Ion Beam Alignment via Convolutional Neural Network
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for aligning high current focused ion beams, such as plasma FIB milling, are time-consuming and often result in unsuitable or undesirable ion beam characteristics, limiting their effectiveness in semiconductor processing.
Innovation Solution
The use of a convolutional neural network to analyze training spot images of ion beam patterns milled into a substrate, allowing for automated adjustment of the ion beam column by processing operational exposure spot images and determining necessary adjustments to lens focus, current, aperture location, or stigmator settings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual alignment methods are used for ion beam column adjustment, then the alignment process can be performed with simple equipment, but the alignment time is excessive and the beam characteristics are unsuitable for high current FIB milling
Solution Approach 1:
The patent replaces manual mechanical alignment operations with an automated optical measurement and control system. A camera captures beam spot images, and a processor automatically analyzes these images to determine alignment parameters and control signals, substituting the mechanical manual adjustment process with an automated optical-mechanical system that achieves both speed and precision
Solution Approach 2:
The system enables self-aligned operation where the ion beam column automatically adjusts its own alignment based on real-time beam spot image analysis. The processor generates control signals that feed back to the beam column components (lenses, deflectors, apertures), creating a self-correcting alignment system that operates autonomously without manual intervention
Solution Approach 3:
The patent implements a closed-loop feedback system where beam spot images are continuously captured, analyzed to determine alignment status, and used to generate corrective control signals. This feedback mechanism allows the system to automatically detect and correct alignment deviations, maintaining precise beam characteristics while operating at high speed
2Manufacturing precision
If conventional manual alignment is used, then the system complexity remains low, but the alignment quality degrades and requires excessive operator time
Solution Approach 1:
The patent employs a multi-functional integrated system where a single apparatus performs beam delivery, beam spot imaging, image analysis, and automatic alignment control. The processor serves multiple functions including capturing beam parameters, analyzing alignment status, generating control signals, and coordinating various beam column components, thereby achieving high precision without proportionally increasing system complexity
Solution Approach 2:
The patent introduces a camera as an intermediary device that optically captures beam spot images, serving as a mediator between the ion beam and the digital processing system. This optical intermediary enables precise non-contact measurement of beam characteristics, translating physical beam properties into digital data for automated analysis and control
3Measurement precision
If automated neural network-based alignment is implemented, then alignment speed and precision are improved, but the computational requirements and processing complexity increase
Solution Approach 1:
The patent replaces complex manual measurement and judgment processes with automated optical imaging and digital image analysis. Instead of requiring operators to visually assess beam characteristics and make subjective alignment decisions, the system uses a camera to capture precise beam spot images and a processor to objectively analyze them, substituting human cognitive processing with automated computational analysis
Solution Approach 2:
The patent transforms physical beam alignment parameters into measurable image parameters through optical imaging. Beam spot position, size, shape, and intensity distribution are converted into pixel-based image data that can be quantitatively analyzed by the processor, enabling precise measurement and control through parameter transformation from the physical domain to the digital domain
Data Source
AI summary
A focused ion beam (FIB) is used to mill beam spots into a substrate at a variety of ion beam column settings to form a set of training images that are used to train a convolutional neural network. After the neural network is trained, an ion beam can be adjusted by obtaining spot image which is processed with the neural network. The neural network can provide a magnitude and direction of defocus, aperture position, lens adjustments, or other ion beam or ion beam column settings. In some cases, adjustments are not made by the neural network, but serve to indicate that the ion beam and associated ion column continue to operate stably, and additional adjustment is not required.