Matter42 offers an AI‑driven research workspace for 2D material scientists to ingest, analyze, and simulate Raman, photoluminescence, and microscopy data. Its multimodal agent preserves project context, performs physics‑informed defect clustering, density estimation, and links results to kinetic Monte Carlo growth simulations, delivering interactive visualizations and reproducible, structured reports.
Funding
Funding not disclosed
Founders
Product
Problem
Researchers working with 2D materials such as transition metal dichalcogenides must manually process large Raman, photoluminescence (PL) and microscopy datasets, segment defect regions, and correlate growth conditions with material quality. This workflow is time‑consuming, error‑prone, and often fragmented across separate analysis tools, slowing discovery and manufacturing optimization.
Solution
Matter42 provides an AI‑driven research workspace that integrates data ingestion, analysis, and simulation for 2D material characterization. Users upload Raman, PL maps, images, tables or documents, and the platform’s research agent parses the files, preserves project context, and can invoke domain‑specific tools via natural‑language prompts. The agent generates interactive spectral visualizations, clusters defect populations, estimates defect densities using calibrated machine‑learning models, and classifies defect families based on peak features. It also links experimental data to kinetic Monte Carlo growth simulations, enabling hypothesis testing of temperature, flux ratio and nucleation parameters. All results are stored alongside the original files, producing structured outputs and reproducible reports that remain accessible within the project’s chat history.
Target Audience
Primary users are scientists and engineers in academic or industrial labs focused on 2D material synthesis, characterization, and process development, particularly those analyzing Raman and photoluminescence spectroscopy data.
Features
- Multimodal data intake supporting LabSpec spectral maps, single spectra, numeric tables, documents and microscopy images
- AI research agent that maintains project context, cites source files, and calls analysis tools via natural‑language prompts
- Physics‑informed clustering and defect quantification pipelines for Raman and PL maps, including linewidth‑based density calibration and defect family ranking
- Automated region segmentation (interior, transition, damaged) with customizable boundary buffers for damaged or PFIB‑processed samples
- Integrated kinetic Monte Carlo CVD growth simulator for MoS2, WS2 and WSe2, providing defect statistics and animated growth trajectories
- Interactive visual outputs (spectral maps, spatial heat‑maps, summary statistics) and structured result export for downstream reporting
- Project‑wide memory that links analysis results to original files, ensuring reproducibility and traceability