Exclusive detection of volatile aromatic hydrocarbons using bilayer oxide chemiresistors with catalytic overlayers

The accurate detection and identification of volatile aromatic hydrocarbons, which are highly toxic pollutants, are essential for assessing indoor and outdoor air qualities and protecting humans from their sources. However, real-time and on-site monitoring of aromatic hydrocarbons has been limited by insufficient sensor selectivity. Addressing the issue, bilayer oxide chemiresistors are developed using Rh–SnO2 gas-sensing films and catalytic CeO2 overlayers for rapidly and cost-effectively detecting traces of aromatic hydrocarbons in a highly discriminative and quantitative manner, even in gas mixtures. The sensing mechanism underlying the exceptional performance of bilayer sensor is systematically elucidated in relation to oxidative filtering of interferants by the CeO2 overlayer. Moreover, CeO2-induced selective detection is validated using SnO2, Pt–SnO2, Au–SnO2, In2O3, Rh–In2O3, Au–In2O3, WO3, and ZnO sensors. Furthermore, sensor arrays are employed to enable pattern recognition capable of discriminating between aromatic gases and non-aromatic interferants and quantifying volatile aromatic hydrocarbon classifications.

Chemical Laboratory Co., Ltd., Japan) were used to prepare the catalytic CeO2 overlayer by ebeam evaporation.
Characterization of materials. Field-emission scanning electron microscopy (FE-SEM, SU-70, Hitachi Co., Ltd., Japan) and high-resolution transmission electron microscopy (HR-TEM, JEM-ARM200F, JEOL, Co., Ltd., Japan) were used to observe the morphology and microstructure of the synthesized materials and gas-sensing films, respectively. The pore size distribution and specific surface area of the Rh-SnO2 powder were analyzed using Brunauer-Emmett-Teller (BET) analysis of nitrogen adsorption isotherms (TriStar 3000, Micromeritics, USA). Field-emission electron-probe microanalysis (FE-EPMA, JXA-8530F, JEOL Co., Ltd., Japan) was used to obtain elemental mappings of the bilayer sensors. The phase and crystallinity of the bilayer gas-sensing films were characterized using X-ray diffraction (XRD, D/MAX-2500 V/PC, Rigaku, Japan) equipped with a CuKα radiation source (λ = 1.5418 Å). The chemical states of the bilayer films were analyzed using X-ray photoelectron spectroscopy (XPS, PHI X-tool, ULVAC-PHI, Japan).
Gas-sensing measurements. Prior to the gas-sensing measurements, the as-prepared CeO2coated Rh-SnO2 bilayer sensors were stabilized by annealing at 450 °C for 2 h. The sensors were covered with a Ni-plated Kovar housing cap and stainless-steel gauze and enclosed in a specially designed acetal chamber. Gas-sensing measurements were conducted in dry air through interval exposure to analyte gases. The atmospheric conditions (dry air or analyte gas) were controlled using mass flow controllers and an automatic four-way valve to ensure a constant gas flow rate of 300 cm 3 ·min -1 . The concentrations of analyte gases were independently controlled by mixing ratios between a synthetic gases and dry air. The test chamber was cleaned by fluxing dry air prior to sensor measurements. The sensor resistance was measured using a digital multimeter (DMM6500, Keithley, Tektronix, Inc., USA) equipped The Pd-SnO2 sensor exhibited low responses to all the analyte gases (S = 1.1-6.2 at 300 °C) ( Supplementary Fig. 9a, e), which hindered VAH detection. In contrast, coating the SnO2 surfaces with Pt or Au catalysts slightly increased the gas responses probably because of the electronic and chemical promotions ( Supplementary Fig. 9b, c, f, g) S1,S2 , whereas the response was enhanced substantially less than that of the Rh-SnO2 sensor ( Fig. 3b and Supplementary   Fig. 9d,h), indicating that coating the SnO2 spheres with Rh is an effective method for enhancing the analyte gas responses.

Supplementary Note 3 | Gas selectivity comparison.
The VAH gas selectivity (SVAH/SA) was calculated based on the ratio of the most stable benzene response to the key interfering ethanol response, and the results are plotted as functions of the sensing temperature ( Supplementary Fig. 12). All the SVAH/SA values of the SnO2 and Rh-SnO2 sensors were low and similar. In contrast, the SVAH/SA values drastically increased when the Rh-SnO2 gas-sensing film was coated with the CeO2 overlayer, suggesting the potential for tailoring VAH selectivity by modulating the catalytic overlayer.

Supplementary Note 4 | Multi-component gas-sensing analysis.
The distinctive features ( Fig. 4 and Supplementary Fig. 14) of the bilayer sensor make it suitable for practical applications. Ethanol, HCHO, and acetone are the major indoor pollutants produced by cleaning supplies, alcoholic beverages, culinary use, and resin decomposition S3,S4 . Ammonia is another representative interfering gas that is not only generated by meat tissues during storage and cooking but also emitted by the chemical industry S5,S6 . CO and CH4 are well-known indoor pollutants generated by the incomplete combustion of gases or fuels S7,S8 . Notably, these six interfering gases are ubiquitous, and most oxide chemiresistors exhibit higher responses to these gases than to aromatic compounds, which may cause gas alarms to malfunction. Therefore, the selective detection properties of the Rh-SnO2 and 0.4CeO2/Rh-SnO2 sensors were evaluated by comparing the gas responses to 5 ppm of pure BTEXS and BTEXS-containing gas mixtures (comprising 5 ppm of BTEXS and 1-5 ppm of interfering gas) ( Supplementary Fig. 14a). As shown in Fig. 4 and Supplementary Fig. 14, the Rh-SnO2 sensor detection characteristics considerably fluctuated depending on the gas concentration, whereas the 0.4CeO2/Rh-SnO2 sensor exhibited well-defined BTEXS gas detection characteristics, even when the gas mixture contained a high interfering-gas concentration, suggesting that these sensors offer a highly reliable and precise solution for selectively detecting aromatic BTEXS gases. To the best of our knowledge, this is the first study wherein aromatic BTEXS vapors were highly selectively and sensitively detected with negligible cross-responses to other representative indoor pollutants such as ethanol, HCHO, acetone, ammonia, CO, and CH4.

Supplementary Note 5 | Catalytic evaluation.
The analyte gas conversions (η) were calculated using Equation (S1) as follows: where [C]in and [C]out are the gas concentrations before and after the reaction at elevated temperatures from 100 to 475 °C, respectively.
The CeO2-induced catalytic oxidation of VAHs and interference gases was investigated using proton transfer reaction-quadrupole mass spectrometry (PTR-QMS 300, Ionicon Analytik GmbH, Austria) ( Supplementary Fig. 17). For this, CeO2 powder (0.1 g) was loaded onto a quartz support and placed in the middle of a tubular quartz reactor (length: 400 mm, inner diameter: 8 mm) in a tube furnace. The reactant gas flow rate was maintained at 200 cm 3 ·min -1 . The concentrations of the outlet gas were determined using an online PTR-QMS.
The drift tube conditions were fixed at 600 V, 80 °C, and 2.3 mbar, and the electric field strength/gas number density ratio (E/N) was set to 136 Td (1 Td = 10 −17 V cm 2 ). H3O + was used as the primary ion.

Supplementary Note 7 | Comparison of sensor resistances.
The sensor Ra values were compared to emphasize the unique bilayer sensor design advantages ( Supplementary Fig. 22). Notably, the intrinsic sensor Ra values were maintained despite the CeO2 overlayer coating, indicating that conduction was hardly affected by the outlying CeO2 overlayer along the lower part of the gas-sensing film near the Au electrodes.
This might be because the distinctive bilayer design separated the sensing and catalytic reactions into independent processes, which enabled both the gas response and selectivity to be tailored without altering the sensor resistance. Therefore, the proposed CeO2-overlayer-coated sensors are versatile, facile, economical, and promising platforms for controlling gas responses and selectivities without affecting sensor resistances.

Supplementary Note 8 | Comparison of response and recovery times.
The response and recovery kinetics (τres and τrecov: the times required to reach 90% of the resistance variation when a sensor is exposed to 5 ppm of benzene and ambient air, respectively) of the single-layer and CeO2-coated bilayer sensors were further calculated based on the sensing transients ( Supplementary Fig. 23).