Volker Deringer

🎓 Applications are open for the IMAT Centre for Doctoral Training at @oxfordchemistry.bsky.social & @oxfordmaterials.bsky.social!

Among many exciting topics, have a look at our projects P10 and P13, both combining MLIPs and #compchem with close experimental collaborations: imatcdt.chem.ox.ac.uk

Congratulations Kim!! 🙂

A selfie of Volker and team members in front of the conference venue on Wednesday evening

#PsiK2025 selfie! Super to see the community so active & so much exciting #compchem research going on. And of course delighted to be here with members of our group 🙂

metatensor logo

metatomic logo

🚨 #machinelearning for #compchem goodies from our 🧑‍🚀 team incoming! After years of work it's time to share. Go check arxiv.org/abs/2508.15704 and/or metatensor.org to learn about #metatensor and #metatomic. What they are, what they do, why you should use them for all of your atomistic ML projects 🔍.

An automated framework for exploring and learning potential-energy surfaces - Nature Communications Machine learning is revolutionising materials modelling but requires high-quality training data. Here, the authors introduce autoplex, an open framework automating exploration and fitting of potential...

🧪🤖 Our first paper on autoplex is published! We describe an automated #compchem framework for building MLIP training datasets, and show a range of application examples. A pleasure to collaborate on this with @molecularxtal.bsky.social & team. Thank you everyone! doi.org/10.1038/s414...

📢pls share
We are hiring! New opening for a W2 Professor in "experimental inorganic chemistry" @unibonn.bsky.social
Deadline Oct. 10 t.co/EqMLA0fCuC

The table-of-contents image for the paper mentioned in the post – showing an atomistic structural model of graphene oxide on the left, and stress–strain plots on the right

Mechanical properties of graphene oxide from machine-learning-driven simulations – now online in ChemComm (@chemcomm.rsc.org)! In this #compchem study, we explore the links between atomistic structure and mechanical behaviour in GO. Congratulations Zak & Bowen 🙂

Read more: doi.org/10.1039/D5CC...

Broadband transient full-Stokes luminescence spectroscopy - Nature A high-sensitivity, broadband, transient, full-Stokes spectroscopy setup is demonstrated, which can detect quickly varying small signals from chiral emitters.

Excited to share in @nature.com today: Broadband transient full-Stokes luminescence spectroscopy - detecting the most subtle changes in light polarization over time with unprecedented sensitivity. Grateful for the team that made this possible!😊 www.nature.com/articles/s41... #chirality #light

Read more about our MLIP distillation preprint in John’s thread! 🙂 #compchem #chemsky

A photo of Georgi, Arun, Johana, and Cecilia, taken outdoors on a sunny day

Congratulations to the group‘s MChem students Georgi, Arun, Johana, and Cecilia on completing their projects & theses! They covered a range of topics across #compchem, ML, and materials chemistry applications – well done and thank you everyone 😀

Now published in @chemicalscience.rsc.org and highlighted as a #ChemSciPicks. Great work by @ffmmgg.bsky.social. Collab with @aicooper.bsky.social

A Universal Foundation Model for Transfer Learning in Molecular Crystals

#compchemsky #chemsky
@unisouthampton.bsky.social @liverpooluni.bsky.social

🚀 After two+ years of intense research, we’re thrilled to introduce Skala — a scalable deep learning density functional that hits chemical accuracy on atomization energies and matches hybrid-level accuracy on main group chemistry — all at the cost of semi-local DFT ⚛️🔥🧪🧬

🎉 DFT-accurate, with built-in uncertainty quantification, providing chemical shielding anisotropy - ShiftML3.0 has it all! Building on a successful @nccr-marvel.bsky.social-funded collaboration with LRM🧲⚛️, it just landed on the arXiv arxiv.org/html/2506.13... and on pypi pypi.org/project/shif...

Distillation of atomistic foundation models across architectures and chemical domains Machine-learned interatomic potentials have transformed computational research in the physical sciences. Recent atomistic `foundation' models have changed the field yet again: trained on many differen...

Distilling atomistic foundation models! ⚗️🧪🤖 In this #compchem preprint, we describe a general (“architecture-agnostic”) approach to creating fast, application-specific MLIP models via synthetic data – led jointly by @jla-gardner.bsky.social & @dft-dutoit.bsky.social arxiv.org/abs/2506.10956

It is very fast! (In particular for those cases where we need hundreds of thousands of atoms and more – "realistic" polycrystalline samples are one example)

Machine-learning-driven modelling of amorphous and polycrystalline BaZrS$_{3}$ The chalcogenide perovskite material BaZrS$_{3}$ is of growing interest for emerging thin-film photovoltaics. Here we show how machine-learning-driven modelling can be used to describe the material's ...

BaZrS3 is an emerging solar-cell material ☀️ In a preprint led by @biancapasca.bsky.social, we describe an ML potential that can tackle the structural complexity of amorphous and polycrystalline BaZrS3. Very happy to see this online! Read more at arxiv.org/abs/2506.01517

Alchemy Frontier Fund 2025 now open (aichemy.ac.uk/aichemy-fron...). Apply for 2-year projects of up to £1.25m to advance the frontiers of AI for chemistry. Deadline = Sept 12th 2025. Webinar to launch the fund on June 18th. For information, email: funding@aichemy.ac.uk @ukri.org #EPSRC #AI

A photo of Volker, Litong, Shixuan, Natascia, and Bianca outside a building at KIT where the workshop was held

It was great to take part in the @cecamevents.bsky.social Flagship Workshop on “Virtual Materials Design” at @kit.edu this week – and to see Litong, Shixuan, Natascia, & @biancapasca.bsky.social present their #compchem research! 😀

Scheme of the GNN architecture of the FlashMD method.

📢 Running molecular dynamics with time steps up to 64fs for any atomistic system, from Al(110) to Ala2? Thanks to 🧑‍🚀 Filippo Bigi and Sanggyu Chong, with some help from Agustinus Kristiadis, this is not as crazy as it sounds. Let us briefly introduce FlashMD⚡ arxiv.org/html/2505.19...

Exploration of crystal chemical space using text-guided generative artificial intelligence - Nature Communications The vastness of chemical space makes discovering new materials challenging. Here, authors propose a generative AI model enabling crystal structures generation from textual descriptions, accelerating m...

The first release of our text-to-crystal model, Chemeleon, is out in @natcomms.nature.com 🌈

Paper: www.nature.com/articles/s41...
Code: github.com/hspark1212/c...

You can sample an inorganic structure in minutes on a laptop thanks to @hspark1212.bsky.social - #CompChem that gives me goosebumps!

A photo of Vancouver on a sunny day, showing a waterfront and high-rise buildings in the background

In Vancouver for a major conference on ceramic & glass technology! Looking forward to presenting some of our recent #compchem work, and to discussing what ML interatomic potentials can do in this exciting area 🙂

ceramics.org/event/16th-p...

graph-pes is John‘s open-source, all-round software package for fitting & fine-tuning graph-based ML interatomic potentials – do try it out, follow him for updates, and share! #chemsky #compchem 🧪

Feldmann Lab at EPFL is hiring at PhD and Postdoc Level in Spectroscopy and Nanomaterials

𝐖𝐄 𝐀𝐑𝐄 #𝐇𝐈𝐑𝐈𝐍𝐆 on the 𝐏𝐇𝐃 & 𝐏𝐎𝐒𝐓𝐃𝐎𝐂 level!
If you are interested in #spectroscopy & #materials science for spin-/optoelectronics, please consider joining our team. Details on the positions and how to apply:
www.feldmannlab.com/open-positions
Sharing with your network is greatly appreciated!🙏:) #EPFL

Great to see this preprint online – a data-driven study of an amorphous metal–organic framework, led by @tcnicholas.bsky.social – thanks to Tom and everyone 🙂 Comments very welcome! arxiv.org/abs/2503.24367

An experimental data library for the full CsPb(ClxBr1−x)3 compositional series A complete series of CsPb(ClxBr1−x)3 mixed-halide perovskites with x = 0–1 in small steps is reported, and their structural and optical properties characterised. A comparison of synthetic approaches s...

Proud to see @Kinga' Master’s project published: 𝐀𝐧 𝐞𝐱𝐩𝐞𝐫𝐢𝐦𝐞𝐧𝐭𝐚𝐥 𝐝𝐚𝐭𝐚 𝐥𝐢𝐛𝐫𝐚𝐫𝐲 𝐟𝐨𝐫 𝐭𝐡𝐞 𝐟𝐮𝐥𝐥 𝐂𝐬𝐏𝐛(𝐂𝐥𝐱𝐁𝐫𝟏−𝐱)𝟑 𝐜𝐨𝐦𝐩𝐨𝐬𝐢𝐭𝐢𝐨𝐧𝐚𝐥 𝐬𝐞𝐫𝐢𝐞𝐬, open-access!⁣⁣
Great #collaboration with co-supervisor & friend @vlderinger.bsky.social, and thanks to @chemcomm.rsc.org for the invitation!
pubs.rsc.org/en/content/a...

Signatures of paracrystallinity in amorphous silicon from machine-learning-driven molecular dynamics - Nature Communications Conflicting theories exist on the structure of amorphous silicon. Here the authors use machine-learning-driven molecular dynamics to show that amorphous Si can accommodate a degree of local paracrysta...

Amorphous silicon has a seemingly random structure – and yet there is more to it, as Louise demonstrates using #compchem & ML approaches. A great collaboration with @dadrabold.bsky.social! Read more in @naturecomms.bsky.social, openly available here: www.nature.com/articles/s41...

Exciting news for #chemsky - you can now follow all of your favourite @rsc.org journals on Bluesky! 🥳🧪⚗

go.bsky.app/QAaTTZ3

Chiral light–matter interactions in solution-processable semiconductors - Nature Reviews Chemistry Endowing chirality onto solution-processable semiconductors unlocks new bounds for their spin and optoelectronic properties. This Review discusses the synthesis, properties, characterization methods, ...

Happy to finally share our review on the broad (and quickly progressing!) field of chiral solution processable semiconductors! #ChemSky ⚗️🧪
www.nature.com/articles/s41...

Title and abstract of "Adaptive energy reference for machine-learning models of the electronic density of states" by Wei Bin How, Sanggyu Chong, Federico Grasselli, Kevin K. Huguenin-Dumittan, and Michele Ceriotti - published as Phys. Rev. Materials 9, 013802 (2025)

📢 Highlighted on #PhysRevMaterials (not yet on 🦋), a little, insightful gem by 🧑‍🚀 Wei Bin, Raymond, Fede & Kevin, relevant for all of you #machinelearning electronic structure properties. Band alignment matters, but it shouldn't, so here's how to deal with it with a self-aligning loss! A short 🧵...

Atomate2: Modular workflows for materials science High-throughput density functional theory (DFT) calculations have become a vital element of computational materials science, enabling materials screening, property database generation, and training of...

🤖 Interested in automated DFT or ab initio calculations for crystals or molecules?

atomate2 could be your package!

doi.org/10.26434/che...

#compchem