New paper in the JCP Becke special issue. We show that Δ-learning can improve the transferability of MLIP potentials by combining an ANI-style MLIP with a DFTB3 baseline. The approach improves transition-state energetics and long-range interactions beyond the ML cutoff. doi.org/10.1063/5.03...

Δ-learning for transferable machine learning interatomic potentials

Machine-learning interatomic potentials (MLIPs) trained by directly learning the total interatomic interaction energies can suffer from limited transferability,

doi.org

Years ago, I gave a lecture on reaction dynamics and was terrified to learn Rudy Marcus was in the audience. To my great relief, he introduced himself beforehand and couldn't have been more gracious. He reminisced warmly about his Ottawa days at the Steacie Institute in the 1940s. A true visionary.

Caltech@caltech.edu · 2mo ago

Nobel laureate Rudy Marcus, Caltech's John G. Kirkwood and Arthur A. Noyes Professor of Chemistry, passed away on July 16, 2026. The prolific chemist was five days shy of his 103rd birthday. www.caltech.edu/about/news/c...

Although LLMs have caused a lot of damage to undergraduate education, there are some helpful aspects for instructions that I talk about in my latest blog post. teachphyschem.blogspot.com/2026/07/wher...

Where LLMs are Useful to Instructors

Large Language Models like ChatGPT, Claude, and Gemini have ruined some parts of university education, but they have also improved my teachi...

teachphyschem.blogspot.com

New post on my Teaching Physical chemistry blog: The Art of the Lecture Slide. teachphyschem.blogspot.com/2026/06/the-...

The Art of the Lecture Slide:

The part of university lecturing that I find most interesting is to figure out how to explain something. What background material is needed...

teachphyschem.blogspot.com

The last chapter of @zeldery.bsky.social 's PhD thesis is now out as a preprint. We train an NNP to predict the difference between DFT and DFTB, giving us near-DFT quality at DFTB cost. Improvement in accuracy, especially at long range. chemrxiv.org/doi/full/10....

New post to my blog Teaching Physical Chemistry on back to back problem sets and quizzes in physical chemistry tutorials. teachphyschem.blogspot.com/2026/04/spli...

Split Tutorials

I taught 2nd year physical chemistry for the first time this year. The course includes an instructor-led weekly tutorial in addition to thre...

teachphyschem.blogspot.com

An obituary for Axel Becke from The Globe and Mail (🎁 🔗): www.theglobeandmail.com/gift/e8c626f... #ChemSky #ChemChat

Scientist Axel Becke showed how to quantify the bonds tying matter together

Theoretical chemist’s insights are now embedded into computer software used to predict the behaviour of molecules from pharmaceuticals to construction materials

theglobeandmail.com

Congratulations to Dr. Phúc Tu for his excellent PhD defence.

Congratulations to Phuc for his new article "Long-Range Interactions in High-Dimensional Neural Network Potentials: A Benchmark Study for Small Organic Molecules", part of a special issue of JPCB, 'At the Cutting Edge of Theoretical and Computational Biophysics" pubs.acs.org/doi/full/10....

Another group summer BBQ in the books. Best wishes to Pohl and Chris on their next steps.

Happy to share Phuc's new preprint in collaboration with Jorg Behler on the integration of our MLXDM dispersion term into 4th generation charge equilibration NNPs chemrxiv.org/engage/chemr...

Long-Range Interactions in High-Dimensional Neural Network Potentials: A Benchmark Study for Small Organic Molecules

Many machine learning potentials (MLPs) rely on representations of the total energy in terms of the positions of the atoms in their local environment, using either a cutoff radius or a limited number ...

chemrxiv.org

In celebration on Canada Day, our group website has been on-shored to www.rowleygroup.ca through @whc.ca

I've written a new article on my log about adding machine learning to an undergraduate computational chemistry course. Dataset and codes are available on github. teachphyschem.blogspot.com/2025/05/intr...

Introducing Machine Learning to Undergraduate Computational Chemistry

Until this year, my computational chemistry courses only covered electronic structure and molecular simulation. Machine learning is now used...

teachphyschem.blogspot.com

Atmospheric chemist @alisonbain.bsky.social works as an assistant professor at @oregonstate.edu, where she measures aerosol particles to understand climate change. Read more: cen.acs.org/environment/... #CENT12 #chemsky 🧪

Opinion from LEK: Recruiting U.S. scientists is a fine idea – except we’re barely supporting Canadian scientists as it is /via @globeandmail www.theglobeandmail.com/opinion/arti...

Opinion: Recruiting U.S. scientists is a fine idea – except we’re barely supporting Canadian scientists as it is

Before we look to take advantage of fleeing American talent, let’s nurture the Canadian talent we have here by increasing R&D spending

theglobeandmail.com

Delighted to share our New and Notable commentary on Lyna (Yun) Luo's excellent new article ion channel rectification in @biophysj.bsky.social: Ion Channel Rectification Made Simple www.cell.com/biophysj/ful...

I was astonished to learn rapamycin was discovered in soil samples from a 1964 Canadian medical expedition to Easter Island and initially developed in Montreal. Suren Sehgal, the Canadian scientist who lead the research didn't have a wikipedia entry, so I wrote one: en.wikipedia.org/wiki/Surendr...

Happy to be part of CHARMM at 45 paper in JPBC pubs.acs.org/doi/full/10....

CHARMM at 45: Enhancements in Accessibility, Functionality, and Speed

Since its inception nearly a half century ago, CHARMM has been playing a central role in computational biochemistry and biophysics. Commensurate with the developments in experimental research and adva...

pubs.acs.org

Happy to see Phuc and Siri's perspective on MLXDM dispersion corrections to NNPs is published in @JPhysChem pubs.acs.org/doi/10.1021/...

Modeling Intermolecular Interactions with Exchange-Hole Dipole Moment Dispersion Corrections to Neural Network Potentials

Neural network potentials (NNPs) are an innovative approach for calculating the potential energy and forces of a chemical system. In principle, these methods are capable of modeling large systems with...

pubs.acs.org

A preprint of our perspective, Modeling Intermolecular Interactions With XDM Dispersion Corrections to Neural Network Potentials is now live. Please send comments! go.shr.lc/3UuvZcU

Don't miss the online seminar on membrane simulation by Dr. Ré Mansbach (Concordia University) and Dr. An Ghysels (Ghent University) April 30 2 PM EDT. Sign up link here: www.cheminst.ca/cic-virtual/...

PTC-Virtual Seminar

The Physical, Theoretical and Computational Chemistry (PTC) subject division is hosting a series of online seminars in the field of physical, theoretical, and computational chemistry. Following in the...

cheminst.ca

Thrilled to have @louisedawe.bsky.social visit @carletonu.bsky.social as the 2025 2025 Don Wiles lecturer.

New preprint on our group's graph neural network classifier to predict of a compound will react with a protein nucleophile. We predict that the ~5% of the ChEMBL database is protein-reactive doi.org/10.26434/che...

Each year Carleton University has a butterfly show in their greenhouse. Easily my favourite day of the year.

Thrilled to host the NMR-themed PTC virtual seminar Oct 3 2 PM ET. Registration and details www.cheminst.ca/cic_virtual/...

Nominations are open for the MGMS Frank Blaney Award for early career faculty in computational chemistry. www.mgms.org/WordPress/pr...

A new blog post in time for the new semester: Obviously Students Don't Read Course Outline (and what do to about it): teachphyschem.blogspot.com/2023/08/obvi...