L-Alanine Supported Autogenous Eruption Combustion Synthesis of Ni/NiO@RuO2 Heterostructure for Electrochemical Glucose and pH Sensor

Safety and quality control are important for long-term storage and preservation of food. Glucose and food pH are the two most common markers for evaluating food quality. Herein, we constructed a Ni/NiO@RuO2 heterostructure-based two-way sensor via a novel eruption combustion pattern (ECP) using non-conventional amino acid as a propellant. This approach has the unique points of interests of in situ doping of oxides and the formation of heterojunctions, providing well-developed pores and high surface areas to enhance the material performance. The Ni/NiO@RuO2 heterostructures have been tested as a bi-functional catalyst for glucose and pH sensing. The sensor exhibits a fast response time of <0.1 ± 0.02 s, a sensitivity of 641.95 ± 0.5 μA mM−1 cm−2 towards glucose with a 0.4 ± 0.08 μM detection limit and a linear response of 0.1 to 5 mM. As a pH sensor, it exhibits an acceptable sensitivity of −41.6 mV pH−1 with a response time of <50 s over a pH range of 2–12. Moreover, this bi-functional sensor based on Ni/NiO@RuO2 performs well when applied to a selection of beverage samples. This study provides a new scalable and low-cost approach to fabricating hetero-oxide nanostructures with controllable heterojunctions for various sensor applications.

Growing global food supply chains require vigilance to monitor food safety as mass production in short periods of time and demand for safer, healthier food can be stored for longer periods of time. 1,2 Glucose is one of the primary ingredients in fermented food or beverages, and its real-time quantification is essential for monitoring the quality and safety of food production. 3 Consumers with diabetes have questions and concerns about whether it is safe for them to eat or drink food which contains a large amount of glucose in the form of sugar. Therefore, the amount of glucose is both a quality control indicator in the food market and a health concern for diabetic end users. 4,5 The sensitive and selective glucose detection is important in fermentation technology, food industry and clinical diagnostics. 3 In food processing, the pH value contains the hallmark of protein denaturation, gelification, microorganism growth and mortality, and the germination or inactivation of bacteria spores. 6 On the other hand, pH monitoring of food and beverages is of critical importance to safeguard consumer safety in line with regulatory requirements, necessitating accurate, reliable and costeffective pH measurement technology. Measuring the pH of food and beverages also provides important insight into product quality, consistency and safety. Current food safety monitoring techniques such as chromatography, mass spectrometry, and fluorescence detection are used only at the end of the food production process. 7 These techniques have certain limitations including safety measures at the end of the process would be expensive when safety concern arises, requiring a large volume of samples, and a sophisticated laboratory with trained personnel. 5,8 Recently, electrochemical biosensors have gained attention to monitor the food quality and ensure safety. 9 However, research on the electrochemical biosensor for food analysis is limited and needs to be explored, especially for real-time glucose detection and pH monitoring in beverages.
The electrochemical sensors measure the electric signals from biomolecules using transduction components by either potentiometric or amperometric techniques. [9][10][11] The detection technique of choice can be specified based on the electrochemical properties of the particular electrode surface. Since their invention in 1962, glucose oxidase enzyme-based and non-enzymatic electrochemical sensors have been widely used in diabetes management. [11][12][13] Because of its rapid and reliable measurement, current research is focused on monitoring food safety using an electrochemical sensor approach. 14,15 Metal oxides with unique electrochemical and biocompatible properties are used to make sensors for food quality monitoring and industrial applications. Also, the solid state metal oxide pH sensors have an advantage over typical glass pH sensors since they are compatible with microfabrication processes. 16 To date, numerous glucose and pH sensors were fabricated based on the noble metals (i.e. Pt, Pd, and Au), and transition metal compounds such as Co/Co 3 O 4 , Cu/CuO, Ru/RuO 2 , IrO 2 , Ni/NiO, WO 3 , ZnO, etc. 15,[17][18][19][20] Concerning Earth abundance, biocompatibility, exclusive redox activity, high stability, and low cost, Ni-based electrodes are promising and suitable alternative to replace conventional noble metals for electrochemical sensor fabrication. 19 Furthermore, NiO's high isoelectric point (IEP = 10.7) is advantageous for binding low IEP biomolecules via strong electrostatic interaction. 21 In this context, mixed transition metal compounds possess additional beneficial properties and have established multifunctionality in various applications including catalysts, sensors, supercapacitors, and others. 20 As food/beverages are complex mixtures, choosing the right combination of electrode materials is a crucial part for the successful detection of glucose in mixtures of inorganic and organic compounds. Interesting properties such as high metallic conductivity, stability in various oxidation states and fast electron transfer kinetics offer RuO 2 great potential as an active component in the development of electrochemical sensors. 22 Constructing RuO 2 heterostructure along with Ni 2+/3+ can introduce distinct reactive sites for catalytic activity. It can also efficiently modulate conduction channels to enhance sensing performance via synergistic effects. 23 Until now, NiO-RuO 2 -based hetero-composites have exhibited enhanced functional activity in oxygen-hydrogen evolution reactions (OER and HER) and energy storage devices applications. 24,25 To our knowledge, there are no reports based on the Ni/NiO-RuO 2 heterostructures for direct glucose and pH sensing applications.
Fabrication of NiO-RuO 2 heterostructures remains challenging owing to their diverse crystalline structure, and the high cost of the Ru component. So far, different synthetic approaches such as z E-mail: kafil.mahmood@tyndall.ie hydrothermal, solvothermal and wet chemical methods have been explored to synthesize different types of heterostructures for glucose detection. For example, Huan Wang et al. fabricated an electrochemical sensor with the NiO-coated CuCo 2 O 4 nanoneedle arrays via the hydrothermal method for non-enzymatic glucose detection and showed high sensitivity of 4140 μA mM −1 cm −2 . 26 Wang and co-workers designed electrospun IrO 2 @NiO core-shell nanowires and investigated their enhanced sensing towards glucose detection. 27 Meanwhile, Lonsdale et al. proposed RuO 2 based solid-state pH sensor for common beverages and attain a sensitivity of 55.3 mV pH −1 . 16 However, inexpensive and scalable synthetic processes are still needed to reduce expensive catalyst costs. The globally-renowned solution combustion (SC) process is highly recommended as it is fast, simple, low-cost, and energy efficient and requires low temperature to synthesize oxide materials. 28,29 Since fuel plays an important role in combustion, the widely accepted organic compounds of urea, citric acid, glycine and hydrazides are used as fuels. [29][30][31] However, the production of noxious gases when fuel ignited must be minimized, which requires environmentally-friendly redox fuels. Amino acids (AA) with carboxylic and amino functional groups provide homogeneous nanoparticle growth with low gas generation when used as fuels in solution combustion synthesis. 32 Among them, the commonly used fuel is glycine comprising -COOH and -NH 2 functional groups for combustion reaction. However, it needs to be further calcined at high temperature due to the formation of carbon impurities in the final product. 30 With the exception of glycine, there is limited research on various amino acid fuels that require further study. In this work, we propose for the first time an easy, and unique L-alanine-activated eruption solution combustion process to fabricate NiO-RuO 2 heterostructures and investigate for the application of non-enzymatic glucose and pH sensors.

Experimental
Reagent grade of Nickel II nitrate hexahydrate (Ni(NO 3 ) 2 ·6H 2 O), Ruthenium III chloride hydrate (RuCl 3 ·xH 2 O), L-alanine (C 3 H 7 NO 2 ), D-glucose (C₆H₁₂O₆), Sodium hydroxide (NaOH), Dopamine (C 8 H 11 NO 2 ), Sodium Chloride (NaCl), Potassium Chloride (KCl), Uric acid (C 5 H 4 N 4 O 3 ), Urea (CH₄N₂O), and Ascorbic acid (C 6 H 8 O 6 ) were purchased from Sigma Aldrich and used without further process. For the solution combustion process, the oxidant to fuel ratio was kept at 1:2. At first, 5 g metal salt equivalent molar of C 3 H 7 NO 2 was dissolved in 25 ml DI water under constant stirring at 40°C for 1 h. Once the amino acid dissolved completely, the Ni(NO 3 ) 2 ·6H 2 O (5 g) was added to the fuel solution. The temperature of the solution was raised to 75 ± 2°C and continued stirring until a highly viscous gel formed. Since the combustion is eruptive, all the mandatory safety precaution was followed during synthesis. Within a few seconds, the gel got ignited and a high flame eruption occurs. Immediately, the convulse was quenched with the addition of 100 ml DI water and ash was decanted by centrifugation at 5000 rpm for 10 min several times with DI water, ethanol and acetone consecutively. The final product was dried at room temperature overnight and directly used for analysis. To synthesise a NiO-RuO 2 heterostructure, Ni: Ru/4.5 g: 0.5 g metal salts were taken and a total 5 g equivalent molar of amino acid was used as the fuel. The remaining experimental conditions were kept the same. Scheme 1 represents the possible redox combustion reaction.

Materials Characterisation
Empyrean Malvern Panalytical X-ray powder diffractometer was used to collect X-ray diffraction patterns with X'pertPro monochromatized Cu Kα radiation (λ = 1.5403 Å). The Raman spectra were collected using a RENISHAW inVia Raman microscope with 514 nm laser excitation. FEI Quanta 650 (UK) Scanning Electron Microscope with an energy-dispersive X-ray analysis (EDS) was used to capture surface morphology and elemental mapping. JEM-2010 microscope (JEOL, Japan) working at 200 kV was employed to obtain transmission electron microscopy images. The Kratos AXIS ULTRA spectrometer was used for X-ray photoelectron spectra measurements with the Al Kα (1486.58 eV) mono-chromatic X-ray at a power of 300 W (20 mA, 15 kV). Construction and peak fitting of synthetic peaks in narrow region spectra used a Shirley-type background and the synthetic peaks were of a mixed Gaussian-Lorenzian type. Relative sensitivity factors used are from the CasaXPS library containing Scofield cross-sections. The C 1 s line at 284.8 eV was used as a charge reference for the peak position correction.

Electrochemical Characterisation
Electrode preparation.-The Ni/NiO and Ni/NiO@RuO 2 modified carbon paper electrode was prepared by mixing 5 mg of catalysts in 2 ml of DI water and Ethanol (1:1) solution containing 20 μl of Nafion (5%) binder. The above mixture was sonicated for 2 h to obtain a homogenous black suspension. Next, 10 μl of catalyst suspension drop cast into 20 mm × 5 mm conductive Toray carbon fibre paper (FuelCell Store, USA, 80 mΩ cm ) of thickness 370 μm. Followed by drying at room temperature overnight the samples were used as working electrodes in sensing applications.
Fabrication of glucose sensor.-The electrochemical analysis was carried out in the CHI600E (CH Instruments, USA) electrochemical workstation. The cyclic voltammetry (CV) measurements were performed with and without the addition of glucose at different scan rates under continuous stirring in a three-electrode configuration, where Ni/NiO or Ni/NiO@RuO 2 modified carbon fibre paper served as the working electrode, platinum wire and Ag/AgCl electrode were connected as the counter and reference electrodes, respectively. The N 2 -saturated 0.1 M KOH solution was used as the supporting electrolyte. The chronoamperometry (CA) measurement was performed at consecutive addition of glucose at its oxidative potential of 0.5 V.
Fabrication of pH sensor.-The pH measurement was performed in SP 300 potentiostat (Biologic, France) with a different pH range of 2-12 at room temperature. The metal oxide solid-state pH sensor was fabricated by drop cast 5 μl of catalyst on a glassy carbon working electrode of 3 mm diameter and dried at room temperature overnight (CHI104, CH Instruments, USA). An open circuit voltage (OCV) was measured for 1500 s, and each result was triplicated to maintain accuracy. The pH was accurately determined using a standard laboratory pH meter (pH5+, Eutech Instruments, calibrated with standard pH 7 and 4). The universal Britton-Robinson buffer (0.04 M H 3 BO 3 , 0.04 M CH 3 COOH and 0.04 M H 3 PO 4 ) solution was prepared with a typical pH value of 1.16 ± 0.02 and 0.2 M NaOH solution was used to adjust the pH from 2 to 12.

Results and Discussion
Amino acids with R-group side chains have specific interests because of their chemical activity and electron donor availability to control the particle size distribution and morphology of nanostructure in the solution combustion (SC) process. 28,33 The typical SC process is an exothermic redox-type reaction and leads to the release of a huge amount of gases during the reaction. 30 The reaction of L-alanine showed an eruptive nature with a rapid increase in pressure. The combustion reaction between Ni(NO 3 ) 2 ·6 H 2 O and L-alanine (C 3 H 7 NO 2 ) is proposed in Scheme 1. Initially, the formation of bi-dentate metal ligands followed dryness and fragmentation under combustion. According to Simmonds et al., the major decomposition product of L-alanine is amine due to the primary decarboxylation process. 34 Subsequent fragmentation reaction of amine leads to the formation of secondary products including nitriles and alkylaldimines. 35 By the loss of two hydrogen ions, the amine is fragmented into corresponding nitriles (CH 3 CN), which are highly flammable. Thereby, an eruptive fire was observed during combustion for a few seconds. On the other hand, the amount of oxygen in the environment is more important to control the crystallinity and phase of the metal oxide nanoparticles. Generally, Ni(NO 3 ) 2 ·6H 2 O decomposition is endothermic and forms the gas phase HNO 3 followed by NiO as the end product as reported by Kumar et al. 36 The proposed combustion of Ni-based materials in presence of L-alanine showed the formation of a metallic nickel-rich NiO phase. This can be explained by the reaction between the gas phase products of HNO 3 and NH 3 in the mixed solution. This reaction is the driving force for the solution combustion process between nickel nitrate and L-alanine precursors, which creates excessive pressure into the solution and the abrupt release of H 2 O, CH 3 CN, N 2 , and H 2 gases leads to insufficient oxygen in the environment. Therefore, the metallic Ni phase is more predominant than the face-centred cubic NiO. However, in the case of Ni/NiO-RuO 2 heterostructure, the formation of HCl in the mixture limits the HNO 3 and NH 3 reaction rate and maintains surplus oxygen access in the environment thus forming the corresponding oxides phases.
The phase and crystallinity of Ni/NiO and Ni/NiO@RuO 2 heterostructure catalysts are examined using XRD analysis. Figure 1a shows the XRD patterns of as-synthesized catalysts where well-resolved diffraction peaks confirm the existence of metallic Ni and NiO phases. All the peaks are indexed to standard JCPDS 065-0380 and JCPDS 065-5745 for Ni and NiO, respectively. The diffraction peaks with high intensity originating at 44 Fig. 1a inset) at 28.2°, (110) 35.2°(101) and 54.4°(211) resemble tetragonal hydrated RuO 2 and can be indexed to JCPDS # 88-0323. 40 The absence of peaks related to carbon or residues in the XRD pattern indicates the complete combustion process. Thereby, the XRD analysis confirms that the samples are composed of mixed Ni/NiO and Ni/NiO@RuO 2 phases. To further confirm the superposition of the Ni, NiO and RuO 2 crystal structures, the Raman spectra shown in Fig. 1b are analysed. A typical Raman spectrum of Ni/NiO shows one phonon (1P) and two phonons (2P) bands at 200-600 cm −1 , and 650-1100 cm −1 , respectively, which are the characteristics of nanocrystalline Ni-O bond vibrations. 38,39 There is no signal for FCC metallic Ni due to the negligible change in polarizability during light-matter interaction. The one phonon-related bands at 206.5, 367.3 and 546.8 cm −1 can be assigned to the zone boundary phonon mode, transverse (TO), and longitudinal (LO) optical modes, respectively. 41 The corresponding 2P-related transverse and longitudinal modes are visible at 693.9 and 1052.2 cm −1 . The intensity of the 2 P bands is relatively high than the 1 P band due to surface disorder or defects. Considering Ni/NiO@RuO 2 heterostructure, the Raman spectrum shows only two broad bands related to 1 P (LO) and 2 P (LO) at 557.3 and 1068.4 cm −1 , respectively for Ni-O bonds. 41 There are no discernible Raman active modes due to crystalline RuO 2 , which should be observed at about 500-700 cm −1 because these bands are super-positioned with Ni-O bands. 42 The absence of low wavenumber bands <500 cm −1 , reveals the formation of low-order crystalline Ni/NiO@RuO 2 hetero-nanostructure.
The specific surface area and porosity of the catalysts were evaluated from nitrogen adsorption-desorption isotherm as shown in Fig. 2. Both nanostructures exhibit type IV isotherm with H3 hysteresis loop, which shows the monolayer-multilayer adsorption followed by capillary condensation uptake at larger pressure. 43 The pristine Ni/NiO showed well-resolved hysteresis when compared to the Ni/NiO@RuO 2 heterostructure, indicating the loose stacking of constituent nanoparticles. The Brunauer-Emmett-Teller (BET) specific surface area of nanostructures are found to be 81.2 and 21.9 m 2 g −1 for Ni/NiO and Ni/NiO@RuO 2 respectively. Our Ni/ NiO heterostructure exhibits a high surface area, which is nearly four times higher than the best-reported values for Ni/NiO synthesized by the eruptive combustion process. 44 The decrease in BET surface area for Ni/NiO@RuO 2 heterostructure indicates that the incorporation of RuO 2 significantly restacks the crystalline structure and gets aggregated, which reduces the pore volume and specific surface area. 45 The corresponding pore size distribution curves are shown in Fig. 2b, revealing the presence of a wide range of mesopores in the samples. Ni/NiO sample reveals a narrow mesoporous structure, whereas Ni/NiO@RuO 2 has disordered mesopores due to more aggregation between the nanoparticles. These results further confirmed that both samples have a large volume of mesopores (2-50 nm), which is highly desirable for superior electrochemical activity. 43 According to Zhang et al. 46 high surface-to-volume gives a large number of surface sites for the adsorption of H + ions, resulting in strong pH sensitivity and a high correlation coefficient with a large surface area. The porous structure, on the other hand, enhances the reaction between H + and electrode, improving the electrochemical properties. Based on the electrical double layer model, increasing the percentage of oxygen in metal/metal oxide surface promotes the pore formation, thus providing larger sensing surface volume. 47 To further explore the surface morphology and elemental distribution of the heterostructures, SEM and EDS mapping studies were performed. Figure 3a displays the as-synthesized Ni/NiO, which is composed of spherical-shaped Ni and NiO nanocrystallites assembled and form Ni/NiO hetero-aggregates. The Ni/NiO nanoparticles are loosely bounded and exhibit a porous structure. An inclusion of RuO 2 generates excess aggregates (Fig. 3c) during the solution combustion process due to more inter-diffusion and surface boundary variations. 48 In both samples, the particles are distributed in the range of 3-10 nm, which is in good agreement with the cumulative particle size observed in BET analysis. The corresponding EDS elemental mapping images shown in Figs. 4b & 4d, envisage an obvious dense Ni and O distribution along with Si from the substrate, which reveals a thick NiO layer is covered over the Ni core. In the case of Ni/NiO@RuO 2 heterostructure, Ni, O and Ru elements are uniformly distributed and a denser distribution is visible because of the formation of NiO and RuO 2 hetero-phase   25,40 In the vicinity of the selected region ( Fig. 4e inset), there are some nanoparticles less than ∼5 nm that are visible close to the RuO 2 surface. It shows a lattice spacing of 0.20 nm and corresponds to the (1 0 1) plane of metallic Ru. 49 This is due to the dehydration of hydrous RuO 2 , which formed RuO 2 and Ru metal nanoparticles. It is found that the RuO 2 nanoparticles are distributed over the Ni/NiO matrix. This result is further verified with selected area electron diffraction (SAED) images as shown in Figs. 4c & 4f). There is a distinct diffraction ring (Fig. 4c) corresponding to Ni and NiO planes, confirming the formation of Ni/NiO hybrid nanostructure. Whereas, for Ni/NiO@RuO 2 heterostructure, the observed SAED pattern is close to the amorphous structure and shows diffraction spots for RuO 2 planes. 50 These results further conclude that Ni/NiO is in low-range crystalline order along with RuO 2 nano-crystallites. It is to be noted that, this type of mesoporous hybrid hetero-nanostructures favours rapid charge and ion transfer during the electrochemical redox process.
The valence states and binding energy (BE) of the constituent elements in the as-prepared hybrid heterostructures are studied using XPS analysis. As shown in Fig. 5a, the XPS survey spectra pin on the Ni, Ru, O, and N elements, which is consistent with the EDS mapping results. In this fabrication process, the amount of Ru is relatively low, which can reduce the overall cost of the catalyst. Considering the actual oxidation state of metals and their chemical environments, the individual core-level electron emissions are measured. As in Fig. 5b, the XPS spectra of Ni 2p electrons from both hybrid nanostructures showed a high binding energy shift from their typical chemical states. This may be due to the eruptive synthesis strategy playing an important role to decide the surface chemical environments. Since the proposed reactions involve both endo/exo-thermic redox processes, the reactant and product interactions lead to varying the metal/metal oxide surface significantly. According to Tomellini, the high energy shift is due to a change in the ionic charge and oxygen coordination induced by cation vacancies. 51 An observed, Ni 2p 3/2 binding energy values are 855.9 and 855.5 eV for Ni/NiO and Ni/NiO@RuO 2 respectively. It is noteworthy that Ni 2p3/2 spectra did not show a metallic Ni signal from samples in contrast to XRD results. The above BE values are well-defined for the hydroxylated Ni/NiO surfaces. 52 Therefore, it can be ascertained that the Ni/NiO surface is covered with chemically reacted hydroxyl groups. The formation of hydrated surfaces can be explained by the surface defects as predicted earlier. [53][54][55] Hydroxyl groups on Ni/NiO surfaces may be formed by the interaction of H 2 O at the defect sites of NiO. Under residual gas atmosphere, the loss of surface oxygen leads to the formation of metallic Ni covered with hydroxylated NiO. 56 The possible reaction is as follows: 57 According to a hypothesis by Zhao et al., to cleave the oxygen from the NiO surface and to form H 2 O requires at least three hydrogen atoms. Because the adsorbed hydrogen trimer can exothermally drag out one oxygen from the NiO surface to form H 2 O. 53 The H 2 O vapour dissociatively adsorbs rapidly on the NiO surface, forms -OH groups following Eq. 3 and leave H ad to adsorb by Ni atom. This means that the formation of Ni(H ad ) and NiO(−OH) is a twostep process according to 1-3, in which reduction is followed by hydrogenation. The formed "Ni 2+ " acts as a process catalyst and promotes more H 2 dissociative ionization over the NiO surface by creating oxygen vacancies (V o ). 58 It can be concluded that our catalyst surface is hydroxylated and thereby the recorded binding energies are higher than the previous reports, suggesting the formation of defective Ni(H ad )/NiO(−OH) or disorder structure. This conclusion is further confirmed by the low-intensity satellite peaks in the Ni 2p XPS spectra. 56 In addition, Fig. 5c shows the Ru 3d core level XPS spectra of Ni/NiO@RuO 2 heterostructure. As can be in Fig. 5c, Ru 3d photoelectron emission is identified as Ru/RuO 2 structure with higher binding energy shift (∼0.5 eV) due to H 2 O interaction as similar to Ni/NiO. The corresponding RuO 2 , 3d 5/2 and 3d 3/2 spin-orbit splitting are found to be 281.5 and 285.7 eV, respectively. 59 By the peak fitting, the near equal intensity of Ru 3d photoelectron emission peaks was found at 280.4 and 284.6 eV, which is concomitantly matched to pure Ru. Also, the 3d 5/2 and 3d 3/2 peaks are separated by approximately 4.2 eV for both (Ru & RuO 2 ) cases further revealing the formation of Ni/NiO@RuO 2 heterostructure. This result is corroborated by the observation that in Ni/NiO, the metallic Ru formation is due to the loss of oxygen at the wateradsorbed RuO 2 surface under residual gas atmosphere during the exothermic combustion process. 60,61 To support the above argument, Ru3p peaks are further deconvoluted into two bands at 461.2 and 463.7 eV, respectively for RuO 2 and Ru-OH as shown in Fig. S2 in SI. 62 As the C 1 s emission strongly interferes with the Ru 3d 3/2 component, the intensity ratio between 3d 5/2 to 3d 1/2 is 1.49 (≈1.5), which reveals that there is a trace of carbon contaminants in the samples. 59 The Peaks at 284.7, 286.2, 287.8 and 289 eV are assigned to binding energies of C 1 s orbitals related to C-C/C=C, C-N/C-O, C=O and O-C=O, respectively. The existence of N 1 s peaks as shown in Fig. 5d further supports the trace of carbon and nitrogen components in the samples. Both the samples show an N 1 s emission at 399.1 and 400.5 eV, which may be due to the C-N and N-O bonding, which is mostly contributed by chemisorbed N 2 . 63 The surface hydrogenation and hydroxylation processes are further validated in terms of core level O 1 s XPS spectra as shown in Fig. 5e. In these Ni/NiO and Ni/NiO@RuO 2 hybrid heteronanostructures, a broad O 1 s peak is observed and confirmed the formation of hydrated oxides. 64 In Ni/NiO, typical O 1 s bands are deconvoluted into three peaks located at 529.9, 531.7 and 533.1 eV may be attributed to lattice oxygen (Ni-O), hydroxylated oxygen (NiO-OH @ 855.9 eV) 65,66 and internally bounded non-dissociated water (Ni-O-H 2 O) or carbonaceous species. 67 Comparing with the spectra of Ni/NiO, the O 1 s peaks slightly shifted (≈0.3 eV) to lower binding energy for Ni/NiO@RuO 2 . The corresponding peak intensity located at 531.4 and 532.9 eV increases due to the superposition of (Ru-OH-Ni-O) hydroxylation and formation of RuO 2 ·H 2 O. 68 These results are in good agreement with XRD and TEM results. Consistent with XPS results, valence band (VB) spectra shown in Fig. 5f provide additional information for the formation of hetero-nanostructures. The predominant Ni-3d orbital state is found at ∼2.7 eV, which is ∼0.7 eV higher than the NiO with oxygen vacancy, further supporting the hydroxylation of the Ni/NiO surface. 69 This can be assigned to the photoionization of divalent nickel 3d 8 final states. The peaks due to O 2p are less pronounced at around 8 eV and may be due to the limited photoionization crosssection of the Al Kα (1486.7 eV) source. It suggests the multielectron process in Ni 2p 3/2 orbitals. 70 Also, the peak above ∼2 eV is visible for both the samples, which are assigned for O 2 s bands. For Ni/NiO@RuO 2 , the Ni 3d band intensity decreased significantly because of hybridization with an outer Ru 4d 4 state. 71 The wellpronounced Ru metal 4p peak is discernible at 44.3 eV along with Ni 3p (68 eV) and O 2 s levels further substantiating the formation of heterostructure. 72 The binding energies in the valence band spectra follow a trend similar to that observed in core level spectra and conclude the formation of lattice-distorted hybrid hetero-nanostructures.
Electro-catalytic oxidation of glucose.-The electrochemical conversion of glucose to gluconolactone is schematically represented in Fig. 6a. Before starting the experiments, the catalysts are electro-activated by continuous cyclic voltammetry in 0.1 M KOH at 50 mVs −1 . Figures S3a & S3b in SI illustrates the consecutive 100 CV cycles of catalysts and the result shows that Ni/NiO@RuO 2 heterostructure leads highly stabilized current response than the Ni/NiO catalyst. The CV of bare carbon paper is negligible as shown in Fig. S3a (inset). Following this, Figs. 6b & 6c represents the electrochemical behaviour of both the catalysts at 20 mVs −1 in the absence and presence of 1 mM glucose. In the blank 0.1 M KOH, the peaks potential at 0.487 and 0.346 V are attributed to the electrochemical ↔ + + Ni Ni 2 3 redox process at the electrode surface. 73 With the addition of 1 mM glucose, the peak current increased and the anodic peak shifted to high potential due to the electro-catalytic oxidation of glucose at the electrode surface. It is noticeable that the glucose oxidation current is much higher for Ni/NiO@RuO 2 because of their synergistic contribution than the Ni/ NiO. The presence of Ni 3+ and added electro-catalytic activity by Ru 4+/3+ should increase the oxidation current significantly. In a negative sweep, two distinct reduction peaks, correspond to the reduction of α-NiOOH →NiO (0.34 V) and γ-NiOOH →NiO (0.22 V) at the electrode surface can be observed, particularly for Ni/NiO catalyst. 74 The decrease in reduction current may be attributed to catalyst consumption during the glucose-to-gluconolactone conversion process. 75 The possible reaction mechanisms are as follows 4-7: 76 As depicted in Figs. S4a & S4b in SI, the glucose oxidation current and peak potential are varying by the potential scan rate measured from 5 to 150 mVs −1 . With the scan rate, the anodic potential shift to positive and the cathodic peak shift to negative, manifesting that dominant diffusion controlled quasi reversible redox process in both catalysts. 78 It can be seen that the anodic and cathodic peak currents are in linear relations to the square root of the scan rate with the regression coefficients as given in Figs. S4c & S4d. With the consecutive addition of glucose from 0.1 to 5 mM, the anodic current in CV plots (Fig. 6d) follows a linear trend (R 2 = 0.9950) as shown in Fig. 6d inset for the Ni/NiO. In the case of Ni/NiO@RuO 2 heterostructure (Fig. 6e), the anodic peak current is comparatively higher than the Ni/NiO catalyst, and the peak current varies linearly (R 2 = 0.9973) (Fig. 6e inset), which indicates the two-way electrooxidation of glucose according to Eqs. 4 & 5. The high concentration deviation can be related to the passivation of the electrode or glucose isomer. 78 In an alkaline medium, Ni 2+ oxidizes into Ni 3+ and works as a catalyst for glucose oxidation by its reversible redox reaction. Typical glucose oxidation reaction takes place according to reactions 6 & 7. During the electrochemical catalytic (EC') process, glucose converts into gluconate and forms hydronium ions, which increase the sensing current in the CV curve in presence of glucose. From the slope at a linear range of the calibration curve, the actual glucose sensitivities are found to be 474.52 ± 1.8 and 641.95 ± 0.5 μA mM −1 cm −2 for Ni/NiO and Ni/NiO@RuO 2 , respectively. The limit of detection (LoD) of the catalysts is estimated from the relation 3σ/ slope, where σ is the standard deviation of the blank electrode. 79 Typical LOD is found to be 1.8 ± 0.3 μM for Ni/NiO and 0.4 ± 0.08 μM for Ni/NiO@RuO 2 heterostructure. The EC' process is further validated by the non-linear nature of the I pa vs scan rate plot as shown in Fig. S5(a) in SI. Meantime, the plot log I pa vs log v shown in Fig. S5(b) also exhibits the slope value of 0.473 and 0.458, which are close to 0.5 for the obvious diffusion-controlled electrooxidation reaction. 78,80 However, the small difference from the predicted theoretical value indicates the kinetic limitation in the glucose oxidation reaction at the catalyst's surface. This infers that the predominant diffusion-controlled process in Ni/NiO@RuO 2 heterostructure oxidises the glucose that increases the anodic current significantly. 80 On the other hand, the higher current may be attributed to the redox process involved in NiO and RuO 2 surfaces that provide well-distributed mesoporous nanoparticles, which improves the charge carrier kinetics for glucose oxidation. 81 The polynomial decrease in the plot of scan rate normalized current density (J pa /v 1/2 ) vs scan rate (v), as shown in Fig. S5(c) demonstrates the irreversible EC' kinetics at the catalyst's surface. Additionally, the rising part of the CV plot at 20 mVs −1 with 1 mM glucose and their extrapolated Tafel plot (E p vs log I p ) in Fig. S5(d)  where i cat and i B are current obtained from chronoamperogram for the catalysts measured with 1 mM glucose and Blank 0.1 KOH solutions, respectively, K cat is the catalytic rate constant, C is the concentration of glucose and t is time. The K cat is found to be 5.396 × 10 2 and 4.079 × 10 2 cm 3 M −1 s −1 for Ni/NiO and Ni/NiO@RuO 2 catalysts, respectively. The observed chronoamperometric results were also used to estimate the diffusion coefficient (D) from the well-known Cottrel Eq. 10, where A is the electroactive area, n, F, C, and t are having their usual meaning as followed in the above Eqs. 8 and 9. 78  Figure 6f and inset shows the Nyquist plots of both Ni/NiO and Ni/NiO@RuO 2 catalysts over the frequency range of 1 Hz to 100 kHz. If the time constants of different processes are distinct, the actual impedance spectra of the system consists of (i) a semicircle, centred on the real axis at high frequencies from which ohmic drop (R o ), double layer capacitance (C dl ), and charge transfer resistance (R ct ) can be deduced and (ii) a straight line with an angle of 45°with the real axis corresponding to the Warberg impedance (semi-infinite diffusion) at low frequencies. 85 The acquired EIS spectra have altered in high frequencies, and there is no well-defined semicircle, simply diffusion process is dominant at low frequencies. These alterations are prompted by the porous nature of the hybrid electrode, which causes potential and concentration distributions at the electrode-solution interface. 86 In other words, the potential distribution on the porous electrode may have an effect on the impedance spectra at relatively high frequency with the respect to the diffusion contribution. 87 This represents the rate-determining steps involved in the glucose oxidation process and can be categorized as diffusion controlled process. 88 These results further reveal the fact that the specific surface area and available mesopores are the reason for the change in the catalytic rate constant and diffusion coefficients.
To precisely assess glucose sensitivity, the applied voltage should be matched with the thermodynamic potential of the catalyst materials. In order to ensure this, the current-time (I-t) response curve is recorded with successive additions of 0.5 mM glucose at varied applied potentials ranging from 0.4 to 0.55 V. Figures 7a &  7b represents the I-t curves of Ni/NiO and Ni/NiO-RuO 2 catalysts at various applied potentials. These results suggest that the current response increases as applied voltage increases. However, at high potential the electrode surface can easily transition to higher oxide states (i.e. γ-NiOOH), which has limited stability in alkaline solutions. 89 Therefore, 0.5 V is chosen as the optimum working potential for the glucose sensor because of its practical versatility. Additionally, in different alkaline conditions, the concentrations of OH − anion play a significant role in the electro-oxidation of glucose. 90,91 The I-t response towards glucose oxidation is performed in 0.1, 0.5, and 1 M KOH solutions to obtain the optimal concentration. Figures 7c & 7d depicts the current response of the catalysts in three different KOH solution for the successive addition of 0.25 mM glucose up to 5 mM at an applied voltage of 0.5 V. All of them show fast current response and good linearity for all three measured KOH concentrations. The corresponding current response as a function of glucose concentration (Figs. 7e & 7f) reveals that the system has good sensitivity, a reasonable detection limit, and a wide linearity. However, hydrated Ni/NiO in highly alkaline environments produces solid solutions of higher oxides with semiconductor properties. 89 Hydrated Ni/NiO is oxidised to the unstable γ-NiOOH phase under high alkaline conditions, and this phase frequently contains K + ions, limiting their stability in alkaline solutions. 92 Therefore, the OH − anion concentration of 0.1 M is maintained for the glucose sensor analysis. Figure 8a shows the I-t response of Ni/ NiO and Ni/NiO@RuO 2 with successive addition of glucose at room temperature. The amperometric response of carbon paper is negligible as shown in Fig. S8 in SI. While Ni/NiO and Ni/NiO@RuO 2 catalysts showed a clear increase in current with a fast response time of <0.15 ± 0.02 s and reached a stable response afterwards as shown in Fig. 8 (a insets). For consecutive addition of glucose, there is an increase in current and indicating good catalytic activity of the heterostructure electrode. At higher glucose concentrations, the formed intermediate oxidation products remain in the catalyst's active sites thus limiting further glucose oxidation and reducing the current response. 93 Table SI in Supporting Information, presents the parametric comparison of the non-enzymatic glucose sensor performance of similar materials in terms of sensitivity, linear range, and LOD. Our Ni/NiO@RuO 2 heterostructure-based sensor shows a fast response time, good sensitivity, reasonable LOD and good linear range.
To utilize the material for real-life sensor applications, specificity or selectivity is one of the important factors that need more attention. Considering food products and biological samples, the coexisted oxidative major compounds are ascorbic acid (AA), dopamine (DA), uric acid (UA), urea (U), NaCl, KCl, CaCl 2 and carbohydrates such as fructose/sucrose (particularly in food products). As shown in Fig. 8b, Ni/NiO have ⩽3% tolerance against interfering species like DA, UA, AA and U, however, the carbohydrates and alkaline salts have a significant influence (⩾15% for sucrose) during glucose oxidation. In the case of Ni/NiO@RuO 2 heterostructure, the tolerance is quite good and it is only ⩽3% against all the interfering compounds except fructose, which has ⩾5%. This result may be due to a change in an isoelectric point of Ni with a Ru hybrid structure. 94 A sharp increase in current is recorded while adding 1 mM glucose, revealing the excellent specificity of the Ni/NiO@RuO 2 heterostructure catalyst towards glucose oxidation. One of the structural benefits of Ni/NiO@RuO 2 heterostructures is that it provides more active sites for glucose electro-catalytic reactions. As the Ni/NiO@RuO 2 heterostructure has defects and active edge sites, it facilitates high mass transport at the catalyst's surface. 95 Particularly, the high specificity is resulting from the synergetic effect of NiO and RuO 2 , which plays a pivotal role in optimizing the kinetics and energetics of the surface catalysis. 96 Therefore, our Ni/NiO@RuO 2 heterostructure shows a high specificity when compared to the Ni/NiO catalyst. Further, to evaluate the reproducibility of the catalysts, the 5 different electrodes are fabricated on carbon paper and an amperometric current is recorded with 1 mM glucose. The typical current response of the electrodes is shown in Fig. 8c, which confirms the excellent reproducibility of both catalysts. The stability of the catalysts is tested by cyclic voltammetry and amperometric analysis after exposure to air at different periods. Figures S9a & S9b in SI illustrates the CV at 20 mVs −1 measured in a week interval for up to 4 weeks, results indicate good stability of the electrode for long-time applications. As compared to Ni/NiO catalyst, the Ni/NiO@RuO 2 showed enhanced and stable catalytic activity even after one month. This result is further validated by amperometric (I-t) response measured after one month in pure KOH (black), with 1 mM G (red) and consecutive addition of 1 mM G (blue) during measurement at ∼50 s are shown in Figs. S9c & S9d. The recorded current shown in Fig. 8d exhibits <±2% variations in the glucose oxidation current after one month. Moreover, successive addition of 1 mM G at 500 s (blue curve Figs. S9c & S9d) shows a sharp increase in glucose oxidation current which indicates that the proposed catalysts are highly stable and can be suitable for long-term glucose monitoring in the food industry.
Detection of glucose in a real sample is desirable to scrutinize the feasibility of proposed sensors for practical applications. Thereby, the catalysts are tested for the detection of glucose in commercial beverages by amperometric analysis with the successive addition of glucose. Initially, CocaCola (Coke Zero), 7up (Zero Sugar) and Coconut Water are purchased from the local market (TESCO, Cork, Ireland) and added to 0.1 M KOH solution with a 1:25 (V: V) ratio to make the electrolyte. 16 The corresponding amperometric response is recorded with the 0.5 mM successive addition of glucose as shown in Figs. S10a & S10b. The results show a notable increase in current, suggesting the practical feasibility of the electrode for glucose monitoring in food products. The calibration plot shown in Figs. S10c & S10d, further reveals the good sensitivity of glucose (Table SII in SI) within a range of 1 to 4 mM. As shown in Table I, the recovery of glucose using both the samples is in the acceptable range of 94% to 120% for the measured soft drinks with 99% correlation accuracy. Therefore, it can be concluded that the Ni/NiO@RuO 2 heterostructure could be an effective alternative catalyst for glucose quantification in food products.
Electrochemical pH sensor.-Typically, the pH response of heterostructures is validated based on the protonation/deprotonation reaction at the electrode surface. During the interaction of H + ions with the Ni/NiO electrode surface, the Helmholtz layer is formed due to surface chemical bonds. Amphoteric characteristics of Ni/NiO@RuO 2 heterostructure, the surface potentials of different magnitudes can be achieved for a wide range of pH values 2-12. Noteworthy, the Ni/NiO provide enough negative charge to attract H + from H 3 O + ions. Meantime, hydrated Ni/NiO and RuO 2 nanostructures surfaces can help to extract the electron from neighbouring OH − ions. The electromotive force (EMF), of Ni/NiO and RuO 2 , are measured based on the surface potential difference against the known Ag/AgCl electrode. The possible redox-couple reactions between lower and higher valency oxidation states are as follows 11-13: 16 18 Both the electrodes are having <50 s response time, which is quite good when compared to the similar metal oxide-based sensor reported so far. 18 It is important to note that, the existence of Ni/NiO and RuO 2 heterostructure leads to a fast response time in both acidic and basic pH regimes. The measured potential response against pH demonstrates excellent linear regressions with high correlation coefficient (R 2 ) values of 0.9964 and 0.9968 as shown in Fig. 9c for Ni/NiO and Ni/NiO@RuO 2 , respectively. To investigate the Nernstian response, three measurements are taken for each electrode and the observed surface potential vs pH results are shown in Figs. S11a & S11b. The average sensitivity values for Ni/NiO and Ni/NiO@RuO 2 are −33.6 and −42.18 mV pH −1 respectively, which are approximately 56.8% and 71.3% of Nernstian response of 59.12 mV. 18 Both electrodes show a sub-Nernstian pH response in the range of 2-12. The reproducibility of the heterostructure   99,100 In contrast, the Ni/NiO-RuO 2 heterostructure exhibits a similar CV response in the reverse trend, as seen in Fig. S12b. Only surface-region associated H + adsorption with less defined Ru (II to III) redox pairs occurs in the measured potential region. 101 The response current increases noticeably with pH, due to the increasing electroactive surface area of the catalyst. Noticeable water oxidation occurs at pH levels greater than 10 due to the presence of more hydroxyl ions in the electrolyte. The voltammetry response of the Ni/NiO-RuO 2 heterostructure indicates that the additional RuO 2 increases the active surface area for H + adsorption by preventing Ni(OH) 2 formation, resulting in high stability in the studied pH range. However, the electrode sensitivity is lower when compared to their monoxides. This is due to the presence of different active phases in the metal/metal oxide surfaces confirmed by our initial XRD, XPS and TEM investigations, which showed the formation of defective heterostructures. These defective oxide phases of Ni/NiO x lead to the formation of hydrated phases, which limits the protonation reaction at the electrode surface. Furthermore, the existence of Ni(H + ), NiO(OH), RuO 2 ·xH 2 O and internally bounded H 2 O molecules varies the intrinsic permittivity (k) of the electrodes and results in low sensitivity. 102,103 During the protonation reaction, the formed defects including oxygen vacancies, free-OH groups, and Ni-Ru interstitials may trap the charge carriers and hinders their transport towards electrochemical active sites. 104 Additionally, NiO is an amphoteric material, which can form surface bonds by interacting with H + and OH − ions. Pre-existence of Ni(H + ) ion in the metal/ metal oxide heterostructure significantly repels the proton adsorption from the electrolyte and the formation of Ni 3d and Ru 3d states leads to the creation of free carrier trap sites thus resulting in the low potential response against pH. 104 Therefore, further improvement of pH sensing is needed for these materials to reach the actual Nernstian response.
To validate the proposed sensor for food quality analysis, the pH response of the sensors is measured in commercial beverages purchased from the local market (Tesco, Cork, Ireland) such as Cranberry Juice (CB), Coconut Water (CW), Milk (M), and Milk of Magnesia (MM). The OCV is measured when freshly prepared electrodes are immersed in 25 ml of solution for 500 s as shown in Figs. S13a & S13b. The pH values are examined with the commercial glass pH meter and presented in Table II. The observed OCV closely agree with the potential measured in B-R buffer solutions with commercial glass pH sensor. As shown in Fig. 9d, Ni/NiO@RuO 2 heterostructure reveals better sensitivity than the Ni/  NiO nanostructure. These results show the great potential of these hetero-nanostructure electrodes in the development of low-cost pH sensors for food analysis. It should be noted that the Ni/NiO electrode takes a much longer time to stabilize in lower pH. Similarly, Ni/NiO@RuO 2 does not reach a stable OCV even at 1500 s as shown in Fig. S13. This may be due to the interaction of the catalyst with different analytes present in the milk. Therefore, further investigation is required to understand the feasibility of the electrode for food pH analysis.

Conclusions
In summary, Ni/NiO and Ni/NiO@RuO 2 heterostructures are synthesized via a one-pot in situ solution combustion process. A strong redox-type eruption combustion mechanism is proposed without any additional precursor for the fabrication of a heterostructured catalyst. With these defect-rich heterostructures, novel glucose and pH sensors are fabricated and validated for real food analysis. Both electrodes show significant electro-catalytic activity toward glucose oxidation. Noteworthy, the Ni/NiO@RuO 2 glucose sensor exhibits a better sensitivity, linearity, fast response and considerably low detection limit than the Ni/NiO. Moreover, it displays excellent selectivity of glucose among the potential interferences and shows good stability. Concurrently, the proposed heterostructures are successfully investigated for pH sensors and showed competitive sensitivity of −41.6 mV pH −1 over a wide pH range of 2-12. At the same time, catalysts are tested in the real environment of commercial soft drinks for both glucose and pH determination. A commendable sensitivity towards glucose and pH is observed, thus authenticating the potential feasibility of the catalysts for food analysis. This study demonstrates a facile, in situ synthesis approach for the mass production of metal/metal oxide heterostructures for further development of low-cost smart sensors for food and bioanalytical applications. Declaration of Competing Interest. The authors declare that they have no known competing financial interests or personal relationships that could appear to have influenced the work described in this paper.

Data availability
Data will be made available on request.

Supporting Information
EDX spectra of Ni/NiO and Ni/NiO@RuO 2 with the elemental compositions, Core level Ru 3p XPS spectra of Ni/NiO@RuO 2 heterostructure, Continuous 100 cycles of CV for Ni/NiO and Ni/NiO@RuO 2 electrodes measured at 50 mVs −1 scan rate, Electrocatalytic oxidation of glucose measured at different scan rates of 5-150 mVs −1 , the corresponding v 1/2 vs I, plot, Scan rate v vs I pa plot, log v vs log I pa , scan rate v vs I pa /v 1/2 plot, the rising part of the CV plot at 20 mVs −1 with 1 mM glucose and (inset) Tafel plot E p vs log I pa . Amperometric (I-t) curves at 0, 0.5, 1.0, 1.5 and 2.0 mM concentrations of glucose for Ni/NiO and Ni/NiO@RuO 2 heterostructure. (Insets) the corresponding I cat /I B vs t −1/2 derivative plot. The linear portion of the (I-t) curve at different concentrations of glucose for Ni/NiO, Ni/NiO@RuO 2 and the corresponding slope of (i.t 1/2 ) vs glucose concentration C plots. Amperometric (I-t) response of pure carbon paper at 0.5 V. The CV curves at 20 mVs −1 were measured at different time intervals for up to 4 weeks for Ni/ NiO, & Ni/NiO@RuO 2 , the corresponding amperometric curves of the catalysts after 15 and 30 d of air exposure without and with 1 mM glucose. The amperometric response of Ni/NiO and Ni/NiO@RuO 2 catalysts in different commercial soft drinks in 0.1 M KOH with successive addition of 0.5 mM glucose, and the corresponding calibration curves. Repeated pH response of the catalyst at different pH values measured with Ni/NiO, Ni/NiO@RuO 2 , and their reproducibility test at pH 3, pH 7 and pH 10. The pH sensing response of Ni/ NiO and Ni/NiO@RuO 2 in real samples. Variation of OCV for Ni/NiO@RuO 2 in low-fat milk.