Fabrication of highly sensitive ethanol sensor based on doped nanostructure materials using tiny chips

Abstract

Doped CuO–Fe₂O₃ nanocubes (NCs) are prepared via a facile wet-chemical process using active reactant precursors with reducing agents in high pH medium (pH > 10). The NCs are totally characterized in detail using various methods, such as FTIR spectroscopy, X-ray photoelectron spectroscopy (XPS), transmission electron microscopy (TEM), dynamic light scattering (DLS), powder XRD, UV/vis spectroscopy, FESEM coupled XEDS, FE-SEM, etc. A thin-layer of NCs is deposited on tiny chips (surface area, ∼0.02217 cm²) to fabricate a selective ethanol sensor with short response time in liquid-phase medium. The fabricated chemi-sensor also exhibits higher sensitivity, large dynamic concentration ranges, long-term stability, and improved electrochemical performance towards ethanol. The calibration plot is linear (r² = 0.9937) over a wide ethanol concentration range (0.1 nM to 0.1 mM). The sensitivity and detection limit are ∼7.258 μA cm⁻² mM⁻¹ and ∼0.08 ± 0.02 mM (SNR, signal-to-noise ratio of 3), respectively. This novel effort establishes a well-organized way of developing efficient nanomaterial-based sensors for toxic pollutants in environmental and health-care fields on large scales.

---

Introduction

The significance of safety for lives as well as the environment has been studied in great detail in terms of using semiconductor sensors for toxic chemical detection via reliable methods.¹ Semiconductor nanostructure materials are very sensitive and efficient due to their smaller spherical size and high active surface to volume ratio as compared to conventional nanomaterials in the micro- or nano-meter ranges. Nanostructure metal oxides have attracted a great deal of attention due to their outstanding properties, including huge active surface area, high stability, quantum confinement effect, and high porosity and permeability (meso-porous nature), which are reliant on the shape and size of the nanocrystal.^2,3 Nanomaterials have attracted wide interest owing to their unique properties and potential application in fabrication of chemical sensors.^4,5 Semiconductor materials have been recognized as promising host nanomaterials for doping transition metals at room temperature. They have revealed a stable morphology and are composed of a number of disordered phases with geometrically-coordinated metal and oxide atoms, piled alternately along the axes.⁶ Transition metals co-doped in semiconductor materials have prompted profound research effort due to their exceptional and outstanding properties as well as versatile applications.⁷ Recently, extensive development has been carried out in research leading to metal-oxide co-doped ZnO-based nanomaterials, actuated by both fundamental sciences and potential advanced technologies.⁸ Doped semiconductor nanostructures exhibit promising uses as field-effect transistors,⁹ UV photo-detectors,¹⁰ gas sensors,¹¹ field emission electron sources,¹² nanomaterials,¹³ nanoscale power generators,¹⁴ and many other functional devices.¹⁵ Transition metal doping of a semiconductor nanostructure is also an effective method to regulate the energy level surface states of ZnO, which can further progress through changes in the doping concentrations in semiconductor materials. Doped nanostructure materials have attracted substantial attention due to their use in catalysis, opto-electronics, bio-chemicals, magnetic materials, photo-catalysis and electronics, their mechanical properties and their prospective applications in various fields. Doped nanostructure materials would be good candidates due to their higher & specific surface area, higher aspect ratio & active surface area, lower-potentials, lower-resistances, higher-catalytic activity, and smart electrochemical as well as optical characteristics. Transition metal-based materials have been recognized as attractive guest-objects for doping into semiconductor materials (as host-objects), exhibiting a stable and controlled morphology containing a significant amount of repeated crystal-phases coordinated with cationic and anionic parts.¹⁶ Nanostructure materials have potential, due to their outstanding and excellent electro-catalytic properties, towards various chemical, physico-chemical, and chemical-sensor applications.¹⁷ In the last decade, widespread progress has been made with transition-metal doped semiconductor materials in advanced sciences and sophisticated technologies.^18,19 Low-dimensional nanomaterials have exciting uses in electronics (i.e., diodes, LED, transistors),²⁰ ultra-violet photo-detectors,²¹ chemi- or immuno-sensors (bio-assays),^22,23 magnetic fields (i.e., para-magnetic),²⁴ doped materials,²⁵ energy generators,²⁶ and various bio-chips and micro-devices.^27,28 Doped nanomaterials are prepared using an effective thermal method to control the surface energy of the metallic states that can promote growth through the change in doping concentration in the semiconductor nanostructure materials. Undoped copper oxide nanomaterials have significant characteristic behaviours with potential applications in the fabrication of nano-electronics, opto-electronics, bio-sensors, bio-chemical chips, field-emission displays, and surface-active properties.^29–31 The guest (CuO) and host nanostructure materials (Fe₂O₃) have a major function in the improvement of accurate, more reproducible, highly electrosensitive & reliable toxic chemical chemi-sensors using tiny chips. Doped nanostructure materials have also attracted considerable attention from researchers, generally to control growth monitoring, owing to the increasing demand in the healthcare and bio-medical fields.^32–34 Semiconductor nanomaterials are being comprehensively investigated owing to their exclusive surface behaviours mediated by large-active surfaces, which can be used to construct perfect chemical recognizing elements in iono-sensors or chemo-sensors. Lately, the development of chemi-sensors using doped nanostructure metal–oxides, active conducting polymers and nanocomposites has been the key motive in the detection or determination of carcinogenic elements and hazardous chemicals.^35–38 Ethanol detection is important in different technological scientific areas, i.e., ecological & environmental monitoring, clinical diagnosis, & various practical industrial applications.^39–41 It has also been attracting increasing attention for toxicity analysis in recent years, particularly for hazardous and carcinogenic chemicals, owing to concerns about eco-environmental protection and health-care fields. Ethanol is a toxic analyte, which is generally applied in various research and industrial laboratories for R&D purposes. Long-term exposure to ethanol can have various consequences, such as health troubles and nervous diseases, and probably cell damage. Therefore, the determination and recognition of ethanol in the liquid phase is a significant task and can be accomplished using doped CuO–Fe₂O₃ nanocubes coupled onto tiny chips. The doped CuO–Fe₂O₃ NCs have interesting behaviours, i.e., large and active-surface area, nontoxicity, chemical instability, elemental natures, and good-conductivity, which offered good-electron communication characteristics that encouraged electron transfer to the target analytes. Typically, it is also demonstrated that transition metal-doped semiconductor nanomaterials (such as meso-porous nanomaterials) can provide huge surface area, higher stability, nano-porosity, and consistency, which could be exploited to develop potential chemical sensors.

Ethanol is extremely toxic and usually has serious consequences in health and the environment, so the detection using a reliable sensing method with prepared doped CuO–Fe₂O₃ NCs using tiny chips is immediately required. The investigation of ethanol detection by doped CuO–Fe₂O₃ NC films on chips is carried out and studied in detail. An easy coating method for the construction of the CuO–Fe₂O₃ NC thin-film with conductor binding-agents is developed for the preparation of nanomaterial films on smart chips. In this approach, doped CuO–Fe₂O₃ NC films fabricated with conducting binders are utilized towards the detection of target carcinogenic analytes using the reliable current vs. voltage (I–V) method. It is confirmed that the fabricated chemical sensor with CuO–Fe₂O₃ NCs on tiny chips using the I–V method is a unique and efficient approach for ultra-sensitive recognition of ethanol in short response-time.

Experimental

Materials and methods

Copper chloride, binders (butyl carbitol acetate and ethyl acetate), iron chloride, sodium hydroxide, and all other chemicals were of analytical grade and purchased from Sigma-Aldrich Company. They were used without further purification. The absorption maximum (λ_max) of the CuO–Fe₂O₃ NCs was investigated using a UV/vis spectrometer, where the band-gap energy (E_bg) was calculated based on this study. The FT-IR spectrum of the CuO–Fe₂O₃ NCs was investigated using a FT-IR spectrophotometer, which was used for the confirmation of metal–oxygen bonds (Cu–O and Fe–O). The XPS measurement of the doped CuO–Fe₂O₃ NCs was carried out with a Thermo-Scientific K-Alpha spectrometer (Germany), for the calculation of the binding energies (keV) of Cu, Fe, and O. The morphology, particle-size, and arrangement of the CuO–Fe₂O₃ NCs were recorded using a FESEM instrument from JEOL (JSM-7600F, Japan). The crystallinity and crystal-patterns of the CuO–Fe₂O₃ NCs were measured using a powder X-ray diffractometer under ambient conditions. Raman-shift was employed to determine the band-shift of the CuO–Fe₂O₃ NCs using a Raman spectrometer with different radiation sources. The size of the CuO–Fe₂O₃ nanocubes was determined using transmission electron microscopy (TEM; JEM-2100F, Japan). The TEM sample was prepared as follows: the synthesized CuO–Fe₂O₃ NCs were dispersed into ethanol under ultrasonic vibration for 2 min, and then the TEM film was dipped in the solution and dried for investigation. The particle size was determined using DLS (photon correlation spectroscopy) with a Zetasizer NanoZS (Zetatrac NPA 152-31A, USA) at 25.0 °C. The CuO–Fe₂O₃ nanocubes were dispersed with 3.0 ml distilled water in order to obtain an accurate scattering intensity before the measurement. An equivalent amount of sample was dispersed into 3.0 ml distilled water at 25.0 °C before analysis. All the samples were analyzed in triplicate (n = 3). The current vs. voltage (I–V) method (two electrodes on a fabricated micro-chip) was used to measure the ethanol ions, using a Keithley-Electrometer from the USA.

Preparation and growth mechanism of CuO–Fe₂O₃ NCs

The term “wet chemical methods” emerged in contrast to conventional and solid-state synthesis methods for nanostructural materials, widely used in the preparation of doped or undoped nanomaterials. The term refers to a group of methods of powder and material production using the liquid phase in one of the preliminary process stages. The wet chemical products of solids in liquid-phase synthesis have much smaller grains (crystallites) and, usually, lower temperature and shorter duration of phase formation. Here, a facile and low-temperature synthesis of CuO–Fe₂O₃ NCs was developed using a wet-chemical process with active reactant precursors, such as copper chloride (CuCl₂), iron chloride (FeCl₃), and sodium hydroxide (NaOH). In a typical reaction procedure, 0.1 M CuCl₂ was dissolved in 50.0 ml deionized (DI) water and mixed with 50.0 ml FeCl₃ solution (0.1 M) under continuous stirring. The pH of the resultant solution was adjusted to over 10.0 by the addition of NaOH and the resulting mixture was shaken and stirred continuously for 10.0 minutes under ambient conditions. After stirring, the solution mixture was then put into conical flux and heated up to 120.0 °C for 6.0 hours. The active solution temperature in reaction medium was controlled manually throughout the process at 90.0 °C. After heating the reactant mixture, the flux was cooled under ambient conditions until it reached room temperature. The final CuO–Fe₂O₃ doped product was prepared, which was washed with DI water, ethanol, and acetone several times sequentially and dried at room-temperature. The as-grown product was left to air dry for a few hours and then used for structural, elemental, morphological, and optical characterizations. The growth mechanism of the CuO–Fe₂O₃ nanocube materials can be explained on the basis of the chemical reactions and nucleation, as well as the growth of the doped nanocrystals. The probable reaction mechanisms (i)–(iv) for obtaining the CuO–Fe₂O₃ nanomaterial are presented below.

NaOH_(aq) → Na_(aq)⁺ + OH_(aq)⁻ (i)

CuCl₂ + NaOH_(aq) → Cu_(aq)²⁺ + OH_(aq)⁺ + Na_(aq)⁺ + 2Cl_(aq)⁻ (ii)

FeCl_3(s) + 2NaOH_(aq) → Fe_(aq)³⁺ + 2OH_(aq)⁻ + 2Na_(aq)⁺ + 3Cl_(aq)⁻ (iii)

Cu_(aq)²⁺ + 2Fe_(aq)³⁺ + 6NaOH_(aq) → CuO·Fe₂O_3(s)↓ + 6Na⁺ + 2H₂O (iv)

The reaction proceeded slowly according to eqn (i) to eqn (iii). During preparation, the pH value of the reaction medium plays an important role in the doped nano-material oxide formation. Over pH 10, when CuCl₂ is hydrolyzed with NaOH solution, copper hydroxide is instantly formed according to eqn (ii). During the whole synthesis process, NaOH operates as a pH buffer to control the pH value of the solution and slowly contribute hydroxyl ions (OH⁻). When the concentrations of the Cu²⁺ and OH⁻ ions are achieved above the critical value, the precipitation of CuO nuclei begins. As there is a high concentration of Fe³⁺ ions [according to reaction (iii)] in the solution, the nucleation of Fe₂O₃ crystals becomes slower due to the lower activation energy barrier of heterogeneous nucleation. Hence, as the concentration of Cu²⁺ increases, a number of larger CuO–Fe₂O₃ crystals with an aggregated cube-like morphology form after the reactions [eqn (iv)]. The shape of the as-prepared CuO–Fe₂O₃ NCs is approximately consistent with the growth pattern of copper oxide-doped iron oxide nanocrystals.^42,43 Then, the solution was washed thoroughly with acetone, ethanol, and water successively and left to dry under ambient conditions. In the NC growth technique, initially the CuO and Fe₂O₃ nuclei growth takes place by self-aggregation, which then re-aggregate and produce CuO–Fe₂O₃ nanocrystals according to the Ostwald ripening method. The nanostructure material crystallizes and re-aggregates amongst itself through van der Waals forces and forms the doped CuO–Fe₂O₃ nanocube morphology, which is presented in Scheme 1. Finally, the as-grown CuO–Fe₂O₃ NCs were fully characterized in detail for their morphological, structural, elemental, and optical properties, and applied for the detection of ethanol as a chemical sensor for the first time. The phases of the air-dried and doped powder materials were examined using XRD. The morphology and cross section of the powders were observed using SEM equipped with XEDS. The evaluation of the band-gap energy and metal–oxygen bond formation (Cu–O and Fe–O) was examined by using UV/vis and FTIR spectroscopy.

Scheme 1 Probable growth mechanism of doped CuO–Fe₂O₃ NCs at low-temperature using a wet-chemical process.

Construction of μ-chips using photolithography method

Electrochemical μ-chips were fabricated using a conventional photolithographic technique, where electrodes and passivation layers are developed on a silicon wafer followed by dicing and packaging.^44,45 Nitrogen-doped silicon wafers are prepared and washed with extra-pure water. In this step, all contaminations on the surface and native SiO₂ layer are removed perfectly. At first, wet oxidation is employed, followed by dry oxidation, where the wafers are annealed in a nitrogen environment. Aluminum is sputtered with an aluminum-1% Si target. Then, the photolithography processes are applied. Resist coating, baking, exposure, and development are employed using materials from Kanto Chemicals, followed by rinsing thoroughly with ionic water. Aluminum is etched with etching solution and the resistance layer is removed perfectly using a plasma etching instrument. Then, the silicon wafers are cleaned using acetone, methanol, and finally plasma. A silicon nitride (SiN) layer is deposited by chemical vapor deposition and then pad electrode surfaces are etched by reactive ion etching. Finally, residual resist layer is removed by plasma etching. After the photolithographic process, platinum is sputtered using SP150-HTS. Then, it is patterned using the lift-off method, in which the wafers are immersed into the remover, and then washed with isopropyl alcohol. The photolithographic process is again utilised, where titanium is sputtered as a binding layer and then gold is evaporated by the deposition method. Finally, the gold layer is patterned using the lift-off method. A parylene passivation layer is formed for the protection of the μ-chip from water. The photolithographic process is performed again for pad protection. Then, the parylene dimer is evaporated using deposition apparatus. The photolithography process is used again for patterning. The parylene layer is patterned by etching. Finally, un-necessary resists are removed using acetone and then the wafer is cleaned with isopropyl alcohol (IPA). The resist is coated on the whole surface of the silicon wafer for protection during the dicing process. The silicon wafer is diced into pieces using dicing apparatus and stored in desiccators when not in use. The resist on the μ-chip surface is removed using acetone and cleaned with isopropyl alcohol (IPA). The opposite side of the chip is roughed using a sandpaper sheet for better adhesion and electrical stability. The μ-chip is bonded with die and packaged with silver paste. It is dried in a drying oven. Pads on the chip are connected to the package through gold wire with a bonding machine. Finally, silicon-based adhesive is put on the periphery of the chip to protect the pads and gold wire from the sample solution. The adhesive is dried for 24 hours at room temperature. The semiconductor smart μ-chips were fabricated on silicon wafer. Aluminum was sputtered to fabricate wiring and bonding pads. Pt–Ti–TiN was sputtered on a thermal oxide of silicon and patterned using photolithography to fabricate a counter electrode (CE). Ti–TiN layers were used for strong adhesion. Au–Ti was sputtered and lithographed, which made a circular working electrode (WE) with a diameter of 1.68 mm in the centre of the μ-chip. After electrode fabrication, the parylene layer was fabricated using the evaporation method as a passivation layer. The wafer was diced into 5.0 mm square μ-chips. This μ-chip was bonded to a package using silver paste. Aluminum pads were connected to the package by gold wire. Finally, adhesive (Araldite, Hantsman, Japan) was put on the periphery of the chip, which prevents the target solution from contacting the pads.

Fabrication of doped CuO–Fe₂O₃ NCs/μ-chip assembly

Construction of the μ-chip using the conventional photolithography method has been already explained in the previous section. Here, a tiny μ-chip is fabricated using the as-grown CuO–Fe₂O₃ NCs with conducting coating agents (ethyl cellulose powder, EC & butyl carbitol acetate solvent, BCA). After that, the CuO–Fe₂O₃ NC-fabricated μ-chip assembly is moved into a heating oven at 65.0 °C for 2 hours to ensure complete drying and uniform film formation. The fabricated NCs/micro-chip and Pt micro-line (on micro-chip) are used as a working and counter electrode, respectively. As-purchased ethanol is used to make the target solution to formulate different concentrations (1.0 nM to 1.0 mM) in 0.1 M phosphate buffer solution with a deionised system (0.1 M PBS is made in deionised water) and used as a selective target toxic analyte. 15.0 μL of the target analyte solution (in phosphate buffer solution, 0.1 M PBS) is dropped onto the NCs/chip during investigation. The current–voltage (I–V, slope of the calibration-curve) is utilized to calculate the target NCs sensitivity. The limit of detection (LOD) is determined from the 3 N/S ratio vs. sensitivity (∼3 × noise vs. Sens.) in the linear portion of the total concentration range of the calibration plot. The current–voltage technique is used to determine the potential using an electrometer in a two-electrode (working–counter) assembly. The CuO–Fe₂O₃ NCs are made and evaluated for the detection of target ethanol in a solution-phase system.

Results and discussion

Evaluation of morphological and elemental properties

FESEM images of the as-grown CuO–Fe₂O₃ NCs are presented in Fig. 1a–c, which show the images of the NCs with the nano-dimensional sizes of the as-grown CuO–Fe₂O₃ NCs. The size of the NCs is calculated as ∼0.11 μm. It is clearly revealed from the FESEM images that the facilely synthesized CuO–Fe₂O₃ NCs are nanostructures in a cubic shape, which are grown in very high density and are almost uniform cubes. When the size of the doped material decreases to the nanometer scale, the surface area is increased significantly; this improved the energy of the system and made re-distribution of Cu and Fe ions possible. The nanometer-sized cube may have tightly packed into the lattice, which is in agreement with published reports.^46,47

Fig. 1 (a–c) FESEM images and (d) elemental analysis of as-grown CuO–Fe₂O₃ NCs under ambient conditions.

The X-ray electron dispersive spectroscopy (XEDS) analysis of these CuO–Fe₂O₃ NCs indicates the presence of copper (Cu), iron (Fe), and oxygen (O) in the pure as-grown nanostructured material, as presented in Fig. 1d. It is clearly displayed that the prepared nanomaterial contained only Cu, Fe, and O elements at 4.99, 90.75 and 4.26 wt% respectively, which is presented in Fig. 1d (inset). No other peaks related with any impurities have been detected in the FESEM-coupled XEDS, which confirms that the nanocubes are composed only of Cu, Fe, and O.

Evaluation of optical and structural properties

The as-grown CuO–Fe₂O₃ NCs are also investigated in terms of their atomic and molecular vibrations. To predict the functional recognition, FTIR spectra fundamentally in the region of 400–4000 cm⁻¹ are investigated under ambient conditions. Fig. 2a displays the FTIR spectrum of the NCs, which shows a band at 532 cm⁻¹. This observed broad vibration band (at 532 cm⁻¹) could be assigned as metal–oxygen (Cu–O and Fe–O) stretching vibrations, which demonstrated the configuration of the CuO–Fe₂O₃ NC material.

Fig. 2 (a) FT-IR spectroscopy, (b) UV/visible spectroscopy, and (c) X-ray powder diffraction of as-grown CuO–Fe₂O₃ NCs under ambient conditions.

The optical properties of the as-grown CuO–Fe₂O₃ NCs are one of the significant characteristics for the assessment of their photo-catalytic activity. UV/visible absorption is a technique in which the outer electrons of atoms or molecules absorb radiant energy and undergo transitions to high energy levels. In this phenomenon, the spectrum obtained due to optical absorption can be analyzed to acquire the energy band-gap of the doped metal-oxides. For UV/visible spectroscopy, the absorption spectrum of the CuO–Fe₂O₃ NC solution is measured as a function of wavelength, which is presented in Fig. 2b. It presents a broad absorption band around 648.0 nm in the visible range between 200.0 to 800.0 nm wavelengths, indicating the formation of CuO–Fe₂O₃ NCs. The band-gap energy (E_bg) is calculated on the basis of the maximum absorption band of the CuO–Fe₂O₃ NCs and is found to be ∼1.9135 eV, according to eqn (v):

where E_bg is the band-gap energy and λ_max is the absorption wavelength (648.0 nm) of the CuO–Fe₂O₃ NCs. No extra peaks associated with impurities and structural defects are observed in the spectrum, which proved that the synthesized microstructure controlled the crystallinity of the as-grown CuO–Fe₂O₃ NCs.⁴⁸

The crystallinity and crystal phase of the as-prepared CuO–Fe₂O₃ NCs were investigated. The powder X-ray diffraction patterns of the doped nanocubes are shown in Fig. 2c. The CuO–Fe₂O₃ NC sample was investigated and exhibited a face-centered cubic shape. Fig. 2c reveals characteristic crystallinity of the CuO–Fe₂O₃ NCs and their crystalline arrangement, which is investigated by powder X-ray crystallography. The peaks were found to match with the CuO phase (tenorite) having face-centered monoclinic geometry [Joint Committee on Powder Diffraction Standards, JCPDS#073-6234]. The major phases indicate the characteristic peaks (hash symbol, #) with indices for 2θ values at 35.5(110), 38.2(111), 53.1(020), 57.1(202), 67.8(113), 74.1(221), and 76.1(−222) degrees. The monoclinic lattice parameters are a = 4.662, b = 3.417, c = 5.118, β = 99.48, point group: C2/c, and radiation: CuK_α1 (λ = 1.5406). These indicate that there is a significant amount of crystalline CuO present in the semiconductor nanomaterial. The reflection peaks are found to match with the iron oxide phase (hematite, α-Fe₂O₃) having rhombohedral geometry [JCPDS#080-2377]. In Fig. 2c, the phases (with indices) represent the major characteristic peaks (star symbol, *) for the as-grown crystalline iron oxide at 2θ values of 24.1(012), 32.5(104), 39.4(006), 40.7(113), 43.4(202), 49.3(024), 54.1(116), 57.6(018), 62.3(214), 64.1(300), 69.2(208), 71.6(1010), and 75.5(220) degrees. The lattice parameters are a = 5.03521, c = 13.7508, Z = 6, point group: R3̄c(167), and radiation: CuK_α1 (λ = 1.5406). These indicate that there is a considerable amount of crystalline iron oxide present in the doped nanostructured material. These results confirmed that there is a large number and amount of crystalline doped CuO–Fe₂O₃ NCs present in the NCs.⁴⁹ The crystallite size was calculated using Debye–Scherrer’s formula, given by eqn (vi):

D = Kλ/(β cos θ) (vi)

where D is the crystal size; λ is the wavelength of the X-ray radiation (λ = 0.15406 nm) for CuK_α; K is usually taken as 0.9; and β is the line width at half-maximum height (FWHM).⁵⁰ The average cross sectional diameter of the CuO–Fe₂O₃ NCs is close to ∼0.11 μm.

Evaluation of binding energy

X-ray photoelectron spectroscopy (XPS) is a quantitative spectroscopic method that determines the chemical states of the elements present within doped materials. XPS spectra are acquired by irradiating a nanomaterial with a beam of X-rays, while simultaneously determining the kinetic energy and number of electrons that get away from the top one to ten nm of the material being analyzed. Here, XPS measurements were carried out for the CuO–Fe₂O₃ NCs semiconductor nanomaterial to investigate the chemical states of Fe₂O₃ and CuO. The XPS spectra of O1s, Fe2p, and Cu2p are presented in Fig. 3a. The O1s spectrum shows a main peak at 531.3 eV in Fig. 3b. The peak at 531.3 eV is assigned to lattice oxygen and may indicate oxygen (i.e., O₂⁻) present in the CuO–Fe₂O₃ NCs.⁵¹ XPS was also used to resolve the chemical state of the doped Fe₂O₃ nanomaterial and its depth. Here, the spin orbit peaks of the Fe2p_(3/2) and Fe2p_(1/2) binding energies for all the samples appeared at around 713.2 eV and 724.1 eV respectively, which is in good agreement with the reference data for Fe₂O₃.⁵² In Fig. 3d, the spin orbit peaks of the Cu2p_(3/2) and Cu2p_(1/2) binding energies for all the samples appeared at around 932.5 eV and 953.1 eV respectively, which is in good agreement with the reference data for CuO.⁵³ XPS compositional analyses evidenced the co-existence of the two single phases of CuO and Fe₂O₃ nanomaterials. Therefore, it is concluded that the wet-chemically prepared doped CuO–Fe₂O₃ NC material has a NC phase containing two materials. Also, this conclusion is noticeably consistent with the XRD data.

Fig. 3 XPS of (a) doped CuO–Fe₂O₃ NCs, (b) O1s level, (c) Fe2p level, and (d) Cu2p level, acquired with MgKα1 radiation.

Particle size analysis using TEM and DLS

Further structural characterization of the CuO–Fe₂O₃ material was carried out and a cube-shaped morphology was found through TEM analysis. As shown in Fig. 4a, the TEM image of the CuO–Fe₂O₃ nanomaterial displayed a cubic shape with an average diameter of 0.11 μm. The TEM observation shows the exact morphology of the CuO–Fe₂O₃ nanostructures assembled in cube-shaped morphology, as found in FESEM, and showed full consistency in terms of shape and dimensions.

Fig. 4 Particle size analysis of CuO–Fe₂O₃ NCs using (a) TEM and (b) DLS.

The size distribution of the CuO–Fe₂O₃ nanocubes was measured using DLS at 25.0 °C. Before investigation, the NCs were well-dispersed into 3.0 ml distilled water in order to obtain an accurate scattering intensity. The dispersed solution of NCs was analyzed in triplicate (n = 3). The size distribution measured using DLS showed an approximately normal distribution (Fig. 4b). The hydrodynamic size of the CuO–Fe₂O₃ NCs was 0.1274 μm in the 0.1001–0.1558 μm range. SEM, TEM, and XRD could be used to determine the morphology and original diameter of the particles. DLS mainly reflects the hydrodynamic size in dispersion media. The hydrodynamic size of the CuO–Fe₂O₃ NCs was measured in distilled water as a stock medium. The observation showed that, owing to the van der Waals forces and hydrophobic interactions with the surrounding medium, the hydrodynamic size is generally larger (∼0.1274 μm) than the original (∼0.11 μm). The zeta potential (particle charge) was also measured by determining the electrophoretic mobility of the nanocubes using the same instrument, a Zetasizer NanoZS, at 25.0 °C. The values were calculated using the Helmholtz–Smoluchowski equation. The CuO–Fe₂O₃ NCs were diluted 10 times with filtered deionized water to obtain an ideal concentration range for optimal measurement (the viscosity values of the dispersion medium were made less than equivalent to water [1.0 mPas] at 25.0 °C). Here, the zeta potential, mobility, conductivity, charge, and dielectric constant were also measured and were found to be 5.7 mM, 0.44 u s⁻¹ V⁻¹ cm⁻¹, 415.0 uS cm⁻¹, 0.056 fC, and 79, respectively.

Application: chemical sensor with NCs/μ-chips assembly

The potential application of the CuO–Fe₂O₃ NCs assembled onto μ-chips as chemical sensors (especially for ethanol) has been investigated for the detection of hazardous chemicals that are not environmentally friendly. The improvement of doping of these CuO–Fe₂O₃ NCs on μ-chips as chemical sensors is in the initial stage and no other reports are available. The NC sensors have advantages such as stability in air, non-toxicity, chemical inertness, electrochemical activity, simplicity to assemble or fabricate, and bio-safe characteristics. As in the case of toxic ethanol sensors, the phenomenon is that the current response in the I–V method for the CuO–Fe₂O₃ NCs considerably changes when aqueous ethanol is adsorbed. The calcined CuO–Fe₂O₃ NCs were applied for modification of a chemical sensor, where ethanol was measured as the target analyte. The magnified view of the internal μ-chip center (sensing area) is presented in Fig. 5a–c. In Fig. 5a–c, the platinum line (PtE) and gold central circle on the micro-chip are employed as CE and WE electrodes (potential sources in 2-electrode system), respectively. In Fig. 5d, the probable detection mechanism of the NCs/chips is presented for the fabricated ethanol-sensor using the I–V method. In Fig. 5e, the expected experimental I–V responses with NCs/chips with reliable conducting coating agents are shown. The fabricated surface of the CuO–Fe₂O₃ NC sensor was made with conducting binders on the μ-chip surface, which is presented in the Fig. 5c and d. The fabricated μ-chip electrode was placed into the oven at low temperature (60.0 °C) for two hours to make it dry and stable, and give a totally uniform surface. The I–V signals of the chemical sensor having a NC doped thin film are anticipated as a function of current versus potential for hazardous ethanol. The real electrical responses to target ethanol are investigated using the simple and reliable I–V technique with the CuO–Fe₂O₃ NC-fabricated μ-chip, and are presented in Fig. 5e. The holding time of the electrometer was set as 1.0 s. A significant amplification in the current response with applied potential is confirmed. The simple and reliable possible reaction mechanism is generalized in Fig. 5d, in the presence of ethanol on the NC sensor surface with the I–V technique. The ethanol is converted to water and carbon dioxide in the presence of the doped nanomaterial by releasing electrons (−6e⁻) to the reaction system (conduction band, C.B.), which improved and enhanced the current responses against potential during the I–V measurement under ambient conditions.

Fig. 5 Schematic diagram of (a) real camera view from top of bared chip, (b) magnified view of CuO–Fe₂O₃ NC fabrication onto chips, (c) expected magnified view of CuO–Fe₂O₃ NCs/chips assembly with conducting coating binders on sensing area of tiny chip, (d) probable reaction mechanism of target ethanol ions in presence of doped NCs, and (e) expected responses of I–V method in the experimental results.

Fig. 6a shows the current responses of un-coated (gray-dotted) and coated (dark-dotted) μ-chip working electrodes with CuO–Fe₂O₃ NCs in the absence of target ethanol. With NCs fabricated on the surface, the current signal is slightly reduced compared to the uncoated NCs/μ-chip surfaces, which indicates that the surface is slightly inhibited with CuO–Fe₂O₃ NCs during the measurement of the I–V curve. The current changes without target analyte (dark-dotted) and with target analyte (blue-dotted), i.e. with ethanol (∼15.0 μL, ∼0.1 μM) added to the CuO–Fe₂O₃ NC-modified μ-chips, are shown in Fig. 6b. A significant current enhancement is exhibited with the CuO–Fe₂O₃ NC-modified μ-chips compared with uncoated μ-chips due to the presence of nanostructures, which give a higher specific surface area, larger surface coverage, excellent absorption and adsorption capability into the porous NC surface towards the target ethanol. A control experiment is performed for the comparison of the current response with various compositions of NCs, such as CuO NCs, Fe₂O₃ NCs, and CuO–Fe₂O₃ NC-embedded μ-chips. Here, the I–V responses of the undoped CuO NC-coated μ-chip, undoped Fe₂O₃-coated NC μ-chip and doped CuO–Fe₂O₃ NC-coated μ-chip are measured by deducting the background current (0.001 mM analyte). In this control experiment, the doped CuO–Fe₂O₃ NC-coated μ-chip exhibits the highest current response compared to the un-doped CuO or Fe₂O₃ NCs, as presented in the ESI section (S1†). This significant change in surface current is monitored for every injection of the target ethanol onto the CuO–Fe₂O₃ NC-modified μ-chips by an electrometer. The I–V responses from the CuO–Fe₂O₃ NC-modified μ-chip surface are investigated for various concentrations (0.1 nM to 1.0 mM) of ethanol, which are shown in Fig. 6c. It shows the current changes of the fabricated μ-chip films as a function of ethanol concentration under ambient conditions. It was also found that from low to high concentration of target analyte, the current responses increased regularly. Potential current changes from lower to higher potential (+0.1 V to +1.3 V) based on various analyte concentrations are observed, which is clearly presented in Fig. 6c. A broad range of analyte concentrations is measured to determine the probable analytical limit, which is calculated from 0.1 nM to 0.1 mM. The calibration (at +0.5 V) and magnified calibration curves are plotted from the various ethanol concentrations, which are presented in the Fig. 6d. The sensitivity is estimated from the calibration curve, which is close to ∼7.258 μA cm⁻² mM⁻¹. The linear dynamic range of this sensor is shown to be from 0.1 nM to 0.1 mM (linearity, R = 0.9937) and the detection limit was considered to be 0.11 ± 0.02 nM [3 × noise (N)/slope (S)].

Fig. 6 I–V responses of (a) uncoated and coated μ-chip with CuO–Fe₂O₃ NCs; (b) in absence and presence of 0.1 μM ethanol with CuO–Fe₂O₃ NCs/chip; (c) concentration variations (0.1 nM to 0.1 mM) of analyte; and (d) calibration plot of CuO–Fe₂O₃ NCs fabricated on μ-chip surfaces. Potential was chosen in +0.1 to +1.5 V ranges. Error limit of I–V measurement was ±0.01. There are three trials done for the same experimental concentration under similar conditions.

The resistance value of the doped semiconductor materials decreased (current increased) with increasing surrounding active oxygen, which are the fundamental characteristics of nanomaterials.⁵⁴ Actually, the oxygen adsorption demonstrates a significant responsibility for the electrical properties of the CuO–Fe₂O₃ NCs on the tiny μ-chips. The oxygen ion adsorption removed the conduction electrons and increased the resistance of the CuO–Fe₂O₃ NCs. Unstable oxygen species (i.e., O₂⁻ & O⁻) are adsorbed on the doped NC surface at room temperature, and the quantity of such chemisorbed oxygen species is directly dependent on the morphological and structural properties. Under ambient conditions, O₂⁻ is chemisorbed, while on the nanocube morphology, O₂⁻ and O⁻ are chemisorbed significantly. For this reason, the active O₂⁻ disappeared quickly.⁵⁵ Here, the ethanol sensing mechanism on the CuO–Fe₂O₃ NCs/μ-chip sensor occurs due to the presence of semiconductor oxides. The oxidation or reduction of the semiconductor NCs occurs, according to the dissolved O₂ in the bulk solution or the surface–air interface of the neighbouring atmosphere, according to eqn (vii–ix).

O_2(diss)(CuO–Fe₂O₃ NCs/μ-chip) → O_2(ads) (vii)

e⁻(CuO–Fe₂O₃ NCs/μ-chip) + O₂ → O₂⁻ (viii)

e⁻(CuO–Fe₂O₃ NCs/μ-chip) + O₂⁻ → 2O⁻ (ix)

These reactions occur in the bulk-system or air/liquid interface or adjacent atmosphere due to the small carrier concentration, which enhances the resistance. The ethanol sensitivity could be attributed to the high oxygen deficiency on the CuO–Fe₂O₃ NCs/μ-chip (e.g. MO_x), as this higher density prompts an increase in oxygen adsorption. The larger the quantity of oxygen adsorbed on the fabricated sensor surface, the larger the oxidizing potential and the faster the oxidation of ethanol would be. The reactivity towards ethanol would have been very large as compared to other fabricated material surfaces under identical conditions.^56,57 When ethanol reacts with the adsorbed oxygen on the exterior/interior of the CuO–Fe₂O₃ NCs/μ-chip layer, it is oxidized to carbon dioxide and water by releasing free electrons (6e⁻) in the conduction band, which is expressed through the following reaction, (x).

CH₃CH₂OH_(ads) + 6O⁻_(ads) → 2CO₂ + 3H₂O + 6e⁻ (C.B.) (x)

In the reaction system, these reactions refer to oxidation of the reducing carriers. This method enhanced the carrier concentration and consequently decreased the resistance on adjacent reducing analytes. The elimination of ionosorbed oxygen amplified the electron concentration on the CuO–Fe₂O₃ NCs/μ-chip and hence the surface conductance is increased in the film.⁵⁸ The reducing analyte (ethanol) gives electrons to the CuO–Fe₂O₃ NCs/μ-chip surface. Consequently, resistance is reduced, and hence the conductance is increased. This is the reason why the analyte response (current) is amplified with increasing potential. Thus, the produced electrons contribute to a rapid increase in conductance of the thick CuO–Fe₂O₃ NCs/μ-chip film. The unusual CuO–Fe₂O₃ NCs regions dispersed on the surface would increase the capability of the nanomaterial to absorb more oxygen species, giving high resistance in ambient air, which is presented in Fig. 7.

Fig. 7 Mechanism of ethanol detection with active CuO–Fe₂O₃ NCs/μ-chip under ambient conditions.

The sensor response time was ∼10.0 s for the CuO–Fe₂O₃ NC-coated μ-chip sensor to achieve saturated steady state current in the I–V plots. The great sensitivity of the μ-chip sensor can be attributed to the good absorption (porous surfaces of the NCs fabricated with binders), adsorption ability, high catalytic activity, and good bio-compatibility of the CuO–Fe₂O₃ NCs/μ-chip. The expected sensitivity of the NC fabricated sensor is better than previously reported ethanol sensors based on other composite or material-modified electrodes.⁵⁹ Due to the perceptive surface area, here the doped nano-materials developed a beneficial microenvironment for toxic chemical detection (by adsorption) and recognition with excellent sensitivity. The prominent sensitivity of the CuO–Fe₂O₃ NCs affords high electron communication features, which improved the direct electron communication between the active sites of the nanosheet-composed microstructures and μ-chips. The modified thin CuO–Fe₂O₃ NCs/μ-chip sensor film had better reliability as well as stability under ambient conditions. The CuO–Fe₂O₃ NCs/μ-chip exhibits several advances in providing ethanol chemical based sensors, and encouraging improvement has been accomplished in the research section.

The sensing selectivity performances (interferences) with other chemicals, like acetone, dichloromethane, methanol, chloroform, ethanol, 4-nitrophenol, methanol, propanol, and butanol (Fig. 8a), were also investigated. Ethanol exhibited the maximum current response according to the I–V system using CuO–Fe₂O₃ NC-fabricated micro-chip electrodes. Therefore, it was specific towards ethanol compared to all other chemicals. Current responses (at +0.5 V) of all interfering analytes converted into percentage (% responses) by deducting the blank current (reading in PBS-only system) are calculated and presented in Fig. 8b. By deducting the current value of the blank solution, it was found that the current value was less than 10% for all chemicals (acetone 4.1%; dichloromethane 3.7%; chloroform 4.3%; 4-nitrophenol 2.7%; methanol 3.5%, propyl alcohol 8.6%, butanol 7.4%, iso-propyl alcohol 5.4%, and blank 0%) compared to target ethanol (90.0%). Therefore, it is clearly demonstrated that the sensor is most selective towards ethanol compared with other chemicals.

Fig. 8 I–V responses of CuO–Fe₂O₃ NC-coated μ-chip are presented for ethanol sensing: (a) selectivity, (b) current responses of analytes at +0.5 V (presented in percentages), (c) control experiment, and (d) reproducibility study. Concentration of ethanol and other chemicals is taken as 0.1 μM for the selectivity study. Delay time: 1.0 s. Potential range: 0 to +1.5 V.

A control experiment was performed with the various nanomaterial fabricated microchips, using individually Fe₂O_3, CuO, and CuO–Fe₂O₃ NCs at 0.001 mM analyte concentration with the I–V method (Fig. 8c). A significant enhancement in current responses was observed for the CuO–Fe₂O₃ NCs/chip compared to undoped Fe₂O₃ and CuO individually. To investigate the reproducibly and storage stabilities, the I–V response for the CuO–Fe₂O₃ NC-coated μ-chip sensor was examined (up to 2 weeks). After each experiment, the fabricated CuO–Fe₂O₃ NCs/μ-chip substrate was washed gently and it was observed that the current response was not significantly decreased (Fig. 8d). The sensitivity was retained as almost the same as the initial sensitivity for up to week (1st to 2nd week); after that the response of the fabricated electrode gradually decreased. A series of six successive measurements of 0.1 μM ethanol in solution yielded a good reproducible signal with the CuO–Fe₂O₃ NCs/μ-chip sensor under different conditions with a relative standard deviation (RSD) of 3.1%. The sensor-to-sensor and run-to-run repeatability for 0.1 μM ethanol detection was found to be 1.8% using the CuO–Fe₂O₃ NCs/μ-chip. To investigate the long-term storage stability, the response of the NC sensor was determined with respect to storage time. The long-term storage stability of the CuO–Fe₂O₃ NCs/μ-chip sensor was investigated under ambient conditions. The sensitivity retained 92% of the initial sensitivity for several days. The above results clearly suggest that the fabricated sensor can be used for several weeks without any significant loss in sensitivity. The dynamic response (0.1 nM to 0.1 mM) of the sensor was investigated from the practical concentration variation curve. The sensor response time is mentioned and investigated using this sensor system under ambient conditions. In Table 1, the performances for ethanol chemical detection based on various modified electrode materials are compared.^60–69 It shows higher sensitivity using the CuO–Fe₂O₃ NCs/chip compared to other nanomaterial-fabricated electrodes with similar target analytes.

Table 1 Comparison of the performances for ethanol detection based on various nanomaterial-fabricated sensors

Materials	Methods	Linear dynamic range, LDR	Sensitivity	Linearity, r^2	Limit of detection, LOD	Response time	Ref.
Ni/Pt/Ti	Potential amperometry	—	3.08 μA mM^−1 cm^−2	—	—	—	60
Ni-doped SnO2 nanostructure	I–V	1.0 nM to 1.0 mM	2.3148 μA cm^−2 mM^−1	0.8440	0.6 nM	10.0 s	61
Pd–Ni/SiNW electrode	Potential amperometry	—	0.76 mA mM^−1 cm^−2	0.9970	10.0 μM	—	62
ZnO–CeO2 nanoparticles	I–V	1.7 mM to 1.7 M	2.1949 μA cm^−2 mM^−1	0.9463	0.6 ± 0.05 mM	10.0 s	63
RuO-modified Ni electrode	Cyclic voltammetry	100–1000 ppm	4.92 μA ppm^−1 cm^−2	—	—	13.0 s	64
CeO2 nanoparticles	I–V	0.17 mM to 0.17 M	0.92 μA cm^−2 mM^−1	0.7458	0.124 ± 0.010 mM	10.0 s	65
Al-doped ZnO nanomaterial	I–V	Up to 3000 ppm	1000 ppm ethanol	—	—	∼8.0 s	66
CuO nanosheets	I–V	Up to 1.7 M	∼0.9722 μA cm^−2 mM^−1	0.7806	0.143 mM	10.0 s	67
Sm-doped Co3O4 nanokernels	I–V	1.0 nM to 10.0 mM	2.1991 ± 0.10 μA cm^−2 mM^−1	0.9065	0.63 ± 0.02 nM	10.0 s	68
Sb2O3–ZnO MFs	I–V	0.17 mM to 0.85 M	5.845 μA cm^−2 mM^−1	0.9989	0.11 ± 0.02 mM	10.0 s	69
CuO–Fe2O3 NCs	I–V	0.1 nM to 0.1 mM	7.258 μA cm^−2 mM^−1	0.9937	0.087 nM	10.0 s	This work

---

Conclusions

CuO–Fe₂O₃ NCs were synthesized using an easy, simple, efficient, reliable, and economical approach using active reducing agents. The optical, elemental, structural, and morphological properties were investigated using FTIR, XRD, XEDS, XPS, FESEM, TEM, DLS, and UV/visible techniques. The CuO–Fe₂O₃ NCs/μ-chip was prepared using a simple fabrication technique and displayed higher sensitivity for chemical sensing. The NCs/μ-chips were efficiently prepared for sensitive and selective ethanol sensing based on chips embedded with doped CuO–Fe₂O₃ NCs with conducting coating binders. The analytical performances of the fabricated ethanol NCs sensors were excellent in terms of sensitivity, detection limit, linear dynamic ranges, and short response time. The CuO–Fe₂O₃ NCs/μ-chips exhibited higher-sensitivity (∼7.258 μA cm⁻² mM⁻¹) and lower-detection limit (∼0.087 ± 0.02 nM), with good linearity and a short response time, and were efficiently utilized as a chemical sensor for ethanol on tiny μ-chips. This novel approach introduced a well-organized route for efficient chemical sensor development for environmental pollutants and health-care fields on a broad scale.

Acknowledgements

This work was supported by the Deanship of Scientific Research (DSR), King Abdulaziz University, Jeddah, under grant no. (130-520-D1435). The authors, therefore, gratefully acknowledge the DSR for technical and financial support.

References

1. N. M. Shaalan, T. Yamazaki and T. Kikuta, Sens. Actuators, B, 2011, 153, 11–16.
2. D. J. Díaz, Book Review of Nanostructure Materials in Electrochemistry, J. Am. Chem. Soc., 2008, 130, 10033–10034.
3. K. Sinkó, N. Hüsing, G. Goerigk and H. Peterlik, Nanostructure of Gel-Derived Aluminosilicate Materials, Langmuir, 2008, 24, 949–956.
4. X. Wang, J. Song, J. Liu and Z. L. Wang, Science, 2007, 316, 102.
5. K. Keren, R. S. Berman, E. Buchstab, U. Sivan and E. Braun, Science, 2003, 302, 1380.
6. D. Chakraborti, J. Narayan and J. T. Prater, Appl. Phys. Lett., 2007, 90, 062504.
7. K. Hara, T. Horiguchi, T. Kinoshita, K. Sayama, H. Sugihara and H. Arakawa, Sol. Energy Mater. Sol. Cells, 2000, 64, 115.
8. Z. L. Wang, J. Phys.: Condens. Matter, 2004, 16, R829.
9. H. T. Ng, J. Han, T. Yamada, P. Nguyen, Y. P. Chen and M. Meyyappan, Nano Lett., 2004, 4, 1247.
10. C. Soci, A. Zhang, B. Xiang, S. A. Dayeh, D. P. R. Aplin, J. Park, X. Y. Bao, Y. H. Lo and D. Wang, Nano Lett., 2007, 7, 1003.
11. Q. H. Li, Y. X. Liang, Q. Wan and T. H. Wang, Appl. Phys. Lett., 2004, 85, 6389.
12. C. J. Lee, T. J. Lee, S. C. Lyu, Y. Zhang, H. Ruh and H. J. Lee, Appl. Phys. Lett., 2002, 81, 3648.
13. M. H. Huang, S. Mao, H. Feick, H. Yan, Y. Wu, H. Kind, E. Weber, R. Russo and P. Yang, Science, 2001, 292, 1897.
14. X. D. Wang, J. H. Song, J. Liu and Z. L. Wang, Science, 2007, 102, 102.
15. Z. Zhang, J. B. Yi, J. Ding, L. M. Wong, H. L. Seng, S. J. Wang, J. G. Tao, G. P. Li, G. Z. Xing, T. C. Sum, C. H. A. Huan and T. Wu, J. Phys. Chem. C, 2008, 112, 9579.
16. Z. H. Cai and C. R. Martin, J. Am. Chem. Soc., 1989, 111, 4138.
17. H. T. Ng, J. Han, T. Yamada, P. Nguyen, Y. P. Chen and M. Meyyappan, Nano Lett., 2004, 4, 1247.
18. C. Soci, A. Zhang, B. Xiang, S. A. Dayeh, D. P. R. Aplin, Y. H. Lo and D. Wang, Nano Lett., 2007, 7, 1003.
19. M. M. Rahman, A. Jamal, S. B. Khan and M. Faisal, J. Phys. Chem. C, 2011, 115, 9503.
20. X. D. Wang, J. H. Song, J. Liu and Z. L. Wang, Science, 2007, 316, 102.
21. M. M. Rahman, M. J. A. Shiddiky, M. A. Rahman and Y. B. Shim, Anal. Biochem., 2009, 384, 159.
22. C. C. Wang, Y. C. Weng and T. C. Chou, Sens. Actuators, B, 2007, 122, 591.
23. M. M. Rahman, A. Jamal, S. B. Khan and M. Faisal, Biosens. Bioelectron., 2011, 28, 127.
24. M. M. Rahman, A. Umar and K. Sawada, Sens. Actuators, B, 2009, 137, 327.
25. D. Chakraborti, J. Narayan and J. T. Prater, Appl. Phys. Lett., 2007, 90, 062504.
26. K. Hara, T. Horiguchi, T. Kinoshita and H. Arakawa, Sol. Energy Mater. Sol. Cells, 2000, 64, 115.
27. Z. L. Wang, J. Phys.: Condens. Matter, 2004, 16, R829.
28. Z. Zhang, J. B. Yi, J. Ding, L. M. Wong, G. Z. Xing, T. C. Sum and T. Wu, J. Phys. Chem. C, 2008, 112, 9579.
29. Z. Zhang, J. B. Yi, J. Ding, L. M. Wong, G. Z. Xing, T. C. Sum and T. Wu, J. Phys. Chem. C, 2008, 112, 9579.
30. M. M. Rahman, Curr. Proteomics, 2012, 9, 272.
31. A. Tao, F. Kim, C. Hess, J. Goldberger, R. He, Y. Sun, Y. Xia and P. Yang, Nano Lett., 2003, 3, 1229.
32. J. J. Vijaya, L. J. Kennedy, G. Sekaran, B. Jeyaraj and K. S. Nagaraja, J. Hazard. Mater., 2008, 153, 767.
33. M. M. Rahman, S. B. Khan, M. Faisal, A. M. Asiri and K. A. Alamry, Sens. Actuators, B, 2012, 171–172, 932.
34. P. P. Sahay and R. K. Nath, Sens. Actuators, B, 2008, 134, 654.
35. Q. H. Li, Y. X. Liang, Q. Wan and T. H. Wang, Appl. Phys. Lett., 2004, 85, 6389.
36. M. M. Rahman, A. Umar and K. Sawada, Sens. Actuators, B, 2009, 137, 327.
37. C. J. Lee, T. J. Lee, S. C. Lyu, Y. Zhang, H. Ruh and J. Lee, Appl. Phys. Lett., 2002, 81, 3648.
38. C. Wongchoosuk, A. Wisitsoraat, A. Tuantranont and T. Kerdcharoen, Sens. Actuators, B, 2010, 147, 392.
39. F. Wang and S. Hu, Microchim. Acta, 2009, 165, 1.
40. M. H. Huang, S. Mao, H. Feick, H. Yan, Y. Wu, H. Kind, E. R. Russo and P. Yang, Science, 2001, 292, 1897.
41. M. M. Rahman, J. Biomed. Nanotechnol., 2011, 7, 351.
42. W. O. Milligan and J. Holmes, X-Ray Diffraction Studies in the System CuO-Fe₂O₃, J. Am. Chem. Soc., 1941, 63, 149–150.
43. E. Sasaoka, M. Hatori, H. Yoshimura, C. Su and M. A. Uddin, Role of H₂O in Oxidation of Spent High-Temperature Desulfurization Sorbent Fe₂O₃ and CuO in the Presence of O₂, Ind. Eng. Chem. Res., 2001, 40, 2512–2517.
44. M. M. Rahman, Reusable and mediator-free cholesterol biosensor based on cholesterol oxidase immobilized onto TGA-SAM modified smart bio-chips, PLoS One, 2014, 9, e100327.
45. M. M. Rahman, Development of mediator-free acetylcholine sensor co-immobilized with acetylcholine oxidase using micro-chips, Curr. Proteomics, 2012, 9, 272–279.
46. H. M. Fan, G. J. You, Y. Li, Z. Zheng, H. R. Tan, Z. X. Shen, S. H. Tang and Y. P. Feng, Shape-Controlled Synthesis of Single-Crystalline Fe₂O₃ Hollow Nanocrystals and Their Tunable Optical Properties, J. Phys. Chem. C, 2009, 113, 9928–9935.
47. S. Ghosh, M. Roy and M. K. Naskar, A Facile Soft-Chemical Synthesis of Cube-Shaped Mesoporous CuO with Microcarpet-Like Interior, Cryst. Growth Des., 2014, 14, 2977–2984.
48. P. M. Rao and X. Zheng, Rapid Catalyst-Free Flame Synthesis of Dense, Aligned α-Fe₂O₃ Nanoflake and CuO Nanoneedle Arrays, Nano Lett., 2009, 9, 3001–3006.
49. Z. Tian, Y. Zhou, Z. Li, Q. Liu and Z. Zou, Generalized synthesis of a family of multishelled metal oxide hollow microspheres, J. Mater. Chem. A, 2013, 1, 3575–3579.
50. H. Zhou and S. S. Wong, A Facile and Mild Synthesis of 1-D ZnO, CuO, and α-Fe₂O₃ Nanostructures and Nanostructured Arrays, ACS Nano, 2008, 2, 944–958.
51. A. F. Carley, P. R. Davies, K. R. Harikumar, R. V. Jones and M. W. Roberts, Reactivity and Structural Aspects of Cesium and Oxygen States at Cu(110) Surfaces: An XPS and STM Investigation, J. Phys. Chem. B, 2004, 108, 14518–14526.
52. X. Deng, J. Lee, C. Wang, C. Matranga, F. Aksoy and Z. Liu, Reactivity Differences of Nanocrystals and Continuous Films of α-Fe₂O₃ on Au(111) Studied with In Situ X-ray Photoelectron Spectroscopy, J. Phys. Chem. C, 2010, 114, 22619–22623.
53. J. Morales, J. P. Espinos, A. Caballero, A. R. Gonzalez-Elipe and J. A. Mejias, XPS Study of Interface and Ligand Effects in Supported Cu₂O and CuO Nanometric Particles, J. Phys. Chem. B, 2005, 109, 7758–7765.
54. S. Pokrel, C. E. Simon, V. Quemener, N. Bârsan and U. Weimer, Investigation of conduction mechanism in Cr₂O₃ gas sensing thick films by ac impedance spectroscopy and work function changes measurements, Sens. Actuators, B, 2008, 133, 78–83.
55. C. Wang, X. Q. Fu, X. Y. Xue, Y. G. Wang and T. H. Wang, Surface accumulation conduction controlled sensing characteristics of p-type CuO nanorods induced by oxygen adsorption, Nanotechnology, 2007, 18, 145506.
56. J. Hagen, Heterogeneous Catalysis: Fundamentals, Wiley-VCH, Weinheim, 1999, pp. 83–206.
57. M. M. Rahman, A. Jamal, S. B. Khan, M. Faisal and A. M. Asiri, Fabrication of Highly Sensitive Acetone Sensor Based on Sonochemically Prepared As-grown Ag₂O Nanostructures, Chem. Eng. J., 2012, 192, 122–128.
58. K. Sahner, R. Moos, M. Matam and J. J. Tunney, Hydrocarbon sensing with thick and thin film p-type conducting perovskite materials, Sens. Actuators, B, 2005, 108, 102–112.
59. K. I. Choi, H. R. Kim, K. M. Kim, D. Liu, G. Cao and J. H. Lee, C₂H₅OH sensing characteristics of various Co₃O₄ nanostructures prepared by solvothermal reaction, Sens. Actuators, B, 2010, 146, 183–189.
60. Y. C. Weng, J. F. Rick and T. C. Chou, A sputtered thin film of nanostructured Ni/Pt/Ti on Al₂O₃ substrate for ethanol sensing, Biosens. Bioelectron., 2004, 20, 41–51.
61. M. M. Rahman, A. Jamal, S. B. Khan and M. Faisal, Highly sensitive ethanol chemical sensor based on Ni-doped SnO₂ nanostructure Materials, Biosens. Bioelectron., 2011, 28, 127–134.
62. B. Tao, J. Zhang, S. Hui and L. Wan, An amperometric ethanol sensor based on a Pd-Ni/SiNWs electrode, Sens. Actuators, B, 2009, 142, 298–303.
63. M. Faisal, S. B. Khan, M. M. Rahman, A. Jamal, A. M. Asiri and M. M. Abdullah, Smart chemical sensor and active photo-catalyst for environmental pollutants, Chem. Eng. J., 2011, 173, 178–184.
64. Y. C. Weng and T. C. Chou, Ethanol sensors by using RuO₂-modified Ni electrode, Sens. Actuators, B, 2002, 85, 246–255.
65. S. B. Khan, M. Faisal, M. M. Rahman and A. Jamal, Exploration of CeO₂ nanoparticles as a chemi-sensor and photo-catalyst for environmental applications, Sci. Total Environ., 2011, 409, 2987–2992.
66. Z. Yang, Y. Huang, G. Chen, Z. Guo and S. Cheng, et al Ethanol gas sensor based on Al-doped ZnO nanomaterial with many gas diffusing channels, Sens. Actuators, B, 2009, 140, 549–556.
67. M. Faisal, S. B. Khan, M. M. Rahman, A. Jamal and A. Umar, Ethanol chemi-sensor: Evaluation of structural, optical and sensing properties of CuO nanosheets, Mater. Lett., 2011, 65, 1400–1403.
68. M. M. Rahman, A. Jamal, S. B. Khan and M. Faisal, Fabrication of Highly Sensitive Ethanol Chemical Sensor Based on Sm-Doped Co₃O₄ Nanokernels by a Hydrothermal Method, J. Phys. Chem. C, 2011, 115, 9503–9510.
69. M. M. Rahman, S. B. Khan and A. M. Asiri, Fabrication of Smart Chemical Sensors Based on Transition-Doped-Semiconductor Nanostructure Materials with μ-Chips, PLoS One, 2014, 9, e85036.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra08224b

---