Faceting of Single-Crystal SrTiO 3 During Wet Chemical Etching
- PDF / 73,621 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 20 Downloads / 242 Views
Faceting of Single-Crystal SrTiO3 during Wet Chemical Etching G. C. Spalding, W. L. Murphy, T. M. Davidsmeier, J. E. Elenewski Department of Physics, Illinois Wesleyan University, Bloomington, IL 61702, USA ABSTRACT We use an Atomic Force Microscope (AFM) to study changes in the surface of single-crystal SrTiO3 etched in HF-based solutions. Attention in this work has been focused upon observations of pyramidal pitting – both because of an interest in avoiding etch pits during substrate preparation prior to heteroepitaxial growth, and because of an interest in micromachining this highly polarizable material. We note that (110) SrTiO3 is surprisingly robust against the formation of pits, while pitting is significant on {100} surfaces. Particular etch rates have been measured, and we discuss anisotropies in the rates of dissolution. These data are combined to extract a macroscopic model describing processes relevant to the most extreme pitting, which we show to be associated with surface defects. INTRODUCTION Strontium titanate crystals are often used as substrates for the heteroepitaxial growth of perovskite superconductors [1] or magnetic oxides [2]. Recent papers have focused on HF-based preparations for the (100) surface of these SrTiO3 substrates [3-5]. However, such preparations, intended to produce a highly-ordered, polished surface, can also lead to faceting, which is extremely undesirable in such applications [6]. While faceting has not been the focus of papers promoting HF-based preparations it seems fitting to revisit [7,8], in a systematic fashion, the pitting of SrTiO3 crystals which accompanies HF-based etching. Notably, we have found that the (110) surface of SrTiO3 is surprisingly robust against the formation of such pits. EXPERIMENTAL DETAILS A paraffin-based etch mask was used because conventional resists exhibited poor adhesion to SrTiO3 in strong HF solutions. Areas exposed for etching were quite large, on the order of a few millimeters. Thus, the shape and orientation of the observed pits were not a consequence of any boundary conditions imposed by the mask; instead, they indicate instabilities native to the surfaces of our particular substrates. After removal of the mask, AFM images were obtained using a Metris 2000-NC (Burleigh Instruments) operating in contact mode. Care was taken to ensure that scans continued well onto surfaces which had been masked during etching, so that underetching would not skew measurements of the etch rate perpendicular to the original surface. In order to increase the total scan area, scans of adjacent regions were stitched together. Our code for such stitching, written in IDL (Research Systems), is available online [9]. The density of pits was calculated simply by counting the number of pits within a typical 70 µm × 70 µm scan field, and dividing by the area of the scan. Because these substrates, obtained from Nikko HiTech (San Jose, CA), are {100} edge-cut, we were able to identify the crystallographic orientation of etch pits by direct observation. O6.4.1
RESULTS
Data Loading...
