Organic semiconductor photocatalysis offers a route to producing hydrogen peroxide using light and dissolved oxygen. In the process, oxygen gains electrons and protons to form H2O2, while another substance is oxidized to supply those electrons. This pairing makes material stability and donor selection central considerations for anyone assessing the chemistry’s potential in energy conversion or other catalytic applications.
A study of three organic semiconductors—epindolidione (EPI), perylene tetracarboxylicdiimide (PTCDI) and poly(3-hexylthiophene) (P3HT)—found that all could support photochemical oxygen reduction in aqueous conditions. The results also show why peroxide production alone does not establish that a material is acting as a stable catalyst: the semiconductor itself may be consumed during illumination.
How organic semiconductor photocatalysis works
When a semiconductor absorbs light, it creates excited charge carriers: electrons and holes. In the reported reaction, electrons participate in reducing dissolved oxygen to hydrogen peroxide. The corresponding holes oxidize an electron donor, which may be an added organic compound or, for some materials, water.
The researchers tested thin films under oxygen in solutions at pH 2, 7 and 12. Formate, oxalate and phenol were examined as sacrificial electron donors; a blank condition contained no added organic donor. Hydrogen peroxide was detected across a range of these conditions, demonstrating activity under acidic, neutral and alkaline conditions. The specific results depended on the semiconductor, pH and donor.
Distinguishing catalysis from material degradation
A key measure in the study was turnover number (TON): the amount of hydrogen peroxide produced relative to the amount of semiconductor that degraded. A TON at or below one indicates that the semiconductor itself can be serving as the reaction substrate. A substantially higher TON indicates that peroxide production exceeds the measured material loss, consistent with a photocatalytic cycle involving another electron donor.
EPI and PTCDI produced higher TONs with oxalate or phenol in some conditions. Across the reported tests, values ranged from 10 to 800 for these donor and material combinations. By contrast, P3HT commonly had a TON of one or less, indicating that it was often oxidized along with peroxide formation. Its rapid peroxide production in some conditions therefore did not mean it was functioning as a durable catalyst.
PTCDI was the only material with a TON above one in the blank condition. Its energy levels were consistent with the ability to oxidize water, allowing water to provide electrons while oxygen was reduced. The reported PTCDI activity for peroxide production from oxygen and water was 2.7 micrograms of H2O2 per milligram of catalyst per hour at pH 7, and 5 micrograms per milligram per hour at pH 12.
Why semiconductor energy levels matter
The three materials differed in their electronic properties: EPI was ambipolar, PTCDI n-type and P3HT p-type. Their energy levels influence which reactions are thermodynamically accessible and how the material behaves under illumination. In particular, a deeper highest-occupied molecular orbital (HOMO) was associated with greater stability and the ability to oxidize water in the tested materials.
That trend is useful for material selection, but it is not a guarantee of long-term durability. All three semiconductors degraded to some extent. The researchers also found that photooxidation of donor compounds could be measured at EPI and PTCDI electrodes under deoxygenated conditions, while peroxide was not detected in those tests. This supported the conclusion that hydrogen peroxide formation in the photocatalytic experiments came from oxygen reduction, rather than direct photooxidation.
Thin films and colloidal crystals
Thin films provide a defined material format, while colloidal particles can expose more of the semiconductor surface to the surrounding solution. To examine this difference, the researchers prepared EPI nanocrystals stabilized in water with surfactants. The colloids maintained a stable size distribution, averaging about 140 nanometers, over one month; some aggregation was visible after 2.5 months.
Under the tested conditions, EPI nanocrystals produced hydrogen peroxide at a rate two to four times faster than equivalent quantities of thin-film EPI. They also gave better TONs at pH 2. The authors attributed the improvement to greater exposed catalytic surface area. However, the blank-condition TON above one at pH 2 may have involved oxidation of the poly(vinylpyrrolidone) stabilizer, so the result does not mean that water alone was necessarily the electron donor in that case.
Practical implications and remaining challenges
The findings identify several factors that should guide further development of organic semiconductor photocatalysts:
- Track degradation alongside product formation. Hydrogen peroxide output should be interpreted together with semiconductor loss to distinguish catalytic turnover from consumption of the active material.
- Match energy levels to the intended reaction. The material must support oxygen reduction and, where water is the electron source, the corresponding oxidation reaction.
- Consider material geometry. Colloidal particles may provide more exposed surface than planar films, although stabilizers and other components can also participate in oxidation.
- Address charge separation and durability. The study reported quantum efficiency below 1%, with low exciton-separation efficiency identified as a major limitation. Donor–acceptor heterojunctions and nanostructuring were proposed as avenues for future investigation.
For technical decision-makers, the central lesson is that peroxide generation rate is only one part of evaluating a photocatalytic material. Reaction pathway, electron-donor chemistry, charge separation and degradation all affect whether a system can sustain useful turnover. Organic semiconductors demonstrate a broad capacity for light-driven oxygen reduction, including water oxidation with PTCDI, but improving efficiency and preventing material oxidation remain important steps toward practical applications.