Office of Naval Research, Europe
Trip Report #3-96
ICFA Beam Dynamics Panel Workshop: 4th Generation Light Sources
Travel Period: 10-26 January 1996
Keywords: FEL; Light Sources; MFEL Program
Contents: 1. Summary 2. Narrative 3. Feedback
1. Summary
The main objective of this international workshop was to prepare the scientific case and
requirements for 4th generation light sources [average spectral brilliance ~1018-20
Photons sec-1mm-2mrad-2(0.1% Bandwidth)-1], and to discuss approaches for developing them.
The author presents his opinions on the Duke/OX-4 XUV FEL (supported by the MFEL Program)
relative to 4th generation requirements, and describes improvements to the Duke system
from discussions at the L.U.R.E. laboratory in Orsay, France.
2. Narrative
The main reason for this trip was to participate in the ICFA Beam Dynamics WORKSHOP ON 4th
GENERATION LIGHT SOURCES. The organizing committee was very selective, limiting the number
of participants to 90. Finally, the workshop had 95 participants from 42 institutes in 12
countries.
The main goal of the workshop was to extend the resounding success of the 3rd
generation light sources to the 4th generation by preparing the scientific case and
requirements for the new generation. The workshop had few plenary sessions; most of the
work was done in seven Working Groups:
Group 1: Scientific opportunities for 4th Generation light sources - VUV/soft X-rays;
Group 2: Scientific opportunities for 4th Generation light sources - hard X-rays
Group 3: Lattice and stability aspects (storage rings)
Group 4: Current, lifetime and time structure (storage rings)
Group 5: Linac sources
Group 6: Storage ring FELs
Group 7: Insertion devices
I participated in the work of Group 6 (70%), Group 2 (10%), Group 4 (10%), and the joint session of Groups 5 & 6 (10%). I presented three talks on the Duke/OK-4 XUV FEL, Coherent X-Ray harmonics generation, and Xray storage ring FELs (new concepts). These presentations have been published in the Proceedings of the Workshop. The main goal of my presentations was to emphasize that the Duke/OK-4 XUV FEL system, funded by the MFEL program, is already a 4th generation light source and is a very attractive alternative to linac based SASE FELs proposed by Stanford and DESY (Germany).
Light source generations are separated by their average spectral brightness: number of photons per second per unit area per unit solid angle per unit bandwidth. Units are: for area 1 mm2, for solid angle 1 mrad2, for bandwidth 0.1% [i.e., Photons sec-1mm-2mrad-2(0.1% Bandwidth)-1]. It is customary to distinguish a new generation if its brightness is 3-4 orders of magnitude higher then that of the previous generation. The 1st generation has brightness ~1012; thus the second generation operates at ~1015 and the fourth generation at a level of ~1018-20. It is very appealing that MFEL's Duke/OK-4 XUV FEL's brightness is ~1026, and therefore could be called a 4th or even 5th generation light source.
The workshop was set up to promote linac SASE FEL (not officially of course). It took a lot of my effort to convey this message on the very last plenary session. As result, the chairman of the Workshop, Prof. J-L. Laclare, said in this closing speech, that "..while we were discussing 4th generation light sources, we have overlook that they exist, like one at Duke University..". After all, they included in the general conclusion the figure I provided.
This is the summary of most important facts that emerged from the workshop: O.
There is an excellent scientific case for 4th generation light source.
1. According to the response from "users" groups 1&2, the most important
figure of merit for applications is the well established one of average spectral
brightness. Peak brightness and photon flux have very limited requests from the user
community. In addition, coherence of the source is a very attractive feature for the most
applications.
2. It is feasible (Group 2&4) to design a 4th generation storage ring with emittance
of 0.2-0.4 nm. Therefore, it is feasible to extend storage ring FEL oscillators into the
X-ray region.
3. 4th generation storage ring synchrotron light sources will have at least the same level
of brightness as linac driven SASE FEL. Only storage ring FELs incorporating high
harmonics generation could bring 3-4 orders high brightness.
4. The VUV range will be dominated by storage ring FELs with brightness ~1026 . Extension
to the soft-X-ray region seems feasible in the near future. 5. Linac drive FELs with 1 ps
- 100 fs pulses are suited mostly for a limited number of ultra-fast processes.
6. The suggested linac driven SASE FEL divided the opinions of the audience: advocates of
these systems (from Stanford and DESY) are very enthusiastic about their prospects, while
opponents see many obstacles on the way from 8 mm SASE FEL (the only ones operational) to
0.5 angstrom SASE FEL. Successful demonstration or failure of intermediate wavelength
operation (in VUV?) could resolve this division. By Laclare's opinion, linac driven FELs
could succeed storage ring sources at best in 15-20 years.
7. Meanwhile, the synchrotron radiation community is expecting results from XUV storage
ring FELs, i.e. from Duke and Delta. You have first hand information from Duke where OK-4
project has no priority and proceeds in slow fashion. The Delta project is in a very
advanced stage, even though they were far behind Duke two years ago. They have all the
chances to beat us if we can not find means to facilitate OK-4 project.
Overall, the meeting was extremely fruitful and informative. The Duke FEL lab and the MFEL program could achieve a historical bench mark with very decent effort. The MFEL program choice to support the first 4th generation light source, namely Duke/OK4 XUV FEL, is and will be within the main-stream of cutting-edge technology in light source applications.
On my way to Grenoble, I stopped in Paris to visit the L.U.R.E. laboratory (Orsay) and to discuss 1) mirror degradation issues at the Super-ACO FEL; and 2) stabilization of the time jitter of the FEL micro pulse.
These issues are relevant to the Duke storage ring OK-4 XUV FEL for MFEL
research. Stability of the mirror reflectivity is critical for continuous runs and high
efficiency of the MFEL applications. Time jitter is crucial for MFEL pump-probe
experiments where sub-picosecond stability is required. The following results emerged from
this visit:
1) Initial mirror degradation could be prevented by gradual increase of the exposure to
severe soft-X-Ray radiation from the FEL wigglers. In practical terms, this can be
accomplished by the gradual increase of the operating current or by manipulation with the
X-ray beam shutter. Vacuum pressure near the mirror surface must be maintained below 10-9
Torr to prevent initial degradation. This information is not published by the Super-ACO
group . As a result, we have modified the design of the OK-4 mirror system to include
additional vacuum gauges and monitors. The X-ray shutter was included in the original
(Novosibirsk) design of the OK-4 optical resonator.
2) Time jitter of FEL micro pulses was suppressed at Super-ACO FEL to picosecond level
using the dissector (produced by Novosibirsk BINP) in "time slit" mode. The
dissector continuously monitors photon density at a particular RF phase and this signal is
used for the feedback of its phase. Super-ACO group uses one dissector for feed-back and a
second for observation. Most interesting was the specifics of the implementation of it for
FEL control and its effect on the FEL stability. So far, Super-ACO FEL could operate
stable with this feed-back, and its radiation is used for material science. Here at Duke,
we have three dissectors and one spare dissector tube, but lack some support
instrumentation (synthesizers, amplifiers, scopes). Therefore. implementation of a similar
technique is very straight forward. It would be useful if Duke XUV FEL will operate in
near UV range (-350 nm), which is not a priority for the MFEL program. Overall, my visit
to Orsay, the only operating storage ring FEL in the western hemisphere, was very
fruitful.
3. Feedback
For further information on this report, please contact:
Dr. Vladimir N. Litvinenko
Duke University FEL Laboratory
PO Box 90319
Duke University
Durham NC 27708-0319
vl@phy.duke.edu
The opinions and assessments in this report are solely those of the author and do not necessarily reflect official US Government, US Navy or ONREUR positions.