Optics: measuring and testing – Inspection of flaws or impurities – Surface condition
Reexamination Certificate
2000-05-04
2003-07-08
Stafira, Michael P. (Department: 2877)
Optics: measuring and testing
Inspection of flaws or impurities
Surface condition
C356S237400
Reexamination Certificate
active
06590645
ABSTRACT:
BACKGROUND OF THE INVENTION
This invention relates in general to systems for analyzing anomalies of sample surfaces, such as those of semiconductors and those of magnetic or optical disks or flat panel displays, and in particular, to a system for detecting and classifying anomalies of such surfaces.
Achieving the best possible financial performance drives the continuous shrinkage of the critical dimensions of integrated circuit devices that are fabricated on surfaces of semiconductor wafers. This shrinkage requires a flatter wafer surface due to limited lithographic depth of focus (DOF) and higher circuit packing density, along with many other factors. Chemical mechanical planarization or polishing (CMP) has become an enabling technology to fulfill these requirements for the semiconductor industry. Since critical defects scale with the design rule, defect detection tools are required to have better sensitivity; the required sensitivity is roughly half the size of the critical dimension. Unlike many other semiconductor processes, CMP is unique in its requirement that the slurry be both chemically active and mechanically abrasive during polishing. The combination of the chemical reaction and abrasive behavior of the CMP process creates a unique set of defects, such as microscratches, chatter marks, slurry residue, etc. The adoption of CMP presents a great opportunity to have a better DOF budget, but poses a unique challenge in the inspection of processed CMP wafers, along with achieving higher chip yield by using appropriate yield management techniques.
Since CMP has become an essential practice in almost all chip manufacturing with a design rule of 0.25 &mgr;m or smaller, there is an urgent need to have a set of metrology tools that can not only detect but also classify those defects. In general, different types of defects have different sources and different impacts on the final device yield. Classifying them in real time will significantly reduce the time-to-results. The shrinkage of the dimensions of devices leads to an increase of the density of chips that can be produced on a wafer; therefore, more value has been produced from each wafer. This places ever-greater demands on yield management and in particular on defect inspection and classification during the CMP process; the loss of a single production wafer leads to a significant revenue loss. Driven by having shorter time for achieving and maintaining high-yield for high-value added products, the ideal defect inspection system should be able to deliver the necessary information to have a more comprehensive solution, which includes detecting all types of defects, classifying them, analyzing them, and recommending corrective actions.
Unlike particles, some of the CMP-induced defects, such as microscratches and chatter marks, cannot be removed by post-CMP cleaning and it is important to sort them out and minimize their occurrence since they may impact yield. For the cleanable defects such as particles, which may have no significant impact on the final yield, the classified defect counts will be used for process control.
Due to the nature of the process, CMP microscratches are very difficult, if not impossible, to avoid. The larger particles in the slurry or fall-on particles from the pad conditioner mainly induce this type of defect. Depending upon their dimensions and locations, these defects may adversely affect the yield of the device. Needless to say, detecting and classifying these defects will be essential for the desired process monitoring and control for achieving the best possible yield of a production line.
Typical CMP-induced defects can be divided into two categories: extrinsic defects, which are the result of the presence of foreign materials, and intrinsic defects, which are the imperfections created on the polished material. Extrinsic defects include slurry residue, surface particles, and embedded particles. Intrinsic defects include such defects as microscratches, chatter marks, water marks, long scratches (continuous and spiral), pits, rip-out and dishing. Due to the complexity of the patterned wafers, additional types of defects are present. These defects are related to the manufacturing process.
As the name implies, residual slurry results from incomplete or improper cleaning after the CMP step. Surface particles could be picked up from anywhere and are not necessarily CMP related. Embedded particles may result from existing surface particles, abraded film particles, flakes or particles from the slurry that are ground in by the down force of the polishing head.
The relative motion of the wafer and the pad produces a microscratch, when larger particles are present in the slurry. In particular, slurry particles larger than 1 &mgr;m in size can be significant contributors to microscratch generation. Other factors that contribute to microscratch generation include unsuitable rinsing and buffing steps following the CMP step. Although buffing with soft pads over hard pads reduces the formation of microscratches, incorrect buffing produces worse results. The dilute-HF dip step, which follows the buffing polish step, can increase the number of micro-scratches, enlarges existing microscratches, and leads to failure of the device.
The Surfscan SP
1
TBI
® wafer inspection system from KLA-Tencor Corporation, the assignee of the present application, has been used for inspecting unpatterned semiconductor wafers. This system has very high sensitivity and can classify large scratches, such as scratches longer than 1500 nm. However, many scratches produced during the CMP processors are smaller than that and may even be smaller than the spot size of the laser beam used in the system. Scratches that are shorter than a predetermined length such as 1500 nm are referred to herein as “microscratches.” Most of the microscratches may be smaller than the spot size of the laser beam used in the Surfscan SPI
TBI
® system, so that they appear as light point defects (LPDs). Therefore, without using a classification method, the user would not be able to distinguish between microscratches from particles or other types of point anomalies such as pits on the wafer surface.
In order to distinguish such microscratches from other point defects such as particles and pits, tools such as scanning electron microscopes or atomic force microscopes have been used. When there are a large number of defects on the wafer surface, such method is time consuming and not practical for use in a production environment. It is, therefore, desirable to provide an improved system for classifying anomalies of a surface in which the above-described difficulties are not present.
SUMMARY OF THE INVENTION
This invention is based on the observation that, by varying the sensitivity by which the anomalies are detected to provide output(s) or by varying a threshold when data on the anomalies are analyzed, more information useful for classifying the anomalies becomes available. By using outputs obtained at two or more different detection sensitivities, or by processing the data on the anomalies using two or more different thresholds, it is possible to distinguish between microscratches and particles, pits or other point defects. At least one classification of the anomalies may then be provided. Preferably, this can be done without requiring a user to scan the sample surface more than one time. This will permit a user to distinguish between microscratches from particles and other point defects and adjust the CMP or cleaning process in real time in an on-line production process or post-processing as required to improve yield.
REFERENCES:
patent: 5699447 (1997-12-01), Alumot et al.
patent: 5864394 (1999-01-01), Jordan, III et al.
patent: 5907628 (1999-05-01), Yolles et al.
patent: 5991699 (1999-11-01), Kulkarni et al.
U.S. patent application Ser. No. 08/933,771, filed Sep. 19, 1997.
“Acoustic Scanners and Modulators,” M. Gottlieb,Optical Scanning, ed. by Gerald F. Marshall, Dekker 1991, pp. 615-685.
“CMP Applications for Sub-0.25&mgr;m Process Technologies,” D. Pramanik e
Akbulut Mustafa
Chen Wayne
Zeng Andrew
Hsue James S.
KLA-Tencor Corporation
Skjerven Morrill LLP
Stafira Michael P.
LandOfFree
System and methods for classifying anomalies of sample surfaces does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and methods for classifying anomalies of sample surfaces, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and methods for classifying anomalies of sample surfaces will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3065164