Characterization Analysis Software for Microstructures Based on an Internet Sharing Platform

Kaixin Sun, Jiasi Wei, Junjie WuJ. Softw 2024 · DOI

Abstract

In the fields of semiconductor manufacturing and precision engineering, analyzing microstructure surface morphology is essential for quality control. Traditional metrology relies on expensive, locally installed commercial software, which hinders collaboration. This paper introduces a cloud-based characterization and analysis platform built on a Browser/Server (B/S) architecture. By integrating standardized metrology algorithms with web-based 3D rendering technologies like Babylon JS, the platform allows users to perform complex analysis and secure data sharing directly through a web browser. Validation using physical standard samples demonstrates high accuracy in measuring step heights, grating periods, and surface roughness.

Introduction

The characterization of microstructures—including step heights, grating periods, and surface roughness—is a cornerstone of modern precision engineering. Historically, researchers have depended on specialized software such as MountainsMap or TrueMap. While powerful, these tools often require costly licenses and significant local hardware resources. Even open-source alternatives like Gwyddion necessitate local installation, creating barriers to seamless data sharing and collaborative research.

To address these challenges, we developed an internet-based sharing platform that centralizes data management and provides cloud-based analysis tools. This approach decouples metrology from specific hardware and proprietary software, fostering a more open and efficient research environment.

System Architecture and Methodology

The platform is designed using a Browser/Server (B/S) architecture, shifting the computational burden from the client to a centralized cloud server.

The software architecture stack, highlighting the use of Laravel, Babylon JS, and PostgreSQL on a CentOS server.

The backend infrastructure utilizes a CentOS Linux server running Apache2 and a PostgreSQL database. The web interface is powered by the Laravel framework, while Babylon JS enables high-performance 3D rendering of surface morphology data within the browser. Algorithmic processing is handled via OpenCV.js. The system supports various data formats (e.g., .npz, .tdms, .txt, .sur, .gwy) and employs a four-tier permission system to ensure data security and integrity.

Metrology Algorithms

The platform implements several standardized evaluation algorithms:

  • Step Height: Adheres to the ISO 5436-1:2000 standard, utilizing a bilateral fitting method with least squares to minimize errors from edge imperfections.
  • Grating Period: Employs a Fourier transform method to identify peak amplitudes in the frequency domain, offering greater robustness against surface impurities compared to center-of-gravity methods.
  • Roughness: Supports both 2D (Ra, Rq) and 3D (Sa, Sq) global surface roughness calculations based on ISO 25178-2006 standards, established via a least-squares datum plane.

Experimental Validation and Results

The software was validated using measurement data from physical standard samples. The platform successfully demonstrated its capability to visualize and analyze complex 3D structures.

3D morphology and cross-sectional analysis of a step sample as rendered by the platform.

Key results from the validation include:

  • Step Height: A nominal 1.8 μm step sample was measured at 1.789 μm.
  • Grating Period: A nominal 3 μm grating sample yielded a calculated period of 2.911 μm.
  • Roughness: A 6 μm Ra standard sample was evaluated at 5.374 μm (Ra) and 6.344 μm (Rq).
  • 3D Roughness: Analysis of a gold-plated sample provided an Sa of 1.387 nm and an Sq of 1.669 nm.

Discussion and Conclusion

The development of this platform proves that cloud-based metrology is a viable alternative to traditional local software. By leveraging Web3D frameworks, dense morphology data can be rendered smoothly, facilitating frictionless collaboration. Centralizing ISO-standard algorithms ensures that all users operate on the same mathematical baseline, reducing inter-laboratory discrepancies.

While the current iteration faces limitations in UI intuitiveness and the breadth of supported proprietary formats compared to established commercial tools, it provides a foundational framework for future expansion. Future work will focus on increasing format compatibility, adding new analysis modules, and optimizing rendering speeds for larger datasets.

Summary updated

BibTeX
@article{sun2024characterization,
  author = {Kaixin Sun and Jiasi Wei and Junjie Wu},
  title = {{Characterization Analysis Software for Microstructures Based on an Internet Sharing Platform}},
  journal = {J. Softw.},
  year = {2024},
  doi = {10.17706/jsw.19.4.98-107},
}
Figure