Beam loss is critical in high-power accelerators. During CAFe’s high-power proton beam commissioning, unexpected irradiation opposite the deflection direction revealed parasitic H${}^{-}$ particles—first observed in a linac. H${}^{-}$ particles were mainly generated by the electron capture reaction between the proton beam and the background gas (H${}_2$). This study uses nuclide analysis, molecular science, and beam experiments to verify this loss mechanism, with key implications for future high-intensity proton accelerator designs.

This paper presents a novel energy recovery proton linear accelerator. In this design, the final GeV-level proton beam is reinjected into the linac from the accelerator exit and decelerated within the same structure to nearly its initial energy. This approach significantly reduces the power consumption of the proton linear accelerator and eliminates the need for a high-power beam dump.

This study presents a comprehensive investigation into the dynamics and optimization of space charge compensation (SCC) in the low-energy beam transport (LEBT) line of a high-perveance proton beam, integrating 3D numerical simulations with time-resolved experimental measurements. These measurements analyze the effects of injected gas properties, pressure levels, and beam current on build-up time, while also considering nonlinear space charge effects that influence SCC efficiency and the beam’s transverse emittance evolution.

Continuous-wave X-ray free-electron lasers (CW FELs) promise transformative capabilities for ultrafast science, but their injectors must remain robust under hardware constraints and unexpected failures. In this work, we present the systematic fault-tolerance analysis of the LCLS-II injector, using multi-objective optimization to explore mitigation strategies for solenoid and cavity degradation. Our study demonstrates how beam quality can be preserved under non-ideal conditions, providing practical guidance for commissioning and reliable operation of future high-repetition-rate FEL facilities worldwide.

When colliding high-intensity high-charge electron beams attract instead of repel.

What happens when two ultra-relativistic electron beams collide? You might expect them to repel and fly apart. But in extreme high-intensity, high-charge conditions, the beams can actually pinch together. This counter‑intuitive effect is driven by intense coherent pair production, which changes the beams’ self-fields and reverses the Lorentz force. This results in a dramatic compression, boosted luminosity, and amplified fields far beyond what has been previously predicted.

In this paper, we report the generation of a 2 GeV electron beam with a peaked energy spectrum using laser wakefield acceleration. This energy was achieved by combining density-transition injection with an optically generated plasma waveguide at the Apollon laser facility, employing a laser energy of 12 J. The use of the waveguide enabled an increase in the peak energy by 300–400 MeV, while preserving the beam charge at the same level, in the tens of pC. These results are highly promising for strong-field QED experiments, which demand both high-energy particles and intense laser pulses.

An upgrade of the SACLA linear accelerator is planned to achieve a kHz-level repetition rate and increased beam brightness. A key component of this upgrade is a new injector based on the existing pulsed DC gun with a thermionic cathode, featuring a novel architecture that combines multistage velocity bunching with a single stage of magnetic compression. Through genetic algorithms-based optimization, the system achieves up to three orders of magnitude in bunch compression while maintaining sub-micrometer emittance levels. The performance of the proposed injector is comparable to that of the state-of-the-art C-band injector based on an RF photocathode gun

Achieving high peak current without degrading beam quality is a challenge for modern XFELs. Bunch compression in the injector stage is required to provide enhanced downstream brightness. However, the process can introduce emittance growth, energy spread, and asymmetries in the longitudinal phase space and charge distribution, which can critically limit FEL performance. This work presents two XFEL injector designs that balance compression with emittance preservation, account for these effects, and provide a pathway toward brighter XFEL operation.

The energy-recovery linac (ERL) can accelerate high-current beams with low emittances and short bunches due to an energy-recycling scheme by using a superconducting cavity with no cavity-wall losses for accelerating cavities. However, the energy recovery of high-current beams needs to be addressed. In this paper, we propose a method for high-current beam tuning of approximately 1 mA under energy-recovery conditions with extremely small beam loss in compact energy-recovery linac (cERL) at KEK by using a collimator and a local fast beam loss monitor. We report successful results at approximately 1 mA, demonstrating 100.0% energy recovery with extremely small beam loss in cERL.

There has been a continuous effort to improve the performance of light sources by pursuing higher photon brilliance and coherence. This trend has been driving the lattice design of storage rings toward a diffraction-limited emittance, which in turn results in a tighter acceptance for the injected beam. Beam injection is one of the most crucial issues for the development of future light sources based on diffraction-limited storage rings. To address the beam injection requirements of next-generation light sources, we have developed an in-vacuum thin eddy-current septum magnet.

The principal planes of a magnetic compound system consisting of N thick quadrupoles is introduced to analyze the transverse phase space diagnostic using the focus scan scheme in plasma-based accelerators. We use the principal planes to analyze the effect of the distance variations between the source, magnets and screen on the imaging energy and apply the conclusions to gas targets with different downramp lengths. A beamline with two quadrupoles is designed and analyzed as an example. The analysis can be applied to beamlines consisting of more quadrupoles and is useful for the transverse phase diagnostic using the focus scan scheme.

Laser-plasma accelerators are a disruptive technology but require improved shot-to-shot stability for many applications. Fluctuations in the longitudinal focal position of the drive laser are a major source of jitter in the accelerated electron beam qualities. We demonstrate high-bandwidth longitudinal focal stabilization beyond the Nyquist frequency limit of a 100 TW-class drive laser, yielding marked improvements in the stability of the generated electron beam charge and spectrum.

Superconducting linear accelerators (linacs) require accurate monitoring of beam dynamics. We developed an improved method for estimating the transverse beam envelope using Beam Position Monitor (BPM) signals. This method has been recognized for decades, but with limited use due to accuracy concerns. To improve, we employed $\cos 2\theta$-type BPMs, which offer improved sensitivity to quadrupole components. This method was applied to heavy ion beams in the superconducting RIKEN linac (SRILAC). Estimated beam envelopes show good agreement with conventional quadrupole scan results, confirming feasibility for accurate, routine, and nondestructive beam monitoring in superconducting sections.

Building on previous experimental and theoretical work, in this article we investigate potential applications of helical targets for medical radioisotope production. Specifically, this study focuses on energy bunching, thanks to the helical coil, of alpha particles generated through proton-boron fusion reactions. By selecting particles with specific energies, it becomes possible to optimize the reaction cross-section for producing desired radioisotopes, demonstrating a promising application of helical targets in nuclear medicine.

The newly developed cryogenic permanent magnet undulator (CUT18) at Taiwan Photon Source employs liquid nitrogen tank cooling to deliver sustainable, vibration-free operation. Using high-remanence Tb-diffused NdFeB magnets and optimized magnetization tilt, CUT18 achieves a deflection parameter (K) above 2. This paper highlights its robust performance under high beam currents and presents a design optimized for magnetic field strength, energy efficiency, and long-term reliability in modern synchrotron light sources.

Ferroelectric devices have progressed to the point that megawatt-class phase shifters and associated cavity tuners are feasible. The design methodology presented for a Ferroelectric Fast Reactive Tuner enables modulation of Mega VAR reactive powers on a sub-microsecond timescale by offering detailed, closed-form component calculations. Performance optimization is addressed through a Figure of Merit, which allows assessment of tuner effectiveness and applicability across a wide range of RF frequencies and reactive power levels. This approach supports rapid feasibility analysis and parameter selection for scenarios defined by required tuning range, cavity frequency, and stored energy.

Accurate long-term particle tracking in modern accelerators with curved, three-dimensional magnetic fields requires robust symplectic methods, which preserve the physical structure of the system’s dynamics. This paper presents a novel algorithm that extends the powerful generating function method—previously limited to straight systems—to arbitrary curved reference trajectories. By combining analytical harmonic field representations with a perturbative solution to the Hamilton-Jacobi equation, our method yields a direct, high-resolution map of the particle dynamics, providing a crucial and efficient tool for designing next-generation circular accelerators with complex magnet configurations.

Tensor-polarized deuteron beams provide a unique probe for studying neutron spin structure at the Electron-Ion Collider in China (EicC) and beyond-Standard-Model physics. This work derives and validates spin tensor transfer matrices to model the evolution of tensor polarization through magnetic elements and during acceleration across depolarizing resonances. We show that polarization losses are negligible at EicC, enabling high-precision spin experiments without dedicated preservation devices. The translational nature of tensor polarization, constrained to a triangular plane, offers new insights for spin tensor direction control in synchrotrons.

We propose a method to calibrate the phase and amplitude of a cavity using the time-of-flight technique with non-relativistic electrons. In this regime, the beam experiences velocity modulation, which alters its longitudinal distribution and can distort calibration, especially for long bunches and high charges. To address this, we introduce a truncation-based correction that suppresses such effects while keeping the procedure simple and practical. The method improves the accuracy and robustness of TOF calibration, providing a reliable basis for accelerator tuning and performance optimization. Its effectiveness has been validated at the HEPS linac.

Nonlinear RF fields in bunchers typically deform the longitudinal phase space (LPS), leading to emittance growth. However, particle tracking simulations of a 125 mA/5 MeV deuteron beam at LIPAc reveal a decrease in longitudinal rms emittance, enabled by a self-linearization mechanism where nonlinear space charge forces counteract RF-induced LPS deformation. To quantitatively describe this effect, we introduce two analytical formulas representing the deformations of phase space and forces acting on the beam and validate them through simulation. These results offer new insights into beam dynamics and suggest novel strategies for mitigating phase space deformation in high-current linacs.

The impedance characteristics of the in-vacuum undulators (IVUs) and associated heat load deposition during beam passage constitute critical considerations in IVU design and operation. This study focuses on the transverse and longitudinal trapped modes and their mitigations, and explores the impedance enhancement introduced by the bulge flexible taper transitions. Laboratory bench measurements are performed to identify the longitudinal and transverse impedance of the IVU experimentally.

Transverse beam instabilities, synchrotron radiation damping, and Landau damping – these phenomena have been thoroughly studied in storage rings where the beam energy is fixed, but much less in booster synchrotrons where the beam energy can vary widely, especially a low-emittance booster meant to serve a 4th generation storage ring. We present numerical studies in which these effects interplay throughout an energy ramp process using the in-house developed tracking code mbtrack2 and the SOLEIL II booster. In critical conditions of the new generation boosters competition between the effects can result in beam instabilities despite the short passage time of the beam in the machine.

Maximizing electron beam energy in laser wakefield accelerators hinges on determining the appropriate combination of laser and plasma parameters. In this work, particle-in-cell simulations coupled with a Bayesian optimization algorithm are employed to identify optimal conditions that yield the highest electron energies for a fixed laser pulse energy. The resulting data are used to derive generalized scaling laws for electron energy and acceleration length as functions of laser energy. These scaling laws are presented alongside the complete set of input parameters, offering a practical framework for the design of laser wakefield acceleration experiments across a wide range of laser systems.

Images on the top and bottom left are the (x,y) and (z,pz) projections, respectively, of the 6D position-momentum phase space density of a proton beam in the LANSCE accelerator. Images on the right were generated by an adaptive latent diffusion model conditionally guided by accelerator RF settings and a 1D vector of beam loss measurements. This paper demonstrates the power of diffusion models to solve extreme ill-posed inverse problems and state-of-the-art ability to generate accurate high-resolution representations of objects with very high variance as seen by the completely different characteristics of the images above which are generated by a single model with varying conditional vectors.

Deep reinforcement learning is a promising technique for particle accelerator optimization. In this contribution, a memory-based policy was trained in simulations and successfully deployed to the injection system of the Super Proton Synchrotron at CERN. A speed-up of one order of magnitude over conventional numerical optimization was achieved.

This article presents the first detailed study of synchrotron beam dynamics incorporating strongly curved canted-cosine-theta (CCT) superconducting magnets. We develop a method to characterize the 3D curved fields of a CCT model developed for a compact carbon ion therapy synchrotron, designed within the HITRI+ project and NIMMS (CERN). Beam dynamics studies are carried out in MAD-X/PTC to ascertain the effects of magnet nonlinearities on long-term beam stability. Insights gained provide new techniques to characterize and correct for the effect of curved magnets, allowing the optimization of both magnet and synchrotron designs, with applications extending to particle physics.