Laser-Induced Graphene-Based Enzymatic Biosensor for Glucose Detection

Laser-induced graphene (LIG) has recently been receiving increasing attention due to its simple fabrication and low cost. This study reports a flexible laser-induced graphene-based electrochemical biosensor fabricated on a polymer substrate by the laser direct engraving process. For this purpose, a 450 nm UV laser was employed to produce a laser-induced graphene electrode (LIGE) on a polyimide substrate. After the laser engraving of LIGE, the chitosan–glucose oxidase (GOx) composite was immobilized on the LIGE surface to develop the biosensor for glucose detection. It was observed that the developed LIGE biosensor exhibited good amperometric responses toward glucose detection over a wide linear range up to 8 mM. The GOx/chitosan-modified LIGE biosensor showed high sensitivity of 43.15 µA mM−1 cm−2 with a detection limit of 0.431 mM. The interference studies performed with some possible interfering compounds such as ascorbic acid, uric acid, and urea exhibited no interference as there was no difference observed in the amperometric glucose detection. It was suggested that the LIGE-based biosensor proposed herein was easy to prepare and could be used for low-cost, rapid, and sensitive/selective glucose detection.


Introduction
Numerous novel and cutting-edge technologies and materials are necessary to satisfy the new trends and requisites of analytical systems as the needs for environmental, biomedical, food and beverage analysis are progressing very quickly. The development of biosensors has evolved as one of the most promising research directions to overcome these challenges. Therefore, biosensor-based techniques have recently started being applied for the determination of different clinically, environmentally and biologically active materials [1][2][3]. In this regard, the design of biosensors in nanoscience/nanotechnology, environmental, medicine and food monitoring has been significantly increased during the past decade for their extensive applications. These advanced technologies have assisted the construction of highly sensitive, selective, customizable, and portable sensors for the determination of various clinically significant materials such as glucose, etc. [4]. The progress of such glucose biosensors has an inordinate significance in diagnosing and controlling diabetes mellitus, which is considered a worldwide public health problem. Diabetes mellitus would increase the risk of heart disease, kidney failure, blindness, postoperative and wound infections [5,6].
Diabetes mellitus has increased worldwide over the past five decades. Diabetes is a medical condition in which patients experience glucose concentration diverging from the normal range of 80-120 mg/dL (4.4-6.6 mM) [7]. In 2019, the International Diabetes Federation (IDF) assessments indicated that approximately 463 million adults have diabetes, and it might rise to 700 million by 2045 [8]. Diabetic patients are required to perform glucose testing several times a day to maintain normal glucose levels. Hence, the rapid quantification of glucose concentration in bodily fluids is vital for diagnosing and treating spectroscopic study was conducted using a micro-Raman spectrometer (JASCO NRS-4100; Laser 532 nm) with a spectral resolution of 2 cm −1 . Data processing/plotting was performed using Origin 9.1 software (OriginLab Inc., Northampton, MA, USA).

Fabrication of LIGE Sensor
A 3-electrode system was designed using AutoCAD software with a 3 mm diameter of working electrode and laser-inscribed to graphene-based electrodes. Kapton ® polyimide tape was pasted onto a paper substrate and cleaned with isopropanol and deionized water. Then, the designed pattern made in graphic software was inscribed on the surface of the Kapton tape using a laser engraving machine (HANLIN 7WLS, 7 W, 450 nm) to form highly conductive graphene electrodes, as shown in Figure 1. The resistance of the graphenebased electrode was optimized by adjusting the laser power intensity (22% of the machine's maximum power), engraving depth (5%), the distance between the laser head and the polyimide substrate (~13 cm). The duration for fabricating a complete LIGE sensor was 2.8 min.

Fabrication of LIGE Sensor
A 3-electrode system was designed using AutoCAD software with a 3 mm dia of working electrode and laser-inscribed to graphene-based electrodes. Kapton ® mide tape was pasted onto a paper substrate and cleaned with isopropanol and deio water. Then, the designed pattern made in graphic software was inscribed on the s of the Kapton tape using a laser engraving machine (HANLIN 7WLS, 7 W, 450 n form highly conductive graphene electrodes, as shown in Figure 1. The resistance graphene-based electrode was optimized by adjusting the laser power intensity (2 the machine's maximum power), engraving depth (5%), the distance between the head and the polyimide substrate (~13 cm). The duration for fabricating a complete sensor was 2.8 min.

Immobilization of GOx/Chitosan Composite on the LIGE
The glucose biosensor was prepared by immobilizing the glucose oxidase an tosan hydrogel homogeneous biocomposite on the LIGE surface. The resulting bi posite could retain the enzyme bioactivity at considerably extreme conditions [35 milligrams of GOx and three milligrams of chitosan were dissolved in 0.5 mL of deio water and stirred for 5 min [36]. Subsequently, 5 µL of the mixture was cast onto th face of the LIGE working electrode. Then, the LIGE sensor was kept in a refrigerat °C for 24 h.

Electrochemical Measurements
All the electrochemical measurements were carried out using PalmSens 4 p ostat (PalmSens, Houten, The Netherlands) at room temperature. The electrochemi dox characteristics of the LIGE were measured by Cyclic voltammetry (CV) with dif concentrations of potassium ferri (III)cyanide (K3[Fe(CN)6]) in 50 mM of phosphate ered solution (PBS). CV measurements were performed at a scan rate of 50 mV/s potential range from −0.8 to +0.8 V. Chronoamperometry (CA) experiments for g detection with LIGE were performed in 50 mM PBS at the fixed applied voltage o for 60 s. The detection principle of glucose is based on the electron transfer mecha GOx reacts with glucose in the presence of O2 and produces gluconolactone and H change in electrical current occurs at the electrode surface during these reactions the electron transfer. Additionally, the resulting current response is proportional number of glucose molecules present in the sample.

Immobilization of GOx/Chitosan Composite on the LIGE
The glucose biosensor was prepared by immobilizing the glucose oxidase and chitosan hydrogel homogeneous biocomposite on the LIGE surface. The resulting biocomposite could retain the enzyme bioactivity at considerably extreme conditions [35]. Five milligrams of GOx and three milligrams of chitosan were dissolved in 0.5 mL of deionized water and stirred for 5 min [36]. Subsequently, 5 µL of the mixture was cast onto the surface of the LIGE working electrode. Then, the LIGE sensor was kept in a refrigerator at 4 • C for 24 h.

Electrochemical Measurements
All the electrochemical measurements were carried out using PalmSens 4 potentiostat (PalmSens, Houten, The Netherlands) at room temperature. The electrochemical redox characteristics of the LIGE were measured by Cyclic voltammetry (CV) with different concentrations of potassium ferri (III)cyanide (K 3 [Fe(CN) 6 ]) in 50 mM of phosphate-buffered solution (PBS). CV measurements were performed at a scan rate of 50 mV/s with a potential range from −0.8 to +0.8 V. Chronoamperometry (CA) experiments for glucose detection with LIGE were performed in 50 mM PBS at the fixed applied voltage of 0.8 V for 60 s. The detection principle of glucose is based on the electron transfer mechanism. GOx reacts with glucose in the presence of O 2 and produces gluconolactone and H 2 O 2 . A change in electrical current occurs at the electrode surface during these reactions due to the electron transfer. Additionally, the resulting current response is proportional to the number of glucose molecules present in the sample.

Optimization of Applied Potential and pH
CA measurements were used to determine the optimal applied potential and pH for glucose detection. The CA potential was optimized by varying the potential from 0.3 V to 1.3 V (5 mM Glucose, pH 7), and the resulting CA current was sampled at 60 s. CA Polymers 2021, 13, 2795 4 of 10 measurements were performed with a LIGE biosensor at 5 mM glucose solution with the applied potential of 0.8 V by varying the pH of the phosphate-buffered solution from a pH of 5 to 9, and the optimal pH was found.

Interference Study
The response of the LIGE biosensor for glucose detection was evaluated in the presence of potential interferences such as 0.1 mM ascorbic acid, 0.1 mM uric acid, and 3 mM urea (pH 7; 5 mM glucose; 0.8 V).

Raman Spectra
In this study, a graphene three-electrode system for electrochemical sensing applications was developed by direct laser inscribing on polymer substrate (Polyimide). The prepared LIGE was characterized with Raman spectra, as shown in Figure 2. The Raman spectrum consists of G band at ca. 1592 cm −1 related to the E 2g phonon of the sp 2 carbon atoms, and D band at ca. 1340 cm −1 corresponds to the disordered grain boundaries [37,38]. Two other bands were observed at 2697 and 2900 cm −1 . The band at ca. 2700 cm −1 is known as the 2D band, an indicator of the number of graphene layers. A sharp peak will appear at ca. 2700 cm −1 for monolayer graphene. Here, the broadened band was observed, which would be attributed to the prepared graphene containing many layers with some defects. The band that appeared at 2900 cm −1 is called an S3 band, which is a second-order peak derived from the D-G peak combination. The band intensity ratio of S3-2D is proportional to the reduction in defects [38]. This Raman spectra result indicated that the obtained black material on polyimide substrate was carbon-based graphene.

Optimization of Applied Potential and pH
CA measurements were used to determine the optimal applied potential and pH for glucose detection. The CA potential was optimized by varying the potential from 0.3 V to 1.3 V (5 mM Glucose, pH 7), and the resulting CA current was sampled at 60 s. CA measurements were performed with a LIGE biosensor at 5 mM glucose solution with the applied potential of 0.8 V by varying the pH of the phosphate-buffered solution from a pH of 5 to 9, and the optimal pH was found.

Interference Study
The response of the LIGE biosensor for glucose detection was evaluated in the presence of potential interferences such as 0.1 mM ascorbic acid, 0.1 mM uric acid, and 3 mM urea (pH 7; 5 mM glucose; 0.8 V).

Raman Spectra
In this study, a graphene three-electrode system for electrochemical sensing applications was developed by direct laser inscribing on polymer substrate (Polyimide). The prepared LIGE was characterized with Raman spectra, as shown in Figure 2. The Raman spectrum consists of G band at ca. 1592 cm −1 related to the E2g phonon of the sp 2 carbon atoms, and D band at ca. 1340 cm −1 corresponds to the disordered grain boundaries [37,38]. Two other bands were observed at 2697 and 2900 cm −1 . The band at ca. 2700 cm −1 is known as the 2D band, an indicator of the number of graphene layers. A sharp peak will appear at ca. 2700 cm −1 for monolayer graphene. Here, the broadened band was observed, which would be attributed to the prepared graphene containing many layers with some defects. The band that appeared at 2900 cm −1 is called an S3 band, which is a second-order peak derived from the D-G peak combination. The band intensity ratio of S3-2D is proportional to the reduction in defects [38]. This Raman spectra result indicated that the obtained black material on polyimide substrate was carbon-based graphene.

Electrochemical Characterization
Before developing the glucose biosensor with LIGE, validating the LIGE sensor towards electrochemical sensing was necessary. The ferri/ferrocyanide (Fe(CN) 6 3−/4− ) redox couple is one of the most widely used electron mediators for electrochemical reactions [39]. The performance of an electrochemical sensor towards an electron mediator was considered most relevant to general biochemical sensing applications. Thus, the electrochemical efficacy of the LIGE sensor was evaluated using cyclic voltammetry responses in different concentrations of ferricyanide redox mediator (K 3  Before developing the glucose biosensor with LIGE, validating the LIGE sensor towards electrochemical sensing was necessary. The ferri/ferrocyanide (Fe(CN)6 3−/4− ) redox couple is one of the most widely used electron mediators for electrochemical reactions [39]. The performance of an electrochemical sensor towards an electron mediator was considered most relevant to general biochemical sensing applications. Thus, the electrochemical efficacy of the LIGE sensor was evaluated using cyclic voltammetry responses in different concentrations of ferricyanide redox mediator (K3[Fe(CN)6]), as shown in Figure 3a. As seen from Figure 3a, the oxidation peaks' current increased from 35.495 to 65.043 µA when the ferricyanide concentration ranged from 0.5 to 2.5 mM. The oxidation peak current showed an excellent linear relationship with different ferricyanide concentrations, as shown in Figure 3b.

Characterization of GOx/Chitosan Immobilized LIGE
Cyclic voltammetry measurement was performed to confirm the LIGE immobilization with GOx/Chitosan. Figure 4 shows the cyclic voltammograms of potassium ferricyanide at bare LIGE and GOx/chitosan composite-modified LIGE. It can be seen that after the immobilization of GOx/chitosan composite onto the LIGE surface, the peak current decreased to 24.325 from 58.336 µA of the bare LIGE. The electron transfer kinetics of [Fe(CN)6] 4− /[Fe(CN)6] 3− is significantly hindered after the LIGE surface was modified with GOx/chitosan. This result confirmed that the GOx/chitosan was successfully immobilized on the LIGE surface.

Characterization of GOx/Chitosan Immobilized LIGE
Cyclic voltammetry measurement was performed to confirm the LIGE immobilization with GOx/Chitosan. Figure 4 shows the cyclic voltammograms of potassium ferricyanide at bare LIGE and GOx/chitosan composite-modified LIGE. It can be seen that after the immobilization of GOx/chitosan composite onto the LIGE surface, the peak current decreased to 24.325 from 58.336 µA of the bare LIGE. The electron transfer kinetics of [Fe(CN)6] 4− /[Fe(CN)6] 3− is significantly hindered after the LIGE surface was modified with GOx/chitosan. This result confirmed that the GOx/chitosan was successfully immobilized on the LIGE surface.

Amperometric Detection of Glucose by the Proposed LIGE
The chronoamperometry technique was employed to detect glucose using GOx/Chitosan coated LIGE sensor at a constant oxidation potential of +0.8 V. Figure 5a depicts the chronoamperometric responses of the LIGE biosensor with glucose concentrations rang-

Amperometric Detection of Glucose by the Proposed LIGE
The chronoamperometry technique was employed to detect glucose using GOx/ Chitosan coated LIGE sensor at a constant oxidation potential of +0.8 V. Figure 5a depicts the chronoamperometric responses of the LIGE biosensor with glucose concentrations ranging from 0 to 10 mM. The current response increased with increasing glucose concentrations. The steady-state current response at 60 s was chosen for the detection of glucose concentration. The amperometric current response of the LIGE biosensor exhibited a linear relationship with the glucose concentrations ranging from 0 to 8 mM, and the current began to level off at a glucose concentration higher than 8 mM as shown in Figure 5b. The linear regression equation was y = 3.05x + 8.54, with a coefficient of determination R 2 = 0.97 and a sensitivity of 43.15 µA mM −1 cm −2 . The limit of detection was calculated according to the 3s a /b criterion, where b was the slope of the calibration curve, and s a was the estimated standard deviation of the y-intercepts of the regression line [3]. The detection limit calculated was 0.431 mM. As seen from Figure 5b, the linear part of the calibration curve includes the normal glucose levels (4.4 to 6.6 mM) in the human blood. Thus, this study could offer a simple approach for the clinical glucose measurement with a disposable LIGE-based biosensor. The performance of the proposed biosensor was compared with other reported glucose biosensors, as shown in Table 1. The developed LIGE-based enzymatic glucose biosensor exhibited good analytical characteristics towards glucose detection such as good linearity and high sensitivity. Moreover, the fabrication and detection procedures of the proposed LIGE-based biosensor were also simple, rapid, and cost-effective.  a GC-glassy carbon electrode; MWCNT-multi-walled carbon nanotubes; Fe3O4/PDA-magnetite/polydopamine; LSGlaser-scribed graphene; PBSE-pyrenebutanoic acid-succinimide ester; PtNPs-platinum nanoparticles; MoS2-molybdenum disulfide; PGE-pencil graphite electrode; CPE-carbon paste electrode; SiO2/Lig-silica/lignin; Cys-cysteine; GA-glutaraldehyde; PPy-polypyrrole; SPCE-screen-printed carbon electrodes; GOD/GOx-Glucose oxidase.

Michaelis-Menten Kinetics
The maximum response current ( ) and the apparent Michaelis-Menten constant ( ) were used to analyze the relationship between chronoamperometric signals and enzymatic reaction. As shown in Figure 5b, when glucose concentration exceeds 6 mM, a response plateau was observed with the characteristic of Michaelis-Menten kinetic mech-

Michaelis-Menten Kinetics
The maximum response current (I max ) and the apparent Michaelis-Menten constant (K app m ) were used to analyze the relationship between chronoamperometric signals and enzymatic reaction. As shown in Figure 5b, when glucose concentration exceeds 6 mM, a response plateau was observed with the characteristic of Michaelis-Menten kinetic mechanism. From the calibration plot (Figure 5b dotted line), the current response showed hyperbolic dependence on glucose concentration and was in good agreement with Michaelis-Menten kinetics [45]. The kinetic parameters, the maximum current generated during the enzymatic reaction (I max ) and the apparent Michaelis constant (K

Optimization of Applied Potential and Buffer pH
Chronoamperometry measurements were used to determine the optimal applied potential and pH for glucose detection with the developed enzymatic LIGE biosensor. Figure 6a shows the chronoamperometric response of the LIGE biosensor at 60 s with different applied potential values ranging from 0.3 to 1.3 V. The results showed that the current increased with increasing applied potential from 0.3 to 0.8 V and currents tended to level off when the potential increased beyond 0.8 V. Thus, 0.8 V was selected as the optimized potential for amperometric glucose detection. Figure 6b illustrates the chronoamperometry current response of the biosensor as a function of the pH of PBS containing 2 mM glucose. The current responses at pH 5, pH 6, and pH 7 were almost similar. Considering the pH of a physiological buffer, pH 7 was chosen for the glucose detection experiments. The results showed that the current increased with increasing applied potential from 0.3 to 0.8 V and currents tended to level off when the potential increased beyond 0.8 V. Thus, 0.8 V was selected as the optimized potential for amperometric glucose detection. Figure 6b illustrates the chronoamperometry current response of the biosensor as a function of the pH of PBS containing 2 mM glucose. The current responses at pH 5, pH 6, and pH 7 were almost similar. Considering the pH of a physiological buffer, pH 7 was chosen for the glucose detection experiments.

Interference Study
The developed LIGE-based enzymatic biosensor was evaluated with possible interferences by comparing the chronoamperometric responses before and after adding some interferents such as ascorbic acid (0.1 mM), uric acid (0.1 mM), and urea (3 mM) in 5 mM glucose. As shown in Figure 7, the chronoamperometric current responses for glucose without and with interferents showed practically no interference. The LIGE was modified with GOx, which is the standard enzyme for biosensors and it has relatively higher selectivity for glucose [48]. Hence, the LIGE biosensor was suggested to possess good selectivity due to the specificity of the GOx enzyme.

Interference Study
The developed LIGE-based enzymatic biosensor was evaluated with possible interferences by comparing the chronoamperometric responses before and after adding some interferents such as ascorbic acid (0.1 mM), uric acid (0.1 mM), and urea (3 mM) in 5 mM glucose. As shown in Figure 7, the chronoamperometric current responses for glucose with-Polymers 2021, 13, 2795 8 of 10 out and with interferents showed practically no interference. The LIGE was modified with GOx, which is the standard enzyme for biosensors and it has relatively higher selectivity for glucose [48]. Hence, the LIGE biosensor was suggested to possess good selectivity due to the specificity of the GOx enzyme.

Conclusions
We developed a simple laser-induced graphene-based enzymatic biosensor for glucose detection. The proposed detection strategy could offer an easy and low-cost route to mass-produce sensitive biosensing electrodes. The chronoamperometric measurements successfully detected the glucose over a linear range from 0 to 8 mM with a detection limit of 0.431 mM. The biosensor response was not affected by interfering compounds (ascorbic acid, uric acid and urea) and demonstrated the high specificity and selectivity of this LIGE biosensor in glucose detection. The proposed LIGE biosensor holds excellent promise in point-of-care diagnosis. Our future study aims to validate the biosensor response in human blood samples for real-life applications.

Stability and Reproducibility of Biosensor
The stability of the developed GOx/chitosan-modified LIGE biosensor was evaluated by measuring the amperometric current response in the presence of 5 mM glucose over 25 days stored at 4 • C in a refrigerator. The biosensor exhibited~90% stability for 10 days, and the response remained approximately 72-85% after 10 days. The reproducibility of the developed biosensor was assessed from the current response of different biosensors prepared independently. In this work, all the measurements were taken from at least three independent sensors (n ≥ 3), and the reproducible signals were obtained with the RSD less than 6%.

Conclusions
We developed a simple laser-induced graphene-based enzymatic biosensor for glucose detection. The proposed detection strategy could offer an easy and low-cost route to mass-produce sensitive biosensing electrodes. The chronoamperometric measurements successfully detected the glucose over a linear range from 0 to 8 mM with a detection limit of 0.431 mM. The biosensor response was not affected by interfering compounds (ascorbic acid, uric acid and urea) and demonstrated the high specificity and selectivity of this LIGE biosensor in glucose detection. The proposed LIGE biosensor holds excellent promise in point-of-care diagnosis. Our future study aims to validate the biosensor response in human blood samples for real-life applications.