The development of new organic semiconductors, crucial for various applications including photovoltaics, displays, transistors, and catalysis, is often a slow and laborious process. Traditional methods involve iterative optimization of known molecules, changing one factor at a time, which becomes increasingly challenging with increasing design space. Donor-acceptor heterojunction materials offer enhanced performance, but the chemical search space expands dramatically. Furthermore, semiconducting properties are influenced by both chemical structure and material morphology/microstructure, making processing conditions critical and leading to variability in reported properties across different laboratories. High-throughput experimentation (HTE) offers a solution by allowing simultaneous variation of multiple factors, but it's often limited to small-scale batch screening, hindering transferability to bulk production. Organic semiconductor heterojunction nanoparticles have shown promise for solar hydrogen production from water, offering advantages over polymeric semiconductors due to their precisely defined structures, ease of purification, and minimal batch-to-batch variation. However, small-molecule organic semiconductors have received less attention as photocatalysts for hydrogen evolution. This article presents an approach to accelerate the discovery of small-molecule nanojunction photocatalysts.

The literature extensively documents the use of organic semiconductors in various applications, highlighting the challenges in optimizing their performance. Studies on high-throughput experimentation for materials discovery show its effectiveness in identifying new functional materials, but the scalability issue remains a significant limitation. Research on organic semiconductor heterojunction nanoparticles for solar hydrogen production reveals their potential, but the control over molecular weight and morphology in polymeric semiconductors poses challenges. Small molecules offer advantages such as precisely defined chemical structures and ease of purification, making them ideal candidates for improved reproducibility and scalability. However, their application as photocatalysts for hydrogen evolution has been limited. This study aims to bridge the gap between small-scale high-throughput screening and large-scale production of small molecule organic semiconductor photocatalysts.

This study employed a three-stage approach: combinatorial synthesis, automated high-throughput screening, and scaled-up flow synthesis. First, a combinatorial library of 26 molecular acceptors (A) was synthesized using a metal-free Hantzsch pyridine condensation reaction. These acceptors were combined with six different electron-rich molecular donors (D) using ultrasonic nanoprecipitation (UNP), generating a library of 186 donor-acceptor hybrids. This library was then tested for sacrificial hydrogen evolution using a high-throughput automated workflow, integrating automated solution preparation, parallel photolysis, and automated gas chromatography analysis. The most active donor-acceptor materials were subsequently scaled-up using a flow-based flash nanoprecipitation (FNP) process. UNP, being suitable for high-throughput screening due to its ease of operation and small material requirements, was paired with an automated screening workflow. The UNP process involved dissolving donor and acceptor molecules in tetrahydrofuran (THF), injecting them into water under ultrasonication, and evaporating the THF to obtain colloidal solutions. The automated screening workflow involved parallel testing of up to 48 samples under uniform illumination. The FNP process, complementary to UNP for large-scale production, used a multi-inlet vortex mixer to rapidly mix an organic solvent (THF) and an antisolvent (water), enabling continuous nanoparticulate material preparation. The optimal donor-acceptor ratio and platinum loading were determined by optimizing the photocatalytic hydrogen evolution rate (HER). Characterizations included scanning electron microscopy (SEM), scanning transmission electron microscopy (STEM), energy-dispersive X-ray (EDX) mapping, cryo-transmission electron microscopy (cryo-TEM), powder X-ray diffraction (PXRD), ultraviolet-visible (UV-Vis) absorption spectroscopy, photoluminescence (PL) spectroscopy, and ultrafast transient absorption spectroscopy. Density functional theory (DFT) calculations were used to investigate the relationship between electronic structure and photocatalytic activity.

The study identified MTPA-CA:CNP147 as the optimal molecular nanojunction photocatalyst, exhibiting a remarkable sacrificial hydrogen evolution rate (HER) of 330.3 mmol h⁻¹ g⁻¹ and an external quantum efficiency (EQE) of 80.3% at 350 nm. This performance surpasses most reported organic photocatalysts. The nanojunctions displayed a one-dimensional nanofibre architecture (~30 nm width, several micrometers length), confirmed by SEM, STEM, EDX mapping, and cryo-TEM. EDX mapping indicated intimate mixing of the donor and acceptor components within the nanofibres. Cryo-TEM revealed crystalline domains within the nanofibres, with lattice spacings matching PXRD data. Photophysical characterizations, including PL spectroscopy and ultrafast transient absorption spectroscopy, demonstrated efficient exciton dissociation and charge separation within the nanojunction. PL quenching and shortened emission lifetime in MTPA-CA:CNP147 compared to CNP147 confirmed efficient exciton dissociation. Transient absorption spectroscopy indicated faster decay of excited-state absorption in the blend compared to single-component CNP147, suggesting efficient charge transfer. DFT calculations showed a volcano-like relationship between HER and binding energy (Eb), with optimal activity observed in the 0.15–0.25 eV range. A similar volcano-like trend was observed with LUMO-LUMO coupling (VLL), while HOMO-LUMO coupling (VHL) showed a monotonic decrease in HER. The study successfully scaled up the synthesis of MTPA-CA:CNP147 using FNP, demonstrating the scalability of the findings from high-throughput screening to bulk production. A 1-liter scale-up resulted in a comparable HER of 305.4 mmol h⁻¹ g⁻¹.

The results demonstrate the effectiveness of combining combinatorial library synthesis, high-throughput automated screening, and scaled-up flow synthesis for the rapid discovery and optimization of molecular nanojunction photocatalysts. The high HER and EQE of MTPA-CA:CNP147 highlight the potential of this approach for developing efficient and scalable organic photocatalysts for hydrogen production. The identified one-dimensional nanofibre morphology plays a crucial role in enhancing charge separation and suppressing charge recombination, resulting in improved catalytic activity. The correlation between binding energy and photocatalytic activity provides a valuable design rule for future development of organic heterojunction photocatalysts. The success of the scale-up process using FNP showcases the feasibility of translating high-throughput screening results to large-scale production, which is a key challenge in materials discovery.

This study presents a materials acceleration platform that successfully integrated combinatorial synthesis, high-throughput screening, and scalable flow synthesis for the discovery of high-performance molecular nanojunction photocatalysts. The optimized MTPA-CA:CNP147 nanojunction showed exceptional performance in sacrificial hydrogen evolution, highlighting the potential of this strategy for various applications. Future research could explore the application of this platform to other photocatalytic reactions and broaden the scope of the combinatorial library to further enhance the performance of these materials. The principles outlined here can be expanded to discover new materials for diverse applications, including drug delivery.

While the study successfully demonstrated the scalability of the photocatalyst synthesis, the synthesis of the molecular building blocks itself was not automated, remaining a time-consuming step. The DFT calculations provided insights into the structure-activity relationships but did not fully capture all factors influencing photocatalytic activity, such as surface hydrophilicity. Further investigation is needed to fully elucidate the roles of these factors and potentially enhance the catalytic performance through targeted modifications.