Die-to-Database Pattern Monitor
As feature sizes shrink well past the limits of optical resolution, the interest and investment in high resolution imaging tools grow significantly. It is not unusual for advanced fabs to employ twenty or more SEM tools and ten or more E-Beam inspection tools. The images they generate are vital to the understanding of yield limiting factors.
Hundreds of thousands of high resolution images are generated every week – and in some cases every day. But these images are significantly underutilized. Their rich information content is typically reduced to a single class code number, and roughly 50% of the images are classified as SEM Non-Visual (SNV) and discarded.
When images are reduced to a single class code and half the images are thrown away, there is a huge loss in the potential value of both the images and the imaging tools.
D2DB-PM not only provides a solution to the problem of image underutilization, but it goes much further by taking Line Monitoring to the next level with the integration of advanced image processing and die-to-database technologies that provide new dimensions and new insights into the yield enhancement workflow. It studies every image in great detail – even if it is an SNV image – and then uploads all of the information into a central database where a historical record of that information is built. The historical record opens a wealth of possibilities for new yield enhancement applications.
Hundreds of thousands of high resolution images are generated every week – and in some cases every day. But these images are significantly underutilized. Their rich information content is typically reduced to a single class code number, and roughly 50% of the images are classified as SEM Non-Visual (SNV) and discarded.
When images are reduced to a single class code and half the images are thrown away, there is a huge loss in the potential value of both the images and the imaging tools.
D2DB-PM not only provides a solution to the problem of image underutilization, but it goes much further by taking Line Monitoring to the next level with the integration of advanced image processing and die-to-database technologies that provide new dimensions and new insights into the yield enhancement workflow. It studies every image in great detail – even if it is an SNV image – and then uploads all of the information into a central database where a historical record of that information is built. The historical record opens a wealth of possibilities for new yield enhancement applications.
Click for Larger
Thorough Analysis of Each Image
Images are thoroughly analyzed and overlaid with their reference design. All critical and consequential features — as defined by built-in rules and by the user’s own rules — are found, measured, and tracked.
Click for Larger
Printed Pattern Database
The CAD layout is a golden reference database of the intended patterns. Over the past decade and a half, use of CAD inside the fab has enabled new opportunities for yield analysis and wafer inspection. But is there an analog equivalent of the CAD layout? That is, is there a database of the printed patterns?
If a database of printed patterns were to exist, it could once again enable new opportunities for process technology development and manufacturing. We call this the Printed Pattern Database.
The printed pattern database is constructed in an intelligent manner that extracts and retains only the patterns of interest within each SEM image. Patterns of interest are identified by a set of parametric search rules that operate in real time on each image. Once extracted, each pattern of interest is assigned a class code corresponding to the rule that identified the pattern. For example, when a tip-to-line pattern is found, it is classified as a tip-to-line. When a tip-to-tip pattern is found, it is classified as a tip-to-tip. This enables the user to query and study the yield impacts of specific types of patterns along with the variations of those patterns (e.g. study the differences in printability of tip-to-line patterns as a function of the gap between tip and line).
If a database of printed patterns were to exist, it could once again enable new opportunities for process technology development and manufacturing. We call this the Printed Pattern Database.
The printed pattern database is constructed in an intelligent manner that extracts and retains only the patterns of interest within each SEM image. Patterns of interest are identified by a set of parametric search rules that operate in real time on each image. Once extracted, each pattern of interest is assigned a class code corresponding to the rule that identified the pattern. For example, when a tip-to-line pattern is found, it is classified as a tip-to-line. When a tip-to-tip pattern is found, it is classified as a tip-to-tip. This enables the user to query and study the yield impacts of specific types of patterns along with the variations of those patterns (e.g. study the differences in printability of tip-to-line patterns as a function of the gap between tip and line).
Inline Real-Time Pattern Monitor
After each inspection and SEM review cycle, D2DB-PM will analyze all of the images taken during the cycle and publish a report detailing the problematic patterns discovered. The report allows stakeholders to be notified quickly and take follow-up action.
Seamless Integration into Production
Despite the power and potential of D2DB-PM, it is surprisingly easy to integrate into the production workflow. It requires no changes to existing Plan of Record (POR) operation. It does not impact production cycle times. It requires no additional overhead. It simply plugs in and plays.
(Note that good contour extraction is dependent on good source image quality.)
(Note that good contour extraction is dependent on good source image quality.)
Zero Waste Policy — Maximize ROI of SEM
All images are fully utilized.
- The rich content of images is not reduced to single class codes.
- SEM Non-Visual (SNV) images that would otherwise be discarded are fully utilized.
- Return-on-Investment (ROI) is maximized from the Fab’s considerable investment in SEM imaging tools.
Main Benefits:
- Realize the full potential of high resolution images from SEM tools.
- Don’t throw away SEM Non-Visual (SNV) images!
- Use the rich information content from each and every image to monitor patterning quality and process drift in production fabs.
- Over time, as more and more images are analyzed, a comprehensive database of patterns is built – all based on real silicon instead of simulations.
Main Features:
- Finds all critical and consequential features with each SEM image, measures the printed fidelity of those features (by comparing contour to design), pulls out a larger pattern centered around each of those features, and tracks these patterns in a true relational database.
- Quickly see which patterns are printing well and which patterns are problematic.
- Determine if good patterns are becoming problematic or if problematic patterns are becoming good – and determine whether such change is produced by a mask or process revision, or some other factor.
- Compare the real silicon problematic patterns against the OPC simulation’s predicted weakpoints to determine the effectiveness of the OPC model.
- Generate care areas for inspection or for targeted SEM review based on the real silicon problematic patterns.
- Perform PWQ/FEM analysis by comparing the fidelity of patterns in each modulation to determine which ones are the best modulations.
Primary Application
Inline Real-Time Pattern Quality Monitor.
Additional Applications
Study patterning issues inside each FEM / PWQ modulation to provide an improved assessment of process window.
Study impact of mask and process revisions.
Compare real-silicon weak patterns with OPC weak patterns to assess OPC model’s strength and weakness.
Evaluate risk of new devices by searching their design for known real-silicon weak patterns.
Technical Requirements
The minimum system requirements are listed below:
Memory and hard drive requirements can vary substantially from customer to customer. Customers who expect to store large quantities of images on the server should allocate appropriate hard drive capacity. Customers who expect to process large numbers of images should allocate additional physical memory. Anchor Semiconductor will help each customer with the appropriate sizing of their hardware.
- Linux 2.6 or later, 64-bit, x86 based processor.
- 16 or more physical cores.
- 128 GB or more physical memory.
- 2 TB or more available hard drive capacity.
Memory and hard drive requirements can vary substantially from customer to customer. Customers who expect to store large quantities of images on the server should allocate appropriate hard drive capacity. Customers who expect to process large numbers of images should allocate additional physical memory. Anchor Semiconductor will help each customer with the appropriate sizing of their hardware.