Gas Sensor Properties of (CuO/WO₃)-CuWO₄ Heterostructured Nanocomposite Materials

Abstract

In this work, we report the evaluation of a (CuO/WO₃)-CuWO₄ heterostructured system as a methanol and acetone gas sensor in different configurations, contrasted with the pure oxides CuO and WO₃. The samples were synthesized using a modified precipitation route followed by a single thermal treatment step to induce multiphase simultaneous crystallization. The structural characterization by XRD showed that all the materials presented the formation of monoclinic CuO and WO₃ and triclinic CuWO₄. No additional phases were detected in the samples, and a reduction in the crystallite size of the CuO phase after the crystallization in the heterostructured system was observed. FE-SEM analysis made it possible to directly observe the morphology and the structures of the samples at the nanometer scale, showing a heterogeneous grain formation and supporting the formation of a heterostructure. UV-Vis DRS was used to study the optical properties of the materials, and the presence of two optical band gaps was successfully determined, which provides further evidence of heterostructure formation via this modified synthesis route. The variation in the resistance of the materials was observed in the presence of methanol and acetone vapors, where the heterostructure exhibited a substantial change in performance in the configuration with 40% copper precursor (Cu40:W60), the sample that presented the highest response as a sensor against these VOCs. To our knowledge, this is the first time that this system has been reported as a gas sensor, using the multiple configurations of the (CuO/WO₃)-CuWO₄ heterostructured system.

1. Introduction

Currently, there is a significant scientific interest in the quantifying pollutant compounds present in different essential natural resources such as water and air. The accelerated development of human activity and industrial development has led to an increase in the emission of highly toxic pollutants to humans. Although these compounds are important in the early stages of industrial processes, when they are released or escape from the processes, they result in potentially harmful and toxic compounds for any living organism. Examples of these gases include ethanol, methanol, acetone, and benzene, among others, generally classified as volatile organic compounds (VOCs). Exposure to concentrations above permissible limits can lead to health problems and even death [1].

Tungsten oxide (WO₃) is among the most cost-effective n-type semiconductors, showing promising performance in many applications, such as solar cells, sensors, photoelectrochemical water splitting, photocatalysis, and superconductivity, among others [2,3,4]. Due to its narrow indirect band gap, between 2.4 and 3.0 eV [5,6], it can absorb visible light more efficiently compared to other commonly studied photocatalytic semiconductors [7,8,9]. However, the electron–hole recombination rate tends to be higher compared to other semiconductors. On the other hand, copper oxide (CuO) is one of the most widely studied p-type oxides. It has an indirect band gap between 1.2 and 2.1 eV [10]. In order to overcome some of the transport disadvantages of many semiconductors and increase their photocatalytic efficiency, heterostructured oxide systems have been developed to reduce charge carrier recombination. Specifically, CuO has been used as a counterpart (p-type) for WO₃ (n-type) in various applications [11,12,13,14,15,16], such as a catalyst in dye degradation processes, organic gas sensors, and solid oxide batteries [14,17,18]. CuO-WO₃ nanofibers (NFs) fabricated via electrospinning method and tested for sarin gas sensor detection demonstrated that the detection performance of WO₃-CuO NFs is better compared to WO₃ and pure CuO and showed excellent selectivity toward trace amounts of sarin gas at room temperature from among several volatile organic compounds [12]. A hydrothermal synthesis of CuO-WO₃ nanocubes for H₂S gas sensing showed substantially greater and faster sensitivity and recovery than those of individual pure samples [19]. The increased surface area of the flower-like morphology, compared to the basic WO₃–CuO composites, led to a higher density of active sites, which are essential for sensor performance. This morphological modification significantly enhanced H₂S detection [16].

Copper tungstate (CuWO₄) could be a key component in enhancing the efficiency of heterojunctions, particularly in CuO/WO₃ systems. It promotes efficient electron transport and reduces recombination at the interface, while minimizing interfacial defects and enhancing charge carrier separation. CuWO₄ can form high-quality interfaces when synthesized via in situ methods, reducing recombination and boosting charge transfer. Studies have shown that CuWO₄-based systems exhibit superior performance in photoelectrochemical cells and gas sensors. CuWO₄ also enhances light absorption and increases the active catalytic surface area, making it a valuable addition to multicomponent systems. Further exploration could lead to significant improvements in material performance [20]. The (CuO/WO₃)-CuWO₄ system has not been previously reported or systematically investigated for gas sensing applications.

In this work, pristine CuO, WO₃, and (CuO/WO₃)-CuWO₄ were synthesized via a modified precipitation step. The materials were characterized using X-ray diffraction (XRD), Fourier transform infrared (FTIR) spectroscopy, scanning electron microscopy (SEM), and UV-Vis diffuse reflectance spectroscopy (UV-Vis DRS). XRD analysis, with the Rietveld refinement step, allowed the phase identification, quantification, and determination of lattice parameters. FTIR spectra provided confirmation of vibrational modes associated with Cu–O and W–O bonds. FE-SEM was used to directly observe the morphology at the nanometer scale, and DRS enabled the estimation of optical band gaps using the Kubelka–Munk model. Finally, the samples were tested as gas sensors by exposing the materials to methanol and acetone vapors in a controlled nitrogen gas flow. This study introduces, for the first time, the evaluation of various configurations of the (CuO/WO₃)-CuWO₄ heterostructure as a gas sensor.

2. Materials and Methods

2.1. Synthesis

The general synthesis process is illustrated in Figure 1. Initially, a stoichiometric quantity of high-purity copper (II) nitrate 3-hydrate (Cu(NO₃)₂⋅3H₂O, 98–103%, PanReac AppliChem ITW reagents, Germany) was dissolved in double-distilled water (DDW). Subsequently, the pH was adjusted to 12 using a 2M sodium hydroxide (NaOH, 98%, PanReac AppliChem ITW reagents) solution, added dropwise. The resulting solution was aged for 0.5 h at 90 °C and 400 rpm. The resulting precipitate was centrifuged and washed with DDW until it reached a pH of 7. The resultant black material was dried for 1 h at 120 °C (A in Figure 1), then was ground into a black powder using an Agate mortar.

For the CuO sample, a portion of the material was heated in a Terrigeno D8-2471 oven (Colombia), with the temperature gradually increasing at a rate of 3 °C/min until reaching 500 °C, and maintained for 2 h. In the case of the heterostructure samples, a portion of the initial black powder was combined with a stoichiometric amount of tungstic acid (B Figure 1) (H₂WO₄, 99%, Sigma Aldrich, USA) to obtain mixtures containing 80, 60, 50, 40, 20, and 0 w/w % of copper precursor. The resulting samples were subjected to the same calcination ramp: 500 °C for 2 h, based on thermal stability and optimized as reported previously [21]. In the case of the WO₃ sample, the same procedure was followed, omitting the copper precursor addition step.

Following the procedure outlined in Figure 1, a set of 7 different samples were obtained, which were labeled as follows in

Table 1. Sample labeling.

Label	Weight Percentage (w-w %)
	Copper Precursor (A)	Tungsten Precursor (B)
WO3	0	100
Cu20:W80	20	80
Cu40:W60	40	60
Cu50:W50	50	50
Cu60:W40	60	40
Cu80:W20	80	20
CuO	100	0

.

2.2. Characterization

The structural characterization of the samples was conducted through XRD analysis to identify the crystal structure and quantify the phase composition. An X’pert Pro Panalytical diffractometer (Amsterdam, The Netherlands) with a real-time multiple strip (RTMS) detector in Bragg–Brentano mode (Cu Kα radiation of λ = 1.542 Å) was used; the 2θ angle range was set between 20 and 80° with a step size of 0.026°. FTIR spectroscopy was used to confirm the presence of characteristic vibrational bands on the surface of the samples. A Shimadzu Prestige-21 IR spectrometer (Kyoto, Japan) with a resolution of 8 cm⁻¹ and 8 runs was used for these measurements. Morphological details were observed through FE-SEM using a Tecnai F20 Super-Twin TMP microscope (United States) operating at an accelerating voltage of 200 kV, and the samples were dispersed in absolute ethanol and deposited on copper grids with 200 Mesh Lacey/Carbon membranes. Particle size distribution was estimated from FE-SEM images using ImageJ (1.53j) software by measuring at least 150 particles per sample. Finally, the determination of the band-gap energy of the samples was performed from the reflectance spectra obtained by UV-Vis DRS with a UV-Vis Cary 5000 spectrophotometer (Malaysia) in diffuse reflectance mode, using the Kubelka–Munk based on the absorbance data obtained.

2.3. Methanol and Acetone Sensing Test

The sensor responses were evaluated by depositing a mixture of the powders dispersed in isopropanol on substrates with interdigitated electrodes. The sensors were introduced in a stainless-steel test chamber for the sensing tests. The experimental bench for the electrical characterization of the sensors allows measurements to be carried out in a controlled atmosphere. Gases from certified cylinders were diluted in nitrogen (60 sccm) to the desired concentrations using mass flow controllers. Sensing measurements were carried out in the temperature range of 150–400 °C, with steps of 50 °C, under a gas (methanol or acetone) total stream of 30 sccm, collecting sensor resistance data in two-point mode under the applied bias of +10 V. A digital multimeter was used for data acquisition. The gas response is defined as the ratio R_a/R_g, where R_a is the electrical resistance of the sensor in dry air and R_g is the resistance at different methanol and acetone concentrations [22].

3. Results and Discussion

3.1. Structural Characterization

The structural properties, crystalline structure, and the phase composition of the samples were analyzed through XRD analysis. Rietveld refinement was conducted using GSAS-II software [23,24]. Figure 2 shows the plots generated from the refinement of the samples. The Rietveld adjustment showed a good agreement between the theoretical model and the experimental data, as confirmed by the low chi-squared (${\chi }^2$) value (

Table 2. Rietveld refinement parameters of the synthesized samples.

	Phase	CuO	Cu80:W20	Cu60:W40	Cu50:W50	Cu40:W60	Cu20:W80	WO3
a (Å)	CuO	4.69	4.69	4.68	4.69	4.69	4.68	-
	WO3	-	7.50	7.30	7.31	7.31	7.33	7.32
	CuWO4	-	4.70	4.70	4.71	4.71	4.70	-
b (Å)	CuO	3.43	3.43	3.43	3.42	3.42	3.40	-
	WO3	-	7.60	7.52	7.53	7.53	7.56	7.54
	CuWO4	-	5.84	5.83	5.84	5.84	5.83	-
c (Å)	CuO	5.13	5.14	5.13	5.14	5.13	5.12	-
	WO3	-	7.65	7.67	7.69	7.69	7.72	7.71
	CuWO4	-	4.88	4.87	4.88	4.88	4.88	-
V (Å3)	CuO	82.52	82.69	82.35	82.44	82.28	81.47	-
	WO3	-	436.05	421.05	423.29	423.29	427.80	425.54
	CuWO4	-	133.95	133.44	134.23	134.23	133.72	-
α (°)	CuO	90						-
	WO3	-	90
	CuWO4	-	91.67	91.66	91.66	91.65	91.66	-
β (°)	CuO	99.52	99.52	99.55	99.54	99.54	99.60	-
	WO3	-	90.42	90.50	90.45	90.42	90.44	90.56
	CuWO4	-	92.51	92.50	92.50	92.50	92.50	-
γ (°)	CuO	90						-
	WO3	-	90
	CuWO4	-	82.78	82.79	82.78	82.78	82.79	-
Chi ${(\mathit{\chi }}^2)$		1.28	1.78	1.85	1.48	1.92	2.81	1.70

). This confirms the quality of the adjustment and the accuracy of phase identification. The CuO and WO₃ contrast samples did not show the presence of secondary phases. All the samples with a copper precursor showed characteristic peaks around 35.5° and 38.8°, attributed to the diffraction planes, (−111), (111), respectively, and indexed to the monoclinic crystalline phase of CuO with PDF No. 01-072-0629 and the C2/c (15) space group. The presence of these characteristic peaks was reported previously [25,26,27].

Additionally, in the samples with a tungsten precursor, characteristic signals were observed around 23.1°, 23.6°, and 24.4° corresponding to the diffraction planes, (002), (020), and (200), respectively, and these signals are consistent with the monoclinic WO₃ phase, PDF No. 01-072-1465 with the P21/n (14) space group system. These results are comparable to those reported by Kang et al. [28]. Finally, in all the heterostructure samples, the CuWO₄ triclinic phase was observed, and characteristic signals around 19.2, 28.3, and 30.5° indexed to the diffraction planes, (100), (−1−11), and (111), respectively, confirm the presence of this phase, which matches PDF No. 01-070-1732 and the P-1 (2) space group, previously reported at this crystallization temperature as α-CuWO₄ [29]. This additional phase is commonly formed due to the coexistence of copper and tungsten precursors. Its formation results from a secondary reaction between these compounds, occurring concurrently with the formation of CuO and WO₃ phases [30]. The presence of these distinctive peaks confirms the coexistence of all the phases in the synthesized samples. Overall, the existence of distinct phases in the samples indicates the formation of a multicomponent system. This coexistence could indicate an interaction between the phases, potentially influencing the material properties [15,20].

The weight percentage of the phases in the synthesized samples, presented in Figure 3, reveals a clear variation in the phase composition among the samples. In Cu80:W20, monoclinic CuO was the majority phase with 88.81%, then a small proportion of CuWO₄ (9.47%) and trace amounts of WO₃ (1.72%). A significant rise in the percentage of the WO₃ phase is seen when the copper precursor concentration decreases, reaching 43.17% in the Cu20:W80 and 100% in the pure WO₃ sample. Conversely, the CuWO₄ phase rises gradually to reach 40.2% in the Cu20:W80 sample, suggesting that the CuWO₄ content stabilizes in the samples with lower CuO content. These differences in phase composition can directly influence the physicochemical properties of materials, particularly their sensitivity as gas sensors, given that the WO₃ phase is known for its good response in this type of application [31,32]. The observed phase evolution suggests that controlling the initial composition is a key factor in tuning the final properties of the materials. In addition, the obtained lattice parameters (Table 2) suggest slight variations in the crystal structure of the phases, which could be related to the generation of structural defects or structural distortions inherent to the synthesis methods used.

Additionally, the average crystallite size (L) was determined using the Debye–Scherrer equation. In Equation (1), the Scherrer constant $K=0.94,$ $\lambda =0.1540598 \mathrm{n}\mathrm{m}$, $\beta$ is the full width at half maximum (FWHM), and $\theta$ is the Bragg angle. In all the samples, the maximum intensity peak of each phase in the diffractogram was selected for crystallite size estimation. In addition, Rietveld L was also estimated in both perpendicular (L⊥) and parallel (L∥) directions with Equations (2) and (3), based on the Lorentzian component parameters (LX and ptec) from GSAS refinement. K and λ correspond to Scherrer parameters [33,34].

$$L={\displaystyle \frac{K\lambda }{\beta cos\theta }}$$

(1)

$$\mathrm{L}\perp ={\displaystyle \frac{1800 \mathrm{K}\lambda }{\pi LX}}$$

(2)

$$\mathrm{L}\parallel ={\displaystyle \frac{1800 \mathrm{K}\lambda }{\pi (LX+ptec)}}$$

(3)

Figure 4A shows the variation in Scherrer and Rietveld L for CuO, WO₃, and CuWO₄ phases between the synthesized samples. The parallel and perpendicular Rietveld L mean values show similar trends to the results obtained by Scherrer. The results show that CuO phase crystallite size decreases with lower copper precursor content, with minimum values observed in Cu20:W80. WO₃ phase crystallite size remains relatively constant between Cu80:W20 and Cu20:W80, with the lowest values found in Cu80:W20 and pure WO₃ samples.

For the CuWO₄ phase, crystallite size increases until reaching a maximum in the Cu40:W60 sample, then decreases in samples with lower copper content, such as Cu20:W80. The Rietveld refinement values provide additional information on particle anisotropy, with a noticeable increase at intermediate compositions, especially at Cu40:W60, where both CuWO₄ and WO₃ reach maximum crystallite sizes. This suggests that crystallite growth is more pronounced in this specific composition, possibly due to phase stabilization or interactions among the constituent oxides. The analysis of crystallite sizes demonstrates the influence of material composition on structural properties, particularly for the WO₃ phase, known for its good gas detection performance [35].

The dislocation density ($\delta$) and microstrain ($\epsilon$) shown in Figure 4B were also determined with Equations (4) and (5), respectively, as follows in [36]. $\epsilon$ represents lattice resulting from imperfections or irregularities in the atomic structure [37]. The CuO and WO₃ phases showed the highest microstrain, possibly due to their nanocrystalline nature and intrinsic properties. This strain could be favorable for applications like gas sensing and photocatalysis, where surface defects and lattice distortions enhance the reactivity and sensitivity [35]. δ is inversely related to crystallite size, affecting both mechanical and electronic properties. CuO shows a higher dislocation density compared to CuWO₄ and WO₃, indicating a higher density of internal defects associated with its smaller crystallite size. These dislocations can enhance mechanical strength and act as charge carrier traps, affecting the electrical properties of CuO and metal oxides in general, especially in semiconductor devices [38,39].

$$\mathrm{\delta }={\displaystyle \frac{1}{L^2}}$$

(4)

$$\mathrm{\epsilon }={\displaystyle \frac{B\cos \theta }{4}}$$

(5)

The vibrational modes $\left(v\right)$ of the synthesized materials were determined through FTIR spectroscopy. The experimental spectra, presented in Figure 5, reveal intense peaks in the region of 500–1100 cm⁻¹. The red shifts (lower wavenumber displacements) and blue shifts (higher wavenumber displacements) presenting the different $v_n$, are attributed to dislocations and structural defects in the materials [40]. This variation is consistent with the structural distortions observed in the XRD analysis. These results provide insight into the bonding environment within the heterostructures. The presence of multiple metal–oxygen stretching modes indicates the coexistence of Cu-O and W-O bonds; the observed vibrational bands are reported and assigned in

Table 3. IR band assignment for the vibrational bands.

	Wavenumber Range (cm⁻¹)	Assignation	Reference
$v_1$	555–560	Metal–oxygen–metal bond vibrations (W-O-W, Cu-O-Cu)	[41,42]
$v_2$	590–605
$v_3$	620–650	W-O-W stretch vibration modes	[14]
$v_4$	670–720	Vibration of O-W-O-W-O flexion	[43]
$v_5$	800–810	O-W-O stretch modes	[14]
$v_6$	870–920	W-O bond stretch mode in the octahedral WO6 structure (CuWO4)	[29,44,45]

. The $v_6$ band, corresponding to W-O stretching in tetrahedral coordination, confirms the successful formation of the ternary structure. This analysis supports the phase identification obtained from the XRD results.

3.2. Morphological Characterization

FE-SEM images (Figure 6) show the morphology of the CuO, WO₃, Cu60:W40, and Cu50:W50 samples. The morphology of the samples varies significantly, and two predominant particle size distributions were observed for the copper- and tungsten oxide-based samples except for Cu80:W20. The CuO sample (Figure 6A) shows a porous, network-like structure composed of densely agglomerated nanoparticles. Sample Cu60:W40 (Figure 6B) shows more compact and well-defined polyhedral grains, attributed to the presence of the CuWO₄ phase, which appears exclusively in samples containing the CuWO₄ phase; agglomerated particles were also observed in the sample. The Cu50:W50 sample (Figure 6B) exhibits granular structures similar to those of Cu60:W40., and this morphology was constant across all the heterostructures samples, except Cu80:W20. These findings confirm that the phase composition directly influences the morphology of the phases as the composition of the system changes, and they also indicate that co-crystallization affects structural features. The sample WO₃ (Figure 6D) shows irregular nanoparticle agglomeration with smooth surfaces. These morphological changes reflect differences in phase distribution, synthesis conditions, and particle growth mechanisms between samples, which influence both texture and particle size distribution. The particle size distributions for the samples are reported in

Table 4. Particle size distributions of the samples.

Sample	Small Particle Size (nm)	Large Particle Size (nm)
CuO	130.8 ± 30.6	-
Cu80:W20	80.9 ± 28.3	-
Cu60:W40	100.5 ± 63.3	407.5 ± 95.2
Cu50:W50	96.3 ± 47.9	408.8 ± 80.6
Cu40:W60	84.9 ± 24.9	407.1 ± 81.2
Cu20:W80	82.6 ± 45.9	426.5 ± 159.3
WO3	99.7 ± 47.5	-

.

The small particle sizes range from 80.9 nm to 130.8 nm, with the CuO sample exhibiting the largest average size (130.8 ± 30.6 nm) and the Cu80:W20 sample showing the smallest average (80.9 ± 28.3 nm). Large particles, on the other hand, exhibit more uniform sizes throughout the samples, ranging in size from 407.1 nm to 426.5 nm. For example, samples Cu60:W40, Cu50:W50, and Cu40:W60 contain large particles with comparable diameters around 407.1 nm, suggesting that particle growth in these compositions may be stabilized. This bimodal particle size distribution behavior might be attributed to the synthesis conditions and the heterogeneity of the samples, with various components and phases leading to the formation of particle sizes of varying sizes. Larger particles may result from the agglomeration or coalescence of smaller crystallites during growth, whereas the smaller particles correspond to isolated crystalline domains [46].

3.3. Optical Properties

The optical properties of the obtained samples were analyzed via UV-Vis DRS. Figure 7 presents the absorbance behavior of the samples, where the CuO sample exhibits the broadest absorption range, which extends approximately from 200 to 800 nm with a strong peak at 300 nm, and undergoes a red shift as the CuO content decreases in the heterostructured samples. In these, the absorption range broadens with increasing CuO content, with the Cu80:W20 and CuO samples showing the widest absorption range. This behavior suggests that CuO has a greater ability to absorb light along the visible spectrum due to its narrower band gap compared to the WO₃ and CuWO₄ phases. The WO₃ sample exhibited a limited absorption range, which could be attributed to the absence of oxygen vacancies or defect states [47], which is consistent with previous reports [48]. Previous studies have shown that the formation of the CuWO₄ phase also induces a red shift in the optical absorption spectrum. This shift is attributed to the narrowing of the band gap, which enhances light absorption at longer wavelengths, particularly in the visible region [20,49]. In general terms, these findings demonstrate that the heterostructure generated by different oxides outperforms the individual oxides in terms of optoelectronic application due to the optimized optical response within the visible spectrum.

The optical band gap energy ($E_g$) of the samples was determined via UV-Vis DRS, using the Kubelka–Munk transform method, which assumes absorption and scattering as first-order phenomena [50,51]. The line segment x-intercept of the ${\left(\alpha hv\right)}^{1/n}$ versus the energy of the incoming photon ($hv$) (Figure 8), derived from Equation (6), was used. In this, C is a proportionality constant, and n is a variable coefficient that changes based on the kind of electronic transition; direct permitted transitions and indirect permitted transitions have n = 1/2 and n = 2, respectively. $h$ is the Planck constant and $\nu$ is the light frequency of the incident light [52,53,54].

$$\alpha hv=C{\left(hv-E_g\right)}^n$$

(6)

The approximate direct and indirect band-gap values of the samples are displayed in

Table 5. Variations in the direct and indirect optical band gaps of the samples.

	Direct Gap (eV)	Indirect Gap (eV)
	$\mathit{E}{\mathit{g}}_1$	$\mathit{E}{\mathit{g}}_1$	$\mathit{E}{\mathit{g}}_2$
CuO	2.44	1.28	-
Cu80:W20	2.68	1.39	-
Cu60:W40	2.43	1.48	1.62
Cu50:W50	2.45	1.48	1.60
Cu40:W60	2.46	1.33	1.93
Cu20:W80	2.72	-	2.04
WO3	2.96	-	2.44

. The identification of different indirect band gaps suggests the coexistence of multiple semiconductor phases. This result, together with structural and morphological evidence, supports the formation of a heterostructure. From the CuO to Cu40:W60 samples, two different band gaps were observed, although three semiconductor phases coexist in the heterostructured samples (CuO, WO₃, and CuWO₄), and only two distinct optical transitions were observed in the UV–Vis DRS analysis. This is attributed to the close proximity of the indirect band gaps of WO₃ (2.44 eV) and CuWO₄ (~2.35 eV [20,55]), which likely leads to overlapping absorption features that cannot be individually resolved in Kubelka–Munk plots [56,57]. As a result, a single indirect transition is observed in that region. In contrast, the CuO phase exhibits a lower band gap (1.28 eV), which is clearly distinguishable. Therefore, the two observed band gaps correspond to CuO ($E_{g1})$ and the combined contribution of WO₃/CuWO₄ ${(E}_{g2})$. The band gaps of some heterostructures were lower than those of the individual oxides after the simultaneous crystallization step, suggesting a potential extension of the system’s light absorption range. This suggests that the heterostructured samples are capable of absorbing visible light [15]. The optical response in the heterostructure is often enhanced by a reduced optical band gap. This narrower band gap makes it easier for electron–hole pairs to form and move between CuO, CuWO₄, and WO₃, which may improve the synergistic optical behavior of the combined oxides [56]. Accordingly, the CuO and Cu40:W60 samples exhibit promising optical performance.

3.4. Gas Sensing Test

The I–V (current vs. voltage) curves presented in Figure 9 exhibit typical ohmic behavior in the CuO, Cu20:W80, Cu40:W60, Cu50:W50, Cu60:W40, and Cu80:W20 samples, indicating low contact resistance and charge transport within the applied voltage range [12,58,59]. A device with this behavior will ensure functionality at low power and with minimum energy consumption. In contrast, the nonlinear behavior of the WO₃ sample at higher voltages is mainly due to a space charge-limited current mechanism. In this case, for low values of voltage, the current exhibits near-ohmic behavior; however, upon increasing the voltage, charge traps within the material limit current flow, resulting in this phenomenon [60,61].

On the other hand, the analysis of resistance as a function of temperature (Figure 10) clearly shows that the resistance decreases with increasing temperature due to the granular nature of the material, which creates multiple potential barriers between grains, as well as depletion layers within the heterostructure channels. The optimum operating temperature of a sensor directly affects its resistivity by balancing the adsorption and desorption of gas molecules on the sensor surface, with higher resistance at moderate temperatures due to oxygen adsorption, whereas at higher temperatures, desorption reduces the sensor response. These findings agree well with metal oxide-based materials, which constitute one of the most studied families of materials in gas sensors due to their high sensitivity and thermal stability [59,62,63].

In the gas sensor analysis, the sample sensors showed a good response to methanol/acetone vapors. Figure 11 shows the resistance profile of all samples exposed to a fixed vapor concentration in N₂ over time. Exposure to methanol (Figure 11A) and acetone (Figure 11B) induces a decrease in resistance, characteristic of the semiconductor’s response to reducing gases. The novel heterostructure system increases the number active sites, which are essential for gas sensing performance, observing a higher response or resistance change in the Cu40:W60 and Cu50:W50 samples. This enhanced behavior is attributed to the multicomponent nature of the sensor, made from metal oxides, which works on the principle of heterostructures between p-type and n-type materials, such as CuO and WO₃-CuWO₄, respectively. In this type of sensor, upon being brought into contact with a p-type material with an n-type material, an electron depletion layer forms in the n-type material, while a hole depletion layer forms in the p-type material, resulting in the formation of a potential barrier at the interface [62].

In contact with oxygen, the sensor surface promotes its adsorption, forming adsorbed oxygen species (${\mathrm{O}}^{2-},{\mathrm{O}}^{-},{\mathrm{O}}_2^{-}$), which subsequently capture electrons from the n-type materials (WO₃ and CuWO₄), leading to an expansion of the depletion layer and an increase in sensor resistance. When exposed to reducing gases such as acetone and methanol, the oxygen ions combine with such gases, releasing electrons, which reduces both the potential barrier and the depletion layer width, thereby decreasing the resistance. The gases react with the oxygen species on the material surface, producing byproducts such as CO₂ and H₂O, among others, depending on the gas’s nature [47,63,64,65]. A schematic representation of this mechanism is shown in Figure 12.

Figure 13 shows how a gas sensor measures the change in resistance from its initial state (Ra) to its state after gas exposure (Rg), enabling the calculation of the sensor response as Ra/Rg for the detected gas. The improved behavior observed in the p–n heterostructure exhibits a faster and more sensitive response and recovery when compared to materials without this type of configuration. For the most responsive material configuration (Cu40:W60), the response to methanol was higher than acetone, 2.2 and 1.7, respectively. This difference is usually reported for semiconductor-based materials [66]. For the CuO nanowires, the ethanol response was reported as 1.5 [67] and 1.4 for methanol [66], and for the WO₃ thin films, the response was 1.5 for ethanol [68]. The enhanced response of the heterostructured system over the pristine WO₃ sample and high WO₃ phase percentage (Cu80:W20) sample shows a superior performance in samples with higher CuO content (p-type component), as previously observed in CuO/WO₃ heterostructures [69].

4. Conclusions

Gas sensing measurements performed on (CuO/WO₃)-CuWO₄ heterostructures show promising results in detecting methanol and acetone vapors. The ohmic behavior observed in most samples through the I-V curves indicates low contact resistance and good electrical conductivity, which is essential for the operation of low-power sensors. The heterostructure samples, especially the Cu40:W60 and Cu50:W50 samples, exhibited an improved gas sensitivity, which was tuned by n-type and p-type sensing mechanisms that promote electron transfer between the p-type CuO and n-type WO₃-CuWO₄ components. This configuration increases the number of active sensing sites, leading to significant changes in resistance upon exposure to reducing gases such as methanol and acetone. The Cu40:W60 sample exhibited excellent sensitivity, especially toward methanol, with a response ratio of 2.2 compared to 1.7 for acetone. These results highlight the potential application of (CuO/WO₃)-CuWO₄ heterostructures in gas sensing due to their high sensitivity, good stability, and rapid response times.