A STUDY OF THE HELIUM -JET RECOIL - TRANSPORT METHOD ' Thesis for the Degree of Ph. D. MTCHTGAN STATE UNIVERSITY KENNETH LEE KOSANKE 1973 1 M ”$3.15.- T" LIBRARY Michigan St; 2 . ' Universi Cy n‘L" v . r ' "- This is to certify that the thesis entitled A STUDY OF THE HELIUM-JET RECOIL-TRANSPORT METHOD presented by Kenneth Lee Kosanke has been accepted towards fulfillment of the requirements for Ph . D degree in Chemistry l- [up . Lime—(1:. . Major professor . Dateliqw [77 3 0-7639 ABSTRACT A STUDY OF THE HELIUM-JET RECOIL-TRANSPORT METHOD By Kenneth Lee Kosanke The Helium-Jet Recoil—transport (HeJRT) system has been studied for the purposes of discovering more of the detail of the mechanism of its operation and to facilitate its further development. The major portion of this thesis is a report of a series of experiments in which the operation of the HeJRT system was defined and in which the effects of varying the operating parameters of the system were recorded. This has provided an experimental basis that has allowed us to make statements on possible mechanisms of operation and the importance of certain features of the system. Of course, also included are a detailed description of our HeJRT system and sections discussing some of the extensions of HeJRT techniques that have resulted, such as those allowing the performance of on-line aqueous chemistry with the system. The initial motivations for the construction of a HeJRT system here was its intended use as a subsystem in an on-line mass identifi— cation system to be used in conjuction with conventional nuclear counting experiments. Our present thoughts on this application are included in the introductory chapter of this thesis. COO-1779-76 A STUDY OF THE HELIUMFJET RECOIL-TRANSPORT METHOD By Kenneth Lee Kosanke A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry Program in Physical Chemistry 1973 J3 a. IT M” Cyfb ACKNOWLEDGEMENTS I sincerely wish to thank Dr. Wm. C. McHarris for suggesting this region of study. His guidance, encouragement, and patience during the experimental work and preparation of this thesis are greatly appreciated. I also wish to thank Dr. W. H. Kelly of the Physics Department and Dr. F. Bernthal and Dr. R. Warner of the Chemistry Department for their help and advice. Their suggestions and advice during this project were very useful. Dr. H. G. Blosser and Mr. H. Hilbert, assisted with the operation of the Michigan State University Sector—Focused Cyclotron. Dr. J. Black, Mr. W. B. Chaffee, Mr. B. Jeltema, Dr. R. E. Doebler, Ms. C. Dors, Ms. J. Guile, Dr. R. E. Eppley, Mr. R. B. Firestone, Dr. R. Goles, Mr. C. Morgan, Dr. L. Samuelson, Mr. M. Slaughter and Dr. R. Todd all deserve special mention for their assistance and advice throughout the course of these experiments. Mr. R. Au, and the cyclotron computer staff have aided greatly in the data aquisition and evaluation through the use of the XDS Sigma 7 computer. Help has also been received from Mr. R. N. Mercer and his staff in the cyclotron machine shop, and from Mr. W. Harder and the cyclotron electronics shop. The cyclotron drafting staff, especially Mrs. M. Blosser, have been very helpful and quick in preparing the drawings for this thesis. Our secretaries Mrs. P. Warstler and Mrs. C. Vanneste have helped in typing this thesis. I thank the National Science Foundation, U. S. Atomic Energy Commission, and Michigan State University for their financial support without which this study would not be possible. Finally, I particularly wish to thank my wife, Bonnie, for her assistance in the operation of the cyclotron and in the collection of data. iv TABLE OF CONTENTS ACKNOWLEDGMENTS . . . . . . . . . LIST OF TABLES . . . . . . . . . LIST OF FIGURES . . . . . . . . . Chapter I. INTRODUCTION . . . . . . I. a. II. II. II. II. II. II. II. II. II. II. II. II. II. II. II. II. General . . . . . . . . Why Nuclides Far From 8 Studied . . . . . . . How Nuclides Far From 8 StUdiEd O O O O O O 0 Stability HELIUM JET RECOIL TRANSPORT SYSTEM n. 0. REFERENCES General . . . . . . . Should be Can be Calculation of Recoil Characteristics Target Assemblies . . Gas Supplies . . . Detector Assembly and Tape Transport . . Cluster Molecules . . Capillary Considerations . . . Total System Efficiency Determination . . Efficiency vs Target Assembly Pressure . . Efficiency vs Detector Assembly Pressure . Efficiency vs Time . Collecting Surface Considerations Aqueous Chemistry On-Line . . . Gas Phase Separations Plasma Chemistry . . Page iv vi vii 20 20 21 26 43 46 55 65 73 77 83 88 95 99 110 112 117 LIST OF TABLES Table Page 1. Reactions and Recoil Characteristics Expected Using the MSU Cyclotron . . . . . . . . . . . . 22 2. Total HeJRT System Efficiency . . . . . . . . . 75 vi Figure 10. ll. 12. 13. 14. 15. 16. 17. 18. LIST OF FIGURES Approximate breakdown of nuclides lying between the neutron drip line and proton stability according to half—life. This figure is taken from [BeI66a]. . . . . . . . . . . . . . . . . . . . . . Graphic display of the usefulness of various techniques for generating radionuclides from targets with Z=55. This figure is taken from [RuG67]. . . . . . . . . . . . . . . . . . . . . . . . Sketch and block drawing of a proposed on—line system employing an electric quadrupole for mass filtering. . . . . . . . . . . . . . Sketch of a proposed on-line system employing a recoil time-of- flight spectrometer for mass identification. . . . . . . . . . . Single target assembly of the HeJRT system . . . . . . . . . . . Gross count rate vs beam current compared for long and short target cylinders. . . . . . . . . . . . . . . . . . . . . . . . . . . Stainless steel to polyethylene capillary coupling sleeve . . . . Target Assembly components mounted on lid of target assembly chamber . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Target assembly chamber on cyclotron beam line. . . . . . . . . Original target assembly configuration. . . . . . . . . . . . . . Recoil collecting/Faraday Cup Hemisphere. . . . . . . . . . . . . Multiple target assembly of the HeJRT system. . . . . . . . . . . Picture of initial multiple target assembly using three targets . Multiple target/capillary assembly proposed for the HeJRT system. Target Assembly Gas Mixing Apparatus. . . . . . . . . . . . . . . Block diagram of HeJRT system set-up for a simple y experiment. . Detector assembly set-up in a search for B—delayed protons. . . Detector assembly set—up in a simple y singles run. . . . . . . vii Page 10 13 17 27 29 31 32 34 35 37 38 4O 41 44 47 48 49 Figure 19. 20. 21. 22. 23. 24. 25a. 25b. 26. 27. 28. 29. 30. Page Overall view of detector assembly area set up for a y singles experiment. . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Tape transport system. . . . . . . . . . . . . . . . . . . . . . 52 HeJRT pumping station on a platform located below the detector assembly area. . . . . . . . . . . . . . . . . . . . . . . . . . 54 Approximate angle of divergence of molecular clusters leaving the HeJRT capillary. . . . . . . . . . . . . . . . . . . . . . . . . 58 Relationship between beam current and relative transport efficiency 61 Relationship between benzene vapor concentration and relative transport efficiency. . . . . . . . . . . . . . . . . . . . . . . 63 Relationship between target assembly pressure and flow rate for 0.055 in. I.D. capillary. Also included are sweep times and an estimate of times in the capillary. . . . . . . . . . . . . . . . 68 Relationship between target assembly pressure and flow rate for 0.034 in. I.D. capillary. Also included are sweep times and an estimate of times in the capillary. . . . . . . . . . . . . . . . 69 Peak areas for transitions in 23Mg and 26Si as a function of target assembly pressure, when using a 0.055 in. capillary. . . . . . . 78 Peak areas for transitions in 23Mg and 2CSi as a function of target assembly pressure, when using a 0.034 in. I.D. capillary . . . . 79 Curves expressing the relationship between detector assembly pressure and total system efficiency for various capillary to tape angles and distances. . . . . . . . . . . . . . . . . . . . . . . . . . 84 An example of the plots recorded as the cyclotron beam was turned on and off in an attempt to discover the short term efficiency build-up in the first moments after the beam is turned on. . . . 89 Composite beam-on and beam-off curves, and the relative efficiency for the HeJRT system in the first few moments after beam is first turned on. . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Figure 31. 32. 33. 34. 35. 36. A comparison of the effectivness of paper tape and aluminized mylar tape as a collecting surface. . . . . . . . . . . . . . . . Sketch of the experimental set-up for our first attempt at perform- ing aqueous chemistry with the HeJRT system. . . . . . . . . . . Spectra recorded in our first successful attempt at performing aqueous chemistry with the HeJRT system. . . . . . . . . . . . . Sketch of the experimental set—up for on-line aqueous chemistry used in our third attempt at performing chemistry with the HeJRT system. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Portions of the spectra recorded using our experimental set-up for on-line aqueous chemistry and comparison with the results from previous runs. . . . . . . . . . . . . . . . . . . . . . . . . . Results supporting the possibility that chemical separations will be possible using "plasma chemistry." . . . . . . . . . . . . . . ix Page 96 100 104 106 108 115 CHAPTER I Introduction I. a. General There remain approximately 2000 unknown nuclides with half- lives in the range between 102 and 10-2 seconds [BeI66a]. This in- cludes all the remaining unknown 8 unstable nuclei, except some superallowed B decays which may have half-lives as short as 10-3 seconds, in addition to proton, neutron, and Y emitters. Figure 1 (taken from [RuG67]) is an approximate breakdown of stable or very long-lived (tl/Z >109 years) nuclides and unknown nuclides lying between the limits of the neutron drip line and proton stability. Further, there is an indication of expected half-lives. Clearly most of these remaining unknown nuclides are expected to have half- 1ives of less than one second and they will outnumber the nuclides presently known. The remainder of this introduction will be devoted to a short discussion of why and how these nuclides should be inves- tigated. l l l l l I l I C) n C 00 l 1 1 1 1 ‘8 1 1 CD CD CD CD C) m ~t m N v- o c? 8 8 I I SSSDXG uonnaN <——————- Figure 1. Approximate breakdown of nuclides lying between the neutron drip line and proton stability according to half—life. This figure is taken from [BeI66a]. 100 ——> Atomic number I. b. Why Nuclides far from 8 Stability Should be Studied B-Delayed Particle Emission Beta-delayed particle emission occurs when as a result of B- decay the nucleus is left in an excited state, one for which the binding energy for the last neutron, proton, or'a particle is negative and Coulomb barrier penetration in the case of protons or a particles competes with y deexcitation. Cases of B—delayed particle emission have been observed for each type particle. These decays are accessible to study using techniques designed for working with activities with half-lives greater than 10"2 seconds because the overall half—life for these processes is determined by the half—life of the B decay. Most of these cases found thus far are in the low-Z region where the energy available for decay and the average level spacings are large; however, now cases are being reported in the higher-Z region. Also found are regions where the binding energy for an a particle or the last proton or neutron is low, such as just above a closed shell. The number of unknown delayed particle emitters should be quite large [BeI66a][BeD65]. Through a careful mapping of these nuclides and consideration of their particle emission energies, further refine— ment of the parameters in the semiempirical mass formula should be possible. Also, as far as these parameters relate to nuclear structure: further refinement there may be possible. Particle Emission In addition to delayed particle emission, proton and a emission begin to compete successfully as modes of decay as one proceeds further from B stability. Many of the least energetic of these decays will have half-lives in the 102-10-2 second range because of the difficulty of Coulomb barrier penetration. A study of these nuclides should yield information on both nuclear masses from the energetics of the decays and on nuclear wave functions from half-lives, which reflect the ability of the emitted particle to penetrate the Coulomb barrier. New Doubly Magic Regions It should be possible to reach the doubly magic 28-28 region using proton reactions and the 50-50 region using heavy ion reactions. Several of the nuclides only one or two nucleons removed from 56Ni and most of the nuclides of the 1008n region remain unknown. A study of these nuclides one nucleon removed from doubly magic will of course involve a determination of their levels. Accordingly, these experi- mentally determined levels can then be used to extend calculations to the levels in nuclides further removed. Transition Region Nuclides and New Deformed Regions It is known in at least one case that practically spherical even- even nuclei may remain spherical as nucleon pairs are added until it very suddenly "collapses" with the addition of another nucleon pair to a shape of stable deformation. This is evidenced by the vibrational nuclear structure suddenly changing to a rotational one with the addi- tion of a nuclear pair. The example of this "collapse" or transition between essentially spherical and permanently deformed nuclei takes place in the beginning of the rare earth region between neutron numbers 88 and 90 [BeI66a]. (A second example of the sudden onset of deforma- tion may be occurring at about 135Nd [EkC72].) Nuclear states in the region of this sudden transition should not be amenable to either purely vibrational or rotational description. There have not been many of the nuclides of this type experimentally investigated and many more experimental studies are needed. Further, expanding decay studies to nuclides further from B stability will yield informa- tion on many more deformed nuclides including many in new deformed regions, such as the one starting in the extremely neutron deficient Xe- and Ba- isotopes [MoH65][ShR6l]. Access to Highly Excited States Because of the high energies associated with B decay far from the stability line, it will be possible to populate highly excited states in this decay. Beta decay will then complement nuclear reaction studies in examining these states. If one also considers the difficulty in obtaining suitable, i.e., stable or long—lived targets for nuclear reaction studies on nuclei far from B stability, it is probable that 8—decay studies will become the primary means of investigating these states. Nuclear Masses from Q—values The semi-empirical mass formula is in a sense a measure of a quantitative understanding of nuclei. Nested in the upwards of 40 parameters necessary to get accurate predictions are liquid drop model effects, shell model effects, and effects from other models. Accurate experimental masses measured in regions chosen to test specific parameters are thereby testing nuclear models and conceivably may lead to a more fundamental composite model. Many of the most interesting mass measurements will be made on nuclei far from B stability and can be made through determination of reaction Q values. Results of Interest in Astrophysics One of the more basic goals of astrophysics is that of developing models that will account for the relative abundances of elements in the universe. Two of the most important types of experimental informa— tion necessary to develop such a model are decay half—lives and cross- sections. Many of the more important half—lives, in particular those necessary to account for the build-up of elements above 209Bi are of nuclei lying far from B stability [BeI66a]. In addition, there are many specific bits of nuclear information missing about nuclear reactions, decay energies, branching ratios, etc., that are necessary to check and extend current astrophysical theories. Much of this needed information is about nuclei lying far from B stability. I. c. How Nuclides Far From 8 Stability Can be Studied There are several difficulties associated with the study of nuclides far from 8 stability which have definitely limited their study using standard nuclear techniques. Accordingly, any serious attempt at extending nuclear investigations to regions far from B stability will have to provide new or improved techniques to over- come these difficulties. The degree of success in overcoming these difficulties will determine the profitability of any attempt at studying nuclei far from B stability. The remainder of this introduc- tion will be devoted to a short discussion of the requirements these difficulties impose on any system to study nuclides far from B stability and how our proposed systems meet these requirements. A discussion of those techniques proposed or being implemented by other researchers studying nuclides far from B stability is beyond the scope of this thesis [MaR72b][AnG66]. The most obvious requirement of any system designed to study nuclides far from B stability is that it must provide a method of generating these nuclei in sizeable number. In a paper prepared for presentation at the International Symposium on Why And How Should We Investigate Nuclides Far Off The Stability Line [RuG67], Rudstam presents a comparison of several possible methods for producing nuclides far from B stability. The processes considered are medium energy (<50 MeV) charged particle (p,d,a) reactions like Qp,xn); high energy proton reactions like (p,2p).spallation (p,2pxn), and proton induced fission; , -‘ g‘fl— heavy ion induced reactions like (HI,xn); thermal neutron fission; and (fi,p) reactions with energetic (l4-MeV) neutrons. The comparison between these processes is made for target material between Z=25 to 75. While Rudstam's comparison is based on estimates of cross-sections and more or less typical beam intensities of different types of machines used to generate radionuclides, a presentation of a summary of his conclusions may be helpful. For the production of moderately neutron-deficient isotopes (510 neutrons deficient) throughout the range of Z=25 to Z=75, medium energy charged particles (p,d,a) should prove to be best by far. However, as one proceeds to extremely neutron deficient isotopes, heavy-ion and spallation reactions should compete well with medium—energy proton reactions for the best production rates, with heavy—ion reaction producing better yields than spallation. For the production of neutron-rich isotopes throughout the range Z=35 to Z=55, thermal fission should prove to be best by far. However, in the region below Z=35, high—energy proton and spallation reactions are favored and in the region above Z=55 high energy proton reactions like (p,2p) are favored. One of Rudstam's figures (that for Z=55) has been included as Fig. 2. Rudstam's figure of merit (FM) is defined by the relation d PM = ID I(x) 0(x) p dx, where I is the particle density, 0 is the reaction cross-section, p is the target density, and d is the thickness. Rudstam assumed the following beam intensities: 10 1m "1 1h 1m 15 1 1 ' 1 1 1 ' Z=55 12 1° _ (pxn) _ f ‘ 1- 1(ftl__ _ (HI,xn) 3 E '6 .‘3 :3 D E (mp) 102- o 1 1 1 1 1 1 1 1° -15 -10 -5 o 5 10 1‘5 Neutron excess Figure 2. Graphic display of the usefulness of various techniques for generating radionuclides from targets with Z=55. This figure 11 Med. energy p or d 100 uA High energy p ' l uA Heavy ions 6 X 1012 HI/sec Thermal n 1014 n/cmzsec High energy n (l4-MeV) 1010 n/cmzsec In those instances where experimental cross—sections were not avail— able, Rudstam used extrapolated or estimated values. A second requirement for the system is that it be capable of pro— viding the radionuclides in a form suitable for study in times approxi— mately equal to their half-lives. If we wish to consider 8 unstable nuclides out to the limits of the neutron drip line and proton stability, then it is necessary to consider times as short as 10.2 seconds (and possibly even as short as 10—3 sec for a few superallowed transitions). To the extent that it is desirable not to count in beam or to use pulsed beam techniques, it is necessary to transport the activities to low background areas. In addition, for charged particle spectro- scopy, it is necessary to accomplish a removal of the radionuclides from the target material in order to make a suitable source for count— ing, all in these short times. Further, in that daughter activities will be rapidly built up, it is necessary to provide for the continuous re— moval of old source material after being counted for some time interval. For each of the possible mechanisms for generating nuclides far from B stability (proton or heavy ion induced reactions or fission), a large number of different product nuclei will result and in most 12 instances the interfering products will greatly out—number the desired products. Thus, an additional requirement of the system is that it provide some means of isolating or at least enriching the desired product nuclei in order to limit interfering radiations. Ideally this would be accomplished using a combination of chemical and isotopic separations. The earlier discussion of Rudstam's paper suggests that it will be possible to produce nuclei far from B stability in good yields using proton beams with energies S50 MeV or 3He beams with energies 575 MeV (with currents 55 uA on target) from the Michigan State University Sector Focused Isochronous Cyclotron to initiate (p,xn) or (3He,xn) reactions. Table l (in Chapter II) provides a listing of some of the reactions that are possible using these beams from the MSU Cyclotron. In most regions of the nuclear chart it should be possible to reach from one to four nuclei beyond those presently reported and several nuclei beyond those for which fairly complete data are available. In addition, it is anticipated that the MSU Cyclotron will be capable of producing heavy ion (carbon, etc.) beams before too long. It may then be possible to extend even further into the extremely neutron-deficient regions. However, the yields for these reactions will be less than those with p or 3He beams. Block and schematic drawings of one of two on—line mass separation systems presently under consideration are shown in Figure 3. At this time the system and the operation of its component parts will only be mmjmfiJD—z NJUFEE FZUZn—Sow 022.2300 m.)— mo... qum maumxs. >I_m_2mwm< mqoaamnaac mo§om<1.1=ao owk<23160 Sketch and block drawing of a proposed on-line system employing an electric quadrupole for mass filtering. Figure 3. l4 discussed briefly. A more detailed discussion of its construction and operation will follow in later sections. In the Helium Jet Recoil Transport (HeJRT) portion of the system a collimated cyclotron beam enters the apparatus, striking a thin target (or a series of thin targets) in an atmosphere of helium. Those nuclei near the back of the target that interact with the beam will recoil out of the target into the helium atmosphere. Some of the characteristics of these nuclear recoils for a few representative targets and beams are listed in Table l. The nuclear recoils are slowed to thermal energies by collisions with the helium atoms. Re- cent work by R. D. Macfarlane [JuH7l] and in this lab suggest the next step is the attachment of the thermal recoils to large molecule clusters (with masses up to 108 amu). These cluster molecules are formed in the plasma generated by the cyclotron beam passing through the helium atmosphere from impurities added to the helium. The nuclear recoils are then removed from the target chamber by differentially pumping the helium containing the molecular clusters out through a polyethylene capillary. The capillary is run through concrete shielding to a low background area about 40 feet away. At the end of the capillary most of the helium will be removed using a one or two stage skimmer. Quite efficient skimming should be possible by just directing the flow coming from the capillary at a coincal orifice [MaR72] [NiJ70]. The molecular clusters with their enormous masses are extremely well collimated (see Section II. g. 15 "Capillary Considerations") and will be passed quite efficiently through the orifice, whereas the helium will have a much larger angle of divergence as it leaves the capillary and will therefore largely be pumped off. If our pumping capacity is not sufficient to reduce the pressure after the first skimmer stage to the low pressures required to operate the electric quadrupole, a second skimmer stage will be used. The molecular cluster will be broken up and the nuclear recoils ionized using a discharge set up in one of the skimmer stages, with the recoils directed into the ion optics associated with the electric quadrupole. With the quadrupole set up properly, only selective masses (efim) will be transmitted through it [PaW58]. At the exit of the quad- rupole the mass separated ions will be deflected either into an electron multiplier such that it will be possible to obtain a mass scan or onto paper or aluminized mylar tape for conventional nuclear counting. It appears this system will be capable of providing mass-separated nuclear reaction products, separated from the target and transported to a low background area, in very short times (2100 msec, Section II.g. "Capillary Considerations"), continuously.and with daughter activities removed. Thus, this system should meet all the requirements mentioned previously, with the exception of providing a chemical separation in addition to the mass separation provided by the electric quadrupole. It now seems that it may be possible to accomplish this chemical separa- tion at the time of attachment of the nuclear recoil to the molecular cluster during thermalization (see Section 11.0. ”Plasma Chemistry"). 16 A second system, and the one we will most probably construct first and patterned directly after R. D. Macfarlane's "MAGGIE" system [JuH7l], is sketched in Figure 4. At this time the system and the operation of its components will be discussed only briefly. A more detailed discussion of its construction and operation will follow in later sections. The first portion of this system as in the first system described is the Helium Jet Recoil Transport (HeJRT) subsystem; its operation was described above. Again, as in the first system described, the flow from the capillary is directed at a conical orifice (helium skimmer) to remove most of the helium from the flow and allowing most of the large molecular clusters with their nuclear recoil atoms attached to pass to the next stage of the system. The next stage of the system is a time-of—flight spectrometer. The molecular clusters with their nuclear recoils still attached are collected on a metal plate just after the helium skimming orifice. The clusters stick to the surface on impact with it. Macfarlane has found that if the prOper impurities were used to generate the molecular clusters and if the collecting surface is heated to about 200°C, the cluster molecule is broken up and driven off, leaving the nuclear recoil bound very weakly to the collecting surface. As the activities on the collecting surface 8 decay, the portion that decay with the proper geometry will recoil from the collecting surface. Some fraction of these 8 recoils will become ionized by shake—off and will be accelerated away from the collecting surface, using an accelerating grid. The 17 .cOHumonHuaovH mmma How uuuoEOHuuwmm uanHwIMOIoaHu HHoowu m wafimoamfim fioumhm mcwalao vmmoeoua m we :oumxm .c madman mmhmEOmHomam H103“. mo msE. mDIE .EBI 0E0 oz.._.nmmaoo mo mammp man so pmanenmnmu mnt mmnwnmam Hwooon Hmnuncn mnH Hag .mGOHmmn udmnonmmwuaonnswc mno> no .mumwnMn leoH .mmnwnmam awn: um muonuummn xx man nuns Hama mquEou on vmnummxm mnm mmaunnnma pmwnmso mo mmoa ozu wcfi>ao>cn maonuommn onn mo meow .nm>mzom moaunnnmm vmwnwno mo mmoa man wcw>ao>afi mCOfinommn sows3 On ucmuxm man mafianwumw on uaamnum on mp3 mnmnh .vMumHH maonnomon mnu now amnu noaamfim on wasozm mdoanommn mmmnu now mwamfim can no>w3om .wmumna muonuomon xx wnn mo some zufi3 mnmaaoo HHHB . COHUSMG HMCOHU inwwm cm use anon on wannmmom on Ame nfi .oanmn mane an pom: mmnmnmco ammo man now momma Hmnm>mm CH .mamoozH zm>nmm mo no naHmH cam mumnmo .%m>nmu mo moanmn mma mnmannmw am nBOGx no: onm3 mvsam> Q wnmnB mmma man mafia: vmananmuow mosam> G mo mammn man no woafianmumw onus muonnommn A§N.mmmv no Amx.mv oanmponm ng .mwme HUM—OHH HflHmO0.0 HUM—NH .O mm.0 Hmmma Cxwammmv mmm >®z Om. Halos Hsimoo.o Haiwo.o om.o nmmom Aeq.mmmv mam >mz mm ”slow Hoinoo.o Heiso.o nN.o amass Ass” v >mz ms .siom ESod ENo.0 sn.o amoom Ass av a >02 om nnmom omm 30.0 .36 EN 38 39.5% mmm >31 2 :uH—OON Sun—000.0 awn—«Nd O.H wmwm Aimammmv GEM >wz mm :5on 380.0 32.0 2.0 5f 3.0.3 as 2 H38 380.0 :1de one 513 ENS a >02 om 31mm 03 God 3.0 fm am? 35.21% mmm can on omm Rod and Ta mnom 2.1me V mm 3: mm own mnod and «To :2? CHM: m is 3 H39: H325.0 £85 3.0 as? Case a >21 om 53 oo: mmnd 34 «S SS :8.me mum >mz E 82 «and $5 Tm nmi $.me V mm 32 mm 0? god omé 04 23 2.1va m 32 we own Sod Rd in 22 case a >31 om was Amafizv Amflnzv ANEU\va A>mzv nonvonm ”ma maowunmm newnmh mHm Hag mHHoomm connomom connommm kwnmam annamm CH umwan :H mo hwnmcm :x Hog mafioomm mo mwcmm asefixmz HmnuHaH manmnonm Conuoaoho :mz oSu menu: venomaxm monumnnmuomnmzo Hwoowm paw maonuomwm .H mHamH 23 .mmaownnmm mmma Hmavm >Honmawxonaam ammaumn wcwnnsooo mcowmwaaoo wawmmoum nmmaosc wsn hp wonmomeEou nmznnsw ma mann mumwnMu onu cw mmwsmn urn wo ommu man cH .wcwmmoum nmmausa he pennanmuww wcwon mwcmn nwmnn wo cowunom wwan m On map Ewan snw3 wmnmwoommm monucwmnnmocs ownmfl %Hnmasowunmm m>m£ wuHSmmn omega .mmanmu man mcwumaommnuxo he vmcwmuno ones mowcmn mzn A:Em\>oz mNHo.QW E\mv E\m wo mosam> 30H zanmaaownnmm nom .mumwnmn man cw mmwcmn mzn :Hmuno on mHnmu man an wmnnoamn wwwma manamoum man com3umn meow mmB cownmaomnmuaw.oma< .wawmmouw owconuomfim aonw mp on mmoa hwnmcm wo monsom hnmeHnm can wwESmmm meownumnnou mnH .mHan msn aw ponnomon odouomw can wo mmma mzn aonw wnwnowwnu Hwoomn man wo mmme mnu now nasouum on meme onus mwnmn mnn 0n mcowuomnnoo .Honqozu wGHHHwnuw was wwwfiosunoz wo moanmu mwcmn mzu wawm: pmcwanonmw onm3 mmwamn Hwoomm S: T: 1 smaanuaoo .n mnnmn 24 high beam energies, low-Z targets, and very neutron—deficient regions. Parameters such as target thickness, target separation, and helium pressure used in setting up the target chamber for a run are based on calculations of the energy for the nuclear recoils. For several reasons these calculations were kept simple and contain many assumptions. The foremost reason was the large uncertainties assoc- iated with the range of very low energy, high mass particles, which makes precise calculation of recoil energy unrewarding. Second, in most instances it is possible to use values in setting up the experi— ment that are on the safe side of the calculated values by factors of 2 or 3 without seriously impairing the Operation of the HeJRT system. As a result of the medium energies of the particle beams generated by the cyclotron, the primary mechanism for the nuclear reactions in- volves compound nucleus formation. It is a simple matter to calculate the recoil energy of the compound nucleus by just considering linear momentum conservation (effects such as the energy transferred to the lattice holding the target nucleus are too small to be considered). The most extreme assumption made in calculating recoil energies is that there are no additional sources of momentum for the recoil as- sociated with the breakup of the compound nucleus. This is clearly not the case and further this source of momentum can be quite large. However, in that the breakup of the compound nucleus will be roughly isotropic, the gross features of the energy distribution of recoils leaving the target will be unchanged. The exception would be the 25 high energy tail corresponding to recoils generated at backside of the target that have had their recoil energies altered in the break- up of the compound nucleus. Maximum recoil ranges in the target and in He(STP) were deter— mined using maximum recoil energies. The ranges were obtained from the range tables of Northcliffe and Schilling [NoL70]. In many instances the maximum recoil energies fall below the lowest reported in these tables. For these cases the ranges reported were extrapolated off the end of the tables. It should be expected that the uncertainties associated with recoil ranges in the target determined by extrapola- tion are large. Within the limits of the tables, Northcliffe and Schilling determine their ranges considering energy losses occurring through both electronic and nuclear stopping mechanisms. However, in the case of very low energy recoils (recoil velocities <108 cm/sec [HaB60]) the energy loss mechanism is primarily through nuclear stopping. In the case of the recoils traveling through the target, the mass of the recoil and the mass of nuclei in the stopping medium are essen— tially equal and the recoil can lose up to half its energy in a single collision. Accordingly, the ranges for these recoils will vary con- siderably recoil to recoil and the calculated ranges can only be con- sidered an average effect. 26 II. c. Target Assemblies A sketch of a target assembly employing a single target is in- cluded as Figure 5. Here the cyclotron beam enters from the left through a Havar window (0.001 in. in practice) separating the vacuum of the beam line from the helium atmosphere (1 to 4 atm pressure, typically run at 3 atm). See Sections II. i. "Efficiency vs. Target Assembly Pressure" and 11. g. "Capillary Considerations" for discus- sions of the effect of He pressure on transport efficiency and on transport times, respectively. The volume between the Havar window and the target is sealed with a foam rubber gasket. This is to prevent those recoils from the window that are thermalized in this volume from entering the bulk helium supply and being swept out of the assembly along with recoils from the target. As a result of interaction with the cyclotron beam, product nuclei recoil from the back surface of the target into the helium. Through collisions with helium atoms the recoils lose their initially high kinetic energies and are reduced to thermal energies. The ap- proximate ranges for some typical recoils carrying a maximum energy are listed in Table 1. By pumping on the capillary, located immed- iately behind the target, a flow of helium down the target cylinder, mounted behind the target, and out the capillary is established. Re- coils in the helium are swept with the helium from the volume behind the target and out the capillary (capillary typically used has an inside diameter of 0.055 in. and is made of polyethylene). SINGLE TARGET ASSEMBLY O. D 0\ FARADAY __ HELIUM {amass} s s m § § 10 NE § :3 ES \: § Us \ \ .J —- \s =\ o-J \ \ 86 s s“\ . 0 § § 8:1 0: 352 \ \*0 S C> 1—14 Nih- § E32; 4 sea \a \ee :2 §E §§\ \§32%§3m§§\ F'L) “Y‘s m . '1 FOAMIRUBBER GASKET %//////////I//////////////// \\\\\\\\\\\\\\\A CYCLOTRON/ Thaw} BEAM {EtKlee‘} 28 A major time consuming step in the process of thermalizing re- coils and transporting them from the target area to low background counting areas is the time necessary to sweep the recoils from the thermalizing volume behind the target into the capillary. Accordingly, it is desirable to keep this thermalizing volume to a minimum by re- ducing the diameter of the target cylinder. To facilitate this, the cyclotron beam is collimated using a water cooled collimator before it is allowed to strike the target. Initially, the target cylinder was intended to help concentrate the most recently thermalized recoils in the flow of helium leaving through the capillary and keep them from diffusing throughout the chamber containing the target assembly. However, it was subsequently learned that it is necessary to attach the recoils to large molecular clusters (cluster of masses up to 108 amu [JuH71]) in order to achieve efficient transport of the recoils through the 40 ft of capillary leading to the low background counting area (see Section II. f. "Cluster Molecule§'[MaR72, WiK72, BoW72]). These molecular clusters are generated from impurities in the helium in the plasma provided by the cyclotron beam passing through the helium behind the target. It is presently intended that the target cylinder help concentrate the molecular clusters as well as the recoils in the flow of helium leav- ing through the capillary. Figure 6 displays the result of an experiment to test the effective- ness of the target cylinder in concentrating suitable molecular clusters 29 35 .rzmmmso 24mm oom_.ooo_.oo¢_.oom_.omol oom owe .ooq .oow .o l ‘1‘! nzmmmao 23m m> mnqm nzaoo Home 80% Elva .LNOOO SSOHO Gross count rate vs beam current compared for long and short Figure 6. target cylinders 30 in the helium flow leaving through the capillary. For this experiment a silicon (quartz) target was bombarded with 30-MeV protons to generate 28? (tl/z = 280 msec). Two different target cylinders were used: one was 1 cm long and the other was 10 cm long. Both cylinders were 2 cm in diameter. The maximum range for the 28F recoils,was less than 1/2 cm for 3 atm helium pressure. After considering the magnitude of the flow of helium leaving through the capillary (120 standard cm3/ second) and the maximum range for the 28P recoils it is apparent that a very small number of, if any, recoils could have diffused out of either target cylinder and not have been swept with the helium into the capillary. Relative efficiencies for the system when run using each of the two cylinders were determined by just comparing the gross Y count rates observed. Spectra were recorded to make certain that peak areas for the 1.78-MeV peak of 28P followed the gross y count rate. The improved efficiency for the system when run using the long cylinder is obvious from Figure 6. It is possible to attribute the improved efficiency of the target assembly when using the 10 cm cylinder to something other than concen- trating suitable molecular clusters in the thermalizing volume. How— ever, upon considering the increase in the relative efficiency of the system using the long cylinder as less beam current was run (at 1.6 uA the long cylinder system was 21% more efficient at transporting recoils than when the short cylinder was used, as compared with 75% more efficient at a 0.2 uA), it does seem likely that the improved transport 31 efficiency for the long cylinder system is related to the generation and/or concentration of suitable cluster molecules in the thermalizing volume. If the transport efficiency improvement were associated with an improvement in the flow characteristics along the long cylinder and into the capillary (which could significantly improve transport efficiency), it would be expected that the transport efficiency im- provement would not be beam current dependent. A more detailed dis- cussion of the shape of the curves presented in Figure 6 will be presented in Section II. f. "Cluster Molecules". The capillary presently being used has an inside diameter of 0.055 in. and, with the exception of a short stainless steel end that attaches to the target cylinder, is made of polyethylene (Clay-Adams Intermedic) tubing. The coupling between the stainless steel and polyethylene capillaries is made by first cutting the polyethylene and grinding the stainless steel tubing in special jigs so that they are cut off perpendicular to their axes. The two pieces of tubing are then brought together by inserting them in a tight fitting sleeve, see Figure 7. BRASS SLEEVE ‘) Figure 7. / ...... .. CAPILLARY CAPILLARY Stainless steel to polyethylene capil- lary coupling sleeve. “W11 ‘«--- -. v- 1 .. .V .3413»); .u - $39.. 5...... 2.. up“ . I ' . v .M9 L e m h c 1v... m m t e 8 r a t f o d i l n o d e t n u o m s t n e n m. m o c Y M m S S a t e 8 r a T D 8 e r u 8 .1 F 33 The requirements and flow considerations in the capillary are discussed in Section II. g-"Capillary Considerations". The beam is stopped and the current mointored by a water cooled Faraday cup located in the helium atmosphere. (While it is not neces- sary to have the Faraday cup in the helium atmosphere, it was conven- ient and a historical carryover from earlier target configurations (discussed later in this section). The target assembly drawn in Figure 5 is pictured in Figure 8. The components are mounted on the lid of the target assembly chamber, pictured in Figure 9, to allow easy access. The initial target configuration used in this lab was patterned after an early configuration of R. D. Macfarlane [MaR69]. Although this configuration has not been used here for two or three years, it does represent a satisfactory target assembly with possibly very short recoil collecting times and has been used recently by other groups [BoW72] [JuH7l]. The basic mechanism for generating and collecting recoils is the same as described above with the exception that recoils ejected from the target are thermalized between the target and a chrome plated aluminum hemisphere and are then swept into the capillary through an orifice in the rear of the hemisphere. The hemisphere serves double duty as recoil collector and as Faraday cup, see Figures 10 and 11. The main reason for changing to our present configuration was that it provided for an easy expansion to a multiple target assembly, which is not possible with the initial configuration. 34 . \ \.. :u,rAT ,1 , fidHHQHohu.wdmu:HH.amum Hmuwwwhamnam Sowaom. \\ .\ \ 1 Target assembly chamber on cyclotron beam line. Figure 9. 36 .The necessity for considering a target assembly using multiple . targets is born out of the short range of recoils in thin targets. Clearly, no higher yield of recoils is to be gained by increasing the target thickness beyond the range of the most energetic recoils created by the cyclotron beam. For most cases this maximum usable target thickness is from 10_5 to 10.4 cm (see Section II. b. "Recoil Charac— teristics"). Even though most experiments will be possible with the HeJRT system using a single target, cases are anticipated where the number of recoils generated from a single target will be a serious limitation. This will be especially true when the HeJRT system is feeding a mass separator. Also, in general it would be wise to have a surplus of recoils generated, making it possible to tolerate larger inefficiencies in other components of an experimental set up. Accord- ingly, it is desirable to consider multiple target assemblies for the HeJRT system. The simplest multiple target assembly considered is sketched in Figure 12. Here the helium exiting the target assembly first sweeps back and forth in front of each of the separate targets in sort of an S pattern. This is accomplished by simply staggering the targets and leaving small spaces, through which the helium can flow, on alternate sides of the target assembly. The thickness of the spacers between the targets must be adjusted to provide sufficient distance in order that recoils from the preceding target are thermalized within the distance to the next target. However, a major portion of the time Figure 11. Recoil collecting/Faraday Cup Hemisphere. 38 .Eonmem Hahwm men wo hwnammmm newnmu mamnnasz .NH unawwm hmxmdo mummnm 240.... .swr , :s.. s\\ Rama” m... W . ///////////////¢/ A \ 2050.196 a8\.v \ mnuomfi NR sass \. .. arias. .. >>Ooz_>> m<>1_m_2mmm< Hmoms. MTEEJD—z 39 elapsing between the generation of the recoils and having them ready for counting in the low background area is the time necessary to sweep the recoils into the capillary. (See Section II. g. "Capillary Con- siderations".) Thus, in an attempt to keep this time to a minimum the separation of the targets should not be any greater than necessary for the recoil to be thermalized before striking the next target. While multiple target assemblies have been used here for some time [GiG7l], a detailed study of their characteristics has not been conducted. It is known that yields from the multiple target assem- blies used are up from single target assemblies; however, the exact relationship between the number or type of targets used and their yield has not been determined. It is also necessary to determine if modifications of the assembly can further improve its efficiency. The initial multiple target assembly used [GiG7l] is shown in Figure 13. The one presently being used is essentially the same but looks more like the sketch in Figure 11. The average time necessary to sweep recoils into the capillary increases with each additional target because as each target is added the total volume from which recoils must be swept is increased. If the half—life of the activity desired is of the order of the time necessary to sweep the recoils from a single target into the capil- lary, then adding additional targets in the manner described above may produce little increase in the total yield. Accordingly, it may be necessary to add additional capillaries as well as targets. Figure 14 is a sketch of a multiple target assembly employing as many . Ilsef. . . 4 i. amy$s¢$3 . . > v S t e g r a t e e r h t g n m y 1 m S c S . a h. t . e . m. a t e l D. 1 t l u m 1 a 1 t 1 n i f O m u t. C .1 P Figure 13. MULTIPLE TARGET/CAPILLARY ASSEMBLY rm . :z 2: 1.13 :EgE sffi>a :mr““ RG T INDER T RCYL CAPILLARIES FOAM RUBBER GASKET /capillary assembly proposed for the HeJRT system. 42 capillaries as targets; however, if the half life restriction is not too severe it may only be necessary to use a capillary for every second or third target. In addition to the half life limiting the number of targets that can be swept with a single capillary there is a second limitation to the total number of targets used. This is a result of the energy degradation occurring in the targets and the helium, shifting the reaction products toward fewer neutrons boiling off. Of course, this restriction can be partially alleviated by using thinner targets, ones approaching the maximum recoil range. 43 II. d. Gas Supplies Initially the helium used in the HeJRT system has a purity 199.99SZ; however, this is typically doped with benzene vapor to a level of 10-20 ppm. This is accomplished using the gas mixing set up shown in Figure 15. The process is a two—step procedure. First, a supply of helium is made up containing about 1 ppt benzene vapor by adding a known amount of liquid benzene into the flow of helium to a storage tank pressurized to 2100 psig. Initially, a compressor was used (as in Figure 15) to pressurize the tank; however, because of the uncertainties about the amount of pump oil this was adding to the helium, the compressor has been replaced with a high pressure regulator on a tank of pure helium. The second step is to take the heavily doped helium from the storage tank and dilute it with pure helium as it is fed to the target assembly. This is accomplished by taking the pres- sure regulated flow from the supply tank through a needle valve and flow gauge and adding it to the pressure regulated flow of pure helium, which also passes through a flow gauge. The final benzene concentra- tion is determined and maintained using the flow gauge readings. For those runs in which the helium is doped with compressed air and water vapor, the compressed air is supplied by the compressor and is fed through the needle valve and flow gauge, while the water vapor is added by just passing the helium supplied to the detector assembly over or through a container of water. It was originally intended that the gas mixing set—up described above be used to add target gases (e.g., Ne for a 20Ne(p,n)20Na Figure 15 Target Assembly Gas Mixing Apparatus. 45 reaction) to the target assembly. This ability was lost when the necessity for adding controlled impurities became apparent. In order to regain the ability to add target gases in addition to impurities, the gas mixing apparatus is being expanded to allow the controlled addition of up to three gaseous components to the primary helium supply. An additional source of motivation for expanding the gas mixing set-up comes from the desire to try more exotic mixtures of impurities in the helium. This is necessitated by the desire to follow up the possibility of performing "Plasma Chemistry" (see Sec- tion II. 0. "Plasma Chemistry"). 46 II. e. Detector Assembly and Tape Transport The physical relation between the components of the HeJRT system and the detector assembly with its tape transport is shown in Figure 16. The capillary is run from the target assembly up into the detector assembly (a distance of :45 ft) where the flow of helium containing cluster-nuclear recoil combinations is directed at a collecting surface. The cluster molecules are accelerated as they are carried along in the helium flow to about the speed of sound in helium at room temperature (see Section II. g. "Capillary Considerations"). Because of their very large mass, up to above 108 amu [JuH7l], this corresponds to very high kinetic energies for the cluster, :1.1 MeV for 108 amu clusters. In turn, these high kinetic energies of the clusters cause them to stick well to collecting surfaces with which they collide. The most fre- quently used collecting surface is l—in. paper tape, which is fed through the detector assembly by a tape drive mechanism. Thus, it is possible to continuously expose detectors to fresh sources or to stop the system repeatedly for half life determinations, etc. The tape transport system is discussed later in this section. Figures 17 and 18 are pictures of two of the possible set-ups for the detector assembly. The first was a set-up used in the search for B—delayed protons. The activities are collected on an aluminum wheel that is being rotated by advancing paper tape in order to expose a fresh collecting surface continuously. A detector telescope has been set up to look for the protons. The second was a set—up used in a simple y singles experiment and is essentially the set-up used in the collection of the efficiency information in Sections II. h. .uamgummxo > madam m. How nnlumm 839nm Hahn: mo amuwmflo Moon .3 9.5mm 5.5ng 9.4 oz.o...m_zm 2.. _|. FL... 239$ . mass m .. mums. sod IAII ozq m3<> “.389. £02.58”. 24mm 5982 205396 595 m $.52 33.. n24 mo._.4n_n5m mszzwm + m: >._n_n_nw 233w: Figure 17, Detector assembly set-up in a search for B—delayed protons. 49 ....¢.....>.kbw~ . ass.mams .A 1. Audi. \- . - ,..s.awsa... mo. .8333. Detector assembly set-up in a simple Y singles run. Figure 18. 50 through II. 1. The activities are collected on paper tape just out of view of a collimated Ge(Li) detector located just outside the detector assembly chamber. The capillary to tape angle and distance has been adjusted so that the tip of the capillary (where activities sometimes collect) is out of View of the detector. In both figures the end of the capillary has been run through a slightly larger diameter metal tube. This has been done to hold the end of the capillary straight to facilitate aiming the capillary and to minimize the problem of activities collecting on the end of the capillary (see Section II. g. "Capillary Considerations").1 An overall view of the detector assembly area is shown in Figure 19. The tape transport system is shown in Figure 20. It is designed to handle up to l-in. tape and has been used with paper and aluminized mylar tape. The tape is advanced using a capstan-pinch roller assembly that is driven by a stepping motor. The accuracy to which the tape speed is known is the accuracy of the pulse generator (iIZ), assuming the tape does not slip or the stepping motor does not miss steps. This is a good assumption providing the pr0per tension is maintained on the tape and the tape speed is not excessive ($10 inches per second). The tape transport system has the ability to move the tape in either direction at speeds up to about 15 inches per second. The recent addition of a l6—channel programmable sequencer allows for quite complex motions of the tape and can also control other functions, such as gating detectors, routing spectra, and operating other me— chanical devices. 51 v s I . .V‘\:-?;_k\'-‘I .'.f.' f. .' .. a... adfiH assow> hansommm nououuonn .._._..._._._.._. . .. 1 . ..___._._._ e ......__............. .§ .5... . . 5...... . ...._ ... . ... ... Overall view of detector assembly area set up for a y singles experiment. Figure 19 . 52 h. o y .m . 1. 8 mm m... n d a“ a. .2 a a... me te .. t nu .Txu so. - c: _....... \. u r. .r\_JNw-.fiu.n “0....- ...w¢.:.u.<~ _. y .. Tape transport system. Figure 20. 53 The pumping station for the HeJRT system is located on a plat- form below the detector assembly area in order to reduce the vibra- tion of the detector assembly by the pumps. The pumping station is shown in Figure 21. The pumping system consists of an :300 £/sec. (measured capacity for helium) Roots blower backed by a 15 cfm mechani- cal pump. The vacuum line to the detector assembly is provided with a Chevron-style cold trap to prevent back streaming of oil from the pumps when solid state detectors are used in the detector assembly. Typically, only the 15 cfm mechanical pump is used for experiments using the HeJRT system, since the lower operating pressure offered by the Roots blower provides no increase in the total transport efficiency of the HeJRT system (see Section II. j. "Efficiency vs. Detector Assembly Pressure"). However, when the HeJRT system becomes the first stage of an on-line mass determination system, the added pumping capacity of the Roots blower will be necessary. When the HeJRT system is ex— panded to the on—line mass system, the additional pumps will also be placed on the HeJRT pumping station platform. Because of the pos- sibility of the pump exhaust containing activities, all pump exhausts are ducted to a filter and pumped out of the building. 54 umsofln muoom a). hilt. - u u H. 55.3.8.4...- p 5 v . ... . .r....\¢..... . . .\\\t\.. ... 5.... .L . rrC.:-:.n ::.h:.. .:.cu. .oo... o..\¢...€.. 34(5)? - a u . (it )up................ git... . \2..L.Cr.. .\ . . 3.2.5.. .. . ..; . may 38 nosed gunman mas _....2.:g 1.: :.:. .:.. 1.:1: .1. 1. 1 W11.: .. _ . i.;, 1:.:,:..1.,11,,;... . y.. 1 1 r11111: 1.1:1: :1: ::11:1;1: 1 :1d/1a__.. ..1 . . : 1;... 1.. I. .1...a.. 1.1.1:... Zyrtfihfitvrm;1....1111131,...1_..:..; 1 .. '11.... .. __.. ....~. . .._.. (..............._1....1 . :1... . ....,_..4:..:....;. .L..11|..Va:..4..1..:.... ... .. . . .1 HeJRT pumping station on a platform located below the detector assembly area. Figure 21. 55 II. f. Cluster Molecules Most simply stated, the role of cluster molecules in the HeJRT system is to aid in the efficient transport of recoil activities through the system and to help hold these activities on collecting surfaces. However, a good deal of detail of the mechanism is not known for certain. It can be argued the details of the mechanism are not important so long as the system works efficiently, which it does. However, we feel that when the HeJRT system is expanded to being part of an on—line mass identification system or when it is used in conjunction with aqueous chemistry and possibly plasma chemistry (see Sections II. m. "Aqueous Chemistry" and II. 0. "Plasma Chemistry") this detailed information will be at least helpful if not in fact necessary. Accordingly, the program to fill in these gaps is continuing. The possible presence of molecular cluster has been considered since Ghiorso [GhA59-69] observed that a fraction of recoils produced in heavy ion reactions on Pb targets apparently had masses as large as 103 amu. In 1962 Macfarlane reported results that indicated the clusters played a role in the collection of recoils when he observed that Tb recoils produced in heavy-ion reactions could be collected electrostatically as either positive or negative ions [MaR62]. Subsequently, Mikheev demonstrated that the efficiency of transport was markedly reduced when extremely pure (non—cluster gEnerating) helium was used [MiV67]. However, the importance of the large cluster molecules was not fully realized until the HeJRT systems grew in length to their present size (capillaries with lengths in the range of 20-200 ft). In these systems the total transport efficiencies were only a few percent, and not until specific impurities were added to the helium did the efficiencies climb back toward 100%. 56 In our system the total transport efficiency increases by a factor of about 20 upon changing from pure helium (purity 299.995%) to helium containing a few parts per million benzene vapor. In addition to the dramatic increase in efficiency observed for our system we also observe the buildup of solid material (cluster molecules) at each of the points where activities are deposited. In all respects the increase in efficiency for the system is tied to the presence of cluster molecules. Those factors that foster the generation of cluster molecules increase the efficiency of the system, and the activities generated in the system "tag along" with the cluster molecules. Thus, it seems apparent that the increase in efficiency is a result of the generation of cluster molecules and the attachment of recoiling activities to these molecular clusters. Macfarlane undertook a study of the masses of the cluster molecules and found them to range upward beyond 108 amu [JuH7l]. Once the importance of having the nuclear recoil attached to large cluster molecules was demonstrated, it was a relatively simple matter to explain why this should have been expected all along. Initially the explanation was that the large mass of the cluster molecules reduced the transverse diffusion rate in the capillary, thus reducing the chance that activity would be lost through collision with the capillary wall. Mac- farlane made the additional observation that the clusters were concentrated nearer the center of the capillary [JuH7l]. After some additional thought, the current concensus of people using HeJRT systems is that it is the focusing effect of laminar flow for massive objects transported in the flow that is the reason the cluster molecules drastically improve the system's efficiency. This focusing effect arises from an imbalance of forces on the large object accelerating in the flow as a result of it being acted on by laminae flowing at different speeds [CoW56]. The 57 direction of the imbalance is toward the laminae with the higher speed, i.e., toward the center of the capillary. For the HeJRT system as it is currently set up, the Reynolds number is just between Reynolds numbers indicating laminar flow and those indicating turbulent flow (see Section II. g. "Capillary Considerations"). However, even if the flow in the center of the capillary tended toward being turbulent, the forcing effect would still be present in the laminar sublayers along the walls of the capillary. An additional benefit resulting from the attachment of recoil activities to massive cluster molecules lies in the relative ease one has in separating clusters (and therefore activities) from the helium used to transport the cluster molecules. Owing to their enormous masses the clusters will diverge only slightly as they exit the capillary into a vacuum, whereas the helium will suffer a much greater divergence. Macfar— lane has been successful in achieving good separations of recoil—carrying clusters from helium using a conical oriface (helium skimmer) with a diameter only slightly larger than the capillary diameter and at a distance of several millimeters from the end of the capillary [JuH7l]. We made an attempt at determining the angle of divergence of the clusters in our system by merely observing the diameter of cluster build—ups at varying distances from the end of the capillary. For this determination the assumption was made that the clusters were not concentrated in the central regions of the capillary and that turbulence created by placing a collect- ing surface in the path of the flow did not affect the paths of the clusters. The result of these assumptions is the determination will yield a larger diameter than would be observed for a properly designed skimmer. The results of this determination are shown in Figure 22 which indicates an angle of divergence of 23° for about 90% of the clusters and an angle of 58 .hHMHHHmmo Hanna map wafi>mma muuumSHo Haasuwaoa mo mocmwum>fic mo mawam mumafixoumm< .NN musmfim I.QN - o .91. ..u .mlm 25 2% - Comem .8 .N m Wm. 23 0% u 3.0me dVOA'IOd "IIW 99 :10 ONE Ih\ \HI it —\ 7 :02. _ :ozi . AmmMFmDJU owoDoomn. 20mk040>8 mfioomm no wozwomeo mo m.52< 59 divergence of 27° for 99% of the clusters. In our case cluster molecules are generated from benzene vapor added to the HeJRT helium supply to a concentration of 10-20 ppm. Sugihara's group at Texas A & M also uses benzene vapor but at a much higher concen- tration. They bubble their helium through a container of benzene [BoW72] such that their benzene concentration is probably in the range of 0.1 to 1%. Macfarlane's group at Texas A & M initially used compressed air and water vapor. They added about 1% compressed air and had an open dish of water in their target assembly (private communication) (the important ingredient in the compressed air most probably was pump oil). More recently Macfarlane has been bubbling his helium through water and adding 2200 ppm CH4 [MaR72a] or using commercial grade helium and water vapor [WiK72]. A group at Simon Fraser University used ethylene in concentra- tions up to 100% as their cluster generating gas [DaH73a]; however, the mechanism for generating the clusters is quite different than we are now considering. Quite obviously the cluster molecules must be primarily composed of the elements making up the impurities added to the HeJRT helium supply. Thus, the clusters would seem to be organic in nature. In addition, the improvement they bring to the total transport efficiency would not seem to depend on their exact composition. One of the things that is not known at this time is the possible importance of trace impurities in the generation of molecular clusters in providing sites for attachment of thermal recoils to the clusters. In our case cluster growth is initiated by the passage of the cyclo— tron beam through the impurity—containing helium. Other researchers report the use of hard ultra-violet [WiK72] and gas discharges [WeC72] to initiate the growth of molecular clusters. Presumably cluster generation is a result of creating free radicals and/or carbonium ions and carbanions 60 within the plasma, formed by the beam passing through the helium from the organic impurities present. These highly reactive species then apparently react chemically to buildup the cluster molecules. Except for the necessity of the cyclotron beam (or some other source of energy input as discussed below), there is no direct information on the mechanism of cluster gener- ation. The suggestion that it involves the formation and subsequent reaction of highly reactive organic species is made only on the basis that it seems logical. The group of Simon Fraser suggests their clusters are generated upon the expansion of ethylene gas from pressures and temperatures above the critical point [DaJ73]. Studies have been conducted to determine the relationship between beam current on target (plasma density) and system performance and between benzene concentration and system performance. In each case the HeJRT system was set up using an aluminum target, the beam was 304MeV protons, a 45-foot long 0.055-in I.D. polyethylene capillary was used, and the activities were collected on moving paper tape such that the source material spent about 1.5 sec in view of the Ge(Li) detector. The detector assembly was pumped to ”7 torr, and the target assembly was maintained at 2 atm. For the series of determinations seeking the relationship between beam current and transport efficiency the helium was doped with :50 ppm benzene vapor. For the series of determinations seeking the relationship between benzene concentration and transport efficiency the beam current was held at very nearly 1 uA. Several spectra were recorded during each series of runs to be certain the gross Y count rate followed the intensity of the 820—keV peak of 2681. The results from the first series of runs are shown in Figure 23. It is quite apparent the amount of activity transported through the system is not linearly dependent on beam current. The production of recoils from GROSS GAMMA OOUNT RATE (COUNTS/see) RELATIVE EFFICIENCY 2000 I8OO .650 I400 IZOO I 000 800 600 400 200 I 00 |.00 0.80 0.60 0.40 0.20 0.I0 0.00 61 ICC 200 400 600 800 1000 I200 I400 I600 IBOO 2000 BEAM CURRENT (NANO AMPERES) I I I I I I I r f T I I l I ICC 200 400 600 800 |000 l200 I400 l600 |800 2000 BEAM CURRENT (NANO AMPERES) Relationship between beam current and relative transport efficiency. Figure 23. 62 the target must be directly proportional to beam current; therefore, it must be the efficiency with which the activity is transported through the system that is not constant. It seems most likely the drop in transport efficiency observed for low beam currents is the result of either not generating enough cluster molecules or not generating cluster molecules that are sufficiently massive. Presumably the reason for this is the drop in the number of free radicals, etc., generated because of the decreasing plasma density following the decreasing beam current. we anticipate performing an additional series of runs using a larger capillary diameter. As a result of using a larger capillary, the flow rate through the HeJRT system will be increased, thus allowing less time for suitable cluster molecules to be built up in the volume behind the target. If we are correct, we expect to see a more drastic dependence of transport efficiency on beam current. The results from the second series of runs are shown in Figure 24. It is quite apparent that the amount of activity transported through the system is independent of the concentration of benzene between the limits shown. It is felt that the slight drop in efficiency to 98% upon going to the lowest benzene concentration may be a result of poor statistics. This is because runs in which the level of benzene in the system was more than two orders of magnitude lower still had total systems efficiencies that were quite high. It takes a couple of weeks of pumping on the HeJRT system to return the system's efficiency to the very low level observed before benzene was first added to its helium supply. Runs were also attempted in which the concentration of benzene was almost to the level of saturation, with the result of no improvement in the system's efficiency. One of the more fundamental questions, and one of the unanswered ones, GROSS GAMMA COUNT RATE (COUNTS/sec) RELATIVE EFFICIENCY 63 2000 I800 I600 I400 I 200 I000 800 600 400 200 I 00 l 1 a L d d I.00 0.80 0.60 0.40 0.20 OJ 0 0.00 80 I20 I60 200 240 280 BENZENE CONCENTRATION (PPM) 320 3 60 400 A l J ‘ 20 40 80 |20 I60 200 240 280 BENZENE CONCENTRATION (PPM) 320 360 400 Relationship between benzene vapor concentration and relative transport efficiency. Figure 24. 64 about cluster molecules in the HeJRT system concerns the nature of the mechanism through which the nuclear recoil becomes bound to the cluster molecule. On the lowest level, it is possible to consider two mechanisms for this. One mechanism for the process is just one of entrapment, in which, as the bulk cluster material condenses out, the nuclear recoil becomes trapped during the process just because of its physical presence in the area. Here "condenses out" should not be taken literally but should be understood to be some form of nucleation or polymerization. A second mechanism for the process would be one in which the nuclear recoil becomes bound to the cluster at some time during the growth of the cluster. Here "bound" can be taken to be either reflecting a physical bond such as a result of van der waals forces or an actual chemical bond. Section II. 0. "Plasma Chemistry" presents some evidence supporting the mechanism involving the formation of a bond between recoils and chemically specific sites in the cluster molecule. As could be expected, considering their probable organic nature, the cluster molecules are soluble in acetone. We routinely use acetone to remove buildups of cluster molecules within the HeJRT system. However, it was not expected that we would find build up of cluster molecules to be soluble in water. There is also some evidence (see Section II. m. "Aqueous Chemistry") that there is a break-up of the cluster molecules when added to aqueous solution, with the break-up freeing the nuclear recoil occuring in quite short times. 65 II. g. Capillary Considerations Our initial experiments with the HeJRT system were run using glass capillaries that were==l meter long. With this length of capillary we were forced to run in an inconvenient location and could not sufficiently shield the detector assembly from the neutrons (and capture-yls) generated in the target assembly. When we were having difficulty obtaining longer glass capillaries and were not looking forward to splicing short ones together, an attempt was made to find a substitute capillary. A stainless steel capillary was to be ordered, but in the interim it was decided to try a polyethylene capillary supported by running it in larger glass capillary. When it was found this worked well we attempted a run in which the polyethylene capillary was strung loosely between the target and detector assemblies to determine whether the capillary could suffer slight bends without losing much transport efficiency. When the HeJRT system still performed well, a series of runs was undertaken to determine how much the capillary could be bent without a serious loss in efficiency. This series of runs culminated in the tying of the capillary in a loose square knot [KOK70] in which the bends were more severe than those expected in using the HeJRT system. The result again was no significant loss in activity transported, and as a result polyethylene capillaries were chosen for our HeJRT system. Our initial capillaries had an inside diameter of 0.022 and 0.034 in. when the total length of the system was only a few feet. Now we typically use 0.055-in. I.D. capillary that is 40-50 feet long. If in the future we are faced with difficulties in transporting very short lived products through the system, we expect to use 0.072-in. I.D. capillary. With the present diameter capillary we often use 28P(t1/2= 270 msec) to study the Operation of the system and have also successfully run 1+08c (t1/2= 182 msec). 66 (A disucssion of the transport times for the HeJRT system is included later in this section.) The capillaries are supplied in 100 foot lengths by Clay-Adams and are their "Intermedic Tubing." The 45-ft. length of the capillary allows us to set up the detector assembly in a conveniently located low background area over the cyclotron shielding wall. While we have run with capillaries longer than 100 feet, no systematic study of the operation of the system has been made as a function of capillary length. There is, of course, the limitation of increasing transport time for longer capillaries of the same diameter. In addition to experiments in which glass and polyethylene capillaries were used, we have also run using teflon capillaries and observed the same level of performance. In the past Ron Macfarlane at Texas A.& M has run using stainless steel capillaries [MaR7l]. It presently appears the operation of the HeJRT system is fairly independent of the material from.which the capillary is formed. Our preference for polyethylene capillary initially was based on its flexibility, but now it is just a historical carryover. We have observed that efficient Operation of the HeJRT system requires the inside bore of the capillary to be smooth and of more or less uniform cross section. Macfarlane also reports (private communication) that localized heating or physically vibrating the capillary can greatly reduce the system's transport efficiency. Another thing to be avoided is discontinuities in the capillary. At present, our system does have one such discontinuity at the point of connection of the polyethylene capillary to the short stainless steel capillary leaving the target assembly. The capillaries are joined as described in section II. c. It has proven to be a relatively simple technique and yet has not seriously affected the transport efficiency. T. T. Sugihara's group, also at Texas A & M, has eliminated this discontin- uity by eliminating the short piece of stainless steel capillary and is 67 running with teflon tubing entered directly into their target assembly [SuT72]. The helium flow rates for our system using a 50 ft. 0.055 in. I.D. polyethylene capillary range from 35.7 cm3/sec (STP) to 167.8 cm3/sec (STP) when the target assembly is pressurized from 1 atm to 3.5 atms, respectively. The relationship between target assembly pressure and flow rate is summarized in Figure 25a. This figure also includes the time necessary to sweep 1 standard cm3 of helium into the capillary and an estimate of the time necessary for activities to traverse the length of the capillary. The sweep time is just the reciprocal of the flow rate. The decrease in sweep time does correspond to a legitimate decrease in the time necessary to get activities into the capillary in that the nuclear recoils will be stopped in the same number of standard cm3. The estimate of the time spent traveling through the capillary is based on an equation presented by H. Dautet, et al. 2 + m 3m1 T: mlr ( 1) [DaH73], namely, where T = transit time in capillary, L = length of the capillary, a = the speed of sound in helium, and m1 = the mach number at the entrance of the capillary. This expression assumes choked flow in the capillary (i.e. the flow reaches sonic velocity at the exit), that the flow is for an ideal gas, and that the flow is an adiabatic process. These are all reasonably good assumptions for our system. Figure 25b presents the same information for our system using a 43—ft. 0.034-in. I.D. capillary. (It should be noted the time spent in the capillary claculated from the above equation is only about one half that obtained using the graphs H. Dautet, et al., have included in their paper.) The total transport time for a HeJRT system is just the sum of the time taken to thermalize the recoils and sweep them into the capillary 68 .humaaammo man ad mmaflu mo mumafiumm no wow mmafiu mmmzm mum movsaosfi omH< .%umHHHQmo .Q.H .sH mmo.o How mums 30am mam whammmum haeaommm ummHMu smmzumn magmsOHumHmm $53 mmDmmwmn. >I_m_>_mmm< wamdfi. oo.» 03 com on. 00.. ..00 :3 =00 . 4. .I :8 m w :8. n m :8. m, I. . a m .lU¢_Mw O :02 :8. db I. ‘- A: on u 1.52mi , 5.3.3540 0H 2. 000.0 “.0 mo_._.m_mw._.oumHHHomo one ca mmafiu mo mumsfiumm so was moafiu qmwam mum vmwdaocfi omH< .humaafiamo .Q.H .ofi wmo.o How many Boam mam musmmmun zanammmm umwumu somSumn afizmsowumamm 2:8 mmDmmmmn. >I_m_>_mmm< Pumas. nan omaw omaw on; b db c d d I 0 m A: M? u Ikozmd >m99.995% which was doped with approximately 250 PPM benzene vapor. The capillary was made of poly- ethylene, had an inside diameter of 0.055 in, and was just under 50 feet in length. The activities were collected on a stationary piece of paper mounted in the chamber pumped to <1000 um pressure. As in the case of the rabbit runs, the targets were bombarded for 30 and 60 sec for the copper and nickel foils, respectively, with a nearly constant current of 30-MeV protons. After the end of the bombardment, the collection of the activities on the paper was continued for an additional minute to assure complete sweeping of activities from the target area. Five minutes after the end of the bombardment a lS-min count of the 7 radiation was taken using the same detector in the same geometry. Also, as in the case of the rabbit run, the photopeak areas were corrected for integrated current and analyzer dead time. In both the rabbit and the HeJRT runs 0.0001-in. foils were used. However, for the efficiency determination the thickness of the target used in the HeJRT system was taken to be the maximum range of the recoils, i.e., the target thickness from.which recoils were able to leave the target and enter the helium atmosphere. These maximum ranges were calculated as outlined in Section II. b. ~"Recoil Characteristics." The efficiencies were determined to be 24% for 532m, 49% for 60Cu, and 60% for 51+mCo. However, it was subsequently learned that the efficiency of the HeJRT system increases as we proceed from 1 atm to 3 atm of helium pressure in the target assembly (see Section II. 1. "Efficiency vs Target Assembly Pressure"). While the next section discusses this increase 75 using Mg and Si, the increased efficiency was confirmed for the Zn, Cu, and Co recoils. Accordingly, the efficiencies for the HeJRT system operating with 3 atm helium pressure in the target assembly are approx- imately 50% for Zn, 75% for Cu, and 90% for Co. Table II. Total HeJRT System Efficiency Efficiengy Isotope (1 atm) (3 atm) 63Zn 24% 50% 6°Cu 49% 75% 5“ch 60% 90% Even allowing for uncertainties of greater than 20% in the measured efficiencies relative to one another, it is clear that the HeJRT system exhibits varying efficiencies for different activities. In that the recoil represents such a small part (in the ppm range) of the cluster with which it is associated, it seems rather doubtful that these varying efficiencies are reflective of differing probabilities for the cluster-plus-recoil to stick to the tape. Similarly, it seems doubtful that these varying efficiencies are reflections of differing probabilities for the clusters- plus recoil to be transported through the system. Thus, as the final alternative it seems likely that these varying efficiencies are reflections of differing abilities of the recoil to form a "bond" with the cluster. [Additional evidence and a discussion of the recoil—cluster bond can be found in Sections II. f. "Cluster Molecules," II. k. "Efficiency vs Time," II. m. "Aqueous Chemistry," and II. 0. "Plasma Chemistry." If 76 the recoil has a difficult time attaching to a cluster or if the attachment is weak possibly resulting in a separation of recoil and cluster as they move through the system, it is likely that some number of the recoils will be traveling through the capillary without being attached to a cluster. Accordingly, these unattached recoils will not be assisted through the system by the clusters and will have an efficiency corres- ponding to the efficiency of the HeJRT system using pure He, 35%. 77 II. 1. Efficiency vs Target Assembly Pressure In the initial experiments when only non-doped helium was used (and the total HeJRT system efficiency was <5% as a result), the efficiency was essentially independent of the pressure in the target assembly. Through the pressure range from 20-in. vacuum to 4—atm there was no obvious or systematic dependence of efficiency on the target assembly pressure. However, when helium doped with small admixtures of benzene is used in the HeJRT system, the effect of target assembly pressure on the overall efficiency is obvious; see Figures 26 and 27. Each point in the figures is the average of three determinations made at that pressure. The HeJRT system was run using the single target assembly described in Section II. c. "Detector Assemblies" using approximately 40-ft poly- ethylene capillaries with 0.055-in. and 0.034-in. inside diameters for Fig 26 and 27, respectively. The helium was doped with benzene to a concentration of about 20 ppm. The experiment was run using as nearly as possible 1.5 uA of 30—MeV protons on an aluminum target foil, initiating for the most part the 27Al(p,2n)2631 and the 27Al(p,om)23Mg reactions. The activities were collected on paper tape moving at 0.5 in./sec past a 4.6% relative efficiency Ge(Li) detector. The activities were deposited on the tape at the edge of the detector's field of view with the capillary at a distance of l in. from the tape and at a 45° angle such that the tip of the capillary was out of View of the detector. The geometry was such that the activities spent 3 to 4 sec in view of the detector. Figure 26 shows the results of runs using a 0.055-in. polyethylene capillary. Peak areas for the 820- and 439-keV transitions in 2631 and 23Mg, respectively, are plotted against target assembly pressures from 0 to 35 PSI gauge pressure. Peak efficiency is reached at roughly 3 atm - 73 0.055 inch Polyethylene Capillary? 3000‘ 439-keV Peak of 23M9 f 7000- 0 I — o 3000 I I l l J l 5’ f> 2400— 820-keV Peak of 2GSI PEAK AREAS |4QO l l 1 l J l 0 5 IO IS 20 25 30 35 TARGET ASSEMBLY PRESSURE (P.S.I. Gauge) Figure 26. Peak areas for transitions in 23Mg and 2681 as a function of target assembly pressure, when using a 0.055 in. capillary. : 6000 PEAK AREAS 79 0.034 Inch Polyethylene Caplllary ) 439-kev Peak of 23Mg 5000 4000 3000 2000— 7 l J l l I l 3" BZO-keV Peak of 263i f l600- '- I400 I200 I000 800 - " l l l l I l 0 5 IO I5 20 25 30 35 TARGET ASSEMBLY PRESSURE (P.S.I. Gauge) Figure 27. Peak areas for transitions in 23Mg and 2CSi as a function of target assembly pressure, when using a 0.034 in. I.D. capillarv 80 and is about 50% greater than at 1 atm. This increase in peak areas with increasing target assembly pressures has to be considered a legitimate efficiency increase and not just the reflection of a half-life effect. [The average transit time for recoils through the HeJRT system decreases for increasing target assembly pressure, see Section II. g. "Capillary Considerations."] If the increase in peak areas with increasing target assembly pressures were a half-life effect, it would be expected that the increase for the 2.1—sec 2681 activity would be much more dramatic than that for the 12.1-sec 23Mg activity. In that the 2381 and 23Mg activities were made simultaneously from the same target, it seems unlikely that this peak area increase could be attributed to anything other than an increase in system efficiency. Further, in that the final cluster-plus-recoil velocity in the capillary is not affected by the target assembly pressure, in the range 0 to 35 PSI gauge pressure under the vacuum pumping conditions used (see Sections II. g. "Capillary Considerations" and II. g. "Effic- iency vs Detector Assembly Pressure), it seems unlikely that the increase in system efficiency is a result of more of the activity sticking to the collecting surface. Additional evidence supporting this is presented in ' where a comparison is Section II. c. "Collecting Surface Considerations' made with a collecting surface with poorer sticking ability. Accordingly, the increase in system efficiency must reflect either a decrease in the number of clusters-plus—recoils lost in their flow through the capillary or an improvement in the conditions fostering the generation of clusters or fostering the attachment of recoils to cluster, possibly a result of increasing plasma density as target assembly pressure in increased. Figure 27 shows the results of runs using a 0.034-in. I.D. polyethy— lene capillary. There is some disagreement between the 2681 and 23Mg 81 curves as to where the maximum efficiency is reached and also as to the degree of the improvement at 3 atm over the efficiency at 1 atm target assembly pressure. The 2681 curve reaches a maximum at roughly 2 atm and is up to about 50% over the efficiency at 1 atm, while the 23Mg curve reaches a maximum at roughly 3 atm and is up to about 100% over the efficiency at 1 atm. It seems likely that this disagreement is a result of poor statistics. If it were the reflection of a half-life effect, the greater improvement in the efficiency would be observed in the shorter 2.1-sec 2681 curve and not in the 12.1-sec 23Mg curve. There is a rather striking difference between the two sets of curves; the runs made using the smaller 0.034-in. I.D. capillary have a much more rapid increase in efficiency in the lower portion of the curves than for the run made using the 0.055-in. capillary. In the discussion of the runs made using the 0.055-in. I.D. capillary it was concluded that the increase in efficiency with increasing target assembly pressure resulted either because of a decrease in the number of clusters-plus-recoils lost in their flow through the capillary or an improvement in the conditions fostering the generation of clusters of the attachment of recoils to clusters. In that the increase in efficiency with increasing pressure is different for the two capillary diameters used, it seems unlikely that the increased efficiency results solely from an improvement in the conditions under which the clusters are formed or under which the recoil is attaching to the cluster. Additional evidence for this comes from the results reported in section II. f. "Molecular Clusters." Here the relative efficiency of the HeJRT system as a function of beam current is discussed. To the first order, plasma density behind the target will be directly proportional to both beam current and helium pressure. If the curves in figures 26 and 27 are corrected for the expected increase in system efficiency as a 82 result of increasing plasma density the major portion of the increase remains. Thus, the improvement in efficiency for the HeJRT system with increasing target assembly pressure must come primarily from an improve- ment in the flow of clusters plus recoils through the capillary. A possible explanation of this effect could lie in the reduction of the mean free path of the helium in the capillary as the pressure in the target assembly, and thus throughout the system, is raised. As the mean free path is reduced, the number of collisions occuring between the recoil-plus-cluster combination and Helium atoms is increased, which will tend to help concentrate the clusters in the center of the capillary. For a discussion of the focusing effects of laminar flow, see reference [Cow 56] and section II. f. "Molecular Clusters." For each of the curves in Figures 26 and 27 it appears the curves are starting to turn downward. Ones first thought might be that this corresponded to the break down of laminar flow occuring at higher pressures in the target assembly. In section II. g. "Capillary Consideration" the Reynolds numbers of the flow were determined and while it was found that this could possibly be the case for the 0.055—in. capillary it most probably was not the case for the 0.034-in. capillary. Accordingly, it is felt that this effect is not a result of the breakdown in laminar flow in the capillary. 83 II. j. Efficiency vs Detector Assembly Pressure In an attempt to determine the relationship between the total system efficiency of the HeJRT system and the pressure at which the detector assembly is held, a series of runs was made in which the system efficiency for collecting paper tape was monitored while the capillary to tape dis- tance and angle were varied along with the pressure in the detector assembly. It was hoped that the drop in total system efficiency as the pressure in the detector assembly was raised could be identified as to the extent of it being a result of a drop in the efficiency for transporting activities through the capillary or a result of a drop in the efficiency for collect- ing activities on the collecting surface. It is expected the efficiencies associated with the generation of activities and cluster molecules and the attachment of activities to clusters would not be affected by varying the pressure in the detector assembly. For these runs the helium was doped with :10 ppm benzene vapor and the pressure across the HeJRT system was held constant at 2 atm. The capillary used was a 0.055-in. J.O. polyethylene capillary. The tape used to collect the activities transported through the system was non— oiled paper tape (computer perforation tape) and was advanced at the rate of 0.25 in.per sec. In order to eliminate any long—term effects from the cyclotron beam current varying, 28F with its short half-life of 280 msec was chosen as the activity to monitor. The 28P was generated using a 354MeV 0.7-uA beam of protons on a 0.05—in. quartz target to initiate a 28Si(p,n)28P reaction. A Ge(Li) detector with a 4.6% relative efficiency was used to monitor the Y count rate. The results were corrected for any small fluctuations in the beam current and were plotted against the pressure in the detector assembly (see Figure 28). The pressure in the .84 4000 vs" AT 90° ‘ 3000 . U I 2000 E . E 0 I000 ('3 I O y : i : 1 4000 W ' I/s“ AT 45° 3000» . " a? 2 2000 _ a E . 9.. I E w E 8 I; I000 ° 0 . N a: cu - a - - ’2' 1 Y : : D 0 II 0 I AT 90° <1 2 . z 4 (9 O I . E E P . - I" AT 45° 3000 - . 2000 b 5:; - a? g . I000- g .0 co - - - . . - O y g l L l 4 0 IOO 200 300 400 DETECTOR ASSEMBLY PRESSURE .(mm Hg) Figure 28. Curves expressing the relationship between detector assembly pressure and total system efficiency for“Various capillary to tape angles and distances. 85 detector assembly was varied by changing pumps and leaking air into the detector assembly. In each of the cases, capillary to tape distance and angle, the initial system efficiency (P l torr) was essentially the same. [See Section II. L. "Collecting Surface Considerations" for a discussion of the initial efficiencies for capillary to tape distance of 1/8 and l in. and angles of 90° and 45°]. However, the pressure at which a reduction in the total system efficiency was first noticed was different for each of the four set-ups. A detector assembly pressure (Pa) of 230 torr was necessary to affect the efficiency in the l/8"-90° set-up, Pa 220 torr for the l/8"-45° set-up, Pa 28 torr for the 1"-90° set-up, and Pa only 26 torr to affect the efficiency of the l"-45° set-up. In each of the cases, as the pressure was further increased the observed count rates fell to a constant minimum value, which corresponded to none of the activity collecting on the tape and was just the result of decay occuring as the helium containing the activities was being pumped from the detector assembly. This was confirmed by the ovservation of an increase in the count rates when the vacuum pump was turned Off (thus allowing a build up of helium containing activities in the detector chamber) followed by an appropriate reduction when the pump was turned on again. If the observed drops in total system efficiency with increasing detector assembly pressure are the result of drops in the collecting efficiency as a consequence of collecting through a more dense atmOSphere, it would be expected that the drops in efficiency would occur at lower pressures and be more severe as the capillary to tape distance was increased, allowing a greater opportunity for the beam of activities to interact with the atmosphere. Similarly, it would be expected that the drops in 86 efficiency would occur at lower pressures as the capillary to tape angle was made more oblique, allowing a greater opportunity for the beam of activities to be deflected from the collecting surface. Both effects are seen in the results reported in Figure 28. If the observed drops in total system efficiency with increasing detector assembly pressure were the result of drops in the transport efficiency as a consequence of changing flow considerations in the capillary, it would be expected that the efficiency would hqld constant until the pressure in the detector assembly was raised beyond the point at which flow of Mach 1 could no longer be achieved, at which time the efficiency would start to drop off and continue to drop off with increasing pressure. Note, the pressure experienced at the exit of the capillary at Mach 1 is no longer achieved in the helium flow will be independent of capillary to tape distance and angle. From H. Dautet's paper [DaH73] this should occur at detector assembly pressures of 235 torr for our system operating as described above. While in each case the total system efficiency did seem to hold constant before starting its drop, the pressure at which the drop started and the severity of the drop was very much dependent on the capillary to tape distance and angle. It is of course necessary to consider that the pressure recorded for the detector assembly is then the pressure experienced at the exit of the capillary. However, it would be expected this error would be largest for the l"-90° case, and this is the case for which the pressure corresponding to the initial drop in efficiency is the highest. Accordingly, it seems the primary factor acting in the drop in total system efficiency with increasing detector assembly pressure is a drop in collecting efficiency. This conclusion is reinforced upon considering the total system efficiencies of 50-75% reported in Section II. m. "Aqueous Chemistry" for collecting 87 activities in aqueous solution under atmospheric pressure. From the results discussed above it is quite apparent that the require- ments of the pumping system for a HeJRT system used alone are minimal. It is only necessary to maintain pressures of 20-30 torr in the low pressure end of the system without loss of efficiency. Only when the HeJRT is used to supply activities to equipment requiring low pressures is it necessary to consider large pumping capacities and helium skimming. 88 II. k. Efficiency vs Time When considering the total HeJRT system efficiency vs time, it is necessary to consider two time scales. The first, measured in seconds, is the short-term efficiency building up when the cyclotron beam is first turned on. This is related to effects such as the time required to build up suitable cluster molecules. The second, measured in hours, is the long- term efficiency falling off as the system is used for long runs. This is related to effects such as cluster molecules collecting in the capillary. Occasionally it seems as if there is a third time scale, measured in minutes, over which the efficiency tends to build up after the cyclotron beam is first turned on. However, this time vs efficiency effect has not been documented. It is felt that this effect may be the result of moisture or oil leaving the target surface when it heats up from the beam striking it. A series of runs was made to characterize the short-term efficiency build-up when the cyclotron beam is first turned on. In these runs 280- msec 28P was the primary activity monitored. The 28F was generated using 30-MeV protons on a 0.05-in. quartz target through the 288i(p,n)28P reaction. The helium was doped with :50 ppm benzene vapor and the pressure of the target assembly was maintained at 3 atm. The pressure in the detector assembly was maintained at 21 torr. The capillary was used 0.055-in. poly- ethylene. The activities were collected On paper tape which was advanced only between runs to supply a fresh surface for the next run. Also, between each run the target assembly was pumped out several times to remove cluster molecules remaining from the previous run. Each run consisted of making an x-y plot of gross y count rate time (using a time base) as the cyclotron beam was turned on and Off (see Figure 29). During the cyclotron on time the beam current was held as nearly constant as possible 89 I l l T I u , I I ' ' C) v N ’ C) ‘N N _ j 2* N st .. 8° 0 _ “Q g D A ‘00 g: LI: 98 (9 g g e " f I- ~82 (z) —0 2 <5: ' E s < -°w _ LIJ —m - U. m g o m tr - -—.._.r (r M 8w : _ >_ 4 Q _ 0 m . z u, 0‘” ~ EL" 2 ‘mg ‘ $2 I- P- U. h .— ' LL E -7 LL] 9 .8 ‘ ~ . I— m ** ~ 0: . .» O) . =~ I 9; <1- ——-"?E,""‘ 5. - qO N l 1 l L 1 A 1- 1 ,. lo 9 m 00 N co :0 <1- ro N — ($1.an AHVHlIBHV) 3mm an00 Figure 29. An example of the plots recorded as the cyclotron beam was turned on and off in an attempt to discover the short term efficiency build-up in the first moments after the beam is turned on. 90 usually 10% and data from only the most constant runs were saved. The beam was turned on and off by raising and lowering a scintillator in the beam line near the cyclotron. The beam passing through the scintillator diverged and was of the wrong energy to be transported through the beam optics. This method of turning on and off the beam was chosen because it was faster than the vacuum valves in the beam line (about 0.2 sec as opposed to more than 1.0 sec for the valve). The count rate signal was taken from an ORTEC rate meter (#441) set for 10% standard deviation (shortest time constant). The output was plotted using a relatively fast Esterline Angus x-y recorder with a maximum slewing speed of 55 in./sec. By comparing the curve recorded as the beam was turned on with the inverted curve recorded as the beam was turned off it should be possible to ovserve the build up of the total HeJRT system efficiency in the early moments after the cyclotron beam is first turned on. This should be possible because the time constants of the rate meter and the x-y recorder will be the same for increasing or decreasing count rates and the growth and decay curves for the activities generated will just be inversions of one another. Accordingly, the present difference in these two curves will correspond to the efficiency of the system. Figure 30 displays composite "beam—0n" and "beam-off" curves taken from a collection of count rate curves like the one in Figure 29. Also shown in Figure 30 is the short term efficiency curve generated from the percent difference between the "beam-0n" and "beam-off" curves. It can be seen that the efficiency rises very rapidly to about 70%, then rises rather slowly, taking almost a minute to reach full efficiency. For these runs the standard target cylinder (see Section II. c. "Target Assemblies"), which is about 4-cm long, had been replaced with one lO—cm long, such that it came to within 2 mm of the Faraday cup. This was 91 HeJRT EFFICIENCY PROFILE (FOR SHORT TIMES AFTER BEAM 0N TARGET) )- I I I I I I 0 ZIOO 9.1 2 90 III 80 “J 70 I; 60 - i E 50 __ EFFICIENCY CURVE GENERATED .. d 40 _ FROM CURVES BELOW .. 0: 30’ ‘ ZO- " IO' "‘ We IZ’ “ 0‘) II" " t IO - ‘ :2) 9 L— BEAM ON CURVE . E:- B BEAM OFF CURVE q 3‘: 7 . t on 6 " ‘5 5 ' LU 4 TRACINGS FROM A COLLECTION OF _‘ E 3 BEAM ON AND BEAM OFF CURVES - l— 2" ‘ Z 2) IF " O o I I l l l l O . IO 2O 3O 4O 50 BO TIME AFTER BEAM 0N (SECONDS) Figure 30. Composit beam-on and beam—off curves, and the relative efficiency for the HeJRT system in the first few moments after beam is first turned on. 92 done to insure that all molecular clusters generated in the volume between the target and the Faraday cup would immediately proceed toward entering the flow through the capillary and would not diffuse out into the bulk helium in the target assembly. This was important because if many of the clusters did enter the bulk helium supply the efficiency of the HeJRT system would not stabilize until the concentration of clusters in the bulk helium supply equilibriated. With the flow rate of 7500 standard cc~min, the rate at which the target cylinder was swept was about once every 1—1/2 sec. Accordingly, any increase in efficiency due to the generation of cluster molecules should be manifested within times of a few seconds (allowing for some mixing occuring between the helium in the cylinder and fresh helium entering the cylinder). Thus, it seems likely that the initial rapid rise in efficiency during the first seconds corresponds to the build up of the cluster molecules necessary for the efficient transport of activities through the HeJRT system. The longer component of the efficiency rise is not so easily explainable. It could be explained if cluster molecules were diffusing out into the bulk helium supply, but this seems extremely unlikely in view of the large helium flow rate and the small separation between the target cylinder and the Faraday cup through which the clusters could diffuse. This possibility will be checked after the construction Of a new target assembly is completed, one in which the helium is fed directly into the target cylinder and which has no place for clusters to diffuse to. If the longer component of the efficiency rise is still present with this new target cylinder, other possibilities such as adsorption of clusters on the cylinder walls or dead spaces occuring in the flow in the target assembly will be looked into. As the HeJRT system is used the efficiency of its Operation tends to 93 decrease; however, it is difficult to characterize this effect quantita- tively. In early runs the drop in total system efficiency typically would manifest itself in a manner of hours of running time. Presently the drop in system efficiency does not manifest itself until many days of running time have elapsed. In fact, the capillary is now typically replaced after a week or so of running time not because of a drop in efficiency for the system but to prevent such a drop from occuring at an inconvenient time in an experiment. Thus, it is difficult to place a high efficiency time limit on the capillaries used in the system as operated today. It is, however, possible to make some empirical and/or qualitative statements concerning this efficiency drop and to present some thoughts on the subject which are not supported by hard experimental evidence. The first concerns the nature of the efficiency drop. The efficiency will hold fairly constant, then after some time will start to drop. Then it will drop quite rapidly to about 10-20% of its initial efficiency, after which the drop is much more gradual. If the system is then rejuvinated as described below, the efficiency will return to essentially its initial value; however, as the system's use is continued the efficiency will again fall and the drop will occur after a shorter period of running than did the initial drop. The reason for the efficiency drop is related to a build-up Of material in the capillary and on the target which is usually quite visable. The effect of the build-up in the capillary presumably is to introduce additional turbulence in the helium flow and thereby to decrease transport efficiency. It seems logical that this turbulence would cause a further build-up, causing the more rapid drop in transport efficiency. The build—up in the capillary is soluble in acetone. Accordingly, by flushing the capillary 94 with acetone it is possible to rejuvinate the capillary. Presumably the build-up on the targets is from the same source; however, it is heated by the cyclotron beam and resembles more a carbon deposit, which is not soluble in acetone. In early runs when benzene was first added to the helium supply to increase the efficiency of the HeJRT system, quite large amounts of benzene were used (higher than 0.1% in some cases). In an effort to reduce the build-up of material in the capillary and on the targets, the concentration of benzene in the system was gradually cut back to the 10-20 ppm used at present. While this may have helped slightly in the case of the build-up on the target it did not seem to bring about a major improvement in the case of the build-up in the capillary. It seems the rate of decay of the system efficiency is related to air in the system. It was observed that the problem of build-up of material in the capillary was particularly bad in those runs in which the target assembly was opened frequently and those runs in which the operation of the HeJRT system using benzene in the helium supply was compared with using air plus water vapor in the helium supply. In both cases, if the pump-out Of the target assembly were not complete, air would remain in the system. In general, little care was taken in these early runs to get a complete pump-out of the target assembly, relying on the flow of helium to purge the system. Since that time care has been taken to get more complete pump-outs, and the target assembly is filled with helium and pumped out 3 or 4 times before bringing the beam into the assembly. This, possibly combined with the low benzene concentrations used at present, have essentially solved the problems of build—up in the capillary and have greatly reduced the problem of build-up on the targets. Accordingly, the problem of the total transport efficiency dropping with time has been essentially eliminated. 95 II. 1. Collecting Surface Considerations In the earliest runs of the HeJRT system in this laboratory the standard collecting surface was the sticky side of masking tape. For the past couple of years, since the addition of the tape transport, the standard collecting surface has become paper tape (1" non-oiled computer perforation tape, Singer P/n 200 1218 tape). The reason for the change to paper tape was for the obvious experimental simplification and has had no effect on the collection efficiency. 0n those occasions when cooled solid state detectors are exposed to the vacuum of the detector assembly, aluminized mylar tape is used to eliminate the problem of the small amount of moisture in the paper tape from entering the vacuum and condensing on the detector. Other groups have used such things as old computer magnetic tape (T. T. Sugihara, Texas A & M University, private communication), metal surfaces, and surfaces smeared with vacuum grease. All things considered, it does not seem to make very much difference what is used as a collecting surface when molecular clusters are used to carry activities through the system. However, a comparison of the collection efficiency of paper tape and aluminized mylar tape indicates the aluminized mylar tape to be a little less efficient. Figure 31 shows the results of the comparison made at a series of different target assembly pressures while looking at the 439-keV peak of 23Mg. Each of the points in the figure is the average of three determinations made at that pressure. The two curves have the same essential shape, but the curve for aluminized mylar is from.9 to 12% below that for paper tape. In each case the detector assembly was pumped to below 10 torr and the capillary to tape distance and angle were 1/2 in. and 90°, respectively. At this time it is felt there is not enough known to present an explanation of the differing collection efficiencies, other 96 If .10 N g a I- {I} a: o: I :5 ‘8 °-~ > as E 002 - 4 #9 it!) (Dc-5 0. e 0 <1 - _ l— E I g 0: a: g j 4 In I— F 3‘. . l—jo I__L n 1 1 8 g o 8 0 o 8 o 8 I‘ (0 ID V” ['0 . svaav was (Asa-6217) 5W2; Figure 31. A comparison of the effectivness of paper tape and aluminized mylar tape as a collecting surface. 97 than to suggest the possibility that it is related to the rougher textured surface presented by the paper tape. A preliminary study has been conducted to determine the effect of varying the tape to capillary distance and the angle on collection efficiency. The collection efficiency was followed by monitoring the gross 7 count rate of activities produced with 30-MeV protons on aluminum. The two principal components of the activity were 23Mg and 2581. For these determinations a 0.055-in. polyethylene capillary was used and the helium flow was doped with 10 ppm benzene vapor. The results are summar- ized below for paper tape. Tape to Capillary Collecting Angle *- Distance Efficiency 90° 1/8 inch ElOO% 90° 1 inch 97.5 i 5.% 45° 1/8 inch 98.5 i 5.% 45° 1 inch 97.5 i 5.% Accordingly, within the limits of the uncertainty of the measurements the collecting efficiency for paper tape is unaffected by changes in capillary to tape angles up to 45° or distances up to l in. Since the drops in efficiency upon changing the angle and distance are so small, if they do exist, none of the in—between values were checked. The capillary to tape angle and distance of 45° and l in. are what is typically used in y singles experiments in order to keep the tip Of the capillary (where activities will Often build up) out of the view of a collimated detector. Activities have been collected on paper tape at distances up to 6 in. with fairly good yields; however, at this distance it is necessary to take into consid- eration the divergence of the cluster molecules leaving the capillary. Approximately 90% of the clusters diverge with an angle of less than 3° 98 (see Section II. F. "Cluster Melecules"). As yet this study has not been extended to the aluminized mylar tape we occasionally use. However, by private communication, the members of R. D. Macfarlane's group at Texas A.& M report the efficiency for collecting on mylar tape drops to about 50% on changing the capillary to tape angle from 90° to 45° When the HeJRT system is operated using molecular clusters to increase transmission efficiency, it is necessary to consider a couple of side effects. The first is a desirable effect; when the desired product of the nuclear reaction is volatile or is naturally a gas, it is not necessary to go to any additional effort (like using a LN cooled collecting surface) 2 to hold the activity after transport through the system. This is evidenced by the presence of the 231l-keV line from 7l—sec 1"0 in all y spectra recorded when using an oxide target in the HeJRT system. The line is present in good yields in runs using both paper tape and aluminized mylar tape, both at room temperature. The second effect is not desirable when performing charged particle spectroscopy; the build-up of cluster molecules on the collecting surface, through which the emitted charged particles must pass will degrade the energy of the particles. The effect on 8 particles is small except when the build-up of clusters occurs over a long period of time as is the case for long-lived activities. However, the effect on the energy of emitted 0 particles is much more serious even when the build-up of clusters is only limited to a few minutes. It is hoped a heated collecting surface can be used in these cases to break up the cluster molecules as they are deposited on the collecting surface, such as R. D. Macfarlane uses in his recoil time—of-flight spectrometer [MaR73]. 99 II. m. Aqueous Chemistry OnéLine The hope that "wet" chemistry could be performed on-line with the HeJRT system arose from the Observation that the efficiency Of the HeJRT system when depositing activities on paper tape did not start to fall Off until the pressure in the detector chamber (box where activities transported through the system are deposited on some col— lecting surface and counted) was raised to above 20—30 torr (see Section II.n. 'Detector Assembly Pressure vs. Efficiency'K This was important because the vapor pressure of water at 20°C is 17.5 torr. Accordingly, it would be possible to do aqueous chemistry under these conditions (collecting chamber pressure 220 torr) and not have the solutions boiling off. The first attempt at chemistry was an on—line attempt in the search for 6"'Ge. The apparatus for this run is shown in Figure 32. The activities were generated using a 3He beam on a natural Zn foil and were transported through a polyethylene capillary to the detector chamber. The detector chamber was pumped so as to maintain a pres— sure 220 torr. The exit of the capillary was directed at an orifice in a sealed container partially filled with concentrated HCl. It was hoped that the activities attached to molecular clusters would enter the H01 container through the orifice and become trapped in the acid solution. There the Ge activities would form the volatile chloride GeClu and be pumped Off to a cold trap, where they could be counted. This first attempt was unsuccessful primarily because of inefficiencies remaining in the HeJRT system at that time, poor alignment between the capillary and orifice to the acid container, 100 .Eovmhm ammom on» new: >pumwemno ascends mcwshomnom um pmauuum umuwm 9:0 pom asupom ampsoewhomxo use mo nupmxm .mm whomwm mine $5304) a Oh a: E58». 9 Basis Ems/Eu 025538 » >4m2mww< Humid... 20mm >m<44Eosa EORM meadows one eyes somwsmmeoo 0cm zepmeono moomsvm OCHHICO pom QSIpOm HmpcmEHhmaxm 9:0 mcwm: 00000000 mepooam may mo mcoflpeom mad I 2123.100 20 wzzbmnfloo NHO I 2123.100 20 oz_.1.om.joo 9.0 I was. 20 02:83:50 __.O A 23m tmmqm Annmczoom \Ammgoomm 0:31 mum—>132 JmZZ