ICFA2000t.GIF (31362 bytes)  

The 19th ICFA Advanced Beam Dynamics Workshop on Future Light Sources

Physics of and Science with

The X-ray Free-Electron Laser

(Arcidosso, Italy, September 10-15, 2000)

Workshop Summary

The19th Advanced ICFA Beam Dynamics Workshop on "The Physics of, and the Science with, X-Ray Free-Electron Lasers" took place in Arcidosso (Italy) from the 10th to the 15th of September, 2000. The Workshop was sponsored by the International Committee for Future Accelerators, the US Department of Energy, the University of California at Los Angeles, the Stanford Linear Accelerator Center, the Deutsches Elektronen-Synchrotron and the Lawrence Berkeley National Laboratory, together with local authorities of the Tuscany, Grosseto and Arcidosso areas. The Workshop's chairmen were M. Cornacchia (SLAC), I. Lindau (SLAC/Lund. Un.) and C. Pellegrini (UCLA). Seventy-five scientists, of which 50 are involved in the physics and technology of accelerators, free-electron lasers and x-ray optics, and 25 in the scientific applications, attended the workshop. There were plenary and parallel sessions and many lively discussions, during and after the regular workshop schedule.

Arcidosso is a medieval town in southern Tuscany, close to the city of Sienna. The meeting took place in the historically evocative scenario of an 11th century castle atop a hill dominating the nearby valley. The castle was restored in 1989, and preserves the atmosphere and raggedness of medieval times.

There were two invited lectures on Monday, September 11, to open the subjects and two summary talks in the afternoon of Friday, September 15. All the other presentations were either informal or in the form of

The Group on "Physics and Technology of the XFEL" with introductory talks by Kwang-Je Kim (ANL) and Jamie Rosenzweig (UCLA), was coordinated by Alberto Renieri (ENEA-Frascati).

The Group on "Science with the XFEL" was coordinated by Mark Sutton (McGill University) with introductory talks by Andreas Freund (ESRF) and Ingolf Lindau (SLAC/Lund. Un.).

These notes reflect the summary talks of the coordinators and the impressions and recollections of the organizers. The American Institute of Physics will publish the proceedings of the Workshop.

Summary of discussions and conclusions of Group 1: Physics and Technology of the XFEL

The main issues that were discussed by the 50 participants in this group were the photo-injector, the production of ultra-short pulses, the effects of wake-fields induced by the electron bunch, the operation at lower charge and emittance, the possibility of harmonic generation and the diagnostics in the undulator. The following is a short summary of the discussions and their conclusions.

It is important to measure the {\it electron bunch emittance}, length and energy spread as a function of charge and not focus exclusively on the standard photo-injector parameters (1 nC charge, 1 pi mm-mrad emittance). The low charge (about 0.2 nC charge, 0.6 pi mm-mrad emittance) option appears as feasible as the standard case used in the LCLS design, and offers the clear advantage of being less vulnerable to the effects of wake-fields. It has not been studied as much as the standard case and requires more work.

One should follow the progress with the new guns currently under study, like the pulsed gun being developed at BNL and Eindhoven, and the Van der Wiel plasma gun.

The studies of the feasibility of {\it electron bunch compression} and/or {\it x-ray pulse slicing and compression} must be continued, given the importance of this option for the experimental program. In particular one should study
the possibility of bunch compression when operating at low charge, and the effect of wake-fields in the two-undulator (seeding) scheme.

Much attention was given to the {\it wake fields} in the undulator vacuum pipe. Comparative estimates were made using different models proposed by A. Agavonov (Levedev Physical Institute, Moscow), A. Novokatsky (Darmstadt
Un.), L.  Palumbo (Rome University) and G. Stupakov (SLAC). The effects have been calculated for the following situation: 15 GeV, undulator length of 100 m, a pipe radius of 2.5 mm, 1 nC charge, 230 fs long bunch. The maximum energy changes along the undulator length, according to different models
and regimes are:

Agafonov model: 210-6 (roughness height: 100 nm, roughness period: 100 micro-m)
Novokatsky: 210-3 (roughness height: 100 nm, roughness period: 0.1 micro-m)
Palumbo: 310-6 (roughness height: 500 nm, roughness period: <10 micro-m)
Stupakov: 10-6 (roughness height: 500 nm, roughness period: 100 micro-m)

In addition, the contribution of the resistive wall effect is about 1.510-4.

Additional contribution will come from vacuum ports, instrumentation, discontinuities.

Since this energy change is of the order of the FEL parameter, it can have a serious and deleterious effect on the LCLS performance. The message from the workshop is that one should be aware of these effects, in particular for the LCLS small gap undulator. Notice that the minimum undulator gap considered for the TESLA X-ray FEL is 12 mm, compared to the present 6 mm of the LCLS.

Possible strategies to reduce the undulator wake-fields effects include reducing the bunch charge, increasing the undulator gap and reducing the undulator length.

The list of recommendations from the workshop on the {\it surface roughness} problem include the enhancement of the analytical models to predict realistic surface roughness conditions and of the numerical simulations to
model realistic randomly distributed surfaces roughness. Experiments should be performed to measure the effect under controlled conditions of surface roughness.

The possibility of operation at a charge different and lower than 1 nC should be studied in all its implications. Different modes of controlling the bunch charge and emittance should also be investigated.

Once the wake-fields and the injector operation at different charges are understood, the system should be re-optimized, including considerations of various types of undulators, planar or helical, and with a gap chosen to
minimize the wake-fields to an acceptable level.

The sub-group on undulator diagnostics reviewed the issues related to the electron and photon beams. The centroid of the electron beam can be measured to mm resolution with rf Beam Position Monitors (BPMs) or Optical
Transition Radiators (OTR). One of the issues is whether the latter can survive the intense electron beam and how the surface quality of the OTR might affect the emitted light. Both questions should be soon be answered
by experiments. On the measurements of the beam profile, there was consensus that saturation makes the scintillators not usable, while OTRs might be useful. It is also important to measure the longitudinal characteristics (bunch length and momentum spread) and the time-resolved slice measurements of emittance and momentum spread. A very promising
technique for measuring very short bunch lengths uses an rf deflector that rotates the beam onto a screen. It was suggested that it might be possible to measure photon pulses down to 10 fs using grating Michelson

One of the outstanding questions concerning the measurements of the x-ray beam is whether one can separate the spontaneous from the FEL radiation and whether the diagnostics can survive the x-ray and electron fluxes. It
was recommended to estimate the damage mechanism with ionization as the dominant mechanism.

Crystalline materials directly impacted by the electron beam may see the space charge field and be subject to damage. It was suggested that an experiment be done at SLAC using the FFTB beam to create a high field
gradient on a crystal similar to that that would occur in the LCLS.

It is very important to have a diagnostic system capable of measuring low charge beams in the linac and undulator.

More detailed studies of the survivability of the detectors and the information they provide are needed.

Some other noteworthy discussions included the following:

1. The wake fields in the undulator could have a strong effect on the harmonics; we need more experimental and simulation work to investigate this possibility.

2. The same wake-fields could limit the possibility of reducing theline width or the pulse length.

3. The X-ray FEL must be optimized including collective effects.

4. A proof of principle of a seeded scheme using High Gain HarmonicGeneration has been done at Brookhaven; the studies of an X-ray FEL usingthis approach should be continued.

Summary of discussions and conclusions of Group 2: Science with the XFEL

About 25 people attended sessions to discuss the possible scientific applications of a x-ray FEL. Because of the recent focus on the first experiments with the proposed Linac Coherent Light Source at Stanford, the discussions were mainly focussed on these proposals. The extension of the characteristics beyond the initial stage and the further developments of
the source were also part of the program.

Six scientific areas were discussed: Atomic Physics, Warm Dense Matter, Femtosecond Chemistry, Imaging/Holography, Bio-molecular Structures and X-Ray Fluctuations Spectroscopy.

New phenomena can be studied in atomic physics. Hollow atoms, where inner core electrons have been removed with outer valence electrons still in place,  appear especially interesting. Non-linear x-ray interactions are of interest, i.e. parametric down-conversion, two-photon absorption and two-photon mixing. Even with an unfocussed LCLS-type beam it is possible to achieve saturation for photo-ionization. With a focussed beam the Compton scattering will saturate.

Warm dense matter (WDM) is a new form of matter, between highly ionized plasma and condensed matter. Though WDM is of great importance in many fields, i.e. laser plasma production, inertial fusion and astrophysics, its basic properties are still basically unknown. With a x-ray FEL beam, WDM can both be created and probed.

In femtosecond chemistry it is of great interest to study bond changes on the time scale characteristic for breaking and forming bonds. This would involve pump-probe experiments where the system is excited with a conventional laser and the structure changes are probed dynamically with the x-ray FEL beam.

The workshop addressed the possibility of imaging and holography of non-crystalline samples and small nano-structures. Bio-fragments and bio-molecules are also an extension of this work. The radiation damage and
the amount of structural information that can be extracted before the molecules fly apart are key issues. For small structures, great advances have been made in computer modeling. It would be desirable to extend the models to bulk samples.

X-ray intensity fluctuation spectroscopy is already being pioneered at third generation light sources and its extension to x-ray FELs, in terms of the time-scales and length-scales, were discussed, together with the possibility of studying a broad range of materials.

There were intense discussions trying to define the most important radiation characteristics. This will of course in many cases depend on the specific experiments, but in general terms the following order was established, in de-creasing order of importance:

1. Beam position stability
2. Beam focusing
3. Synchronization for pump-probe
4. Shorter pulses
5. Smoother pulses
6. Reduced pulse to pulse intensity fluctuations

M. Cornacchia, I. Lindau, C. Pellegrini