2005 ATF Newsletters

Feb | Dec
 

    Much progress has been made since the last ATF Newsletter was written in August of 2004. Excellent data has been obtained in three successfully completed experiments, a new PhD thesis was defended, and another thesis is in the process of being written. Upgrades of ATF components provide more experimental possibilities, better quality results and are a good basis for attracting new experimental groups.This 10 page newsletter includes a brief report from each of the three completed experiments: Nonlinear Components in Relativistic Thomson Scattering, Stimulated Dielectric Wakefield Accelerator and Optical Diffraction-Transition Radiation Interferometry Beam Diagnostics. It also includes a review of the status of the Chicane Compressor and related Beam Diagnostic, CO₂ laser system upgrade to near terawatt level, the photo injector upgrade to Gun IV with asymmetric filed tuning and new record beam brightness, and details of the completed upgrade of the ATF computer control system.

Back to Top

    We have observed nonlinear Thomson scattering in counter-propagation of the 65 MeV electron and the CO₂ laser beams. Namely, at high laser intensity, an electron can absorb multiple laser photons before emitting a single photon of higher energy. Pulse duration and energy of the laser were estimated to be 15-20 ps (FWHM) and 4 J, respectively. Figure 1 shows energy spectra of Thomson x-rays simulated by the computer code CAIN. The black line is the spectrum at the interaction point. Energy of x-rays produced by the linear (single-photon) Thomson scattering process is limited at 6.5 keV but the nonlinear process extends the spectrum to higher energy. The red line shows the spectrum of the x-ray signal on a detector outside the vacuum beamline after its attenuation in air and on a Beryllium window used for a vacuum seal. To cut off the x-rays produced in linear Thomson scattering, a Silver filter is used. The green line in Figure 1 shows the spectrum of Thomson x-rays reaching the detector filtered by a foil.

Figure 1. Simulated energy spectra of Thomson x-rays: (black) at the I.P.; (red) on the detector; (green) filtered by a foil

Observation of a signal on the Si diode detector after passing through the Ag foil was the first indication and quantitative estimate of the nonlinear x-ray yield.

Figure 2. Transverse profiles of Thomson x-rays (simulation): (a) the linear process;
(b) the second order nonlinear process.

In addition to the spectral shift, the nonlinear Thomson scattering process has another characteristic pattern in angular distribution. Figure 2 shows transverse profiles simulated by CAIN for (a) linear Thomson scattering and (b) the second order nonlinear process, where we assumed the laser is linearly polarized and the polarization plane is along the x-axis. One can notice that the second harmonic has the minimum intensity along their e-beam axis where a linear component has the maximum.

Figure 3. Transverse profiles of x-rays (experiment): (a) without filter; (b) with filter.

In our experiment, transverse profiles of Thomson x-rays are observed using a luminescent Kodak screen and a CCD camera. Figure 3(a) shows an x-ray profile experimentally observed without the Ag filter, while Fig. 3(b) is taken with the foil filter. One can see a two-lobe pattern appear in Fig. 3(b). When the laser polarization changes to the orthogonal, the 2-D x-ray distribution rotates 90⁰.

Figure 4 shows (a) simulated histogram of 2-D x-ray distribution behind the filter and (b) processed experimental pattern from Figure 3(b). The distance between the two peaks in the experimental distribution appears consistent with the simulation. Note that the simulation does not take into account a smearing of the picture due to angular divergence of the electron beam focused at the interaction point.

Figure 4. 2-D normalized occurrence distribution of x-rays in the plane of detector:
(a) simulation; (b) experimental distribution after background subtraction.

Back to Top

    In this experiment, the successful superposition of wake fields excited by 50MeV bunches which traveled ~50cm along the axis of a cylindrical waveguide that is lined with alumina was demonstrated. The bunches were prepared by splitting the laser pulse incident on the rf photocathode and inserting an optical delay. Wake fields from two short (5-6psec) 0.15-0.35nC bunches were superimposed and the energy losses of each bunch was measured as the bunch separation was varied so as to encompass approximately one wake field period (~21cm). A spectrum of 40 TM_0m eigenmodes was excited by the bunch. The energy loss of the second bunch exhibited a narrow resonance with a 4mm (13.5psec) footprint [Fig.1]. A substantial retarding wake field (2.65MV/m·nC for just the first bunch) was developed because of the short bunches and the narrow vacuum channel diameter (3mm). This should be compared with a related experiment [1] done at Argonne National Laboratory, where the bunch spacing was not varied, and a much weaker wake field (~0.1MV/m·nC for the first bunch) having ~10 eigenmodes was observed after a train of much longer bunches.

    Fig.1 Measured difference in energy losses (bars) normalized per 1 m between the 2^nd and 1^st bunches vs. the bunch spacing. Both bunches have the same rms-length z - 6.0 ± 0.43psec. The solid line represents the theoretical fit (L is the wake field period); it has a piece-wise character because the charges change from one experimental point to another (200-340pC).
    Experimental studies of the high- frequency spectrum of wake field radiation set up by a relativistic electron bunch in a channel surrounded by a dielectric resulted in the development of a technique [2] to measure bunch RMS-length in the psec range and below, and eventually in the fsec range. The proposed technique a) is nondestructive, b) is routine to calibrate (no other independent technique is required to perform calibration), and c) can be a single shot technique. The overall accuracy is proportional to 1/z², with z = being the RMS length. Measurement of the millimeter-wave spectrum will determine the RMS bunch length in the psec range. This was done using a series of calibrated narrow-bandwidth mesh filters. For the experiment at the ATF with the range of filtered frequencies up to 120GHz, one may determine the RMS-length z3-4 psec with an accuracy 1.5%; the minimum RMS-length one may expect to resolve (accuracy 100%) z 450 psec. Using the same dielectric-lined structure but changing to the range of filtered frequencies up to 300GHz, one could resolve the RMS length z 190 psec.
[1] Phys. Rev. ST Accel. Beams 3, 101302, (2000), J.G. Power, et al.
[2] AIP Conf. Proc. 737, 421, (2004), S. V. Shchelkunov, et al.

Back to Top

   We have recently (January, 2005) achieved success in imaging the far field interference pattern of optical diffraction-transition radiation (ODTRI) produced by the interaction of the 50 MeV ATF beam with an interferometer consisting of a 5thick copper micromesh (750 lpi) and a aluminized silicon mirror oriented at 45 degrees with respect to the beam. The observed interference of ODR from the mesh and OTR from the mirror are sensitive to the horizontal and vertical rms beam divergences at the respective beam waists. The ODTRI technique [NIMB, 201, 153-160 (2003)] overcomes the inherent limitation introduced by beam scattering in the first foil of a conventional two foil OTR interferometer. We have compared the divergences obtained using ODTRI, standard OTRI employing a 0.7 micron Al front foil and a multi-position beam imaging technique combined with transport code calculations for a standard ATF beam tune. All three techniques agree to within experimental error thus establishing the validity of the ODTRI divergence measurement technique.

The two images above show ODTR (left) and OTR (right) interference patterns. The images were obtained by integrating over 480 and 360 seconds respectively. The pulse repetition rate was 1.5 Hz. A 16 bit cooled CCD imaging camera (Apogee Instruments, Alta E47+), standard relay optics, whose field of view of approximately 0.1 rad or 10/, where is the Lorentz factor, and a 650nm filter with a 10nm band pass were used to obtain both images. The left and right scans below show horizontal line scans (faint lines) taken through the centers of each of the corresponding ODTRI and OTRI images, compared with simulation code calculations (dark lines). The fitted values of horizontal rms beam divergences obtained are 0.3±0.03 (ODTRI) and 0.35±0.03 (OTRI). The multi-foil simulation results produce a value of 0.31 for the horizontal beam divergence, which is excellent agreement with both of these values.

*email: rfiorito@umd.edu

Back to Top

After a period of trials and errors, the UCLA-ATF chicane commissioning is well on its way. The electron beam was propagated through the chicane, with minimum dispersion, and the focusing properties were measured in agreement with the model. The helium-cooled THz sensitive bolometer was installed to look at the coherent radiation at the junction between the 3rd and 4th chicane magnets. The bolometer was operated successfully last week, and very strong CER (Coherent Edge Radiation) signal was observed.

A sharp dependence of CER amplitude on the bunch compression was measured, and the spectrum was characterized. At the point of maximum compression (peaked CER), a structure in the e-beam is observed behind the bend. Detailed data analysis is coming.

Figure 1: Signal from bolometer [volts] VS. RF phase offset between gun and linac.

Figure 2: “Beam breakup” due to over compression shown on the BPM in the dispersion region (FPOP-UP1)

Back to Top

   Since the last status report of the ATF CO₂ laser in the ATF Newsletter of February 2004, much progress has been made toward bringing the CO₂ laser in line with the increasingly challenging demands of ATF user experiments on advanced laser acceleration (STELLA LWFA, LACARA) and radiation sources (Thomson scattering, EUV).

Figure 1.

Modification of the front end and the amplifier chain allowed us to achieve reliable outputs at shorter pulse duration (down to 10 ps) and higher peak power (close to 1 TW). The first successful demonstration of this new regime is the observation of nonlinear Thomson scattering reported in this newsletter.

Figure 2.

Figure 1 shows the present optical configuration of the ATF CO₂ laser system. A 10 ns pulse, gated by a Pockels cell from the CO₂ oscillator output, is sent for pre-amplification. This compensates for the relatively low efficiency of the short ps pulse switching on the Kerr cell down the road. After preamplification, we perform optical semiconductor switching by combining CO₂ and YAG lasers on a Germanium Brewster window at a much higher CO₂ power then we did previously. The 200 ps CO₂ laser pulse produced on the Germanium switch is further truncated to 10-15 ps inside the Kerr cell. The present configuration allows for fast alteration between two regimes of laser operation: a 200 ps or 10 ps pulse. Such changing of regimes is useful to synchronize the laser with a picosecond electron beam and to study energy- or power-related nonlinear effects, etc. High-pressure regenerative and final amplifiers pictured in Figure 2 provide sufficient bandwidth and gain to amplify ps pulses to ~10 J of the output energy, which is delivered to experiments.

Figure 3.

We continue efforts to improve reliability and overall performance of the ATF CO₂ laser system. Presently, we are replacing our 15-year old CO₂ oscillator, with a new commercial device shown on Figure 3. The combination of the new oscillator and regenerative amplifier, another recent upgrade, allows production of Gigawatt CO₂ laser pulses at the repetition rate of the ATF linac. This will speed-up laser/e-beam synchronization and co-alignment tasks and benefit several ATF experiments that require modest laser power for microbunching the electron beams (Resonant PWFA, PASER). Our long-term plans include upgrade of the ATF YAG laser to enable slicing of 2-3 ps CO₂ pulses with the second harmonic of YAG radiation as has been described in the February 2004 ATF Newsletter.

Back to Top

    The original 1.6 Cell electron gun was replaced with a widely used modification GUN IV. The original gun was installed at the ATF in 1996 and achieved record beam brightness with 0.8 micron normalized emittance for a 0.5 nC beam. The main purpose of the upgrade was to eliminate multipactoring due to migration of Mg from the cathode to other gun surfaces that resulted in unstable operation of the gun.
There are few distinct features of the new gun:

1. “all-copper” cathode with RF seal ring designed at the BNL Source Development Laboratory;
2. laser port of the increased aperture that allows to achieve a 4 mm laser spot on the cathode;
3. asymmetric tuning, with the field reaching 160 MV/m in the half cell while the full cell operates at 110 MV/m gradient;
4. improved vacuum pumping utilizing laser ports in the half cell.

   We were unable to manufacture a good cathode with a Mg plug in time for installation. An all-copper cathode was installed instead of the Mg one traditionally used at ATF. The Mg plug in the cathode offers about an order of magnitude improvement in quantum efficiency up to 0.5% but typically limits the gradient. Installation of a “simple” copper cathode allowed the ATF to supply a stable beam to users and achieve another record in beam brightness due to high gradient: 0.9 micron for a 0.8 nC beam.
    The beam energy from the gun reached 5.8 MeV (a stable beam operation was ~5.5 MeV). Field balance was measured with the network analyzer suggesting a misbalance of 1.3:1 (half-cell/full-cell). We confirmed with Parmela simulations that higher gradient is more important then 1:1 field balance for the emittance. Attempts to balance the gun with RF tuners located in the full cell lead to the reduction of a sustained filed and output beam energies. Hopefully these studies will continue to achieve a brightness corresponding to 1m @ 1 nC.
    We are in the process of installing a Mg cathode that would offer multiple beam capabilities required by some ATF experiments. These experiments need bunch separation equal to or shorter then the RF period. Reduction of available laser energy as a result of splitting the Nd YAG laser beam to generate such beam structure requires a high quantum efficiency cathode.

Back to Top

   In mid-June 2004, failure of a primary cooling fan caused a power supply to overheat in our venerable VAX-based computer control system which had been in service for 14 years. The resulting power supply spikes lead to a cascade of damage, affecting the CPU, a memory board, the front panel bulkhead, on-board diagnostics, the time-of-year clock, Ethernet interface, one disk drive, the CAMAC serial highway driver and a CAMAC crate controller. While the VAX was repairable, parts for the CAMAC highway driver were no longer available, leaving ATF without the means to control any of its equipment. Fortunately, work was almost complete on the planned migration from the VAX to the new Intel-based replacement system as announced in the previous ATF User Meeting. This serial highway driver problem arose at the same time ATF planned the summer 2004 shutdown to change the electron gun, thus adding additional urgency to an already very busy summer.
    I am pleased to report that work on the migration was indeed completed by the end of the summer shutdown and the new computer control system was commissioned and has been fully operational since then. Our new system changed many items, including:

- Processor: Single VAX CPU --> 8-way Intel Pentium
- Data communications: 5 MHz CAMAC byte-serial highway --> 100 MHz Ethernet
- Crate control: L-2 serial crate controllers --> Ethernet controller
- Operating system: VMS --> Linux
- Application language: Fortran, C --> C++
- Graphics, database development tools: Vsystem/VMS --> Vsystem/Linux

   The new host machine has many state of the art features, but those of most interest to our users are redundant hot-swap power supplies, hot-swap RAID mass storage and uninterruptible power supplies/power conditioning on all components of the control system.
    The new system is substantially faster than the old one; database builds, for example, now take 30 seconds, compared to 17 minutes using the old equipment. Both our operators and experimenters have remarked on the change in the responsiveness of the new system.
While many changes were introduced to upgrade the control system, the guiding requirement to minimize any impact to our users was always kept in mind. To that end, the entire set of ~800 window displays was ported to the new system. ATF users see exactly the same displays as on the old control system with no need for any retraining. In addition, all services such as network socket connections to the database, Mathcad interface, frame grabbers, etc. work the same under the new system as before. Also, all database channel names were preserved.
    We have been very pleased with the success of this upgrade and the new conveniences and opportunities it affords us and our users. We are confident the power and potential of our new system will bring new services to our users and will report on them in future issues of the newsletter.
    Special thanks to the US DOE's Dave Sutter and Bruce Strauss who recognized our request to upgrade the control system and provided the funding to do so.

Back to Top

For previously published ATF Updates, see ATF Newsletter Home Page:

ATF Newsletters

General ATF information is at the ATF Home Page:

ATF home page

Descriptions of the ATF Experiments are at the ATF Experiments Home Page:

ATF experiments

See also ATF Beam Time Schedules Home Page:

ATF Experiments Schedules

Back to Top

Last Modified: December 3, 2007
Please forward all questions about this site to: Vitaly Yakimenko

One of ten national laboratories overseen and primarily funded by the Office of Science of the U.S. Department of Energy (DOE), Brookhaven National Laboratory conducts research in the physical, biomedical, and environmental sciences, as well as in energy technologies and national security. Brookhaven Lab also builds and operates major scientific facilities available to university, industry and government researchers. Brookhaven is operated and managed for DOE’s Office of Science by Brookhaven Science Associates, a limited-liability company founded by Stony Brook University, the largest academic user of Laboratory facilities, and Battelle, a nonprofit, applied science and technology organization.

Privacy and Security Notice  | Contact Web Services for help