Laser Cross-Sectioning With AI Material Detection for Microelectronics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining material composition in microelectronic devices are often inaccurate and time-consuming, relying on trial-and-error due to a lack of mechanistic understanding of laser-matter interaction, and struggle with high-throughput material detection and precise machining of complex layered structures.
Innovation Solution
A system utilizing a multilayer fully connected neural network trained on confocal images to predict material composition, integrated with a femtosecond laser and CO2 gas delivery system for synchronized machining, enabling fast and precise material detection and machining of microelectronic parts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional methods (FIB milling, mechanical etching, chemical etching) are used for material detection, then measurement precision is improved, but productivity deteriorates due to time-consuming processes
Solution Approach 1:
The patent replaces mechanical and chemical methods with laser-based ablation and confocal optical imaging. The laser system ablates material while the confocal microscope captures surface topology images, eliminating the need for mechanical FIB milling or chemical etching processes while maintaining detection accuracy and dramatically improving throughput.
Solution Approach 2:
The patent creates optical copies (confocal images) of the material surface to analyze surface topology characteristics. Instead of physically examining the material through destructive methods, the system captures optical representations and analyzes image features to identify material composition, enabling non-destructive high-speed detection.
2Productivity
If ultrashort pulsed laser is used for material ablation, then productivity is improved through fast machining, but measurement precision deteriorates due to lack of mechanistic understanding leading to trial-and-error approaches
Solution Approach 1:
The patent implements a feedback loop where confocal images are captured during laser ablation, surface topology parameters are extracted and analyzed by a neural network to predict material composition, and this information feeds back to adjust laser parameters. This closed-loop system eliminates trial-and-error by using real-time optical feedback to guide the ablation process and identify materials accurately.
Solution Approach 2:
The patent dynamically changes laser parameters (power, pulse duration, scanning speed) based on material composition predictions from the neural network. The system adjusts these parameters in real-time to optimize both ablation efficiency and detection accuracy for different materials encountered during the process.
3Measurement precision
If advanced deprocessing equipment is used to access internal IC structures, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent creates a universal system that combines laser ablation and confocal imaging into a single integrated platform capable of both material removal and material identification. This multi-functional system replaces multiple specialized devices (FIB millers, chemical etching stations, microscopes) with one system that performs both functions, reducing device complexity and cost.
Solution Approach 2:
The patent merges the material removal function (laser ablation) with the material detection function (confocal imaging) into a single integrated system. The laser and confocal microscope are combined in one setup, allowing simultaneous or sequential ablation and imaging without requiring separate equipment stations, thereby simplifying the overall device 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
The system provides accurate and efficient material characterization, enabling real-time feedback for optimized machining and reducing the need for multiple samples, thus improving the throughput and accuracy of material detection and machining processes.
Implementation Method 1
Ultrashort pulsed (USP) laser which offers a thermal material ablation
Implementation Method 2
The laser system is configured to controllably etch a specimen
Implementation Method 3
The imaging system is configured to capture image data of the etched surface of the specimen
Data Source
AI summary
Systems and methods for detecting a material composition of a specimen and for cross-sectioning of the specimen. The system includes an imaging system, a femtosecond laser source, and optionally, a synchronized CO2 injection system. The imaging system is configured to capture image data of a surface of the specimen that has been etched by the laser. A machine learning model is applied to determine a predicted material composition of the specimen based at least in part on the image data. The machine learning model is trained to receive as input the image data and/or one or more quantified surface texture parameters determined from the image data and to produce as output an indication of a predicted material composition. A laser-based milling system is configured to use these material composition detection mechanisms to automatically determine when the laser system has milled through a first layer of a specimen and reached a second layer, and to adjust the operation of the milling system in response. The CO2 injection system can be used to provide fast, clean, high aspect ratio cross-sectioning of microelectronic parts for providing high-precision and high-throughput machining for material removal (e.g., for intrusive inspection of electronic components).