CuO nanothorn arrays on three-dimensional copper foam as an ultra-highly sensitive and efficient nonenzymatic glucose sensor

Abstract

A CuO nanothorns/Cu foam (NTs-CuO/Cu foam) was synthesized using a low-cost and facile method. The morphology and composition of the NTs-CuO/Cu foam were characterized using SEM, TEM and XRD. Copper foam as the current collector played a key role in the formation of the NTs-CuO/Cu foam. The CuO nanothorns were freely grown on copper foam, and can make contact with the underneath conductive copper foam directly. The NTs-CuO/Cu foam was used as an electrocatalyst for the detection of glucose in an electrochemical sensor. The CuO nanothorns/Cu foam electrode shows an extremely high sensitivity of 5.9843 mA mM⁻¹ cm⁻² and a low detection limit of 0.275 μM based on a signal to noise ratio of 3. Due to its excellently high sensitivity, stability and anti-interference ability, the NTs-CuO/Cu foam will be a promising material for constructing practical non-enzymatic glucose sensors.

---

Introduction

Diabetes has become one of the three major threats to human health. Thus, looking for a rapid and accurate method to detect glucose content is of importance. Glucose sensors can be divided into glucose oxidase based sensors and non-enzymatic glucose sensors. However, there exist some disadvantages for an enzymatic electrode such as the poor stability and complicated immobilization procedures. The pH, humidity and temperature have effects on the stability of the enzyme to some degree. Besides, the activities of the enzyme decline further due to the tedious immobilization procedures.¹ The performance of an electrocatalyst depends on the intrinsic properties of the electrode materials to a great extent. It is a pity that most of these electrode materials can adsorb intermediates, which causes instability and lowers activity in the course of glucose electrocatalytic oxidation.^2,3 Therefore, it is extremely urgent to seek appropriate electrode materials with high performance and enzyme-less glucose sensors.

Many researchers have focused on the investigation of metals and metal oxides. With high conductivities, large specific surface areas and electrocatalytic performances, noble metals (e.g. Au, Pt and Pd) and their alloys (Pt–Au, Pt–Pd and Au–Pd) have been widely used for non-enzymatic glucose detection.^4–11 Given that there are some drawbacks for noble metals such as the high cost, poor stability and low sensitivity, transition metals (Co, Ni and Fe) and their oxides (Cu₂O, NiO) have been developed to construct enzyme-free glucose sensors.^12–16 Owing to its appropriate redox potentials, good electrochemical activity and excellent stability, cupric oxide (CuO) is an important p-type semiconductor with a band gap of 1.2 eV and has been widely applied in sensors, catalysts, lithium ion batteries and solar cells.^17–21 Thus, CuO can be selected as an excellent candidate for an active electrode material. Particularly, gaining a uniform morphology of the CuO nanostructures has become very critical.

It is noted that two-dimensional (2D) nanostructured materials are increasingly being developed in the fields of electronics, optoelectronics, catalysis and biotechnology. They show a high surface-to-volume ratio, more active sites as well as having excellent thermal and chemical stabilities.^22–24 Despite these numerous applications, only a few studies on the growth of Cu(OH)₂ and CuO tubes arrays have been published. According to the literature, Cu(OH)₂ tubes arrays can be formed with a copper foil immersed in an aqueous solution containing NaOH and ((NH₄)₂S₂O₈).^25,26

In this study, we developed a low-cost and facile method to synthesize CuO nanothorns supported on Cu foam (denoted as NTs-CuO/Cu foam). Copper foam was not only used as the current collector but also as the substrate for the CuO formation. The CuO nanothorns were freely grown on the copper foam, and can contact with the underneath conductive copper foam directly. And the as-prepared material was used as an electrocatalyst for the detection of glucose. The NTs-CuO/Cu foam shows a higher sensitivity and lower detection limit compared with other CuO nanomaterials^27,28 due to its high surface-to-volume ratio and the high conductivity of the Cu foam as the current collector, which can facilitate the charge and mass transfer. Moreover, the Cu foam with its open framework provided a large amount of anchoring sites for the deposition of CuO NWs during the synthesis of the NTs-CuO/Cu foam. Due to its excellently high sensitivity, stability and anti-interference ability, the NTs-CuO/Cu foam will be a promising material for constructing practical non-enzymatic glucose sensors.

Experimental section

Reagents and apparatus

Glucose, uric acid (UA) and ascorbic acid (AA) were supplied by Sigma-Aldrich. Maltose, sucrose, lactose, KOH, NaOH and (NH₄)₂S₂O₈ were purchased from Beijing Chemical Co (Beijing, China). Human serum samples were provided by the School Infirmary of Northeast Normal University. All of these reagents were of analytical grade and used without further purification. Ultrapure water (18.2 MΩ cm) produced using a Milli-Q system was used as the solvent throughout this work.

Scanning electron microscopy (SEM) images were obtained on an XL-30 ESEM FEG scanning electron microscope with an energy dispersive X-ray (EDX) analyzer at an accelerating voltage of 20 kV. Transmission electron microscopy (TEM) characterization was implemented using a H-600 transmission electron microscope. XRD patterns were obtained on a D8 Advance diffractometer with an area detector operating under a voltage of 20 kV using Cu Kα radiation (λ = 1.54 Å). Cyclic voltammetric and amperometric experiments were performed with a CHI660A electrochemical workstation with a three-electrode system using the NTs-CuO/Cu foam as the working electrode, a platinum wire as the auxiliary and an Ag/AgCl electrode as the reference electrode. All experiments were carried out at room temperature.

Preparation and characterization of the NTs-CuO/Cu foam

Arrays of CuO nanothorns on copper foam were prepared via anodization of copper foam followed by thermal treatments.²⁹ The Cu foam (99.9%, Jilin X-metal material, China) was cut into 0.5 cm × 0.5 cm slices. The anodization was implemented by applying a constant current density of 20 mA cm⁻² for 100 s in a three-electrode setup in 2 M KOH aqueous solution. Copper foam, platinum foil and a saturated calomel electrode acted as the working electrode, counter electrode and reference electrode, respectively. Finally, a blue film appeared. The anodized copper foam was washed with distilled water and ethanol three times and dried in the air. The pretreatment Cu foam was placed into an alumina boat. After purging with N₂ for 0.5 h, the furnace was heated to 200 °C for 1 h under a nitrogen atmosphere. As a result, CuO nanothorns were formed on the copper foam. The formation mechanism of the NTs-CuO/Cu foam is shown in Scheme 1.

Scheme 1 Schematic illustration for preparation of the NTs-CuO/Cu-foam.

Live subject statement

The authors state that all experiments were performed in compliance with the relevant laws and institutional guidelines. The institutional committee of the Changchun Institute of Applied Chemistry has approved the experiments with live subjects. The authors also state that informed consent was obtained for any experimentation with human subjects and Changchun Institute of Applied Chemistry is committed to the protection and safety of human subjects involved in research.

Results and discussion

Preparation and characterization of the NTs-CuO/Cu foam

Fig. 1a–f display the morphologies of the Cu foam, NTs-Cu(OH)₂/Cu foam and of the NTs-CuO/Cu foam at different magnifications. It was clear that the copper foam with its three-dimensional open network framework (Fig. 1a and b) provided a high volumetric specific surface area and more active sites for electrolyte diffusion. For the NTs-Cu(OH)₂/Cu foam (Fig. 1c and d), the nanothorn-like Cu(OH)₂ had lengths of around 4 μm, and they aligned densely on the entire surface of the copper foam. For the NTs-CuO/Cu foam, the SEM and TEM images (Fig. 1e and f) show that the nanothorn-like CuO almost covered the whole Cu foam. The NTs-CuO/Cu foam was further characterized using EDX as shown in Fig. 2. It can be seen that peaks for Cu and O exist in the image, indicating the formation of the NTs-CuO/Cu foam.

Fig. 1 SEM images of Cu foam (a and b), NTs-Cu(OH)₂/Cu foam (c and d), and NTs-CuO/Cu foam (e and f). The inset in (f): corresponding low magnification TEM image.

Fig. 2 SEM image of NTs-CuO/Cu foam and corresponding EDX element mapping of Cu and O.

As shown in Fig. 3a, the Cu foam displayed an XRD pattern with three strong reflection peaks at 43.4°, 50.5° and 74.1°, which can be indexed to the (111), (220) and (220) planes of the Cu structure (JCPDS no. 04-0836). The XRD pattern of the NTs-Cu(OH)₂/Cu foam (Fig. 3b) is assigned to the orthorhombic Cu(OH)₂ phase (JCPDS no. 35-0505), confirming that the Cu(OH)₂ nanothorns have been successfully in situ grown on the copper foam. The XRD patterns of the NTs-CuO/Cu foam (Fig. 3c) with two strong reflection peaks at 35.4° and 38.9° can be indexed to the (210) and (211) planes of CuO compared with the standard file (JCPDS no. 44-0706), which demonstrated that the Cu(OH)₂ nanothorns had been converted to CuO nanothorns completely after the heating treatment.

Fig. 3 XRD patterns of Cu foam (a), NTs-Cu(OH)₂/Cu foam (b), and NTs-CuO/Cu-foam (c).

Electrochemical characteristics of the NTs-CuO/Cu foam electrodes

NTs-CuO/Cu foam, a platinum wire and a saturated calomel electrode were used as the working electrode, counter electrode and reference electrode, respectively. 10 μL of glucose was added to 4 mL of 0.1 M NaOH solution. The NTs-CuO/Cu foam based non-enzymatic electrode was characterized using cyclic voltammetry (CV) curves between the potentials of 0 V and 0.8 V in alkaline solution (0.1 M NaOH) at a scan rate of 50 mV s⁻¹ and the solution was saturated with N₂, as depicted in Fig. 4a. There was no obvious current peak from the naked Cu foam (curve a) or from the Cu(OH)₂ NTs/Cu foam (curve c) in the absence of glucose. However, a dramatic current peak was observed for the CuO NTs/Cu foam (curve e). With injecting 1 mM of glucose to the electrolyte, a weak oxidation peak appeared at ca. +0.55 V for the Cu foam (curve b) and an oxidation peak appeared at ca. +0.5 V for the Cu(OH)₂ NTs/Cu foam (curve d). For the CuO NTs/Cu foam (curve f), the oxidation peak was in the potential range from 0.3 V to 0.6 V. However, the CVs of the CuO NTs/Cu foam exhibited the larger and more obvious glucose oxidation peak current at 0.5 V than the naked Cu foam and Cu(OH)₂ NTs/Cu foam, implying the excellent catalytic properties of the NTs-CuO/Cu foam electrode toward glucose oxidation. Meanwhile, the typical amperometric responses of the naked Cu foam, Cu(OH)₂ NTs/Cu foam and NTs-CuO/Cu foam to glucose were measured using successive injections of 50 μM of glucose into 0.1 M NaOH solutions at 0.5 V. The amperometric responses of the NTs-CuO/Cu foam were obviously enhanced, indicating that the NTs-CuO/Cu foam electrode possessed excellent electrocatalytic performance for glucose oxidation.

Fig. 4 (a) CVs of Cu foam (a and b), NTs-Cu(OH)₂/Cu foam (c and d), NTs-CuO/Cu foam (e and f) in 0.1 M NaOH in the absence (a, c and e) and presence of 1 mM glucose (b, d and f) at a scan rate of 50 mV s⁻¹. (b) CVs of the CuO NTs/Cu foam in the absence and presence of glucose with different concentrations (from the bottom to top: 0–6 mM). Scan rate: 50 mV s⁻¹. (c) CV curves of the CuO NTs/Cu foam in 0.1 M N₂-saturated NaOH at different scan rates (from the top: 20–200 mV s⁻¹). (d) The plot of electrocatalytic current of glucose at 0.5 V versus scan rate.

Compared with the Cu(OH)₂ NTs/Cu foam and Cu foam, the potential of the oxidation peak was more negatively shifted for the CuO NTs/Cu foam. What’s more, the current density of the oxidation peak for the NTs CuO/Cu foam increased more obviously than those of the Cu(OH)₂ NTs/Cu foam and Cu foam. Based on the above advantages, we can infer that the CuO NTs/Cu foam showed a high catalytic performance as a non-enzymatic glucose sensor.

Fig. 4b depicts the CVs of the NTs CuO/Cu foam in 0.1 M NaOH solution in the absence and presence of glucose. With the different concentrations of glucose ranging from 0 to 6 mM, a remarkable change was seen in the CV curves. A sharp increase of the oxidation peak at +0.5 V was observed with increasing glucose concentrations, which was assigned to the irreversible glucose oxidation due to the conversion of Cu(II) to Cu(III).^30,31

Fig. 4c and d show the CVs of the NTs CuO/Cu foam in 0.1 M NaOH solution at different scan rates. The redox peak currents are linear with the square root of the scan rate in the range of 20–200 mV s⁻¹, revealing a diffusion-controlled electrochemical process.

Amperometric sensing of glucose

Many factors may influence the current response, such as the NaOH concentration and the applied potential. Alkaline electrolyte was necessary for the electrooxidation of the carbohydrates on the Cu substrates due to its electrocatalytic effect being mediated by the Cu(OH)₂/CuO(OH) redox couples according to previous reports.³² As depicted in Fig. S2,† the effect of NaOH concentration on the current response of the NTs CuO/Cu foam electrode to glucose was investigated, with different concentrations of NaOH ranging from 0.1 mM to 1 M. It is clear that the oxidative current increases with the increase of NaOH concentration and no obvious oxidative current was observed when the concentration of NaOH is less than 0.1 M. However, a high alkaline concentration was bad for the practical application of the NT-CuO/Cu foam. As a result, 0.1 M NaOH was selected as the supporting electrolyte.

The effect of the applied potential on the amperometric response of the NTs-CuO/Cu foam to 0.1 mM glucose is shown in Fig. 5a. The current increased when the applied potential ranged from 0.3 to 0.6 V. The maximum response was obtained at 0.5 V, and 0.5 V was an optimal working potential for the amperometric measurement.

Fig. 5 (a) Amperometric responses of the NTs-CuO/Cu foam at various potentials (vs. Ag/AgCl) in 0.1 M NaOH solution with successive additions of 50 μM glucose. (b) Typical amperometric response and (c) the corresponding calibration curve of the NTs-CuO/Cu foam at 0.5 V vs. Ag/AgCl with the successive addition of glucose from 0.5 μM to 500 μM in 0.1 M NaOH solution. Inset in (b): amperometric response towards 0.5 μM per μM glucose. (d) Amperometric response of the CuO NTs/Cu foam towards the addition of 1 mM of glucose and various 0.1 mM interfering species in the 0.1 M NaOH solution at 0.5 V vs. Ag/AgCl.

The typical amperometric responses of the NTs-CuO/Cu foam to glucose were determined using successive injections of glucose into 0.1 M NaOH solution at the optimum potential. The NTs-CuO/Cu foam exhibited a fast response to the change in glucose. After the addition of the glucose, oxidation current increased at once and achieved a steady-state current within 5 s.

A typical current–time plot with successive injections of glucose in 0.1 M NaOH solution was recorded, as displayed in Fig. 5b and c. The linear concentration was in the range of 0.5 μM to 2 mM with a correlation coefficient of 0.9974, a high sensitivity of 5.9843 mA mM⁻¹ cm⁻² and a low detection limit down to 0.275 μM at a signal-to-noise ratio of 3. The performance of the fabricated sensor in this work was compared with other CuO-based nonenzymatic sensors. As depicted in Table 1, it was apparent that the NTs-CuO/Cu foam sensor had a higher sensitivity and lower detection. The excellent performance can be attributed to the merits of copper foam such as its high surface-to-volume ratio and more active sites as well as its excellent thermal and chemical stabilities.

Table 1 Comparison of the analytical performance of our proposed NTs-CuO/Cu foam sensor with other CuO based non-enzymatic glucose sensors

Type of electrode	Sensitivity (μA mM^−1 cm^−2)	Detection limit (μM)	Linear range (mM)	Reference
a Nanoleaf-shaped copper oxide.b CuO nanowires.c Multiple-walled carbon nanotubes.d Hierarchical MFI zeolite with micro/meso pore structures.
CuO NLs^a	26.6	5	0.1–3	31
CuO nanoflowers	404.53	4	0.004–8	33
CuO nanospheres	709.52	1	0–2.55	34
CuO nanourchins	2682	1.52	0.1–3	27
CuO nanobelts	582.0	<1	0.01–7.3	28
CuO NWs^b	1886.3	0.05	0.002–3.56	35
CuONL/MWCNTs^c	664.3	5.7	0–0.9	36
CuxO/Cu	1620	49.0	—	37
CuO/TiO2	79.9	1	0–2.0	38
meso-MFI^d–xCuO	85.35	0.37	0.0005–1.84	39
CuO nanothorns	5984.26	0.276	0.0002–2	This work

---

Repeatability, reproducibility and stability of the sensor

The repeatability of the NTs-CuO/Cu foam was examined at a glucose concentration of 50 μM glucose, and the relative standard deviation (RSD) for six detections was 2.52%, showing an acceptable repeatability. The RSD of the current signals for the measurement of 50 μM glucose at six independently prepared sensors was 1.30%, which proved the excellent reproducibility of the sensor preparation. When the NTs-CuO/Cu foam was not in use, it was stored at 298 K. 99.7% of the initial response of the NTs-CuO/Cu foam remained after two weeks, indicating the good stability of the as-prepared sensor.

Interference tests

The selectivity of the NTs-CuO/Cu foam for glucose was also investigated since the co-existing electroactive species might have effects on the detection of glucose in real sample analysis. UA, AA, lactose, sucrose and maltose are the major interferents normally co-existing with glucose in human blood serums. Taking into account the fact that the concentration of glucose is at least 30 times that of the interferents in human blood serum, the interference test was carried out using the successive injection of 1.0 mM glucose and 0.10 mM interfering species. As seen in Fig. 5d, the amperometric responses of NTs-CuO/Cu to 1.0 mM glucose were hardly affected by the additions of 0.10 mM UA, AA, lactose, sucrose and maltose. Therefore, we can infer that the NTs-CuO/Cu can show high selectivity for glucose detection.

Human serum sample measurements

The resulting sensor demonstrated extremely high sensitivity and selectivity toward the determination of glucose. To evaluate the application prospects, the proposed NTs-CuO/Cu foam sensing system was applied to the determination of glucose in human blood serum. 10 μL of serum sample was added to 4 mL of 0.1 M NaOH solution, and the current response was recorded at 0.5 V. Recovery testing was carried out to demonstrate the validity of the proposed method. The obtained recoveries of the proposed method ranged from 95.79% to 97.96%. For comparison, the concentrations of glucose in these samples were also detected using a spectrophotometric method. The results tested using both methods showed a good agreement. These results indicated that the proposed NTs-CuO/Cu foam had an excellent accuracy for glucose sensing, and could be applied to the determination of glucose in human serum samples (Table 2).

Table 2 Determination of glucose in human serum samples using the NTs-CuO/Cu foam electrode

Samples	Determined values	Reference values	RSD (n = 3) (%)	Recovery (%)
1	5.29	5.40	3.79	97.96
2	10.25	10.70	5.02	95.79

---

Conclusions

In summary, we successfully synthesized CuO nanothorns on copper foam using a simple two-step method. Copper foam worked as not only the current collector but also as the substrate for the CuO formation. There exist some advantages such as the high surface-to-volume ratio and more active sites as well as the excellent thermal and chemical stabilities. The NTs-CuO/Cu foam was used to construct a non-enzymatic glucose sensor. The newly developed sensor displayed good catalytic activity for glucose oxidation with extremely high sensitivity, a low detection limit and a wide linearity range. In addition, the sensor also showed excellent stability and good selectivity to glucose detection. Therefore, the NTs-CuO/Cu foam will be a promising material for constructing practical non-enzymatic glucose sensors.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (21275135, 21405146).

Notes and references

1. P. Holt-Hindle, S. Nigro, M. Asmussen and A. Chen, Electrochem. Commun., 2008, 10, 1438.
2. Y. Li, Y. Y. Song, C. Yang and X. H. Xia, Electrochem. Commun., 2007, 9, 981.
3. X. Chen, H. Pan, H. Liu and M. Du, Electrochim. Acta, 2010, 56, 636.
4. C. Wang, L. Yin, L. Zhang and R. Gao, J. Phys. Chem. C, 2010, 114, 4408.
5. K. E. Toghill, L. Xiao, M. A. Phillips and R. G. Compton, Sens. Actuators, B, 2010, 147, 642.
6. W. Lu, X. Qin, A. M. Asiri, A. O. Al-Youbi and X. Sun, Analyst, 2013, 138, 417.
7. C. Guo, Y. Wang, Y. Zhao and C. Xu, Anal. Methods, 2013, 5, 1644.
8. C. W. Kung, Y. H. Cheng and K. C. Ho, Sens. Actuators, B, 2014, 204, 159.
9. Z. You, K. Shen, Z. Wu, X. Wang and X. Kong, Appl. Surf. Sci., 2012, 258, 8117.
10. Y. Li, J. Yao, Y. Zhu, Z. Zou and H. Wang, J. Power Sources, 2012, 203, 177.
11. A. Sun, J. Zheng and Q. Sheng, Electrochim. Acta, 2012, 65, 64.
12. K. Khun, Z. Ibupoto, X. Liu, V. Beni and M. Willander, Mater. Sci. Eng., B, 2015, 194, 94–100.
13. J. Wang, W. Bao and L. Zhang, Anal. Methods, 2012, 4, 4009.
14. X. Cao and N. Wang, Analyst, 2011, 136, 4241.
15. R. Khan, R. Ahmad, P. Rai, L. W. Jang, J. H. Yun, Y. T. Yu, Y. B. Hahn and I. H. Lee, Sens. Actuators, B, 2014, 203, 471.
16. Q. Liu, Y. Yang, H. Li, R. Zhu, Q. Shao, S. Yang and J. Xu, Biosens. Bioelectron., 2015, 64, 147.
17. J. Chen, K. Wang, L. Hartman and W. Zhou, J. Phys. Chem. C, 2008, 112, 16017.
18. X. Zhang, G. Wang, W. Zhang, N. Hu, H. Wu and B. Fang, J. Phys. Chem. C, 2008, 112, 8856.
19. Y. Feng and X. Zheng, Nano Lett., 2010, 10, 4762.
20. X. Gao, J. Bao, G. Pan, H. Zhu, P. Huang, F. Wu and D. Song, J. Phys. Chem. B, 2004, 108, 5547.
21. Y. Liu, L. Liao, J. Li and C. Pan, J. Phys. Chem. C, 2007, 111, 5050.
22. X. Zhuang, Y. Mai, D. Wu, F. Zhang and X. Feng, Adv. Mater., 2015, 27, 403.
23. Q. Zhang, Y. Wang, Z. W. Seh, Z. Fu, R. Zhang and Y. Cui, Nano Lett., 2015, 15, 721.
24. Y. Chen, C. Tan, H. Zhang and L. Wang, Chem. Soc. Rev., 2015, 44, 2681.
25. W. Zhang, X. Wen, S. Yang, Y. Berta and Z. L. Wang, Adv. Mater., 2003, 15, 822.
26. J. Xu, W. Zhang, Z. Yang and S. Yang, Inorg. Chem., 2008, 47, 699.
27. S. Sun, X. Zhang, Y. Sun, S. Yang, X. Song and Z. Yang, ACS Appl. Mater. Interfaces, 2013, 5, 4429.
28. T. Soejima, H. Yagyu, N. Kimizuka and S. Ito, J. Phys. Chem. B, 2006, 110, 1632.
29. Z. Zhuang, X. Su, H. Yuan, Q. Sun, D. Xiao and M. M. Choi, Analyst, 2008, 133, 126.
30. N. Q. Dung, D. Patil, H. Jung and D. Kim, Biosens. Bioelectron., 2013, 42, 280.
31. S. Liu, Z. Wang, F. Wang, B. Yu and T. Zhang, RSC Adv., 2014, 4, 33327.
32. P. F. Luo, F. Z. Zhang and R. P. Baldwin, Anal. Chim. Acta, 1991, 244, 169.
33. X. Wang, C. Hu, H. Liu, G. Du, X. He and Y. Xi, Sens. Actuators, B, 2010, 144, 220.
34. E. Reitz, W. Jia, M. Gentile, Y. Wang and Y. Lei, Electroanalysis, 2008, 20, 2482.
35. P. Ni, Y. Sun, Y. Shi, H. Dai, J. Hu, Y. Wang and Z. Li, RSC Adv., 2014, 4, 28842.
36. Z. Yang, J. Feng, J. Qiao, Y. Yan, Q. Yu and K. Sun, Anal. Methods, 2012, 4, 1924.
37. C. Li, Y. Su, S. Zhang, X. Lv, H. Xia and Y. Wang, Biosens. Bioelectron., 2010, 26, 903.
38. S. Luo, F. Su, C. Liu, J. Li, R. Liu, Y. Xiao, Y. Li, X. Liu and Q. Cai, Talanta, 2011, 86, 157.
39. J. Dong, T. Tian, L. Ren, Y. Zhang, J. Xu and X. Cheng, Colloids Surf., B, 2015, 125, 206.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra24579f

---