Crowd-Sourced Data and Analysis Tools for Advancing the Chemical Vapor Deposition of Graphene: Implications for Manufacturing

Joshua A. Schiller, Ricardo J. Toro, Aagam Shah, Mitisha Surana, Kaihao Zhang, Matthew J. Robertson, Kristina M. Miller, Kevin Cruse, Kevin Liu, Bomsaerah Seong, Seol Chae, Ian Foster, Ben Blaiszik, Ben Galewsky, Darren Adams, Daniel S. Katz, Placid M. Ferreira, Elif Ertekin, Sameh Tawfick

TL;DR

Gr-ResQ is an open-source, crowd-sourced platform designed to standardize and accelerate the production of high-quality graphene. By providing a searchable database of over 300 synthesis parameters and automated analysis tools for Raman and SEM data, it aims to move graphene manufacturing from "black magic" lab techniques to a predictable, data-driven industrial process.

Abstract

Industrial production of graphene by chemical vapor deposition (CVD) requires more than the ability to synthesize high-quality material in a lab. Progress is currently hindered by hundreds of coupled synthesis parameters and poor communication of experimental data in literature. We report on Gr-ResQ, a platform featuring a crowd-sourced database of CVD synthesis recipes and automated analysis tools. By capturing ~300 parameters and providing standardized processing of Raman and SEM data, Gr-ResQ enables machine-learning models to advance synthesis and facilitates cyber-physical collaborations. This holistic approach aims to accelerate CVD recipe discovery and production control for graphene and other 2D materials.

The Reproducibility Crisis in CVD

Chemical Vapor Deposition (CVD) is the primary method for growing large-area graphene, but it is notoriously sensitive. A growth run can fail due to minor variations in furnace geometry or trace water vapor (dew point). Despite this sensitivity, critical parameters are rarely reported consistently in scientific literature.

An analysis of 21 highly cited papers revealed that data such as tube dimensions or ambient humidity were missing in over 90% of cases. Furthermore, the community lacks a central repository for "negative" results from failed experiments, which are essential for training predictive models. Gr-ResQ addresses this by providing a centralized home for the "rich but dispersed" knowledge of the graphene community.

The Gr-ResQ Platform

Hosted on nanoHUB, the Gr-ResQ platform integrates several modules to move graphene manufacturing toward a data-driven industrial process. It captures approximately 300 parameters across the entire CVD lifecycle, including heating, annealing, growth, and cooling phases.

The typical CVD process involves multiple stages, each with numerous parameters that influence the final graphene quality.

Automated Analysis and Data Integration

To remove subjective bias and create a "ground truth" for quality, Gr-ResQ provides automated tools for processing experimental results:

  • Image Analysis: Uses OpenCV-based template matching and Sobel filters to calculate graphene coverage and domain alignment in SEM images.
  • Raman Spectroscopy: Automatically fits Lorentzian curves to identify layer counts and defect densities (D, G, and G' peaks).
  • Dimensionality Reduction: Employs t-SNE (t-distributed Stochastic Neighbor Embedding) to visualize how high-dimensional recipes cluster, helping identify patterns like the influence of pressure on experimental similarity.
  • Cyber-Physical Integration: Connects with the Operating System for Cyberphysical Manufacturing (OSCM), allowing researchers to request experiments from remote labs and receive live-measured process data.

Quantifying the Reporting Gap

A primary contribution of this work is quantifying the information gap in current literature. Many parameters with massive impacts on gas kinetics and nucleation are frequently ignored by researchers.

Analysis of highly cited literature shows that critical parameters like Tube Length and Dew Point are unreported in nearly all cases.

Conclusion and Future Directions

Gr-ResQ establishes a framework for FAIR (Findable, Accessible, Interoperable, and Reproducible) data in the 2D materials community. By assigning DOIs to recipe groups via the Materials Data Facility (MDF), it allows researchers to receive academic credit for data contributions. While currently focused on graphene, the platform is designed to expand to other materials like transition metal dichalcogenides (TMDCs), accelerating the transition of low-dimensional materials from laboratory curiosities to scalable industrial products.

Summary updated

Abstract

Industrial production of graphene by chemical vapor deposition (CVD) requires more than the ability to synthesize large domain, high quality graphene in a lab reactor. The integration of graphene in the fabrication process of electronic devices requires the cost-effective and environmentally-friendly production of graphene on dielectric substrates, but current approaches can only produce graphene on metal catalysts. Sustainable manufacturing of graphene should also conserve the catalyst and reaction gases, but today the metal catalysts are typically dissolved after synthesis. Progress toward these objectives is hindered by the hundreds of coupled synthesis parameters that can strongly affect CVD of low-dimensional materials, and poor communication in the published literature of the rich experimental data that exists in individual laboratories. We report here on a platform, “Graphene – Recipes for synthesis of high quality material” (Gr-ResQ: pronounced graphene rescue), that includes powerful new tools for data-driven graphene synthesis. At the core of Gr-ResQ is a crowd-sourced database of CVD synthesis recipes and associated experimental results. The database captures ∼300 parameters ranging from synthesis conditions like catalyst material and preparation steps, to ambient lab temperature and reactor details, as well as resulting Raman spectra and microscopy images. These parameters are carefully selected to unlock the potential of machine-learning models to advance synthesis. A suite of associated tools enable fast, automated and standardized processing of Raman spectra and scanning electron microscopy images. To facilitate community-based efforts, Gr-ResQ provides tools for cyber-physical collaborations among research groups, allowing experiments to be designed, executed, and analyzed by different teams. Gr-ResQ also allows publication and discovery of recipes via the Materials Data Facility (MDF), which assigns each recipe a unique identifier when published and collects parameters in a search index. We envision that this holistic approach to data-driven synthesis can accelerate CVD recipe discovery and production control, and open opportunities for advancing not only graphene, but also many other 1D and 2D materials.
BibTeX
@inproceedings{schiller2020crowd,
  author = {Joshua A. Schiller and Ricardo J. Toro and Aagam Shah and Mitisha Surana and Kaihao Zhang and Matthew J. Robertson and Kristina M. Miller and Kevin Cruse and Kevin Liu and Bomsaerah Seong and Seol Chae and Ian Foster and Ben Blaiszik and Ben Galewsky and Darren Adams and Daniel S. Katz and Placid M. Ferreira and Elif Ertekin and Sameh Tawfick},
  title = {{Crowd-Sourced Data and Analysis Tools for Advancing the Chemical Vapor Deposition of Graphene: Implications for Manufacturing}},
  booktitle = {ACS},
  year = {2020},
  doi = {10.1021/acsanm.0c02018},
}
Figure