Method for manufacturing semiconductor devices and method and its apparatus for processing detected defect data

• US 6,841,403 B2
• Filed: 01/21/2003
• Issued: 01/11/2005
• Est. Priority Date: 09/20/2001
• Status: Expired due to Term

• Current Assignee: Hitachi High-Tech Company Limited (Hitachi, Ltd.)
• Original Assignee: Hitachi, Ltd.
• Inventors: Takagi, Yuji, Okabe, Takafumi, Maeda, Shunji, Tanaka, Maki, Noguchi, Minori, Shishido, Chie
• Primary Examiner(s): PERT, EVAN T
• Application Number:
  US10/348,747
• Publication Number:
  US 20030138978A1
• Time in Patent Office: 721 Days
• Field of Search: 438/5, 438/8, 438/14, 700/121, 702/35, 702/81, 702/83, 702/84, 702/110, 702/117, 702/118
• US Class Current: 438/14
• CPC Class Codes: H01L 21/67276 Production flow monitoring, e.g. for increasing throughput program-control systems per se G05B19/00, e.g. total factory control G05B19/418
• CPC Class Codes: H01L 22/20 Sequence of activities consisting of a plurality of measurements, corrections, marking or sorting steps
• CPC Class Codes: H01L 2924/00 Indexing scheme for arrangements or methods for connecting or disconnecting semiconductor or solid-state bodies as covered by H01L24/00
• CPC Class Codes: H01L 2924/0002 Not covered by any one of groups H01L24/00, H01L24/00 and H01L2224/00

Abstract

Disclosed herein is a method for manufacturing semiconductor device and a method and apparatus for processing detected defect data, making it possible to quickly infer or determine a process and related manufacturing equipment that causes defects in a fabrication line of semiconductor devices, take remedy action, and achieve a constant and high yield. The method of the invention comprises quantitatively evaluating similarity of a defects distribution on a wafer that suffered abnormal occurrence of defects to inspection results for wafers inspected in the past, analyzing cyclicity of data sequence of evaluated similarity, evaluating relationship between the cyclicity of defects obtained from the analysis and the process method according to each manufacturing equipment in the fabrication line, and inferring or determining a causal process and equipment that caused the defects.

Claims

1. A method for processing detected defect data comprising:

a step of generating a defects distribution on a semiconductor substrate in order to investigate the cause of the defects by processing detected defect data obtained by inspecting semiconductor substrates processed until a predetermined sub-process, with an inspection equipment, for defects or suspected defects occurred to the semiconductor substrates in a fabrication line consisting of a plurality of sub-processes arranged in sequence, each sub-process being performed by one or a plurality of manufacturing equipments;

a step of quantitatively evaluating defect features of the defects distribution on a semiconductor substrate generated by said step of generating a defects distribution;

a step of generating data of the defect features quantitatively evaluated by said step of quantitatively evaluating defect features for the manufacturing equipments that processed the semiconductor substrates and the sub-processes that applied to the semiconductor substrates; and

a causation inferring step of evaluating cyclicity of the defect features data for the manufacturing equipments that processed the semiconductor substrates and the sub-processes that applied to the semiconductor substrates, generated by said step of generating data, thereby inferring a causal manufacturing equipment that caused the defects.

2. A method for processing detected defect data according to claim 1, wherein said step of generating a defects distribution includes a step of specifying in advance a defects distribution on a semiconductor substrate.

3. A method for processing detected defect data according to claim 1, wherein said step of quantitatively evaluating defect features includes a step of evaluating similarity of the defects distribution on a semiconductor substrate generated by said step of generating a defects distribution to defects distributions respectively observed on a plurality of semiconductor substrates inspected in the past.

4. A method for processing detected defect data according to claim 1, wherein, in said step of evaluating similarity, the similarity is evaluated with regard to defects distributions of nearly the same defect type.

5. A method for processing detected defect data according to claim 1, wherein said causation inferring step includes a step of identifying a manufacturing equipment whose cycle of processing semiconductor substrates nearly corresponds to said cyclicity of the defect features data sequence.

11. A method for processing detected defect data according to claim 1, wherein said causation inferring step includes a step of identifying a manufacturing equipment whose cycle of processing semiconductor substrates nearly corresponds to said cyclicity of the defect features data sequence.

6. A method for processing detected defect data comprising:

a step of generating a defects distribution on a semiconductor substrate in order to investigate the cause of the defects by processing detected defect data obtained by inspecting semiconductor substrates processed until a predetermined sub-process with an inspection equipment for defects or suspected defects occurred to the semiconductor substrates in a fabrication line consisting of a plurality of sub-processes arranged in sequence, each sub-process being performed by one or a plurality of manufacturing equipments;

a step of quantitatively evaluating defect features of the defects distribution on a semiconductor substrate generated by said step of generating a defects distribution;

a step of generating data of the defect features quantitatively evaluated by said step of quantitatively evaluating defect features for the manufacturing equipments that processed the semiconductor substrates and the sub-processes that applied to the semiconductor substrates; and

a causation inferring step of evaluating correlations between said defect features and said manufacturing equipments that processed the semiconductor substrates for each sub-process that applied to the semiconductor substrates, based on the defect features data for the manufacturing equipments that processed the semiconductor substrates and the sub-processes that applied to the semiconductor substrates, generated by said step of generating data, and selecting a sub-process of high correlation, thereby inferring a causal manufacturing equipment that caused the defects.

7. A method for processing detected defect data according to claim 6, wherein said step of generating a defects distribution includes a step of specifying in advance a defects distribution on a semiconductor substrate.

8. A method for processing detected defect data according to claim 6, wherein said step of quantitatively evaluating defect features includes a step of evaluating similarity of the defects distribution on a semiconductor substrate generated by said step of generating a defects distribution to defects distributions respectively observed on a plurality of semiconductor substrates inspected in the past.

9. A method for processing detected defect data according to claim 6, wherein, in said step of evaluating similarity, the similarity is evaluated with regard to defects distributions of nearly the same defect type.

10. A method for processing detected defect data according to claim 6, wherein, in said step of specifying a defects distribution, one or a plurality of regions formed by a closed curve are specified as discrete defects distributions on a semiconductor substrate and said step of quantitatively evaluating defect features evaluates defects features for each region thus specified.

12. A method for processing detected defect data according to claim 6, wherein, in said step of specifying a defects distribution, a defects distribution specified in advance on a semiconductor substrate is a defects distribution merged with inspection results obtained in the past for specific semiconductor substrates.

13. A method for processing detected defect data comprising the following steps:

a step of evaluating similarity of a defects distribution within a chip or shot specified in advance on a semiconductor substrate to a defects distribution within each of the remaining chips or shots on the semiconductor substrate by processing detected defect data obtained by inspecting semiconductor substrates processed until a predetermined sub-process with an inspection equipment for defects or suspected defects occurred to the semiconductor substrates in a fabrication line consisting of a plurality of sub-processes arranged in sequence, each sub-process being performed by one or more manufacturing equipments;

a step of evaluating how the similarity evaluated is distributed over the semiconductor substrate.

14. An apparatus for processing detected defect data comprising:

a portion for generating a defects distribution on a semiconductor substrate in order to investigate the cause of the defects, based on detected defect data obtained by inspecting semiconductor substrates processed until a predetermined sub-process, with an inspection equipment, for defects or suspected defects occurred to the semiconductor substrates in a fabrication line consisting of a plurality of sub-processes arranged in sequence, each sub-process being performed by one or a plurality of manufacturing equipments;

a portion for quantitatively evaluating defect features of the defects distribution on a semiconductor substrate generated by said portion for generating a defects distribution;

a portion for generating data of the defect features quantitatively evaluated by said portion for quantitatively evaluating defect features for the manufacturing equipments that processed the semiconductor substrates and the sub-processes that applied to the semiconductor substrates; and

a causation inferring portion for evaluating cyclicity of the defect features data for the manufacturing equipments that processed the semiconductor substrates and the sub-processes that applied to the semiconductor substrates, generated by said portion for generating data, thereby inferring a causal manufacturing equipment that caused the defects.

15. An apparatus for processing detected defect data according to claim 14, further comprising a display unit for displaying a defects distribution on a semiconductor substrate generated by said portion for generating a defects distribution, defect features data for the manufacturing equipments that processed the semiconductor substrates, generated by said portion for generating data, and information about a causal manufacturing equipment or sub-process that caused the defects, inferred by said causation inferring portion.

16. An apparatus for processing detected defect data comprising:

a portion for generating a defects distribution on a semiconductor substrate in order to investigate the cause of the defects, based on detected defect data obtained by inspecting semiconductor substrates processed until a predetermined sub-process, with an inspection equipment, for defects or suspected defects occurred to the semiconductor substrates in a fabrication line consisting of a plurality of sub-processes arranged in sequence, each sub-process being performed by one or a plurality of manufacturing equipments;

a portion for quantitatively evaluating defect features of the defects distribution on a semiconductor substrate generated by said portion for generating a defects distribution;

a portion for generating data of the defect features quantitatively evaluated by said portion for quantitatively evaluating defect features for the manufacturing equipments that processed the semiconductor substrates and the sub-processes that applied to the semiconductor substrates; and

a causation inferring portion for evaluating correlations between said defect features and said manufacturing equipments that processed the semiconductor substrates for each sub-process that applied to the semiconductor substrates, based on the defect features data for the manufacturing equipments that processed the semiconductor substrates and the sub-processes that applied to the semiconductor substrates, generated by said portion for generating data, and selecting a sub-process of high correlation, thereby inferring a causal manufacturing equipment that caused the defects.

17. An apparatus for processing detected defect data according to claim 16, further comprising a display unit for displaying a defects distribution on a semiconductor substrate generated by said portion for generating a defects distribution, defect features data for the manufacturing equipments that processed the semiconductor substrates, generated by said portion for generating data, and information about a causal manufacturing equipment or sub-process that caused the defects, inferred by said causation inferring portion.