ISABELLE ANTIPROTON OPTION*

C. Baltay Columbia University, New York, New York R. Chasman, H.W.J. Foelsche, H. Hahn, M. Month, and A. van Stienbergen Brookhaven National Laboratory Upton, New York

Summary: The possibility of colliding antipro­ A somewhat different way of obtaining a high antiproton tons in ISABELLE has been investigated. are intensity is to use antiprotons originating in decay and again avoid the constraints of Liouville's accelerated on the second harmonic to 200 GeV in one theory.4 The antihyperons would be produced at a tar­ of the rings and are extracted into a bypass during get in the vicinity of the and would pro­ flat top. The 30 GeV antiprotons from a target placed duce antiprotons by decay in flight inside the aperture. at a low-β point in this bypass are transported to the However, in order to reach high intensities with this second ring where they are injected and matched into a method, injection times of the order of days are re­ bucket of a 10th harmonic rf system. This procedure quired. Furthermore, very high radiation levels would is repeated until all the buckets are filled. The result near the injection point.5 antiprotons are then accelerated to 200 GeV. Protons are now stacked in the first ring in the opposite di­ rection, accelerated to 200 GeV and made to collide In the ISABELLE antiproton option none of the with the antiprotons. luminosities up to above-mentioned technically difficult methods are ~ 1029 cm-2sec-1 seem feasible. used. A relatively high intensity of circulating antiprotons (~ 1 mA) is obtained in the following way: Introduction Protons (~ 10 A) are accelerated to 200 GeV in one of the rings by a second harmonic rf system and, while The attractiveness for intersecting storage rings still bunched, extracted in a single turn into a by­ such as ISABELLE would be further enhanced if they pass. The 30 GeV antiprotons from a target placed at a would provide interactions of a variety of particles low-β point in this bypass are transported to the sec­ besides -proton interactions. Ever since the ond ring, where they are injected and stacked in a ISABELLE Summer Study some design effort has been directedbucke t of a tenth harmonic rf system. This procedure towards exploring such possibilities. As a is repeated until all ten buckets are filled, where­ result several options have emerged and special care after the antiprotons are accelerated to 200 GeV. Protons was taken in the basic ISABELLE design to allow for are now stacked in the first ring in the opposite their addition at a later time at minimum cost. The direction accelerated to 200 GeV and made to collide antiproton option, to be described here, involves acceleratingwith the antiprotons. and storing a ~ 1 mA antiproton beam and colliding it at 200 GeV with 10 A ISABELLE proton beam. Antiproton Intensity A pp luminosity of the order of 1029 cm-2sec-1 is ob­ tained. The antiproton yield, for a primary proton energy of 200 GeV, has recently been measured.6 It has a Choice of Method plateau around 30 GeV with d2σ/d dp 0.5 per inter¬ action/sr (MeV/c). For small forward angles (<10 mrad), Various methods of accumulating large antiproton the number of 30 GeV antiprotons that can be stacked in densities in storage rings have been studied in the the ISA ring is given by: past. C d2σ ISA ΝΡ ΔΩΔPηstη . (1) The most straightforward mechanism for -production d dp ΜP B is to target protons outside the storage ring and to inject the produced antiprotons into the ring. In Here Νp is the number of protons in the ISA, this case the number of antiprotons collected in the ring depends on the production cross section as well as MP is the number of proton bunches, on the longitudinal and transverse acceptances of the ring. In previous studies, such a procedure alone has CISA is the ISA circumference, resulted in rather small intensities. To increase those, several methods have been suggested involving B is the proton bunch length at 200 GeV repeated injection made possible by transverse phase-space (antiproton bunch length at 30 GeV), damping or, in other words, by overcoming the limitations imposed by Liouville's theorem. Budker2 Δ (in steradians) is the solid angle at the proposed to damp the transverse oscillations of the target, inside of which antiprotons will be antiprotons by means of Coulomb collisions with a high accepted by the ISA ring, quality beam of traveling with the antiprotons. This is technologically complex and not very Δp (in GeV/c) is the momentum spread which will efficient for the ISABELLE parameters. Recently, be accepted in the ring at 30 GeV, van de Meer3 put forward the idea of damping betatron oscillations by detecting and compensating statistical ηt is the target efficiency including the effect variations of the beam center. This so-called sto­ of interactions within the target, and chastic cooling is being investigated experimentally at the ISR, but its usefulness is limited by rf noise ηs is the stacking efficiency. problems. The solid angle ΔΩ of the acceptable beam is given by:

ΔΩ = πa'b', *Work performed under the auspices of the U.S. Atomic Energy Commission.

572 where a' and b' are half angles of divergence. These Extraction and Targeting of Primary Proton Beam are related to the horizontal and vertical betatron acceptances, Ah and Av, of the machine by The 200 GeV proton beam will be extracted in the south experimental insertion (Figs. 1 and 2). The quadrupole Ah AV a' = and b' = , in this insertion will be rearranged to πa πb give slowly varying β-functions in the central part a and b being the effective half sizes of the source (βh ~ 12 m, βv ~ 25 m at the center) and to leave a at the target. The effective source dimensions must free space of 100 m there. In the beginning of this include an allowance for the divergence of the beam 100 m drift space, a 1 mrad horizontal kicker over the finite length, t, of the target. It is ad­ will be installed. The time between proton bunches is vantageous, in principle, to make the physical width about 4 μsec so that only very moderate demands are put of the target (or the proton beam size) as small as on the kicker rise time. After a 10 m drift space fol­ lowing this kicker, the beam will be sufficiently dis­ possible to get close to the limits amin = a'/2 and bmin = b,/2. However, practical considerations limit placed to enter the array of 3 septum magnets listed in Table I. They are followed by a 24 m drift space at the the effective source dimensions to a = √2amin and b = √2b leading to end of which the beam will enter a 60 mrad achromatic min +½ bend, followed by a triplet that will focus the beam at ΔΩ = √2 (AhAv) the target 5 m away. The target is located at a dis­ tance of ~ 3.5 m from the ISA ring, which leaves ade­ t quate space for shielding. Substituting in Eq. (1), one gets 2 The beam spot size at the target is only ½ 2 d σ cISA √2 (AhAv) ~ 0.25 × 0.25 mm corresponding to Β = 0.60 m. The ΝΡ d dp ρηsηt. (2) target will most likely explode from the exposure to MΡB t 6 × 1014 protons and will have to be replaced for each proton pulse. Betatron and momentum acceptances are closely re­ lated to the available aperture, which should be as Table I. Septum Magnets large as possible for maximum For antiprotons the available aperture can be made larger than that assumed for protons, considering the low intensity Septum Unit 1 2 3 (~ 1 mA). Space-charge effects will be negligible in this case and one can let the beam fill the vacuum chamber horizontally. Furthermore, beam loading prob­ Length (m) 1 1.5 1.5 lems do not exist. One can use a low cost, rf acceler­ Deflection (mrad) 1.8 4.0 4.0 ating system of higher frequency and voltage than what Peak field (kG) 12 18 18 is used for the proton beam. This in turn will make it Gap height (cm) 1.5 1.5 1.5 possible to stack the antiprotons azimuthally rather Septum thickness (mm) 2.5 5 5 than in momentum space, gaining both in available aper­ Peak current (kA) 15 22.5 22.5 ture and stacking efficiency. At the same time, the Current density (kA/cm2) 40 30 30 voltage can be made high enough so that the rf system will be able to accommodate as large a momentum spread as the vacuum chamber will accept. Antiproton Beam Transport

Protons in the ISA are accelerated on the second Antiprotons emerging from the target will be fo­ harmonic frequency, resulting in two circulating bun­ cused by a strong, large aperture, room temperature ches. On flat top, during which the protons are ex­ triplet located 3 m downstream from the target. There­ tracted, 70% of the particles in one bunch lie within after, they will enter a 21 mrad bending magnet, which 1/10 of the circumference. The antiprotons, produced will separate them from the protons that were trans­ by the protons, will have the same bunch structure. mitted by the target. The protons will continue into a One of the antiproton bunches will be injected, in the beam dump while the antiprotons will go into an achro­ second ring, into a bucket of a 10th harmonic rf sys­ matic bend from which they will emerge parallel to the tem with an efficiency of 70%. Ten ISA proton pulses ISA ring. Two quadrupole doublets will now match the (the second bunch from each ISA pulse will be dumped) antiproton beam into a transport channel that runs in are required to fill the ten buckets with antiprotons. parallel with the ISA ring and which simulates an ISA octant, with room temperature magnets. The horizontal aperture of 8 cm is subdivided in the following way at 30 GeV, assuming a momentum spread Antiproton Injection and Acceleration -6 of 2.5% and a horizontal acceptance Ah, = 2π × 10 m.rad (which is matched to the production cone): The antiprotons will be injected into the ISA by a fast kicker at the injection insertion ordinarily used Betatron amplitude: 1.9 cm for protons. (The rise time of this kicker has to be Phase oscillation amplitude [Χp (Δp/p)]: 4.3 cm ~ 150 nsec which is well within the present state of Sagitta: 1.3 cm art.) They will be captured and matched into a bucket Reserve: 0.5 cm of a 10th harmonic (1.12 MHz) rf system. The peak vol­ tage per turn required for matching is ~ 150 kV, which Total: 8.0 cm is also suitable for 2 min acceleration to 200 GeV, using a synchronous phase of 5º. Because of the low A 4.4 cm long iridium target has an optimum tar­ antiproton intensity, the rf system can have high im­ get efficiency of ηt = 1/3. With these numbers and pedance resulting in low power demands. taking Av = Ah and CISA/Mp B = 5, one gets

ΔΩ = 2 × 10-4 and 8 × 10-5 ΝΡ

573 References Luminosity 1. ISABELLE Design Study, in press. The expected luminosity for collisions has been calculated. It is limited by beam-beam inter­ 2. G.I. Budker, Sov. J. Atomic Energy, 22, 438 (1967). action and it has been assumed that the maximum per­ missible tune shift for the antiprotons is Δν = 0.005. 3. S. van der Meer, Report CERN/ISR-PO/72-31, (1972). For a 1.7 mrad horizontal crossing geometry with βv = 2.5 m and βh = 5 m at the intersection point, 4. G.K. O'Neill, Princeton-Pennsylvania Accelerator assuming a 10 A proton beam with an emittance of Report PPAD 415D (1961). 0.4 π 10-6 m.rad at 30 GeV, one gets 5. K. Hübner, Report CERN/ISR-TH73/-19 (1973).

28 -2 -1 L = 8 × 10 cm sec 6. W.F. Baker, et al., NAL-Pub-74/13-EXP (submitted to Phys. Rev. Lett.) (1974). Acknowledgments

The authors are greatly indebted to Dr. G.K. Green for valuable discussions.

Fig. 1.

Fig. 2.

574 DISCUSSION Andrei Kolomensky (Lebedev Institute): What is the main Chasman: At 30 GeV, where the yield reaches a plateau limitationfor increasing the luminosity: Is it a tune proc­ so that the intensity is really proportional to the momentum ess or something else? bite. Renate Chasman (BNL): It's the tune. The antiprotons Marshal King (Rutherford): The bypass presumably is will see a very dense proton beam and will experience the superconducting. Have you thought about the effects of ordinary beam-beam interaction. scattering downstream of the target, particularly in com­ parison with the dipole value? Kolomensky: So this figure 0.005 is the main limitation? Chasman: No, the entire bypass from the target down­ Chasman: Yes. stream will be at room temperature. James Allaby (CERN): At what energy are the antiprotons injected?

575