Experimental determination of valence band maxima for SrTiO3, TiO2, and SrO and the associated valence band offsets with Si(001)
Experimental determination of valence band maxima for SrTiO3, TiO2, and SrO and the associated valence band offsets with Si(001)
复制标题
DOI：
10.1116/1.1768525
复制
发表时间：
2004-07-01
影响因子：
1.4
通讯作者：
Gutowski, M
中科院分区：
文献类型：
--
作者：
Chambers, SA;Droubay, T;Gutowski, M
我们解决的问题，准确测定的价带顶（VBM）的SrTiO 3（001）单晶和外延薄膜，以及TiO 2（001）SrTiO 3和SrO外延薄膜。这些测量是至关重要的，在确定这些氧化物与Si的异质结中的价带偏移。本文分析了三种不同的方法：（1）用高斯加宽的理论态密度拟合X射线光电子价带谱，（2）寻找跨越X射线光电子价带前沿线性部分的回归线与价带顶和费米能级之间的背景的交点，（3）用高斯加宽的理论态密度拟合X射线光电子价带谱，（4）用高斯加宽的理论态密度拟合X射线光电子价带顶和费米能级之间的背景。以及（3）确定前沿处的高分辨率紫外光发射强度变为零的能量。我们发现，当与密度泛函理论结合使用时，方法I产生物理上不合理的结果，因为后者不能以足够的精度预测这些氧化物中价带的详细形状。相比之下，方法2和3给出了物理上合理的结果，相互之间具有良好的一致性。方法1与方法2和3之间的VBM的差异为0.4- 0.6eV，取决于氧化物。方法2和3产生最可靠的VBM，提供的实验进行了足够的能量分辨率。(C)2004年美国真空学会。
We address the issue of accurate determination of the valence band maximum (VBM) for SrTiO3(001) single crystals and epitaxial films, as well as TiO2(001) anatase and SrO epitaxial films. These measurements are of critical importance in determining valence band offsets in heterojunctions of these oxides with Si. Three different methods are analyzed: (1) fitting a Gaussian broadened theoretical density of states to the x-ray photoelectron valence band spectrum; (2) finding the intersection of a regression line that spans the linear portion of the x-ray photoelectron valence band leading edge with the background between the valence band maximum and the Fermi level; and (3) determining the energy at which high-resolution ultraviolet photoemission intensity at the leading edge goes to zero. We find that method I yields physically unreasonable results when used in conjunction with density functional theory because the latter does not predict the detailed shape of the valence bands in these oxides with sufficient accuracy. In contrast, methods 2 and 3 give physically reasonable results that are in good mutual agreement. The difference in VBM between method 1 and methods 2 and 3 is 0.4-0.6 eV, depending on the oxide. Methods 2 and 3 yield the most reliable VBM, provided the experiments are carried out with adequate energy resolution. (C) 2004 American Vacuum Society.