磁吸附原子之原子力與基板改造的第一原理計算

Abstract

啟發自二維建造原子自旋與分子磁鐵，我們提出針對三維建造原子尺度自旋進行電腦 計算研究。我們與IBM 矽谷研究中心掃描穿隧顯微鏡實驗室合作，將表面磁原子結構 與表面絕緣層都進行三維建造。該表面是在銅(100)上覆蓋單層氮化銅。此設計使自旋免 受傳導電子遮蔽，又容許穿隧電流從掃描探針通到表面來探測自旋，更進一步對個別原 子進行建構、探測與改造。過去已研究二維原子自旋的耦合與磁異向性。沿伸表面自旋 到第三個維度主要原因有：一、單分子磁鐵是一成熟領域，惟其分子組成不易改造。而 大自旋與大磁異向性的分子磁鐵多為三維結構。二、現行技術僅能造出最大20 奈米見 方的單層氮化銅。要持續增加磁原子數目，勢必要往垂直方向發展。三、原子自旋建在 單層氮化銅上，其分子軌域受下方傳導電子影響較大。我們在2010~2011 的年度研究計 劃中，已尋找到最近用來量測單一自旋鬆弛時間的鐵銅二聚子最穩定的三維結構，我們 正在計算其磁異向性。 密度汎函理論搭配投影綴加波法將用來研究此系統的原子力與雙層氮化銅成長。搬 動一顆原子所需的力對於建造三維磁結構是一重要的物理量，我們將研究單顆銅原子與 有或無原子空缺的單層氮化銅表面的作用力。一層一層長晶雙層氮化銅的機制也會以電 腦計算模擬，在單層與雙層氮化銅表面上，其磁原子的分子軌域也會被作比較。

Inspired by two-dimensional engineered atomic spins (1~10 atomic spins) out of the scanning-tunneling-microscope (STM) moving-atom technique and single molecular magnets, we propose computational studies of three-dimensional (3D) engineered atomic spins. We will collaborate with the STM Lab at IBM Almaden Research Center in engineering, three-dimensionally, both the magnetic atoms and their underlying substrate, where the specially designed substrate surface consists of one insulating copper-nitride monolayer on top of the Cu(100) surface (to be called CuN surface latter), and such a surface provides the opportunity to preserve the spins of magnetic adatoms from being screened by the underlying conduction electrons while at the same time allowing enough tunneling current from an STM tip to probe the spin excitations. Moreover, the magnetic atoms on this surface can be constructed, probed, and manipulated one atom at a time. Previous studies of two-dimensional engineered spins include the exchange coupling and magnetic anisotropy. There are three major reasons to explore the third dimension in the surface-engineered spin systems: First of all, single molecular magnets have formed a quite mature topic despite that atoms in these molecules can not be easily manipulated. Most of the large-spin and large-magnetic-anisotropy molecular magnets indeed form 3D structures rather than planar ones. Secondly, the present experimental technique in fabricating the monolayer CuN on Cu(100) can extend the monolayer CuN area up to 20nm20nm, so the magnetic adatoms will need to be constructed not only along the plan but also along the perpendicular direction if one wants to increase the number of magnetic atoms. Finally, the previously-built atomic spins are placed on one monolayer CuN on top of the Cu(100) surface, where the molecular orbitals of the atomic spins are still substantially affected by the underlying conduction electrons. Thus more than one monolayer CuN is needed to the isolate molecular orbitals. During our 2010~2011 NSC project, we have found the most stable structure of this Fe-Cu dimer, where the Fe-Cu dimer is the recently built in experiments for relaxation time measurement of a single spin.We are currently calculating its magnetic anisotropy energy. Density functional theory (DFT) in the projector augmented wave basis (PAW) will be used to study the magnetic properties, atomic forces of such novel systems. The force needed to move an atom on a surface is an important quantity that determines how one can build a 3D magnetic structure. We will study the force between a single Cu to the CuN surface with/without missing atoms. The mechanism of growing more than one monolayer will be studied computationally through layer-by-layer deposition. The molecular orbitals of magnetic atoms when being placed on top of the thicker CuN will be calculated and compared with those on the monolayer CuN.