Berkeley Lab

ATAP Newsletter, February 2015

Click here if you’d like to receive our bimonthly newsletter.


Leemans_Wim_Headshot_2014_150_captionedThe past week brought exciting news to ATAP: the DOE funded two areas of our research through their accelerator stewardship program. One grant will partner us with Paul Scherrer Institute and Varian Medical Technologies to apply our superconducting magnet expertise to beam delivery for cancer therapy. Supported by the other grant, Lawrence Livermore National Laboratory and the University of Michigan will work with us in an ambitious project to increase the performance of ultrafast fiber lasers a thousandfold. To learn more, read “Grants Give Particle Accelerator Technologies a Boost” in the February 17 issue of LBNL’s newsletter, Today at Berkeley Lab.

We invite you to read on for other recent ATAP achievements in science and service.

  • The Advanced Photoinjector Experiment (APEX), important to LCLS-II and other next-generation facilities for photon science, is running again with much-improved parameters.
  • Three articles show how our Superconducting Magnet Program not only contributes to the future of high-energy physics, most recently with high-field quadrupole magnets for an LHC upgrade, but is also leveraging its expertise on behalf of other DOE priorities: prototyping a superconducting undulator for FEL-based light sources, and a complex and challenging magnet for the ion source of the Facility for Rare Isotope Beams, a flagship nuclear-science facility being built by Michigan State University.
  • NDCX-II, an LBNL facility for fusion energy science and high-energy-density physics, has demonstrated the final key piece of its suite of capabilities: neutralized drift compression.

Finally, as part of our commitment to safety, health, and environmental protection as the foundation of everything we do, on February 23 we will take our annual Safety Day to review both physical and procedural aspects of EH&S throughout our Division.


APEX is Back and Better than Ever

The previous issue of this newsletter described how the Advanced Photoinjector EXperiment (APEX) had passed readiness reviews and was returning to operation after a failure had caused a shutdown. Now APEX is not only back in operation, but better than ever.

APEX is dedicated to the development and test of a new concept for a high-repetition rate, high-brightness electron injector, optimized to operate at the performance required by an x-ray FEL user facility. The VHF gun of APEX became the baseline for the injector of LCLS-II; as one of its major contributions to that facility, LBNL is responsible for the design, construction and commissioning of the LCLS-II injector.

Success will dramatically benefit not only the performance of future generations of FELs in which high repetition rates (> 10 kHz) are required, but also high- repetition-rate ultrafast electron diffraction (UED), a structural dynamics probe, as well as ultrafast electron microscopy (UEM), and inverse Compton scattering (ICS) applications. APEX was initially funded by LBNL Laboratory-Directed R&D funds and then supported as an Office of Basic Energy Sciences research program.

The core of the system is the VHF Gun. Based on a new concept developed at LBNL, it is a normal-conducting, continuous wave (CW) RF gun in which the electrons are generated by laser-induced photoemission on high-quantum-efficiency photocathodes, then accelerated by the cavity fields (~20 MV/m) to more than 750 keV.

APEX_polishing_250 While APEX was recovering from an RF distribution failure, we took the opportunity to upgrade the VHF Gun in ways that provided a cascade of benefits. It now operates at lower rf power and thus lower temperature; and also lower pressure (better vacuum), which improves cathode lifetime. Undesirable “dark current” was vastly improved by cleaning and state-of-the-art polishing. These improvements will directly benefit the Linac Coherent Light Source II project at SLAC National Accelerator Laboratory, as the VHF Gun will be the design basis of the LBNL-provided injector for LCLS-II. Logo of Advanced Photoinjector Experiment at LBNL


APEX recently recovered from a serious failure in its RF distribution system that forced the experiment to shut down for about five months, impacting both the project cost and schedule. However, the failure opened up several opportunities to improve the performance of APEX — particularly the VHF Gun. First, we identified and corrected the weakness in the RF distribution system design that caused the failure. Further, opening the VHF Gun cavity (which was designed to be opened) let us improve relevant characteristics of the gun — actions that resulted in a cascade of benefits.

The VHF Gun cavity is composed of two main separate copper parts (the “anode” and the “cathode” halves). RF continuity between the parts is ensured by the mechanical pressure exerted on them by the external stainless steel shell of the gun. The contact surfaces between the two halves were re-machined in the LBNL shops to improve their flatness and hence their mutual contact. The operation was successful and resulted in an increased cavity quality factor Q of ~15%.

Because of the higher Q, the power required to obtain the nominal accelerating fields inside the cavity was reduced by the same amount, reducing the gun operation temperature from about 100° C to about 70° C.

Lowering the temperature allowed the operating vacuum to be improved from about 0.8 nTorr to about 0.3 nTorr. The decreased pressure has the benefit of lengthening the lifetime of the reactive semiconductor cathodes, one of the important parameters for a gun operating in a high-repetition- rate application.

Opening the gun cavity also allowed reduction of the “dark current” (emission when the gun should be off). Dark current consists of undesired electrons that the high electric fields in the cavity generate through field emission. Minimizing it is important because dark current can propagate downstream; there, especially in CW machines, it can trigger quenching in superconducting structures, degradation of the magnetic performance in permanent magnet undulators, and undesired radiation doses that can activate and damage components.

Although before the shutdown, APEX’s 350 nanoamperes of dark current already met the LCLS-II requirement (<400 nA), we greatly desired to decrease it even further. Opening the VHF Gun allowed us to use a cleaning technique based on solid CO2 (dry ice) that had been shown to be very effective in removing particulates, large clusters of molecules that can stick to the cavity walls and serve as a source of dark current. Additionally, we used a state of the art technique to polish the anode and cathode areas to a mirror finish, removing surface irregularities that, like particulates, could locally enhance the electric field and induce field emission. The use of such techniques turned out to be extremely effective, reducing dark current at the nominal operation power by more than 3 orders of magnitude: from 350 nA down to 0.1 nA.

The next important goal for APEX is the completion of Phase II of the project, where a small linac will be added to accelerate the beam to several tens of MeV. This energy level will make space-charge forces negligible, and allow reliable measurement of the beam brightness, which is the ultimate beam quality parameter and the gun’s last remaining milestone for LCLS-II.

The APEX team, February 17, 2015. Back row, l-r: Houjun Qian, Rick Lellinger, Eric Norum, Daniele Filippetto, Matt Johnson, Toby Kramasz, Greg Portmann, Daniela Leitner, Fernando Sannibale, Carl Cork, and Michael Dickinson. Central row, l-r: Ken Baptiste, John Staples, Greg Harris. Front row, l-r: Jennifer Doyle, Ruixuan Huang. Not pictured: Mike Chin, John Corlett, Stefano De Santis, Larry Doolittle, Gang Huang, Slawomir Kwiatkowski, Vladimir Moroz, Max Vinco, Steve Virostek, Russell Wells, Max Zolotorev.


A High Gradient Toward the LHC’s Future

Modern accelerators are not built so much as rebuilt: ingeniously upgraded in large and small ways as new ideas and technologies come along, thus maximizing the scientific returns on the original investment. The Large Hadron Collider at CERN had not even begun operation when people began thinking about how to improve it. The first major upgrade, long envisioned and now an official project called the “Hilumi-LHC”, will greatly increase the machine’s luminosity — a measure of the possible collision rate at the points where the two beams interact, and thus of the scientific productivity.

The Hilumi-LHC requires, among other improvements, stronger quadrupole (focusing) magnets to reduce the beam size in the interaction regions. This work fell to Brookhaven National Laboratory, Fermilab, and LBNL as part of “LARP,” the LHC Accelerator Research Program. It is one of several roles that LBNL has played in the LHC and its upgrades over many years.

The High-gradient Quadrupole (HQ) magnets for the Hilumi-LHC, designed and built in part by ATAP’s Superconducting Magnet Program (SMP) as part of a multi-institutional collaboration, have demonstrated a magnetic-field gradient of 192 teslas per meter (T/m) in a 120 mm bore — more than strong enough for the upgrade. The magnets, designated HQ02a and HQ02b and tested at Fermilab and CERN, respectively, are made with niobium-tin (Nb3Sn) superconductor rather than the niobium-titanium (NbTi) used in the present interaction-region quadrupole magnets.

Nb3Sn has much higher field-strength limits, but because the heat treatment needed to bring out its superconducting potential also renders it brittle, new ways of assembling the magnets had to be found. SMP has been a leader in making practical accelerator-style magnets from this material. The Conductor Development Program, which SMP manages on behalf of DOE, helped develop the Nb3Sn wire used in the magnets. Involved with every aspect of the HQ magnets, SMP fabricated the superconducting cable; designed and built the mechanical support structure, using concepts developed at LBNL; and played key roles in coil winding and quench protection. To aid in understand magnet performance and the results, LBNL developed advanced diagnostics for assessing field quality and quench behavior.

The coils are assembled and pre-loaded to a low stress in the shell at room temperature by using pressurized water filled stainless steel bladders. Then the aluminum shell of the LBNL-developed mechanical support structure places further load upon the coils during cooldown to 4 kelvin. This design approach was demonstrated to be a superior loading option for Nb3Sn coils, keeping them from being overstressed during assembly and cooldown as might have occurred with other support structures under consideration. Therefore it has been adopted as the baseline design for the actual LARP quadrupole magnets.

As shown below, both HQ02a and HQ02b reached over 80% of their short-sample current limit (a theoretical performance ceiling set by the cable with which they are made) within their first or second quench, exhibiting a field gradient of no less than 165 T/m, and in three tests achieving more than 192 T/m. The Hilumi-LHC will operate the interaction-region quadrupole magnets at 75-80% of the short-sample current limit, so this is a great success.

Tests of HQ02a and HQ02b show performance that even in the worst cases was in excess of the needs of the Hilumi-LHC  and in most cases far exceeded the need. Tests of HQ02a and HQ02b show performance that even in the worst cases was in excess of the needs of the Hilumi-LHC (lower dashed line) and in most cases far exceeded the need. The data points are quenches (transitions from superconductivity to normal conductivity in part of the magnet, requiring its energy to be removed). Quenches are thought to result from local movement of part of the superconductor.As shown in these “training curves,” such magnets achieve their full field over the course of several quenches — a sort of break-in process that beds the superconducting coils down into a highly stable position. A trained magnet can be relied upon to deliver a certain magnetic field in an operating accelerator.


The operational concerns for an accelerator magnet include not only the strength and uniformity of the magnetic field, but also the speed with which the field can be ramped up to its full value. For HQ02a and HQ02b, we added a tape of SAE 316 stainless steel, 8 mm wide and 0.025 mm thick, into the core of the cable. This component suppressed eddy-current heating in the strands of the cable, so that, as shown below, much higher ramp rates could be obtained without quenching the magnet.

HQresults_450px A core with a tape of stainless steel allows a magnet to be ramped up faster to a higher current (and therefore field) before quenching, as shown in this comparison of quench current vs. ramp rate for HQ02 and the coreless HQ01a.

Encouraged by the results from HQ02a and HQ02b, LARP is presently testing the final magnet in the series, HQ03, at FNAL, and expects to report the results in a few weeks. Now that the coil parts have been standardized along with the coil fabrication processes, the magnet is expected to be the best in the series. The coils in HQ03 were all made from the same batch of cable, and thus the same wire, so they should behave similarly, providing a clearer picture of the magnet performance.

The HQ03 magnet is the last of the HQ series. LARP is now focusing its efforts on the development of the MQXF quadrupole magnets — with a bore diameter of 150 mm, larger in diameter than the developmental HQ units — for actual use in the project.

LBNL played key roles at every stage of the HQ program: (left to right) design of the actual superconducting wire; making cables from it with our in-house R&D cabling machine; designing the support structure; assembling coils and magnets; and instrumentation and test (performed at Fermilab and CERN).

To Learn More…

Visit the Superconducting Magnets page of the ATAP website. Those who want technical background on this magnet may wish to read these relevant journal articles:

H. Bajas (CERN) et al., “Cold test results of the LARP HQ02b magnet at 1.9K,” in Proceedings of the 2014 Applied Superconductivity Conference (Charlotte, North Carolina, US, 10-15 August 2014), IEEE Transactions on Applied Superconductivity PP, 99 (preprint, 18 December 2014), 2378375.

F. Borgnolutti et al., “Fabrication of a second-generation of Nb3Sn coils for the LARP HQ02 quadrupole magnet,” in Proceedings of MT-23, the 23rd International Conference on Magnet Technology (Boston, Massachusetts, US, 14-19 July 2013), IEEE Transactions on Applied Superconductivity 24, 3 (June 2014), 6605499.

G. Chlachidze (FNAL) et al., “Performance of HQ02, an optimized version of the 120 mm Nb3Sn LARP quadrupole,” in Proceedings of MT-23, the 23rd International Conference on Magnet Technology (Boston, Massachusetts, US, 14-19 July 2013), IEEE Transactions on Applied Superconductivity 24, 3 (June 2014), 4003805.

A. Godeke et al., “A review of conductor performance for the LARP high-gradient quadrupole magnets,” Supercond. Sci. Technol. 26, 095015 (2013).

X. Wang et al., “Multipoles induced by inter-strand coupling currents in LARP Nb3Sn quadrupoles,” in Proceedings of MT-23, the 23rd International Conference on Magnet Technology (Boston, Massachusetts, US, 14-19 July 2013), IEEE Transactions on Applied Superconductivity 24, 3 (June 2014), 6626579.


Leveraging Superconducting Magnet Expertise for FELs

ATAP’s Superconducting Magnet Program is about to begin coil fabrication on a superconducting undulator (SCU) for free-electron-laser facilities. The technology development will yield significant performance improvements for existing and future FELs. The project is a joint effort with Argonne National Laboratory and SLAC National Accelerator Laboratory.

In an FEL, an electron beam goes through an undulator — an array of powerful magnets of alternate polarity, spaced at a certain period. The undulator makes the electron beam change direction rapidly, giving off synchrotron light. Due to the periodic nature of the undulator field and microbunching of the electron beam, the resulting photon beam is coherent and hence powerful and intense, and its energy can reach into the hard-X-ray spectrum. Users in a variety of fields are already using FELs for discovery science, and new facilities such as Linac Coherent Light Source II at SLAC are being built to meet their needs.

The trend in these science-oriented FEL facilities is toward obtaining the electron beam from linacs with a repetition rate so high that it is often described as a continuous wave. The high repetition rate gives a higher data rate for user science and also allows a single linear accelerator to drive a “farm” of FELs. SCUs — particularly SCUs built with the very high performance niobium-tin (Nb3Sn) superconductor — can provide a significantly higher magnetic field than a traditional undulator with the same period which is based on permanent magnets. The result is a significant cost savings because

  • The undulator can be much shorter (by some tens of meters, compared to a conventional unit, for a hard-X-ray FEL)
  • For a given output wavelength, less electron beam energy is required; and
  • For a given beam energy, a broader energy spectrum (tuning range) can be achieved.

SCUs also provide a natural mechanism for “tapering” by changing the current in consecutive sections, resulting in access to extremely powerful photon beams.

SCUs for FELS (and for storage-ring-based synchrotron light sources) have been investigated for years, but building them — especially for x-ray FELs, where the undulator serves as the FEL amplifier and must meet exacting criteria — is challenging. This project takes advantage of deep LBNL expertise in undulator technology, as well as leadership in the design and fabrication of superconducting magnets and in working with Nb3Sn. The latter two capabilities originated with work on behalf of high-energy physics and are now being leveraged for other fields, including FELs.

LBNL deliverables include a 1.5 m long Nb3Sn SCU prototype that meets LCLS-II requirements, as well as development of a method for actively tuning SCUs. The project is proceeding well, with conductor and undulator cores fabricated and coiling winding about to commence. All major components of the active tuning scheme have been developed, and detailed Monte-Carlo simulations based on measured machining tolerances suggest we can attain the stringent magnetic-field quality requirements of an FEL with only a modest number of active correction loops. Our ANL collaborators are complementing this work with a 1.5 m niobium-titanium (NbTi) SCU prototype and a test cryostat that will be used to test both devices. Testing is anticipated this summer.

STop:  conceptual  layout  of an FEL line based on SCUs.<br />  Bottom:  the machined 1.5 meter long core that will be used for half of the LBNL niobium-tin SCU prototype.
Top: conceptual layout of an FEL line based on SCUs. Bottom: the machined 1.5 m core that will be used for half of the LBNL Nb3Sn undulator prototype.


To Learn More…

P. Emma, N.R. Holtkamp, and H.-D. Nuhn (SLAC); D. Arbelaez, J.N. Corlett, S.A. Myers, S. Prestemon, and D. Schlueter (LBNL); C.L. Doose, J.D. Fuerst, E. Gluskin, Q.B. Hasse, Y. Ivanyushenkov, M. Kasa, G. Pile, and E. Trakhtenberg (ANL), “A plan for the development of superconducting undulator prototypes for LCLS-II and future FELs,” in unrefereed Proceedings of the 2014 Free Electron Laser Conference (Basel, Switzerland, 25-29 August 2014), presentation THA03.


Rare Isotopes Start with an Unusual Magnet

At one of the flagship facilities of the DOE Office of Nuclear Physics — the Facility for Rare Isotope Beams (FRIB) that is being built at Michigan State University — everything begins with a high-performance source of heavy ions. Leveraging experience with LBNL’s VENUS ion source, the FRIB team is developing a 28-GHz electron cyclotron resonance (ECR) source. ATAP’s Superconducting Magnet Program has designed and is building a key component: an advanced magnet with a sextupole field within three solenoids.

The source provides beams of ions as heavy as uranium. It must be capable of high current and high charge states (13.5 pµA and 33+, respectively, for uranium). This difficult task was a natural for LBNL, where an ECR source called VENUS (Versatile ECR Ion Source for Nuclear Science) had already been built and is now operating at the Nuclear Science Division’s 88-Inch Cyclotron.

The Superconducting Magnet Program in ATAP plays a key role in the FRIB ECR source project, implementing the combination of strong magnetic fields needed for plasma confinement: a 2-tesla sextupole field in the plasma chamber with a solenoidal field profile (4 T – 0.8 T – 3T) produced by three solenoids. After a successful design review in September 2014, construction is now getting underway. SMP staff led by Helene Felice and Soren Prestemon are working closely with the Front End team at FRIB, led by Eduard Podzeyev, and the FRIB ion source group, led by Guillaume Machicoane, to ensure proper integration of the ECR magnet in its environment.

The design and realization of the magnetic fields is crucial because the performance of an ECR ion source relies primarily on the plasma density and confinement time. The plasma is produced by ECR heating of the electrons. To build up the plasma density, strong confinement is required. The strength of the confinement field has to increase with the ECR heating frequency. High intensity sources require correspondingly high frequencies and thus high magnetic fields. The combination of the solenoidal and sextupolar fields will provide a closed iso-B surface in the magnet aperture of at least 1.75 T.

FRIB_fields Left: Profile of solenoidal field required to operate at 28 GHz. Right: Iso-B surface at 1.85 T inside the magnet aperture that provides plasma confinement.


The task was a natural match for SMP, with their depth and breadth of experience in the science and technology of magnets relevant to accelerators. The program offers expertise “from melt to magnet” (that is, at all stages from the metallurgy of superconducting wire to the construction and testing of magnets) and an unequalled degree of integration of computerized design tools. It is also notable for deep two-way collaboration with other researchers and industrial partners.

All coils will be wound from niobium-titanium (NbTi), a familiar product used widely in accelerator applications. The support structure will be based on the “key and bladder” concept, which has been extensively demonstrated on accelerator-type magnets. It will allow tunable pre-loading of the magnet to minimize conductor motion during magnet excitation. Such motion could result in quenches (sudden losses of superconductivity, which would trigger shutdowns). This technique uses water-inflatable bladders (removed after assembly) and load keys inside the structure to provide compression to the sextupole at room temperature. Differential thermal contraction of the support structure components during cool-down to the operating temperature (4.2 K) completes the preload. Thanks to this concept, the assembly of the sextupole is fully reversible, allowing repair or replacement of a sextupole coil if necessary.

Left: View of the magnetic system: 3 solenoids around a sextupole. The pole pieces of the sextupole are made of magnetic steel to reinforce the field in the aperture. Right: Iso-view of the assembled magnet. Shown in pink is the aluminum mandrel on which the solenoids are wound. Left: View of the magnetic system: 3 solenoids around a sextupole. The pole pieces of the sextupole are made of magnetic steel to reinforce the field in the aperture. Right: Iso-view of the assembled magnet. Shown in pink is the aluminum mandrel on which the solenoids are wound.

The complete and tested magnet is expected to be delivered to FRIB in 2016.


Neutralized Drift Compression of Ion Beams: NDCX-II Demonstrates Final Feature

Understanding materials under conditions of high temperature and pressure is important in many areas of astrophysics (for example, the study of warm dense matter and planetary cores). It provides foundational data to the quest for inertial fusion energy, and could lead to the discovery of novel materials and compounds. NDCX-II (the second-generation Neutralized Drift Compression Experiment at Berkeley Lab) is an ion accelerator, supported by the DOE Office of Fusion Energy Science, with unique capabilities for user science in these fields.

We reached its energy goal of 1.2 MeV in October 2014. Since then, the NDCX-II team has commenced neutralized drift compression experiments, demonstrating that all aspects are working.

NDCX-II circa February 2015 ndcx-II-distribution_300px

Left: Photo of NDCX-II in its present configuration. Right: Time trace from a pulse of 1.2 MeV lithium ions compressed to 2.5 ns with a spot size of 1.4 mm (inset). The ion pulse is now compressed 200-fold in time and 60-fold in space compared to results from October 2014, marking a significant advance in the physics of intense ion beams and validating key design features of NDCX-II.


Ions repel each other due to space charge forces, which makes it difficult to pack a great many ions into a short pulse and focus them into a small spot on a target. As the final stage in acceleration, NDCX-II gives the ion pulses a “velocity tilt” (accelerating the rear of the pulse harder than the front), then neutralizes the ions. As the pulses move toward the target, they become shorter because the rear particles are catching up with the front ones (the “drift compression”), and since they have been neutralized, they are no longer repelling each other.

These techniques have been developed by the Fusion Science group at LBNL in order to reach peak currents of tens of amperes during short pulses. Development involved ongoing, close collaborations with colleagues from Lawrence Livermore National Laboratory and the Princeton Plasma Physics Laboratory.

Pulse compression from an initial ~500 nanoseconds to 2.5 ns (a factor of 200) has been achieved with a spot size of 1.4 mm for lithium ions at 1.2 MeV. The repetition rate is two pulses per minute. This demonstration of longitudinal and lateral focusing validates the basic design goals of NDCX-II.

Further pulse compression and focusing with increasing peak currents is planned for this spring. The intense, short ion pulses from NDCX-II now enable exciting research in discovery plasma science in support of the mission of DOE’s Office of Fusion Energy Science.

To Learn More…

Visit the Fusion Science and NDCX-II page of the ATAP website; contact Peter Seidl ( or Thomas Schenkel of the Fusion Science and Ion Beam Technology Program in ATAP.


Safety: Always the Bottom Line

ISMCircleOfSafety With safety as a top priority in ATAP, we are holding a Safety Day throughout the Division. This year’s Safety Day will be Monday, February 23rd, 2015.Safety Day is a chance to step aside from our busy routines and prioritize both physical safety and the EH&S training and assurance aspects of our jobs. Each program’s Safety Coordinator will organize their Safety Day, developing tasks together with supervisors and QUEST Team leaders.

These are among the likely Safety Day activities for all:

  • Assessing the safety of our work areas as part of a QUEST Team.
  • Meeting with our QUEST Teams to focus on electrical safety (a particular area of emphasis Labwide).
  • Completing online safety training and Job Hazard Analyses (JHAs) as needed.

Additionally, designated Activity Leads will be developing Activities in the new Work Planning and Control system. Those who have been assigned to WPC Activities will be reviewing and accepting these entries.

If you have any questions or would like more information, please contact ATAP EH&S Coordinator, Pat Thomas at or x6098 or Division Deputy, Operations, Asmita Patel at or x7021.

At the end of this important day that hopefully will give us a fresh perspective on our safety practices and how to improve them, a debrief will take place among safety coordinators, Pat, Asmita and ATAP Director Wim Leemans. Observations, findings and action items will be shared across the Division.