An Optical Streaking Method for Measuring Femtosecond Electron Bunches

Conference · Wed Dec 14 00:00:00 EST 2011

OSTI ID:1032758

Ding, Yuantao; Bane, Karl L.F.; Huang, Zhirong

The measurement of the ultra-short electron bunch length on the femtosecond time scale constitutes a very challenging problem. In the x-ray free electron laser facilities such as the Linac Coherent Light Source, generation of a sub-ten femtoseconds electron beam with 20pC charge is possible, but direct measurements are very difficult due to the resolution limit of the present diagnostics. We propose a new method here based on the measurement of the electron beam energy modulation induced from laser-electron interaction in a short wiggler. A typical optical streaking method requires a laser wavelength much longer than the electron bunch length. In this paper a laser with its wavelength shorter than the electron bunch length has been adopted, while the slope on the laser intensity envelope is used to distinguish the different periods. With this technique it is possible to reconstruct the bunch longitudinal profile from a single shot measurement. Generation of ultrashort x-ray pulses at femtoseconds (fs) scale is of great interest within synchrotron radiation and free electron laser (FEL) user community. One of the simple methods is to operate the FEL facility at low charge. At the Linac Coherent Light Source (LCLS), we have demonstrated the capability of generating ultrashort electron-beam (e-beam) with a duration of less than 10 fs fwhm using 20 pC charge. The x-ray pulses have been delivered to the x-ray users with a similar or even shorter pulse duration. However, The measurement of such short electron or x-ray pulse length at the fs time-scale constitutes a challenging problem. A standard method using an S-band radio-frequency (rf) transverse deflector has been established at LCLS, which works like a streak camera for electrons and is capable of resolving bunch lengths as short as 25 fs fwhm. With this device, the electrons are transversely deflected by the high-frequency time-variation of the deflecting fields. Increasing the deflecting voltage and rf frequency are the right direction to achieve a better resolution. For example, by choosing an X-band transverse deflecting cavity, the expected resolution for LCLS beam with 4.3 GeV is about 1 fs rms. Typically the rf breakdown threshold and the power source availability prevent going to even higher voltage and frequency. With the highly-developed laser techniques, we can choose to streak the beam at optical frequencies. By jumping from rf to optical frequency, the wavelength is shortening by 4 to 5 orders. With an electron bunch length shorter than half period of the laser, we can apply the similar rf deflecting or zero-phasing method for e-beam bunch length measurements using a high-power laser. A short wiggler is required to provide interaction between the electron and the laser. For example, to measure the e-beam at the order of 1 m rms length, a laser with its wavelength of 10 {mu}m may be considered. For a typical few GeV e-beam, the wiggler period has to be large to satisfy the resonance condition. Also, if the e-beam is longer than one laser period, the different modulation periods will overlap and we cannot distinguish them. So this method is so far limited by the achievable long-wavelength laser power. To get an effective modulation on an e-beam of 4.3 GeV, the required laser power is about a few tens GW. In this paper we propose to adopt a high-power Ti:Sapphire laser (wavelength of 800 nm), and use the slope in the intensity envelope to distinguish the different modulation periods. First an ultrashort electron beam interacts with the Ti:Sapphire laser in a wiggler, where the electron energy is modulated at the same periods of the laser. If the laser pulse is long and the short electron bunch is overlapped (in time) with the middle part of the laser, such as the setup at LCLS laser heater, the different energy modulation periods on the electron beam will be overlapped on the energy profile. In this conditionwe typically have a double-horn distribution of the energy profile, and the electron-bunch length information cannot be retrieved. But if the laser pulse (with a Gaussian temporal shape) is relatively short, we can synchronize the e-beam with the laser at the slope region of the intensity envelope, and the amplitude of each energy modulation period will be different. Then these electrons pass through a dispersive section such as a spectrometer, after that this periodically-modulated energy profile can be observed in a transverse screen. By properly choosing the laser parameters, each modulation period will show as a separate streak on the screen. This modulation period in energy dimension correlates to the laser wavelength in time dimension. Since the laser wavelength is a known parameter, no additional calibration is needed. This provides a single shot, self-calibrated method for ultrashort electron bunch length measurements.

• Research Organization: SLAC National Accelerator Lab., Menlo Park, CA (United States)
• Sponsoring Organization: USDOE Office of Science (SC)
• DOE Contract Number: AC02-76SF00515
• OSTI ID: 1032758
• Report Number(s): SLAC-PUB-14576; TRN: US1200551
• Resource Relation: Conference: Contributed to the 33rd International Free Electron Laser Conference (FEL 2011), 22-26 Aug 2011. Shanghai, China
• Country of Publication: United States
• Language: English

Related Subjects

43 PARTICLE ACCELERATORS
AMPLITUDES
AVAILABILITY
BEAM BUNCHING
BREAKDOWN
CALIBRATION
DIMENSIONS
DISTRIBUTION
ELECTRON BEAMS
ELECTRONS
FREE ELECTRON LASERS
LASERS
LIGHT SOURCES
LINEAR ACCELERATORS
MODULATION
RESOLUTION
RESONANCE
SHAPE
STREAK CAMERAS
SYNCHROTRON RADIATION
WAVELENGTHS
Astrophysics
ACCPHY
XFEL