Terahertz control of relativistic electron beams for femtosecond bunching and laser-synchronized temporal locking

Beatriz Higuera-Gonz\'alez; Christopher T. Shaw; Connor D. W. Mosley; Daniel S. Lake; Darren M. Graham; David A. Walsh; Graeme Burt; James K. Jones; Joseph T. Bradbury; Laurence J. R. Nix

REVIEW 3 major objections 5 minor 1 cited by

Terahertz control of relativistic electron beams for femtosecond bunching and laser-synchronized temporal locking

T0 review · 3 major / 5 minor · reviewed 2026-08-05 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper shows that laser-generated terahertz pulses can imprint a steep, laser-synchronized energy chirp on 35.5 MeV electron bunches, compressing them to 15 fs duration or into picosecond-spaced bunch trains, and passively locking their

desk verdict Real experimental advance in THz chirping of 35.5 MeV bunches, but the 15 fs compression and 25 fs locking are idealized-model predictions, not measurements. read the letter →

arxiv 2508.20685 v1 pith:YQX7KA6P submitted 2025-08-28 physics.acc-ph physics.app-phphysics.optics

Morgan T. Hibberd , Christopher T. Shaw , Joseph T. Bradbury , Daniel S. Lake , Connor D. W. Mosley , Sergey S. Siaber , Laurence J. R. Nix , Beatriz Higuera-González

show 7 more authors

Thomas H. Pacey James K. Jones David A. Walsh Robert B. Appleby Graeme Burt Darren M. Graham Steven P. Jamison

This is my paper

classification physics.acc-phphysics.app-phphysics.optics PACS 41.75.Fr29.27.-a42.65.Ky

keywords THz-drivencompressionelectronbunchtemporallockingtrainsdielectric-linedwaveguiderelativisticbeamslasersynchronizationjittersuppression

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper reports a way to compress relativistic (35.5 MeV) electron bunches to femtosecond durations and synchronize them to a high-power laser using laser-generated terahertz (THz) pulses. The authors show experimentally that a 0.39 THz pulse in a dielectric-lined waveguide imprints a steep, laser-synchronized energy chirp on the bunch; sending the measured time-energy distribution through a modeled magnetic chicane compresses a 400 fs, 2 pC bunch by a factor of 27 to 15 fs rms, or turns a 2.5 ps, 30 pC bunch into a train of roughly 37 fs micro-bunches spaced about 2 ps with up to 50 A peak current. The same mechanism passively locks the compressed bunch arrival time to the drive laser: because the THz carrier phase is tied to the laser envelope and the chicane makes arrival time nearly independent of injection phase, jitter simulations predict 25 fs rms electron-to-laser jitter despite 200 fs laser timing jitter. The significance is practical: this chirp is produced in an 8 mm interaction length with a 16 MeV/m gradient, where a 3 GHz RF system would need an unfeasible gradient or more than 20 MeV of acceleration capability, opening a route to laser-synchronized femtosecond beams for FELs, ultrafast electron diffraction, and plasma-accelerator injection.

What carries the argument

The load-bearing object is the THz-driven longitudinal chirp produced in a dielectric-lined waveguide (a rectangular copper structure with fused-quartz liners whose LSM11 mode is phase-velocity matched to the relativistic electron beam). A 0.39 THz, quasi-monochromatic pulse in this mode accelerates one part of the bunch and decelerates another, creating a time-energy slope of about 0.35 MeV/ps over 8 mm. A subsequent magnetic chicane with negative temporal dispersion converts that slope into compression; because the chirp repeats every 2.56 ps, the same mechanism generates micro-bunch trains. Temporal locking emerges because the THz carrier phase is fixed to the drive-laser pulse, so any sh

What would settle it

Measure the longitudinal phase space of the compressed bunch after a real magnetic chicane with a THz streaking or transverse-deflecting cavity on a single-shot basis: if the rms bunch duration is not near 15 fs, or if the shot-to-shot arrival time relative to the THz pulse is not near 25 fs rms while the drive laser jitters by about 200 fs, the central claims fail. The model's weakest link could also be probed by recording the RF and laser jitters simultaneously instead of using the independently estimated values in the Methods.

Watch

Extended reading notes

Core claim

At a 35.5 MeV radio-frequency linac, the authors used laser-generated 0.39 THz pulses in a dielectric-lined waveguide to imprint a time-energy chirp on electron bunches. For 2.5 ps, 30 pC bunches, the multi-cycle THz field produced a periodic local chirp of 0.325 MeV/ps, which a modeled chicane with D = -3.1 ps/MeV compresses into a train of about 37 fs rms micro-bunches spaced roughly 2 ps with up to 50 A peak current. For 400 fs, 2 pC bunches injected at the THz zero crossing, the chirp reached 0.345 MeV/ps, and a matched chicane (D = -3.01 ps/MeV) compresses the bunch by a factor of 27 to 15 fs rms. Because the THz carrier phase is locked to the drive-laser envelope and the chicane makes

Load-bearing premise

The 15 fs compression and 25 fs temporal locking assume an ideal first-order magnetic chicane with no collective effects (space charge, coherent synchrotron radiation, wakefields) adding energy spread at the few-femtosecond scale, and the jitter result uses independently estimated RF and laser jitters rather than values measured on the same shots.

Editorial extensions

If this is right

At 35.5 MeV with 2 pC charge, a 15 fs rms compressed bunch is projected from measured time-energy distributions plus a matched chicane, about an order of magnitude shorter than previous THz-driven compression results.
A single 30 pC bunch can be shaped into a train of micro-bunches with roughly 2 ps spacing and about 37 fs rms duration, tuneable through the injected linear chirp, with up to 50 A peak current per micro-bunch.
The compressed bunches' arrival time tracks the drive laser, so a 200 fs laser timing jitter is reduced to 25 fs rms electron-to-laser jitter in simulation, relaxing demands on laser-accelerator synchronization.
The THz chirp is produced in 8 mm at 16 MeV/m; an equivalent 3 GHz RF chirp would require an unfeasible gradient or diverting more than 20 MeV of acceleration, so the scheme is compact.
If slice energy spread and THz gradient are improved as the paper extrapolates, micro-bunch peak currents can approach the kiloampere regime typical of FEL injectors.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If the passive-locking mechanism holds in a real chicane, the same self-correcting synchronization concept could extend to other laser-driven high-frequency structures where the drive phase is envelope-locked, not only THz.
A direct single-shot measurement of both laser and bunch arrival times would separate the model's two claims; the jitter prediction could be tested even if the compressed bunch is longer than 15 fs.
Because the micro-bunch spacing depends on the injected chirp as well as the THz period, intentionally varying the linac chirp could provide fast, tuneable spacing control without changing the fixed 0.39 THz source.
The measured slice energy spreads (8 keV for long bunches, 10 keV FWHM for short bunches) set the floor for further compression; reducing them or raising the THz gradient could reach the kA peak-current regime the authors extrapolate.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Short version: this is a genuinely useful experimental paper with an overreaching abstract. The measured THz-driven energy modulation at 35.5 MeV is clean and well-characterized; the 15 fs compression and 25 fs locking numbers are predictions from an idealized transport model, not measurements, and the paper mostly says so in the main text but not in the abstract.

The new part is the energy regime and charge: previous THz compression demonstrations (Snively and Zhao, both 2020) were at few-MeV scale. Here they show multi-cycle 0.39 THz pulses imprint a strong, stable chirp on 30 pC, 35.5 MeV bunches, and even on 2 pC, 400 fs bunches. The calibration work is solid—using the spectral splitting threshold to get the THz gain model-free, and extracting 8–10 keV slice energy spread from the narrow spectral peaks. That gives a real measure of the best-case compressed duration.

The soft spots are where the stress-test note lands. The compression and temporal-locking results come from applying a first-order chicane dispersion to the measured time-energy distribution. There is no CSR, longitudinal space charge, or wakefield term in the transport map. At the claimed peak currents (~130 A for the single-bunch case, 50 A in the microbunches), CSR in a chicane with D≈−3 ps/MeV could produce correlated energy spread comparable to the 8–10 keV slice spread, which would push the compressed bunch length well beyond 15 fs. The jitter analysis also uses independently estimated jitter inputs (Table 1) rather than simultaneous measurements, so the 25 fs locking is a model result. That said, the 50-shot spectral agreement in Fig. 4a gives some confidence in the jitter model, but it does not directly validate arrival time.

The authors are honest in the methods, but the abstract and conclusion overstate. “Capable of compression” is doing a lot of work—no actual compression is demonstrated. This is a strong concept paper, but it should be read as such.

I'd send it to review, with the expectation that the authors either add a realistic treatment of collective effects or explicitly frame the compression numbers as idealized upper bounds. It deserves a serious referee; the experimental core will be useful to the community either way.

Referee Report

3 major / 5 minor

Summary. The paper reports experiments at the CLARA linac in which multi-cycle THz pulses (~0.39 THz, up to ~120 µJ) generated by a PPLN wafer stack are coupled into a dielectric-lined waveguide to imprint a strong energy chirp on 35.5 MeV electron bunches. Two configurations are studied: long chirped bunches (2.5 ps rms, 30 pC) produce multi-cycle periodic energy modulation, from which the authors extract the injected chirp and an 8 keV rms time-slice energy spread, and then model compression into a train of 37 fs rms micro-bunches with ~50 A peak current after a chicane with D = -3.1 ps/MeV; short bunches (400 fs rms, 2 pC) are driven with a single sub-cycle chirp of 0.345 MeV/ps, leading to a predicted compression to 15 fs rms after a matched chicane. The paper also presents a jitter model combining estimated RF, THz, and injection jitter sources, concluding that compressed bunches would be temporally locked to the drive laser with 25 fs rms arrival-time jitter despite 200 fs THz-source jitter. The experimentally measured quantities are the THz-modulated energy spectra and their dependence on THz phase and energy; the compression and locking results are predictions from a first-order transport model.

Significance. If the compression and locking predictions hold, this is a substantial advance: it would extend THz-driven manipulation to fully relativistic, high-charge beams and demonstrate a passive synchronization mechanism that suppresses the intrinsic jitter of high-power laser systems. The experimental demonstration of a 16 MeV/m THz gradient in a dielectric-lined waveguide at 0.39 THz and the model-free extraction of the time-slice energy spread from the splitting threshold are solid and valuable contributions. The concept of using the multi-cycle THz field to create tunable micro-bunch trains from a single RF bunch is also interesting. However, the headline quantitative results (15 fs bunch length, 37 fs micro-bunches, 25 fs locking) are not measured; they are produced by an idealized first-order chicane model that neglects collective effects and uses independently estimated jitter inputs. The paper would be strengthened by clearly separating measured from modeled claims and by addressing the robustness of the predictions.

major comments (3)

[Methods: Compression and bunch train modelling; Charge density distribution evolution] The compressed bunch lengths (15 fs single bunch, 37 fs micro-bunches) are obtained by applying a first-order temporal dispersion D to the measured or fitted time-energy distribution. The stated symplectic map U_f = f(U_i,t_i,{α_k}), t_f = g(...) contains no CSR, longitudinal space charge, wakefields, or higher-order dispersion terms. At the claimed compressed parameters (2 pC into 15 fs rms; 5 pC per micro-bunch into 37 fs rms), the peak currents are ~50-135 A depending on profile, and through a chicane with |D|~3 ps/MeV the CSR-induced correlated energy spread can be comparable to or larger than the measured 8-10 keV slice energy spread. This would degrade the compression and, in turn, the temporal-locking analysis that assumes this compression. Please provide quantitative estimates of these effects (or justify their neglect for the specific chicane and charge) and adjust the claims ac
[Electron-to-laser temporal locking; Table 1] The 25 fs rms arrival-time jitter is the output of a simulation in which the jitter inputs (t_THz = 200 fs, t_inj = 200 fs, RF amplitude/phase jitters) are 'independently determined or estimated' and are not measured simultaneously with the presented data. In particular, t_THz is estimated from previous CLARA electro-optic measurements (ref. 37), not from an interleaved measurement in this experiment. The statement 'we find an arrival time jitter of 25 fs with respect to the THz source' is therefore a conditional prediction, not an experimental result. The authors should either measure the electron-laser arrival-time jitter directly (e.g., via THz streaking of the compressed bunch) or clearly label the 25 fs value as a simulation-based estimate with a stated sensitivity to the assumed input jitters.
[Femtosecond bunch compression; Fig. 3] For the single-bunch compression case, the injected bunch parameters (400 fs rms, residual chirp 20 keV/ps, slice energy spread 10 keV FWHM) are obtained from a model fit that assumes a Gaussian temporal profile, linear chirp, and Gaussian slice energy spread. Unlike the long-bunch case, where the chirp and slice spread are extracted model-free from the splitting threshold, the short-bunch input parameters are not independently measured. The good agreement of the modelled timing scan (Fig. 3b) is encouraging but does not validate the compressed bunch length under non-Gaussian tails or higher-order chirp. A sensitivity analysis (e.g., varying the slice energy spread and chirp shape within the fit uncertainty) or a direct bunch-length measurement is needed to support the 15 fs claim.

minor comments (5)

[Abstract and Conclusions] The abstract and conclusions present '15 fs duration' and '25 fs rms arrival-time jitter' in a way that may be read as measured results. Suggest adding explicit qualifiers such as 'predicted' or 'modeled' in these summary statements to distinguish demonstration of the concept from measurement of the compressed beam.
[Fig. 4] The statement that the modelled spectra 'closely match' the measured 50-shot energy fluctuations is not quantified. Provide a quantitative metric (e.g., RMS difference or correlation) and state whether the jitter amplitudes in Table 1 were tuned to achieve this match or were fixed a priori.
[Methods, Table 1] For each jitter entry, please specify the source of the estimate (previous measurement, accelerator model, assumed) and the reference. Currently, several entries cite only [36,37] collectively, which is imprecise.
[Methods, 'Terahertz generation and transport'] Typographical errors: 'T erahertz' in the section heading and 'LPWF A' / 'LPFWA' inconsistencies in the introduction. Also, the missing space in 'T Hz' in several places should be corrected.
[Eq. (1)] The general transfer map is useful, but the implementation in the paper uses only first-order linear dispersion. It would help to state explicitly that higher-order terms and collective forces are neglected, and to list the specific functional forms of f and g used in the simulations.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: compression and jitter results are derived from measured/modelled inputs via standard transport, not preimposed.

full rationale

The paper's central quantitative claims (15 fs compressed bunch, 37 fs microbunches, 25 fs laser-locking) are model outputs obtained by applying first-order magnetic-chicane dispersion to experimentally characterised time-energy distributions. The Methods state: 'The bunch chirp and slice energy spread used in the modelling were obtained directly from THz phase scans using the narrow-peak splitting threshold' and 'the subsequent bunch compression of a chicane was modelled through applying a first-order temporal dispersion to the as-measured electron-beam time-energy distribution.' The compression predictions therefore follow from measured chirp and slice energy spread via standard transport equations; no parameter is fitted to the claimed compressed bunch length. The temporal-locking analysis uses independently estimated rms jitters (Table 1) and symplectic propagation to compute arrival-time jitter relative to the THz source; the 25 fs result is an output of that error-propagation model, not an input. Self-citations (refs 13, 34, 35) support the THz source and waveguide hardware, and refs 36/37 provide accelerator parameters and previous timing measurements; none of these citations carry the burden of the compression or locking predictions, which are self-contained given the quoted measured distributions and standard accelerator physics. The single-bunch model does assume a Gaussian temporal profile and linear chirp, making the 15 fs value model-dependent, but that is an accuracy/verification concern, not circularity.

Assumptions & free parameters 9 free parameters · 5 assumptions · 0 invented entities

The central predictions rest on a compact set of measured parameters (chirp, slice energy spread, THz gain) and on modeling assumptions (ideal chicane, no collective effects, estimated jitter values). The measured parameters are well cross-checked, but the unchicane and jitter assumptions are the weakest links.

free parameters (9)

injected bunch duration (short configuration) = 400 fs rms
Fitted from optimized model match to measured THz phase scan (Fig. 3b).
residual linear chirp (short configuration) = 20 keV/ps
Fitted from model to measured spectra.
time-slice energy spread (short configuration) = 10 keV FWHM
Fitted from spectral width; directly determines the 15 fs compressed bunch length.
peak THz energy gain (short configuration) = 133 keV
Measured directly from spectral peak positions; used in compression model.
injected bunch chirp (long configuration) = 70 keV/ps
Extracted from THz phase scan.
time-slice energy spread (long configuration) = 8 keV rms
Determined from narrow spectral peak width.
peak THz energy gain (long configuration) = 104 keV
Calibrated from splitting threshold.
chicane temporal dispersion D = -3.01 ps/MeV (short), -3.1 ps/MeV (long)
Chosen to optimize compression for the measured chirp.
jitter estimates (Table 1) = tTHz 200 fs, tinj 200 fs, etc.
Estimated from prior CLARA measurements (refs 36,37) and used as inputs to the jitter simulation.

assumptions (5)

standard math Symplectic charge density transformation across accelerator elements; linear transport approximation.
Invoked in Methods 'Charge density distribution evolution'.
domain assumption THz pulse modeled as a sinusoidal multi-cycle modulation with top-hat envelope, phase-locked to the drive laser.
Used in 'Terahertz-frequency bunch trains' and 'Femtosecond bunch compression'; supports the chirp and compression calculations.
domain assumption The magnetic chicane is treated as a first-order linear temporal dispersion with no nonlinear or collective effects.
Stated in Methods 'Compression and bunch train modelling'; not justified against CSR or space charge.
domain assumption Injected short bunch has a Gaussian temporal profile with linear chirp and Gaussian time-slice energy spread.
Used in 'Femtosecond bunch compression' for the optimized model.
domain assumption Jitter sources (Table 1) are independent, Gaussian, with rms values estimated from prior CLARA measurements.
Used in Methods 'Temporal-locking and time-jitter modelling'.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Terahertz control of relativistic electron beams for femtosecond bunching and laser-synchronized temporal locking." pith.science (2026). https://pith.science/paper/YQX7KA6P

@misc{pith2026250820685,
  author       = {Pith},
  title        = {Pith review of: Terahertz control of relativistic electron beams for femtosecond bunching and laser-synchronized temporal locking},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YQX7KA6P}},
  note         = {Machine review of arXiv:2508.20685}
}

read the original abstract

Femtosecond relativistic electron bunches and micro-bunch trains synchronised with femtosecond precision to external laser sources are widely sought for next-generation accelerator and photonic technologies, from extreme UV and X-ray light sources for materials science, to ultrafast electron diffraction and future high-energy physics colliders. While few-femtosecond bunches have been demonstrated, achieving the control, stability and femtosecond-level laser synchronisation remains critically out of reach. Here we demonstrate a concept for laser-driven compression of high-energy (35.5 MeV) electron bunches with temporal synchronisation to a high-power (few-TW) laser system. Laser-generated multi-cycle terahertz (THz) pulses drive periodic electron energy modulation, enabling subsequent magnetic compression capable of generating tuneable picosecond-spaced bunch trains with 30 pC total charge and 50 A peak currents, or to compress a single bunch by a factor of 27 down to 15 fs duration. The THz-driven compression simultaneously drives temporal-locking of the bunch to the THz drive laser, providing a route to femtosecond-level synchronisation, overcoming the timing jitter inherent to radio-frequency accelerators and high-power laser systems. This THz technique offers compact and flexible bunch control with unprecedented temporal synchronisation, opening a pathway to unlock new capabilities for free electron lasers, ultrafast electron diffraction and novel plasma accelerators.

Figures

Figures reproduced from arXiv: 2508.20685 by the authors.

**Figure 1.** Concept for bunching and temporal locking. a Schematic diagram demonstrating the extreme slope in accelerating gradient of high-frequency 0.39 THz fields compared to few-GHz RF fields. Inset showing THz-driven chirping of an electron bunch injected at zero-crossing phase. b Overview of the concept showing the 35.5 MeV electron bunches from the CLARA RF linac injected into the dielectric-lined waveguide (DLW) for THz… view at source ↗

**Figure 2.** Multi-cycle energy modulation for micro-bunch trains. a The measured THz-modulated electron energy spectra for varying THz pulse energy using approximately linearly-chirped 2.5 ps rms injected bunches. The narrow-peak splitting threshold at 15% of the maximum THz energy is highlighted in red, with the corresponding b measured time-delay scan (inset showing the extracted bunch chirp), c modelled fitting to the modula… view at source ↗

**Figure 3.** Sub-cycle THz interactions for single-bunch compression. For 400 fs rms injected bunches, the a measured and b modelled energy spectra as a function of THz-electron injection time. Selected spectral projections c without THz interaction and with THz d zero-crossing, e decelerating and f accelerating phase, with g-j the corresponding modelled time-energy density distributions. From the zero-crossing phase with maximu… view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: Laser-driven temporal locking of femtosecond bunch trains and single bunches. a, Observed electron energy spectra (50 shots) using the CLARA long-bunch configuration with the highest energy THz modulation, and the corresponding modelled spectra including jitter in the…

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

Controlling external injection in laser-plasma accelerators with terahertz frequency bunch manipulation
physics.acc-ph 2026-04 unverdicted novelty 5.0 of 10

Terahertz-frequency bunch manipulation enables sub-10 fs compression and temporal locking for external injection into LWFA, yielding simulated GeV acceleration with 0.2% energy jitter and spread.

Reference graph

Works this paper leans on

37 extracted references · 36 canonical work pages · cited by 1 Pith paper

[1]

Guo, Z. et al. Experimental demonstration of at- tosecond pump–probe spectroscopy with an X-ray free-electron laser. Nat. Photon.18, 691–697 (2024)

work page 2024
[2]

Salen, P. et al. State-of-the-art electron beams for compact tools of ultrafast science. Ultramicroscopy 268, 114080 (2025)

work page 2025
[3]

Mirzaie, M. et al. All-optical nonlinear Compton scattering performed with a multi-petawatt laser. Nat. Photon. 18, 1212–1217 (2024)

work page 2024
[4]

Wu, Y. et al. High-throughput injection–acceleration of electron bunches from a linear accelerator to a laser wakefield accelerator. Nat. Phys. 17, 801–806 (2021)

work page 2021
[5]

Maxson, J. et al. Direct Measurement of Sub-10 fs Relativistic Electron Beams with Ultralow Emit- tance. Phys. Rev. Lett.118, 154802 (2017)

work page 2017
[6]

Zhao, L. et al. Few-femtosecond electron beam with terahertz-frequency wakefield-driven compres- sion. Phys. Rev. Accel. Beams21, 082801 (2018)

work page 2018
[7]

Kim, H. W. et al. Towards jitter-free ultrafast electron diffraction technology. Nat. Photon. 14, 245–249 (2020)

work page 2020
[8]

Qi, F. et al. Breaking 50 femtosecond resolution bar- rier in MeV ultrafast electron diffraction with a dou- ble bend achromat compressor. Phys. Rev. Lett.124, 134803 (2020)

work page 2020

Show all 37 references

[9]

Yang, Y. et al. Sub-5-fs compression and synchro- nization of relativistic electron bunches enabled by a high-gradient-magnet and low-jitter photoinjector. Preprint available at: arXiv:2508.03946 [physics.acc- ph]

arXiv
[10]

Pompili, R. et al. Femtosecond timing-jitter between photo-cathode laser and ultrashort electron bunches by means of hybrid compression. New J. Phys. 18, 083033 (2016)

2016
[11]

Lautenschlager, B. et al. Arrival time stabilization at FLASH using the bunch arrival corrector cavity (BACCA). Proc. IPAC’21, 2194-2196 (2021)

2021
[12]

Winkler, P. et al. Active energy compression of a laser-plasma electron beam. Nature 640, 907 (2025)

2025
[13]

Hibberd, M. T. et al. Acceleration of relativistic beams using laser-generated terahertz pulses. Nat. Photon. 14, 755–759 (2020)

2020
[14]

& Gover, A

Curry, E., Fabbri, S., Maxson, J., Musumeci, P. & Gover, A. Meter-scale terahertz-driven acceleration of a relativistic beam. Phys. Rev. Lett.120, 094801 (2018)

2018
[15]

Xu, H. et al. Cascaded high-gradient terahertz- driven acceleration of relativistic electron beams. Nat. Photon. 15, 426-430 (2021)

2021
[16]

Tang, H. et al. Stable and scalable multistage terahertz-driven particle accelerator. Phys. Rev. Lett. 127, 074801 (2021)

2021
[17]

Yu, X.-Q. et al. Megaelectronvolt electron accelera- tion driven by terahertz surface waves. Nat. Photon. 17, 957-963 (2023)

2023
[18]

Nanni, E. A. et al. Terahertz-driven linear electron acceleration. Nat. Commun. 6, 8486 (2015)

2015
[19]

Walsh, D. A. et al. Demonstration of sub-luminal propagation of single-cycle terahertz pulses for par- ticle acceleration. Nat. Commun. 8, 421 (2017)

2017
[20]

Zhang, D. et al. Segmented terahertz electron accel- erator and manipulator (STEAM). Nat. Photon.12, 336–342 (2018)

2018
[21]

Zhang, D. et al. Femtosecond phase control in high- field terahertz-driven ultrafast electron sources. Op- tica 6, 872-877 (2019)

2019
[22]

Zhang, D. et al. Cascaded multicycle terahertz- driven ultrafast electron acceleration and manipula- tion. Phys. Rev. X10, 011067 (2020)

2020
[23]

Nix, L. J. R. et al. Terahertz-driven acceleration of subrelativistic electron beams using tapered rectan- gular dielectric-lined waveguides. Phys. Rev. Accel. Beams 27, 041302 (2024)

2024
[24]

Huang, W. R. et al. Terahertz-driven, all-optical electron gun. Optica 3, 1209–1212 (2016)

2016
[25]

Ying, J. et al. High gradient terahertz-driven ultra- fast photogun. Nat. Photon. 18, 758–765 (2024)

2024
[26]

Kealhofer, C. et al. All-optical control and metrology of electron pulses. Science 352, 429-433 (2016)

2016
[27]

Zhao, L. et al. Terahertz streaking of few- femtosecond relativistic electron beams. Phys. Rev. X 8, 021061 (2018)

2018
[28]

Zhao, L. et al. Terahertz oscilloscope for recording time information of ultrashort electron beams. Phys. Rev. Lett.122, 144801 (2019)

2019
[29]

Li, R. K. et al. Terahertz-based subfemtosecond metrology of relativistic electron beams. Phys. Rev. Accel. Beams22, 012803 (2019)

2019
[30]

Wang, Z. et al. Femtosecond diagnostics of ultrashort electron beam using a pinhole-based THz-driven de- flector. Nucl. Instrum. Methods Phys. Res. A1040, 167279 (2022). 8

2022
[31]

Song, Y. et al. MeV electron bunch compression and timing jitter suppression using a THz-driven res- onator. Nucl. Instrum. Methods Phys. Res. A1047, 167774 (2023)

2023
[32]

Snively, E. C. et al. Femtosecond compression dy- namics and timing jitter suppression in a THz-driven electron bunch compressor. Phys. Rev. Lett. 124, 054801 (2020)

2020
[33]

Zhao, L. et al. Femtosecond relativistic electron beam with reduced timing jitter from THz driven beam compression. Phys. Rev. Lett. 124, 054802 (2020)

2020
[34]

Mosley, C. D. W. et al. Large-area periodically-poled lithium niobate wafer stacks optimized for high- energy narrowband terahertz generation. Opt. Ex- press 31, 4041 (2023)

2023
[35]

Dalton, P. J. et al. Cryogenically cooled periodi- cally poled lithium niobate wafer stacks for multi- cycle terahertz pulses. Appl. Phys. Lett.125, 141101 (2024)

2024
[36]

Angal-Kalinin, D. et al. Design, specifications, and first beam measurements of the compact linear accel- erator for research and applications front end. Phys. Rev. Accel. Beams23, 044801 (2020)

2020
[37]

Walsh, D. A. et al. Demonstration of an electro-optic spectral interferometry longitudinal profile monitor at CLARA. Proc. IPAC’23, 4855-4858 (2023). 9 Methods Electron beam. The Compact Linear Accelerator for Research and Applications (CLARA) test facility at Dares- bury Labo...

2023

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.

[1] [1]

Guo, Z. et al. Experimental demonstration of at- tosecond pump–probe spectroscopy with an X-ray free-electron laser. Nat. Photon.18, 691–697 (2024)

work page 2024

[2] [2]

Salen, P. et al. State-of-the-art electron beams for compact tools of ultrafast science. Ultramicroscopy 268, 114080 (2025)

work page 2025

[3] [3]

Mirzaie, M. et al. All-optical nonlinear Compton scattering performed with a multi-petawatt laser. Nat. Photon. 18, 1212–1217 (2024)

work page 2024

[4] [4]

Wu, Y. et al. High-throughput injection–acceleration of electron bunches from a linear accelerator to a laser wakefield accelerator. Nat. Phys. 17, 801–806 (2021)

work page 2021

[5] [5]

Maxson, J. et al. Direct Measurement of Sub-10 fs Relativistic Electron Beams with Ultralow Emit- tance. Phys. Rev. Lett.118, 154802 (2017)

work page 2017

[6] [6]

Zhao, L. et al. Few-femtosecond electron beam with terahertz-frequency wakefield-driven compres- sion. Phys. Rev. Accel. Beams21, 082801 (2018)

work page 2018

[7] [7]

Kim, H. W. et al. Towards jitter-free ultrafast electron diffraction technology. Nat. Photon. 14, 245–249 (2020)

work page 2020

[8] [8]

Qi, F. et al. Breaking 50 femtosecond resolution bar- rier in MeV ultrafast electron diffraction with a dou- ble bend achromat compressor. Phys. Rev. Lett.124, 134803 (2020)

work page 2020