Time‐resolved microwave conductivity is used to compare aqueous‐soluble organic nanoparticle photocatalysts and bulk thin films composed of the same mixture of semiconducting polymer and non‐fullerene acceptor molecule and the relationship between composition, interfacial surface area, charge‐carrier dynamics, and photocatalytic activity is examined. The rate of hydrogen evolution reaction by nanoparticles composed of various donor:acceptor blend ratio compositions is quantitatively measured, and it is found that the most active blend ratio displays a hydrogen quantum yield of 0.83% per photon. Moreover, it is found that nanoparticle photocatalytic activity corresponds directly to charge generation, and that nanoparticles have 3× more long‐lived accumulated charges relative to bulk samples of the same material composition. These results suggest that, under the current reaction conditions, with ≈3× solar flux, catalytic activity by the nanoparticles is limited by the concentration of electrons and holes in operando and not a finite number of active surface sites or the catalytic rate at the interface. This provides a clear design goal for the next generation of efficient photocatalytic nanoparticles.
11 June 2018 in The Journal of Physical Chemistry C
The complex interplay between local chemistry, the solvent microenvironment, and electrified interfaces frequently present in electrocatalytic reactions has motivated the development of quantum chemical methods that can accurately model these effects. Here, we predict the thermodynamics of the nitrogen reduction reaction (NRR) at sulfur vacancies in 1T'-phase MoS2 and highlight how the realistic treatment of potential within grand canonical density functional theory (GC-DFT) seamlessly captures the multiple competing effects of applied potential on a catalyst interface interacting with solvated molecules. In the canonical approach, the computational hydrogen electrode is widely used and predicts that adsorbed N2 structure properties are potential-independent. In contrast, GC-DFT calculations show that reductive potentials activate N2 toward electroreduction by controlling its back-bonding strength and lengthening the N-N triple bond while decreasing its bond order. Similar trends are observed for another classic back-bonding ligand in CO, suggesting that this mechanism may be broadly relevant to other electrochemistries involving back-bonded adsorbates. Furthermore, reductive potentials are required to make the subsequent N2 hydrogenation steps favorable but simultaneously destabilizes the N2 adsorbed structure resulting in a trade-off between the favorability of N2 adsorption and the subsequent reaction steps. We show that GC-DFT facilitates modeling all these phenomena and that together they can have important implications in predicting electrocatalyst selectivity for the NRR and potentially other reactions.