Study of Dislocation Mobility in 4H SiC by X-Ray Transmission Topography, Chemical Etching and Transmission Electron Microscopy
Study of Dislocation Mobility in 4H SiC by X-Ray Transmission Topography, Chemical Etching and Transmission Electron Microscopy
复制标题
X射线透射形貌、化学刻蚀和透射电子显微镜研究4H SiC中的位错迁移率
DOI：
10.4028/www.scientific.net/msf.457-460.355
复制
发表时间：
2004
期刊：
影响因子：
--
通讯作者：
B. Pichaud
中科院分区：
文献类型：
--
作者：
H. Idrissi;M. Lancin;G. Regula;B. Pichaud
位错是由悬臂弯曲和压缩退火引起的。它们由化学蚀刻所示的有缺陷的半回路组成。通过蚀刻和XRTT验证了它们在样品中的不对称传播。基于这一特点，提出了一种成核和滑动机理。在550℃时测量了位错速度和应力指数。碳化硅的结构缺陷因其对电性能的影响而受到广泛的研究。对于位错动力学，通过塑性实验和透射电镜观察获得了一些信息[1-8]。这种方法表明，硅核的肖克利部分位错比碳核的肖克利部分位错具有更高的迁移率[3,4]，但对于90°部分[9]的从头计算仍然存在矛盾。此外，无论核心是什么，速度都不是直接测量完全或部分位错的。因此，我们通过化学蚀刻和x射线透射形貌（XRTT）进行直接测量，并通过透射电镜确定位错核心的性质。这项工作涉及该研究的第一步，包括在550,700和1050°C下激发特定的滑动系统来引入位错。在550°C时测量它们的速度作为应力的函数。我们使用由CREE提供的4H-SiC晶圆，通过改进的Lely方法沿[11-20]方向生长。这样的取向确实是最方便引入可控位错的。XRTT使用配备Ag Kα源的Lang装置进行。样品与x射线敏感薄膜结合，用光束扫描给定的衍射族平面，该衍射族平面由其法向g定义。为了表征扩展缺陷，选择了不同的g。在XRTT图像上，暗对比度是由于位于材料中的扩展缺陷引起的应变。图1：生长[11-20]4H-SiC: XRTT图像（g = 1-101）显示a)螺位错(1)基位错（2)微管(3)和b）亚晶界；c)沿[11-20]观看的HRTEM图像。a) c) b) Materials Science Forum Online: 2004-06-15 ISSN: 1662-9752, vol .457-460, pp 355-358 doi:10.4028/www.scientific.net/MSF.457-460.355©2004 Trans Tech Publications Ltd， Switzerland版权所有。未经Trans Tech Publications Ltd， www.scientific.net的书面许可，不得以任何形式或任何方式复制或传播本文的部分内容。使用JEOL 2010F在通过常规机械研磨和离子稀释获得的薄片上实现了200 keV的高分辨率TEM （HRTEM）。XRTT图像（图1a）显示了与基面法向的螺旋位错，其Burgers向量b等于c，基底位错的b为1/3[1-210]或1/3[-2110]，b=1/3[11-20]的位错没有对比。在4cm晶圆内仅检测到两个微管（图1a）。它们平行于c轴。他们的汉堡向量b = 5c，是由微管对比度[10]的厚度推断出来的。XRTT还检测到一些亚晶界（图1b）。此外，通过HRTEM（图1c）评估的晶体结构被认为具有非常好的质量，因为在几百纳米以上的薄箔中没有检测到层错。变形实验步骤：将平行六面体形状的样品（5x20x0.25 mm）切割，最大面平行于（11-20），长度与[2-201]错位1.7°（图2a）。由于滑移系统[1-210]（0001）和[-2110]（0001）的施密特因子均为0.43，因此选择这种几何结构是为了在（0001）滑动面上获得较大的分解剪切应力σr。用金刚石尖沿平行于位错长度的x方向在样品表面划伤位错形成核。在室温下，试样沿试样宽度方向以悬臂模式弯曲（图2b），并在应力下退火。图2:a)试样的几何形状，显示了滑动面、划痕和弯曲轴方向；B)悬臂弯曲原理的截面图。研究了三种变形温度（550°C, 700°C, 1050°C）。之所以选择最低，是因为目前还没有报道单晶SiC在如此低温下的塑性变形[5,7]。由于SiC在1100℃以上的化学计量量越来越小，所以选择了最高。通过激光反射在试样表面的位移测量，记录局部曲率半径，得到应力沿试样长度变化的σr(x)。应力沿试样宽度(y)是恒定的。它在整个深度中减小，在中性面之外改变其符号。样品采用XRTT成像。然后在Al2O3坩埚中用热KOH（500°C）蚀刻12分钟。在XRTT图像上，对比度对应于位于样品中任何深度的扩展缺陷。相反，蚀刻只能显示表面出现的缺陷。结果：值得注意的是，无论变形温度如何，与划痕位置相比，XRTT图像对比度都是不对称的。在图3a中，划痕下方的对比度与未变形样品中观察到的对比度相似（图1a）。在划痕上方观察到的与基面平行的直线与划痕形成的位错相对应。虽然在样品边缘形成核的位错向划痕移动，但它们并不妨碍迁移率的估计。在550℃、700℃和1050℃时，位错主要发生在σr(x)大于（53±2）MPa、（21±2）MPa、（10±2）MPa的区域。这些特征通过化学蚀刻得到证实（图3b）。与出现的位错相对应的单个蚀刻坑从未被观察到。因此，我们假设XRTT图像中的直线和中性平面a) b) 356碳化硅和相关材料2003
Dislocations are introduced by bending in a cantilever mode and annealing under compression. They consist of faulted half loops as shown by chemical etching. Their asymmetric propagation in the sample is attested both by etching and XRTT. Based on such a feature, a nucleation and glide mechanism is proposed. The dislocation velocity and the stress exponent are measured at 550°C. Introduction Structural defects in SiC are being widely studied because of their influence on electrical properties. As for the dislocation dynamics, some information were obtained from plasticity experiments followed by transmission electron microscopy (TEM) observations [1-8]. Such an approach shows that Shockley partial dislocations with silicon core have a higher mobility than those with a carbon core [3,4] but there is still a contradiction with ab-initio calculation for the 90° partial [9]. Moreover, the velocity was not directly measured for perfect or partial dislocations whatever the core. Thus we are carrying direct measurements by chemical etching and X-Ray Transmission Topography (XRTT), the nature of the dislocation core being determined by TEM. This work deals with the first step of that study which consists in exciting specific gliding systems to introduce dislocations at 550, 700 and 1050°C. Their velocity is measured as a function of stress at 550°C. Characterization of the as-grown material We used 4H-SiC wafers provided by CREE, grown along the [11-20] direction by a modified Lely method. Such an orientation is indeed the most convenient to introduce controlled dislocations. XRTT was performed using a Lang set up equipped with an Ag Kα source. The sample is bind to a X-Ray sensitive film and scanned by the beam for a given diffracting family plane defined by its normal g. To characterize the extended defects, different g were chosen. On XRTT images, the dark contrast is due to the strain induced by the extended defects located in the material. Fig. 1 : As-grown [11-20] 4H-SiC: XRTT images (g = 1-101) showing in a ) screw dislocations (1) basal dislocations (2) a micropipe (3) and in b) sub-grain boundaries ; c) HRTEM image viewed along [11-20]. a) c) b) Materials Science Forum Online: 2004-06-15 ISSN: 1662-9752, Vols. 457-460, pp 355-358 doi:10.4028/www.scientific.net/MSF.457-460.355 © 2004 Trans Tech Publications Ltd, Switzerland All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of Trans Tech Publications Ltd, www.scientific.net. (Semanticscholar.org-11/03/20,15:23:17) 2 Title of Publication (to be inserted by the publisher) High Resolution TEM (HRTEM) was realized at 200 keV using a JEOL 2010F on thin foils obtained by conventional mechanical grinding and ion thinning. The XRTT images (Fig.1a) reveal screw dislocations normal to the basal plane whose Burgers vector b is equal to c and basal dislocations whose b is either 1/3[1-210] or 1/3[-2110], those whose b=1/3[11-20] being out of contrast. Only two micropipes within the 4 cm wafer were detected (Fig.1a). They are lying parallel to the c axis. Their Burgers vector, b = 5c, is deduced from the thickness of the micropipe contrast [10]. Some sub-grain boundaries are also detected by XRTT (Fig.1b). Additionally, the crystallographic structure evaluated by HRTEM (Fig.1c) is assumed of very good quality because stacking faults were not detected in thin foils over a few hundred nanometers. Deformation experiments Procedure: Parallelepiped shaped samples (5x20x0.25 mm) were cut with their largest face parallel to (11-20) and their length 1.7° misoriented from [2-201] (Fig.2a). This geometry was chosen to get a large resolved shear stress σr on the (0001) gliding plane since the Schmidt factor is 0.43 in both slip systems [1-210](0001) and [-2110](0001). Dislocations were nucleated by scratching with a diamond tip the sample surface in the direction x parallel to their length. The samples were bent at room temperature around the direction of the sample width in cantilever mode (Fig. 2b) and annealed under stress. Fig. 2: a) geometry of the sample showing the glide plane, the scratch and the bending axis directions; b) cross section view of the cantilever bending principle. Three deformation temperatures (550°C, 700°C, 1050°C) were investigated. The lowest was selected because plastic deformation of monocrystalline SiC has yet been reported at such a low temperature [5,7]. The highest was chosen since SiC becomes less and less stoechiometric beyond 1100°C. The stress varying along the sample length, σr(x) was reached by recording the local radius of curvature via the displacement measurement of a laser reflection on the sample surface. The stress is constant along the sample width (y). It decreases throughout the depth, changing its sign beyond the neutral plane. The samples were imaged by XRTT. They were then etched with hot KOH (500°C) for 12 minutes in an Al2O3 crucible. On XRTT images, the contrast corresponds to extended defects located at any depth in the sample. On the contrary, etching only reveals defects emerging at the surface. Results: It is noteworthy that the XRTT image contrast is asymmetric as compared to the scratch position whatever the deformation temperature. In figure 3a, the contrast below the scratch is similar to the one observed in un-deformed samples (Fig.1a). The straight lines parallel to the basal plane which are observed above the scratch correspond to the dislocations nucleated from it. Though dislocations nucleated at the edges of the sample moved towards the scratch, they did not prevent mobility estimations. The dislocations mainly developed in the area where σr(x) is beyond (53 ± 2) MPa, (21 ± 2) MPa, (10 ± 2) MPa at 550°C, 700°C and 1050°C respectively. Such features are confirmed by chemical etching (Fig.3b). Individual etch pits corresponding to emerging dislocations have never been observed. Thus, we assume that the straight lines in XRTT images and Neutral plane a) b) 356 Silicon Carbide and Related Materials 2003