A New Approach Using Artificial Substrates for Growth of High-Quality Precipitate-Free HTS Thin Films, Toward Electronic

  • PDF / 254,063 Bytes
  • 7 Pages / 612 x 792 pts (letter) Page_size
  • 10 Downloads / 191 Views

DOWNLOAD

REPORT


C4.4.1

A New Approach Using Artificial Substrates for Growth of High-Quality Precipitate-Free HTS Thin Films, Toward Electronic Device Applications 1

K. Endo,

2,3

4

4

P. Badica, H. Sato, H. Akoh

1

Nanoelectronics Research Institute (NeRI), National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Japan. 2 Institute for Materials Research, Tohoku University, Sendai, Japan. 3 National Institute of Materials Physics (INCDFM), Bucharest, Romania 4 Correlated Electron Research Center (CERC), National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Japan. ABSTRACT High quality thin films of HTS have been grown by MOCVD on substrates with artificial steps of predefined height and width. The surface of the films grown on the steps having width equal to the ‘double of the migration length’ of the atomic species depositing on the substrate is totally free of precipitates: precipitates are gathered at the step edges where the free energy is lowest. The method has several advantages: it is simple, universal (it is independent of the materials, substrates, deposition technique or application) and allows control of precipitates segregates so that the quality and growth conditions of the films are the same as for the films grown on conventional substrates. The method is expected to result in new opportunities for the device fabrication, design and performance. As an example we present successful fabrication of a mesa structure showing intrinsic Josephson effect. We have used thin films of Bi-2212/Bi-2223 superstructure grown on (001) SrTiO3 single crystal substrates with artificial steps of 20µm. INTRODUCTION For devices or integration purposes, component thin films should fulfill several requirements: a) A certain relationship between substrate - film or film - film is necessary (e.g. lattice matching, wetting and chemical stability and compatibility) b) A certain morphology/roughness is necessary (optimum growth conditions should be found and knowledge of the growth mechanism are useful in this respect) c) Controlled properties, usually uniform and reproducible have to be obtained d) Relatively large clean surfaces are desired; no precipitates-segregates e) Fabrication cost should be low. Sometimes this is not a trivial problem especially for the multicomponent materials such as high-Tc superconductors (HTS), manganites or other electronic materials. For example, in-situ, as-prepared BSCCO thin films have shown excellent quality (high Tc and Jc and low and uniform morphology [1]), but on the surface of the film Curich precipitates-segregates are usually found. This is a serious problem toward (layered)

C4.4.2

device fabrication and/or integration! Currently, to avoid precipitates-segregates several methods are available: a) introduction of buffer layers b) changes in the chemical composition of the films c) complex heat treatments/ post-annealings However, these methods are not always convenient since they might influence the quality of the films. To solve this problem