GSI Helmholtzzentrum für Schwerionenforschung GmbH

Where the Heavy Ions Come From

The wide range of experimental possibilities that exist at GSI is based on a multi-stage accelerator facility consisting of a whole system of linear and circular accelerators and beam pipes which interconnect both the accelerators and the experiments. This report focuses in greater detail upon the accelerator technology referred to in preceding chapters. It also attempts to give some idea of the complexity involved in simultaneously providing up to five experiments with different types of ion beams, each possessing a different energy.

Plan view of the accelerator and experimentation facility. Ions are injected by the North and South injectors (left) and then pass through the adjoining Wideröe structure or, alternatively, originate in the high-charge injector before being fired into the Alvarez structure at about 5% of the speed of light. By the time it has reached the end of UNILAC, the ion beam has reached 16% of the speed of light. Part of the beam is then diverted to the adjoining experiment hall for experiments, while the remainder is transferred to the heavy ion synchrotron SIS for further acceleration. There, the ions reach up to 90% of the speed of light, before being directed to experiments at the fragment separator FRS, the ESR, or the target hall..

View of the UNILAC tunnel. In the foreground, the two old injectors with the ion sources can be seen to the left and right. The Wideröe and Alvarez tanks are situated in the tunnel behind the injection point. The single resonator structures developed by Christoph Schmelzer and his colleagues are at the back..
Animation as Mpeg-File with 1164460 Bytes

Ion beams required for experimental work at GSI are generated by an accelerator facility consisting of three components: the linear accelerator UNILAC, the heavy ion synchrotron SIS, and the experimental storage ring ESR. UNILAC serves two functions: it provides ions for a program of experiments at low energies and acts as a pre-accelerator for the SIS. The high-energy beams from the SIS can then be directed either to the experiments set up in the target hall, the fragment separator FRS, or the ESR. The ESR makes it possible to "cool" the ion beams and use them for internal ring experiments. Alternatively, the ESR beams can also be extracted for use in experiments in the target hall, or returned to the synchrotron for further acceleration. In short, we are talking about a very versatile labyrinthine-like system.

UNILAC, which was completed in 1975, was GSI's first accelerator. Thanks to preparatory work carried out by a study group at the Institute of Applied Physics in Heidelberg, it was possible to begin construction of the accelerator shortly after GSI was founded. At the time, important arguments for choosing this type of accelerator included its simple energy variation and the fact that injection could be easily achieved from a range of ion sources. These characteristics also proved very useful when the installation was later expanded.

In order to accelerate the atoms of the beam as efficiently as possible, they must be put into a highly ionized state. In other words, they must be stripped of as many electrons as possible. This state of affairs is achieved in the ion source, for example, by means of a high-current gas discharge. Back in those days, it was assumed that a maximum of ten electrons could be stripped from uranium, which has 92 electrons in all. In other words physicists believed they could "manufacture" uranium with a surplus charge of ten. The structure of UNILAC was designed with these considerations in mind.

Selecting the desired isotope
Located at the beginning of the beam path are the injectors (North and South). These contain the ion sources from which the ionized atoms are extracted with the help of an electrical field. Next comes an electrostatic preliminary acceleration, in which the ions "fall" through a potential difference of up to 320,000 volts. In the succeeding deflection of the beam into the main accelerator, the desired isotope is selected from the natural mixture of isotopes of the element concerned.

A so-called "Wideröe structure" makes up the first stage of UNILAC. The structure is named after Rolf Wideröe, a Swiss physicist of Norwegian descent, who invented the principle of the radio-frequency accelerator in 1928. It consists of four electrode structures together measuring about 30 meters in length. Each structure is situated in a copper-plated steel tank and contains about 130 accelerating electrodes - the "drift tubes." The length of the electrodes is so chosen that the radio-frequency electrical fields, which have an operating frequency of 27 MHz, always have the correct polarity when the ions are located between the electrodes. In this way, the ions are continually "nudged" forward through the fields. Some of the electrodes also contain magnetic lenses which guide the beam along the axis of the accelerator.

Accelerating electrodes Wideröe structure

Just like links in a chain: the 130 accelerating electrodes of the Wideröe structure are neatly arranged one after the other in the four copper-plated steel tanks. When the ions emerge from the metal cylinders, they are subjected to the force of a high-voltage field which accelerates them until they enter the next cylinder. .
Animation as Mpeg-File with 43319 Bytes

All ion types leave the Wideröe structure at an equal, fixed speed of about 16,000 km/s, which is a good 5% of the speed of light. This speed corresponds to an energy of 1.4 mega-electron- volts per nuclear particle (1.4 MeV per nucleon). The ions then pass through a supersonic gas beam, in the process of which, they undergo such violent collisions that many more electrons are stripped off than initially was the case in the ion source. In fact during this "stripper" phase, uranium ions can reach charge states of around 28+: in other words, they have lost 28 electrons. Without this additional ionization stage, it would have been necessary to make UNILAC much longer in order to reach the desired final energy. Just as is the case in the ion source, the collisions with target atoms of the gas beam produce different charge states, only one of which is selected for further acceleration.

So-called "Alvarez structures," which were invented in 1946 by the US physicist and later Nobel laureate Luis W. Alvarez, then take over the job of accelerating the ions in UNILAC's 55-meter-long second stage. Acceleration occurs in four tanks, each of which is 13 m long and has over 150 drift tubes or accelerating gaps. The polarity reversal of the electrical field located between the drift-tube electrodes occurs at an operating frequency of 108 MHz - or four times as fast as in the Wideröe structure. The specific exit energy ultimately amounts to 11.6 MeV per nucleon for all ions, a value corresponding to 16% of the speed of light, or almost 50,000 km/s.

The third stage of UNILAC is a specially developed structure, comprising a sequence of fifteen single resonators. Each of these resonators has only one accelerating gap, and is powered by an independently controllable radio-frequency generator. As a result, the ion speed can be adjusted (this is not the case in the first and second stage structures). By accelerating or decelerating the ion beam, it can be given any desired final energy in a range extending from 2 to 18 MeV per nucleon.

Three new techniques for the high-charge injector
In the mid-Eighties, UNILAC's second and third stages were converted in such a way that it became possible to switch over to a different final energy in less than 15 milliseconds. Thanks to this improvement, injection energies for the heavy ion synchrotron, and for the low-energy experimental program can now be selected independently. With the installation of an additional injector at the Alvarez stage - the "high-charge injector" - in 1992, it also became possible to change over to a different type of ion in less than 15 milliseconds. Consequently, the type of accelerated ion can be selected independently for UNILAC and SIS experiments. This new injector, which is operated in parallel to the Wideröe component, enables the charge state of 28+ for uranium to be set directly using a special ion source known as an electron cyclotron resonance source (ECR).

In order to accelerate the ions to the 1.4 MeV per nucleon needed for the Alvarez structure, they are first injected into a "radio-frequency quadrupole" (RFQ) structure of the type invented in the late Seventies by Kapchinsky in Moscow, and subsequently built for GSI at the University of Frankfurt. The next point on the journey is formed by an IH structure of the kind first employed by Morinaga and colleagues in Munich in 1978. The IH structure has since been further developed at GSI for use with slow, heavy ions. The combination of these three new techniques has made it possible to build very compact heavy ion linear accelerators.

Soon after completion of UNILAC, GSI began to formulate plans to build an accelerator for ion beams with considerably higher energies. It was evident that only a synchrotron ring accelerator would suffice for heavy ions. A synchrotron ring is a vacuum ring in which the ions, held on a circular path by magnetic fields, are accelerated to near light speed over hundreds of thousands of revolutions. This type of accelerator gets its name from the fact that both the magnetic field and the frequency of the accelerating electrical field rise synchronously with the increase in speed.

View along the SIS accelerator ring, which can accelerate the ions coming from the UNILAC to 90% of the speed of light. The red sections are the deflecting magnets; the yellow sections the focusing magnets. The beam is injected into the ring from the left.
Animation as Mpeg-File with 1440556 Bytes

The SIS, which has a circumference of 216 meters, is just such a "synchrotron." Orbiting ions are held to their circular path by 24 bending magnets and focused by 36 magnetic lenses. In order both to make the acceleration as efficient as possible and to achieve a high energy, the ion beam is passed through a thin carbon foil which ionizes it still further. The foil, which is situated in the section between UNILAC and the SIS, can increase the surplus charge on uranium ions from 28+ to 72+. A specific energy of about 1000 MeV (= 1 GeV) per nucleon is achieved at the maximum possible magnetic bending force of the SIS magnets. In fact, it is even possible to achieve completely ionized ions when light elements such as neon are involved. As a result, specific energies of up to 2 GeV per nucleon, which is equivalent to about 90% of the speed of light, are possible.

Acceleration in the synchrotron takes place in two radio-frequency structures diametrically opposed on either side of the ring. In these, the ions "fall through" a potential difference of 15,000 volts during each revolution and thereby attain maximum energy over the course of some hundreds of thousands of revolutions. The frequency increases from 800 kHz to a maximum of 5.6 MHz in step with the increase in speed. During the acceleration phase, which lasts for about a second, the ions cover many thousands of kilometers. Since each uranium ion still has about 20 electrons bound to its nucleus, it is vital that the tube in which the ions are orbiting be evacuated to an extremely high vacuum - up to 10^-10 torr. Otherwise, they might lose electrons by collisions with gas molecules in the tube and thus deviate from their path. This means that the vacuum in the SIS ring accelerator must be a thousand times better than that in UNILAC.

After acceleration to the desired energy, the beam is either extracted quickly - in less than a microsecond - or slowly - over a period of up to ten seconds. Fast extraction is carried out if the beam is to be transported into the storage ring or used for special experiments. Slow extraction is used for most experiments in the fragment separator FRS and in the large target hall.

Experiments with individually selectable parameters
In conjunction with a special technique for the magnet power supplies, the SIS control system enables physicists to set a different final energy, a different kind of extraction ("fast" or "slow"), or a different intensity for each acceleration cycle. At present, up to sixteen different settings can be stored and then recalled in a selectable sequence. Thanks to this flexibility, several experiments can be run in parallel, each with individually selectable parameters. Moreover, with the installation of the new high-charge injector, a further parameter has been added - the second type of ion. The dimensions of the Experimental Storage Ring (ESR) are the direct result of its coupling to the SIS. At 108 meters, its circumference is half that of the SIS accelerator. The storage ring has a hexagonal shape with two longer straight sections - one for the electron cooler, the other for a gas target. The presence of the target enables physicists to conduct experiments inside the ring. The hexagon is made up of six bending magnets, each with a deflection angle of 60 , and twenty magnetic lenses for focusing. It is possible to store fully ionized uranium ions with a specific energy of up to 560 MeV per nucleon, or neon ions with up to 830 MeV per nucleon in the ring.

The beam properties of the stored ions can be substantially improved using an "electron cooling" process, which is described in the chapter on atomic physics. Here, an electron beam traveling at the same speed as the ion beam is deflected into the ring, where it accompanies the ion beam for two meters, before being deflected out again. The electrons have precisely the desired speed and direction of the ion beam. Should any individual ion deviate from this velocity, it is accelerated or decelerated by the electrons over the course of many revolutions until it possesses the desired value. Statistically, deviations from the ideal values can be described in terms of a temperature which decreases with time. This is actually why the process is referred to as "electron cooling" of the ion beam.

The cooling effect shrinks the diameter of the ion beam from several centimeters to a few millimeters within a fraction of a second. At the same time, and depending on the number of ions in the ring, the uncertainty in momentum can be reduced to a thousandfold of its original value. The "cooled" beams are then utilized for internal ring experiments or extracted from the ESR for experimental work in the target hall. Not only is the quality of these beams superior to anything achieved previously, they can also be extracted quickly, injected back into the SIS, and accelerated to even higher energies as fully ionized atoms.

The transverse gas target, which can be bombarded repeatedly with circulating ions, is located in the second of the ESR's longer sides. The target is used for both nuclear and atomic physics experiments in the ring itself. Since the beam is heated up by collisions with the gas atoms, it is important that electron cooling runs simultaneously. The vacuum pressure in the torus of the ESR is about 10^-11 torr, i.e., yet another order of magnitude lower than the pressure in the SIS. When subjected to electron cooling, the typical life of the orbiting ion beam is about ten hours for neon ions, and 30 to 60 minutes for heavy ions. It was primarily thanks to such long lifetimes that physicists were finally able to experiment with fully ionized and cooled uranium- ion beams.

The ESR's ion lens system is distinguished by its ability to handle a wide range of particle angles and momenta. For example, two beams intersecting at a small angle can be stored simultaneously in one ring. In addition, the relative speed can be chosen in such a way that the mutual repulsion of the nuclei is overcome with the result that nuclear reactions occur. Moreover, in a process which will soon be improved by the introduction of stochastic cooling, "hot" secondary beams from the fragmentation of ion beams can be collected and cooled. Stochastic cooling works by measuring and correcting the path deviations of individual ion bunches.

The entire accelerator system is controlled and monitored from the main control room.

Depending upon experimental requirements, UNILAC can provide up to seven different elements in each operating period. These go to supply experiments at UNILAC, the SIS, and the ESR. The choice of injector - i.e. whether the Wideröe structure or the high-charge injector is used - depends upon what intensities and elements are required. This is true for both low-energy and high-energy experiments. As a rule, the new injector is used for very expensive elements, such as enriched isotopes. In the normal course of affairs, a main experiment and a so-called "parasitic" one (or sometimes even two such experiments) are run at UNILAC and at the SIS. In addition, there is also the beam for the ESR. In other words, it is not uncommon for five experiments to be supplied with beams in parallel. The accelerator facility is in operation for 5,000 to 6,000 hours per year. Periods of five to six weeks of 24-hour-a-day operation alternate with two to three weeks of maintenance, repair, and conversion work. The accelerator facility and the beam transport systems are monitored and controlled by a shift supervisor and two operators in a central control room for all the accelerators. Specialists are on call round the clock for more complicated malfunctions..

Future development of the accelerators
Development work currently in progress on UNILAC and the SIS is aimed at further improving operating reliability, as well as the precision with which individual settings can be reproduced for both the accelerators and the beam transport system. This is the most urgent task, especially in view of the tumor therapy project. In addition, a program to increase intensity has begun at the SIS. The aim of this project is to inject so many ions into the SIS that a further increase becomes impossible due to the mutual repulsion of the charged ions. In other words, the SIS is to be filled up to the space-charge limit, and this is to be done for all ions. Although such a feat is already possible for light ions, researchers need a 500-fold improvement if they are to reach this limit for very heavy ions such as uranium.

The diagram shows the development over time of the ion currents accelerated in the SIS (in units of particles per accelerator pulse) for the various elements. The period covered is from 1994 to 1998. The green bars show the situation at the end of 1994; the red broken line the maximum possible values which physicists hope to attain as a result of the program to increase intensity. This program comprises improvements to the ion sources, installation of an SIS electron cooler, and a new high-current injector.

The central component of this program to increase intensity is a new high-current injector for UNILAC. This injector has been designed to deliver a short, intensive beam-pulse to the SIS about once every second. In order to accomplish this, it has proved necessary to equip UNILAC with instrumentation and monitoring equipment for extremely high currents. This upgrading work is due to be completed by the end of 1998.

Last update: 12. Nov. 2010 by C.Bisignano