XuanyuZhu | CompChem

Computational Physical Chemistry researcher | WebXR developer | VR molecular universe @

[Paesani Lab @PaesaniLab]

🚨The pirates presented at the 2026 @GordonConf on Water & Aqueous Solutions in Holderness, NH!💧

A fantastic week of exciting science, great discussions, and reconnecting with friends old and new. #WaterGRC #CompChem

Four of our PhD students presented their research🎉:
Henry — Many-Body Effects Determine the Propensity of Halides for the Air–Water Interface
Xuanyu — Many-Body Effects for Solvated Ion Pairs
Suman — Confinement-Controlled Autoionization of Nanoconfined Water
Zitong — Computational Investigation of Supercooled Water Using Many-Body Simulations and 2D IR–Raman Spectroscopy

We’re incredibly proud of them for representing the group and sharing their work with the water community.

Well done, pirates! 🏴‍☠️👏
#PhDLife #ManyBody #WaterScience
@UCSanDiego @UCSDPhySci @UCSDChemBiochem @HDSIUCSD @SDSC_UCSD

[Paesani Lab @PaesaniLab]

🚨 Greetings from the 2026 Gordon Research Conference on Water and Aqueous Solutions in New Hampshire!

It’s great to reconnect with old friends, meet new people, and of course spend the week talking about #water. 🌊

Our group will be presenting posters on #datadriven #manybody simulations of ionic aqueous interfaces and ion pairs, nanoconfined water reactivity, and simulations of the 2D IR spectra of liquid water. 😎

Looking forward to a fun week of science and good conversations! 🏄‍♀️

@UCSanDiego @UCSDPhySci @UCSDChemBiochem @HDSIUCSD @SDSC_UCSD @GordonConf

[PhilipZ/dev @philip1xyz]

Very 2026 moment: my Claw agent diagnosed a live Signal reply-context blind spot, hot-patched runtime live, restarted gateway, and validated with quoted-message traffic. signal-cli wasn’t the bottleneck—context mapping was. Reply refs now arrive as structured ReplyTo* context.

[Paesani Lab @PaesaniLab]

🚨 New preprint on @ChemRxiv! 🚨

Can a molecular model trained ONLY on the many-body expansion (#MBE) derived from CCSD(T) reference data predict bulk ion hydration thermodynamics without fitting to hydration free energies or bulk solution structure? 🤔

We show that our #datadriven #manybody MB-nrg potentials can! 😎

👉 https://t.co/d9yJVZxZAe

Method
A staged alchemical cycle that turns on the physics step-by-step:
✅ auxiliary vdW → charge → polarization → explicit 2B → explicit 3B
✅ then remove the reference vdW
✅ finite-difference thermodynamic integration
and #PIMD mass-scaling for #NuclearQuantumEffects.

Results
🎯 Correct monotonic weakening of hydration with increasing ion size (cations + anions).
🎯 Close agreement with experiment when comparing within-series reference differences (ΔΔG).
🎯 #NuclearQuantumEffects are modest but slightly improve the trends.

Take-home message
Hydration thermodynamics naturally emerges from a rigorous representation of many-body interactions without requiring fitting to bulk solvation data. 🚀

Huge congrats to Suman (@SSuman_24) for leading this project! 🙌

🙏 Big thanks to @NSF for funding and to @ACCESSforCI and @NERSCfor providing computational resources!

@UCSanDiego @UCSDPhySci @UCSDChemBiochem @HDSIUCSD @SDSC_UCSD

[Paesani Lab @PaesaniLab]

🚨 New on @ChemRxiv: Computer simulations with MB-pol reveal fascinating #water behavior in #CNTs, modulated by carbon polarizability & tube stacking!

👉 https://t.co/bqvhTH9ZKs

💧 In the tiniest (10,0) #CNT, #water lines up single-file. In mid-sized (15,0) #CNTs, it forms 5-molecule rings. In the largest (20,0) #CNTs, two distinct layers emerge.
⚡ Including carbon polarizability causes #water to tilt more toward CNT walls, especially in the larger tubes.
💻 Our MB-pol #datadriven #manybody potential gives a deeper view of confined #water, resolving #IR fingerprints directly from simulations.
💨 When #CNTs are stacked, #water dynamics accelerate: hydrogen bonds break ~20–25% faster, and molecules switch layers ~20% more often, indicating significant inter-tube coupling.

🤝 This is the results of a wonderful collaboration with the Kirchner group at @UniBonn. We were delighted to host Leonard, who led this exciting study.

🙏 Huge thanks to @NSF and the German Academic Exchange Service (@DAAD_Germany) for supporting this work and making this collaboration possible!

@UCSanDiego @UCSDPhySci @UCSDChemBiochem @HDSIUCSD @SDSC_UCSD

[Paesani Lab @PaesaniLab]

And that’s a wrap for #SMLQC2025! 🖥️

Huge congratulations to @EtBuVu! 🙌

@UCSanDiego @UCSDPhySci @UCSDChemBiochem @HDSIUCSD @SDSC_UCSD

[Paesani Lab @PaesaniLab]

We’re at #SMLQC2025! ⚛️
(https://t.co/NhJdAlBQ1I)

Ethan, Xuanyu, Xinmao, and Francesco presented our latest work on #datadriven #manybody potentials for #biomolecules 🧬 and our development of permutationally invariant fourier series (PIFS) as a robust mathematical basis that ensures stability beyond training data and enables simulations of #water at extreme conditions.🧊

Interested in learning more about our work on #datadriven #manybody potentials for large, covalently bonded molecules? Check out our recent papers:
👉 https://t.co/pxVRWgRiVB
👉 https://t.co/4ovRZxiN1b
👉 https://t.co/CS4dnXNg6I

@UCSanDiego @UCSDPhySci @UCSDChemBiochem @HDSIUCSD @SDSC_UCSD

[XuanyuZhu | CompChem @philipzxy]

very important

[Inter-Parliamentary Alliance on China (IPAC) @ipacglobal]

📢 Statement on the 26th Anniversary of the Persecution of Falun Gong Practitioners / 就法輪功學員遭中共壓迫26周年聲明

[Paesani Lab @PaesaniLab]

🚨 New paper alert! 🚨

👉 https://t.co/CS4dnXMIha

We just introduced in @JCIM_JCTC permutationally invariant Fourier series (PIFS) as a new basis for #datadriven #manybody potentials!

#MachineLearning potentials can hallucinate 😱, predicting unphysical energies in regions beyond their training set. PIFS eliminates these holes while preserving the accuracy in regions covered by the training data.

Using #water as a test case 🌊, our new MB-pol(PIFS) potential matches the accuracy of MB-pol, the current "gold standard" for water simulations, across gas and liquid phases, but exhibits significantly improved stability under extreme thermodynamic conditions.

Out with blind extrapolation ❌. In with physics-grounded predictions. 🏄‍♀️

@UCSanDiego @UCSDPhySci @UCSDChemBiochem @HDSIUCSD @SDSC_UCSD

[Shen Yun @ShenYun]

"Growing up in the West, the persecution always felt a bit distant because nothing actually happened to me personally. But a few years back, during a show in Holland...

It was intermission and a Chinese lady came up to the pit and I don't remember exactly what the acknowledgement was, but she acknowledged our Erhu player at the time. I asked her who the lady was, and she responded, we were in the same labor camp together.

In that moment, the weight of our mission felt very real. We weren't just here to entertain. We have a mission of reviving Chinese culture and sharing the stories of practitioners in China whose voices cannot be heard.

For someone who feels strongly about this mission, I was deeply hurt when I heard about the bomb threats towards our company. As a performing arts group our mission to revive traditional Chinese culture and raise awareness about the persecution of Falun Dafa came from a place of faith and compassion. To see people go as far as endangering the well-being of our community, our families, and even our audience members... It's egregious."

At the Lincoln Center press conference, Principal Violist Rachel Chen speaks about the bomb threats, how Shen Yun gave her pride in her heritage, and why it became her spiritual home.

[XuanyuZhu | CompChem @philipzxy]

@PaesaniLab https://t.co/H9ByDUXdQz

[XuanyuZhu | CompChem @philipzxy]

https://t.co/H9ByDUXdQz

[Paesani Lab @PaesaniLab]

🚨 Xuanyu (@philipzxy) just wrapped up his talk at #APSSummit25! 🎉

He presented his latest work on permutationally invariant Fourier series (PIFSs) and their role in eliminating unphysical energy “holes” in #datadriven #manybody potentials. PIFSs enhance stability and accuracy, pushing the boundaries of what’s possible in CCSD(T)-level #compchem simulations! 🖥️

If you missed Xuanyu's talk, you can check out his preprint on @ChemRxiv:
👉 https://t.co/geqVRyYx8m

@UCSanDiego @UCSDPhySci @UCSDChemBiochem @HDSIUCSD @SDSC_UCSD

[Paesani Lab @PaesaniLab]

🚨 Xuanyu (@philipzxy) just wrapped up his talk at #APSSummit25! 🎉

He presented his latest work on permutationally invariant Fourier series (PIFSs) and their role in eliminating unphysical energy “holes” in #datadriven #manybody potentials. PIFSs enhance stability and accuracy, pushing the boundaries of what’s possible in CCSD(T)-level #compchem simulations! 🖥️

If you missed Xuanyu's talk, you can check out his preprint on @ChemRxiv:
👉 https://t.co/geqVRyYx8m

@UCSanDiego @UCSDPhySci @UCSDChemBiochem @HDSIUCSD @SDSC_UCSD

[PhilipZ/news @philxyz]

New York Times Continues Unfounded Attack on Falun Gong Beliefs, Community https://t.co/k3ARyTt7Nw

[Paesani Lab @PaesaniLab]

Bioluminescent water-air interface at Scripps Pier @UCSanDiego @Scripps_Ocean (Photo credit: @philipzxy, https://t.co/JTzxRJUuZv)

[XuanyuZhu | CompChem @philipzxy]

@PhysInHistory Precision upto 29.9792 +- 0.0001