Hollow AuPt alloy nanoparticles as an enhanced immunosensing platform for detection of multiple analytes

Abstract

Novel hollow AuPt alloy nanoparticles (hAuPt NPs) with a rough surface are prepared by a simple one-step galvanic displacement reaction between Co nanoparticles and a mixture of HAuCl₄ and K₂PtCl₄ under mild conditions. The synthesized hAuPt NPs are decorated with graphene oxide reduced by thionine or thiol functionalized ferrocene, which are utilized as electrochemical immunosensing probes for simultaneous detection of multiple analytes in a signal run. Using carcinoembryonic antigen (CEA) and alpha-fetoprotein (AFP) as model analytes, this proposed immunosensing shows wide linear range and low detection limits to 0.011 and 0.015 ng mL⁻¹ for CEA and AFP, respectively. The proposed immunosensor for the detection of serum samples is consistent with ELISA. These results suggest that the hAuPt NPs can be used as an immunosensing platform for detection of multiple analytes.

---

1. Introduction

Electrochemical immunosensors are a high-efficiency screening tool for the detection of tumor markers due to their advantageous features, such as simplicity of construction, high sensitivity, low cost, and easy miniaturization.^1–4 In order to amplify electrochemical signals and thus realize high sensitivity, some promising approaches are to employ enzymes as amplified signal reporters.^5,6 Unfortunately, the performance of the enzyme-based sensors is limited by the immobilization techniques of the enzyme and the experimental conditions such as pH and temperature, which greatly affect the bioactivities of the enzymes.⁷ Moreover, the natural enzymes are not only expensive but also unstable at the electrode surface due to their intrinsic nature. Therefore, there is an urgent need to develop alternative enzymeless materials for biomarker detections.

It is well known that nanostructured platinum are usually used as catalysts for H₂O₂ detection with electrochemical methods.^8,9 It is because that the unfilled d state of Pt can facilitate O–H bond breaking and weaken the adsorption of oxygenated intermediates on the surface of Pt, which can catalyze the hydrolysis of H₂O₂ to produce the oxygen. The catalytic oxidation process is listed as follows:^10–12

Pt + H₂O₂ ↔ Pt(OH˙_ads)₂

Pt(OH˙_ads)₂ ↔ Pt + 2OH˙_ads

2OH˙_ads → O₂ + 2H⁺ + 2e⁻

However, the high d-band centre energy of Pt entity would have strong bonding with its adsorbed O and OH species and then reduced the whole oxygen reduction reaction kinetics.¹³ Combining Pt with other metal elements to form bimetallic nanostructures could be efficient to tune the catalytic activity of Pt. Recent reports have demonstrated that Pt–Au alloy nanoparticles have unique effects on catalysis. The nanoscale Au could lead to reducing binding energy between Pt and O through increasing the orbital overlap between neighboring atoms, down-shifting the d-band center,^14–16 by which oxygen reduction reaction kinetics was enhanced. Thus, Au–Pt alloy nanoparticles would have higher electrocatalytic activity than Pt toward H₂O₂ oxidation. This implies that the nanostructured Au–Pt alloy can be used as a low-cost and stable substitution for peroxidase oxidase.

Additionally, a number of previous works showed that the exceptional physical and chemical properties of bimetallic nanoparticles are strongly determined not only by the composition but also by the morphology. Hollow metallic structures can provide high surface areas and large pore volumes as well as high electrical conductivity, which are important factors for electrochemical applications.¹⁷ In general, hollow AuPt alloy nanoparticles (hAuPt NPs) can be prepared by coating the surface of colloidal particles (silica, polystyrene spheres, or Ag nanoparticles) with Au and Pt layer, followed by the selective removal or of the colloidal templates through calcinations, wet chemical etching, or galvanic displacement reaction.^18–21 Sastry group has demonstrated a transmetallation reaction involving hydrophobic silver nanoparticles and hydrophobized AuCl₄⁻ and PtCl₆²⁻ to obtain hydrophobic hAuPt NPs.²² Huang et al. prepared hAuPt NPs with TiO₂ as a template in the presence of citric acid, and the template was removed during the formation of hAuPt NPs.²³ It is noted that these methods are always involved in two-step seed-mediated growth method under a complicated process and are difficult to purify. Therefore, one-step synthesis of hollow structured nanoparticles with design multiple composition under mild conditions is a great challenge for the development of multifunctional materials.

Herein, hAuPt NPs were prepared by a one-step galvanic displacement between Co nanoparticles and a mixture of AuCl₄⁻ and PtCl₄²⁻ under mild conditions. The syntheses of such hAuPt NPs with rough surface guided by galvanic replacement reaction have never been reported. In order to further increase the versatility of the hAuPt NPs, the thionine (or thiol functionalized ferrocene) reduced graphene oxide (Thi-rGO or Fc-rGO) were combined with the hAuPt NPs to form hAuPt NPs decorated Thi-rGO nanocomposite (hAuPt-Thi-rGO) or hAuPt NPs decorated Fc-rGO nanocomposite (hAuPt-Fc-rGO). A unique nonenzymatic sandwich-type electrochemical immunosensing platform for multiple analytes detection was fabricated using hAuPt-Thi-rGO and hAuPt-Fc-rGO as electrochemical immunosensing probes. Enhanced sensitivity was achieved by employing the enzymeless properties of hAuPt NPs to catalyze H₂O₂, which would further increase the electrochemical signals. And the high conductivity of graphene can promote electron transfer between the solution and electrode. The Thi and Fc could be transduced to current signals determined by the concentration of antigens.

2. Experimental Section

Synthesis of hAuPt NPs

The hAuPt NPs were synthesized according to the literature with a little modification.²⁴ In a typical synthesis, CoCl₂·6H₂O (8.5 mg) and trisodium citrate dehydrate (20 mg) were added in 50 mL water accompanied by vigorous mechanical stirring under a nitrogen atmosphere at room temperature for 30 min. The color of the solution immediately changed from colorless into dark brown when 10 mL NaBH₄ (1 mg mL⁻¹) was added. After 10 min, 2 mL K₂PtCl₄ (2 mM) and 1.2 mL HAuCl₄ (2 mM) were synchronously dropwise added to the above solution. Subsequently, the solution was centrifuged at 8000 rpm for 15 min and dissolved in 10 mL water.

3. Results and discussion

To prepare hAuPt NPs, a facile one-step galvanic displacement between Co nanoparticles and a mixture of HAuCl₄ and K₂PtCl₄ was conducted at room temperature. It can be seen in the image that the centre portions of the hAuPt NPs are lighter than their wall edge (Fig. 1A), demonstrating the formation of the hollow interior structures.²⁵ The insets of Fig. 1A showed the High-resolution transmission electron microscope (HRTEM) images of hAuPt NPs. The external surface of the nanosphere is irregular as shown in Fig. 1B, which could greatly increase the surface area and probably improve the catalytic performance. The hAuPt-Thi-rGO and hAuPt-Fc-rGO were fabricated by self-assembling the hAuPt NPs and Thi-rGO (or Fc-rGO) via electrostatic interaction. Morphologies of the hAuPt-Thi-rGO and hAuPt-Fc-rGO nanocomposites were characterized by Transmission electron microscopy (TEM) as shown in Fig. 1C and D. It was clearly observed that the hAuPt NPs were absorbed on the surface of rGO. The graphene provided a large surface area for the assembly of the hAuPt NPs at the top and bottom of the nanosheets. No aggregation of rGO sheet was observed. To further analyze the chemical composition of the hAuPt-Thi-rGO and hAuPt-Fc-rGO nanocomposites, X-ray photoelectron spectroscopy (XPS) measurements were performed. As shown in Fig. 1C and E and O elements can be found in GO,²⁶ and N element was arisen in the synthesized Thi-rGO.²⁷ The characteristic peaks for Pt 4f, Au 4f, C 1s, N 1s and O 1s core level regions could be obviously observed at the hAuPt-Thi-rGO nanocomposite. The N 1s core level was mainly derived from thionine and the Au 4f (83.9 eV and 87.6 eV)²⁸ and Pt 4f (71.2 eV and 74.5 eV)²⁹ doublet were consistent with Au⁰ and Pt⁰. Fig. 1F showed the Pt 4f, Au 4f, C 1s, N 1s, S 2p, O 1s and Fe 2p core level regions of the hAuPt-Fc-rGO nanocomposite. The Fe 2p core level region (710 eV and 727 eV) was originated from ferrocene.³⁰

Fig. 1 Typical TEM (A) and SEM (B) images of hAuPt NPs, TEM images of (C) hAuPt-Thi-rGO and (D) hAuPt-Fc-rGO, survey XPS spectra of (E) GO, Thi-rGO, and hAuPt-Thi-rGO and (F) GO, Fc-rGO, and hAuPt-Fc-rGO. Inset in Fig. 1A show the HRTEM images of hAuPt NPs.

Recently, the excellent catalytic ability of nanostructured Co₃O₄ towards glucose oxidation has been certified and exploited for nonenzymatic glucose detection.³¹ This implies the possibility using nanostructured Co₃O₄ as a low-cost and stable substitution for glucose oxidase. The principle of electrocatalytic glucose by Co₃O₄ was listed as follows:³²

2Co₃O₄ + H₂O ↔ 3Co₂O₃ + 2H⁺ + 2e⁻

Co₂O₃ + H₂O ↔ 2CoO₂ + 2H⁺ + 2e⁻

2CoO₂ + C₆H₁₂O₆ → C₆H₁₀O₆ + Co₂O₃ + H₂O

It is because that both Co_III and Co_IV have strong oxidative abilities while being reduced to Co_II and Co_III, respectively. The theoretical equilibrium potentials of Co_IV to Co_III conversion (φ′ = 0.909 V vs. SCE) and Co_III to Co_II conversion (φ′ = 0.48 V vs. SCE) are high, suggesting their strong ability to oxidize glucose. The morphology and structure of the Co₃O₄ nanocubes anchored graphene composite (Co₃O₄/graphene) were examined by TEM and Scanning electron microscopy (SEM) as shown in Fig. 2A and B, respectively. The Co₃O₄/graphene nanocomposite was fully and uniformly covered by the Co₃O₄ nanocubes, which are about 100 nm in length. The XRD pattern of the Co₃O₄/graphene nanocomposite appeared ten obvious diffraction peaks (Fig. 2C), which coincide with the (111), (220), (311), (222), (400), (422), (511), (440), and (531) planes in the standard Co₃O₄ spectrum (JCPDS 42-1467), and the diffraction peak (002) can be indexed into the graphene.³¹ Detailed compositional analysis of the Co₃O₄/graphene was further characterized by XPS. As shown in Fig. 2D, Co 2p, O 1s, and C 1s can be found in the Co₃O₄/graphene composite. The Co 2p and O 1s were derived from cobalt oxide. The C 1s (284.8 eV) peaks observed is related to graphitic carbon in graphene.³² The Co 2p_3/2 and Co 2p_1/2 peaks at 780.2 and 795.6 eV were consistent with the spin–orbit peaks of Co₃O₄, further confirming the formation of Co₃O₄ on the surface of graphene (inset in Fig. 2D).^33,34 These results indicate that Co₃O₄ nanoparticles and graphene were integrated successfully.

Fig. 2 Typical TEM (A) and SEM (B) images, XRD pattern (C), and XPS spectrum (D) of Co₃O₄/graphene composite. Inset: high-resolution Co 2p XPS spectrum of Co₃O₄/graphene composite.

Using electrochemical immunosensor for multiplexed targets detection, the most important two issues are designing new immunosensing substrate and new electrochemical probes which possess distinguishable electrochemical signals to improve the performance of the immunosensor. The schematic illustration of the stepwise fabrication process of the immunosensor was shown in Scheme 1. The hAuPt NPs with rough surface were synthesized in one-step with the guidance of galvanic displacement reaction, and a simple approach for assembling nanoparticles onto the Thi-rGO or Fc-rGO via electrostatic interaction was also successfully achieved. The hAuPt-Thi-rGO and hAuPt-Fc-rGO nanocomposites were used as electrochemical immunosensing probes. The hAuPt NPs can not only immobilize lots of antibodies, but also catalyze H₂O₂ to improve the electrochemical signal due to their large surface area and excellent catalytic capacities.^35,36 The Thi-rGO and Fc-rGO can provide electrochemical signal and promote the electron conductivity on the surface of electrode.^29,37 The chitosan coated Co₃O₄/graphene was utilized to build immunosensing substrate, because it integrates both the enzymeless properties of Co₃O₄, conductivity of graphene, film-forming ability of chitosan and large active sites to immobilizing antibody. As a result, high sensitivity and wide linear range of the proposed electrochemical immunosensor was obtained. With the sandwich-type assay format, the antibody–antigen immunocomplex was formed on the surface of the chitosan coated Co₃O₄/graphene. The electrochemical signals were come from the redox probes thionine and ferrocene, and the hAuPt NPs and Co₃O₄ nanoparticles were used to amplify the electrochemical signals.

Scheme 1 The fabrication process of the immunosensor.

In order to monitor the build process of the immunosensor, cyclic voltammetry measurements were performed in 0.1 M phosphate buffer solution (PBS) containing 5 mM [Fe(CN)₆]^4−/3− (pH 7.0) (Fig. S1A†). The bare GCE (curve a) displayed a pair of distinct redox peaks due to the oxidation and reduction of the redox couple Fe(CN)₆^4−/3−. The peak current of GCE obviously increased with the modification of chitosan coated Co₃O₄/graphene (curve b), suggesting that it can enlarge electrochemical response and enhance the sensitivity of the electrode. After the electrode was activated by 1% glutaraldehyde (curve c), immobilized with capture anti-CEA and anti-AFP (curve d), blocked with BSA (curve e) and incubated in a solution with 3 ng mL⁻¹ of carcinoembryonic antigen (CEA) and alpha-fetoprotein (AFP) (curve f), successively, the peak currents decreased step by step owing to the poor conductivity of proteins. In addition, the electrochemical impedance spectroscopy (EIS) measurements were also used to monitor the electrode modified procedure (Fig. S1B†). With chitosan coated Co₃O₄/graphene modified on the electrode surface, the electron transfer resistance (R_et) decreased (curve b, 73 Ω) than a bare GCE (curve a, 289 Ω). After the electrode was activated by 1% glutaraldehyde, an increase in R_et was observed (curve c, 124 Ω) and a further increase was noticed when the electrode was immobilized with antibodies (curve d, 351 Ω), blocked with BSA (curve e, 378 Ω), and incubated with 3 ng mL⁻¹ of CEA and AFP (curve f, 546 Ω). These results were consistent with the CV's, demonstrating that the electrode was successfully modified and can be used in the next steps. To monitor the signal amplified performance of the sensor (Fig. S1C†), 5 mM H₂O₂ and 5 mM glucose was added into the electrolyte. When the immunosensor was simultaneously incubated with anti-CEA-hAuPt-Thi-rGO and anti-AFP-hAuPt-Fc-rGO in PBS (pH = 7), two pairs of redox waves for Thi and Fc appeared (curve a). After the addition of 5 mM H₂O₂ and 5 mM glucose into the electrolyte, an obviously amplified current response was obtained (curve b). The result indicated that the signal was greatly amplified because of the excellent catalytic performance of hAuPt NPs and Co₃O₄ to H₂O₂ and glucose. The catalytic processes were listed as follows:

Thi_ox/Fc_ox + 2H⁺ + 2e⁻ → Thi_red + Fc_red

The effects of incubation time and pH value of electrolyte were shown in Fig. S2.† With increasing incubation time, the current responses were increased and then started to level off at 40 min (Fig. S2B†), showing a saturated binding in the immunoreaction. With increasing pH value of electrolyte from 5.0 to 7.0, the current response was increased first and then decreased at pH = 7.0 (Fig. S2A†). Hence, an incubation time of 40 min and pH = 7.0 of electrolyte was optimal and selected for the immunoassay.

To test the repeatability of the immunosensor, five freshly prepared modified electrodes were incubated with 10 ng mL⁻¹ CEA and AFP. All five electrodes exhibited similar current response behavior, results inter-assay relative standard deviation (RSD) is 3.7% for CEA and 3.9% for AFP, and intra-assay RSD is 2.2% for CEA and 2.4% for AFP. This certified that the repeatability of the immunosensor for CEA and AFP was acceptable. Additionally, the immunosensor retained 91% of its initial response after a storage period of two weeks (stored at 4 °C). This showed that the stability of the present immunosensor is good. In order to characterize the specificity of the immunosensor, 100 ng mL⁻¹ BSA, glucose, UA, IgG, or AA were mixed with 3 ng mL⁻¹ CEA and AFP, respectively. The current responses of 3 ng mL⁻¹ CEA and AFP with and without interferential substance were shown in Fig. S3.† The RSD was less than 5%, indicating that the present immunosensor possesses good selectivity for CEA and AFP. The cross-reactivity of the proposed electrochemical immunosensor was examined by the following two control tests: (i) single analyte, CEA or AFP; and (ii) simultaneously monitored, using the mixture of anti-CEA-hAuPt-Thi-rGO and anti-AFP-hAuPt-Fc-rGO as signal probes. CEA and AFP with 0.3 and 30 ng mL⁻¹ were assayed in tests (i) and (ii). When CEA or AFP was present, the corresponding current signal was obvious, in contrast, the corresponding responses were negligible (Table S1†), indicating that the simultaneous detection exhibited low interference with each other.

To evaluate the performance of the proposed immunosensor, different concentrations of CEA and AFP were detected. Under optimal conditions, SWV measurements were carried out in pH 7.0 PBS, containing 5 mM H₂O₂ and 5 mM glucose. The SWV currents of the multiplexed immunoassay increased as the logarithm concentrations of CEA and AFP increased in the linear detection range (LDR) from 0.03 to 100 ng mL⁻¹ (Fig. 3A) for both CEA (Fig. 3B) and AFP (Fig. 3C). The limits of detection (LOD) reached 0.011 ng mL⁻¹ for CEA and 0.015 ng mL⁻¹ for AFP at a signal-to-noise ratio of 3, and the sensitivities were 15.402 μA (log ng mL⁻¹)⁻¹ for CEA and 16.093 μA (log ng mL⁻¹)⁻¹ for AFP, respectively.

Fig. 3 SWV responses (A), and calibration curves for different concentration of CEA (B) and AFP (C) in PBS (pH 7.0), containing 5 mM H₂O₂ and 5 mM glucose.

Signal amplification is a crucial important factor for a successful electrochemical immunosensor. To verify the advantages of the hAuPt NPs, hollow Pt NPs (hPt NPs) and core–shell nanoparticles (Au@Pt NPs) instead of hAuPt NPs were used to fabricate the electrochemical immunosensing probes. The characteristics of the hPt NPs, hPt-Thi-rGO, hPt-Fc-rGO, Au@Pt NPs, Au@Pt-Thi-rGO, and Au@Pt-Fc-rGO were demonstrated in Fig. S4 and S5 in ESI.† The changes of currents using hAuPt-Thi-rGO and hAuPt-Fc-rGO as immunosensing probes were significantly higher than those of the other two signal probes (Fig. 4A and B). And the amperometric response change of the hAuPt NPs, hPt NPs, and Au@Pt NPs were monitored by chronoamperometry when 5 mM H₂O₂ added into 0.1 M PBS (pH = 7.0) (Fig. 4C), the hAuPt NPs exhibited greater current shift than hPt NPs and Au@Pt NPs. This may be due to hAuPt NPs would have higher electrocatalytic activity than Pt toward H₂O₂ oxidation. Furthermore, the hAuPt NPs provides a large surface area to volume ratio leading to increase of antibody loading and the immunoreaction probability, thus improved the sensitivity. However, in the XPS graph of the Au@Pt NPs core–shell structures (Fig. S5F†), only characteristic Pt can be observed, because the Au core was enwrapped compactly by Pt. Thus, the electrocatalytic activity of Au@Pt NPs is lower than hAuPt NPs. In order to verify the effect of Co₃O₄, the immunosensing substrate without Co₃O₄ was used to fabricate the electrochemical immunosensor. The proposed immunosensor using the Co₃O₄/graphene as immunosensing substrate exhibited higher sensitivity than the immunosensor using the substrate without Co₃O₄ (Fig. S6†). These results suggested that much greater signal amplification can be obtained when hAuPt NPs and Co₃O₄ used as the electrochemical immunosensing platform.

Fig. 4 Electrochemical responses of the multiplexed immunoassay with various signal probes, (a) hAuPt-Thi-rGO and hAuPt-Fc-rGO, (b) hPt-Thi-rGO and hPt-Fc-rGO, and (c) Au@Pt-Thi-rGO and Au@Pt-Fc-rGO toward various concentrations of (A) CEA and (B) AFP standards, using chitosan coated Co₃O₄/graphene as immunosensing substrate, (C) comparison of the electrocatalytic activities of (a) hAuPt NPs, (b) hPt NPs, (c) Au@Pt NPs by chronoamperometry.

Furthermore, the proposed immunosensor was validated by assaying five clinical serum samples. The obtained results were compared with those obtained by ELISA. Each human serum samples were analyzed for five times. The relative errors between the two methods were 2.82% to 5.31% and 2.45% to 3.62% for CEA and AFP (Table 1), respectively, indicating no signification differences between the results provided by the two methods. Therefore, the proposed immunosensor could have the potential application for simultaneous determination of CEA and AFP in clinical diagnostics.

Table 1 Assay results of clinical serum samples using the proposed and reference methods

Sample no.	Present method (ng mL^−1)		ELISA (ng mL^−1)		Relative (%)	Error
	CEA	AFP	CEA	AFP	CEA	AFP
1	1.94	1.12	1.88	1.09	3.19	2.75
2	1.39	1.25	1.35	1.22	2.96	2.45
3	26.18	1.43	24.86	1.38	5.31	3.62
4	0.73	0.76	0.71	0.74	2.82	2.70
5	2.46	1.65	2.37	1.60	3.80	3.13

---

4. Conclusions

In summary, novel hAuPt NPs with rough surface were prepared and were used to fabricate enhanced immunosensing probes for multiple analytes detection in electrochemical immunosensing. Highlights of this work can be summarized as follows: (1) the hAuPt NPs was fabricated in a one-step method under moderate conditions; (2) the sensitivity of the electrochemical immunosensing using hAuPt NPs as immunoprobes were much higher than those using hPt NPs and Au@Pt NPs due to its excellent catalytic properties and high surface-to-volume ratio; (3) such immunosensor used a single working-electrode for simultaneous multiple analytes determination in a single run. It can be foreseen that the hAuPt NPs could be widely used as an immunosensing platform for multiple targets.

Acknowledgements

This research was financed by grants from the National Natural Science Foundation of China (21273153), Beijing Natural Science Foundation (2132008), the Project of Construction of Innovative Teams and Teacher Career Development for Universities and Colleges Under Beijing Municipality (IDHT20140512) and the Project of the Construction of Scientific Research Base by the Beijing Municipal Education Commission.

Notes and references

1. H. L. Qi, C. Ling, Q. Y. Ma, Q. Gao and C. X. Zhang, Analyst, 2012, 137, 393.
2. D. Chen, H. B. Feng and J. H. Li, Chem. Rev., 2012, 112, 6027.
3. H. S. Yin, Z. N. Xu, Y. L. Zhou, M. Wang and S. Y. Ai, Analyst, 2013, 138, 1851.
4. Y. Liu, Y. Liu, H. B. Feng, Y. M. Wu, L. Joshi, X. Q. Zeng and J. H. Li, Biosens. Bioelectron., 2012, 35, 63.
5. M. J. Moehlenbrock and S. D. Minteer, Chem. Soc. Rev., 2008, 37, 1188.
6. M. J. Coonney, V. Svoboda, C. Lau, G. Martin and S. D. Minteer, Energy Environ. Sci., 2008, 1, 320.
7. M. Emami, M. Shamsipur, R. Saber and R. Irajirad, Analyst, 2014, 139, 2858.
8. J. H. Han, H. Boo, S. Park and T. D. Chung, Electrochim. Acta, 2006, 52, 1788.
9. A. L. Morais, J. R. C. Salgado, B. Sljukic, D. M. F. Santos and C. A. C. Sequeira, Int. J. Hydrogen Energy, 2012, 37, 14143.
10. J. Xuan, L. P. Jiang and J. J. Zhu, Biosens. Bioelectron., 2005, 21, 74.
11. I. Katsounaros, W. B. Schneider, J. C. Meier, U. Benedikt, P. U. Biedermann, A. Cuesta, A. A. Auer and K. J. J. Mayrhofer, Phys. Chem. Chem. Phys., 2013, 15, 8058.
12. S. H. Chang, M. H. Yeh, J. Rick, W. N. Su, D. G. Liu, J. F. Lee, C. C. Liu and B. J. Hwang, Sens. Actuators, B, 2014, 190, 55.
13. G. R. Zhang, D. Zhao, Y. Y. Feng, B. S. Zhang, D. S. Su, G. Liu and B. Q. Xu, ACS Nano, 2012, 6, 2226.
14. M. H. Shao, A. Peles, K. Shoemaker, M. Gummalla, P. N. Njoki, J. Luo and C. J. Zhong, J. Phys. Chem. Lett., 2011, 2, 67.
15. H. J. Li, H. X. Wu, Y. J. Zhai, X. L. Xu and Y. D. Jin, ACS Catal., 2013, 3, 2045.
16. Y. Yamauchi, A. Tonegawa, M. Komatsu, H. J. Wang, L. Wang, Y. Nemoto, N. Suzuki and K. Kuroda, J. Am. Chem. Soc., 2012, 134, 5100.
17. I. S. Park, K. S. Lee, J. H. Choi, H. Y. Park and Y. E. Sung, J. Phys. Chem. C, 2007, 111, 19126.
18. L. H. Lu, H. J. Zhang, G. Y. Sun, S. Q. Xi, H. S. Wang, X. L. Li, X. Wang and B. Zhao, Langmuir, 2003, 19, 9490.
19. V. G. Pol, H. Grisaru and A. Gedanken, Langmuir, 2005, 21, 3635.
20. J. Chai, F. H. Li, Y. W. Hu, Q. X. Zhang, D. X. Han and L. Niu, J. Mater. Chem., 2011, 21, 17922.
21. H. M. Song, D. H. Anjum, R. Sougrat, M. N. Hedhili and N. M. Khashab, J. Mater. Chem., 2012, 22, 25003.
22. P. R. Selvakannan and M. Sastry, Chem. Commun., 2005, 1684.
23. R. Huang, A. M. Zhu, Y. Gong, Q. G. Zhang and Q. L. Liu, Ind. Eng. Chem. Res., 2013, 52, 7432.
24. J. Chai, F. H. Li, Y. W. Hu, Q. X. Zhang, D. X. Han and L. Niu, J. Mater. Chem., 2011, 21, 17922.
25. H. P. Liang, Y. G. Guo, H. M. Zhang, J. S. Hu, L. J. Wan and C. L. Bai, Chem. Commun., 2004, 1496.
26. N. Liu and Z. F. Ma, Biosens. Bioelectron., 2014, 51, 184.
27. J. M. Han, J. Ma and Z. F. Ma, Biosens. Bioelectron., 2013, 47, 243.
28. Q. Gao, J. M. Han and Z. F. Ma, Biosens. Bioelectron., 2013, 49, 323.
29. X. L. Jia, X. Chen, J. M. Han, J. Ma and Z. F. Ma, Biosens. Bioelectron., 2014, 53, 65.
30. L. J. Bai, R. Yuan, Y. Q. Chai, Y. Zhuo, Y. L. Yuan and Y. Wang, Biomaterials, 2012, 33, 1090.
31. Z. S. Wu, W. C. Ren, L. Wen, L. B. Gao, J. P. Zhao, Z. P. Chen, G. M. Zhou, F. H. Li and M. Cheng, ACS Nano, 2010, 4, 3187.
32. A. M. Cao, J. S. Hu, H. P. Liang, W. G. Song, L. J. Wan, X. L. He, X. G. Gao and S. H. Xia, J. Phys. Chem. B, 2006, 110, 15858.
33. B. Varghese, C. H. Teo, Y. Zhu, M. V. Reddy, B. V. R. Chowdari, A. T. S. Wee, V. B. C. Tan, C. T. Lim and C. H. Sow, Adv. Funct. Mater., 2007, 17, 1932.
34. X. Chen, X. L. Jia, J. M. Han, J. Ma and Z. F. Ma, Biosens. Bioelectron., 2013, 50, 356.
35. J. Luo, P. N. Njoki, Y. Lin, L. Y. Wang and C. J. Zhong, Electrochem. Commun., 2006, 8, 581.
36. Y. B. Zhou, G. Yu, F. F. Chang, B. N. Hu and C. J. Zhong, Anal. Chim. Acta, 2012, 757, 56.
37. N. Liu, Z. M. Liu, H. L. Han and Z. F. Ma, J. Mater. Chem. B, 2014, 2, 3292.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra10262b

‡ These authors contributed equally to this work.

---