Method of OPC model building, information-processing apparatus, and method of determining process conditions of semiconductor device

An information-processing apparatus for forming optical proximity correction (OPC) models for quantifying an optical proximity correction process, based on selected design data and on OPC-processed image data of a pattern formed on a wafer using a mask based on a design pattern that is formed based on said design data, said image data having been acquired by an electron microscope according to selected acquisition conditions, said apparatus comprising:

• a data processing unit;
• and
• a data storage unit;
• wherein,
• said data processing unit is programmed to
• i) perform a first comparison of areas of said image data and said design data;
• ii) classify results of said first comparison and correlate said results with said acquisition conditions;
• and
• iii) store said results of said first comparison in said storage unit, together with said correlated acquisition conditions;
• said data processing unit is further programmed to set comparison areas with respect to each area on said design data to which an OPC pattern is added;
• said data processing unit acquires size differences between said image data and said design pattern based on a first comparison for each comparison area;
• said data processing unit classifies sizes of said OPC pattern according to said acquired size differences between said image data and said design pattern for each comparison area;
• said data processing unit constructs a database which stores a result of said classification;
• and
• said data processing unit is further programmed to:
• superpose said image data and said design data to achieve matching;
• set a coordinate system for a specific region;
• calculate an approximate curve that approximates a borderline of said image data in said specific region;
• integrate said approximate curve, thereby calculating the area of said image data;
• calculate a gradient of said approximate curve at specific coordinates;
• perform a second comparison of said calculated gradient with a gradient of a corresponding position in said design data;
• and
• correlatively classify and store results of said second comparison and conditions for acquiring said image data in said storage unit.

A method of determining process conditions for fabrication of a semiconductor device by utilizing optical proximity correction (OPC) modeling, comprising:

• selecting an OPC pattern and a first set of process conditions from a database created by OPC modeling;
• fabricating a semiconductor device by forming a pattern on a substrate based on said selected first set of process conditions;
• inspecting whether or not said fabricated semiconductor device falls within a preset specification;
• and
• determining a second set of process conditions according to the result of said inspection;
• wherein,
• said database correlatively classifies and stores results of previous comparisons of an area of OPC-processed electron microscope-based image data with a corresponding area of a design pattern, together with corresponding conditions for acquiring said image data;
• and
• said database is constructed by,
• providing design data;
• forming a design pattern based on said design data;
• using an electron microscope to acquire image data of a pattern formed on a wafer using a mask based on said design pattern;
• setting comparison areas with respect to each area on said design data to which an OPC pattern is added;
• acquiring size differences between said image data and said design pattern based on a first comparison for each comparison area;
• classifying sizes of said OPC pattern according to said acquired size differences between said image data and said design pattern for each comparison area;
• and
• constructing said database, which stores a result of said classifying, and
• said acquiring and classifying steps are performed by a data processing unit that:
• superposes said image data and said design pattern to achieve matching;
• sets a coordinate system for a specific region;
• calculates an approximate curve that approximates a borderline of said image data in said specific region;
• integration-processes said approximate curve, thereby calculating the area of said image data;
• calculates a gradient of said approximate curve at specific coordinates;
• performs a second comparison of said calculated gradient with a gradient of a corresponding position in said design pattern;
• and
• correlatively classifies and stores results of said second comparison and the conditions for acquiring said image data in said storage unit.