AI-ready design of realistic 2D materials and interfaces with Mat3ra-2D

Vsevolod Biryukov; Kamal Choudhary; Timur Bazhirov

cond-mat.mtrl-sciMar 29, 2026

AI-ready design of realistic 2D materials and interfaces with Mat3ra-2D

Authors: Vsevolod Biryukov, Kamal Choudhary, Timur Bazhirov

Organizations: Exabyte Inc. (Mat3ra.com), Walnut Creek, CA 94596, USA · Department of Materials Science and Engineering, Johns Hopkins University, Baltimore, MD 21218, USA · Department of Electrical and Computer Engineering, Johns Hopkins University, Baltimore, MD 21218, USA

Abstract

Artificial intelligence (AI) and machine learning (ML) models in materials science are predominantly trained on ideal bulk crystals, limiting their transferability to real-world applications where surfaces, interfaces, and defects dominate. We present Mat3ra-2D, an open-source framework for the rapid design of realistic two-dimensional materials and related structures, including slabs and heterogeneous interfaces, with support for disorder and defect-driven complexity. The approach combines: (1) well-defined standards for storing and exchanging materials data with a modular implementation of core concepts and (2) transformation workflows expressed as configuration-builder pipelines that preserve provenance and metadata. We implement typical structure generation tasks, such as constructing orientation-specific slabs or strain-matching interfaces, in reusable Jupyter notebooks that serve as both interactive documentation and templates for reproducible runs. To lower the barrier to adoption, we design the examples to run in any web browser and demonstrate how to incorporate these developments into a web application. Mat3ra-2D enables systematic creation and organization of realistic 2D- and interface-aware datasets for AI/ML-ready applications.

Figures & tables

Figure 1: Interface transformation stages: define, refine, and build for selecting and constructing a strain-matched interface with recorded metadata.

Panel of Figure 2 — Figure 2: SrTiO 3 (110) slab construction with termination control: (a) bulk crystal from the reference data repository, (b) and (c) alternative surface terminations obtained with the same Miller indices, layer count, and vacuum by changing the termination formula.

Panel of Figure 3 — Figure 3: Ge(001)/Si(001) interface without strain matching: (a) and (b) film and substrate slabs defined independently with Miller indices (001), layer count, and vacuum; (c) combined interface after applying a target interfacial gap and optional lateral shift.

Panel of Figure 4 — Figure 4: Graphene/Ni(001) interface construction workflow: (a) Starting graphene monolayer film, (b) Ni(001) slab substrate, (c) ZSL analyzer output showing strain–size trade-off for candidate configurations (each point represents a commensurate match ranked by strain percentage and interface area), (d) Selected interface structure with commensurate matching applied. The workflow follows define–refine–build: define film and substrate slabs, refine by enumerating and ranking matches, then build the selected configuration with recorded metadata.

Figure 5: The JupyterLite environment hosting the transformation notebooks. The left panel shows the file system browser with the notebooks listed, the right panel shows the editor with the Introduction notebook open. The Introduction notebook contains the index and allows to navigate to the other ones. The notebooks are organized by the M-CODE [ 21 ] tag for quick navigation.

Figure 6: Example material transformation notebook opened in JupyterLite. The notebook exposes editable input cells and executable workflow steps so that users can modify parameters, run the build procedure (via Run > Run All Cells menu item), and inspect intermediate results directly in the browser. The example shows the interface construction workflow. The input parameters contain the film and substrate slab configurations, the interfacial gap, and the lateral shift. The settings for the strain matching algorithm - e.g. the maximum area constraint - are also shown.

Figure 7: Final output produced by an example notebook in JupyterLite. The resulting structure can be previewed after execution, illustrating how the notebooks serve as interactive and reproducible templates for realistic structure generation. The example shows the final result of the Gr/Ni(001) interface construction workflow. The final structure is saved in the uploads folder and can be loaded back into the (other) notebook(s) environment for further analysis or reuse.

M-CODE	Name	Notes
P-3D-CRY	Load Material from Standata	Input: material identifier from the reference repository. Result: material structure. Example: Ni bulk.
P-3D-CRY	Create Supercell	Inputs: bulk crystal and replication matrix. Result: crystal supercell. Example: Si 3x3x3 supercell.
P-2D-MNL	Create Monolayer	Input: layered bulk crystal and cleavage settings. Result: isolated 2D monolayer. Example: graphene monolayer.
P-2D-SLB-S	Create Slab	Input: bulk crystal, Miller indices, layers, vacuum, termination. Result: a slab structure. Example: SrTiO 3 (110) slab with SrTiO termination.
P-1D-NWR	Create Nanowire	Input: bulk crystal, orientation, and cross-section settings. Result: finite-width nanowire. Example: Si(001) nanowire.
P-0D-NRB	Create Nanoribbon	Input: monolayer, cut direction, length, and width. Result: finite nanoribbon. Example: graphene armchair nanoribbon.

Table 1: The list of notebooks for material transformations available at jupyterlite.mat3ra.com . The notebooks are organized according to the M-CODE [ 21 ] tag corresponding to the resulting structures produced. The Name column includes the name of the notebook. Notes explain the (adjustable by user) input parameters and the example structure shown in the last column.

Materials Science

Materials science studies how composition, structure, processing, and defects determine material properties and how to design useful materials.

See all 301 papers

Robust and Efficient AI Frameworks for Scalable Material Design and Property Prediction
Sep 15, 2026Kishalay DasGraph SSL Materials Property Prediction
Property Prediction of Stacked Bilayer Materials: A Multimodal Learning Approach
May 31, 2026An Vuong, Minh-Hao Van, Chen Zhao +1Materials Property Prediction Materials Science
University of Arkansas · Baylor University
MatMind: A Structure-Activity Knowledge-Driven Generative Foundation Model for Materials Science
Jun 5, 2026Zhan'ao Yao, Boxuan Zhang, Jingyuan Shu +10Materials Property Prediction Crystal Structure Generation
State Key Laboratory of High Performance Ceramics, Shanghai Institute of Ceramics, Chinese Academy of Sciences, 1295 Dingxi Road, Shanghai, 200050, China. · Center of Materials Science and Optoelectronics Engineering, University of Chinese Academy of Sciences, Beijing, 100049, China. · State Key Laboratory of Multimodal Artificial Intelligence Systems, Institute of Automation, Chinese Academy of Sciences, 95 Zhongguancun East Road, Beijing, 100190, China. +5

Sep 15, 2026J/K move · Enter open · S save

Robust and Efficient AI Frameworks for Scalable Material Design and Property Prediction

Kishalay Das

Graph SSL Materials Property Prediction Crystal Structure Generation Multimodal Graph Learning Conditional Generative Modeling

Open paper AskPDF arXiv

This thesis develops robust and efficient AI frameworks for accelerating crystalline materials discovery by addressing both major stages of the materials-design pipeline: crystal property prediction and crystal structure generation. Motivated by the high computational cost of Density Functional Theory (DFT) and the limited availability of labeled materials data, the thesis explores graph representation learning, pretraining, multimodal learning, and generative modeling for scalable materials design. For property prediction, the thesis first introduces CrysXPP, which learns transferable crystal representations through unsupervised graph autoencoding, reducing dependence on large property-labeled datasets. It then proposes CrysGNN, a large-scale self-supervised graph pretraining framework that captures atomic connectivity, chemical attributes, and global structural information and transfers this knowledge to downstream property predictors through knowledge distillation. CrysMMNet further enriches crystal representations by jointly modeling graph structure and textual descriptions, thereby incorporating both local chemical and global structural knowledge. For crystal generation, the thesis introduces TGDMat, a text-guided joint diffusion framework that jointly models lattice parameters, atomic types, and atomic coordinates while incorporating textual structural knowledge during denoising. This enables the generation of more valid and stable periodic materials while also supporting conditional generation from natural-language descriptions. Overall, the thesis establishes a unified AI-based framework for data-efficient property prediction and controllable crystal generation, demonstrating how graph learning, multimodal representations, and generative models can reduce computational cost and improve the scalability of materials

Abstract

Materials Science

Explore similar work

Robust and Efficient AI Frameworks for Scalable Material Design and Property Prediction

Property Prediction of Stacked Bilayer Materials: A Multimodal Learning Approach

MatMind: A Structure-Activity Knowledge-Driven Generative Foundation Model for Materials Science