International team establishes new record for solar hydrogen production

Novel semiconductor technology helps US-led team to a new high in electrolysis yield

Research conducted by engineers from Germany and the US has resulted in the new record for efficiency in solar hydrogen production from a photoelectrochemical (PEC) process. The team reports 16.2 per cent solar-to hydrogen (STH) efficiency, an increase in the previous record of 14 per cent recorded in 2015.

The researchers, from the US Department of Energy’s National Renewable Energy Laboratory in Colorado, Helmholtz-Zentrum Berlin, TU Ilmenau, Fraunhofer ISE and the California Institute of Technology, were building on research originally led by John Turner of NREL (who was also involved in the new research) in the 1990s. Turner developed a concept device made from gallium indium phosphate (GaInP2) grown on top of gallium arsenide (GaAs) which was capable of splitting hydrogen from water using only sunlight as the energy source. This device set an STH efficiency record of 12.4 per cent which stood for many years.

The new device has a different architecture, although it is also based on light-absorbing two-layer (tandem) semiconductor stacks immersed in an acid water solution that conducts electricity and thus acts as an electrolyte as well as a source of hydrogen. In a paper in Nature Energy, the team explains that this one is grown the other way up; rather than starting with the substrate layer and depositing the most active layer on the top, this PEC cell is grown from top to bottom. The team replaced conventional GaAs with indium gallium arsenide (InGaAs), which boosted efficiency considerably, they say. They also added a very thin “window layer” of aluminium indium phosphide (AlInP) on top of the device, followed by an additional layer of GaInP2. These additional layers perform two functions: they eliminate defects at the surface that limit efficiency, and they protect lower layers from the corrosive acidic electrolyte.

The new PEC cell is called a inverted metamorphic multi-junction (IMM) device, and despite its success it still needs refinement before it can become commercially viable, the team warns. The problem is that the cost of hydrogen is still too high; the US Department of energy has set a target of less than $2 per kilogram for commercially-viable methods. To meet this, further improvements are needed in cell efficiency and lifetime. The NREL team is actively pursuing this target, it says, and is particularly looking at increasing the lifespan of the device. One option might be to connect the device to a conventional electrolyser rather than immersing it in an acidic electrolyte, although a techno-economic analysis commissioned by DOE has shown that submerged devices have the potential to produce hydrogen at a lower cost.