of PAn ®lm as a direct result of the enzymatic reaction between horseradish peroxidase (HRP) and PAn in the presence of H2 O2 , a novel PAn-mediated HRP sensor has been fabricated. The electrochemical SPR biosensor, unlike a usual binding assay with SPR, can a€ord a larger SPR response, and can also be reused by reducing the PAn ®lm electrochemically to its reduced state. This method opens up a new route to the fabrication of SPR biosensor. Ó 2001 Elsevier Science B.V. All rights reserved. Keywords: Surface plasmon resonance; HRP; Polyaniline; H2 O2 ; Biosensor

1. Introduction Surface plasmon resonance (SPR) is a very sensitive optical technique based on the excitation of surface charge±density wave, which can be produced at the interface between a metal (such as Au, Ag) and a dielectric, and propagates along the interface. Any change in the optical properties of the dielectric layer adjacent to the metal layer in¯uences the excitation of the plasmon, thus forming the basis of SPR sensing. At present, this technique has been used successfully to detect changes in the refractive index or the thickness of thin ®lms besides the application in the analysis of biomolecular interactions including antigen±antibody and receptor±ligand binding [1±5]. However, the maximum sensitivity that can be obtained in the above applications is limited by the amount of analyte that can directly bind or adsorb to the surface of the SPR substrate [6]. Electrochemical reaction occurring at the electrode surface is a heterogeneous process. Therefore it is possible to detect the electrochemical processes by using SPR technique [7±15]. Previously we combined SPR with electrochemical method to monitor the electro* Corresponding author. Tel.: +86-431-5682801; Fax: +86-4315689711. E-mail address: [email protected] (S. Dong).

chemical growth process of conducting polymer, and its electrochemical doping/dedoping behavior [15]. Here we attempt to further extend its application, o€er a novel PAn-mediated horseradish peroxidase (HRP) sensor. We demonstrate that SPR is sensitive to the change in the conductivity of PAn ®lm. Upon oxidation or reduction, PAn ®lm can change its conductivity by several orders of magnitude, which thus leads to a large change of SPR response due to the change of the ®lm dielectric property. Based on the SPR response to the change of the optical dielectric property of the PAn ®lm by enzymatic catalytic reaction, a novel electrochemical SPR biosensor is presented in this paper. This approach can also be extended by using other oxidoreductases, and applied to the measurements of di€erent substrates.

2. Experimental 2.1. Reagents Aniline (Aldirich, 99.7%) was used after distillation. 1,2-diaminobenzene (Aldirich, 99%) was puri®ed by sublimation before use and stored in the dark. Na2 SO4 (99%), citric acid (99.5%), disodium hydrogen phosphate (99.5%), and hydrogen peroxide (30% vol%) (Beijing Chemical) were used without further puri®ca-

1388-2481/01/$ - see front matter Ó 2001 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 8 - 2 4 8 1 ( 0 1 ) 0 0 1 7 0 - 9

490

X. Kang et al. / Electrochemistry Communications 3 (2001) 489±493

tion. Stock solutions of hydrogen peroxide were freshly prepared each day. Horseradish peroxidase (EC1.11.1.7, Type x, 275 units mg 1 ) was obtained from Sigma. All the aqueous solutions were prepared with doubly distilled deionized water. Ethanol used is of spectrophotometric grade. 2.2. SPR and electrochemical experiments Electrochemistry and SPR measurements were performed with a home-built electrochemical SPR system using Kretschmann optical con®guration. As described previously [15], a special feature of the setup was the electrochemical SPR cell, designed specially to drive SPR and electrochemistry `in situ'. The gold ®lm (thickness 46 nm) on the glass microscope slide (n ˆ 1:562 at 650 nm) was used as both the surface plasmon medium and the working electrode. The glass slide evaporated gold was pressed on a BaK7 cylindrical prism (n ˆ 1:566 at 650 nm) via refractive index matching liquid. The gold surface of the slide was covered using a silicone rubber sheet with a hole for the electrolyte contact. The exposed gold surface area was 0:8 cm2 . The linearly p-polarized light from a laser diode (650 nm, 5 mW) was directed through prism onto the gold electrode. SPR scanning-angle curves of the re¯ectivity versus incidence angle and SPR kinetic curves of the re¯ectivity versus time at a ®xed kinetic angle were recorded. The cell was provided with a saturated Ag/AgCl reference electrode and a Pt counter electrode. The three-electrode system was connected to EG&G PARC-175 universal programmer, PAR-174A polarographic analyzer, PAR-173 potentiostat, PAR-178 electrometer probe and an X±Y recorder for the electrochemical control and measurement.

3. Results and discussion 3.1. SPR responses of PAn ®lm before and after electrochemical redox switches It is known that PAn ®lm is highly conducting at low pH solution. However, in order to use PAn ®lm with ¯avoproteins it is essential to work in neutral or weakly acidic condition to avoid the destruction of the enzyme. The workable pH range for PAn ®lm grown from sulfuric acid solution is determined by the pKa (5.5) of the emeraldine form of the polymer, so that the ®lm is suitable for the operation at pH 5 [16,17]. In the citrate/ phosphate bu€er solution of pH 5, electrochemical characteristic of PAn ®lm on the gold electrode is shown by cyclic voltammogram in Fig. 1. It can been seen that PAn ®lm at pH 5 shows only one redox couple …Ea ˆ 3:6 V; Ec ˆ 0:2 V† in contrast to the well-separated two redox couples at low pH. Such behavior is similar to the previously reported characteristic of PAn ®lm on carbon electrode [16±18]. Many works [16±19] previously discussed the conductivity change of PAn ®lm with potential in a bu€er solution of pH 5, and demonstrated that there exists a fast and reversible transition in the conductivity around ‡0:28 V corresponding to a redox transition between a conducting state, emeraldine, and an insulating state, pernigraniline. Such e€ect is related to the anion transport to o€set the buildup of the positive charge generated by the oxidation of the ®lm. Fig. 2 shows representative SPR spectra obtained by scanning angle for bare gold (curve 1), and the PAn ®lm before (curve 2) and after (curve 3) electrochemical oxidation in the bu€er solution of pH 5. The SPR angular scanning curve for the oxidized PAn shows a shift to

2.3. Preparation of enzyme sensor PAn ®lm was electrochemically deposited on the gold electrode surface by holding potential at ‡0:9 V (versus Ag/AgCl) for 30 s in 0.05 M H2 SO4 =0:05 M aniline solution. The substrate was then rinsed with distilled water and ethanol, subsequently blown dry with nitrogen. The adsorption and immobilization of HRP on the top of PAn ®lm were carried out as following: PAncoated gold electrode was ®rst held at open circuit for 10 min in 0.1 M citrate/0.2 M phosphate/0.1 M Na2 SO4 bu€er solution (pH 5.0) containing 0.025 M 1,2-diaminobenzene and 2.5 mg/ml HRP to allow adsorption of the enzyme onto the PAn ®lm surface. Then poly (1,2diaminobenzene) ®lm containing HRP was deposited by potentiostatic control at ‡0:4 V (versus Ag/AgCl) for 4 min. Finally the resulting ®lm was washed with the bu€er solution of pH 5, and stored in the bu€er solution.

Fig. 1. Cyclic voltammograms (curve a) and potential dependence of the re¯ectance change …DR† at ®xed angle 62.6°A (curve b) for PAnmodi®ed gold electrode in 0.1 M citrate/0.2 M phosphate/0.1 M Na2 SO4 bu€er solution of pH 5. The scan rate was 50 mV/s.

X. Kang et al. / Electrochemistry Communications 3 (2001) 489±493

Fig. 2. SPR scanning-angle curves of re¯ectivity versus angle of incidence. PAn ®lm was electrochemically deposited on gold electrode by holding potential at 0.9 V for 30 s in 0.05 M H2 SO4 =0:05 M aniline solution. The ®lm was then washed with water and ethanol, and transferred to the 0.1 M citrate/0.2 M phosphate/0.1 M Na2 SO4 bu€er solution (pH 5) to record SPR curves. Curves 1, 2 and 3 correspond to the results obtained for bare gold electrode (as a comparison), the reduced PAn ®lm (after holding potential at 0 V for 5 min) and oxidized PAn ®lm (after holding potential at 0.9 V for 5 min), respectively.

higher angle of the minimum re¯ectance relative to that of the bare gold, as expected for a transparent thin ®lm. However, the SPR curve for the reduced PAn ®lm also shows very distinct changes in both the resonance depth and the width, in addition to the shift in the minimum re¯ectance angle. This change in shape of the SPR curve is ascribed to the change in the imaginary part of the dielectric constant, eimag , from zero value after the oxidation (transparent ®lm with a real dielectric constant) to non-zero value before the oxidation (absorbing ®lm with non-zero imaginary part of the dielectric constant). Optical constants are related to ®lm conductivity. At optical frequencies, a conducting material can be characterized by a complex dielectric constant, e. For light of wavelength k, incident on material with conductivity r, the dielectric constant can be expressed as follows [12]: e ˆ ereal ‡ i…2rk=c†; where c is the speed of light, the imaginary part of the dielectric constant is eimag ˆ 2rk=c. Hence the change in ®lm conductivity can cause a larger change of eimag . In our previous paper [15], the optical constants of the PAn ®lm had been extracted by non-linear least-squares ®tting to a four-layer Fresnel optical mode. The values were found to be 1.51 and 1:5 ‡ 0:052i for the oxidized and reduced forms, respectively. It can been seen from these optical constants and Fig. 2 that the e€ect of the ®lm conductivity on the imaginary part of the dielectric constant is much larger with respect to that of the real part of the dielectric constant. In SPR, the shift in the

491

resonance minimum angle can readily be associated with the changes of ereal and ®lm thickness. The changes in both the resonance depth and the width are mainly determined by the eimag [20]. Therefore, the change in ®lm conductivity mainly in¯uences the change in shape of the SPR curve. In Fig. 2, we observe that the potential switching between the reduced and oxidized forms of PAn causes a very large change of the re¯ectivity at ®xed angle 62.6° (marking DR in Fig. 2) relative to the shift of the resonance minimum (marking Dh in Fig. 2). The observed re¯ectance at 62.6° for the reduced PAn ®lm is larger than that of the oxidized PAn ®lm due to the change of the ®lm conductivity. In SPR measurement, two modes, recording the change of the resonance minimum angle and the re¯ectance change at the ®xed angle, can be chosen to study di€erent surface processes. Obviously, in the present system, the latter is favorable to pursue sensitive SPR response. Fig. 1 displays the cyclic voltammogram (curve a) of PAn ®lm and the typical SPR kinetic curve (curve b) taken by in situ monitoring of the change of the re¯ectance with the potential at 62.6°. We observe that the change of the re¯ectance shows a bimodal change around the CV peak potential, corresponding to the transformation between the oxidized and reduced states of PAn ®lm. The previous works have demonstrated that the redox transformation of PAn ®lm can lead to a large change of SPR signal due to the change of the ®lm conductivity resulting in a change in the imaginary part of dielectric constant of PAn ®lm. 3.2. SPR responses of enzyme switch It has been previously demonstrated that direct chemical communication between PAn and HRP in the presence of H2 O2 is possible without the need for any other mediator being added [10,16,17]. HPR is readily adsorbed onto PAn surface while PAn acts as an e€ective redox enzyme mediator. In the presence of H2 O2 , the reaction of HRP with H2 O2 is followed by oxidation of the PAn ®lm with the oxidized form of the enzyme, so that the PAn is switched from the reduced state to the oxidized state. The overall catalyzed reaction scheme is shown as follows: H2 O2 ‡ HPR…red† ! HPR…compound I† ‡ H2 O HRP…compound I† ‡ PAn…red† ‡ H‡ ! HRP…compound II† ‡ PAn…ox† HRP…compound II† ‡ PAn…red† ‡ H‡ ! HRP…red† ‡ PAn…ox† where HRP (red) is the ferric form of the enzyme, HRP (compound I) is the oxidized form of the enzyme, HRP (compound II) is the intermediate form of the enzyme. The redox transformation of the PAn ®lm as a result of the enzymatic reaction enables us to construct a novel

492

X. Kang et al. / Electrochemistry Communications 3 (2001) 489±493

electrochemical SPR biosensor, and applies to determine the substrate concentration quantitatively. Present sensor was fabricated by immobilizing HRP in an electropolymerized ®lm of poly (1,2-diaminobenzene) grown on top of a PAn base ®lm. Poly (1,2-diaminobenzene) ®lm was chosen because it can be electrochemically polymerized at a potential where PAn is conducting and it produces a highly active enzyme ®lm. The immobilization process of HRP on the top of PAn ®lm has been described in Section 2. This process can also be monitored by SPR measurement. After immobilizing HRP with poly (1,2-diaminobenzene), whether the oxidized ®lm or the reduced ®lm, scanning angle curves all remain unchanged in the shape (similar to curves 2 and 3 in Fig. 2). However, the angle of minimum re¯ectance occurs at a shift of 1.2° to higher angle, indicating the formation of poly (1,2-diaminobenzene) ®lm containing HRP on PAn ®lm surface. To observe the response of SPR to enzyme switch, we carried out the following experiments. The PAn/HRP electrode was initially held at the potential of 0 V for 3 min, and then was set in the potentiometer mode. After H2 O2 was added to the electrochemical SPR cell, the change of the re¯ectance with time at the ®xed angle 63.8° was recorded, and simultaneously the potential change was recorded by chronpotentiometry to monitor the redox transformation of the ®lm. Fig. 3 displays the results of such an experiment. As shown in Fig. 3, at zero time, the electrode potential positions at 0 V versus Ag/AgCl, the mediator is full reduced, no change in the re¯ectance is observed; Upon addition of H2 O2 , the electrode potential shifts positively and the re¯ectance produces a very distinct change, indicating the concentration of the reduced form of PAn near the electrode surface rapidly decreases through the enzymatic reaction. The potential value at a di€erent time can re¯ect

Fig. 3. SPR kinetic curve (curve 1) and chronopotentiometric curve (curve 2) of HRP/PAn ®lm during the enzymatic reaction in the presence of 1 mM H2 O2 .

the fraction of the oxidized or reduced forms of PAn ®lm. We note that when the potential approaches the formal potential of the PAn ®lm (around 0.28 V), the chronpotentiometric curve appears as a potential plateau at which the fraction of the oxidized form in the PAn ®lm approaches 0.5. In addition, we observe that the re¯ectance change becomes slow in the latter period. This may be ascribed to the decrease of driving force as a result of the transition from the conducting state to the insulating state. The above experiments demonstrated that SPR is sensitive to the transformation between the oxidized and reduced states of PAn by the enzymatic reaction. This redox transformation is shown to be dependent on the H2 O2 concentration. Fig. 4 depicts the SPR kinetic curves at di€erent H2 O2 concentrations. In the absence of H2 O2 , the re¯ectance remains unchanged due to no redox transformation of PAn. After adding H2 O2 , the re¯ectance rapidly changes indicating the redox transformation. This transformation is dependent on H2 O2 concentration. When a lower concentration of H2 O2 is used, a smaller transformation rate is obtained. For 0:5 mM H2 O2 , the transformation is completed in less than 100 s. The slopes of the SPR kinetic curves in the linear regions can re¯ect the rate of this transformation, and thus can be used to produce calibration curve. Fig. 5 shows the relationship between H2 O2 concentration and the slope of the curve in Fig. 4. Here the linear relationship was obtained by using the change rate rather than the time for complete transformation. This allows measuring low levels of H2 O2 in less than 100 s with the SPR biosensor. The inset in Fig. 5 clearly shows that H2 O2 concentration of the order of 10 uM

Fig. 4. SPR kinetic curves at various concentrations of H2 O2 . The angle was ®xed at 63.8°. After adding an aliquot of various concentrations of H2 O2 to 0.1 M citrate/0.2 M phosphate/0.1 M Na2 SO4 bu€er solution (pH 5), SPR kinetic curve of the re¯ectivity versus time was recorded. Between each measurement, the sensor was washed and reproduced at the potential of 0 V versus Ag/AgCl.

X. Kang et al. / Electrochemistry Communications 3 (2001) 489±493

493

amplify the chemical information between PAn and HRP into the SPR optical signal. This method provides a new route to the fabrication of SPR biosensor because the other oxidoreductases can all be used as immobilization enzyme to transform PAn ®lm. This work has shown that the SPR biosensor poses a number of potential advantages. First, a larger SPR signal can be obtained in the SPR measurement than in the direct binding assay of SPR. Second, the SPR measurement does not require electrochemical instrument to operate. We can simply employ the chemical transformation of the redox states in PAn ®lm by the enzymatic reaction. And after each measurement, it can also be reused by reducing the PAn ®lm electrochemically to its reduced state.

Fig. 5. Relationship between H2 O2 concentration and the slope of DR t curve in Fig. 4. The inset shows an ampli®ed calibration plot at low H2 O2 concentration.

can be detected. In addition, this biosensor can be reused by electrochemically reducing the PAn ®lm to its reduced state. In a set of experiments as shown in Fig. 4, the sensor was washed between each measurement and re-reduced at 0 V versus Ag/AgCl. The results show good reproducibility, and no detectable loss in activity over this period of up to 2 h. For the PAn/HRP sensor, whatever concentration of H2 O2 was used, the amplitude of the re¯ectance change after complete transformation between two redox states is always the same (see Fig. 4), and is the same as the amplitude on the PAn-modi®ed gold electrode obtained by the electrochemical potential scan (Fig. 1 curve b). Hence, the SPR response does not depend on the changes of both the refractive index of H2 O2 and the concentration of HRP intermediates. Poly (1,2-diaminobenzene) is a permselective insulating ®lm under this condition. Here the ®lm serves the dual purpose of immobilizing the enzyme onto the surface of the PAn ®lm and providing selectivity against interference in solution [16,17]. The poly (1,2-diaminobenzene) ®lm does not in¯uence the re¯ectance change due to its excellent chemical stability. 4. Conclusions In this paper, we describe a new electrochemical SPR biosensor using PAn as a mediator to transduce and

Acknowledgements The support for this study by the National Science Foundation of China is gratefully acknowledged.

References [1] L.S. Jung, J.S. Shumaker-Parry, C.T. Campbell, S.S. Yee, M.H. Gelb, J. Am. Chem. Soc. 122 (2000) 4177. [2] R. Georgiadis, K.P. Peterling, A.W. Peterson, J. Am. Chem. Soc. 122 (2000) 3166. [3] M. Malmqvist, Nature 361 (1993) 186. [4] J. Homola, S.S. Yee, G. Gauglitz, Sens. Actuators B 54 (1999) 3. [5] S. Toyama, A. Shoji, Y. Yoshida, S. Yamauchi, Y. Ikaariyama, Sens. Actuators B 52 (1998) 65. [6] S. Claire, S. Andrea, S. Katrina, K. Ste, Anal. Chem.. [7] D.G. Hanken, R.M. Corn, Anal. Chem. 69 (1997) 3665. [8] Y. Iwasaki, T. Horiuchi, M. Morita, O. Niwa, Surf. Sci. 427±428 (1999) 195. [9] Y. Iwasaki, T. Horiuchi, M. Morita, O. Niwa, Sens. Actuators B 50 (1998) 145. [10] S. Koide, Y. Iwasaki, T. Horiuchi, O. Niwa, E. Tamiya, K. Yokoyama, Chem. Commun. (2000) 741. [11] F. Mirkhalaf, D.J. Schi€rin, J. Electroanal. Chem. 484 (2000) 182. [12] R. Georgiadis, K.A. Peterlinz, J.R. Rahn, A.W. Peterson, J.H. Grassi, Langmuir 16 (2000) 6759. [13] T.T. Ehler, J.W. Walker, J. Jurchen, Y. Shen, K. Morris, B.P. Sullivan, L.J. Noe, J. Electroanal. Chem. 480 (2000) 94. [14] D.D. Schlereth, J. Electroanal. Chem. 464 (1999) 198. [15] X. Kang, Y. Jing, G. Cheng, S. Dong, Anal. Chem., submitted. [16] P.N. Bartlett, Y. Aastier, J. Chem. Soc. Chem. Commun. (2000) 105. [17] P.N. Bartlett, P.R. Birkin, J. Wang, Anal. Chem. 70 (1998) 3685. [18] J. Ye, R.P. Baldwin, Anal. Chem. 60 (1988) 1979. [19] E. Wang, A. Liu, Anal. Chim. Acta 252 (1991) 53. [20] Z. Salamon, H.A. Macleod, G. Tollin, Biochim. Biophys. Acta 1331 (1997) 117.