Topological superconductors have garnered significant attention due to their potential for realizing topological quantum computation.However,a universal computational tool based on first-principles calculations for pr...Topological superconductors have garnered significant attention due to their potential for realizing topological quantum computation.However,a universal computational tool based on first-principles calculations for predicting topological superconductivity has not yet been fully developed,posing substantial challenges in identifying topological superconducting materials.In this paper,we present a numerical method to characterize the superconducting spectrum and topological invariants of two-dimensional(2D)slab systems using first-principles band structure,implemented in the open-source software WannierTools.To more accurately model the superconducting proximity effect,we integrate a phenomenological theory of SC pairing decay module into the program.Our approach can be applied to classical superconductor-topological insulator(SC-TI)heterostructures,SC-semiconductor heterostructures,and intrinsic topological superconductors.The program’s validity is demonstrated using the topological crystal insulator SnTe,the Rashba semiconductor InSb,and the superconductor NbSe2 as examples.We anticipate that this tool will accelerate the discovery of topological superconductor candidates.展开更多
Machine learning(ML)offers considerable promise for the design of new molecules and materials.In real-world applications,the design problem is often domain-specific,and suffers from insufficient data,particularly labe...Machine learning(ML)offers considerable promise for the design of new molecules and materials.In real-world applications,the design problem is often domain-specific,and suffers from insufficient data,particularly labeled data,for ML training.In this study,we report a data-efficient,deep-learning framework for molecular discovery that integrates a coarse-grained functional-group representation with a self-attention mechanism to capture intricate chemical interactions.Our approach exploits group-contribution concepts to create a graph-based intermediate representation of molecules,serving as a low-dimensional embedding that substantially reduces the data demands typically required for training.Using a self-attention mechanism to learn the subtle but highly relevant chemical context of functional groups,the method proposed here consistently outperforms existing approaches for predictions of multiple thermophysical properties.In a case study focused on adhesive polymer monomers,we train on a limited dataset comprising only 6,000 unlabeled and 600 labeled monomers.The resulting chemistry prediction model achieves over 92%accuracy in forecasting properties directly from SMILES strings,exceeding the performance of current state-of-the-art techniques.Furthermore,the latent molecular embedding is invertible,enabling the design pipeline to automatically generate new monomers from the learned chemical subspace.We illustrate this functionality by targeting several properties,including high and low glass transition temperatures(Tg),and demonstrate that our model can identify new candidates with values that surpass those in the training set.The ease with which the proposed framework navigates both chemical diversity and data scarcity offers a promising route to accelerate and broaden the search for functional materials.展开更多
基金supported by the National Key R&D Program of China(Grant No.2023YFA1607400)the National Natural Science Foundation of China(Grant Nos.12274154 and 12274436)+1 种基金the Science Center of the National Natural Science Foundation of China(Grant No.12188101)the Center for Materials Genome,China.
文摘Topological superconductors have garnered significant attention due to their potential for realizing topological quantum computation.However,a universal computational tool based on first-principles calculations for predicting topological superconductivity has not yet been fully developed,posing substantial challenges in identifying topological superconducting materials.In this paper,we present a numerical method to characterize the superconducting spectrum and topological invariants of two-dimensional(2D)slab systems using first-principles band structure,implemented in the open-source software WannierTools.To more accurately model the superconducting proximity effect,we integrate a phenomenological theory of SC pairing decay module into the program.Our approach can be applied to classical superconductor-topological insulator(SC-TI)heterostructures,SC-semiconductor heterostructures,and intrinsic topological superconductors.The program’s validity is demonstrated using the topological crystal insulator SnTe,the Rashba semiconductor InSb,and the superconductor NbSe2 as examples.We anticipate that this tool will accelerate the discovery of topological superconductor candidates.
基金supported by the U.S.Department of Energy,Office of Science,Basic Energy Sciences,Materials Sciences and Engineering Division.
文摘Machine learning(ML)offers considerable promise for the design of new molecules and materials.In real-world applications,the design problem is often domain-specific,and suffers from insufficient data,particularly labeled data,for ML training.In this study,we report a data-efficient,deep-learning framework for molecular discovery that integrates a coarse-grained functional-group representation with a self-attention mechanism to capture intricate chemical interactions.Our approach exploits group-contribution concepts to create a graph-based intermediate representation of molecules,serving as a low-dimensional embedding that substantially reduces the data demands typically required for training.Using a self-attention mechanism to learn the subtle but highly relevant chemical context of functional groups,the method proposed here consistently outperforms existing approaches for predictions of multiple thermophysical properties.In a case study focused on adhesive polymer monomers,we train on a limited dataset comprising only 6,000 unlabeled and 600 labeled monomers.The resulting chemistry prediction model achieves over 92%accuracy in forecasting properties directly from SMILES strings,exceeding the performance of current state-of-the-art techniques.Furthermore,the latent molecular embedding is invertible,enabling the design pipeline to automatically generate new monomers from the learned chemical subspace.We illustrate this functionality by targeting several properties,including high and low glass transition temperatures(Tg),and demonstrate that our model can identify new candidates with values that surpass those in the training set.The ease with which the proposed framework navigates both chemical diversity and data scarcity offers a promising route to accelerate and broaden the search for functional materials.
