Integrated Modeling of Semi-conductor Wafer Slicing Using a Wiresaw
Research and Development in Modern Wiresaw Manufacturing Technology
Principal Investigator: Dr. Imin Kao, Department of Mechanical Engineering
Co-PI: Dr. Fu-Pen Chiang, Dr. Vish Prasad, Department of Mechanical Engineering
Dr. Kedar Gupta, Dr. Mohan Chandra, Mr. Jon Tabott, GT Equipment
Sponsors: National Science Foundation and DoE through SBIR Phase II grant from GT Equipment
Wiresaw is an emerging technology for wafer production in semiconductor materials as well as other crystals because of its ability to cut single crystalline and polycrystalline crystals with large diameter and produce wafers from thick (a few mm) to very thin sizes (300 microns) with a small kerf loss and high yield. The wiresaw is being commonly used for the production of Si wafers in photovoltaic (PV) industry and has the potential to find widespread use in various crystals with large diameters and ceramics.
Figure 1: Industrial wiresaw
Schematic of Modern Wiresaw
A schematic of the wiresaw is shown in Figure 2. The figures shows a typical setup of the industrial wiresaw with continuous wire feed and abrasive slurry to cut specimen and substrates. In the figure, a steel wire is stretched with high tension (typ. 20 N) and pushed into a working material. A slurry with a fraction of abrasive particles, such as SiC,. is trapped in the cutting zone between the wire and the cutting material. The abrasives squeezed and sheared between the wire and the cutting material penetrate-roll and penetrate-scratch the surface to generate cracks and remove material from the workpiece -- a free abrasive machining (FAM) process.
Figure 2: Schematic of wiresaw
Modeling and Control of Wiresaw Manufacturing Process
The wiresaw cutting, however, is a poorly understood phenomenon and no model exists for simulation and design. This has reduced the design and manufacturing of wire saw cutting processes to a black art with no consistent theoretical methodology. The objectives of the ongoing research in wiresaw and wafer manufacturing are to model and simulate the underlying principles of this process and to develop control tools to achieve more accurate and efficient manufacturing. An innovative rolling-indenting model, which will give rise to a more systematic theory and simulation, of the cutting process and experimental validation of the cutting theory have been developed. Based on the theoretical development, a stiffness control scheme can be employed to improve the quality of the manufacturing processes. The model will be able to predict stress, vibration and deflection characteristics of the wire as well as the substrate. Measurements of stresses, residual stresses, warpage, and total thickness variation to validate the model are also conducted with the unique expertise at Stony Brook using the laser speckle technology, modified moire interferometry, and photoelasticity.
Figure 3: Rolling-Indenting Model of wiresaw manufacturing process
Our goal is to build US-based technology for wiresaw manufacturing, a critical technology that is dominated by Europe and Japan to this date. SEMATECH, a semiconductor R/D consortium, anticipates that the microelectronics fabrication will be based on 12'' wafers at the turn of the century. Wiresaw will be the only viable technology for slicing 12'' wafers. We are at the cross road of this critical technology and research with theoretical development and experimental validation is the only way to guarantee our success in building the needed US-based technology.
The research in modeling and control of wiresaw conducted at the Manufacturing Automation Laboratory (MAL) includes the following.
A diamond impregnated wiresaw has been purchased to conduct experiments and to provide data for validation of theoretical models. Experiments are also performed concurrently at the GT Equipment Technologies, Inc. (GTi), our industrial partner for the DoE and NSF sponsored projects, to make the research results more relevant. In collaboration with our industrial partner, we have successfully sliced single and poly crystalline silicon, alumina (aluminum oxide), and quartz with high surface quality. We have also successfully sliced 12'' silicon wafers for microelectronics fabrication. Continuing investigations and extensions in slicing other crystals such as III-V compound crystals, InP, silicon carbide, and ceramics are currently being developed. Though slicing of crystal is the first post-growth manufacturing process, we are also expanding research to wafer preparation and production.
1. Develop fundamental understanding and modeling of wiresaw cutting processes
To understand fully the interaction between the wire and the workpiece such as polycrystalline and single crystalline silicones and other materials in the wiresaw process, the effects of individual abrasives or groups of grains on the material removal rate from the workpiece surface are studied from the viewpoint of the elastoplastic indentation of single abrasive and linear elastic fracture mechanics. Photoelasticity will be used to verify the indentation analysis and simulate the cutting process.
2. Control and improve the wiresaw manufacturing processes
To develop a better understanding of the wire saw operation, an accurate prediction of vibration and deflection of wire in the cutting is necessary. from the point of view of the surface quality of wafer being cut. The vibration of wire, modeled as a constant tension string, has been studied taking account of damping effect of the slurry.
3. Establish US-based wiresaw technology
Currently there is no US-based technology in this growing and critical manufacturing process. Based on the research results, a commercial wiresaw unit, cutting large diameter ingots of hard and brittle materials will be designed and made available in the future through the industry partner GT Equipment. This will give an edge to US industry to compete in the wiresaw market and will also promote ancillaries for sensors, consumable, and other supplies.
Wiresawing is the latest technology for slicing large diameter wafers used in semiconductor as well as photovoltaic industries. In order to fine tune wiresawing to the modern day requirements of efficiency and productivity, it is necessary to understand the process. With this objective, an integrated model of wiresawing is being developed in our laboratory. Also experiments to validate the model and study the exact boundary conditions are being conducted
Follow the links below for more information.
Please direct any questions to:
Date revised - 02/03/99