3-D Atomistic Kinetic Monte Carlo Simulations of Point Defect Incorporation During CVD Diamond Film Growth

1996 ◽  
Vol 441 ◽  
Author(s):  
C. Battaile ◽  
D. J. Srolovitz ◽  
J. E. Butler

AbstractThe incorporation of vacancies and H atoms into { 100}- and { 111 }-oriented diamond films during CVD growth in an atmosphere of H, H2, CH3, and C2H2is simulated atomistically. The growing films are represented in three dimensions by a diamond cubic lattice, and the temporal evolution of the surfaces is accomplished by a kinetic Monte Carlo method. The dimer bonding of diamond atoms on the { 100} surface is treated explicitly. Growth begins on { 100}(2xl):H and { 11 }:H surfaces containing 288 and 300 atoms, respectively, in the surface plane, and the growth of approximately sixty atomic layers (18,000 atoms) is accomplished at each of fifteen substrate temperatures between 800 and 1500 K. The growth rates depend on the combined contributions to growth from CH3and C2H2, whereas the point defect concentrations depend on the ratio of C2H2to CH3growth. The maximum { 100} growth rate is achieved at approximately 1200 K. Point defect concentrations are low at temperatures below 1300 K, but become high above 1300 K. Growth efficiency, defined as the ratio of growth rate to defect concentration, is maximum for both film orientations at substrate temperatures in the vicinity of 1100 to 1200 K.

1994 ◽  
Vol 340 ◽  
Author(s):  
J. Deppe ◽  
J. V. Lill ◽  
D. J. Godbey ◽  
K. D. Hobart

ABSTRACTThe temperature dependence of germanium surface segregation during growth by solid source SiGe molecular beam epitaxy was studied by x-ray photoelectron spectroscopy and kinetic Monte Carlo (KMC) modeling. Germanium segregation persisted at temperatures 60ºC below that predicted by a two-state exchange model. KMC simulations, where film growth, surface diffusion, and surface segregation are modeled consistently, successfully describe the low temperature segregation of germanium. Realistic descriptions of MBE must follow the physical rates of the growth, surface diffusion, and surface segregation processes. A specific, step mediated exchange mechanism is also considered and shown to lead to surface segregation. While this model of Ge segregation seems possible, more work is necessary to obtain a consistent set of energy barriers.


1999 ◽  
Vol 584 ◽  
Author(s):  
M. Fearn ◽  
M. Sayed ◽  
J. H. Jefferson ◽  
D. J. Robbins

AbstractWe report the development of an atomistic scale Kinetic Monte Carlo model of silicon CVD growth. By employing a variable time step algorithm, simulations have been performed over a range of time scales, enabling direct comparison with experimental data. The validity of using the kinetic theory of gases for evaluating steady state incoming particle fluxes within the model is demonstrated by comparison with computational fluid dynamics simulations. The model is applied to study hydrogen desorption rates from Si(001) and the dependence of silicon growth rate on substrate temperature, with results found to be in good agreement with experimental data. An experimentally observed decrease of growth rate with increasing H2 partial pressure is also reproduced by the model and shown to be caused by a decrease in silane adsorption on a hydrogen-rich surface.


2007 ◽  
Vol 994 ◽  
Author(s):  
Ning Kong ◽  
Taras A. Kirichenko ◽  
Gyeong S. Hwang ◽  
Foisy C. Mark ◽  
Steven G. H. Anderson ◽  
...  

AbstractWe report that arsenic diffusion can be enhanced and retarded by surrounding interstitial rich and vacancy rich environments created by Si point defect engineering implant. The enhancement and retardation can be attributed to the dominant arsenic interstitial diffusion mechanism during post-implant anneal. Kinetic Monte Carlo simulations with newly implemented models show good match with experiments. Our study suggests the importance of arsenic interstitial mechanism and a possible approach for n-type ultra shallow junction fabrication.


2009 ◽  
Vol 615-617 ◽  
pp. 73-76 ◽  
Author(s):  
Massimo Camarda ◽  
Antonino La Magna ◽  
Francesco La Via

Using joined super-lattice Kinetic Monte Carlo simulations, continuous modelling and recent experimental data on the homoepitaxial growth of 4H Silicon Carbide we study the transition between monocrystalline and polycrystalline growth in terms of misorientation cut, growth rate and temperature. We compare these optimally calibrated results both with previous continuous models and literature data. We demonstrate that this study was, indeed, necessary to correctly reformulate the phase diagram of the transition.


Sign in / Sign up

Export Citation Format

Share Document