15th International Particle Accelerator Conference,Nashville, TN JACoW Publishing

ISBN: 978-3-95450-247-9 ISSN: 2673-5490

doi: 10.18429/JACoW-IPAC2024-MOPG48

OVERVIEW OF R&D ACTIVITIES

IN THE PRODUCTION OF HIGH ENERGY PHOTON BEAMS
FOR FUTURE USER EXPERIMENTS BEYOND 25 keV AT THE EUXFEL
Content from this work may be used under the terms of the CC BY 4.0 licence (© 2024). Any distribution of this work must maintain attribution to the author(s), title of the work, publisher, and DOI.

Y. Chen1∗ , F. Brinker1 , W. Decking1 , I. Inoue2,3 , T.-Y. Long1 , J.-W. Yan4 , Z.-H. Zhu1†
1 Deutsches Elektronen-Synchrotron DESY, Notkestr. 85, 22607 Hamburg, Germany
2 RIKEN SPring-8 Center, 1-1-1 Kouto, Sayo, Hyogo 679-5148, Japan
3 University of Hamburg, CFEL, Luruper Chaussee 149, 22761 Hamburg, Germany
4 European XFEL, Holzkoppel 4, 22869 Schenefeld, Germany

Abstract
thermore, demonstrated a unique capability of the facility in
Scientific opportunities with very hard XFEL radiation combining a high energy linac and long flexible undulators
demands dedicated facility development towards FEL oper- for achieving decent lasing intensities of even harder X-rays.
ation in the sub-ångström regime. Very hard X-rays provide Interested readers are referred to [1] for more details.
capabilities of high Q-range coverage and high penetration,
and also allow access to the K-edge spectroscopy of high-
MILESTONES
𝑍 elements. Production of such X-rays using FELs takes
advantage of general FEL characteristics such as large coher- Proof-of-principle experiments have been performed in
ence, short pulse option, variable pump-probe delay control the past years at the EuXFEL, aiming to explore the ma-
and higher brightness compared to conventional storage ring chine performance in the sub-ångström radiation regime. In
sources. R&D activities in the characterization and pro- Table 1, a summary on the working progress is given. As
duction of high energy photon beams beyond 25 keV has shown, FEL lasing at 30 keV in the fundamental mode is
been launched since 2021 at the EuXFEL. Photon beams of made possible since 2021. The highest achievable photon
30 keV have been produced, characterized and delivered to energy so far is about 30 keV with a SASE intensity of about
experimental hutches. In this paper, we give an overview of 340 µJ (see [8] for details). Figure 1 shows, for example,
the overall development. Obtained results will be discussed. the measured SASE spectrum at about 24 keV (left) and
30 keV (right) using a Hard x-ray single-shot spectrome-
INTRODUCTION ter [9]. The insets illustrate the FEL beam spots observed
downstream on the imagers. Dedicated photon beam line
The generation of femtosecond pulses of very hard X-rays studies were also carried out, through which the 30 keV FEL
(i.e., at least 12.4 keV in photon energy) at coherent light beams were successfully transported to the experimental
source facilities opens up unprecedented scientific opportuni- hutches with reasonably high transmission rates. A first test
ties in probing matters and materials at extremely fine spatial user-experiment was conducted at the SASE1 undulator line
(atomic-scale) and temporal (femto- and attoseconds scale) using 24 keV FEL beams. Overall, great progress was made.
resolutions. Great progresses have been made at existing During these experiments more physical and / or technical
X-ray Free-Electron Laser (FEL) facilities worldwide [1–5]. challenges have been unveiled. Two major challenges are
Towards even harder X-rays beyond 25 keV (i.e. 0.5 Å in described in the next section.
photon wavelength), more physical and technical challenges
must be confronted and overcome in order to push forward CHALLENGES
the frontier of delivering such X-rays to the user experiments,
inspiring discoveries in the fundamental fields [6]. The accessible photon wavelength via the Self-Amplified
The European XFEL (EuXFEL) is a hard X-ray FEL based Spontaneous Emission (SASE) process depends on electron
on a high-electron-energy superconducting linear acceler- beam energy, undulator parameter and undulator period. An
ator [1]. It was commissioned in 2017. The first lasing inverse scaling of the photon wavelength with the square of
was reported in May 2017. Shortly after that, the facility the beam energy conceptually defines shorter wavelengths
started its user operation. At a nominal electron beam en- at given higher beam energies. The undulator parameter
ergy of 14 GeV, stable, high-intensity SASE performance is is proportional to the product of undulator period and the
achieved at photon energies up to 20 keV in user runs. In undulator magnetic field. Given a fixed undulator period, a
2020, it was reported, for the first time, that lasing signals at larger undulator gap results in weaker magnetic field, which
25 keV were observed [7]. The FEL performance in terms of is characterized by a smaller undulator parameter, resulting
highest achievable photon energies expected in the original in a shorter photon wavelength. Reducing the undulator
technical design report has been surpassed. This has, fur- period naturally shifts the emission regime towards shorter
photon wavelengths. In addition, the lasing performance of
∗ [email protected] an FEL depends on the achievable qualities of the electron
† now at SLAC National Accelerator Laboratory beams in the undulators.

MOPG48 MOPG: Monday Poster Session: MOPG

392 MC2.A06 Free Electron Lasers
 15th International Particle Accelerator Conference,Nashville, TN JACoW Publishing
ISBN: 978-3-95450-247-9 ISSN: 2673-5490 doi: 10.18429/JACoW-IPAC2024-MOPG48

Table 1: Milestones of Proof-of-Principle Experiments at the European XFEL Generating Very Hard X-Rays
X-ray Beamline Photon energy, Eph Pulse energy, Jph Transport to exp. hutches Exp. time
n/a [keV] [µJ] n/a Month/Year
SASE1 25 lasing signal [1] no Apr. 2020
SASE1 24 810 yes Oct. 2021

Content from this work may be used under the terms of the CC BY 4.0 licence (© 2024). Any distribution of this work must maintain attribution to the author(s), title of the work, publisher, and DOI.
SASE2 30 340 no Oct. 2021
SASE2 30 40 [2] yes Nov. 2022
SASE1 30 200 no Oct. 2023
SASE2 30 120 yes Nov. 2023
[1] Lasing signals were seen on an FEL imager downstream the undulators. No SASE tuning was carried out for increasing the obtained intensity.
[2] This was for dedicated photon beamline studies; Low requirement was made intensity-wise and thus no dedicated FEL tuning was carried out.

Figure 1: Measured SASE spectrum at about 24 keV (left) and 30 keV (right).

The FEL gain length increases with the increase of the 1 mm in diameter. The emittance measurement is usually
transverse emittance and uncorrelated energy spread of the done using on-crest phasing of the first accelerating module
electron bunch, and decreases with an increase in the bunch and the third harmonics module of the injector. A detailed
peak current. The obtained gain length should be sufficiently description can be found in [10].
small so that the SASE process at a resonant wavelength
Figure 2 shows long-term statistics of the measurement
can grow within a given undulator length until saturation.
data during emittance optimization in the injector. A multi-
An undulator length of typically 20 times the obtained gain
quadruple scan method is used [11]. As shown, an average
length allows for the generation of high-power X-rays at a
emittance of about 0.38 µm is usually obtained after the
requested wavelength. Note that longer undulators would
optimization. Specifically, for the FEL experiments as listed
also add more challenges for an accurate alignment.
in Table 1, the optimized emittance was in the range from
The main challenges for extending the operational regime 0.31 µm to 0.34 µm. Using such low-emittance beams, it
of the EuXFEL towards higher photon energies in the fun- turned out the FEL had lased at 30 keV with good intensities.
damental mode mainly include: (A) further improvement in
the electron bunch quality and (B) further improvement in Despite the obtained results with 30 keV lasing in the fun-
the undulator alignment. Better bunch quality shortens the damental, there is still much room for improvement in the
FEL gain length. Well-aligned undulators allow the SASE transverse bunch emittance. Numerical studies have been
process to grow adequately until saturation. carried out to explore different working scenarios and con-
ditions. In [12], an optimal transverse distribution of the
electron bunch, namely, truncated-Gaussian (TG), is used for
Low-Emittance Electron Beams
optimizing a charge packet of 100 pC in the injector. Such
Reducing the transverse emittance of the electron bunch a TG distribution can efficiently linearize the space-charge
can be the most efficient way to access to higher photon fields during the photoemission and reduce the space-charge
energies (shorter wavelengths) for an existing SASE FEL. induced emittance contribution [13]. It is worth mentioning,
A prerequisite requirement thus lies on the emittance opti- that an optimized emittance of about 0.23 µm is obtained
mization for the low energy beams in the photoinjector. At at 100 pC following this approach, indicating another 30%
the EuXFEL, the emittance optimization is conducted for reduction compared to the presently measured emittance
250 pC electron bunches energized to 130 MeV. A typical op- in the injector. Furthermore, numerical studies in [14] re-
timization approach considers the peak cathode accelerating garding the emittance optimization for an superconducting
gradient (up to 60 MV/m), the radio-frequency (rf) gun phase radio-frequency (SRF) gun based continuous-wave (CW)
and the gun solenoid strength. The cathode drive laser pulse photoinjector, have demonstrated that the overall emittance
has a temporal Gaussian distribution of 5–7 ps in full width of a 100 pC bunch can be reduced to its thermal emittance
half maximum and a transverse uniform or semi-truncated level (i.e. the fundamental limit for a smallest possible emit-
Gaussian distribution with a beam shaping aperture size of tance), under the conditions that the bunch has a longitudinal

MOPG: Monday Poster Session: MOPG MOPG48

MC2.A06 Free Electron Lasers 393
 15th International Particle Accelerator Conference,Nashville, TN JACoW Publishing

Content from this work may be used under the terms of the CC BY 4.0 licence (© 2024). Any distribution of this work must maintain attribution to the author(s), title of the work, publisher, and DOI. ISBN: 978-3-95450-247-9 ISSN: 2673-5490 doi: 10.18429/JACoW-IPAC2024-MOPG48

Figure 2: Long-term measurement history of transverse projected emittance of the electron bunch in the injector at 250 pC.

flat-top shape and a transverse TG shape. The required tem- alignment of the last few undulator cells. To further improve
poral and transverse shapes can be principally realized by the undulator alignment accuracy, both electron-beam based
shaping the cathode drive laser pulses. This promising ap- alignment [15] and photon-beam based alignment [16] meth-
proach will also be explored in the future experiments. ods are developed and implemented. A technical goal is set
to reduce the misalignment error to a level below 10 µm.
Undulator Alignment
The hard X-ray undulator line of the EuXFEL consists USER PERSPECTIVES
of 37 cells. Each cell is composed of a 5 m long undulator Hard X-ray FEL pulses with photon energy above 25 keV
and a 1.1 m long intersection. Well-aligned undulators are from EuXFEL will open a door for various new science
of great importance to the overlap of photon and electron cases. Firstly, the high-energy XFEL will broaden the scope
beams and thus for the FEL performance. This is crucial as of diffraction experiments. Although serial femtosecond
approaching to the emission regime at 25+ keV, where the crystallography is one of the most popular experiments at
tolerances to machine uncertainties become much tighter. current FEL facilities, the targets of these experiments have
Figure 3 shows the measured SASE pulse energies along mostly been limited to macromolecular crystals. These crys-
the undulators. The data are extracted from the experiments tals possess large unit cell parameters, resulting in a suffi-
in Oct. 2021 and Nov. 2023 (see Tab.1). As illustrated, the cient number of diffraction spots on the detector to determine
final pulse energy by the undulator cell 31 was limited to crystal orientation even when selecting low photon energies
about 0.12 mJ in Experiment 2023 (red bar), and it was ob- (∼10 keV). The very high-energy X-ray pulses from EuXFEL
served that, the last few cells downstream cell 31 did not will expand the accessible scattering vector (Q) range and
contribute to the lasing process. Pulse-energy-wise, this extend the capabilities of damage-free structure determina-
result is consistent with the intensity level that was achieved tion to more general samples, such as organic materials and
at the same location in Experiment 2021 (blue bars). The organic-inorganic hybrid materials, both known for their
difference is that the last few cells did contribute in the 2021 high radiation-damage sensitivity. The broad Q-range ac-
experiment and raised the final intensity to about 0.34 mJ cessible with high-energy FELs is also advantageous for
by the cell 37. It is empirically known that the optimized detailed structure analysis beyond simple determination of
electron bunch qualities for these two experiments were sim- atomic configurations. Measurement of diffraction intensity
ilar, as are the obtained peak currents downstream the last across a wide Q-range allows for imaging of density distri-
bunch compression stage. Assuming reasonable beam trans- butions of valence electrons. By combined with external
port through the main linac until the undulator entrance, the pumping sources, such as femtosecond optical lasers, the vi-
observed discrepancy in the lasing performance between sualization of valence electrons in action will be observable.
these two experiments was very likely attributed to the mis- The high-energy XFEL also offers significant benefits
for spectroscopic studies. The photon energies it provides
can cover the K-edges of high-𝑍 elements including 4d and
5d transition metals, which are essential for material sci-
ence. These elements serve as the foundation for many
man-made catalysis processes and strongly correlated mate-
rials that exhibit unique quantum phases. The time-resolved
measurements of charge dynamics within these systems via
absorption/emission spectroscopy and inelastic scattering
techniques will provide a route to find out the detailed mech-
anisms underlying the emergence of their properties.

ACKNOWLEDGEMENTS
Figure 3: SASE pulse energies along the undulator cells as We thank all the colleagues in the European XFEL and
measured in the experiments 2021 and 2023. DESY operation teams.

MOPG48 MOPG: Monday Poster Session: MOPG

394 MC2.A06 Free Electron Lasers
 15th International Particle Accelerator Conference,Nashville, TN JACoW Publishing
ISBN: 978-3-95450-247-9 ISSN: 2673-5490 doi: 10.18429/JACoW-IPAC2024-MOPG48

REFERENCES
[10] F. Brinker, "Commissioning of the European XFEL Injec-
[1] W. Decking et al., "A MHz-repetition-rate hard X-ray free-
tor", In Proceedings of the International Particle Accelera-
electron laser driven by a superconducting linear accelera-
tor Conference (IPAC’16), Busan, Korea, 8–13 May 2016.
tor", Nat. Photonics 2020, 14, 391–397. doi:10.1038/
doi:10.18429/JACoW-IPAC2016-TUOCA03

Content from this work may be used under the terms of the CC BY 4.0 licence (© 2024). Any distribution of this work must maintain attribution to the author(s), title of the work, publisher, and DOI.
s41566-020-0607-z
[11] S. Meykopff and B.Beutner, "Emittance Measurement
[2] E. Prat et al., "A compact and cost-effective hard X-ray free- and Optics Matching at the European XFEL", In Pro-
electron laser driven by a high-brightness and low-energy ceedings of the 16th International Conference on Ac-
electron beam", Nat. Photonics 2020, 14, 748–754. doi: celerator and Large Experimental Physics Control Sys-
10.1038/s41566-020-00712-8 tems, Barcelona, Spain, 8-13 Oct 2017. doi:10.18429/
[3] P. Emma et al., "First lasing and operation of an ångstrom- JACoW-ICALEPCS2017-THPHA116
wavelength free-electron laser", Nat. Photonics 2010, 4, 641– [12] Y. Chen, F. Brinker, W. Decking, M. Scholz, and L.
647. doi:10.1038/nphoton.2010.176 Winkelmann, "Perspectives towards Sub-Ångström Work-
[4] T. Ishikawa et al., "A compact X-ray free-electron laser emit- ing Regime of the European X-ray Free-Electron Laser with
ting in the sub-ångström region", Nat. Photonics 2012, 6, Low-Emittance Electron Beams", Applied Sciences 2021, 11,
540–544. doi:10.1038/nphoton.2012.141 10768. doi:10.3390/app112210768

[5] H. Kang et al., "Hard X-ray free-electron laser with [13] Y. Chen et al., "Budgeting the emittance of photoemitted
femtosecond-scale timing jitter", Nat. Photonics 2017, 11, electron beams in a space-charge affected emission regime
708–713. doi:10.1038/s41566-017-0029-8 for free-electron laser applications", AIP Advances 2020, 10,
035017. doi:10.1063/1.5129532
[6] The Workshop on Scientific Opportunities with Very Hard
XFEL Radiation, Jan 18—20, 2023, Hamburg, https:// [14] D. Bazyl, Y. Chen, M. Dohlus, and T. Limberg, "CW
indico.desy.de/event/33463/overview Operation of the European XFEL: SC-Gun Injector Op-
timization, S2E Calculations and SASE Performance",
[7] European XFEL News, "European XFEL reaches world arXiv:2111.01756. doi:10.48550/arXiv.2111.01756
record photon energies", 8 Apr. 2020.
[15] M. Scholz, W. Decking, and Y. Li, "Beam Based Align-
[8] Y. Chen et al., "Virtual commissioning of the European XFEL ment in all Undulator Beamlines at European XFEL", In
for advanced user experiments at photon energies beyond 25 Proceedings of the International Free Electron Laser Con-
keV using low-emittance electron beams", Journal of Physics: ference (FEL’19), Hamburg, Germany, August 2019. doi:
Conference Series 2023, 2420, 012026. doi:10.1088/ 10.18429/JACoW-FEL2019-THP002
1742-6596/2420/1/012026
[16] W. Freund et al., "First measurements with the K-
[9] N. Kujala et al., "Hard x-ray single-shot spectrometer at the monochromator at the European XFEL", Journal of Syn-
European X-ray Free-Electron Laser", Review of Scientific chrotron Radiation 26, 1037 (2019). doi:10.1107/
Instruments 2020, 91, 103101. doi:10.1063/5.0019935 S1600577519005307

MOPG: Monday Poster Session: MOPG MOPG48

MC2.A06 Free Electron Lasers 395