Elastic Tensors from Pairwise Energy Frameworks in Molecular Crystals The mechanical properties of molecular crystals are fundamentally important in their industrial applications across pharmaceuticals, agrochemicals, energetic materials and other areas. Despite this, complete measurement or even computational prediction of elastic tensors for molecular crystals is anything but commonplace. The absence of rapid, reliable and broadly applicable methods in this endeavor frequently leads chemists to rely on intuitive ideas and examination of pairwise intermolecular interactions such as hydrogen- or halogen-bonds in order to rationalize the mechanical behavior of molecular crystals. Such perspectives are widespread in contemporary literature, but the extent to which these notions yield reliable and quantifiable insight is itself relatively unexplored. We propose a simple approximation, the Equilibrium Pairwise Model (EPM), compatible with any method to predict intermolecular interaction energies, that directly and efficiently yields an estimate of the complete elastic tensor. The protocol can be performed for any given molecular crystal structure, even those directly from experiment (i.e., without geometry optimization), and is guaranteed by construction to yield a positive-definite result─in contrast to conventional methods where computing valid elastic tensors for molecular crystals can prove challenging even for well-established and otherwise accurate model chemistries. We examine the accuracy of this protocol, along with other classical and contemporary methods, against experiment and periodic (plane-wave) density functional theory calculations to assess their reliability and accuracy. Through examination of the failures and successes, we aim to provide chemical insight into the kinds of materials where the model and, more broadly, thinking based on pairwise intermolecular interactions can reliably explain mechanical or other material properties and where they should be avoided.

How well can you predict mechanical properties (elastic tensors) for molecular crystals *just* looking at pairwise interaction energies? Surprisingly well! With a big asterisk... Read all the ways it fails and where it works in JCTC now:

pubs.acs.org/doi/full/10....

I think Fermions are probably still cursed too tbh

Come work with my colleagues at Curtin! New postdoc opportunity modelling leaching of precious metals with glycine, a challenging intersection of a lot of chemical problems - particularly for computational chemistry!
staff.curtin.edu.au/job-vacancie...

Erik Svensson Grape - awardee - Struchkov Prize 2026

I am incredibly honored to recieve this prize!

Greatly looking forward to attending @iucr2026.bsky.social

-
I have many people to thank for all the learning and exciting research over these past years. First and foremost though, is my crystallography mentor and great friend
@keninge.bsky.social

Awesome, congratulations!

Looking forward to checking these out!

Just published in JOSS: 'Open Computational Chemistry (OCC) - A portable software library and program for quantum chemistry and crystallography' https://doi.org/10.21105/joss.09609

Screenshot of the matbench discovery leaderboard as of 14.01.2026, showing a PET based model in the top position

Not going to make a big deal out of a benchmark table, but PET just got the top spot on matbench-discovery.materialsproject.org. And don't be fooled by the huge parameters count, it's faster and can handle larger structures than eSEN-30M 🚀. Kudos to 🧑‍🚀 Filippo, Arslan and Paolo!

Just archived today - "RobInHood: A Robotic Chemist in a Fume Hood" - where we squeezed a robot arm into a standard-format fume hood to carry out a range of chemistry operations. Potentially scalable to any lab that has fume hoods. Well done Louis, Franciso et al. chemrxiv.org/engage/chemr... 🤖🧪

One more week to apply to work with me - trying to predict the crystals on Titan!

A title image describing a recently publish scientific paper. The title is To help publicize your article, Crystal structure of diacetylene unveiled by X-ray and neutron diffraction, Raman spectroscopy and periodic DFT' and there is a brief description of the paper that reads 'The crystal structure and thermal expansion behaviour of diacetylene were determined via diffraction methods and supplemented by spectroscopic techniques. The similarities to acetylene's structure and implications for Titan in situ exploration are discussed.' There is also an image that depicts the crystal structure and its unit cell

📢 One less molecular mystery - the crystal structure of diacetylene is solved, published in @iucrj.iucr.org today! Fantastic work by Larissa, and also I think will be the first structure deposit in @ccdc.cam.ac.uk from wombat! Read the OA paper here journals.iucr.org/m/issues/202...

Great to see the work in Vitaly’s group on descriptors for comparing crystal structures get recognised, and in an interesting article to boot!

Super excited to get this preprint online with @graemeday.bsky.social !

With SAUCE, we significantly improve on the efficiency of random structure searching for crystals containing multiple symmetry-independent molecules!

chemrxiv.org/engage/chemr...
#compchem #csp

Still a few more weeks to apply for a sunny postdoc opportunity with a pretty cool guy (me) in a really warm place (Perth) and predict the organic crystals on a pretty cold place (Titan), in space. staff.curtin.edu.au/job-vacancie...

Congratulations!

Elastic Tensors from Pairwise Energy Frameworks in Molecular Crystals Mechanical properties of molecular crystals are fundamentally important to their in- dustrial utility as pharmaceuticals or agrochemicals, energetic materials and more. Yet, complete measurements of e...

Excited to share our new pre-print, with features sneakily already available in CrystalExplorer + OCC: Elastic Tensors from Pairwise Energy Frameworks in Molecular Crystals doi.org/10.26434/che...

High-Pressure Crystallization and Compression of Isobutyronitrile The structural response of a high-pressure phase of isobutyronitrile up to 6.12 GPa was studied by using high-pressure single-crystal X-ray diffraction, periodic density functional theory (DFT), and C...

New paper out, High-pressure crystallisation of Isobutyronitrile pubs.acs.org/doi/10.1021/..., nice results from the UWA and Synchrotron team that show this small molecule crystallises into the same structure as at low temperatures. There's some lovely calculations that explain why this may be.

👀

Thanks! I'm looking forward to getting my teeth into it. Unfortunately, at least for my research where it's overwhelmingly C, H, and N, titanium isn't a particularly important element :( maybe I'll find a way...

Still another month or so to apply to work with me: predicting the crystals and their properties on Titan!

Actually their webgpu support is better than I'd thought, I was wrong. I'll get it working if I can!

As far as I know ggml doesn’t have webgpu acceleration, but in principle it should absolutely be possible as there are plenty of other wasm and webgpu combinations! Yes I think given my current implementation is single core only and quite slow in wasm it’s not the memory that’ll bottleneck first!

Small enough to conveniently run directly in the browser via WASM… peterspackman.github.io/mlip.cpp/

Great news! Can vouch that it’s a very impressive and stable (and a small model at only about 14MB).

PET-MAD as a lightweight universal interatomic potential for advanced materials modeling - Nature Communications PET-MAD is a fast and lightweight universal machine-learning potential, trained on a small but diverse dataset, that delivers near-quantum accuracy in atomistic simulations for both organic and inorga...

📢 PET-MAD is here! 📢 It has been for a while for those who read the #arXiv, but now you get it preciously 💸 typeset by @natcomms.nature.com Take home: unconstrained architecture + good train set choices give you fast, accurate and stable universal MLIP that just works™️ www.nature.com/articles/s41...

This is all using the PET-MAD model (github.com/lab-cosmo/pe...). Wouldn't call my code production ready or anything but it's nice to see what's possible in a web browser.

The latest in my wasm experiments is getting an MLIP running, so you can watch some molecules (or crystals, if you have more patience) wiggle, all running locally in your web browser on your phone, computer, tablet, wherever! peterspackman.github.io/mlip.cpp/

I've got a post-doctoral research position available at Curtin University (Western Australia) starting 2026.
The project involves computational prediction of crystal structures, their growth, and their properties for organic minerals on Titan (moon of Saturn).

staff.curtin.edu.au/job-vacancie...

If that sounds like something you or someone you know would be interested in, feel free to reach out to peter.spackman@curtin.edu.au with any questions you might have. Or chat to me here!

Broadly, interactions of small hydrocarbons, nitriles etc. improved sampling methods for crystal structure prediction, and modelling to predict stabilities in growth conditions.

We've got a really nice group of people here at Curtin - and I think this is great opportunity to join us!