Mat3ra

The AI-ready collaborative platform driving the #DigitalTransformation of #materials and #chemicals R&D through #modeling and #machinelearning.

[Mat3ra @Mat3ra_com]

Twitter

Welcome to the Era of Digitally-Engineered Mat3rials!

Thank you @WSJ + @newsient for sharing the news about our $3M seed round, solid growth + how our platform accelerates #RnD of #materials and #chemicals!

https://t.co/zGkMJVlk3m

[Mat3ra @Mat3ra_com]

📐 𝗥𝗲𝗹𝗲𝗮𝘀𝗲 𝗛𝗶𝗴𝗵𝗹𝗶𝗴𝗵𝘁: 𝗕𝗮𝗻𝗱 𝗢𝗳𝗳𝘀𝗲𝘁𝘀 𝗮𝘁 𝗦𝗲𝗺𝗶𝗰𝗼𝗻𝗱𝘂𝗰𝘁𝗼𝗿 𝗜𝗻𝘁𝗲𝗿𝗳𝗮𝗰𝗲𝘀

Release 2026.10.29 adds two new examples that calculate the valence band offset at a semiconductor heterojunction with the platform's valence band offset workflow.

Each one builds a strained superlattice and the matching bulk references, then lines up the bulk valence band maxima through the average electrostatic potential across the interface:
▪️ Si/Ge (001), following Van de Walle and Martin (Phys. Rev. B, 1986): 0.765 eV, against 0.74 eV in the publication
▪️ GaAs/AlAs (110), following Hinuma et al. (Phys. Rev. B, 2014): 0.471 eV, against the published 0.45 eV

Explore Release 2026.10.29: https://t.co/Hh0t0rYvld

👉 Try it for free at https://t.co/PyKuoYljVO

#Mat3ra #BandOffset #Semiconductors #DFT #Heterojunction #MaterialsScience

[Mat3ra @Mat3ra_com]

⚛️ Mat3rials 3xplorer.

𝗜𝗻𝘁𝗲𝗿𝗳𝗮𝗰𝗲 𝗯𝗲𝘁𝘄𝗲𝗲𝗻 𝗚𝗮𝗹𝗹𝗶𝘂𝗺 𝗡𝗶𝘁𝗿𝗶𝗱𝗲 𝗮𝗻𝗱 𝗔𝗹𝘂𝗺𝗶𝗻𝘂𝗺 𝗢𝘅𝗶𝗱𝗲 (𝗚𝗮𝗡(𝟬𝟬𝟭)/𝗔𝗹𝟮𝗢𝟯(𝟬𝟬𝟭)), 𝘀𝘁𝗿𝗮𝗶𝗻-𝗺𝗮𝘁𝗰𝗵𝗶𝗻𝗴 𝘀𝘂𝗽𝗲𝗿𝗰𝗲𝗹𝗹.

Available online in our materials bank:
https://t.co/DC6nTCcSNU

#mat3rials3xplorer #materialsscience #mat3ra #Semiconductors #GaN

[Mat3ra @Mat3ra_com]

🔌 𝗥𝗲𝗹𝗲𝗮𝘀𝗲 𝗛𝗶𝗴𝗵𝗹𝗶𝗴𝗵𝘁: 𝗦𝗰𝗵𝗼𝘁𝘁𝗸𝘆 𝗕𝗮𝗿𝗿𝗶𝗲𝗿 𝗛𝗲𝗶𝗴𝗵𝘁

Release 2026.10.29 adds a new Quantum ESPRESSO workflow that calculates the Schottky barrier height at a metal/semiconductor contact.

It shares the electrostatic potential steps with the band offset workflow and lines up the metal Fermi level with the semiconductor band edges through the averaged potential across the interface. The job page reports the p-type and n-type barrier heights as separate results.

Explore Release 2026.10.29: https://t.co/Hh0t0rYvld

👉 Try it for free at https://t.co/PyKuoYljVO

#Mat3ra #SchottkyBarrier #Semiconductors #DFT #QuantumESPRESSO #MaterialsScience

[Mat3ra @Mat3ra_com]

🚀 𝗠𝗲𝗲𝘁 𝗤𝘂𝗮𝗻𝘁𝘂𝗺 𝟯𝘅𝗽𝗹𝗼𝗿𝗲𝗿 (𝗤𝟯): 𝗮𝗻 𝗔𝗜-𝗻𝗮𝘁𝗶𝘃𝗲 𝗲𝗻𝗴𝗶𝗻𝗲 𝗳𝗼𝗿 𝗾𝘂𝗮𝗻𝘁𝘂𝗺 𝗺𝗮𝘁𝗲𝗿𝗶𝗮𝗹𝘀 𝗺𝗼𝗱𝗲𝗹𝗶𝗻𝗴

A quantum simulation engine written from scratch by AI agents, under the direction and supervision of domain scientists. Developed in weeks rather than years, and open source (Apache 2.0) as part of https://t.co/y0ccElQz9E.

What Q3 does:
⚛️ Plane-wave pseudopotential DFT: norm-conserving, ultrasoft and PAW, with spin polarization, Hubbard U, dispersion corrections, forces, stresses and structural relaxation
✅ Reproduces established reference implementations and runs faster on a single core in representative benchmarks
🖥️ Scales across GPU-accelerated supercomputers
🔁 Runs unmodified Quantum ESPRESSO pw.x inputs and reads UPF pseudopotentials

Mat3ra is part of the team behind the effort, with co-authors from Berkeley Lab/NERSC, UC Berkeley, Stanford, SLAC, USC, Johns Hopkins, AFRL and NVIDIA.

👉 Read more:
https://t.co/VA42SN9ChF

📄 Preprint: https://t.co/OHazB1CHLR
💻 Code: https://t.co/wbB2tvHlAA

#Mat3ra #Quantum3xplorer #CAMSAI #DFT #OpenSource #AIforScience #MaterialsScience

[Mat3ra @Mat3ra_com]

🚀 𝗠𝗮𝘁𝟯𝗿𝗮 𝗣𝗹𝗮𝘁𝗳𝗼𝗿𝗺 𝗥𝗲𝗹𝗲𝗮𝘀𝗲 𝟮𝟬𝟮𝟲.𝟭𝟬.𝟮𝟵

Our latest platform release introduces new capabilities for computational materials research.

Key additions include:
▪️ New workflow: Schottky barrier height at metal/semiconductor contacts
▪️ Valence band offsets for Si/Ge and GaAs/AlAs interfaces
▪️ Graphene on substrates, including p-doping on α-quartz
▪️ Point defects in graphene and GaN
▪️ More examples that reproduce published results: SrTiO₃(011) surface energies and a Pt₄ cluster on monolayer MoS₂

The release also includes interface improvements, bug fixes, and updates for developers.

📄 Also new: our paper "AI-ready design of realistic 2D materials and interfaces with Mat3ra-2D"
https://t.co/yqAY3NrENw

Explore the complete release: https://t.co/Hh0t0rYvld

👉 Try it for free at https://t.co/PyKuoYljVO

#Mat3ra #PlatformRelease #MaterialsScience #ComputationalMaterials #DFT #Semiconductors #Graphene

[Mat3ra @Mat3ra_com]

⚡Newsletter 2026/10. Platform Release and News! - https://t.co/KQPGzOBXwb

[Mat3ra @Mat3ra_com]

⚛️ Mat3rials 3xplorer.

𝗗𝗶𝘀𝘀𝗼𝗰𝗶𝗮𝘁𝗶𝘃𝗲 𝗮𝗱𝘀𝗼𝗿𝗽𝘁𝗶𝗼𝗻 𝗼𝗳 𝗢𝘅𝘆𝗴𝗲𝗻 𝗼𝗻 𝗮 𝗣𝗹𝗮𝘁𝗶𝗻𝘂𝗺 (𝟭𝟭𝟭) 𝘀𝘂𝗿𝗳𝗮𝗰𝗲 (𝗣𝘁(𝟭𝟭𝟭)–𝟮𝗢), 𝗿𝗲𝗹𝗮𝘅𝗲𝗱 𝟯×𝟯×𝟭 𝘀𝗹𝗮𝗯 𝘀𝘂𝗽𝗲𝗿𝗰𝗲𝗹𝗹.

Available online in our materials bank:
https://t.co/9BZghAM0R6

#mat3rials3xplorer #materialsscience #mat3ra #Catalysis #SurfaceScience

[Mat3ra @Mat3ra_com]

😄 Materials Meme!

[Mat3ra @Mat3ra_com]

⚡Newsletter 2026/10. Platform Release and News! - https://t.co/knIWxsUrIl

[Mat3ra @Mat3ra_com]

📄 𝗡𝗼𝘄 𝗮𝘃𝗮𝗶𝗹𝗮𝗯𝗹𝗲 𝗼𝗻 𝗡𝗮𝘁𝘂𝗿𝗲.𝗰𝗼𝗺: 𝗔𝗜-𝗿𝗲𝗮𝗱𝘆 𝗱𝗲𝘀𝗶𝗴𝗻 𝗼𝗳 𝗿𝗲𝗮𝗹𝗶𝘀𝘁𝗶𝗰 𝟮𝗗 𝗺𝗮𝘁𝗲𝗿𝗶𝗮𝗹𝘀 𝘄𝗶𝘁𝗵 𝗠𝗮𝘁𝟯𝗿𝗮-𝟮𝗗

AI/ML models for materials are mostly trained on ideal bulk crystals, while real devices are dominated by surfaces, interfaces and defects. Realistic structures have to be built and documented, not just picked from a database.

Mat3ra-2D is our open-source framework for generating realistic 2D materials, slabs, interfaces, defects and disorder.

What's inside:
🧱 Define → Refine → Build workflows: every structure keeps its provenance, including the source materials and build parameters (Miller indices, layers, vacuum, termination, strain, shift)
🌐 23 reusable notebooks that run in any browser, no installation: https://t.co/3hktocdoG8
📦 Representative structures in the public Materials Explorer (https://t.co/tqJ4ONcWGJ), archived on Figshare: https://t.co/4dNK6oTrn8
🔗 Open source, interoperable with pymatgen and ASE

It is the structure-generation layer that comes before property calculations and ML training.

Authors: Vsevolod Biryukov, Kamal Choudhary (Johns Hopkins University) and Timur Bazhirov. Supported in part by NIST. Open access.

👉 Read the paper:
https://t.co/yqAY3NrENw

#MaterialsScience #OpenScience #AIforScience #Mat3ra

[Mat3ra @Mat3ra_com]

💎 𝗠𝗮𝘁𝗲𝗿𝗶𝗮𝗹 𝗼𝗳 𝘁𝗵𝗲 𝗠𝗼𝗻𝘁𝗵: 𝗚𝗮𝗔𝘀(𝟬𝟬𝟭)/𝗔𝗹𝗔𝘀(𝟬𝟬𝟭) 𝗜𝗻𝘁𝗲𝗿𝗳𝗮𝗰𝗲

This model represents a strain-matched interface between two closely lattice-matched III-V semiconductors, combining the GaAs(001) and AlAs(001) surfaces in a single supercell. GaAs/AlAs heterostructures are a foundational system in semiconductor physics, long used to study band alignment, quantum confinement, and superlattice behavior at the atomic scale.

The GaAs(001)/AlAs(001) system can be used to explore structure–property relationships in III-V semiconductor heterostructures and to study how atomic-scale interface configurations influence electronic and optoelectronic device behavior — relevant to applications such as high-electron-mobility transistors (HEMTs), quantum-well lasers, and distributed Bragg reflectors.

Explore this featured GaAs/AlAs interface model on the Mat3ra platform:
https://t.co/Z1pmrSSMDB

#MaterialOfTheMonth #Semiconductors #GaAs #IIIV #MaterialsScience #Mat3ra

[Mat3ra @Mat3ra_com]

⚛️ Mat3rials 3xplorer.

𝗜𝗻𝘁𝗲𝗿𝗳𝗮𝗰𝗲 𝗯𝗲𝘁𝘄𝗲𝗲𝗻 𝗧𝗶𝘁𝗮𝗻𝗶𝘂𝗺 𝗡𝗶𝘁𝗿𝗶𝗱𝗲 𝗮𝗻𝗱 𝗦𝗶𝗹𝗶𝗰𝗼𝗻 (𝗧𝗶𝗡(𝟬𝟬𝟭)/𝗦𝗶(𝟬𝟬𝟭)), 𝘀𝘁𝗿𝗮𝗶𝗻-𝗺𝗮𝘁𝗰𝗵𝗶𝗻𝗴 𝘀𝘂𝗽𝗲𝗿𝗰𝗲𝗹𝗹.

Available online in our materials bank:
https://t.co/bqdo6Baf0t

#mat3rials3xplorer #materialsscience #mat3ra #Semiconductors #DFT

[Mat3ra @Mat3ra_com]

🚀 𝗠𝗮𝘁𝟯𝗿𝗮 𝗮𝘁 𝘁𝗵𝗲 𝗚𝗲𝗻𝗲𝘀𝗶𝘀 𝗠𝗶𝘀𝘀𝗶𝗼𝗻 𝗛𝗮𝗰𝗸𝗮𝘁𝗵𝗼𝗻, 𝗮𝗹𝗼𝗻𝗴𝘀𝗶𝗱𝗲 𝗚𝗼𝗼𝗴𝗹𝗲, 𝗔𝗪𝗦 𝗮𝗻𝗱 𝗢𝗽𝗲𝗻𝗔𝗜

Three days. One national lab. Some of the biggest names in tech, all working on one goal: AI that accelerates science.

From October 14 to 16, Argonne National Laboratory hosted the Genesis Mission Hackathon, and Mat3ra took part as an Industrial Partner, alongside Google, AWS, Microsoft, OpenAI and IBM Research.

What made it exciting:
⚛️ Genesis Mission teams worked with technical experts to bring their scientific workflows onto the DOE Genesis Mission Platform
🧪 Materials discovery is one of the places where this pays off, and we have three selected Genesis Mission projects, including AlphaFilm for thin film discovery
🤝 Our Founder and CEO, Timur Bazhirov, represented Mat3ra in the room

AI, data and computing are coming together to speed up science, and Mat3ra is right in the middle of it.

👉 Read the full story:
https://t.co/kUxTvkXeVu

#GenesisMission #DOE #MaterialsScience #AIforScience #Mat3ra

[Mat3ra @Mat3ra_com]

🚀 𝗪𝗲'𝗿𝗲 𝗛𝗶𝗿𝗶𝗻𝗴 𝗮𝘁 𝗠𝗮𝘁𝟯𝗿𝗮

Let's make Iron Man's Digital Materials R&D experience a reality together.

☀️ Computational Materials Scientist
https://t.co/OUCIrLghxi
Bridge materials/chemistry, data science, and computer science: apply nanoscale modeling across large sets of materials, surfaces, and reaction pathways; organize simulation and experimental data; build AI approaches on top of it.

🧑‍🤝‍🧑 Community Evangelist
https://t.co/gH1dXQ6a1U
For outgoing scientists/engineers who want to grow a vibrant global community: identify pathways for community growth, prepare case studies and technical content, deliver it, and measure KPIs.

All open positions: https://t.co/tAcEbpTLyt

#Mat3ra #Hiring #JobOpenings #MaterialsScience #ComputationalMaterials

[Mat3ra @Mat3ra_com]

🔬 𝗨𝗽𝗰𝗼𝗺𝗶𝗻𝗴 𝗪𝗲𝗯𝗶𝗻𝗮𝗿: 𝗘𝗹𝗲𝗰𝘁𝗿𝗼𝗻𝗶𝗰𝘀 — 𝗕𝗮𝗻𝗱 𝗢𝗳𝗳𝘀𝗲𝘁 + 𝗦𝗰𝗵𝗼𝘁𝘁𝗸𝘆 𝗕𝗮𝗿𝗿𝗶𝗲𝗿

Join us October 28 for Mat3ra 2D, Season 2, Episode VII: Electronics — Band Offset + Schottky Barrier.

This session demonstrates how to compute electronic interface properties — band offset and Schottky barrier — using Quantum ESPRESSO from JupyterLab notebooks on the Mat3ra platform. We'll cover practical workflows for obtaining and interpreting band alignment and carrier-injection barriers from atomistic simulations, with examples including HfO₂(001)/Si(001) and MoS₂/Au(111) interfaces.

Presenter: Vsevolod (Seth) Biryukov, Scientific Software Engineer
Host: Timur Bazhirov
Language: English

🗓️ October 28, 2026 — 07:30 PDT / 10:30 EDT / 14:30 GMT / 15:30 CET / 16:30 EET / 20:00 India / 22:30 Singapore / 23:30 Japan
📍 Online via Zoom

Register to attend: https://t.co/rTWilpzJkw

#Mat3ra #Webinar #MaterialsScience #DFT #QuantumESPRESSO #Semiconductors

[Mat3ra @Mat3ra_com]

📚 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵 𝗪𝗲 𝗟𝗼𝘃𝗲: 𝗘𝗹𝗲𝗰𝘁𝗿𝗼𝗻 𝗜𝗺𝗮𝗴𝗶𝗻𝗴 𝗶𝗻 𝗦𝗲𝗺𝗶𝗰𝗼𝗻𝗱𝘂𝗰𝘁𝗼𝗿𝘀 & 𝗗𝗶𝘀𝗰𝗼𝘃𝗲𝗿𝘆 𝗟𝗼𝗼𝗽

𝗔 𝗡𝗲𝘄 𝗖𝗹𝗼𝘀𝗲-𝗨𝗽 𝗩𝗶𝗲𝘄 𝗼𝗳 𝗘𝗹𝗲𝗰𝘁𝗿𝗼𝗻𝘀 𝗜𝗻𝘀𝗶𝗱𝗲 𝗦𝗲𝗺𝗶𝗰𝗼𝗻𝗱𝘂𝗰𝘁𝗼𝗿𝘀
Berkeley Lab combined atomic-scale imaging (scanning tunneling microscopy) with Quantum Monte Carlo simulations to directly observe how electrons interact with defects in 2D semiconductors like molybdenum diselenide — some defects act like "potholes," others like "speed bumps," as electrons shift between an ordered Wigner solid and a disordered Fermi liquid state. Published in Nature.
Authors: Mike Crommie, Feng Wang, et al. (UC Berkeley / Berkeley Lab, with Flatiron Institute, UC, University of New Mexico, Hofstra University, Arizona State University, and NIMS Japan)
🔗 Source: https://t.co/Dks9YXeeEK

──────── • ────────

𝗜𝗻𝘁𝗿𝗼𝗱𝘂𝗰𝗶𝗻𝗴 𝗗𝗶𝘀𝗰𝗼𝘃𝗲𝗿𝘆 𝗟𝗼𝗼𝗽
A new AI startup founded by four Google AI veterans — Jeff Dean, Sanjay Ghemawat, Quoc Le, and Oriol Vinyals — aiming to automate the experimental loop of scientific and engineering research, running thousands of experiments in parallel with frontier AI models and compute infrastructure.
Authors: Jeff Dean, Sanjay Ghemawat, Quoc Le, Oriol Vinyals
🔗 Source: https://t.co/AlTzPgrwdU

#AIforScience #MaterialsScience #Semiconductors #Research

[Mat3ra @Mat3ra_com]

⚛️ Mat3rials 3xplorer.

𝗜𝗻𝘁𝗲𝗿𝗳𝗮𝗰𝗲 𝗯𝗲𝘁𝘄𝗲𝗲𝗻 𝗦𝗶𝗹𝗶𝗰𝗼𝗻 𝗖𝗮𝗿𝗯𝗶𝗱𝗲 𝗮𝗻𝗱 𝗚𝗮𝗹𝗹𝗶𝘂𝗺 𝗡𝗶𝘁𝗿𝗶𝗱𝗲 (𝗦𝗶𝗖(𝟬𝟬𝟭)/𝗚𝗮𝗡(𝟬𝟬𝟭)), 𝘀𝘁𝗿𝗮𝗶𝗻-𝗺𝗮𝘁𝗰𝗵𝗶𝗻𝗴 𝘀𝘂𝗽𝗲𝗿𝗰𝗲𝗹𝗹.

Available online in our materials bank:
https://t.co/UuyoMf7SCA

#mat3rials3xplorer #materialsscience #mat3ra #Semiconductors #DFT

[Mat3ra @Mat3ra_com]

🗓️Upcoming Webinar on Band Offset and Schottky Barrier, October 28th, Tokyo Materials Cafe, and AI 2 Science Happy Hour - https://t.co/TnCawBBZT8

[Mat3ra @Mat3ra_com]

📰 𝗜𝗻𝗱𝘂𝘀𝘁𝗿𝘆 𝗡𝗲𝘄𝘀: 𝗟𝗶𝗴𝗵𝘁-𝗦𝘁𝗲𝗲𝗿𝗶𝗻𝗴 𝗖𝗵𝗶𝗽𝘀 & 𝗔𝗜-𝗖𝗼𝗻𝘁𝗿𝗼𝗹𝗹𝗲𝗱 𝗙𝘂𝘀𝗶𝗼𝗻

𝗖𝗮𝗹𝘁𝗲𝗰𝗵'𝘀 𝘁𝗶𝗻𝘆 𝗻𝗲𝘄 𝗰𝗵𝗶𝗽 𝗰𝗮𝗻 𝘀𝘁𝗲𝗲𝗿 𝗹𝗶𝗴𝗵𝘁 𝗶𝗻 𝟳𝟰 𝗾𝘂𝗮𝗱𝗿𝗶𝗹𝗹𝗶𝗼𝗻𝘁𝗵𝘀 𝗼𝗳 𝗮 𝘀𝗲𝗰𝗼𝗻𝗱
A new all-optical device uses a nanoscale silicon metasurface — pillars smaller than the laser's wavelength — to redirect one light beam with another via the optical Kerr effect. The result: light steering nearly a million times faster than conventional methods, opening the door to faster photonic communications, computing, and sensing.
🔗 Source: https://t.co/uZMjIUi5Av

──────── • ────────

𝗔𝗜 𝗰𝗮𝗻 𝗻𝗼𝘄 𝗰𝗼𝗻𝘁𝗿𝗼𝗹 𝗳𝘂𝘀𝗶𝗼𝗻 𝗽𝗹𝗮𝘀𝗺𝗮 𝗳𝗮𝘀𝘁𝗲𝗿 𝘁𝗵𝗮𝗻 𝗵𝘂𝗺𝗮𝗻𝘀 𝗰𝗮𝗻 𝗿𝗲𝗮𝗰𝘁
Princeton researchers built PACMAN, a machine-learning framework that monitors and controls fusion plasma roughly 50x faster than a human operator can react. Tested on the DIII-D tokamak, it predicted a damaging instability 200 milliseconds in advance and prevented it — a meaningful step toward stable, continuous fusion reactions.
🔗 Source: https://t.co/HaPMm93HFV

#IndustryNews #Photonics #FusionEnergy #AI #MaterialsScience

[Mat3ra @Mat3ra_com]

🗓️Upcoming Webinar on Band Offset and Schottky Barrier, October 28th, Tokyo Materials Cafe, and AI 2 Science Happy Hour - https://t.co/3aSUN1Y4dq