Voxel Mesh Material Processing Simulation for Accurate Etch Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Computer-Aided Design (CAD) tools for semiconductor fabrication lack efficient simulation capabilities, particularly in accurately modeling material growth and removal processes without requiring physical device fabrication, leading to inaccuracies and increased development cycles.
Innovation Solution
A fabrication simulator utilizing a voxel mesh and Hamilton-Jacobi solving technique to simulate material processing, incorporating conformal surface meshes and stability operations like Lax-Friedrichs and weighted essentially non-oscillatory schemes to improve numerical stability and accuracy in modeling crystalline growth and etch processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If physics-based models and computer simulations are used to predict experimental results, then development cycles are reduced and resources are saved, but manufacturing precision and reliability of predictions deteriorate due to inaccuracies in modeling material growth and removal processes
Solution Approach 1:
The simulation domain is divided into a voxel mesh grid, where each voxel represents a discrete volume element. This segmentation allows the complex material processing simulation to be broken down into manageable computational units, enabling accurate tracking of material growth and removal at each voxel level while maintaining overall simulation efficiency.
Solution Approach 2:
The patent implements direction-dependent rates for each voxel based on its orientation and position within the structure. This local quality approach allows the simulation to account for anisotropic material properties and direction-specific growth/etch rates, significantly improving prediction accuracy for crystalline structures where growth characteristics vary by crystallographic direction.
Solution Approach 3:
The simulation performs preliminary calculations of direction-dependent rates and flux values for each voxel before executing the main material processing simulation. This preliminary action prepares the computational framework in advance, enabling more accurate and efficient prediction of material evolution without requiring iterative physical experimentation.
2Ease of manufacture
If conventional CAD tools are used for semiconductor fabrication simulation, then device fabrication can proceed with standard tools, but simulation accuracy deteriorates due to lack of efficient modeling capabilities for material growth and removal
Solution Approach 1:
The patent creates a virtual copy of the physical semiconductor structure in the form of a voxel mesh representation. This digital twin allows the simulation of material processing operations on the virtual model, providing accurate predictions of material growth and removal without requiring physical trial-and-error experimentation, thus improving manufacturing precision while maintaining ease of manufacture.
Solution Approach 2:
The simulation dynamically updates voxel states and material properties during the modeling process, adjusting parameters such as growth rates, etch rates, and material densities based on local conditions. This parameter adaptation enables the model to accurately reflect real material behavior under varying process conditions, significantly improving simulation accuracy for material processing.
3Measurement precision
If physical device fabrication and metrology are performed to determine operating parameters, then measurement accuracy is maintained, but resource consumption and development time increase
Solution Approach 1:
The simulation system performs self-validation by comparing simulated results with target specifications and automatically adjusting process parameters to achieve desired outcomes. This self-service capability reduces the need for external physical metrology measurements and iterative fabrication cycles, thereby decreasing resource consumption while maintaining parameter determination accuracy through virtual experimentation.
Solution Approach 2:
The patent replaces physical measurement and fabrication systems with a computational simulation system based on Hamilton-Jacobi equations and voxel-based modeling. This substitution eliminates the need for repeated physical device fabrication and metrology measurements, significantly reducing material consumption and resource usage while providing accurate predictions of material processing outcomes.
4Productivity
If numerical simulation methods are used to model material processing, then computational efficiency is improved, but numerical stability and accuracy deteriorate due to errors in modeling crystalline growth and etch processes
Solution Approach 1:
The simulation incorporates stability operations and correction mechanisms that anticipate and compensate for numerical errors before they propagate through the calculation. By applying cushioning algorithms that dampen oscillations and correct deviations in advance, the system maintains numerical stability and accuracy in modeling crystalline growth and etch processes while preserving computational efficiency.
Solution Approach 2:
The patent implements feedback mechanisms that continuously monitor simulation results and adjust computational parameters to maintain numerical stability. The simulation evaluates the accuracy of material processing predictions at each step and applies corrective adjustments, ensuring reliable and stable results while maintaining high computational productivity through efficient algorithms.
Data Source
AI summary
Among other things, one or more techniques for simulating a process operation of a process tool are provided. In an embodiment, a voxel mesh is defined to represent a structure and a conformal surface mesh is defined for the voxel mesh. Direction dependent rates are determined for voxels in the voxel mesh using the conformal surface mesh. The voxel mesh is updated based on the direction dependent rates. The defining of the conformal surface mesh, the determining of the direction dependent rates, and the updating of the voxel mesh are iterated to simulate the process operation. A parameter of the voxel mesh is determined after simulating the process operation.