We provide a detailed account of the design and implementation of the LCLS-II photoinjector laser system, highlighting its major subsystems and innovations that enable the generation of high-energy, high-quality X-ray pulses on the attosecond timescale, crucial for advancing X-ray Free Electron Laser (XFEL) research.

We report on the operation of the first high peak and average power single-pass terahertz free-electron laser (THz FEL) based on a high-brightness photoinjector for developing a THz source for pump-probe experiments at advanced x-ray FELs (XFELs). Electron beams with 17 MeV/c momentum and charge up to 2.4 nC were used to generate narrow-band 3 THz radiation with pulse energies of more than 0.1 mJ. The novelty of the realized method is that the undulator radiation by electron beams is much longer than the radiation wavelength, with FEL starting from coherent beam contribution rather than shot noise. This proved to be not a simple scaling of the XFEL mechanism due to the significant role of the bunching factor in the considered case.

We present a comprehensive overview of the development of the LCLS-II photoinjector laser system, emphasizing its key components and advancements in producing high-quality, high-energy, attoseconds X-ray pulses for X-ray Free Electron Laser science.

This paper presents a comprehensive technical overview of the Linac Coherent Light Source II (LCLS-II) photoinjector laser system, its first and foremost component. The LCLS-II photoinjector laser system serves as an upgrade to the original LCLS at SLAC National Accelerator Laboratory. This advanced laser system generates high-quality laser beams to power the LCLS-II, contributing to the instrument's unprecedented brightness, precision, and flexibility. Our discussion extends to the various subsystems that comprise the photoinjector, including the photocathode laser, laser heater, and beam transport systems. Lastly, we draw attention to the ongoing research and development infrastructure underway to enhance the functionality and efficiency of the LCLS-II, and similar X-ray free-electron laser facilities around the world, thereby contributing to the future of laser technology and its applications.

We present a thorough description of the LCLS-II photoinjector laser system, an instrument underpinning ultrafast X-ray sciences. Our presentation will highlight key components and advancements in generating high-quality, high-energy, ultrashort X-ray pulses in X-ray Free Electron Lasers.

Abstract Advanced experiments using THz pump and X-ray probe pulses at modern free-electron lasers (FELs) like the European X-ray FEL require a frequency-tunable (from 0.1 THz to 30 THz), high-power (> 10 microjoule), narrow-band (~ 1–2%) THz source maintaining the repetition rate and pulse structure of the X-ray pulses. This paper reports the first results from a THz source, that is based on a self-amplified spontaneous emission (SASE) FEL operating with a central wavelength of 100 micrometers. The THz SASE FEL prototype is currently under development at the Photo Injector Test facility at DESY in Zeuthen (PITZ) and uses the same type of electron source as the European XFEL photo injector. The first proof-of-principle experiments were done at PITZ using an LCLS-I undulator to generate high-power, high-repetition-rate THz SASE FEL radiation. Electron bunches with a beam energy of ~ 17 MeV and a bunch charge of up to several nC are used to generate THz pulses with a pulse energy of several tens of microjoules. For example, for an electron beam with a charge of ~ 2.4 nC, 65 microjoules were measured at a central wavelength of 100 micrometers. These proof-of-principle experiments pave the way for a tunable, high-repetition-rate THz source providing pulses with energies in the millijoule range.

For LCLS-II two undulators were installed at SLAC, one for soft and one for hard x-rays. Before the superconducting linac gets turned on the copper linac is providing beams at 120 Hz to these two beam destinations. The 120 Hz can be split in many different ratios between soft and hard via a pulsed magnet. To get an optimized beam for the quite different photon energies the pulsed linac components like modulators and RF can provide many different beam parameters, mainly energies and bunch lengths for the two undulator lines. How this was implemented with timing setups of triggers and finally after the split the necessary matching of the transverse phase space will be discussed. PULSED BEAM MANIPULATION The two undulators have their optimal intensity performance around 10-12 GeV for the Hard X-Ray line (HXR), and 4-6 GeV for the Soft X-Ray line (SXR). Since the copper linac is a pulsed at 120 Hz, each pulse can have different trigger assignments and the RF can have different amplitudes and phases. This enables beams with different energies like in the past [1], different bunch lengths, and different charges at a few different rate ratios, for example one beam can have 0, 1, 10, 30, or 60 Hz, while the other beam gets the rest of the 120 Hz.

An accelerator-based THz source for pump-probe experiments at the European XFEL is under development at the Photo Injector Test Facility at DESY in Zeuthen (PITZ). For the proof-of-principle experiments an LCLS-I undulator is planned to be installed downstream of the PITZ accelerator. The fields of the undulator module 26 have been re-measured at DESY in Hamburg and the results are consistent with earlier SLAC measurements. A model for 3D field reconstruction based on the undulator magnetic measurements has been developed. It includes also a horizontal gradient of the vertical field. Tracking of the 17 MeV/c beam has revealed that the transverse gradient will lead to a significant off-axis trajectory in the horizontal plane. This offset has to be corrected with a steering coil, the design of which is also presented. The performance of the THz generation with the correction coil is discussed as well.

There is a high demand for intense THz sources since “many excitation mechanisms of matter resonate in the terahertz regime” especially for condensed matters. “Accelerator-based THz sources provide the wide tunability together with high intensity and repetition rates beyond 100 kHz, that will enable broad application at the European XFEL to the most interesting scientific problems in the field” [1]. Supported by European XFEL a proof of principle study is started at the Photo-Injector Test Facility located at DESY in Zeuthen site (PITZ). Since PITZ and European XFEL electron sources are identical the X-ray and THz radiation can be produced with identical bunch train structure so that for every X-ray pulse a corresponding THz pulse can be provided for the pump-and- probe experiments.

During commissioning and operation of the Linac Coherent Light Source (LCLS) x-ray Free Electron Laser (FEL) at the SLAC National Accelerator Laboratory electron and x-ray beam size, shape, centroid motion have been studied. The studies, sources, and remediation are summarized in this paper.

Establishing good initial quantum efficiency (QE) and reliable in-situ cleaning for copper cathode in the RF gun is of critical importance for the RF gun operations. Recent studies on the SLAC RF gun test bed indicated that the pre-cleaning (plasma cleaning) in the test chamber followed by copper cathode exposure to air for cathode change leads to a very low initial QE in the RF gun, and also demonstrated that without the pre-cleaning good initial QE >4×10−5 can be routinely achieved in the RF gun with the cathodes of QE <1×10−7 measured in the test chamber. QE can decay over the time in the RF gun. The in-situ laser cleaning technique for copper cathodes in the RF gun is established and refined in comparison to previous cleaning at the linac coherent light source, resulting in an improved QE and emittance evolutions. The physics of the laser cleaning process is discussed. It is believed that the reflectivity change is one of the major factors for the QE boost with the laser cleaning.

The Linac Coherent Light Source (LCLS) has used three copper photocathodes since its commissioning in 2007. Two of three copper cathodes had low initial quantum efficiency (QE) (<1×10) in the LCLS radio frequency (RF) gun. The two cathodes were exposed to the plasma cleaning in the cathode test chamber before installation in the RF gun. Recent studies at the SLAC RF gun test bed at the Accelerator Structure Test Area (ASTA) reveals that the pre-cleaning in the test chamber followed by cathode exposure to air for installation in the gun is the major factor leading to the low initial QE. All four cathodes, without the plasma pre-cleaning prior to the installation in the gun, have demonstrated initial QE>4×10 at the ASTA. Systematic studies also demonstrate that high-power RF gun operation provides an initial QE boost. In-situ laser cleaning for three new cathodes in the RF gun is extensively investigated, and a robust laser cleaning procedure is established at the ASTA with improvements of previous cleaning recipe for the LCLS cathode. The QE was shown to reproducibly evolved to >1×10 from about 4×10 immediately following the laser cleaning over ~3 weeks, a time much shorter than a few months for the previous laser cleaning for the present LCLS cathode. The intrinsic emittance of copper cathodes is recovered to the normal value within 1-2 days following the laser cleaning, much shorter than 3 weeks for previous laser cleaning for the present LCLS cathode. The experimental results at the ASTA, including comparison with the previous cleaning for the present LCLS cathode, are presented in the paper. Physics of the laser cleaning process and the evolution of the QE is discussed. INITIAL QE WITHOUT IN-SITU CLEANING IN THE RF GUN An RF gun test bed located at the SLAC’s Accelerator Structure Test Area (ASTA) has been constructed [1] to study photocathodes for the Linac Coherent Light Source (LCLS) injector cathode operations. The beamline of the ASTA gun test bed duplicates the existing LCLS injector gun system [2], consisting of a chirp-pulse-amplifier laser tripled to 253 nm wavelength, LCLS-type RF gun, a solenoid for emittance compensation, one pair of magnet correctors, a Faraday cup to measure the bunch charge, and a YAG screen to measure beam size and intrinsic emittance. Similar to the LCLS injector, the drive laser is configured for normal incident injection to the photocathode surface using a 45 in-vacuum mirror. The final electron beam energy from the RF gun is about 5.5 MeV. The Cause for Low Initial QE in the RF Gun The LCLS located at the SLAC National Accelerator Laboratory has been successfully operated for users for about 5 years. Its copper-based photo-injector has produced an ultra-low emitance and ultra-fast electron beam for the x-ray free electron laser (XFEL). Since its commissioning in 2007, three identical copper cathodes have been used for the LCLS injector operations with different initial quantum efficiency (QE) values. As illustrated in Fig.1, the first and third (present) LCLS cathodes had unexpectedly very low initial QE [3], about 5×10, while the second one had 6.5×10 of initial QE as expected. Lately, it is realized that both the first and third LCLS cathodes were exposed to the plasma cleaning in the test chamber before cathode exposure to air for installation in the RF gun, while the second one did not have this cleaning process. Very low QE measured in the test chamber drives to proceed to plasma cleaning for the cathode prior to the installation in the gun. During the cathode installation in the LCLS RF gun, the cathodes have to be exposed to air for about 3 minutes for the cathode change due to the lack of loadlock system. The recent observations at the ASTA RF gun reveal that the laser-cleaned areas are much more susceptible to the air exposure than the non-cleaned areas do. The ASTA RF gun is vented to nitrogen and then exposed to air for about 3 minutes to mimic the LCLS cathode change before its vacuum starts to be pumped down. Figure 2 (left) and (right) shows the QE maps before and after the RF gun vacuum venting to air, respectively. Before the RF gun vacuum venting, areas A, B, C, D, E, F, G and H on the cathode have been cleaned by the intensive laser, while the circled center area is not exposed to any laser cleaning. In Fig. 2 (left), bunch charge from areas A, B and C have been evolved to about 7500 units for a given laser energy, equivalent to 1×10 of QE, while the cathode center area has about 3500 units of the bunch charge for the same laser energy, equivalent to about 4×10 of QE. After gun vacuum venting, bunch charge productions from the previously cleaned areas A, B, C, D, E, F, G, and H were dropped to 1500-2000 units shown in Fig. 2 (right), equivalent to about 1-2×10 of QE, but the QE of the cathode center area still remains unchanged, at 4×10. The observations indicate that the cleaned or activated surface is susceptible to the air-exposure, revealing the pre-cleaning is the cause for low initial QE measured in the RF gun. Total four cathodes are characterized, which are not exposed to the pre-cleaning prior to the installation in the ASTA RF gun. All four ASTA cathodes have good initial QE in the RF gun ranging from 4×10 to 8×10, as illustrated in Fig. 1. We conclude that the pre-cleaning ____________________________________________ *The work is supported by DOE under grant No. DE-AC02-76SF00515. Proceedings of FEL2014, Basel, Switzerland THP030 Electron Bunch Generation and Manipulation ISBN 978-3-95450-133-5 769 C op yr ig ht © 20 14 C C -B Y3. 0 an d by th e re sp ec tiv e au th or s process activates the cathode surface, extremely susceptible to the contaminations, such as air-exposure. Without plasma cleaning in the test chamber, high initial QE can be routinely achieved in the RF gun. Figure 1: Original QE of cathodes in the LCLS and ASTA RF guns. Figure 2: QE map before (left) and after (right) the gun vacuum venting. Before the gun venting, areas A to H are processed by laser cleaning while the center area is not cleaned by intensive laser. QE Impact from the High-Power RF Processing For LCLS operations all three photocathodes were characterized in a test chamber prior to the installation in the RF gun. The test chamber utilizes a broadband UV light source followed by a narrow-band monochromator to select the desired photon energy. The photoemission from the cathode under 2.6 kV/m of electric field is measured with nanometer. The QE measured in the test chamber is typically on the order of 1×10 at the desired photon energy of 4.91 eV (253 nm). The cathodes with the low QE observed in the test chamber are then directly installed in the ASTA RF gun. Surprisingly, following high-power RF processing all four cathodes in the gun have demonstrated an initial QE >4×10 under normal operation conditions, two orders of magnitude higher than in the test chamber. The Schottky effect enhances QE by less than an order of magnitude for a copper work function between 4.3-4.7 eV. The two orders of magnitude of QE enhancement in the RF gun therefore cannot be explained by the Schottky effect alone. It is noticed that the high-power RF operation boosts QE at the ASTA RF gun, as shown in Fig. 3. For a new cathode installed in the ASTA RF gun, the initial QE is 3×10 for the first day of electron beam operation. Then the QE increases by 50% after a few days of RF operations. It is logical to assume the RF processing conditions the cathode by removing surface contamination. In combination with the Schottky effect, a much higher QE is observed relative to measurements in a test chamber. Figure 3: QE map for a new cathode in the ASTA RF gun: 3×10 of peak QE for first day with turn-on electron beam (left), 4.5×10 for a week later (right). IN-SITU LASER-ASSISTED CLEANING DEVELOPMENTS IN THE ASTA RF GUN Laser-based cleaning techniques have been widely used to clean metal photocathodes, such as copper and Mg, for more than two decades. A high-intensity laser beam, interacting with the metal cathodes, may ablate the cathode surface, removing surface contamination and possibly changing the cathode reflectivity, thereby resulting in a QE increase. However, the laser cleaning may change cathode’s QE uniformity, thereby electron beam emittance, and also generate unwanted dark current or even deteriorate cathode’s crystal quality, if the laser cleaning process is too aggressive. The laser cleaning was performed for the present LCLS cathode in July 2011 [4]. The QE was evolved to 1×10 from 3×10 over a few months following the laser cleaning. Since then, 1×10 of the QE is essentially unchanged for three years to date for 24/7 users operation. The emittance was recovered to the normal value within three weeks following the laser cleaning. Although the previous LCLS laser cleaning was successful, two major concerns still remain. One concern is the reproducibility of the laser cleaning for different spots on the same cathodes and different cathodes. The other is the need to reduce emittance-recovery time and QE-evolution time following the laser cleaning with refinements of laser leaning process. These concerns are particularly important to deliver reliable and high quality electron beam for a high-impact machine for users, like the LCLS, which drives further studies at ASTA. THP030 Proceedings of FEL2014, Basel, Switzerland ISBN 978-3-95450-133-5 770 C op yr ig ht © 20 14 C C -B Y3. 0 an d by th e re sp ec tiv e au th or s Electron Bunch Generation and Manipulation Improvements of Laser Cleaning at the ASTA

The injector is the low energy part of a linac, where space charge and non relativistic kinematic effects may affect the electron beam quality significantly, and in the case of single pass systems determines the brightness in the downstream components. Following the increasing demand for high repetition rate user facilities, a normal conducting, high repetition rate (1 MHz) RF gun operating at 186 MHz has been constructed at LBNL and is under operation. In the current paper, we report on the status of the beam dynamics studies. For this, a multi-objected approach is used, where both the transverse and the longitudinal phase space quality is optimized, as quantified by the transverse emittance and the bunch length and energy spread respectively. We also report on different bunch charge operating modes, as well as the effect of different gun gradients.

A new laser laboratory has been constructed at SLAC to test and characterize photocathode gun physics and develop diagnostics for ultrafast FEL applications. At the heart of the laboratory is a dual-purpose Ti:Sapphire oscillator/regen laser that can deliver either 1.5-3.5 ps or 25 fs pulse durations. The primary objectives of the photocathode research are to identify reliable Cu cathode cleaning recipes and to produce high quantum efficiency with low beam emittance. The ultrafast applications program is presently aimed at developing spectral- encoding systems for shot-to-shot pulse arrival time diagnostics with 10's of fs timing resolution. In this paper we review the laser system and update status of the physics programs.

The Linac Coherent Light Source is an x-ray freeelectron laser at the SLAC National Accelerator Laboratory. It produces coherent soft and hard x-rays with peak brightness nearly ten orders of magnitude beyond conventional synchrotron sources and a range of pulse durations from 500 to <10 fs. The facility has been operating at x-ray energy from 500 to 10,000 eV. Users have expressed great interest in doing experiments with xrays near the carbon absorption edge at 284 eV. We describe the operation and performance of the LCLS in the newly established regime between 250 and 500 eV.

We report on the performance and the operational experience of the LCLS RF gun copper photocathodes used during the LCLS run I, II, III and IV. We discuss the problems of cathode surface contamination and our experience with methods to remove such contamination. Techniques to obtain high quantum efficiency (QE) while preserving the low emittance quality are discussed. Furthermore, we will present the current status of the installed cathode, its quantum efficiency and the typical injector emittances of the extracted beam.

It is widely believed that a drive laser with uniform temporal and spatial laser profiles is required to generate the lowest emittance beam at the photoinjector. However, for a given 3 ps smooth-Gaussian laser temporal profile, our recent simulations indicate that a truncated-Gaussian laser spatial profile produces an electron beam with smaller emittance. The simulation results are qualitatively confirmed by later analytical calculation, and also confirmed by measurements: emittance reduction of similar to 25% was observed at the linac coherent light source (LCLS) injector with a truncated-Gaussian laser spatial profile at the nominal operating bunch charge of 150 pC. There was a significant secondary benefit-laser transmission through the iris for the truncated-Gaussian profile was about twice that compared to the nearly uniform distribution, which significantly loosens the laser power and quantum efficiency requirements for drive laser system and photocathode. Since February 9, 2012, the drive laser with the truncated-Gaussian spatial distribution has been used for LCLS routine user operations and the corresponding free electron laser power is at least the same as the one when using the nearly uniform spatial profile.