3.1 Production of 18F
In this section details of the production of 18F at RIKEN are described. The procedure is almost the same as that used in PET (Positron Emission Tomography) [23,24].
The threshold energy of the l80(p,n)l8F reaction is 25 MeV and the cross section has a sharp peak around 5 MeV. The energy of the proton beam suitable for the efficient production of l8F in a thick target is about 15 MeV. The thick target yield of the 18F upon bombardment of a 1 HA proton beam for lhour is 2.2 GBq (60 mCi).
Figure 2 shows the target container, which is installed at the end of the line of the AVF cyclotron in RIKEN. Made of titanium and of cylindrical form, it has a 15 mm diameter incidence window made of 40 pm thick Havar foil. Behind it is 3 mm thick silver plate. The 1.8 cm3 of 180 water (concentration 95 %, Rotem Industries) is transferred from a reservoir to the target container by means of a mechanical pump. When the container has filled with the l80-water, it is pressurized to 2 MPa with Ar gas to suppress boiling during the bombardment Boding of the target water could occur most probably near the middle of the Havar foil, blocking contact of the foil with the water and resulting in breakage of the foil due to overheating. The container is cooled by cooling water flowing through the back space and by He gas passing over the Havar foil. Weak focusing of the proton beam is preferable; the beam diameter is about 10 mm to prevent boiling. As the l8F produced is deposited on a graphite rod as described below, the size of the positron source does not depend on the proton beam diameter.
3.2 Spot 18F Source Electro-Deposition a Graphite Rod
The 18F produced solution is transferred for electro-deposition from the target container to a measurement hall some distance away through a 10 m long Teflon capillary, pushed by He gas pressurized to 15 atm. On the way, the 18F solution is temporarily stored in a reservoir on the capillary to separate and remove radioactive l3N2to a balloon. Hie amount of l3N2 (half-life lQmin.) produced via the 160(p,a)13N reaction will be large if the concentration of
water is low.
The electro-deposition and source-supply system is shown in figure 3 . It has two rotatable disks. One serves as a turntable which carries twelve electro-deposition cells, made of platinum. It is sufficient to set a single cell on the turntable and use it repeatedly, if the liquid is collected after each electro-deposition. The cell acts as a cathode as well as a vessel for the18F solution. A second disk is set perpendicularly to the first disk, carrying twelve graphite rod anodes of 5mm diameter on the side. When a cell is filled with the 18F solution, the turntable rotates to set it at the position for electro-deposition, directly below the source window of the slow positron beam apparatus. Then the second disk descends until the end of a graphite rod dips 0.2 mm below the surface of the solution. The position of the liquid surface is detected by monitoring the electric current through the graphite rod.
A charge carrier is not required for electro-deposition of activities higher than 2 GBq(50mCi) of l8F. The P+ emitted from 18F ionizes H20 molecules and provides
AVF Cyclotron H21!0(p,n)lsF
sufficient conductivity. For example, the deposition current for 5.6 GBq (150mCi) of l8F solution (in 1.8 cm3) with an applied voltage of 200V was 7mA, which was almost the same as that for the Nal9F (F 2ppm) solutioa The recovered H2I80 water without the carrier impurity is more easily reusable than that containing carrier.
The dependence of the collection yield on the electro-deposition time is shown in Figure 4. The data was obtained from the electro-deposition performed with l8F of activities 2.0-6.4GBq (55-173mCi). Electro-deposition for 20 minutes with an applied voltage of 200V is sufficient to collect nearly all the l8F on a graphite rod of 5mm diameter.
Following electro-deposition, the vertical disk is raised to the home position and rotated by 180°. It is then raised further to maneuver the graphite rod close to the titanium window of the slow positron beam apparatus. The whole procedure, controlled by a personal computer, takes 25 minutes including the time for the liquid transfer. The source is placed outside the beam apparatus and positrons are introduced into the chamber through a titanium window. The titanium window cuts the low energy positrons and increases the polarization of the beam at the expense of the beam intensity. As (3) indicates, N^^ for Ps-BEC is not greatly affected by this arrangement.
Half of the positrons emitted from the 18F come out of the graphite rod. This indicates that the deposited l8F atoms stay near the surface. The fraction of the back scattered positrons is expected to be low in such a light material as graphite (back scattering of positrons reduces the spin polarization of the beam). A low back scattering source on a graphite rod is thus a promising candidate for a spin polarized positron source, in spite of its relatively low P+end point energy.
Measurement of a few days duration is possible by changing the 18F charged graphite rods successively. Immediately after a batch of irradiated l?0 water is transferred and the electro-deposition starts, the next batch of 180-water can be loaded to the target container thus ensuring continuous production. As the cyclotron vault and the measurement hall are separated by some distance, the measurement is free from the radiation background of the proton bombardment
With a small cyclotron (commonly used for PET) producing a proton beam current of 50 pA, a positron source of l00GBq (3 Ci) and about 107 slow positrons per second can be obtained with a tungsten moderator. We expect that at least 2 TBq (60 Ci) of l8F can be electro-deposited on a 5mm diameter graphite rod This estimate is based on the observation that 97% of 2 GBq (60 mCi) l8F contained in 2 ppm Na19F solution was electro-deposited. The 19F atoms, the density of which is three orders of magnitude more than that of 2 GBq l8F atoms, must have been electro-deposited on the graphite rod with the same fraction as the l8F atoms .
The positron emitting radioisotope is a source suitable for the production of a spin polarized slow positron beam. Intense beams can be made by using short halflife RI sources in a beam apparatus linked to an accelerator or a reactor.
For Ps EEC, the electro-deposited ,8F source is a promising candidate. The reasons are that the production of high intensity 18F is relatively easy, almost all the l8F atoms are collected on a spot area of the graphite rod, and almost all the positrons emitted into forward hemisphere come out of the graphite rod. The relatively low helicity of the l8F is not necessarily a drawback since it is the product of the spin polarization and the intensity which is the crucial quantity in the realization of Ps BEC.
We are grateful for the support by the Nuclear Cross-over Research Funds.
 S. R. Swaminathan and D.M Schiader, Appl. Surf Sci. 116 (1997) 151
 J. Van House and P. W. Zitzewitz, Phys. Rev. A 29 (1984) 96
 J. Major, in Positron Beams and their applications, edited by P. Coleman, (World Scientific, Singapore, 2000), pp. 259-306
 D. W. Gidley, A. R. Koymen, and T. W. Capehart, Phys. Rev. Lett. 49 (1982) 1779
 P. M. Platzman and A. P. Mills, Jr., Phys. Rev. B 49 (1994) 454
 D. B. Cassidy and J. A. Golovchenko, in this volume
 R. J. Jaszczak, R. L. Macklin, and J. H. Gibbons, Phys. Rev. 181 (1969) 1428
 P. D. Ingalls, J. S. Schweizer, and B. D. Anderson, Phys. Rev. C, 13 (1976) 524
 G. J. F. Legge and I. F. Bubb, Nucl. Phys. 26 (1961) 616
 F. Bubb, J. M. Poate, and R. H. Spear, Nucl. Phys. 65 (1965) 655
 T. J. Ruth and A. P. Wolf, Radiochimica Acta, 26 (1979) 21
[ 13] T. S. Stein, W. E Kauppila and L. O. Roellig, Rev, Sci. Instrum., 45 (1974)951
 K. G. Lynn, M. Weber, L. O. Roellig, A. P. Mills, Jr. and A. R. Moodenbough, Atomic Physics with Positrons, eds. J. W. Humberston and E. A. G. Armour, NATO ASI Series B; Physics (Plenum Press, New York 1987), pp. 161-174
 R. Xie, M. Petkov, D. Becker, K. Canter, F. M. Jacobsen, K. G. Lynn, R. Mills and L. O. Roellig, Nucl. Instr. And Meth. B 93 (1994) 98
 M. Hirose, M. Washio and K. Takahashi, Appl. Surf Sci., 85 (1995) 111
 Y. Itoh, Z. L. Pen, K.H. Lee, M. Ishii, A. Goto, N. Nakanishi, M. Kase, and Y. Ito, Appl. Surf. Sci., 116 (1997) 68
 F. Saito, Y. Nagashima, T. Kurihara, I. Fujiwara, R. Iwata, N. Suzuki, Y. Itoh, A. Goto, and T. Hyodo, Nucl. Instr. And Meth. A 450 (2000) 491
 K. G. Lynn and F. M. Jacobsen, Hyperfine Interactions, 89 (1994) 19
 A. P. Mills, Jr., Proceeding of RIKEN Symposium on Development of Spin Polarized Slow Positron Beam and its Applications, (1996) 86
 P. Decrock, M. Huyse, G. Reusen, P. VanDuppen, D. Darquennes, T. Delbar and P. Lipnik, Nucl. Instr. And Meth. B 70 (1992) 182
 T. Kumita, M. Chiba, R. Hamatsu, M. Hrose, T. Hirose, H. Iijima, M. Irako, N. Kawasaki, Y. Kurihara, T. Matsumoto, H. Makabushi, T. Mori, Y. Takeuchi, M. Washino, and J. Yang, Appl. Surf. Sci., 116 (1997) 1
 R. Iwata, T. Ido, F. Brandy, T. Takahashi, and A. Ujiie, Appl. Radiat Isot. 38 (1996) 87
 M. Guillaume, A. Luxen, B. Nebeling, M. Argentini, J. C. Clark, and V. W. Pike, Appl. Radiat. Isot. 42 (1991) 749
 F. Saito, N. Suzuki, Y. Itoh, A. Goto, I. Fujiwara, T. Kurihara, R. Iwata, Y. Nagashima, and T. Hyodo, Rad. Phys. and Chem. 58 (2000) 755
This page intentionally left blank
Was this article helpful?