Hydrogen could be an environmentally friendly alternative to conventional fossil fuels, particularly if it is electrochemically produced from ordinary seawater in the so-called hydrogen evolution reaction (HER), for example in polymer electrolyte membrane cells (PEMCs). However, before this can happen, scientists need to make advanced catalysts that decrease the overpotential of HER. The catalysts also need to be inexpensive, which is not the case for Pt – the most electroactive catalyst for hydrogen fuel cells known today

A group of chemical compounds called dichalcogenides could come into their own here. These easily processed materials, which have the chemical formula MX2, where M is a transition metal (such as Mo or W) and X is S, Se and Te, go from being indirect bandgap semiconductors in the bulk to direct bandgap semiconductors when scaled down to monolayers. This property could make them ideal in a variety of optoelectronics device applications such as light-emitting diodes and solar cells.

Good catalyst for the HER

A team led by Manish Chhowalla has now found that a new crystal structure of WS2 is a good catalyst for the HER. This structure, known as the 1T phase (which is metallic), has an octahedral geometry and has never been studied before. Researchers previously only looked at the so-called (semiconducting) 2H phase, which has a trigonal prismatic geometry.

A monolayer of WS2 is very similar in structure to that of graphene, except that the unit layer is composed of one tungsten atom sandwiched between two layers of sulphur atoms via ionic bonds. As a consequence, a monolayer of WS2 is three atoms thick, as opposed to one-atom thick graphene (which is made up entirely of carbon atoms). Whether or not the unit layer is 1T or 2H depends on the coordination of the W atoms.

Chhowalla and colleagues were able to obtain the 1T phase by chemically exfoliating (or shaving off) layers of WS2. Graphene can be produced in this way too. The process induces a phase transition between the 2H and 1T phases.

Producing hydrogen with very low overpotential

Electron microscopy images of the materials reveal that the basal planes of the exfoliated WS2 nanosheets are not only composed of the 1T phase but that this phase is highly strained too, explained team member Damien Voiry. "Some bonds are under tensile strain, while some are under compression forming a zigzag pattern," he said. "In our experiments, we found that the as-exfoliated WS2 can be used to produce hydrogen with very low overpotential."

The overpotential is the difference between the thermodynamic potential and the potential at which a reaction occurs at the surface of a given material. In other words, it is the extra energy that needs to be supplied to the system to induce the reaction.

The researchers also found that the energy of hydrogen absorbed on the 1T phase depends on the applied strain and that a positive strain significantly reduces the absorption energy. Indeed, calculations by the team predict that an ideal strain of 2.7% would lead to the most hydrogen being absorbed on the WS2 nanosheets. And that is not all: another advantage of this 1T phase is that it improves the kinetics of the HER reaction thanks to the fact that is it highly conducting (because it is metallic).

Engineering strain and structure

"Our work points to new ways in which the catalytic activity of TMDs might be improved by simply engineering the strain and structure of these materials," Voiry told nanotechweb.org. "The attraction of working with these exfoliated materials is that they are very efficient, even though they are single layer. Only a small amount of material is thus needed to cover a large surface – a non-negligible advantage compared to platinum."

The 1T phase of chemically exfoliated WS2 is also highly conductive, so we can imagine developing conducting electrodes made of this material with a very high surface area to increase the density of the catalytic active sites, he added. These electrodes could be then used in PEMCs connected to a solar cell, for example.

Although calculations show that the optimal strain is 2.7% for hydrogen absorption, the material produced by Chhowalla's team is not uniform, with some regions under compression and therefore not active. To overcome this problem, the researchers say that they now plan to work on further increasing the strain in the basal planes of the nanosheets and to increase the ratio of the 1T phase in the exfoliated materials. "We also envisage combining WS2 with other conductive materials such as graphene or carbon nanotubes to improve the kinetics of the HER reaction, with faster electron transport to the catalytic active sites," said Voiry.

The present research is described in Nature Materials.

Further reading

Nanoparticles for hydrogen production (Apr 2011)
Exposed nanofilm edge states make good HER catalyst (Jul 2013)
Nanocomposite targets hydrogen generation (Mar 2009)