A Model for Heterogeneous Nucleation and Growth of Silicon Nanoparticles on Silicon Dioxide from Disilane

  • PDF / 76,157 Bytes
  • 6 Pages / 612 x 792 pts (letter) Page_size
  • 61 Downloads / 216 Views

DOWNLOAD

REPORT


A Model for Heterogeneous Nucleation and Growth of Silicon Nanoparticles on Silicon Dioxide from Disilane

William T. Leach, Jian-Hong Zhu, and John G. Ekerdt Department of Chemical Engineering, The University of Texas at Austin, Austin, Texas Supika Mashiro, Junro Sakai, and Takayuki Kawshima Anelva Corporation, Tokyo, Japan ABSTRACT A model is presented that describes silicon nanoparticle deposition in terms of disilane decomposition on silicon dioxide, adatom diffusion, nucleation, nanoparticle growth and coalescence. Total nanoparticle densities are output as a function of time, and segregation of nanoparticles into subsets with common size allows size distributions to be reported for all times during the simulation. Model parameters are fit to low pressure chemical vapor deposition data with disilane pressures ranging from 5×10-4 to 5×10-3 Torr and surface temperatures from 510 to 570 °C. Simulations are used to explain how growth pressure and surface temperature influence incubation time, nanoparticle density and size distribution. INTRODUCTION Silicon nanoparticle arrays on dielectrics are increasingly common in semiconductor devices such as nanocrystals based flash memory [1], and intensive research is underway to enable chemical vapor deposition (CVD) of such arrays with better size and density control. This paper reports experimental and theoretical work that attempts to outline mechanisms responsible for trends in incubation time, nanoparticle density and size distributions. Theoretical works describing the initial stages of thin film growth [2-5] were adapted to explain CVD of silicon nanoparticles on SiO2 using disilane. The previous works describe island formation and growth in terms of Si adatom formation from the vapor, surface diffusion, nucleation, and diffusive and epitaxial growth of nanoparticles. Kinetic equations unique to the disilane/SiO2 system were derived that can be solved iteratively to give a time dependent nucleation rate, and nanoparticles formed at all times are tracked as they grow and eventually coalesce. The model described is set apart from previous works is several ways. First, it explains CVD as well as physical vapor deposition (PVD), thus details concerning selectivity of the vapor towards the substrate are addressed. Second, nanoparticles of all sizes are accounted for at all times in the simulation, thus no size averaging is needed and the size distribution can be analyzed for any time during the simulation. Efforts are made to describe temperature and pressure dependent nanoparticle growth trends, and a PVD seeding step is described that yields high density nanoparticles with a narrow size distribution.

A6.3.1

THEORY The atomic scale formulation of nucleation put forth by Walton is utilized to describe the requirements for stability of islands [5]. A population of loosely bound adatom clusters is assumed to exist on the surface in quasi-equilibrium with the adatom population. The largest of these unstable clusters is the critical nucleus, and thermal rearrangement of a