Nanostructured metal oxide semiconductors have received great attention used as the chemiresistive layer of gas sensor to detect the volatile organic compound recently. As indispensable complementary parts for dominative n-type semiconductors, the p-type metal oxides based gas sensors fail to be studied sufficiently, which hampers their practical applications. In this work, the p-type delafossite CuCrO ₂ nanoparticles were synthesized, characterized, and tested for gas sensing, followed by the first principles calculations to simulate the generation of chemiresistive signal. The hydrothermal synthesis time of CuCrO ₂ nanoparticles is optimized as 24 h with a higher proportion of oxygen vacancies but a smaller size, which is confirmed by the microscopy and spectrum characterization and allows for a prevailing gas sensitivity. Meanwhile, this CuCrO ₂ gas sensor is proven to perform a higher selectivity to n-propanol and a low detection limit of 1 ppm. The adsorption sites and charge variations of dehydrogenation at the gas-solid interface predicted by the theoretical analysis are claimed to be crucial to such selectivity. It is an innovative approach to understand the chemiresistive gas sensing by evaluating the preference of charge transfer between the sensor and target gaseous molecule, which provides a new route to precisely design and develop the advanced sensing devices for the diverse applications.

Most of the current bioinspired sensory neuron (BSN) focus on the single-mode sensory data, due to the problems of power supply and integration complexity when involving multimodal sensing signals. This work presents a bimodal BSN consisting of a piezoelectric nanogenerator and a photosynaptic transistor to implement the tactile-visual fusion. The PNG serving as a tactile sensor can effectively drive and modulate the photosynaptic transistor. Since there is no additional energy supply and moving parts, this bimodal BSN has low consumption but high integration level. The strengthened and modulated synaptic plasticity of the sensory neuron based on the synergistic effect of pressure and light stimuli are demonstrated. The pattern recognition results simulated by artificial neural networks confirm that multiple sense fusion with higher recognition accuracy outperforms the single sense.