In Situ Oxidation of Cu2O Crystal for Electrochemical Detection of Glucose

The development of a sensitive, quick-responding, and robust glucose sensor is consistently pursued for use in numerous applications. Here, we propose a new method for preparing a Cu2O electrode for the electrochemical detection of glucose concentration. The Cu2O glucose electrode was prepared by in situ electrical oxidation in an alkaline solution, in which Cu2O nanoparticles were deposited on the electrode surface to form a thin film, followed by the growth of Cu(OH)2 nanorods or nanotubes. The morphology and electrocatalytic activity of a Cu2O glucose electrode can be tuned by the current density, reaction time, and NaOH concentration. The results from XRD, SEM, and a Raman spectrum show that the electrode surface was coated with cubic Cu2O nanoparticles with diameters ranging from 50 to 150 nm. The electrode exhibited a detection limit of 0.0275 mM, a peak sensitivity of 2524.9 μA·cm−2·mM−1, and a linear response range from 0.1 to 1 mM. The presence of high concentrations of ascorbic acid, uric acid, dopamine and lactose appeared to have no effects on the detection of glucose, indicating a high specificity and robustness of this electrode.


Introduction
Monitoring the concentration of glucose is essential for diabetes diagnosis [1], fermentation [2], chemical synthesis [3], and other industrial applications. In recent years, continuous effort has been made to develop electrochemical [4] and other methods including optical [5], acoustic [1], chemiluminescence [6], fluorescent nanogels [7], and near-infrared spectroscopy [8], for glucose detection. The electrochemical method is one with proven advantages in terms of accuracy, sensitivity, ease of operation, and portability. In 1962, Clark and Lyons proposed the concept of the glucose enzyme electrode and prepared the first glucose electrode [9]. In order to avoid the interference of background O 2 in the sample, electron acceptors such as ferrocene derivatives are used to specify the electron transfer in GOx recycling [4,10]. More recently, a virgin electrode without mediators was used to regenerate GOx (flavin adenine dinucleotide) directly on the surface of the electrodes [11][12][13]. While enzymatic sensors are routinely used in blood glucose detection, their applications to on-site and in situ detection, as often requested by research and engineering practice, are hindered by the limited stability of GOx and unsatisfactory sensitivity due mainly to deficient electron transfer. To overcome these weaknesses, non-enzymatic glucose sensors [14] that directly oxidize glucose at the surface electrode are pursued, and numerous materials have been developed, such as palladium nanoparticles coated functional carbon nanotubes [15], Ti/TiO 2 nanotube arrays/Ni composites [16], electrospun palladium (IV)-doped copper oxide composites in the form of nanofibers [17], nanoporous Au [18], and Cu nanoclusters/multi-walled carbon nanotubes (MWCNT) composites [19]. Compared with Pt and Au, Ni and Cu exhibit higher efficiency in the electrochemical detection of glucose because of the formation of Ni(II)/Ni(III)and Cu(II)/Cu(III) redox pairs [20,21]. CuO-based glucose sensors are also attractive due to their rich arability and low toxicity [22]. The oxidation of glucose on a CuO-based glucose electrode has not been fully elucidated [23,24]; however, it is suggested that Cu(II) forms Cu(III), which is capable of oxidizing glucose to gluconolactone and is reduced to Cu(II) after the glucose oxidation. The conventional method to fabricate a CuO glucose electrode consists of several steps, in which the first step is the hydrothermal synthesis of CuO nanoparticles, typically at approximately 100 • C [25,26]. The resulting CuO nanoparticles are then mixed with eutectic polymer binders, such as Nafion, Polyvinylidene fluoride (PVDF) and chitosan, forming an "ink". Finally, the ink is coated onto the virgin noble electrode to form a thin layer [25,27,28]. The sensitivity of such an electrode is affected by the presence of binders with no oxidation capabilities.
Although nanomaterials such as Carbon nanotubes (CNT) and graphene are blended with CuO nanoparticles [28] giving an improved conductivity, a direct contact of CuO with glucose and an accelerated electron transfer remains desirable for improved sensitivity and responsive speed. In situ fabrication methods have been explored in previous studies by chemical or electrochemical oxidation under different conditions [29][30][31][32]. Here, we present a new method for an in situ fabrication of a Cu 2 O-based glucose electrode, in which a Cu 2 O electrode was prepared by electrooxidation of metallic copper as an anode in alkaline solution, forming a Cu 2 O layer which consists of particles with diameters of 30-150 nm on the surface of the Cu electrode. We expect the circumvention of the "ink" not only simplifies the preparation process but also, and more importantly, allows direct contact of glucose with the electrode and facilitates the electron transport, thereby promising an improved sensitivity and responsiveness. The sensitivity and specificity of the detection, as well as the stability of the electrode, were examined. The effects of the preparation conditions on the morphology and electrocatalytic activity of a Cu 2 O glucose electrode were investigated to understand the fabrication mechanism of high catalytic activity Cu 2 O nonenzymic glucose sensors. The results reveal that the morphology and crystal phase are determined by the ratio of current density and NaOH concentration (J/c). As J/c increases, the morphology changes from Cu 2 O nanoparticles to Cu(OH) 2 nanotubes. Cu 2 O nanoparticles are deposited on the electrode surface, forming a thin film, followed by the growth of Cu(OH) 2 nanorods or nanotubes; among them, Cu 2 O nanoparticles exhibit higher activity. These results provide a rapid and efficient method, using mild conditions, for the in situ synthesis of a Cu 2 O electrode with good stability and repeatability for electrochemical detection of glucose.

Preparation of the Cu 2 O Electrode
The Cu 2 O electrode was synthesized via an in situ electrical oxidation method in alkaline solution. First, a disk Cu electrode with a diameter of 3 mm was polished using Al 2 O 3 powders of 300 nm and 50 nm in diameter, respectively. After washing in ultrasonic baths of ethanol and distilled water consecutively and drying in a N 2 atmosphere, the Cu electrode was implemented as the working electrode on CHI 852D (CH Instruments, Shanghai, China) with a 3-electrode system, in which the counter electrode was a Pt electrode and the reference electrode was Ag/AgCl, and oxidized for 300 s at a constant current density of 1.415 mA/cm 2 (0.1 mA for a φ3 disk electrode) in 3 M NaOH solution. The obtained Cu 2 O electrode was washed with distilled water and dried in a N 2 atmosphere.

Characterization Apparatus
A field-emission scanning electron microscope (FESEM) (Merlin, Carl Zeiss Jena, Germany) was used to observe the microstructures of Cu 2 O glucose electrode. The applied voltage was 15 kV. The crystalline phase of the structures was identified by X-ray diffraction (S2, Rigaku, Tokyo, Japan) with Cu Kα radiation. Raman spectroscopy was performed with a Raman microscopy system (LabRAM HR, HORIBA Jobin Yvon, Paris, France). A YAG laser served as an excitation source, and the applied wavelength was 532 nm.

Electrochemical Measurements
Electrochemical tests including cyclic voltammetry (CV) and Amperometric (i-t) measurements were conducted on CHI 852D (CH Instruments, Shanghai, China) with a 3-electrode system, in which the working electrode, the counter electrode, and the reference electrode were the Cu 2 O electrode synthesized as mentioned in Section 2.2, Pt electrode and Ag/AgCl (saturated) electrode, respectively.

Characterization of the Cu 2 O Glucose Electrode
The as-prepared Cu 2 O glucose electrode was characterized by XRD, SEM and a Raman spectrum, as shown in Figure 1.
working electrode on CHI 852D (CH Instruments, Shanghai, China) with a 3-electrode system, in which the counter electrode was a Pt electrode and the reference electrode was Ag/AgCl, and oxidized for 300 s at a constant current density of 1.415 mA/cm 2 (0.1 mA for a ϕ3 disk electrode) in 3 M NaOH solution. The obtained Cu2O electrode was washed with distilled water and dried in a N2 atmosphere.

Characterization Apparatus
A field-emission scanning electron microscope (FESEM) (Merlin, Carl Zeiss Jena, Germany) was used to observe the microstructures of Cu2O glucose electrode. The applied voltage was 15 kV. The crystalline phase of the structures was identified by X-ray diffraction (S2, Rigaku, Tokyo, Japan) with Cu Kα radiation. Raman spectroscopy was performed with a Raman microscopy system (LabRAM HR, HORIBA Jobin Yvon, Paris, France). A YAG laser served as an excitation source, and the applied wavelength was 532 nm.

Electrochemical Measurements
Electrochemical tests including cyclic voltammetry (CV) and Amperometric (i-t) measurements were conducted on CHI 852D (CH Instruments, Shanghai, China) with a 3-electrode system, in which the working electrode, the counter electrode, and the reference electrode were the Cu2O electrode synthesized as mentioned in Section 2.2, Pt electrode and Ag/AgCl (saturated) electrode, respectively.

Characterization of the Cu2O Glucose Electrode
The as-prepared Cu2O glucose electrode was characterized by XRD, SEM and a Raman spectrum, as shown in Figure 1.   [27]. Peak 4 and peak 5 are related to the transitions of Cu(III)/Cu(II) and Cu(II)/Cu(I), respectively [24]. In the presence of glucose, the CV profile also contains peak 1 and peak 2, while peak 3 disappears and is replaced by a shoulder peak around 0.5 V (vs. Ag/AgCl). The height of this peak is related to the glucose concentration.
cubic phase of Cu2O (JCPDS card no.77-0199, Fm3m, a0 = b0 = c0 = 4.258 Å). The copper metal substrate (JCPDS card no.03-1005, Fm3, a0 = b0 = c0 = 3.6077 Å) is also present. Figure 1b shows the Raman spectrum of the Cu2O glucose electrode. The characteristic peaks at 108.8, 148.7, 218.3, 415, 520 and 630 cm -1 demonstrate the formation of Cu2O on the Cu electrode. It can be seen from Figure 1c that cubic nanoparticles with diameters of 30-150 nm are dispersed evenly on the electrode surface. As can be seen from Figure 1d, only Cu and O are displayed in the energy-dispersive X-ray spectroscopy (EDS) analysis; it is concluded that the cubic phase crystals on the Cu electrode are Cu2O, as described elsewhere [23,33,34]. Figure 2a shows the cyclic voltammetry of the Cu2O electrode prepared in a 50 mM NaOH solution at a scan rate of 50 mV/s, in the absence or presence of glucose. Here, five distinctive redox peaks can be observed on the Cu2O electrode. Peak 1 is assigned to the transition of Cu(0)/Cu(I), while peak 2 includes two transitions of Cu(0)/Cu(II) and Cu(I)/Cu(II). Peak 3 is related to the transition of Cu(II)/Cu(III) in alkaline solution [27]. Peak 4 and peak 5 are related to the transitions of Cu(III)/Cu(II) and Cu(II)/Cu(I), respectively [24]. In the presence of glucose, the CV profile also contains peak 1 and peak 2, while peak 3 disappears and is replaced by a shoulder peak around 0.5 V (vs. Ag/AgCl). The height of this peak is related to the glucose concentration.

Electrochemical Behavior of the Cu2O Electrode
Marioli and Kuwana [35] and Chen et al. [36] have attributed the difference in the cyclic voltammograms obtained in the absence and presence of glucose to the transition of Cu(II)/Cu(III). First, glucose turns to an enol form upon isomerization in an alkaline medium, then, the active intermediate is oxidized to gluconolactone by Cu(III) followed by the hydrolysis of gluconolactone and forms gluconic acid. The above-mentioned process is shown in Figure 2b. Marioli and Kuwana [35] and Chen et al. [36] have attributed the difference in the cyclic voltammograms obtained in the absence and presence of glucose to the transition of Cu(II)/Cu(III). First, glucose turns to an enol form upon isomerization in an alkaline medium, then, the active intermediate is oxidized to gluconolactone by Cu(III) followed by the hydrolysis of gluconolactone and forms gluconic acid. The above-mentioned process is shown in Figure 2b. Cyclic voltammetry curves were performed at different scan rates of 20-90 mV/s in the presence of 3 mM glucose and 50 mM NaOH solution, as shown in Figure 2c,d. It is shown that the intensity of the shoulder peak near 0.5 V (vs. Ag/AgCl) increases as the scan rate increases. Moreover, a linear relationship between current intensities of the peak at 0.5 V and the square root of scan rates is observed (Figure 2d). According to the Randles-Sevcik equation, good fit linearity reflects a sufficiently rapid electron transfer reaction rate, indicating the diffusion-controlled progression.

Amperometric Detection of Glucose
We then determined the current density at different glucose concentrations in 50 mM NaOH solution and obtained the results shown in Figure 3. Cyclic voltammetry curves were performed at different scan rates of 20-90 mV/s in the presence of 3 mM glucose and 50 mM NaOH solution, as shown in Figure 2c,d. It is shown that the intensity of the shoulder peak near 0.5 V (vs. Ag/AgCl) increases as the scan rate increases. Moreover, a linear relationship between current intensities of the peak at 0.5 V and the square root of scan rates is observed (Figure 2d). According to the Randles-Sevcik equation, good fit linearity reflects a sufficiently rapid electron transfer reaction rate, indicating the diffusion-controlled progression.

Amperometric Detection of Glucose
We then determined the current density at different glucose concentrations in 50 mM NaOH solution and obtained the results shown in Figure 3.  The amperometric J-t curve in Figure 3a shows that the current density responds to the change of glucose concentrations quickly. The current becomes stable within 3 s after each injection of fresh glucose solution. Figure 3b shows an almost ideal linear response of the current intensity within the glucose concentration range from 0.1 to 1.0 mM. The working voltage is 0.5 V vs. Ag/AgCl, which is higher than the standard potential of the oxygen evolution reaction (0.204 V vs. Ag/AgCl). Therefore, small oxygen bubbles accumulated on the surface of the electrode at a very slow rate. The unreleased small bubbles result in increased noise. Figure 3b shows the linear dependency of the current response versus different glucose concentrations. The sensitivity of this Cu 2 O electrode is 2524.9 µA·cm −2 ·mM −1 , when glucose concentrations are in the range of 0.1 to 1 mM. The limit of detection (LOD) of the Cu 2 O electrode is 2.57 µM, which is calculated by substituting the standard deviation of the background signal with the mean standard deviation of the fitting results (S/N = 3).  Figure 3e,f, from which it is concluded that these interferents have little effect on the detection of glucose since their physiological concentrations are AA (30 µM) [37], DA (0.14 nM) [38], UA (300 µM) [39] and Lac (1.5 µM) [40], respectively. The prepared Cu 2 O electrode can detect glucose in the presence of other traditional contaminants in their physiological concentrations. Table 1 details the Cu 2 O electrodes and other metal-based electrodes utilized under alkaline conditions reported elsewhere, as well as their performance in terms of sensitivity, linear response range, detection limit, response time and cycles times. We profiled the data listed in Table 1, as shown in Figure S1. It can be concluded that the Cu 2 O electrode developed in the present study is superior in terms of both sensitivity and durability with regard to recycle times. The high sensitivity, being close to Cu 2 O/MoS 2 , we believe, is attributed to the circumvention of polymer binder, which leads to direct electron transport, as evidenced by the reduction of the overpotential [27]. The response time is a function of the thickness of the Cu 2 O layer and can be tuned by the current density and electro-oxidation time. This is advantageous for manipulating the structure of the electrode to meet the given requirement of sensitivity and response time. The electrode developed by the present study has the highest durability in terms of cycles times, which, we believe, benefits from the in situ generation of Cu 2 O nanostructures on the surface of the metallic copper electrode. Such high durability makes it promising for long-term detection.

Mechanism of In Situ Deposition of Cu 2 O
We also examined the effect of the fabrication conditions on the microstructure and electrochemical activity. The morphology and responsive behavior of the yielded Cu 2 O electrodes were determined, respectively. As shown in Figure 4a The current density and the NaOH concentration may also significantly affect the morphology of the Cu 2 O electrode. It is observed that the current intensity and the NaOH concentration significantly affect the morphology of the Cu 2 O electrode. Here, we establish a model to elucidate this interesting phenomenon. The current density, which is denoted as J, is equivalent to the rate of Cu + /Cu 2+ formation during the electro-oxidation reaction. The concentration of NaOH, which is denoted as c, is related to the rate of consuming Cu + and Cu 2+ . In the case of a low value of J/c, the Cu 2+ or Cu + , produced by electro-oxidation is immediately consumed to form nanoparticles due to excessive OH − at the surface of the electrode. This prevents further nucleation and results in small polyhedral nanoparticles depositing on the surface of electrode, as shown in Figure 4a (e.g., 5M-0.1, 5M-0.3, 5M-0.5, and 3M-0.1). The polyhedral nanoparticles are face-centered cubic Cu 2 O. When the value of I/c increases, the rate of Cu + /Cu 2+ formation gradually reaches that of Cu + /Cu 2+ consumption, resulting in the formation of incomplete nanotubes (akin to a nanosheet), as shown in Figure 4a (e.g., 3M-0.3-300). If we further increase the value I/c by increasing the current intensity or reducing the concentration of NaOH, the intact nanotubes composed of Cu(OH) 2 form on the surface of the Cu electrode (e.g., 3M-0.5-300), as shown in Figure 4a. According to the above discussion, the mechanism of forming different Cu 2 O nanostructures is illustrated in Figure 4d.
The oxidation time is vital for catalytic activities by determining the depth of the active material, which also relates to mass transfer resistance during the glucose oxidation reaction [61][62][63]. Figure 4c shows the relationship between sensitivity and charge consumption for each electrode. According to the similar catalytic activity at the optimum oxidation time and the fabricating mechanism shown in Figure 4d, the first step is the formation of the Cu 2 O film, followed by the reaction between Cu 2 O and OH − to produce Cu(OH) 2 nanorods or nanotubes. As a result, the electrodes prepared under different conditions contain a layer of Cu 2 O film. It could be inferred that cubic Cu 2 O exhibits a higher catalytic activity than orthorhombic Cu(OH) 2 and the peak sensitivity is 2524.9 µA·cm −2 ·mM −1 .

Mechanism of In Situ Deposition of Cu2O
We also examined the effect of the fabrication conditions on the microstructure and electrochemical activity. The morphology and responsive behavior of the yielded Cu2O electrodes were determined, respectively. As shown in Figure 4a, their fabrication conditions were used to label the electrodes. For instance, 1M-0.1-100 indicates that the Cu2O electrode is fabricated in a condition of 1.0 M NaOH, 0.1 mA current and 100 s electrooxidation time. The SEM images of the 3M-0.1-300, 3M-0.1-1200, 5M-0.1-300 and 5M-0.1-1200 electrode ( Figure S2) reveal that the electrooxidation time has little effect on the morphology of the electrodes. Therefore, we fixed the electro-oxidation time as 300 s for further discussion. To clarify, only the concentration of NaOH and the current density are labelled to indicate samples, as shown in Figure 4a.

Conclusions
In this paper, Cu 2 O electrodes for glucose detection were prepared in situ by an electrochemical oxidation method. The electrode exhibits a high sensitivity of 2524.9 µA·cm −2 ·mM −1 for glucose detection, with a linear response range of 0.1-1 mM and a detection limit of 0.0275 mM. Moreover, the electrode showed good stability and repeatability. The electrodes fabricated in situ under different conditions were characterized by SEM, XRD, and a Raman spectrum. The mechanism relating preparation condition, morphology and catalytic activity was clarified. Here, we describe a promising method and underlying mechanism to fabricate cubic Cu 2 O electrodes with reduced resistance and