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: September 30, 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:
The results presented for this period give much greater emphasis and details on the FEM-based analysis and supporting infrastructure for capturing geometry and gridding the structures to be analyzed, vis a vis the compact modeling aspects of the project. The key contributions are as follows:
a) the calculations of stress dependent behavior related to deposited thin films are being tested based on use of the commercial ABAQUS tool, with emphasis on understanding of the materials and process dependences.
b) an initial prototype integration of electrostatic-electromechanical simulations using the ProPHLEX tool has demonstrated the viability of this environment as a test-bed for advanced algorithmic development, including the testing of support tools needed for geometry modeling and gridding.
c) the integration of geometric modeling support, based on level-set algorithms, within the 2D gridding tool CAMINO has been demonstrated and is being applied in support of multiple tool access and further demonstration of data persistence for a realistic process flow.
d) the 3D algorithmic support for geometry modeling, again based on level-set approach, has been captured in a stand alone server prototype that will be beta-site tested by MicroCosm.
Based on industrial feedback about critical factors in determining stress in deposited layers, the emphasis has been retargeted to development of models for stress gradients in thin films deposited uniformly on bare wafers. This work will subsequently be extended to include non-uniform topographic effects in 2D and 3D. Layer-by-layer deposition and relaxation effects are being investigated using commercially available tools to determine their limitations. The "model" system being studied is tungsten where mainly growth/intrinsic stresses (as opposed to thermal stresses) are dominant. Resultant stress state can be quite different as each incrementally deposited layer is allowed to relax/equilibrate to some extent before the next incremental layer is deposited. Initial ABAQUS simulations show that the stress gradient generated from layer-by-layer simulations is smaller than would be expected from an all-at-once simulation. Recent feedback from a group in Kyoto University (Japan) suggests that Molecular Dynamics (MD) simulations of sputtering can possibly reflect the density variation in a deposited thin film on an atomic scale and thereby obtain the residual stress directly. The viability of obtaining and using this more basic level of data is also being pursued.
There are important multi-physics interactions and numerical requirements that need to be addressed in dynamic response simulation of MEMS systems. A methodology has been devised (to be completed in stages) that addresses these challenges in realizing a complete electromechanical solver. (Results for Step 1 are key achievement) Six steps have been identified:
A 3D level-set boundary movement algorithm using an oct-tree grid has been implemented and is now being tested. This oct-tree grid supports adaptive refinement and derefinement. Adaptation criteria can be distance, curvature, and externally supplied functions. The code can be used for 3D etching and deposition simulation on the level-set grid. The MircoCosm group is actively interested and has agreed to be a test site, in spite of the very early stage of this prototype version. A key example that they are interested in is KOH etching where the etch rates depend on crystal orientation. Another important application is APCVD oxide deposition where the deposition rates have surface curvature dependence.
In terms of user support issues, the code is connected to the Centric Visualizer. Contour plots on field values on oct-tree can be directly rendered. Parts of the code are being incorporated in CAMINO so that it can perform 3D gridding and level-set boundary movement on given geometries.
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:
A two mask process flow for micromachining fatigue test structures has been developed. The test structures consist of 50 micron by 600 micron metal beams suspended over a backside etched gap in the silicon substrate. Two different versions with either a single beam or a six beam array have been designed. The necessary masks have been fabricated and the process has been tested and optimized. The first set of 2 micron thick aluminum fatigue test structures have been fabricated and are ready to be tested. We have taken note of some small drawbacks in the first generation masks and will alleviate those in a second generation mask set to facilitate backside alignment.
The available piezo-driven testing equipment has been evaluated for feasibility of performing fatigue tests. The necessary modifications and additions for fatigue testing have been designed and are being machined.
All our efforts have been concentrated on designing a process flow for micromachining fatigue test samples, thus the actuated membrane concept for stress measurement during film deposition has temporarily been put on hold until the first quarter of '98.
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:
To date, we are in the midst of fabricating the new MEMS test structures and example devices. Several devices (i.e. switch and membrane devices of various geometries) were included on the layout for testing and comparison modeled performance and actual MEMS devices. Behavior of these devices can also be directly correlated to the test structures as a further gauge of materials properties.
During the next quarter we will test, measure and evaluate the completed material properties test structures. Several options for dynamic measurement of the performance of the test structures are under consideration. Current methods anticipate electrostatically actuation while optically measuring the deflection of the structure. This method can be employed to measure both relative deflection as well as resonance. Two possible systems available for this dynamic measurement are now being considered. One method is a pair of optical receivers whose outputs are differentially connected. In conjunction with a laser source, this method can be used to measure the deflection of a surface. The other method is an interferometric system designed to measure deflection by measuring the interference pattern of a laser source. These resources are available at Stanford but currently not tailored to our needs. With working devices, a systematic method can be developed to characterize their dynamic performance. In-situ test measurement capabilities are still being considered and test structures have been included on the test chips.