Release 2026.9.24

Release 2026.9.24

Release 2026.9.24 introduces multiple fixes and new capabilities. We highlight (1) a new example reproducing the registry, separation and work of adhesion of graphene on Ni(111); (2) a new tutorial calculating the formation energy of a boron vacancy in hexagonal boron nitride; and (3) new REST API endpoints for uploading files and changing the type of an entity set.

Feature

• [AX] Specific Example for Graphene/Ni(111) film XY position and total energy calculations
• [AX] SE for the Vacancy defect in h-BN: formation energy
• [AX] SE for the Vacancy-Interstitial defect pair in SnO: formation energy and DOS
• Add a REST API endpoint for uploading files (to dropbox)
• Add a REST API endpoint for changing an entity set type (ordered material sets)

Improvement

• Toggle Relaxation has Version and Build for the App
• [CLI] Update default version of python to 3.11

Bugfix

• Fix Cyrillic Г vs Greek Γ typo in persisted k-path/q-path point labels
• Fix JupyterLab workflow notebook URLs

Developers

• [AX] Create NB for Custom Python Calculation
• [Webinar] Thermodynamics II

Highlight 1: Graphene/Ni(111) Interface

A new notebook reproduces the structure and energetics of graphene on Ni(111) following the review by Dahal and Batzill (Nanoscale, 2014) and the calculations of Lahiri et al. (New J. Phys., 2011). It relaxes each candidate registry - fcc, hcp, hollow and bridge - to obtain the graphene-metal separation and the work of adhesion. A fast tier uses the MACE-MP machine-learned force field and runs in minutes; a precise tier submits platform jobs with the LDA functional used in the publication:

Highlight 2: Boron Vacancy in Hexagonal Boron Nitride

A new tutorial calculates the formation energy of a neutral boron vacancy in h-BN with Quantum ESPRESSO and compares the result against the QPOD database of quantum point defects in 2D materials (Bertoldo et al., npj Computational Materials, 2022). It builds on the existing tutorial that creates the defective structure:

Highlight 3: File Uploads via the REST API

The external REST API now supports uploading files to the account's object storage ("Dropbox"), which compute nodes read from. Text files travel inside the request, while larger or binary files - model checkpoints, archives, pseudopotentials - go to a signed URL straight to storage. The example notebooks use this to ship custom scripts and data files with a workflow. A second new endpoint changes the type of an entity set, for example to an ordered material set:

‍

Try the new functionality online at https://platform.mat3ra.com/

‍