By Zhirong Huang, SLAC National Accelerator Laboratory, and Timur Shaftan, Brookhaven National Laboratory
2024 has become a significant year in the light source community (Fig. 1) with two premier projects finishing their commissioning and starting user operation. APS-U, the brightest storage ring source ever, was successfully commissioned within two months and began sending photons to its first set of beamlines in June. A new superconducting-accelerator-based FEL, LCLS-II, has finished commissioning and started user experiments in June. LCLS-II has opened the new page of XFEL development, transitioning from conventional FELs based on copper linacs, such as LCLS, PAL, SACLA and SwissFEL operating at 102 pulses per second to eventually 106 pulses per second, surpassing most powerful EuXFEL by two orders of magnitude.
Figure 1: The map of light sources worldwide (https://lightsources.org/lightsources-of-the-world/). Orange pins on the map represent members of the lightsources.org collaboration.
The growth of light sources continues with HEPS commissioning this month and Diamond, SOLEIL and SPRING-8 entering the construction phase. Another continuous-wave, high-repetition FEL, SHINE, is expected to generate first light in 2025.
As the community continues to grow and expand on the capabilities, it makes sense to look at the basic figures of merit of light source operations and discuss the outlook to the future. These two topics are the focus of our newsletter. The recent Basic Research Needs workshop (BRN) hosted by DOE BES (https://science.osti.gov/bes/Community-Resources/Reports), participated by the authors, provided insights into the future developments and needs of the light source community and informed us on the priority directions in accelerator physics and technology. We quote the BRN-2023 (Fig. 2) conclusions throughout this paper.
Figure 2: DOE-BES Basic Research Needs Workshop 2023 brochure, highlighting five priority research directions
We begin by presenting the two main performance parameters of any light source, taking performance as the rate of scientific output of the facility, i.e. both the number of publications and patents, as well as their depth and impact in the field. The first one is the facility capacity. The second one is the extent of capabilities.
The capacity of beamlines at the light source defines the throughput of the samples from the processing to characterization and analysis. This equates to the productivity of the facility as the number of publications is a strong function of the number of operating beamlines. In searching for opportunities for expanding the beamline suite, the storage ring light sources strive to multiply available straight sections or build canted IDs in a single straight, while XFELs are planning undulator farms with beam switchyards to take advantage of high repetition rate of the superconducting linear accelerators. A multiplexed XFEL enables a synchrotron-like mode of operation, serving a broad community with tailored X-ray properties and optimized beamlines for distinct scientific areas. (BRN direction “Tailor and control beams with unprecedented precision and speed to probe complexity in matter”).
The capabilities at a light source define the depth of the research, as well as open new opportunities for cross-cutting analysis of the samples. While some beamlines do not benefit from the new technologies or new capabilities, including the next levels of brightness or special timing modes, the instruments that aim to achieve high spatial, temporal or energy resolution ultimately depend on the advanced capabilities at the light source.
The major aim of evolutionary advances of the modern light sources are the quality and the intensity of radiation, both translated into a single quantity known as “brightness”. Brightness is defined as “the phase-space density of the photon flux, evaluated in the forward direction and at the center of the source” [KJK] and is a product of the radiation intensity per unit of photon energy divided by the beam size and divergence in the transverse planes. As presented in the conclusions of the recent RBN workshop, reaching theoretical limit of brightness, given, on one hand, by the capabilities of particle accelerators and, on another, by the diffraction or Fourier-transform limit of the radiation phase-space is the priority research direction in the next decade. (BRN direction “Realize next generation capabilities that achieve theoretical performance limit”).
The current storage rings and XFELs (including APS-U and LCLS-II) have not reached the theoretical performance limit. The next-generation storage rings will aim at even smaller emittances (to reach the so-called diffraction limit at hard X-ray energies). The XFELs have reached diffraction limit in most of the X-ray energies, but lack the temporal coherence (hence not Fourier-transform limited) especially in the hard X-ray regime. Laser-seeding and self-seeding schemes have been developed, but challenges remain in extending the wavelength reach or satisfying the stability requirement (see below). X-ray regenerative amplifier FEL and XFEL oscillators use crystal cavities to feedback the monochromized X-ray pulses and to interact repetitively with high-repetition rate bunches from the accelerator. They promise to deliver near-Fourier-transform limited hard X-rays with much better brightness and stability.
Now we switch to the metrices of light source operations. There are three of them, from the reliability, to the stability of beams and, finally, to the intensity of light.
Reliability is critically important for securing productivity of the facility. As modern light sources serve 1,000 to 4,000 users per year, uninterrupted operations enable successful experimental programs to flow within the allocated time. Both Mean Time To Repair (MTTR) and Mean Time Between Failures (MTBF) are important, as frequent small interruptions hinder quality of the measured data, while infrequent, but long, ones lead to canceling scheduled experiments. As the user time at a beamline may range from a week to 4 hours, the oversubscription rates are high (x2.4 for NSLS-II and x4.5 for LCLS on average), and user teams may await their scheduled beam time for months. If the experiment is flawed by machine issues, the quality of the data becomes inadequate, which adversely affects the user program and tarnishes the facility renommée on a long run.
Beam stability in both position and intensity is vital in assuring quality of the measured data at synchrotron light sources. The advent of low-emittance lattices offers beam sizes of ~5 µm in both planes, which constrains the magnitude of permissible jitter to the hundredth of a nanometer level. Poor beam stability in terms of orbit vibration and drifts devalues the efforts invested in low-emittance machine design, and limits energy and spatial resolution of high-performance experiments at the facility. Satisfying the stability requirements calls for strict attention to the related aspects of the machine design. The components of success include the choice of low-noise site for facility construction, the balanced design of the machine girders and magnet supports, tight tolerances on power supply and RF stability, as well as low-vibration beam transport from the source point to the detector.
Another component of the stability program is the orbit feedback. Broadband powerful feedbacks stabilize the electron beam orbit in the source points and, when integrated with the beamline X-BPMs and mirror or grating positioning system, hold X-ray beams to their nominal positions preserving high resolution of imaging or spectroscopy.
In addition to transverse stabilities, XFELs have the additional challenges of energy and temporal stabilities. XFELs based on self-amplified spontaneous emission have spiky spectra due to noise startup and large intensity fluctuations after a monochromator. Seeding and cavity-based XFELs will tremendously benefit the spectral stability. As XFEL pulses are very short (tens of femtoseconds down to attosecond), the time jitters can have significant impact on time-resolved experiments. Hence, longitudinal feedback and controls are critical for XFELs. X-ray pump and X-ray probe methods with intrinsic synchronization must be developed to overcome jitters at these extreme timescales.
Radiation intensity is another metric of the light source operations. High intensity benefits all experimental programs in the facility and especially the programs based on the coherent techniques gain in performance faster than the linear scaling of the intensity. On the other hand, high beam intensity impacts on the electron beam quality due to the collective effects, which, in turn diminishes benefits of high beam current. Yet another limitation comes from the resilience of user samples to the intense beams of undulator radiation that damages the sample matter or tissue at a high rate.
In addition to the three metrices described above, operating cost is one of the factors influencing the choice of technologies and constraining performance of the facility. Therefore, the size of the electric bill comes into consideration as the limiting factor in relation to the size of the accelerator systems and the magnitude of the electrical systems.
These two performance parameters and three metrices motivate the R&D directions in the world of light sources. These directions have similarities and differences between the rings and FELs. As mentioned above, the recent BRN outlined specific R&D priorities related to new capabilities of the light sources, new technologies in accelerator engineering and new approach in modeling and data analysis.
In terms of the new capabilities, the high brightness and coherent flux are important for the future high-performance experiments. Increasing energy, temporal and spatial resolution requires high brightness and substantial photon flux on a microscopic scale. The purity of radiation wavefronts becomes the specific objective for chemical, biological and material science experiments, as well as for manufacturing and metrology.
Enabling high brightness requires development of low-emittance lattices for the storage rings and low-emittance injectors for FELs. Establishing long undulators with a short period and low field errors will maximize the brightness, especially at the hard X-ray range, and boost performance of premier beamlines at the light sources.
The new cutting-edge designs of accelerators call for new simulation tools and deep understanding of the underlying physics in regimes never approached before. Powerful simulation packages exist in the areas of beam dynamics, wavefront propagation, FEL- and space-charge-dominated regimes and as the community progresses towards pm-range emittances and 10s of kA peak currents, these simulations will need to be expanded and benchmarked with respect to the experimental data (BRN direction “Understand scientific mechanisms that limit system performance and utilization”).
Separately, as AI/ML is firmly entering in our daily life, similar algorithms are being routinely used in machine tuning, fast commissioning, detection of trends and characterization of faults in operations. The ongoing revolution in the digital domain motivates expansion of applications of AI/ML in accelerator design, optimization and operations, which is a priority direction as per BRN (“Accelerate advanced modeling, real-time feedback, fully-integrated co-design, and physical–digital fusion”).
Among key factors limiting productivity of light sources is the power consumption. With the cost of electric power trending up while the push for green technologies strengthening, the 1-10MW facilities will have to find more efficient ways to operate. Using alternative sources of energy, e.g. via solar panels, has emerged in the past decade as a partial solution. Another, more powerful, solution is to switch from electro- to permanent magnets, which offers large gains in reducing hundreds of power supplies, the cooling systems and electrical distribution.
Furthermore, in the age of rapid evolution of electronics, obsolescence of accelerator components has become a significant issue. Facing challenges of computers, BPMs, power supply controllers, instrumentation electronics “aging” and losing support from chip manufacturing motivates a concerted effort from light source laboratories to transition to common standards of accelerator technology and establishing common platforms for diagnostics, RF, vacuum and machine controls.
Aside from large scale facilities, usually hosted by national labs or premier universities, a new generation of light sources with the focus on universities and industrial applications has come up. Development of compact and versatile accelerator-based sources is important for expanding the synchrotron science beyond major centers and for inviting communities of professors and students in smaller university campuses. On the other hand, establishing reliably operating facilities tailored to production of high-tech devices or commodities with high throughput is a priority research direction. Examples of such industrial applications include producing or diagnosing microchips at highest possible density of transistors via EUV or Xray FEL machines (BRN direction “Lead innovation in new materials, system design, and advanced fabrication as a foundation for integration of technologies in accelerator-based facilities”).
Finally, workforce development and education of new generations of accelerator physicists and engineers are thought to be a critical direction for the light-source community at large. With rapidly increased level of sophistication of light source accelerators, the future staff will need to exhibit detailed knowledge of classical mechanics and electrodynamics complemented by deep insight into engineering disciplines such as mechanics, electrical circuits and RF systems, heat transfer, coding and digital methods. BRN explicitly mentioned this need as a cross-cutting direction for synchrotron and FEL communities.
Concluding, we congratulate teams from APS-U and LCLS-II with completion of these ambitious projects and successful machine commissioning.
[BRN] https://science.osti.gov/-/media/bes/pdf/brochures/2024/24-G00737-BRN-ABI-brochure-Final.pdf
[KJK] https://www.osti.gov/servlets/purl/6806336