Hydrogen sensing with nanostructures

Citation

(2011), "Hydrogen sensing with nanostructures", Sensor Review, Vol. 31 No. 1. https://doi.org/10.1108/sr.2011.08731aab.002

Hydrogen sensing with nanostructures

Article Type: Nanosensor update From: Sensor Review, Volume 31, Issue 1

It is often argued that some of the greatest prospects for nanosensors lie in applications where present technology is deficient or even nonexistent and a good example is hydrogen sensing. This gas has attracted much recent attention due to its potential role in a perceived future “hydrogen economy” but no existing sensors can determine low concentrations selectively and at low cost. Accordingly, many groups are working on hydrogen-responsive nanosensors. These are based on a range of different approaches and as yet it is unclear which offers the greatest commercial prospects. A selection of research activities involving differing technologies, reported in 2009 and 2010, are considered below.

In 2010, workers from the Korean Ewha Womans University and Yonsei University reported a high-sensitivity hydrogen sensor based on single-walled carbon nanotubes (SWCNTs), modified with Pd nanoparticles. This showed a fast response time (7 s) at a hydrogen concentration of 10,000 ppm and a very low-detection limit of 10 ppm at room temperature. A different approach is being pursued by a group from the Spanish Rovira i Virgili University. Here, a self-ordered array of WO₃ (tungsten oxide) nanodots, ∼50 nm in diameter and each composed of a few nanocrystals (diameter ∼9 nm) was prepared and tested for its response to hydrogen, CO, ethanol vapour and relative humidity. The sensor was highly sensitive to hydrogen and exhibited a linear response over the range 5-20 ppm. Recent research by the University of Florida has investigated another approach to hydrogen sensing: coating semiconducting nanowires, fabricated from GaN, InN or ZnO, with catalytic metals (Pt and Pd). Nanowires biased at small voltages show large changes in currents on exposure to hydrogen at concentrations in the ppm range and InN nanobelts and nanorods were capable of detecting hydrogen down to 20 ppm after coating with metallic catalysts. Importantly, the sensors exhibited no response at room temperature to other gases and vapours which included O₂, N₂O, CO₂ and C₂H₅. A collaborative project involving groups from the University of Central Florida, the Technical University of Moldova and the Russian Academy of Sciences has studied the hydrogen response of 100 nm diameter ZnO nanowires deposited on a SiO₂-coated silicon substrate. It was found that individual nanowire-based sensors displayed a response (ΔR/R) of about 34 per cent to 100 ppm hydrogen at room temperature (22 °C). UV irradiation rather than thermal desorption was used to enhance the sensor’s recovery time (t_90percent−t_10percent) which was about 2 s at 10 ppm hydrogen. Research at the University of Kentucky involves yet another approach: a Schottky diode device based on a 5−100 nm thick film of SnO₂ nanoparticles beneath an interdigitated platinum electrode. A hydrogen concentration of 100 ppm led to the current changing by a factor of 168. The sensor reached half of its full-scale response in <10 s and has good selectivity, with a low response to CO and none to CH₄.

A group from the US Argonne National Laboratory (ANL) have been investigating the two different mechanisms whereby thin palladium films respond to hydrogen. They have demonstrated that, in discontinuous films, exposure to hydrogen leads to closing of gaps between the palladium clusters associated with absorption-induced swelling, causing a reversible increase in conductivity. In the case of continuous films, the resistivity reversibly increases due to conversion of the palladium into less conductive palladium hydride. The crossover between these two mechanisms occurs at a film thickness of ∼5 nm for hydrogen concentrations below the lower explosion limit.