Contractor: Stanford University
Agreement Number: F30602-96-2-0308
Project Number: HJ1500-3221-0599
DARPA Order Number: N-7-0504/AO E117/12
Contract Period: September 13, 1996 - September 12, 1999
Date: July 22, 1997
Sponsor: Department of the Air Force
Technical POC: Robert W. Dutton
Report Prepared by: Robert W. Dutton
a) enhanced gridding capabilities compatible with VLSI TCAD.
b) object-oriented design of layered access to model definition that
allows fast prototyping of new and realistic physical and constitutive
models.
c) enhanced element technology for discretization of
multi-physics system of equations.
d) fully-integrated MEMS fabrication and device simulation with
detailed material and stress analyses monitored by the test MEMS
structure discussed below.
e) development of lumped models for specific MEMS applications of
interest to DARPA.
PROGRESS:
Physical effects of the nonlinear dynamics for MEMS microwave (RF) switches have been achieved using a lumped model representation that includes: electrostatics, bending, stretching, residual stress, inertia, squeeze film damping and several surface effects. In addition to achieving a unified formulation that is successfully applied to analysis of transient characteristics and geometrical scaling, this model provides key guidance in directing our physical modeling efforts towards the areas of residual stress analysis in MEMS and in determining the underlying causes that come from process dependences--especially those of deposition and subsequent thermal processing. Two different test devices are being used to evaluate the modeling approach:
a) structures have been fabricated in the MUMPS process and the effects for polysilicon switches will be emphasized,
b) switches that use the MAFET process will also be considered in the context of aluminum materials. In the area of supporting tool development for FEM-based MEMS analysis, a new set of tools and formulations for the topographic changes (using the level-set algorithm) and analysis of the viscoelastic behavior and creep (drawing heavily on Stanford expertise in computational mechanics) are now being strongly leveraged in this project.
In looking towards the specific goals on the tools side, details of the computation and modeling of stress behavior will be emphasized. Currently, the simulation of residual stress evolution even assuming some initial stress state is still in its infancy. The initial goal is to get a good stress relaxation model for polysilicon and PSG (and possibly aluminum and tungsten). The challenge is to create some test structures that will demonstrate stress gradients, edge effects from complex patterning, and curvature dependence that actually change with different annealing recipes. Evolution of microstructure with anneal complicates the constitutive model of polysilicon. As a first step we will identify (in collaboration with our experimental partners) a process that can create stable materials and then simulate stress evolution.
Most of the MEMS structures of interest, including the RF switch, are geometrically complicated, electromechanically coupled, and inherently three-dimensional. It is critical for accurate and efficient prototyping of MEMS to account for the dynamic response of the system. With the objective of creating a unified representation to test gridding, geometrical modeling support, and dynamic behavior modeling, two paths are being taken simultaneously to develop a computational prototype of the RF switch. First, the commercial simulation based design system MEMCAD (Microcosm) is being used to investigate the effectiveness of elastostatic and electrostatic modeling. Second, using various commercial tools including ProPHLEX (COMCO), Spectrum (Centric), AML (TechnoSoft), and HyperMesh (AltAir Computing), the ability to do dynamic modeling, use hp-adaptive finite element methods, include complex material models, and integrated automatic mesh generation is being investigated.
Based on very strong positive feedback from other groups in the Composite CAD community, the adaptive gridding capabilities demonstrated based on use of the level-set method for simulation of thin-film fabrication are being pushed aggressively to create a prototype for distribution to other interested users. Important applications of the new gridding capabilities directly support our own modeling of residual stress effects during growth. The key challenge now under investigation and active development is to create a flexible application programming interface (API) that can consistently treat different algorithmic needs for geometry and gridding through a common, unified representation.
To provide a consistent API for static/moving geometry and structured mesh, the current undertaking is to combine the functionalities of oct/quad-tree based meshing (CAMINO) and level-set boundary movement.
Using the same oct/quad-tree internal grid as a reference grid, tetrahedralization can be performed for structured mesh, or level-set boundary evolution equation can be solved for boundary movement. This is an important step towards providing consistent geometry/field service for client solvers such as ALAMODE.
Develop extraction schemes for materials properties using MEMS test structures, in combination with the CAD tools developed above, to support predictive modeling. The scope of material characterization includes the use of multilayer and graded structures, and reliability issues such as plastic yield and fatigue. This closed-loop process of test structures, simulation and extraction/validation of models is the key link between groups in CAD/Material Science/ MEMS developers.
APPROACH:
a) mechanical characterization of MEMS materials with dependencies on
processing and microstructure.
b) reliability study including fracture and fatigue.
c) feedback to the constitutive models in the FEM-based solver in Task 1.
PROGRESS:
We have developed and tested a new analysis technique for extraction of unstrained lattice parameters of crystalline thin films. Applications of this technique are extraction of residual stress states of poorly characterized blanket films, and determination of the coefficient of thermal expansion of thin films without their removal from the substrate. Our technique does not require knowledge of the elastic parameters of the thin film material, and does not require any micromachining.
Two types of bulk micromachined uniaxial fatigue test structures have been designed, and a mask set is being fabricated. Uniaxial fatigue structures were chosen over torsional fatigue structures for ease of fabrication and actuation. A high frequency piezoelectric device will be used for actuation. External actuation has two advantages over built-in actuation: large loads, and the ability to select strain, load, and frequency independently.
Design and objective implementation of MEMS structures for in-situ materials parameter extraction (mentioned above), which will be applicable to MEMS devices of interest to DARPA. The experimental measurements made with these structures will facilitate the critical evaluation of models, parameters and overall simulation accuracy for MEMS devices. The geometry and process flow design will support not only the application side (such as micromachined RF switches) but also the underlying need to quantify and understand MEMS device limits resulting from materials and process dependences.
APPROACH:
a) design and implementation of in-situ test structures. b) design and implementation of example prototype MEMS device. c) measurement, parameter extraction, modeling and testing.
PROGRESS:
Using our own insight and information from literature sources, the basic designs for the test structures were created. The details of how to best realize these structures have been considered and the mask layout completed. There has been a thorough review from project members in the modeling, material science and fabrication segments. Each segment verified that the designs are both practical and that the designs address their needs. The review was also an opportunity to consider non-idealities that may effect the overall accuracy of the results.
The structures on the mask set include a variety of devices for determination of Young's Modulus and film stresses. The bilayer structures will allow us to measure the influence of multiple layers in a controlled manner. Failure analysis structures as well as in-situ structures for testing process influences have been included with the overall design. The different structures that have been selected will allow a comparison of the nominal stress measured between structures. We have included designs that take advantage of the enhanced z-axis (out of plane) resolution in anticipation of the acquisition of an interferometer system.
Fabrication issues also require careful consideration. Once the layers are drawn it is necessary to have an appropriate process flow to create the structures. Electrostatically driven elements require a conductive underlayer for actuation. The sacrificial material must be planar in order to guarantee the flatness of the test structures. It was decided to use a support material over the edges of the thin film structures to accurately determine both the dimensions and boundary conditions. Since this support material is intended to be much thicker than the thin film of interest, it is critical that the etch chemistry of the support be highly selective. Dry etching has been selected as the preferred release mechanism to avoid stiction problems associated with wet release.
Task 1: (Dutton group)
a) Assessment of target material system (poly, PSG, aluminum) for detailed
stress analysis.
b) Demonstration of FEM-based modeling of RF switch based on commercial tools.
c) Definition of architecture for merged geometry/grid API.
d) Initial demonstration of CAMINO (gridder) in server-based architecture.
Task 2: (Bravman group)
a) Complete fabrication of fatigue test structures.
b) Modify test equipment for low-cycle fatigue test requirements.
c) Conduct proof-of-concept fatigue tests.
d) Evaluate actuated membrane concept for stress measurement during
film deposition.
Task 3: (Kovacs group)
a) Fabricate test structures
b) Establish final plan for dynamic testing
c) Extract material properties from test structures
d) Re-evaluate structures for improvements