SPRINTCAD QUARTERLY SUMMARY
July 1 - Sept. 30, 1996
ORGANIZATION:
Stanford University
SUBCONTRACTORS:
none
PRINCIPAL INVESTIGATORS:
Robert W. Dutton, dutton@gloworm.Stanford.EDU, (650) 723-4138
Kincho H. Law, law@cive.Stanford.EDU, (650) 725-3154
Krishna Saraswat, saraswat@sierra.Stanford.EDU, (650) 725-3610
Peter Pinsky, pinsky@ce.Stanford.EDU (650) 723-9327
TITLE OF EFFORT:
"SPRINT-CAD"---Industry-Networked TCAD using Shared Parallel Computers
RELATED INFORMATION:
The URL for Stanford TCAD projects is: http://www-tcad.stanford.edu
The URL for the Sprint-CAD projects is:
http://www-tcad.stanford.edu/sprintcad/
OBJECTIVE:
First-time capabilities to bridge solid modeling, FEM-based
parallel computation of fabrication processes and electrical analysis
of the resulting IC structures will be developed. Models needed
to represent diffusion, etching, deposition, oxidation and stress
analysis resulting from a sequence of process steps necessary in
the creation of electrical devices will be developed. This effort
will provide a radically new HPC framework for technology-based 3D
process/device modeling as well as realistic benchmarks to test HPC
architectures and software.
APPROACH:
We will build, integrate and test TCAD modules based on
an object-oriented approach that both develops and uses information
models in support of CFI-based standards. The modules and software
engineering methodology will be designed specifically to exploit
parallel computers and library components. The 3D process simulation
modules will utilize HPC platforms and provide new functional
capabilities for "computational prototyping" of the following key
technology fabrication steps: 1) deposition/etching module---of special
interest are CVD and plasma assisted processes that result in high
aspect ratio structures such as trenches and filling/planarization
of structures for metal interconnects. Algorithmic work focuses on
geometric manipulations and surface evolution. 2) thermal/stress
analysis module---that can solve nonlinear constitutive models for
key process steps involving growth of dielectric layers and impurity
redistribution as well as the resulting stress fields. Advanced
formulations for finite elements are being developed that support:
parallel computation, adaptive gridding and domain decomposition.
PROGRESS:
The reduced SWR 0.3 procedural interface (PI) had been designed
to support multiple 3D gridders. The first application is to wrap
existing 3D gridding tools developed at Stanford to support
the geometry/field services of two different PDE solvers (ALAMODE
from Stanford and PROPHET from Lucent), and hence to evaluate
the complexity and performance of an open environment (plug-and-play)
for physical definition and numerical discretization.
The 3D oct-tree based gridder, CAMINO developed at Stanford,
has been used to implement the field services in SWR 0.3 and
connected to ALAMODE and PROPHET. 3D diffusion with adaptive
gridding according to the impurity gradients have been
demonstrated in both solvers. Domain decomposition using the
spectral method has been prototyped for parallelization.
In the dial-an-operator regime, since the stiffness of the
coupled system of equations is determined by the user, numerical
stability of the simulation demands more robust time stepping and
nonlinear iterative schemes. Although it is unlikely to have
a universal solution for all types of PDEs, the reactive-diffusive
systems can be reasonably treated with TR-BDF2 time discretization
and nonlinear convergence validation heuristics. The algorithm
has been tested using a transient-enhanced diffusion (TED) model
with very stiff system of equations from perturbation very close
to equilibrium and has shown good stability behaviors.
High density plasma (HDP) processing has many advantages for
anisotropic dry etching and air-bridge dielectric deposition
on high aspect ratio structures. However, its modeling
involves many mechanisms that cause both etching and deposition
depending on the ion distribution, sputtering probability
and source viewing angles. For boundary movement schemes, it
is very hard to employ rule-based algorithms on discontinuously
bent edges for degenerate conditions, since neighboring segments may
flip sign in boundary velocity. The level-set method can be
easily extended for modeling HDP systems, since simultaneous
etching/deposition can be explicitly accounted for in the
Hamilton-Jacobi equation of the level set function.
Introduction of the Eulerian representation of the interface
(such as the level-set method) will require either frequent
transformation between boundary types or posing interface conditions
such as segregation directly on the Eulerian grid. This is
critical for robust and efficient oxidation simulation since
the diffusion equation is solved at every time step. However,
although element base functions can be extended to high orders
to improve accuracy, the conventional finite-element method can
hardly follow discontinuous fields within the element without
employing expensive shock-tracking methods. Also, the physical
origin of segregation arises from variation of the chemical affinity
in different materials, and is not similar to the shock formation
mechanism in hydrodynamics which requires the source velocity
equal to the front velocity. Finite-element base functions employing
step functions with delta functions as their derivatives have
been developed. Together with the level-set boundary representation,
the interface conformity constraint on volume grids is eliminated
for modeling segregation effects.
RECENT ACCOMPLISHMENTS:
- The 3D oct-tree based gridder (CAMINO) has been connected with
both process simulation tools PROPHET and ALAMODE using the reduced
SWR 0.3 interface.
- Time stepping and nonlinear iterative schemes in ALAMODE have been
improved with inclusion of TR-BDF2 discretization and nonlinear
convergence validation heuristics.
- The level-set boundary movement algorithm has been extended to
simulate high density plasma (HDP) etching. Specific features
such as simultaneous deposition/etching have been well captured.
- Finite-element base functions for discontinuous fields have
been developed. Together with the level-set boundary representation,
the interface conformity constraint on volume grids is eliminated
for modeling segregation effects.
FY-'97 PLANS:
- Parallelization of ALAMODE and the geometry/field server functions
through the static domain decomposition techniques.
- 3D conformal gridding tools to provide 3D elements that satisfy the maximum
principle in various mass lumping schemes for the diffusive-reactive systems.
- Validation of the 1-2-3D geometry/field server using fully adaptive schemes
for semi-discrete finite-element based TCAD tools.
- Benchmark the overall SPRINT-CAD modules in terms of a deep
submicron MOS technology. The thermal module will demonstrate 3D
analysis of locally oxidized isolation and shallow junction diffusion
steps. The deposition/etching module will demonstrate first-time
functionality of a 3D server-based architecture and implementation.
TECHNOLOGY TRANSITION:
ALAMODE (A LAyered MOdeling Environment) has been released to
the SRC/National Labs CRADA projects, and other related DARPA/SRC
efforts on transient-enhanced diffusion modeling. A suite of
operators, element technology and model examples is available
together with a couple of field interface specifications.
A 60-page user manual is also available in both postscript and html
formats.
Edwin C. Kan
kan@gloworm.stanford.edu
CIS-X 334, Stanford University, Stanford, CA 94305
Office: (415) 723-9796
Fax: (415) 725-7731
Date prepared: 1/30/97