|
| 第36期出刊日:2018.7.12 |
鐵電物質中鐵電相極化與反極化共存之謎 |
文/ 凝態科學中心朱明文研究員與 物理系郭光宇教授
鐵電相早在二十世紀就被透徹研究,意指物質中存在可受外加電場翻轉的自動極化,且自動極化具永久性,因這些功能性,鐵電相成為高度實用的材料,並被廣泛應用於消費性電子中。2015年一月美國Rutgers大學的物理學家發表一種鐵電氧化物 – (Ca,Sr)3Ti2O7 (CSTO) (圖一),反直覺地,其自動極化是由兩個不具極化的晶體結構扭曲序化因子組來而來,材料中更存在大量頭對頭(HH)與尾對尾(TT)極化介電壁。頭對頭介電壁必須由電子屏蔽,而尾對尾壁則須由電洞屏蔽,然而,自然界中不存在電子電洞共存的單一物質。因此,CSTO一發表,全球相關學界都關注。
我們於2015年二月由Rutgers大學取得第一手材料,運用暗場電子顯微術、聚焦電子束繞射、原子級電子顯微能譜術、微觀量子力學計算和巨觀蘭道相理論分析投入研究,歷經兩年,成功解開CSTO為何如此特殊的謎團,刊登於指標性物理學期刊 <Physical Review Letters> [2017年10月,119卷157601文],並獲主編選為封面創下臺灣凝態物理學界來的第一次。
CSTO的奧妙之處,在於其二維晶體結構本身就允許於一反極化序化因子(圖一),這個結構特性早在1980年代就由晶體學家利用群論推論出來。我們於2016年初除確認應有的極化原子排列外,居然還發現1980年代推論的反極化排列,然而,極化–反極化共存從未出現於任何文獻中。這個弔詭的問題,意外地在蘭道相變化理論中找到線索。首先,蘭道相變化理論明確指出,相變化的發生必定遵循某對稱性破壞的路徑,且由序化因子來扮演巨觀尺度下對稱性破壞的角色,因此,只要能把蘭道相變化物理學與群論整合起來,那必定可以找到CSTO是否真的允許極化–反極化序化因子共存。在這個想法的指引之下,以群論加蘭道相變化理論仔細推敲CSTO,果然發現過去從未被報導過的的多重、多維度序化因子共存巨觀特性,將此結論與微觀晶體對稱穩定度計算做比較,赫然發現巨觀特性結論居然與當代的微觀計算吻合,更進一步在聚焦電子束繞射、原子級影像中找到吻合度的實驗證據(圖二),如此由巨觀科學指引當代微觀原子級分析,不只像是物理學的文藝復興,其結論更衍生到極化–反極化共存對應的特殊介電區拓墣:頭對頭與尾對尾極化介電壁其實是由一數奈米寬的反極化介電區構成(圖三),因此,頭對頭、尾對尾介電壁不會發生靜電發散效應,也就無須任何相關載子來作靜電屏蔽。
原始論文詳見:https://journals.aps.org/prl/issues/119/15
|

|
圖一、 (a) and (b) Ferroelectric (FE)-A21am crystal structuresalong b¢ and a¢ projections, respectively. White arrows, the antiparallel Ca1/Sr1 and Ca2/Sr2 displacements (black arrow, P). Gray arrows, the b¢-oriented antipolar Ca2/Sr2 displacements. Dashed gray lines, centered lines for revealing the off-center Ca/Sr displacements. (c) The a¢-oriented polarization (black) of an individual perovskite slab in (a) and the atomistic contribution of each crystallographic site. (d) Group-theoretical analysis of the symmetry tree. The symbols in parentheses depict the primary order-parameter directions in 2-, 4-, and 1-dimensional irreps, respectively. Solid (dashed) lines, reported (otherwise) transition pathways.
|

|
圖二、 (a) Calculated phonon dispersion of pristine Ca3Ti2O7 (CTO) that shows the identical antipolar distortion tendency to the CSTO counterpart (inset, a¢-projected HAADF of CTO). (b) The Convergent-beam electron diffraction (CBED) patterns along respective b¢-projection with the BF (gray margined) embedded in the center of the WP. m, mirror. The BF and WP symmetries in (b) agree with the 2mm point-group of A21am.
|
 |
圖三、 (a), (b), and (c) , , and irreps with primary order parameters of P (black arrows), R (red), and T (blue) in the basal-plane vector spaces of (a,a), (0,a), and (0,a), forming 4-domain topologies by the degeneracy labeled on edges. u (dashed lines), secondary order parameter of ferroelastic strain at the DWs. Gray (black) uc, c-projected parent-tetragonal (FE-orthorhombic) lattice. (d) P5 with effective (0,0,b,b). Green arrow, primary order parameter of b¢-oriented antipolar Ca/Sr displacements (A, double-headed for the antipolar nature). The 8 domains (labels on edges) form a 4-domain topology considering the ab-degeneracy. (e) Domain topology upon -, -, -, and P5-irrep condensations. P and A, intertwined macroscopic order parameters in the 4 domains (1-4; otherwise eight domains upon P reversals). (f) The HH and TT domains with a generically sandwiched antipolar-A, Néel-type DW by the inherent domain topology such as 1-2-3 or 1-4-3 in (e). (g) A cross-sectional view of the coexisting HH and TT domains in (f). (h) The HAADF evidence for (g), with the n = 3 defect also showing P5- type antipolar distortion within the DW (yellow).
|
 |
| |
|
|