Tie-of-flight Identification of Ions in CESR and ERL Eric Edwards Departent of Physics, University of Alabaa, Tuscaloosa, AL, 35486 (Dated: August 8, 2008) The accuulation of ion densities in the bea pipe of an electron bea accelerator can disturb the bea s dynaics. Efforts to clear this ion cloud fro the bea pipe rely on assuptions about its coposition, and this work concerns the construction of a detector to deterine the concentration of different ion species and charge states in the CESR and ERL bea pipes. This report describes the detector s assebly and its test with a low energy electron bea. I. INTRODUCTION The plan for construction of an Energy Recovery Linac (ERL) at Cornell necessitates a thorough study of the effects of ion densities that accuulate in the bea pipe of such an accelerator. In an ERL, the desire to achieve unprecedentedly controlled bea dynaics renders these densities undesirable, and they need to be understood and avoided where possible. These densities result fro scattering by the electron bea on the dilute gas of the bea pipe. The interaction between accuulated ion densities and the electron bea has a nuber of deleterious effects on the bea, and the reader is referred elsewhere for detailed investigations (e.g., see [1, 2]). In soe anner the ion densities that will accuulate ust be reoved, and the particular coposition of the ion cloud both deterines the efficacy of the clearing ethods and gives inforation necessary to copute the effects of ions that cannot be reoved. This work concerns the assebly of a detector to characterize the concentration of different species and charge states in the bea pipe. In particular, this project encopasses the assebly and initial calibration of a tie-of-flight detector that can be ounted on an accelerator bea pipe to perfor in-situ easureents of the ion densities. The entire syste contains three ain eleents: the detector, the easureent syste, and a test assebly or low energy bea source. This paper describes the assebly and calibration of these three eleents through testing with the low energy electron bea. II. THEORY A general theory for tie-of-flight easureents is outlined here for copleteness. A tie-of-flight schee operates by iparting an equivalent potential V to a nuber of particles with asses 1, 2,..., N and charge q 1,q 2,..., q N then easuring the corresponding velocity distribution of the particles. For a particle of ass and charge q, 1 2 v2 = qv (1) 2qV v =. (2)
2 So v q and this relation will allow identification of the relative concentration of ions using easureents of the velocity distribution of the ion cloud. The velocities are easured by letting the ions drift along a tube of length l and easuring the corresponding tie-offlight. For, 2qV v = where v = l so that (3) t t = l 2V q. (4) Ions with ass-to-charge ratio r 1,r 2,..., r i have a tie-of-flight t 1,t 2,..., t i respectively, and we perfor the easureents to count the nuber n j of signals in the interval t j = [t j δ, t j + δ]. We then expect, where N = j n n j counts all ion signals, li j N = ρ N j where ρ j is the relative concentration of the ion with ass-to-charge ratio r j. Siulations of the ion cloud in the bea pipe deterine the proportion p of the ions that are easured N by the detector N = pn total, and we obtain N j = ρ j for the nuber of ions in the bea p pipe of ass-to-charge ratio r j. These siple relations encapsulate the general theory necessary to perfor tie-of-flight easureents and identify the ions - the rest concerns details of easureent. For exaple, we ll need a well-defined drift length l and zero of the tie-of-flight t 0 in order to differentiate between ions. The resolution in the tie easureents t necessary to resolve ions differing by can be written, to first order in, as 1 2 ( + )v2 = qv (5) l t + t = 2qV ( + ) (6) ( ) 1 1+ 2 2qV l =(t + t) ( 1+ 1 2 ) 2qV l =(t + t) t = ( l 2 2qV (7) (8) ). (9) Table I lists exeplative tie-of-flights for ions expected in easureents on CESR and ERL. TABLE I: Saple of tie-of-flights of various ions assuing each olecule is once ionized, q=1.6 10 19 C, and fro inserting the detector specifications V = 400 Volts, l = 5.08 c. Ion Mass/Ion (kg) TOF (µs) H 2 O 2.99 10 26 0.78 Ne 3.32 10 26 0.82 Ar 5.32 10 26 1.04 N 2 7.31 10 26 1.21
3 III. EXPERIMENTAL SETUP The TOF spectroeter is analyzed by a test assebly. This test assebly generates a low energy electron bea to ionize the dilute gas and provide an ion source for the detector. The detector contains the drift tube for the tie-of-flight schee and produces signals when the ions arrive at the end of the drift tube. The easureent apparatus encopasses an array of electronics used to receive the incoing signals fro the detector and operate sufficiently rapid easureents. A. Test assebly FIG. 1: A sketch of the test assebly setup. The test assebly, as shown in Fig. 1, constitutes the first phase of operation of the detector to test its functionality and to calibrate it before installation on the CESR ring. It is designed to be a low energy electron bea source to scatter on either the dilute gas that is in the vacuu syste inadvertently or on gases leaked into the chaber for the test. It consists of a tungsten filaent that is heated by a current to eit electrons, a leak valve, and a region for the electrons to ionize the gas. The filaent is operated between 2-2.5 Aps and biased at voltages < 100V. This biasing is accoplished by floating the ground of the filaent power supply while isolating it fro the ground of the chaber walls by using an isolating transforer. The energy iparted to the electrons propels the into an ionization region to scatter on the dilute gas (at pressure approxiately 10 6 Torr). The ions are shielded fro the filaent bias by a grounded plate ounted in front of the filaent. The resultant ion cloud is used to test and calibrate the detector and resebles the ion cloud in an accelerator such as CESR or the ERL. B. Detector Coponents of the detector as shown in Fig. 2 are the gate ring, the focus ring, the drift tube, and an electron ultiplier. The gate ring gates the flows of ions into the drift tube by aintaining a positive voltage and quickly switching to ground or negative voltage.
4 FIG. 2: A scheatic of the detector showing, fro left to right, the gate ring, focus ring, drift tube, and channeltron. This switching constitutes the t 0 of the ties-of-flight of the ions. A grounded plate, though not part of the detector per se, is ounted in front of the gate ring to siulate the bea pipe in CESR or the ERL. It is assued the ions begin as a hoogenous ixture in the region outside of this ground plate. The focus ring operates at -200V and focuses the ions into the drift tube once the gate is switched. The drift tube is 5.08c and operates per the tie-of-flight easureent schee described above. The electron ultiplier, in our case a Channeltron, operates at -2kV and, via secondary eission processes, aplifies the individual ion signals. C. Measureent apparatus The Channeltron produces a signal of aplitude 10-30 V for an individual ion arrival, and this signal is run through a pre-aplifier to transfor it into a narrow, negative pulse of aplitude 2-5 V. The gate ring voltage is controlled by a siple operational aplifier circuit that switches between a supply at positive voltage V + and a negative voltage V. As shown in Fig.3, the circuit receives a pulse of width w fro a pulser and switches fro V + to V for a tie interval of duration w. The arrival of the pulse at the gate ring circuit deterines the t 0 of the tie-of-flight easureents. This signal is siultaneously sent to a tie-to-aplitude converter as the start signal. The stop signal is then the ion signal as detected on the Channeltron. The tie-to-aplitude converter produces a pulse of voltage V out = trange t ion t range V ax where t range and t ion are the selected range of the tie-to-aplitude converter and [t stop t start ], respectively (V ax is the axiu output voltage of the tie-to-aplitude converter). Thus we deterine the tie-of-flight t stop t start = t ion = Vout V ax t range precisely fro the voltage produced by the tie-to-aplitude converter. To investigate different regions of the tie doain, we can delay the start signal fro the pulser, using an Ortec gate and delay generator odule, by a tie t delay and the tie-of-flight is then t delay + t ion. A coputer running an Ortec progra, MCA32, operates as a ulti-channel analyzer that easures and counts the pulses of aplitude V out produced by the tie-to-aplitude converter. In this anner, we construct the desired histogra that, through a sequence of relations fro the voltage to the ass-to-charge ratio, provides the inforation necessary to reconstruct the relative concentrations of ion species in the bea pipe. The histogras that have been obtained so far have been inconclusive. As seen in Fig. 4, a Gaussian type peak develops, but it cannot be concluded if this lineshape corresponds to interference or unresolved ion signals.
5 FIG. 3: Diagra for the easureent process. FIG. 4: Photo of the histogra readout in MCA32 on a coputer taking inputs fro the tie-toaplitude converter. IV. FUTURE WORK Future work for this project consists of copletion of the test assebly phase and installation on CESR and the ERL. The present experiental setup does not yet produce histogras that allow us to ake conclusions about the character of the ion densities in the test assebly. Three odifications are discussed here that ay be ade to iprove the resolution and to conclude the test assebly phase. First, the low energy bea of the test assebly produces electrons that can be attracted by the positive voltage on the gate ring, which will not be the case in either CESR or ERL. These electrons have a long ean-free path inside the test assebly-detector chaber so that soe nuber arrive at the gate ring voltage and are attracted into the detector resulting in noise. This proble can be avoided, however, by ionizing in a region and then clearing
6 the electrons. The solution requires only slight odification of the current test assebly structure and provides a pure ion source to the detector, which will reseble ion densities in CESR and the ERL ore closely. Second, radio-frequency noise currently disallows starting the tie-to-aplitude converter on the ions signals, and reflections affect the shape of the gate ring voltage (sears the edges of the square wave). There exist various ethods for eliinating or reducing RF noise that need to be ipleented. If one can decouple the signal on the collector wire fro the noise, then that is ost desirable. But if the noise persists, one ight run the Channeltron at higher voltages to increase the weakest ion signals to a threshold of, say, 15V. Then the RF noise proble need not be totally eliinated, only reduced so that the noise has aplitude < 15V and can be reliably filtered fro the ion signals. Third, it would be helpful to enhance the switch rate to reduce abiguity in t 0. V. ACKNOWLEDGMENTS I a pleased to thank Dr. Georg Hoffstaetter of Cornell University for offering this work as a Laboratory for Eleentary Particle Physics Research Experience for Undergraduates project. I would like to thank Michael Ehrlichan and Dr. Val Kostroun of Cornell for their assistance and Dr. Rich Galik for his organization of the LEPP REU progra. Finally I would like to thank the generosity of the entire CLASSE staff and faculty. The pervasive willingness to help and interest in our success fro the CLASSE personnel ade the work both possible and enjoyable. This work was supported by the National Science Foundation REU grant PHY-0552386 and research co-operative agreeent PHY-0202078. [1] G.H. Hoffstaetter, M. U. Liepe, Nuclear Instruents and Methods in Physics Research A 557, 205-212 (2006) [2] G.H. Hoffstaetter, C. Spethann, Physical Review ST-AB 11, 014001 (2008)