One-step electrochemical fabrication of a nickel oxide nanoparticle/polyaniline nanowire/graphene oxide hybrid on a glassy carbon electrode for use as a non-enzymatic glucose biosensor

Abstract

We propose here a novel non-enzymatic glucose biosensor composed of a nickel oxide nanoparticle/polyaniline nanowire/graphene oxide hybrid composite on a glassy carbon electrode (NiONP/PANiNW/GO/GCE). The composite, prepared by mixing aniline with graphene oxide (GO) together, was transferred onto the surface of a bare glassy carbon electrode (GCE). This was then immersed in deoxygenated 50 mM NiCl₂ solution and electrodeposited at −0.8 V for 400 s to obtain the modified electrode, NiONP/PANiNW/GO/GCE. We characterized the morphology and electrochemical performance of the modified electrode using scanning electron microscopy (SEM) and cyclic voltammetry (CV), respectively. We found that the NiONP/PANiNW/GO/GCE exhibits higher electrocatalytic activity for glucose oxidation than a nickel oxide nanosheet/graphene oxide modified glassy carbon electrode (NiONS/GO/GCE) in alkaline solution. The sensitivity of the sensor towards glucose oxidation is 376.22 μA mM⁻¹ cm⁻² with a linearity range of 2 μM to 5.560 mM and a detection limit of 0.5 μM (S/N = 3). The sensor selectively detects glucose in the presence of common interfering species such as ascorbic acid, uric acid and dopamine. Furthermore, we examined the applicability of this modified electrode as a sensing probe for the detection of glucose concentration in fetal bovine serum. We conclude that the highly selective and sensitive NiONP/PANiNW/GO/GCE based non-enzymatic glucose sensor has the potential to be applied to the accurate measurement of glucose levels for various practical purposes such as clinical diagnosis and food analysis, etc.

---

1. Introduction

Accurate measurement of glucose levels is essential for various purposes, especially in clinical diagnosis,¹ chemical testing² and food analysis,³ etc. Meanwhile, diabetes, one of the most devastating diseases, is considered a serious threat to human health worldwide in the 21st century.⁴ Aside from diabetes, glucose levels are an important indicator of other health conditions such as trauma, stroke and other acute conditions requiring intensive care management.⁵ A simple, inexpensive, portable and easy-to-use method for regular monitoring of blood glucose will be a boon to diabetic patients. Although there are several glucose measuring kits, for example the commercially available glucometer based on glucose oxidase, maintaining their activity is a difficult task due to their enzymatic properties. Designing a glucose biosensor that exhibits high sensitivity, selectivity, accuracy and low sample-consumption is a priority. Glucose oxidase (GOx) based biosensors lack accuracy because of the instability of bioenzymes at varying temperatures.⁶ However, a non-enzymatic glucose biosensor may overcome such limitations and perform sensitively and selectively.⁷ In the past decade, many non-enzymatic glucose biosensors have been developed.^8–10

Non-enzymatic electrochemical glucose biosensors are always attractive alternatives to enzymatic biosensors, because of their stability.¹¹ Noble metal based materials, such as Pt,¹² Pd,¹³ Au and their metal alloys,¹⁴ have been explored as catalysts for non-enzymatic glucose detection. However, the high cost associated with noble metal based materials has limited their application. In recent years, several types of nanomaterials, e.g., CuO,¹⁵ MnO₂,¹⁶ NiO_x¹⁷ and CoO_x,¹⁸ have been extensively explored as sensing materials for developing non-enzymatic glucose biosensors. Among them, nickel oxide (NiO) nanostructure based materials have been widely used in non-enzymatic glucose sensors due to their excellent electro-catalytic properties and chemical stability.^19,20 Graphene–NiO nanoparticles (NiONPs),^21,22 graphene–Ni(OH)₂ ²³ and graphene oxide (GO)–electrospun NiO nanofiber²⁴ composites are known for their exceptional thermal, chemical and mechanical properties, high specific surface area and excellent conductivity. To improve the performance of electrochemical non-enzymatic glucose sensors, the above nanomaterials have been functionalized using conductive polymers, such as polyaniline (PANi), polypyrrole and polythiophene, etc.^25,26

PANi has previously been used as a typical conducting polymer in a glucose biosensor.²⁷ Owing to conjugation in the PANi molecule, its electrons possess a high degree of transferability.²⁸ By altering PANi's electronic structure by “mixing” it with other materials, one can improve its magnetic properties, optical properties, electrical conductivity and structural features remarkably.²⁹ Lee et al.³⁰ reported an electrochemical enzymatic glucose sensor prepared by modifying GOx with functionalized PANi and multiwalled carbon nanotubes. The resulting sensor exhibited high sensitivity (4.34 μA mM⁻¹), good reproducibility and a detection limit of 0.11 μM (S/N = 3).

This study reports an electrochemical non-enzymatic glucose sensor based on three-dimensional (3D) NiO nanoparticles (NiONPs) and PANi nanowire (PANiNW). We electrochemically fabricated a NiONP/PANiNW composite in one-step onto the surface of a GO-modified glassy carbon electrode (GCE), to develop a non-enzymatic glucose sensor (NiONP/PANiNW/GO/GCE). The modified electrode was characterized using scanning electron microscopy (SEM), elemental mapping, X-ray diffraction and elemental analysis. The new sensor possesses excellent selectivity and sensitivity with good results. Furthermore, the NiONP/PANiNW/GO/GCE demonstrated glucose detection in a real sample, by measuring glucose levels in fetal bovine serum.

2. Experimental

2.1 Reagents

Graphene oxide (GO) was purchased from Nanjing XFNano Materials Technology Company (Nanjing, China). Glucose, uric acid (UA), ascorbic acid (AA), dopamine (DA) and nickel(II) chloride hexahydrate were purchased from Sigma (Shanghai, China) and used as supplied. All other chemicals used were of analytical reagent grade, and deionized water (18.2 MΩ cm⁻¹) was used for all experiments.

2.2 Apparatus

Electrochemical measurements were performed using a computer-controlled electrochemical analyzer (CHI 832, Chenhua, Shanghai, China) composed of a conventional electrochemical cell containing one as-prepared glassy carbon (GC) electrode as the working electrode, one platinum wire as the counter electrode, and an Ag/AgCl electrode (KCl-saturated) as the reference electrode. The morphologies of the various modified GC electrodes were directly examined using a field emission scanning electron microscope (SEM) (S-4800, Hitachi, Japan) with a special GC electrode (3 mm) for SEM. The X-ray diffraction (XRD) analysis was conducted on a Rigaku D/max 2500 X-ray diffractometer with Cu Kα radiation (λ = 1.54 Å) at a scan rate of 5° per min. All experiments were conducted at room temperature.

2.3 Preparation of the NiONP/PANiNW/GO/GCE

GO aqueous dispersion (5 mL, 1 mg mL⁻¹) was mixed with aniline (5 mL, 5 mM), and the mixture was ultrasonicated for 60 min, centrifuged for 10 min at 10 000 rpm (12 300 × g), and subsequently redispersed in water. Prior to modification, the bare GCE was polished first using emery paper and then with aqueous slurries of alumina powder (0.3 μm and 0.05 μm) on a polishing cloth, and finally rinsed with doubly distilled water in an ultrasonic bath for 10 min. Next, the GC electrode surface was casted with 5 μL aniline/GO hybrid and dried under ambient atmosphere to fabricate the aniline/GO/GCE electrode. The NiONP/PANiNW/GO/GCE was structured using an electrodeposition method: (i) the aniline/GO/GCE was first electrodeposited in a deoxygenized 50 mM NiCl₂ solution with 0.1 M KCl using a chronoamperometric technique at 0.8 V for 400 s, and the polymerization of the aniline was completed during the chronoamperometry process; (ii) the electrode was then rinsed with doubly distilled water for 30 seconds. To compare the performance of the NiONP/PANiNW/GO/GCE, NiONS/GO/GCE was used as the control, which was fabricated in the same way mentioned above.

2.4 Electrochemical measurements

The sensing strategy for glucose is illustrated in Scheme 1. Cyclic voltammetry (CV) was performed to characterize the different modified electrodes in 0.1 M NaOH solution containing different concentrations of glucose at a scan rate of 100 mV s⁻¹. The chronoamperometric responses of the NiONP/PANiNW/GO/GCE were recorded at +0.6 V by continuously injecting glucose solution into a 10 mL 0.1 M NaOH solution. The measurements were conducted under stirring at room temperature.

Scheme 1 Schematic illustration of the enhanced electrochemical detection strategy for glucose, based on NiONP/PANiNW/GO deposited onto a GCE.

3. Results and discussion

3.1 Characterization of the modified electrode

We examined the morphology of the electrodes using SEM. Fig. 1 presents the SEM images of the GO/GCE (A), NiONS/GO/GCE (B) and NiONP/PANiNW/GO/GCE (C and D). The surface of the GO/GCE presents a wrinkle structure, which is the typical morphology of GO. As shown in Fig. 1, the structures of the NiO nanomaterials deposited on the GO/GCE and aniline/GO/GCE surfaces are different. On the GO/GCE surface, a 2D structure of NiONS (with ∼200 nm width and ∼20 nm thickness) was deposited, while a 3D dendritic structure of PANiNW (25 ± 4 nm in width as shown in Fig. 1D, n = 10) with dense NiONPs was generated on the surface of the GO/GCE. We found that the 3D structure of the NiONP/PANiNW/GO/GCE is more homogeneous and small, because the presence of PANiNW in the hybrids made the NiONPs more regular in shape and more compact. Furthermore, the elemental mapping, as shown in Fig. S1,† describes the elemental distribution at the surface of the modified electrodes. Compared to the NiONS/GO/GCE, the Ni element is more homogeneous, while the N element exists at the surface of the NiONP/PANiNW/GO/GCE, implying that PANiNW was formed at the surface. Furthermore, the XRD pattern in Fig. S2† shows the PANI and NiO crystalline structure. A broad peak is observed at 2θ = 20.31° for PANI, while the peaks at 2θ = 43.85° and 79.49° correspond to the (2 0 0) and (2 2 2) planes of the cubic phase of NiO, respectively.³¹ According to the above results, the two components in the hybrids, PANiNW and NiONPs, exhibit a metal chelation interaction, which results in the stable existence of NiONPs within the PANiNW. In addition, it is worth noting that van der Waals forces between PANiNW and NiONPs are an important factor that makes it possible to manipulate the nanoparticles into a close-packed and ordered array with a much larger active surface area than that of the NiONS/GO/GCE. This makes the NiONP/PANiNW/GO/GCE more suitable for electrocatalysis.

Fig. 1 Typical SEM images of the GO/GCE (A), NiONS/GO/GCE (B) and NiONP/PANiNW/GO/GCE (C and D).

Fig. S3† and Table 1 present the elemental analysis of the hybrids deposited on graphene. The results imply the presence of N in the NiONP/PANiNW/GO/GCE hybrid (17.66%), suggesting an abundance of PANiNW on the surface of the modified electrode. Based on energy-dispersive X-ray spectroscopy (EDS) analysis, we found the atomic proportions of Ni in the NiONS/GO/GCE and NiONP/PANiNW/GO/GCE hybrids to be 6.02% and 1.51%, respectively. Evidently, both electrodes contain less Ni, with the NiONP/PANiNW/GO/GCE having the least, indicating that the nanoparticles deposited on the surface of the modified electrode act as Ni nanocatalysts.

Table 1 Elemental analysis of the modified electrode using EDS

Modified electrode	Element	Atomic%^a
a Atomic% was adjusted with the exclusion of substrate.
NiONS/GO/GCE	C	76.64
	O	17.33
	Ni	6.02
NiONP/PANiNW/GO/GCE	C	67.44
	N	17.66
	O	13.40
	Ni	1.51

---

3.2 Electrochemical behaviour

Fig. 2 illustrates the cyclic voltammograms (CVs) of the GO/GCE (a), PANI/GO/GCE (b), NiONS/GO/GCE (c) and NiONP/PANiNW/GO/GCE (d) in 0.1 M NaOH, respectively. As is evident from the figure, there are no oxidation or reduction peaks for the GO/GCE (curve a), or the PANI/GO/GCE in alkaline solution (curve b). However, the NiONS/GO/GCE (curve b) has a couple of reversible redox peaks: the peak currents are 198.33 μA and 58.48 μA, respectively. Curve d, corresponding to the NiONP/PANiNW/GO/GCE, shows a pair of well-defined redox peaks with the peak currents at 603.23 μA and 292.12 μA, which are several times that of the NiONS/GO/GCE. The electrochemically active surface area (ESA) can be calculated based on the Rendles–Sevcik formula:

i_p = 0.4463 n^1.5F^1.5ADv^0.5C/RT

n: the number of electrons transferred in the redox reaction; F: Faraday constant; A: ESA; D: diffusion coefficient; v: scan rate; C: concentration; R: perfect gas constant; T: temperature. The ESA was calculated as 0.131 and 0.378 cm² for the NiONS/GO/GCE and NiONP/PANiNW/GO/GCE, as compared to the geometric area of 0.0707 cm² (0.3 cm diameter). This result indicated that the NiONP/PANiNW/GO/GCE was more electrochemically active than the NiONS/GO/GCE under the same process of electrochemical deposition, because of the existence of PANi.

Fig. 2 CVs of the GO/GCE (a), PANI/GO/GCE (b), NiONS/GO/GCE (c) and NiONP/PANiNW/GO/GCE (d) in 0.1 M NaOH at a scan rate of 100 mV s⁻¹.

The electrochemical reaction mechanism of NiO in basic solution has been well studied. The redox peaks, shown in Fig. 2, are attributed to the transition between NiO and NiOOH:^32,33

NiO + OH⁻ ↔ NiOOH + e⁻ (1)

Therefore, we can infer that NiO can retain its redox properties even after immobilizing it on the GCE surface containing PANiNW.

3.3 Amperometric detection of glucose

To check whether the NiONP/PANiNW/GO/GCE electrochemically catalyzes the oxidation of glucose, we recorded CVs of the tested electrodes in the presence of 2 mM glucose in 0.1 M NaOH. As shown in Fig. 3A, no catalytic currents are observed for the GO/GC electrode (curve a) or the PANI/GO/GCE (curve b) upon the addition of 2 mM glucose in 0.1 M NaOH. However, catalytic characteristics are observed when the anodic current is increased and the cathodic current disappears for the NiONS/GO/GCE (curve c) and NiONP/PANiNW/GO/GCE (curve d). The increase in the anodic current mainly originates due to the doped nickel in the modified electrode during glucose oxidation. Importantly, we found that the catalytic current of glucose oxidation by the NiONP/PANiNW/GO/GCE is two times that of the NiONS/GO/GCE, indicating that the former possesses higher catalytic activity with lower nickel content.

Fig. 3 (A) CV of the GO/GCE (a), PANI/GO/GCE (b), NiONS/GO/GCE (c) and NiONP/PANiNW/GO/GCE (d) in 0.1 M NaOH containing 2 mM glucose at a scan rate of 100 mV s⁻¹. (B) CV of the NiONP/PANiNW/GO/GCE in 0.1 M NaOH containing glucose of varying concentrations at a scan rate of 100 mV s⁻¹: (a) 1 mM; (b) 2 mM; (c) 5 mM.

To investigate the applicability of the NiONP/PANiNW/GO/GCE as a sensing element in non-enzymatic glucose detection, we recorded the CVs of the NiONP/PANiNW/GO/GCE in 0.1 M NaOH containing glucose of varying concentrations (Fig. 3B). We observed that the oxidation peak current corresponding to the transformation from NiO to NiOOH increases significantly with an increase in glucose concentration, while the reduction peak current disappears accordingly. This may be caused by greater NiOOH consumption with increasing glucose concentration (eqn (2)), which in turn accelerates the oxidation of NiO to NiOOH (eqn (1)), suggesting excellent electrocatalytic activity of the hybrids for glucose oxidation.³²

NiOOH + glucose → NiO + glucolactone (2)

3.4 Sensitivity study

Next, we monitored the chronoamperometric responses of the NiONP/PANiNW/GO/GCE to successive additions of glucose to the stirred 0.1 M NaOH at an applied potential of +0.6 V (Fig. 4A). We found that the non-enzymatic glucose sensor responds quickly to the additions of glucose and the oxidation current reaches its steady-state within 5 s. The oxidation current increases with increasing concentration of glucose in the range from 2 μM to 960 μM, following the linear equation: y = 0.03596x + 3.676 (R² = 0.9936), and from 960 μM to 5.560 mM following the linear equation: y = 0.02553x + 13.39 (R² = 0.9939). The detection limit was calculated to be 0.5 μM (signal-to-noise ratio of 3). The sensitivity exhibited by our sensor to glucose oxidation is 376.22 μA mM⁻¹ cm⁻². The low value of the detection limit, compared to some previously reported sensors,^34,35 suggests the ability of the new sensor to detect low levels of glucose in samples, thus confirming the improved sensitivity of the NiONP/PANiNW/GO/GCE based non-enzymatic glucose sensor, along with its wide linear range of detection.

Fig. 4 (A) Chronoamperometric responses of the NiONP/PANiNW/GO/GCE to successive additions of glucose in 0.1 M NaOH at +0.6 V. Inset: amplified calibration plot corresponding to the lower concentration range. Calibration plots between the current and glucose concentration in the range from 2 μM to 960 μM (B), and from 960 μM to 5.560 mM (C).

3.5 Selectivity study

Interfering reagents often pose challenges in glucose detection. To examine the selectivity of the NiONP/PANiNW/GO/GCE, we subjected the sensor to glucose determination at +0.6 V in the presence of interfering reagents such as AA, DA and UA (Fig. 5). We found that the NiONP/PANiNW/GO/GCE produces negligible current signal ratios upon the addition of 0.1 mM interfering reagents in the test solution containing 0.1 mM glucose, suggesting the excellent selectivity (or anti-interference effect) of the NiONP/PANiNW/GO/GCE. The excellent selectivity can be attributed to high isoelectric point (10–11) of NiO. Due to the higher isoelectric point, the surfaces of the NiONPs remain negatively charged in NaOH solution, while the interfering reagents (UA, AA and DA) become negatively charged in NaOH solution due to proton loss.³⁶ Consequently, the negatively charged NiONP/PANiNW/GO surface strongly repels the negatively charged molecules, thus reducing the electrooxidation of interfering reagents and resulting in improved selectivity.

Fig. 5 Amperometric responses of the NiONP/PANiNW/GO/GCE to successive additions of 0.1 mM glucose, 0.1 mM UA, 0.1 mM AA, 0.1 mM DA and then 0.1 mM glucose at +0.6 V in 0.1 M NaOH.

Furthermore, we investigated other influences from common co-existing substances and found that most ions and common substances at high concentration cause negligible change: Na⁺, K⁺, Cl⁻, NO₃⁻, SO₄²⁻ (>300 fold); Ca²⁺, Zn²⁺, Mg²⁺ (150 fold). Therefore, the NiONP/PANiNW/GO/GCE exhibits enhanced selectivity for glucose detection.

3.6 Application to real samples

We investigated the applicability of the proposed method for real sample analysis by directly analyzing glucose levels in fetal bovine serum. The ability for the amperometric determination of glucose concentration in fetal bovine serum, based on the repeated responses (n = 5) of the diluted analytes, was measured. The samples were spiked with a specified concentration of glucose. Using the standard addition method, we measured the glucose concentration in pharmaceutical preparations, as well as the recovery rate of the spiked samples. Table 2 lists the results.

Table 2 Method recoveries for the determination of glucose in fetal bovine serum (n = 5)

Spiked (mM)	Found (mM)	Recovery (%)	RSD (%)
0.50	0.502	100.3	1.3
2.0	1.97	99.2	1.9
5.0	5.2	100.4	1.7

---

4. Conclusions

In summary, we have developed a novel electrochemical non-enzymatic glucose sensor based on a NiONP/PANiNW/GO/GCE probe, which demonstrates high sensitivity and improved selectivity. The new sensor exhibits selective detection of glucose in a linear concentration range of 2 μM to 5.560 mM at a tolerance limit of 0.5 μM. Furthermore, the NiONP/PANiNW/GO/GCE exhibits an enhanced ability to suppress the background current from common physiological interferents. To check the applicability of the proposed sensor, the NiONP/PANiNW/GO/GCE was utilized to detect glucose levels in fetal bovine serum, and the sensor performed satisfactorily. Based on our findings, we can conclude that the NiONP/PANiNW/GO/GCE based non-enzymatic glucose sensor is a promising analytical platform for glucose detection in real samples.

Acknowledgements

The authors thank greatly the support from the National Natural Science Foundation of China (21405132, 21405131, 21575122) and Young Scholars Fund of Yantai University (Grant HY13B19, HY13B21).

References

1. P. Martinkova and M. Pohanka, Anal. Lett., 2015, 48, 2509.
2. C. Nagesh and R. P. Paily, Micromachines, 2014, 5, 722.
3. D. Albanese, A. Sannini, F. Malvano, R. Pilloton and M. Di Matteo, Food Analytical Methods, 2014, 7, 1002.
4. D. Espes, M. Martinell, H. Liljeback and P. O. Carlsson, Curr. Diabetes Rep., 2015, 15, 1.
5. G. V. Bochicchio, K. M. Bochicchio, M. Joshi, O. Ilahi and T. M. Scalea, Ann. Surg., 2015, 252, 597.
6. S. Alwarappan, C. Liu, A. Kumar and C. Z. Li, J. Phys. Chem. C, 2010, 114, 12920.
7. C. Karuppiah, S. Palanisamy, S. M. Chen, V. Veeramani and P. Periakaruppan, Sens. Actuators, B, 2014, 196, 450.
8. L. Ozcan, Y. Sahin and H. Turk, Biosens. Bioelectron., 2008, 24, 512.
9. T. T. Baby and S. Ramaprabhu, J. Nanosci. Nanotechnol., 2011, 11, 4684.
10. A. Pandya, P. G. Sutariya and S. K. Menon, Analyst, 2013, 138, 2483.
11. S. Masoomi-Godarzi, A. A. Khodadadi, M. Vesali-Naseh and Y. Mortazavi, J. Electrochem. Soc., 2014, 161, B19.
12. G. Chang, H. H. Shu, Q. W. Huang, M. Oyama, K. Ji, X. Liu and Y. B. He, Electrochim. Acta, 2015, 157, 149.
13. J. S. Ye, C. W. Chen and C. L. Lee, Sens. Actuators, B, 2015, 208, 569.
14. X. H. Niu, M. B. Lan, C. Chen and H. L. Zhao, Talanta, 2012, 99, 1062.
15. F. Sun, R. H. Zhu, H. W. Jiang, H. L. Huang, P. Liu and Y. J. Zheng, Electroanalysis, 2015, 27, 1238.
16. Y. Wang, W. S. Bai, F. Nie and J. B. Zheng, Electroanalysis, 2015, 27, 2399.
17. S. Liu, B. Yu and T. Zhang, Electrochim. Acta, 2013, 102, 104.
18. C. Y. Guo, X. Zhang, H. H. Huo, C. L. Xu and X. Han, Analyst, 2013, 138, 6727.
19. X. Luo, Z. J. Zhang, Q. J. Wan, K. B. Wu and N. J. Yang, Electrochem. Commun., 2015, 61, 89.
20. H. Liu, X. L. Wu, B. Yang, Z. J. Li, L. C. Lei and X. W. Zhang, Electrochim. Acta, 2015, 174, 745.
21. R. M. Kottegoda, N. H. Idris, L. Lu, J. Z. Wang and H. K. Liu, Electrochim. Acta, 2011, 56, 5815.
22. B. Yuan, C. Xu, D. Deng, L. Liu, Q. Zhang, S. Ji, L. Pi, D. Zhang and Q. Huo, Electrochim. Acta, 2013, 104, 78.
23. P. Subramanian, J. Niedziolka-Jonsson, A. Lesniewski, Q. Wang, M. Li, R. Boukherroub and S. Szunerits, J. Mater. Chem. A, 2014, 2, 5525.
24. Y. Q. Zhang, Y. Z. Wang, J. B. Jia and J. G. Wang, Sens. Actuators, B, 2012, 171, 580.
25. S. Park, H. Boo and T. D. Chung, Anal. Chim. Acta, 2006, 566, 46.
26. G. Wang, X. He, L. Wang, A. Gu, Y. Huang, B. Fang, B. Geng and X. Zhang, Microchim. Acta, 2013, 180, 161.
27. W. W. Tang, L. Li and X. P. Zeng, Talanta, 2015, 131, 417.
28. O. A. Raitman, E. Katz, A. F. Buckmann and I. Willner, J. Am. Chem. Soc., 2002, 124, 6487.
29. A. G. Macdiarmid and A. J. Epstein, Synth. Met., 1995, 69, 85.
30. K. P. Lee, S. Komathi, N. J. Nam and A. I. Gopalan, Microchem. J., 2010, 95, 74.
31. B. H. Shambharkar and S. S. Umare, J. Appl. Polym. Sci., 2011, 122, 1905.
32. Y. Zhang, X. Xiao, Y. Sun, Y. Shi, H. Dai, P. Ni, J. Hu, Z. Li, Y. Song and L. Wang, Electroanalysis, 2013, 25, 959.
33. B. Yuan, C. Xu, D. Deng, Y. Xing, L. Liu, H. Pang and D. Zhang, Electrochim. Acta, 2013, 88, 708.
34. C. Li, Y. Liu, L. Li, Z. Du, S. J. Xu, M. Zhang, X. Yin and T. Wang, Talanta, 2008, 77, 455.
35. M. Shamsipur, M. Naja and M. M. Hosseini, Bioelectrochemistry, 2010, 77, 120.
36. M. Li, X. Bo, Z. Mu, Y. Zhang and L. Guo, Sens. Actuators, B, 2014, 192, 261.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra14970g

---