When the researchers used the chloride version, the results were poor. Chloroplatinate precursors leave behind partially reduced platinum–chlorine species stuck to the platinum surface. These residues block the very sites where hydrogen should form — a process chemists call "poisoning" — and severely suppress hydrogen production in organic semiconductor systems that generate a low photopotential, meaning a weak driving force from absorbed light.

The bromide and iodide precursors behaved far better. Both reduce more readily during photodeposition, which largely avoids the poisoning problem and leaves the platinum surface clean and active.

Iodide gives an extra boost

The researchers went one step further. They deliberately modified the platinum surface with iodide ions and measured a clear increase in the rate of hydrogen evolution. The result was the 17% apparent quantum yield at 700 nm — a number that describes how efficiently absorbed photons convert into chemical product. At a wavelength of 700 nm, near the red edge of visible light where many photocatalysts struggle, this figure places the material among the best performers of its type.

The study, published in 2026, shows that the kinetics of platinum photodeposition — how fast the metal deposits under illumination — and the final photocatalytic performance both hinge on the halide ligand in the precursor.

A design parameter hiding in plain sight

The broader lesson extends beyond one material. Co-catalyst surface chemistry, the authors argue, is a critical and often overlooked design parameter in photocatalytic systems. Researchers have spent years tuning the light-absorbing semiconductor while treating the platinum helper as a generic add-on. This work suggests the helper's surface deserves the same scrutiny.

For the organic nanoparticle platform specifically, the practical implications are direct: choosing bromide or iodide precursors, and adding an intentional iodide coating, converts a poisoned catalyst into a highly active one — without changing the semiconductor itself.

Caveats and next steps

As with any laboratory result, some context is warranted. The quantum yield figure applies at a specific wavelength (700 nm) and under controlled laboratory conditions, not yet at the system level relevant to industrial hydrogen costs. The study examined one family of organic semiconductor nanoparticles, and the authors themselves frame the work as uncovering design principles rather than delivering a finished technology. Whether iodide-modified platinum surfaces remain stable over months of operation, and how the approach scales beyond small photoreactors, remain open questions.

Still, the mechanism is now clearer, and it points to a simple, actionable lever. In a field where cost targets are unforgiving and every percentage point of efficiency counts, knowing which halide to reach for — and why — is the kind of insight that compounds.

Publication: Arnau Bertran et al., "Halide Effects on Platinum Co-Catalysts Govern Photocatalytic Hydrogen Evolution in Organic Semiconductor Nanoparticles," ACS Energy Letters (2026). DOI: 10.1021/acsenergylett.6c01928.