Plasma Etching Endpoint Detection via Multi-Spectral Index Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for detecting the end point of a plasma etching process fail to accurately identify the etching end-point for features with different dimensions, such as narrower holes, leading to incomplete etching or over-etching, and struggle with detecting etching end-points in shallow-junction S/D regions, resulting in potential damage to junctions.
Innovation Solution
The method involves collecting time-dependent intensity data from multiple spectral lines of etching products in the plasma emission spectrum, calculating and plotting indices such as Lm(t) and Ls(t) to enhance the etching-through response, allowing for precise detection of etching end-points, especially for features with different dimensions and shallow-junction regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional end-point detection methods (single spectral line monitoring) are used, then the etching end-point for larger features can be detected, but the etching end-point for narrower features cannot be accurately detected
Solution Approach 1:
The invention segments the detection approach by dividing the spectral analysis into multiple components: monitoring multiple spectral lines (e.g., CN at 387nm and CO at 483nm) separately and analyzing their individual time-dependent intensity curves. This segmentation allows distinct etching events for different feature sizes to be resolved independently, enabling accurate detection of end-points for both narrow and wide contact holes
Solution Approach 2:
The invention transitions from monitoring a single spectral line to analyzing multiple spectral lines simultaneously, adding a dimension of spectral multiplicity. By plotting multiple intensity curves Iλ1(t), Iλ2(t), ..., Iλn(t) and their ratios, the system creates a multi-dimensional detection space where different etching events manifest as distinct signals, enabling comprehensive detection across varying feature dimensions
2Ease of operation
If fixed etching period is used, then the process is simple to control, but incomplete etching or over-etching occurs due to varying etching conditions
Solution Approach 1:
The invention implements real-time feedback by continuously monitoring the time-dependent intensity of spectral lines during the etching process. The system analyzes the evolution of intensity curves Iλ(t) and their ratios to detect etching end-points dynamically, providing feedback that automatically determines when to stop etching. This eliminates fixed timing constraints and adapts to varying etching conditions, ensuring complete etching without over-etching
3Device complexity
If single spectral line monitoring is used, then the detection method is simple, but the etching-through signal is not clear for shallow-junction regions
Solution Approach 1:
The invention merges information from multiple spectral lines by analyzing both the individual intensity curves and their ratios. By combining the signals from different etching products (e.g., CN and CO), the system enhances the clarity of etching-through signals for shallow-junction regions. The ratio analysis Iλ1(t)/Iλ2(t) provides additional contrast that makes weak signals more detectable, improving measurement precision without excessive complexity
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
This approach significantly enhances the detectability of etching end-points, enabling precise control over the etching process, reducing over-etching, and preventing damage to shallow-junction regions, while allowing for better selectivity and etching profile management.
Implementation Method 1
Another way to determine when to stop a plasma etching process is based on the optical emission spectroscopy (OES) of the plasma. Such a method typically includes monitoring the intensity of one spectral line of one etching product of the layer under the etching target layer in the emission spectrum of the plasma.
Data Source
AI summary
A method of detecting an end point of a plasma etching process for etching a first layer on a second layer is described, the first layer producing a first etching product and the second layer a second etching product. Time-dependent intensity [Ij=1 to m(t)] of a number “m” (m≧1) of spectral line(s) of the first etching product in emission spectrum of the plasma and that [Ii=1 to n(t)] of a number “n” (n≧1) of spectral line(s) of the second etching product in the emission spectrum are collected, wherein “m+n≧3” is satisfied. One index ofLm(t)[=∏i=1,j=1n,mIi(t)Ij(t)],Ls(t)[=∑i=1,j=1n,mIi(t)Ij(t)],Lm′(t) {=d[Lm(t)]/dt} and Ls′(t) {=d[Ls(t)]/dt} is calculated in real time and plotted with the time. An etching end-point is identified from the plot of the one index with the time.