Bias-Optimized Hydrogen Sensing in a Mo-Electrode Pd/SnO2 Thin-Film Sensor with Integrated Microheater

TL;DR

A new hydrogen sensor using a Mo-electrode Pd/SnO2 thin film with a microheater is developed for reliable low-ppm detection at moderate temperatures.

Contribution

A bias-optimized hydrogen sensor with integrated microheater and thermal insulation for low-ppm detection is introduced.

Findings

The sensor achieves a limit of detection of approximately 6.4 ppm.

A concentration-normalized figure of merit is introduced to assess low-concentration hydrogen sensing performance.

The sensor exhibits stable and repeatable operation in the low-ppm regime.

Abstract

Hydrogen is a key energy carrier for fuel cell vehicles and hydrogen energy systems. However, its colorless and odorless nature, combined with a wide flammability range, poses significant safety risks in the event of leakage. Accordingly, compact and reliable hydrogen sensors capable of low-ppm detection at moderate operating temperatures are essential for early-stage safety monitoring. In this study, a bias-optimized hydrogen gas sensor based on a Pd-functionalized SnO2 thin film with Mo electrodes and an integrated microheater is designed, fabricated, and systematically characterized. The sensor employs a Mo-based vertical microheater and a multilayer thermal insulation stack, enabling thermally efficient and stable operation at 250–280 °C with low power consumption. The electrical and sensing properties of the SnO2 layer are optimized by controlling the oxygen partial pressure during reactive sputtering and post-deposition annealing. The Pd catalytic layer promotes hydrogen dissociation and spillover, resulting in pronounced resistance modulation through surface redox reactions and interfacial charge transport effects. By systematically optimizing the sensing bias voltage, a clear trade-off between sensitivity enhancement and electrical noise is identified, which allows stable and repeatable operation in the low-ppm regime. The sensor response follows a power-law dependence on hydrogen concentration, and an automated measurement platform is employed to evaluate repeatability and statistical performance. Based on baseline noise analysis and concentration-dependent resistance variation, a limit of detection of approximately 6.4 ppm is achieved. Furthermore, a concentration-normalized figure of merit that combines response magnitude and concentration dependence is introduced to quantitatively assess low-concentration hydrogen sensing performance. These results demonstrate that the proposed Mo-electrode Pd/SnO2 thin-film sensor, enabled by bias-optimized operation and integrated thermal control, provides a robust and scalable platform for safety-critical hydrogen leak detection.

Figures