Growth of heteroepitaxial GaSb thin films on Si(100) substrates

  • PDF / 680,582 Bytes
  • 7 Pages / 612 x 792 pts (letter) Page_size
  • 87 Downloads / 262 Views

DOWNLOAD

REPORT


Edward Adams, Mark Lavoie, and Stephen Mongeon IBM Corporation, Essex Junction, Vermont 05405 (Received 10 February 2004; accepted 12 April 2004)

The heteroepitaxial growth of GaSb thin films on Si(100) and GaAs(100) substrates is presented. The growth technique involves the use of atomic Ga and Sb species, which are provided by thermal effusion and radio frequency sputtering, respectively. The crystalline quality of the heteroepitaxial GaSb film on the Si substrate is high despite the larger lattice mismatch. Epitaxial quality is determined by high-resolution x-ray diffraction and Rutherford backscatter spectrometry channeling. Atomic-force microscopy is used to monitor the evolution of surface morphology with increasing film thickness. Transmission electron microscopy shows the formation of stacking faults at the Si/GaSb interface and their eventual annihilation with increasing GaSb film thickness. Annihilation of stacking faults occurs when two next-neighbor mounds meet during the overgrowth of a common adjacent mound.

I. INTRODUCTION

The “6.1 Å family” of semiconductor materials, GaSb, InAs, and AlSb, continues to attract considerable attention due to the prospects of producing both novel highspeed electronic and opto-electronic devices.1,2,3 Current thinking is that devices will be fabricated on either GaSb or GaAs wafers. Small-diameter wafers of GaSb are available but naturally conductive as a result of an intrinsic p-type defect. Larger-diameter GaAs wafers are available; but thick buffer layers are required as a result of the large lattice mismatch of 7.8 % between the two materials. The purpose of this investigation is to consider the growth of GaSb on Si wafer substrates. The principal advantage of this marriage is the potential of combining the unique properties of the “6.1 Å family” of semiconductor materials with the highly developed technology for information handling in Si. The existence of siliconon-insulator (SOI) or high-resistivity Si substrates can potentially solve the device-isolation problem. Some additional advantages of Si are substrate cost; the availability of large diameter, high-quality substrates; and a highly developed processing technology. Additionally Si has excellent mechanical properties in terms of strength and thermal conductivity relative to GaAs or GaSb. The growth of III-V materials on Si has been a longterm goal of many because of the possibility of increased a)

Address correspondence to this author. e-mail: [email protected] DOI: 10.1557/JMR.2004.0307 J. Mater. Res., Vol. 19, No. 8, Aug 2004

http://journals.cambridge.org

Downloaded: 01 Apr 2015

device performance by joining materials of these two families on a single chip. The object of practically all these research efforts has involved the growth of GaAs on a Si wafer.4,5 The typically recognized problems of this effort includes a large lattice mismatch, difference in surface bonding (polar versus non-polar), and differences in the thermal expansion coefficient. Investigations have involved schemes such as stra