Bacteria-Templated NiO Nanoparticles/Microstructure for an Enzymeless Glucose Sensor

The bacterial-induced hollow cylinder NiO (HCNiO) nanomaterial was utilized for the enzymeless (without GOx) detection of glucose in basic conditions. The determination of glucose in 0.05 M NaOH solution with high sensitivity was performed using cyclic voltammetry (CV) and amperometry (i–t). The fundamental electrochemical parameters were analyzed and the obtained values of diffusion coefficient (D), heterogeneous rate constant (ks), electroactive surface coverage (Г), and transfer coefficient (alpha-α) are 1.75 × 10−6 cm2/s, 57.65 M−1·s−1, 1.45 × 10−10 mol/cm2, and 0.52 respectively. The peak current of the i–t method shows two dynamic linear ranges of calibration curves 0.2 to 3.5 µM and 0.5 to 250 µM for the glucose electro-oxidation. The Ni2+/Ni3+ couple with the HCNiO electrode and the electrocatalytic properties were found to be sensitive to the glucose oxidation. The green chemistry of NiO preparation from bacteria and the high catalytic ability of the oxyhydroxide (NiOOH) is the good choice for the development of a glucose sensor. The best obtained sensitivity and limit of detection (LOD) for this sensor were 3978.9 µA mM−1·cm−2 and 0.9 µM, respectively.


Introduction
Nowadays, amperometric glucose sensors are relevant for use in blood sugar monitoring with reliable sensitivity and selectivity in the health care industry. In the clinical field it was estimated that 2.8% of the world population, around 171 million people, were affected by diabetes in 2000. It will be projected to be 4.4% by the year 2030, approximately 366 million people [1]. The diagnosis of diabetes has become a far more sophisticated branch of science with increasing self-testing kits. Almost 85% of the entire biosensor market is the commercial glucose biosensor which makes diabetes a model for the development of new biosensors [2]. Biosynthesis of nanoparticles (NPs) using biological molecules and microorganisms has rapidly emerged as nanobiotechnology, i.e., biotechnology combined with nanotechnology [3]. Green nanotechnology could be a best alternative for the nanomaterial synthesis with meticulous nature with the help of variety of biological molecules, such as proteins, carbohydrates, and polyphenols [4]. NPs with different shapes, such as spherical, triangular, octahedral, cluster, and amorphous crystalline have been successfully synthesized using microorganisms. Pseudomonas stutzeri, Scheme 1. Pictorial representation of the reaction mechanism.
Thus, the couple at half-wave potential (Ep/2) = 0.45 V and ∆Ep = 0.08 V values exhibited a Ni 3+ /Ni 2+ redox with a quasi-reversible process. During the successive scan the anodic peak current (Ipa) beyond the +0.6 V increased successively, which may be attributed to the electro-oxidation of OH − ions to O2 with HO • radicals as intermediates. This has also been checked with the increasing OH − ion concentration of supporting electrolyte for this system. It is observed that the anodic peak potential (Epa) and cathodic peak potential (Epc) shifted in the negative direction of potentials (lower), and the peak currents (Ipa and Ipc) were shifting to higher values when the scan cycle number increased, which might be due to the rapid development of NiOOH by adsorptive behavior of nucleation from Ni(OH)2. This is the indication of active sites of Ni 2+ /Ni 3+ produced on the surface of HCNiO/GC electrode, which is in agreement with the similar study of Wang et al. in the Ti/TiO2 Thus, the couple at half-wave potential (Ep/2) = 0.45 V and ∆Ep = 0.08 V values exhibited a Ni 3+ /Ni 2+ redox with a quasi-reversible process. During the successive scan the anodic peak current (Ipa) beyond the +0.6 V increased successively, which may be attributed to the electro-oxidation of OH − ions to O2 with HO • radicals as intermediates. This has also been checked with the increasing OH − ion concentration of supporting electrolyte for this system. It is observed that the anodic peak potential (Epa) and cathodic peak potential (Epc) shifted in the negative direction of potentials (lower), and the peak currents (Ipa and Ipc) were shifting to higher values when the scan cycle number increased, which might be due to the rapid development of NiOOH by adsorptive behavior of nucleation from Ni(OH)2. This is the indication of active sites of Ni 2+ /Ni 3+ produced on the surface of

Effect of Scan Rates in the Blank Supporting Electrolyte
After a well production of Ni(OH) 2 /NiOOH couple by CV scanning at 0.05 V¨s´1, the modified electrode was transferred to a fresh 0.05 M NaOH aqueous solution for further studies. Figure 2 shows the cyclic voltammograms of HCNiO/GC-modified electrode at various scan rates (ν) obtained in 0.05 M NaOH solution. The peak currents of anodic (I pa ) and cathodic (I pc ) are linearly proportional to the wide range of scan rates between 50 and 5000 mV¨s´1, indicating an immobilized surface-controlled electrode process [41]. This adsorption-controlled process was confirmed by the double logarithmic (log I p vs. log ν) plot of CV data with the slope values greater than 0.5 (0.58 and 0.68 for oxidation and reduction peaks respectively, Figure 2b). In addition, peak currents (I pa and I pc ) are linearly proportional to the square root of scan rates (ν 1/2 ) at the same range of scan rates as shown in Figure 2c. Regression squared coefficient, (R 2 > 0.99 for both anodic and cathodic) exists for the linearity plot of I p vs. ν 1/2 , which indicated that the diffusion controlled process also involved. This is because the huge availability of OH´ion diffusion transport from the quiescent supporting electrolyte (liquid) to the vicinity of electrode surface film (solid) and vice versa (diffusion Ø adsorption), during the reaction shown in Equation (3) surface-controlled electrode process [41]. This adsorption-controlled process was confirmed by the double logarithmic (log Ip vs. log ν) plot of CV data with the slope values greater than 0.5 (0.58 and 0.68 for oxidation and reduction peaks respectively, Figure 2b). In addition, peak currents (Ipa and Ipc) are linearly proportional to the square root of scan rates (ν 1/2 ) at the same range of scan rates as shown in Figure 2c. Regression squared coefficient, (R 2 > 0.99 for both anodic and cathodic) exists for the linearity plot of Ip vs. ν 1/2 , which indicated that the diffusion controlled process also involved. This is because the huge availability of OH − ion diffusion transport from the quiescent supporting electrolyte (liquid) to the vicinity of electrode surface film (solid) and vice versa (diffusion ↔ adsorption), during the reaction shown in Equation (3), below [40].   However, the amount of Ni(OH)2 spontaneously formed at the surface due to the interaction of H2O or OH − was found to be extremely limited since no obvious redox peaks can be observed in the initial CV scan and at the lower applied potential energy than the overpotential of the modified electrode. Further reaction processes are occurring, as shown in the Equation (2) in Section 3.1. When increasing the scan rates, the anodic peak potential (Epa) was shifted to the more positive direction and the cathodic peak (Epc) moved towards the negative potential as shown in Figure 2a and inset a'. Even though the peak current increased remarkably with increasing scan rates, but the ratio of the anodic to cathodic peak current (Ipa/Ipc) was above unity. It indicates that the Ni(II) to Ni(III) hydrous oxide transformation process is a quasi-reversible reaction [42,43]. This is due to the charge interaction between the positively-charged Ni, Na, and negatively-charged OHions in the environment [44]. However, the amount of Ni(OH) 2 spontaneously formed at the surface due to the interaction of H 2 O or OH´was found to be extremely limited since no obvious redox peaks can be observed in the initial CV scan and at the lower applied potential energy than the overpotential of the modified electrode. Further reaction processes are occurring, as shown in the Equation (2) in Section 3.1. When increasing the scan rates, the anodic peak potential (E pa ) was shifted to the more positive direction and the cathodic peak (E pc ) moved towards the negative potential as shown in Figure 2a and inset a'. Even though the peak current increased remarkably with increasing scan rates, but the ratio of the anodic to cathodic peak current (I pa /I pc ) was above unity. It indicates that the Ni(II) to Ni(III) hydrous oxide transformation process is a quasi-reversible reaction [42,43]. This is due to the charge interaction between the positively-charged Ni, Na, and negatively-charged OHions in the environment [44]. Figure 3 shows the effect of glucose addition on the above situation. The electroactive species of the Ni(OH) 2 /NiOOH, redox couple generated on the modified HCNiO/GC electrode was tested with (peak X in Figure 3a") and without (peak Y in Figure 3a") 1 mM glucose in 0.05 M NaOH solution. It can be seen that an increased I pa from "78 µA (blank NaOH) to "114 µA (glucose-mixed NaOH solution) were produced at the scan rate of 50 mV¨s´1. This excellent electrocatalytic ability is due to the rich production of NiOOH in the HCNiO material by virtue of the enhanced surface area and special morphological nature on the substrate, glassy carbon electrode (GCE) [42]. Here, the cathodic peak (I pc ) decreased at higher than 50 mV¨s´1 sweep rate but the enhancement of the glucose oxidation peak (I pa ) was found (Figure 3a") and there was no reduction peak at the lower than 15 mV¨s´1 sweep rate. This is due to the change in the Ni 2+ /Ni 3+ concentration ratio. The redox reaction occurring on the electrode surface (Ni 2+ ØNi 3+ ) is restricted due to the limited OH´diffusion rate with the presence of glucose ions, which in turn limits the production of Ni(OH) 2 from NiOOH, resulting from the direct electrocatalytic oxidation of glucose to gluconolactone, as showed in Equation (4). The simplest proposed reaction mechanism and its rate constant for the above processes are shown in the following Scheme 2 and Equation (4) Figure 3 shows the effect of glucose addition on the above situation. The electroactive species of the Ni(OH)2/NiOOH, redox couple generated on the modified HCNiO/GC electrode was tested with (peak X in Figure 3a") and without (peak Y in Figure 3a") 1 mM glucose in 0.05 M NaOH solution. It can be seen that an increased Ipa from ~78 µA (blank NaOH) to ~114 µA (glucose-mixed NaOH solution) were produced at the scan rate of 50 mV·s −1 . This excellent electrocatalytic ability is due to the rich production of NiOOH in the HCNiO material by virtue of the enhanced surface area and special morphological nature on the substrate, glassy carbon electrode (GCE) [42]. Here, the cathodic peak (Ipc) decreased at higher than 50 mV·s −1 sweep rate but the enhancement of the glucose oxidation peak (Ipa) was found (Figure 3a") and there was no reduction peak at the lower than 15 mV·s −1 sweep rate. This is due to the change in the Ni 2+ /Ni 3+ concentration ratio. The redox reaction occurring on the electrode surface (Ni 2+ ↔Ni 3+ ) is restricted due to the limited OH − diffusion rate with the presence of glucose ions, which in turn limits the production of Ni(OH)2 from NiOOH, resulting from the direct electrocatalytic oxidation of glucose to gluconolactone, as showed in Equation (4). The simplest proposed reaction mechanism and its rate constant for the above processes are shown in the following Scheme 2 and Equation (4) [42,45]:

Effect of Scan Rates and Electrocatalytic Effect in the Basic Glucose Solution
In the above expressions the rate constants k1 (forward-anodic) and k−1 (reverse-cathodic) are obviously potential-dependent for the redox reaction, ° is a standard-rate constant, η is the overpotential, and other parameters are of their usual meaning. The peak potentials (Epa) shifted towards the positive direction (higher Epa), which indicated that the diffusion limitation of glucose in the catalytic process ( Figure 3a). The result might show glucose diffusion from the bulk solution phase to the surface of modified electrode in which NiOOH (Ni 3+ active sites) formation and glucose oxidation simultaneously occurred. Adsorption of glucose is more competitive than OH − on the surface of the electrode when the presence of glucose and OH − ions diffused from the solution. Hence, the higher potential was required to form more Ni 3+ sites from already-covered Ni 2+ sites on the surface for the presence of glucose at the higher scan rates [45][46][47]. This demonstrated the certainty of electrocatalytic oxidation of glucose. In addition, there was no poisoning effect on the surface by the consistent CV run on this system. The effect of scan rates was almost similar for the absence and the presence of 1 mM glucose in 0.05 M NaOH system (Figures 2 and 3). The potential sweep rates between the ranges of 50 to 3500 mV/s were used for the presence of glucose as in Figure 3. It can be seen in Figures 2a and 3a, the potential shifting (both Epa and Epc) occurs upon increasing sweep rates in both the cases. However, the oxidation peak (Epa) broadening started in the case of glucose solution ( Figure 3a). Eventually, at higher than 800 mV/s scan rates, a wider broadening arose that indicated the limitation was due to the charge transfer kinetics and which is Scheme 2. Glucose oxidation reaction mechanism under basic conditions on Ni(OH) 2 /NiOOH system. By simply, Ni pIIq and Ni pIIIq`Glucose In the above expressions the rate constants k 1 (forward-anodic) and k´1 (reverse-cathodic) are obviously potential-dependent for the redox reaction, k˝is a standard-rate constant, η is the overpotential, and other parameters are of their usual meaning. The peak potentials (E pa ) shifted towards the positive direction (higher E pa ), which indicated that the diffusion limitation of glucose in the catalytic process ( Figure 3a). The result might show glucose diffusion from the bulk solution phase to the surface of modified electrode in which NiOOH (Ni 3+ active sites) formation and glucose oxidation simultaneously occurred. Adsorption of glucose is more competitive than OH´on the surface of the electrode when the presence of glucose and OH´ions diffused from the solution. Hence, the higher potential was required to form more Ni 3+ sites from already-covered Ni 2+ sites on the surface for the presence of glucose at the higher scan rates [45][46][47]. This demonstrated the certainty of electrocatalytic oxidation of glucose. In addition, there was no poisoning effect on the surface by the consistent CV run on this system. The effect of scan rates was almost similar for the absence and the presence of 1 mM glucose in 0.05 M NaOH system (Figures 2 and 3). The potential sweep rates between the ranges of 50 to 3500 mV/s were used for the presence of glucose as in Figure 3. It can be seen in Figures 2a and 3a, the potential shifting (both E pa and E pc ) occurs upon increasing sweep rates in both the cases. However, the oxidation peak (E pa ) broadening started in the case of glucose solution ( Figure 3a). Eventually, at higher than 800 mV/s scan rates, a wider broadening arose that indicated the limitation was due to the charge transfer kinetics and which is also associated with the charge propagation in the surface film. A very slow e´transfer occurs in the complex situation of the film through Glu´, Ni 3+ , Ni 2+ , and regenerated Ni 3+ when abrupt increasing the sweep rate. This could barrier for the diffusion of ions and migration of es. It could also involve the chemical reactions between the ions in the solution (OH´, Glu´) and the nickel active sites on the surface (i.e., EC' mechanism). In addition, the polarizability of the ions might interfere for free movement (in and out) through the film. This similar diffusion process and the rate of the reaction have been already reported for other nickel-modified electrodes [43,44,48]. Figure 3b-c show the linear relationships for adsorption controlled and diffusion controlled processes with their regression (R 2 > 0.99) and slope values greater than 0.5 (I pa~0 .55 and I pc~0 .81), respectively [49]. This same range of values was obtained in the case of the absence of the glucose experiment ( Figure 2). This diffusion process is the rate-determining step (RDS) of the total redox process on the film when the presence of glucose, OH´ions and counter Na + ions (for charge neutralization of the film) [41,43,49]. The surface coverage concentration (Г) of HCNiO/GC was evaluated from the following Equation (5) where A is the area of the GCE (0.071 cm 2 ), n is the number of electrons involved in the redox reaction, Q is the charge obtained by integrating the anodic peak area of CV at low scan rate (ν = 10 mV/s) for HCNiO/GC, and F is the Faraday constant (96,500 C/mole). It is assumed that all of the immobilized redox centers are electro-active species on the voltammetry time scale. The values of surface coverage for 2 µL and 10 µL coated HCNiO/GC electrodes were 1.45ˆ10´1 0 mol/cm 2 and 1.25ˆ10´9 mol/cm 2 , respectively, which correspond to the presence of a monolayer of effective surface species. also associated with the charge propagation in the surface film. A very slow e − transfer occurs in the complex situation of the film through Glu − , Ni 3+ , Ni 2+ , and regenerated Ni 3+ when abrupt increasing the sweep rate. This could barrier for the diffusion of ions and migration of es. It could also involve the chemical reactions between the ions in the solution (OH − , Glu − ) and the nickel active sites on the surface (i.e., EC' mechanism). In addition, the polarizability of the ions might interfere for free movement (in and out) through the film. This similar diffusion process and the rate of the reaction have been already reported for other nickel-modified electrodes [43,44,48]. Figure 3b-c show the linear relationships for adsorption controlled and diffusion controlled processes with their regression (R 2 > 0.99) and slope values greater than 0.5 (Ipa ~ 0.55 and Ipc ~ 0.81), respectively [49]. This same range of values was obtained in the case of the absence of the glucose experiment ( Figure 2). This diffusion process is the rate-determining step (RDS) of the total redox process on the film when the presence of glucose, OH − ions and counter Na + ions (for charge neutralization of the film) [41,43,49]. The surface coverage concentration (Г) of HCNiO/GC was evaluated from the following Equation (5) where A is the area of the GCE (0.071 cm 2 ), n is the number of electrons involved in the redox reaction, Q is the charge obtained by integrating the anodic peak area of CV at low scan rate (ν = 10 mV/s) for HCNiO/GC, and F is the Faraday constant (96,500 C/mole). It is assumed that all of the immobilized redox centers are electro-active species on the voltammetry time scale. The values of surface coverage for 2 µL and 10 µL coated HCNiO/GC electrodes were 1.45 × 10 −10 mol/cm 2 and 1.25 × 10 −9 mol/cm 2 , respectively, which correspond to the presence of a monolayer of effective surface species.

Kinetic Studies of HCNiO/GC Electrode in the Basic Glucose Solution
Moreover, the adsorption process confirmed by the plot of the scan rate normalized current function (I/ν 1/2 ) vs. scan rate (ν). Current function increased with scan rate, as shown in Figure 4a. It is expected for catalytically-coupled adsorption of an electrochemical-chemical (EC') process. Steady linearity was found at higher ν in both the presence and absence of glucose cases, which indicates that the current function is independent of ν. This consistent fact revealed that the Ni(OH) 2 /NiOOH transition is the quasi-reversible process in the NaOH/glucose solution [44,50]. In order to obtain information on the rate determining step (RDS), the Tafel slope "b", was determined from the Figure 4b, plot of E p vs. log ν, using the following Tafel equation (Equation (6)), valid for a totally irreversible diffusion-controlled process: where "a" and "b" are Tafel constants 2.3RT αF log i o and´2 .3RT αF , respectively; and i and i o are current and exchange current, respectively [51]. The partial derivative of E vs. log ν plot is BE B log ν , equal to the Tafel slope b/2. Slope b/2 obtained for this work is 0.0573 V/decade, so, b = 0.1146 V/decade. This slope indicates that a one electron transfer process is the RDS (slow) with the electronic transfer coefficient (α = 0.52) for glucose solution. This one electron transfer (rate-limiting step) process was also confirmed by the another type of Tafel plot (E pa vs. log I), which were drawn using the data from the rising part of the steady-state current-voltage (I-E) curve for electrocatalytic oxidation of glucose recorded after 30 s of polarization at the desired potential at a scan rate of 5 mV/s in 0.05 M NaOH. A slope of 0.114 V/decade (i.e., 8.77 (V/decade)´1) is obtained as in Figure 4c, indicating the one electron transfer process to be the rate-limiting step with the transfer coefficient of α = 0.52 [52]. The value of diffusion coefficient (D, cm 2 /s) were obtained by the Randles-Sevcik equation (Equation (7)) for irreversible systems using the slope of scan rates study (I p vs. ν 1/2 ) with 1 mM glucose solution in 0.05 M NaOH [53]: I p " p2.99ˆ10 5 q α 1/2 A C˚D 1/2 ν 1/2 where C* is the bulk concentration of glucose in 0.05 M NaOH in terms of molar (mol/cm 3 ) solution.
In this study the obtained D value is 1.75ˆ10´6 cm 2 /s from the slope value of Figure 3c, which is correlated with the reported literature; A is the electrode surface area (cm 2 ) and ν is the scan rate (V¨s´1). The rate constant of the heterogeneous electron transfer (k s ) reaction between basic glucose solution and electro-generated Ni(II)/Ni(III) couple on the solid surface can be calculated by the CV technique [50]. Here, this system involved a second-order rate constant k s (M´1¨s´1) because OHá nd Glu´ions are necessary for the oxidation. It may be affected by specific adsorption, surface solvent layer, nature of the electrode material itself (GCE), and the film on the electrode surface. There is no bond formation and bond dissociation in this case [53,54]. The rate constant (k s ) was calculated using the following Nicholson and Shain's interpreted equation (Equation (8)) [50]: where I cat is the catalytic oxidation peak current with known glucose concentration (C) and I pa is the diffusion current of oxidation peak without glucose in 0.05 M NaOH. The linearity plot of (I cat /I pa ) vs.  Figure 5a-d, respectively. When increasing the concentration of glucose from 0.5 to 1.5 mM, the rate constant also increased gradually, and then decreased, starting at 2 mM glucose, as shown in the Figure 5d. This implied that the limitation of the rate constant occurred which could be attributed to the saturation of the active sites by the passive layer of the basic glucose molecules on the surface of the electrode [55]. It resulted that there may not be sufficient active sites for the adsorption of higher concentrations of glucose. In addition, some other chemical complications, such as the glucose intermediate occurring, could be a barrier to reach the catalytic surface of the electrode for the diffusion of glucose molecules. A high concentration of OH´might be required for the oxidation of a high concentration glucose, and this part of work is under progress along with this study [50,52]. The average value of the rate constant for the concentration range of 0.5 mM to 2 mM glucose solution was 57.65 M´1¨s´1, which is in agreement with the reported literature [47,50]. on the surface of the electrode [55]. It resulted that there may not be sufficient active sites for the adsorption of higher concentrations of glucose. In addition, some other chemical complications, such as the glucose intermediate occurring, could be a barrier to reach the catalytic surface of the electrode for the diffusion of glucose molecules. A high concentration of OH − might be required for the oxidation of a high concentration glucose, and this part of work is under progress along with this study [50,52]. The average value of the rate constant for the concentration range of 0.5 mM to 2 mM glucose solution was 57.65 M −1 ·s −1 , which is in agreement with the reported literature [47,50].  on the surface of the electrode [55]. It resulted that there may not be sufficient active sites for the adsorption of higher concentrations of glucose. In addition, some other chemical complications, such as the glucose intermediate occurring, could be a barrier to reach the catalytic surface of the electrode for the diffusion of glucose molecules. A high concentration of OH − might be required for the oxidation of a high concentration glucose, and this part of work is under progress along with this study [50,52]. The average value of the rate constant for the concentration range of 0.5 mM to 2 mM glucose solution was 57.65 M −1 ·s −1 , which is in agreement with the reported literature [47,50].   Figure 6a shows the CVs of the 2 µL HCNiO (Г = 1.45ˆ10´1 0 mol/cm 2 )-coated GC electrode recorded in different glucose concentrations (0.10 mM to 2 mM) in 0.05 M NaOH solution at a 50 mV/s sweep rate. The linearity plot obtained for the current versus concentration with R 2 = 0.99 from the CV data is shown in the inset (Figure 6a'). The catalytic current of a HCNiO/GC electrode increased towards the glucose oxidation and the enhancement of the anodic peak current attributed to the profound conductivity and large active surface area of the material coated on the electrode.  Figure 6a shows the CVs of the 2 µL HCNiO (Г = 1.45 × 10 −10 mol/cm 2 )-coated GC electrode recorded in different glucose concentrations (0.10 mM to 2 mM) in 0.05 M NaOH solution at a 50 mV/s sweep rate. The linearity plot obtained for the current versus concentration with R 2 = 0.99 from the CV data is shown in the inset (Figure 6a'). The catalytic current of a HCNiO/GC electrode increased towards the glucose oxidation and the enhancement of the anodic peak current attributed to the profound conductivity and large active surface area of the material coated on the electrode.

Amperometric i-t Detection of Glucose
The lower quantity of glucose oxidation can be achieved using i-t method, in which potential at 0.52 V was selected. Figure 7a,b show the amperometric sensing of glucose by successive addition of a low range of glucose concentration of 0.2 to 10 µM in a continuously-stirred 0.05 M NaOH solution. Figure 7c shows its two different calibration plot. Moreover, Figure 8a showed the current response for a higher concentration range of 0.5 to 500 µM glucose, in which the current increment was limited up to 250 µM. The current response decreased gradually after 250 µM concentration, which may be due to the diffusion dominance and adsorption difficulties of glucose ions onto the small volume of the active sites when compared to the high glucose concentration. Figure 8a' is the linearity plot of 0.5 to 250 µM. The fast steady-state current achieved (95% of all within 3 s) and stable response upon the successive additions of 10 µM glucose with the reliable RSD value (1.14% ± 4.32%) for the 13 additions is shown in Figure 8a". The limit of detection (LOD) was 0.9 µM measured by the standard deviation of the blank run in 0.05 M NaOH by the i-t technique (n = 10) and its signal was also observed by this method. The HCNiO/GC sensor exhibits much higher sensitivities of 3978.9 and 1232.4 µA mM −1 cm −2 for the linear ranges 0.2 to 3.5 µM and 0.5 to 250 µM, respectively, compared to other nickel-based GCEs reported in the literature [23]. The analytical parameters are listed in Table 1 for a comparison study. The stability and reproducibility of the sensor checked by tests every week, and a number of experiments on the same day, indicated that there was almost no change in current density for more than two weeks' time and up to 25 tests on the same day.

Amperometric i-t Detection of Glucose
The lower quantity of glucose oxidation can be achieved using i-t method, in which potential at 0.52 V was selected. Figure 7a,b show the amperometric sensing of glucose by successive addition of a low range of glucose concentration of 0.2 to 10 µM in a continuously-stirred 0.05 M NaOH solution. Figure 7c shows its two different calibration plot. Moreover, Figure 8a showed the current response for a higher concentration range of 0.5 to 500 µM glucose, in which the current increment was limited up to 250 µM. The current response decreased gradually after 250 µM concentration, which may be due to the diffusion dominance and adsorption difficulties of glucose ions onto the small volume of the active sites when compared to the high glucose concentration. Figure 8a' is the linearity plot of 0.5 to 250 µM. The fast steady-state current achieved (95% of all within 3 s) and stable response upon the successive additions of 10 µM glucose with the reliable RSD value (1.14%˘4.32%) for the 13 additions is shown in Figure 8a". The limit of detection (LOD) was 0.9 µM measured by the standard deviation of the blank run in 0.05 M NaOH by the i-t technique (n = 10) and its signal was also observed by this method. The HCNiO/GC sensor exhibits much higher sensitivities of 3978.9 and 1232.4 µA mM´1 cm´2 for the linear ranges 0.2 to 3.5 µM and 0.5 to 250 µM, respectively, compared to other nickel-based GCEs reported in the literature [23]. The analytical parameters are listed in Table 1 for a comparison study. The stability and reproducibility of the sensor checked by tests every week, and a number of experiments on the same day, indicated that there was almost no change in current density for more than two weeks' time and up to 25 tests on the same day.    GC: glassy carbon; rGO: reduced graphene oxide; MWCNT: multi-walled carbon nanotube; GNS: graphene nanosheet; DNA: deoxyribonucleic acid; LOD: limit of detection.
The interference test was also done to check the selectivity. The main interferents and the same oxidation potential for glucose detection in serum or other biological samples are dopamine (DA), ascorbic acid (AA), and uric acid (UA). Figure 9a expressed the selectivity by the presence of interferences of AA, UA, and DA for glucose detection. This sensor shows a better response to glucose than other interferents including AA, DA, and UA, which demonstrated good selectivity. The good result was obtained for real samples in 0.05 M NaOH, as shown in Figure 9b. This sensor could be used for the screening purposes without the invasion for the patients who suffer diabetes mellitus. The interference test was also done to check the selectivity. The main interferents and the same oxidation potential for glucose detection in serum or other biological samples are dopamine (DA), ascorbic acid (AA), and uric acid (UA). Figure 9a expressed the selectivity by the presence of interferences of AA, UA, and DA for glucose detection. This sensor shows a better response to glucose than other interferents including AA, DA, and UA, which demonstrated good selectivity. The good result was obtained for real samples in 0.05 M NaOH, as shown in Figure 9b. This sensor could be used for the screening purposes without the invasion for the patients who suffer diabetes mellitus.

Reagents
NiCl 2¨6 H 2 O, NaOH, urea, methanol, and ethanol were obtained from Choneye Pure Chemicals (Taipei, Taiwan), yeast from Becton Dickinson Biosciences (Taipei, Taiwan), ammonium sulfate, Tris-HCl were purchased from J.T. Baker (Chu-Bei, Taiwan), glucose from Sigma-Aldrich China Inc. (Shanghai, China), and Milli-Q (MQ-18.2 MΩ¨cm at 25˝C) water were used. All chemicals were used without further purification, analytically pure, and all electrochemical experiments were carried out at ambient temperature unless otherwise stated anywhere in this paper.

Electrochemical Measurements
Cyclic voltammetry studies were performed on a CHI627C Electrochemical Analyzer (CH Instruments, Inc., Austin, TX, USA). The conventional three-electrode system was used throughout the electrochemical experiments consisted of a bare GCE (geometric area 0.071 cm 2 ) or a modified GCE as the working electrode, a platinum wire as the auxiliary electrode, and Ag/AgCl (3 M NaCl) as the reference electrode against which all potentials were measured in this paper. For steady-state amperometric measurements, the working potential was set at 0.52 V and the solution was stirred gently with a magnetic stirrer. A digital pH meter (SUNTEX TS-1, Suntex Instruments Co., Ltd., Xinbei, Taiwan) for pH measurements and a personal computer were used for data storage and processing.

Preparation of Hollow Cylinder NiO Nanostructured Material
The synthesis procedure followed the method developed by our group [57]. Briefly, a homogeneous mixture was prepared from 0.1 M NiCl 2¨6 H 2 O, 20 mL MQ water, 1.0 M urea solution, and 40 mL bacterial broth in a screw-capped Teflon tube. Bacterial culture of Sporosarcina pasteurii was carried out using 20 g/L yeast extract, 10 g ammonium sulfate, and 20.48 g Tris-HCl (pH of 8.5) in 1 L sterilized MQ water. The tube was then kept for one day in a mechanical shaker at 35˝C. The obtained precipitate was separated by centrifugation at 3700 rpm for 15 min. Then the precipitate was washed several times with water and, consecutively, with ethanol, dried in an air oven for 6 h at 50˝C, followed by calcination at 550˝C for 6 h. The green precipitate of bio-inorganic nickel compound changed to a black color. The hollow cylinder of microbacterial shape with nanocell wall-like structured NiO (HCNiO) compound was confirmed with the help of transmission electron microscopy (TEM) and scanning electron microscope (SEM) (see Figure 10 for TEM and SEM of bacterial shaped HCNiO).
Tris-HCl were purchased from J.T. Baker (Chu-Bei, Taiwan), glucose from Sigma-Aldrich China Inc. (Shanghai, China), and Milli-Q (MQ-18.2 MΩ·cm at 25 °C) water were used. All chemicals were used without further purification, analytically pure, and all electrochemical experiments were carried out at ambient temperature unless otherwise stated anywhere in this paper.

Electrochemical Measurements
Cyclic voltammetry studies were performed on a CHI627C Electrochemical Analyzer (CH Instruments, Inc., Austin, TX, USA). The conventional three-electrode system was used throughout the electrochemical experiments consisted of a bare GCE (geometric area 0.071 cm 2 ) or a modified GCE as the working electrode, a platinum wire as the auxiliary electrode, and Ag/AgCl (3 M NaCl) as the reference electrode against which all potentials were measured in this paper. For steady-state amperometric measurements, the working potential was set at 0.52 V and the solution was stirred gently with a magnetic stirrer. A digital pH meter (SUNTEX TS-1, Suntex Instruments Co., Ltd., Xinbei, Taiwan) for pH measurements and a personal computer were used for data storage and processing.

Preparation of Hollow Cylinder NiO Nanostructured Material
The synthesis procedure followed the method developed by our group [57]. Briefly, a homogeneous mixture was prepared from 0.1 M NiCl2·6H2O, 20 mL MQ water, 1.0 M urea solution, and 40 mL bacterial broth in a screw-capped Teflon tube. Bacterial culture of Sporosarcina pasteurii was carried out using 20 g/L yeast extract, 10 g ammonium sulfate, and 20.48 g Tris-HCl (pH of 8.5) in 1 L sterilized MQ water. The tube was then kept for one day in a mechanical shaker at 35 °C. The obtained precipitate was separated by centrifugation at 3700 rpm for 15 min. Then the precipitate was washed several times with water and, consecutively, with ethanol, dried in an air oven for 6 h at 50 °C, followed by calcination at 550 °C for 6 h. The green precipitate of bio-inorganic nickel compound changed to a black color. The hollow cylinder of microbacterial shape with nanocell wall-like structured NiO (HCNiO) compound was confirmed with the help of transmission electron microscopy (TEM) and scanning electron microscope (SEM) (see Figure 10 for TEM and SEM of bacterial shaped HCNiO). (a)

HCNiO/GC Electrode Preparation and Its Activation by CV Cycle
Prior to each experiment, the GCE was first polished with gamma alumina in water slurry using a polishing cloth and rinsed thoroughly with MQ water and ethanol. Then the required amount of HCNiO material was ultrasonically dispersed in methanol:water (20:1) solution to achieve a 10 mg/mL uniform ink. Finally, 2 µL of the ink was drop-casted onto the GCE and dried before the electrochemical experiments to study the properties of HCNiO/GCE. Then the electrode (HCNiO/GC) was conditioned between +0.2 to +0.7 V in 0.05 M NaOH, to attain the stable, well-defined peaks of Ni(OH)2/NiOOH film on the modified GCE by cyclic voltammetry, and the optimum 54 cycles developed as in Figure 1; its characteristic peak potentials were identified with the reported literature [10,55].

Conclusions
A modified GCE with bacteria-template HCNiO was successfully employed for the enzymeless detection of glucose in basic solutions. Electrochemical characteristic parameters were obtained for this well-developed redox couple on the HCNiO/GC by CV method. The detection of glucose was performed by CV and amperometric i-t techniques and LOD was found as 0.9 µM. The excellent electrocatalytic activity of HCNiO/GC sensor was stable, reproducible, and sensitive towards the detection of glucose. It is noticed that the electrode was stable for the wide range of potential scan rates up to 5 V/s with the linearity curves. Limitation of second order rate constant for the higher concentration of glucose solution was observed. The low cost of the preparation and high catalytic ability of the oxyhydroxide (NiOOH) were achieved for the electro-oxidation of glucose.