SLAC AND ACCELERATORS
SLAC 40th ANNIVERSARY CELEBRATION
October 2, 2002
Today is a birthday celebration and speeches on such occasions should be a tribute to the birthday child. The child is a laboratory, but the tributes belong to the people who built it, run it and transform it. I will celebrate the accelerator part of the story.
Since the start of the linac, SLAC has produced a string of revolutions and evolutions in accelerators (Figure 1) that have advanced research in high-energy, nuclear and condensed matter physics; and in chemistry, biology and environmental science. The revolutions come with the SPEAR colliding-beam storage ring, the SPEAR synchrotron light source, the SLAC Linear Collider and the future x-ray laser. These have changed or will change the way many people do their science. The evolutions include the increase to the Linacís energy; the line of colliding-beam storage rings from SPEAR to PEP to the B-Factory; and the NLC R&D following on from the SLC. I will tell something of how these came to be, and apologize in advance for mentioning only a few of the large number of people who contributed.
The linac has been a remarkably long-lived machine and, while much has changed outside the linac tunnel, down below the same accelerator sections are in place as were there in 1966. The energy of the machine is today three times what it was then. Two technologies are responsible for this. The first is the development of ever-higher powered klystrons that deliver the radio-frequency power to the accelerator. Our original 20 MW peak-power klystrons have grown to put out 60 MW and still higher-powered tubes are under development. Our innovative Klystron Department is responsible for this under the leadership of only four people in 40 years (Figure 2). The Department not only supplied tubes for SLAC, but also, when the need arose, built special purpose tubes for the DESY Laboratory in Germany, for the CERN Laboratory in Switzerland, for the KEK Laboratory in Japan, for the Institute of High Energy Physics in China, and many others.
The second technology is a bit of RF magic called the SLAC Energy Doubler, or SLED. The energy of the accelerator depends on the RF power, which is why we make ever-higher powered klystrons. But itís easier to make a klystron with a long pulse than one with a higher peak-power output. David Farkas and Perry Wilson (Figure 3) developed a technique widely used around the world to take a long pulse out of a klystron and to turn it into a shorter pulse with much higher peak power. The klystrons that now put out 60-MW drive the SLED system, which turns that into 200 MW to be pumped into the linac.
The first of the revolutions comes with the SPEAR colliding-beam storage ring. Like the linear-accelerator technology, the colliding-beam work at SLAC has its roots in the old High-Energy Physics Laboratory on the Stanford campus. The construction of the first serious colliding-beam system began there in 1958 (Figure 4). It was an electron-electron collider, not an electron-positron collider as we have now, but it was the pioneer facility and until recently held the record for circulating current in an electron machine, 0.6 Amp. This facility, first proposed by Gerard K. OíNeill of Princeton, and built at the High-Energy Physics Lab by a Princeton-Stanford collaboration, is the grandfather of all electron colliding-beam machines.
The SLAC colliding-beam story goes back to 1963, when Panofsky lured me to SLAC with an Associate Professor position and carte blanche to do what I had been trying to convince him to do Ė build a big electron-positron collider.
I want to pause for a bit and say something about Pief Panofsky. Building a world-class laboratory on a green field site is a very difficult job. Pief could do it because, besides management experience and influence in Washington, he had a deep knowledge of physics, a vision of its future, outstanding good taste in the selection of problems, and a talent for picking good people. He set the style of the place and a good style it has been.
It did not take long to design the machine, thanks to the arrival of John Rees in 1965 (Figure 5), but it did take a long time to get the funding. Thus, we had a great deal of time to do the design and plan the construction in detail. When we finally began the construction, it took only 20 months. It was a Spartan facility (Figure 6) with no permanent buildings but had all of the features of todayís electron-positron colliding beams including what I think was the first real computer control system, low-beta collision points, sophisticated beam-monitoring systems, global and local orbit-correction schemes, etc. All of the electron storage ring colliders since then have been extensions of SPEAR, except for the new two-ring asymmetric machines such as our B-Factory.
With SPEAR came something that turned out to be as revolutionary in many areas of science as SPEAR was in high-energy physics. Four people are responsible for starting what has become the SSRL Division of the laboratory (Figure 7). Sebastian Doniach of Stanfordís Applied Physics Department and William Spicer of the Electrical Engineering Department visited me one day and told me they could revolutionize solid-state physics if I would only bring the x-rays produced by the circulating beams in the storage ring out of the ring so that they could use them. These beams were one million times the intensity of what could be obtained with conventional x-ray tubes and with that extra intensity they could do things that had never been done before. I, in turn, asked two of our people here at SLAC, Gerry Fischer and Ed Garwin, if I should believe them. Fischer and Garwin convinced me that they were correct and, since it was a different time, I just did it. I didnít ask anyone, I didnít tell anyone, I didnít request a formal change order, and I didnít rebaseline the project. Letting the beams out cost $25,000 (Figure 8) and it did revolutionize solid-state physics, structural biology, some areas of chemistry and environmental science.
The scientific output of the synchrotron-radiation program here from 1972 to 1976 convinced the rest of the world of the importance of this system, and there are now dozens of such machines all over the world. At first SPEAR was a shared facility between high-energy physics and x-ray science. SPEAR was turned over to what was by then SSRL (Figure 9) in 1988 when a more modern colliding-beam facility for high-energy physics was built in Beijing in the same energy range. SPEAR will continue to be a highly productive generator of multidisciplinary science for many years, thanks to its most recent upgrade which, because of the importance of the facility to biologists, was half paid for by the National Institutes of Health.
SLACís first evolutionary step in colliding-beam storage rings was the PEP project built as an extension of SPEAR technology. However, it didnít start that way. In its original conception, PEP was to include a second ring of superconducting magnets storing a circulating beam of protons to make an electron-proton colliding beam facility. Lawrence Berkeley National Laboratory would do the superconducting magnet ring and SLAC would do the electron-positron ring (Figure 10).
It turned out to be too expensive for its time and PEP, the Proton-Electron-Positron machine, became PEP, the Positron-Electron-Project. However, the important collaboration with Lawrence Berkeley Laboratory stayed in place to everyoneís mutual benefit and grew later with the B-Factory and the linear collider, both of which also included Lawrence Livermore National Laboratory. The initial PEP design showed how to build a proton-electron collider and the DESY laboratory later built the first one with a much more advanced super-conducting magnet system than could ever have been built at the time PEP was designed. PEP and the B-Factory did demonstrate a new management technology Ė multi-laboratory projects bringing together a range of talents not available at one place. That too has been a model for other things such as the spallation neutron source, a collaboration of Oak Ridge, Argonne, Brookhaven, Los Alamos and LBNL and Jefferson Lab.
Though the B-Factory is not next chronologically, it is next in the extension of the colliding-beam storage ring technology. B-meson physics had become very important and some of the most important problems required the production of huge numbers of B-mesons in a fashion that was simply not possible with the one-ring colliding-beam system. Pier Oddone had the brilliant insight that, if you could make a two-ring machine with unequal energies, the kind of experiments that were not possible with the one-ring machine could be done (Figure 11). The suggestion was enough to make Jonathan Dorfan pull together a group of our accelerator physicists to see whether such a thing could in fact be done. It could and, after some interesting times in Washington, SLAC began the construction of the B-Factory in 1994.
The B-Factory and its sister facility at the KEK laboratory in Japan have been incredibly productive. I think itís fair to say that our machine awed the accelerator physics community with its very rapid turn-on and rise in luminosity to pass its designed performance 12 months after startup. We look forward to continued increases in its performance and to a long and productive career in physics.
Revolution number three is the development of the linear collider (Figure 12). With the construction of the huge 27-kilometer circumference LEP electron-positron storage ring collider at the CERN laboratory, it became clear that the line pioneered by SPEAR had reached its practical end. The size of colliding-beam storage rings for electrons scales as the square of the energy, and so going up by a factor of ten from the energy of LEP would require a machine that was 2700 kilometers in circumference Ė a financial impossibility, never mind the accelerator issues.
Maury Tigner of Cornell, Alexander Skrinsky of Novosibirsk, and I had independently been thinking about how to keep the electron-positron colliding beam line going. We met at a conference and found that we had all been thinking about this problem and had come to the same conclusion Ė one could do better with two linear accelerators firing beams at each other. We worked out the general parameters and at SLAC a linear collider design began using our one existing linac and a trick. We could accelerate both the electrons and positrons in the same linac pulse and then use magnets (Figure 13) to bring the beams into collision at the end of the linac.
At the start-up of the facility in 1987, there were a host of new problems to be faced, several of which had to do with the magnet system that brought the beams in to collision. These were solved and then the real work began on stabilizing the linac to a degree far beyond anything dreamed of before. Data began coming in for the experiments slowly and it wasnít until 1992 that the data rate began to climb, thanks to the work of the SLC Task Force led by Nan Phinney (Figure 14). At the end of SLC operation, it produced as many physics events in an hour as it did in a year at the beginning.
The SLCís success made the rest of the world believe that this crazy notion could actually work and so began an unprecedented worldwide collaboration to carry out the critical R&D necessary to realize an actual two-linac collider (Figure 15). The collaboration began smoothly with DESY in Germany, the KEK laboratory in Japan, and SLAC in the U.S., working on critical R&D issues. One of the first joint projects was the final focus test beam here where the world collaboration led by David Burke produced a beam focused down to a size of 70 nanometers, much smaller than anything yet produced for an integrated circuit.
There is now agreement among the worldwide high-energy physics community that the construction of a 500-GeV machine, ten times the energy of the present SLAC linac, should begin as soon as possible. Where once all were working together, there is now a rivalry for the site mixed up with the still on-going cooperative R&D. Of course, I think that the solution to the site problem is simple. Since the technology was born here at SLAC, it should grow up here. The correct site is California. However, others may not agree with me and I expect that there will be a site selection some time in the next two to three years for the first accelerator project that is built by a true worldwide collaboration.
This is, however, not the end of the linear collider story for, once again, we are attempting to repeat what was done with SPEAR, revolutionizing both high-energy physics and x-ray science. The aim is to produce an x-ray laser with a hundred times the average power of our storage ring and ten billion times the peak power (Figure 16). The success of this project requires the marriage of the ultra-precise control of tiny beams in a linear accelerator developed for the linear collider to the undulator and x-ray technology of the standard storage ring source. I believe it will be as revolutionary as was SPEAR itself. The DOE has approved it in principle, is supplying design money, and construction is scheduled to start in 2005.
Accelerators are the glue that has held together all the parts of the laboratory. The revolutions and evolutions are what bring the 3000 scientist-users here each year. Some worry that with the very large and very costly 500 GeV Linear Collider, we have reached the end of the line. But, at any time in the laboratoryís 40-year history, the question of how long we can productively continue our program has always had the same answer Ė 10 years. But, the genius of SLACís staff has been continued innovation - evolutions and revolutions that have always added another decade before the last has run out.
Let me conclude by observing that the SLAC staff has built great accelerators that have benefited many areas of science. This kind of work has not stopped. In various corners of the laboratory people are working on using lasers to accelerate electrons; making accelerator structures out of silicon, as integrated circuit chips are fabricated; creating plasmas to serve as accelerators, etc. Some of these systems may work out and some of them may not. But, if any of them does, we will then be able to have ten to a hundred times the energy of the NLC for an affordable price. How long will our programs be on the science frontier?
I think the answer will be the same at SLACís 50th birthday Ė 10 years.