A method of enhancing detection of special distribution defects

Abstract

A method for enhancing special distribution defect detection, comprising: performing defect scanning on a defect scanning machine to generate a wafer defect distribution map; performing graph algorithm processing on the defect distribution map to determine whether the defect distribution map has a special distribution defect map; for a wafer without the special distribution defect map, performing inspection and judgment according to a first rule according to OOC / OOS values set by a defect control specification of a corresponding process; for a wafer with the special distribution defect map, performing graph algorithm processing, extracting and classifying the special distribution defect map, matching different OOC / OOS values, and performing inspection and judgment according to a second rule. The present application further increases the definition of the special distribution defect map and matches different OOC / OOS values to perform inspection and judgment on the wafer with the special distribution defect map. The present application not only has strong timeliness and can discover the special distribution defect map in time, but also has high sampling pertinence, enhances sampling inspection of the special distribution defect group, and increases sample quantity.

Description

Technical Field

[0001] The present invention relates to the field of semiconductor manufacturing technology, and in particular to a method for enhancing special distribution defect detection. Background Art

[0002] In semiconductor chip manufacturing, establishing a rational and efficient defect monitoring system enables real-time monitoring of wafer defects caused by process or tool anomalies, and promptly identifies the causes of defects. Such a defect monitoring system plays a vital role in reducing defects and improving yield.

[0003] Currently, the conventional defect monitoring process is as follows: first, a defect scan is performed on the defect scanning machine to generate a wafer defect distribution map (DefectMap). Based on the out-of-control (OOC) and out-of-specification (OOS) values ​​set by the defect control specifications of the corresponding process, it is determined whether to trigger an alarm. Wafers that have not triggered an alarm will directly enter the next stage, while wafers that have triggered an alarm will undergo defect electron microscope image sampling inspection to confirm the defect image.

[0004] See also Figure 5 , Figure 5 The defect number trend chart of the existing defect monitoring process is shown. Obviously, the defect monitoring process has the following disadvantages: (1) poor timeliness: for wafers with special distribution defects, the system may automatically release them because the number of defects does not trigger OOC / OOS, resulting in the failure to perform defect image detection in the first place, such as Figure 5 The edge cluster defect shown is 5. (2) Lack of targeted sampling: When a wafer with a slight special defect pattern triggers OOC due to a high total defect count, the relatively small sample size during sampling affects defect judgment, such as Figure 5 The central cluster defect 6 is shown.

[0005] Seeking a wafer defect monitoring method that is highly timely, targeted, and does not miss minor special defect patterns has become one of the technical problems that technicians in this field urgently need to solve.

[0006] Therefore, in response to the problems existing in the existing technology, the designers of this case, relying on their many years of experience in this industry, actively researched and improved it, and thus came up with the present invention, a method for enhancing the detection of special distribution defects. Summary of the Invention

[0007] The present invention aims to provide a method for enhancing the detection of special distribution defects in view of the defects in the conventional defect monitoring process in the prior art, such as poor timeliness and lack of targeted sampling.

[0008] To achieve the object of the present application, the present application provides an enhanced special distribution defect detection method, which comprises the following steps:

[0009] Step S1 is performed: performing defect scanning on a defect scanning machine to generate a wafer defect distribution map;

[0010] Step S2 is performed: performing graph algorithm processing on the wafer defect distribution map to determine whether the wafer defect distribution map has a special distribution defect map;

[0011] Step S3 is performed: for the wafers without the special distribution defect map, performing random sample inspection judgment according to the first rule based on the exceeding control value and the exceeding specification value set by the defect control specification of the corresponding process; for the wafers with the special distribution defect map, performing graph algorithm processing, extracting and classifying the special distribution defect map, and then matching different exceeding control values and exceeding specification values for various types of special distribution defect maps, and performing sample inspection judgment of the special distribution defect map according to the second rule.

[0012] Optionally, the enhanced special distribution defect detection method further comprises the following steps:

[0013] Step S4 is performed: when the judgment is performed according to the first rule, if the wafers without the special distribution defect map do not trigger an alarm, the wafers without the alarm are released in batches; if the wafers without the special distribution defect map trigger an alarm, a failure action plan is executed to perform random sample inspection judgment on the wafers without the special distribution defect map;

[0014] Step S5 is performed: when the judgment is performed according to the second rule, if the wafers with the special distribution defect map do not trigger an alarm, the wafers without the alarm are released in batches; if the wafers with the special distribution defect map trigger an alarm, a failure action plan is executed to perform sample inspection on the wafers with the special distribution defect map.

[0015] Optionally, the wafers without the alarm, i.e., the wafers with the defect distribution map or the wafers with the special distribution defect map, are less than the exceeding control value and the exceeding specification value set by the defect control specification of the corresponding process.

[0016] Optionally, the wafers with the alarm, i.e., the wafers with the defect distribution map or the wafers with the special distribution defect map, are greater than or equal to the exceeding control value and the exceeding specification value set by the defect control specification of the corresponding process.

[0017] Optionally, the sample inspection judgment performed by the enhanced special distribution defect detection method is sample electron microscope image sampling detection to confirm the defect image.

[0018] Optionally, the special distribution defect map is a cluster type of type one.

[0019] Optionally, the special distribution defect diagram is a line type of type 2.

[0020] Optionally, the special distribution defect diagram is an arc type of type three.

[0021] Optionally, the special distribution defect diagram is a fan-shaped type four.

[0022] In summary, the method for increasing special distribution defect detection described in the present invention further adds the definition of special distribution defect maps on the basis of conventional sampling detection, and matches different OOC / OOS values ​​to inspect and judge wafers with special distribution defect maps. It is not only highly timely and can promptly discover special distribution defect maps and determine whether sampling inspection is needed according to the defect distribution type, but also has highly targeted sampling, enhances sampling inspection of special distribution defect groups, and increases sample size. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 FIG2 is a flow chart of a method for enhancing special distribution defect detection according to the present invention;

[0024] Figure 2 FIG. 1 is a diagram showing the mechanism principle of the method for enhancing the detection of special distribution defects according to the present invention;

[0025] Figure 3(a) to Figure 3(d) Shown are various special distribution defect images extracted by the method for enhancing special distribution defect detection according to the present invention;

[0026] Figure 4(a) to Figure 4(d) The above is a quantity trend diagram of various special distribution defect diagrams;

[0027] Figure 5 The figure shows the defect quantity trend of the existing defect monitoring process. DETAILED DESCRIPTION

[0028] In order to explain the technical content, structural features, achieved objectives and effects of the present invention in detail, the following embodiments will be described in detail with reference to the accompanying drawings.

[0029] See also Figure 1 、 Figure 2 , Figure 1 FIG. 4 is a flow chart of a method for enhancing special distribution defect detection according to the present invention. Figure 2 The diagram shows the mechanism principle of the method for enhancing the detection of special distribution defects of the present invention. The method for enhancing the detection of special distribution defects includes:

[0030] Executing step S1: performing defect scanning on a defect scanning machine to generate a wafer defect distribution map;

[0031] performing step S2: performing a graphic algorithm on the wafer defect distribution map to determine whether the wafer defect distribution map has a special distribution defect map;

[0032] performing step S3: for the wafer without the special distribution defect map, performing a random sample inspection judgment according to the out of control value (OOC, Out Of Control) and the out of specification value (OOS, Out Of Specification) set by the defect control specification of the corresponding process according to the first rule; for the wafer with the special distribution defect map, performing a graphic algorithm, extracting and classifying the special distribution defect map, and then matching different out of control values (OOC, Out Of Control) and out of specification values (OOS, Out Of Specification) for various types of special distribution defect maps, and performing a sample inspection judgment of the special distribution defect map according to the second rule.

[0033] More specifically, the method for enhancing special distribution defect detection further comprises:

[0034] performing step S4: when performing the judgment according to the first rule, if the wafer without the special distribution defect map does not trigger an alarm, the wafer without the special distribution defect map is released in batches; if the wafer without the special distribution defect map triggers an alarm, a loss action plan is executed, and a random sample inspection judgment is performed on the wafer without the special distribution defect map;

[0035] performing step S5: when performing the judgment according to the second rule, if the wafer with the special distribution defect map does not trigger an alarm, the wafer without the special distribution defect map is released in batches; if the wafer with the special distribution defect map triggers an alarm, a loss action plan is executed, and a sample inspection judgment is performed on the wafer with the special distribution defect map.

[0036] As a person skilled in the art, it is easily understood that in step S4, step S5, the number of wafers with a defect distribution map or with a special distribution defect map that does not trigger an alarm is less than the out of control value (OOC, Out Of Control) and the out of specification value (OOS, Out Of Specification) set by the defect control specification of the corresponding process. Similarly, the number of wafers with a defect distribution map or with a special distribution defect map that triggers an alarm is greater than or equal to the out of control value (OOC, Out Of Control) and the out of specification value (OOS, Out Of Specification) set by the defect control specification of the corresponding process. The method for enhancing the detection of special distribution defects of the present application performs sample inspection judgment, i.e., sample electron microscope image sampling detection, to confirm the defect image.

[0037] Obviously, the method for enhancing the detection of special distribution defects of the present application does not execute steps S1-S5 in strict order, and within the scope of understanding of a person skilled in the art, the order of some steps can be changed to achieve the purpose of the present application.

[0038] In order to more directly disclose the technical solutions of the present application and highlight the beneficial effects of the present application, the specific steps and detection processes of the method for enhancing the detection of special distribution defects will be described in conjunction with specific embodiments. In the specific embodiments, the special distribution defect map, the out of control value, and the out of specification value are only examples and should not be considered as limiting the technical solutions of the present application.

[0039] Please refer to Figure 3(a) to Figure 3(d) , Figure 4(a) to Figure 4(d) , and in conjunction with Figure 1 , Figure 2 , Figure 3(a) to Figure 3(d) The various special distribution defect maps extracted by the method for enhancing the detection of special distribution defects of the present application are shown. Figure 4(a) to Figure 4(d) The number trend graphs of the various special distribution defect maps are shown. The method for enhancing the detection of special distribution defects includes:

[0040] Step S1 is performed: defect scanning is performed on a defect scanning machine to generate a wafer defect distribution map;

[0041] Step S2 is performed: the wafer defect distribution map is processed by a graph algorithm to determine whether the wafer defect distribution map has a special distribution defect map;

[0042] performing step S3: for the wafers without special distribution defect map, performing random sample inspection judgment according to the out of control (OOC) and out of specification (OOS) values set by the defect control specification of the corresponding process according to the first rule; for the wafers with special distribution defect map, performing graph algorithm processing, extracting and classifying the special distribution defect map, and then matching different out of control (OOC) and out of specification (OOS) values for various types of special distribution defect map, and performing sample inspection judgment of the special distribution defect map according to the second rule.

[0043] Non-limitingly, for example as shown in FIG. 3(a), for the wafers with special distribution defect map, performing graph algorithm processing, extracting and classifying the special distribution defect map, the special distribution defect map is cluster type 1 of type one, and the first OOC / OOS value is set.

[0044] For example, as shown in FIG. 3(b), for the wafers with special distribution defect map, performing graph algorithm processing, extracting and classifying the special distribution defect map, the special distribution defect map is line type 2 of type two, and the second OOC / OOS value is set.

[0045] For example, as shown in FIG. 3(c), for the wafers with special distribution defect map, performing graph algorithm processing, extracting and classifying the special distribution defect map, the special distribution defect map is arc type 3 of type three, and the third OOC / OOS value is set.

[0046] For example, as shown in FIG. 3(d), for the wafers with special distribution defect map, performing graph algorithm processing, extracting and classifying the special distribution defect map, the special distribution defect map is fan type 4 of type four, and the fourth OOC / OOS value is set.

[0047] More specifically, the method for enhancing special distribution defect detection further comprises:

[0048] performing step S4: when the wafers without special distribution defect map do not trigger an alarm, the wafers without special distribution defect map are released in batches according to the first rule; if the wafers without special distribution defect map trigger an alarm, a loss action plan is executed, and the wafers without special distribution defect map are subjected to random sample inspection judgment;

[0049] Execute step S5: When judging according to the second rule, if the wafer with the special distribution defect pattern does not trigger an alarm, the wafer that does not trigger an alarm is batch released; if the wafer with the special distribution defect pattern triggers an alarm, the out-of-control action plan is executed, and sample inspection and judgment are performed on the wafer with the special distribution defect pattern.

[0050] As a specific embodiment, the wafer with a special distribution defect map does not trigger an alarm, that is, the number of wafers with type 1 cluster type 1 defects is less than a first OOC / OOS value; the number of wafers with type 2 line type 2 defects is less than a second OOC / OOS value; the number of wafers with type 3 arc type 3 defects is less than a third OOC / OOS value; and the number of wafers with type 4 fan type 4 defects is less than a fourth OOC / OOS value. The wafer with a special distribution defect map triggers an alarm, that is, the number of wafers with type 1 cluster type 1 defects is greater than or equal to a first OOC / OOS value; the number of wafers with type 2 line type 2 defects is greater than or equal to a second OOC / OOS value; the number of wafers with type 3 arc type 3 defects is greater than or equal to a third OOC / OOS value; and the number of wafers with type 4 fan type 4 defects is greater than or equal to a fourth OOC / OOS value.

[0051] When judging according to the second rule, if the wafer with a special distribution defect pattern triggers an alarm, the out-of-control action plan is executed, and a sample inspection is performed on the wafer with a special distribution defect pattern, that is, a sample electron microscope image sampling test is performed to confirm the defect image.

[0052] Obviously, the method of increasing special distribution defect detection described in the present invention further adds the definition of special distribution defect maps on the basis of conventional sampling detection, and matches different OOC / OOS values ​​to inspect and judge wafers with special distribution defect maps. It is not only highly timely, and can promptly discover special distribution defect maps, and determine whether sampling inspection is needed according to the defect distribution type, but also has highly targeted sampling, enhances sampling inspection of special distribution defect groups, and increases sample size.

[0053] In summary, the method for increasing special distribution defect detection described in the present invention further adds the definition of special distribution defect maps on the basis of conventional sampling detection, and matches different OOC / OOS values ​​to inspect and judge wafers with special distribution defect maps. It is not only highly timely and can promptly discover special distribution defect maps and determine whether sampling inspection is needed according to the defect distribution type, but also has highly targeted sampling, enhances sampling inspection of special distribution defect groups, and increases sample size.

[0054] Those skilled in the art will appreciate that various modifications and variations may be made to the present invention without departing from the spirit or scope of the present invention. Therefore, if any modification or variation falls within the scope of protection of the appended claims and their equivalents, the present invention is considered to cover such modifications and variations.

Claims

1. A method for enhancing the detection of special distribution defects, characterized in that: The method for enhancing special distribution defect detection includes: Executing step S1: performing defect scanning on a defect scanning machine to generate a wafer defect distribution map; Executing step S2: performing graphic algorithm processing on the wafer defect distribution map to determine whether the wafer defect distribution map has a special distribution defect map; wherein the special distribution defect map includes type 1 cluster type, type 2 line type, type 3 arc type, and type 4 fan type; Execute step S3: for the wafers without special distribution defect maps, perform random sample inspection and judgment according to the first rule based on the out-of-control values ​​and out-of-specification values ​​set by the defect control specification of the corresponding process; for the wafers with special distribution defect maps, perform graphic algorithm processing, extract and classify the special distribution defect maps, and then match different out-of-control values ​​and out-of-specification values ​​for various types of special distribution defect maps, and perform sample inspection and judgment of the special distribution defect maps according to the second rule.

2. The method for enhancing special distribution defect detection according to claim 1, characterized in that: The method for enhancing special distribution defect detection further comprises: Executing step S4: when judging according to the first rule, if the wafer without the special distribution defect pattern does not trigger an alarm, the wafer without the special distribution defect pattern is batch released; if the wafer without the special distribution defect pattern triggers an alarm, executing the out-of-control action plan and performing random sample inspection and judgment on the wafer without the special distribution defect pattern; Execute step S5: When judging according to the second rule, if the wafer with the special distribution defect pattern does not trigger an alarm, the wafer that does not trigger an alarm is batch released; if the wafer with the special distribution defect pattern triggers an alarm, the out-of-control action plan is executed and sample inspection is performed on the wafer with the special distribution defect pattern.

3. The method for enhancing detection of special distribution defects according to claim 2, wherein: The number of wafers that have not triggered an alarm, that is, the number of wafers having a defect distribution map or a special distribution defect map, is less than the out-of-control value and the out-of-specification value set by the defect control specification of the corresponding process.

4. The method for enhancing special distribution defect detection according to claim 2, wherein: The wafer trigger alarm means that the number of wafers having a defect distribution map or a special distribution defect map is greater than or equal to the out-of-control value and the out-of-specification value set by the defect control specification of the corresponding process.

5. The method for enhancing special distribution defect detection according to claim 1, wherein: The method for enhancing the detection of special distribution defects performs sample inspection and judgment, that is, performs sample electron microscope image sampling and detection to confirm the defect image.

Images