Youth Innovation Session

02 October

16:00 - 16:15

0 hr 15 Mins

Identifying the best photocatalysts for Green Hydrogen Production using computational screening

Converting solar energy into valuable fuels has been established as an intriguing strategy to shift traditional fossil fuels towards cleaner, more sustainable energy sources,. In this regard, green hydrogen generation via H2O (or H2S splitting) has consistently drawn attention in both academia and industry. However, seeking robust photocatalysts to achieve higher visible-light conversion efficiency, quantum yields, and selectivity remains a pressing challenge, as a priori, there are thousands of materials that need to be investigated.

Theoretical calculations and high-throughput modeling approaches are becoming indispensable tools for the design of materials with desired properties, as they provide an understanding of the connection between their structural features and their performance, while also providing physical insights into the key properties and mechanisms, thus complementing the experiments and guiding the effective design of photocatalysts.

The team at the RICH center has been working on using computational modeling to guide the design of the best photocatalysts for green hydrogen production. We have been using quantum calculations and machine learning as supporting tools for understanding and searching novel materials for clean energy applications, focusing on hydrogen evolution reaction (HER) from H2S splitting and water splitting. We started the process by calculating the simple key properties any photocatalytic material considered for hydrogen production should have. After it, we used some other screening procedures until we shortlisted the ones that can have an outstanding performance, some of them never tested experimentally. We are currently synthesizing and testing them at the RICH lab, with the goal of scaling up the best ones for testing at industrial conditions.

This contribution specifically focuses on the following interconnected themes:

• Understanding the performance of cadmium-sulfide photocatalyst for H2S splitting.
• Screening single transition metal doped sulfide-based co-catalyst for H2S splitting and H2O splitting.
• Machine Learning-assisted large-scale screening metal sulfide photocatalysts for hydrogen generation

First, we investigated the adsorption and dissociation of H2S on perfect and defective CdS surfaces using DFT calculations. This marks the beginning of our journey in computational modeling, aiming at understanding the relationship between the fundamental properties of the photocatalysts and their performance in photocatalysis. We analyze the electronic properties and surface reactivity in order to provide quantum insights into the photocatalytic activity of CdS-based materials for hydrogen evolution.

Then, we extended our DFT calculations to screen possible transition metals as co-catalysts doped on CdS surfaces for HER. We analyzed the structural stability, electronic properties and adsorption energies of various surfaces to identify promising candidates for future experimental testing. Our results not only provide a fundamental understanding of the role of transition metal doping in enhancing the HER activity of CdS, but also offer valuable guidance for the rational design and optimization of transition metal-doped photocatalysts for hydrogen evolution.

Finally, we employed ML algorithms to assist in a large-scale screening (high-throughput method) for metal sulfide photocatalysts with optimal stability and activity in HER. Geometric, electronic, and energyrelated features were used for training. We were able to predict the structural stability and H adsorption energies over +1,000 unique adsorption structures and identify 10 potential metal sulfide lattices with optimal stability, band gap, and catalytic activity for HER. These candidate materials can be further studied by computational simulations or experiments, and potentially used in photocatalytic HER applications. Such integration of different levels of modeling techniques provides a direct link between the molecular structure of photocatalysts and their catalytic performance that can assist in the optimal design of the synthesis of desired photocatalysts.

The modeling work also applied to some of the experimental works developed at the Research and Innovation Center on CO2 and Hydrogen (RICH) in a complementary manner. These studies have demonstrated the crucial relationships between the fundamental properties of photocatalysts and their stability, catalytic activity and selectivity. Key descriptors have been identified for the design and synthesis of proper semiconductors, aiming at inspiring further simulation and experimental works of tailoring photocatalysts design for HER so that the best ones are scaled up for green hydrogen production directly using solar energy.

Furthermore, the integration of theoretical calculations and machine learning techniques has proven to be a powerful approach to accelerate novel materials discovery at a lower research cost. 

Session Panelists:

Dr. Lourdes Vega

Director of the Research and Innovation Center on CO2 and Hydrogen (RICH)

Khalifah University

Dr. Yuting Li

post-doctoral researcher

Khalifah University

Dr. Daniel Bahamon Garcia

senior research scientist

Khalifah University

Dr. Samar Al Jitan

post-doctoral researcher

Khalifah University

Reem Al Sakkaf

research engineer

Khalifah University

Jaber Almarri

BSc student

Khalifah University