SimuSEM Software
SEM simulation software
AMAG nanometro's SimuSEM software simulates scanning electron microscope (SEM) images and electron interactions with solids.
SimuSEM's core simulation engine is NIST’s validated JMONSEL simulator, augmented with much faster performance, many additional materials, and numerous other improvements. SimuSEM provides a comprehensive 3D visualization system using the Blender application as an intuitive GUI.
SimuSEM can run on virtually any Windows PC or Linux system, including in-house servers and workstations, cloud servers, desktop PCs, or even laptops, with performance that scales with the number of available CPU cores.
Intro to SimuSEM
ONO Intentional Defect Array
AMAG7 XSEM
Buried Copper Grating
Use Cases
SimuSEM is very flexible, capable of valuable simulation support in many roles:
- Virtual metrology / application exploration & understanding
- Metrology Gaps / Limits / Feasibility studies
- Fundamental understanding of signal generation
- Model Based Metrology algorithm development
- Hybrid metrology comparisons involving SEM
- Faux image generation for AI training
- SEM imaging condition optimization
- Calibration of analytical models
- HV-SEM/Overlay exploration
- Ability to quantitatively study detailed profile influences on SEM signal, or any/all of the stochastic components involved in roughness measurement, key industry knowledge gaps
Key Use Case: Cost Reduction of Metrology and Inspection Ownership
Organizations using SimuSEM can achieve an increase in metrology/inspection throughput, reduction of production cycle-time, and an overall improvement in process control by optimizing current metrology and inspection recipes. SimuSEM is a virtual SEM that can improve metrology and defect inspection by resolving issues, optimizing, and assisting with new recipe creations without adding to cycle time, for a small fraction of the cost of the purchase of another metrology tool.
SimuSEM's Core Simulation Engine
NIST JMONSEL
JMONSEL — Java MONte Carlo Simulator for Secondary Electrons — is a 3D electron beam simulator developed between 2010 and 2012 at NIST by Dr. John Villarrubia, and funded by SEMATECH AMAG, to enable limits simulation studies for defect and critical dimension SEM metrology.
NIST JMONSEL is a full Monte Carlo simulator that tracks primary electrons as they enter material, scatter, lose energy, and generate secondary and back-scattered electrons. It monitors simulated electrons that exit the material and are captured by the detector (a software counter element). Electron yields are found at any point designated as a target pixel. Physical models in JMONSEL are the best-known models in the literature in the energy ranges used, open-source, and completely transparent in their documentation, definition, implementation and execution. The physical models were programmed by NIST, with a decade of validation data and wide acceptance by the semiconductor industry.
SimuSEM JMONSEL
SimuSEM's core simulation engine is an augmented version of JMONSEL, with simulation speed that is orders of magnitude faster than NIST JMONSEL. SimuSEM supports many additional materials beyond those supported by NIST JMONSEL.
Sample geometry definitions are created within the SimuSEM GUI, rather than with a text script. The result is a user-friendly SEM simulator with rich and reliable visualization. SimuSEM supports a wide range of simulation use cases, with more complex sample geometries and materials than other simulators on the market.
Compared to NIST JMONSEL, SimuSEM greatly improves utility, productivity, flexibility, visualization, accessibility, and achievable complexity of designed features while improving simulation speed and scalability. Thus SimuSEM can be applied to most conceivable realistic simulation scenarios.
SimuSEM's GUI
SimuSEM's GUI is built on the Blender open-source 3D modeling application. Blender is widely used for entertainment-quality animation, visual effects, interactive 3D applications, and 3D printer target design. It is an excellent 3D Cartesian space for a physical simulation GUI, providing the user with a powerful and rewarding sample generation workbench.
Blender's 3D interface allows SimuSEM to scale, move and rotate primitive and user-defined shape objects and Boolean operators for sample design with outstanding visualization.
Key SimuSEM Features
- Simulations run in the background, with results displayed in SimuSEM's GUI for analysis and detailed viewing.
- Discretely-defined height map objects for rough features and edges.
- A growing database of materials, with a path to adding more, including proprietary materials.
- Adjustable beam energies, beam spot sizes or profiles and tilt beam, user-defined discrete detectors with tunable energy windows, and other SEM variants.
- User-defined roughness, new materials, and other augmentations.
- Results images for SE, BSE or user-defined energy windows, viewable top-down and in 3D.
- Energy, angle & detector hit histograms/maps, and energy absorption maps.
- Fully viewable electron trajectories in 2D & 3D.
Papers
- ASMC 2024: Coming of Age in Computational SEM
- SPIE 2024: Simulating SEM imaging of via bottoms
- SPIE 2023: Simulating HV-SEM imaging of HAR and buried features
- SPIE 2022: Simulating Process Subtleties in SEM Imaging
Presentations
- ASMC 2024: Coming of Age in Computational SEM
- Video: SimuSEM GUI
- Video: Buried grating overlayer
- Video: SimuSEM runtime
- Video: HAR profile setup
- Video: ONO IDA
- SPIE 2024: Simulating SEM imaging of via bottoms
- Video: Via Bottom Setup
- SPIE 2024: Quantitative analysis of cross-section SEM spatial distortion artifacts
- SPIE 2023: Simulating HV-SEM imaging of HAR and buried features
- SPIE 2022: Simulating Process Subtleties in SEM Imaging