A Study of Semiconducting Silicides

An Experimental and Theoretical Study of Semiconducting Silicides and Germanides for Silicon Based Electro-Optic Applications

Terry Golding

Wayne Holland

Arup Neogi

Department of Physics and Materials Science

University of North Texas

Denton, Texas, 76203

Draft White Paper

Executive Summary

We propose a program to investigate the synthesis, materials, and electro-optical properties of semiconducting silicides and germanides – specifically FeSi2-xGex, FeMSi2 (M = Os, Cr, W) and FeMSi2-xGex.

FeSi2 has recently been shown to be optically active in the technologically important 1.3 -1.5m wavelength range. Twelve other silicides have also been identified to be semiconducting, with bandgaps ranging from 0.07 to 2.3eV. Theoretical work also suggests that related germanides will also be semiconducting.

The aim of the program is the ‘holy grail’ for silicon integrated optoelectronic devices – to synthesize silicon based alloys that are optically active, i.e. possess direct bandgaps - thus enabling silicon based bandgap engineering schemes, and silicon based optoelectronic devices.

The experimental program will be assisted by theoretical studies of ab initio density-functional studies to determine suitable alloy compositions and elements for alloying.

The materials will be synthesized by molecular beam epitaxy and by ion beam impantation techniques. Alloy composition, crystalline structure, phase, and strain will be studied using x-ray diffraction and Raman spectroscopy and the bandstructure of the alloys will be determined by photoluminescence, FTIR and temperature dependent magnetostransport studies.

Introduction

The necessity for direct and tunable bandgap semiconductors, such as the II-VI and III-V compounds for opto-electronics and the dominance of Si based microelectronic circuits has resulted in considerable efforts over the past three decades to monolithically integrate these compounds with Si. However, this approach has met with limited success at best, and it is becoming increasingly evident that such a task may be technologically unworkable. An alternative approach may be to explore further the properties of semiconducting silicides. Thisapproach has been made all the more attractive by the success of FeSi2 for electro-optical applications, having a direct band gap of 0.83-0.87 eV at room temperature. For this reason it is by far the most studied. The silicides have high thermal stability, ability to withstand chemicals normally encountered during fabrication process, and the ability to form a passivating layer of SiO2 by thermal oxidation fulfill many requirements imposed by fabrication technology and device compatibility. While the vast majority of silicides are metallic (several for example CoSi2, NiSi2, and TiSi2, have been employed as interconnects and gates in MOS structures in silicon integrated circuits) ‘there are however ~ 13 silicides that are semiconducting with bandgaps ranging from 0.07-0.12eV (hexagonal MoSi2, WSi2 and ReSi2) to 2.3eV (Os2Si3). Of these FeSi2 and OsSi2 have isocrystal structures, possibly permitting bandgap engineered schemes similar to the II-VI and III-V systems as are Ru2Si3 and Os2Si3. Of these four FeSi2 and OsSi2 have the same crystal structure and fesi2 has been studied extensively.

For epitaxial growth and controllable alloying possibilities we must have the same crystal type and stoiciometry between any two binaries. This leads us to the following possible combinations.

FeSi2 and OsSi2. Space group Cmca

CrSi2, MoSi2, and WSi2. Space group P6222

Ru2Si3 and Os2Si3. Space group Pbcn

Germanides. Much less studies. This is what is known.

Know Os2Si3 and Os2Ge3 are both Pbcn as is Ru2Si3 and Ru2Ge3.

Os2Si3 and Os2Ge3 is not determined if they are misible.

Ru2Si3 and Ru2Ge3 are misible.

Os2Ge3 has bandgap 0.87 eV (quasidirect)

OsGe2 – OsGe2

Ternary Silicides

Ternary silicide based compounds with the metal or silicon atoms partially substituted by other metals or germanium respectively.

Ternary compounds in this couple are most attractive for energy gap engineering. However, they have not been synthesized or investigated yet.

They are attractive fo flexible regulation of the fundamental electronic properties of silicides and for epitaxial matching of semiconducting silicide films.

He

Be

Na

0.78eV

Zn

Xe

Ba

Ir

Table I { adapted from [Borisenko (2000)]}: Listing of all the known semiconducting silicides with bandgap and structure type. Note: The group VI silicides all have the same crystal and stociometry type, CrSi2. FeSi2 and OsSi2, and Ru2Si3 Os2Si3.

Systematic study of the ternary silicides is at its very beginning.

Germanides

All of the related germanides are also semiconductors.While a priori estimated possibilities to form ternary silicides by substituting silicon atoms withgermanium look rather simple, only a few have been experimentally obtained and investigated.

Crystal structure

b-FeSi2

The crystal structure of FeSi2 is a base centered orthorhombic in the D(18,2h) (Cmca) space group having 48 atoms per unit cell and with lattice parameters a =0.9863 nm, b = 0.7884 nm, and c = 0.7791 nm. The unit cell has two equivalent Fe sites, each occupied by 8 atoms as well as two inequivalent Si sites with 16 atoms in each.

OsSi2

Materials Synthesis

Materials Characterization

Theoretical Program