Real-time oxygen tracking shows hematite photocurrent overstates water splitting at low currents

Photoelectrochemists commonly read current as a count of product. The phys.org report calls that assumption widespread, and the Imperial results challenge it ^[8]. If every charge crossing a hematite surface ended up in an oxygen molecule, current and oxygen would be the same number. The Imperial measurements say they are not, at least at low current. In that range, much of the charge went into competing surface reactions ^[5].

The dark run is the experiment's control. The switch still appeared with the light off ^[7], so illumination drops out of the explanation. The team attributes the behaviour to hematite itself ^[7].

Past the switch, once enough positive charge had built up at the surface, oxygen-production efficiency rose to around 80% ^[6]. Even at that best reported point, roughly a fifth of the charge went somewhere other than oxygen ^[13]. The phys.org account does not give the efficiency below the switch, the current at which it occurs, or what the competing surface reactions produce.

Catching the gap took oxygen detection sensitive enough for slow reactions ^[9]. "Its sensitivity allowed us to measure oxygen at reaction rates where current alone would have given an incomplete picture," said Eisner, now at Queen Mary University of London ^[10][3]. Ifan Stephens of Imperial's Department of Materials credited the mix of methods. "It is amazing to see how multiple techniques together, in this case electrochemistry, operando optical spectroscopy and electrochemical mass spectrometry, can make us rethink our mechanistic understanding of a reaction," he said ^[11]. The paper's title ties oxygen selectivity to the reaction order of water oxidation on hematite photoanodes ^[4].

The thing this doesn't tell you is how much a working device is affected. The losses sit at low current ^[5]. So the error in a current-only oxygen figure depends on where in its operating range a photoanode runs. I'd expect low-current comparisons between materials to be the most distorted and high-current device figures less so, on the condition that the switch holds across the hematite preparations other groups make. That condition carries weight because hematite is one of the most widely studied materials in solar water splitting ^[2].