3D ZnO/ZIF‐8 Hierarchical Nanostructure for Sensitive and Selective NO2 Sensing at Room Temperature

Gas sensors based on semiconductor metal oxides (SMOs) have gained widespread attention for Internet of Things applications; however, high operating temperatures and low gas selectivity limit their applications. Recently, metal–organic frameworks (MOFs) have demonstrated potential in enhancing gas selectivity through the physical filtration of gas molecules based on their kinetic diameters. However, their application has been predominantly limited to simplistic nanostructured sensors. These sensors exhibit inherently inferior gas sensor performance compared to three‐dimensional nanostructure gas sensors. In this study, a highly periodic, 3D hierarchical ZnO/ZIF‐8 nanostructure is fabricated for photoactivated gas sensing at room temperature. Under UV illumination, the gas sensor exhibited a 17‐fold enhancement in gas response toward 0.1 ppm NO2 compared to pristine ZnO. In addition, the ZIF‐8 coating selectively increased the NO2 gas response compared to ethanol, acetone, and toluene gases, thereby improving the gas selectivity. The gas response improvement by the ZIF‐8 layer coating, which has not been achieved by previous studies, is based on enhanced photoactivation by the solid interaction between ZIF‐8 and ZnO. These results provide a systematic background for controlling the layer thickness of SMO‐MOF nanostructures and the catalytic role of MOF in photoactivated gas sensing.


Introduction
The development of digital infrastructure through the Internet of Things (IoT) technology has resulted in a significant demand for accurate and reliable gas sensors.3] Specifically, the device integration of these sensor chips in IoT systems has become increasingly important because the realtime monitoring of gas species is required.[6][7][8][9] With the advancement of nanotechnology, SMOs have been extensively studied as gas-sensing materials, demonstrating their potential as gas-sensor devices.However, implementing SMObased gas sensors in IoT systems present persistent challenges, such as the requirement for high operating temperatures exceeding 200 °C, which increase the power consumption of sensor devices and potential thermal degradation of electronic components. [10]Various strategies, such as the use of alternative energy sources or the design of heterogeneous materials for low-temperature gas sensing, have been adopted to address this issue. [11,12][15] This approach allows the detection of gas species at low temperatures by adjusting the wavelength of the light source to generate electron-hole pairs in metal oxides.To enhance photoactivation, various synthesis methods for sensing material, such as adding Gas sensors based on semiconductor metal oxides (SMOs) have gained widespread attention for Internet of Things applications; however, high operating temperatures and low gas selectivity limit their applications.Recently, metalorganic frameworks (MOFs) have demonstrated potential in enhancing gas selectivity through the physical filtration of gas molecules based on their kinetic diameters.However, their application has been predominantly limited to simplistic nanostructured sensors.These sensors exhibit inherently inferior gas sensor performance compared to three-dimensional nanostructure gas sensors.In this study, a highly periodic, 3D hierarchical ZnO/ZIF-8 nanostructure is fabricated for photoactivated gas sensing at room temperature.Under UV illumination, the gas sensor exhibited a 17-fold enhancement in gas response toward 0.1 ppm NO 2 compared to pristine ZnO.In addition, the ZIF-8 coating selectively increased the NO 2 gas response compared to ethanol, acetone, and toluene gases, thereby improving the gas selectivity.The gas response improvement by the ZIF-8 layer coating, which has not been achieved by previous studies, is based on enhanced photoactivation by the solid interaction between ZIF-8 and ZnO.These results provide a systematic background for controlling the layer thickness of SMO-MOF nanostructures and the catalytic role of MOF in photoactivated gas sensing.
[22] Cho et al. fabricated a highly ordered, periodic 3D TiO 2 porous nanostructure and detected NO 2 gas at room temperature using the proximity-field nanopatterning (PnP) technique. [20]This 3D nanostructure exhibited excellent NO 2 gas-sensing performance under ultraviolet (UV) illumination at room temperature because of its increased light absorption efficiency owing to lightscattering effects.Our previous studies have also shown that this periodic nanostructure is favorable for gas sensing because of its efficient gas diffusion compared with other nanostructures, including colloidal particles and inverse opal structures. [21,22][22][23][24] However, gas selectivity cannot be controlled by modifying the structural factors; it relies solely on the intrinsic properties of the sensing material.
A newly proposed molecular sieve method, involving the use of nanoporous membranes as physical gas filters, can enhance the gas selectivity of a sensor. [25]This method improves the selectivity by regulating the diffusion of gas molecules that are larger than the membrane pores by physically filtering them based on their molecular diameters.[28][29][30] Yao et al. demonstrated the gas selectivity enhancement effect of a MOF layer on a Au-ZnO nanorod gas sensor. [31]ZIF-8 and ZIF-DMBIM were coated on the surface of the ZnO nanorods, improving gas selectivity by suppressing the gas response toward large gases than their pore size while maintaining the gas response toward small gases.However, previous studies have been limited to simple metal oxide structures because of difficulties in controlling the shape of the MOF layer on the metal oxide surface.Furthermore, most MOF-based gas sensors are activated by high-temperature thermal energy, possibly because of low light absorption efficiency of bare nanostructured metal oxides, resulting in limited gas response under photoactivation.
In this study, we present a novel approach for fabricating a 3D ZnO/ZIF-8 hierarchical nanostructure (3D ZnO/ZIF-8 HNS) for room-temperature gas sensing under UV illumination.This synthesis approach results in nanostructures with enhanced both gas sensitivity and selectivity.The 3D ZnO nanostructure was fabricated using the PnP technique, and a ZIF-8 layer was formed on the 3D ZnO structure as a gas filter.The thickness of ZnO was controlled at 30, 50, and 70 nm, whereas that of ZIF-8 was varied by changing the conversion time from 10-60 min to optimize the HNS (porosity, electrical properties, and structural stability).The fabricated 3D ZnO/ZIF-8 HNSs were exposed to 0.1 ppm of NO 2 gas under UV illumination (365 nm) under dark conditions, which exhibited 124% gas response with 3D ZnO (50 nm)/ZIF-8_30 min condition.Furthermore, the gas selectivity of the 3D ZnO/ZIF-8 HNS was evaluated by measuring the gas response of four target gases with different kinetic diameters (i.e., 3.3, 4.3, 4.7, and 5.7 Å for NO 2 , ethanol, acetone, and toluene, respectively).The ZIF-8 layer improves the gas selectivity toward NO 2 gas compared to other target gases with kinetic diameters larger than the pore size of ZIF-8 (3.4 Å).The gas-sensing characteristics of the 3D ZnO/ZIF-8 HNS gas sensor were further investigated through electrical resistance and ultraviolet-visible (UV-Vis) spectroscopy to verify the gas-response enhancement effects on NO 2 gas.The results confirmed that the ZIF-8 layer enhanced the photoactivation of ZnO.

Morphological and Structural Characterization
Figure 1a shows a schematic of the concept of the 3D ZnO/ZIF-8 HNS-based gas sensor.The sensor interdigitated electrodes (IDEs) were patterned with an active area of 0.8 mm Â 0.8 mm and interspacing of 5 μm.The 3D ZnO/ZIF-8 HNS was fabricated using 3D polymer templates defined by the PnP technique (Figure S1, Supporting Information).Briefly, a 355 nm filtered laser exposed a photopolymer through a conformal phase mask, generating light diffraction and interference patterns through the Talbot effect. [32]The periodically distributed light intensity in the photopolymer film created a 3D periodic polymer nanostructure via cross-linking.To ensure efficient gas diffusion, we fabricated the 3D nanostructure with height of %6 μm, as confirmed in our previous study. [20]he 3D ZnO nanostructure was fabricated by atomic layer deposition (ALD) on a 3D polymer nanostructure and by removing the polymer templates through annealing.Then, a ZIF-8 layer was created on the surface of the ZnO nanostructure using a post-synthesis conversion method to cover the ZnO surface while restricting the random creation of ZIF-8 crystals at untargeted sites.Specifically, the 3D ZnO nanostructure was treated with a solution containing 2-methylimidazole (2-MIM) such that a small portion of ZnO exposed to the ligand solution was converted to ZIF-8 crystals.35] Figure 1b,c show the experimental concept of the 3D ZnO/ ZIF-8 HNS.Our previous research on 3D TiO 2 nanostructures reported the geometric advantages of 3D periodically ordered nanostructures for gas responses.A highly ordered 3D nanostructure can enhance light absorption efficiency through light-scattering effects, which improves the gas response under photoactivation. [20]The periodic micro-and nanopores in the 3D nanostructure ensured efficient gas diffusion into the nanostructure. [22][22] With an excellent gas response based on structural advantages, we fabricated a HNS by coating a ZIF-8 layer on the surface of ZnO to improve gas selectivity.A thin ZIF-8 layer containing pore windows of size 3.4 Å was used as molecular sieve membranes that can sharply separate gas molecules based on their kinetic diameters and enhance the gas selectivity of the sensor (Figure 1c). [33]Our approach, which uses periodically ordered 3D HNS, can result in a high gas response while simultaneously improving gas selectivity.Therefore, this approach offers a promising solution to the challenges encountered by conventional photoactive gas sensors.Figure 1d shows representative scanning electron microscopy (SEM) images of the 3D ZnO/ZIF-8 HNS (ZnO (50 nm)/ZIF-8_30 min).An ordered periodic nanostructure was observed despite the ZIF-8 layer coating.The pore size of the 3D ZnO/ ZIF-8 HNS was approximately 400 nm, as shown in Figure 1e.For a precise observation of the 3D ZnO/ZIF-8 HNS, a cross-sectional transmission electron microscopy (TEM) image was carefully obtained, which revealed uniform formation of a ZIF-8 layer on the surface of the ZnO film (Figure 1f ).
Surface conversion to the ZIF-8 layer was confirmed using energy-dispersive X-Ray spectroscopy (EDS) analysis and X-Ray photoelectron spectroscopy (XPS).The uniform distribution of nitrogen from 2-MIM of the ZIF-8 ligand confirms the uniform coating of the ZIF-8 layer along the ZnO structure (Figure 1g).In addition, the surface coverage of the ZIF-8 layer was investigated through XPS analysis of 3D ZnO, 3D ZnO/ZIF-8 HNS, and argon-etching-treated 3D ZnO/ZIF-8 HNS (Figure 1h,i).Figure 1h shows the O1s spectra of the 3D ZnO and 3D ZnO/ZIF-8 HNSs obtained with/without etching.The surface oxygen ion peak (532.0 eV) appeared for all conditions; however, the oxygen ion (O 2À ) inside ZnO (530.1 eV) was not detected after ZIF-8 coating. [36]When the etching process exposed the surface of the ZnO/ZIF-8 HNS, O 2À in the ZnO peak was detected, implying that the ZnO structure remained after ZIF-8 coating and ZIF-8 successfully covered the surface of ZnO.The N1s spectra of the HNSs confirmed that the nitrogen from 2-MIM of the ZIF-8 ligand was pyrrolic N (400.9eV) and pyridinic N (399.0eV) after ZIF-8 coating (Figure 1i). [37]o investigate the effect of the thicknesses of ZnO and ZIF-8 on the gas-sensing performance of the 3D ZnO/ZIF-8 HNS, the thickness of ZnO was finely tuned using the ALD technique with nanometer-scale precision to achieve three targeted thicknesses, that is, 30, 50, and 70 nm.Moreover, the thickness of ZIF-8 was regulated by changing the conversion time (10, 20, 30, and 60 min).The thickness of ZnO affects the porosity, changes in electrical resistance, and structural stability.In contrast, the thickness of ZIF-8 modulates the gas diffusion process with a gas-filtering effect.
The morphologies of the 3D ZnO/ZIF-8 HNSs with different ZnO and ZIF-8 thicknesses were confirmed through SEM (Figure 2a).The 3D periodically ordered nanostructures were maintained under all synthesis conditions, even after ZIF-8 conversion.When the thickness of ZnO was 30 nm, a longer conversion time damaged the structure because of the sacrificial conversion reaction of Zn from the ZnO scaffold.In contrast, for a ZnO thickness of 70 nm, a longer reaction time reduced the void space of the nanostructure owing to the increased volume of the ZIF-8 film.This resulted in a significant increase in the mass transfer resistance, which inhibited gas diffusion into the deeper side of the nanostructure.The thickness of ZIF-8 was measured using high-resolution (HR-TEM) (Figure S3, Supporting Information).As the ZIF-8 conversion time increased, the thickness of ZIF-8 increased within the range of 10-50 nm.In the initial stage of ZIF-8 conversion, the ZIF-8 nanoparticles grew on the ZnO surface, leaving parts of the ZnO surface uncovered by the ZIF-8 layer.As the conversion time increased, the ZIF-8 layer covered the ZnO surface with a linearly increasing layer thickness.
Figure 2b shows the collected thickness for ZIF-8 different ZnO thicknesses and conversion times.ZnO with a thickness of 30 nm resulted in a thin ZIF-8 layer compared to other ZnO thicknesses, due to insufficient Zn source with structural damage.Notably, the thickness of ZIF-8 for 50 nm-thick ZnO was slightly larger than that for 70 nm-thick ZnO owing to ligand molecule diffusion into the nanostructure impeded by the lower porosity of 3D HNS.The growth rate of ZIF-8 slowed after 1 h of conversion because of the thick ZIF-8 film, which hindered the access of ligand molecules to the internal structure and reduced the supply of Zn ions to the ZnO surface.

Gas-Sensing Performance of the 3D ZnO/ZIF-8 HNS
The gas-sensing performance of the 3D ZnO/ZIF-8 HNSs with different layer thicknesses were measured after exposure to 0.1 ppm NO 2 gas at room temperature.A UV LED (wavelength:365 nm) was placed overhead of the sensing area to facilitate a photoactivated gas-sensing mechanism.Figure 3a shows the NO 2 gas sensing results of the 3D ZnO and 3D ZnO/ZIF-8 HNSs with different ZnO thicknesses and ZIF-8 conversion time of 30 min under UV illumination and dark conditions.Notably, ZnO and ZnO/ZIF-8 exhibited enhanced gas responses under UV illumination, demonstrating the possible photoactivation of the 3D ZnO nanostructures for gas detection.Under ZnO (30 nm) and ZnO (30 nm)/ZIF-8_30 min conditions, UV exposure enhanced NO 2 gas response from 1% to 10.5% and from 2% to 47.3%, respectively.Comparing the ZnO gas sensors under UV light, as the internal porosity and surface electron depletion region increased, thinner ZnO layers showed higher gas responses; the gas responses of the sensors with ZnO layer thicknesses of 30, 50, and 70 nm were 10.5%, 7.33%, and 2.5%, respectively.This observation is consistent with our previous results on 3D TiO 2 nanostructures. [20]In contrast, after ZIF-8 layer coating, the gas response showed an optimal point at the ZnO layer thickness of 50 nm, and the NO 2 gas responses of sensors with ZnO layer thicknesses of 30, 50, and 70 nm were 47.3%, 124%, and 56.9%, respectively.As shown in Figure 2c, the 30 nm thick ZnO structure collapsed during the ZIF-8 conversion process, degrading the gas response.For a ZnO layer with 70 nm thickness, low porosity associated with the ZIF-8 layer interferes with gas diffusion into the nanostructure, resulting in decreased gas responses.The 3D ZnO (50 nm)/ZIF-8 composite showed a 17-fold higher gas response than ZnO under UV illumination.This result was measured presumably because of the concentration of gas molecules in the pores of ZIF-8, which provided a large surface area for gas molecules to adsorb. [39]To investigate the ZIF-8 thickness effect on gas responses, NO 2 gas responses under UV illumination for all synthesis conditions were measured (Figure 3b).As the ZIF-8 thickness increased, the NO 2 gas responses increased for a conversion time of 30 min.Over this conversion time, the gas responses decreased owing to the thick ZIF-8 layer, inhibiting gas diffusion into the ZnO surface through the ZIF layer.Interestingly, the ZIF-8 coating reduced the gas response because ZIF-8 possesses active sites for metal oxides.Thus, the existence of an optimal ZIF-8 thickness implies that ZIF-8 plays a catalytic role in the gas responses with enhanced effects.
Figure 3c shows the gas selectivity measurement results for 3D ZnO and ZnO (50 nm)/ZIF-8 with different ZIF-8 conversion times to determine the effect of ZIF-8 thickness on gas selectivity.The target gases were selected based on their kinetic diameters among generally considered air pollutants to confirm the molecular sieving effects of the ZIF-8 layer.Four target gases, including NO 2 , ethanol, acetone, and toluene, have different kinetic diameters with 3.3, 4.3, 4.7, and 5.7 Å, [40,41] which are comparable to the pore size of ZIF-8 (3.4 Å). [42] To compare the gas response toward NO 2 gas with those toward other gases, the concentrations of the other target gases were adjusted to 1 ppm.The gas selectivity enhancement effects of the ZIF-8 layer coating on metal-oxide-based gas sensors have been reported previously. [25,26,31,33,40]However, no studies have investigated the effects of structural factors in 3D nanostructure on the sensing performance by controlling the thicknesses of ZnO and ZIF-8.In addition, most previous studies utilized ZIF-8-coated metal oxide gas sensors under thermal activation mechanisms.This study is the first to demonstrate the effect of the ZIF-8 layer on metal oxide gas sensors under a photoactivation mechanism.We mainly focused on the effects of the thickness of the 3D ZnO/ZIF-8 HNS layer on the gas response and selectivity under photoactive conditions.The thickness of the ZIF-8 layers was determined at different conversion times.For a meaningful comparison, a ZnO thickness of 50 nm, which exhibited the highest gas response, was selected to obtain a reasonable gas response value.With the optimized conversion conditions, the gas response to NO 2 of 3D ZnO (50 nm)/ZIF-8_30min was enhanced compared to that of the 3D ZnO (50 nm)-based gas sensor.The optimized ZIF-8 layer thickness can enhance the gas selectivity with different degrees of gas response enhancement.In addition, the optimal ZIF-8 thickness differed for NO 2 and other target gases.The other target gases showed the highest gas response for the thinner ZIF-8 film (20 min conversion time) because their large kinetic diameter prevented diffusion into the thicker ZIF-8 layer.The gas selectivity enhancement is more clearly displayed in Figure 3d, which shows the gas responses of the 3D ZnO and 3D ZnO/ZIF-8_30 min nanostructures with different ZnO thicknesses.After ZIF-8 conversion, the NO 2 gas response sharply increased; however, the other target gases did not show such gas response increments.As the thickness of ZnO increased, the gas selectivity of NO 2 gas improved because the larger portion of the ZIF-8 layer in the internal pores of the nanostructure.As shown in Figure 2a, the ZIF-8 layer on the thicker ZnO layers can close the internal pores.In other words, the incident gas molecules must cross the ZIF-8 layer to diffuse into the bottom of the nanostructure.As a result, the gas selectivity of thicker ZnO exhibited better gas selectivity than thinner ZnO.Even for larger gas species, such as ethanol and toluene, the ZnO thickness of 70 nm had no gas response.Figure 3g shows the normalized NO 2 gas responses to the other target gases.The gas selectivity enhancement effects are shown in the normalized graph after the ZIF-8 conversion.The ZIF-8 layer decreased the responses of the target gases, except for NO 2 gas.In addition, even at a ZnO thickness of 30 nm, the gas sensor exhibited gas selectivity enhancement due to the ZIF-8 layer.Consequently, the gas selectivity of the 3D ZnO/ZIF-8 HNS was enhanced by the ZIF-8 layer.From the gas selectivity results in Figure 3, we infer that the ZIF-8 layer improves the gas selectivity by further improving the NO 2 gas response compared with other larger target gases.Most previous studies used the ZIF-8 layer as a gas filter membrane to maintain or reduce gas responses to even small gas molecules because ZIF-8 covers the active site of the metal oxide and requires an additional diffusion path.However, we observed a gas response enhancement in the 3D ZnO/ZIF-8 HNS after the ZIF-8 layer coating, which is an abnormal behavior in the ZIF-8 layer coating for gas sensors.To elucidate the origin of the gas response enhancement, we measured the gas-sensing properties of the 3D ZnO (50 nm)/ZIF-8_30 min gas sensor for NO 2 .
Figure 4a shows multiple pulses of NO 2 at various concentrations (< 0.1 ppm) under UV illumination.Even at a low concentration of 20 ppb, the 3D ZnO (50 nm)/ZIF-8_30 min sample can detect NO 2 with a gas response of 22.3%.With multiple gas responses in the low concentration range, the theoretical detection limits of 3D ZnO and 3D ZnO (50 nm)/ZIF-8_30 min were calculated by extrapolating the linear relationship until the signalto-noise ratio was equal to 3 (Figure 4b).The gas responses of 3D ZnO and 3D ZnO (50 nm)/ZIF-8_30 min exhibited a linear relationship with gas concentration, with the slope of 0.487 and 6.69 ppm À1 , respectively.Subsequently, the calculated theoretical detection limits were 33.1 ppb and 375 ppt, respectively, which indicates that the ZIF-8 layer coating can improve the detection limit to below ppb level.The outstanding gas-sensing performance of the 3D ZnO (50 nm)/ZIF-8_30 min gas sensor for a low concentration range was compared with another previous result using a ZnO-based room-temperature gas sensor.3D ZnO (50 nm)/ZIF-8_30 min was more sensitive to NO 2 concentrations in a sub-ppm range with a higher slope, demonstrating that the ZIF-8 layer can enhance gas selectivity and sensitivity with comparable effects to other catalysts.

Role of the ZIF-8 Layer on Gas-Sensing Properties
The specific catalytic effect of the ZIF-8 layer on gas sensing under UV illumination was investigated in terms of photoactivation.Figure 5a shows the electrical resistance of the gas sensor under UV illumination and dark conditions as a function of the ZIF-8 conversion time.As the ZIF-8 layer was created on the surface of ZnO, the electrical resistance without UV illumination increased owing to the reduction in ZnO thickness by ZIF-8 conversion.However, under UV illumination, the electrical resistance of the gas sensor decreased after the ZIF-8 layer coating, implying that the ZIF-8 layer enhances the photoactivation of ZnO with a large number of photogenerated electron-hole pairs.In addition, the light absorption characteristics of 3D ZnO (50 nm)/ZIF-8 were investigated using UV-Vis spectroscopy (Figure 5b).The absorbance of the 3D nanostructure was increased by the ZIF-8 layer coating at 365 nm, which was used for activation by the UV LED.Light absorption enhancement by ZIF-8 can be observed in a 2D film ZnO/ZIF-8 gas sensor with the same conversion conditions, which exhibits gas response enhancement by ZIF-8 under UV illumination, even with a 2D film morphology (Figure S4, Supporting Information).We previously reported the light absorption enhancement of this type of nanostructure.However, the light absorption was further enhanced by the ZIF-8 layer coating, contributing to excellent gas response enhancement.Figure 5c shows a schematic of the proposed enhancement mechanism with electrical-band alignments.These types of photoactivation enhancement in ZnO/ ZIF-8 composites were also observed in photocatalyst applications with enhanced photoactivation. [34,35,43,44]The photoactivity enhancement effect in ZnO/ZIF-8 composite was explained by the interfacial charge transfer between organic ligand to metal (LMCT). [45]The photo-activated electrons were transferred from ligand (2-methylimidazole) to Zn ion, which enhance the photoactivation.In addition, studies using TiO 2 /ZIF-8 composites for photocatalyst reported that ZIF-8 decoration can suppress the interfacial charge recombination, which resulted in photoactivity enhancement. [46,47]This study represents the first demonstration of the effect of ZIF-8 in light-activated gas sensor with the significant gas response enhancement.The ZIF-8 layer assisted the photoactivation of the ZnO with an enhanced gas response, validating the capability of ZIF-8 as a photoactive additive.

Conclusions
We fabricated a 3D ZnO/ZIF-8 HNS as a highly sensitive and selective light-activated gas sensor operating at room  temperature.A 3D nanostructure fabricated using the PnP method is an ordered and periodic nanostructure that offers effective gas diffusion and enhanced light absorption.By converting the ZnO surface into ZIF-8, 3D ZnO/ZIF-8 HNS were successfully synthesized while preserving their ordered and periodic nanostructures.The thickness of the ZnO layer was varied (30, 50, and 70 nm) and the conversion time of ZIF-8 was adjusted (10, 20, 30, and 60 min) to optimize the gas filter effect without compromising the gas responses.The 3D ZnO/ZIF-8 HNS, particularly the ZnO (50 nm)/ZIF-8_30 min gas sensor, exhibited a gas response of 124% toward 0.1 ppm NO 2 under UV illumination.Notably, the gas response of the HNS was 17 times higher than 3D ZnO gas sensor.In addition, the 3D ZnO (50 nm) /ZIF-8(30 min) HNS exhibited an excellent theoretical detection limit of 375 ppt.Interestingly, the conversion of ZIF-8 enhanced the gas responses, but a thick ZIF-8 layer after 30 min of conversion time reduced the gas responses owing to gas diffusion interference.Among the target gases with various kinetic diameters (NO 2 , ethanol, acetone, and toluene), the 3D ZnO/ZIF-8 HNS showed a selective gas response to NO 2 , which has a smaller kinetic diameter than the ZIF-8 pores.Additionally, the electrical resistance changes and UV-Vis spectral analysis demonstrated that the presence of the ZIF-8 layer results in an increased resistance difference between the dark and UV light conditions as well as enhanced light absorption in the UV region, contributing to the unique gas response enhancements.This study is the first to propose a 3D ordered and periodic hierarchical metal oxide/ MOF nanostructure, which can provide a research background for manufacturing nanostructured hierarchical composites.Unfortunately, the growth of ZIF-8 within the 3D structure led to an unexpected increase in the response/recovery time (>1000 s), along with narrowing of the internal pores.These critical drawbacks can be addressed through subsequent investigations by adjusting the pore size and height of the 3D nanostructures.This study provides valuable insights into the engineering of HNS and the role of MOF layers in gas sensing and offers a promising approach for future gas sensor development.

Experimental Section
Fabrication of a 3D ZnO HNS on an IDE Substrate: Pt/Ti (thicknesses of 150 nm/30 nm) interdigitated electrode patterns (IDEs) were deposited on a SiO 2 /Si substrate (300 nm/150 μm) with an active sensing area of 0.8 mm Â 0.8 mm.The IDE-patterned substrates were treated with air plasma (45 sccm, 40 motors, 40 W) for 2 min (CUTEMP, Femtoscience).A thin layer of photoresist (PR) (SU-8 2, Microchem) was coated using the spin-casting method (3000 rpm, 30s, %2 um) as an adhesion layer.The PR-coated wafer was then soft-baked at 65 °C for 2 min and 95 °C for 3 min to remove the solvent.After soft baking, the PR-coated wafer was exposed to UV light for 1 min with a photomask that blocked the exposure of the active sensing area and was hard-baked at 180 °C for 10 min.A development process was then conducted to remove the uncrosslinked polymer from the active sensing area.SU-8 10 spincoated (2400 rpm, 30s, %6 um) on the adhesion layer and soft-baked at 65 °C for 1 h and 95 °C for 1 h.The 3D polymer template was fabricated using the PnP technique with a conformal phase mask.The soft-baked film was exposed to a pulsed Nd:YAG laser (355 nm) beam through a conformal phase mask with a surface relief grating consisting of square arrays of hole patterns with a periodicity of 600 nm, diameter of 480 nm, and depth of 400 nm.The substrate was post-baked at 60 °C for 6 min, then the uncrosslinked moiety was dissolved using an SU-8 developer for 30 min.A thin ZnO layer (30, 50, and 70 nm) was conformally deposited on the 3D polymer nanostructure using ALD at 90 °C (Mini-ALD, NexusBe Co., Ltd).Diethylzinc (IChem.Co. Ltd.) was used as the precursor and H 2 O was used as the reactant.The thin-shell thickness of ZnO was controlled at 30, 50, and 70 nm, with a deposition rate of 0.72 Å per cycle.Then, the 3D polymeric template with ZnO thin film was annealed at 500 °C to crystallize ZnO and remove the polymer template.
Conversion of ZIF-8 on a ZnO Nanostructure: The ZIF-8 layer was coated using a ligand-exchange method with ZnO as the Zn ion source.In a mixed solution of DI:DMF (2.5:13.5),0.08 g of 2-MIM (Sigma-Aldrich, USA) was dissolved in a mixed solution of DI:DMF (2.5:13.5) in a total volume of 32 mL.The ligand solution and 3D ZnO nanostructures were transferred to a Teflon-lined autoclave and heated in an oven at 60 °C for 10, 20, 30, and 60 min.The final 3D ZnO/ZIF-8 HNSs were cleaned with ethanol and deionized water to remove residual ligands and dried at room temperature for 24 h.
Material Characterization: The morphology of the 3D ZnO/ZIF-8 HNS was investigated by field-emission scanning electron microscopy (FE-SEM, SU 5000, Hitachi).The microstructures were investigated using TEM (Tecnai F20, FEI Company).The crystallinity of the sensors was measured using XRD(Ultima IV, RIGAKU) with a Cu-Kα radiation source (wavelength 1.5418 Å).The chemical bonding and binding energies of the sensor materials were investigated by XPS using a K-alpha system (Thermo VG Scientific) with an Al-Kα X-Ray source.The absorbance was measured using a UV-Vis spectrophotometer (SolidSpec-3700, Shimadzu).
Gas Response Measurements: The responses to the target gases were measured using a quartz tube.The gas flow was controlled using a mass-flow controller to maintain a constant flow rate of 1000 sccm.The sensor resistance was measured using a Keithley 2401 instrument with a DC bias voltage of 1 V.To provide UV light, the sensor with an operating voltage of 3.2 V was equipped with a UV LED (365 nm).The gas responses were measured by the MUX system (Keithley 7001) connected to the electrical line for measuring the gas responses.The MUX system could switch the sensor response in 100 ms for each channel of the sensor; thus, the response of each sensor was constructed at an interval of approximately 1 s.

Figure 3 .
Figure 3. Gas-sensing results of the 3D hierarchical ZnO/ZIF-8 structure.a) Gas response results of 3D ZnO and 3D ZnO/ZIF-8 HNS under dark and UV light (365 nm) condition at room temperature with different ZnO film thickness.ZIF-8 conversion time is conditions for 30 min.b) Gas response results of the 3D ZnO/ZIF-8 HNS with different ZnO film thicknesses and ZIF-8 conversion times.c) Gas selectivity analysis of the 3D ZnO/ZIF-8 HNS with different ZIF-8 conversion times (ZnO film thickness: 50 nm).d-f ) Gas selectivity analysis of 3D ZnO/ZIF-8 HNS for four target gases (NO 2 , ethanol, acetone, and toluene) with different ZnO film thickness (ZIF-8 conversion time: 30 min).g) Normalized gas responses of 3D ZnO/ZIF-8 HNS for target gases.All the gas response measurements were under UV light (365 nm) at room temperature.

Figure 4 .
Figure 4. a) Sensing performance of 3D ZnO (50 nm)/ZIF-8_30 min toward different concentrations of NO 2 .b) Linear fit of the responses as a function of NO 2 concentration.c) Comparison of gas-sensing properties with those in other previous ZnO-based RT gas sensors.