Organic Hydrogen Sensors for Potential Use in Safety-Critical Environments

TL;DR

The paper introduces a metal-free organic hydrogen sensor based on $Alq_3$ in a vertical-stack architecture designed for safety-critical, oxygen-depleted environments. The sensor exhibits a linear, high-sensitivity response to hydrogen across a wide concentration range, with a relative signal up to 3.5% at 100% vol H2, and its rise/fall kinetics can be tuned by an external magnetic field. Mechanistic evidence points to a combination of bulk diffusion of H2 within the organic layer and interfacial modulation of the bottom-electrode barrier, supported by comparative studies with 4CzIPN and DFT analyses showing modest changes in optical properties but notable structural and energetic changes at the molecular level. The work lays groundwork for scalable, low-cost hydrogen sensing suitable for real-time monitoring in fuel cells and related applications, though practical deployment will require improved encapsulation to mitigate humidity effects and temperature-induced mobility changes.

Abstract

Accurate monitoring of the hydrogen concentration is critical for optimizing fuel cell performance, minimizing purge losses, and reducing long-term degradation. Conventional hydrogen sensors often rely on catalytic materials and face limitations such as the need of oxygen purging when operated in fuel cell environments. Here, we report the discovery of a novel hydrogen-sensing mechanism based on organic molecules, without the use of catalytic metals. The sensor is based on a typical vertical stack geometry, containing $\mathrm{Alq_3}$ as active organic material. Upon exposure to hydrogen, the device shows an increase in resistivity, yielding a reliable sensor signal that varies linearly with hydrogen concentration, temperature, and humidity, and exhibits a relative response of up to 3.5 % at 100 %vol hydrogen. By exposing the sensor to an external magnetic field, the rise and fall times of the sensor response were found to be tunable. This novel organic sensor demonstrates sensitivity across a wide range of hydrogen concentrations under fuel cell-relevant conditions and beyond. This new class of hydrogen sensors with high miniaturization potential and cost efficiency paves the way for real-time hydrogen monitoring and advanced control strategies in fuel cells, the chemical industry, or energy storage applications.

Figures (20)

• Figure 1: a Chemical structure of the hydrogen-sensitive organic compound $\mathrm{Alq_3}$. Additionally, the band alignment for the $\mathrm{Alq_3}$-based $\mathrm{H_2}$ sensor is shown. b Demonstrator chamber together with the used sensor, placed on a printed circuit board (PCB). On the right-hand side, an optical microscopy image of the full layer stack after the deposition of the epoxy resin is shown. The structured electrode with a width of the bars of 50µ m can be seen. c Raw sensor response for various bias voltages and hydrogen concentrations. d Sensor response corrected by the background signal using a spline function (IRSQR). e Linear behavior for the sensor response towards hydrogen concentration.
• Figure 2: a Sensing behavior of $\mathrm{Alq_3}$ compared to the common conor-acceptor type emitter molecule 4CzIPN. $\mathrm{Alq_3}$ exhibits a much more stable response to the recurring hydrogen concentration. b Current and photocurrent response of an $\mathrm{Alq_3}$ based sensor with the marked regions shown in c. In c and d, the background was subtracted using a spline function. Even without radiative recombination, after the photcurrent drop, the sensor response remains stable. e Molecular structure for the 4CzIPN molecule, with the corresponding band alignment in f.
• Figure 3: a Schematic of the two investigated devices: Stack PE and Stack PE+$\mathrm{Alq_3}$. b Response of both devices to hydrogen exposure: the resistance decreases for the Stack PE, while it increases for the Stack PE+$\mathrm{Alq_3}$. c Rise and fall times of both devices, showing that the Stack PE exhibits nearly five times longer rise and fall times than the Stack PE+$\mathrm{Alq_3}$.
• Figure 4: a Impact onto the sensor response of the active layer thickness of the $\mathrm{Alq_3}$ layer, which was varied from 20nm to 60nm, in 20nm steps, and b shows the difference in sensor response upon various encapsulation methods, choosing either a structured or closed electrode, and encapsulation either by epoxy resin with or without a microscope slide (cf. Figure S8).
• Figure 5: Evaluation of rise and fall times of the sensor response in panels a–c, with panel a showing a single cycle of the sensor response to hydrogen. In all three diagrams, the sensor response was inverted and normalized, such that the signal increases during the rise time and decreases during the fall time. However, the raw sensor signal itself decreases under hydrogen exposure due to the increase in resistance. In b the determination of the rise time is depicted, while the fall time evaluation is shown in c. Please note that the sensor response was reversed for this demonstration. The $t_{90}$, and $t_{10}$ times were determined for a relative change between $0\mathrm{\% \, vol}$ and $100\mathrm{\% \, vol}$$\mathrm{H_2}$. In d, the rise and fall time for several measurements and devices are shown. In e, the dependence of the rise and fall times on the application of an external magnetic field is compared for increasing and decreasing hydrogen concentrations. f illustrates the effect of the magnetic field on the sensor signal determined at $100\mathrm{\% \, vol}$, $50\mathrm{\% \, vol}$, and $0\mathrm{\% \, vol}$$\mathrm{H_2}$, respectively.