7.iv. Strategic plan for the facility for the next five years.

Transcription

1 7.iv.Strategicplanforthefacilityforthenextfiveyears. Executive Summary ) New Instrumentation Initiatives... 4 a. Macromolecular Femtosecond Crystallography (MFX) Instrument in Hutch b. Serial Sample Chamber (SSC) for CXI... 5 c. XCS Upgrade... 6 d. Beam Sharing... 7 e. Detector Program... 9 f. Rapid Data Analysis ) New Beam Capabilities a. Multi-Color X-Ray Beam Modes b. Soft X-Ray Self-Seeding c. Soft X-Ray Polarization d. Electron Energy Stability e. XTCAV ) Efficiency a. Linac Infrastructure Mission Readiness b. Commonality of Controls/DAQ c. Prioritization Process, WBS Structure d. Cross Training of Staff ) Support Facilities a. Laser b. Sample Preparation c. X-ray Optics Metrology Laboratory d. Other Laboratories ) Science Initiatives a. Roadmap to Single Particle Imaging b. Matter under Extreme Conditions and High Energy Density Science ) Roadmap to LCLS-II Operations a. Definition of the Instrumentation Needs for LCLS-II Operations b. Strategic R&D in Support of LCLS-II Operations c. Construction and Implementation of Instrumentation in Support of LCLS-II Operations Future 7.iv pg. 1

2 ExecutiveSummary The world s first hard X-ray free-electron laser (FEL), the Linac Coherent Light Source (LCLS), achieved lasing on April 10, 2009, started user assisted commissioning on October 1 of the same year, and became a dedicated user facility in August Every six months a new instrument completed commissioning and by November 2012 all six instruments were in user operation. The science achieved at the LCLS was extremely impressive from the very beginning, leading to a rapidly growing user demand. To keep pace with this demand and to maximize the impact of these new science opportunities, LCLS chose a start-up strategy of rapid growth in operations and a support mantra of ensuring the success of every experiment. Last summer, a subcommittee of the Basic Energy Science Advisory Committee (BESAC) issued a report on future X-ray Light Sources that was approved by the Basic Energy Science Advisory Committee on July 25. The report states that despite intense international competition, an exciting window of opportunity exists for the U.S. to provide a revolutionary advance in X-ray science by developing and constructing an unprecedented X-ray light source that should provide high repetition rate, ultra-bright, transform limited, femtosecond X-ray pulses over a broad photon energy range. To address these recommendations, and in close dialogue with BES, SLAC proposed major modifications to the LCLS-II upgrade. The modified LCLS-II project will include a new superconducting (SC) linac capable of producing intense electron pulses at up to 1 MHz repetition rate and energies up to 4 GeV. This SC linac will feed a variable gap soft X-ray undulator producing high rep-rate X-ray pulses in the kev range. Also included in the project is a new variable gap tender/hard X-ray undulator that will replace the existing LCLS undulator. When driven by the SC linac, it will serve an energy range of 1-5 kev at high rep-rate, and when driven by the existing 120 Hz Cu linac it will reach photon energies up to ~25 kev. This upgrade puts LCLS on the path of continuing leadership as the premier XFEL facility and builds a platform that will allow future capacity and capability increases in both high rep-rate and high peak power XFEL science. As stated in the BESAC report, anticipated experiments include time-resolved physics and chemistry studies to record movies of how bonds break and form, how energy flows at the molecular level, how charge is transferred in nanoscale electronic devices, and how complex solids with correlated electrons function. Furthermore, the new LCLS sources will enable the exploration of new approaches to study ultrafast (magnetic) data storage devices, and pump-probe imaging of biological and other soft material systems that are critical to reveal mechanisms of energy (charge) transfer in biological molecules and biomimetic systems. With these tremendous opportunities, our goal can only be to keep LCLS at the forefront of FEL science for the decades to come. In order to achieve this goal, LCLS will streamline user operation to a support model that is efficient and scalable to LCLS-II. Under our plan the funding for core X-ray operations will be reduced by ~20% over the next two years and the core accelerator operation funds will be reduced by ~10% over three years. The funds made available by these reductions will be invested in strategic R&D and improvements that set us on the path to operate LCLS-II once first light is achieved. In addition, we will make major investments in accelerator operation mission readiness, a program that will replace some of the aging linac infrastructure to reduce the risk of failures and ensure long-term operation efficiency gains. In order to ensure continued scientific excellence we will invest in facility instrumentation and standardized platforms, focus on efficiency gains by scheduling and cross training, increase LCLS dependence on user resources for experiment preparation and execution, and take advantage of growing user experience and proficiency with XFEL experiments. Future 7.iv pg. 2

3 To carry out both short-term and long-term strategic initiatives we are working in close collaboration with our constituents, in particular the scientific user community, the Department of Energy, SLAC, and Stanford University. We ask for advice and feedback from our LCLS Scientific Advisor Committee (SAC) and the SLAC Science Policy Committee (SPC), and use our LCLS Proposal Review Panel (PRP) for the ranking of the user science proposals. SLAC has recently launched several initiatives to strengthen the relationships of LCLS within the Laboratory, in particular with the Accelerator Directorate, SSRL and SLAC Photon Science. These working groups and retreats will help to identify areas of growth and strategic hires important to LCLS. The annual SSRL/LCLS User Meeting and associated workshops are used to exchange ideas and information and engage the user community in the LCLS science strategy, and we will start the LCLS-II workshop series with a series of workshops in the week of September 22-26, In the following we describe the initiatives and actions that we will undertake over the next 5 years. Specifically we will present our new instrumentation initiatives, new beam capabilities, efficiency, support facilities, science initiatives and finally the roadmap to LCLS-II operations. Under this 5-year plan, LCLS will continue to improve operations and instrumentation to deliver high impact science in the near term while we are building up the infrastructure to utilize the unique properties provided by LCLS-II with facilities matched to the grand challenge scientific questions in the long term. We believe this will ensure that LCLS remains the leader in X-ray FEL science in the near term as well as the in the decades to come. Future 7.iv pg. 3

4 1)NewInstrumentationInitiatives Six instruments in separate hutches exist at LCLS. They have been serving a wide range of scientific areas since 2009 by allowing extreme flexibility and multiple configurations. As the LCLS facility matures, some scientific areas are arising as higher demand and higher impact. A quest for higher efficiency of operation and increased beamtime through multiplexing leads to the need for more hutches with more dedicated capabilities. The construction of new hutches and the development of more multiplexing techniques are key to the future of LCLS. a.macromolecularfemtosecondcrystallography(mfx)instrumentinhutch4.5 The Macromolecular Femtosecond Crystallography (MFX) instrument is planned to expand the unique scientific capability of LCLS by building a new x-ray instrument inside a new hutch (Hutch 4.5) at the end of a dedicated new branch of the LCLS beam distribution system (for schematic layout see Fig. 1.a.1). MFX addresses a growing need and demand in structural biology, especially in Femtosecond Crystallography, by delivering a flexible atmospheric pressure system capable of studying crystals of various sizes using various sample delivery techniques. LCLS will strategically invest in a new beam endpoint to enhance to the capabilities and functionality of the facility. Partnerships will be sought to fund an endstation capable of fully utilizing the beam delivered by the new beamline. The MFX beamline and endstation will augment the LCLS facility and its existing six instruments. The MFX project, managed as an X-ray Improvement Project (XIP) under the LCLS directorate, will be built over a period of 3 years, and will be phased into operation by FY17 if external funding arrives as planned. The MFX instrument concepts have been developed based on the experience gained by LCLS operations, in consultation with external user groups, SLAC Photon Sciences and SSRL and was encouraged by the LCLS Science Advisory Committee. MFX will be optimized for hard x-ray diffraction and scattering studies of crystals and other samples at atmospheric pressure, addressing basic structural biology problems for radiation-sensitive samples as well as hard-to-crystallize macromolecules and complexes. The MFX instrument will make use of the diffract-before-destroy technique primarily for structural biology studies where the extremely short pulses of LCLS will be used to obtain damage-free structures of biological systems. The instrument is intended to be operated at atmospheric pressure with a versatile platform for a variety of sample delivery techniques and experimental geometries. X-ray diffraction will be the primary tool in MFX, but other techniques such as imaging and spectroscopy will also be possible. Provisions will be made for the addition of a versatile pump laser system for the study of dynamics under laser excitation. Other types of dynamics will also be ideally measured at MFX, with for example the study of rapid reactions using room temperature fast mixing sample delivery systems such as jets. The instrument design will emphasize versatility but within the context of an instrument optimized for structural biology studies at atmospheric pressure. Funding was provided by DOE-BER in April 2014 for the initiation of the MFX project. This funding was provided to Prof. Soichi Wakatsuki in the sum of $2,000,000 under FWP# , entitled SLAC Future 7.iv pg. 4

5 Mesoscale Integrated Biology Pilot Project Concept Paper. This funding was provided to initiate the creation of an integrated platform for research on complex biological phenomena related to development of the MFX station at the SLAC Linear Coherent Light Source (LCLS). Fig. 1.a.1: Schematics of the planned MXF hutch located in a new hutch located in the downstream section of the current XCS hutch. b.serialsamplechamber(ssc)forcxi For the majority of its experiments, the CXI instrument uses a detector with a central hole to measure the forward scattering patterns from weakly scattering samples. As a consequence, the majority of the photons pass unchanged through the detector, only to terminate their course on some target such as a viewing screen or a beam dump. Effectively, CXI dumps photons into a beam dump 120 times per second for most of its operating conditions. This feels like a bit of a waste of good photons. To this end, an XIP project was undertaken to design, build and install a new sample chamber in an empty space between the two existing detector chambers utilizing the CXI 1 micron sample environment. The spent beam from the existing chamber, after passing through the hole of the first detector, can be refocused using lenses into this new sample chamber. The beam can then be reused for a second independent experiment in a parasitic manner that does not affect the primary experiment. The design of this Serial Sample Chamber (SSC) is shown below. This parasitic experimental system will allow for a significant increase in available beamtime. This is expected to be particularly useful for expanding the user base for serial femtosecond crystallography by Future 7.iv pg. 5

6 allowing for crystal screening runs with the spent beam, but it might also profit other instruments as CXI will use the beam more efficiently. The Serial Sample Chamber (see Fig. 1.b.1) will also be an invaluable tool to develop techniques toward single particle imaging. Fig. 1.b.1: The left of the schematics shows the serial sample chamber located downstream of the one-micron focus sample chamber shown to the right. c.xcsupgrade An ultrafast laser system will be installed on the XCS instrument to increase its portfolio of experiments by performing optical-laser-x-ray pump-probe experiments. The system will provide short (35 fs and below) pulses that can take advantage of the time sorting techniques that are now available at LCLS. The initial system would comprise a compact version of the standard commercial oscillator and regenerative amplifiers that are employed in the other hutches. This system would be capable of making 3.5 mj, fs pulses that are synchronized to the x-rays using the standard SLAC timing-synchronization system. The laser would be installed in a separate laser room (see Figure 1). However, the space constraints in the FEH require a modification of the existing hutch to enable a laser room in the vicinity of the Hutch 4 sample location (see Fig. 1.c.1 below). The diagonal wall near the XRT will be changed to a right-angle protrusion to encapsulate the laser room in the original Hutch 4 area. The laser enclosure will retain space for an optional fiber broadening system to compress the pulses to the sub-10 fs regime as well as for a future OPA, laser beam diagnostics, and time-tool beam generation. Additionally, a time-tool chamber will be installed in the x-ray beamline to allow for time sorting down to sub-10 fs precision. Future 7.iv pg. 6

7 This laser system will significantly expand research capabilities and dramatically increase scientific throughput of the XCS instrument. This enhancement will allow a new class of pump probe experiments at LCLS, namely those requiring large sample to detector distance in SAXS and WAXS geometry (highresolution Bragg peak deformation, SAXS measurements in biomolecules or polymers, to name a few). Together with an x-ray split-and-delay (S&D), an optical laser would dramatically extend probe options at LCLS. It would for example, provide novel tools to study non-equilibrium/stimulated dynamics within nanosecond to picosecond time windows, which is one of key areas in modern material science (glassy dynamics, aging, stress/strain relaxation, etc.). Such unique combination of S&D type of XPCS experiments with pump and probe techniques allows studying of dynamics in a length- and time-scale regime not currently accessible anywhere. Fig. 1.c.1: Location of the XCS (Hutch 4) laser room and laser system. In addition to an ultrafast laser system, the XCS x-ray beam transport will be modified to permit the delivery of pink beam into Hutch 4 along the identical beam trajectory as the large offset monochromator system. This will be accomplished via the addition of an offset mirror pair in the x-ray transport tunnel. This enhancement will increase scheduling flexibility, in particular to relieve the heavily oversubscribed XPP instrument and in turn increase the amount of multiplexed experiments performed. d.beamsharing Since LCLS only has one undulator, it has historically operated in a serial manner. Efficient scheduling of experiments has been further complicated by the practical fact that removing the pink beam instrument and pipe takes too much time therefore pink beam experiments in XPP must be scheduled opposite of either AMO or SXR experiments. In an effort to try and drive efficiency and provide more available beam Future 7.iv pg. 7

8 time, we explored techniques for multiplexing the beam so that two or more experiments can be run at the same time. The first technique involved placing very thin diamond crystals in the XPP monochromotor. In this case, the required narrow spectral band of the FEL pulse is sent to the XPP experiment. About 80% of the remaining spectral power transmits through the crystal and can be used in one of the experiments in the Far Experimental Hall (FEH). This was first demonstrated when a monochromatic experiment was performed in XPP and an imaging experiment was performed in CXI simultaneously. Another way of multiplexing the beam is related to the fact that the large laser systems and sometimes complex target alignment schemes demand low repetition rate (1 shot every 7 minutes) operation. In this case, by rapidly moving the MEC mirror into and out of the beam, we can run experiments in XCS or in CXI during the 6 minutes between MEC shots. If the XPP instrument is in monochromatic mode, this mirror moving technique can be combined with the transmission through the mono thus allowing three experiments to be run at the same time. Both forms of multiplexing were made available to users in run 8 and 20% of the experiments scheduled in run 9 received beam time because of this capability. Both techniques require a reasonable amount of coordination between the mutiplexed experiments, as all experiments must be at the same wavelength to within the FEL bandwidth. Furthermore, all experiments are delivered the same temporal pulse width. While we intend to increase the level of multiplexing, this required coordination of beam parameters makes it impossible to multiplex some experiments. Another way to share the beam is related to CXI only. This involves re-collecting the non-scattered X- rays and re-focusing those X-rays into a second experimental chamber downstream of the main CXI chamber. This allows two imaging experiments to be run simultaneously in CXI as long as both experiments require the same beam parameters In the future we want to make these multiplexing schemes even more capable by adding the thin crystals to the XCS monochromator and adding a laser and the capability to get pink beam into XCS. This will mitigate the scheduling problems mentioned above related to pink beam experiments in XPP by allowing some of those experiments to be performed in XCS. Finally, as discussed above, we are underway on a project to build an in air imaging station (MFX) in the back part of the XCS hutch. Beam to this instrument will be steered to this hutch using a mirror at the same position as the MEC mirror which will allow this instrument to also be multiplexed. Fig. 1.d.1 below shows how the beam can be multiplexed once we have thin crystals in both monochromators and MFX on line. In this configuration, if there were two compatible mono experiments in XPP and XCS, we could then use the mirror to quickly switch the remaining spectral components of the pink beam to MFX, CXI, and/or MEC. Future 7.iv pg. 8

9 Fig. 1.d.1: Schematic layout of the hard x-ray hutches at LCLS benefitting from current and future beam sharing capabilities. e.detectorprogram The users of LCLS perform a wide variety of science experiments employing an assortment of techniques for which a suite of detectors with a variety of parameters are needed. To provide this in a manner, which uses the fewest resources and results in the shortest time-to-science, all the new LCLS detectors will be built around a common platform known as epix. They will share part of the circuit chip, firmware, software, printed circuit boards, and mechanical components. Not only will this approach produce more cameras with fewer resources, but it will significantly reduce the technical and schedule risk. At present, four members of the epix family are envisioned over the next few years: epix100, epix10k, epixs, and epixm. Along with the fccd, all the epix detectors are capable of being scaled to higher frame rates in the LCLS II era. One outstanding need of the LCLS, hard x-ray science community is a detector with small pixels and low noise for applications, such as x-ray photo-correlation spectroscopy and fluorescence spectroscopy using dispersive crystal geometries. To meet this need, epix100, the first member of this new family of detectors, has been designed. The prototype circuit chip has demonstrated performance that meets all specifications. The design of the science-size, production circuit chip has been completed and recently has been submitted for fabrication. Over the next few years, several megapixels of epix100 cameras will be built and deployed for science operations. CSPAD detectors have been used in the overwhelming majority of hard x-ray science experiments performed at the LCLS to date. However, even with the redesigned second generation circuit chip, the performance of CSPAD limits many of the user experiments. For example, both femto-second serial crystallography and time-resolved pump-probe experiments require a detector with large dynamic range and the ability to record weak signals. Specific characteristics that must be improved include noise and maximum signal. epix10k will be significantly better in both these parameters and offer other critical advantages as well. The goal is to cut the noise in half relative to CSPAD, which will allow weak diffraction peaks and other low-intensity signals to be more clearly seen. In order to record bright diffraction peaks without attenuating the LCLS beam, the maximum signal that epxi10k can measure is 4 times greater than CSPAD. A further critical enhancement relative to CSPAD is the incorporation of auto-ranging for the gain in a pixel-by-pixel basis and occurs automatically in real time for each pulse. epix10k is the second member of the epix detector family and is expected to supersede the CSPADs in all the hard x-ray instruments as about ten megapixels worth of cameras are built and deployed in the hutches in 2015 and Future 7.iv pg. 9

10 Silicon drift detectors (SDDs) have been used in a few LCLS user experiments, but with mixed results at least partially due to their very limited channel count, while at synchrotrons and in many other fields SDDs have proven to be a critical technology. epixs, where the S is for spectroscopy is an adaptation of epix100, that has larger pixels and lower noise. It will be capable of offering SSD-like energy resolution, but with hundreds to thousands of channels, essentially creating an SSD for FELs, where channel count and not count rate is the defining parameter. The cost per epixs channel is expected to be 100 times lower than for SDDs. LCLS has recently acquired a two-plane pnccd system from the Max Planck Society, which has seen heavy use in four different hutches since its arrival in the fall of Given the supply of spare parts and the historical rate of module damage, this camera is expected to be a mainstay of the LCLS soft x-ray program for two to three years before becoming non-functional. A pair of technologies is being pursued as a replacement to the pnccd and to complement its performance: epixm and the fccd from LBNL. epixm is a monolithic CMOS sensor with a fully-depleted bulk, which is a technology first developed at Stanford. epixm will offer very low noise, very small pixels, and is expected to cost significantly less per pixel than detectors based on hybrid architectures or custom sensors. The second generation Fast CCD from LBNL has been used in one characterization experiment and one user experiment at the LCLS in early Active discussions are underway to acquire a camera for dedicated use at LCLS and to collaborate on future designs matched to the specific needs of the LCLS. Pixel Pitch Chip Size (pixels) Frame Rate Noise (r.m.s.) Maximum Signal CSPAD 110 m 185x Hz 1,100 ev 20 MeV 2 epix m 352x Hz 225 ev 800 kev 1 Gains epix10k 100 m 176x192 1 khz 650 ev 80 MeV 2 Auto Ranging fccd x Hz ~100 ev 200 kev Table 1.e.1 Expected initial performance specifications of the existing detectors. 2DElectron/IonDetector Over the past two decades momentum resolving spectrometers have transformed electron and ion spectroscopy. Momentum spectroscopy has enabled a new class of experiments in both laser based ultrafast sciences, and x-ray spectroscopy. The fundamental concept of these spectrometers is that the arrival time and position of each fragment is measured. From this data the momentum vector of each particle can be calculated. Science examples requiring such a detector range from Coulomb explosion imaging of large molecules for supporting alignment information in diffractive imaging experiments to time-resolved photoelectron imaging of nanoscale targets. So far, there exists no suitable detector technology for this class of experiments with fourth generation light sources such as LCLS. Future 7.iv pg. 10

11 Fig. 1.e.1: Basic detector design: The thin silicon sensor will be bump bonded to the ASIC readout logic. We are developing a revolutionary detector for electron and ion spectroscopy based on silicon technology. The concept of the new detector is that a charged particle impinging on the detector will create electron-hole pairs in a thin silicon sensor. The generated charges will be read out with high timing precision. The basic detector design is shown in Fig. 1e.1. A thin silicon sensor will be bump bonded to fast read-out electronics based on an Application-Specific Integrated Circuit (ASIC). As the electron mobility in silicon is much higher than the hole mobility, the sensor will be designed so that the electrons travel through the sensor to the read-out electrode and the holes travel to the entrance electrode. We have started designing a first prototype of the detector funded by the SLAC LDRD program. The anticipated performance of the detector has a spatial resolution of 100 microns and a time resolution of a few hundred picoseconds. The prototype detector chip will have 48x48 pixels. The first prototype detector shall have an array of 2x2 chips resulting more than 9000 particle detection channels. We anticipate having a first functional prototype in mid FY15. Test measurements and iterative improvements of the detector chip are planned for FY15 and FY16. By the end of FY16 we will perform benchmark experiments showcasing the performance of the new detector with a science application. Upon successful demonstration of the new detector we are planning to apply for follow up funding and scale the system to larger chip sizes. Further, we plan to concurrently deploy the new detector technology for LCLS experiments were multi-hit capabilities are required. f.rapiddataanalysis The ability to rapidly process data is critical to the success of LCLS experiments and for the efficient use of beamtime. This is due to the various complexities associated with experiments at a free electron laser: x-ray pulse fluctuations, relatively small allocation of beamtime for each experiment, and sample mutation issues. Experimenters need the information obtained from data analysis to make an educated decision on how to proceed with their experiment. The LCLS has developed two tools to accomplish this task. The first tool, the online Analysis and Monitoring Implementation (AMI), is a user-configurable, GUIbased analysis that does not require any user coding or preparation. The second tool is the interactive psana (Photon Systems ANAlysis) software. The features of interactive psana include: Support for both C++ and python; Command line interface Ability to capture commonly used algorithms in reusable modules that can be chained together in a serial fashion; Support for calibrating images using standard tools; Future 7.iv pg. 11

12 A new Data Description Language (DDL) that allows automatic code generation for both C++ and python data access; Ability to run the same software offline and online (with real-time plot display); Ability to analyze data parallelizing over events (up to thousands of cores). The psana framework is becoming the main analysis framework for LCLS because of its unique advantages: it is simple to use, free and open-source, able to use the same code for real-time monitoring and for offline analysis, and works both interactively and in batch mode. A few critical projects are required to make it widely adopted: Increase the speed through local and Message Passing Interface (MPI) parallelization Increase the flexibility through XTC (online data file) indexing Add the ability to display real-time plots Add more powerful algorithms driven by LCLS physics. User training by the instrument engineers in the use of ipsan Adoption of more general detector handling APIs with a reduced set of access functions We will execute these projects in the next 2 years. Future 7.iv pg. 12

13 2)NewBeamCapabilities In the following section we describe different beam modes that were developed over the last years. The techniques described here are in varying stages towards standard user operation and several LCLS scientists and users have started to employ these capabilities for their experimental programs. Activities to fully develop their operation will be conducted during MD times and priorities of those developments are set in the joint LCLS division meetings to reflect the scientific program identified by LCLS. a.multi-colorx-raybeammodes During the past two years various methods were developed to deliver two color and two-bunch x-ray beams: A method for SASE beams using two undulators is depicted in Fig. 2.a.1. Using this approach, the undulator system is split into two independent systems that amplify the FEL pulses under slightly different conditions, thereby generating pulses with two different energies. It is possible to tune the photon energy, pulse delay and pulse duration. A typical spectrum is shown in Fig. 2.a.2. The isase method interleaves the K values of alternating undulators (Fig. 2.a.3). It is thereby possible to generate a beam containing two distinct photon energies with zero delay with respect to each other. A dual photon energy pulse with adjustable delay can be generated using the double slotted foil which splits one electron beam pulse into two by spatially masking out a part of the electron pulse energy spectrum in a high dispersive region. This method is illustrated in Fig.2.a. 4. A dual bunch can be generated using a double drive laser pulse at the photocathode that generates the FEL beam. This is accomplished using a pulse splitter/stacker as part of the drive laser system. Multi-ColorSchemes Fig. 2.a.1: Generating dual pulses using undulators set to different K values. Color, pulse delay and duration are tunable. Future 7.iv pg. 13

14 Fig. 2.a.2: Two-color beam on SXR spectrometer (18 fs pulse duration, 1.5 kev, 19 ev energy separation). Fig. 2.a.3: Two-color generation using the isase method. Alternating K1 and K2, instead of simply detuning gives a Two-Color scheme instead of simple bandwidth reduction. For K1 = K2 it is the regular SASE. K1 K2 gives 2,3,4 colors configurations, depending on the phase advance. Fig. 2.a.4: Generating dual pulses using undulators set to different K values and slotted foil application. Color, pulse delay and duration are tunable. The distance between slots controls the delay between pulses. Future 7.iv pg. 14

15 TwocolorbeamdeliveryusingtheHXRSSsystem Some of these methods can be used in combination with HXRSS to deliver dual pulses which are separated in time and/or energy. One particularly intriguing method is to take advantage of intersecting crystallographic planes of the HXRSS diamond crystal that provides the opportunity to generate two seeding energies by taking advantage of the relatively broad energy spectrum of SASE FEL beam that is transformed into the seed. The hard x-ray self-seeding device was conceived to deliver a very narrow seeded spectrum using a particular Bragg or Laue diffraction plane of a thin diamond crystal. The seeded photon energy is controlled by crystal pitch and the electron beam energy. Due to strong user interests in two-color FEL operation, the LCLS accelerator team used the crystal yaw angle to bring two diffraction planes within the SASE bandwidth (about 20 ev) at the photon energy 8.45 kev for simultaneous two-color seeding. The color separation can be completely controlled by yaw angle alone without adjusting any other machine configuration (see figure below). The technique is scalable to any seeded photon energy ranging from 7.1 kev to 9.5 kev. Fig. 2.a.5 illustrates the arrangement of the crystallographic planes of the diamond crystal, Fig. 2.a.6 schematically displays the arrangement of the seeding crystal with respect to the FEL beam (left) and shows the dual energy pulses (right). Fig. 2.a.5: Photon Energy as functions of diamond crystal yaw angle for accessible crystallographic planes. Future 7.iv pg. 15

16 Fig. 2.a.6: Left: Orientation and rotational movements of the diamond crystal with respect to the FEL beam. Right: Dual photon energy pulse spectra for different crystal yaw angles. SeededBunchFELOperationforHigh-Intensity2-ColorX-Rays A recent development for the Hard X-ray Self Seeding beam program is the delivery of seeded 2 bunch 2 color X-rays (also for SASE operation). Two electron pulses are generated at the photocathode, accelerated and propagated through the undulator system and HXRSS systems. SASE or Seeded 2-color pulses with adjustable delay have been achieved with peak power comparable to standard operation (this is an improvement of 10 over other methods). This 2-color scheme enables new types of pump-probe and imaging experiments at hard x-rays! Typical energy separation of the two pulses is about 80 ev, with pulses of ~ 20 fs and 180 J per pulse. Figs. 2.a.7 and 2.a.8 show examples the seeded spectra. This new mode is currently transitioning to operations and we conducted a first successful user experiment. Fig. 2.a.7: Spectrometer image of a seeded 2 Bunch 2-Color x-rays beam. Future 7.iv pg. 16

17 Fig. 2.a.8: Integrated lineout of a seeded 2 Bunch 2-Color x-ray beam. b.softx-rayself-seeding The goal of the SXRSS project was to design, construct, and test a FEL soft x-ray self-seeding device. Two institutions, SLAC and LBNL, contributed to the design, engineering, fabrication, and testing. The design is similar to the HXRSS system. Instead of using a crystal, a ruled grating provides the capability of diffracting the incoming x-ray beam to select the seeding wavelength (Fig. 2.b.1). The design provides a wavelength range of 500 ev to 1 kev for seeding in the soft x-ray photon energy range with a design resolving power of Fig. 2.b.1: Soft x-ray Self seeding System (M1, M2, M3: X-Ray Mirrors, G: Grating, lower part illustrates the arrangement of the electron chicane bending magnets). Future 7.iv pg. 17

18 Recently, the R&D phase of the project came to completion. Seeding at 860 ev has been demonstrated with a relative bandwidth of less than 2x10-4. The optimization has been done at this photon energy. In additional seeding was observed across the design wavelength range (0.5-1keV). Currently, work is performed to transition the SXRSS system to normal operation to prepare for upcoming user demands. The goal is to be able to provide seeded beams in the Soft x ray range to user with the beginning of the upcoming run 10 (October 2014). Main tasks are streamlining the operation of the system, optimization of seeding across the entire design energy range and increasing the peak brightness at all energies. We expect to expand the energy range up to 1200 ev and possibly as low as 300 ev. A part of the performance optimization is to develop effective undulator tapering strategies, which are also required to increase pulse energies towards the terawatt goal. To aid all of these studies as well as the user program, a new shot-by-shot soft x-ray spectrometer is under study. In the longer term (3-5 years), we are considering the possibility of two-color soft x-ray selfseeding, which will require additional hardware (for timeline see Table 2.b.1). Time FY14 FY15 FY16-FY17 FY18 and beyond Project Phase Transition to Operation Expansion of energy range (300 ev ev) Development of two-color soft x-ray self-seeding Adopt system to LCLS-II parameters Table 2.b.1: Planned timeline of the Soft X-Ray Self Seeding program at LCLS c.softx-raypolarization A polarizing undulator (Delta undulator) and an up-beam phase shifter will be installed at the end of the LCLS undulator line in place of U33 and commissioned by the end of CY2014. The LCLS FEL produces intense x-ray radiation from a high energy electron beam with fixed, linear polarization. The addition will be compact, similar to the existing LCLS undulator segments, but can be operated in one of three modes to either produce fully tunable polarized x-ray FEL beams, or boost fundamental or second harmonic linearly polarized FEL output. Controlling the polarization state of FEL radiation has broad applications for probing valence charge, spin, bonding dynamics, phonons and emergent phenomena on fundamental time and spatial scales. The 3.2-m-long Delta undulator is expected to produce polarized radiation intensities at the GW-level with degrees of circular polarization close to the 90% level at soft x-ray photon energies up to about 1.5 kev. Fig. 2.c.1 shows performance estimates for the basic setup. The degree of polarization can be further enhanced by reverse tapering of the LCLS planar undulators and/or through collimation of the linear polarized light from the LCLS planar undulators. The system will be operational from early FY15 until the decommissioning of the LCLS facility. With the phase shifter the DELTA undulator can be used to investigate the so-called cross-undulator configuration in order to provide rapid polarization switch. Shot-by-shot polarization diagnostics will be developed and installed in Future 7.iv pg. 18

19 the LCLS. The Delta undulator will later be used, together with up to 2 additional undulators of the same type, as polarizers for the LCLS-II SXR beamline (for timeline see Table 2.b.2). Time FY14 FY15 FY16-FY18 Project Phase R&D Phase Commissioning and Transition to Operations Routine Operation Table 2.b.2: Timeline of the planned Soft X-Ray Polarization program at LCLS. Fig. 2.c.1: Estimated LCLS Delta Polarized Radiation Properties. d.electronenergystability LCLS beam energy stability was previously found to be outside the acceptable limits for stable operation of seeded FEL operation. Following recent linac configuration changes and regular maintenance, the energy stability is now found to meet or approach the requirements to stabilize self-seeded FEL intensity to the empirically nominal 20% RMS level. The maximum relative RMS energy jitter requirements are found to be 2.5e-3 for HXRSS and ~5e-3 to 6e-3 for SXRSS. As shown in Fig. 2.c.2, RMS energy jitter for hard x-ray operation has been reduced from >5e-3 to 2.4e-3, reducing the HXRSS power fluctuations from >72% to 20%, as expected. The impact of energy jitter (< 7e-3) on the recently developed SXRSS is still being evaluated. Future 7.iv pg. 19

20 In the short term (3 months), efforts will be concentrated on ensuring regular delivery and verification of the optimized setup. This includes the deployment of newly developed linac jitter analysis tools for operations to quickly identify maintenance issues, as well as modifying controls to allow the increased flexibility required to regularly apply the new linac phasings used to further reduce jitter. As the above corrections leave zero tolerance for general linac performance, hardware improvements are still recommended in the long term (1-2 years), both to relax operational constraints and to further improve general performance. Prior pulse forming network upgrades have shown mixed results and are being evaluated. High-level RF correction via fast feedback through LLRF modulation is being alternatively considered. Fig. 2.c.2: Two-dimensional histograms (top) of the LCLS seeded FEL pulse energy, less SASE contributions, versus relative beam energy with one-dimensional histograms (bottom) of relative beam energy. Plots shown for HXRSS at 8.3 kev in May 2012 (left) and April 2014 (right) demonstrating a factor of >2 reduction of energy jitter. e.xtcav An X-band transverse deflecting cavity (XTCAV) was installed downstream of the LCLS undulator beamline and tested in early 2014 for user operation. This device measures the electron bunch timeenergy phase space distribution. Since it is located after the undulator, time-resolved FEL lasing effects (electron energy loss and energy spread increase) can be measured. By comparing images between FELon and FEL-off conditions, we can reconstruct the X-ray temporal profile for each lasing shot without interrupting FEL operation. A schematic depiction of the system is given in Fig. 2.e.1. Future 7.iv pg. 20

21 Fig. 2.e.1: Diagnostics layout of the X-ray temporal measurement. The deflector and camera can work at 120 Hz so each shot can be recorded. Presently the number of pixels in the energy (vertical) direction is limited to < 350 pixels at this rate. This is acceptable for hard X-ray modes where the full vertical range is not required. For soft X-rays, to avoid image truncation, the full ROI is preferred, but the camera must be acquired at 60Hz or less for beam-synchronous acquisition. The best temporal resolution measured is about 1 fs rms for soft X-rays (800 ev), and about 4 fs rms for hard X-rays (8 kev). As an example, the temporal profile reconstruction of soft X-ray pulses is illustrated in figure 2.e.2. The electron beam energy is 4.7 GeV with FEL operating at a resonant photon energy of 1.0 kev. We first suppressed the lasing process by perturbing the electron horizontal trajectory at the beginning of the undulator and then recorded hundreds of what we refer to as lasing-off or baseline longitudinal phase space images on the screen. Panel a in figure 2.e.2 shows a typical single-shot baseline image. Its projection onto time gives the electron bunch current profile. Next, we restored the electron trajectory and recorded the lasing-on images for normal operation (a single shot is shown in panel b). On comparing panel b with a in Fig. 2.e.2, one can clearly see the time-resolved energy loss and energy spread growth due to the lasing process. Fig. 2.e.2: Temporal profile reconstruction. The electron bunch charge is 150 pc with an energy of 4.7 GeV to produce photons at 1.0 kev. The measured single-shot longitudinal phase space images are shown in a: lasing off and b: lasing on. Comparing the lasing-off with the lasing-on images, we reconstruct the X-ray power profile as shown in c from the time-dependent energy loss (blue curve) and energy spread growth (red curve). The electron current profile (green-dashed line) is also shown in c. The example shown here is measured just after saturation with a total X-ray pulse energy of 1.5 mj. The bunch head is to the left in these plots and throughout. To record XTCAV data from the user side, in each machine configuration, MCC needs to perform calibrations and also suppress lasing for the user to record baseline (non-lasing) images. A configuration for the LCLS DAQ system is currently being finalized at which point more detailed documentation on data acquisition procedure and analysis will also be posted. A detailed summary of the XTCAV system is described in a recent publication by Behrens et al. in the journal Nature Communications. Future 7.iv pg. 21

22 3)Efficiency a.linacinfrastructuremissionreadiness The Mission Readiness (MR) program was developed during the past year to address accelerator operational issues such as high maintenance cost, obsolescence, operational and safety risks. It focusses on SLAC s infrastructure that serves the LCLS Accelerator and FEL. The high level goal is to ensure that the accelerator and associated infrastructure is capable to deliver the x-ray beam to the LCLS users and to optimally enable the LCLS science program while ensuring maintenance of the accelerator can be done efficiently. We anticipate completing the proposed plan within an 8-10 year time frame. Priorities were established based on risk and anticipated Return of Investment. The Return of Investment is defined as reduced operational cost of system and the prevention of future exponential cost growth typical for the maintenance of end of life systems. Added benefits are improved performance and new capabilities that are typically associated with modern equipment. To develop the Mission Readiness Program, the critical accelerator infrastructure has been evaluated to identify high risk items. Each item has been assigned a number of risk levels in a variety of categories: program, cost, ES&H, technical and likelihood of failure. A relative ranking of all items was performed by combining all risk categories. Additional ranking was established by evaluating the Return of Investment. Project Mitigated risk FY14 FY15 FY16 FY17 FY18 Klystron Modulator Upgrade Addresses maintainability and performance of klystron modulators, future maintenance program uses significantly less staff. Beam Switch Yard Pump Station Replacement PPS System Upgrades, I/O devices and cables Removes risk of unrepairable failures, reduces required staff for repairs and maintenance, and addresses oil leaks into vacuum system Significant reduction of failure rate, reduced maintenance staff, ensures compatibility with future LCLS expansion. Linac LLRF System Phase Replacement of obsolete, maintenance intensive LLRF system with modern electronics, reduces labor effort for maintenance and repair. Use of commercial equipment to replace mix of SLAC built and obsolete commercial hardware. VMDD CNC Mill 0.32 Replacement of inoperable machine, support of LCLS mission, reduced personnel effort to carry out manufacturing program (benefits mechanical and klystron support program) HPRF Test Stands Current high power RF test stands are at the end of life, appropriate repairs frequently impossible due to unavailable obsolete Future 7.iv pg. 22

23 hardware. Upgraded system allow automation of test procedures, thereby reducing required staff for operation. Brazing furnace rebuild 0.25 Preserves unique SLAC capability for brazing large structures METS Pit Furnace 0.6 Replaces furnace chamber Klystron Solenoid Power Supplies Klystron Modulator PFN Upgrade Mopdern implementation of power supply and controls distribution, eliminates single point of failure Replacement of End of Life components. Linac LLRF System Phase Replacement of obsolete, maintenance intensive LLRF system with modern electronics, reduces labor effort for maintenance and repair. Use of commercial equipment to replace mix of SLAC built and obsolete commercial hardware. PPS Relays and Interlooks Significant reduction of failure rate, reduced maintenance staff, ensures compatibility with future LCLS expansion. BSY Cold Cathode Gauges 1 Replace existing obsolete hardware after pump stations are upgraded. BSY Fast Valves 1 Replacement of obsolete device. Acc. Cooling Water System 3.0 Replacement of many decades old water cooling components that are corroding and deteriorating. Linac Control System Camac replacement 2.3 Replacement of obsolete controls hardware with modern TCA standard equipment. Totals Table 3.a.1: Five Year Mission Readiness Plan to address high and medium risk items. During the past year, we conducted a number of reviews of the proposed plan, both external and internal. Recommendations were implemented as appropriate. All projects were assigned a responsible engineer to develop cost, detailed scope and schedule. We now begun to implement a number of these projects using our well established Accelerator Improvement Project formalism. The total budget to complete all identified projects is approximately M$ 55. In FY2014, we plan to spend about M$ 5 to launch the highest ranking projects. During the following years, this effort will ramp up to Future 7.iv pg. 23

24 ~ M$10 (FY15-17) and subsequently ramp down annually until completion. Table 3.a.1 provides an overview of our current plan for the next 5 years. To fund this effort, priority decisions are being made both in the Accelerator and LCLS Directorate. Both directorates are developing a strategy to reduce the core operational cost by implementing measures to improve overall efficiencies. A significant fraction of realized saving will be made available to conduct the MR projects. No additional funding beyond the overall LCLS operational budget is available for this work. The MR Return of Investment will allow SLAC to fund strategic Research and Development activities that are needed to ensure the success of future light source projects at SLAC. b.commonalityofcontrols/daq The LCLS Photon Controls and Data Systems (PCDS) department provides a common platform for controls, data acquisition, data format and data analysis across all LCLS instruments, with two exceptions. The first exception is the fragmentation of the Python scripts developed by the various instruments to interface to the EPICS slow controls framework. Existing systems originated from a central code base but have diverged over time, with many improvements in different areas that could have been better unified across instruments. This is currently being addressed through an operational improvement project (OIP) to unify the various interactive Python tools, to prevent future duplication of effort when making improvements to such tools. The OIP is on track to complete this effort by the end of the summer shutdown in FY14. The second exception is the fragmentation of analysis tools (psana, pyana, Matlab, CASS, AMI) adopted by the different experiments. LCLS has created a working group, composed of PCDS staff, beamline scientists, and instrument engineers, to help guide the process of making the LCLS data analysis systems more user friendly. One of the key goals of this group is to identify the analysis tools that should be supported in the long term. The working group has proposed that the interactive psana (ipsana) framework become the main development focus of the LCLS data analysis team. Significant effort has already been devoted to expanding the capabilities of the interactive psana framework, up to the point that this tool is already capable of replacing CASS and much of AMI functionality. Two additional OIPs, one to add indexing capability to the raw data files in order to increase data access speed, and one to add parallel processing capabilities in order to optimize CPU usage, will complete the effort of making ipsana the main LCLS analysis tool by the end of FY14. The DAQ system currently provides two user interfaces to manage data acquisition: a Python API, invoked by the Python command line, and a Graphical User Interface. The former is mainly used by XPP, XCS and MEC instruments, and the latter by AMO, SXR and CXI instruments. Both interfaces have their respective advantages and therefore both will continue to be supported by PCDS. c.prioritizationprocess,wbsstructure LCLS has created a process for identifying potential enhancements to LCLS capabilities. In particular the process consists of creating proposals for enhancement projects with defined scope, schedule, and budget; socializing and evaluating the proposals; approving enhancement projects and capturing them within the LCLS Work Breakdown Structure; and tracking the progress of these projects. All enhancements to LCLS capabilities take place within this process. Future 7.iv pg. 24

25 Enhancement proposals are created by LCLS staff (beamline and laser scientists, instrument engineers, detector scientists, PCDS staff, etc.). These proposals are formally presented to LCLS department heads at the Joint Divisions Meeting, where they are discussed and prioritized. Approved projects are entered into the LCLS WBS with a charge number, a CAM and WBS manager, and a project leader, who is responsible for the project. The CAM and the leader work together during the planning, implementation, and completion phases of the project. d.crosstrainingofstaff Technical staff to reduce total size, improve capabilities, and broaden coverage Controls and DAQ to help support experiments Laser scientists to help support experiments Beamline scientists to reduce dependence on specialized support staff o Controls - install and control motors, user equipment, trouble shoot, etc. o Laser - setup and configure basic schemes, trouble shoot, etc. Users to reduce dependence on LCLS staff o Train on basic controls interface o Analysis o Basic mech/vac Future 7.iv pg. 25

26 4)SupportFacilities a.laser NEH The core NEH laser systems are stable and require only routine maintenance and incremental upgrades such as replacing older hardware, e.g. to improve timing jitter performance, and adding new or improved diagnostics. The mid-ir OPA setup on the general use table has been very successful in reducing set-up time and providing consistent output and diagnostics of the mid-ir source. Similar mobile, well-defined configurations will be developed for the optical THz source and for the <10 fs hollow fiber source. Further improvements in THz conversion efficiency by cryogenic cooling of LiNbO3 crystals is being developed in the RLL, and this will be engineered for deployment in the LCLS hutches. FEH An increasing number of experiments in CXI require higher pulse energy and changes in wavelength using OPAs. Currently each of these configurations must be set up from scratch and disassembled after the experiment. To increase operational efficiency, we will develop stable, mobile configurations (along the lines of the mid-ir setup) for a multipass amplifier and an OPA stage. In MEC, a second phase of upgrade to the femtosecond laser system is planned for late FY14. The output of the nanosecond laser system will be used to pump a large aperture Ti:sapphire crystal configured as an additional amplifier for the existing femtosecond laser. This is expected to increase the compressed pulse energy to >8 J in 40 fs, for >200 TW peak power. Such high energy beams are subject to beam quality degradation, and a commercial deformable mirror and wavefront sensor will be used to increase the Strehl ratio of the focused beam on target. The nanosecond laser will still be able to be used to drive shocks in targets. To improve the energy stability and the beam profile, we propose to replace the flashlamp pumped Nd:YLF front end with a diode-pumped amplifier chain. LCLS also will evaluate improvements in the temporal shaping of the fiber seed to provide more flexibility and operational efficiency in setting the temporal shape of the nanosecond pulses. Numerous improvements in the diagnostics and controls of both laser systems are planned or underway. XCS and XPP have similar X-ray focusing and diagnostic capabilities. XPP is one of the two most oversubscribed hutches at LCLS, and there would be a considerable operational advantages to move some of the XPP experiments to XCS to avoid the beam-pipe changeover required to switch between XPP and the FEH hutches. The majority of XPP experiments use optical lasers, so a laser system is needed in XCS to enable this program of offloading XPP experiments. The XCS laser system would require the same basic performance as the oscillator/regen system in CXI, but will need to fit into a smaller footprint due to space constraints. Eventually it is anticipated that this system will be upgraded with a multipass amplifier and all of the wavelength conversion systems used in other hutches. Finally the planned hutch 4.5 MFX instrument is being designed with space reserved for an optical laser system, but to date the requirements for this system are still being defined. Timing The existing optical timing distribution has been very successful in maintaining low jitter and drift for the optical pulses delivered to the LCLS. There are some challenges maintaining the system without direct involvement by collaborators at LBNL, and some components of the system have no readily available spares. Furthermore there is no backup system to switch to in case of a catastrophic failure. To address Future 7.iv pg. 26

27 these issues, the LCLS timing group has developed a timing distribution and synchronization system based on RF distribution rather than fiber. This system is currently in use for the two LCLS photoinjector lasers and is currently undergoing commissioning at MEC and XPP. It is being installed in parallel with the LBNL fiber system for redundancy and cross-checking of timing drift. During FY14-15 this system will be installed in parallel at the rest of the LCLS hutches (CXI, AMO, SXR, XCS in that order). So far the SLAC system has shown improved timing jitter (<50 fs RMS). OpticalSetupLaboratory The existing RLL is being repurposed as a cleanroom for the assembly of beamline components. To partially replace the RLL, a new optical setup laboratory in B950 is nearly complete, and it will have the LCLS-standard oscillator/regen/multipass, identical to those used for hutch operations. This lab will serve the same purpose as as the RLL of providing laser light to test users samples and detectors prior to beamtime. In addition the lab has been configured to allow further engineering and development of the mid-ir source and the engineered version of the optically-pumped THz source. Additional laser lab space in B750 is being refurbished to provide space for R&D efforts on optical sources, such as the cryo-cooled THz source, <10 fs sources, short-pulse UV sources, HHG sources, etc. b.samplepreparation Our level of understanding of sample injectors is now advanced enough that we may communicate our specific needs to potential manufacturers. Standard etched microfluidic device construction, hard lithography, 3D printing and micro injection molding are all actively being pursued with the expectation of commercial availability by at least one of these methods in a 1 to 2 year time frame. Submicron drop aerosolizers can also be produced in-house for use in single particle imaging, SPI, experiments. However, if LCLS is to eventually be able to image single molecules, a new type of injector that produces even smaller droplets, under 100nm will be required. A FEL simulator station consisting of an optical laser and hit rate counter, will be added to the ICL by summer 2015 for off-line testing of liquid jets and gas phase injection. In order for users to gain experience with the same sample delivery system as is used at the beamlines, a second sample delivery system will be installed in the ICL in the near term. Scheduled remote access will allow users to operate the ICL sample delivery system from their home institutions. In addition to the ICL, the Sample Characterization Lab, SCL, will also assist users with off-line preparation. The SCL is still in the planning phase. The location, general layout and equipment needs have been agreed upon; architectural drawings will be the next step. When completed, the SCL will house the equipment necessary to characterize a sample and evaluate that sample s potential to provide useful data at the FEL. The Soft X-ray Department, SXD, and the in-air station hutch 4.5, will present the highest demand for new sample delivery equipment in the near to mid-term and FEL upgrades and expansions will continue that demand in the long term. As an example, all six endstations presently need some type of liquid target injector; three have similar or shared LCLS equipment available, MEC is pursuing one through collaboration, SXR has one available through collaboration with an outside group, and AMO does not yet have any liquid injector. We will gain efficiency by bringing equipment in-house and integrating it with a standard set of hardware and controls for similar tasks. LCLS II will bring new challenges for sample delivery -high speed pulsed injectors will be desired to match the high X-ray rep rate. This is currently in development as a collaboration with Stanford University. Pulsed sample delivery will also be the enabling step for some types of experiments to be run remotely. Pulsed sources will use far less sample and will allow users to screen very many submicroliter Future 7.iv pg. 27

28 sample volumes to find optimal conditions for extended runs to collect full data sets. Both screening and regular beamtime may then require little on-site handling and could possibly be run remotely from the user s home institution. A similar remote mode of operation is employed at synchrotron protein crystallography beamlines and eventually will be the norm for most structural protein crystallography runs at LCLS. c.x-rayopticsmetrologylaboratory LCLS requirse state of the art optics to deliver a variety of photon beam capabilities while preserving the characteristic properties of the XFEL. In order to realize the necessary upgrades to the existing optic systems as well as develop the instrumentation capabilities for the LCLS-II source, LCLS will build an optical metrology lab in the clean room facilities currently under construction in the NEH. The lab will start with the basic instrumentation to properly characterize and install the upgrade of the Hard X-Ray Offset Mirrors (HOMS) and eventually contain the full suite of tools to measure across the full PSD. d.otherlaboratories Sample prep/wet lab in 757 FEH mezzanine sample prep/wet lab Set - up space in FEH ControlsTrainingLab In order to increase the capabilities and independence of users, LCLS will build an offline controls training center. The center will contain a full control system infrastructure and hutch interface that will provide users the opportunity to familiarize themselves with controls and DAQ interfaces as well as setup and operate common functions. Cleanroom In order to develop and support the broad suite instrumentation at the LCLS, large clean assembly space with a range of specification is required. Future 7.iv pg. 28

29 5)ScienceInitiatives The MFX instrument described in section in Section 1 is the result of a collaborative science initiative that was created at SLAC with the formation of the Bio-Imaging Working Group headed by Soichi Wakatsuki. Besides members from LCLS, the Bio-Imaging Working Group includes members from the Structural Molecular Biology (SMB) Division at SSRL, the SLAC Photon Science Directorate and Stanford University. In the following we list two other science initiatives that will result in new instrumentation and/or science programs at LCLS. a.roadmaptosingleparticleimaging Since its inception, the LCLS has been recognized as an x-ray source with the potential to support transformative techniques for structure determination. The goal of imaging single particles has long been held as one of the most important such advances. To date, a number of impressive studies have been conducted and LCLS now seeks to accelerate the process of refining this technique and supporting its routine application by a broad user community interested in applications ranging from materials science to biology. In March 2014, the LCLS and SLAC hosted a workshop that brought a wide cross-section of the user community together with local experts to determine a suitable path toward the fulfillment of this goal. With the support of the LCLS user community, the facility will pursue a program of open-access, precompetitive aimed at the development of instrumentation, algorithms, and software that will enable advanced imaging experiments to occur more rapidly and reliably. We anticipate that this program will last between three and five years, meeting technical milestones that will be selected in conjunction with input from the user community to support key user science cases. The program will begin during the LCLS s 10 th user experiment run with input from the user community provided during the 2014 LCLS Users Meeting. b.matterunderextremeconditionsandhighenergydensityscience The Matter under Extreme Conditions (MEC) Instrument receives funding from both Fusion Energy Sciences and Basic Energy Sciences. The MEC short pulse laser system is undergoing an upgrade from 4 to 25 TW in phase 1 and then to 200 TW in phase 2. At these higher pulse energies, investigations of high intensity laser-matter interactions will become feasible. This project includes the construction of three multipass amplifiers and cross-polarized wave generation (XPW) as well as the procurement of a new pump laser and deformable mirror. In addition, radiation safety activities are required because the laser system will be able to generate high energy electrons and x-rays. A radiation interlock system will exclude personnel from the hutch when high intensity laser beams are present in the target chamber. In addition, there will be measurements of the radiation inside and outside the MEC hutch first at and then at W/cm 2. In the present status, the phase one upgrade has been constructed and is now under commissioning. A formal program for optical-laser-only beamtime has started. An announcement was made at the SLAC High Power Laser Workshop, October 1 and 2, After the first call for proposals, 6 proposals were received. The proposals are reviewed by a subset of the MEC proposal review panel (PRP). During the LCLS accelerator shutdown from early August to mid-october, 2014, three optical-laser-only beamtimes will take place. It is anticipated that there will be a call for optical-laser-only proposals once a year. The optical-laser-only experiments will occur during each LCLS accelerator shutdown and potentially during Future 7.iv pg. 29

30 the LCLS runs. This program will be especially valuable during the extended shutdowns necessary for LCLS-II installation. The highest pressure achievable at MEC is limited by the pulse energies provided by the nanosecond laser system. These drive laser pulses result in a pressure of about 1.5 Mbar. In principle, the optical beams can be focused tightly, but members of the MEC PRP have been skeptical about whether the shocks under these conditions are sufficiently uniform. A project will be initiated to characterize and improve the nanosecond laser foci and the uniformity of the shock waves generated by these laser beams. Measurements will be made of the wavefront errors in the nanosecond laser beams, and corrections made to the optical wavefronts. In addition, Velocity Interferometer System for Any Reflector (VISAR) observations will be made of the shock waves produced by the different drive laser conditions. Phase contrast imaging has the potential to provide sub-micron resolution imaging of materials under dynamic compression. In collaboration with the University of Dresden, an optimized phase contrast imaging setup will be constructed. The MEC department is contributing engineering effort as well as procuring a detector for this project. For two years, this new assembly for phase contrast imaging will stay at MEC instrument. After that, the components paid for by the University of Dresden will be shipped to the European X-FEL. Replacement components will then be fabricated for the MEC Instrument. While the 200 TW upgrade to the MEC short pulse laser is currently underway, the MEC science community has already begun a push for even higher peak intensities. This idea was presented and received strong support at the High Power Laser workshop at the 2013 LCLS Users Meeting. As a result, a White Paper was developed and presented to FES that discussed the scientific possibilities and some of the possible scenarios for getting PW peak power laser pulses and the LCLS X-ray pulses onto the same target. Because PW lasers are large and capable of generating significant amounts of hazardous radiation when interacting with a target, the two biggest issues to resolve when considering the PW upgrade are: where is there space for the laser and how can people be shielded from the radiation hazard. There are numerous possible solutions to these two problems, but all require either new or modified conventional facilities. LCLS is currently planning an engineering study to better determine the cost and schedule implications of the various possible solutions during FY15. Further workshops will also be held to better define the scope of such a project, including a determination of which type of PW laser system, high energy (>250 J, <250 fs) or short pulse (>40J, <40fs), would provide the greatest scientific reach. Future 7.iv pg. 30

31 6)RoadmaptoLCLS-IIOperations The LCLS-II project will include a new superconducting (SC) linac capable of producing intense electron pulses at up to 1 MHz repetition rate and energies up to 4 GeV. This SC linac will feed a variable gap soft X-ray undulator producing high-rep-rate X-ray pulses in the kev range. Also included in the project is a new variable gap tender/hard X-ray undulator that will replace the existing LCLS undulator. When driven by the SC linac, it will serve an energy range of 1-5 kev at high rep rate, and when driven by our existing 120 Hz Cu linac it can reach photon energies up to ~25 kev. The six LCLS instruments will undergo modifications and enhancements to operate with the enhanced beam properties delivered by the LCLS-II project. The definition and execution of the changes to the LCLS scientific instruments will follow a three step process: 1. Definition of the instrumentation needs 2. Conducting strategic research and development activities to enable the technologies for instrumentation 3. Construction and implementation of enhanced instrumentation The following sections outline the tactics associates with these steps. a.definitionoftheinstrumentationneedsforlcls-iioperations LCLS-II will expand the scientific capabilities of the LCLS facility with higher repetitions rates and increased photon energy. This will both better serve the current scientific community and engage new users at LCLS. Alignment between the community s needs and the facility s instrumentation is critical for enabling the best science. This alignment requires dialogue. A weeklong series of workshops will take place September 22-26, 2014 to introduce the possibilities of LCLS-II to the scientific community and define instrumentation development priorities. Each daylong workshop will be co-organized by members of specific scientific communities and an associated LCLS staff scientist. The workshops include: Dynamics of atomic and molecular systems excited by both optical and x-ray radiation. Chemical dynamics and reactivity on surfaces and in solution phase systems. Dynamics of electronic structure of materials with element specificity and spin and orbital sensitivity High energy x-rays for probing bulk materials and short range structure Biology using both scattering and spectroscopic probes to study the structure of biological systems Non-linear interactions of x-rays with matter enabled by ultrahigh intensity x-ray pulses from LCLS-II b.strategicr&dinsupportoflcls-iioperations We have identified four areas of strategic R&D in support of LCLS-II operations: x-ray detectors, x-ray data systems, optical lasers and superconducting undulator technology. The specific details of these R&D initiatives will be honed following the LCLS-II instrumentation workshops. Nevertheless, generic tactics of these areas can be formulated. The following initiatives will be conducted over the next 3 years: Future 7.iv pg. 31

32 X-raydetectordevelopments The properties of the new LCLS-II source, in particular the high repetition rate (up to 1 MHz) and extended photon energy range (250 ev - 25 kev), will provide a means to address important scientific questions in a broad range of disciplines. Experimenters will use various x-ray scattering and spectroscopic techniques and this in turn will drive a wide range of detector requirements. When one compounds to this fact the need to span two orders of magnitude in photon energy, it becomes evident that a suite of detectors is needed. The following goals will serve as the criteria driving the development of this detector suite: Meet the specifications defined by the scientific needs Minimize development costs of the detector suite Minimize the resources and costs to operate and maintain the detector suite to ensure a sustainable operations model We envision the suite to include two flavors of detectors (spectroscopy, imaging) for each of three photon energy regimes [soft ( ev), tender/hard (1 15 kev) and very hard (15-25 kev)]. This leads to a total of 6 tailored detector systems. Spectroscopicdetectors: The key characteristics of spectroscopic detectors is very low-noise performance, good quantum efficiency and high spatial resolution. X-ray spectrometers are typically operative in a dispersive geometry, mapping photon energy onto a spatial dimension. For this reason, there is a coupling of the detector spatial resolution and the achieved energy resolution. In addition, the detected x-ray patterns are weak and detecting single photons that may be shared across multiple pixels is crucial. These requirements are particularly challenging in the soft x-ray regime. Imagingdetectors In contrast to spectroscopic detectors, imaging detectors are intended to resolve patterns with a wide range of intensity distribution, which can include very intense and very weak Bragg peaks. This class of detectors is characterized by large pixel well depths, modest spatial resolution and low noise. Imaging detectors are large in size, i.e. many megapixels, to sample a large range of reciprocal space. All of the abovementioned detectors can operate in two modes: accumulating and pulse-by-pulse. In the first category the experimental parameters do not change over the accumulation time that is long compared to the pulse-to-pulse time. These include: timing jitter between pump and probe, sample state, FEL intensity, etc. Therefore detectors sought to work in accumulating mode can be read out at a speed slower than the machine repetition rate but have to provide stable performance over the accumulation time and large dynamic range. In the second category the experimental parameters change on a pulse-bypulse basis requiring detectors with a readout speed matched to the machine repetition rate and fast feedback from beam diagnostic for data sorting and binning. These detectors should implement the possibility to work on Region of Interest (ROI) mode as well as triggering and veto mechanisms. In order to minimize development costs and ensure efficient operations of the detector suite, a modular approach will be used to maximize the commonality between systems. The detector systems will be divided into 3 sub-components: sensor and front-end electronics, readout electronics, and communications firmware. We will develop a common framework that standardizes the readout Future 7.iv pg. 32

33 electronics and communications firmware. This approach will minimize development costs and ensure efficiency in the integration and operation. The sensors and front-end electronics will be tailored for each application. In some cases, different technologies will be explored. For example, CCD and CMOS technology for soft x-rays, hybrid pixel arrays for tender/hard x-rays and high-z material based hybrid pixel arrays for very hard x-rays. Partner laboratories will be pursued for the development of various sensors and front-end electronics. These partnerships will be explored and defined over the next 18 months. X-raydatasystems The high repetition rate (1-MHz) and, above all, the potentially very high data throughput (100GB/s) generated by LCLS-II requires sophisticated data acquisition and storage system and increased data processing and data management capabilities. The main challenge will be developing high density, highthroughput, peta-scale storage systems that allow concurrent access from thousands of user submitted jobs. Additional critical capabilities include the deployment of a trigger/veto system, upgrading the SLAC network connection to the DOE Energy Sciences Network (ESNet), and expanding bandwidth and capacity of the long-term tape archive. DataAcquisition Two main modifications to the current system will be required for operating at high repetition rates: moving the event builder from online to offline and adding the ability to aggregate contributions from multiple events in the readout nodes. These changes are required for running at 1-kHz or above, independent of throughput. Changes necessitated by increase in data throughput are: network link upgrades (from 10Gb Ethernet to Infiniband or to 40Gb Ethernet) and online cache upgrades. Both changes are discussed later in the storage and network sections. The deployment of a trigger/veto system for LCLS-II may be required. While the benefit of vetoing events based on the event data is potentially very large (factors are common for single particle imaging and nanocrystallography experiments), it will be challenging to implement an effective trigger/veto system. Vetoing based on beam parameters is not effective, as most x-ray pulses are of good quality. Setting veto parameters based on the event quality for a given experiment is difficult because users themselves often do not know which online feedback parameters might determine event quality, and are very suspicious of anything that could filter their data on-the-fly. In addition, a veto system is useless for experiments with hit rates close to 100%, which may very well represent the majority of future LCLS- II experiments. DataStorage The primary limitation for high-throughput experiments at LCLS-II will be storage technology: the existing technology is too slow, and even a simple analysis task for one experiment would take weeks to complete. Spindle-based systems will become cheaper and denser, but not much faster or easier to manage. Solid state systems will also become cheaper, but the current trend for commercial systems is to optimize IOPS instead throughput and scalability, which are the two key aspects for a system hosting LCLS-II science data. In addition, current commercial systems come with a significant premium on the cost of the flash memory, making a multi-petabyte system prohibitively expensive. The SLAC tape system is approaching its limits in both overall storage capacity (~20+PB) and throughput. We are already Future 7.iv pg. 33

34 seeing limitations when handling data from ongoing experiments and concurrent users requests to restore files from tape. DataNetwork The secondary limitation is network technology. LCLS would greatly benefit from having a 100Gbps connection between SLAC and ESNet. The primary usage for such a link would be the ability to offload part of the LCLS science data processing to some of the large DOE computing centers. For example, current LCLS data acquisition rates are up to 5 times the 10Gb link between SLAC and the National Energy Research Scientific Computing Center at LBL (NERSC). It would take weeks before users are able to analyze, at NERSC, data they acquired at LCLS. With regard to the local network, Infiniband would be superior to Ethernet for building a high throughput network for LCLS-II, especially under high congestion conditions, but Infiniband has limitations due to its short maximum distance and its cost. It is more expensive than Ethernet to connect devices that are not within 10m of each other, and significantly more expensive to connect devices that are separated by more than 300m. DataManagement LCLS has developed a powerful data management system that handles both the automatic workflows of the data from the online cache to the various storage layers, and the users requests (e.g. restoring data from tape) through a web portal. Some aspects of the current system, such as checksum calculations, HPSS interface, and lack of prioritization, will become bottlenecked at higher data volumes and will need to be upgraded. DataProcessing The increase of data processing capabilities doesn't present any particular challenge on the hardware side, even when considering the deployment of hybrid processing solutions (e.g. GPUs). The addition of parallel capabilities to the analysis frameworks is already being developed for LCLS-I. DataFormat LCLS DAQ is currently writing the raw data as XTC and users can request that their data be translated to HDF5. The translation step will become bottlenecked in the future and LCLS-II should target only one data format for standardization. HDF5 is becoming the de-facto standard for storing science data at light source facilities, but, in order to effectively replace XTC in LCLS, a couple of critical format features are required. These features, namely the ability to read while writing and the ability to consolidate multiple writers into a consistent virtual data set are currently missing in HDF5, however, the HDF group claims that these features could be added, given enough development resources. Summary This is a summary of the critical projects required to build a data system able to handle the LCLS-II requirements: 1. Move the event builder from the online to the offline and introduce the ability to aggregate contributions from multiple events in the readout nodes. Both changes are software-only, are incremental from what is currently in operation, and are required for running at 1kHz or above, independently of throughput. The online monitoring framework (AMI) and the offline framework (psana) will need to be adapted to the new paradigm. 2. Develop a custom, solid state, online cache (DAQ recorders) and offline storage (users data analysis) to solve the storage challenge. No vendor is currently offering high density, petascale, Future 7.iv pg. 34

35 solid state storage solutions. SLAC has previously worked on peta scale flash-based systems and we believe it's possible to build scalable, peta scale, solid state storage by aggregating commercial off-the-shelf components. This project presents medium risk with significant R&D and development effort. The same technology could be used to build custom recorders for both the online cache and the fast feedback system. The prototype of this project would be an inexpensive, high density (>100TB per rack unit), fast (10GB/s per 100TB), flash-based storage chassis. 3. Upgrade the LCLS local area network. The current system uses 10Gb Ethernet from the readout nodes to the online cache and Infiniband (IB) from the online cache to the offline and within the offline system. Introducing IB or 40Gb Ethernet from the readout nodes to the online cache would be investigated. The final solution will be based on actual space constraints. 4. Add data management capabilities at high throughput. This includes upgrading the tape system, since we already see limitations when handling data from ongoing experiments and concurrent user requests to restore files from tape. The data management framework should also be upgraded, because the current checksum calculations, HPSS interface and lack of prioritization are destined to become bottlenecked at higher data volumes. 5. Increase data processing capabilities. This includes improving the parallel capabilities of the analysis frameworks, introducing a different mechanism for generating HDF5 files, and deploying a hybrid processing solution (e.g. GPUs or many-core solutions). 6. Develop a new timing system. The existing EVG/EVR system was not designed for MHz operations and a new system will be required. Competency exists for this effort at SLAC and due to the flexibility required over time, especially on the experiment side, this system should be provided by SLAC, ideally through an RED/ICD collaboration. 7. Investigate a veto system. A veto signal could be delivered to the front-end electronics (EuXFEL approach), to the readout nodes, to the online cache or in the fast feedback layer. In general, a veto in the front-end electronics reduces the throughput requirements on the DAQ components, while a veto in the deeper layers provides cheaper/larger buffers and more time to reach a decision. 8. Investigate replacing POSIX-compliant file systems (currently Lustre and Gluster). This requires exploring easier to maintain, easier to scale up, faster technologies. Big Data industry has been moving to distributed storage models. Large particle physics experiments have moved to XROOTD but that may not be the right solution for LCLS, because of the lack of strong security features and the inability to operate with HDF5. The following projects were implemented by partnerships between SLAC and other institutions: 1. HDF5 upgrade: Some critical features are missing from the HDF5 API. The HDF group believes these features are useful in general, not just for LCLS, and that they could, and should, be added to the API. The group needs additional resources to work on these features. 2. Offline computing: Data centers built towards data intensive systems could help offload the LCLS/SLAC offline computing system. General support for LCLS offline analysis would require, at a minimum, ~50 PB tape storage, a dedicated ~10 PB of disk storage and ~100 teraflop processing farm with an aggregate throughput to the storage above 10 GB/s per PB. Other key requirements would be the ability for LCLS users to manage their data through the LCLS tools and workflows, and the ability to use their SLAC account (or a federated account). 3. ESNet link upgrade: The ability to offload computing capabilities assumes an upgrade to a faster 100Gbps connection between SLAC and ESNet. Future 7.iv pg. 35

36 OpticalLaserSystems Significant needs for laser R&D will be driven by the requirements of LCLS-II, principally the development of MHz, kw-class ultrafast lasers for pump-probe experiments and relativistic beam conditioning. Such laser sources would have immediate application in laboratory-based experiments in material and chemical science, and enable technology for other areas of DOE mission interests such as accelerator stewardship. Furthermore, kw-class lasers would drive fundamental and applied research relevant to the Stanford faculty, in areas such as the production of unique micro- and nano-structured materials, high precision metrology, and XUV photonics, enabling SLAC to expand its scope of scientific research through new and strengthened collaboration with campus groups. The existing LCLS has required significant laser engineering to maintain its high level of uptime and reliability, expert scientific support of complex experiments, and innovative development efforts to provide unique wavelengths and delivery systems. LCLS-II will require these same capabilities along with the considerably challenging requirement of operation at high repetition rate, up to 1MHz for the photoinjector and 100kHz to 1MHz for the experimental beamlines. To fully realize the potential of LCLS- II, significant R&D will be required to develop appropriate laser sources. The two initial laser development objectives for LCLS-II are for the photoinjector lasers systems and for pump-probe experimental laser systems. The key features of the photoinejctor laser will be 1 MHz rep-rate with variable rep-rate capability, >50W average power in the IR to produce >5 W in the UV, with a flat top spatial profile, flat top temporal profiles with varying duration over 10s of picseconds, and variable timing of individual pulses for beam spreader, >98% uptime. No commercially available lasers exist that can meet all of the requirements of the LCLS-II photoinjector, and the LS&T group will either develop an appropriate system at SLAC or work with a commercial laser company to develop lasers systems to our specifications. Operating at 1 MHz with ~50 J pulse energy, this laser system also would have immediate applications in photoemission spectroscopy applications in chemical and material science as well as applications for SSRL pump/probe experiments. The majority of experiments at LCLS involve a millijoule level pump/probe laser. While most experiments do not need such high energy pulses directly on sample, many do require nonlinear conversion to wavelengths that cover the range from VUV through THz. These conversion processes are inherently inefficient and require mj level pulses to generate nj to J level pulses at the sample as required by most experiments. To achieve comparable capabilities to realize the full potential of LCLS-II, R&D is required to develop lasers capable of providing mj level, 10 fs pulses at up to 100 khz for pump/probe experiments. Similar efforts are underway at DESY for pump/probe lasers for FLASH-II and the European XFEL, and we will leverage these developments in our efforts for LCLS-II. A 100 khz, kw-class laser with fs laser pulses represents a two orders-of-magnitude increase in the average power of existing commercial ultrafast laser systems. Such a laser would become a key asset for SLAC, both for conducting further R&D for LCLS-II, e.g. through the development of wavelength extensions (UV, THz, mid-ir, etc.) for this system, and by enabling numerous other research programs in PSD, AD, and other programs on Stanford campus. Just as with the conventional ultrafast lasers used in LCLS, access to this new high-rep-rate kw laser will benefit users and beamline scientists by enabling them to test samples, detectors, and optical techniques in preparation for LCLS-II beamtimes. Future 7.iv pg. 36

37 SuperconductingUndulatorTechnology Undulators serve as the primary source of radiation for modern storage rings (SR) and Free-Electron Laser (FEL) light sources. The advent of permanent magnet undulators (PMU) in the early 1980 s ushered in an era of 3 rd generation light sources, and since then improvements in undulator technology have served to continually enhance the flux and brightness provided to experiments by SR and FEL beamlines. A comparison of different undulator technologies clearly shows superconducting undulator (SCU) technologies, with much higher magnetic fields, as the leading candidates for future FEL-based x- ray sources (see Fig. 6.b.3). In addition, SCUs have no permanent magnet material, and therefore are orders of magnitude less sensitive to radiation dose, which is a serious risk for PMUs using a high-power electron beam. In addition, with the very high peak fields available, SCUs will offer more compact and more efficient FEL designs in the future, clearly surpassing cryogenic permanent magnet undulators (CPMUs) and in-vacuum devices with similar beam stay-clear dimensions. This project is in collaboration with SLAC, ANL, and LBNL, and will demonstrate the viability of superconducting undulator (SCU) technology for FELs, by building, measuring, testing, and correcting two 1.5-m long prototype SCU s using two different technologies. The first technology is based on conductors of NbTi, the chosen technology at ANL, and the second is Nb3Sn, the choice at LBNL. The Nb3Sn conductor holds promise for the highest fields, but high-temperature processing of the conductor adds some technical risk. We will conduct independent cryogenic tests on each of the two undulators using the same 2-m long cryostat, designed and built at ANL and based on an article installed at the APS, verifying the magnetic performance of each undulator using the same cryostat and magnet measurement bench. The FEL-specific undulator parameters, such as the magnetic field strength for the chosen period and gap, and the magnetic field quality, field integrals, and FEL phasing over the undulator, must be demonstrated. Fig. 6.b.3: Peak magnetic fields vs. undulator period length for four different technologies (blue-solid: PM-NdFeB, cyan-dotted: PM-NdFeB-in-vacuum, green-dashed: SCU-NbTi, red-dash-dotted: SCU-Nb 3 Sn). All have the same 5.7-mm vacuum chamber gap for comparison and a 7.5-mm magnetic gap, except the in-vacuum gap at 5.7 mm. Future 7.iv pg. 37

38 A magnetic tuning technique will be developed and applied to each undulator as needed. The project design and execution will rely heavily on already well-established designs of the undulator cryostat, undulator magnet core, and magnetic measurement systems developed and implemented for shorter undulators at the APS. It will also employ a unique magnetic tuning approach developed at LBNL. The APS SCU magnetic measurement system incorporates Hall-probe and rotating coils sensors modified for the smaller vacuum pipe diameter; the system will furthermore incorporate a pulsed-wire technique developed at LBNL. The project deliverable, in 1.5 years (i.e., July 2015) from project start, will be two fully functional superconducting undulators that meet LCLS-II Hard X-Ray FEL undulator specifications and demonstrate this enabling technology for long-term use in future light source projects, and possibly as a replacement for the LCLS-II PMU baseline design, if time allows. Prototype undulator parameters are listed in Table 1, showing the slightly different period for each technology, based on the LCLS-II FEL parameters. Parameter Symbol NbTi Nb 3 Sn Unit Undulator magnetic full gap g m 8.0 mm Vacuum chamber full gap (internal) g vc 5.7 mm Prototype undulator segment length L seg 1.5 m Undulator period u mm Peak magnetic field B max T Table 6.b.1: SCU prototype parameters for each technology. c.constructionandimplementationofinstrumentationinsupportoflcls-iioperations The LCLS instrumentation suite will undergo enhancements to make optimal use of the new sources realized with the LCLS-II project. The specific details of these enhancements will be defined in the series of workshops described above. However, generic needs in the following areas are identified already: X-ray optics and safety systems: enhanced to handle high average power. Optical lasers: high repetition rate systems. X-ray detectors: high repetition rate and accumulating systems. Data systems: increased throughput to accommodate high repetition rate detectors. Controls systems: updated triggering system for high repetition rate operation. As detailed previously, some of these areas require research and development before construction and implementation of the respective instrumentation can occur. The activities should conclude in the 2-3 year time frame. Subsequent to this, LCLS will pursue an aggressive implementation of these technologies. This implementation is planned for the FY16-FY20 time frame as shown in Table 6.c.1. Future 7.iv pg. 38

39 Table 6.c.1: Implementation plan for improvement programs for LCLS-II operations. Future 7.iv pg. 39