An antifouling electrochemical immunosensor for carcinoembryonic antigen based on hyaluronic acid doped conducting polymer PEDOT

Abstract

Nonspecific binding is a critical issue in protein immunoassays, and the construction of sensing interfaces that can dramatically resist nonspecific protein adsorption in complex biological media is of great significance. Herein, an antifouling electrochemical immunosensor for a tumor biomarker, carcinoembryonic antigen (CEA) was developed based on a novel conducting polymer composite, poly(3,4-ethylenedioxythiophene) (PEDOT) doped with hyaluronic acid (HA). The electrodeposited PEDOT/HA composite had a porous microstructure and strong hydrophilicity, and it also possessed many carboxylic groups on its surface that can be used for the immobilization of the capture probe CEA antibodies. The CEA immunosensor based on PEDOT/HA exhibited high sensitivity to CEA over a wide linear range of concentrations from 1 pg mL⁻¹ to 0.1 μg mL⁻¹, with a detection limit of 0.3 pg mL⁻¹. Moreover, the developed immunosensor is highly specific and antifouling, and it can be used for the assay of CEA in real human serum samples without suffering from strong nonspecific protein adsorption.

---

1. Introduction

In recent years, with more and more attention being paid to the early diagnosis of diseases, numerous immunosensors based on accepted biomarkers and indicators of serious cancers have emerged dramatically.¹ Carcinoembryonic antigen (CEA) is one of the most significant tumor biomarkers associated with pancreatic cancer, gastric cancer, colon cancer, colorectal cancer, breast cancer, and so on.^2,3 Currently, several conventional immunoassay procedures have been developed to detect tumor markers including the radioimmunoassay, the enzyme-linked immunosorbent assay (ELISA),⁴ the polarization assay,⁵ the fluorescence immunoassay,^6,7 and the electrochemical immunoassay. Among these methods, the electrochemical immunoassay is of particular interest because of its rapid response, low cost, high sensitivity, and simple instrumentation. However, achieving sensitive, fast and selective detection of CEA in real human blood serum remains challenging, due to the existence of a large number of nonspecific proteins in human serum. Therefore, it is necessary to design and fabricate a protein-resistant electrochemical immunosensor capable of reducing nonspecific adsorptions of cells and proteins in human blood serum.^8,9

The key factor in preventing nonspecific adsorption of proteins is generally believed to be the surface hydration. In recent years, zwitterionic polymers, synthetic polypeptides, PEG-based polymers, and short-chain molecules, which can tightly bound interfacial hydration layer, have been reported as low fouling materials with excellent antifouling ability.^10–15 As a hydrophilic anionic polysaccharide,⁸ hyaluronic acid (HA) is also endowed with outstanding hydrophilicity due to the composition of a repeated disaccharide containing carboxyl (–COOH) groups, hydroxy (–OH) groups and amide groups (CO–NH), which has been demonstrated to be antifouling towards protein in previous studies^16,17 and has received great applications in medical and cosmetic industries.¹⁸ However, HA is non-conductive, immensely restricting its application in electrochemical assays.

A possible approach to expand the electrical application of HA involves packing it with certain conductive materials, especially conducting polymers. Poly(3,4-ethylenedioxythiophene) (PEDOT) is one of the most promising conducting polymers. Since the discovery of PEDOT in the 1980s,¹⁹ it has received much attention in both academia and industry. Owing to its unique properties, such as ease of synthesis, high conductivity (1 × 10¹ to 1 × 10² S cm⁻¹),²⁰ and good environmental stability,²¹ it has been widely applied in the fields of electrochemical sensors, electrochromic display devices, supercapacitors, organic light emitting diodes, drug delivery systems, neural prostheses, etc.^19,22–24 PEDOT can be polymerized by chemical or electrochemical oxidation methods from 3,4-ethylenedioxythiophene (EDOT) monomer. During polymerization, the positively charged PEDOT backbones need to be neutralized by negatively charged ions from various chemicals, such as sodium dodecyl sulfate (SDS),²⁵ sodium poly(styrene sulphonate) (NaPSS),²⁶ ionic liquids (ILs)²⁷ and so on. Since HA possesses carboxylic groups that are negatively charged in neutral solution, it may be incorporated into PEDOT through polymerization. Considering that fact that HA can enhance the interfacial hydration and PEDOT has very high conductivity, the combination of these two components may generate a new material with unique properties from both components.

Herein, a novel composite of PEDOT doped with HA was synthesized through a simply electrochemical method. The prepared PEDOT/HA composite was found to be of highly porous microstructure and exhibited good antifouling property. Based on the excellent properties of the PEDOT/HA, an electrochemical immunosensor for the detection of the tumor biomarker CEA was developed through the immobilization of CEA antibody (anti-CEA) onto the PEDOT/HA composite surface. The developed CEA immunosensor was highly sensitive and capable of sensing in complex media with low fouling.

2. Experimental

2.1 Materials

Poly(sodium-p-styrenesulfonate) (PSS), 3,4-ethylenedioxythiophene (EDOT), bovine serum albumin (BSA), lysozyme (Lys) and monoethanolamine were purchased from Aladdin Reagents (Shanghai, China). HA (M_W = 100 kDa) extracted from rooster comb, 1-ethyl-3-(3-(dimethylamino)propyl)-carbodiimide (EDC) hydrochloride and N-hydroxysulfosuccinimide (NHS) were purchased from Sigma-Aldrich (Beijing, China). Carcinoembryonic antigen (CEA) antibody and antigen, alpha fetoprotein (AFP) antigen was purchased from Beijing Xinkezhongjing Biological Technology. Phosphate buffered saline (PBS, pH 7.4, containing 0.9% NaCl) was used as buffer for all related experiments. The human serum sample (serum was centrifuged for 10 min at 10 000 rpm, and then the collected supernatant was diluted by 0.2 M pH 7.4 PBS) was obtained from The Eighth People's Hospital of Qingdao, China. All aqueous solutions were prepared using millipore water produced by a Milli-Q water purifying system.

2.2 Instruments and measurements

Electrochemical polymerization was performed with a CHI660E electrochemical workstation (Shanghai CH Instruments Co., China), and other electrochemical characterization and detection was performed with a GAMRY Reference 3000 Potentiostat (Gamry Instruments Co., USA). A conventional three-electrode system was used, with an Ag/AgCl (3.0 M KCl) electrode as the reference electrode, a platinum wire electrode as the auxiliary electrode, and a bare or modified glassy carbon electrode (GCE, 3.0 mm in diameter) as the working electrode.

The fabrication process of the immunosensor was monitored using cyclic voltammograms (CV). Differential pulse voltammetry (DPV) experiments were used to detect CEA in PBS (0.2 M, pH 7.4) solution containing 5.0 mM [Fe(CN)₆]^4−/3−and 0.1 M KCl with the potential scanning between 0 V and 0.4 V at a step size of 2 mV s⁻¹ and pulse size E of 25 mV.²⁸ Scanning electron microscope (SEM) was performed with a JEOL JSM-7500F SEM instrument (Hitachi High-Technology Co., Ltd, Japan). Energy dispersive X-ray spectroscopic (EDS) was performed with an Oxford X-Max 50 (Oxford Instruments, England). Contact angle measurement was performed with a JC2000 Instrument (Shanghai Zhongchen instrument Co., China). Fourier transform infrared spectroscopy (FTIR) was recorded using the BRUKER TENSOR 70 spectrometer (Bruker Optics, Germany).

2.3 Electrodeposition of hydrated PEDOT/HA nanocomposite

GCE was polished with 0.3, 0.05 mm alumina slurries in sequence and ultrasonically washed in distilled water, ethanol and ultrapure water for about 3 min, respectively. HA doped PEDOT (PEDOT/HA) was electrodeposited on GCE surface from an aqueous solution containing 0.02 M EDOT and 2.0 mg mL⁻¹ HA, at a constant potential of 1.0 V for 150 s. After thoroughly washed with water and PBS, the prepared PEDOT/HA/GCE was immersed in PBS (10 mM, pH 7.4) and stored for later use. For comparison, commonly studied PEDOT/PSS was prepared similarly but using PSS (2.0 mg mL⁻¹) instead of HA.

2.4 CEA immunosensor fabrication

The fabrication process of the CEA immunosensor using the prepared biocompatible PEDOT/HA is schematically illustrated as Scheme 1. In brief, the –COOH groups of PEDOT/HA/GCE were activated, in a PBS buffer solution (10 mM, pH 6.0) containing 50 mg mL⁻¹ NHS and 50 mg mL⁻¹ EDC for 1 h at 37 °C, for the covalent attachment of CEA antibody.²⁹ The activated PEDOT/HA surface was then rinsed with deionized water to remove unreacted NHS and EDC. The covalent attachment of CEA antibody (1.0 μM, 3 h) on PEDOT/HA/GCE surface was based on the reaction of activated carboxylic sites on PEDOT/HA/GCE surface with amine groups. The obtained anti-CEA/PEDOT/HA/GCE was rinsed with deionized water to remove the physically adsorbed CEA antibody, and the un-reacted activated carboxylic groups were finally treated with monoethanolamine.

Scheme 1 Schematic illustration of the fabrication process of the CEA immunosensor.

3. Results and discussion

3.1 Characterization and property of materials

3.1.1 SEM characterization. The surface morphologies and microstructures of PEDOT/HA and PEDOT/PSS were investigated using SEM. As shown in Fig. 1, the PEDOT/PSS composite showed a generally uniform and smooth film (Fig. 1A and B). Clearly, the PEDOT/HA composite showed a highly rough and porous morphology (Fig. 1C and D), which may be ascribed to the presence of micellar anionic in the HA solution, which guided the growth of PEDOT.³⁰ The unique microstructure of PEDOT/HA greatly increased the specific surface area of the modified electrode and provided large surface area for the attachment of CEA antibodies.

Fig. 1 SEM images of (A and B) PEDOT/PSS composite and (C and D) PEDOT/HA composite electrodeposited on electrode surfaces for 150 s. (B) and (D) are of higher magnifications.

3.1.2 FTIR and EDS characterization. The PEDOT/HA composite was characterized with FTIR spectroscopy (Fig. 2). Band at 1635 cm⁻¹ is originated from the stretching modes of C=C bonds in thiophene rings.²⁷ Moreover, the (N–H, O–H) and COO–H bonds of the anionic group of HA show stretching vibration peaks at about 3440 cm⁻¹ and 1384 cm⁻¹,³¹ respectively. To further verify the successful preparation presence of PEDOT/HA, its components were further analyzed elementally by EDS. The total elemental analysis of a SEM sample (Fig. S1†) confirmed that HA and PEDOT were efficiently existed. The high content of nitrogen is solely originated from the HA, and the sulphur is derived from the EDOT. Judging from the FTIR, SEM and EDS characterization above, it can be concluded that the PEDOT/HA composite materials have been successfully prepared.

Fig. 2 FTIR spectra and functional group analysis of PEDOT/HA composite.

3.1.3 The antifouling analysis in real human serum samples. To assess the antifouling property of PEDOT/HA/GCE (PEDOT/PSS/GCE as control), CV measurements based on the electrochemical redox of [Fe(CN)₆]^4−/3− in serum were employed.³² Before the CV measurement, both modified electrodes were soaked in 0%, 1%, 5%, 10%, 20%, 50% and 100% serum (V/V) for 30 min in sequence. As shown in Fig. 3, along with the increase of the concentration of serum, the peak current of the PEDOT/PSS/GCE significantly decreased (Fig. 3A), mainly owing to the nonspecific adsorption of proteins onto the electrode surface that blocked the electron transfer of the redox probe [Fe(CN)₆]^4−/3−. While for the PEDOT/HA/GCE, slight changes in the peak currents were observed (Fig. 3B). The performances of these two different modified electrodes in 100% serum have also been evaluated (Fig. S2†). As summarized in Fig. 3C and Fig. S2,† the peak current changes caused by the serum biofouling (nonspecific protein adsorption) at the PEDOT/HA/GCE were much smaller than that at the PEDOT/PSS/GCE, which indicated that the prepared PEDOT/HA composite has excellent antifouling property, and it can significantly reduce the nonspecific adsorption of proteins from human serum.

Fig. 3 CV responses of PEDOT/PSS/GCE (A) and PEDOT/HA/GCE (B) in different serum concentrations. From outer to inner, the serum concentrations are 0, 1, 5, 10, 20, 50 and 100 percent, respectively. (C) Corresponding reductive peak currents of PEDOT/PSS/GCE (a) and PEDOT/HA/GCE (b) with different serum concentrations. The CV curves are obtained in PBS (0.2 M, pH 7.4) solution containing 5.0 mM [Fe(CN)₆]^4−/3− and 0.1 M KCl, in the presence of different concentration. Error bars represent the standard deviations of the reductive peak currents obtained from three repeated determinations.

Moreover, well defined CV curves with clear redox peaks in Fig. 3A and B proved that the PEDOT/HA, just like the PEDOT/PSS, possessed good conductivity. Additionally, Fig. S3† showed the excellent long-term stability of the PEDOT/HA modified electrode. Therefore, on basis of the outstanding antifouling property, prominent conductivity, and excellent long-term stability, the PEDOT/HA can be used as a suitable substrate for the development of antifouling electrochemical sensors and biosensors.

3.2 Construction and sensing application of the CEA immunosensor

3.2.1 Contact angle characterization. For the construction of the CEA immunosensor based on the PEDOT/HA composite, the strategy is to covalently immobilize CEA antibody onto the PEDOT/HA surface through the reaction between amine groups of the antibody and carboxylic groups of the HA (HA doped PEDOT). In order to prove that free carboxyl groups (–COOH) are existed on the surface of the PEDOT/HA composite, the contact angles of the PEDOT/HA surface at a series of pH values were measured. As shown in Fig. 4, with the increase of the pH value from 2 to 11, the contact angle gradually reduced, indicating that the hydrophilicity of the PEDOT/HA surface increased accordingly. For the PEDOT/HA electropolymerized from aqueous solution containing EDOT and HA, the possibly charged group on its surface is only the carboxylic group of the HA. If there are free carboxylic groups on the PEDOT/HA surface, along with the increase of pH value, the carboxylic groups will gradually become deprotonated and carry more charges, which can then increase the hydrophilicity of the PEDOT/HA surface. Therefore, according to the observed contact angle decrease (or hydrophilicity increase) associated with the increase of pH value, it is reasonable to conclude that there are free carboxylic groups on the surface of the PEDOT/HA composite.³³

Fig. 4 Contact angles of the PEDOT/HA surface measured at different pH values. Insets show pictures of the contact angle of 46° (pH = 2), 28° (pH = 7) and 19° (pH = 11).

3.2.2 Electrochemical characterization. For the fabrication and test of the immunosensor, the bare GCE was modified with PEDOT/HA and CEA antibody, and then incubated with CEA, and this process was monitored using CV. As shown in Fig. 5, the peak current signal of the PEDOT/HA/GCE was significantly larger than that of the bare GCE, indicating that the PEDOT/HA composite has good conductivity. When CEA antibodies were immobilized onto the PEDOT/HA/GCE surface, the peak current signal of the anti-CEA/PEDOT/HA/GCE decreased, owing to the fact that the insulating CEA antibodies prevented the electron transfer to some degree. As expected, there was a further decrease in the peak current signal after the specific binding of target (CEA) to the immunosensor (anti-CEA/PEDOT/HA/GCE), as the insulating protein layers on the electrode surface can retard the charge transfer.^34,35 Therefore, a CEA immunosensor can be successfully fabricated based on the electrochemical current signal change associated with the specific antibody–antigen recognition.

Fig. 5 The cyclic voltammograms of the bare GCE (a), PEDOT/HA/GCE (b), anti-CEA/PEDOT/HA/GCE (c), and CEA (0.1 ng mL⁻¹)/anti-CEA/PEDOT/HA/GCE (d) in PBS (0.2 M, pH = 7.4) solution containing 5.0 mM [Fe(CN)₆]^4−/3− and 0.1 M KCl with the potential range from −0.2 to 0.6 V at a scan rate of 100 mV s⁻¹.

3.2.3 Sensing performance of the immunosensor. To achieve high sensitivity, DPV was used for the real-time detection of CEA based on the current changes induced by antibody–antigen interaction in 0.2 M PBS (pH 7.4) containing 5.0 mM [Fe(CN)₆]^4−/3− and 0.1 M KCl. As shown in Fig. 6A, with the increase of CEA concentration, the current (I) decreased accordingly, and the change in current (ΔI)³⁶ (Fig. 6A, inset) increased accordingly. It was found that the change in current (ΔI) showed a satisfactory linear correlation with the logarithmic value of target concentrations from 1 pg mL⁻¹ to 0.1 μg mL⁻¹ (Fig. 6B), with a limit of detection (LOD) of 0.3 pg mL⁻¹ (S/N = 3). This high sensitivity may be attributed to the good biocompatibility of the prepared hydrophilic PEDOT/HA composite, which is suitable for CEA antibodies to retain their bioactivity and binding affinity toward CEA.

Fig. 6 (A) DPV responses of the immunosensor toward different concentrations of CEA from 0 to 10⁻⁶ g mL⁻¹ in 0.2 M PBS (pH 7.4) containing 5.0 mM [Fe(CN)6]^4−/3− and 0.1 M KCl. Inset shows the DPV curves after background subtraction (B) changes in current of the immunosensor as a function of analyte concentration. Inset shows the corresponding calibration curve. Error bars represent the standard deviations of the change in current (ΔI) obtained from three repeated determinations.

3.2.4 Selectivity of the immunosensor. For practical applications, biosensors are generally used to detect their target molecules in multicomponent solutions, and therefore their selectivities need to be assessed. To investigate the selectivity of the immunosensor, the signal response (change in current) of the prepared immunosensor was assessed in various solutions containing BSA, Lys, AFP, or target antigen, either separated or combined. As shown in Fig. 7, the prepared immunosensor showed negligible signal responses to BSA, Lys and AFP even at very high concentrations, demonstrating excellent selectivity to its target CEA antigen. This result may be ascribed to the following two reasons: (i) the strong specific binding between CEA antibody and antigen, and (ii) the antifouling property of the PEDOT/HA composite, which can effectively reduce the nonspecific protein adsorption. Furthermore, the reproducibility of CEA immunosensor was also investigated. Five immunosensors independently fabricated were used to detect 0.1 ng mL⁻¹ CEA, and a relative standard deviation (RSD) of 4.43% was obtained, indicating good reproducibility.

Fig. 7 Responses of the CEA immunosensor to BSA, Lys, AFP, target antigen, and a mixture of all the above substances, respectively. Error bars represent the standard deviations of three repeated determinations.

3.2.5 Analysis in real human serum samples. In order to evaluate the feasibility of the developed immunosensor for possible applications, the recoveries of different concentrations of CEA in real human serum were detected with the prepared immunosensor. As summarized in Table 1, the recoveries were between 98.7–105.6% and the RSD was between 3.92–7.64%. These results suggested that the developed immunosensor can be used to detect CEA in real human serum samples.

Table 1 Recovery tests of CEA in serum

Samples	Serum samples (ng mL^−1)	Added (ng mL^−1)	Found (ng mL^−1)	Recovery (%)	RSD (%) (n = 5)
Sample 1	2.52	7.5	10.58	105.6	7.64
Sample 2	2.52	47.5	49.37	98.7	7.31
Sample 3	2.52	97.5	102.93	102.9	3.92

---

4. Conclusions

A novel PEDOT/HA composite can be successfully prepared through electrochemical polymerization. The prepared PEDOT/HA exhibited good conductivity and antifouling ability owing to its excellent hydrophilicity, and it possessed many carboxylic groups on its surface that can be used for the immobilization of biomolecules. Taking advantage of the unique properties of PEDOT/HA, such as conductivity, porous microstructure (large surface area), hydrophilicity and modifiability (containing carboxylic groups), an ultrasensitive, highly selective, and low fouling electrochemical immunosensor for a typical tumor marker CEA was constructed, and it showed promising potential in real clinical applications. It is expected that the prepared novel conducting polymer PEDOT/HA can be used for the development of other low fouling electrochemical sensors and biosensors.

Acknowledgements

This research was supported by the National Natural Science Foundation of China (21275087, 21422504), the Natural Science Foundation of Shandong Province of China (JQ201406), and the Taishan Scholar Program of Shandong Province of China.

Notes and references

1. N. Liu, Z. Liu, H. Han and Z. Ma, J. Mater. Chem. B, 2014, 2, 3292–3298.
2. P. Gold and S. O. Freedman, J. Exp. Med., 1965, 121, 439–462.
3. Y. Qiu, D. Deng, Q. Deng, P. Wu, H. Zhang and C. Cai, J. Mater. Chem. B, 2015, 3, 4487–4495.
4. R. M. Lequin, Clin. Chem., 2005, 5, 2415–2418.
5. D. M. Jameson and J. A. Ross, Chem. Rev., 2010, 110, 2685.
6. A. D. Quach, G. Crivat, M. A. Tarr and Z. Rosenzweig, J. Am. Chem. Soc., 2011, 133, 2028–2030.
7. W. Zhao, W. P. Zhang, Z. L. Zhang, R. L. He, Y. Lin, M. Xie, H. Z. Wang and D. W. Pang, Anal. Chem., 2012, 84, 2358–2365.
8. C. Blaszykowski, S. Sheikh and M. Thompson, Biomater. Sci., 2015, 3, 1335–1370.
9. C. Blaszykowski, S. Sheikh and M. Thompson, Chem. Soc. Rev., 2012, 41, 5599–5612.
10. S. Jiang and Z. Cao, Adv. Mater., 2010, 22, 920–932.
11. A. D. White, A. K. Nowinski, W. Huang, A. J. Keefe, F. Sun and S. Jiang, Chem. Sci., 2012, 3, 3488–3494.
12. Q. Shi, Y. Su, W. Chen, J. Peng, L. Nie, L. Zhang and Z. Jiang, J. Membr. Sci., 2011, 366, 398–404.
13. X. Luo, Q. Xu, T. James and J. J. Davis, Anal. Chem., 2014, 86, 5553–5558.
14. C. Leng, H. C. Hung, S. Sun, D. Wang, Y. Li, S. Jiang and Z. Chen, ACS Appl. Mater. Interfaces, 2015, 7, 16881–16888.
15. M. C. Lukowiak, S. Wettmarshausen, G. Hidde, P. Landsberger, V. Boenke, K. Rodenacker, U. Braun, J. F. Friedrich, A. A. Gorbushina and R. Haag, Polym. Chem., 2015, 6, 1350–1359.
16. S. Yamanlar, S. Sant, T. Boudou, C. Picart and A. Khademhosseini, Biomaterials, 2011, 32, 5590–5599.
17. R. Huang, X. Liu, H. Ye, R. Su, W. Qi, L. Wang and Z. He, Langmuir, 2015, 31, 12061–12070.
18. X. Liu, R. Huang, R. Su, W. Qi, L. Wang and Z. He, ACS Appl. Mater. Interfaces, 2014, 6, 13034–13042.
19. J. A. Vigil, T. N. Lambert and K. Eldred, ACS Appl. Mater. Interfaces, 2015, 7, 22745–22750.
20. A. Elschner, S. Kirchmeyer, W. Lovenich, U. Merker and K. Reuter, PEDOT: principles and applications of an intrinsically conductive polymer, CRC Press, 2010.
21. D. Yoo, J. Kim and J. H. Kim, Nano Res., 2014, 7, 717–730.
22. S. Bhandari, M. Deepa, A. K. Srivastava, A. G. Joshi and R. Kant, J. Phys. Chem. B, 2009, 113, 9416–9428.
23. Y. Lin, K. Liu, C. Wang, L. Li and Y. Liu, Anal. Chem., 2015, 87, 8047–8051.
24. X. Luo, C. L. Weaver, D. D. Zhou, R. Greenberg and X. T. Cui, Biomaterials, 2011, 32, 5551–5557.
25. N. Sakmeche, S. Aeiyach, J. J. Aaron, M. Jouini, J. C. Lacroix and P. C. Lacaze, Langmuir, 1999, 15, 2566–2574.
26. W. Wang, M. A. Ruderer, E. Metwalli, S. Guo, E. M. Herzig, J. Perlich and P. Müller-Buschbaum, ACS Appl. Mater. Interfaces, 2015, 7, 8789–8797.
27. M. Döbbelin, R. Marcilla, C. Pozo-Gonzalo and D. Mecerreyes, J. Mater. Chem., 2010, 20, 7613–7622.
28. Y. Wu, Z. Dou, Y. Liu, G. Lv, T. Pu and X. He, RSC Adv., 2013, 3, 12726–12734.
29. R. Q. Zhang, S. L. Liu, W. Zhao, W. P. Zhang, X. Yu, Y. Li, A. J. Li, D. W. Pang and Z. L. Zhang, Anal. Chem., 2013, 85, 2645–2651.
30. N. Sakmeche, S. Aeiyach, J. J. Aaron, M. Jouini, J. C. Lacroix and P. C. Lacaze, Langmuir, 1999, 15, 2566–2574.
31. S. S. Hong, J. Chen, J. G. Zhang, Y. C. Tao and L. Y. Liu, Chin. Chem. Lett., 2004, 15, 811–812.
32. S. M. Majedi, H. K. Lee and B. C. Kelly, Anal. Chem., 2012, 84, 6546–6552.
33. X. Shi, J. Wen, Y. Li, Y. Zheng, J. Zhou, X. Li and H. Z. Yu, ACS Appl. Mater. Interfaces, 2014, 6, 21788–21797.
34. S. He, Q. Wang, Y. Yu, Q. Shi, L. Zhang and Z. Chen, Biosens. Bioelectron., 2015, 68, 462–467.
35. A. Dey, A. Kaushik, S. K. Arya and S. Bhansali, J. Mater. Chem., 2012, 22, 14763–14772.
36. J. Das, I. Ivanov, L. Montermini, J. Rak, E. H. Sargent and S. O. Kelley, Nat. Chem., 2015, 7, 569–575.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra19169j

---