TRIUMF: Canada's national laboratory for particle and nuclear physics

A proposal for an intense radioactive beams facility ISAC Oct 19, 1995

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


51833-TRI-95-01.pdf [ 62.55MB ]
JSON: 51833-1.0228648.json
JSON-LD: 51833-1.0228648-ld.json
RDF/XML (Pretty): 51833-1.0228648-rdf.xml
RDF/JSON: 51833-1.0228648-rdf.json
Turtle: 51833-1.0228648-turtle.txt
N-Triples: 51833-1.0228648-rdf-ntriples.txt
Original Record: 51833-1.0228648-source.json
Full Text

Full Text

I SAC A PROPOSAL FOR AN INTENSE RADIOACTIVE BEAMS FACILITY Postal Address: TRIUMF Publications Office 4004 Wesbrook Mall Vancouver, B.C. Canada V6T 2A3 October 19, 1995 TRI-95-1 Contents 1 INTRODUCTION 1 2 SCIENTIFIC MOTIVATION 5 2.1 Introduction . . . . . 5 2.2 Nuclear Astrophysics .... 6 2.2.1 Introduction . . ... 6 2.2.2 Nuclear reactions of astrophysical interest 9 2.2.3 Hydrogen burning . . . . . . 10 2.2.4 The solar neutrino problem 10 2.2.5 The hot pp chain ..... 11 2.2.6 Explosive nucleosynthesis 11 2.2.7 Heavy elements . . . . . 17 2.2.8 Experimental equipment 18 2.3 Fundamental Symmetry Tests 20 2.3.1 Neutral atom traps . . . 21 2.3.2 PN C in francium isotopes 23 2.3.3 Symmetry tests . . 25 2.3.4 TRIN AT progress . . . . . 29 2.4 Nuclear Physics . . . . . . . . . . 31 2.4.1 Properties of nuclear states 31 2.4.2 Radioactive decay modes . 34 2.4.3 Fusion reactions . . . 36 2.4.4 Coulomb excitation . 37 2.4.5 Magnetic moments 39 2.5 Materials Science ...... 41 2.5.1 Introduction . . . . . 41 2.5.2 Mossbauer spectroscopy 42 2.5.3 PACS .......... 44 2.5.4 Nuclear stimulated desorption 46 2.5.5 Nuclear orientation . . 47 2.5.6 Radiotracer diffusion 49 2.5.7 Ion beam modification 49 2.5.8 Thin layer activation 50 2.6 Biomedical Applications 2.6.1 Introduction . . 0 0 2.6.2 On-line separation 2.6.3 Therapy . 0 • 2.6.4 Implantation . . . 3 RADIOACTIVE BEAMS AND FACILITIES 3.1 Introduction ............ . 3.2 Production of Radioactive Species . 3.2.1 Projectile fragmentation 3.2.2 The ISOL method ... 3.2.3 Radioisotope production 3.3 General Concept of ISAC . . . 3.4 Projected Intensities . . . . . 3.5 ISAC in the World Situation . 3.5.1 Overview ...... . 3.5.2 PFM laboratories . . . 3.5.3 !SOL-based laboratories 3.5.4 Comparison . . . . . . 4 THE PROPOSED FACILITY 4.1 Introduction .... 4.2 The ISOL Facility .. 4.3 The Accelerator . . . 4.4 Experimental Areas . 5 THE ISOL SYSTEM 5.1 Overview . . . . . 5.2 Beam Line 2A . . 5.3 The Target Area 5.3.1 General 5.3.2 Target facility . 5.3.3 Remote handling 5.3.4 Target/ion-source module 5.4 Production Targets . . . . . . 5.4.1 General considerations 5.4.2 Existing systems .. 5.4.3 Target analysis ... 5.4.4 High-current systems 5.5 Ion Sources . . . . . . . . . 5.5.1 General ...... . 5.5.2 Surface and thermal cavity . 5.5.3 Plasma . . . . . . . . . . . 11 52 52 54 54 54 57 57 57 58 58 61 62 63 64 64 64 66 68 73 73 73 75 77 79 79 82 82 82 83 85 86 87 87 88 88 89 89 89 90 90 5.5.4 Electron cyclotron resonance . 5.5.5 Laser . . . . 5.5.6 Multi-cusp . . . . 5.6 Ion Beams . . . . . . . . 5.6.1 Beam diagnostics 5. 7 High Voltage System 5.8 Vacuum System . 5.9 Off-Line Source 6 ACCELERATOR 6.1 Specifications . . . . . 6.2 The Need for Stripping . 6.3 Conceptual Design . . . 6.3.1 General description . 6.3.2 RFQ ........ . 6.4 Beam-matching and Stripper Section 6.5 Drift-tube Linac . . . . . . .. 6.6 Beam Dynamics Calculations 7 CONTROLS 7.1 Overview .......... . 7.2 Central Controls Philosophy 7.3 Implementation .. .... . 7.4 Integration with User's Requirements 8 RADIOLOGICAL SAFETY 8.1 General Considerations .. 8.2 Radioactivity Production . 8.3 Radioactivity Distribution 8.4 Target Shielding ..... . 9 COST ESTIMATES AND SCHEDULE 1ll 91 92 92 92 95 96 96 97 101 101 101 102 102 103 109 111 114 119 119 119 120 121 123 123 123 125 126 129 EXECUTIVE SUMMARY The last decade has seen a growing worldwide interest in the possibility of generating beams of unstable nuclei for use in a variety of applications ranging from nuclear physics and nuclear astrophysics through atomic physics, condensed matter physics, medicine, etc. This interest has been generated by the considerable improvements which have occurred over the past 20 years in the fields of heavy-ion acceleration, ion sources and, more particularly, in on-line production and separation of unstable nuclear isotopes (ISOL), in the development of high-resolution isotope separators, and in advances in experimental techniques, such as ion trapping. Coupling very intense production sources of unstable elements to efficient accelerator structures should lead to the general availability of radioactive beams for a wide range of nuclei far from the line of stability. The high energy (500 MeV) and high intensity (> 100 pA) proton beam from the TRI-UMF cyclotron makes TRIUMF the prime choice for a radioactive beams facility in North America. The present proposal is based upon the experience which has been gathered at TRIUMF in the operation of its test facility (TISOL) for the on-line production of isotopes, its expertise in ion sources, its experience in handling high-intensity proton beams, and on the development of RFQ accelerators. The goal is to accept proton beams of up to 10 pA on a production target, separate the relevant species in an appropriate ISOL separator, and to accelerate the accepted ions up to an energy of 1.5 MeV ju. Products from the target/ion source will be available over a very wide mass range due to the high energy of the proton beam. The acceleration of the extracted ions will be accomplished by continuous ( CW) RFQ ana two-stage linac structures, will accept ions with qjA ~ 1/30, and could be expanded to higher energies in the future. The primary focus of the subatomic physics programme for ISAC will be: • Studies of fundamental interactions in particle physics using ,8-decaying nuclei; • Studies of the Standard Model using parity violation in atomic systems such as selected Cs and Fr isotopes; • Studies of nuclear reactions of interest in astrophysics; • Studies of nuclear reactions below the Coulomb barrier. Other fields of reasearch will include: • Studies of condensed matter structures via radioactive impurities; • Production of selected isotopes for medical imaging. The cost of the ISAC facility has been estimated at $46,000,000, inclusive of salaries and buildings; it could be in full operation by the middle of 1999. By then the ISAC facility would be a unique instrument which would provide Canadian physicists with the best radioactive beams facility in the world. It also paves the way to the ultimate facility ISL (ISospin Laboratory) as defined by the United States nuclear physics community in their long-range planning. Section 1 INTRODUCTION During the last decade there has been a growing world-wide interest in the possibility of using radioactive ion beams for a variety of fundamental studies in pure and applied science. The possibility of producing intense beams of radioactive nuclei with extreme neutron to proton ratio (N /Z), compared with naturally-occurring isotopes, has opened a new era in nuclear science. Some examples where the use of a radioactive ion beam (RIB) could be of great interest are: • In nuclear astrophysics, enormous pr~r_ess in our understanding of nucleosynthesis, particulary in explosive astrophysical scenarios, can be expected. All elements except hydrogen have been synthesized by complex nuclear reaction and decay processes, beginning with the first few minutes of the Big Bang, which produced primarily 4 He and traces of other light elements, and including the massive stars, which produce a wide range of elements, including all heavy species, and end their lives in spectacular supernova explosions. Except for a few cases where long-lived elements can be employed in experiments, explosive processes, in which interactions involving radioactive species predominate, have had to rely on theoretical estimates of the rates of processes that determine their evolution. RIB facilities will provide the opportunity to measure the probability for formation of many elements at many different stages of stellar evolution. • In nuclear physics it will be possible for the first time to test isospin dependence in a large variety of nuclear systems far from stability. Interesting (and sometimes extreme) nuclear shapes (or deformations) predicted many years ago with a variety of nuclear models can be expected. Tests of these models will provide a strong constraint to our understanding of the nuclear interaction. • Atomic and nuclear physics have been interconnected since the beginning of modern science. Atomic physics techniques are commonly used to measure many properties of the nucleus, including spin, electromagnetic moments, and the charge radius. In a similar way, nuclear physics techniques utilizing nuclei far from stability can be used to test fundamental issues in atomic physics, such as the range of validity of quantum electrodynamics (QED) . With the combination of a mass separator, beams 1 of radioactive nuclei of high intensity, and an electromagnetic trap , a new generation of extremely precise measurements can be achieved. • Precision tests of the Standard Model of elementary particles utilizing the intense RIBs expected to become available will be complementary to and, in some specific areas , competitive with those performed at the high-energy frontier. • The high sensitivity and accuracy of techniques used in nuclear science can be applied in many other areas of research. Unstable nuclei can be used for doping new generations of semiconductor material. An RIB can be used as a probe to obtain information, on an atomic scale, about the environment of implanted ions. This complements the use of muons in the study of the dynamics of certain magnetic phenonmena on a very short time scale, which is currently a very active field of research at TRIUMF. • The medical sciences can benefit from the production, in large quantities, of extremely pure positron emitters, largely for use with the PET (Positron Emission Tomography) technique. Also, radioactive nuclides which emit high LET (Linear Energy Transfer) radiation, can be produced as RIBs and attached to compounds that have specific uptake in tumorous tissue, thereby providing more effective treatment of cancer. • The development of nuclear physics and its appliactions is strongly related to the development of particle accelerators. The study of an RIB facility at TRIUMF has triggered the development of an RFQ (RadioFrequency Quadrupole) accelerator ca-pable of capturing and accelerating, in a continuous mode, heavy ions with very-low charge-to-mass ratio. The high energy (500 MeV) and high current (> 100 pA) of the TRIUMF H- cyclotron makes it a very good choice for the driver accelerator for an RIB facility in North America. The high energy of the protons permits the use of a thick target so that, in contrast with the use of a low-energy accelerator, a much wider mass range of isotopes can be produced via the larger number of reaction mechanisms available at the higher energy. Low-energy proton beams can only produce intense RIBs of nuclei near the valley of stability. Furthermore, high-energy protons have a smaller energy deposition density than low-energy proton or heavy-ion beams. This puts fewer constaints on the choice of target material, thereby allowing higher production intensity for specific isotopes. An H- cyclotron can produce multiple extracted beams simultaneously, which allows one beam line to be dedicated to RIB production. The recent development of an intense polarized ion source at TRIUMF means that nearly 100% of the operating time will be available for RIB production, since adequate beam will be available even during the polarized beam schedule. With the existing TISOL facility (established after extensive study in 1984/85 of the possibility of accelerating radioactive ions), TRIUMF has gained valuable expertise in the domain of RIBs. The studies that resulted in the TISOL facility have been the prototype for a number of RIB proposals from various laboratories around the world. 2 We are at the point where mature developments in the field of isotope separation on-line and the technological expertise of the meson factories can be combined to produce a very attractive RIB facility with beam intensities large enough to provide for many new scientific applications. This can be achieved on a relatively small scale, and with moderate investments, if existing facilities are used. TRIUMF is an ideal site for a world-class RIB facility. The main scientific motivation for an RIB facility is described in Section 2. RIB pro-duction methods are described in Section 3 which places TRIUMF in the world context. In Section 4 a description of the entire facility including specifications, rationale for the techni-cal aspects incorporated, and general layout will be presented. Details of the ISOL facility are given in Section 5, while the accelerator, controls system and safety are described in Sections 6, 7 and 8, respectively. Schedules and cost estimates are presented in Section 9. 3 4 Section 2 SCIENTIFIC MOTIVATION 2.1 Introduction Nuclear reactions determine the energy release and nucleosynthesis that occurs in the birth, evolution and death of stars. In recent years, the focus of investigations in nuclear astro-physics has turned to explosive events (novae, supernovae, X- and 1-ray bursts, etc) for which reaction rates involving unstable nuclear species must be known. A primary moti-vation for the establishment of an accelerated radioactive beams facility at TRIUMF has been the demand for cross section information needed to establish such reaction rates. The measurement of cross sections of astrophysical interest will be a major part of the !SAC program. Very precise measurements of the properties of specific states in selected nuclei have had a dramatic impact on the knowledge of the fundamental properties of the weak interaction. Recent advances in the development of neutral atom traps have made it feasible to contem-plate experiments involving fundamental symmetry tests with trapped neutral radioactive atoms. Specific initiatives of this kind are proposed for !SAC and will form a major part of the program in the low-energy experimental area. The nucleus is a complicated many-body quantum system. Our understanding of the diverse facets of nuclear structure is firmly based on a multitude of sophisticated techniques of nuclear spectroscopy, whereby the properties of many states in a wide range of nuclei have been investigated in detail. The comparison of these measured properties with theoretical predictions based on the best available models of nuclear structure is the means by which, in an iterative way, understanding of the field has been greatly expanded. With the ISAC facility the range of nuclei subject to investigation will be significantly expanded, thereby testing the predictive powers of the models. In addition, the availability of the wide variety of intense radiaoctive beams possible with the ISAC facility will provide unique opportunities in the fields of nuclear medicine and, in particular, materials science. This section describes briefly some aspects of the proposed experimental program at ISAC. The scope of this program is very broad and yet any subdivision results in topics which in some cases overlap significantly. In addition, any description of the program in 5 terms of a truncated list of topics inevitably omits aspects which it is now known will be important. Most significantly, at a facility which will provide so many new opportunities, it is the unexpected ones that may provide the most dramatic results. 2.2 Nuclear Astrophysics with Radioactive Beams 2.2.1 Introduction It is believed that the universe in which we live was created in a hot Big Bang which occurred some 15-19 billion years ago. From an initial singularity of pure energy, the universe has been expanding and cooling ever since. The initial "soup" of massless particles quickly devolved into photons, leptons and a quark-gluon plasma which soon condensed into neutrons and protons. As the universe continued to expand and cool, a brief period of rapid nucleosynthesis occurred and the free neutrons were bound into helium nuclei (and a trace of deuterium and lithium) before they could decay. After this era of nucleosynthesis no further element synthesis could occur 11ntil the coalescence of the products of primordial nucleosynthesis into stars provided the high temperature and density required for thermonuclear reactions to begin. The nebulae, galaxies, clouds of dust and gas, and stars that we observe today provide the material and sites for the synthesis of all of the elements, with the exception of helium, most of which was produced in the Big Bang. There are various processes by which elements heavier than helium are produced. They are illustrated in Fig. 2.1. Normal stellar evolution produces elements up to the iron region through fusion of nuclei in the stellar core, and heavier nuclei through successive neutron captures in the s-process. Stars somewhat more massive than the sun will end their evolution with the collapse of an unstable iron core into a neutron star or a black hole and, in so doing, will synthesize additional heavy elements through the r-process during the resultant supernova explosion. The remnants of stellar evolution (neutron stars, black holes and white dwarfs) can, in certain circumstances, accumulate hydrogen and helium on their surfaces. Explosive burning of this material can occur resulting in the rapid burning of the accumulated matter in the rp-process. In all of these explosive events reactions involving radioactive nuclei play a prominent role. Our knowledge of the universe has been considerably increased by the very strong links between nuclear physics and astrophysics. Pioneering work by Gamow, Bethe, Hoyle, Fowler and others provided the foundation for the field of nuclear astrophysics. In spite of very important progress in our understanding (recognized by Nobel prizes to Bethe in 1967 and to Fowler in 1983) major problems and puzzles remain. The theoretical and experimental study of these outstanding questions presents a major challenge to nuclear physics today. 6 82 r a Stable ~ Unstable 126 r process 82 50 Figure 2.1: Chart of the nuclides showing the paths for various nucleosynthesis processes. Nuclear astrophysics deals with two essential problems in our description of the universe. They are: 1. Energy production in astrophysical sites ranging from hydrostatic burning in stars such as our sun to spectacular explosive events, such as novae, supernovae, and X- or 1-ray emitters. 2. The production of the elements and their isotopes in their relative abundances as ob-served in our region of the universe by observation of the sun, moon and meteorites, or as observed in more distant regions through the study of the radiation emitted by stars. Nucleosynthesis began in the Big Bang and is continuing today during the life and death of stars. The processing and reprocessing of matter in various astrophys-ical sites can modify the observed abundances of selected nuclides. The observation of abundances at these sites can provide important information about the processes that are taking place. Nuclear physics and astrophysics play complementary roles in describing the properties of such sites. Input from research in both experimental and theoretical nuclear physics is required for the study of these problems. The large number of questions to be answered can be classified into two categories: 1) the determination of cross sections for nuclear reactions at energies (temperatures) relevant for diverse astrophysical sites; and 2) the determination of nuclear properties, such as mass, lifetime, decay modes, etc, for a large range of nuclei, including many at the extreme limits of stability. Our understanding of the universe is based upon observation. The traditional source of information is from the analysis of electromagnetic radiation at wavelengths ranging from radiofrequencies to the X- and 1-ray region. These radiations originate from a diversity of sources (galaxies, the interstellar medium, stars of various types and at various stages of evolution, including novae and supernovae), and even from the early universe. More recently, advances in optical astronomy and space technology have provided a wealth of new data at infrared, ultraviolet and X-and 1-ray wavelengths that increase dramatically the amount of information available from cosmic sources. The analysis of electromagnetic radiation from astrophysical sources provides information on the characteristics of these sources as well as on the temporal evolution of the universe and its various constituents. These observations are an essential source of information on the operation of nuclear reactions at these sites, and have permitted the construction of paths for stellar evolution, including energy production. Isotopic abundance information has tradi-tionally been derived from the analysis of optical spectra but, more recently, the discovery of 1-ray emission from the interstellar medium, nearby galaxies (such as the Large Magellanic Cloud), or from supernovae (such as SN1987a), has provided an important additional source. These abundance data provide additional constraints on the operation of nuclear reactions at astrophysical sites. Additional abundance data comes from analysis of the very small amount of matter directly accessible to study: the planets and meteoritic material of our solar system; and cosmic rays, energetic particles mostly of galactic origin. The discovery that some minute 8 inclusions in otherwise ordinary meteorites have anomalous isotopic composition has been a source of intense interest in recent years. These inclusions are believed to be relic material from the formation of our solar system and raise important questions about the nature of the astrophysical processes that were operational in this region of the galaxy at the time of the birth of the sun. 2.2.2 Nuclear reactions of astrophysical interest Most nucleosynthetic reactions occur in hot dense astrophysical sites in which trace amounts of nuclei with masses greater than that of helium are in thermal equilibrium with large amounts of hydrogen or helium, or both. The rate at which protons or a particles are captured by a heavier nucleus is determined essentially by the overlap of the tail of the Maxwell-Boltzmann distribution of relative energy, which decreases exponentially with in-creasing energy, and the probability for the reaction to occur, which increases only slowly with energy because of the Coulomb repulsion between the charged particles. In hydrostatic burning temperatures are relatively low and , except for the very lightest nuclei, even for stable nuclei the reaction cannot be studied in the laboratory at the very low energies at which the reaction occurs in stars. The reliable extrapolation of measured cross sections to these lower energies relies on careful experimentation combined with the best available the-oretical description of the reaction process. In explosive burning scenarios, where reactions involving unstable nuclei are of great importance, temperatures are much higher and, at the correspondingly higher interaction energies, cross sections are much larger. This makes it possible to contemplate using beams of radioactive nuclei to study such reactions in the laboratory. Until recently, detailed calculations involving unstable nuclei in explosive modelling have relied heavily on theoretical estimates of the rates of the many nuclear reactions taking place. A few long-lived nuclei , such as 7Be, 22 Na and 26 AI, have become available as targets. Also, indirect determination of the properties of the nuclear states of many radioactive nuclei has been accomplished through nuclear reaction or nuclear decay studies. This nuclear information can then be used to provide more reliable calculations for the rates of reactions involving these nuclei. Although valuable contributions have been, and will continue to be, made by these indirect studies, it is obviously preferable to determine the rate for a reaction directly where this is possible. Direct measurement of rates for nuclear reactions involving unstable nuclei is the new frontier of research in experimental nuclear astrophysics. The first step in cross section measurement using an accelerated radioactive beam was taken at Louvain-la-Neuve with the study of the 13N(p,1)140 reaction (see Section 2.2.6). In the following section we shall investigate selected reactions of importance in several astophysical processes and indicate how these can be investigated using accelerated radioactive beams. 9 2.2.3 Hydrogen burning The nuclear products of the Big Bang were primarily hydrogen and helium. Any study of the evolution of the universe and of its components requires an accurate modelling of the formation and evolution of stars composed initially mainly of hydrogen and helium, and with initial masses up to 100 M0 or more. The evolution of the stellar core begins with the fusion of hydrogen into helium through the pp chain in the lightest stars. As stellar mass increases, the CNO cycle becomes a more and more important secondary source of energy release and nucleosynthesis in hydrogen burning. In more massive, second generation stars some hydrogen may also be consumed in NeNa and MgAl cycles. These cycles are not significant for energy production, but do influence abundances of isotopes in this mass range. More importantly, the MgAl cycle might contribute to the production of 26 AI, a radionuclide of great ntrrent interest in 1 -ray astronomy and cosmochf'mistry. 2.2.4 The solar neutrino problem and 7Be(p,, )8 B In the observation of the capture of solar neutrinos on 37Cl only about one quarter of the expected flux has been found [1, 2]. This missing solar neutrino flux has constituted the solar neutrino problem [1] which has been with us for more than 20 years. The neutrinos detected by the 37Cl detector are of relatively high energy and result largely from the ,8-decay of 8B which is in a weak branch of the pp chain [1] through which hydrogen is converted into helium. More recent measurements at the Kamiokande light water detector have shown a discrepancy of about one half of the expected flux of even higher-energy neutrinos that are also associated primarily with the ,8-decay of 8 B. Recent measurements of the flux of the low-energy neutrinos coming from the p(p,e+v)d reaction in the SAGE [3] and GALLEX [4] experiments (measured by neutrino capture on 71 Ga) lead to fluxes which are about 60-80% of those expected. Because of these energy-dependent deficiencies, the solar neutrino problem is very difficult to resolve and has led consequently, among other ideas, to speculations about the nature of neutrinos. In particular eigenmasses and matter-induced resonances (the MSW effect), and other new physics beyond the standard model of particle physics, have been proposed. The Kamiokande and 37 Cl experiments as well as several others approved and proposed (including the Sudbury Neutrino observatory (SNO) now under construction) are sensitive primarily to the 8 B neutrino flux from the sun. This neutrino flux results, in the standard solar model, from a weak branch in the pp chain p(p,e+v)d, d(p,{)3He via 3He(a,,fBe leading to 7Be(p,{)8 B. The 7Be(p,{)8B reaction rate has been determined experimentally with radioactive targets and via Coulomb dissociation, but there is a considerable variation in the values of the astrophysical S-factor at zero energy ( S17(0)) deduced from the various experiments [5]-[12]. Because the value of S17(0) remains open to discussion, and because of the importance of this value for the interpretation of the results of present and proposed solar neutrino experiments, it is proposed that the cross section for the 7Be(p, 1 )8B reaction be measured in inverse kinematics with a 7Be beam from ISAC. This approach allows for the determination of the cross section relative to the well-known Rutherford cross section, 10 which avoids the normalization problem associated with 7 Be target measurements [13]. The radioactive isotope 7 Be is produced in large quantities at TRIUMF, often as an unwelcome byproduct. During each running period on beamline 2C samples of 7 Be of the order of 1017 to 1018 atoms could be obtained. Alternatively, the production of large amounts of 7Be using the 13 MeV TR13 proton cyclotron via 7Li(p,nfBe is possible. An experiment has been proposed at TRIUMF (E730) whose objective is a measurement of the p(7 Be,{ )8B reaction with a precision of about 5% in the combined statistical and systematic errors. The experiment is designed to be simple, but with as much cross checking as possible. The basic idea is to implant recoiling 8B particles (together with the 7Be beam) into an adequate substrate on a tape, move the collected particles to a counter, and detect the f3 and a particles from the decay of 8B. During implantation, the recoiling elastic protons can be measured within the forward hemisphere of the beam. A heavier nucleus (e.g. 132Xe) may be used in the target to measure all cross sections relative to the well-known Rutherford cross section, as well as to determine contaminations in the beam. Initial estimates show that with a beam current of 10 nA of 7Be and 100 h of run time, cross sections down to 25 nb for the p(7Be,{)8B reaction can be measured. This is the cross section at Ecm ~ 200 keY. The yield would be about 430 events for a 1018 atoms/cm2 H2 target assuming a conservative 30% detection efficiency. This would require a total integrated beam of 5.8 x 1016 7Be ions at the lowest attainable energy point. (The calculated yield into 471" is 1 event/(nb x 1018 atoms/cm2 H2 x 1015 atoms 7Be).) The experiment will require abw lute measurements at a only few high-cross section points where a relatively thin target is required; measurement of the excitation function can be carried out with a relatively thick target (up to 1019 /cm2). A complete description of this experiment can be found in the detailed research proposal for experiment E730 which has been approved with high priority by the TRIUMF Experiment Evaluation Committee. 2.2.5 The hot pp chain The Big Bang produced an extremely small fraction, by mass, of isotopes beyond helium. As a consequence, the first stellar objects were likely very massive (106 M0 ) [14]. Because of the near absence of CNO isotopes, the CNO cycle, even in its "hot" form (see Section 2.2.6), would not stabilize such objects. However, a hot version of the pp chain could operate, which would lead to a gradual build-up of CNO material [15]. This burning mode could also develop in nova explosions resulting from the accretion of material from a companion star onto the surface of a white dwarf. Some of the key reactions in the hot pp chain are listed in Table 2.1. 2.2.6 The hot CNO cycle and explosive nucleosynthesis At temperatures between 7x107 and 3x108 K, the CNO cycle starts to include proton capture reactions on radioactive isotopes such as 13N, 17F and 18F. The hot CNO cycle is displayed in Fig. 2.2. The onset of the hot CNO cycle occurs when the rate for the 11 Table 2.1: Some reactions important for the hot pp cycle Reaction t~~~m (s) t product ( S) 1/2 7Be( a, 1 )11 C 4.61 X 106 1223 8B(a,p)11 C 0.770 1223 sB(p,I)9C 0.770 0.127 9C( a,p )12N 0.127 0.011 11C(p,1 )12N 1223 0.011 13N (p, 1 )140 raction exceeds the rate for ,8-decay of 13N. In recent years the rate for this critical reaction has been fairly-well determined through particle-transfer reactions, resonant capture of 13N in hydrogen targets at Louvain-la-Neuve [17, 18], and Coulomb dissociation of high-energy 140 beams in the field of a heavy nucleus [19]. Sites for this cycle include novae and the outer shells of supernovae of type II (SNII) which are the result of collapse of the cores of massive stars. At temperatures T8=2-3 (T8 = 108 K) numerous proton capture reactions in the A=20-40 mass region are also taking place [20]. For even higher temperatures the CNO cycle opens up to masses of A=20 and beyond via reactions on radioactive isotopes like 150( a,1 )19Ne and 18Ne( a,p )21 Na. Simultaneously, additional proton capture reactions on higher-mass isotopes up to 56Ni start to occur with a net flow of isotopes into this mass region and beyond (depending on temperature). Possible stellar sites of this reaction chain (known as the rp-process [21]) are in thermal runaways of hydrogen burning occurring on the surface of mass-accreting white dwarfs or neutron stars in close binary systems. Figure 2.3 shows the reaction network involved in the rp-process for a high temperature case (X-ray burst). The reaction rates during explosive hydrogen burning in these networks must be known, if we are to reliably predict the production of long-lived isotopes such as 22 Na, 26 Al, 44Ti, 54Mn, 56Co, 57Co, and 65Zn. On the basis of large-scale network calculations, which include stellar evolution, it is possible to make predictions for the 1-ray flux from the decay of these isotopes. In the rp-process and the silicon burning stage of SNII's, many radioactive isotopes of mass greater than A=40 at the N =Z line are involved. The reaction flow proceeds to 56Ni and partially beyond this isotope to masses up to A=100. Capture reactions for these isotopes can often be estimated by statistical Hauser-Feshbach calculations, but some experimental checks are essential in determining the mass fluxes and the N and Z distribution of the seeds to the r-process. Some key reactions to be determined directly in the context of explosive nucleosynthesis are listed in Table 2.2. 12 Onset of rp-Process (p,"tl + ~· ' , Figure 2.2: The hot CNO cycle network of nuclear reactions including the break-out path for the rp-process (from Ref. [16]). Example: 13N(p,,) 140 A knowledge of the reaction rate for the 13N(p, 1 )140 reaction for temperatures up to 109 K (T9 = 1) is vital for understanding hydrogen burning in the hot CNO cycle and the conditions under which breakout into the rp-process may occur. Proton capture on 13N should be dominated at low energies by a resonance at 526 keV due to the Ex = 5.173 MeV 1-, T = 1 first excited state in 140. In addition, a direct capture contribution is also expected. The resonance parameters have been fairly well determined by a combination of particle-transfer reactions, resonant capture of 13N in hydrogen targets at Louvain-la-Neuve [17, 18], and Coulomb dissociation of high-energy 14 0 beams in the field of a heavy nucleus [19]. The non-resonant component of the cross section has been calculated by several groups, either separately or as part of a calculation of the total cross section. References to theo-retical studies can be found in Refs. [17, 18, 19]. There are differences among the various 13 Zn Cu (p,y) ~ (~·.,) Se As Ge GO 61 c a 60 56 s~\ ~e, 59, I ~s s'a s'7 s'e' • [/ 'Y Sr Rb Kr 6!) Br 6~ 66, &.7, &'5 '_1_, G6 ' ~ ' G2, &3, ,'/ G1r, [/ -, • G1t 6~" ,;'3 Gl-t ' [/ G1o" ,-, 61 62 63 ~19 GO 61 G2 / y .,{ 1/ ?9 EJ7 .. 7/ 7!1 .,., 78 1/ ~ 73 /4 7~ 7G 77 7{ 70, ~~, 73 74 7!1 76 / .-, • r.'u ~o' 711 712 73 74 7S ' / ,'a' 69 70 71 72 ?:J 74 &'7 60 (i!) 70 71 72 7:\ li6 67 GU ti9 70 71 7Z (i5 6G G7 68 (i9 70 71 6'1 65 GG G71 GB 69 70 63 G'l G5 GG 67 Gn (j!) Figure 2.3: The rp-process network of nuclear reactions for a temperature of T=l.5 x 109 K, p=106 g/cm3 • calculations, as is apparent from Table 2.3, both in the magnitude of the S-factor and in its energy dependence. Breakout from the CNO cycle via the 13N(p,/) 140 raction begins at about T8 = 1. The peak temperature for a nova outburst is near T8 = 2.4 [21]. It is shown in Ref. [22] that the cross section that determines the reaction rate at this temperature is uncertain by about ±30 %, due mainly to the uncertainty in the direct capture cross section. With a thin gas-jet target (1018 atoms/cm2) and an estimated peak cross section of 330 p;b, a beam of 1010 ions/son target, and a detection efficiency of 10% (or better), the shape of the yield curve could be measured over the region from 400 to 1000 ke V within a reasonable time(~ 100 counts per day at the extreme energies). Details are given in Ref. [22]. In the pioneering experiment at Louvain-la-Neuve, 13N beams of up to 4 x 108 particles/s were available [18]. The upgraded facility ARENAS3 at Louvain-la-Neuve (see Section 3.5.3) will have a beam intensity of 1.6 x 1010 particles/s (26], while the projected beam intensity for ISAC is 1.9 x 1011 particles/s (see Table 3.1 in Section 3.4). Thus, the yield curve should be measurable over this extended energy range even if a detection efficiency lower than 10% 14 Table 2.2: Some reactions important for explosive nucleosynthesis Reaction t~j~m (s) tproduct ( ) 1/2 s 13N(p,1 )t4Q . 598 70.6 1sO(a, 1 )19Ne 122 17.2 18f(p,a )1so 6585 122 18f(p,1 )19Ne 6585 17.2 19N e(p, 1 )20N a 17.2 0.448 21 N a(p, 1 )22Mg 22.5 3.86 31 S(p, 1 )32Cl 2.57 0.298 39Ca(p,1 )40Sc 0.86 0.182 43Sc(p,a )4°Ca 1.40x 104 stable ssNi(p,1 )s6cu 0.21 ? s6Ni(p,1 )s7cu 5.25x105 0.223 Table 2.3: S-factors in keV·b from theoretical direct capture calculations E(keV) Ref. [17] Ref. [23, 24] Ref. [25] 0 0.344 0.818 -100 0.324 0.733 0.612 200 0.305 0.658 0.571 300 0.287 0.590 0.529 400 0.270 0.529 0.488 500 0.254 0.475 0.448 600 0.239 0.426 0.410 700 0.225 0.382 0.374 800 0.212 0.342 0.341 900 0.200 0.307 0.312 1000 0.188 0.276 0.287 15 is available initially. Example: U>o(a,1) 19Ne When temperatures of about 108 K (T 8 = 1) are exceeded in stellar explosions, mass flow and energy production in the hot CNO cycle become restricted by the ,8-decay rates of 140 and 150 with half-lives of 70.6 s and 2.03 min, respectively. As a result, most of the mass previously residing in the stable CNO isotopes is transformed to 150 and, to a lesser extent, 140. As helium is always present in hydrogen-rich matter (from primordial nucleosynthesis and previous hydrogen burning), a capture on 150 can occur, followed by a rapid 19Ne(p,1) 20Na reaction. This leads into the A 2: 20 mass region from which there is no back flow of material into the CNO region and signals the beginning of the rp-process (see Fig. 2.2). Our present nuclear knowledge suggests that the mass flow into heavier isotopes, and thus the subsequent energy production, is controlled by the 150( a, 1 )19Ne reaction. In the relevant temperature region near T 8 = 3, the reaction rate is expected to be dominated by a resonance at Ecm = 504 keY, with resonances at 851 and 1076 keY becoming important only at higher temperatures [27]. Because of low Coulomb penetrability for the a particles, these resonances are narrow and their widths are dominated by the 1 width so that the reaction rate is simply proportional to the a width (fa). Since the ,8-decay rate is independent of stellar temperature while the a-capture reaction rate depends strongly on temperature, the reaction rate dependence defines a line on a stellar p-T phase diagram where the rates are equal. In order to determine the position of such a line for a given resonance, and to estimate the yield in an a-capture experiment, the value of r a must be known. For the 504 ke V resonance, we can take r a = 9 J.Le V based on the 15N(6Li,d)19F analog transfer reaction [28]. For the 851 keY resonance no value has been estimated; an arbitrary value of r a = 0.5 meV is taken. For the 1076 keY resonance we take fa = 24.5 meV [29, 30]. Since novae are typically in the temperature range T8 =2-3.5 and density range p = 102-104 g/cm2 , only the 504 keY resonance is of any importance in controlling break-out from the CNO process into the rp-process in novae. These same widths can be used to estimate reaction yields in a radioactive beam exper-iment. For a pure helium target and an 150 beam of 1011 ions/s, we could expect a yield of 1 event/hr for the 504 keY resonance, with an 150 beam energy of 0.154 MeV ju in the laboratory. The other two resonances are at 0.226 and 0.287 MeV /u and would yield 35 and 740 events/hr, respectively. Clearly, the higher-energy resonances would have to be measured first. This experiment illustrates the necessity for intense beams at relatively low energy from ISAC for the study of some reactions of astrophysical interest. Examples: 19Ne(p,1) 20Na, 18Ne(a,p)21Na The 19Ne(p,1) 20Na reaction, which is in the break-out path of the hot CNO cycle and therefore crucial in the determination of its total energy output, is dominated in the astro-physically interesting energy region by a state at Ex=L6 MeV in 20Na with a centre-of-mass resonance energy of 0.45 MeV in the 19Ne+p system. The details of nuclear structure indicate that the state has been assigned a J7r=1+ by transfer reactions [31, 32]. In addition, below 16 the resonance, direct capture contributions, which are likely to give about one hundredth of the yield, are significant. However, these can be measured above the resonance where the direct capture yield is higher. Recent .8-delayed proton decay experiments on 20Mg [33] show no evidence of a proton decay of this resonance state, which puts the spin/parity assignment for this state into question. Therefore, a spin/parity assignment of J1r =3+ has been proposed [34]leading to a higher resonance strength than for the 1 + assignment. However, a recent measurement at the Louvain-la-Neuve radioactive beams facility sets upper limits which are below what was predicted for the 3+ case [35]. By bombarding a hydrogen gas target (see Section 2.2.8) with a 8.55 MeV 19Ne beam of 1010 /s in inverse kinematics this resonance can be observed directly. The integrated resonance strength W/ has been estimated to be 6 meV [32] resulting in yields of about 400 events per hour into 47r. With a detector of 10% efficiency it would be feasible to detect such a yield in a short time and progress to measurement of higher-lying resonances and the direct capture component of the 19Ne(p,1)20Na reaction. The 18Ne(a,p) 21 Na reaction, which is part of a different break-out path from the hot CNO cycle, has a high Q-value of Q=8.14 MeV with the proton channel open at Ex=5.50 MeV in 22Mg. A state at Ex=8.55 MeV is known in 22Mg, but is not likely to be the only one in the energy region of interest. Because this reaction has particle channels for the ingoing and outgoing wave, its yield could be higher than that of the competing 150(a,1)19Ne reaction. Several proton groups are emitted in the decay of 22 Mg corresponding to excited states in 21 N a. However, those going to low-lying states will dominate because of the Coulomb barrier in the break-up of 22 Mg. These groups can be easily identified in an array of silicon surface barrier detectors. For proton groups to higher-lying states the detection of the recoil nucleus is possible, as little momentum is transferred from the proton to the recoiling 21 Na nucleus. 2.2. 7 Production of heavy elements Elements with masses greater A=60 are produced largely by neutron capture on lighter seed nuclei. Two reaction processes have been identified, the slow neutron capture process (s-process) and the rapid neutron capture process (r-process) proceeding at quite different stellar sites. Both processes produce about the same number of isotopes. For the production of the far less abundant proton rich isotopes, which cannot be reached by either process, the so-called p-process is responsible. This process is a non equilibrium photodisintegration and rearrangement of the seeds- and r-process nuclei. The s-process Although the s-process proceeds,along the line of nuclear stability, there are some long-lived isotopes where a branching takes place depending on the half-life and neutron capture rate of the specific isotope (e.g. 155Eu, t 1t 2=4.94 y ). While neutron capture cross sections on stable isotopes are known to about 4%, capture rates on unstable isotopes are estimated with errors assumed to be 30%. These critical isotopes carry therefore the biggest uncertainties in calculations of the s-process abundances [36]. The high-yield isotope separator proposed for 17 ISAC is ideally suited to produce targets of long-lived isotopes for neutron capture measure-ments. Typically isotopic fluxes of greater than 109 fs are required for reasonable collection times of days (1014 atoms). The r-process In the r-process isotopes are formed very far away from the line of stability in a high neutron flux environment (probably the hot neutrino bubble of an SNII [37]). If a thermal equilibrium situation is reached, the abundances of isotopes depend simply on neutron separation energies (nuclear masses) and half-lives [38]. A major experimental effort has gone into determining these masses and half-lives. The ISAC separator, with its proposed very-high yields, should assist in extending the measurement of masses farther from the valley of stability. For somewhat lower fluxes and during freeze-out from the process, neutron capture rates and specific separation energies near N =50, 82, and 126 become even more important because of shell effects. The determination of neutron capture data could be carried out via transfer reactions in inverse kinematics. Such measurements require the development of higher initial charge states from the separator to increase the mass range of ISAC. In addition, during freeze-out isotopic concentrations are redistributed somewhat by ,8-delayed neutron emission [39], which can be investigated experimentally with ISAC. The p-process The light, proton rich stable isotopes of many elements with masses A=60-200 are shielded from both the s- and r-process. Though the existence of the lightest of these isotopes can be explained by o- and p-capture from explosive burning, the cosmic site for the higher mass isotopes is not firmly established. Nuclear cross sections in this mass region are largely un-known and must be approximated by statistical methods (Hauser-Feshbach). Reaction cross sections and nuclear properties such as binding energies and half-lives could be determined with the ISAC high-flux isotope separator. 2.2.8 Experimental equipment for measuring capture reactions Proton-capture reactions of astrophysical interest produce 1 event per 1010 to 1012 incident beam particles on targets of 1018 nuclei/cm2• Isobaric stable beams of higher intensities than the beam of interest may also be incident on the target. For this reason a very sensitive detection system with high detection efficiency is required. A {-ray, recoil-particle detection system is proposed for the ISAC experimental area. Such a detector would be useful also for the study of other reactions (see, for example, Section 2.4.3). References [1] J.N. Bahcall and M. Pinsonneault, Rev. Mod. Phys. 64 (1992) 885. 18 [2] N. Hata, preprint in Proceedings of the Solar Modeling Workshop, Seattle, March, 1994. [3] P. Anselmann et al., Phys. Lett. B314 (1993) 445. [4] J .N. Abdurashitov et al., Phys. Lett. B328 (1994) 234. [5] C.W. Johnson et al., Ap. J . 392 (1992) 320. [6] A. Dar and G. Shaviv, preprint, 1994. [7] B.W. Fillipone et al., Phys. Rev. C28 (1983) 2222. [8] T. Motobayashi et al., Phys. Rev. Lett. 73 (1994) 2680. [9] K. Langanke and T.D. Shoppa, Phys. Rev. C49 (1994) R1771. [10] P. Descouvement and D.Baye, Nucl. Phys. A567 (1992) [11] A. Csoto et al., preprint, 1994. [12] F.C. Barker, Nucl. Phys. A588 (1995) 693. [13] F.C. Barker and R.H. Spear, Ap. J. 307 (1986) 847. [14] G.M. Fuller, S.E. Woosley and T.A. Weaver, Ap. J. 307 (1986) 675. [15] M. Wiescher et al., Ap. J. 343 (1989) 352. [16] S. Kubono, Comments Astrophys. 16 (1993) 287. [17] P. Decrock et al., Phys. Rev. C48 (1993) 2057. [18] Th. Delbar et al., Phys. Rev. C48 (1993) 3088. [19] G. Bauer and H. Rebel, J. Phys. G: Nucl. Part. Phys. 20 (1994) 1. [20] L. van Wormer et al., 1993, preprint. [21] R.K. Wallace and S.E. Woosley, Ap. J. Suppl. 45 (1981) 389. [22] J.D. King, in Proceedings of the /SAC Workshop, Lake Louise, Alberta, February, 1994, TRIUMF report, to be published. [23] P.B. Fernandez et al., Phys. Rev. C40 (1989) 1887. [24] P.V. Magnus, E.G. Adelberger and A. Garda, Phys. Rev. C49 {1994) R1755. [25] C. Iliadis, private communication. [26] Nuclear Physics European Collaborative Committee (NUPECC), Report of Study Group, European Radioactive Beam Facilities, 1993. 19 [27] P.V. Magnus et al., Nucl. Phys. A506 (1990) 332. [28] Z.Q. Mao, H.T. :Fortune and A.G. Lacaze, Phys. Rev. Lett. 74 (1995) 3760. [29] P.V . Magnus et al., Nucl. Phys . A470 (1987) 206. [30] K. Langanke et al., Ap. J. 301 (1986) 629. [31] S. Kubono et a/., in Proceedings of the First Conference on Radioactive Beams, eds. W.D. Myers, J.M. Nitschke and E.B. Norman, World Scientific, 1990, p. 220. [32] L.O. Lamm et al., Nucl. Phys. A510 (1990) 503. [33] J. Gorres et al., Phys. Rev. C46 (1992) R833. [34] B. A. Brown et al., Phys. Rev. C48 (1993) 1456. [35] M. Wiescher, private communication. [36] K. Wisshak et al., in Proceedings of the Second Symposium on Nuclei in the Kosmos, eds. F. Kappler and K. Wisshak, Institute of Physics, Bristol, 1993, p. 203. [37] K. Takahashi et al., in Proceedings of the Conference on Nuclei far from Stability, Institute of Physics, Bristol, 1993, p. 839. [38] W.A. Fowler, G.E. Caughlan and B.A. Zimmermann, Ann. Rev. Astron. Ap. 5 (1967) 525. [39] K.L. Kratz et al. , J . Pbys . G: Nucl. Part. Phys. Suppl. 14 (1988) S331. 2.3 Fundamental Symmetry Tests With Trapped Neu-tral Atoms It is a fashionable belief held by many, especially particle physicists, that, although semilep-tonic decay data in complex nuclei have played an important role in the historic development of the theory of the electroweak interaction, all the important "action" has now shifted to the high-energy colliders. We propose tests of the Standard Model (SM) in complex nuclei produced by the ISAC facility which are complementary to and, in specific areas, competitive with those performed at higher energies. The low-energy experiments can be carried out at relatively low cost, and they benefit from high statistical accuracy and the multiplicity of atomic and nuclear quantum states. Laser trapping of neutral alkali atoms has been perhaps the most-rapidly-advancing sub-field in atomic physics during the past decade, with the promise of revolutionizing other fields of physics (ultra-high precision atomic clocks, Bose-Einstein condensate, isotope separation, quantum noise reduction, precision measurements of special relativity, etc.). We propose to study symmetry properties of the Standard Model using isotopically pure atoms confined in 20 space by magneto-optical forces. Our intent is to couple the copious production of radioactive alkali atoms from TISOL with these recent developments in laser and magnetostatic, neutral-atom-trapping technology to produce dense, cold, highly-polarized samples. This technology holds the promise of revolutionizing precision measurements of the weak interaction , both the weak neutral current via measurements of atomic parity violation on radioactive (hence, scarce) atoms and the weak charged current, by making a new class of fi-decay experiments possible. We propose to study symmetry properties of the Standard Model (SM) using isotopically-pure atoms confined in space by magneto-optical forces . Extensions of the SM might show up at low energies if experiments of sufficiently high accuracy (typically a few times 10-3 for non-vanishing observables) can be carried out. For some searches it may be an advantage to be at low energy since the interference amplitudes for established (W, zo) physics and new physics are both real; this is not the case for experiments near the zo pole. Furthermore, the multiplicity of well-defined nuclear quantum states may allow one to select a suitable J'-decay transition which allows the most sensitive test of a specific extension of the SM. Nuclear uncertainties which might obscure the new physics of interest can often be controlled to sufficient accuracy. In J1 decay one can select fast transitions to keep the influence of nuclear matrix elements of recoil order at a negligible level [1] . All the J'-decay experiments contemplated for ISAC are thus of the allowed or super-allowed type. At an accuracy level below 1%, anticipated in the future for studies of parity-nonconservation (PNC) in atoms, knowledge of the distributions of u- and d-quarks (or protons and neutrons) over the nuclear volume becomes important [2, 3]. These nuclear effects can in principle be tested independently by performing PNC experiments with a series of isotopes. 2.3.1 The potential of neutral atom traps The upgraded isotope separator proposed for the ISAC facility is potentially a very bright source of mass-separated alkali atoms. In a test experiment with TISOL in early 1994 a 7 g/ cm2 uranium target and a surface ion source were used to measure the production rates of a range of Fr isotopes from 203Fr to 221 Fr. The yields are shown in Fig. 2.4. They show a peak of the order of 108 Fr atoms/sf pA of beam protons. Isotopes with half-lives down to 5 ms were detected. With improved beam optics and shielding of the isotope production region thicker targets and higher beam currents can be used. Production rates of 1010 atoms/s can then be envisaged for half-lives exceeding 1 m (see Fig. 2.4 and Table 3.1) . The very high production rates of alkali atoms can be exploited by coupling the isotope separator to a neutral atom trap. The combination of a very thin carbon degrader foil followed by a thin, heated (Ta/Y) catcher tube is one of several feasible neutralization schemes. The principle of laser trapping is shown schematically in Fig. 2.5. Laser light with its wavelength detuned to the red side of a resonance is absorbed preferentially by atoms moving against the light. Since re-emitted light is multidirectional, the net impulse from each absorption can be used to slow the atom's momentum by "' 2 x 10-5 of the average momentum at room temperature. For initial cooling of the low-velocity tail of the Maxwellian distribution, the magneto-optical or Zeeman-optical trap (ZOT) [4] has 21 Francium Yields 10 1 ~~~~~~~~~~~+ 200 205 210 215 220 225 A Figure 2.4: Yield of francium isotopes measured recently with TISOL. The gap for A=215-217 is caused by sub-ms half-lives. The yields were severely limited by the target thickness of 7 gjcm2 because of insufficient shielding around the target. been effective to obtain temperatures of typically 200 J.LK. Magnetic field gradients, and six counterpropagating light beams of opposite circular polarization, produce a net. force directed towards the trap centre (see Fig. 2.5). The Zeeman splitting implies that u+ light with = +1 (u- light with = -1) is preferentially absorbed on the left (right) side of the trap center. (Note that the helicity of the opposing light beams is the same, whereas B changes sign.) This results in a net restoring force directed towards the trap center. A quantitative model to describe the dependence of the fraction of trapped atoms on laser power and beam diameter has been developed by Lindquist, Stephens and Wieman [5]. The trapping efficiency can be increased greatly when atoms are allowed to bounce many times from the walls. This requires wall coatings with short sticking times ( < 100 J.LS), low outgassing rates, and durability in the presence of intense ionizing radiation [6]. A second step in neutral atom cooling, which may not prove necessary for some applica-tions, is "optical molasses" [7], again using six red-shifted light beams, but no magnetic field. This works so well that the low temperatures achieved ("' 1J.LK) require a rather exotic expla-nation [8]. By optical pumping of a particular hyperfine transition with circularly polarized light from a single laser, all the population can be transferred to a single hyperfine state of maximum F corresponding to almost complete atomic and nuclear polarization. Finally, the cooled atoms can be held in space without lasers by utilizing magnetic forces on the atom's 22 10 J=O -J=1 --v force-B(O , I D ,' , , , ... 'B , -- -v force c 0 B•O --.. force B>O Figure 2.5: Schematic of a one-dimensional Zeeman optical trap for neutral atoms (top) and realization in three dimensions (bottom). magnetic moment to produce the magnetic equivalents of either a (static) Penning (9] or a (dynamic) Paul trap (10]. For states with maximum F complete nuclear and atomic polarization can be achieved (although the existence of a field gradient and the small, but finite size of the trapped volume implies that all spins are not maximally aligned in space). To achieve close to 100% population of a particular hyperfine state (maximum nuclear polarization) we must limit the number of trapped atoms to about 1010 atoms with limitations imposed by reabsorption of scattered, unpolarized photons (radiation trapping) [11]. 2.3.2 PNC in francium isotopes The main interaction in atoms is the electromagnetic interaction between the electrons and the nucleus, which can be thought of as the exchange of virtual photons. The unification of electromagnetism and the weak force in the Standard Model implies that virtual zo's are also exchanged between the electron and the nucleus. Since the weak interaction violates parity, zo exchange mixes atomic states of different parity, which allows electromagnetic transitions between states that are otherwise parity forbidden. The first PN C experiments in atoms could have provided early confirmation of the Glashow-Weinberg-Salam theory of the electroweak interaction via its prediction of these 23 "weak neutral currents" , yet the results were incorrect (see the review in Ref. [12]). The PNC situation has changed dramatically during the past decade, with the latest result for 133Cs [13, 14) competitive with recent LEP results [15). The calculations of atomic matrix elements needed to extract the weak charge are now at the 1% level for alkali atoms such as Cs [16) and further improvements may be possible in the near future. With the improved luminosity of neutral atom traps compared with the atomic beam method, more accurate results on more than one isotope may become available in the future. PNC in atoms arises in the SM from ,_zo interference and is sensitive to new heavy particles that affect the radiative corrections. At the tree level (i.e. without radiative corrections) the weak charge Q!:;ee = -N + Z(1- 4sin20w) determines the weak mixing angle Ow. Since we know the factor (1 - 4 sin20w) to be small (e.g. by inferring sin20w from the mass of the W and zo) the radiative corrections for PNC (about 6%) are largely independent of the weak mixing angle. If one parametrizes radiative corrections involving the exchange of particles heavier than the W and zo according to the parameters S and T of Peskin and Takeuchi [17), Marciano and Rosner find [18) that atomic PNC is primarily sensitive to S, the weak isospin-conserving "new physics". For Cs Qwe33C s) = -73.20 - o.8S - o.oo5T. In Fr the predicted value of the weak charge depends on the extension to the SM in a form very similar to that for 133Cs and Qw(87+N Fr) = -128.22- 1.255- 0.003T + 0.986(136- N)(1 + 0.008T). This insensitivity to T is in contrast to high energy experiments (e.g. at LEP /LEPII) which always involve linear combinations of S and T (T is related to a comparison of neutral and charged currents at q2 = 0, whereas S involves the q2 dependence of the neutral current self-energy). Thus PNC in atoms tests "new physics" via radiative corrections in a way that is orthogonal to LEP or SLC physics (see Refs. [15, 18]). Atomic PNC is also sensitive to potential new physics that does not couple to the vector bosons. The LEP and SLC data, although extremely sensitive to interactions which directly affect the zo properties, are rather indifferent to new interactions which are 90° out of phase with the zo amplitude at the pole. Among the latter are extra Z' bosons with small zo_ Z' mixing [19), leptoquarks (new gauge bosons predicted by grand unified theories) [20), and new four-Fermi contact terms generated by models with composite fermions [21]. It was shown by Langacker [21] that PNC results of 1% accuracy can provide limits on these interactions which are significantly better than those obtainable at LEP, SLC or HERA. For example, HERA can hope to probe the fermion compositeness scale up to A = 5 TeV whereas the corresponding limit from a 1% PNC measurement for Cs or Fr is A= 22 TeV. For leptoquarks with strong (electromagnetic) couplings the mass limits are about 0.3 TeV from HERA and 0.9 TeV from 1% PNC data. The unique properties of trapped neutral atoms, i.e. compact size of 1 mm, high density (up to 1010 atoms), and nearly complete polarization, are expected to have a dramatic impact on future atomic PNC measurements. Of the two techniques used to detect atomic PNC 24 (optical rotation, or highly forbidden nS-+(n+ 1 )S transitions with deliberately introduced St'irk mixing) only tlw latter is suitahlf' for measurements on relatively few isotopic atoms cooled in ZOT's. It is necessary to interrogate a large column density of atoms with laser light at the parity-forbidden transition frequency in a power buildup cavity; the present experiments are limited by statistical accuracy. The Cs PNC experiments at Colorado [13] have achieved an experimental error of 2% using an atomic beam with a typical size of 0.5x2.5 cm2, a thermal velocity of 2.7x104 cm/s, and an intensity of 1014 jcm2js. This implies a target thickness of about 2x109 Cs atomsjcm2• With stationary trapped atoms, target densities more than two orders of magnitude larger might be achievable for stable isotopes. The challenge then is to take full advantage of the trapping technologies by fully loading the traps with radioactives, with a goal of 109 atoms trapped in a 1 mm volume. With demonstrated on-line trapping efficiencies for radioactive species of 10-4 to 10-s and trapping lifetimes of 20 s achieved at Berkeley and SUNY Stony Brook [22, 23), the projected intensity of 5 x 1010 21°Fr atoms/s for ISAC (see Table 3.1) would enable 107 to 108 atoms to be trapped; improvements in on-line trapping efficiency are being actively pursued in the community, with a goal of achieving the 5% (40%) achieved (extrapolated) by Stephens and Wieman in an off-line stable Cs trap [6). In Fr the parity-violating effect has been predicted to be 18 times larger than in Cs [24), allowing for a more precise test of the weak charge and the atomic physics calculations. The adaptation of neutral atom traps to the nS-+(n+ 1 )S transition/Stark mixing technique of measuring PNC in atoms is being actively pursued by Wieman's group at Boulder and in an experiment (E1303) at LAMPF on Cs isotopes [25]. The production, isotopic separation and trapping of large numbers of Cs or Fr atoms involving the widest possible range in N for a given Z would permit measurement of relative differences in Qw which are much less sensitive to the theoretical estimate of the atomic matrix element. When more accurate data on a series of isotopes become available, the measured variation of the weak charge will test the theoretical evaluation of < /sPN > provided one uses a sensible model of the neutron distribution in different isotopes [2, 3]. 2.3.3 Symmetry tests in nuclear f3 decay Parity violation was initially discovered in the weak interaction by measurements of the asymmetry in direction of emission of {3's with respect to the nuclear spin [26]; efforts continue to elucidate whether Nature is truly completely left-handed. The search for additional terms in {3 decay, beyond the well known vector and axial-vector couplings, sets interesting limits on non-Standard Model physics. For {3-decay studies, traps promise highly-polarized samples in a localized volume with virtually zero source thickness. An exciting development in trapping technology is the "Time-averaged Orbiting Potential" [TOP] trap recently developed at Boulder [27]. A magnetostatic quadrupole field is coupled with a uniform magnetic field rotating in the horizontal plane. The sample's polarization will adiabatically follow the rotating field, providing the possibility of a powerful tool for eliminating systematic effects in the {3 and recoil detection. Precision tests of the SM from weak interaction experiments in complex nuclei have been 25 discussed in an excellent recent review by Deutsch and Quin [1]. The nuclear ,8-decay probability has been formulated by Jackson, Treiman and Wyld [28] without assumptions with respect to parity, charge conjugation, and time reversal, and including the possibility of non-standard scalar and tensor couplings. The transition probablity dW can then be written as dW In this expression dW0~ contains the coupling strength, Coulomb and phase space factors, r = .j1 - ( aZ)2 is a Coulomb factor, < J > is the vector polarization of the parent nuclear state, iJ the ,B spin, p and E are momenta and energies of ,B or neutrino. In the SM the Fierz interference term band the T-violating parameters D and R vanish, whereas the ,B- v correlation a, the ,8- and v-asymmetry parameters A and B, respectively, ,B polarization G , and polarization asymmetry correlation Q' depend on angular momentum and on the ratio of matrix elements A = 9AMGT I gv MF. The parameters D and R measure the imaginary parts of products of coupling terms. A summary of the current status for all measurable quantities is given in Ref. [1]. Using the ISAC facility, we would concentrate mainly on correlation experiments involv-ing the spins of parent nucleus and emitted e±, and the momenta of e± and the recoiling nucleus. Neutral atom traps offer new opportunities because: i) high (i.e. very close to 100%) polarizations of the parent nucleus can be achieved; ii) all particles are emitted from a well-localized volume; and iii) momenta of particles lenving the source volnmf' arf' free of distortions because of the negligible source thickness. The most relevant parameters are thus the spin correlation coefficients A, B and D, and, perhaps at a later stage, when positron polarimetry is added, the R parameter. Because of the special properties of the neutral atom trap all types of measurements that have been performed with cold or ultracold neutrons could be pursued with nuclear positron emitters. Independent measurements for mirror transitions having mixed Fermi and Gamow-Teller transitions such as 19Ne [29], 21 Na [22] and 37K [30] are especially interesting because several parameters, if measured, could allow a comprehensive test of the consistency of augmenting the SM with new physics of a certain type. In Table 2.4 we show predictions of the SM for the mixed Gamow Teller-Fermi ground state decay of 1.23 s 37K assuming that 9AMGT I gv MF = - 0.584. The ft value is presently known only to an accuracy of 0. 7%, although a new measurement with better than 0.1% accuracy is planned at TASCC [31]. The same group has recently measured the 19Ne half-life with a precision of 0.05%, comparable to the precision of ft values for several superallowed o+ -+ o+ ,B decays [1]. It is seen that statistically significant deviations of ~0.2% from the SM values in Table 2.4 would be indicative of new physics outside the SM. The significance of measurements of D, R, A and Ra (the latter being the ratio of the longitudinal polarization of ,B's emitted parallel and antiparallel to the nuclear polarization 26 Table 2.4: Observables in 37 K positron decaya observable a A B G Q' valueb 0.6609 -0.5729 -0.7764 1.0000 -0.5729 relative errore 0.14% 0.06% 0.12% 0.00% 0.16% aobservables not shown (b, D, R) vanish in the Standard Model. bcalculated with the Standard Model assuming GAMaT/GvMF = -0.584. conly the errors arising from an assumed 0.1% error in the ft value are shown. [32]) for the 37K decay has been discussed by Deutsch [33]. His conclusions are that it will be difficult to improve the limits for an imaginary phase between axial and vector couplings established from D results for neutron and 19Ne ,8-decays, at least in the first stages of the experiment. The accuracy required for D would be a few 10-4 , comparable to calculable electromagnetic final state interaction effects. The R term requires measurement of the transverse polarization of the positron. Since lm(C4CT) is a.lready known to be very small from 8Li decay [34], a measurement of R to ±0.01 in 37 K would usefully constrain Im(CACs). The final state contribution toR from the Coulomb interaction for a point-like nucleus is Rfinal ~ a.ZmeAfpf3 ~ -0.014 in 37K; this can be calculated to sufficient accuracy for a non-point-like nucleus [35], and the result can be tested by measuring R in 21 Na, where A has opposite sign and Rfinal ~ +0.023. Both measurements of A (absolute) and Ra (relative) in 37K can sensitively probe the mass M2 of a predominantly righthanded W boson postulated in manifestly left-right symmetric models. In Fig. 2.6 (from Ref. [33]) the impact of an 0.3% A measurement and of an 0.6% Ra measurement is shown together with the allowed region due to all known beta decays (the SM value ( = b = 0 is excluded mainly due toft and A for the neutron). The dashed regions are excluded by considerations of the energetics of SN1987 A [36] (line labelled "SN") and by plausible assumptions on the Higgs sector [1]. The 37K data are insensitive to the mixing angle (, but could provide very useful constraints on M2 in the mass region ~ 420 GeV. The detection of positrons is made difficult by the fact that the fraction of unpolarized atoms on the walls has to be accurately known to interpret measured asymmetries, and that tracking of particles is inaccurate because of multiple scattering. Innovative schemes are needed to utilize walls during the ZOT trapping stage while eliminating their influence during the detection of positrons and recoils. Recoil detection using microchannel plates will allow the measurement of neutrino correlations. From direction and momentum of the positron and the relative time-of-flight of the recoil nucleus, neutrino direction and momentum can be evaluated. This would make possible studies of the .8 - v correlation both in the decay of 37K and in the Fermi o+ -. o+ decay of 38mK, providing limits on the tensor and scalar terms in f3 decay that can be translated, for instance, into limits on the existence of vector and scalar leptoquarks, new gauge bosons predicted by grand unified theories [37]. 27 O:(M,JM2)2 M2 020 r-----------------,(GeVJc2) 90%CL 200 0.10 250 0.08 CIA lA• 0.3% 300 0.06 K37 0.04 crR0 1Ra•0.6'1o 0.02 0 -0.01. -0.02 0.02 004 t Figure 2.6: Exclusion plot of the mass of a predominantly right-handed boson W 2 versus mixing angle (from Ref. [33]). The curves for 37K are constraints imposed by future mea-surements of the asymmetry A and the relative polarization Ra. The allowed region from all available beta decay data is also shown. The hatched regions "H" and "SN" are excluded as referred to in the text. 28 Preliminary experiments on the f3 asymmetry A can be performed with present TISOL beams. To achieve the statistical accuracy necessary for a final measurement of A, and for any of the coincidence experiments, will require at least 105 trapped 37K or 38mK atoms. At presently-achieved trapping efficiencies, this will require the larger yields projected from ISAC. 2.3.4 The TRIN AT facility Progress has been made in developing a TRIUMF Neutral Atom Trap (TRINAT) facility. Trapping of stable 39K and 41 K isotopes in an off-line "vapor-cell" [6] Zeeman-optical trap has been demonstrated. Trapping of these isotopes (and of 37K) is made somewhat complicated by the fact that their nuclear magnetic moments are small, so that the hyperfine splittings of the excited P ~ levels are only slightly larger than the naturallinewidth, and there are no true closed "cyclir:g" transitions. To avoid pumping all the atoms to either of the two hyperfine-split S1 ground states, it is necessary to use laser light at two different frequencies with 2 roughly equal power, i.e. a separate trapping beam for both the F=1 and F=2 42Sl ground 2 states. In a scheme patterned after that of T. Walker of Wisconsin [38] we split the light from the Ti:Sapph and shifted it with commercial acousto-optic modulators. Optimizing the trapping of 41 K (7% of natural potassium) is useful because its hyperfine structure is almost identical to that of 37K. Neutralization of the Fr+ beam from TISOL has also been demonstrated. The strategy is to stop the ion beam in a hot thin foil with a low work function, producing neutrals at thermal energies which can be trapped. Both on- and off-line experiments are proceeding to develop a neutralizer fast enough for the 1.2 s half-life of 37K and compatible with efficient injection into the Zeeman optical trap. The resulting facility will be housed in a clean room at a TISOL beamline, with a goal of first trapping of radioactives in the autumn of 1995. References [1] J. Deutsch and P. Quin, in Precision Tests of the Standard Electroweak Model, ed. P. Langacker, World Scientific, to be published. [2] S.J. Pollock, E.N. Fortson and L. Wilets, Phys. Rev. C46 (1992) 2587. [3] B.Q. Chen and P. Vogel, Phys. Rev. C48 (1993) 1392. [4] E. Raab et al., Phys. Rev. Lett. 59 (1987) 2631. [q] K. Lindquist, M. Stephens and C. Wieman, Phys. Rev. A46 (1992) 4082. [6] M. Stephens and C. Wieman, Phys. Rev. Lett. 72 (1994) 3787. [7] S. Chu et al., Phys. Rev. Lett. 55 (1985) 48. 29 [8] see special issue: Laser Cooling and Trapping of Atoms, eds. S. Chu and C. Wieman, J. Opt. Soc. Amer. B6 (1989) 11. [9] V. Bagnato et al., Phys. Rev. Lett. 58 (1987) 2194. [10] E. Cornell, C. Monroe and C.E. Wieman, Phys. Rev. Lett. 67 (1991) 2439. [11] T. Walker, D. Sesko and C.E. Wieman, Phys. Rev. Lett. 64 (1990) 408. [12] E.N. Fortson and L.L. Lewis, Phys. Rep. 113 (1984) 289. [13] M.C. Noecker, B.P. Masterson and C.E. Wieman, Phys. Rev. Lett. 61 (1988) 310; and references therein. [14] D. Cho, Bull. Amer. Phys. Soc. 39 (1994) 1082. [15] P. Langacker, in 30 years of Neutral Currents, Santa Monica, California, February 1993, preprint. (16] S.A. BlundelL W.R. Johnson and J. Sapirstein, Phys. Rev. Lett. 65 (1990) 1411. [17] M.E. Peskin and T. Takeuchi, Phys. Rev. Lett. 65 (1990) 964. [18] W.J. Marciano and J.L. Rosner, Phys. Rev. Lett. 65 (1990) 2963. [19] G. Altarelli et al., Phys. Lett. B245 (1990) 669. [20] W. Buchmiiller and D. Wyler, Phys. Lett. B177 (1986) 377. [21] P. Langacker, Phys. Lett. B256 (1991) 277; P. Langacker, M. Luo, and A.K. Mann, Rev. Mod. Phys. 64 (1992) 87. [22] Z-T. Lu et al., Phys. Rev. Lett. 72 (1994) 3791. [23] G. Gwinner et al., Phys. Rev. Lett. 72 (1994) 3795. [24] V.A. Dzuba, V.V. Flambaum "Calculation of Parity Non-Conservation in Francium" paper 2D-5. 145th International Conference on Atomic Physics, Boulder (1994), and V.A. Dzuba, V.V. Flambaum and O.P. Sushkov, submitted to Phys. Rev. A. [25] LAMPF experiment E1303, D.J. Viera and C.E. Wieman, spokesmen. [26) C.S. Wu et al., Phys. Rev. 105 (1957) 1413. [27) W. Petrich et al., submitted to Phys. Rev. Lett. [28) J.D. Jackson, S.B. Treiman and H.W. Wyld, Phys. Rev. 106 (1957) 517; --, Nucl. Phys. 4 (1957) 206. 30 [29] A.L. Hallin et al., Phys. Rev. Lett. 52 (1984) 337; -, Phys. Rev. Lett. 52 (1984) 1054(E). [30] T. Walker, private communication. [31] E. Hagberg, private communication. [32] P.A. Quin and T.A. Girard, Phys. Lett. B229 (1989) 29. [33] J. Deutsch, in Proceedings of the /SAC Workshop, Lake Louise, Alberta, February, 1994, m press. [34] M. Allet et al., Phys. Rev. Lett. 68 (1991) 572; --, PANIC93 contributions 4.8, p. 121 and 4.16, p. 132. [35] P. Vogel and B. Werner, Nucl. Phys. A404 345 (1985). [36] A. Burrows, R. Gandhi and M.S. Turner, Phys. Rev. Lett. 68 (1992) 3834. [37] E.G. Adelberger, Phys. Rev. Lett. 70 (1993) 2856; -, Phys. Rev. Lett. 71 (1993) 469E. [38] R.S. Williamson III et al., submitted to Phys. Rev. A; T. Walker, private communica-tion. 2.4 Nuclear Physics with ISAC 2.4.1 Known properties of nuclear states It seems appropriate to initiate the detailed discussion of the nuclear physics opportunities of ISAC by a brief review of the diverse nature of the nuclear spectroscopic information currently available and the extent to which that information depends on the proximity of the nucleus in question to the valley of stability. To illustrate the broad generalizations specific reference will be made to the known properties of nuclei of mass A=9, 73 and 209, chosen simply to be representative of the situation in light, medium and moderately-heavy nuclei. Nuclear ground states The atomic mass of the ground state of a nucleus is a basic property of importance both to an understanding of the many-body quantum system and in defining the role that a nucleus may have in measuring other nuclear properties of interest. The simplest aspect of this latter role is illustrated in Fig. 2. 7 which is a chart of the nuclides in which each nucleus that is known to be stable, or sufficiently long-lived to be found in nature, is represented by a solid square. The experimental data for known masses [1] supplemented by theoretical predictions of the masses of nuclei further from stability [2, 3], plotted as a function of Z and N, define 31 100 NUMBER OF PROTONS z so so PROTON DRIP LINE NUMBER OF NEUTRONS, N 100 1SO Figure 2. 7: Chart of the nuclides. The stable isotopes are represented by black squares. The grey area represents unstable isotopes that have already been synthesized and identified. a complicated three-- dimensional mass surface in which the values for the stable nuclei define the bottom of the valley of stability. The A=9 nuclei are an example of light isobars for which the masses of all nuclei are known to the limits of particle stability and in two cases beyond (9 B and 9 He). The situation is quite different in the case of A=73 nudei. Here the masses on the p-rich side are known probably to the limit of particle stability (4, 5). On then-rich side, however, there are seven nuclei (extending to ~~V 50 ) which are all predicted to be particle stable but for which no mass measurement exists. All of the A=209 nuclei are unbound to the emission of an a particle and this decay mode dominates the known decays of the most p-rich isobars. Nothing (mass or any other property) is known of the n-rich A=209 beyond ~~9Tlt 28 despite predictions that there are probably sixteen more n-stable isobars extending to ~~Tb144 . The masses of the unstable nuclei in any chain of isobars have a dramatic impact on their decay modes and particularly the half-lives (h). For the nuclei from 73Kr to 73Ni for which 2 . half-lives are known, the only decay modes available are (3+ (and electron capture) or {3-decay. The shorter half-lives further from stability reflect the increasing energy available for decay. These known or predicted half-lives must be considered when evaluating the techniques available for the study of these nuclei utilizing radioactive beams, particularly for a facility based on an isotope separator. For the nuclei 9C and 73Kr there is ample energy available for (3+ decay to proton-unstable 32 states in 9 B and 73Br. The importance of ,8-delayed proton decay as a technique for the study of p·· ri ch nuclei is discussed in Section 2.4.2. Similarly, 9 Li is a ,8-delayed neutron emitter and, from the predicted masses, one might anticipate this mode will be prominent in future studies of the decay of 73Co. On t.he proton-rich side of stability all of the known A =209 nuclei exhibit a decay with Q values increasing with Z such that for ~g9 Ac120, h "' 0.1 s. . 2 On the n-rich side of stability, theoretical predictions of the ,a--decay half-lives [6] would suggest that there are possibly eight unknown isobars (extending to ~g9Ta136 ) with half-lives longer than 0.1 s. Any significant assessment of the understanding of structure of a nuclear state requires measurement of its spin ( J) and parity ( n'). These properties are known for the ground states of all stable nuclei and are assumed to be J'lr = o+ for the lowest level in all even-even nuclei. For the ground states of all other unstable nuclei, measurements of J1r are largely concentrated near the valley of stability. At A=9, 73 and 209 the distribution of known values of J1r(gs) is given with reference to the stable nucleus by Z = 4 ± 1, Z = 32~i and Z = 83~i, respectively. Within the context of the static properties of nuclear ground states one should mention the ongoing effort to place even more restrictive limits on the electric dipole moments of nuclei and the rapidly-emerging prospects for measurements of the nuclear weak charge (Qw) (see Section 2.3.2). The potential importance of ISAC in this field will be the ability to extend the measurements to select specific nuclei of interest further from the valley of stability. Nuclear isomeric states In the odd-A nuclei there are many examples of such states, which are the result of the last unpaired nucleon occupying a low-lying single-particle orbit from which 1 decay must be of multi pole order 3 or greater. In addition to the possibility of using beams (or targets) of unstable nuclear ground states, a radioactive beam facility would permit studies of nuclear structure based on such isomeric states. Nuclear excited states In addition to the ground state and the possible existence of isomeric levels, each nucleus normally exhibits a rich spectrum of short-lived excited states. Below the thresholds for particle emission the states are discrete with well-defined excitation energy (Ex) and spin and parity ( J1r) and generally decay by electromagnetic transitions. In many cases the low-lying spectrum of unbound states is approximately discrete but the increase in level density and strong interaction decay width leads eventually to a continuum of states at higher excitation energies. Even in this region of strongly overlapping levels the simple modes of nuclear excitation can be revealed by an appropriate choice of nuclear reaction. As in the study of any many-body quantum system, detailed measurements of a wide range of properties of excited nuclear states has played a vital role in the evolution of our understanding of nuclear structure. 33 Our information on the "known" nuclei furthest from stability is extremely limited, par-ticularly with regard to spin and parity, quantities of crucial importance to understanding nuclear structure. There is a general tendency for more detailed information to be available on the proton-rich side of the valley of stability largely as the result of studies of reac-tions induced by the fusion of stable nuclei followed by nucle<?n evaporation, predominantly neutrons. 2.4.2 Radioactive decay modes For most radioactive nuclei the dominant decay mode for n-rich isotopes is by (3- -decay and for p-rich by (3+ -decay or electron capture. Close to the valley of stability these decays occur primarily to bound states of the final nucleus. Techniques for the study of the (3-decay spectra and the (3-delayed 1-ray spectra have been established over many years and provide the basis for a substantial body of precise spectroscopic data. The (3+ -delayed proton decays of both ~C3 and ~~Kr37 have been mentioned in Section 2.4.1. Studies of this decay mode have played a prominent role in extending our knowledge of proton-rich nuclei towards the proton drip-line. Recent progress is summarized in two review articles [7, 8] with reference to important results based both on activities produced by proton-induced spallation at ISOLDE and by fusion evaporation and fragmentation reactions induced by heavier ions. Critical factors in the thick-target ISOL techniques are the time required for the release of the activity and the efficiency of ionization. As was the case with the initial studies of this decay mode, (3+ -delayed proton decay remains important because of the uniqueness of the signal and the efficiency with which it can be detected. It has been particularly significant for the decay of nuclei with Z > N for which the Fermi (3+ -decay populates a proton-unstable isobaric analogue state. The potential range of experiments has been significantly extended by specialized techniques, including one used to deduce the very short lifetimes of the proton-emitting states [9] and another used to extract important information on (3+ - v correlations [10, 11). In addition to the emission of single protons, several nuclei with Tz = N;z ::; -2 are known to be delayed 2p emitters [7) and one (i~Ar13) has been identified as a (3-delayed 3p emitter [12). The low production cross sections for nuclei far from the valley of stability require the highest-possible beam currents on the production target. In Section 2.4.1 attention was drawn to the large number of n-rich nuclei which are predicted to lie within the neutron drip-line but about which nothing is known. The majority of these nuclei (including ~~Co46 ) are expected to exhibit 13--delayed n emission, a mode of decay well known for ~Li6 [13]. Recent experimental progress in the study of f3n-decays is also summarized in Refs. [7, 8]. As in the case of the f3p emitters there are now several known cases of 13--delayed multiple n emitters. In some cases the studies of f3n emission are limited by the energy resolution compatible with reasonably efficient n detection. In this regard some progress has recently been made utilizing time-of-flight techniques [14]. Given the very large number of nuclei which could be investigated using f3n spectroscopy and a growing interest inn-rich nuclei (see Section 2.2.7, for example), it is expected that studies of this decay mode will be important in the future. 34 Studies of 11-delayed a emission have played a very prominent role in the early phases of the TISCH, experimental program [15, 16, 17] . This decay mode has been observed for many light nuclei, both p-rich and n-rich, with significant contributions in such fields as: accounting for the cosmic 160/12C ratio [1.5]; parity violation in the strong interaction [18]: the spectrum of solar neutrinos following 8B 11-decay [19]; and studies of 11+ -v correlations [20]. For studies of the reactions 7Be(p, 1 )8 B and 19Ne(p, 1 )20Na, which are of great importance in nuclear astrophysics (see Sections 2.2.4 and 2.2.6), the 11a-decays of the final nuclei provide very convenient means of detection. Example: 17Ne(l1+pa) decay In the TISOL "Red Giant" experiment, the spectrum of a particles following the 11-decay of 16N was measured [15, 16]. Simultaneous R- and K-matrix fits to this spectrum and to existing 12C(a,1)160 and 12C(a,a)12C data put much tighter constraints on the E1 component of the 12C( a, 1 )160 reaction, at energies of interest in helium burning in the cores of red giant stars, than are possible from consideration of the a-capture data alone. However, the E2 component has not been so well determined. The 11-delayed proton decay of 17Ne (t 1 = 109 ms) into unbound states of 160 may provide an opportunity to improve our knowl~dge of this component. The 11-delayed proton decay of 17Ne populates both the 7.117 MeV 1- and 6.197 MeV 2+ states in 160 [21, 22] but the branching ratios are largely unknown. It is the tails of these two sub-threshold states that determine the E1 and E2 components, respectively, of the 12C( a , 1 )160 reaction at energies appropriate for helium burning. The isotope separator for ISAC will provide the high-purity, high-intensity beams of 17Ne needed for the study of the l1pa-decay of this nucleus. Experience in detecting and analyzing this complicated decay mode has already been obtained in the study of the decay of 9 C (TRIUMF experiment E682). Example: Possible l1+p-decay of g~Se31, g~Kr35 and ~~Sr35 Explosive burning of hydrogen via the rp-process (see Section 2.2.6) can produce nuclei of masses up to A= 100 [23]. The rp-process proceeds through nuclei with N ~ Z. Such nuclei lie near the proton drip-line in the region A = 65-73. A proposal to populate astrophysically-important nuclear levels in the key nuclei 65 As, 69Br and 73Rb via 11-decay of the precursers 65Se, 69Kr and 73Sr has been approved (TRIUMF experiment E726). The simultaneous detection of delayed protons and 1 rays from these states will provide essential nuclear information needed to predict quantitatively the energy production and nucleosynthesis of elements beyond iron for explosive hydrogen-burning scenarios. Production yields for the precursors have been estimated primarily from results obtained at ISOLDE [24] and are ~ 1/s/ 11-A of proton beam. This experiment will be initiated using TISOL, but will benefit greatly from the higher :fluxes available from the ISAC isotope separator, since TISOL intensities are marginal, at best. Light nuclei have also provided the first examples of 11-delayed d- and t-decay [7]. The 11--decay of 6 He to the unbound states of a+ dis a topic of current interest [25, 26] and the 35 subject of TRIUMF experiment E678 [27]. The first known example of direct proton radioactivity [28] is the isomer 53:2jCo26· This decay mode is even more distinctive than {3p and analysis of recent results suggests that it may provide unique spectroscopic information. Another decay mode for which there are possible candidates [29], but as yet no experimental evidence, is direct 2 proton (2He) decay. These would involve nuclei for which the ground state is bound to single proton emission but unbound to the simultaneous emission of 2 protons. Such nuclei would be far from stability and would require the ISAC target facility for their production. 2.4.3 Fusion reactions with neutron halo nuclei There is strong evidence that lighty-bound neutron-rich nuclei such as 11 Li and 11 Be have an extended neutron distribution or halo. For example, 11 Li is well described as a 9 Li core surrounded by two valence neutrons lying on average far from the core. This has a variety of consequences, including a prediction that a halo nucleus has strong low-lying El transitions. As a result, halo nuclei are easily polarizable in heavy-ion collisions. This should increase the probability of fusion of such a halo nucleus with a target nucleus at energies below the fusion barrier [30, 31]. The halo nucleus deforms in the electric field of the target nucleus, with its core being repelled by the target charge. Since the valence neutrons contact the target in advance of the core, the fusion cross section is enhanced for energies below the Coulomb barrier. Figure 2.8, taken from Ref. [32], shows the calculated enhancement of the fusion cross section as a function of energy for different Li isotopes incident on a 208Pb nucleus. 104 f I I . I . I I • I I II~ r r 102 ~ F r 10° ~ 'Li f .r 'Li r 'Li 10·2r 'Li f -o- IILi -4r 10 20 25 30 35 E c.m. (MeV) Figure 2.8: Calculated fusion cross sections (in mb) for Li isotopes on 208Pb [32]. While stable nuclei have rms radii for neutrons and protons that are nearly identical, 36 neutron-rich nuclei are predicted to have a neutron skin whose thickness is proportional to the di fference in the neutron and proton Fermi energies [33). One could study the the relatiYe value of the cross sections for a chain of isotopes at a single center-of-mass energy to see the effect of a growing neutron skin. The availability of intense beams of light neutron-rich ions from ISAC should permit wide-ranging studies of fusion at and below the Coulomb barrier to be carried out with light to medium weight nuclei. For example, with 6 - 11 Li projectiles at 1.5 MeV ju, the classical Coulomb barrier energy can be reached for targets up to 64Ni (and somewhat beyond). However, with somewhat heavier projectiles, such as 21 - 30Na, higher energies are needed (above 2 Mev /u) to cover the same range. An initial experiment might be carried out with a carbon or aluminium foil target. Assuming a 12C target of 1 mg/cm2 and a beam of 106 particles/s and a cross section of 1 mb, we could expect about 0.6 counts/s with a high-efficiency recoil detector (see Section 2.2.8). 2.4.4 Coulomb excitation studies Recently, some remarkable correlations between nuclear observables have been discovered which, being simple and nearly universal, hint at deep underlying origins [34, 35). In Fig. 2.9 we show a plot of yrast energies E( 4t) vs. E(2t) which reveals that the data for all collective non-rotational nuclei lie on a single straight line satisfying the equation E( 4t) = 2.0E(2t) + t: where t: = 156 ± 10 keV. This is the equ-ation of an anharmonic oscillator. Also shown are the harmonic oscillator (R4; 2 = E( 4t)/E(2t) = 2.00) and rotor (R4; 2 = 3.33) limits as well as a least-squares fit to the data. The nuclei involved span a range of structures from nearly harmonic vibrators to /'-soft rotors to soft transitional nuclei, with a variety of intermediate cases. Yet all satisfy the same model with constant anharmonicity t:. This is totally unexpected and raises two questions. Firstly, is this correlation a feature only of those nuclei we already have available for study or is it a generic property of all nuclei, and what are the general properties of nuclear structure and interactions that give rise to it? Secondly, can simple plots of one collective observable against another, such as this, provide new signatures of structure; can they perhaps provide the first evidence for radically new types of structure that may exist far from stability [36)? These questions can only be answered by extensive measurements on new nuclei using accelerated radioactive beams. A second correlation that provides a signature of structure involves a plot of B(E2: Of -+ 2t)jA, with A in Weisskopf units, against R4; 2 = E(4t)/E(2t). This correlation splits into two tracks, one for nuclei in a sperical vibrator to rotor transition, and one for nuclei in a 1'-soft to rotor transition [35]. Previously, very thorough and detailed studies of entire transition regions have been necessary to disentangle these two types of phase transitions. Now we see the measurement of only E(2t),E(4t) and B(E2) in one or two nuclei can suffice. 37 2rJOO 3 .33 • / .··2 .00 2000 • . •. · - 1500 > cu .!!:: -. ... 1000 CaJ . . Z=30-02 500 /_ .. ··· . 2.05<R<3.15 . .. ·· . . . . 0 . '·. 0 250 500 750 1000 1250 E2· (keV) Figure 2.9: Correlation of E( 4t) with E(2t} . The harmonic oscillator (R412 = 2.00) and rotor (R412 = 3.33) limits are shown as well as a least-squares fit to the data [34]. An extensive programme of low-energy Coulomb excitation experiments proposed for ISAC [37] is designed to excite only the 2i and 4i states and to measure only the B(E2 : Oi --+ 2i) value. The low energy assures the excitation of only the lowest state or two, giving very clean spectra. Since the spectra are simple and scattered radioactive ions are in a narrow cone, an efficient, compact Nal(Tl) detector can probably be used. For a 105 particles/s of A = 40(100) nuclei incident at 1.4 MeV ju on a 1 rng/crn2 target of 12 C, the calculated counting rate for a 500 keV 2i --+ ot transition in the photopeak for a 2.5 x 2.5 ern Nai(Tl) detector located close to the target is 2800(180) counts/hr, with very low background. With this system extended isotopic chains can be studied sequentially under identical conditions. For nuclei with half-lives of the order of 1 ns or less, a single detector near the target is sufficient. For nuclei with half-lives of a few ns, de-excitation occurs in flight and a linear array of 4 or 5 detectors spanning 20- 30 ern would be required. In this case, B(E2) is obtained from the decay curve, and it is not even necessary to monitor the beam intensity. Initial experiments will be on light nuclei with A ~ 30, and possibly on selected heavier species (such as Kr or Xe) which might be available as multiply-charged ions. 38 2.4.5 Magnetic moments in mirror nuclei Several methods exist for the creation of nuclear polarization of long-lived nuclear levels. Among the most extensively used are low temperature nuclear orientation and laser tech-niques which are, however, limited to specific cases. For example, low temperature nuclear orientation is limited to implanted nuclei in thermal equilibrium with the surrounding lat-tice. An alternative method is tilted foil polarization (38] in which the nuclei are oriented in the hyperfine field resulting from the surrounding electrons which are oriented in flight by a tilted foil. Even though the expected polarization is small, the method is universal and easy. So far the latter method has been applied to quadrupole moment measurements of excited states [39], parity violation experiments (40], and ground state g-factor measurements [41]. In particular, the measurement of g-factors of ground states of mirror nuclei, and the sums and differences of such moments, can provide direct information on the isovector and isoscalar components of the nuclear current. Such nuclei can be produced and accelerated by a facility such as ISAC, polarized with the tilted multi-foil technique, and their g-factors measured using the ,8-NMR technique (41]. The net polarization of a beam passing through a single foil is very small, so that a multi-foil stack is used to increase the polarization. Nuclear polarizations of 0.02 to 0.07 can be expected for low-spin ground states. After passing through the last foil, the ion is stopped in a catcher lattice of cubic symmetry free of internal fields, cooled to a low temperature, and placed in a "holding" magnetic field (:::::: 0.1 T) in the direction of the nuclear polarization, in order to preserve the polarization along a well-defined axis during the nuclear decay. The combination of catcher temperature and holding field is chosen to make the depolarization time long compared with the nuclear lifetime. Beta particles from the ground state under study are detected at 0° and 180° to the direction of the nuclear polarization and the asymmetry determined. Nuclei implanted into a cubic environment experience a simple Zeeman splitting asso-ciated with the holding field (Bext)· Under the influence of an applied RF field at right angles to Bext, resonant absorption will result in the destruction of the ,B asymmetry. The resonance frequency can be determined to high precision and is a measure of the nuclear g-factor. Thus the magnetic moment of the isoptope under study can be extracted without detailed interpretation of the observed symmetry. A programme has been initiated at ISOLDE using a high-voltage platform to accelerate beams up to 520 keV. With this system a nuclear polarization of ::::::0.2% has been observed for 23Mg ions implanted into Pt at 4 K, using 3 carbon foils at 70° (42]. ISAC could be used to produce higher beam energies than are available at ISOLDE to investigate the enexplored regions of T= ~ nuclei in the sd-shell, and of T= ~ nuclei in the f-shell, by measuring magnetic moments of mirror nuclei. At the higher beam energies, increased polarization should be obtained by using more foils. Theoretical considerations show that, under certain conditions, such enhancement could be considerable [42]. Finally, the polarization data can be analyzed to provide information on ,8-decay matrix elements (for example, the ratio of Gamow-Teller to Fermi matrix elements). 39 References [1] G. Audi and A.H. Wapstra, Nucl. Phys. A565 (1993) 1. [2] A.H. Wapstra, G.Audi and R. Hoekstra, At. Data and Nucl. Data Tables 39 (1988) 281. [3] P.E. Haustein, At. Data and Nucl. Data Tables 39 (1988) 185. [4] J.M. D'Auria et al., Phys. Lett. B66 (1977) 233. [5] F. Mohar et al., Phys. Rev. Lett. 66 (1991) 1571. [6] A. Staudt et al., At. Data and Nucl. Data Tables 44 (1990) 79. [7] B. Jonson and G. Nyman, in Handbook of Nuclear Decay Modes, Volume 2, ed. D. Poenaru, CRC Press, Boca Raton, Florida, 1993. [8] A.C. Mueller and B.M. Sherrill, Annu. Rev. Nucl. Part. Sci. 43 (1993) 529. [9] J.C. Hardy et al., Phys. Rev. Lett. 37 (1976) 133. (10] D. Schardt and K. Riisager, Zeit. Phys. A345 (1993) 265. (11] E.G. Adelberger, Phys. Rev. Lett. 70 (1993) 2856. (12] D. Bazin et al., Phys. Rev. C45 (1992) 69. (13] G. Nyman et al., Nucl. Phys. A510 (1990) 189. (14] R. Harkewicz et al., Phys. Rev. C44 (1991) 2365. (15] L. Buchmann et al., Phys. Rev. Lett. 70 (1993) 726. [16] R.E. Azuma et al., Phys. Rev. C50 (1994) 1194. (17] M. Dombsky et al., Phys. Rev. C49 (1994) 1867. (18] K. Neubeck, H. Schober and H. Waffler, Phys. Rev. ClO (1974) 320. (19] J. Napolitano, S.J. Freedman and J. Camp, Phys. Rev. C36 (1987) 298. (20] E.T.H. Clifford et al., Nucl. Phys. A493 (1989) 293. (21] M.J.G. Borge et al., Nucl. Phys. A490 (1988) 287. (22] J.C. Hardy et al., Phys. Rev. C3 (1971) 700. (23] L. van Wormer et al., Ap. J. 432 (1994) 326. (24] ISOLDE User's Guide, ed. H.-J. Kluge, CERN 86-05 (1986) and updates. 40 (25] K. Riisager et al., Phys. Lett. B235 (1990) 30. [26] C. Borcea et al., Nucl. Phys. A565 (1993) 158. (27] D. Anthony et al., to be published. (28] X. Xu et al., Chin. J. Nucl. Phys. 15 (1993) 18. (29] B.A. Brown Phys. Rev. C43 (1991) R1513. (30] W. van Oertzen et al., Z. Phys. A326 (1987) 463. (31] P.R. Stelson, Phys. Lett. B205 (1988) 190. (32] N. Takigawa, M. Kuratoni and H. Sagawa, Phys. Rev. C47 (1993) R2470. (33] D. Hirata et al., RIKEN Report AF-NP-158. [34] R.F. Casten, N.V. Zamfir and D.S. Brenner, Phys. Rev. Lett. 71 (1993) 227. [35] N.V. Zamfir and R.F. Casten, Phys. Lett. B305 (1993) 317. (36] J. Dobacewski et al., Phys. Rev. Lett. 72 (1994) 981. (37] R.F. Casten et al., Letter of intent to the TRIUMF Experiment Evaluation Committee, July, 1994. (38] H.G. Berry and M. Hass, Ann. Rev. Nucl. Part. Sci. 32 (1982) 1. (39] M. Hass et al., Phys. Rev. C43 (1991) 2140. [40] C. Braude et al., Z. Phys. A336 (1990) 133. [41] W.F. Rogers et al., Phys. Lett. B177 (1986) 293. [42] M. Hass, Letter of intent to the TRIUMF Experiment Evaluation Committee, July, 1994. 2.5 Materials Science 2.5.1 Introduction The interdisciplinary field of materials science embraces such diverse subjects as physics, chemistry, engineering, metallurgy and biomedical science. The ability to synthesize new materials with specific, desired properties has stimulated the rapid development of powerful, sophisticated techniques for their characterization and study. A major impetus for this intense activity has been the potential for commercial applications. An equally important factor has been the need to elucidate the behaviour of these substances and to understand 41 the fundamental interactions and reaction dynamics of the processes by which they are produced. In recent years the use of radioactive ion beam technology has been an important new technique for the study of materials. By removing the serious limitation imposed on many nuclear methods by the restricted range of suitable radioisotopes, ISAC will open up new vistas for a wide range of applications. These applications include: Mossbauer spectroscopy; perturbed angular correlation spectroscopy; nuclear stimulated desorption; low temperature nuclear orientation; radiotracer diffusion applications; ion beam modification of materials; and thin layer activation. This is not an exhaustive list, as we fully expect other modes of experimentation to benefit from ISAC capabilities. Before proceeding with individual applications, we offer some remarks concerning their general characteristics and operational implications. 1. We may differentiate between a primary use and a secondary use of any given isotope. In a primary use, advantage is taken of some feature of a radioactive isotope which does not occur for stable isotopes of that species, thus creating a totally new mode of experimental study. In a secondary application, the radioactive and stable isotopes are subjected to similar conditions, but the position of the former is monitored by its radioactivity. 2. Applications can be classified as on-line or off-line. This is an important operational classification, for it dictates (for on-line work) the provision of a suitable infrastructure around the installation (UHV, cooling, high voltage manipulation, etc) . 3. We must distinguish between research-oriented and production (industrial) applica-tions, whose demands from the target-separator system are often contradictory to one another. 2.5.2 Mossbauer spectroscopy on line (MSOL) The applications of Mossbauer spectroscopy (MS) in solid state physics, chemistry, biology and materials technology using conventional radioactive sources are now well established. This is an especially efficient method for studying ion implantation, radiation damage, and modifications of materials by ion beams. MS studies of implanted radioactive nuclei were initiated in the mid-sixties and have been extensively covered in numerous surveys [1, 2, 3, 4, 5]. The possibility of extending the use of this technique to isotopes obtainable only with high intensity radioactive beams is especially attractive. From the quantities measured in Moss bauer spectra, such as the isomer shift, the quadrupole splitting and the magnetic hyperfine splitting, one derives, respectively, the density of elec-trons, the components of the electric field gradient tensor, and the hyperfine magnetic field at the nucleus. From the magnitude of the resonant effect, the mean square vibrational amplitude of the implanted probe nucleus in a host matrix is determined. From the second order Doppler shift, the mean square velocity of the nucleus at its residence site can be determined and, in some cases, its relaxation and mode of diffusion can be extracted from 42 the shape of resonant lines. The analysis of the spectra as a function of temperature pro-vides information about the chemical state and electronic structure of implanted atoms, their position in the matrix, the configuration of associated lattice defects, characteristics of the phonon spectra, and the magnetic properties of the material. MS applies to both crystalline and non-crystalline (amorphous) solids, and to highly dispersed or colloidal materials. The application of MS is limited to specific nuclei near the stability line with atomic masses A > 40 and with isomeric states with Ex < 200 keV and life-times ranging from 10-6 to 10-10 s. Also, a stable or very long-lived counterpart isotope for the absorber is always required. In spite of these limitations, there have been more than 100 MS resonances discovered. New resonances may still be added to the list, and some nuclear parameters may still be refined, with the advent of radioactive beam facilities. MSOL opens a large field of solid state and materials science applications. In addition, because of the decreasing availability of separated enriched isotopes for making Mossbauer sources and absorbers, and the increasing cost of irradiations and subsequent radiochemical processing, the use of on-line techniques promises to be more and more attractive. Isotope separator implantation of stable resonant isotopes has been extensively studied using conversion electron Mossbauer spectroscopy (CEMS) [6, 7]. Relatively high implanted doses of 1013 - 1015 ions/cm2 yielding concentrations of"" 10-2 to 1 at.% are needed even in most favorable cases of 57Fe and 119Sn. Experiments with 151 Eu and 197 Au using rv 1016 -1017 ions/ cm2 have also been reported. CEMS is also very useful in the study of beam-induced modifications in materials containing resonant isotopes. Thanks to its inherent depth selectivity, CEMS makes it possible to determine how a particular property (e.g. valence state, chPmical composition or orientation of magnetic moments of Mosshauer atoms) varies with the depth below the surface. Off-line isotope separator implantation of long-lived Mossbauer source nuclei has been studied extensively [4, 5]. These studies provided a wealth of experimental data and continue to play a leading role in understanding the microscopic nature of the lattice sites occupied by implanted ions, the behaviour of lattice defects as a function of temperature, and the local chemistry of the impurity-host system in conditions far from thermal equilibrium. However, the radioactive species used in these studies have to undergo sometimes fairly complex irradiation and chemical processing before they are inserted into the ion source. On-line techniques allow for implantation of short-lived isotopes (100 ms < h < 1 h), and also offer intensity advantages for longer-lived isotopes. Several years ago, ~n on-line Mossbauer spectroscopy set-up was installed at ISOLDE [8]. Initially, mass-separated beams of 109 ions/s of 119In+ (h = 2.1 min) and 119Sn+ (h = 38.5 h) were produced by proton-induced fission in a uraniu~ target; other beams prod~ced by spallation reactions on a variety of target materials have also been used. Table 2.5 lists beam parameters for the Mossbauer experiments at ISOLDE. Very low doping levels of the order of 108 - 1010 ions/cm2 can be used, which is essential in investigations of semiconductors. With this technique, spectra can be obtained in periods as short as 1 min, which allows various time-dependent phenomena to be studied. Recently, this technique has been used mostly in studies of impurities and defects in metals and semiconductors [4, 5, 8]. In particular, implantation-produced defects in III-V and II-VI semiconductors, of high technological interest, and the mechanism of the 43 Table 2.5: ISOLDE beams used for Mossbauer spectroscopy [8] Radioact. Half-life Moss b. Target Reaction Ion Intensity Isotope (min) Isotope Mat. Source (ions/s) lllhnCd 2.4 11l:ISn Sn spallation plasma 5xl07 119Jn 2.1 119Sn uc fission W-surf lx109 119Xe 5.8 119Sn La spallation plasma 8x108 121Xe 39 121Sb La spallation plasma 3x109 12sxe 0.95 12sTe La spallation plasma 6x109 unusually fast diffusion of Sb donors in n-type Si have been studied. The in-beam recoil-implantation technique, where excited Moss bauer nuclei are implanted directly into a target, has also been developed. In this case, Mossbauer spectra are measured during the lifetime of excited Mossbauer states (lo-s -10-10 s). In particular, the Coulomb recoil implantation Mossbauer effect (CRIME) has been applied in Stanford, Erlangen and Oak Ridge for 57Fe, 61 Ni, 73Ge, 158Gd, 161Gd, 172Yb, 174Yb and 176Yb [9]. Considerably improved CRIME experiments are now being conducted using a pulsed beam of 40 Ar at 110 MeV at the heavy-ion VICKSI accelerator facility of the Hahn-Meitner Institut in Berlin [10] or a 57Fe beam at 200 Mev at the UNILAC accelerator at GSI in Darmstadt [11]. 2.5.3 Perturbed angular correlation spectroscopy (PACS) Typically in PACS a radioactive probe, which decays by successive 1-ray transitions, is in-corporated into the system under study. During the lifetime of the intermediate nuclear state, the interaction between the electric quadrupole (or magnetic dipole) moment of the nucleus and the local electric field gradient (or magnetic hyperfine field) can perturb the known angular correlation for the delayed 1 - 1 coincidences. A measurement of the per-turbed correlation then yields information on the perturbing interaction, which can in turn be related to the electronic structure of the material. While PACS reveals information similar to that obtained from other "environmental" techniques, such as nuclear magnetic resonance (NMR) and Mossbauer emission (ME) spec-troscopy, it has many advantages. The sample quantities can be very small (typically 1-100 mg) and there is essentially no limitation to the choice of reaction vessel material or the form of the sample. There is no dependence on temperature nor sensitivity to vibrations. No magnetic or radiofrequency field is required (as in NMR) and small nuclear quadrupole interactions are more easily resolved than in ME. The entire hyperfine spectrum is nor-mally detected and the signal is a constant which depends only on nuclear properties. This means that the sum of all inequivalent probe sites must always add up to 100%, even in heterogeneous systems (although problems can arise due to insufficient time resolution). Unlike NMR and ME, there is essentially no way in which the observed linewidth in the 44 Fourier-transformed spectra (damping of the oscillations in the perturbation spectrum) can be influenced by instrumental means; thus small site inhomogeneities can be detected with high reliability. Reorientational motions of the electric field gradient ( efg) can generally be identified unequivocally and are easily distinguished from static inhomogeneous efg distribu-tions. Finally, the isotopes that are suitable for PACS complement those amenable for ME and NMR in a useful way. As mentioned earlier, a serious limitation to the applicability of PACS, however, has been the availability and cost of suitable isotopes. ISAC will eliminate this restriction, extending accessibility to a much wider range of probe nuclei. The following is a list of candidates for PACS studies· 44Ti 99Mo 99Rh 100Pd 111 Ag 111mCd 111 In 115Cd 116mSb 131mTe 133Ba . ' ' ' ' ' ' ' ' ' ' ' 133Ce 14oLa 1ssTb 112Lu 111Yb 11sw 1s1Hf 1s1w 199mHg 204pb 2o4Bi This list can be ' ' ' ' ' ' ' ' ' ' . expected to be augmented as ISAC expands its menu. Furthermore, while the low-energy separated beam can be implanted to only a relatively-small depth, the accelerated beam can be implanted much deeper, and with a programmable profile. Applications of PACS to materials science is already extensive [12], embracing many aspects of solid state physics [13), including the study of surfaces [14]. Here we list two proposals, the first concerned with basic physics, the second an application to the field of semiconductors. Example: Quadrupole interactions in highly-oriented pyrolitic graphite (HOPG) A fundamental understanding of the interaction between the impurity probe and host material is essential for the interpretation of PACS data. Here ISAC can play an important role, for example, in the study of electric quadrupole interactions, by making measurements possible which should be able to isolate the contribution to the efg from the probe itself. Cap-italizing on the wide variety of probes available with ISAC, a systematic study of quadrupole interactions in HOPG is proposed for the purpose of testing current theoretical predictions of the efg at various impurity sites. Calculations incorporating complete hybridization of all electronic orbitals, previously considered intractable, are now feasible as a result of the massive increase in computing power that has occurred in recent years. HOPG is a synthetic form of graphite in which the mosaic spread of the compressed c-axes is typically less than 1 degree. Thi-s pseudo-single crystal nature of the material offers several advantages for the study of quadrupole interactions [15, 16]. As well as the probes listed above it will also be possible to use polarized beams, a factor which will enormously enhance the possibilities of PACS. Example: Semiconductor studies PACS has proven to be an invaluable tool in the study of semiconductors [17], yielding information on impurity-defect configurations, the annealing of radiation damage after heavy ion implantation, and on the hydrogen passivation of acceptor dopants. Measurements with the following probes and semiconductors are proposed: 111 In, 116mSb , nsmsb: lnP, lnAs, lnSb, GaAs, GaP, GaSh (binary III-V systems) 45 1nmcd, nscd, n1Cd: CdS, CdSe, CdTe CdSiP2, CdGeP2, CdGeAs2, CdSnP2 204mpb, 131mTe: PbS, PbSe, PbTe (III-V alloys) (binary II-VI systems) (ternary II- IV-V2 systems) The extent and annealing behaviour of the radiation damage is connected with the binding properties of these compounds. PACS affords a significant advantage by allowing measure-ments to be performed in situ. Since doping is the most widely used method for imparting required properties to semiconductors, it has been pointed out [18] that the investigation of transition element impurity centres, particularly those with strong electron localization, is important both for theoretical and experimental reasons. As can be seen from the candi-date isotope list, there are several elements on the list suitable for such studies. Thus the application of PACS to semiconductors at ISAC could in itself be an on-going long range program. 2.5.4 Nuclear stimulated desorption There are two generically distinct relationships between nuclear phenomena and atomic-scale induced effects on surfaces and thin films. Firstly, the dynamics of a nuclear reaction (primarily the recoil of the nucleus) may affect the position of the atom. Secondly, a nu-clear reaction (or decay) may serve as an analytical indicator of the location of the atom, or molecule, in question. In nuclear stimulated desorption (NSD) both of these aspects com-bine in ;tn essential way [19] when two consecutive decays (e.g. weak decays or isomeric transitions) are used. The first of these decays causes the nucleus (i.e. the atom or molecule containing it) to desorb from the surface onto which it had been placed; the second serves to determine the position of the daughter and thereby the characteristics of the primary desorption. The basic premise is that the probability of desorption, as well as its temporal and angular dependence, and the characteristics of the desorbed species contain information about the surface and about the substrate-adsorbate interaction. Quite independently of the inherent significance of the primary, desorption-stimulating decay, its net result is an outgoing flux of radioactive atoms. If no other mechanism (such as thermal desorption) is effective, then essentially only the radioactive daughters leave the surface at all times. This constitutes an ultra-clean flux of radioactive markers, which can be put to use in different applications [20]. These applications include: the marking of designated surfaces for ·reference in etching; the implantation of monoenergetic charged particle sources for thickness measurement and monitoring; and the study of surface and bulk diffusion of selected atomic species. Nuclear stimulated desorption offers an alternative, or complementary, technique to PACS, when one studies sites on a surface, as in the following example. Instead of using the isotope 46 111mCd which is deposited on a surface in the course of the PACS experiment, one uses the isotope 107Cd, with the very same collection and re-evaporation methods. Upon its electron capture decay (h = 6.5 h), 107Cd recoils with an energy of 8.7 eV. This energy is high enough to cause ~t to desorb from the surface. However, this desorption probability, and its departure from azimuthal isotropy, depend crucially on the nature of the site in which the decaying atom resides. By collecting the recoils, and measuring the subsequent decay of the 107mAg activity (h = 44 s), the site-specific signature (or its time-dependent mix) can be directly determin~d. Although this involves a completely new set-up, it provides the relevant information in a much shorter time, and relies on a rather straightforward and intuitive theoretical analysis. One of the isotopes which is most suitable for application in a variety of NSD scenarios is 47Ca. However, the only practical method for obtaining it in a usable, carrier-free form is to produce it in an on-line mass separator and implant it in a foil (e.g., platinum) from which it can be subsequently evaporated. Its long half-life (4.5 d) actually permits its utilization at sites far removed from the actual production facility, and for an extended period of time. Existing facilities (ISOLDE) produce 47Ca through its precursor 47 K, but in quantities too small for many interesting applications. The projected intense primary beam available at ISAC will be ideal for the implementation of experiments based on this particular isotope. 2.5.5 Low temperature nuclear orientation Nuclei under the influence of an ordering interaction will become oriented when kT is smaller than the interaction energy involved. The ordering interaction can be either magnetic or electric. Magnetic ordering is usually obtained via the hyperfine interaction in magnetic materials or, in some cases, by applying large laboratory fields. There are, however, materials in which the electric quadrupole interaction is sufficiently large to produce significant nuclear orientation. The degree of orientation of an ensemble of radioactive nuclei in bulk matter can be obtained by measuring the directional anisotropy of the emitted radiation, which depends both on the details of the decaying nucleus and the characteristics of the local field. 'When the ordering field is known, the measurement yields information concerning the nuclear moments and decay scheme parameters. Conversely, knowledge of the nuclear parameters allows deduction of the local fields, which are of general interest for the study of the solid state. Commonly, the observed transition is a 1-ray, but a- and .8-decays can also be observed. The temperature needed to orient nuclear spins is typically of the order of 10 mK and can be obtained using a dilution refrigerator. A very important refinement of the nuclear orientation (NO) technique, is nuclear mag-netic resonance of oriented nuclei (NMRON). In this method, once a certain degree of nuclear order has been obtained, magnetic resonance can be used to alter the substate population distribution and can be detected by the resulting change in the anisotropy pattern of the radiation. NMRON is extremely sensitive, and can measure the ordering interaction very precisely and independently of other operational parameters (e.g. temperature). The combi-nation of NO with NMRON and its variants (for example, pulsed NMRON) is a very potent analytic tool [21]. 47 Conventional NO and NMRON are limited to isotopes having half-lives in excess of a few hours. Even with "top-loading" dilution refrigerators the lower limit required to arrive at the proper operational conditions is about two hours. Much shorter-lived isotopes can be studied by combining a dilution refrigerator with an isotope separator and an on-line accelerator. The isotope of interest is produced and implanted directly into a pre-cooled host specimen, thereby permitting the utilization of half-lives as low as 10 s. Since NO is a singles measurement, it can be efficiently carried out for quite weak radioactive sources. It can, in fact, often be done in conjunction with other nuclear studies, such as ; - ; and f3 - ; correlation, electron conversion and ;-ray linear polarization measurements. Another advantage is the possibility of implanting a variety of isotopes into a given specimen, and extracting information both on condensed matter physics [22] and a variety of nuclear physics phenomena [23, 24, 25]. On-line NO facilities currently operate in a number of laboratories (for example, NICOLE at CERN and KOOL at Louvain-la-Neuve). However, the projected high intensity of the primary proton beam for ISAC (10 11-A) makes the proposed facility extremely competitive. This is further enhanced by the possibility of obtaining polarized nuclei using the tilted foil technique (see Section 2.4.5). An example of the inherent advantage of direct on-line implantation is the study of high-temperature superconducting materials. One could study, in principle, the antiferromagnetic phase by conventional NO methods if it were not for the problems encountered in sample preparation. Even when suitable long-lived radioactive probes are available (e.g. 54Mn or 6°Co) crystal growers are naturally reluctant to incorporate them into their system, risking its contamination. This is obviously not the case in on-line experiments. Not only is the choice of isotopes much wider, because of the removal of the practical limitation on half-life, but the implanted species normally decays quickly enough to leave a valuable crystal available for subsequent studies. Furthermore, copper, which is of particular interest in many systems, has no radioactive isotopes suitable for conventional NO. Also of interest are low dimensional systems of magnetic materials , where the ordering field and the spin-lattice relaxation time give information about the electronic magnetization and magnon energy gap, respectively. Magnetic semiconductors are of topical interest as well, and could be investigated by this technique. A more speculative, yet extremely interesting application, would be the investiga.tion of advanced materials with small structure, such as magnetic nanostructures (in particular quantum wells and magnetic semiconductor superlattices) and very thin films of ordered magnets. An inherent problem in the utilization of conventional NMR for such studies is the insufficient number of available spins. A 20 A thick film of 1 mm2 of iron, for example, has approximately 6 x 1013 nuclei, while the number required for NMR is typically at least 1016 • By contrast, the the NMRON technique is sensitive to about 4 x 104 x t 1 spins, where t1 is the isotope half-life in seconds. For very short-lived isotopes, with half-fives in the rang~ of seconds, only about 105 spins are thus required. To apply this approach successfully calls for a very precise manipulation of the separated beam; it must be strictly focussed onto the millimeter scale area, and its energy must be finely adjusted in the range of a few ke V to allow efficient implantation at the proper depth. From the point of view of material science, the general rule would be to implant an 48 isotope of the host constituent, unless one studies specifically an impurity problem. Thus, a number of isotopes of copper (59Cu, 60Cu, 61 Cu), as well as an isotope of bismuth (213Bi) are likely candidates to be used as probes for studying the antiferromagnetic phases of high-temper~tture superconducting materials. Similarly, 51 Mn could be used for studying Mn-based magnetic materials and magnetic semiconductors, in bulk or as nanostructures, while 53Fe and 55Co would be suitable as probes of the corresponding iron and cobalt magnetic materials. 2.5.6 Radiotracer diffusion Diffusion studies with the radioactive tracer technique, using either sample sectioning or some depth sensitive radiation counting, require a certain solubility of the diffusing element for the evolvement of the characteristic diffusion profile. In cases where impurity atoms are strongly held at the sample surface, these methods are inapplicable. Rather, deep implan-tation is required, as the initial stage of such studies. The depth of implantation increases monotonically with the implanted species energy, and so does the width of the implantation profile. However, the ratio between the two, which determines the degree of localization of the implanted species, actually improves with energy. The system of isolated impurity atoms at a specific depth lends itself perfectly to the investigation of diffusion even for immiscible systems. Such systems, produced by forced alloying, are of increasing technological interest. Their systematic study at high implantation depth and very low concentration could lead to a better understanding of the processes occurring in materials, with practical applications. The potential of this proposed diffusion technique has been demonstrated in detailed stud-ies of astatine diffusion in alkali metals, even though that specific system is of no particular technological significance. The diffusion mechanisms can be studied further by measuring the a spectra emerging from systems with implanted a-emitting isotopes, such as 205Fr in K. Reliable diffusion coefficients for such systems can, in fact, be extracted from detailed measurements of the a-particle spectra as a function of time and temperature. When this is coupled with monitoring the behavior of the daughters of francium (the corresponding astatine isotopes) and by repeating this for an entire range of available isotopes, the diffu-sion cor;:.; tants can be determined accurately over an unprecedented range of more than ten orders of magnitude. The typical requirements for this experiment are beams in the energy range of 20 ke V to 100 MeV for the isotopes 199 At (h = 7 s) and 209 At (h = 5 h), and of 1 MeV to 100 MeV 2 2 for 217 At (h = 30 ms). The intensities required for the short lived isotopes are typically 105 2 ions/s. Measurements for the long-lived 209 At isotope can actually be carried out off-line, so that brief collections from a high intensity beam may be used. 2.5. 7 Ion beam modification of materials It is now well established that ion implantation can be used to modify the surface and near-surface regions of almost any material [26]. In fact ion beams can provide the most 49 versatile and sophisticated family of methods for tailoring the surface properties of all classes of material (semiconductors, metals, ceramics, polymers, etc.) . Their usefulness arises from the particle energies they can carry, which are much greater than conventional thermal energies or interatomic bond energies. Consequently, non-equilibrium modifications can be achieved, the results of which are rapidly quenched into place. Applications in the semiconductor industry, where ion beam technology has rapidly re-placed diffusion techniques, include the following: highly controlled and reproducible doping with P and As ions to form n-type Si and with B ions to form p-type; high-resolution ion beams for the manufacture of very large scale integration (VLSI) devices; and deeply buried 0 implants for the fabrication of silicon on insulator (SIO) devices. A comprehensive list of applications to other fields can be found in Refs. [27, 28]. To demonstrate the wide range of these applications one might mention a few: carbon and nitrogen implantation into titanium alloys to reduce the wear of orthopaedic prostheses; chrome implantation into copper alloys to increase the resistance to corrosion in batteries; metallic implantation into polymers to enhance electric conductivity for use in microelectronics. This field is in its infancy. With ISAC there are essentially limitless possibilities to investigate with the variety of ions that will be available. A systematic search for new methods of modifying and treating materials will undoubtedly bring huge dividends. 2.5.8 Thin layer activation The efficiency and reliability of machine parts and industrial equipment is significantly af-fected by such degradation processes as wear and corrosion. A reliable and on-line measure of such degradation can result in substantial savings in time and money during the develop-ment of machine components a.nd lubricants, or static parts subject to corrosion. Moreover, on-line monitoring may be used to minimize costly downtime and unscheduled interruptions during a component's lifetime. Thin layer activation (TLA) is a technique for achieving this aim. Activation with protons, deuterons or a particles will dope a thin layer at a pre-determined surface of the component with radio-isotopes. This induced activity is monitored in situ by 1-ray detectors during the wear or corrosion process. There exist, however, many cases where proton activation is either impossible or impractical. This is particularly true for parts made of plastics, aluminum and many ceramics. Furthermore, insulators are susceptible to radiation damage as well as to ionic discharge during the activation process. When activating by protons, the activation profile can be controlled somewhat by adjusting the proton beam energy and incident angle, but the resultant profile is largely dependent on the production cross-section of the reaction in question. A possible solution to this problem is to implant the radioactive species themselves, instead of activating the material already present. Only long-lived isotopes of sufficiently light elements are suitable, because of the limited range and of the damage induced by heavier species. The two best candidates are 7Be and 22Na. The dose required by implantation is roughly 104 times lower than in the case of activation, and the doping profile can be controlled more directly and precisely . Any material can be doped by this technique. 50 A possible method for doping with 7Be might be recoil implantation via the H(1Li,7Be)n reaction [29]. A much better technique is direct implantation with an accelerated radioactive beam. This eliminates essentially all the problems associated with recoil implantation, and allows the user to create precise, pre-defined doped layers in any material. A classic example of this approach is the case of near-surface doping of ceramic materials. As a quantitative guiding case, we consider a spot of 1 cm2 which has to be doped to a depth of 2 microns. This would require radioactive beams with an energy in the range of 1 to 3 MeV. Implantation requirements are typically in the range of 1011 to 1013 nuclei. Establishing a routinely-available facility at !SAC for doping special components would be of direct use to many industrial users. References [1] H. de Waard and L. Niesen, in Mossbauer Spectroscopy Applied to Inorganic Chemistry, Vol. 2, ed. G.J. Long, Plenum, New York, 1987, p. 1. [2] H. de Waard, Hyp. Int. 40 (1988) 31. [3] G. Langouche, in Mossbauer Spectroscopy Applied to Inorganic Chemistry, Vol. 3, ed. G.J. Long, Plenum, New York, 1989, p. 445. [4] D.L. Williamson et al., in Hyperfine Interactions of Defects in Semiconductors, ed. G. Langouche, Elsevier, Amsterdam, 1992, Chapter 1. [5] G. Langouche, Solid State Phenomena 27 (1992) 181. [6] J.A. Sawicki, Mat. Sci. Eng. 69 (1985) 501. [7] J.A. Sawicki, in Industrial Applications of the Mossbauer Effect, eds. G.J. Long and J.G. Stevens, Plenum, New York, 1986, p. 83. [8] G. Weyer, Hyp. Int. 29 (1986) 249. [9] P.B. Russell et al. Nucl. Phys. A210 (1973) 133. [10] Y. Yoshida ef al., Phys. Rev. Lett. 61 (1988) 195. [11] P. Schwalbach et al., Phys. Rev. Lett. 64 (1990) 1274. [12] A. Lerf and T. Butz, Angew. Chern. Int. Ed. Eng. 26 (1987) 110. [13] "Hyperfine Interactions of Radioactive Nuclei", ed. J. Christiansen, in Topics in Current Physics, Vol. 31, Springer, 1983. [14] T. Klas et al., Phys. Rev. Lett. 57 (1986) 1068. [15] B. Kastelein, H. Postma and J. Andriessen, Hyp. Int. 75 (1992) 315. 51 [16) P. W. Martin et al., Hyp. Int. 77 (1993) 315. (17) S. Unterricker, Hyp. Int. 60 (1990) 709. [18) E. E. Omel'yanovskii and V. I. Fistul, Transition Metal Impurities in Semiconductors, Adam Hilger Ltd., Bristol and Boston, 1986. [19) I. Kelson, J. Phys. D: Appl. Phys. 20 (1987) 1049. [20) I. Kelson et al., J. Appl. Phys. 69 (1991) 1147. [21) M. Le Gros, A. Kotlicki and B. G. Turrell, Hyp. Int. 77 (1993) 131. [22) M. Le Gros, A. Kotlicki and B. G. Turrell, Hyp. Int. 77 (1993) 203. [23) "Proceedings of the International Symposium on Nuclear Orientation and Nuclei far from Stability", eds. B. I. Deutch and L. Vanneste, Hyp. Int. 22 (1985). [24) "Proceedings of the 1st International Conference on On-Line Nuclear Orientation", eds. N. J. Stone and J. Rikovska, Hyp. Int. 43 (1988). [25] "Proceedings of the 2nd International Conference on On-Line Nuclear Orientation", ed. K. S. Krane, Hyp. Int. 75 (1992). [26] G. Dearnaley, Nucl. Instr. Meth. B40/41 (1989) 731. [27] P. Sioshansi, Materials Engineering, February, 1987. [28) J. D. Destefani, Inc.-Metal Progress 39 (1988) 1. [29] J. Asher et al., Nucl. Instr. Meth. 178 (1980) 293. 2.6 Biomedical Applications 2.6.1 Introduction At present, radionuclides are produced mainly through the use of accelerators and reactors. In most cases low-energy accelerators such as cyclotrons produce high quality radionuclides that need little or no separation from undesirable isotopes. However, radionuclides produced via high-energy reactions are for the most part contaminated by isotopes that are stable or radioactive, both of which present problems in the final use of the product. Reactor-produced nuclides are of very low specific activity which shortens their useful life time. ISAC with its isotope separator and accelerator has great potential for the production and separation of radionuclides that are of interest in the biosciences. All indications are that the future growth area for radionuclide use in medicine will be in therapeutics. The production of monoclonal antibodies, or their fragments, labelled with a radiotoxic isotope is a highly promising strategy for directing radiation specifically 52 Table 2.6: Physical properties for therapeutic radionuclides Physical Property Comments ·----·- !----·-- ·--·- - ·----·- -a.) Physical T long enough to localize in tumour (6-200 h) b) Gamma ray energies and intensities little or no "}'-rays (140 keV may be an advantage) c) Parent-daughter relationships stable decay products d) Ratio of penetrating to non- same as b) above penetrating emissions e) Production how available is the radionuclide? f) Particle radiation (/3, a , IC and relatively high linear energy transfer (LET) auger electrons) Table 2. 7: Chemical properties for therapeutic radionuclides Chemical Properties Comments a.) Stability of ra.dionuclide-protein bond necessary to transport radionuclide to proper site b) Specific activity number of labels per molecule obtainable c) Retention of immunological nature as does nature of antibody change function of amount of carrier with activity labelling? d) N onra.dioa.ctive carrier related to b) and c) above as well as what burden does metal ion contamination result in? by biochemical means to cancer cells wherever they are growing in the body. The use of ra.diola.belled, monoclonal antibodies has received considerable attention for the detection of tumour tissue [1, 2] . Chelating techniques [3, 4, 5] have made possible the use of metals for labelling, which overcomes many of the problems associated with the use of 131 I labelling. Of the many variables that are associated with tumour antibody imaging and therapy, the physical and chemical properties of the potential radiolabels are the easiest to assess . Tables 2.6 and 2. 7 list the properties and the criteria. sought for therapeutic nuclides. With the ability to attach radionuclides that emit high LET radiation to compounds that have specific uptake in tumorous tissue, the treatment of cancers will be refined to the choice of ra.dionuclide as well as a vehicle for delivery. The roadblocks to this· development include the fact that, for the most part, the radiotoxic nuclides are produced in reactors and thus are inherently of low specific activity. The low specific activity limits the useful shelf-life at best and at worst prevents effective labelling of the nuclide to the desired compound, usually a peptide fragment. 53 2.6.2 On-line separation Isotope separators that are located at the source of production offer the opportunity to produce and separate the desired radiotoxic species in high yield and purity. There has been relatively little work in this area, with the ISOLDE facility at CERN being the only on-line separator available until recently. The projects that hold the most promise for ISAC include: the production and separation of radiotoxic nuclides for use in therapy; and exploration of the feasibility of using the accelerator for generator production via ion implantation. The implantation work can take the form of investigating a number of generator systems that could be more efficient if the parent nuclide is held in a matrix more suitable for the separation of two species (that is daughter from parent). Also, this technique could be used to explore the possibility of making positron sources for calibrating positron tomographs. 2.6.3 Therapy Therapy radionuclides include the a-emitting nuclides 211 At and 225 Ac. Their production would be through the use of a thorium carbide target, for example, to produce 211 Rn as a precursor for 211 At. The 225 Ac would be produced directly. Candidate radionuclides for radioimmunotherapy 47Sc 67 Cu 11 As goy 1o5Rh 1o9pd 111 Ag 131 I 142pr 149pm 153Sm 159Gd 166Ho 111Lu 186/1ssRe 194Ir 199pt 211 At Reactor-produced radionuclides that potentially could be prepared via ISOL 105Rh 109pd 111 Ag 142pr 149pm 153Sm 159Gd 166Ho 177Lu 186Re 1ssRe 194Ir 199 Au Several groups have demonstrated the radiotoxic characteristics of 211 At in vitro and in vivo. The labelling of complex molecules with 211 At has also been performed at a number of centres. However, the availability of 211 At is still limited to the very few locations possessing accelerators capable of extracting a particles with an energy of about 30 MeV. Because of its short half-life (7.2 h), 211 At cannot be shipped very far without losing large quantities to decay. We explored the possibility of producing 211 Rn (h = 15 h) with the intent 2 of investigating the preparation of 211 Rn/211 At generators via an on-line isotope separator using the TISOL facility [6]. Using a uranium oxide target and 500 MeV protons, production rates of 1.5 x 107 nuclei/sf pA have been observed [7]. 2.6.4 Implantation The implantation work would follow up the work begun at ISOLDE with the 81 Rb/81mKr generator system. The 81 Rb would be implanted into a number of substrates that would serve as a support for the Rb ions, allowing the 81mKr to escape from the matrix; since the decay product is a noble gas, it should be easily washed off. Other generator systems to be explored include the 68Ge/68Ga, 82Sr/82Rb and 62 Znj62Cu systems. 54 lmplantion of 68Ge creates the possibility that very small sources of positron emitters could be made into a wide variety of configurations for the calibration of positron tomographs. There are other possible radionuclides that may have utility as sources for PET calibration, such as 85 Kr which has a .514 keY 1-ray which can be used to test scatter and random corrections in positron tomographs. The need to separate this isotope from other krypton isotopes is based on the desire to form a point-like source of the smallest possible volume. References (1] D.A. Scheinberg, M. Strand and O.A. Gavson, Science 215 (1982) 1511. (2] D.J. Hnatowich et al., Science 220 (1983) 613. (3] B.A. Khaw et al., Nucl. Med. 23 (1982) 1011. (4] S.M. Larson et al., J. Nucl. Med. 24 (1983) 123. [5] B.A. Rhodes et al., J. Nucl. Med. 27 (1986) 685. (6] T.J. Ruth, M. Dombsky and J.M. D'Auria, Bull. Amer. Chern. Soc. (August, 1993). (7] G. Beyer, H.L. Ravn and Y. Huang, Int. J. Appl. Radiat. Isot. 35 (1984) 1075. 55 56 Section 3 RADIOACTIVE BEAMS AND FACILITIES 3.1 Introduction The radioactive beams facility proposed for TRIUMF is intended to address the science pro-gram outlined in Section 2 and capitalizes on experience gained from the present TISOL facility. Although there are several radioactive beams facilities in the world either in operation, under construction or proposed [1], TRIUMF has the opportunity for timely construction of the next major facility and the only one utilizing an existing high-current, high-energy proton production beam. Such a facility would be unique in the world in terms of the range and intensity of available radioactive projectiles. The technology required for this next generation facility is generally well known and takes advantage of the present thick-target isotope separator system in operation at TRIUMF. New developments are required in the handling of the highly-radioactive production targets and in the initial stages of the acceleration of the radioactive ions. TRIUMF is well positioned to resolve these developmental issues from extensive experience gained over the last 20 years with high current production targets, and with the experienced accelerator and ion source groups which are available. 3.2 Production of Radioactive Species Radioactive nuclei can be produced using fission, spallation, fragmentation, fusion evapora-tion, deep inelastic collisions and nucleon transfer reactions. Many different facilities using a variety of techniques have been constructed and used over the last 25 years as described by Mueller and Sherrill [2]. Each production method has its range of applicability and limits of production. The methods which have had the greatest success in producing the widest range of radioactive species are high energy, light ion nuclear reactions (such as proton induced spallation, fragmentation and fission) and heavy ion projectile fragmentation utilizing very energetic projectiles. Selection and separation of the species of interest from other products 57 and the delivery of a usable beam to an experimental area is accomplished by: 1) the Isotope Separation On-Line (ISOL) technique, in which radioactive species are produced in a target and subsequently transferred to an ion source; and 2) the Projectile Fragmentation Method (PFM). The latter is characterized by a peripherial interaction of the stable projectile with the target nucleus that leaves the radioactive fragment with much of the initial momentum and a small angular spread. The ISOL approach produces radioactive beams of greater use-fulness to the scientific program described above. Also, it is more suitable for the production of such beams at a facility like TRIUMF. The PFM technique will be described briefly to illustrate the complementary nature of these two approaches. 3.2.1 Projectile fragmentation method (PFM) In the PFM approach, the energy of the incident projectile is transferred to a recoiling fragment of either the target or of the projectile. This technique is most applicable with energetic heavy ions as a production beam. The beam passes through a thin target from which energetic radioactive products can escape. These nuclei are then captured in an optical system downstream from the production target, mass- and energy-analyzed, and transported to an experimental area. Production and transport delays in this method are very short ( < 1 ps) which makes short-lived species readily available. However, the phase space of the beam produced is determined by the nuclear reaction kinematics and is usually very large. This makes good energy resolution and purification of the beam difficult to achieve. Cooling and deceleration could be contemplated but might lead to difficulties comparable with the acceleration of ISOL beams. This approach has been applied successfully at a number of facilities around the world [2, 3). Beams of various radionuclei have been produced with intensities up to 109 /s. In general, the optimum production energy of these beams is in the range 50-2000 MeV fu. The variety of available beams in the low Z region is quite large; however, it is limited in the medium and heavy mass region. 3.2.2 The ISOL method In this method, radioactive species are produced in a target material by nuclear reactions and then transferred continuously to a suitable ion source where they are ionized, extracted, and subsequently mass-analyzed to form a radioactive beam. Many different target materials and transfer methods (both physical and chemical) have been used to optimize the speed, efficiency and selectivity of this transfer process for different radionuclei. Production target materials can take the form of powders, foils, pellets or liquids and must often be maintained at high temperatures (~ 2000° C) to enable the escape and transfer of nonvolatile species. Delay times are inherent to these processes and short-lived species (T 1 < 1 ms) are not available by this production method. The techniques involved in using2 the ISOL method (thin and thick target) have been developed to the stage that radioactive beams of about 80 of the 92 naturally occurring elements can be formed. A more detailed review of these techniques can be found elsewhere [4, 5). 58 The ISOL method has been applied using many different production beams ranging from thermalrwutrons, to light ions, and to low- and high-energy protons . Each type of beam has its advantages and disadvantages in terms of available intensities, reaction cross sections and energy deposition in the target [6]. In general, it has been found that high energy protons (0.5-1 GeV) are most effective for the production of a broad range of radionuclides due to a combination of high cross sections, available intensities and low dE/dx in target materials. Reaction cross sections and proton range increase with energy and a proton beam > 1 Ge V would seem to be the optimum production source. However, beam intensities presently available at these energies are limited and the high-intensity 500 MeV beam at TRIUMF represents one of the most attractive production beams available in the world. This beam energy allows the use of very thick targets, which in turn leads to relatively high radioactive ion intensities(::::::: 1011 /s after the separator). Slight variations of this method include the thin-target, gas-jet approach which provides a long separation between the target and the ion source, but uses only a thin layer of the target for production [7]. This method can produce radioisotopes not available with the standard !SOL approach because of the chemical properties of the element of interest. The beams produced by the ISOL method are usually singly-charged, positive ions which are efficiently produced in the ion sources. The emittance is usually very good because the ions have been thermalized and are extracted from the source in a controlled manner at a precisely fixed energy. The energy of the beams produced, however, is limited by the voltage (~ 60 kV) that can be applied to the target and ion source in the hostile environment of the production region. To obtain beams of higher energy some form of acceleration is required. The options for acceleration include tandems, cyclotrons, synchrotrons and linacs [7, 8]. A tandem is well matched to the DC low-energy beams available from an JSOL and has easily varied energy and low energy spread. Unfortunately, only negative ions can be accelerated and these must be stripped in the high voltage terminal. The efficiencies of negative ion production and losses on stripping limit the application of this accelerator option. A cyclotron presents the attractive feature of simultaneous acceleration and mass separation of the beam. However, the final beam energy is proportional to the square of the charge-to-mass ratio of the accelerated ions and high-charged-state ions are required to realize this option. This would require the development of efficient high-charge-state ion sources or some form of preacceleration to an energy where stripping to higher charge states can be accomplished efficiently. A linac, preceded by an RFQ bunching and preacceleration section, seems the most attractive option for acceleration of presently available ISOL beams. The RFQ has a large acceptance for low-velocity ISOL beams and produces a beam than can be stripped and efficiently matched to a following conventional or superconducting linac. Energy variability is available in principle and the energy range can be extended by the addition of accelerating sections. The high acceptance and transmission of this accelerator option for presently available ISOL beams have made it the preferred choice for the ISAC facility. TISOL An example of a thick-target, on-line isotope separator is the TISOL facility in-stalled on the medium intensity ( < 10 pA) beam line at TRIUMF [9, 10]. Its original 59 Figure 3.1: A schematic vertical representation of the upgraded TISOL facility. purpose was to provide a location to allow research and development studies related to the installation of a full accelerated radioactive beams facility ISAC and to allow establishment of a modest experimental program involving radioactive beams. The vertically-positioned TISOL is a thick-target separator based upon methods pio-neered at ISOLDE at CERN. Figure 3.1 presents a schematic representation of the present TISOL layout; complete descriptions can be found elsewhere [10]-[12]. Two types of ion source are currently in use at TISOL: a heated surface source; and an ECR (electron cyclotron resonance) system. The surface-source ionizer is simply a material (metal) with a relatively-high work function, and the elements ionized are those with rela-tively low ionization potentials. In general, metal foils of either Ta, Re or W are used inside a graphite cylinder which is heated by resistance heating up to about 2000° C; radioactive products can diffuse from the hot thick target through a small graphite chimney into the 10mzer. The ECR ion source consists of two quartz tube liners, with the inner one serving as the plasma chamber with the RF coupled in radially. An iron yoke is used in the generation of the magnetic field, with the ECR mirror condition being axially produced by an iron ring in the middle of the source. Radial electron confinement is provided by a SmCo hexapole (surface field 0.52 T). The source is presently operated at 6.4 GHz, but has also been used 60 'H r'jie I ..... Li Be 8 c N 0 , Ne lill,.:t 4 • ..... ,.,.. ..... • IY Na W& Al Si p s Cl Ar lill,.u II llli..t:t 14 II It l'..oillll ll...oill K Ca Sc Ti v Cr MD Fe Co Ni Cu Zn Ga Ge Aa Se Br Kr liii,.Jt 10 II II ., .. • • ., • .. 10 .._,I • D .. • ..... Rb Sr y Zr Nb Mo Tc Ru Rh Pd AI Cd In Sn Sb Te I Xe lill,.n lh..• It 40 41 .. ., 4o4 .. 41 ., .. ...... 10 II II ., ...... c. Ba La Hf Ta 'W Re Oa lr PL Au Ha Tl Pb Bi Po AL Rn IIi.. IS II ., ,. " ,. ,. ,. ., ,. ,. 10 tl • a .. • ..... Fr Ra Ac ..,.., .. .. 104 1115 IDI 111'7 IDI Ce Pr Nd Pm Sm Eu Gd Tb D7 Ho Er Tm Yb Lu .. •• ID tl u ., .. tl .. " .. •• ..,.,. ,. LANTHANIDES ACTINIDES Th Pa u Np Pu Am Cm Bk Cf £8 Fm Md No Lr ID tl tl a .. .. .. " .. " IDO Ill 101 ID:t ~ IONIZED 'WITH TlSOL ECR SOURCE ~ IONIZED 'WITH TlSOL SURFACE SOURCE Figure 3.2: A standard representation of the periodic table of the elements with an indication of the elements for which ion beams have been generated at the TISOL facility using the ion source indicated. up to a maximum frequency of 10 GHz. The inner quartz tube can be removed together with the coupling piece to the target assembly. To reduce heat problems and resultant sparking, the plastic insulation at the coupling end is cooled by a closed freon circuit. The target unit can be a pulled away from the coupling. 3.2.3 Radioisotope production Target materials used on the TISOL facility include powders of SiC, SeC, UO/C, ZrC, Nb, LaC2 , MgO, A1N, CaO and NaA1Si04 (zeolite) and foils ofTa, Hf, Zr, Ti and Mo. In general, the release of a particular element from a particular target matrix varies as a function of the species involved, the form of the target, the temperature of the target, etc. The half-life of the released isotope is also a factor in the observed production yields. Complete details on the production of specific radioisotopic ion beams can be found in Refs. [10]-[12] . Figure 3.2 displays the standard periodic table of the elements with an indication of what beams have been produced using which ion source at the TISOL facility. 61 3.3 General Concept of ISAC To effectively exploit the research opportunities presented in section 2, and to fulfill its role as the prototype of the next generation facility, the ISAC facility will be required to produce beams of radioactive heavy ions with a variety of properties. The most important feature of ISAC will be the available radioactive beam intensity which will allow a greater range of key experiments than possible with present facilities. With more intense beams, less sensitive experimental techniques can be applied. A significant number of projectiles should be available in sufficient intensity to allow the possibility of studies of reactions with relatively low cross sections. A benchmark intensity of 1010 particles/s has been deemed sufficient for studies of the low cross-section, radiative-proton-capture reactions of interest to astrophysics. The highest priority in all ISAC design matters will be minimizing losses or maximizing beam intensity available to the user. There should be a wide variety of exotic, radioactive and isomeric probes available both for the study of their inherent properties and for use in a variety of studies. The requirements of the program indicate a need for a wide range of species across the chart of nuclides from A '-'-' 6 ur io A .::::; 240. Beam contamination poses one of the greatest problems in many of the planned studies. For example, in the search for, and study of, rare exotic nuclides, the overwhelming flux of ions from the target system presents a problem of discrimination, which sometimes must be of the order of 1 part in 1012 . In some instances, and for certain elements, careful design of the target material and of the ionization system may provide chemica.] selectivity and isobaric purity. In other cases, the experimental approach may provide the necessary discrimination. Complete isobaric mass resolution is not attainable in all regions of nuclear masses. Given these considerations, the isotopic purity of the beam should be of the order of 1 in 105 , and an isobaric purity of up to 1 in 104 should be attainable in selected cases, and as needed by the experiment. Special techniques may be required to achieve ultra-high purity when necessary for critical experiments. A transverse beam emittance of the order of 0.2511" mm·mr (normalized) should be ac-cepted by the accelerator and a longtitudinal emittance of the order 100 keV·ns should be available. Such beam properties are competitive with the properties of stable beams and give the experimenter the opportunity to exploit useful experimental approaches, such as time-of-flight mass identification. The proposed radioactive beams facility consists of a thick-target, on-line isotope separa-tor followed by a linear accelerator. A wide range of radioactive species at both the proton and neutron limits of stability will be produced by target fragmentation, spallation, or fis-sion reactions. Final radioactive ion-beam intensities as high as 1011 /s for species close to stability will be possible, given the availability of at least 10 J-LA of production beam and . targets of thickness~ 100 gfcm2• Based on present technology of thick-target ISOL devices, isotopes from about 80 elements could be produced in the form of ion beams; a conservative estimate is that 1000 beams could be available. Estimates of expected radioactive beams intensities are given in Section 3.5.4. The low-energy ion beam from the separator can be directed either into the experimental area or captured (for qf A ;::: 1/30) and accelerated by a 62 radiofrequency quadrupole (RFQ) linear accelerator, followed by several stages of drift tube linear (DTL) accelerators. 3.4 Projected Intensities of Radioactive Beams At !SAC the desired radioactive beams will be produced by spallation, fission and fragmen-tation of the target using 500 MeV protons. The final radioactive beam intensity I is given by where a is the production cross section; </> is the proton beam intensity; N(i is the target thickness; E1 is the overall efficiency resulting from release efficiency of the target, transmis-sion efficiency of the transfer tube, and decay losses in the target/ion source system; E2 is the ionization efficiency of the ion source; and TJ1 reflects loss due to stripping in the acceleration system (if used). Other losses are considered insignificant. A more thorough discussion of these concepts can be found in Refs. [4, 13]. Cross sections for spallation processes can be calculated using the semi-empirical formulas of Silberberg and Tsao [14], which is a useful approach when measurements do not exist. The diffusion efficiency of the target release and effusion through the transfer tube depends on the chemistry of the released element, the elemental content of the target matrix, the elemental content of the transfer tube, and the temperatures in the environment. A more complete discussion can be found elsewhere [13, 15]. Observed ionization efficiencies for a range of elements in different types of !SOL ion sources are available [13, 16]. Table 3.1 presents estimates of projected beam intensities from ISAC for selected radionuclides of interest. These were chosen partially as a result of the interests of the scientific program discussed in Section 2, and partially as a result of the comparison presented in the NUPECC report of European facilities . Species heavier than A = 30 are not shown in the medium-energy regime as they may not be accelerated (unless they can be easily multiply-charged), but species up to A = 240 will be available in the low-energy area. The target thickness corresponds to that in which the beam energy is reduced from 500 to 300 MeV, but is not longer than 20 em, the maximum length to be used. The cross sections were calculated using the latest versions of the semi-empirical codes of Ref. [14]. Wherever possible, expected intensities were compared with those observed from TISOL or ISOLDE to assess their validity. Efficiencies for producing multiply-charged ions from the ISOL ion source were taken from studies at TISOL using its ECR ion source and these are used, where needed, for species of A > 30 in the RFQ. If stripping were to be used, the residual intensity of the beam, as a function of Z, would be as shown in Fig. 3.3 [17]. 63 0 t:;:: -' ' -· ·· - Proton Rich --Stable . . ' \ ' ' ' ' Neutron Rich \ \ \ \ \ L ' \ Io~~~~--~~~~~·~~· ~- ~~ 0 20 40 60 80 I 00 z Figure 3.3: The effect of stripping on beam intensity. 3.5 ISAC in the World Situation 3.5.1 Overview At present there are at least four operating PFM high-energy radioactive beam facilities, and one operating accelerated-beam facility of the ISOL type. For most laboratories the approach is to adapt older, but available, accelerator structures to become operational quickly and for a reasonable cost. Additional information on some of these facilities is given below, while more detailed material can be found in Refs. [1, 2, 5, 13, 18, 19]. 3.5.2 PFM radioactive beam (RB) facilities Given the nature of ISAC (low-energy, !SOL-based) only minimal information will be given on PFM facilities; it is presented in Table 3.2. Additional information can be found in Refs. [2, 3, 20, 21]. 64 Table 3.1: Projected beam intensities (particles/s) for ISAC for 10 J.LA of protons on target. PRODUCTION EFFICIENCIES EXPECTED RATES Beam T.~. Target Thickness Rate (.1a (.2b f1 X f2 Low ~ (g/cm2) % % % energy 11 He 122 ms uc2 122 3.2xl0!1 30 10 3 9.6x 107 sLi 842 ms Ta 121 1.7x1010 40 95 38 6.7x109 11Li 9 ms Ta 121 2.1 X 107 0.32 95 0.30 6.3x104 7Bed 52 d c 45 9.3x1011 80 1.0 0.80 7.4x 109 11C 20m MgO 29 4.1x1011 80 10 8.0 3.3x 1010 1sc 747 ms MgO 29 4.7x108 30 10 3.0 1.4x 107 13N 10m Zeolite 33 2.0x 1012 80 30 24 4.8x 1011 1sN 624 ms Zeolite 33 4.1 X 109 40 30 12 4.9x108 140 71s Zeolite 33 6.7x1011 60 30 18 1.2x1011 150 2.0 m Zeolite 33 2.1 X 1012 60 30 18 3.8x1011 220 2.3 s Pt 121 5.8x107 6.0 30 1.8 l.Ox106 17F 65 s SiC 45 2.8x1011 58 1.0 0.58 1.6x109 19Ne 17 s Zeolite 33 l.Ox1011 80 30 24 2.4x1010 26Ne 162 ms uc2 122 5.6x109 80 30 24 1.3x108 20Na 446 ms SiC 45 4.5x1010 24 95 23 l.Ox1010 30Na 53 ms uc2 122 3.9x108 40 95 38 1.5x107 26m AI 6 s SiC 45 2.5x1012 15 10 1.5 3.8x 1010 34Ar 884 ms Ti 90 1.2x109 40 40 16 1.9x108 46Ar 88 s vc 90 1.1 X 108 40 40 16 1.8x107 37K 1.2 s TiC 50 6.0x 109 40 90 36 2.1x109 64Ged 64 s Zr02 92 5.3x 107 2.5 1.3x 106 73Se 7h Zr02 92 2.5x 1011 80 10 8.0 2.0x 1010 74Kr 12m ZrC 92 5.2x109 80 40 32 1.7x109 91Kr 8.6 s uc2 122 3.9x 1010 72 40 28 1.1 X 1010 lOscd 56 m Sn 108 2.9x1011 8.0 40 3.2 9.3x 109 111 In 2.8 d Sn 108 1.2x1012 80 40 32 3.8x 1011 lOssn 10m TeCl4 61 1.4x109 8.0 50 4.0 5.6x 107 132Sn 40 s uc2 122 2.0x109 4.9 50 2.4 4.8x107 119Xe 5.8 m LaC2 103 2.5x1010 80 60 48 1.2x 1010 142Xe 1.2 s uc2 122 3.9x1010 44 60 26 l.Ox 1010 121cs 2.3 m La 110 7.4x1010 80 95 76 5.5x 1010 16oyb 4.8 m Ta 121 7.0x1010 0.13 9.1 X 107 210Fr 3.0m ThC 121 6.6x 1010 80 95 76 5.0x 1010 aEstimated release efficiency from the target (based upon present technology). bEstimated ion source efficiency (based upon present technology). ccalculated stripping efficiency. dRequires development. 65 'lc Medium energy 0.9 8.6x107 0.85 5.7x109 0.85 5.4x104 0.7 5.9x109 0.6 2.0x 1010 0.6 8.4x106 0.55 1.9x 1011 0.40 2.7x108 0.50 6.0x1010 0.50 1.9x1011 0.40 4.0x105 0.48 7.8x108 0.45 1.1 X 1010 0.35 4.5x 107 0.43 4.3x 109 0.35 5.2x 106 0.38 1.4x1010 0.34 6.5x107 0.20 3.6x106 3.5.3 ISOL-based accelerated radioactive beam facilities Louvain-la-Neuve, Belgium Using the K=30 CYCLONE 30 high-intensity (500 J.LA) proton cyclotron as the production system and the K-110 CYCLONE cyclotron as the booster, the first accelerated RB (13 N) was produced here; they now have 6 additional beams available (6He, 11 C, 18F, 18•19Ne and 35Ar). An upgrade to the facility, the ARENAS3 project, has been funded; it includes a new booster accelerator to cover the energy range between 0.2 and 0.8 MeV fu for (6.5 > Afq > 13). The K= 30 will still be used initially as the production machine, which limits the number of possible beams. HRIBF, Oak Ridge National Laboratory, USA This facility is under construction and the production system is expected to be tested later this year. The K=105 ORIC proton cyclotron is the production facility and, following charge exchange, negative ion beams from the ISOL system will be accelerated with a 25 MV tandem accelerator. Beams up to mass 80 will be accelerated from about 0.5-5 MeV /u. The first accelerated beams are expected to be 17F, 32 Cl or 64Ga. Due to the low energy of the production accelerator, the number of radioactive beams is limited. SPIRal, GANIL, France Energetic heavy ion projectiles will be used to produce a wide range of products primarily via projectile fragmentation in a graphite target, but also via target fragmentation with different target materials. A K=260 cyclotron will be the booster aml the final RB will exhibit energies from 2 to 29 MeV/u. A test system (SIRa) is in operation to develop a method to handle the high power density that results from the short range of the heavy ion production beam. ISOLDE, CERN, Switzerland The ISOLDE facility has been the premier thick-target !SOL device in the world for about 26 years, initially using 600 MeV protons from the synchrocyclotron and now using 1 GeV protons from the CERN PS-Booster. A wide range of radioactive ion beams in relatively high intensities is available. These beams have been used in an extensive science program mainly in nuclear studies far from stability and in condensed matter physics. Very recently a proposal (REX-ISOLDE) has been approved to install a system similar to the linear accelerator system at Heidelberg. This would lead to the acceleration of ions with A/ q < 9; coupling of this system to an electron beam ion source is being studied to extend the range of accelerated masses. INS, University of Tokyo, Japan A prototype facility for theE-ARENA area of the Japanese Hadron Project is presently under construction at INS. It will make use of the existing K=68 cyclotron as the production accelerator capable of delivering 10 J.LA of 40 MeV protons on target, as well as some other light ions. Acceleration will be done with a split-coaxial RFQ followed by an interdigital-H linac. The output energy for species with Afq < 30 is from 200 to 800 keV /u. This is the first RB facility employing a linac; while 100% transmission is expected, the duty factor is only 30%. It has also been proposed that 66 Table 3.2: PFM radioactive beam facilities Facility Country GSI Germany ( Gesellschaft fiir Schwerionenforschung) GANIL France (Grand Acccelerateur N ationale D 'Ions Lourds) RIKEN Japan (Inst. of Physical and Chemical Research) NSCL U.S.A. (National Superconducting Cyclotron Laboratory) LNS Italy (Laboratorio Nazionale del Sud) ADRIA Italy (Laboratori Nazionali di Legnaro) 8 upgrades planned or proposed bradioactive beams not yet produced 67 Beam Energy Acceleration System (MeV /u) 0.5-2 LINAC (20 MeV /u) 50-2000 Synchronotron 30-100 2 K = 400 cyclotrons 20-135 K = 540 cyclotron8 50-200 K = 540 cyclotron8 n.a.b K = 800 cyclotron n.a.b LINACS the prototype be moved to the KEK laboratory at Tsukuba where 500 MeV protons at 2 pA are available. Excyt, Catania, Italy In this proposed facility, energetic heavy ions would be the pro-duction beams and an !SOL-type device would reside on a 150 kV platform. Negative radioactive ions would be produced and accelerated (A < 40) using an available 15 MV Tandem. RNB, Moscow, Russia In connection with the Moscow meson factory, a radioactive beam project is believed under construction to utilize the 500 pA of 600 MeV protons as the production beam. A linac based on RFQ and interdigital-H structures will produce RB's (with Ajq < 60) with energy up to 6.5 MeV ju. PIAFE, Grenoble, France Thermal neutrons from the ILL (Institut Laue Langevin) reactor would be used to produce radioactive fission products from a 235U target placed close to the core. Mass selected ions would be injected into the SARA accelerator complex consisting of a K=88 injector cyclotron and a K=160 separated sector cyclotron. The aim is to produce RB's with energies from 2 to 10 MeV ju. Considerable Rand D is needed to resolve some of the technical problems connected with this proposal. Others A few other laboratories or groups (ANL, LAMPF, RAL, ISL) are developing proposals or are involved in related projects. While there are several operating RB facilities using the PFM, there is only one facility producing low-energy RB's with the ISOL approach. There are some facilities planned, pro-posed, or under construction using the ISOL method, but several of these have challenging technical problems which must be overcome to make them useful facilities. Given the exten-sive experimental programs possible at such facilities, and the large number of laboratories around the world interested in this field, it is expected that accelerated radioactive beams will provide the basis for important and fruitful research for many years. 3.5.4 Con1parison with existing facilities Table 3.3 provides a comparison of the projected production of selected radioactive species for facilities existing or under construction in Europe and the U.S. with the production expected with the proposed ISAC facility. The projected intensities come from the recent NUPECC report [13], from the HRIBF proposal [22], from Ref. [23] for GANIL, and from the present proposal. Other facilities under construction are not included in this comparison because of uncertainties in, or absence of, projected intensities. 68 Table 3.3: Projected production per second of selected radioactive species8 Beam T1 Arenas3 b HRIBFC REX-ISOLDEd SPIRale 2 Oak Ridge GANIL 8He 122 ms NA NA 5.2x108 1.5 X 109 8Li 842 ms NA NA 2.2x109 D 11 Li 9 ms NA NA 3.0x104 D 7Be 52 d 7.5x108 NA 2.8x1010 D 11C 20m 7.5x109 2.0x108 1.9x 1010 D 13N 10m 1.6x 1010 NA 1.4x 1010 D 1sN 624 ms NA NA D 140 71 s NA 5.7x108 5.8x109 D 220 2.3 s NA NA 6.0x 106 D 17p 65 s NA 4.5x108 3.2x108 D 19Ne 17 s 7.0x109 NA 4.6 X 1010 3.8x 108 26Ne 162 ms NA NA 2.2x 106 D 20Na 446 ms NA NA 2.0x109 D 30Na 53 ms NA NA 2.6x104 D 34Ar 884 ms 1.8x109 NA 4.6x 108 1.1x107 64Ge 64 s NA 7.2x108 D 73Se 7h NA 8.0x108 1.7x1010 D 74Kr 12m NA NA 3.2x109 1.1 X 106 91Kr 8.6 s 4.4x 108 NA 2.4 X 109 (1. 7 X 108 ) wscd 56 m NA NA 1.6x109 D wssn 10m NA NA 5.8x107 D 132Sn 40 s NA NA 7.0x106 D 119Xe 5.8 m NA NA 142Xe 1.2s NA NA 6.6x107 (6.1 X 106) 121Cs 2.3 m NA NA 4.4xl010 D 16oyb 4.8 m NA NA D 210Fr 3.0 m NA NA 2.0x109 S D NA = not yet available; D = difficult to extract from target 8 Extracted from target/ion-source system, prior to acceleration. ISACr 9.6x107 6.7x109 6.3x104 7.4x109 3.3x1010 4.8x 1011 4.9x108 1.2x 1011 l.Ox 106 1.6 X 109 2.4x 1010 1.3x 108 l.Ox 1010 1.5x 107 1.9 X 108 1.3 X 106 2.0 X 1010 1. 7x 109 1.1x1010 9.3x 109 5.6x 107 4.8x107 1.2x1010 l.Ox1010 5.5xl010 9.1 xl07 5.0x1010 bFrom NUPECC Report [13]; assumes 10 pA intensity and K = 30 cyclotron. cFrom Ref. [22]; at this time beams have not yet been produced. dFrom NUPECC Report [13]; assumes 2 pA intensity. eFrom GANIL Newsletter No. 53, Dec, 1994; assumes~ 5 kW fcm3 (1 ppA) on target and on-line ECR ion source. f Assumes 10 pA intensity. SBased on yield observed at ISOLDE. 69 References [1] J.M. D'Auria, Nucl. lnst. Meth., in press. [2) A.C. Mueller and B.M. Sherrill, Annu. Rev. Nucl. Part. Sci. 43 (1993) 529. [3) G. Munzenberger, Nucl. Inst. Meth. B70 (1992) 265. [4) H.L. Ravn and B. Allardyce, in Treatise on Heavy Ion Science, Vol. 8, ed. D.A. Bromley, Plenum Press, New York, 1989, p. 363. [5) H.L. Ravn, Nucl. lnst. Meth. B70 (1992) 107. [6) H.L. Ravn et al., Nucl. Inst. Meth. B88 (1994) 441. [7) W.L. Talbert, Jr., in Proceedings of the Workshop on the Science of Intense Radioactive Ion Beams, Los Alamos National Laboratory, April, 1990, LANL Report LA-11964-C, p. 171. [8] L. Buchmann et al., Nucl. lnst. Meth. B26 (1987) 151. [9) K. Oxorn et al., Nucl. lnst. Meth. B26 (1987) 143. [10) M. Dombsky et al., Nucl. lnstr. Meth. A295 (1990) 291. [11] M. Dombsky et al., Nucl. lnstr. Meth. B70 (1992) 125. [12) L. Buchmann et al., Nucl. lnst. Meth. B62 (1992) 521. [13) Nuclear Physics European Collaborative Committee (NUPECC), Report of Study Group, European Radaioactive Beam Facilities, 1993. [14) R. Silberberg and C.H. Tsao, Astrophys. J. Suppl. 25 (1973) 315; --,Naval Research Laboratory Report {1973). [15] R. Kirchner, NucL Instr. Meth. B70 (1992) 186. [16] ISOLDE User's Guide, ed. H.-J. Kluge, CERN 86-05 (1986) and updates. [17] P. Bricault., private communication. [18] "Proceedings of the International Workshop on the Physics and Techniques of Secondary Nuclear Beams, Dourdan (France), March, 1992, eds. J.F. Bruandet, B. Fernandez and M. Bex, Edition FronW~res, Gif-sur-Yvette, France, 1992. [19] H.L. Ravn, in Proceedings of the Second International Conference on Radioactive Nu-clear Beams, Louvain-la-Neuve, Belgium, August, 1991, ed. Th. Delbar, Adam Halger, Bristol, 1992, p. 85. 70 [20] I. Tanihata, Nucl. Phys. A522 (1991) 275. j? i] ''Pr Y eeding:, of the Third International Conference on Radioactive Nucle<lr Beams'', East Lansing, Michigan, May 1993, ed. D.J. Morrissey, Edition Frontieres, Gif-sur-Yvette, France., 1993. [22] "A Proposal for Physics with Exotic Beams at the Hollifield Heavy Ion Research Facil-ity", eds. J.D. Garrett and D.K. Olsen, ORNL Report, 1991. [23] A. Villari et al., Ganil Newsletter No. 53, December, 1994. 71 72 Section 4 THE PROPOSED FACILITY 4.1 Introduction A schematic diagram of the ISAC facility design, developed over the past year, is shown in Fig 4.1 and the current spf>cifications an' given in Table ,l.l. The facility will be situated on a new, extracted-proton beam line (2A) to the North of the accelerator building. It consists of a thick-target ISOL, a two-stage linear accelerator (linac) and two experimental areas: a low-energy area utilizing the ~ 60 ke V beams from the separator; and a high-energy area utilizing beams from the accelerator. Each of these subsections of the facility will be described briefly below and more detailed technical descriptions of the ISOL facility and linear accelerator are given in Sections 5 and 6 of this proposal. The initial design located ISAC on an existing beam line ( 4A), which minimized the cost. However, in the course of the development of this design, numerous practical constraints and limitations were encountered which were caused by the location chosen. The decision to locate the facility on a new proton beam line gives the possibility of ultimately higher beam currents, provision of two target stations, and increased flexibility in the hot cell and containment area design, at a small increase in cost for a new extraction system and primary proton beam line. 4.2 The ISOL Facility The thick-target ISOL is the source of separated radioactive beams for ISAC and therefore is crucial to the successful operation and performance of the proposed facility. It is made up of a production beam, a target/ion-source complex, and a mass separator. Flexible target handling and containment of radioactivity are major considerations in the present design. 73 -B-.. lrnporial o s to m a • !10 root Scalo 0 10 I~ Yet••s -~-~-~-~-~--.J----· Figure 4.1: The proposed ISAC facility. ~ YALLT ) X A 500 MeV, 100 J.LA production beam will be extracted from the TRIUMF cyclotron and delivered to the target area. The target facility presents the most difficult technical problems in the ISOL design. High levels of radioactivity will be produced by the interactions of the protons with the production target and beam dump making effective handling techniques a requirement. Two target stations are planned; however, only one will be built initially. The target stations will be housed in a new heavily-shielded building which incorporates hot cell and storage facilities. All highly-activated components such as the production target, beam dump, ion sources and initial focusing devices will be located in this building along with the primary radiation shield. Services required to operate the target area components will also be housed in this building. A repair centre consisting of a hot cell, a warm cell and decontamination and storage facilities is directly connected to the target area. The highly-active components will be constructed as modules which can be accessed ver-tically by an overhead crane. Required repairs or modifications will be made by transferring a module from the target area to the repair facility. Mobile radioactivity is expected to be produced within the target facility; therefore, the circulation of air in the building is controlled and filtering is provided. This approach to the target facility design, described in detail in Section 5.3, is based on successful experience at TRIUMF and the other operating meson factories. For the thick production targets, a novel water-cooled target system was designed during a previous study to accommodate proton beams up to 100 J.LA in several different types of target material. Such target designs continue to be of interest but should not be required for the initial 10 J.LA operation anticipated here. Provision has been made for a variety of ion sources. Ion beams from these sources will be extracted at (nominally) 60 kV and delivered to the mass separator by an electrostatic matching section. The mass separator (resolving power of ::; 10, 000) bends in the horizontal plane and provides a separated beam to an electrostatic beam transport system which delivers this beam to either the low-energy experimental area or to the accelerator, both of which are located at ground level. 4.3 The Accelerator The ISAC accelerator must be capable of accepting ion beams produced by the ISOL sys-tem and accelerating them to 1.5 MeV /u with minimal losses. The chosen solution is a low-,8 radiofrequency quadrupole (RFQ) preaccelerator followed by a room temperature IH struc:ture linac. A carbon foil stripper is placed between acceleration stages to increase the charge-to-mass ratio of the ions injected into the IH linac. A detailed description of the proposed accelerator and the matching optics between stages is given in Section 6 of this proposal. The initial section of the accelerator is a 35 MHz RFQ which captures and bunches the de beam delivered by the ISOL system and accelerates it to 150 keV ju. This low frequency RFQ, technically the most challenging part of the accelerator, must operate CW and accelerate ions with charge-to-mass ratios down to 1/30. It may require a ±60 kV bias to ensure that 75 Table 4.1: Specifications of ISAC - = PRODUCTION SYSTEM Production Accelerator Projectile Energy Intensity Beam Size Time Structure Target System proton 500 MeV :S100 J.LA :S 1 cm2 ~cw Form foils, powders, molten metals Length ~20 em Power Deposited ~4 kW Temperature Range 25- 2600°C EXTRACTION SYSTEM Ion Sources Types Separator Systems Type Extraction Energy Mass/Mass Resolving Power Transmission Vacuum Surface, Laser, CUSP, ECR Plasma (FEBIAD, Bernas-Nier) Magnetic for mass analysis :S60 kV A <240/ R :S 10,000 >95% <2 x 10-7 torr ACCELERATION SYSTEM First Stage Type Input Energy Input qfA Emittance Transverse Longitudinal Output Energy Transmission Second Stage Type Input q/A Output Energy Range Resolution ll.E / E Energy Increment Transmission RFQ LINAC, Room Temperature 2 keV ju 1/30; (A> 30 for q > 1) :S 0.2571" mm·mr (normalized) 10071" keV·ns 150 keV ju > 90% LINAC- RT ~ 1/6 0.15-1.5 MeV fu < 0.1% <0.02 MeV/u >95% excluding stripper effects 76 all ions enter the RFQ with the design energy of 2 keV ju. For singly-charged ions the mass range which can be accelerated (with full efficiency) is limited to =:; 30. The acceleration of higher masses will require higher-charge-state ions (up to +4) to be delivered by the ISOL facility. The beam from the RFQ is passed through a matching section containing a carbon foil stripper, which increases the q/ A of the ions to a minimum of 1/6, before injection into the first tank of the IH linac. This linac accelerates the ions to 1.5 MeV /u and is split into four tank sections. The beam is variable in energy (from 0.15-1.5 MeV /u) and is delivered to the high-energy experimental area. Sufficient drift length has been left between the accelerator exit and the experimental target stations to allow the insertion of a buncher to alter the time structure of the delivered beam. 4.4 Experimental Areas and Facilities The concepts for the experimental areas and facilities are shown in Fig. 4.2. The low-energy area contains experimental stations for studies using ion beams directly from the mass separator. There are, in general, several types of experimental programs that define the characteristics of the individual stations: nuclear astrophysics measurements; nuclear decay studies (including perturbed angular correlations); on-line nuclear orientation studies; atomic and hyperfine interaction laser spectroscopy; measurements using trapped neutral atoms or ions; collection of long-lived radionuclides (for off-site studies or potential commercial use); and various materials science studies (both surface and bulk) using implantation and ultra-high-vacuum techniques. Provision has been made for a number of such experimental stations with beam delivered through electrostatic transport elements. The ion optical designs presented in Fig. 4.2 are only conceptual, although they have been patterned after the working optical designs for the ISOLDE-PSB area. The actual design of the experimental area to be constructed will include detailed ion-optical modeling to define the character of the beam lines. The experimental stations are envisioned to be largely modular in character so that, at least for the more modest apparatus, it will be feasible to change the station make-up to respond to changes in experimental requirements. The high-energy experimental area is located on a momentum-analysed-beam line at the end of the accelerator. A switching magnet will deliver the beam to one of three experimental stations as indicated schematically in Fig. 4.2. This high-energy area will provide space and equipment for nuclear astrophysics and nuclear physics reaction studies, and for materials science studies requiring penetrating radioactive beams. One station, located on the direct beam from the switching magnet, will house a recoil separator for studies of reactions of astrophysical interest. A general purpose scattering facility is envisaged an one side the recoil separator while, on the opposite side, a general-purpose chamber will be available into which, for example, a tilted foil array could be placed to polarize the beam. 77 Imperial Scale 0 5 10 20 30 50 Feet ·-.~·~·r-.... ~ .......... ~ .... ••• 40 .. Figure 4.2: Facility layout diagram. Section 5 THE ISOL SYSTEM 5.1 Overview The ISOL system consists of the primary production beam, the target/ion source, the mass separators, and the separated beam transport system. These systems together act as the source of radioactive beams to be provided to the accelerators or the low-energy experimental areas. A similar system has been developed at ISOLDE, CERN over the last 25 years, first at the SC [1] using 600 MeV protons and now using 1 GeV protons from the PS Booster [2]. The beam current is, however, limited to ~ 3 J.LA. At TRIUMF we intend to utilize the > 10 J.LA proton beam available from the cyclotron to produce a significant increase in the intensities of many radioactive species over those available at present. The proposed ISOL facility layout is shown in Fig. 5.1. Protons of 500 MeV energy with intensities in the region of 10-100 J.LA are extracted from the TRIUMF cyclotron. The target station will contain a production target and ion source, a beam dump, proton beam monitoring, and the first section of the ion beam transport. TRIUMF has many years of experience in handling the operational radiation fields and radioactive components associated with high current meson production targets. The production of radioactive beams from thick heav ',.' targets has the additional problem of the creation of large inventories of radioactivity, some of which is mobile, in the target/ion-source area. The handling of this activity and the target area components will be critical to the facility operation. A strategy has been adopted in which the target station is contained in a heavily-shielded building which is connected directly to a hot cell facility. This approach is based on the successful experience at TRIUMF of vertically servicing modular components embedded in a dose-packed radiation shield, coupled with the requirement for quick access to the production target and of containment of any mobile activity. Careful design of both the modular components and the remote-handling systems will be required to ensure the operational viability of this system. 79 00 0 ' '---"" -• W/~ -I=" = = I I I I I / 'Y_"f ~~ ~~ ~~ .~-I \____] I .----c ! i \j L.....-...J ,--p I I ~ l r r 1- -11: m 1-~~ 'it\- 1- ~- ~- A 1- ~-~ ~- r- - ..... f- 1-F ali~JOl~a1n';,ri 9" -- -f I L p I L -- -v -- --tHtHW n_ _r--~0~\J "-! [LJ bf;;;;l:l rt) f=\J~ \J L_ = L_ L_ c=::J t.t:= Figure 5.1: The ISOL facility. The effective operation of the ISOL system is crucial to the overall ISAC facility perfor-mance. It is therefore essential that the many technical issues inherent to operation with high proton beam current be resolved. Flexibility in the target/ion-source region will be required to meet present and future demands of the experimental program and must be incorporated into the facility design. The target/ion-source module will be the key component of the ISOL system. This module must be serviced, modified and exchanged on a regular basis to satisfy the varying demands of the physics program. Its design will address many of the difficult problems associated with the production of intense radioactive beams. These include high voltage services, containment of radioactivity, accommodation of various target/ion-source combi-nations, radiation-hard components, and ease of remote handling. A prototype module has been designed which resolves the majority of these problems; however, this will have to be fabricated and tested before a final design is established. The target/ion source represents the most technically-difficult aspect of a high-intensity radioactive beam facility. It is expected that existing target designs will accommodate 10 pA beam intensities and that the available intensities of many radionuclides can be expected to scale with the proton beam current. The design of production targets capable of withstanding proton beam intensities up to 100 pA without compromising the yield of radioactive isotopes will be a future challenge. Several approaches to the dissipation of the power deposited in such targets by the proton beam have been investigated and a realistic solution for the removal of heat from the target container seems at hand. The heat transfer within the target material itself, however, is highly target dependent and it is clear that 100 pA operation will be limited to only a few target materials. Some of the problems may have to be addressed near the 10 pA level but, in general, heat will have to be supplied to the target system to maintain the required temperatures. The development of high intensity targets will be the subject of ongoing research and development at the ISAC facility. Experience at operational on-line separators clearly shows that there is not a universal target/ion-source combination for the production of all required radioactive species. Several types of ion sources will be accommodated within the ISOL design. In addition, new ion-source developments must be anticipated and flexibility must be provided in the system design to allow a reasonable possibility of their successful implementation. Initial ISAC operation anticipates the use of singly-charged ions for which efficient sources have already been proven. The efficient production of multiply-charged ions would significantly enhance the beam capabilities of ISAC; therefore, the development of such sources will be pursued vigorously. Ion beams from most sources contain many species which are not of interest to a particular experiment and in fact provide intolerable backgrounds if not removed from the beam. In many cases, these unwanted species are several orders of magnitude more numerous than those of interest. A mass separator is an essential requirement in the beam transport between the ion source and the experiment. The quality of mass separation required will depend on the particular experiment and the production source. In some cases, high mass resolutions will be required to remove contaminations from the beam, while, in others, high acceptance will be of more importance. A medium resolution 81 separator is proposed here and will satisfy many of the experimental requirements. The mass separator entrance is at the proton beam elevation and separation is done in the horizontal plane. The separated beams are transported by an electrostatic system to either a low-energy experimental area or to the accelerator. The vertical separation previously designed for the 4A option is being modified to accommodate the present target geometry and horizontal orientation. 5.2 Beam Line 2A Proton beams to be delivered to the ISAC production targets will be extracted from the beam-line 2A port of the TRIUMF Cyclotron. This can be done independently of other extracted beams by inserting a stripping foil at the appropriate location in the machine. The extraction of beams from this port was considered many years ago and the properties of 400-500 MeV extracted beams were determined at that time; beam properties for 200-500 MeV extraction have been calculated recently. Variable energy beam extraction requires a combination magnet to be placed at the exit port which will direct the different energy beams to a common transfer line. A combination magnet is available which is capable of handling 400-500 MeV beams which is the energy range of most interest for production at the ISAC target. The section of beam line 2A within the cyclotron vault accepts the beam from the com-bination magnet and produces a doubly achromatic double waist after a 60° bend. At this location, the beam direction is to the North and parallel to the cyclotron vault wall. The double waist is reproduced outside the vault wall by means of two quadrupole doublets which act as a unit section. Beyond this point, the beam is transported in a tunnel by identical unit sections. The beam may be directed to a target station by placing a dipole at a waist location or at the symmetry location between the unit section doublet pairs. In the present layout, the beam is provided to two target stations by a symmetric 30° bend. A quadrupole triplet is placed after · the bend to focus the beam onto the production target. An alternative design has also been considered; it directs the beam to two target stations by placing 20° bends at the waist points of successive unit sections. This option aligns the target stations in the North-South direction. 5.3 The Target Area 5.3.1 General Isotope generating targets proposed for ISAC must withstand bombardment by 10 pA of 500 MeV protons. This proton bombardment will generate very high operating and residual radiation fields. A viable approach must be provided to handle and service the highly radioactive components near the production targets. Also, the facility must incorporate 82 radiation shields to reduce operating fields outside the containment building to biologically acceptable levels for the maximum expected beam currents. The ISAC target-handling concept and ISAC target facility is based on fifteen years of experience a! the three operating meson factories -- PSI, LAMPF, and TRIUMF. The meson production target and beam stop areas of these facilities have power dissipation and radiation levels similar to, or greater than, those expected at ISAC. Meson factory experience shows that the correct approach to handling components in high-current, thick-target areas is to place them in tightly-shielded canyons within a large target shield. Both PSI and TRIUMF access the components vertically and do most repairs in dedicated hot cells. The target facility design [3, 4] must address three important complications that are not encountered in the meson factory target areas. These are: the containment of large amounts of mobile radioactivity; the high voltage required for beam extraction; and quick routine replacement of short-lived target systems. In the present design these issues are resolved by placing the target in a sealed self-contained module which can be transferred directly to a hot cell facility for maintenance operations. 5.3.2 Target facility A plan view of the proposed target facility is included in Fig. 5.1. The target stations are located in a sealed building serviced by an overhead crane. The target maintenance centre contains a hot cell, warm cell, decontamination facilities and a radioactive storage area needed to keep the target area components operational. It also houses the high voltage terminal, vacuum, water and other service systems. It is separated from the target area to allow personnel access during all beam operations a.nd has a controlled and monitored access. Beam-line elements near the target are installed inside a large T-shaped vacuum chamber surrounded by close-packed iron shielding. This general design eliminates the air activa-tion problem associated with high current target areas by removing all the air from the surrounding area. The front-end design breaks naturally into four modules: an entrance module containing beam diagnostics, an entrance collimator and pump port; a beam dump module containing beam diagnostics, pre-dump collimator, and a beam dump; a target modul.: containing the target, ion source, and extraction electrode; and an ion-beam-line, front-end module. The vacuum design seeks to eliminate the need for radiation-hard vacuum connections at beam height by using a single vessel approach. The front-end components, with their integral shields, are inserted vertically into a single large vacuum vessel. Most vacuum connections are situated where elastomer seals may be used and only two beam-height connections exist - the proton beam entrance and the ion beam exit. Figure 5.2 is a section through the target facility, showing that the target modules are housed within a heavily-shielded containment building. Steel shielding is placed close to the target and this is surrounded by concrete shielding. When target station modules are removed, the shielding forms a canyon. The target is placed deep in the canyon to lower both the operating and residual radiation fields at the canyon top. The operating levels 83 are reduced, allowing long term use of nonradiation-hard materials. The residual fields are reduced, allowing personnel access for maintenance. All components installed in the canyon are mounted on the bottom of shielding plugs that extend to the canyon top. Services, such as power and water, connect to the components at the canyon top. Below this level , joints and connections are minimized and radiation hard. High residual radiation fields will make manual in situ maintenance of component parts below the canyon top impossible. Such operations are carried out in a hot cell. SECT10N A·A Figure 5.2: A cross section of the target hall through the target. A large void is left in the shielding immediately above the target station. The target hall crane operates in this space, which connects to the maintenance facilities. Sufficient height is provided to allow front-end modules to be lifted free of the shield canyon and transported to the hot cell. One meter of boron loaded concrete, mounted on rollers, can be positioned above the target station. This easily removed shield reduces the thickness and the cost of the target hall's roof and reduces operational fields in the void area. This building design places the target station in a heavily shielded bunker that is directly connected to the hot cell facilities and has many advantages when considering speed of servicing or access and contamination control. 84 5.3.3 Remote handling The ISAC targets are expected to be less robust than TRIUMF's high-current targets due to the extreme operating conditions required. Also, different target/ion-source combinations will be used to produce the required radioactive beams. Therefore, an effective remote handling and servicing system will be required to bring about quick and frequent target charges. All modules in the target area will have high levels of residual activity and will be potentially contaminated with mobile activity. Both aspects are considered in the handling design. Target component maintenance involves disconnecting services and craning the module to the hot cell. Rolling aside a concrete shield gives the overhead crane access to the modules. Flaskless transport is permitted by the thick building walls provided personnel are excluded from the area. Target module transfers to the hot cell must therefore be done completely remotely. The connection and disconnection of the target module services can be done manually or remotely. Manual option is proposed for at least the initial years of ISAC operation as this will be less costly than a completely remote system. However, lengthy cool-down times may be required after extended operation at 100 pA. Residual fields and the requirement to turn off the proton beam will make speed and simplicity of this operation mandatory. Connection designs will allow conversion to a completely automated system with only a minimum of redesign and refitting. The mobile contamination produced in the target area is normally contained within the target module. Contamination of the interior of the target building is considered possi-ble since the modules will be transported without a containing flask. This building must therefore be considered as an extension of the hot cell complex and all entrances must be controlled and provided with appropriate contamination control. The air within the building must be controlled and HEPA-filtered. The interior surfaces will be designed to be easily decontaminated. All fluid drains will go to sump tanks for monitoring before disposal. Two cranes are provided for handling operations. The target hall crane covers the target station, the module storage area, and the hot and warm cells. The maintenance centre crane covers the loading bay, the warm cell and most of the personnel-accessible areas. A module storage area is located between the hot cell and the target station. An array of pumped and shielded storage silos is provided to store modules under vacuum. One silo is provided with the necessary services for the testing and preconditioning of target modules before installation. This area is accessible during beam operation; servicing and testing of modules will therefore be possible during beam production. The hot cell provides facilities to remotely maintain, replace, decontaminate, or inspect the highly-radioactive components removed from the target area. It is a conventional design with concrete shielding walls, lead glass viewing windows, and sealable roof ports to allow crane access to the cell. Personnel access to the top of the cell is possible, if required. The hot cell bay is provided with direct actuated master slave manipulators and a transfer port to the connecting low-level warm cell. The mechanical bay includes remote viewing, service equipment, and an elevating turntable to support and position the component being serviced. 85 A floor pit allows the removal of containment vessels from the target modules. The hot cell is kept under negative pressure by its own HEPA-filtered air handling system. A multi-purpose warm cell allows handling of components with low radiation levels and limited removable contamination. A pair of heavy duty master-slave manipulators operate over the shielding wall. Several large radiation shielding windows provide viewing into the warm cell. Personnel access to the warm cell is through a locked-gate shielding maze where contamination monitoring is provided. Much of the work in the warm cell will be hands-on. Typical jobs include sealing shipping casks and temporary storage pigs, transferring new components into the hot cell, and the manual repair of equipment having low radiation levels. The warm cell also serves as the hot cell maintenance access bay. A support annex houses the remote handling control room, offices, personnel change rooms, radiation safety monitoring equipment, and target hall entry air-locks. The equip-ment needed to control the remotely-operated cranes, viewing systems and other devices is in the control room. Cameras are mounted in strategic locations throughout the building and on the cranes. An air-lock is provided for truck transport of equipment into the target hall. 5.3.4 The target/ion-source module In the present target area design, the target, ion source and extraction system are all con-tained within a single remotely handleable module. The design of this module will be critical to the successful operation of the ISAC facility as the difficult problems of target/ion-source servicing, ion-beam extraction and high-voltage isolation must be resolved for a highly-radioactive environment. This module will contain the majority of the radioactivity pro-duced by the proton beam and containment of this activity will have to be a major design feature. The targets and ion sources are expected to be changed frequently during ISAC operations. Therefore, handling of the module will have to be a relatively simple and quick procedure. A section through the target/ion-source region as presently conceived is shown in Fig. 5.3. The target module is completely contained in a surrounding vacuum enclosure with a single exit hole to allow the extraction of the ion beam. It is inserted into an external vacuum enclosure. Windows will be provided for the entry and exit of the proton beam and a special radiation-hard valve will seal the ion beam exit hole when required. Service connections to the module, vacuum pumps and the high-voltage insulator are all shielded from the production target by 2 m of steel. Conventional nonradiation-hard components can be used at this location. The target, ion source and extraction system are mounted on a central shield plug which carries the required high voltage services. This plug is isolated from the grounded vacuum enclosure by a 5 em gap. This gap provides both high voltage isolation and the required pumping channel to the ion source region. Vacuum pumps will be integral to the module to avoid the use of large vacuum valves and to maximize pumping speed. The target, ion source and extraction system are envisaged to be modular components which can be connected to, or disconnected from, the shield plug by a manipulator in the 86 hot cell. All such connections will have to be radiation hard. The alignment of the compo-nent modules will be critical to the source operation and will be performed in the hot cell. Sufficient space has been provided at the base of the shield plug to accommodate the largest envisioned ion source and extraction system. Beam from the ion source will be injected into a four-quadrupole matching section. • • • • 4• • • • 4 • • • • 4 • • 4 • • • Figure 5.3: A section through the target/ion-source region. 5.4 Production Targets 5.4.1 General target considerations Of obvious importance is the ability of the target to tolerate the harsh thermal and radiation environment associated with the proton production beam. Targets for the proton production of a wide range of radioactive species have been developed principally at facilities using "thick" targets, notably ISOLDE [5] and TISOL [6]. Targets capable of handling more intense proton beams will be designed using the same basic concepts. 87 To achieve optimum production of short-lived species, the length 9f the target (and thus the production rate) must be balanced with the requirement for fast release of the product activities. In practice, targets meeting these dual requirements have thicknesses of approxi-mately one interaction length . For :>11 ch targets, even modr>st production cross sections give high yields, essentially taking advantage of the long range uniquely associated with energetic protons. Other light-ion projectiles can be used, but intense production beams of energetic protons (up to 100 J.tA) are more readily available. Target temperature is important. Even relatively volatile activities are released more readily at elevated target operating temperatures that sometimes approach target material stability limits. Other target material characteristics that are important for good release are: diffusion rates (always higher for elevated temperatures); desorption rates (element specific); and physical limits, such as vapor pressure limitations, sintering limits (for powder targets), or chemical stability. When the production beam intensity is increased significantly above the presently avail-able experience level, as proposed for the ISAC facility, a new set of target-specific issues must be considered. At high production beam intensities, heating of the target material through beam energy losses may require the use of forced target cooling, even for targets that normally operate at high temperatures. Thus, transport of heat within the target material and subsequent heat extraction is of concern to the design of any thick target. 5.4.2 Existing target systems In general, solid targets consist of refractory metallic foils operated at high temperature (around 2000°C) and are preferred for systems where the release of desired activities demands high temperature operation or is governed by surface desorption. Liquid targets are suitable for release of activities more volatile than the target material, and are usually limited to operating temperatures not much higher than the target melting point. Both metal (solid) and liquid targets are relatively dense, and beam heating from intense production beams can pose serious heat removal problems, while at the same time maintaining the required operating temperatures. Powder targets are mainly of refractory chemical compounds. These targets present the widest range of operating temperatures, depending on the chemical or physical stability of the material employed. They are also usually of low density; therefore, heating from intense production beams is reduced compared with dense targets and external heating is required. However, the transport of induced heat from the interior of the target is inefficient due to inherently-low thermal conductivity, which can result in large thermal gradients existing within the target material itself. Indeed, this aspect of powder targets may impose constraints on the tolerable maximum production beam current. 5.4.3 Target analysis tools Two types of analysis are needed for evaluation of candidate target systems: analysis of beam-induced heating within the target; and analysis of the thermal characteristics of the 88 target, including the efficacy of cooling and external heating schemes. For the beam heating analysis, codes based on Monte Carlo techniques have been devel-oped which can provide beam energy loss profiles within the target system. The two codes extensively used in the literature are LAHET [7) and FLUKA [8) (and updated versions). LAHET can be linked to the MCNP code [9) (and updated versions) to provide tracking of interactions of primary and secondary particles down to thermal energies, and includes energetic fission interactions in high-Z target materials. The FLUKA code, used at CERN, is comparable in application. Both codes can quickly predict beam energy deposition profiles to statistical levels of a few per cent. Thermal analysis is provided by the finite-element approach of the ANSYS code [10), which can be used both to evaluate the mechanical strains of a system, and to determine the thermal profiles generated from the beam energy losses by means of heat transport networks. 5.4.4 High-current target systems A target system proposed for a 100 J.tA facility, and based on a credible extension of previously-reported studies [11, 12), is described in detail in a TRIUMF Design Note [13). This target concept has the following attractive features: it is mechanically simple; it pro-vides for selection of operating temperature range; it can be designed to provide uniform axial temperatures; it uses simple water-cooling techniques; and it is expected to be relatively inexpensive to construct. For ISAC, it is anticipated that present production target designs will be adequate for initial operation, with minor modifications to accommodate internal beam heating. However, high current designs such as that described above, or as being developed at the Rutherford Appleton Laboratory [14), may prove very beneficial and will continue to be persued in the development of the ISAC facility. 5.5 Ion Sources 5.5.1 General considerations The target and ion source of an ISOL determine the type and quality of the radioactive beam. Certain characteristics are common to all ion sources operating on-line. The initial amount of the ionized species is determined not by a flow rate from a supply bottle, but by the production cross section for that species in the ISOL target. Maximizing the yield of a desired product is achieved by minimizing the losses incurred from its initial production to its extraction as a radioactive beam. Clearly, an ISOL ion source m:ust be closely coupled to the production target to minimize losses from decay of short-lived products during their diffusion from target to source. Such close proximity further requires the ion source be thermally rugged and radiation hard. Sources must be constructed of materials with high melting points and low vapour pressures, limiting the choice to the refractory metals and ceramics. The hostile environment also dictates that the ion source be both simple and 89 small. Volumes and surfaces must be kept to a minimum both for the sake of economy and to minimize delaying surface interactions of short-lived products. Simplicity of construction ensures that a radioactive source can either be easily repaired by remote manipulation or economically thrown away after its inevitable breakdown. Fortunately, past experience at ISOL facilities has produced several types of sources that meet these criteria. 5.5.2 Surface ionization and thermal cavity ion sources The simplest and most efficient of the ISOL ion sources is the surface ionization source. In surface ionization, an electron is transferred when an atom (or molecule) contacts a hot surface. The atom subsequently leaves the surface as a positive or negative ion. For positive ionization, the surface must have a work function greater than the atom's ionization potential; for negative ionization, a low surface work function and a high atomic (molecular) electron affinity are required. The degree of ionization is dependent on the work function and ionization potential (electron affinity) difference and also scales with surface temperature. For positive ionization, refractory metal surfaces such as Ta, Re, W or Ir, at temperatures >2000°C provide ionization efficiencies of up to 95% for the alkali elements. Elements with ionization potentials <7 eV, such as the alkali earths, rare earths, Al, Ga, In and Ra are ionized to varying degrees depending on the surface and temperature employed. For negative ionization, a LaB6 surface (work function = 2. 7 e V) achieves ionization efficiencies of about 40% for Cl, Br, I and At. In theory, it may also be possible to ionize molecules with high electron affinity. Somewhat higher positive ionization efficiencies can be obtained with high temperature cavity ion sources. In these sources, ionization is enhanced beyond the predictions of simple surface ionization due to the formation of an electron plasma near the walls of a hot W cavity. This can allow ionization of elements with work functions of up to 8 eV in W (Re) cavities heated to 2700 K (2400°C). The theory of thermionic sources has been extensively surveyed [15] . The chief advantages of surface sources are their simplicity and high efficiency. In practice, a source consists simply of a refractory metal tube connected to the production target and resistance heated to the desired temperature. Some chemical selectivity is possible by the appropriate choice of ionizing surface and temperature. Surface sources produce the best quality beams for purposes of mass resolution; since ions leave the surface with thermal energies, the beam emittance and energy spread of the resulting beam are the lowest of any ISOL ion source. The disadvantage of surface sources is their limitation to elements with ionization potentials <7 eV. 5.5.3 Plasma ion sources To ionize elements with ionization potentials > 7 eV most ISOL facilities use sources employ-ing a plasma generated by an arc discharge in a support gas. Though many different designs of plasma source have been used [16, 17], the forced-electron-beam, induced-arc-discharge (FEBIAD) source such as employed at GSI [18] and ISOLDE [19] is currently the most popu-lar. In general, electrons from a heated cathode are accelerated in a gas-filled chamber where 90 a magnetic field guides them to an anode grid, thus generating a plasma. Products from the target are ionized by electron impact as they diffuse into the plasma chamber. Unlike earlier sources, FEBIAD sources have the advantage of operating with support gas pressures on the order of 10-5 torr. This reouced operating pressure allows the use of calibrateo test leaks as a source of support gas and also reduces the current density (and emittance) of the extracted beam. Unlike surface sources, plasma sources are not chemically selective; any element diffusing into the plasma chamber will be ionized. Chemical selectivity is achieved by the proper choice of target material or by controlling the temperature of the transfer line from the target to select elements of a particular volatility. Plasma sources are more com-plex than surface sources: a support gas, a magnetic field, cathode heating and an electron accelerating potential are required. Ionization efficiency is generally determined from the ionization of the support gas. Efficiencies of 23% (Kr) and 48% (Xe) have been reported for the GSI FEBIAD source as well as a 76% efficiency for Hg ionization [18] . Efficiencies are expected to show a mass dependence and scale as UionAt, where Uion is the cross section for electron impact ionization. However, yields of light elements such as C, N and 0 have been less than expected. Arc discharge sources operate at high temperatures and are constructed of refractory materials. Since the carbides, nitrides and oxides of the refractory elements are usually themselves refractory, direct chemical reactions with the heated cathode or plasma chamber walls significantly reduce the light-element yields. 5.5.4 Electron cyclotron resonance (ECR) ion sources In order to efficiently ionize the light elements, some ISOL facilities have successfully used a different type of plasma ion source. Recently, ECR ion sources have started to be used on-line, first at the TISOL facility at TRIUMF [20] and subsequently at the Belgian Radioactive Ion Beam Facility [21]. ECR sources show good efficiencies for producing C, N and 0 ions as well as molecular species such as Nt, No+, co+ and COt. ECR sources consist of a plasma chamber coupled to the target. However, the plasma is generated by microwaves (in the GHz range) introduced through a wave guide. A magnetic field keeps the electrons in cyclotron resonance while a second solenoidal field can be used to confine the plasma. Due to the plasma confinement and a lower operating temperature, wall interactions are minimized. Ionization efficiencies of 10% (C), 27% (N), 55% (0), 31% (Ne) and 65% (Xe) have been measured [22]. ECR sources are not chemically selective. In theory, any element introduced into the ECR plasma chamber would be ionized. However, in practice, only the volatile gaseous elements have (as yet) been produced on-line. An advantage of ECR sources is their ability to produce multiply-charged ions. Due to higher electron density, low operating pressures and magnetic plasma confinement, ECR sources can efficiently produce ions with charge states up to 8+. This feature can enhance the options available for acceleration of ion beams but can also be a disadvantage if low yield products are distributed over many charge states. The chief disadvantage of ECR sources is their relatively-high beam emittance and energy spread. While highly efficient, ECR sources may not be the first choice for high-resolution mass separation. 91 5.5.5 Laser ion sources Very recently, the ISOLDE group has tested an ion source that promises both good ionization efficiency and the highest degree of chemical selectivity yet observed [23]. The ISOLDE laser ion source is essentially a surface ionization source (consisting of a 1 mm dian wfer hr·;1tcd cavity coupled to the production target) into which three tunable dye laser beams are directed simultaneously. The laser frequencies are chosen such that two lasers first excite the desired product element via atomic states allowing the third laser to ionize it. Since each element has its own distinctive energy level structure, the tuned laser source offers an exceptionally high degree of chemical selectivity. In the study of Sn, Tm, Yb and Li laser ionization, a thousand-fold selectivity increase in Tm ionization and a 30% enhancement of Sn ionization efficiencies was observed. Laser sources are still in the development stage and their chief disadvantage is the low laser duty cycle. 5.5.6 Multi-cusp ion source It has been suggested [24] that a plasma source using a multi-cusp magnetic field could efficiently produce multiply-charged ions. Currently, a TRIUMF-LBL collaboration is de-signing such a source for on-line testing. The source will employ permanent magnets for the multi-cusp field and a heated cathode for plasma generation. It is expected that good beam quality and selective multiply-charged ion extraction can be achieved with good beam quality. If successful, this source would offer a practical alternative to the ECR sources. 5.6 Ion Beams The design of the radioactive beam system requires that a number of fundamental issues be addressed and resolved. These issues include the following: extraction from the ion source; matching of the source to the mass separator; beam emittances and currents to be transported; multiple beam requirements from the mass separator; required mass resolution; and transport from the separator to the user. There is no solution which simultaneously satisfies all possible requirements so that critical choices have to be made. Extraction of the beam from the ion source will be dependent on the source used, the conditions expected in the source during operation and the philosophy taken by the designer. The extraction system used at ISOLDE is a single independently-mounted electrode. The alignment of this electrode to the source is critical to the beam extraction and five inde-pendent degrees of motion are required to achieve this. The design of a reliable mechanism for providing such alignments in the environment of the proposed target system will be a difficult task. Other options, with less elaborate mechanical requirements, therefore have to be investigated. Beam emittances and currents are dependent on the particular source and target system being used. Large emittance beams require a high-acceptance beam transport system and are not amenable to high mass resolution. High current beams may be subject to space charge effects which may introduce aberrations and cause beam losses in the transport system. 92 Typical ion sources such as surface or plasma sources give beam emittances of 2-207r mm·mr at 60 kV. ECR sources have much larger emittances (801r mm·mr) due to the magnetic field in the extraction region. JJ) Figure 5.4: A moderate-resolution mass separator and beam transport line from a surface ionization source. Transport and matching optics must be provided between the ion source and the mass analyzer to tailor the beam for the separator magnets and to compensate for changes in the beam extraction. The requirement of multiple beams from a single mass separator brings with it the implication that this transport system be electrostatic, as magnetic systems transport only a limited mass range. This in turn implies that the ion beams will not be space-charge compensated due to the removal of space-charge compensating electrons by the focusing elements; space-charge effects at the entrance of the mass separator where the beam must be small will be maximized. The transport of high current beams from the source to the separator may therefore be compromised. The angular divergence of the beam to be transported is limited by the position and size 93 Table 5.1: Specifications for the mass separator Bend radius Total bend angle Magnet gap Entrance angle Exit angle Maximum field strength Focal plane dispersion Mass resolving power Correction elements 62 em 2 X 112 = 224° 7.6 em full gap 28° 28° 10 kG 6cmj% 10,000 (for 2.571' mm·mr) 4 sextupoles + 4 octupoles of the first lens. Therefore, to increase the acceptance of the transport system requires a special optical design with large elements close to the ion source. If large phase space beams are to be transported, care must be taken to avoid or compensate aberrations which lead to beam loss at the entrance to the separator. High mass resolution obviously has implications on the separator design but also requires high precision and stability in the transport line from the ion source. This must be taken into account when designing this system. For example, focusing fluctuations in an electrostatic system due to variations in the ion beam current have an effect on the separator performance. To avoid such complications and the expense of high quality optical elements, the injection line to the mass separator should be kept as short as possible. The present design of one source and beam layout is shown in Fig. 5.4 for a surface ionization source. A multi-electrode fixed extraction system is attached directly to the source. The extraction system will be optimized for each individual source type to give good source brightness and small angular divergence. Space has been allowed in the target module design to incorporate additional optical elements such as steering devices or other lens systems. Four electrostatic quadrupoles are used to match the beam properties between the extraction system and the mass separator. The mass separator in Fig. 5.4 consists of two antisymmetric QQQDDQQQ systems each with a total bend angle of 112°. It will have a source-defining entrance aperture. Aberrations arc corrected by four sextupole and four octupole magnets placed near each dipole. This separator will have a dispersion at the focal plane of 6 emf% in !:l.M / M and a mass resolving power of 10,000. A movable slit system will be placed on the focal plane to select the mass to be transmitted. The second bend section could be maintained at an elevated potential to provide additional beam purification [25] if this is required; however, this possibility has not been incorporated in the present design. The specifications of the separator are given in Table 5.1. This presents the expected optical performance with low-emittance beams. Actual performance will be dependent on precision and stability of the elements and the ion beam emittance. 94 Transport of the separated beams will be done entirely with electrostatic elements as the beam currents will be low and tuning will be dependent only on the beam acceleration voltage. A goo achromatic vertical bend will allow transport of the beam to grade level where a similar 90° bend will dirert it into a switchyard arrangement which will deliver the beam either to a low-energy experimental area or to a matching section preceeding the RFQ. 5.6.1 ISOL beam diagnostics The purpose of the ISOL beam diagnostics is to give the necessary information for adjusting the beam lines from the ion source to the separator, selecting the appropriate ion species in the two stages of the separators, adjusting the beam line from the separator to the RFQ, and adjusting the beam lines to the low-energy experiments. Preliminary considerations for the beam diagnostics can be found in Ref. [26). The ISAC facility will operate with beams of 102 to 1010 ions/s. It is impossible to cover this intensity range with one beam diagnostic device; therefore, either multiple diagnostic elements must be used, or a more indirect approach must be taken for low-intensity beams. The sensitivity of the beam diagnostic elements used at ISOLDE covers the range from 106 to 1011 ions/s (0.2 p.\ -20 nA) [27). When less intense beams are used the system is set up with a more intense pilot beam of the same or near mass. The same procedures will be employed at the ISOL facility and the sensitivity of the beam diagnostic elements used at ISOLDE will be sufficient here. The basic beam observation unit for the ISOL system will be a standardized unit similar to those at ISOLDE [28) which combine a wire scanner and a Faraday cup. The wire scanners are mounted at 45° and use V-shaped wires with a goo angle. Thus, the horizontal and vertical beam profiles are measured with the same mechanism. The spatial resolution is determined to 0.5 mm, mainly by the diameter of the wire. The standard wire scanners cannot be made radiation-hard, because of a pre-amplifier, which must be placed directly on the moving wire. Therefore, in high-radiation areas before the separator, the beam profile will be measured by moveable slits, placed in front of a special, radiation-hard Faraday cup. The same slits will also be used to trim the beam emittance from the ion source before entry to the separators. Other radiation-hard Faraday cups will be placed immediately after the ion source, to make it possible to verify its operation at any time. The diagnostics at the separator focal planes will become radioactive and must be designed and built for very high reliability. Horizontal slits will be used to select the appropriate ions. Faraday cups will be placed behind the slits to analyze for different ion species in the heam. Horizontal wire scanners, with enough stroke to cover a mass range of ±6%, will be placed just before the slits and will be used for scanning the mass range of the desired ion beam, and to monitor a reference beam while running low intensity beams. A removable vertical wire grid will be used to measure the vertical beam profile. An emittance-measuring apparatus will be placed before the separator, after the emittance-defining slits. Another will be placed just before the RFQ. Each will consist of two moveable slit plates and a Faraday cup. The slit plates will contain horizontal and vertical slits, and g5 will be moved at 45° with respect to those planes. The slit width and the distance between the slit plates will be chosen to achieve a resolution in the emittance measurement of about 0.27r mm·mrad at 60 kV. Several of the ion sources will be equipped with a very-small, controlled leak, through which a known gas, such as xenon, can be introduced to produce a reference beam, for testing and calibrating the separator and beam transport system. 5. 7 High Voltage System The high voltage systems which service the target stations will be housed in a Faraday cage located in the target maintenance center where radiation levels will be sufficiently low to allow access during beam operation. High voltage services from the Faraday cage will be routed to the target area through an electrically-shielded service maze. Devices required for the operation of the target and ion source will be located within the cage and powered through an isolation transformer. Control of these devices will be via computer optical links. Closed-circuit water systems and any required pumping systems will be located at ground potential and fed though insulating sections to the Faraday cage. The high voltage isolation will be designed for 100 kV operation although 60 kV is the present operational voltage specified for ISAC. This will give some flexibility in responding to future needs and provide a design margin for the entire high voltage system. The extraction voltage must be very stable for operation with the high-resolution mass separator and the high voltage power supply will have to maintain ±1 V in 60,000 V. This stability is well within the limits of present technology. 5.8 Vacuum System High pumping speeds are required in the target/ion-source region to remove background gases. This is difficult to achieve if pumps must be placed in a non-radioactive environment and separated from the target by close-packed shielding. In addition, many of the gases to be pumped will be radioactive and precautions must be taken to contain this radioactivity. The transport lines for the mass separated beams are long and a good vacuum must be maintained over the entire length to avoid beam losses due to interactions with the residual gas. Turbomolecular pumps are presently proposed to provide the requirement of high pump-ing speeds for all gases. In the present design of the target module, these pumps are located at the top of the module and pump the target/ion-source region through the 5-cm-high voltage gap which has a conductance of about 1000 f-/sec. The turbo backing pumps are sealed pumps which are exhausted via a compressor to an evacuated waste volume. The gases collected in this waste volume may be stored for a period of time to allow decay of the short-lived species before controlled release to the atmosphere. The pipe length between the turbo pump and the backing pump is made especially long to provide a delay time in which very-short-lived species may decay, thereby reducing the radiation dose to the backing 96 pump. The backing pumps and the waste volumes will be located in shielded regions of the service building and will be monitored with radiation detectors. Separate pumps will be used to evacuate the system to a rough vacuum. These pumps will be vented through the target. area HEPA filter system since only low levels of radioactivity are expected during this operation. The amount of radioactive gas pumped in the mass separator region is expected to be much lower than in the target region but still significant. The vacuum system for this region will be similar to the target area systems; however, all pumps and waste volumes will be located close to the separator. Beyond the mass separator the amount of radioactive gas to be pumped will be small and pumps will be exhausted directly to the atmosphere through one of the HEPA-filtered air handling systems. To minimize the loss of beam particles due to scattering in the residual gas, a 10-7 torr vacuum will have to be maintained in the beam transport between the production target and the accelerator or the low-energy experimental area. This implies the use of high-vacuum components and procedures and the ability to mildly bake (100°C) the system after assembly. For practical purposes, the system will be divided into a series of individual vacuum sections separated by inline gate valves. Each section will have a dedicated high-vacuum pumping system and gauges. Hot cathode gauges will be used for high-vacuum measurements and convectron gauges for rough vacuum. Roughing will be provided by a centralized roughing system except for the production and experimental target areas. A central dry nitrogen venting system will be provided and integrated with the vacuum controls. 5.9 Off-Line Source An off-line ion source will be required to provide stable beams to the ISAC accelerator and low-energy areas. This source will be situated as shown in Fig. 4.2 (see Section 4.4) so that it can inject beams into the ion-beam switchyard and will be used both for tuning the accelerator and for providing test and set-up beams to the experimental areas. A number of different sources will have to be accommodated on the high voltage platform to provide a broad range of possible stable beams. Only moderate mass separation will be required and a replica of the first section of the on-line separator will be used for this purpose. Beam diagnostics and emittance measurements will be provided in the ion transport line before injection into the beam switchyard. References [1] H.L. Ravn and B. Allardyce, in Treatise on Heavy Ion Science, Vol. 8, ed. D.A. Bromley, Plenum Press, New York, 1989, p. 363; H.L. Ravn, Phys. Reports 54 (1979) 201. [2] E. Kugler et al., Nucl. lnstr. Meth. B70 (1992) 41. 97 [3] J .L. Beveridge et al., in Proceedings of the Third International Conference on Radioactive Nuclear Beams, ed. D.J. Morrissey, Editions FronW~res, Gif-sur- Yvette, France, 1993, p. 55; J.L. Beveridge and G.S. Clark, TRIUMF Design Note TRI-DN-93-15, May, 1993. [4:j J.L. Beveridge, G.S. Clark and C. Mark , TRIUMF Design Note TRI-DN-94-2, 1994. [5] ISOLDE User's Guide, ed. H.-J. Kluge, CERN 86-05 (1986), and updates. [6] K. Oxorn et al., Nucl. Instr. Meth. B26 (1987) 143. [7] R.E. Prael and H. Lichtenstein, User Guide to LCS: The LAHET Code System, Los Alamos National Laboratory report LA- UR-89-3014, 1989. [8] A.A. Arnio et al., CERN report TIS-RP7168, 1986. [9] Group X-6, Los Alamos National Laboratory report LA-7396-M Revised, April, 1981. [10] ANSYS User's Guide for Revision 5.0, Swanson Analysis Systems, Inc., Houston, PA, 1993. [11] T.W. Eaton et al., Nucl. lnstr. Meth. B26 (1987) 190. [12] W.L. Talbert, H.-H. Hsu and F.C. Prenger, Nucl. lnstr. Meth. B70 (1992) 175. [13] W.L. Talbert and T.A. Hodges, TRIUMF Design Note TRI-DN-94-1, 1994. [14] J.R.J. Bennett, private communication, 1993. [15] R. Kirchner, Nucl. Instr. Meth. A292 (1990) 203. (16] K.O. Nielsen, Nucl. Instr. Meth. 1 (1957) 289. [17] I. Chavet and R. Bernas, Nucl. lnstr. Meth. 51 (1967) 77. [18) R. Kirchner, Nucl. lnstr. Meth. 186 (1981) 275. [19] S. Sundell et al., Nucl. Instr. Meth. B70 (1992) 160. (20] M. Dombsky et al., Nucl. Instr. Meth. A295 (1990) 291. [21] P. Decrock et al., Phys. Rev. Lett. 67 (1991) 808. (22] V. Bechtold, H. Dohrmann and S.A. Sheikh, in Proceedings of the 7th Workshop on ECR Ion Sources, Julich, Germany, 1986, p. 248. (23] V.I. Mishin et al., Nucl. Instr. Meth. B73 (1993) 550. (24] M.R. Schubaly, Nucl. Instr. Meth. B26 (1987) 195. 98 [25] H. Wollnik, in Proceedings of the Workshop on the Production and Use of Intense Ra-dioactive Beams at the ISL, Oak Ridge, Oct . 7-10, 1992, ed. J.G. Garrett , Report Conf-92/0121, p. 213. [26] D. Reistad and Y. Yin, TRIUMF Design Note TRI-DN-93-29, November, 1993. [27] H. Ravn, private communication, August, 1993. [28] E. Kugler and G.-J. Focker, private communication, November, 1993. [29] "TRIUMF 5-Year Expansion Plan", March 10, 1995. 99 100 Section 6 ACCELERATOR 6.1 Specifications The specifications for an accelerator to provide beams primarily for a research program in nuclear astrophysics are given in Table 6.1. The assumptions are that singly-charged ions with masses up to 30 u and a fixed energy of 60 keY, irrespective of iou mass, will be delivered at the output of the on-line separator. For most of the experiments described in Section 2, some further acceleration must be provided. Thus, for example, the nuclear astrophysics studies require ion energies of up to 1.5 MeV ju, while somewhat higher energies would be desirable for nuclear structure studies. Because the radioactive ions are produced continuously in the target/ion source, it is desirable, in view of the relatively-low intensities of some of the separated ion species, to require continuous ( CW) rather than pulsed operation of the accelerator. 6.2 The Need for Stripping Because the ions extracted from the source are most abundant in the singly-charged state, it is for these relatively-low charge-to-mass-ratio particles that the accelerator must be de-signed, at least in the initial stage. If the entire acceleration were carried out on ions in this charge state, a total effective acceleration voltage of at least 45 MV would be required to meet the specifications listed in Table 6.1. To limit the accelerator size here, as at other heavy-ion accelerators, the beam will be passed through a stripper to increase the ion charge-to-mass ratio at an intermediate point in the acceleration process. The mean charge state q8 of the equilibrium charge-state distribution of an ion beam with atomic number Z and energy W, after passage through a foil, is given by the semi-empirical relation [1) q, = Z [ 1- (o.9 + 0.0769 ~) exp ( -83.275:,)] , (6.1) where /3 is the usual relativistic form of the ion velocity, A is the ion mass number, 1 = 0.4 77, 101 Table 6.1: Basic specifications for the ISAC accelerator. INPUT BEAM Energy 60 keV Ion mass A:::; 30 Ion charge ±1 Beam current < 1JlA DC Beam emittance :::; 0.257r mm·mrad (normalized) (within 100% contour) ACCELERATED BEAM Output energy range 0.15-1.5 MeV ju Resolution .6.E / E < 0.1% Duty factor 100% and W is in MeV. If the stripper is located at a position corresponding to an ion energy W.,, and the input and output energies of the accelerator are Wi and W0 , respectively, then the total effective acceleration voltage Vr for that component of the accelerated beam with charge state q3 is given by (6.2) where qi is the input ion charge state (usually 1 in this application). In Fig. 6.1 Vr is plotted as a function of the beam energy at the stripper for the case of a singly-charged 30Na input beam with Wo and Wi as given in Table 6.1. This neutron-rich ion beam has, according to Eq. 6.1, a mean charge after stripping of +5, and represents about the lowest particle charge-to-mass ratio for which the post-stripper accelerator must be designed. From Fig. 6.1 we see that the effective acceleration voltage has a minimum at 13 MV when the stripper foil is located at the point corresponding to an ion beam energy of 130 keV ju. The penalty paid for reducing the accelerator size by intermediate stripping is of course a reduction in beam intensity. Fig. 6.2 shows a plot of the expected intensity ratio, after stripping, as a function of atomic number for stable as well as neutron-rich and proton-rich ions with half-ives greater than 20 ms. 6.3 Conceptual Design 6.3.1 General description The block diagram in Fig. 6.3 illustrates the two-stage linear accelerator (linac) that would satisfy the ISAC specifications in Table 6.1. A radiofrequency quadrupole (RFQ) operating 102 20 19 18 > ~ u ... 17 " ~ 1 V Tfoit{E s) 16 .... .. -~ IS ~ c (,1,; D • a a a 14 13 a' 12 0 0.05 0.1 I z =111 A=30 ...d / ~~ ...rll v ~ ~ 0.15 0.2 Es A 0.25 lao CDC1JY atllripper Me V/u 0.3 0.35 0.4 Figure 6.1: Total voltage required to accelerate singly-charged 30Na ions to 1.5 MeV /u, as a function of the beam energy at a carbon foil stripper. at 35 MHz provides the initial acceleration of the singly-charged ion beam delivered from the mass separator. As is the case for all fixed-frequency linacs, the RFQ in this case is designed for a specific input ion velocity. However, as a consequence of the fixed extraction voltage at the ion source, the ions are delivered from the mass separator with velocities that are mass dependent. To accommodate the RFQ input requirements, it is necessary to operate it with a DC bias, adjustable within approximately ±60 kV, so that the ion input energy can in all cases be 2 keY fu (for A:::; 30). After acceleration to 150 keY /u in the RFQ, the beam passes through a matching and stripper section, where its charge-to-mass ratio is increased to ?:..l/6 by passage through a thin carbon foil, before being injected into the 70 MHz drift-tube linac stage and accelerated to energies up to 1.5 MeV /u. To permit variation of the output beam energy, the 48 accelerating gaps of the drift-tube stage are divided into eight independently-driven sections of six gaps each. Quadrupole doublets between sections provide transverse focusing of the beam. 6.3.2 RFQ RF electric fields both focus and accelerate the ion beam in a RFQ. Because these fields are distributed continuously along the length of the RFQ electrodes, a strong focusing channel can be realized, even for quite low velocity ions with small charge-to-mass ratios. At the same 103 I . 0 - • ' • • • w·' 0 4 8 z a Proton Rich -stable • Neutron Rich • • • • • 12 16 20 Figure 6.2: Intensity of the most-probable-charge-state component in a 150 keY /u ion beam after passing through a carbon foil stripper, as a function of ion atomic number. time, variation of the axial field component to give a prescribed beam dynamics behavior (bunching and acceleration) can be achieved by changing the size of undulations in the quadrupole electrode surfaces [2]. The transverse focusing strength in a RFQ is characterized by a parameter B , given by (6.3) where V is the inter-vane voltage, q is the ion charge state, M c2 is the ion rest energy, r0 is the average aperture radius, and A is the free-space wavelength corresponding to the operating frequency. Stable ion transport requires that 0 < B < 17.92 [3]. Values at or near either extreme, however, correspond to zero or very-low beam acceptance and are therefore impractical. Inter-vane voltages are generally chosen as high as practical, limited either by a sparking limit or, in the case of high duty factor RFQs, possibly by a cooling limit; r0 must be chosen large enough to · accommodate the beam emittance. Therefore, in designing the RFQ for the singly-charged ions of the ISAC beam, A in Eq. 6.3 is the only free parameter available to compensate for the low qfM and maintain B within the stable range. Contrarily, the physical size of the accelerator, both in length and diameter, increases with A so that , in general , the highest possible operating frequency is preferred. 104 Input £:.2 keV/u q/A>1/30 T RFQ !=160 keV/u, q/A>I/G Matching & Stripping Section Output E<l.D NeV/u T JH DTL I" B.i! " 3.95 " "I 6.5 PI "I 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 ~ ~ ~ ~~~ ~liiJU ~ $ $ $ n ~ 1 +l-80./ /~rv ~ 6 6 6 (1)6 6 (I) 36 MHZ, ... 100 kW 70 NHz. B Amputlen, Total power ... 160 kW c ron .tripper Figure 6.3: Block diagram of an accelerator for ISAC. Taking singly-charged mass 30 ions as the worst case design requirement, we are led to an operating frequency choice of 35 MHz. With V = 85 kV, and r0 = 0.86 em this gives a design value of B = 3. This is somewhat lower than is conventionally used but seems, from initial beam dynamics calculations, to provide adequate focusing. The vane voltage corresponds to a fairly conservative 1.1 times the gap-dependent Kilpatrick sparking criterion; with the above choice for r0 the normalized RFQ acceptance is 0.511" mm ·mrad. z Figure 6.4: One quadrant of an RFQ unit cell. The undulations of the RFQ vane surface are characterized by a modulation factor m defined in Fig. 6.4. From the beam dynamics point of view, once V and the operating frequency are chosen, the RFQ design is specified by two independent parameters, the mod-ulation factor m(z) and the synchronous phase </>11 (z), where z is the axial distance along the accelerator. In the original LANL RFQ design procedure [4], the RFQ was divided into four sections (radial matcher, shaper, gentle buncher, and accelerator section) with different design laws for parameter values and z dependence in each section. The injected DC beam is adapted to the time-dependent focusing forces of the RFQ in the radial matcher, where the transverse focusing is gradually increased, by increasing B over a distance of several {3)... In the shaper </>11 and m are increased linearly while B is held constant; in the gentle buncher adiabatic bunching is achieved by increasing m and <l>s in a manner that keeps the longitudinal small-amplitude oscillation frequency and the spatial length of the separatrix constant. In the final section, where most of the acceleration is accomplished, all parameters are held constant. The LANL code RFQUIK aids in the parameter selection according to 106 the above design criteria. A modification of the RFQ design procedure, for the situation when space charge is not significant, has been introduced by Yamada [5). In this case the gentle buncher and accelerator sections are augmented by prebuncher and booster sections. A rapid beam compression is accomplished in the prebuncher section by a relatively-rapid ramping of </>8 from, typically, -88° to -60° while keeping the separatrix area constant (constant longitudinal acceptance) and allowing some increase in the RF defocusing parameter with /3. Bunching is completed in the buncher section where m and </>a are inaeased rapidly, subject to the continued constraint of constant separatrix area, as well as to a specified maximum value for the RF defocusing parameter, to preserve the desired beam envelope size. In the booster, which is a transition section between the buncher and accelerator sections, the accelerating field is raised as high as possible by reducing the aperture to a size, within a specified safety factor, to be no larger than necessary to accommodate the estimated beam envelope. With the aid of the code GENRFQ, non-spacecharge RFQ designs can be generated that have significantly-smaller longitudinal emittance growth and are somewhat shorter, in general, than designs obtained through the LANL approach [6]. A further reduction in longitudinal emittance is realized in the ISAC RFQ by modifying the shaper section so that it becomes, in effect, a discrete buncher. Five cells at the beginning of the shaper, with a modulation factor m = 1.01, generate a small, local, axial electric field to provide a sinusoidal, time-dependent velocity modulation in the beam. This is followed by an 80 em drift with no vane modulation ( m = 1) in which the beam bunches during a 1/4 synchrotron oscillation. Fig. 6.5 illustrates the parameter variation along the length of the RFQ. The strong focusing in the RFQ channel leads to exiting beams that are highly convergent in one plane (xz, say) and divergent in the orthogonal plane (yz). To ease the design requirements of the following magnetic beam transport elements, the convergence and divergence of the beam can be reduced somewhat by tapering the focusing parameter B. In this case B is reduced from 3 to 2 over a length of 1.25 m at the output end of the RFQ. The taper is gradual enough to avoid inducing mismatch betatron oscillations, and has no effect on the longitudinal motion. From a mechanical point of view, the low operating frequency of the RFQ dictates that some form of resonant, semi-lumped parameter structure be used to generate the required high RF voltages on the RFQ electrodes, rather than a resonant cavity, as used at higher frequencies. Various semi-lumped parameter RFQ structures have been built and studied at several laboratories [7]. The structure proposed for the ISAC accelerator is a variant of the 4-rod structure first developed at the University of Frankfurt [8]. The distributed capacitance of the RFQ electrodes, in this case four rods or vanes of small cross section, in combination with the inductance of the split ring supports as illustrated in Fig. 6.6, forms the resonant circuit. Calculations using the 3-dimensional electromagnetic field code MAFIA combined with model measurements have been used to optimize the structure dimensions. For our current 35 MHz design, the rings have a rectangular cross section (15 em axially by 8 em radially), a mean diameter of 43.5 em, and are spaced at 40 em intervals along the RFQ electrodes. The 107 (A) (B) (C) ~ .850 4-..-J--.L....--1--,,L-....J---L-..I.-....1--1....--t .7 ~ N ~ W~65 .6~ .680 .595 .510 .425 .340 .255 .170 .085 I I I~ t: 1..:! -~I~ lilt .. ,_ it! /:/I 0 ~,;; , Q. bore radius a(z) .5 .4 .3 .2 .1 _ ... . 000 .0 0 75 150 225 300 375 450 525 600 675 750 length (c:m) -20 ---------· .£.;:)U f,(z) I m(z) I -30 I I 2.25 -40 •-;;;-I-.;. 2.00 IE -50 I >r t-.8 1.75 ,.s -60 /t: 1:! 1.50 -70 ,..g I -6 -80 /t us " --... -90 1.00 0 75 150 225 300 375 450 525 600 675 750 length (c:m) 0.00 .150 " W(z) ll(z) ... " " .125 " , -0.01 , ,..... ~,~ / , .. .100 ~ ! , -o.02 .r . ," .075 !I •,S. , u F" 0 ~/ ~ , .050 -0.03 , -, .: , .025 , , ; -o.04 --- .000 0 75 150 225 300 375 450 525 600 675 750 length (c:m) Figure 6.5: Design parameter variation along the RFQ length: a) aperture a(z) and acceler-ation factor A(z); b) modulation index m(z) and synchronous phase </>,(z); c) kinetic energy per mass unit W(z) and RF defocusing factor ~(z) . 108 Figure 6.6: One model of a split-ring, 4-rod RFQ. theoretical specific shunt impedance z& for this structure, when enclosed in a 1 m diameter cylindrical tank, is 500 kO·m, where Z6 is defined in terms of the peak electrode voltage V, and power dissipation per unit length PL, by Table 6.2 summarizes the parameters of the current reference design for the ISAC RFQ. 6.4 Beam-matching and Stripper Section As noted in Section 6.3.1, the beam from the RFQ passes through a thin carbon foil to strip additional electrons from the ions before injection into the drift-tube linac. To minimize the effect of scattering in the foil, and thus to limit transverse emittance growth, we need a strongly-converging beam in both the xz and yz planes. Four quadrupole magnets provide sufficient degrees of freedon1 to transform the RFQ output beam to the required double waist and small beam spot size, as illustrated in Fig. 6.7(a) [9]. With allowance for drift spaces to accommodate diagnostic devices, this section of the beam line is 1.8 metres long. To match the stripped beam to the drift-tube linac requires a further four quadrupoles and, because the energy spread in the RFQ output beam leads to 109 Table 6.2: ISAC RFQ design parameters. INPUT /OUTPUT Particle mass :::; 30 u Particle charge ±1 Input energy 2 keV /u Output energy 150 keV ju VANE PARAMETERS Focusing factor (B) 3 Aver age aperture ( r0 ) 0.88 em Minimum aperture (a) 0.46cm Modulation factor ( m) 1-2.5 STRUCTURE Type 4-rod split-ring Ring diameter 0.434 m Ring spacing 0.4 m Tank diameter 1m Overall length 8m RF Frequency 35 MHz Specific shunt impedance 500 kO·rn Vane voltage 85 kV RF power ~100 kW BEAM DYNAMICS Transverse acceptance 0.557!' mm·mrad (normalized) Input emittance 0.257!' mm·mrad (100% contour) Output emittance 0.257!' mm·mrad (98% contour) Longitudinal emittance 1 (keV ju)·ns Synchronous phase ( <l>s) -90° :S </>s :S -20° Transmission 92% (of DC input beam) 110 20 . 0 II ~5. DtD Horiz. Long Vrrt Lrngth• IBDD.DD 11 (A) 20 . 0 11 ~5. Ora Vtrt Length• 1750.00 11 (B) Figure 6.7: Transverse and phase envelopes along the beam line from: a) the RFQ to the stripper foil; and b) the stripper foil to the drift-tube linac. ( Q is a quadrupole magnet and G is a buncher gap.) some debunching, a rebuncher cavity is required as well. The component layout and beam profile for this part of the beamline is shown in Fig. 6. 7(b ). The buncher in this case would be a double gap, folded A./4 coaxial resonator operating at 35 MHz with an effective gap voltage of 65 k V and a synchronous phase angle of -90°. 6.5 Drift-tube Linac While the RFQ is very effective in capturing, bunching and providing some acceleration of the DC beam delivered from the mass separator, it is in the much more efficient drift-tube linac that the major part of the acceleration is accomplished. The so-called IH (Interdigital H mode) linac [10] (see Fig. 6.8) is chosen in this case because it offers both very-high shunt impedance (i.e. high efficiency) for the low-velocity ions being dealt with here, and small transverse dimensions relative to the wavelength corresponding to its operating frequency. The high shunt impedance is due, in part, to the use of thin cylindrical shells as drift 111 Figure 6.8: Simplified cross section and end views of a typical IH drift-tube linac. tubes which, as a consequence, results in relatively low capacitive loading of the structure. However, this means that quadrupole magnets , often included in linac drift tubes to com-pensate for the defocusing in the accelerating gaps and to transport the finite-emittance beam, cannot be used in this case. Instead, quadrupole doublets or triplets are included periodically between groups of drift tubes along the length of the linac. A common design approach for IH linacs is the combined zero degree synchronous par-ticle design [11], in which RF defocusing is minimized by designing groups of cells for a synchronous phase of oo, followed by a few cells with a negative synchronous phase to main-tain phase focusing, and then by a quadrupole lens for transverse focusing. Designs with up to 30 accelerating gaps between transverse focusing elements are possible with this scheme, but flexibility in achieving energy variability is limited. The proposed ISAC DTL uses a more conventional linac design, somewhat similar to that being used at the Institute for Nuclear Study (INS) in Japan [12], with all accelerating gaps designed for a constant negative synchronous phase angle of -25°. With this choice, quadrupole doublets at six-cell intervals1 are sufficient to compensate for the RF defocus-ing corresponding to a 3.0 MeV /111 accelerating gradient, and to transport the beam (with normalized transverse emittance ~ 0.57r mm·mrad) through the channel defined by the 2 em diameter drift-tube aperture. Using eight groups of six cells provides an output energy of 1.64 MeV /u (slightly higher than the specification in Table 6.1). Energy variability is 1 We consider two half drift tubes and the included accelerating gap as a cell. 112 Table 6.3: Drift-tube linac design parameters. IONS Charge/mass ratio (q/ A) ?. 1/6 Input energy 150 keY /u Output energy 1.64 MeV /u STRUCTURE Type IH Number of cells/section 6 Number of sections 8 Drift distance between sections 36 em Drift-tube diameter 2.8 em Drift-tube aperture diameter 2.0 em RF Frequency 70 MHz Power ~ 150 kW Eacc 3 MV/m ¢3 -25° FOCUSING Periodicity 6 cells-Quad Doublet-6 cells Quadrupole length 12 em Quadrupole aperture diameter 4cm Quadrupole gradient 2.6-4.4 kG/em accomplished by selectively tuning on the RF drive of the separately-driven, six-cell units, starting with the lowest-energy one, and partially powering the last powered unit. In the current reference design, the total length of the eight-unit DTL is 6.5 m, and the total RF power requirement is estimated to be 150 kW at 70 MHz. To minimize the effects of beam debunching in drift spaces, the inter-unit spaces have been kept small (0.36 metres in this case) and just large enough to accommodate the quadrupole doublets along with some beam diagnostics. (n view of the small inter-unit drift spaces, it is not practical to have each six-cell unit accommodated in a separate vacuum tank. Instead, the present conceptual design calls for one, or perhaps two, tanks divided internally into either four or eight independent RF sections. The basic DTL design is summarized in Table 6.3. 113 6.6 Beam Dynamics Calculations Self-consistent particle-tracking computations through the RFQ, stripper/matching section, and drift-tube linac have been carried out using two computer codes: PARMTEQ and LINCALC. The latter, a new code written at TRIUMF, is equivalent to the linac generating and particle tracking code PARMILA (13], but has the ability to do time-dependent particle tracking. ~+---~----~----._---+ 30 -30 ~~~~om~>,·-~--·-~-r----~---+ -0.4 -o.2 0.0 0.2 0.4 X{CW) (A) ~+---~~--~----~----+ 30 0" I o r -30 cJ.UJS) • .25 • ~ ~+---~~---r----~----+ -o.4 -o.2 0.0 0.2 0 .4 Y(CW) (B) 11.0 +---~-----.1....----..L----+ ~.5 e 60.0 .... .... 5!1.5 (C) Figure 6.9: Scatter plots in the xx' (a) and yy' (b) phase-space projections at the RFQ input, and a plot in E, </>space at cell91 (c), illustrating bunching at the end of the shaper section. We start with a monoenergtic 2 keV ju, singly-charged mass 30 ion beam. A random distribution of 5000 particles in the 4-dimensional transverse phase space, with a normalized emittance of 0.257r mm·mrad (matched to the RFQ input), is used as the particle coordinate input for the code PARMTEQ which then calculates the particle motion through the RFQ. Scatter plots in the xx' and yy' phase space projections of the beam at the RFQ input, and a plot in E,</> space at cell 91, illustrating the beam bunching at the end of the shaper, are shown in Fig. 6.9. A similar set of plots at the end of the RFQ (see Fig. 6.10) shows that there is virtually no transverse emittance growth through the RFQ and that the longitudinal 114 emittance also kept quite low at 1 (keV /u)·ns [14]. () () 5 5 c c c c !i 0 !i 0 )(' f IL -5 -5 '• ,. 0.25 " W•~RAO •. • 0.26 tr t.M.A•~RAO -() -() -0.6 -0.4 -0.2 0 .0 02 0 .4 0.6 -0.4 -0.2 0 .0 0.2 0 .4 X(Ciol) Y(Ciol) 4 .58 () 1 • 33 " KEV•NSEC 4 .56 "' 8 ... 4 .54 ~ ~ 4.52 8 6 . Q ... ;;r .. .... ::; 4 4.50 J il' 4 .411 i 2 4 .46 0 -35 -30 -25 -20 -15 -() -5 0 4 .46 4 .411 4.50 4.52 4.54 4 .56 .(35 MHZ OEG) [N£RGY(IoiEV) Figure 6.10: Scatter plots in the xx', yy' and E, </>phase-space projections at the end of the RFQ for the input shown in Fig. 6.9. The calculated beam coordinates at the RFQ output were then used as input for the low-energy beam transport routines of LINCALC, which tracked the particles throught the matcher/stripper section [9]. Straggling and scattering in the stripper foils, assumed here to have a thickness of 9 pg/cm2 , account for some emittance growth, both longitudinally and transversely. Based on semi-empirical estimates of the worst-case scattering and straggling scenario (that is, for proton-rich ions near the drip line), we have assumed for these calcula-tions that the energy-straggling and angular-scattering distributions are Gaussian with 1u widths of 8 mrad and 26 keV, respectively. The phase-space plots before and after the stripper (see Fig. 6.11) show that the effective transverse emittance (that is, the area of the best-fit ellipse in phase space enclosing 98% of the particles) is increased by about 75%, and the effective longitudinal emittance is nearly quadrupled. A reduction in this emittance growth would be possible if the added complexity of a gas stripper, and about a 20% reduction in mean charge, were acceptable. Results of the final stage of beam-dynamics calculations through the drift-tube linac, us-ing the LINCALC code, are shown in Fig. 6.12. Here the calculated phase-space plots for a 1.5 MeV fu 30Na beam at the output of the linac show final transverse and longitudinal emit-tances of 0.9471' mm·mrad and l11r (keV /u)·ns, respectively. Overall transmission through the linac, excluding losses at the stripper, is 86%. 115 BEFORE STRIPPER 40+-----~----~----~-----T 20 ,..... c c i 0 --)( n. -20 £N -= .25 7T MM•MRAD -40+-~---.-----r-----.-----+ -0.2 -0.1 0.0 X(CM) 0.1 0.2 40+-----~----~----_.-----+ 20 0 -20 . . -404------.-----r-----.-----+ -0.2 -0.1 0.0 0 .1 0.2 y 4.7 +-------~------_.-------+ 4.6 ,.... > ~ 4.~ --...., 4.4 £ = 32 n KEV•NSEC AFTER STRIPPER 40+-----~----~----_. ____ -T 20 -20 . . £N = .46 7T MM•MRAD -404-~---.-----r----~-----+ -0.2 -0.1 0.0 0 .1 0.2 X(CM) 40+-----~----~----_.------r • 20 0 -20 £N = .52 7T MM•MRAD -40+---~-.-----r----~------+ -0.2 -0.1 0 .0 0 .1 0.2 Y(CM) 4.7 +-------~---------1.--------+ 4.6 > ~ 4.5 ..... ...., 4.4 £ = 125 n KEV•NSEC ... .3 +--------.-------r-------+ 4 . .3 +---------.-------~-------+ -150 -1.30 -110 -90 -150 -1.30 -110 ;(.35 MHZ OEG) ;(.35 MHZ OEG) -90 Figure 6.11: Phase-space plots before and after the stripper for a 30Na beam. 116 X-PX ~+-----~----~----~----+ I) 5 0 c ! 0 ';( Q. -5 '> -I) '• "' 0.52 1r MM•IIARAO -~+-~--.-----r-----r-----+ -1.0 -0.5 0 .0 0.5 1.0 X(CM) rJ>-E !i0.5 +-__._ __ ..___.__;_~_._ __ ...___._--t !iO.O 411 .5 !I 49.0 ;:; 48.5 480 '. ' "' 335 n KEV•NSEC 4 7.5 +----.--,----.--,----r--.----r---+ 145 150 155 160 165 170 175 180 '115 .(70 MHZ DEG) Y-PY ~+--~-~~~~--~---+ 0 c ! 0 r -5 -I) '• "' 0.94 n' MM•MRAD -~+-~,__-,----,---,----,----~ -0.6 -0 4 -0.2 0 .0 0.2 04 0.6 Y(CM) ENERGY HISTOGRAM Ill 8 ~ 8 6 0 "' ~..... 4 :i IX ~ 2 Figure 6.12: Phase-space plots and energy histogram for a 1.64 MeV /u 30Na beam at the output of the drift-tube linac. References [1] E. Baron and C. Ricaud, in Proceedings of the European Accelerator Conference, Rome, 1988, World Scientific, New Jersey, 1988, p. 839; with minor modifications to Eq. 6.1 by P. Bricault. [2] I.M Kapchinski and V.A. Teplyakov, Prib. Tekh. Eksp. 2 (1970) 19. [3] D.A. Swensen, in Proceedings of the 10th Linear Accelerator Conference, Montauk, New York, 1979, Brookhaven National Laboratory Report, BNL-51134, 1979, p. 129. [4] K.R. Crandall, R.H. Stokes and T.P. Wangler, in Proceedings of the 10th Linear Accel-erator Conference, Montauk, New York, 1979, Brookhaven National Laboratory Report, BNL-51134, 1979, p. 205. [5] S. Yamada, in Pmceedings of the 1981 Linear Accelerator Conference, Santa Fe, New Mexico, 1981, p. 316. [6] J. Staples, in Proceedings of the Conference on Heavy Ion Accelerators and their Appli-cation to Inertial Fusion, Institute for Nuclear Science, Tokyo, 1984, p. 379. [7] A. Schempp, CERN Accelerator School, 1992, CERN Report, 92-03. 117 [8] A. Schempp et. al., in Proceedings of the 1984 Linear Accelerator Conference, Seeheim, Germany, GSI Report GSI 84-11, 1984, p. 100. [9] S. Koscielniak, TRIUMF Design Note TR-DN-95-3, 1995. [10] U. Ratzinger, in Proceedings of the 1990 Linear Accelerator Conference, Albuquerque, New Mexico, Los Alamos Report LA-12004-C, 1990, p. 525. [11] U. Ratzinger, in Proceedings of the 1991 Linear Accelerator Conference, San Francisco, California, IEEE Conference Record 91 CH3038-7, 1991, p. 567. [12] M. Tomizawa et al., in Proceedings of the 1993 Linear Accelerator Conference, Wash-ington, D.C., IEEE Conference Record 93CH3279-7, 1993, p. 1786. [13] G.P. Boicourt, in Proceedings of the Workshop on Linear Accelerator and Beam Optics Codes, 1988, Los Alamos Report LA-UR-88-1544, 1988, p. 1. [14] S. Koscielniak, TRIUMF Design Note TR-DN-95-4, 1995. 118 Section 7 CONTROLS 7.1 Overview The ISAC control system will be an extension to the existing TRIUMF Central Control System. It will provide operators and users of the facilities with a clear and comprehensive view for monitoring and allow control to be managed in a straight-forward manner. The system will be based on networked computers communicating via ethernet and will use serial CAMAC as the primary control system bus. X terminals will form the backbone of the user interface to the control system. The X terminals may be located anywhere on the TRIUMF site to which the network extends, including close to machine or experimental equipment. The control system will provide a suite of software components from which application programs can be created. It will be possible for application programs to display their results in textual or graphical form on an X terminal and to have the same data sent to a particular data acquisition system for logging as part of an experiment. The system will support commissioning of devices and the day-to-day operations of the completed facility. A similar but separate system will support the development test stands. 7.2 Central Controls Philosophy General purpose 32 or 64 bit computers now form the backbone of most control systems. Dedicated embedded computers are used to provide intelligent control for sub-systems, such as RF, monitors, targets, vacuum systems, etc. The 32/64 bit computers typically allow the use of software tools such as, high-level languages (FORTRAN, C, C++), on-line databases (ORACLE, SYBASE, etc.), windowing user interfaces (e.g. Motif and the X Window sys-tem), and network communications between programs executing on different computers. Accelerator operators responsible for the operation of the facility typically monitor the facility in a number of ways via an operator interface. The interface comprises: dedicated displays of device or system parameter values (LED displays, analogue meters, etc); TV monitor displays of device or system parameter values; and workstation or X window terminal textual and graphical displays of device or system parameter values. 119 Similarly, the operators control the behavior of the facility via an operator interface comprised of: dedicated manual knobs, sliders, buttons, etc; keyboards for typing ASCII commands into a computer; and one or more windowing terminal(s), with an attached mouse. In recent years the trend in operator interfaces has been to move away from the use of dedicated displays and controls to the use of general purpose workstations or X termi-nals. Workstations and X terminals offer an adaptable and efficient platform for displaying information. The control system for ISAC will be based on these recent advances in computer tools and technology. This will allow a cost-effective control system to be created that will enhance the effectiveness of the facility while remaining flexible and easy to maintain. 7.3 Implementation The ISAC control system will be implemented as an extension to the existing TRIUMF Central Control System (T-CCS). This extension will be constructed in phases as each of the beam-line, accelerator or experimental areas is installed and will provide a robust, well-tried control system for early commissioning procedures and later routine operations. As part of these extensions, the T -CCS will be enhanced to provide the appropriate displays to ISAC operators in a form that has a similar look and feel as the existing T-CCS. These displays will be provided on X terminals in the ISAC control room, or on any other X terminals authorized to operate the facility. Operators will monitor and control the facility via the X terminal displays. The current T-CCS is a multiple-processor, distributed system that allows the TRIUMF cyclotron operators to control the cyclotron and its primary beam lines. It provides: an op-erator interface, whereby all primary-beam facilities are controlled; an equipment interface to all controlled/monitored hardware in the primary-beam facilities; computers, instrumen-tation and software to manage both interfaces and provide required closed-loop operations; spare equipment to handle hardware module failures; and hardware and software diagnostic facilities to help characterize and identify problems. The T-CCS is composed of a set of older 16 bit Data General Nova computers, which are in the process of being phased out, a group of 32 bit DEC VAX computers dedicated to operational aspects of the T -CCS, and a second group of VAX computers dedicated to the development of beam dynamics and T-CCS software and hardware. All of the above computers are interfaced to a large CAMAC system based on the Hytec System-Crate architecture. In this architecture, a standard CAMAC crate is used with a non-standard crate controller (termed the Executive Crate Controller) to become an Executive Crate. Each Executive Crate Controller can manage up to 8 standard CAMAC branches by way of Branch Coupler modules plugged into it. The CAMAC standard provides for two types of branches, parallel or serial. Each parallel branch can have up to 7 standard CAMAC crates located along its length, while serial branches can have up to 62 crates. The CAMAC crates located along each branch have up to 23 slots available in them, in which 120 CAMAC standard modules can be installed. These modules directly monitor and control most equipment in the TRIUMF cyclotron and its primary beam lines. The existing T-CCS employs 3 Executive Crates with 10 branches distributed among them. Any computer needing to access modules located in a particular CAMAC crate must be interfaced to the appropriate Executive Crate. Application programs running in that VAX are then able to read and write data to and from the required CAMAC modules. Presently, the Nova computers are interfaced to only one Executive Crate, but some VAX computers are interfaced to all three. The operator interface to the T-CCS is presently provided by: dedicated knobs, buttons and displays on a traditional operators console; old remote console configurations comprising button and thumbwheels for device selection, LEDs and lights for monitoring, and buttons for slewing and control; and some recently-installed X terminals replacing traditional hardware listed above. Operators use an X terminal to run application programs on the VAX computers. These programs allow the operators to monitor and control the behavior of the cyclotron or its beam lines. Most of these application programs incorporate a windowing interface. Interface building tools are used that assist programmers in creating, modifying and updating control displays. 7.4 Integration with User's Requirements In addition to having a dedicated data acquisition system for their experiments, experi-menters using the ISAC facility will likely need: to view a summary of the state of the cyclotron, BL2A or the TSAC facility ; to view historical information (logged data) regarding the state of these facilities; to control beam-line devices in order to improve the "quality" of the beam for their experiments; and to have control-system information readily available to their data acquisition system. The ISAC control system will provide these services to users and make them available via an intuitive windowing interface on X terminals. To access the services of the ISAC control system, users will be required to "sign-on" to certain VAX computers and run specific user-application programs. These programs will provide a windowing interface to the ISAC control system (such as those listed above). The form of the user interface will be designed in collaboration with the users of the facility. Each application program will likely require a different form for its displays. For example, monitoring a split-plate protect monitor may require vector displays, monitoring beam-line vacuum may require a digital "chart recorder", and tuning beam-line elements may need push buttons for change of state, slew buttons for value changes, and textual display for present value. The ISAC control system will provide a set of high-level software components with which new user application programs can be constructed. These tools will improve the integrity of the control system by re-using components that have already been well-tested through previous use, provide some level of protection against users trying to make inappropriate changes, and provide a level of security by only allowing users to execute applications which 121 they are authorized to run. 122 Section 8 RADIOLOGICAL SAFETY 8.1 General Considerations TRIUMF has considerable experience of problems and constraints arising from residual radi-ation fields produced by 100 J.LA, 500 MeV proton beams. The operating shield requirements and residual radiation fields around the ISAC source targets will be less than those found at the meson production targets and residual beam dump on the main beam line by the lower intensity factor. The most significant difference in the ISAC source targets will be the quan-tities of 'loose' radioactivity that will be deliberately driven off and/or unavoidably released from the source targets. The disposition, containment and handling of this radioactivity is the overriding design constraint for the ISAC facility. Section 8.2 gives estimates and descriptions of the elemental and mass distributions of species produced in selected thick targets bombarded by 500 MeV protons. The targets selected are r~presentative of those already in use at the TISOL facility. Section 8.3 describes the internal disposition of the dominant species emerging from the target, the most important mechanisms for their transport, and the constraints this will place on access to various components during operation or for maintenance after shutdown. Section 8.4 defines the operating dose-equivalent rate criteria and the methods for estimating the beam-line and source-target shield requirements. The expected ISAC accelerator shield and access-control requirements are also described briefly. 8.2 Radioactivity Production in ISAC Source Targets Figure 8.1 shows the basic mass distributions of spallation reaction products from thin scandium, niobium, lanthanum and bismuth bombarded by 200 and 400 MeV protons. The distributions are based on Silberberg and Tsao's [1] empirical fit to the experimental data 123 available in the early 1970's. This empirical description can be written in the form { [z- Z0 (A At)]Ls} a= a0 (At. E)exp[-P(E)(At- A)Jexp - Zw(A) O.ry~. 139La 209Bi 10 3 93Nb '"0 <1.> >- . 10 2 c ·-0 ...c 10 1 u (/) (/) 0 10 ° ~ c 0 ........ 10-1 0 --0 a.. (/) 1 o-2 0 100 200 A Figure 8.1: Proton spallation reaction mass chain yields for 200 and 400 MeV protons on Sc, Nb, La and Bi. Plotted in Fig. 8.1 is the product of the mass chain yield factors a0 and exp[-P(E)(At-A)]. Ignored are the charge distribution (the second exponential) and various nucleon-pairing and ad-hoc correction factors for particular elements, all of which are of order unity, at least at the peak of the charge distribution. 124 The charge distributions from spallation reactions are quite independent of target mass and bombarding proton energy, at least for proton energies below 500 MeV. They are only weakly dependent on product mass varying, in width (1/10 value) from 1.2 charge units at A-= 10 to 1.0 charge unit at the minimum at A=40, and rising again to 1. 7 charge units at A=200. The centroid of the distribution is seldom further from the mean stability line than one charge unit and is usually within a fraction of a charge unit, with a small average bias toward the neutron-deficient side. For higher masses, up to lead and bismuth, the dominant mechanism remains spallation but high-excitation fission begins to contribute, at least at the higher proton energies. For the highest-mass targets (thorium and uranium) the dominant de-excitation mecha-nism is fission, with a tendency to more equal mass binary fission products than for thermal or low-energy neutron fission. However, there remains a significant probability ( ~0.2±0.1) that spallation (or so-called peripheral) reactions populate the species near, and to the neutron-deficient side of, the quasi-stability line. Some of these products are very volatile o emitters; radon and, possibly, polonium could be difficult to contain. The estimation of the yields of the four unique a -decay chains, and the multiple sub-chains not coupled by {3 decay from spallation of uranium and thorium, will require careful analysis, especially those involving the longer-lived radon and polonium species. For targets that degrade the incident 500 MeV protons to 300 MeV (80-125 g/cm2 ) approximately 50% of the incident protons suffer nuclear collisions; degrading the proton energy further to 100 MeV (55-85 g/cm2) adds less than 20% to the proton collision fraction. Thus for lOJ.LA of 500 MeV protons incident on a ~100 g/cm2 target, the total reaction yield is approximately 0.5 X 6.24 X 1013 = 3 X 1013 s-1 • Because the number of decays per residual species is of order unity, this is also the saturated total source strength in the target. Not all species saturate, of course, but a substantial fraction have half-lives shorter than the expected bombardment times for the production targets, yielding individual source strengths between 50% and 100% of saturation. The assumption of saturation is a con-servative but useful working approximation for most purposes. 8.3 Radioactivity Distribution from ISAC Targets Because the objective is to release radioactive species from the production target, the system must accommodate release of most of that inventory. Only in those cases where the elemental species required are the most volatile would significant relief be available. The yield from the target, both desired and unavoidable, is dependent on thermal diffusion, the thermodynamics of surface binding, and the distribution of ionization charge states in the source. The neutral species will simply diffuse through the extraction port until they come into contact with a wall cool enough that they can stick to it. The dwell period on such a surface is given by 125 the phenomenological expression [2] T = T0 exp( cQ /T), where Q is the sticking enthalpy of the volatilized species on the wall surface, T is the wall temperature, c is a normalizing constant, and T0 is an essentially-universal transit time for uninhibited traversal of an interatomic distance. For cQ :» T the sticking time increases very rapidly with decreasing temperature allowing the possibility of holding species with high Q values indefinitely. Because some of the targets can emit significant quantities of target material and the Q values can be substantially changed by even a one-atom surface layer, the actual sticking factors found in practise are mostly empirical. Volatile species generally migrate from warm and/or low-Q surfaces to cooler/higher-Q surfaces. Beyond the immediate vicinity of the ISAC ion source, the transport will be governed by the accelerated beam characteristics. At the first momentum selection element most of the species with the wrong momentum and q/ A ratio will be buried in scraper slits. The total radioactivity deposited at this point could be a significant fraction of the target im·entory, at least for the volatile fraction. The migration of radioactivity from the beam scrapers and slits will depend on the diffu-sion of the buried species back to the surface. This will depend on many factors, including the deca:y modes of the buried radioactivity and the radiation damage to the bombarded surface. It will be necessary to back the vacuum pumping systems by isolated, shielded cold traps between the high-vacuum and rough-vacuum stages in order to concentrate any medium-to-low-volatility species in isolated traps. The roughing stage will need to pump into ballast volumes to hold any high-volatility species, such as noble-gases, for assaying before release, and to guard against anomalous releases from malfunctions of the system. 8.4 Production Target Shielding The shield attenuation around targets bombarded by 500 MeV protons can be estimated using the Moyer model [3] and the empirically-determined normalization and relaxation length parameters for the usual shield materials, namely iron and concrete. The dose-equivalent per proton incident on any target (except hydrogen) at a field point a distance r and angle (} from the incident beam direction through a shield of thickness t is given by D.E. = Hr- 2 exp( -{3{}) exp( -±). The variable His given by Ho f E, where Ho is an empirical normalization constant equal to 3 x 10-13 Sv per proton for proton energy E in GeV, and f is a correction factor (the fraction of protons suffering non-elastic nuclear collisions before being stopped by ionization) used to extrapolate from higher proton energies. The quantity /3=2.3 radian-1 is the empirical field point angle dependence factor. It is not significantly proton-energy-dependent, because most of the angular divergence in the hadron cascade occurs at the end of the cascade, when the 126 10 5 '---L...--L...----..IL...-----..IL...------11.3....----II-----I 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Concrete thickness t. m Figure 8.2: Dose attenuation in concrete for 1 MeV 1 rays. internuclear cascade hadrons have momentum of the same order as the intranuclear hadrons in the target nuclei. Finally, .A is the shield relaxation length, equal to 0.2 m for iron and to 0.5 m for standard concrete. For regions of substantial potential occupancy the objective total dose-equivalent rate is usually 1 11Sv /h. Limited regions of low occupancy may be shielded to operating dose rates of 10 11Sv /h, if any significant cost is involved. For regions where personnel can be rigorously excluded during operations, the dose-equivalent rate could be tolerated in the 100-1000 11Sv /h range, limited by roof and/or sky shine into occupied areas. The secondary channel from the ion source and transport line to the ISAC accelerator will require shielding at points where significant radioactive beam deposition takes place, such as targets, beam collimating elements and beam spill points. The penetrating radiation component will be high-energy 1 radiation. Presuming 1% of the target inventory estimated above, deposited at a downstream point, and emitting 1 MeV photons, the bare-target dose rate would be 3 X 1013 X 10-2 X 1 -------- = 43 mSv at 1 m, 47r1002 X 5.5 X 104 since a photon energy flux of 5.5 x 104 MeV /cm2 /s delivers a dose-equivalent rate of 1 mSv /h. To reach the objective 10 11Sv /h dose-equivalent rate at 2 m distance for uninhibited access (but not general occupancy), the surrounding shield must provide attenuation of the order 2 x 10-4 • This requires approximately 0.8 m of concrete as can be seen in the attenuation 127 curve for 1 MeV photons in Fig. 8.2. The attenuation of concrete for different photon energies is shown in Fig. 8.3, indicating the need either for a flexible shield arrangement that can be supplemented simply, or an accurate estimate of the 'Y-ray spectrum. 1 6 5 L..-----lE...-----1'--------'----1.---'-----"----' 0 2 3 4 5 E1 (MeV) 6 7 Figure 8.3: Dose attenuation for 1 m of concrete (density= 238 g/cm2). References [1] R. Silberberg and C.H. Tsao, Astrophys. J. Suppl. 25 (1973) 315 and 25 (1973) 335. [2] J.H. de Boer, "The Dynamic Character of Absorption, 2nd Edition," Oxford University Press, Oxford, U.K., 1968. [3] R.H. Thomas and G.R. Stevenson, "Radiological Safety Aspects of the Operation of Proton Accelerators," IAEA Technical Report Series No. 283, Vienna, 1988. 128 Section 9 COST ESTIMATES AND SCHEDULE Preliminary cost estimates (see Table 9.1) are based on: the current conceptual design of the facility; costs of a similar prototype accelerator at the Institute for Nuclear Study in Japan; and on cost estimates made for a larger ISAC facility proposed for beam line 2A. The ISOL facility is estimated at $5.0 M, including the mass separator, remote handling, and ion-beam transport system. The estimate of $7.5 M for the accelerator includes the off-line source required for tuning and commissioning. An allowance of $3.6 M has been made for controls, safety, experimental equipment, and administrative overhead. The estimated total cost of $18.1 M, exclusive of building and salaries, is somewhat larger than the $14.4 M indicated in the TRIUMF Five-Year Plan [1]. However, several items have been included which were not part of the original cost estimate. The conventional construction design and costing was done in some detail with the assis-tance of consulting engineers. The ISAC building is estimated at $9.7 M including shielding, cranes, air-handling systems, services, and a 15% contingency. It is estimated that the ISAC project could be completed in five years, after construction approval (see Table 9.2). The science program would, however, begin earlier, as low-energy beams from the ISOL system would be available approximately three years after the start of construction. Accelerated beams from the RFQ could be ready in the fifth year, if this were not in conflict with the construction of the higher-energy section of the accelerator. References [1] TRIUMF's Five-Year Plan, 1995- 2000, July, 1994. 129 Table 9.1: Cost estimates for ISAC (in units of 1,000 1994 dollars) . Item Cost Target test stand 500 Target area 1,200 Mass separator 800 Remote handling 1,500 Ion beam lines 1,000 RFQ 3,000 Linear accelerator 4,000 Off-line source 500 Controls and safety 1,800 Experimental facilities 1,000 Overhead/ administration 800 BL2A 2,000 Total 18,100 Target maintenance building 3,900 Experimental/technical support 3,500 Services 2,300 Total 9,700 130 ,-------- ---- ----- - --------- ---------- ----- - - -- ·----- ------ -- ----------------------- - - - ---------------- - --- , TRIUMF ISAC SUMMARY SCHEDULE 95/10/26 ID Task Name _____ ----+-'=-.:::....L...;::_.c...~atr -d§ir-1]air~91~tr 3 I atr .1 atr }Tcitr-~IT~~JcitrQ~ltrW~ir3TQB I atr 1 I atr ~~tr 3 I atr • f"§i~]"atr ~~tr 3 1 BUILDING FACILITY - -2 Experimental Hall Constru ction - ----· 3 Occupancy 4 Target Hall & 2A Constructi on -~ - . 5 Occupancy -6 TARGET AREA ··- -7 design ... 8 fabricate/install/commission -· 9 60 kev beam operati on 10 RFQ 1----11 design & prototype 12 fabricate/instaWcommi SSIOil ----. -13 150 kev/u beam ope ration .. 14 IH LINAC --. -- --16 design --16 fabricatefmstalllcommissio n . ---17 1.5Mev/u Beam Ope ration I t ' ~. - ---l ~'-' . > . ..;<e;;; • . ;,. ' . , • · -• 01/01 [, -. -s .... .. ,. ·;, .. . -----_ .. -- - --~ ---- ~~ - -:-1 • 05101 I c-----"l ; ·· c----~-::---::--:--:-::--~ - ---- -::-:-.-.. -_--. J - • 01/01 ! ~, ----·------------., ,-_ ;-;~7--:-- - -~--- .-- --- - ---:-~- =----::---::: ,; :,;:- ! £;~ #~. J • ~ r--:-~-- --·:-·-.. . -----·-·-] :+ 7~.' .. 04101 • 01/01 ---'-------'-------- - ------------- -------- ------- - ------- -----------'-------Project:TRIUMF - ISAC !----- - - ----- -- -Task Progress Milestone --- - - - ------- ----------- - ------ -- --------- --- ----------------------~ ~ -~ -.---:-~. ::=\J Summary Rolled Up Task • Rolled Up Milestone ........... Rolled Up Progress I . I - --------------- - --------------- -----------


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items