Reversible protonic ceramic cells(RePCCs)hold promise for efficient energy storage,but their practicality is hindered by a lack of high-performance air electrode materials.Ruddlesden-Popper perovskite Sr_(3)Fe_(2)O_(7...Reversible protonic ceramic cells(RePCCs)hold promise for efficient energy storage,but their practicality is hindered by a lack of high-performance air electrode materials.Ruddlesden-Popper perovskite Sr_(3)Fe_(2)O_(7−δ)(SF)exhibits superior proton uptake and rapid ionic conduction,boosting activity.However,excessive proton uptake during RePCC operation degrades SF’s crystal structure,impacting durability.This study introduces a novel A/B-sites co-substitution strategy for modifying air electrodes,incorporating Sr-deficiency and Nb-substitution to create Sr_(2.8)Fe_(1.8)Nb_(0.2)O_(7−δ)(D-SFN).Nb stabilizes SF’s crystal,curbing excessive phase formation,and Sr-deficiency boosts oxygen vacancy concentration,optimizing oxygen transport.The D-SFN electrode demonstrates outstanding activity and durability,achieving a peak power density of 596 mW cm^(−2)in fuel cell mode and a current density of−1.19 A cm^(−2)in electrolysis mode at 1.3 V,650℃,with excellent cycling durability.This approach holds the potential for advancing robust and efficient air electrodes in RePCCs for renewable energy storage.展开更多
Ceramic electrochemical cells(CECs)are promising devices for clean and efficient energy conversion and storage due to their high energy efficiency,more extended system durability,and less expensive materials.However,t...Ceramic electrochemical cells(CECs)are promising devices for clean and efficient energy conversion and storage due to their high energy efficiency,more extended system durability,and less expensive materials.However,the search for suitable materials with desired properties,including high ionic and electronic conductivity,thermal stability,and chemical compatibility,presents ongoing challenges that impede widespread adoption and further advancement in the field.Artificial intelligence(AI)has emerged as a versatile tool capable of enhancing and expediting the materials discovery cycle in CECs through data-driven modeling,simulation,and optimization techniques.Herein,we comprehensively review the state-of-the-art AI applications for materials design and optimization for CECs,covering various material aspects,database construction,data pre-processing,and AI methods.We also present some representative case studies of AI-predicted and synthesized materials for CECs and provide insightful highlights about their approaches.We emphasize the main implications and contributions of the AI approach for advancing the CEC technology,such as reducing the trial-and-error experiments,exploring the vast materials space,discovering novel and optimal materials,and enhancing the understanding of the materials-performance relationships.We also discuss the AI approach’s main limitations and future directions for CECs,such as addressing the data and model challenges,improving and extending the AI models and methods,and integrating with other computational and experimental techniques.We conclude by suggesting some potential applications and collaborations for AI in materials design for CECs.展开更多
基金supported by the Research Grants Council,University Grants Committee,Hong Kong SAR(Project Number:N_PolyU552/20)supported by the National Nature Science Foundation of China(22209138)Guangdong Basic and Applied Basic Research Foundation(2021A1515110464).
文摘Reversible protonic ceramic cells(RePCCs)hold promise for efficient energy storage,but their practicality is hindered by a lack of high-performance air electrode materials.Ruddlesden-Popper perovskite Sr_(3)Fe_(2)O_(7−δ)(SF)exhibits superior proton uptake and rapid ionic conduction,boosting activity.However,excessive proton uptake during RePCC operation degrades SF’s crystal structure,impacting durability.This study introduces a novel A/B-sites co-substitution strategy for modifying air electrodes,incorporating Sr-deficiency and Nb-substitution to create Sr_(2.8)Fe_(1.8)Nb_(0.2)O_(7−δ)(D-SFN).Nb stabilizes SF’s crystal,curbing excessive phase formation,and Sr-deficiency boosts oxygen vacancy concentration,optimizing oxygen transport.The D-SFN electrode demonstrates outstanding activity and durability,achieving a peak power density of 596 mW cm^(−2)in fuel cell mode and a current density of−1.19 A cm^(−2)in electrolysis mode at 1.3 V,650℃,with excellent cycling durability.This approach holds the potential for advancing robust and efficient air electrodes in RePCCs for renewable energy storage.
文摘Ceramic electrochemical cells(CECs)are promising devices for clean and efficient energy conversion and storage due to their high energy efficiency,more extended system durability,and less expensive materials.However,the search for suitable materials with desired properties,including high ionic and electronic conductivity,thermal stability,and chemical compatibility,presents ongoing challenges that impede widespread adoption and further advancement in the field.Artificial intelligence(AI)has emerged as a versatile tool capable of enhancing and expediting the materials discovery cycle in CECs through data-driven modeling,simulation,and optimization techniques.Herein,we comprehensively review the state-of-the-art AI applications for materials design and optimization for CECs,covering various material aspects,database construction,data pre-processing,and AI methods.We also present some representative case studies of AI-predicted and synthesized materials for CECs and provide insightful highlights about their approaches.We emphasize the main implications and contributions of the AI approach for advancing the CEC technology,such as reducing the trial-and-error experiments,exploring the vast materials space,discovering novel and optimal materials,and enhancing the understanding of the materials-performance relationships.We also discuss the AI approach’s main limitations and future directions for CECs,such as addressing the data and model challenges,improving and extending the AI models and methods,and integrating with other computational and experimental techniques.We conclude by suggesting some potential applications and collaborations for AI in materials design for CECs.