Autonomous Materials Synthesis Strategies Towards Accelerated Discovery of Novel Functional Nanomaterials for Green Energy Production

Abstract

Electrocatalysts designed by alloying two or more elements (e.g., transition metals) showed promising enhancement in catalytic activity and durability due to synergistic effects from alloying, such as strain engineering or electron exchange. The potential of the alloying approach is constrained by standard synthesis methods (e.g., wet chemistry and thermal decomposition techniques) due to the governing thermodynamics rules (e.g., miscibility of the different alloying elements). That renders the synthesis of high entropy alloys (HEA), with more than 3 elements, to be complicated and tedious to achieve by those conventional methods.In the presented talk we are presenting different strategies to synthesize metastable high entropy alloys (HEA) and their outstanding catalytic performance towards OER, in correlation to their unique chemistries and structures. Rapid cycling of temperature from room 25 °C to highly elevated levels (1700 °C), in conjunction with ultra-rapid cooling, within milli-second range have been demonstrated by different methods including rapid Joule heating, Intense light flashing, and microwave plasma shock. Our results show successful synthesis of non-noble metals HEA NPs, possessing higher catalytic activity than IrO2 catalyst for OER. The nature of the alloying elements dictates OER activity by promoting different oxidation states of the catalytically active transition metals (e.g., Fe, Ni and Co). In addition, a phenomenal in-situ chemical welding of HEA NP to carbon support, yielded HEA catalysts with two orders of magnitude longer durability than IrO2. Our developed techniques resulted in formation of super-strong metal-carbide bonds “chemically welding” HEA NP to the catalyst support.Synthesis of various HEA catalyst and their electrocatalytic performance screening has been carried out by using liquid handler robots and multi-axes robotic arm systems. Moreover, to optimize the design of materials (structure and chemistry), we are using active learning techniques to evaluate the achieved performance results by applying Gaussian process, followed by applying Bayesian Optimization techniques to propose novel chemistries and structures with a higher probability to achieve further improved performances. The presented the study does not open the door only towards developing novel catalysts for energy production. However, it presents a strategy with multiple successfully demonstrated platforms towards accelerated discovery of novel materials. The high-throughput nature of the synthesis protocols presented in this talk can produce tens of thousands of different novel chemistries, at the same time taken by conventional methods to synthesize just a few, at a minimal energy consumption, cost, and manpower.