Selective light-activation of sensing regions in hybrid Au-graphene-TiO2 chemiresistive gas sensor

Controlling the sensing activity of novel gas sensors is a desirable and simultaneously challenging issue. In this work, ink-printed graphene-based hybrid sensors are demonstrated, and the possibility of selective activation of gas-sensitive components with light is presented. Graphene flakes were mixed with TiO 2 nanoparticles and decorated with Au nanoparticles to produce sensors of photocatalytic and plasmonic properties (Au-G-TiO 2 ). The sensors decorated with Au exhibited enhanced selectivity to NO 2 and a detection limit of 10 ppb under UV light (275 nm) by DC resistance measurements. Activating the plasmon resonance with green light (515 nm) in Au improved sensor sensitivity to NO 2 compared to dark conditions but did not enhance response and recovery time constants as UV light. Low-frequency resistance noise measurements confirmed that the Au-G-TiO 2 sensor exhibits ∼1/ f spectrum shape with a characteristic Lorentzian at 100–200 Hz, ascribed to Au decorating the defect sites in graphene. Noise spectra revealed that different parts of the sensing surface are activated by irradiation. Green light induces charges in the vicinity of Au, whereas UV light excites parts with TiO 2 . These phenomena generate Lorentzians in noise power spectra of intensity and position dependent on NO 2 concentration, making the noise features a sensitive and selective sensing platform. The sensors exhibited long-term stability of their performance with reproducible DC and noise features for more than four months (resistance baseline shift of 5 %), showing the perspectives of their practical applications. • Au decoration enhances NO 2 detection to 10 ppb limit. • Au-decorated sensor generates characteristic Lorenztian in 1/ f noise spectra. • The intensity and position of Lorentzian depend on NO 2 concentration. • Green and UV light activate selectively regions in Au-graphene-TiO 2 sensors. • Sensors exhibit time stability and reproducible DC and noise characteristics.