Improving Efficiency and Productivity with a Data-Driven Tool Architecture
Axcelis and Cimetrix Collaboration Improves Process Control and Predictive Preventative Maintenance
By Larry Bourget, Axcelis Technologies and Dave Faulkner, Cimetrix
Since the introduction of the 300mm wafer size, data requirements for semiconductor manufacturing equipment have increased dramatically. Equipment suppliers today are required to support more than a dozen SEMI® standards related to factory automation and a host of commonly used substrate-handling components such as robots and vacuum system hardware.
Equipment suppliers now need to support a new suite of “Equipment Engineering Capabilities” (EEC) including: e-Diagnostics, data collection, recipe management, data quality, fault detection and classification, run-to-run control and predictive maintenance. The key underlying factor for most of these features is access to data. More and better quality data is needed from manufacturing equipment to support the real-time decision making that will optimize fab productivity.
Data-driven initiatives by industry organizations, such as the International SEMATECH Manufacturing Initiative’s (ISMI) 300mm next generation factory (NGF) program, also focus on the accessibility of high-quality data and the use of this data to improve efficiency and productivity. In addition, factories are also requiring data storage and access on and off the tool for future performance analysis.
Unfortunately, most existing equipment-control software was developed prior to these new requirements. Tool-control software purchased or built more than five years ago is probably already outdated.
The industry needs next-generation, data-driven tool architectures to provide access to the data demanded. These types of architectures work by feeding high-speed diagnostic and processing data to factory interfaces and to on-tool services such as databases and graphical user interfaces. When fully implemented, this approach can provide significantly improved process-control performance, resulting in higher device yields and tool reliability.
Equipment suppliers, however, face many integration challenges including compatibility between old and new technology, distributing information across multiple computers and operating systems and accessing data from a wide variety of sources. Yet, with limited software resources and a competitive marketplace forcing equipment suppliers to continuously improve their process technology, investing in creating a data-driven tool architecture can be a challenging endeavor.
Axcelis Technologies, a leader in semiconductor processing equipment, and Cimetrix, a leader in connectivity and tool-control software, tackled these challenges recently while undertaking an 18-month program to jointly develop a new tool-control software framework. During this time, Cimetrix’s CIMControlFramework™ was created and adapted based on the specific needs of Axcelis and their new Integra RS photoresist dry strip cluster tools. The focus of the project was to improve the efficiency of 300mm substrate handling and factory automation in the Integra RS, which is capable of running more than 8,000 wafers per day.
There are two primary objectives of an effective next-generation data-driven architecture. First, an adequate supply of high-frequency, high-quality data is required. Second, data must be gathered and presented to users in a way that is useful, as well as easily and quickly understandable.
To address the data-supply problem, the two companies opted to use the SEMI-specified equipment data acquisition (EDA) standards, also known as Interface A. Although issued several years ago and now being built into an increasing number of EDA-compliant tools, many equipment makers still have not revised their tool architectures to take advantage of EDA’s high-speed data collection capabilities. These standards, designed to improve process monitoring and control, provide a virtual fire hose of nearly real-time data for use by equipment and process engineers. To date, only the largest chip makers have developed programs to make use of this new data.
For Axcelis, the difference in data output between previous products and the new EDA-optimized version was impressive. Earlier software typically output around 300 data points at 3-5 Hz. Thanks to the new software, however, the Integra RS outputs more than 2,500 data points at 100 Hz with nominal speeds typically around 30-40 Hz based on the data change rate. Additional data points can easily be added without compromising this performance.
To interpret this high-speed data stream, Axcelis ported the Integra tools’ data into an Oracle® database, which provides detailed analysis of the systems’ performance and status. Cimetrix also plans to offer CIMControlFramework with MySQL and other databases. Other key software decisions included the use of a Microsoft .NET service-oriented architecture to provide the flexibility to easily adapt to changes in tool architecture and subsystem requirements over time.
Although data-analysis tools are still being enhanced, Axcelis and Cimetrix have already developed an array of applications for process control, fault detection and classification, and predictive preventative maintenance. For process control, these tools allow tracking of vital statistics and provide warnings when process conditions begin to approach the high or low end of specifications. This data helps diagnose potential problems and schedule predictive preventive tool maintenance, which is expected to result in increased tool uptime. By tracking plasma ignition times, for instance, the companies can now detect defective microwave power supplies and replace them before they actually fail.
Developing a new data-driven tool architecture is not an easy undertaking. However, as this successful collaboration demonstrates, it is achievable and bears significant benefits, enabling the real-time decision making necessary to achieve optimized productivity and efficiency. With the result of this joint development project, the ability to meet the industry’s stringent tool performance and reliability requirements of the future is available today.
Larry Bourget is director of Integra Product Development for Axcelis Technologies Inc. in Beverly, Massachusetts. Dave Faulkner is executive vice president at Cimetrix, Inc. in Salt Lake City, Utah.
July 2, 2009