Report of the
First International Review Committee
for the
SPring-8 Compact SASE Source Project

February 3-4, 2005

Committee Members: Marie-Emmanuelle Couprie, John Galayda, Jerry Hastings, Kwang-Je Kim (Chair), Shin-ichi Kurokawa, Won Namkung, Jochen Schneider

EXECUTIVE SUMMARY

The SPring-8 Compact SASE Source (SCSS) is an innovative project for generation and use of intense, coherent, short-pulse X-ray beams. Although its goals are similar to other X-ray free-electron laser (FEL) projects in the USA and Europe, the SCSS is unique in its compactness of design and in its co-location with SPring-8, the world's leading third-generation X-ray synchrotron radiation facility. The project is ambitious in schedule and technologically challenging. However, members of the project team have track records for innovative solutions to difficult problems. The accumulated knowledge in accelerator technology and X-ray science at the SPring-8 site is also an important asset. The Committee therefore agrees that the SCSS project can be completed as scheduled and recommends the prompt start of the project.

The scientific case for the SCSS is in the formative stage, due largely to the revolutionary nature of the source. However, the SPring-8 scientists have extensive experience in making use of coherent X-ray beams, leading to recent success in coherent X-ray imaging, which will be one of the key application areas of the SCSS. The project team recognizes the importance of boosting discussions among scientists through workshops, and has already held successful workshops in chemistry, medicine, biology, and atomic physics. The Committee endorses the plan to hold many more workshops to further stimulate the engagement of leading scientists.

The design of the accelerator-undulator-beamline system is mature. The SCSS injector, being based on a DC gun, is unique among X-ray FEL injectors, aiming for high stability and reliable operation of high-brightness electron beams. The injector will be operational for the planned 250-MeV test experiment to be carried out this year. R&D items for the injector include beam collimating to suppress unwanted tails, and confirmation of low jitter. The high-gradient C-band structure is one of the critical elements in achieving the compactness of the SCSS facility. The C-band system has been in development since the mid-1990s as a part of the KEK linear collider effort. Prototypes of necessary technological ingredients for the SCSS have been demonstrated. The in-vacuum design of the undulator is essential for reaching the 1-� goal with 6- to 8-GeV electron beams. The in-vacuum undulators have been operating successfully at SPring-8. R&D items for the undulator system include extension of the previous study of radiation damage at 2 GeV to the higher energies and more stringent requirements of the SCSS project, minimizing the dark current in the undulator area, and estimating the wakefield effect in the small gap device, including possible frequency-dependent resistivity. The design of the X-ray beamline is in the early stage, awaiting further definition of the experimental program. Important progress has been made in X-ray optics, particularly in the understanding of wavefront distortions from optical elements. A detailed strategy for alignment has been worked out, a critical element of which is the laser alignment of RF beam position monitors (BPMs). A major step towards satisfying the stringent tolerances required for the SCSS project was achieved with the demonstration of 4-micron resolution with the test laser alignment system. The RF BPM was originally developed for the KEK linear collider project and has achieved 20-nm resolution and stability.

The SCSS project benefits from its co-location with SPring-8, especially in the areas of manpower sharing, development of coherent X-ray optics techniques, the accelerator control system, and novel use of the SCSS linac electron beam at SPring-8. The Committee recommends that the SCSS include provisions for 8-GeV operation to allow a possible SCSS operating mode involving injection of the SCSS linac beam to the SPring-8 ring. International collaboration with laboratories pursuing similar projects will be beneficial for all parties. Topics for collaboration include high-brightness gun development, beam diagnostics and alignment strategy, stability issues, shot-to-shot X-ray beam characterization, synchronization issues, and detectors.

To play a leading role in the worldwide X-ray FEL efforts, it is important that the SCSS project start in 2006. Timely construction and commissioning of the 250-MeV facility is critical for the success of the project. Sufficient resources for developing the scientific instrumentation should be allocated from the start of the project. Success of the SCSS project will be a milestone event in the advance of technology for X-ray FELs, stimulating progress in X-ray science worldwide.

SCIENTIFIC CASES

　The science pursued at third-generation synchrotron radiation facilities is exciting and includes applications in many different fields of research. In most cases equilibrium states of matter are studied with atomic resolution in space. There is a clear trend to investigations of matter on nanometer-length scales and to spectromicroscopy. Focusing of hard X-ray beams down to about 50 nm has been achieved at various laboratories. The storage rings and the beamline technology are mature and offer very reliable research opportunities. However, the technology is getting close to its theoretical performance limit, and the facilities will put more emphasis on sample environment and detectors.

X-ray sources making use of energy recovery linacs (ERLs) will further improve the beam quality and mainly widen the potential of X-rays for studying equilibrium states of matter. A prototype ERL with superconducting RF is under development at Cornell University, USA, for proof of principle. The project just received funding from the National Science Foundation. A proposal for a hard X-ray user facility should be available in three years.

Free-electron lasers based on the principle of self-amplified spontaneous emission (SASE) offer hard X-ray beams with peak brilliance about 10 orders of magnitude higher than the best storage ring facilities available today and about 7 orders of magnitude higher than expected for ERL facilities. Proof of principle has been achieved at DESY in Hamburg. Using a linac in TESLA technology for 100- and 32-nm wavelength radiation, the shortest pulse duration was about 20 femtoseconds. In the first step of FEL development for hard X-rays the pulse duration will be of the order of 10 or 100 femtoseconds. When applying proposed seeding schemes combined with harmonic generation with external lasers, one should be able to achieve pulse durations of the order of 1 femtosecond and below with improved longitudinal coherence.

Whereas ERLs are a very interesting evolutionary extension of the opportunities offered by modern storage ring facilities, the potential of X-ray free-electron lasers is revolutionary in character. For the first time it will become possible to study nonequilibrium and new states of matter with atomic resolution in space and time. The scientific case for hard X-ray FELs has been discussed in some detail in context of the LCLS project at Stanford, USA, and the European XFEL Facility in Hamburg, Germany, as well as for the project pursued at Bates Laboratory at MIT, Boston, USA; the BESSY FEL at Berlin, Germany; and the FERMI project at Trieste, Italy. Essentially seven research fields have been identified:

• atomic, molecular and cluster phenomena,
• plasma physics,
• condensed matter physics,
• materials science,
• ultrafast chemistry,
• life-sciences, structural biology, and
• quantum optics and non-linear processes.　

• understanding of wave-front distortions from optical elements,
• X-ray mirrors with unprecedented roughness and figure error to minimize wavefront distortions in collaboration with Osaka University, and
• nearly perfect synthetic diamond monochromators jointly with industry.

The SPring-8 scientists have extensive experience in making use of coherent X-ray beams, leading to recent success in coherent X-ray imaging, which will be one of the key application areas of the SCSS. The present beamline design is in the early stage and awaits further definition of the experimental program. Promising concepts for ultrafast timing experiments, single-event measurements, and X-ray intensity interferometry have been presented at the review.

TECHNICAL COMPONENTS

Injector

The injector consists of

• a 500-kV "pulsed DC" gun with a 3-mm-diameter single-crystal cerium hexaboride cathode, producing a 1-ampere pulse of duration 1.6 μs,
• a pulsed chopper, reducing the pulse duration to 1-2 ns,
• a 238-MHz, 210-kV pre-buncher cavity inducing a 250-keV energy chirp centered on ~ 400 keV,
• a 476-MHz booster cavity accelerating the beam to about 1.1 MeV and compressing to 23 amps,
• an S-band pre-buncher and S-band linac, accelerating the beam to 50 MeV, and
• a chicane energy filter providing a 10-ps, 40-amp output beam.

The DC gun has been constructed and tested. Measurements of emittance have been carried out, yielding an estimate of the thermal emittance: 0.6 mm-mrad. Space-charge effects seem to have increased the projected emittance of the pulse to 1.1 mm-mrad, and stray capacitance in the gun load resistors caused the pulse duration of the gun voltage to be longer than required for SCSS. There is some emittance increase near the head and tail of the gun pulse, possibly due to space-charge effects. The data demonstrate that, with additional collimation, an emittance less than 1 mm-mrad should be achievable at the desired current.

The choice of a thermionic cathode with very low current in the early stages of acceleration offers some significant advantages. There should be no significant fast fluctuations of beam current within the bunch and no significant variations in current density across the beam profile.

Discussion of the start-to-end simulation did not cover assumptions about wakefields from the chopper aperture and RF cavities for RF bunching and acceleration. Incorporation of short-range transverse wakefields is important to determine alignment tolerances in the initial stages.

Linac

The SCSS adopts the C-band RF frequency for the main linac, which has been extensively developed for the linear collider at KEK. The accelerating gradient of higher than 30 MV/m is enough for a compact and economical accelerator of 6-8 GeV in the available space at the SPring-8 site. The proposed beam energy is also adequate for the 0.1-nm XFEL radiation combined with the in-vacuum undulators. The proposed C-band subsystems are state-of-the-art components. The integrated test will be demonstrated through the prototype system of 250 MeV for 60-nm radiation by November of this year. The major subsystems in the main linac are the accelerating section, klystron, modulator, and pulse compressor. Some special characteristics are noted as follows.

• Accelerating section: The choke-mode accelerating section has a unique feature of suppressing the wakefield effects by intense electron beams. The input coupler has also a feature of reducing breakdowns and providing symmetric RF fields in the structure. In view of the multibunch operation option, the Committee strongly recommends keeping this excellent choke-mode scheme for the SCSS project.
• Klystron: The reliability has been demonstrated for more than 10,000 hours at 50 MW, 2.5 ms, and 45% efficiency. The committee is confident and recommends the klystrons for the project.
• Pulse compressor and PM-AM modulation: This innovative concept is capable of providing flat RF pulses to accelerating sections with four times higher RF power, i.e., a beam energy gain of about two.
• Klystron modulator: The smart modulator adopts an inverter power supply and a compact oil tank that is available from the commercial utility market. The reduction of the modulator volume and costs is so dramatic that the SCSS will benefit greatly from this new development.
• Support and alignment system: The SCSS developed the Roller Cam Mover system with repeatability of about 0.1 mm; it is necessary for precise alignment of the whole system. The ceramic support stand is also developed for integration of the whole system.
• Cavity BPM: For beam-based position alignment, the SCSS group developed and tested the improved RF BPM with 20-nm resolution

The SCSS project team has developed or demonstrated most of the critical subsystems required for a linear accelerator of greater than 35 MV/m. The Committee is very impressed with the current project status, and the Committee strongly recommends the C-band accelerator to the SCSS main linac.

Undulator

The in-vacuum undulator adopted in the SCSS project is unique in the world. Due to its shorter period and higher magnetic field, the in-vacuum undulator allows the generation of 1- SASE with electron energy of 6-8 GeV, twice or three times smaller than in other X-ray FEL projects.

The SPring-8 insertion devices team was a pioneer in the concept and development and is a leader in the field of in-vacuum undulators. Besides the first devices installed in KEK and SPring-8, with the remarkable 25-m-long in-vacuum undulator on the 8-GeV synchrotron radiation facility, the group has shared its expertise with other facilities around the world such as the ESRF and the Swiss Light Source. The productive collaboration with industry (Sumitomo group) established many years ago makes the technology reliable. It supposes special magnet arrangement techniques (shimming not being allowed) while maintaining a high field quality, particular treatment to the magnets, and specific mechanical design.

From this choice arise new specific difficulties.

First, the electrons from the beam halo might directly damage the magnets, so particular care should be devoted to control the electron beam purity, to reduce and monitor the dark current in the undulator area, and to implement fast electron beam stops. In addition, the efforts to understand the magnet resistance to radiation, which have been already extensively studied in collaboration with PAL, should continue further. In particular, the studies should be extended to higher energy to meet the stringent requirements of the SCSS project.

Second, the effect of the wakefield in the small-gap device should be carefully studied, including possible frequency-dependent resistivity. Possible modifications to the beam dynamics should be driven; the 250-MeV SCSS prototype will provide some interesting results.

Third, the alignment of the electron beam in the undulator is more critical than on other 1- SASE facilities. The SCSS has proposed a double alignment strategy ased on laser alignment of the different components and cross checked with undulator radiation analysis of each individual undulator segment. See the section "Diagnostics and Commissioning" below for further comments on alignment.

The success of the in-vacuum undulator approach for short wavelength free-electron lasers can contribute significantly to the FEL community and have a great impact on the future design of compact FEL sources around the world. It is therefore particularly relevant that this strategy has been adopted by the leaders in this field.

X-ray Optics

XFEL radiation has many unique aspects that will enable quality new science when compared to undulator radiation from current third-generation storage ring sources. In particular, its peak brightness will enable time-resolved studies of unprecedented temporal resolution at the femtosecond level. This extraordinary peak brightness derives from the very large flux, the remarkably short pulse durations, and the fully coherent transverse phase space of the photon beams. The challenge for X-ray optics is to maintain and or manipulate the transverse phase space of this beam and develop X-ray optics that are the analogs of beam splitters, delay stages and polarizers, etc. that are used routinely at optical wavelengths.

The SCSS team is a world leader in these developments. The team of Dr. Ishikawa has developed the X-ray phase plates now in common use at third-generation light sources to manipulate the photon polarization. His team has demonstrated a variety of unique applications of X-ray optics in understanding the coherence properties of synchrotron radiation, and the extension of these ideas to the SCSS beam follows in a natural way. The SCSS team has the capability to provide not only optics to split/combine/delay the X-ray beams, but it is also developing, in collaboration with Osaka University, mirror optics that are world's best in both figure and roughness that will be used to focus the SCSS beam and provide unprecedented peak powers in the X-ray regime. In the area of X-ray optics, the SCSS team is the match for any in the world.

Diagnostics and Commissioning

Presentations on diagnostics and commissioning placed emphasis on RF beam position monitors and X-ray intensity measurements. The ability to open undulator gaps will be exploited for diagnostic purposes. A partial list of controllable parameters in the undulator includes:

• K of the undulator,
• pathlength (phase) adjustment between undulators,
• correctors for the Earth's magnetic field,
• steering correctors, and
• RF BPM position.

Adjustments to achieve best performance will be based on measurements of the position of the electron beam using RF BPMs, observation of position and profile of the electron beam using optical transition radiation monitors, and on-axis (and near-axis) intensity of monochromatized X-rays from the undulators. A system for direct measurement of undulator K was mentioned but not described in detail.

A method for alignment of the RF BPMs with a laser was presented. Iris targets are inserted into alignment features on the RF BPMs. The targets are illuminated with a collimated helium-neon laser, and the position of the light passing through the iris is measured with an imaging detector. The stated goal was to align the BPMs to a few microns. Results from a test of a single BPM and iris were presented, showing resolution and repeatability of a few microns in the measurement.

K and inter-undulator phase will be measured and corrected by observation of a higher harmonic of the spontaneous undulator radiation (3rd or 5th). Observation of the X-ray intensity profile will be used to co-align the X-ray beams to about 0.8 micro-radian.

C-band resonant RF beam position monitors will be used for measurement of electron beam position with sensitivity of the order of a micron.

Diagnostic measurements will be made on single undulators and pairs of undulators to achieve the necessary tolerances. Laser alignment of the BPMs is likely to be necessary on a daily basis.

The sensitivity of RF BPMs is adequate for the purpose of alignment. The sensitivity of the laser alignment technique has been measured. Continued testing should be carried out with more than one iris target in use. It is worthwhile to devise a test to confirm that the RF BPM position is not affected by insertion of the iris target. It is also worthwhile to verify that the laser spot profile is sufficiently stable over time, since a change in center-of-intensity of the laser spot could affect the alignment process. The effect of damaged or tapered undulators should be considered in interpreting the meaning of X-ray intensity measurements. The measurement may be affected by grazing-incidence reflection of X-rays if there are surfaces sufficiently close to the beam path.

The sensitivity of the lasing process to peak field in the undulator is used to set a tolerance of 6X10^-4on K. This tolerance translates into 1- to 2-micron-level adjustability and stability for gap control. This is practical but challenging. It will also be necessary to eliminate or compensate energy jitter in the electron beam to a precision comparable to the relative K tolerance.

Synergy with SPring-8

The unique aspects of FEL radiation provide the opportunity for complementary studies in the broad areas of materials, chemical, and biological sciences. As with the Linac Coherent Light Source (LCLS) and European XFEL (EXFEL), the SCSS is co-located with a world-class synchrotron radiation source. This co-location has many advantages spanning both the accelerator and X-ray science and instrumentation.

For the SCSS, as an example, the SPring-8 control system and the personnel responsible for its design, upgrade, and maintenance will provide the controls for the SCSS. The proposed design of the SCSS affords a unique option compared with the LCLS and the EXFEL. The location of the accelerator will be chosen so that pulses from the SCSS linac can be injected into the SPring-8 storage ring. It is here, for example, that a common control system is not only cost saving, but also provides the natural coordination for the SCSS accelerator as an injector for special operations of the SPring-8 storage ring.

In the area of X-ray science and instrumentation, the SPring-8 program will be critical in the early development phases of the SCSS program. Again, there is a natural integration from the leadership of Dr. T. Ishikawa, who has significant responsibilities for both the SPring-8 science and instrumentation and is the project leader for the SCSS proposal.

The undulator magnet is another obvious synergy. Here Dr. Kitamura is responsible for insertion device development for SPring-8. His world leading development in small-gap in-vacuum magnets is critical to the production of very high-energy undulator radiation at SPring-8, and it forms the basis of the design for the SCSS undulator. Dr. Kitamura and his team are part of the SCSS project.

Of course there are the obvious synergies in the areas of accelerator operations and maintenance beyond controls, in vacuum, power supplies, RF and so on. These developments will further engage the accelerator physics staff of both the new project and SPring-8 for years to come.

Finally, as the SCSS comes into operation, all of the infrastructure of the SPring-8 campus is available to the SCSS; the housing, food services, open space in close proximity for development of labs, offices, and so on.

In summary, there are significant advantages to locating the SCSS at the SPring-8 site that cannot be found elsewhere in Japan.

SCHEDULE AND MILESTONES

With the SPring-8 hard X-ray and the New Subaru soft X-ray storage ring facilities operating, together with the planned FEL for hard X-rays (SCSS) and the soft X-ray prototype facility for wavelengths down to 60 nm under construction, a Synchrotron Radiation Science Center of enormous potential for scientific and technological breakthroughs can be realized at the SPring-8 site. The Center will be highly attractive for scientists from many different disciplines including the optical laser community. This community is familiar with femtosecond pulses, coherence, and extreme intensities in the optical wavelength range and will transfer these concepts to studies of new states of matter with atomic resolution in space. The whole field of laser physics is progressing very fast, and it is important that the free-electron lasers are realized in due time. At present there are four projects for hard X-rays in the world that are either under construction, approved, or close to approval.

• The Linear Coherent Light Source (LCLS) at SLAC in Stanford, USA, is under construction and is expected to provide first X-ray beam at the end of 2008.
• The Pohang Free Electron Laser Facility, making use of the existing linac, is approved. Construction is expected to start in 2006, and first X-ray beam in the wavelength range of 0.3 to 0.7 nm is anticipated in 2009.
• The European XFEL Facility is expected to start construction in the fall of 2006 and provide first X-ray beams in 2012.
• The SCSS at SPring-8: The Committee strongly recommends the start of this project in 2006.

The start of SCSS construction in 2006 is important in order to assure that Japan will play a leading role in the worldwide FEL efforts. Timely construction and commissioning of the 250-MeV facility is critical for the success of the SCSS.

The SCSS as well as the lower-energy prototype facility will be realized in close collaboration with Japanese industry. Without knowing the commitments and the potential of the industries involved in the realization of the project, it is difficult to judge if the proposed three-year construction time is realistic. The project schedule of the SCSS is ambitious but we feel it is feasible in view of the competence of the project team and the accumulated knowledge on the SPring-8 site. The success of the SCSS relies strongly on the performance record of the three principle investigators: Tetsuya Ishikawa, Hideo Kitamura, and Tsumoru Shintake. They are well known for their successes and innovative solutions to difficult problems. The review committee was also impressed by the other members of the design team, including a number of young scientists.

TOPICS FOR INTERNATIONAL COLLABORATION

The committee recognizes that the SCSS project is one of the four major X-FEL projects in the world; the other three being the LCLS in the USA, the PAL X-FEL in Korea, and the European X-FEL project in Germany.

Since the full-scale SCSS project has yet to be approved and funded by the Japanese government, the project can learn much from forerunners, LCLS and X-FEL, and at the same time all three X-FEL projects can reap benefits from each other by establishing a well-organized collaboration scheme. The Committee recommends that the SCSS project try to establish a well-organized collaboration scheme with the LCLS and the TESLA FEL, and also recommends having close contact with the PAL project.

Possible areas of the collaboration are listed below:

• High-brightness gun development
• Beam diagnostics and alignment strategy
• Stability issues
• Shot-to-shot X-ray beam characterization
• Synchronization issues
• Detectors

The Committee also recognizes the importance of the project's close collaboration with KEK, High Energy Accelerator Research Organization, in Japan, especially on the electron linac. The electron linac of the SCSS project is based on development of the C-band linac technology that has been developed at KEK for linear colliders. The Committee suggests the first step of this collaboration be the signing of a specific MOU under the umbrella of the general RIKEN-KEK accelerator collaboration.

Top