In Situ Electrical Detection of Methane Oxidation on Atomically Thin IrO2 Nanosheet Films Down to Room Temperature

Activation of the CH bond is the first step in converting methane into valuable chemicals. Herein, the successful induction and electrical detection of a methane oxidation reaction are reported at room temperature using IrO2 nanosheets, a 2D form of IrO2. A clear decrease in electrical resistance upon exposure to methane is observed by using atomically thin IrO2 nanosheet films. The resistance decrease disappears upon simultaneous exposure to oxygen, suggesting that methane is oxidized by consuming the lattice oxygen of the IrO2 nanosheets and that the oxygen vacancies generated are recovered by oxygen in the atmosphere. The resistance decrease is observed even at 300 K, indicating the high methane oxidation ability of the IrO2 nanosheets. These results are confirmed by a shift of the Ir 4f peaks in ambient pressure X‐ray photoelectron spectra. Furthermore, deposition of amorphous carbon, that is, methane oxidation products, on IrO2 nanosheets is also confirmed by Raman scattering spectroscopy after prolonged methane exposure at high temperatures in the absence of oxygen. This study demonstrates the ability of IrO2 nanosheets to oxidize methane at least down to 300 K and is an important example of the usefulness and simplicity of chemical reaction monitoring using electrical resistance changes.


Introduction
Since the depletion of petroleum resources has begun to be recognized, methane, the main component of natural gas, has been attracting attention as a useful raw material of chemicals for substituting petroleum partly due to innovations in shale gas extraction technology. [1][2][3] In addition, from a perspective of fossil fuels recycling, the conversion reactions of methane are significant for C1 chemistry, which uses a group of molecules possessing one carbon atom. [4][5][6] To realize the conversion reactions of methane into useful chemicals, its C-H bond must first be broken by oxidation/dehydrogenation. Considerable research on substances exhibiting high C-H activation ability that could break the C-H bonds of methane has been undertaken. [7][8][9][10][11][12][13] Recently, the (110) surface of iridium dioxide (IrO 2 ) was reported to exhibit very high methane oxidation activity. [14] Liang and Weaver et al. obtained the IrO 2 single-crystal surface by oxidizing bulk Ir single crystals and conducted methane exposure at low temperatures, followed by temperature-programmed desorption (TPD) analyses to examine its reactivity. They confirmed that intact molecular desorption of methane occurred below 150 K; however, desorption of oxidation products (CO, CO 2 , and H 2 O) and recombinative desorption were detected above 400 K. In other words, a C-H bond scission reaction could be considered to occur at least in the low temperature range down to 150 K, suggesting that IrO 2 possesses the high activation ability of methane C-H bonds even at low temperatures.
To confirm whether such a reaction is actually occurring, in situ monitoring of the chemical reaction would be informative. For this purpose, many spectroscopic methods have been used, including the analysis of chemical bonds using X-ray photoelectron spectroscopy (XPS), [15][16][17][18][19][20] that of vibrational energies of adsorbates/surfaces using Raman scattering spectroscopy, [21][22][23] and that of vibrational energies of adsorbates using infrared spectroscopy. [24][25][26] Technical difficulties still exist for observation under actual reaction conditions, especially for methods based on operation under vacuum, such as XPS.
By contrast, in this paper, we report the successful observation of methane oxidation by IrO 2 using a simple method of in-situ measurement of electrical resistance. This method is the simplest among various detection techniques because it only requires applying an electrical voltage and measuring an electrical current (or vice versa). Moreover, this method does not require a vacuum and can easily acquire information on catalytic cycles under realistic conditions. Considering the increasing attention to renewable energies, their fluctuation should be included, unlike in most of current reaction systems operated under steadystate conditions. [27,28] Therefore, the time resolution of operando techniques becomes an important factor for monitoring the reaction systems under dynamically changing conditions. The development of X-ray monochromators that enable a rapid energy scan as well as the technical progress of related devices like detectors have contributed to shortening the time resolution of X-ray based techniques. [29][30][31] The timescale of the related techniques can be reduced to sub-millisecond regime thanks to the high brilliance of the X-ray sources in synchrotron facilities. [32,33] While the techniques based on X-ray absorption enable gaining detailed information on chemical states, they require such large facility to achieve the time resolution in the (sub-) millisecond scale. By contrast, the present technique that measures the time evolution of an electrical resistance can easily reach a microsecond-scale resolution in the laboratory. Although each resistance change should be related to a specific change in surface chemistry using chemical analyses such as XPS, the electrical technique can be applied after the initial calibration for the fast and facile monitoring of the dynamics/kinetics of target reactions.
However, in the target reaction of this study, changes in electrical resistance due to chemical reactions on the surface cannot be detected because they are masked by the high bulk electrical conductivity of metallic IrO 2 . Here, instead of using single crystal IrO 2 , ultrathin films with an extremely high surface-to-volume ratio were fabricated using atomically thin IrO 2 nanosheets. [34][35][36] The bulk IrO 2 , which exhibits methane oxidation ability, pos-sesses a rutile structure. [14] By contrast, IrO 2 nanosheets were obtained from liquid-phase exfoliation of parent layered crystals. Therefore, the structure of IrO 2 nanosheets reflects that of the parent crystal (K 0.2 IrO 2 nH 2 O). [36] This layered compound consists of 2D IrO 2 layers with intercalated potassium atoms that are extracted during the exfoliation process via cation exchange. The purely 2D structure in nature distinguishes it from 3D counterparts such as rutile. Therefore, the IrO 2 nanosheet used in the present study is like an allotrope of IrO 2 , as 2D graphene obtained from layered compound graphite possesses a different structure from that of 3D diamond. IrO 2 nanosheets are an ultimately thin IrO 2 derivative with a thickness of only three atoms and are strongly affected by surface reactions. Consequently, a clear change in electrical resistance was observed upon exposure to methane. By reducing the thickness of IrO 2 nanosheet thin films to their monolayer limit, the resistance change was confirmed even at 300 K. The chemical states of the substrate were also studied using spectroscopic methods, which clearly show that the methane reaction occurs at room temperature. These findings indicate that, similar to rutile IrO 2 , IrO 2 nanosheets also have a very high methane C-H activation ability. Notably, this study is the first to report on methane activation using IrO 2 nanosheets. Figure 1a shows the measurement schematic of the in-situ electrical resistance measurements. The electrical response of IrO 2 nanosheet films deposited using layer-by-layer deposition was measured by introducing a controlled gas flow in a closed chamber. Figure 1b shows a measurement result of an electrical resistance change upon methane exposure, where the measurement was performed at 473 K to facilitate the methane oxidation reaction. The sample used for the measurement is a monolayer film, which is most affected by surface phenomena because of its ultrahigh surface-to-volume ratio. An atomic force microscopy (AFM) image of the film is shown in Figure 1c. The lateral size of each nanosheet is <1 μm, and a partial overlap between nanosheets ensures an electrical current path bridging the inter-electrode spacing of 50 μm. The thickness is ≈1 nm outside the overlap, indicating that the film is a monolayer. Here, a background flow of nitrogen at 400 mL min −1 is introduced before methane is introduced at 5 mL min −1 from 1500 to 3000 s. A decrease in electrical resistance with methane exposure can be clearly seen.

Results and Discussion
In contrast to the observed decrease in electrical resistance, an increase in resistance has long been known to be a change associated with adsorption of molecules on solid surfaces. [37][38][39] When the film thickness of a conductor becomes approximately the same order of magnitude as the mean free path of electrons, electrical conductivity begins to decrease from the bulk value because of surface scattering of electrons. When atoms or molecules are adsorbed on the surface, specular reflection of conducting electrons at the surface is changed to diffuse scattering, leading to a further decrease in electrical conductivity. Based on the surface scattering theory formulated primarily by Fuchs [37] and Sondheimer, [38] the change in electrical resistance can be explained by the change in the ratio (0 ≤ ≤1) at which specular reflection occurs. [39] This concept has been widely used to explain the result of the electrical resistance change by molecular adsorption on metal thin films. [40][41][42][43] Notably, the adsorption of molecules induces a resistance increase regardless of whether the molecules are exposed to methane itself or oxidation reaction products. Consequently, if an increase in resistance is observed, the change should not be considered a signal of electrical detection of the methane oxidation reaction.
Conversely, Figure 1b shows a decrease in electrical resistance that is, an increase in electrical conductivity which is opposite to the increase in electrical resistance with molecular adsorption predicted by the surface scattering theory. Therefore, a phenomenon other than an increase in surface scattering by surface adsorbates is the cause of the resistance decrease associated with methane exposure. Because Figure 1b shows the results without the introduction of oxygen, the methane reaction with lattice oxygen of the IrO 2 nanosheets is one of possible reasons for the observed resistance decrease. The room-temperature electrical resistivity of IrO 2 single crystals is higher than that of Ir single crystals that is, 5.2 μΩ cm for Ir; [44] 35 and 49 μΩ cm for [011] and [001] orientations of IrO 2 , respectively. [45] The consumption of lattice oxygen can be considered to be the reduction process of IrO 2 to Ir, which leads to a decrease in electrical resistance (Figure 1d).
On the (110) surface of rutile IrO 2 , rows of coordinately unsaturated (cus) Ir atoms are separated by rows of bridging O atoms, [14] where methane is initially activated by chemisorption onto the cus-Ir atoms. Meanwhile, the IrO 2 nanosheet are thought to possess the same structure as other transition metal oxide nanosheets that consist of three atomic layers where the center metal layer is sandwiched by the top and bottom O layers. [34] Therefore, the perfect crystal of the nanosheet possesses no cus-Ir atoms on its surface. Initial activation centers that provide chemisorption sites for methane may be the edges and in-plane O vacancies of the nanosheet, where cus-Ir atoms are present. Chemisorbed methane molecules at these activation centers are oxidized by subtracting O atoms adjacent to the cus-Ir atoms, leading to the generation of additional O vacancies (i.e., cus-Ir atoms). As such, the methane oxidation reaction is expected to proceed from the initial activation centers to the surrounding sites.
As shown in Figure 1b, the resistance did not return to its original level after methane exposure. The resistance decrease in the monolayered IrO 2 nanosheet films is attributed to the reduction of IrO 2 nanosheets. The recovery of resistance to its original value requires the re-oxidation of the reduced nanosheets. The trace amount of the residual oxygen in the chamber may lead to sluggish oxidation of the reduced nanosheets. In addition to the reduction/re-oxidation process, temperature changes by the generation of reaction heat contributed to the resistance changes. According to the temperature dependence of the samples (see Section S3, Supporting Information), the resistance of the monolayer films should decrease by the reaction heat generated during methane oxidation. After the methane introduction is stopped, the resistance should increase. Although the observed resistance changes appear to follow these trends, the resistance of the monolayer film did not recover to its initial value after methane exposure was stopped for a long time, which is not attributable only to the temperature changes. The observed resistance change is related to the redox kinetics of the nanosheets.
Notably, the resistance already started to increase during methane exposure in Figure 1b. Observed changes in resistance may be due to competing reactions of the nanosheets with introduced methane and residual oxygen, and/or the adsorption of reaction products, which leads to resistance increase by surface electron scattering.
To support the notion that the observed resistance decrease was due to the reduction of IrO 2 by methane, the change in electrical resistance altered via the introduction of oxygen in addition to methane was investigated. A native thin-oxide layer is spontaneously formed at the Ir surfaces; [46] thus, the (partially) reduced surface can be expected to be re-oxidized by exposure to oxygen gas. If this condition holds for IrO 2 nanosheets, the simultaneous introduction of oxygen should make the resistance decrease difficult to observe because the re-oxidation of the surface that is, the recovery of lattice oxygen occurs immediately during the reaction with methane. As shown in Figure 2a, the change in electrical resistance due to methane exposure is measured under the introduction of oxygen diluted with nitrogen at various concentrations. The response to methane exposure is reduced to ≈1/7 at an oxygen flow rate of 2 mL min −1 compared to that without the introduction of oxygen. When the oxygen flow rate is further increased to 10 mL min −1 , the response to methane exposure is virtually unobservable. Figure 2b summarizes the resistance decrease attributed to the reaction with methane as a function of oxygen flow rate. Notably, in Figure 1b, a gradual recovery of electrical resistance can be observed after methane exposure is stopped. As discussed above, this recovery may be due to unintentional re-oxidation by residual oxygen in the measurement chamber or partly due to the temperature lowering after stopping the generation of reaction heat. Without the intentional introduction of oxygen gas, the rate of re-oxidation is slow, and the resistance does not recover to its original value within the observation period. When oxygen is intentionally added, as shown in Figure 2a, the recovery speed of the electrical resistance after stopping methane exposure is much faster. These experimental facts support the notion that the decrease in electrical resistance of atomically thin IrO 2 nanosheet films represents the electrical detection of the methane oxidation reaction. In addition, the facts strongly suggest that catalytic methane oxidation by co-exposure to oxygen is achieved on IrO 2 nanosheets, as has been confirmed on single-crystalline IrO 2 (110) surfaces by XPS measurements. [47] The over-recovery of the resistance after stopping methane exposure was observed clearly with co-exposure to oxygen at 2 mL min −1 (Figure 2a). A similar behavior was also discernible in the data obtained at 10 mL min −1 . Several mechanisms that account for resistance increases in the present system; first, the re-oxidation of reduced IrO 2 nanosheets leads to resistance increase. Although the over-recovery case was not likely because the sample was heated overnight under the mixed flow of nitrogen and oxygen (see the Supporting Information for the detailed experimental procedure). Second, endothermic processes such as desorption of adsorbates lowers the temperature of the sample, thereby resulting in resistance increase. However, even if such endothermic processes occur after methane exposure is stopped, the temperature lowering that could occur should be counteracted by the effect of reaction heat generated during methane exposure and the subsequent re-oxidation process. Third, the presence of additional adsorbates increases resistance through enhanced scattering of conduction electrons. Methane exposure generates additional adsorbates (methane itself and/or reaction products), which can remain on the surface even after the methane exposure is stopped. The third mechanism might explain the gradual resistance decrease after the over-recovery, being interpreted with the gradual desorption of the additional adsorbates by reaction with oxygen.
While the data presented to date were measured at 473 K, Figure 3a shows the results of measurements at various temper-  Figure 1(b). The arrows indicate the direction of the temperature sweep. The inset shows the AFM image of each sample. b) Schematic diagram showing the difference in ratio of reduced areas attributed to the difference in nanosheet coverage. c) Confirmation of a resistance increase in the thick film of IrO 2 nanosheets (thickness: ≈9 nm). The experimental conditions are the same as that in Figure 1(b). The inset shows the AFM image of the film. d) Schematic diagram of an increase in surface electron scattering due to methane exposure.
atures. The vertical axis indicates the rate of resistance change, with the negative sign corresponding to a decrease in resistance. Here, two samples, which were monolayer films but had different nanosheet coverage, were used, and a double sweep of temperature (the temperature being raised and then lowered) was performed. For the sample with relatively high coverage (Sample #1; see the AFM images shown in Figure 3a), the change rate drops to be within the noise floor as the temperature is lowered below 373 K. Conversely, for the sample with relatively low coverage (Sample #2), the decrease in electrical resistance can be observed even at 300 K in the backward temperature sweep after removing surface contaminants in the forward sweep.
The difference in resistance change with nanosheet coverage can be attributed to the effect of overlap regions between each nanosheet. As shown schematically in Figure 3b, nanosheets located below the overlap regions are not directly exposed to methane; thus, the reduction of nanosheets due to methane oxidation does not occur. Consequently, the resistance decrease in the sample with a high coverage may be more difficult to observe because the relative fraction of the reduced section is smaller. Moreover, the presence of overlap regions is synonymous with the presence of areas of large film thickness. The data presented so far were obtained using an ultrathin film consisting of a monolayer film but containing several overlapping portions. By con-trast, Figure 3c shows the results of a study using a sample with a large film thickness (the detailed data of which are shown in the Supporting Information). In these data, an increase in electrical resistance was observed. Since the fraction of the top surface where nanosheet reduction occurs decreases in thick films, the increase in electrical resistance due to the increase in surface scattering of conduction electrons can be readily observed (Figure 3d). In the overlapping regions of the monolayer-based films, the increase in electrical resistance should be seen at the same time as that in the sample with a larger film thickness. Consequently, the resistance decrease effect due to nanosheet reduction is offset by the increase in surface scattering, with the methane response being less visible. These discussions indicate the desirability of using monolayer films with the smallest possible inter-sheet overlap for the in situ observation of methane oxidation reactions using electrical resistance monitoring.
Next, as a more direct evidence of the reaction taking place, change in oxidation states of Ir associated with the reaction was confirmed using ambient-pressure XPS (AP-XPS). Electrical experiments confirmed a decrease in electrical resistance due to exposure to methane in the absence of oxygen (reduction of IrO 2 ) and a much less significant resistance decrease in the presence of oxygen (redox of IrO 2 ). To confirm this, spectral changes of corelevels of IrO 2 nanosheet monolayer films by methane or oxygen exposure were studied using AP-XPS. Figure 4 shows the Ir 4f spectra measured in the presence of oxygen or methane (information on the other peaks is summarized in the Supporting Information; see Section S6, Supporting Information). The experimental procedure was as follows; first, the sample was heated from room temperature to 473 K in oxygen (p = 1400 Pa) for removing surface contaminants, as was done in the electrical experiments. AP-XPS experiments were performed at 473 K in the presence of oxygen. Then, after cooling down to 307 K, the oxygen gas flow was stopped, and methane gas (1400 Pa) was exposed on the cleaned IrO 2 nanosheet, which was then subjected to stepwise heating to 373 and 473 K in the presence of methane. Finally, heating in oxygen at 473 K was performed once again.
In an oxygen atmosphere, the core-level spectrum consisted of Ir 4f 5/2 and Ir 4f 7/2 doublet peaks of IrO 2 , as well as a shake-up satellite, which originates from the excitation of valence electrons into unoccupied Ir d states of the core-hole final state. [48] When methane was introduced at the sample temperature of 307 K, new peaks are observed at a lower binding energy than the IrO 2 main peaks; these peaks are attributed to reduced Ir atoms, [49][50][51] indicating that a part of the IrO 2 nanosheets was reduced at room temperature. However, the main IrO 2 peaks remained at room temperature under the methane atmosphere, suggesting that only minor sites in IrO 2 , such as the rims of nanosheet domains, had been reduced by methane at 307 K.
After heating above 373 K, a substantial increase (decrease) in the intensity of the Ir (IrO 2 ) peaks could be clearly seen. Moreover, O 1s spectra of the sample showed that the peak intensity of the lattice oxygen in IrO 2 nanosheets was considerably decreased when the sample was heated in methane (see Figure S8 in the Supporting nformation), demonstrating the nearly complete reduction of IrO 2 nanosheets by methane at higher temperature. The AP-XPS results are consistent with the electrical resistance measurements that show a slight decrease in resistance at 300 K Figure 4. a) The Ir 4f AP-XPS spectrum of the IrO 2 nanosheet monolayer film on SiO 2 in the presence of oxygen (1400 Pa) at 473 K. Then, after cooling to 307 K, oxygen gas is evacuated, and methane gas (1400 Pa) is exposed on the cleaned IrO 2 nanosheet, followed by stepwise heating to 373 K and 473 K in the presence of methane. (b-d) The AP-XPS spectra of the monolayer film during the reaction with methane as a function of temperature. After the reaction with methane, the sample is oxidized again in an oxygen atmosphere. e) Ir 4f of the re-oxidized sample in oxygen (1300 Pa) at 473 K. All spectra was measured at a photon energy of 720 eV. The solid and dashed red lines indicate the main and shake-up satellite doublet peaks of IrO 2 , respectively; the solid blue lines are the Ir peaks and the thin gray lines at the bottom correspond to the background levels.
and a significant resistance decrease at temperatures above 373 K due to the reaction of the nanosheet with methane. The reduced nanosheets were re-oxidized by heating to 473 K in the presence of oxygen ( Figure 4e); its Ir 4f spectrum consist of the IrO 2 peaks. These spectral changes depending on the reaction condition are in good agreement with electrical resistance changes observed in the electrical experiments and support the notion of the reversible redox reaction of IrO 2 nanosheets by methane and oxygen.
As further support for the occurrence of the reaction, the reaction products were spectroscopically tracked. The decomposition of methane on metallic films is widely used for the chemical vapor deposition of graphene. [52,53] Therefore, it is expected to result in the deposition of graphitic or amorphous carbon in addition to reaction products desorbed as gas (CO, CO 2 , and H 2 O). Raman scattering spectroscopy is frequently used as a spectroscopic detection method for graphitic substances. Figure 5 shows the Raman scattering spectra of IrO 2 nanosheet thin films before and after exposure to methane. The formation of amorphous carbon deposits is a competitive process with the formation of reaction products that desorb in gaseous form. Consequently, methane exposure was conducted for a long time (overnight) Figure 5. Raman scattering spectra of IrO 2 nanosheet ultrathin films before and after exposure to methane, normalized using the Si-related peak around 1000 cm −1 . Amorphous carbon (a-C) peaks are discernible at ≈1300-1600 cm −1 .
at 473 K to provide sufficient deposition to be detectable using Raman scattering spectroscopy. These spectra were normalized to the Si-related peak at ≈1000 cm −1 . The presence of graphitic deposits can be identified by the presence of a peak at ≈1300-1600 cm −1 . [54][55][56] Two peaks in this wavenumber range can be observed before methane exposure and possibly formed during the process of removing the dispersant (tetrabutylammonium) from the IrO 2 nanosheet dispersion. The intensities of these peaks substantially increase after methane exposure. Because the peak at ≈2700 cm −1 , which is characteristic of graphite, is not visible, the deposits possess an amorphous-carbon-like structure rather than highly crystalline graphite. [56] The carbon species remaining on the surface can also be observed in the C 1s AP-XPS spectra measured in the methane atmosphere (see Figure  S8 in the Supporting information). Confirmation of the presence of such amorphous carbon deposits again supports the notion that methane oxidation reaction actually occurred on the IrO 2 nanosheets.

Conclusion
In this study, we succeeded in electrically detecting the methane oxidation reaction on IrO 2 at 300 K using 2D nanosheets sensitive to surface phenomena. The electrical resistance of atomically thin IrO 2 nanosheet films decreased upon methane exposure, the resistance decrease diminishing with the coexistence of oxygen. These results suggest that the reduction of IrO 2 nanosheets by methane oxidation caused the resistance decrease and that the consumed lattice oxygen was recovered in the presence of oxygen. This notion was confirmed by the shift of the Ir 4f peak position due to the sequential exposure to methane and oxygen, as visualized using AP-XPS. In addition, Raman scattering spectroscopy showed that amorphous carbon was deposited on the surface of the IrO 2 nanosheet films after prolonged exposure to methane at high temperatures, thus confirming that the C-H bond cleavage of methane actually occurred.
Electrical resistance changes associated with chemical reactions can be an extremely simple platform for monitoring surface chemical reactions. Surface reactions mostly cause an increase in electrical resistance owing to increased surface scattering of conduction electrons by adsorbed atoms/molecules. However, al-though the surface scattering is increased by the adsorption of reaction products, the increase also occurs when the reactants (methane molecules in this study) are adsorbed without any reaction. In contrast to the resistance increase, via which distinguishing between reactants and reaction products is fundamentally difficult, a clear decrease in electrical resistance was observed in this study. To achieve catalytic cycling, for methane oxidation on IrO 2 , lattice oxygen defects generated during the reaction must be recovered. In this study, the disappearance of the resistance decrease with the coexistence of oxygen was observed, indicating that the catalytic cycle was realized in this system. Thus, chemical reaction monitoring using changes in electrical resistance can provide insight into reactions through the combined exposure to the gases involved in the catalytic cycle.
The applicable sample size for this method is limited by the current electrode fabrication technologies. In general, lithographic resolution determines the smallest limit, which is in the sub-100 nm range. Adopting nanogap formation technologies [57] can further reduce the limit down to a single nanometer scale. For larger samples, no upper limit basically exists because lithographic techniques can be easily scaled up. However, these considerations are based on planar device technologies, which are easily applicable to samples deposited on a flat surface such as thin films grown on a wafer (as in the present study) and nanoparticles supported by a flat insulating substrate. For other forms such as pelletized powders, lithographic techniques are hard to be applied, and the detection limit of the electrical current flowing through the samples determines the size limit of the interelectrode spacing.
Chemical reaction monitoring systems, such as the one presented in this study, are applicable to heterogeneous catalytic systems in general. In particular, the high time resolution of the present method makes it suitable for monitoring reaction systems operated under dynamically changing conditions. The results of this study are an important example of the usefulness of simple chemical reaction monitoring based on changes in electrical resistance and are also significant because they provide an experimental protocol for obtaining important insights into the catalytic cycle.

Experimental Section
Sample fabrication processes and measurement conditions were briefly described here, and the details are outlined in the Supporting Information.
IrO 2 nanosheet films were deposited on a SiO 2 (285 nm thick)/Si substrate using layer-by-layer deposition. Photolithographic processes were conducted to define the edge of the films for measuring the film thickness. The substrate with the photolithographically patterned photoresist window was immersed sequentially in an aqueous solution of polycation (polyethyleneimine; PEI) and a colloidal suspension of IrO 2 nanosheets, followed by a photoresist liftoff process. The thickness of the IrO 2 nanosheet films was controlled by the concentration of the PEI solution. Heat treatment in air was performed to remove the PEI and dispersant (tetrabutylammonium). Finally, Au electrodes (30 nm thick) with a Cr adhesion layer (1 nm thick) were formed using vacuum deposition through a shadow mask. The inter-electrode spacing was designed to be 50 μm.
The surface morphology and thickness of the nanosheet films were measured using AFM in the dynamic force mode (Hitachi High Technologies AFM-5200S). In the present study, monolayer and 9 nm thick films were used for electrical measurements. The electrical measurement chamber was evacuated using a rotary pump, and the base pressure of ≈50 Pa was believed to affect the measurement as residual air. Gas flow rates during electrical measurements were controlled using a mass flow controller. Raman scattering spectra were acquired using a Raman microscope (Raman-DM, Nanophoton) at an excitation of 532 nm.
AP-XPS measurements of the IrO 2 nanosheets on the SiO 2 in the presence of methane and oxygen gases were performed at the beamline BL07LSU of SPring-8. [58] The details of the AP-XPS system were described elsewhere. [59,60] The binding energy of the XPS spectra was referenced at the Fermi energy of a gold foil attached to the samples. Peak fitting of the obtained XPS spectra was conducted using the UNIFIT software. In the fitting procedures, the weighted Shirley and Tougaard background was included, and each peak of the Ir 4f core-levels was fitted by the doublet Doniach-Sunjic functions. The C 1s and O 1s spectra shown in the Supporting Information were fitted using the Voigt functions.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.