Skip to content

An x-ray regenerative amplifier free-electron laser using diamond pinhole mirrors

MetadataDetails
Publication Date2019-08-30
JournalNew Journal of Physics
AuthorsH P Freund, P J M van der Slot, Yu Shvyd’ko
InstitutionsArgonne National Laboratory, University of Twente
Citations31
AnalysisFull AI Review Included
  • Core Innovation: Proposes a high-gain, low-Q Regenerative Amplifier Free-Electron Laser (RAFEL) utilizing a six-mirror resonator composed entirely of diamond crystal Bragg reflectors, departing from traditional low-gain XFELO designs.
  • Outcoupling Strategy: Utilizes a pinhole (optimum radius of 135 µm) in a 100 µm-thick diamond mirror to achieve 90% power extraction, circumventing the physical impossibility of using transmissive diamond crystals at hard x-ray energies.
  • Target Output: Predicts 21.2 µJ pulse energy (880 MW peak power per pulse) at the fundamental 3.05 keV photon energy, with an average power of 21.2 W at a 1 MHz repetition rate.
  • Harmonic Generation: Simultaneously produces a 3rd harmonic output at 9.15 keV, yielding 0.067 µJ pulse energy and 67 mW average power from a parasitic single-pass generation.
  • Beamline Integration: Designed for compatibility with the LCLS-II HXR undulator, utilizing 11 segments (47.18 m total length) to achieve saturation within 15-25 passes.
  • Performance Stability: Reduces typical SASE shot-to-shot power fluctuations from 10-20% down to ~3% rms pass-to-pass fluctuations, with spectral linewidth contracting from 3.7 × 10-4 to 6.0 × 10-5 at saturation.
  • Simulation Framework: Validates the design using a modified MINERVA/OPC simulation chain, where the Optics Propagation Code was adapted to handle Bragg reflection from diamond crystal mirrors.
ParameterValueUnitContext
Fundamental Photon Energy3.05keVPrimary output wavelength
3rd Harmonic Photon Energy9.15keVParasitic harmonic output
Repetition Rate (frep)1.0MHzLCLS-II electron bunch rate
Electron Bunch Duration24fsFlat-top temporal profile
Bunch Charge24pCCalculated from 1000 A peak current
Peak Current1000APer electron bunch
Synchronous Cavity Length (Lvac)299.7924580mRoundtrip zero-detuning length
Cavity Detuning Tolerance± 7.2µmLimited by electron bunch length (~7.2 µm)
Pinhole Radius (Optimum)135µmFor 90% outcoupling efficiency
Diamond Mirror Thickness100µmFirst downstream mirror (C6)
Undulator Length (Optimum)47.18m11 HXR undulator segments
Inter-Mirror Distance (L)130mDistance framing the undulator
RMS Energy Spread0.0125%Nominal LCLS-II value (ΔE/E = 1.25 × 10-4)
Output Pulse Energy (Fundamental)21.2µJAt peak of detuning curve
Peak Output Power (Fundamental)880MWPer pulse
Average Output Power (Fundamental)21.2WLong-term average at 1 MHz
3rd Harmonic Pulse Energy0.067µJSaturated output
Average Power (3rd Harmonic)67mWLong-term average
Output Power Fluctuations~3% rmsPass-to-pass stability
Saturated Linewidth (Δλ/λ)6.0 × 10-5-Fundamental, at saturation
Initial Linewidth (Pass 1)3.7 × 10-4-SASE-like before coherence develops
M² Beam Quality Factor~1-High purity, optically guided mode
  1. Simulation Platform Integration: Linked the MINERVA 3D time-dependent FEL simulation code with the Optics Propagation Code (OPC), modifying OPC to handle Bragg reflection from diamond crystal mirrors rather than standard free-space Gaussian propagation.
  2. Beamline Configuration: Modeled the RAFEL on the LCLS-II HXR undulator line, using 11 undulator segments with a total optimized length of 47.18 m to maintain high-gain operation.
  3. Pinhole Outcoupling Design: Evaluated transmissive outcoupling (ruled out due to impractical ~5 µm crystal thickness) and selected pinhole outcoupling through a 100 µm-thick diamond crystal (C6) with an optimized 135 µm radius for 90% extraction.
  4. Steady-State Optimization: Conducted time-independent sweeps to optimize hole radius and undulator length, determining that 11 segments and 135 µm hole radius maximized saturated output power.
  5. Cavity Synchronization: Calculated the zero-detuning cavity length (Lvac = c / frep = 299.7924580 m) and mapped the detuning curve, confirming gain over a ±7.2 µm window corresponding to the electron bunch length.
  6. Time-Dependent Simulation: Modeled start-up from shot noise on the electron beam, tracking pulse energy, spectral linewidth, and 3rd harmonic growth over multiple passes until saturation at 15-25 passes.
  7. Harmonic Analysis: Quantified parasitic 3rd harmonic generation (9.15 keV) as a function of fundamental pulse energy, noting rapid growth after the fundamental exceeds 1 µJ.
  • Applicable Materials: This RAFEL architecture requires ultra-high-purity, low-absorption Optical Grade Single Crystal Diamond (SCD) for all six Bragg reflector mirrors. The critical requirements—withstanding high thermal/radiation loads, sustaining sub-meV Bragg reflectivity, and being processable into 100 µm-thick pinhole substrates with sub-µm tolerances—align directly with 6CCVD’s premium SCD offerings. For any proposed transmissive variants or beamline diagnostics, our Polycrystalline Diamond (PCD) plates provide a robust, scalable alternative.
  • Customization Potential: The paper specifies non-standard diamond geometries, including 100 µm-thick pinhole mirrors with precise 135 µm drilled apertures. 6CCVD delivers custom laser cutting, shaping, and ultra-precision drilling for thin-film diamond components. Furthermore, our internal metalization capabilities (Ti, Pt, Au, W) are available for any hybrid mirror-mount assemblies, cryogenic cooling interfaces, or rf cavity integration required for beamline hardware.
  • Diamond Thickness & Wafer Range: The SCD thickness range required (0.1 µm - 500 µm) for the pinhole Bragg reflectors falls squarely within 6CCVD’s manufacturing envelope. We can produce free-standing diamond membranes down to sub-micron thicknesses or thicker stress-free crystals up to 500 µm, polished to Ra < 1 nm to minimize scattering losses at x-ray wavelengths.
  • Engineering Support: The design of next-generation XFELO and RAFEL beamlines demands diamond crystals with exceptional crystallographic perfection, controlled strain fields, and well-characterized rocking curves. 6CCVD’s in-house PhD engineering team provides end-to-end material consultation, from crystal orientation selection and Bragg-plane alignment to thermal load modeling, assisting researchers in specifying the optimal diamond grade for synchrotron and FEL applications.
  • Call to Action: For custom specifications, diamond membrane fabrication, or material consultation on x-ray optics and FEL beamline components, visit 6ccvd.com or contact our engineering team directly.
View Original Abstract

Abstract X-ray free-electron lasers (FELs) rely on SASE due to the lack of seed lasers and the difficulty in obtaining mirrors. Progress in diamond crystal Bragg mirrors enables the design of x-ray FEL oscillators. Regenerative amplifiers (RAFELs) are high gain/low-Q oscillators that out-couple most of the optical power. An x-ray RAFEL based on the LCLS-II at SLAC using a six-mirror resonator out-coupling 90% or more through a pinhole in the first downstream mirror is analyzed using the MINERVA simulation in the undulator and OPC for the resonator. Results show substantial powers at the fundamental (3.05 keV) and 3rd harmonic (9.15 keV).