Radiofrequency (RF) electron guns operating at high accelerating gradients offer a pathway to producing bright electron bunches. Such beams are expected to revolutionize many areas of science: they could form the backbone of next-generation compact x-ray free-electron lasers or provide coherent ultrafast quantum electron probes. We report on the experimental demonstration of an RF photoemission electron source supporting an accelerating field close to 400~MV/m at the photocathode surface. The gun was operated in an RF transient mode driven by short $\sim 9$~ns X-band (\SI{11.7}{\giga\hertz}) RF pulses. We did not observe any major RF breakdown or significant dark current over a three-week experimental run at high accelerating fields. The demonstrated paradigm provides a viable path to forming relativistic electron beams with unprecedented brightness.

We have demonstrated generation and transport of a patterned electron beam from a Diamond Field-Emitter Array (DFEA) cathode in a radio frequency (rf) gun. DFEAs are arrays of micrometer-scale pyramids with nanometer-scale tips. They can be fabricated with base widths ranging from 3 μm to 25 μm and pitches as small as 5 μm. They have an inherent 1:0.7 base to height ratio. DFEAs operate as field-emitter cathodes and potentially produce intrinsically shaped electron beams, which are of interest for a number of accelerator applications. We report on the results of a recent experiment in which a beam, consisting of several beamlets, was produced from a DFEA cathode in an rf gun and transported 2.54 m along a beam line. A macrobunch charge of 60 pC was measured at a cathode field gradient of 15.1 MV/m.

Collinear wakefield acceleration has been long established as a method capable of generating ultrahigh acceleration gradients. Because of the success on this front, recently, more efforts have shifted towards developing methods to raise the transformer ratio (TR). This figure of merit is defined as the ratio of the peak acceleration field behind the drive bunch to the peak deceleration field inside the drive bunch. TR is always less than 2 for temporally symmetric drive bunch distributions and therefore recent efforts have focused on generating asymmetric distributions to overcome this limitation. In this Letter, we report on using the emittance-exchange method to generate a shaped drive bunch to experimentally demonstrate a TR≈5 in a dielectric wakefield accelerator.

Argonne Wakefield Accelerator (AWA), has been commissioned and in operation since last year. It can provide beam of several bunches in a train of nanoseconds and 10s of nC with energy up to 70 MeV. In addition, the AWA can accommodate various beamlines for experiments. One of the proposed experiments is to use the AWA beam as a diagnostics for time resolved high density material, typically a target with high Z and time dependent, imaging experiments. When electron beam scatters after passing through the target, the angular and energy distribution of beam depend on the density and thickness of the target. A small aperture is used to collimate the scattered electron beam for off axis particles, and the target image will be detected by imaging screen. By measuring the scatted angle and energy at the imaging plate would yield information of the target. In this paper, we report on the AWA electron imaging (EI) system setup, which consist of a target, imaging optics and drift. The AWA EI beam line was installed on June, 2016 and the first test run was performed on August, 2016. This work will have implication on the high energy density physics and even future nuclear fusion studies. The details of AWA EI experiment setup, results, analysis and discussions are presented here.

We report on the experimental generation of relativistic electron bunches with a tunable longitudinal bunch shape. A longitudinal bunch-shaping (LBS) beam line, consisting of a transverse mask followed by a transverse-to-longitudinal emittance exchange (EEX) beam line, is used to tailor the longitudinal bunch shape (or current profile) of the electron bunch. The mask shapes the bunch's horizontal profile, and the EEX beam line converts it to a corresponding longitudinal profile. The Argonne wakefield accelerator rf photoinjector delivers electron bunches into a LBS beam line to generate a variety of longitudinal bunch shapes. The quality of the longitudinal bunch shape is limited by various perturbations in the exchange process. We develop a simple method, based on the incident slope of the bunch, to significantly suppress the perturbations.

The Argonne Wakefield Accelerator beamlines have stringent vacuum requirements (100 picotorr) necessitated by the Cesium telluride photoinjector. In direct conflict with this, the structures-based wakefield accelerator research program sometimes includes worthy but complex experimental installations with components or structures unable to meet the vacuum standards. A proposed chamber to sequester such experiments safely behind a thin beryllium (Be) window is described and the results of a study of beam-quality issues due to the multiple scattering of the beam through the window are presented and compared to GEANT4 simulations via G4beamline. Three thicknesses of Be foil were used: 30, 75 and 127 micron, probed by electron beams of three different energies: 25, 45, and 65 MeV. Multiple scattering effects were evaluated by comparing the measured transverse rms beam size for the scattered vs. unscattered beam. The experimental results are presented and compared to simulations. Results are discussed along with the implications and suggestions for the future sequestered vacuum chamber design. MOTIVATION FOR USE OF BE WINDOWS AT AWA The Argonne Wakefield Accelerator (AWA) beamlines have a demanding vacuum environment to preserve the drive gun Cs2Te photocathode. Such photocathodes require vacuum pressures on the order 10−10 torr. The strict vacuum requirements have a large impact on experimental design, severely limiting material choices to those that are UHV compatible. In addition, the UHV requirement prevents easy access to experimental structures after installation and usually prohibits the possibility of altering experimental setups within the timeframe of an experiment. Thus, the experiment must work "as installed". If it does not, at best, the consequences can include lengthy downtime while the experimental area is vented and equipment is uninstalled, modified, cleaned, and re-installed. Once this occurs, the offending sector of the beamline must be pumped to attain UHV vacuum pressure before the operations and the experiment may resume. This can take several days to more than one week. Compromise: Separate vacuum regimes One way of easing the vacuum requirement and allowing quick and easy access to make changes to the experimental setup is to place a vacuum chamber sequestered behind a Be window at the end of the beamline (discussed previously in [1]). The vacuum requirement in the "dirty" vacuum chamber can be relaxed to 10−8 torr, which can be attained in a matter of hours with much fewer restrictions. Of course, there is a cost: beam quality suffers due tomultiple scattering as the beam passes through the Be window. An electron beam traveling through matter primarily interacts with the nuclei via the Coulomb force. Electrons experience many mostly small deflections as they scatter multiple times within the media. The distribution of scattering events is described by Moliere’s theory. The details of the theory are beyond the scope of this paper. However, it is important to point out that the predictions of scattering theory become increasingly less reliable as the foil thickness is decreased. Hence, it is important to gather some experimental data in order to understand what to expect. The foils used in these studies are very thin: 127, 75 and 30 micron. The studies described here were designed to develop guidance that can be used in simulations and planning for experiments using such an installation in the near future. It was hoped to use the results of these studies to develop guidance to be used in planning such installations in the future by trying to measure the effects on the beam transverse size and understand how well it is matched to numerical and analytical predictions. AWA has already had some experience with this limiting effect of the increase in beam transverse size and emittance due to a Be window. Two experiments (one involving an RF choke cavity and another involving a photonic-bandgap (PBG) structure [2]) come to mind. Both devices had an aperture I.D.=6 mm and also required the beam to be moved within the aperture from an on-axis position to offaxis without significant beam loss inside the structure. In other words, a tightly focused beam was required with a fairly constant transverse size much less than the aperture I.D. Performing these experiments with a beam scattering through a Be window was indeed a challenge. TESTING SEVERAL THIN BE FOILS AT AWA A motorized actuator was equipped with a custom Be foil holder designed to hold foils of three different thicnesses, 127 μm, 75 μm, and 30 μm probed by electron beams of three different energies: 25, 45, and 65 MeV. The laser pulse length was 6 ps FWHM. ISBN 978-3-95450-180-9 Proceedings of NAPAC2016, Chicago, IL, USA WEPOB20 2: Photon Sources and Electron Accelerators T12 Beam Injection/Extraction and Transport 937 Co py rig ht © 20 16 CC -B Y3. 0 an d by th e re sp ec tiv e au th or s Diagnostics: Two YAG(Ce) scintillator screens for beam spot size and an ICT to measure charge. The first YAG captured the initial size of the beam before scattering, and was located at the foil z-position (on the same actuator). The second YAG was located 88 cm downstream. A quadrupole triplet was used to focus the beam to a small spot at the foil position. Thus the two factors that could be studied were foil thickness and beam energy by observing the increase in transverse beam spot size rmsx and rmsy. See Fig. 1 for a schematic of the experimental setup. Figure 1: Schematic of the experimental setup. The electron beam propagated from the left, charge measured with an ICT (not pictured) before the foil. The beam was focused by 3 quadrupoles. The image was recorded at the two YAG screens located 88 cm apart, YAG1 at the Be foil position and YAG2 88 cm downstream. After recording the initial beam image at YAG1, A systematic comparison of beam spot sizes for similar beam conditions on the YAG2 screen with and without the different Be foils in place was performed. The beam energy was varied by turning off RF cavities in the beamline and re-tuning the beam. The beam energy was 65 MeV, 45 MeV and 25 MeV. In each case the mean charge was about 1.5 nC, but the charge varied from less than 1 to more than 2 nC due to laser jitter. Three quadrupoles located about 1.5 m upstream from the Be foil were used to focus the beam to a small spot on the YAG screen at the position of the foil. Then the beam imagewas captured using cameras interfaced through a framegrabber, there and at the second YAGwith and without the Be foils. The images of the beam spot was analyzed (See Fig. 2). Projections from fits of the intensity distributions were analyzed to calculate the transverse spotsizes rmsx and rmsy. Each table presents the mean results of data taken for one of 3 beam energies, after cut on charge. Charge jitter due to laser intensity fluctuations.was from 0.6 nC to 2.3 nC. Data was cut to include the range from 1.1-2.2 nC. The initial beam size was made very small to ensure that the beam would not be clipped at the window which has an aperture of 1 cm diameter. The YAG size is 50 mm diameter. The smallest rms sizes extracted have a larger error due to the reduction in pixels available for the fit routine (poor resolution). It is hoped to repeat the experiment with a more tightly focused camera to reduce this source of error. Experimental results are presented in Fig. 3 (25 MeV data), Fig. 4 (45 MeV data) and Fig. 5 (65 MeV data). Figure 2: A representative example of the 45 MeV intensity distributions with the projections from fits for the initial spot at YAG1 and the spots at YAG2 for case of each Be foil and no foil. The effect of scattering on the beam size was quite pronounced, resulting in transverse sizes that were as much as 5 times the un-scattered beam size. However, the case of the 65 MeV beam with 30 micron foil seems promising. Figure 3: 25 MeV results for the 3 Be foils, no foil, and the initial spot at YAG1. Figure 4: 45 MeV results for the 3 Be foils, no foil, and the initial spot at YAG1. Figure 5: 65 MeV results for the 3 Be foils, no foil, and the initial spot at YAG1. COMPARISON OF SIMULATION RESULTS AND EXPERIMENT RESULTS The goal of the simulations was to see how well the code could predict the effect of multiple scattering in terms of the experimentally measured transverse spot sizes. The simulations employed G4beamline [3], a particle tracking code WEPOB20 Proceedings of NAPAC2016, Chicago, IL, USA ISBN 978-3-95450-180-9 938 Co py rig ht © 20 16 CC -B Y3. 0 an d by th e re sp ec tiv e au th or s 2: Photon Sources and Electron Accelerators T12 Beam Injection/Extraction and Transport which provides an interface to GEANT4, which does the multiple scattering calculation. According to the GEANT4 reference manual, the multiple scattering algorithm is based on the Lewis theory, which is more complete than Moliere’s theory [4]. The simulations were simplified to the case of a Gaussian beam with the initial energy and transverse size at the screen set to reflect the average from the data. However, not all the beam parameters are well known. For both simulations, a Gaussian beam was assumed and the experimentally measured parameters beam energy (65, 45, and 25 MeV), mean charge (1.5 nC), and initial spot-size at the first YAG screen were the inputs. A comparison of the simulation and data results is shown in Fig. 6 below. Some of the discrepancies are quite large, the closest match being for the 25MeV, 30 μm foil case. In most other cases the simulation overestimates the scattering effects. The sources of error most likely include the initial beam distribution ( the real beam is not a simple Gaussian), energy spread, laser jitter, and other unknown details of the particle distribution, as well as the multiple scattering algorithm in the simulation.In addition, another source of error is the poor resolution in the experimental measurement of the smaller spot siz

The AWA L-band, high-charge photoinjector for the 70 MeV drive beamline has been operating for almost 3 years at the Argonne Wakefield Accelerator (AWA) facility. at Argonne National Laboratory (ANL). The gun operates at high-field (85 MV/m peak field on the cathode) and has a high quantum efficiency (QE) Cesium telluride photocathode with a large area (30 mm diameter). It produces highcharge, short pulse, single bunches (Q > 100 nC) as well as long bunch-trains (Q > 600 nC) for wakefield experiments (high peak current). During the first two years of operation, photocathode performance was evaluated and areas of improvement were identified. After study, consideration and consultation, steps were taken to improve the performance of the photocathode. So far, in total, three photocathodes have been fabricated on-site, installed and operated in the gun. Improvements made to the photocathode plug, vacuum system, and gun operation are detailed. The results include vastly improved conditioning times, better cathode performance, and QE above 4% for over 11 months. THE ARGONNE WAKEFIELD ACCELERATOR (AWA) DRIVE PHOTOCATHODE GUN The AWA L-band drive gun for the new 75 MeV drive linac has been commissioned and is operating. The 1.3 GHz photo-injector operates at high gradient (85 MV/m). The 31 mm dia. Cesium telluride photocathode, specifically designed for the production of high charge, is fabricated on-site. The method of fabrication used at AWA was based on and developed from methods published by researchers at LANL and INF-LASA and described in detail elsewhere [1, 2]. Using those sources for guidance, the AWA Cs2Te photocathode is fabricated in a UHV chamber with a base pressure of 1.5 × 10−10 Torr and transported to the drive gun for installation in a UHV load-lock chamber. The photoinjector generates high-charge, short pulse, single bunches (Q > 100 nC) and bunch-trains (Q > 450 nC) for wakefield (and other) experiments. The photocathode requirements at AWAwere determined by the drive beam parameters. The AWA drive beam parameters are summarized in Table 1 [3]. Table 1: AWA Drive Beam Cathode Operating Parameters Cathode peak RF field >85 MV/m RF pulse length ≈7.6 μs Average dark current <5 nC/RF pulse

We report on investigations into the fundamental surface emission parameters, the geometric field enhancement factor (β) and the work function (φ), by making both field emission and Schottky-enabled photoemission measurements. The measurements were performed on a copper surface in the Tsinghua University S-band RF gun in two separate experiments. Fitting our data to the models for each experiment indicate that the traditionally assumed high value of β(≈50-500) does not provide a plausible explanation of the data, but incorporating a low value of φ at some sites does. In addition, direct measurements of the surface conducted after the experiment show that β is on the order of a few, consistent with our understanding of the electron emission measurements. Thus we conclude that the dominant source of electron emission in high gradient RF cavities is due to low φ sites, as opposed to the conventionally assumed high β sites. The origin of low φ at these sites is unclear and should be the subject of further investigation.

We present and discuss the results from the experimental generation of high-charge annular (ring-shaped) electron beams at the Argonne Wakefield Accelerator (AWA). These beams were produced by using laser masks to project annular laser profiles of various inner and outer diameters onto the photocathode of an RF gun. The ring beam is accelerated to 15 MeV, then it is imaged by means of solenoid lenses. Transverse profiles are compared for different solenoid settings. Discussion includes a comparison with Parmela simulations, some applications of high-charge ring beams, and an outline of a planned extension of this study.

We report on dark current and Schottky-enabled photoemission from a copper photocathode surface. Field-emitted dark current is a major gradient-limiting factor in RF cavities. Field emission is generally attributed to geometrical projections on the bulk surface whose field enhancement factor (beta) and the emitting area (A(e)) can be extracted from the Fowler-Nordheim (FN) plot. Measurements were made at Tsinghua S-band RF gun facility in two separate experiments. Using the traditional FN formula for RF fields we discovered that field enhancement factor (beta) alone cannot explain the full data set. Instead, we found that a low work function at some sites is required. In addition, surface analysis of the cathode after the experiment shows that geometric beta indicated would be < 10. Thus we conclude that low work function sites with a small beta are responsible for dark current emission and subsequent breakdown in high-gradient structures. The origin of these sites is unclear but could be due to defects or impurities.