About me
Plasma/Accelerator Physicist at UCLA • AI and Machine Learning
Los Angeles, CA
I’m Chaojie Zhang, a plasma and accelerator physicist at UCLA. My research has focused on turning plasma wakefield acceleration from proof-of-principle demonstrations into reliable technology for future colliders and compact light sources, combining experiments, large-scale simulations, and machine-learning-based modeling. The rapid progress of large language models has drawn my interest to AI itself: how the technology is developing, and how it can be applied to frontier scientific research.
I received my B.S. and Ph.D. in Engineering Physics from Tsinghua University. During my Ph.D. I developed femtosecond relativistic electron probing (FREP), which produced the first direct images of plasma wakefields and received the John Dawson Thesis Prize. Since 2017 I have been at UCLA, working on plasma accelerators and kinetic plasma instabilities, and leading experiments at SLAC’s FACET-II and Brookhaven’s Accelerator Test Facility.
More about my work is in my CV, publications, and research projects.
Research Highlights
Plasma Wakefield Transformer
As PI of the E304 experiment at FACET-II, I demonstrated a plasma "dual transformer" that decouples energy gain from quality preservation—converting a low-quality drive beam into a new, ultra-bright beam with 2× higher energy (>20 GeV) and 10× higher brightness. This approach achieves sub-0.5% energy spread while maintaining the extreme brightness needed for X-ray free-electron lasers, and enables novel staging architectures that could bypass the quality-preservation challenge plaguing conventional multi-stage designs. Published in Nature Communications (2025).
AI/ML-Driven Virtual Diagnostics
I developed physics-informed "virtual diagnostics" that use machine learning to reconstruct the longitudinal phase space of femtosecond electron bunches from plasma wakefield accelerators—information impossible to measure directly. This ML-driven technique was critical to analyzing E304 results and is now being adopted by collaborators in other PWFA experiments. Beyond diagnostics, it opens the door to ML-enabled optimization and autonomous control of plasma accelerators. Published in Nature Communications (2025).
Femtosecond Relativistic Electron Probing (FREP)
I invented FREP during my Ph.D., using ultrashort relativistic electron bunches from a laser wakefield accelerator to probe plasma wakefields—capturing the first-ever snapshots of these microscopic, transient, near-light-speed structures. This breakthrough enabled the discovery of plasma wake reversal and earned the John Dawson Thesis Prize. FREP has since become an essential diagnostic at the frontier of plasma acceleration. Published in PRL (2017, Editors' Suggestion).
Probing the Hierarchy of Kinetic Instabilities
As PI of the AE98 experiment at BNL, I led the first direct measurement of the thermal Weibel instability—a fundamental kinetic instability predicted decades ago but with no conclusive experimental validation in laboratories. We mapped the self-generated magnetic fields in laser-ionized plasmas, revealing the growth and saturation of this instability and bridging laboratory and astrophysical plasma physics. Published in PNAS (2022, highlighted by DOE) and PRL (2020).
Let’s Connect
I’m glad to hear from people working on accelerator physics, plasma science, or AI and computing systems. You can reach me through the email link below.
news
| Nov 28, 2025 | Our plasma "dual transformer" work is published in Nature Communications |
|---|---|
| Jul 21, 2024 | Presented the latest results on plasma wakefield acceleration experiments from FACET-II as a plenary speaker at AAC24 in Naperville, IL. |
latest posts
| Sep 18, 2026 | 理解 Transformer |
|---|---|
| Jul 16, 2026 | The Fourth Generation: Where the X-Ray Light Sources Stand in 2026 |
| Jul 11, 2026 | Laser-Driven Proton Therapy: Strong Physics, Hard Road to the Clinic |
selected publications
- Nat. Commun.
- Phys. Rev. Lett.2017Femtosecond probing of plasma wakefields and observation of the plasma wake reversal using a relativistic electron bunchSummary: First demonstration of femtosecond relativistic electron microscopy for visualizing plasma wakefields.