Fabrication of Flexible Arrayed Lactate Biosensor Based on Immobilizing LDH-NAD+ on NiO Film Modified by GO and MBs

We proposed the flexible arrayed lactate biosensor based on immobilizing l-lactate dehydrogenase (LDH) and nicotinamide adenine dinucleotide (NAD+) on nickel oxide (NiO) film, and which the average sensitivity could be enhanced by using graphene oxide (GO) and magnetic beads (MBs). By using GO and MBs, it exhibits excellent sensitivity (45.397 mV/mM) with a linearity of 0.992 in a range of 0.2 mM to 3 mM. According to the results of electrochemical impedance spectroscopy (EIS), the electron transfer resistance of LDH-NAD+-MBs/GPTS/GO/NiO film was smaller than those of LDH-NAD+/GPTS/GO/NiO film and LDH-NAD+/GPTS/NiO film, and it presented the outstanding electron transfer ability. After that, the limit of detection, anti-interference effect and bending test were also investigated.


Introduction
Beyond glucose, blood lactate monitoring is also greatly important in critical care medicine [1]. In several biochemical processes that involved muscle movement, lactate plays an important role, and it is the key metabolite in the anaerobic glycolytic pathway [2,3]. When the energy in tissues is insufficient from aerobic respiration, an increase in lactate concentration will occur from the anaerobic metabolism [4]. The normal range of lactate concentration in human blood is from 0.5 mM to 1.5 mM [2]. The lactate monitoring technology has been developed and used in terms of clinical diagnostics, sports medicine, and food analysis [1][2][3][4][5].
L-lactate dehydrogenase (LDH) is an enzyme, which presents throughout the tissues such as blood cells and heart muscle, and it has a high catalytic activity for conversion of lactate with an aid of a coenzyme (nicotinamide adenine dinucleotide, NAD + ). Recently, LDH and NAD + have widely been used to catalyze reaction of lactate, and which are mainly on the purpose of lactate monitoring. Moreover, the catalytic reaction mechanism of LDH could be described by the following equations [2]: 1.
The silver paste, used as the conductive wires and reference electrodes, was printed on flexible PET substrate (30 mm × 40 mm) by using screen-printing technique, and was then baked at 120 • C for 30 min in the oven. The screen-printing technique could make traditional glass reference electrode replaced by silver reference electrodes, and it also could achieve the miniaturization of the device.

2.
NiO films were deposited on the ends of silver conductive wires by using radio frequency (R.F.) sputtering system. The flow rates of Ar and O 2 gases were controlled at 10 sccm and 3.8 sccm, respectively. The sputtering power and working pressure were set at 50 watts and 3 mTorr, respectively. The deposition time was 50 min. Furthermore, a pre-sputtering of target in pure Ar atmosphere was done for 10 min to avoid target contamination.

3.
Epoxy was printed on part of silver conductive wires by using screen-printing technique, and was then baked at 120 • C for 90 min in the oven. Epoxy was used as an insulation layer to prevent silver conductive wires from corroding from test solution; besides, it was also used to define the areas of sensing windows (1.77 mm 2 ). The uncovered silver conductive wires could be connected to potentiometric measurement system, which offered a favor to measurement signal transmission. The potential signals of six sensing windows and two reference electrodes could be obtained by potentiometric measurement system. 4.
GO powders were added into deionized (D.I.) water to prepare 0.1 wt %, 0.3 wt % and 0.5 wt % GO solutions, and which were uniformly dispersed by using an ultrasonic vibrator. After that, each 2 µL GO solution was dropped onto NiO film of flexible arrayed pH sensor, and it was left at room temperature for one day.

5.
Each 2 µL GPTS-toluene mixture (the volume ratio set at 1:2) was dropped on GO layer, and then it was left in the oven at 120 • C for one hour. 6.
Different amounts of MB solutions (10 mg/mL), such as 0.25 mL, 0.50 mL, 0.75 mL and 1.00 mL, were used to investigate the impact of MB contents on average sensitivities of the biosensors.
a. MB solution was drawn out and put into test tube by using micropipettor, and then the suspension in MB solution was drawn out with the aid of external magnetic field. b.
MBs were thrice cleaned with PBS. c.
MBs and EDC solution (10 mg/mL in PBS solution) were mixed and ultrasonicated for 30 min. EDC was used as a carboxyl activating agent to bind MB and enzyme. d.
EDC solution in mixture was drawn out with the aid of external magnetic field so that MBs were obtained.

7.
1 mL enzyme solution, composed of 1 mg LDH, 1 mg NAD + and 100 µL PBS, was added into test tube containing MBs. LDH-NAD + -MBs composite solution was stirred by the stirrer at 4 • C for 8 h. Each 2 µL LDH-NAD + -MBs composite solution was dropped on GPTS layer, and was then stored in refrigerator at 4 • C for one day. 8.
After that, the flexible arrayed lactate biosensor was completed, the schematic diagram of which was shown in Figure 1b. Moreover, Figure 2a,b showed the photos of flexible arrayed lactate biosensor under unfolded state and bended state.

Measurement
Except measuring the limit of detection, all of the measurements were operated between 0.2 mM and 3 mM lactate solutions because it was the best linear region and also fell within the normal range of lactate concentration in human blood. Therefore, the lactate biosensors were respectively immersed into 0.2, 0.7, 1.3, 2 and 3 mM lactate solutions, and then their response voltages were measured by using potentiometric measurement system. The potentiometric measurement system consisted of the instrumentation amplifiers (LT1167), a data acquisition (DAQ) card and a computer equipped with LabVIEW software. All of the measurement environments were kept at room temperature. All of the average sensitivity and linearity were calculated by calculation program of Origin 7.0.
The electrochemical impedances were analyzed by using EIS. The electrochemical instrument consisted of a potentiostat (SP-150, Bio-Logic Science Instruments, Seyssinet-Pariset, France), a computer with EC-Lab ® software and three-electrode setup. The three-electrode setup is made up of a working electrode (sensing film), a platinum (Pt) counter electrode and a commercial silver/silver chloride reference electrode, and the above electrodes are all contacted with electrolyte. A potentiostat with the aid of a computer can supply a stable AC sinusoidal signal by a commercial silver/silver chloride reference electrode, resulting in an electric field that can be generated between working electrode and counter electrode. All of the experiments for EIS used a small AC signal at frequency ranging from 10 kHz to 100 mHz. Moreover, the area and distance of film and Pt counter

Measurement
Except measuring the limit of detection, all of the measurements were operated between 0.2 mM and 3 mM lactate solutions because it was the best linear region and also fell within the normal range of lactate concentration in human blood. Therefore, the lactate biosensors were respectively immersed into 0.2, 0.7, 1.3, 2 and 3 mM lactate solutions, and then their response voltages were measured by using potentiometric measurement system. The potentiometric measurement system consisted of the instrumentation amplifiers (LT1167), a data acquisition (DAQ) card and a computer equipped with LabVIEW software. All of the measurement environments were kept at room temperature. All of the average sensitivity and linearity were calculated by calculation program of Origin 7.0.
The electrochemical impedances were analyzed by using EIS. The electrochemical instrument consisted of a potentiostat (SP-150, Bio-Logic Science Instruments, Seyssinet-Pariset, France), a computer with EC-Lab ® software and three-electrode setup. The three-electrode setup is made up of a working electrode (sensing film), a platinum (Pt) counter electrode and a commercial silver/silver chloride reference electrode, and the above electrodes are all contacted with electrolyte. A potentiostat with the aid of a computer can supply a stable AC sinusoidal signal by a commercial silver/silver chloride reference electrode, resulting in an electric field that can be generated between working electrode and counter electrode. All of the experiments for EIS used a small AC signal at frequency ranging from 10 kHz to 100 mHz. Moreover, the area and distance of film and Pt counter

Measurement
Except measuring the limit of detection, all of the measurements were operated between 0.2 mM and 3 mM lactate solutions because it was the best linear region and also fell within the normal range of lactate concentration in human blood. Therefore, the lactate biosensors were respectively immersed into 0.2, 0.7, 1.3, 2 and 3 mM lactate solutions, and then their response voltages were measured by using potentiometric measurement system. The potentiometric measurement system consisted of the instrumentation amplifiers (LT1167), a data acquisition (DAQ) card and a computer equipped with LabVIEW software. All of the measurement environments were kept at room temperature. All of the average sensitivity and linearity were calculated by calculation program of Origin 7.0.
The electrochemical impedances were analyzed by using EIS. The electrochemical instrument consisted of a potentiostat (SP-150, Bio-Logic Science Instruments, Seyssinet-Pariset, France), a computer with EC-Lab ® software and three-electrode setup. The three-electrode setup is made up of a working electrode (sensing film), a platinum (Pt) counter electrode and a commercial silver/silver chloride reference electrode, and the above electrodes are all contacted with electrolyte. A potentiostat with the aid of a computer can supply a stable AC sinusoidal signal by a commercial silver/silver chloride reference electrode, resulting in an electric field that can be generated between working electrode and counter electrode. All of the experiments for EIS used a small AC signal at frequency ranging from 10 kHz to 100 mHz. Moreover, the area and distance of film and Pt counter electrode were kept constant, and all of the measurement environments were kept at room temperature. In potentiometric measurement system, the potential signals of silver reference electrodes were sent to a computer equipped with LabVIEW through LT1167 and DAQ. Therefore, we could confirm the stability of reference potential. However, in EIS measurement, in order to confirm the stability of the electrochemical reaction, we used a commercial silver/silver chloride reference electrode instead of our silver reference electrodes.

Characterization of NiO Film
The surface morphology of NiO sensing film was characterized by using a field emission scanning electron microscope (FE-SEM, S4800-I, Hitachi, Tokyo, Japan) equipped with energy dispersive X-ray spectroscopy (EDX) detector with X-ray mapping capability in National Chung Cheng University, Taiwan. Figure 3 showed the surface morphology of NiO film, and it was dense and porous. Besides, from SEM and EDX results, we could know some information about NiO film. The thickness of NiO film was about 355 nm, and the elemental ratio of Ni/O was about 38.50/61.50 = 0.63. electrode were kept constant, and all of the measurement environments were kept at room temperature. In potentiometric measurement system, the potential signals of silver reference electrodes were sent to a computer equipped with LabVIEW through LT1167 and DAQ. Therefore, we could confirm the stability of reference potential. However, in EIS measurement, in order to confirm the stability of the electrochemical reaction, we used a commercial silver/silver chloride reference electrode instead of our silver reference electrodes.

Characterization of NiO Film
The surface morphology of NiO sensing film was characterized by using a field emission scanning electron microscope (FE-SEM, S4800-I, Hitachi, Tokyo, Japan) equipped with energy dispersive X-ray spectroscopy (EDX) detector with X-ray mapping capability in National Chung Cheng University, Taiwan. Figure 3 showed the surface morphology of NiO film, and it was dense and porous. Besides, from SEM and EDX results, we could know some information about NiO film. The thickness of NiO film was about 355 nm, and the elemental ratio of Ni/O was about 38.50/61.50 = 0.63. Besides, the performances of NiO film, including mobility, sheet hole concentration and resistivity, were also measured by Hall effect measurement system (HMS-3000), which came from the Precision Instrument Development Center, National Yunlin University of Science and Technology, Taiwan. The measurement parameters were set at a current of 1 mA, the thickness of 355 nm and a magnetic flux density (B) of 0.55 T. The mobility, hole concentration and resistivity of NiO thin film were 1.275 × 10 1 cm 2 /V•S, 1.782 × 10 19 cm −3 and 2.748 × 10 −2 Ω•cm, respectively.

Optimization of Flexible Arrayed Lactate Biosensor by Varying GO Content
In this study, GO was used to modify the surface of NiO film because of its high specific surface area, high electron transfer ability and high affinity for specific biomolecules [10][11][12]. Moreover, 0.1, 0.3 and 0.5 wt % GO solutions were used for the processing of film surface modification.
The response voltage was measured by potentiometric measurement system. The average sensitivity and linearity of flexible arrayed lactate biosensor based on LDH-NAD + /GPTS/NiO film were 38.218 mV/mM and 0.992, respectively, as shown in Table 1. Moreover, the experimental results of lactate biosensors based on LDH-NAD + /GPTS/GO/NiO film were shown in Table 1. The average sensitivity and linearity of flexible arrayed lactate biosensor modified by 0.1 wt % GO were 39.237 mV/mM and 0.998, respectively. The average sensitivity and linearity of flexible arrayed lactate biosensor modified by 0.3 wt % GO were 40.018 mV/mM and 0.995, respectively. However, the average sensitivity and linearity of flexible arrayed lactate biosensor modified by 0.5 wt % GO were 37.146 mV/mM and 0.994, respectively. From Table 1, we could know that the optimal GO Besides, the performances of NiO film, including mobility, sheet hole concentration and resistivity, were also measured by Hall effect measurement system (HMS-3000), which came from the Precision Instrument Development Center, National Yunlin University of Science and Technology, Taiwan. The measurement parameters were set at a current of 1 mA, the thickness of 355 nm and a magnetic flux density (B) of 0.55 T. The mobility, hole concentration and resistivity of NiO thin film were 1.275 × 10 1 cm 2 /V·S, 1.782 × 10 19 cm −3 and 2.748 × 10 −2 Ω·cm, respectively.

Optimization of Flexible Arrayed Lactate Biosensor by Varying GO Content
In this study, GO was used to modify the surface of NiO film because of its high specific surface area, high electron transfer ability and high affinity for specific biomolecules [10][11][12]. Moreover, 0.1, 0.3 and 0.5 wt % GO solutions were used for the processing of film surface modification.
The response voltage was measured by potentiometric measurement system. The average sensitivity and linearity of flexible arrayed lactate biosensor based on LDH-NAD + /GPTS/NiO film were 38.218 mV/mM and 0.992, respectively, as shown in Table 1. Moreover, the experimental results of lactate biosensors based on LDH-NAD + /GPTS/GO/NiO film were shown in Table 1. The average sensitivity and linearity of flexible arrayed lactate biosensor modified by 0.1 wt % GO were 39.237 mV/mM and 0.998, respectively. The average sensitivity and linearity of flexible arrayed lactate biosensor modified by 0.3 wt % GO were 40.018 mV/mM and 0.995, respectively. However, the average sensitivity and linearity of flexible arrayed lactate biosensor modified by 0.5 wt % GO were 37.146 mV/mM and 0.994, respectively. From Table 1, we could know that the optimal GO content was 0.3 wt % for flexible arrayed lactate biosensor, and which response voltages to different lactate concentrations were shown in Figure 4.  Figure 4.

Optimization of Flexible Arrayed Lactate Biosensor by Varying MB Content
After finding the optimal GO content, MBs were used to enhance enzyme-immobilization ability and catalytic ability because of their biocompatibility, high specific surface area, less toxicity, physicochemical stability, easy functionalization, magneto-controlled location and transport, and an excellent contact between biocatalyst and substrate [13][14][15]. Moreover, 0.25, 0.50, 0.75 and 1 mL MBs were respectively mixed into enzyme solution to deposit on the optimal GO layer.
The response voltage was measured by potentiometric measurement system. The average sensitivity and linearity of flexible arrayed lactate biosensor based on LDH-NAD + /GPTS/GO/NiO film were 40.018 mV/mM and 0.995, respectively, as shown in Table 1. On the other hand, the experimental results of lactate biosensors based on LDH-NAD + -MBs/GPTS/GO/NiO film were shown in Table 2. The average sensitivity and linearity of flexible arrayed lactate biosensor modified by 0.25 mL MBs were 42.531 mV/mM and 0.987, respectively. The average sensitivity and linearity of flexible arrayed lactate biosensor modified by 0.50 mL MBs were 43.595 mV/mM and 0.993, respectively. The average sensitivity and linearity of flexible arrayed lactate biosensor modified by 0.75 mL MBs were 45.397 mV/mM and 0.992, respectively. However, the average sensitivity and linearity of flexible arrayed lactate biosensor modified by 1.00 mL MBs were 43.606 mV/mM and 0.992, respectively. The average sensitivity was increased with increasing MB content but the increase had a limit. The average sensitivity was decreased when MB content was over 0.75 mL. From Table 2, we could know the optimal MB content was 0.75 mL, and which could achieve the highest average sensitivity (45.397 mV/mM), as shown in Figure 5.

Optimization of Flexible Arrayed Lactate Biosensor by Varying MB Content
After finding the optimal GO content, MBs were used to enhance enzyme-immobilization ability and catalytic ability because of their biocompatibility, high specific surface area, less toxicity, physicochemical stability, easy functionalization, magneto-controlled location and transport, and an excellent contact between biocatalyst and substrate [13][14][15]. Moreover, 0.25, 0.50, 0.75 and 1 mL MBs were respectively mixed into enzyme solution to deposit on the optimal GO layer.
The response voltage was measured by potentiometric measurement system. The average sensitivity and linearity of flexible arrayed lactate biosensor based on LDH-NAD + /GPTS/GO/NiO film were 40.018 mV/mM and 0.995, respectively, as shown in Table 1. On the other hand, the experimental results of lactate biosensors based on LDH-NAD + -MBs/GPTS/GO/NiO film were shown in Table 2. The average sensitivity and linearity of flexible arrayed lactate biosensor modified by 0.25 mL MBs were 42.531 mV/mM and 0.987, respectively. The average sensitivity and linearity of flexible arrayed lactate biosensor modified by 0.50 mL MBs were 43.595 mV/mM and 0.993, respectively. The average sensitivity and linearity of flexible arrayed lactate biosensor modified by 0.75 mL MBs were 45.397 mV/mM and 0.992, respectively. However, the average sensitivity and linearity of flexible arrayed lactate biosensor modified by 1.00 mL MBs were 43.606 mV/mM and 0.992, respectively. The average sensitivity was increased with increasing MB content but the increase had a limit. The average sensitivity was decreased when MB content was over 0.75 mL. From Table 2, we could know the optimal MB content was 0.75 mL, and which could achieve the highest average sensitivity (45.397 mV/mM), as shown in Figure 5.

Analysis of Electrochemical Impedances for Different Films
In this study, the electrochemical impedances of modified and unmodified films were analyzed to investigate their electron transfer abilities. The electrochemical impedance spectroscopy (EIS) is broadly employed to the descriptions of fundamental electrochemical and electronic processes [19]. The Nyquist diagram and its equivalent circuit for electrochemical biosensor are shown in Figure 6a,b. R et and R s are the electron transfer resistance and solution resistance, respectively. C dl is the double layer capacitance. Z d is the diffusion impedance. Figure 6a showed a curve that consists of many complex impedances as function of frequency. In Cartesian coordinates, the real part and imaginary part are plotted on the X and Y axes, respectively. The left semicircle in high frequency is attributed to the electron transfer process, and its diameter is R et . The right arc in low frequency is attributed to the diffusion process.

Analysis of Electrochemical Impedances for Different Films
In this study, the electrochemical impedances of modified and unmodified films were analyzed to investigate their electron transfer abilities. The electrochemical impedance spectroscopy (EIS) is broadly employed to the descriptions of fundamental electrochemical and electronic processes [19]. The Nyquist diagram and its equivalent circuit for electrochemical biosensor are shown in Figure 6a,b. R et and R s are the electron transfer resistance and solution resistance, respectively. C dl is the double layer capacitance. Z d is the diffusion impedance. Figure 6a showed a curve that consists of many complex impedances as function of frequency. In Cartesian coordinates, the real part and imaginary part are plotted on the X and Y axes, respectively. The left semicircle in high frequency is attributed to the electron transfer process, and its diameter is R et . The right arc in low frequency is attributed to the diffusion process.

Analysis of Electrochemical Impedances for Different Films
In this study, the electrochemical impedances of modified and unmodified films were analyzed to investigate their electron transfer abilities. The electrochemical impedance spectroscopy (EIS) is broadly employed to the descriptions of fundamental electrochemical and electronic processes [19]. The Nyquist diagram and its equivalent circuit for electrochemical biosensor are shown in Figure 6a,b. R et and R s are the electron transfer resistance and solution resistance, respectively. C dl is the double layer capacitance. Z d is the diffusion impedance. Figure 6a showed a curve that consists of many complex impedances as function of frequency. In Cartesian coordinates, the real part and imaginary part are plotted on the X and Y axes, respectively. The left semicircle in high frequency is attributed to the electron transfer process, and its diameter is R et . The right arc in low frequency is attributed to the diffusion process.

Detection Range of Flexible Arrayed Lactate Biosensor
According to the above, we could know that the optimal GO and MB contents were, respectively, 0.3 wt % and 0.75 mL for flexible arrayed lactate biosensor. Next, we also investigated its detection range, as shown in Figure 8. The response voltage was measured by potentiometric measurement system. We could observe that the response voltage was decreased with an increase in lactate concentration until about 15 mM. There was no significant variation on response voltage between 10 mM and 15 mM, and their error bars were inter-overlapped. After that, we could estimate its lower limit of detection by a point of intersection for two lines, where one line was the response voltage for pure PBS (not containing lactate), and another line was the trend line of experiment data. Figure 9 showed that the lower limit of detection is about 2 μM.

Detection Range of Flexible Arrayed Lactate Biosensor
According to the above, we could know that the optimal GO and MB contents were, respectively, 0.3 wt % and 0.75 mL for flexible arrayed lactate biosensor. Next, we also investigated its detection range, as shown in Figure 8. The response voltage was measured by potentiometric measurement system. We could observe that the response voltage was decreased with an increase in lactate concentration until about 15 mM. There was no significant variation on response voltage between 10 mM and 15 mM, and their error bars were inter-overlapped. After that, we could estimate its lower limit of detection by a point of intersection for two lines, where one line was the response voltage for pure PBS (not containing lactate), and another line was the trend line of experiment data. Figure 9 showed that the lower limit of detection is about 2 µM.

Anti-Interference Effect of Flexible Arrayed Lactate Biosensor
In order to confirm the anti-interference ability of flexible arrayed lactate biosensor, the various substances were added into the test solution when monitoring. In this study, we used some substances that are commonly present in the human body, such as urea, uric acid, ascorbic acid and glucose. At the beginning of the experiment, we immersed the flexible arrayed lactate biosensor into the test solution containing the lactate, and which concentration was 1.3 mM. The response voltage was measured by potentiometric measurement system. Then, 4.3 mM urea, 0.01 mM uric acid, 0.06 mM ascorbic acid and 5 mM glucose were sequentially added into the test solution by using micropipettor because they were common substances in the human body. In addition, these concentrations belonged to the normal range in the human body. After that, 3 mM solution was added into the test solution by using micropipettor. The result of the anti-interference effect was shown in Figure 10. Even though these substances were added into 1.3 mM solution, the response voltage kept within the reasonable range. There was no obvious variation on response voltage, and which fell within the response range of 1.3 mM lactate solution. However, when 3 mM solution was added into the test solution, the response voltage was varied significantly. It exhibited excellent

Anti-Interference Effect of Flexible Arrayed Lactate Biosensor
In order to confirm the anti-interference ability of flexible arrayed lactate biosensor, the various substances were added into the test solution when monitoring. In this study, we used some substances that are commonly present in the human body, such as urea, uric acid, ascorbic acid and glucose. At the beginning of the experiment, we immersed the flexible arrayed lactate biosensor into the test solution containing the lactate, and which concentration was 1.3 mM. The response voltage was measured by potentiometric measurement system. Then, 4.3 mM urea, 0.01 mM uric acid, 0.06 mM ascorbic acid and 5 mM glucose were sequentially added into the test solution by using micropipettor because they were common substances in the human body. In addition, these concentrations belonged to the normal range in the human body. After that, 3 mM solution was added into the test solution by using micropipettor. The result of the anti-interference effect was shown in Figure 10. Even though these substances were added into 1.3 mM solution, the response voltage kept within the reasonable range. There was no obvious variation on response voltage, and which fell within the response range of 1.3 mM lactate solution. However, when 3 mM solution was added into the test solution, the response voltage was varied significantly. It exhibited excellent

Anti-Interference Effect of Flexible Arrayed Lactate Biosensor
In order to confirm the anti-interference ability of flexible arrayed lactate biosensor, the various substances were added into the test solution when monitoring. In this study, we used some substances that are commonly present in the human body, such as urea, uric acid, ascorbic acid and glucose. At the beginning of the experiment, we immersed the flexible arrayed lactate biosensor into the test solution containing the lactate, and which concentration was 1.3 mM. The response voltage was measured by potentiometric measurement system. Then, 4.3 mM urea, 0.01 mM uric acid, 0.06 mM ascorbic acid and 5 mM glucose were sequentially added into the test solution by using micropipettor because they were common substances in the human body. In addition, these concentrations belonged to the normal range in the human body. After that, 3 mM solution was added into the test solution by using micropipettor. The result of the anti-interference effect was shown in Figure 10. Even though these substances were added into 1.3 mM solution, the response voltage kept within the reasonable range. There was no obvious variation on response voltage, and which fell within the response range of 1.3 mM lactate solution. However, when 3 mM solution was added into the test solution, the response voltage was varied significantly. It exhibited excellent anti-interference ability, which would be attributed to excellent selectivity of flexible arrayed lactate biosensor.
anti-interference ability, which would be attributed to excellent selectivity of flexible arrayed lactate biosensor.

Stability of Flexible Arrayed Lactate Biosensor under Bending
Moreover, we also investigated another important issue, i.e., the stability of flexible arrayed lactate biosensor after repeated bending. The flexible arrayed lactate biosensor was bended under U shape, as shown in Figure 2b. After that, the flexible arrayed lactate biosensor was unfolded and immersed in 1.3 mM lactate solution, and which response voltage was measured by potentiometric measurement system. The above steps were repeated 20 times. The results were shown in Figure 11. We observed the variations on response voltage of flexible arrayed lactate biosensor after each bending. After the flexible arrayed lactate biosensor was bended 20 times, there was almost no variation on response voltage. The variation of response voltage was around 10 mV, as shown in Figure 11.

Stability of Flexible Arrayed Lactate Biosensor under Bending
Moreover, we also investigated another important issue, i.e., the stability of flexible arrayed lactate biosensor after repeated bending. The flexible arrayed lactate biosensor was bended under U shape, as shown in Figure 2b. After that, the flexible arrayed lactate biosensor was unfolded and immersed in 1.3 mM lactate solution, and which response voltage was measured by potentiometric measurement system. The above steps were repeated 20 times. The results were shown in Figure 11. We observed the variations on response voltage of flexible arrayed lactate biosensor after each bending. After the flexible arrayed lactate biosensor was bended 20 times, there was almost no variation on response voltage. The variation of response voltage was around 10 mV, as shown in Figure 11.

Stability of Flexible Arrayed Lactate Biosensor under Bending
Moreover, we also investigated another important issue, i.e., the stability of flexible arrayed lactate biosensor after repeated bending. The flexible arrayed lactate biosensor was bended under U shape, as shown in Figure 2b. After that, the flexible arrayed lactate biosensor was unfolded and immersed in 1.3 mM lactate solution, and which response voltage was measured by potentiometric measurement system. The above steps were repeated 20 times. The results were shown in Figure 11. We observed the variations on response voltage of flexible arrayed lactate biosensor after each bending. After the flexible arrayed lactate biosensor was bended 20 times, there was almost no variation on response voltage. The variation of response voltage was around 10 mV, as shown in Figure 11.    Table 3 showed the comparisons of sensitivity and their concentration ranges for various lactate biosensors. In the literature [20], they immobilized the lactate oxidase (LOD) and manganese dioxide nanoparticles (MnO 2 NPs) into layer-by-layer poly(dimethyldiallylammonium chloride) (PDDA) films to fabricate (PDDA/MnO 2 /PDDA/LOD) n multilayer films, and which were used for the lactate biosensor based on enzyme field effect transistor (ENFET). The sensitivity was 16.84 mV/mM. In the literature [21], they proposed NAD + -dependent ENFET-based lactate biosensor.

Comparisons of Various Lactate Biosensors
The aminosiloxane-functionalized gate interface was modified with pyrroloquinoline quinone (PQQ), which acted as a catalyst for the oxidation of NADH. NAD + and LDH were covalently linked to the PQQ monolayer. The sensitivity was 26 mV/dec, and lower limit of detection was 0.1 mM. In the literature [22], they reported a microsensor, and which used an ENFET with a platinum microelectrode and combined potentiometric and amperometric techniques. The sensitivity could achieve 20 mV/mM. In the literature [23], they developed a lactate biosensor based on electrolyte-insulator-semiconductor using a nanostructured Si 3 N 4 surface modified by a polyacrylic acid (PAA) layer, and which could covalently link to the NH 2 groups of LDH. The sensitivity was 49.7 mV/dec. Compared with the other literatures [20][21][22][23], the flexible arrayed lactate biosensor based on NiO film modified by GO and MBs exhibits excellent sensitivity though its concentration range was narrower than the other literatures [20][21][22][23].

Conclusions
The flexible arrayed lactate biosensor based on immobilizing LDH-NAD + on NiO film modified by GO and MBs exhibits excellent sensitivity (45.397 mV/mM) with a linearity of 0.992. Besides, it had a great anti-interference ability, and its lower limit of detection was estimated to be about 2 µM. According to EIS results, the electron transfer resistance of LDH-NAD + -MBs/GPTS/GO/NiO film was smaller than those of LDH-NAD + /GPTS/GO/NiO film and LDH-NAD + /GPTS/NiO film, and it presented the outstanding electron transfer ability. Moreover, after the flexible arrayed lactate biosensor was bended 20 times, there was almost no variation on response voltage.