NO₂ Gas Sensing Behavior of ZnO/Ag/AG₂O Thin Films Prepared by Pulsed Laser Deposition

Hazardous gases are commonly divided into oxidizing and reducing gases, and nitrogen dioxide (NO₂) is among the most important oxidizing gases because of its toxicity and its role in atmospheric photochemical processes that contribute to ground-level ozone and acid rain [1, 2]. Low-cost resistive gas sensors are still appealing for environmental monitoring due to their sensitivity to surface adsorption processes, ease of fabrication, and integration with electronic circuits [3, 4]. The target gas and adsorbed oxygen species interact in these sensors to alter the carrier concentration and potential barrier at the surface of the semiconductor, resulting in a detectable change in resistance [5-7].

The temperature-dependent performance of the %Ag sample reveals a distinct operating regime that is affected by thermal activation of the surface. It is hypothesized that high silver concentration may act as a physical barrier or create deep electron traps, such as Schottky barriers, thereby reducing the response [8]. Due to its high specific surface area, open transport channels, and compatibility with silicon technology, porous silicon is a useful substrate for gas sensing [9-13]. Due to its n-type semiconducting behavior [14], wide direct band gap (~3.37 eV) [15], chemical stability [16], and ease of preparation by several thin-film routes including pulsed laser deposition (PLD) [17], zinc oxide (ZnO) is also a well-established gas-sensing material [18-20]. Combining ZnO with a porous substrate is therefore an effective strategy for increasing the density of active sites available for gas adsorption [21, 22]. A further route to improving ZnO-based sensors is the addition of a catalytically active second phase [23]. Adding Ag is of particular interest because it can modify the adsorption chemistry [24], promote spillover-type effects [25], and change the electronic barrier structure at the ZnO surface [26, 27]. Recent studies on Ag-decorated ZnO for NO₂ sensing have shown that moderate Ag loading can enhance the response substantially, but the improvement is not monotonic with Ag content [28-32]. Despite this general trend, the role of Ag addition in PLD-grown ZnO/Ag/Ag₂O thin films remains insufficiently clarified, especially when Ag loading was varied systematically on a PS [33, 34]. The aim of the present work is therefore to examine the effect of nominal Ag content on the structural, morphological, and NO₂ gas-sensing properties of ZnO/Ag/Ag₂O thin films prepared by PLD. The discussion focuses on the internal consistency of the measured data, the relationship between structure and sensing behavior, and the identification of the Ag content that provides the best performance within the studied range.

3.1 X-ray diffraction analysis

According to the assignments, the standard reference cards are JCPDS 36-1451 for hexagonal wurtzite ZnO and JCPDS 43-1038 for cubic Ag₂O. It was carefully considered that metallic Ag (JCPDS 04-0783) was an alternative candidate. The XRD patterns of the ZnO/Ag/Ag₂O films are shown in Figure 1. The prepared film exhibits the characteristic reflections of hexagonal wurtzite ZnO at 2θ ≈ 31.7°, 34.6°, 36.5°, 47.7°, and 56.8°, corresponding to the (100), (002), (101), (102), and (110) planes, respectively [28, 36]. A broad feature centered near 28.5° is assigned to the porous silicon substrate.

1.png

Figure 1. X-ray diffraction (XRD) patterns of ZnO/Ag/Ag₂O thin films with nominal Ag contents of 0%, 5%, and 10% deposited on porous silicon

In the 5% Ag and 10% Ag films, additional peaks appear near 32.7°, 37.8°, and 54.5° [37]. These additional reflections are confidently assigned to the cubic Ag₂O phase by more reasonably assigning these peaks to Ag₂O-related reflections than to additional ZnO planes.

The corrected peak positions and assignments are shown in Table 1. In contrast to the Ag₂O-related peaks, which provide somewhat larger values up to 20.6 nm, ZnO peaks are associated with crystallite sizes (D) roughly within range (5.5–12 nm). In Table 2 and Figure 2, the recalculated lattice parameters from the ZnO (100) and (002) reflections are detailed [38]. The values change only slightly across the series (a = 3.2255 Å and c = 5.1731 Å), indicating that Ag addition does not strongly distort the ZnO lattice within the resolution of the present XRD data.

2.png

Figure 2. Variation of recalculated lattice parameters with nominal Ag content

Table 1. Corrected X-ray diffraction (XRD) peak positions and phase assignments for ZnO/Ag/Ag₂O thin films

Note: D = crystallite size, FWHM = full width at half maximum. D is the lattice parameter of the hexagonal wurtzite ZnO structure.

Table 2. Recalculated lattice parameters of ZnO/Ag/Ag₂O thin films from the ZnO (100) and (002) reflections

Note: a and c are the lattice parameters of the hexagonal wurtzite ZnO structure.

The principal structural effect of Ag addition is therefore the appearance of a secondary Ag-containing phase rather than a large shift of the ZnO host lattice [39].

The sensing mechanism of the chemical state was selected by the optimal sample that was found to be 5% Ag before and after thermal treatment at the working temperature (250 ℃) with the target gas ambient and NO₂. According to reference [40], the core-level Ag 3d₅/₂ spectra reveal that at RT. Predominantly, silver exists as Ag⁺ in the Ag₂O state with a binding energy of 367.8 eV, aligning with the XRD findings.

At 250 ℃, a highly dynamic induced redox equilibrium was enhanced by the sensing mechanism, whereas the co-existence of Ag₂O and metallic Ag clusters drives two complementary processes, Electronic Sensitization and Chemical Sensitization [41].

The crystallite sizes were estimated using the standard Scherrer equation; we have updated the text to confirm that the observed FWHM was corrected for instrumental broadening using a crystalline standard. Furthermore, a clarifying statement was added to the text emphasizing that, because the Scherrer equation does not decouple size from macrostrain effects, these values are considered order-of-magnitude estimates rather than absolute physical sizes and were added to the crystallite size values in Table 1.

As detailed in Table 2 and Figure 2, the calculated lattice parameters from the ZnO (100) and (002) reflections exhibit a distinct composition-dependent behavior. For the 5% Ag film, the lattice parameters show only minor variations (a = 3.2291 Å and c = 5.1731 Å) compared to the undoped ZnO. This suggests that at moderate concentrations, Ag predominantly segregates to form a secondary Ag₂O phase without severely distorting the host ZnO lattice. However, increasing the nominal Ag content to 10% induces a significant anisotropic structural distortion, characterized by a drastic expansion of the c-axis (c = 5.1784 Å) and a noticeable contraction of the a-axis. This pronounced shift indicates that excessive Ag loading generates substantial interfacial stress and macrostrain between the ZnO matrix and the heavily agglomerated Ag₂O secondary phase. Furthermore, it may also reflect the forced interstitial incorporation or substitution of the larger Ag⁺ ions (compared to Zn²⁺) into the ZnO crystal lattice at higher concentrations.

3.2 Field-emission scanning electron microscopy morphology

Figure 3 presents top-view FE-SEM images at low and high magnification. A cross-sectional and top-view FE-SEM image of the bare porous silicon substrate before thin-film deposition has been investigated. The bare substrate exhibits an average pore diameter around 45 nm and a cross-sectional layer thickness up to 3.2 µm. All samples retain the porous architecture of the Ag/Ag₂O substrate, with pore openings on the order of micrometers and thin separating walls. The undoped ZnO/Ag/Ag₂O film shows an open porous network with comparatively fine surface coverage.

3.jpg

Figure 3. Field-emission scanning electron microscopy (FE-SEM) images of ZnO/Ag/Ag₂O thin films with nominal Ag contents of (A) 0%, (B) 5%, and (C) 10% at two magnifications

The 5% Ag film exhibits more uniform decoration of the pore walls by faceted grains, while the porous structure remains clearly accessible. In contrast, the 10% Ag film appears denser, with larger aggregates and more local pore coverage. From a gas-sensing perspective, the 5% Ag morphology is the most favorable because it combines accessible pores with an increased density of surface-active grains. The FE-SEM observations, therefore, agree with the electrical measurements, where the 5% Ag film gives the highest NO₂ response.

3.3 Fourier-transform infrared spectroscopy analysis

Table 3 summarizes the bands observed, and Figure 4 displays the FTIR spectra, where a = 0%, b = 5%, and c = 10% of the added Ag ratio. The wide band observed at 3367-3392 cm⁻¹ is attributed to the O-H stretching of surface hydroxyl groups and/or adsorbed water [42].

Table 3. Tentative Fourier-transform infrared spectroscopy (FTIR) band assignments for ZnO/%Ag/Ag₂O thin films

The band near 1688–1695 cm⁻¹ is attributed to the bending vibrations of molecular water [43]. The bands in the 2882–2886 and 1059–1084 cm⁻¹ regions are tentatively attributed to surface vibrations related to C–H and C–O [44], as well as residual carbonaceous species. The low wavenumber band at 455–476 cm⁻¹ is consistent with a Zn–O vibration [45].

In addition, Ag shows an extra band near 1175 cm⁻¹. This feature is regarded as a potential band related to the surface rather than conclusive evidence of a particular metal-oxygen bond. Therefore, the FTIR results corroborate the existence of ZnO and common surface adsorbates; however, they should not be overinterpreted beyond that threshold.

4.png

Figure 4. Fourier-transform infrared spectroscopy (FTIR) spectra of ZnO/Ag/Ag₂O thin films with nominal Ag contents of 0%, 5%, and 10%

3.4 NO₂ sensing performance

Dynamic resistance transients for 50 ppm NO₂ at 200, 250, and 300 ℃ are shown in Figure 5. For all compositions, the resistance increases during NO₂ exposure and decreases when clean air is reintroduced, which is the expected behavior of an n-type oxide exposed to an oxidizing gas. The response is reversible over the recorded cycles, although the recovery behavior depends strongly on temperature and Ag content.

5.png

Figure 5. Dynamic resistance transients of ZnO/Ag/Ag₂O sensors toward 50 ppm NO₂ at 200, 250, and 300 ℃

Figure 6 and Table 4 summarize sensitivity, response time, and recovery time. The undoped film shows its maximum sensitivity at 250 ℃ (31.0%). The 5% Ag film also reaches its optimum at 250 ℃, but with a much larger sensitivity of 68.0%. This value is about 2.19 times that of the undoped film at the same temperature and about 2.64 times that of the 10% Ag film at 250 ℃. The 10% Ag film exhibits distinct behavior, with its sensitivity increasing gradually from 13.3% at 200 ℃ to 28.2% at 300 ℃. This suggests that the optimal response moves toward a higher operating temperature as the Ag loading increases.

6.png

Figure 6. Sensitivity of ZnO/Ag/Ag₂O sensors toward 50 ppm NO₂ as a function of operating temperature

Table 4. Sensitivity, response time, and recovery time of ZnO/Ag/Ag₂O sensors toward 50 ppm NO₂

Response and recovery times should be interpreted together with sensitivity rather than separately. For the undoped film, the shortest recovery time (30 s) occurs at 300 ℃ rather than at the maximum sensitivity. For the 5% Ag film, the response time remains 20 s at all three temperatures, while the recovery time decreases from 60 s at 250 ℃ to 40 s at 300 ℃. The 10% Ag film shows the slowest overall activity behavior at 300 ℃ with time (30 s response and 70 s recovery). Within the tested range, 5% Ag therefore provides the best overall trade-off between response magnitude and dynamic behavior.

3.5 Sensing mechanism

Resistance characteristics and formal time criteria are critical for an objective comparison of gas-sensing dynamics. The formal definitions of response and recovery times clarify the baseline metrics, and standard deviations from multiple repeated cycles are shown in Table 4. The sensing behavior can be interpreted in terms of adsorption-induced modulation of the electron depletion layer in ZnO [31, 32]. In the ambient air, oxygen adsorbs on the surface and captures electrons from the conduction band.

In the 200–300 ℃ range, O-species become increasingly important. When NO₂ is introduced, the oxidizing gas further withdraws electrons either directly or through reactions involving preabsorbed oxygen, which widens the depletion layer and increases the sensor resistance. The higher NO₂ response (68%) were observed, which allows electrons to be captured directly from the ZnO matrix conduction band. The introduction of the Ag/Ag₂O phase boosts this pathway by forming local p-n heterojunctions.

These heterojunctions undergo a massive depletion width modulation exclusively in the presence of strongly oxidizing gases, rather than non-specifically shifting general adsorption behavior [46]. Ag addition can enhance this process in two complementary ways. First, Ag/Ag₂O-related surface sites may promote catalytic adsorption and activate surface oxygen.

Second, the interface between ZnO and the Ag-containing phase can increase electronic sensitization and strengthen the modulation of the interfacial barrier [47]. The superior performance of the 5% Ag film suggests that moderate Ag loading increases the density of active sites without destroying the open porous network. The weaker response is in line with the larger Aggregates seen by FE-SEM at 10% Ag, which probably lessens the benefit of porosity and introduces less favorable conduction pathways. The sensing mechanism is driven by the formation of local ZnO/Ag₂O heterojunctions.

In clean air, the band alignment causes electrons to flow from ZnO/Ag₂O, generating a wide depletion region and a high baseline resistance. When exposed to the strongly oxidizing NO₂ molecules extract additional electrons from both phases, further widening the space-charge layer and triggering a sharp increase in resistance. This directly justifies why the silver-modified films display markedly higher baseline changes compared to pristine ZnO [48].