Highly Sensitive and Flexible Copper Oxide/Graphene Non‐Enzymatic Glucose Sensor by Laser Direct Writing

Accurate and convenient detection of human blood glucose levels is an effective method for early diagnosis of diabetes and prevention of complications. The flexible and wearable electrochemical glucose sensor with low cost, fast responsiveness, good stability, reliability, and high sensitivity has attracted much attention in monitoring glucose concentration. The preparation of a conductive layer with catalytic activity on a flexible substrate is the key to making a wearable glucose sensor. Here, graphene composite materials sintered with copper oxide (CuO) nanoparticles are successfully prepared on a polyimide film by laser direct writing method and fabricated a flexible non‐enzymatic glucose sensor using laser‐engraved graphene (LEG) as a conductive electrode. The CuO/LEG sensor exhibits a high sensitivity of 619.43 μA mm−1 cm−2 in 0–3 mm glucose and 462.96 μA mm−1 cm−2 in 0–8 mm glucose. In addition, the CuO/LEG sensor shows good reproducibility, high anti‐interference capability, and long‐term stability. It also presents good bending stability, which can maintain 82.40% initial current after 100 times bending. Moreover, the CuO/LEG sensor has an obvious step‐ampere response in the detection of sweat samples, indicating the great potential of wearable sweat sensors.


Introduction
According to the 2019 International Diabetes Federation (IDF) Global Diabetes Map, there are currently 463 million diabetic patients in the world. The vast majority of them are type 2 diabetes on CuO nanoparticles by colloidal method. Liu et al. [31] reported a non-enzymatic glucose sensor constructed by Cu nanowires synthesized by using one-pot hydrothermal method.
More and more studies have shown that the combination of Cu-based nanomaterials and other electroactive nanomaterials (such as metal nanoparticles, [28] graphene and its derivatives, [32] carbon nanotubes [33] ) can enhance the electrochemical catalytic performance of the sensor. The synergistic effect of graphene and Cu-based nanomaterials can expand the electroactive catalytic area and enhance the electron transfer ability, thereby improving the performance of electrochemical sensors. [32,34,35] Zahra Khosroshahi et al. [36] developed a non-enzymatic sensor based on 3D graphene foam decorated with Cu-xCu 2 O nanoparticles. Jiang et al. [21] reported a novel, stable, and sensitive non-enzymatic glucose sensor based on a Cu-nanoparticle-modified graphene edge nanoelectrode. Cheng and coauthors demonstrated a new strategy to fabricate the functional circuits on 3D freeform surfaces by intense pulsed light-induced mass transfer of zinc nanoparticles. [37] However, the combination of Cu-based nanomaterials and graphene is usually achieved by chemical methods. The chemical synthesis methods may require high-temperature treatment, hydrothermal reaction, or other complex technological processes. [38][39][40] In addition, nanocomposite materials usually need to be modified on hard substrates, such as glassy carbon electrodes, [41] silicon dioxide, [42] etc., which limits the application of sensors to wearable devices.
Compared with traditional non-enzymatic sensors, flexible non-enzymatic glucose sensors have broader application prospects in health monitoring. The preparation of a conductive layer with catalytic activity on flexible substrates is the key to fabricate a flexible glucose sensor. As an important technical means of flexible electronics, laser direct writing (LDW) [43] has been used for sintering, rapid patterning, [44] and synthesis of nanoparticles from ion precursors. [45] Zhang et al. [34] prepared a flexible nonenzymatic glucose sensor by depositing Cu nanoparticles (NPs) on graphene formed by LDW. Cheng and coauthors demonstrate the Au/Ni/LIGbased non-enzymatic glucose sensors with high sensitivity of 3500 A mm −1 cm −2 and wide linear range of 0-4 mm @ 0.5 V or 0-30 mm @ 0.1V. [46] LDW can be used to easily fabricate Cu and graphene nanocomposite materials on the flexible substrates. It can avoid complicated processes and harsh conditions required in the preparation of Cu or graphene by chemical methods. LDW can directly sinter Cu NPs on laser-engraved graphene (LEG), increasing the electron transfer ability of the Cu catalyst. The synthesis process does not require the use of binders or other additives. [36] This greatly saves the preparation time, reduces the preparation cost of the composite material, and increases the commercialization potential of the sensor.
In this article, we propose a LDW method for quick modification of CuO NPs on electrodes, and fabricate a flexible wearable sensor with a wide glucose detection range as well as high sensitivity. A Cu layer is first reduced by the laser, and then melted into particles and sintered on the resulting graphene electrode by a second irradiation of the laser, followed by the oxidation of the Cu particles into CuO. In this way, the CuO/LEG electrodes are generated by LDW method directly, which have no need for postprocedures and extra fabrication assistant. The obtained sensor shows good electrocatalytic activity for glucose. The glucose sensor shows two linear relationships between current density and glucose concentration, with the sensitivity of 619.43 A mm −1 cm −2 in ranging from 30 M to 3 mm and the sensitivity of 315.16 A mm −1 cm −2 in ranging from 3 to 8 mm. The CuO/LEG sensor has an obvious step-ampere response in the detection of sweat samples, indicating the great potential for sweat sensor. The sensor maintains good sensitivity over a wide detection range. At the same time, it also shows good selectivity, reproducibility, long-term stability, and bending stability.

Characterization of Cu/LEG, CuO 1 /LEG, CuO 5 /LEG, and CuO 20 /LEG Electrodes
The schematic diagram of the preparation of CuO/LEG electrode is shown in Figure 1a. Briefly, a Cu film is first reduced by the laser from a precursor solution, and then irradiated by the laser again with the focus adjusted to the Cu-polyimide (PI) interface, resulting in the melted Cu particles sintered on the resulting graphene electrode. Finally, the obtained Cu/LEG is oxidized to CuO/LEG electrode. The surface morphology of the Cu/LEG, CuO 1 /LEG, CuO 5 /LEG, and CuO 20 /LEG electrodes is characterized by scanning electron microscopy (SEM) (Figure 2a,b and Figure S1a-c, Supporting Information). The images show that Cu or CuO NPs are evenly distributed on the graphene sheet surface. This microstructure has a large specific surface area and provides abundant catalytic sites for the oxidation reaction of glucose.
The morphological and structural characteristics of electrodes are further investigated using transmission electron microscope (TEM). Figure 2c and Figure S1d-o, Supporting Information, shows TEM images of electrodes under four oxidation conditions and corresponding elemental mapping of Cu and O. They described the distribution of Cuand oxygen element before and after electrode oxidation. The elemental mapping of CuO 1 /LEG ( Figure S1i, Supporting Information) shows that the oxygen elements aggregate into a spherical shape. These results indicate that the Cu NPs may become CuO NPs after oxidation. Figure S2a, Supporting Information, illustrates the X-ray photoelectron spectroscopy (XPS) spectra of electrodes under four oxidation conditions. Compare to the Cu/LEG sample, the oxygen content in the oxidized sample is greatly increased. The oxidation reaction of Cu to CuO by ultraviolet (UV) ozone cleaning machine is further confirmed by XPS spectra of Cu 2p. The degree of oxidation reaction of Cu NPs is investigated by elemental line scans of nanoparticles with different oxidation conditions. The nanoparticles used for elemental line scans are marked with circles in the TEM images ( Figure S1d,g,j,m, Supporting Information). Figure 2e shows the O/Cu intensity ratio increases with the increase of oxidation time, which confirmed   that the oxidation time is proportional to the degree of oxidation within 20 min. Figure S2b, Supporting Information, shows the Raman spectra of Cu/LEG electrodes under four different oxidation conditions. It clearly shows three characteristic peaks, a D peak (1341.9 cm −1 ) due to disordered structure and defects of graphitic carbon, a G peak (1589.1 cm −1 ) generated by hybridization of graphite carbon atoms, and a week 2D peak (2681.5 cm −1 ) associated with second-order zone-boundary phonons. As shown in Figure 2f, the week peak located at 288 cm −1 is originated from CuO NPs. These results indicate that Cu-base nanoparticles and LEG do exist in the composite.

Optimal Oxidation Conditions of Electrode for Glucose Detection
The electrocatalytic properties of the LEG and CuO 1 /LEG electrodes in 0.1 m NaOH solution are investigated by cyclic voltammetry (CV) using a routine three-electrode configuration. The CV profiles of LEG and CuO 1 /LEG electrodes are measured separately in 0.1 m NaOH alkaline medium with the absence and presence of 2 mm glucose at a scan rate of 50 mV s −1 ( Figure  S3a, Supporting Information). There is no oxidation peak in the curves of the LEG and CuO 1 /LEG, indicating that no oxidationreduction reaction occurs in the absence of glucose. The LEG electrode shows a small current change with the addition of 2 mm glucose while the CuO 1 /LEG electrode shows a significant oxidation peak ranging from 0.35 to 0.65 V (vs Ag/AgCl). Compared with the LEG electrode, the anode current of CuO 1 /LEG electrode has a significant increase. The enhanced electrocatalytic performance can be attributed to the CuO NPs on the surface of the electrode, which can increase the contact area with the solution.
The amperometric response of electrodes with different oxidation conditions to glucose was explored by chronoamperometry. The Figure S3b, Supporting Information, exhibits the typical amperometric response of the CuO 1 /LEG electrode under different glucose concentration of 0.01 mm for 6 times, 0.02 mm for 7 times, 0.1 mm for 8 times, and 0.5 mm for 14 times at +0.50 V (vs Ag/AgCl), respectively. When glucose is added to the solution, the current of the sensor increases immediately and remains stable quickly, and finally a stepped curve is formed. Figure S3c, Supporting Information, shows the calibration curves of electrodes with various oxidation times obtained by the same test process. The sensors under different oxidation times exhibit a good linear range for 0-1.5 mm glucose, so it is very suitable for sweat detection, as the glucose concentration range of human sweat is 0-1.5 mm. [47] The sensitivity of various electrodes is shown in

Electrochemical Performance of the CuO 1 /LEG Electrode
Electrochemical impedance spectroscopy measurement is conducted in order to evaluate the charge transfer activity of the LEG and CuO 1 /LEG electrode. The Nyquist plots are shown in Figure 3b. All Nyquist plots consisted of a semicircle in the high frequency region and a straight line in the low frequency region. The diameter of the semicircle associates with the charge transfer resistance (R ct ), indicating the conducting capability. The R ct of LEG electrode and CuO 1 /LEG electrode are 93.8 and 70.4 Ω, respectively. The CuO 1 /LEG electrode has the smaller semicircle diameter, indicating the better electron transfer behavior of the CuO 1 /LEG electrode, which is beneficial for the glucose electrocatalysis. Figure 3c shows the CV curves of the LEG and CuO 1 /LEG electrode under a 0.1 m KCl solution containing 5 mm K 3 [Fe(CN) 6 ] and 5 mm K 4 [Fe(CN) 6 ], which indicates the increase in the charge transfer rate is due to the presence of CuO NPs. The cathodic peak current of ipc = 0.53 mA is increased to ipc = 1.41 mA, indicating that CuO NPs can improve the conductivity of the electrode. Figure 3d shows the response of the CuO 1 /LEG electrode to different concentrations of glucose. The oxidation current increases with the increase of the glucose concentration, which further confirms the catalytic effect of the CuO 1 /LEG electrode towards glucose. The CV curves of the CuO 1 /LEG electrode in 0.1 m NaOH solution with 2 mm glucose at different scan rates (25-225 mV s −1 ) are shown in Figure 3e. It can be found from the curves that the current density of the anode peak increases with the increase of the scan rate. Figure S4a, Supporting Information, exhibits a great linear dependence between the current density of anodic peaks and square root of the scan rate with a correlation coefficient of 0.995, revealing that the electro-catalytic process on the CuO 1 /LEG electrode is controlled by diffusion. [48] The mechanism for CuO based non-enzymatic glucose electrodes is mainly explained by the semiconductive properties of the CuO film and the adsorption of hydroxyl ions (Figure 1b). The increase anodic potential of CuO (p-type semiconductor) can promote an electronic cloud distortion on oxygen atom and lead to the adsorption of OH─. The energy accumulated about OH─ contributes to a great tendency for oxidation reactions. Thus, the CuO─OH first oxidizing glucose to gluconolactone, and then the gluconolactone is converted into gluconic acid by hydrolytic action. The catalytic mechanism can be described as follows: [49] CuO The electrochemical active surface area (ECSA) is used to probe the active sites on the surface of the electrode. [50] The value of the ECSA is difficult to calculate. Owing to the proportional relationship between ECSA and the double-layer capacitance, it can be evaluated by the double-layer capacitance (C dl ) obtained by the CV curves. [51,52] As shown in Figure 3f, the CV curves of CuO 1 /LEG electrode are tested at various scan rates (25-150 mV s −1 at 25 mV s −1 intervals) in the potential range from 0.0 to 0.2 V in 0.1 m NaOH solution, respectively. Figure S4b, Supporting Information, presents the capacitive current is proportional to the scan rate. It can be drawn from the figure that the C dl is 2.25 mF cm −2 , which is better than other electrodes. [52] These results indicate that the CuO 1 /LEG electrode has more catalytic active sites.

Amperometric Analysis of the CuO 1 /LEG Glucose Sensor
The applied potential is important to evaluate the amperometric response of the glucose sensor. The effect of applied potential to the CuO 1 /LEG glucose sensor is investigated by chronoamperometry. Figure 4a shows the amperometric response curve of CuO 1 /LEG glucose sensor (in 0.1 m NaOH with the continuous addition of 0.1 mm glucose) at different potentials of 0.45, 0.5, 0.55, 0.6, and 0.7 V, which exhibit the sensitivity of 1.61, 1.31, 1.58, 1.26, and 0.70 mA mm −1 cm −2 , respectively. The calibration curves of the CuO 1 /LEG glucose sensor at different potentials are show in Figure S5, Supporting Information. It can be found that the curve at 0.55 V displayed the maximum current response. Therefore, 0.55 V was chosen as the suitable working potential for detecting glucose in the subsequent study. At the same time, Figure 4b presents a fast and sensitive response (≈3 s) when the glucose concentration continuously increases, indicating the CuO 1 /LEG glucose sensor has a fast response time. Figure 4c shows the detailed amperometric response curve of the CuO 1 /LEG glucose sensor to the stepwise addition of glucose in ranging from 10 to 500 M in 0.1 m NaOH solution under the applied potential of 0.55 V. The inset of Figure 4c presents the current response at low glucose concentrations. The current density increases clearly while the concentration of glucose increases. Figure 4d shows the current density of the CuO 1 /LEG glucose sensor increased monotonically with glucose concentration increment. The corresponding calibration graphs are shown in Figure S6, Supporting Information. For glucose concentration ranging from 0 to 8 mm, the sensor exhibits a sensitivity of 462.96 A mm −1 cm −2 with R 2 = 0.98 (see Figure S6a, Supporting Information). It is noted that different detection ranges may allow sensors to be applied to different actual samples, such as sweat, blood, urine, and tears. For example, the range of 0-1.5 mm may be suitable for sweat sensing, the range of 0-0.8 mm may be suitable for tear sensing, and the range of 2.8-5 mm may be suitable for urine sensing. [53] Hence, considering potential sweat application, the calibration curve for glucose concentration of 0-3 mm is calculated and shown in Figure S6b, Supporting Information. The sensor exhibits a higher sensitivity of 619.43 A mm −1 cm −2 with R 2 = 0.99. In the process of glucose oxidation, the high glucose concentration might result in the aggregation of intermediates and hindered the diffusion of the oxidized products. [54,55] It provides strong evidence for the change in sensitivity of different concentration intervals. The limit of detection (LOD) was about 49 nM. LOD was calculated by 3 × the standard deviation of the background current divided by the slope of the calibration curve. The excellent LOD was obtained owing to abundant CuO NPs on the graphene surface and the synergistic effect of CuO NPs and graphene. The electrocatalytic performances of the CuO 1 /LEG glucose sensor compared with previous CuO based non-enzymatic glucose electrodes and the recently reported glucose sensors are in Tables S1 and S2, Supporting Information, respectively. The sensor we prepared exhibited good comprehensive performance in the aspects of linear range, sensitivity, LOD, and response time. Moreover, the combination of versatile fabrication and full electrode integration presents great potential to be applied in flexible wearable device.
Selectivity is one of the most important parameters for the evaluation of the performance of a sensor. Some electroactive species and other sugars might have jammed effect on the detection of glucose. [56] The selectivity of the CuO 1 /LEG glucose sensor is assessed by chronoamperometry at the potential of 0.55 V. Figure 5a shows the selectivity of the glucose sensor toward 0.5 mm glucose in the presence of 0.05 mm fructose, 0.05 mm maltose, 0.05 mm lactose, and 0.05 mm sorbitol. Likewise, Figure 5b shows the selectivity of the sensor toward 0.5 mm glucose with the existence of 0.05 mm L-ascorbic acid (AA), 0.05 mm Dopamine acid (DA), 0.05 mm Uric acid (UA), 0.05 mm sucrose, and 0.5 mm NaCl. Figure 5c shows that the addition of interfering species produces negligible current response ratio relative to glucose. Sorbitol (8.26%) has the strongest ability to interfere with glucose detection, and DA (0.79%) is the weakest. Considering that the concentration of interfering species is much smaller than that of glucose in actual measurement, these results confirm that the CuO 1 /LEG glucose sensor has good selectivity in the detection of glucose.
Reproducibility and stability are important performance indicators to assess the sensing ability of the glucose sensor in practical application. Figure 5d shows the reproducibility of the CuO 1 /LEG glucose sensor. The reproducibility of the CuO 1 /LEG glucose sensor is evaluated by measuring the amperometric re-sponse of 8 equally fabricated sensors toward 2 mm glucose in 0.1 m NaOH. The relative standard deviation (RSD) is 4.60%. The CuO 1 /LEG glucose sensor can be used in mass production owing to its excellent reproducibility. Figure 5d shows the stability of the CuO 1 /LEG glucose sensor. The stability of the CuO 1 /LEG glucose sensor is assessed by measuring the amperometric current response of the same sensor toward 2 mm glucose for 11 times in 0.1 m NaOH. The RSD is 4.47%, revealing a good repeatability. To measure the long-term stability of the glucose sensor, the CuO 1 /LEG glucose sensor is kept in air at room temperature condition over 42 days, and the amperometric experiment is carried out every week. As shown in Figure 5e, the current response of the glucose sensor still maintains 79.8% of the original value after 6 weeks. The above results imply that the CuO 1 /LEG glucose sensor is very promising for glucose detection.
As a flexible glucose sensor, the change in sensor performance after bending is very important. Figure 5f shows the normalized amperometric response of the CuO 1 /LEG glucose sensor affected by multiple bending cycles (curvature radius: 7.8 mm). It can be observed that the repeated bending can reduce the glucose detection ability of the sensor. The decrease in amperometric response may be due to the shedding of the graphene layer on the sensor surface during the bending process. The signal of the sensor shows a 30.3% loss of the initial signal after 500 cycles of bending. These results show the good anti-fatigue ability of the CuO 1 /LEG glucose sensor under continuative mechanical stress.

Determination of Glucose in Sweat Fluid
It is important that the glucose sensor can be applied in sweat environment without signal interference. [57] Figure 6a shows the amperometric response of the CuO 1 /LEG sensor to the successive addition of 0.05 mm glucose in sweat for 2 times, 0.05 mm glucose in 0.1 m NaOH for 2 times, 0.1 mm glucose in sweat for 2 times, and 0.1 mm glucose in 0.1 m NaOH for 2 times. Figure S7, Supporting Information, shows the corresponding calibration curve of the CuO 1 /LEG glucose sensor in sweat environ-ment. As can be seen from the Figure S7, Supporting Information, the current response was not obvious for the first four additions of glucose, which may be caused by the interference of substances in the real samples. [57] From the fourth addition of glucoses, the current increases with the increase of glucose concentration, indicating our sensor has the ability to response to glucose concentration in the sweat environment. Whether glucose is soluble in sweat or sodium hydroxide has little effect on the determination of glucose. It shows that the sensor can be used for sweat detection. Since the glucose sensor needs to react in an www.advancedsciencenews.com www.advsensorres.com alkaline environment, 2 L of 1 m NaOH is added dropwise to the sensing area and dried before the test. [58] At first, the background current for the sensor is collected. After the sweat is in full contact with the sensor, the amperometric response of the sensor is measured. As shown in Figure 6b, the sweat is collected from volunteer who did the 20 min cycling exercise. Then, the collected sweat is filtered and further purified. After that, 0.1 m sodium hydroxide is added to the sweat solution to get the pH = 12 solution environment, which is finally utilized for CuO 1 /LEG glucose sensor analyzing. The current increases significantly after exposure to sweat, indicating the potentiality of the CuO 1 /LEG glucose sensor to further develop into a wearable sensor.
All of the experiments for the human sweat detection by the CuO 1 /LEG glucose sensor are performed in compliance with the relevant laws and institutional guidelines. The experiments have been approved by the Ethics Committee at East China University of Science and Technology (Project number: 2017-10-01).

Conclusion
We have successfully fabricated a high-performance flexible CuO 1 /LEG glucose sensor on a polyimide film using LDW technology. LEG composite materials sintered with CuO particles exhibit high conductivity, electrocatalytic activity in glucose oxidation, and large specific surface area. The CuO 1 /LEG sensor shows two linear relationships between current density and concentration, with the higher sensitivity of 619.43 A mm −1 cm −2 in ranging from 30 M to 3 mm and the lower sensitivity of 315.16 A mm −1 cm −2 in ranging from 3 to 8 mm. In addition, it also exhibits excellent bending stability and anti-interference performance. The sensor has good application potential for the commercial production of high-performance flexible wearable glucose sensors.

Experimental Section
Experimental Material: Cu (II) nitrate trihydrate was obtained from Sinopharm Chemical Reagent Co., Ltd (Shanghai, China). D-Sorbitol, KCl, NaCl, NaOH were purchased from Shanghai Titan Technology Co., Ltd. Ethylene glycol, sucrose, glucose were purchased from Shanghai Zhonghe Chemical Technology Co., Ltd. Formic acid, lactosum, anhydrous, maltose, UA, AA were purchased from Shanghai Macklin Biochemical Technology Co., Ltd. DA was purchased from Shanghai Chuangsai Technology Co., Ltd. Deionized water (18.2 MΩ cm) was used for all solution preparations.
Preparation of the Cu Layer onto Polyimide Film: A Cu nitrate solution was prepared by adding 2.5 g of Cu (II) nitrate trihydrate into 2.5 g of ethylene glycol and stirring for 15 min. After that, the obtained solution was put into water at 100°C for 15 min, then the color of the solution changed from dark blue to dark green with a lot of bubbles generated. After cooling the solution to room temperature, the copper precursor solution was finally prepared by adding 20 L of formic acid into the above solution and sonicating for 5 min.
ThePI was cleaned by alcohol and put into UV ozone cleaning machine (Kunshan Sanwright New Energy Technology Co., Ltd.) for 10 min. It allowed active groups to be generated on the surface of the PI film to improve the surface hydrophilicity.
The Cu precursor solution was coated evenly onto a piece of PI film by a Spin coater, and the rotating speed was maintained at 150 rpm for 5 s. Then the PI film with coating was loaded into the laser engraving machine (Chengde Yineng Trading Co., Ltd.) with the wavelength of 445 nm for reducing Cu layer (8 mm × 8 mm). The laser focus was adjusted to the coating over the PI film. The laser power for reduction was 4200 mW and the scanning rate was 6000 mm min −1 . The reduced Cu layer and PI was washed by deionized water and dried by nitrogen flow.
Fabrication of the Glucose Sensor: The cleaned PI film with copper layer was put into the laser engraving machine. The laser focus was adjusted to the surface of PI film. Three LEG electrodes were engraved around the Cu layer. The laser power for reduction was 4200 mW and the scanning rate was 6000 mm min −1 . The three electrodes were the working electrode in the middle, the reference electrode on the right, and the counter electrode on the left, respectively. The Cu layer was ablated a second time to form Cu NPs on the LEG of the working electrode. In order to increase the sensing ability of the electrode, the PI film was placed in a preheated UV ozone cleaning machine to oxidize for 1, 2, 5, 10, and 20 min, respectively. For simplification, the oxidized electrode was called CuO 1 /LEG electrode, CuO 2 /LEG electrode, CuO 5 /LEG electrode, CuO 10 /LEG electrode, and CuO 20 /LEG electrode, respectively. Then conductive silver glue (Shenzhen sinew electronic material Co., Ltd.) was applied on the reference electrode. Finally, the surface of the silver paste coating was oxidized to silver chloride by adding 100 L of 0.5 m FeCl 3 solution for 30 s.
Materials Characterization: The microstructure, morphology, and composition of the as-prepared Cu/LEG and CuO/LEG sensors were characterized by SEM (Hitachi S-4800) and TEM (JEM-2100, JEOL). Raman spectra were recorded on Laser Micro-Raman Spectrometer invia reflex (using a 532 nm laser as excitation source). XPS measurements were done on a Thermo Scientific ESCALAB 250Xi.
Electrochemical Characterization: Electrochemical characterizations of the working electrode were performed on an electrochemical workstation (Chenhua CHI660E, China) in a standard three-electrode cell. The CuO-NPs/LEG sensor was used as the working electrode, with a platinum mesh counter electrode and Ag/AgCl reference electrode. Electrochemical characterizations of the obtained glucose sensor were also performed on an electrochemical workstation (Chenhua CHI660E, China). All the measurements were carried out in 0.1 m NaOH solution at room temperature. Amperometric measurements were performed by consecutive addition of known amount of glucose into the NaOH solution under constant stirring. The glucose concentration in the solution gradually increased from 0.01 to 8 mm. The catalytic performance of the sensor was evaluated with CV and chronoamperometry.
Statistical Analysis: The correlation between XPS intensity and feature was performed by XPS Peak Fit software after the calibration of C peak. Data presentation: The several sets of data were collected for each position, and the continuous variables were expressed as the mean. Software used for statistical analysis: All statistical analyses and graphing were performed using Origin.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.