Characterization of the Cu-Ni thin film electrode
Wafer (diameter - 2.5 inch) scale Cu-Ni electrodes were prepared by RF sputtering thin film deposition methods (see experimental methods section). Fig. 1(a) shows the schematic illustration of a thin film deposition chamber depicting Cu and Ni targets powered on simultaneously to realize co-sputtered Cu-Ni thin film. Optical image of the Cu-Ni thin film deposited on 2.5-inch Si wafer is displayed in Fig. 1(b). The scanning electron microscopy (SEM) micrograph of the deposited Cu-Ni thin film Fig. 1(c) shows the morphology of the uniformly formed thin film with tightly packed nano grains. High magnification SEM micrograph further confirms the presence of tightly packed grains with varying sizes of 10 nm to 20 nm forming multiple grain boundaries Fig. 1(d). Such grain boundaries are provided highly electrochemical active sites for different electrochemical reactions such as CO2 electrochemical conversion35. The cross-section of Cu-Ni thin film displays the contrast to the substrate and confirms the grown homogeneous film of ~150 nm thickness. In addition, the energy-dispersive X-ray spectroscopy (EDS) confirms the presence of only Cu and Ni in the thin film Fig. 1(f). The signal for Si atoms is obtained due to substrate. The EDS results confirm the purity of the thin film as other atoms were not detected. A Cu/Ni atomic ratio of about 5.5/1 was observed from EDS analysis. The thickness of the films was also measured at different points of Si wafer, and these results further confirm the growth of homogeneous Cu-Ni thin film at wafer-scale as minor variation (~ 4%) in height measured at different points was observed Fig. 1(g).
Electrochemical activity of Cu-Ni electrode for glucose sensing:
Glucose sensing capabilities of Cu-Ni thin films were characterized by performing cyclic voltammetry (CV) experiments via sweeping potential between −0.5 V and 1.0 V with the scan rate of 20 mV/s in the presence and absence of glucose (Fig. 2(a)) in 0.1M NaOH (pH−13) electrolyte. All potentials are reported with respect to the Ag/AgCl (saturated KCl) electrodes. In the absence of glucose, an exceedingly small oxidation peak around ~ 0.5 V was observed. This peak may be attributed to the oxidation of Cu/Ni atoms during the electrochemical experiment in NaOH solution. The CV collected in the presence of 1 mM glucose concentration, resulted in significant oxidation current in the region of 0.5 - 0.7 V. This is directly associated with the glucose oxidation reaction12,24,36,37. To uncover, if the observed performance is associated with the individual Cu or Ni atoms or whether the co-sputtered Cu-Ni thin film exhibits a synergistic effect, we performed glucose-sensing experiments for Cu- and Ni-based electrodes under similar experimental conditions and collected CVs (Fig. S1 and S2). We compare the glucose oxidation current at 0.65 V, normally used as an operational potential for Cu and Ni electrodes, and 1.0 V (Fig. 2(b)). At 1.0 V, oxidation of adsorbed hydroxyl species occurs, therefore a slight change in oxidation current is observed after glucose addition to the Cu-Ni electrode. Interestingly, the Cu-Ni electrode shows more than one order of magnitude higher current density than the pristine Ni electrode at 0.65 V potentials (glucose oxidation reaction potential). We also observed significantly higher oxidation currents for Cu-Ni comparison to the Ni (5 times higher) and Cu (2.3 times higher) electrode at 1.0 V potential, respectively. Electrochemical impedance spectroscopy (EIS) (Fig. S7.) evidence that the charge transfer resistance Rct (26.15 Ω) and equivalent series resistance Re (6 Ω) of Cu-Ni thin film electrodes are significantly smaller than those of the individual bulk Cu (Rct ~47, Re ~12 Ω) and sputtered Ni (Rct ~213, Re ~50 Ω) thin film electrodes. Thus, fast charge transfers due to minimum Rct results in the superior performance Cu-Ni electrode among all studied electrodes. These results suggest that Cu-Ni co-sputtered thin film exhibits synergistic activity toward glucose, therefore resulting in superior performance. When Cu and Ni work synergistically for glucose sensing, multiple electrochemical reactions occur simultaneously37. Initially, Cu and Ni atoms are oxidized and then participate in glucose oxidation reactions as expressed in reaction 1 and 237.
NiO + CuO + 2OH− → NiO(OH) + CuO(OH) + 2e− (1)
NiO(OH) + CuO(OH) + 2Glucose → NiO + CuO + 2gluconolactone (2)
In addition, CV experiments were performed at various scan rates (5, 10, 20, 50, 150, and 200 mV s−1), and the magnitudes of anodic peaks were plotted for the square root of scan rates (Fig. 2(c) and (d)). The anode and cathode peaks gradually increase with respect to the change in the scan rates (Fig. 2(c)). A positive shift in anode peak and a negative one in cathode peak with respect to the increment in scan rate have also been observed. Here it should be noted that a relation between the square root of the scanning rate and magnitudes of anode/cathode peak provides insight into the reaction mechanism30,37–39. CV curves of Cu (Fig. S3) and Ni (Fig. S4) electrodes with varying scan rates are provided in supporting information for comparison. The linear regression equation of oxidation peak current is j (mA)= 0.421x + 1.243 (R2 – 0.997) where x stands for ν1/2 (Fig. 2(d)) indicating linear relation between the square root of scan rate and oxidation current, confirming that the glucose oxidation at Cu-Ni electrodes is a diffusion-controlled and fast reaction kinetics reaction process. The diffusion coefficient (D) for the transfer of glucose at the time of electrocatalysis is estimated using the Randles-Sevcik equation40 as shown below (Eq. 1)

where Ip is the peak current density in amperes, n is the electron transfer number (usually 1), A is the electrode surface geometrical area (in cm−2), C is the bulk concentration (in mol cm−3), and is the scan rate (in V s−1) and D is the diffusion coefficient. Diffusion coefficients was determined from the slopes of Ip vs v1/2, for Cu-Ni (Fig. 2(d)), Cu (Fig. S5) and Ni (Fig. S6) electrodes, respectively. The obtained diffusion coefficient values of Cu-Ni, Cu and Ni electrodes are 5.82, 0.05 and 0.001 cm2 s−1, respectively. This is further evidence that co-sputtered Cu-Ni thin film electrode favors higher glucose transfer resulting superior performance than individual Cu and Ni electrodes.
The amperometric measurements of Cu-Ni thin film electrode is displayed in Fig. 3(a). The results were obtained at 0.65 V potential in 0.1 M NaOH with increasing concentration of glucose (0.01–10 mM, with 0.01, 0.02, 0.05, 0.1, 0.2, 0.5, 1, 2, 5 and 10mM successive injections). The Cu-Ni thin film electrode exhibited a rapid response when glucose was introduced and the current saturated within 3–5 seconds. The inset of Fig. 3(a) displayed a magnified version of amperometric curve at lower concentrations which clearly shows a significant amount of current increase with the addition of 10 µM. The linear correlation between glucose concentration and the current was displayed in Fig. 3(b). The respective correlation coefficients (R2) were high, 0.9704, at a lower concentration. The amperometric measurements of Cu and Ni thin film electrodes are presented in Fig. S8 (a) and Fig. S9(a) and respective calibration plots presented in Fig. S8 (b) and Fig. S9(b). A wide linear range from 0.01 mM to 2.0 mM for glucose sensing behavior was observed for Cu-Ni thin film. In addition, the detection limit (LOD) of Cu-Ni thin film electrode was 1.86 μM which is 13 times lower than the Cu electrode and 50 times lower than Ni electrode and a high sensitivity 3517.2 μA mM−1 cm−2 which is 2.2 times greater than Cu and 19 times greater than Ni electrodes was achieved. This result shows that the proposed sensor has excellent sensitivity and very low detection limit. A composite material made of Ni-Cu/CNT/FTO39 for non-enzymatic glucose sensor displayed a linear range between 0.02 – 4.5 mM, a detection limit of 2 μM, and a sensitivity of 1836.5 μA mM−1 cm−2. Additionally, the electro-analytical performance of Cu-Ni electrode was compared with other Cu- and Ni-based catalysts tested in NaOH for glucose sensing as summarized in Table 1. Our Cu-Ni thin film electrode exhibit competitive performance with high sensitivity (3517.2 μA mM−1 cm−2) and a low detection limit (1.86 μM). This performance can be attributed to the synergistic activity of Cu and Ni atoms and the presence of a high density of grain boundaries (electrochemically active sites).
Table 1.Comparison of the glucose sensing characteristic of Cu-Ni thin film electrode with other Cu and Ni based electrodes.
Sample
|
Working potential (V)
|
Detection limit
|
Sensitivity (µA mM–1 cm–2)
|
Linear range
|
Ref.
|
Cu/G
|
0.4
|
2.47
|
145.52
|
0.01–0.2
|
41
|
Cu2O/GWs/CFP
|
0.55
|
0.21
|
|
0.0005–5
|
39
|
Cu2O/PtE
|
0.55
|
26
|
507
|
0.1–2.5
|
36
|
CuNPs/GP/GCE
|
0.5
|
0.2
|
607
|
0.005–1.4
|
42
|
Cu-CuO nanowire
|
0.3
|
500
|
NA
|
0.3 — 12
|
43
|
CuO nanorods-graphite
|
0.6
|
4
|
371.4
|
8
|
44
|
CuO nanospheres
|
0.6
|
1
|
404.5
|
2.6
|
45
|
CuO nanoplatelets
|
0.55
|
0.21
|
3490.7
|
3.5
|
46
|
Cu nanocluster- CNTs
|
0.65
|
0.2
|
251.4
|
2.5
|
47
|
NiO/GNS/GCE
|
0.5
|
0.2
|
|
4.2
|
48
|
Ni/NiO-Nafon-rGO/SPE
|
0.55
|
1.8
|
|
6.4
|
49
|
NiO nanosheet
|
|
2
|
3400
|
0.18
|
50
|
Ni-Cu/CNT/GCE
|
|
0.025
|
2633
|
0.025 - 800
|
51
|
Ni-Cu/CNT/FTO
|
|
2
|
1836.5
|
20 - 4500
|
38
|
Cu
|
0.65
|
23.4
|
1574.1
|
0.01 - 2.0
|
This work
|
Ni
|
0.65
|
91.67
|
182.6
|
0.01 - 2.0
|
This work
|
Cu-Ni
|
0.65
|
1.86
|
3517.2
|
0.01 - 2.0
|
This work
|
It is most important for a biosensor to discriminate against electroactive interfering species. In typical physiological sample, interfering species, such as ascorbic acid (AA), dopamine (DA), uric acid (UA) co-exists along with the glucose41. However, the glucose concentration (3-8 mM) is significantly higher in comparison to the concentration of AA (~23 µM), UA (0.13–0.46 mM) and DA (~0.1 mM) in normal conditions. Therefore, we performed CA experiments in the presence of 0.1 mM glucose and 0.02 mM of each interreference. The addition of glucose results in an immediate and significant increase in the glucose oxidation current. As shown in Fig. 3(c), the Cu-Ni thin film catalyst displays excellent selectivity for glucose oxidation; the corresponding oxidation current change is 270 μA.cm−2 upon adding 0.1 mM glucose, which greatly exceeds those recorded for the 0.02 mM concentration of interfering species, i.e., 33 μA.cm−2 for AA, 12 μA.cm−2 for DA and 45 μA.cm−2 for UA. The signal for glucose oxidation overshadowed the response for interreference despite high working potential resulting in higher current. Considering that in case of diabetic condition, glucose concentration in human blood samples is much higher than in the normal case. The obtained slight response for the interference will be negligible. These results suggest that Cu-Ni electrodes exhibit a reliable anti-interference characteristic and they can be used for practical biotechnology applications. For comparison, the effect of interferences on Cu (Fig. S10 (a)) and Ni thin film (Fig. S10 (b)) electrodes are presented in supplementary information.
Reproducibility and long-term stability of Cu-Ni electrodes
A non-enzymatic glucose sensor must show excellent reproducibility, repeatability, and long-term stability. The reproducibility of the Cu-Ni sensors was examined by performing CV for eight different electrodes in 0.1 mM Glucose 0.1M NaOH solution. We exposed the Cu-Ni electrode to continuous 100 CV cycles in the range of -0.5 to 1.0 V and in the presence of 0.1 mM glucose. Fig. 4(a) shows the recorded current at 0.65 V after 20 cycles alternatively. Initially, the glucose oxidation current increases after the first cycle and attains maximum value after 40 cycles, and afterward the current slightly decreases. This phenomenon can be attributed to the formation of new active sites as a result of oxidation/reduction process. After 100 cycles, the measured peak current density remains higher in comparison to the initial current recorded for the first cycle confirming the stability and reproducibility of the Cu-Ni electrodes. Moreover, electrochemical impedance spectroscopy (EIS) spectra were also recorded before and after 100 cycles (Fig. 4(b), Fig. S11). The similar charge transfer resistance (~ 26 Ω) with other circuit components confirmed the stability of the electrode. Seven different electrodes from the same wafer were tested to explore the reproducibility of electrodes (Fig. 4(c)). The results confirm that there is slight variation (~ 7%) from one sample to another suggesting homogeneity of the film and reproducibility of the electrodes. The repeatability was also examined by testing the glucose sensing performance of individual Cu-Ni electrode for 75 days. The recorded current and shape of the CVs remain identical confirming the excellent repeatability of the Cu-Ni sensors. While the sample was stored at room temperature, it shows excellent durability for the examined period (75 days) with only 10% drop in activity after this duration (Fig. 4(d)).