Polysilicon Characterization via TEM Frequency Spectrum Analysis

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods struggle to effectively characterize the quality of polysilicon in 3D NAND memory devices at the early stages of manufacturing, leading to potential defects and increased costs due to delayed detection of low-quality wafers.

Innovation Solution

A method utilizing transmission electron microscopy (TEM) image processing to extract and analyze the frequency spectrum of polysilicon structures, identifying crystal grain orientations and calculating the polysilicon crystallization ratio (PCR) to assess quality, allowing for early decision-making on wafer processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of time

If traditional characterization methods are used, then manufacturing cost is reduced, but detection timing is delayed to back-end processes

Engineering Contradiction:

Improvedetection timingVSAvoidmanufacturing cost

Core Design Contradiction:

Loss of timeVSEase of manufacture

Solution Approach 1:

The patent applies preliminary action by performing polysilicon crystallization characterization at the front-end of manufacturing using TEM imaging and Fourier transform analysis. This early detection approach identifies quality issues before they propagate through the manufacturing process, enabling preventive actions rather than corrective actions at the back-end, thus resolving the contradiction between early detection and manufacturing cost.

Inventive Principle:

Principle #10Preliminary action

2Measurement precision

If TEM image processing with Fourier transformation is used, then polysilicon quality detection precision is improved, but measurement complexity increases

Engineering Contradiction:

Improvepolysilicon quality detection precisionVSAvoidmeasurement complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces complex physical measurement systems with computational image processing. By using Fourier transformation of TEM images, the system achieves high-precision polysilicon crystallization characterization without requiring complex experimental apparatus. This substitution of mechanical/physical measurement complexity with computational analysis resolves the contradiction between detection precision and measurement complexity.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Measurement precision

If comprehensive frequency spectrum analysis is performed, then characterization accuracy is improved, but processing time increases

Engineering Contradiction:

Improvecharacterization accuracyVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent extracts only the relevant frequency spectrum components from the complete Fourier transform of TEM images. By identifying and analyzing only the frequency components corresponding to polysilicon crystal structures rather than processing the entire frequency spectrum, the method achieves accurate characterization while significantly reducing processing time, thus resolving the contradiction between accuracy and processing time.

Inventive Principle:

Principle #2Taking out (Extraction)

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 early detection of low-quality wafers, reducing manufacturing costs by characterizing polysilicon quality before the back-end of the manufacturing process, ensuring better electrical performance and device reliability.

Implementation Method 1

The image data is in a spatial domain and is generated by transmission electron microscopy (TEM)

Methodology Applied

Scientific EffectTransmission electron microscopy: Electron Beam

Implementation Method 2

To extract the frequency spectrum of the image data in the frequency domain, in some examples, the method includes performing a Fourier transformation on the image data of the polysilicon structure

Methodology Applied

Scientific EffectFourier transformation:

Implementation Method 3

To transform the selected subset of the frequency spectrum to the spatial domain to construct the first spatial image for the first crystals of the first orientation, in some examples, the method includes performing an inverse Fourier transformation on the selected subset of the frequency spectrum

Methodology Applied

Scientific EffectInverse Fourier transformation:

Data Source

PatentUS11467084B2Methods for polysilicon characterization

Publication Date: 2022.10.11 YANGTZE MEMORY TECH CO LTD

AI summary

Aspects of the disclosure provide methods for polysilicon characterization. The method includes receiving image data of a polysilicon structure formed on a sample substrate. The image data is in a spatial domain and is generated by transmission electron microscopy (TEM). Further, the method includes extracting frequency spectrum of the image data in a frequency domain. Then, the method includes selecting a subset of the frequency spectrum that corresponds to characteristic of first crystal grains that are of a first orientation, and transforming the selected subset of the frequency spectrum to the spatial domain to construct a first spatial image for the first crystal grains of the first orientation.