SEM Pattern Dimension Measurement via Model Profile Matching

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional methods for measuring fine patterns on semiconductor substrates lack accuracy in determining dimensions at specific positions on cross-sectional shapes, especially for transistor gate wiring patterns, and require numerous pre-prepared models, making them costly and time-consuming.

Innovation Solution

A method and apparatus using a scanning electron microscope to create an image profile from a secondary electron image, matching it with prestored model profiles to determine dimensions at desired positions on the cross-sectional shape, reducing the need for numerous pre-prepared models and enabling accurate measurements with fewer models.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional image profile methods are used to measure pattern dimensions, then the measurement process is simple, but the measurement precision at specific cross-sectional positions cannot be achieved

Engineering Contradiction:
Improvemeasurement precisionVSAvoiddevice complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent introduces a database of model profiles as an intermediary between the measured image profile and the actual pattern dimensions. The model profiles serve as a mediator that contains pre-calculated dimensional information for various cross-sectional shapes, allowing accurate measurement without complex real-time calculations. This resolves the contradiction by providing precise measurements through the intermediary model database rather than through complex measurement algorithms.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent performs preliminary action by pre-calculating and storing model profiles with known dimensions and cross-sectional shapes in a database before actual measurements are taken. This preliminary preparation allows the measurement system to simply match the measured image profile against pre-computed models, achieving high precision without complex real-time processing. The preliminary action of creating model profiles resolves the contradiction between precision and complexity.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If numerous pre-prepared models are used to achieve accurate measurements, then the measurement precision improves, but the device complexity and time required increase

Engineering Contradiction:
Improvemeasurement precisionVSAvoidtime required
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent uses copying by creating simplified model profiles that represent various pattern cross-sections. Instead of requiring complex physical models or extensive real-time calculations, the system copies the essential dimensional characteristics into pre-computed model profiles stored in a database. This copying approach allows rapid matching during measurement, reducing time requirements while maintaining precision through the use of accurate model representations.

Inventive Principle:
Principle #26Copying

3Manufacturing precision

If cross-sectional shape information is not considered in measurements, then the measurement process is simpler, but the manufacturing precision for controlling pattern dimensions is compromised

Engineering Contradiction:
Improvemanufacturing precisionVSAvoiddevice complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent employs model profiles as an intermediary that encapsulates cross-sectional shape information. Rather than requiring complex real-time analysis of cross-sectional geometry, the system uses pre-computed model profiles as a mediator between the measured image and the required manufacturing specifications. This intermediary approach provides manufacturing precision by incorporating cross-sectional information without adding complex processing steps, as the models are pre-prepared and stored.

Inventive Principle:
Principle #24Intermediary (Mediator)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Enables accurate measurement of pattern dimensions at desired positions on the cross-sectional shape, reducing the need for multiple models and improving measurement efficiency, even with low signal-to-noise images.

Implementation Method 1

irradiating an electron beam so as to be converged on the sample and allowing the electron beam to scan

Methodology Applied
Scientific EffectElectron beam irradiation: Electron Beam

Implementation Method 2

obtaining a secondary electron image of a sample by picking up an image of the sample using a scanning electron microscope

Methodology Applied
Scientific EffectSecondary electron detection: Photoelectric Effect

Data Source

PatentUS7408155B2Measuring method and its apparatus
Publication Date: 2008.08.05 HITACHI HIGH TECH CORP
  • US7408155B2 patent drawing
  • US7408155B2 patent drawing
  • US7408155B2 patent drawing

AI summary

A method for measuring a dimension of a pattern formed on a sample using a secondary electron image obtained by picking up an image of the sample using a scanning electron microscope includes: obtaining a secondary electron image of a sample by picking up an image of the sample using a scanning electron microscope; creating, using the secondary electron image, an image profile of a pattern whose dimension is to be measured, within the obtained secondary electron image; retrieving a model profile that matches best with the created image profile from a plurality of model profiles prestored that are obtained from respective secondary electron images of a plurality of patterns, the cross sections of the plurality of patterns being of known shapes and dimensions and being different in shape; and obtaining a dimension of the pattern using information of the retrieved model profile.