Pengchao Zhang

PostDoc, PhD @Tsinghua_Uni. Bachelor @2024_HUST. #AI4Science, #compchem, #enhancedsampling, #MachineLearning.

[Pengchao Zhang @Pengchao1997]

Our @J_A_C_S study (https://t.co/Y1OuKtjnO8) shows nanobubbles strongly prefer staying at the TiO₂–water interface. Detaching a ~3 nm N₂ nanobubble requires ~150 kJ/mol, making spontaneous release unlikely—highlighting why bubble removal often needs growth or external forces.

[Pengchao Zhang @Pengchao1997]

Nanobubble stability is governed by interfacial ions. TiO₂ induces OH⁻ accumulation, creating a more negative zeta potential. More negative ζ → higher dissolution barrier → longer-lived nanobubbles. Simulations + experiments agree.

[Pengchao Zhang @Pengchao1997]

Our @J_A_C_S study (https://t.co/Y1OuKtjnO8) shows nanobubbles strongly prefer staying at the TiO₂–water interface. Detaching a ~3 nm N₂ nanobubble requires ~150 kJ/mol, making spontaneous release unlikely—highlighting why bubble removal often needs growth or external forces.

[Pengchao Zhang @Pengchao1997]

TiO₂ doesn’t always promote nucleation. The anatase (101) surface increases nucleation barriers, but in saline environments, the salting-out effect lowers the barrier—making TiO₂–NaCl systems most favorable for nanobubble formation.

[Pengchao Zhang @Pengchao1997]

🔥 The payoff: the rate-determining first N₂ hydrogenation drops to ~0.6 eV (ΔG ≈ 0.15 eV), finally matching experiments. Hydronium ions, H-bond networks, and local electric fields team up to make NRR easier than we thought.

[Pengchao Zhang @Pengchao1997]

🚀 Why does theory keep underestimating electrochemical N₂ reduction? Because water isn’t just a background. In our new work, we show that explicit solvent fundamentally rewires nitrogen reduction reaction at single-atom catalysts. Solvent ≠ spectator.
https://t.co/GOtvKpuJkr

[Pengchao Zhang @Pengchao1997]

💧 We used deep potential molecular dynamics with enhanced sampling to watch NRR happen in real water. No static snapshots, no implicit solvent—just long-time dynamics, proton shuttling, and a fully alive catalyst–electrolyte interface.

[Pengchao Zhang @Pengchao1997]

Extreme Nanoconfinement Reshapes the Self-Dissociation of Water
https://t.co/Hq1Drv399M

[Pengchao Zhang @Pengchao1997]

This paper is now available in Langmuir. @ACS_Langmuir https://t.co/1kcJgyPZuk

[Frank Noe @FrankNoeBerlin]

BioEmu now published in @ScienceMagazine !!

What is BioEmu? Check out this video:
https://t.co/PAj96iKvR7

[Pengchao Zhang @Pengchao1997]

The data sets for the deep Wannier (DW) and deep potential long range (DPLR) models are available at https://t.co/0tEGDPb9Mm

[Pengchao Zhang @Pengchao1997]

🚀 Based on Voronoi CV-based OPES DPLRMD, our new study reveals how electric fields & air-water interfaces switch glycine tautomerism mechanisms @JCIM_JCTC.
🔗 Article: https://t.co/qAqTyX7hRv
📂 Data: https://t.co/EG8GfOrOtW
#compchem #machinelearning #enhancedsampling

[Pengchao Zhang @Pengchao1997]

Results:
🔄 ​​Mechanism switch:​​
🔌 Fields flip dominant path:
Intra-PT → OH⁻-mediated Inter-PT
⚡ ​​10x kinetic boost​​
Field accelerates [Z]→[N] 10×
🌊 ​​Interface surprise:​​
Partial solvation speeds Intra-PT 5×
(while suppressing ion pair)

[Pengchao Zhang @Pengchao1997]

Method:
🖥️ ​​DPLR-MD model​​
Trained long-range electrostatics
Deep Wannier centroids for polarization
📊 ​​Voronoi CVs​​
OPES sampling for rare events
Captures Proton Transfer pathways
⚡ ​​Field integration
Explicit E-field in periodic box
Validated vs DFT (RMSE: 1 meV/atom)

[Pengchao Zhang @Pengchao1997]

Background:
⚡️ ​​Electric fields are EVERYWHERE in biology!​​
From enzyme electrostatics to membrane potentials → but how do they control amino acid proton transfer?
🔍 Glycine tautomerism = key model
💧 Solvent-driven PT pathways known
❓ UNKNOWN: Field/interface roles

[Pengchao Zhang @Pengchao1997]

🎬 ​​The above trajectory is derived from the Voronoi CV-based OPES DPLRMD under a electric field of 10 mV/Å (Trajectory is translated as a whole with N as the box center)
🔑Color: H⚪ C⚫ N🔵 O🔴 Wannier centroid🟡
🔌 10 mV/Å field boosts [Z]→[N] by 10x via OH⁻-mediated path.

[Machine Learning in Chemistry @ML_Chem]

Modulation of Electric Field and Interface on Competitive Reaction Mechanisms #machinelearning #compchem https://t.co/D7T7tPgYnz

[Pengchao Zhang @Pengchao1997]

@GroupParrinello @TrizioEnrico @PeilinKang @NatComputSci @IITalk Congratulations!

[Parrinello Group @GroupParrinello]

The follow-up work by @TrizioEnrico and @PeilinKang is out on @NatComputSci!🧨

A semi-automatic method to extensively sample "everywhere and everything" (metastable and transition states) "all at once" (in a single simulation) 🔥

📑: https://t.co/J8ScIuFo7W

Short🧵 @IITalk

[Pengchao Zhang @Pengchao1997]

@ACS_Langmuir At oil–water interfaces, hydronium ions are repelled into the bulk water, while hydroxide ions are enriched in the interfacial layer due to stabilization by positive charges from interfacial water and decane molecules.

[Pengchao Zhang @Pengchao1997]

Check out our latest preprint on double-layer distribution of hydronium and hydroxide ions at air/oil-water interfaces after water autoionization (https://t.co/U3UFugdkxO). #compchem #machinelearning #enhancedsampling (1/2)

[Pengchao Zhang @Pengchao1997]

This paper is now available in Langmuir. @ACS_Langmuir https://t.co/1kcJgyPZuk

[Pengchao Zhang @Pengchao1997]

@ACS_Langmuir This research investigates the behavior of hydronium and hydroxide ions at air/oil–water interfaces using enhanced-sampling molecular dynamics simulations with NNPs. The study finds that a stable ionic double-layer forms near the air–water interface. @ACS_Langmuir

[Pengchao Zhang @Pengchao1997]

@ACS_Langmuir The input files needed to reproduce the research are available at https://t.co/J4PSj3IN1H.