海歸學者發起的公益學術平臺
分享信息,整合資源
交流學術,偶爾風月
許多功能材料的特徵是其電子能帶結構具有特定圖案,例如,Dirac材料的特徵是能帶的線性交叉,拓撲絕緣體的特徵是「墨西哥帽」圖案,有效自由電子氣的特徵在於拋物線分散。
為了成功找到這些材料的特徵圖案,手動檢查少量材料的電子能帶結構比較容易做到的。然而現代電子能帶結構資料庫中的數據量不斷增加,手動查找已不切實際。為了解決這個問題,瑞典Nordita、KTH皇家理工學院和斯德哥爾摩大學的Alexander V. Balatsky教授等,提供了一個在線搜索工具,用於在有機材料資料庫(Organic Materials Database)中查找含有某些圖形圖案的電子能帶結構,從而可以找到具備特定圖案電子能帶的候選材料。該工具通過在線高通量計算,能夠在幾秒鐘內從資料庫幾千個能帶結構集合中、在26,739個有機晶體於費米面附近的十個電子能帶內,找到用戶指定的圖形圖案。該工具可對無法手動檢查的大量能帶結構進行自動在線分析,適用於任何別的電子能帶結構資料庫,而且免費提供原始碼。
該文近期發表於npj Computational Materials4: 46 (2018) ,英文標題與摘要如下,點擊左下角「閱讀原文」可以自由獲取論文PDF。
Online search tool for graphical patterns in electronic band structures
Stanislav S. Borysov, Bart Olsthoorn, M. Berk Gedik, R. Matthias Geilhufe & Alexander V. Balatsky
Many functional materials can be characterized by a specific pattern in their electronic band structure, for example, Dirac materials, characterized by a linear crossing of bands; topological insulators, characterized by a 「Mexican hat」 pattern or an effectively free electron gas, characterized by a parabolic dispersion. To find material realizations of these features, manual inspection of electronic band structures represents a relatively easy task for a small number of materials. However, the growing amount of data contained within modern electronic band structure databases makes this approach impracticable. To address this problem, we present an automatic graphical pattern search tool implemented for the electronic band structures contained within the Organic Materials Database. The tool is capable of finding user-specified graphical patterns in the collection of thousands of band structures from high-throughput calculations in the online regime. Using this tool, it only takes a few seconds to find an arbitrary graphical pattern within the ten electronic bands near the Fermi level for 26,739 organic crystals. The source code of the developed tool is freely available and can be adapted to any other electronic band structure database.

擴展閱讀
npj: 金屬玻璃中的軟區—梯度原子堆垛結構與塑性
npj: 硫族金屬元素化物超級原子體
npj: 機器學習添視覺—材料缺陷快分析
npj: 數據驅動:裂縫擴展的微機械變量識別
npj: 相場模擬—再現複雜的金屬材料點腐蝕
本文系網易新聞·網易號「各有態度」特色內容
媒體轉載聯繫授權請看下方