avatar University of Washington Inc Services

Pages

  • Page 1


  • Page 2

    INTRODUCTION Last year the Nuclear Physics Laboratory (NPL) officially became the Center for Experimental Nuclear Physics and Astrophysics (CENPA), with an expanded mandate. CENPA includes the activities of the former NPL and in addition fosters collaborative work among the members of the NPL and others in the University of Washington Physics Department and elsewhere. CENPA pursues a broad program of research in nuclear physics, astrophysics and related fields. Research activities are conducted locally and at remote sites. The current program includes "in-house" research on nuclear collisions and fundamental interactions using the local tandem Van de Graaff and superconducting linac accelerators, as well as local and remote non-accelerator research on fundamental interactions and user-mode research on relativistic heavy ions at large accelerator facilities in the U.S. and Europe. A good 7Be(p,g)8B Phase I data taking run has been completed and the data are being analyzed. We used a 100 mCi target fabricated on a newly designed post backing with a breakaway washer, eliminating unwanted target tails. All important sources of systematic error were measured, including loss of 8B due to backscattering out of the target. We expect to meet our goal of 5% precision on the cross section and the astrophysical S-factor. The SNO detector has been running in a production mode with pure heavy water in the acrylic vessel since November 1999. Analysis of the data taken since that time has been directed towards a measurement of the rate of charged-current interactions of 8B neutrinos, and will be completed soon. Both the SNO collaboration and the scientific community await this milestone with interest. A new collaboration of CENPA members with the University of Mainz, Kernforschungszentrum Karlsruhe, and other institutions to carry out a large-scale experiment on tritium beta decay has formed. The objective is a direct kinematic measurement of the mass of the electron antineutrino with 0.5-eV sensitivity. The notion of "large extra dimensions" has recently attracted a great deal of attention, particularly as a solution of the gravitational hierarchy problem. For example, the "true" Planck mass could be lowered to about 1 TeV if two of the extra 7 dimensions of string theory had sizes of around 1 mm. This would show up as a violation of the gravitational inverse-square law for separations less than a milimeter. We recently used a novel torsion balance instrument to test the inverse-square law down to 0.2 mm and found no evidence for anomalies, which implies an unification mass of > 3.5 TeV. The big news in ultrarelativistic heavy ion physics this year is first data from the RHIC collider. Experimental results have been pouring in from all four experiments. The UW event-by-event program has pioneered a number of novel analysis techniques that are now being brought to bear on the first batch of STAR data. We have seen for the first time substantial dynamical fluctuations in event-wise mean transverse momentum, possibly signaling new QCD effects in Au-Au collisions, and charge- or isospin-dependent correlation structures possibly connected to novel structure on the hadronic freezeout surface formed during rapid traversal of the QCD phase boundary. A study of collision-induced multifragmentation of C60 has been completed. Multifragmentation into three or more fragments each with three or more carbons in each fragment is found to be an important reaction channel for large deposition energies. An odd-even dependence of fragment yields implies sequential decay of chain or ring multifragmentation products. As always, we encourage outside applications for the use of our facilities. As a convenient reference for potential users, the table on the following page lists the capabilities of our accelerators. For further information, please contact Prof. Derek W. Storm, Executive Director, Nuclear Physics Laboratory, Box 354290, University of Washington, Seattle, WA 98195; (206) 543-4080, or storm@npl.washington.edu. Further information is also available on our web page: http://www.npl.washington.edu . We close this introduction with a reminder that the articles in this report describe work in progress and are not to be regarded as publications or to be quoted without permission of the authors. In each article the names of the investigators are listed alphabetically, with the primary author, to whom inquires should be addressed, underlined.


  • Page 3

    Derek Storm, Editor storm@npl.washington.edu (206) 543-4085 Barbara Fulton, Assistant Editor


  • Page 4

    TANDEM VAN DE GRAAFF ACCELERATOR A High Voltage Engineering Corporation Model FN purchased in 1966 with NSF funds; operation funded primarily by the U.S. Department of Energy. See W.G. Weitkamp and F.H. Schmidt, "The University of Washington Three Stage Van de Graaff Accelerator," Nucl. Instrum. Meth. 122, 65 (1974). Some Available Energy Analyzed Beams Ion Max. Current Max. Energy Ion Source (particle m A) (MeV) 1H or 2H 50 18 DEIS or 860 3He or 4He 2 27 Double Charge-Exchange Source 3He or 4He 30 7.5 Tandem Terminal Source 6Li or 7Li 1 36 860 11B 5 54 860 12C or 13C 10 63 860 * 14N 1 63 DEIS or 860 16O or 18O 10 72 DEIS or 860 F 10 72 DEIS or 860 * Ca 0.5 99 860 Ni 0.2 99 860 I 0.01 108 860 * Negative ion is the hydride, dihydride, or trihydride. Additional ion species available include the following: Mg, Al, Si, P, S, Cl, Fe, Cu, Ge, Se, Br and Ag. Less common isotopes are generated from enriched material.


  • Page 5

    [UW NPL Home Page] ANNUAL REPORT Center for Experimental Nuclear Physics and Astrophysics University of Washington May, 2001 Sponsored in part by the United States Department of Energy under Grant #DE-FG03-97ER41020/A000. This report was prepared as an account of work sponsored in part by the United States Government. Neither the United States nor the United States Department of Energy, nor any of their employees, make any warranty, expressed or implied or assumes any legal liability or responsibility for accuracy, completeness or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe on privately-owned rights. Introduction Accelerator Beams available (Much of this material is in PDF format. Download Adobe Acrobat Reader) Detailed Table of Contents Chapters: 1. Fundamental Interactions (329kB) 2. Neutrino Physics (1913kB) 3. Nuclear and Particle Astrophysics (203kB) 4. Ultra-Relativistic Heavy Ions (444kB) 5. Atomic and Molecular Clusters (53kB) 6. Electronics, Computing and Detector Infrastructure (58kB) 7. Accelerator and Ion Sources (71kB) 8. Nuclear Physics Laboratory Personnel (43kB) 9. List of Publications from 2000-2001 (88kB) [UW NPL Home Page]


  • Page 6

    UW CENPA Annual Report 2000-2001 v Contents 1 Fundamental Interactions 1 1.1 Eöt-Wash data acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Millimeter-scale test of the gravitational inverse square law . . . . . . . . . . . . . . 2 1.3 Feetback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.4 Simple parallel computing with Mathematica . . . . . . . . . . . . . . . . . . . . . . 6 1.5 Data analysis and signal calculations for the short-range experiment . . . . . . . . . 7 1.6 Gravity’s gravity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.7 Measurement of Newton’s constant G . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.8 emiT: time reversal violation in neutron beta decay, preparations for a second run . 11 1.9 Dead layers of proton detectors for the emiT experiment . . . . . . . . . . . . . . . . 13 1.10 4 He(α, γ)8 Be . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 1.11 Search for a permanent electric dipole moment in liquid 129 Xe . . . . . . . . . . . . 16 1.12 f t value of the 0+ → 0+ decay of 32 Ar: a measurement of isospin breaking in a superallowed decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2 Neutrino Physics 19 2.1 SNO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.1.1 Status of the SNO solar neutrino analysis . . . . . . . . . . . . . . . . . . . . 19 2.1.2 Verifying event building in SNO data . . . . . . . . . . . . . . . . . . . . . . 21 2.1.3 Channel status verification for SNO data . . . . . . . . . . . . . . . . . . . . 22 2.1.4 Muons and muon induced spallation neutrons in SNO . . . . . . . . . . . . 23 2.1.5 SNO operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 2.1.6 The SNO data acquisition system . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.1.7 Time stamp validation in the SNO experiment . . . . . . . . . . . . . . . . . 26 2.1.8 Variation of neutrino flux with (local) solar position . . . . . . . . . . . . . . 27 2.1.9 Neutrino events with solar flares revisited . . . . . . . . . . . . . . . . . . . 29 2.2 SNO/NCD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30


  • Page 7

    vi 2.2.1 Neutral current detectors in SNO . . . . . . . . . . . . . . . . . . . . . . . . . 30 2.2.2 NCD data taking and analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 31 2.2.3 NCD electronics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 2.2.4 Overview and status of the NCD DAQ software . . . . . . . . . . . . . . . . . 34 2.2.5 First Deployment of a Neutral Current Detector in the Sudbury Neutrino Observatory: The CHIME Engineering Run . . . . . . . . . . . . . . . . . . . 36 2.2.6 The NCD laser welding equipment . . . . . . . . . . . . . . . . . . . . . . . . 37 2.2.7 NCD deployment equipment progress . . . . . . . . . . . . . . . . . . . . . . 39 2.3 Lead as a neutrino detector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 2.3.1 Lead perchlorate as a neutrino detection medium . . . . . . . . . . . . . . . . 40 2.3.2 Supernova neutrino detection using lead perchlorate . . . . . . . . . . . . . . 42 2.4 SAGE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 2.4.1 SAGE summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 2.4.2 Verification of the technology for production of an intense 37 Ar neutrino source from a Ca target irradiated in a nuclear reactor . . . . . . . . . . . . . 45 2.5 MOON (Mo Observatory Of Neutrinos) for low energy neutrino physics . . . . . . 46 3 Nuclear Astrophysics 47 3.1 Reanalysis of α+α scattering and β-delayed α decay from 8 Li and 8 B . . . . . . . 47 3.2 β-delayed alpha spectra from 8 Li and 8 B decays and the shape of the neutrino spectrum in 8 B decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 3.3 7 Be(p,γ)8 B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 3.4 7 Be(p,γ)8 B Phase II . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 3.5 WALTA: The Washington Large-area Time-coincidence Array . . . . . . . . . . . . . 53 4 Relativistic Heavy Ions 54 4.1 HBT physics at STAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 4.1.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 4.1.2 Pion phase space density from STAR HBT analysis . . . . . . . . . . . . . . 56 4.1.3 A fast algorithm for finite-size and -resolution Coulomb correlations . . . . . 57


  • Page 8

    UW CENPA Annual Report 2000-2001 vii 4.1.4 Elliptic flow effect on pion interferometry analysis in ultra-relativistic heavy- ion collisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 4.2 Event by event physics at STAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 4.2.1 Summary: EbyE physics at RHIC . . . . . . . . . . . . . . . . . . . . . . . . 60 4.2.2 STAR < pt > fluctuation analysis . . . . . . . . . . . . . . . . . . . . . . . 61 4.2.3 STAR mt ⊗ mt correlation analysis . . . . . . . . . . . . . . . . . . . . . . . 62 4.2.4 STAR η and φ correlations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 4.2.5 Multi-particle azimuthal correlations and jet production in central collisions √ of Au on Au at sNN = 200 GeV in a Monte-Carlo model . . . . . . . . . 64 4.2.6 Power-law structure of minimum-bias multiplicity distributions . . . . . . . . 66 4.2.7 Multiplicity fluctuations in STAR . . . . . . . . . . . . . . . . . . . . . . . . 67 4.3 Event by event physics at the SPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 4.3.1 Summary: EbyE physics at the SPS . . . . . . . . . . . . . . . . . . . . . . . 69 4.3.2 NA49 < pt > fluctuations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 4.3.3 NA49 pt ⊗ pt correlations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 4.3.4 NA49 multiplicity fluctuations . . . . . . . . . . . . . . . . . . . . . . . . . . 72 4.4 Scale dependence of global-variables fluctuations . . . . . . . . . . . . . . . . . . . . 74 4.5 K0Long detection in STAR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 4.6 Particle-production mechanisms at RHIC . . . . . . . . . . . . . . . . . . . . . . . . 77 4.7 Particle identification at STAR/RHIC . . . . . . . . . . . . . . . . . . . . . . . . . . 79 4.8 Defining a generic event-by-event data summary format . . . . . . . . . . . . . . . . 80 5 Atomic and Molecular Clusters 81 5.1 High energy fragmentation of C60 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 5.2 Search for gas phase dianions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 6 Electronics, Computing and Detector Infrastructure 83 6.1 Status of advanced object oriented real-time data acquisition system . . . . . . . . . 83 6.2 Shaper board electronics development . . . . . . . . . . . . . . . . . . . . . . . . . . 84 6.3 Laboratory computer systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85


  • Page 9

    viii 6.4 Electronic equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 7 Van de Graaff, Superconducting Booster and Ion Sources 88 7.1 Van de Graaff accelerator operations and development . . . . . . . . . . . . . . . . . 88 7.2 Superconducting Linac . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 7.3 Tandem terminal ion source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 7.4 Cryogenic operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 8 Nuclear Physics Laboratory Personnel 94 9 List of Publications from 2000-2001 97


  • Page 10

    UW CENPA Annual Report 2000-2001 1 1 Fundamental Interactions 1.1 Eöt-Wash data acquisition E. G. Adelberger, B. R. Heckel, C. D. Hoyle, D. J. Kapner, S. Merkowitz,∗ U. Schmidt† and H. E. Swanson The current data acquisition code is written in Microsoft Visual C++ . The User Interface which includes data plots and control panels is built with National Instruments Lab Windows CVI. A custom driver was written for the data acquisition hardware that reads sixteen 16-bit ADC’s, switches among sixteen channels of a temperature-sensor multiplexer, and otherwise controls the experiment. This combination can run at an interrupt rate of up to 20 Hz where each interrupt is a burst of four at a 240 Hz rate. This burst of four is averaged to filter any 60 or 120 Hz component of the signal. These can be further averaged up to the period between two samples. Data is continually displayed whether or not it is being written to the disk. Two of the experiment data stations run variations of this code under the Windows 98 operating system. The Eöt-Wash II data station: The data acquisition program for the Eöt-Wash II experiment was upgraded last year but remains a Turbo Pascal program running under the DOS operating system. A second computer houses the DSP based controller for the turntable speed. A Stanford Research Systems (SRS) function generator provides the clock which determines the turntable speed. It communicates with the acquisition program by serial link. The Eöt-Wash III data station: The general features of this program have been previously re- ported.1 Since then communication with the DSP turntable speed controller has been fully inte- grated into the acquisition program. The DSP generates the timing for both sampling the data and the turntable speed. It now includes code to drive stepping motors for the fiber attachment’s angle and z position, and the screws which adjust the rough heights of the mounting legs. Precise control of the tilt of the apparatus is achieved by varying the temperature of these legs as described in Section 1.3. The Short-Range data station: The Turbo Pascal program and 486 based computer used for the Rotwash experiment and first Short-Range measurement has been replaced by a pentium class computer running the software described above. It writes a compatible data file and has the same functionality as the old program such as automatic calibration of the attractor angle readout. Three phase locked (SRS) function generators control the sample period, and the periods of the attractor and calibration turntables. These can all be set to integer multiples of the pendulum period. GPIB communication with the function generators and lockin amplifiers is integrated into the code. An independent Lab Windows program communicates with a Newport stepping motor controller to set the fiber attachment’s angle, and x, y, and z positions. Another reads a capacitance meter measuring the pendulum - screen capacitance. In the future these functions will be added to the acquisition program. ∗ Presently at NASA/GSFC, Code 661, Greenbelt, MD 20771. † Presently at Phyrikalisches Institut, Heidelberg, Germany. 1 Nuclear Physics Laboratory Annual Report, University of Washington (2000) p. 13.


  • Page 11

    2 1.2 Millimeter-scale test of the gravitational inverse square law E. G. Adelberger, J. H. Gundlach, B. R. Heckel, C. D. Hoyle, D. J. Kapner, A. Kopp,∗ U. Schmidt,† D. Spain‡ and H. E. Swanson . Since our last report,1 we have made significant progress toward testing gravity at distances less than 1 mm with specially designed torsion balances. Our experiments are principally motivated by higher-dimensional string theories which predict deviations from the gravitational inverse-square law at short distances.2 These deviations are typically parameterized as an addition of a Yukawa term to the Newtonian potential, m1 m2 V (r) = −G( )(1 + αe−r/λ ). (1) r We have achieved the best constraints to date on the parameters α and λ for short distances (see Fig. 1.2-1). These results have recently been published.3 Figure 1.2-1. Constraints on the strength, α, and range, λ, of a new interaction of the form given in Eq. (1). The heavy line labeled Eöt-wash is from this work. See C.D. Hoyle et al.3 for references regarding the theoretical predictions and other experimental limits. Our apparatus consisted of a disk-shaped torsion pendulum containing 10 holes evenly spaced about the azimuth. The pendulum was suspended above a rotating attractor of similar geometry. A gravitational or Yukawa interaction between the holes produced a torque on the pendulum that varied periodically at 10 times per attractor revolution (higher harmonics were present as well). The attractor was composed of two stacked disks. The holes in the lower disk were rotated relative to ∗ REU summer student, Lawrence University, Appleton, WI 54912. † Presently at Phyrikalisches Institut, Heidelberg, Germany ‡ REU summer student, Indiana University, Bloomington, IN 47405. 1 Nuclear Physics Laboratory Annual Report, University of Washington (2000) p. 11. 2 see, for example, N. Arkani-Hamed, et al., Phys. Lett. B 429, 263 (1998). 3 C. D. Hoyle et al., Phys. Rev. Lett. 86, 1418 (2001).


  • Page 12

    UW CENPA Annual Report 2000-2001 3 those in the upper disk in such a way as to suppress Newtonian torques, while leaving the signature of a short-range interaction unchanged. A thin, beryllium-copper electrostatic screen separated the pendulum and attractor. Our published data3 constrained the Yukawa interaction modeled by the two extra dimension scenario to have a range less than 190 µm, corresponding to a unification scale greater than 3.5 TeV. Testing gravity at even shorter length scales requires improvements in two main areas. First, the pendulum and attractor geometries must be optimized for the range of interest. Second, the separation between the pendulum and attractor must be made as small as possible. In the last year, we have concentrated much effort into these two areas. To optimize the geometry, a detailed numerical integration code was developed. This code is described separately in this report (see Sec. 1.5). The optimization resulted in the design shown in Fig. 1.2-2. This new design has 26-fold symmetry. The Newtonian torques are suppressed to a very high degree and effects due to a short-range interaction are enhanced relative to the previous data. Several innovations have been applied to achieve the smallest possible pendulum/attractor sep- aration. A copper bellows in series with the torsion fiber reduced the vertical “bounce” amplitude to around 1 µm, a factor of 10 improvement. An improved magnetic damper has also reduced the simple pendulum, “swing,” mode. In addition, a thinner (10 µm thick) electrostatic screen has been implemented. The active components of the pendulum and attractor are made of titanium and molybdenum respectively. These materials can be made very flat with careful machining and grinding techniques, reducing problems due to surface irregularities. Finally, we devised improved capacitive techniques for leveling and aligning the disks. We are presently taking data with the new design and expect to have much improved results this year. Pendulum-attractor separations of less than 0.1 mm should be attained. Figure 1.2-2. The pendulum and rotating attractor of our second-generation experiment. The shaded disks are the active components. The mirrors used for the optical readout system and aluminum support frame are also shown. The electrostatic screen has been omitted for clarity.


  • Page 13

    4 1.3 Feetback E. G. Adelberger, B. R. Heckel, U. Schmidt∗ and H. E. Swanson Changing the tilt on our high precision torsion balances causes a twist of the fiber that the pendulum is attached to. This effect, labeled “tilt-feed through,” is of the order of 1% in our experiments. The tilt of the laboratory floor typically changes by 1 µrad per day, resulting in a daily false signal of the order of 10 nrad. The biggest correction to our previous measurement1 of the composition- dependent test body accelerations was for laboratory tilt, and the uncertainty of the measured tilt feed-through was a major contribution to our systematic error budget. To eliminate the tilt effect, we developed an active leveling system. Our apparatus rests on three feet. Two of them are active ones, which can change their height using temperature dependent expansion to compensate for the laboratory tilt. Fig. 1.3-1 shows a cross section through one foot. The expanding and shrinking components of one foot consists of two lead rings (1) which are soldered to a copper disk (2). Thermal energy can be pumped into or out of the copper disk by a peltier element, which is also thermally coupled to a brass block. The brass block is held at constant temperature by circulating water from a temperature-stabilized reservoir. Two G10 rings (5) thermally isolate the lead rings from the laboratory floor and a stainless steel disk on top, on which the apparatus rests. The peltier element and the brass block are clamped to the copper disk by one bolt (6). Specially formed G10 pieces provide thermal isolation between the bolt, the copper disk and the brass block. 5 4 1 2 1 6 3 5 102 mm Figure 1.3-1. Cross section of a foot of the Eöt-Wash III rotating torsion balance. For details see text. We measure the tilt with two perpendicular Applied Geomechanics Inclinators (AGIs). The AGIs are mounted on the rotating top of the apparatus near to where the prehanger of the pendulum is attached. The analog signals of the AGIs are digitized by the data acquisition system. The Eöt-Wash II rotating torsion balance rests on similar feet to the one shown in Fig. 1.3-1. Due to the heat capacity and the heat resistance of the copper disk and the lead rings, the response of the expansion of the lead rings to a change in the heat flux provide by the peltier element is ∗ Presently at Phyrikalisches Institut, Heidelberg, Germany. 1 S. Baeßler et al., Phys. Rev. Let. 83, 3585 (1999).


  • Page 14

    UW CENPA Annual Report 2000-2001 5 delayed by 14 seconds. Also the heat conductivity of the peltier element, which tends to bring back the temperature of the copper disk to the temperature of the heat bath, has to be taken into account. Therefore we need a model that predicts the response of a foot to a change in heat input to the peltier element. Fig. 1.3-2 shows the model used for calculating the response of the Eöt-Wash II feet. This model poses no problem for a real time calculation because it can be solved analytically. The delayed response of the feet together with the 8 sec integration time of the AGIs leads to a low-pass behavior of the leveling system. Therefore its response to fast changes of the tilt caused by imperfections of the turn tables ball bearing is limited. The tilt, caused by the imperfections of the bearing is periodic in the turntable angle. This allows us to express this tilt in Fourier coefficients of the turntable angle. Once these Fourier coefficient are calculated, they can be fed forward to compensate for the imperfections of the bearing. The simplified flow chart of the feed-back loop of Eöt-Wash II is shown in Fig. 1.3-2. turntable angle stored boundary conditions data provide by the data aquisition raw AGI 1 AGI 1 AGI X calibration raw transformation AGI Y AGI 2 AGI 2 into lab frame simple thermal model calibration calculation AGI X AGI Y of fourier I R2 coefficients projection Σ correction on foot A R1 C1 C2 and foot B Σ ∆t ∆ΤΑ ∆TB calculate feed update fourier T1A forward signal coefficient at from fourier first zero crossing I : heat flux peltier element T2A solve simple coefficients after 2 turns C1 : heat capacity copper disk thermal model C2 : heat capacity lead rings hfA for foot A R1: heat resistance peltier element to solve model for foot B R2: heat reststance copper disk T1A inverse DAC calibraton to peltier T2A peltier element element hfA function power update DAC supply Figure 1.3-2. Thermal model and flow chart of the simplified feetback loop implemented at the Eöt-Wash II rotating torsion balance With the feetback switched on, the typical 1 µrad tilt per day is reduced to 20 nrad.2 For the Eöt-Wash III rotating torsion balance we optimized the foot design using finite element analysis. The response time of the optimized design is 6.7 seconds compared to 14 seconds for the old one. Also we used the results of the finite element analysis to make a more realistic numerical model of the heat flux inside a foot and therefore for the time dependence of the expansion. This numerical model can be solved in real time and is used in the feetback loop implemented at Eöt- Wash III. In order to measure the performance of the new feetback system we have to wait until all changes of the Eöt-Wash III apparatus are finished and we are able to run the feetback with temperature stabilized AGIs. 2 Nuclear Physics Laboratory Annual Report, University of Washington (2000) p. 14.


  • Page 15

    6 1.4 Simple parallel computing with Mathematica U. Schmidt∗ Beginning with Version 3.0, Mathematica allows a user to connect a Mathematica front end to a Mathematica kernel running on a remote computer. On Windows based PCs, connecting to a remote kernel requires one to type commands on both PCs, making this feature not very useful. In addition, the file access takes place on the remote computer. Therefore, all needed input files have to be copied to the remote computer or the code has to be changed in order to access the input file through the network. Also, for large time consuming calculations it is desirable to have batch queues executing jobs automatically one after the other. To solve these problems and to provide a basis for simple parallel computing I wrote the code for a server for Mathematica with the following features: • Start a Mathematica kernel and provide automatic remote access to it. • Start a modified Mathematica kernel, for which the file access is linked to a file server and provide automatic remote access to it. • Act as a file server to enable remote file access for Mathematica kernels. • Provide batch queues, which execute batchjobs automatically one after the other using local or remote kernels provided by a Mathematica server. For data exchange between kernels on different computers the TCP/IP protocol is used. Beside the file server feature the code is written completely platform independent in Mathematica. This allows one to run the same code including file access on any platform without any modification. Also, the server provides FTP-like file exchange between computers. The server requires user identification to retain security. To avoid password sniffing while using an unsecured public network, user identification is scrambled each time with a different key. Tasks like calculating function values for different parameters can be easily split in a number of batch jobs. These batch jobs can be placed in queues and executed on different computers simultaneously. This code provides a very simple, but for many tasks, an effective way to do parallel computing. The feature was used extensively in computing the expected response of our short-range torsion balances. ∗ Presently at Phyrikalisches Institut, Heidelberg, Germany.


  • Page 16

    UW CENPA Annual Report 2000-2001 7 1.5 Data analysis and signal calculations for the short-range experiment E. G. Adelberger, B. R. Heckel and U. Schmidt∗ The procedures to calculate the signals of our short range torsion pendulum experiment,1 to analyze the raw data and to calculate limits of the deviation of gravity from Newton’s law were all written in Mathematica. Calculation of the torque on the pendulum induced by the source: The pendulum torque arises from the interactions between the missing masses in the cylindrical holes in the pendulum and attentors. For the purpose of the torque calculation we can think of sets of cylinders with the dimensions and locations of the holes in our pendulum and source. The torque induced in the pendulum cylinder set by gravitational interaction with the source cylinder sets is equivalent to the torque induced by our source with holes in our pendulum with holes. The total torque is given by the sum over all pairs of source and pendulum cylinders. The locations of a pendulum cylinder and a source cylinder, together with the horizontal component of the gravitational force between the two cylinders, leads to the torque of this pair. The value of the horizontal force components is given by a 6-dimensional integral. In the case of Newton’s 1/r-potential, 4 of the 6 dimensions can be integrated analytically. The remaining two dimensions have to be integrated numerically. In the case of a Yukawa interaction, only two dimensions can be calculated analytically. The Fourier coefficients of the total torque can be determined by calculating the total torque for different source angles. Each Fourier coefficient was calculated on a grid of different vertical separations and different horizontal offsets between source and pendulum. For the values of arguments between grid points, cubic spline interpolation was used. Also the gradients at the boundaries of the grid were calculated and used to improve the cubic spline interpolation. In addition to the values of these Fourier coefficients the maximal possible error of the values due to the numerical integration and the interpolation was calculated. Extraction of the Fourier coefficients of the torque from the raw data: Each data run was divided in an adequate subset of data (the Fourier coefficients of interest have to be orthogonal). These subsets were fitted using the relevant Fourier coefficients and fiber drift terms. Bad subsets, where the data were affected by vibrations (small earth quakes) or fiber slips, were eliminated using the change in the value of the free torsion amplitude as a cut criterion. The measured value for each Fourier coefficient was calculated from the mean of the values of the subsets while the error was determined by the scatter of these values. Calculation of the upper limits on Newton’s law violating interaction: The 10ω, 20ω and 30ω measured Fourier coefficients were fitted simultaneously using a multidimensional fitting routine. The independent parameters were the horizontal offsets and the vertical separation between the source and the pendulum. Most fitting parameters like the offset of the z (vertical)-stage or the overall calibration were constrained by independent measurements. These constraints were included in the χ2 calculation. The χ2 minima as a function of the fit parameters were calculated on a grid of different coupling strengths α and Yukawa lengths λ. From the known probability of the underlying χ2 -distribution, the single-sided 95% confidence level was determined by interpolation for each λ value. ∗ Presently at Phyrikalisches Institut, Heidelberg, Germany. 1 C. D. Hoyle et al, Phys. Rev. Let. 86, 1418 (2001).


  • Page 17

    8 1.6 Gravity’s gravity E. G. Adelberger, J. H. Gundlach, B. R. Heckel, U. Schmidt∗ and H. E. Swanson The theory of general relativity requires that gravitational binding energy undergoes the same acceleration as all other forms of mass and energy in the presence of a gravitational field. Lunar laser ranging measurements have confirmed that the earth and moon fall toward the sun with the same acceleration to 5 parts in 1013 , although their gravitational binding energies differ by 4 parts in 1010 . To use the earth-moon system as a test of the acceleration rate of gravitational binding energy, it is first necessary to confirm that the materials that comprise the earth and moon undergo the same gravitational acceleration. Our Eöt-Wash II torsion balance has been operating with four test bodies: two that have a composition similar to that of the earth’s core and two that have a composition similar to that of the moon (and earth’s) crust. As the apparatus rotates in the lab, we look for a composition dependent acceleration of the test bodies toward the sun. In last year’s report,1 we noted three important upgrades to the Eöt-Wash II rotating torsion balance. The first was the implementation of thermally controlled Pb leveling feet for the apparatus. Signals from tilt monitors on the rotating apparatus were used to generate a feedback signal to the Pb feet to hold the rotation axis vertical to within 10 nrad. A non-vertical rotation axis gives rise to a spurious fiber twist that leads to systematic errors. The leveling feet removed both the daily tilt of the laboratory floor and most of the wobble created by the turntable bearing. The second upgrade was the implementation of small bellows in series with the torsion fiber to isolate and damp vertical (bounce) modes of the pendulum. The third upgrade was the implementation of a real time data acquisition system. The new acquisition system had a feature that would remove energy from the torsional mode of the fiber if the torsion amplitude exceeded a threshold. Ion pump bursts and seismic events would give the pendulum a torsion amplitude too large to be useful for analysis, causing long sections of data to be discarded. With the automatic damping routine enabled, very little data needed to be rejected. In 2000-2001, we accumulated data with the upgraded Eöt-Wash II apparatus. Analysis of this data is underway. The results should reduce our statistical errors by the square root of 2 without adding to the systematic uncertainties. The 6.8 magnitude earthquake of 2/2001 caused the Eöt-Wash II torsion fiber to break, providing a natural conclusion to this experiment. ∗ Presently at Phyrikalisches Institut, Heidelberg, Germany. 1 Nuclear Physics Laboratory Annual Report, University of Washington (2000) p. 14.


  • Page 18

    UW CENPA Annual Report 2000-2001 9 1.7 Measurement of Newton’s constant G J. H. Gundlach and S. M. Merkowitz∗ The gravitational constant, together with h̄ and c, is one of the fundamental and universal constants in Nature. Its value must be determined by experiment. Due to the weakness and non-shieldability of gravity, G is nowadays the least-well measured constant. In addition, several measurements conducted in the last decade deviated considerably from the accepted value, so that the 1998 recommended value for G was assigned an uncertainty of 1500 ppm. We have developed a new torsion balance method to measure G. This method was designed to eliminate the largest sources of systematic error that effected previous measurements. A detailed description of our method can be found in several previous reports and in published papers.1 We measure the angular acceleration of a torsion balance mounted on a continously rotating torsion balance. The rotation rate of the turntable is changed so that the torsion fiber itself remains untwisted. Our method is therefore independent of most torsion fiber properties which may have led to a systematic bias in previous measurements. Our pendulum consists of a thin vertically hanging plate. The plate-pendulum makes this measurement practically independent of the details of its mass distribution. This simple arrangement reflects the biggest single reduction in systematic uncertainty compared to previous measurements. The attractor spheres are located on a separate coaxial turntable. This turntable is operated at a higher rotation rate. The angular velocity difference to the pendulum turntable is kept constant, so that the gravitational acceleration on the pendulum occurs with a constant and high frequency. Gravitational accelerations due to objects in the lab can be eliminated. Figure 1.7-1. Angular acceleration Fourier spectrum. The signal of interest(≈6.37 mHz) is over four orders of magnitude above the random background. The room fixed gravitational background (≈1.7 mHz)is cleanly separated. The additional sharp peaks at≈13 mHz are the expected higher harmonic signals. The spectrum represents 10 hours of data. ∗ Presently at NASA/GSFC, Greenbelt, MD 20771. 1 J.H. Gundlach and S.M. Merkowitz, Phys. Rev. Lett. 85, 2869 (2000).


  • Page 19

    10 We have completed two data sets consisting of six three-day runs each. After every run the spheres were moved to different positions on the turntable to subtract out interactions with the turntable itself. The spheres were reoriented to average over density fluctuations and non-sphericity of the spheres. The second measurement was done with a different set of four spheres. Our largest uncertainty was due to the attractor mass metrology. We used Invar micrometers that were fabricated in our machine shop. We calibrated the micrometers before and after each distance measurement with Invar standards that were in turn calibrated at NIST. The vertical spacing between the spheres was measured with small gauge blocks. The signal frequency was set to ≈6.37 mHz. Systematic checks with exaggerated turntable speeds, magnetic fields and thermal gradients were conducted. The torsion balance turntable angle was numerically differentiated to yield angular acceleration. The data were subdivided and fitted. The scatter of the individual fit values determined the statistical error. The data analysis was tested with numerous simulations. Our value for G is: G = (6.674215 ± 0.000092) × 10−11 m3 kg−1 s−2 . Figure 1.7-2. Results of two data sets taken with different spheres. Each data point is the combination of a pair of attractor configurations that together eliminate accelerations due to the attractor turntable. Combining the three pairs taken with different sphere orientations averages over sphere-density and shape imperfections. The uncertainties are statistical. Using the LAGEOS satellite results2 we compute the mass of the Earth to be: M⊕ = (5.972245± 0.000082) × 1024 kg. Our results are published in ref 1. A detailed publication intended for Physical Review D is in preparation. 2 J.C. Ries et al., Geophys. Res. Lett. 19, 529 (1992).


  • Page 20

    UW CENPA Annual Report 2000-2001 11 1.8 emiT: time reversal violation in neutron beta decay, preparations for a second run H. P. Mumm, A. W. Myers, L. P. Parazzoli, R. G. H. Robertson, T. D. Steiger,∗ K. M. Sundqvist, T. D. Van Wechel, D. I. Will and J. F. Wilkerson The emiT experiment is a search for time-reversal (T) invariance violation in the beta decay of free neutrons. Both CP (charge conjugation - parity) and explicit T violation have been observed in the neutral K meson system. However, some 36 years since the discovery of CP violation, neither CP nor T violation have been observed in any other system and possible origins are still not well understood. Although CP(T) violation in the Kaon system can be accommodated within the standard model of particle physics, both baryogenisis and attempts to develop unified theories indicate that additional sources are required. The standard model predicts T-violating observables in beta decay to be extremely small (Sec- ond order in the weak coupling constant) and hence these are beyond the reach of modern ex- periments.1 However, potentially measurable T-violating effects are predicted to occur in some non-standard models such as those with left-right symmetry, exotic fermions, or lepto-quarks.2,3 Thus a precision search for T-violation in neutron beta decay provides an excellent test of physics beyond the Standard Model. The emiT experiment probes the T-odd P-even triple correlation between the neutron spin and the momenta of the neutrino and electron, D σn · Pe × Pν , in the neutron beta-decay distribution. The coefficient of this correlation, D, is measured by detecting decay electrons in coincidence with recoil protons from a polarized beam of cold (2.7 meV) neutrons. Four electron detectors (plastic scintillators) and four proton detectors (large-area PIN diode arrays) are arranged in an alternating octagonal array concentric with the neutron beam. The protons produced in the decay of free neutrons have a relatively low energy (≤ 751 eV). While this allows for a delayed coincidence trigger between the proton and electron (eliminating much of the background) it increases the complexity of the detection scheme. During the first run, high voltage related problems stemming from higher than expected energy loss in the PIN diodes (refer to the following section) led to damaged electronic components, high voltage related backgrounds and a non-symmetric detector. Systematic effects were less effectively canceled due to the lack of full detector symmetry and a more complex data analysis scheme was required. The result, D = −0.1± 1.3× 10−3 , represents a small improvement over the current world average.4 To assure that the second run is not affected by these problems, a number of major detector upgrades are in progress at CENPA. The goals of these modifications are to increase the reliability of the proton detectors, reduce dead time and reduce systematic effects through improved charac- terization of the neutron beam. To accomplish the goal of increased detector reliability the majority of the proton electronics has been isolated through the use of analog fiber-optic links. This change ∗ Presently at Cymer, Inc. San Diego, CA 92172-1712. 1 M. Kobayashi and T. Maskawa, Prog. Theor. 49, 652 (1973). 2 P. Herczeg, Progress in Nuclear Physics, W.-Y.P Hwang, ed., Elsevier Sciences Publishing Co. Inc. (1991) p. 171. 3 E.G. Wasserman, Time Reversal Invariance in Polarized Neutron Decay, Ph.D. thesis, Harvard University, (1994). 4 L. J. Lising et al, Phys. Rev. C 62, 055501 (2000).


  • Page 21

    12 also lowers the capacitance of the high voltage system, which will reduce damage due to discharges along the proton paddles. Both systems were completed during the last year. The high dead time seen during the first run was partly due to high voltage related backgrounds and partly due to the data acquisition system. In order to detect the low energy decay protons, they are accelerated and focused through a potential of approximately 30 kV. Various parts of the focusing assembly create sufficiently high fields that electron emission can take place. The electrons ionize absorbed gasses (most likely Hydrogen) in the ground plane. These ions are then accelerated back toward the detectors. We believe that this is the source of most of our high voltage related background. We are in the process of a complete design review of the focusing system, and the tests made to date indicate that it will be possible to completely eliminate this source of background. In addition, the thresholds of the ADC boards have been sharpened, allowing better background rejection and operation at lower, more stable voltages. Finally upgrades to the DAQ software will allow closer monitoring of the detector status, significantly increased data rates, and will improve the capability for real time data analysis. We have also made a decision to replace the PIN diodes with Surface Barrier detectors based on comparative studies of PIPS, PIN and Surface Barrier detectors. Surface barrier detectors have proven to be robust, and modifications to the detector that allow their use are nearing completion. We have developed plans to better characterize the neutron beam, and have developed procedures to obtain accurate polarization, flux and magnetic field maps that will be used to better understand systematic effects. In addition, an upgrade to the NIST reactor will result in a factor of approximately 1.8 higher neutron flux. It is expected that emiT will resume collecting data around the first of the year 2002, likely reaching the goal of D < 5 × 10−4 during the following summer.


  • Page 22

    UW CENPA Annual Report 2000-2001 13 1.9 Dead layers of proton detectors for the emiT experiment M. Bhattacharya, H. P. Mumm, K. M. Sundqvist and J. F. Wilkerson We have been conducting studies of dead layers on semiconductor detectors used for proton detec- tion in the emiT experiment. Dead layers are layers of inactive material present on the surface of detectors. They are responsible for energy loss of incident particles before they are measured in the active portion of the detector. They inhibit low energy particle measurements by creating a natural energy threshold which particles must overcome in order to be detected. The emiT experiment detects protons (in coincidence with the betas) from the decay of free neutrons. In order to detect the low energy protons (maximum kinetic energy of 760 eV) the emiT experiment uses a high voltage grid in front of the detectors to accelerate the protons. The applied voltage (−35 kV for emiT’s first run) must compensate for the energy lost by the protons in the dead layer of the detectors. With the application of such high voltages care must be taken to prevent breakdowns as the resulting discharges could damage the detectors and associated electronics. It is therefore crucial to characterize these dead layers such that we may determine the lowest voltage required for the protons to be detected. We measure dead layers in a tabletop vacuum box using 3.18 MeV alpha particles from a collimated 148 Gd source to make energy loss measurements as function of angle of incidence. At normal incidence, energy loss is a minimum as the particles take the shortest path through the dead layer. By making energy measurements at varied angles of incidence, one can determine the thickness of a dead layer by observing the energy difference in measured pulse heights. To accomplish this, our detector is mounted in front of the source on a micrometer arm which allows the angle of incidence to be varied without having to break vacuum. PIN diode detectors were used in emiT’s original experimental run. They have been measured with typical energy losses in their dead layers of 20 keV (corresponding to 25 µg/cm2 of Silicon). However, their dead layer was found to increase over the course of the run. We have decided to use surface barrier detectors for emiT’s second run. These are factory specified with 40 µg/cm2 gold dead layer. Dead layers for new and unused surface barrier detectors were measured in our setup and we found the measured dead layers to be consistent with these specifications. It would be extremely valuable to know how dead layers change on detectors over time, perhaps as the result of radiation damage. This has been the primary motivation for these measurements, and it is hoped that in the near future we will be able to correlate how dead layers increase with detector usage.


  • Page 23

    14 4 1.10 He(α, γ)8 Be R. Hazama, K. A. Snover and D. W. Storm We are in the process of completing the analysis of our high-precision measurement1 of the 4 He(α, γ)8 Be reaction. The goal of this experiment is the determination of the isovector M1 decay width and the isovector E2 / isovector M1 mixing ratio for a precision test of CVC and second-class currents in the mass-8 system. Using three large NaI photon detectors placed at 40◦ , 90◦ , and 140◦ in the laboratory, in conjunction with a gas cell and the superconducting linac beam, we measured angular distributions for the 4 He(α, γ) reaction as a function of excitation energy in 8 Be. We have reported an earlier measurement2 based on the same ideas. The differential cross section is assumed to be given by an R-matrix calculation with an α-α formation channel for the 2+ doublet in 8 Be near 17 MeV and isovector and isoscalar photon transitions from the resonant states to the 2+ state near 3 MeV as well as to the tails of the resonance doublet. Alpha widths for these states are obtained from other measurements, and the isovector and isoscalar E2 and M1 transition widths are obtained from fits to our data. Since the isovector M1 transition to the 3-MeV state is dominant and is also the width we are most interested in, we characterize the other transitions in terms of their mixing ratios relative to the isovector M1 amplitude. For the transitions to the tails of the 17-MeV doublet, we have determined the relative isovector M1 amplitude from a measurement of the spectrum.3 We assume other gamma widths to the tails of the 17-MeV doublet are negligible.2 The remaining three amplitudes (isovector and isoscalar E2 and isoscalar M1) for transitions to the 3 MeV state are obtained relative to the isovector M1 amplitude from the shapes of the excitation curves over the 2+ resonance doublet. The magnitude of the isovector M1 width will be obtained from the overall normalization. Last year we reported discrepancies in the absolute efficiency of the photon detectors deter- mined from 10 B(3 He,pγ)12 C and 12 C(p,γ)13 N measurements, both of which produce photons of approximately 15 MeV. Last year we also described how we obtained energy dependent detector response line shapes from about 8 to 15 MeV. We revised some of the parameters describing these lines to get better fits, and re-determined the efficiencies. We find that for all three detectors, the efficiencies determined by repeated measurements of the same reaction were consistent. However, the efficiencies (at 15 MeV) obtained for all three detectors from the 12 C(p,γ)13 N reaction were 4.0±0.5% lower than those obtained with the 10 B(3 He,pγ)12 C reaction. Since the 12 C(p,γ) reaction cross sections were originally obtained using the 10 B(3 He,pγ) reaction to determine the detector efficiency,4 this discrepancy is puzzling. We obtain relative yields from the measurements of 4 He(α, γ)8 Be by fitting the data with a convolution of the detector response and an R-matrix calculation of the spectrum shape. Adjustable parameters are the normalization and an energy offset. The efficiencies we have determined then enable us to compare the yields in the different detectors. The detectors at 40◦ and 140◦ are not quite at symmetric angles in the cm system, but are close enough (about 37◦ and 137◦ in the cm) that the relative yields are not sensitive to the mixing ratios in the angular distributions. Thus comparing these yields is a check on the relative efficiencies of the two detectors. In one of the 1 Nuclear Physics Laboratory Annual Report, University of Washington (2000) p. 2. 2 L. DeBraeckeleer, et al., Phys. Rev. C 51, 2778 (1995). 3 Nuclear Physics Laboratory Annual Report, University of Washington (1999) p. 6. 4 R. E. Marrs, E. G. Adelberger, and K. A. Snover, Phys. Rev. C 16, 61 (1977).


  • Page 24

    UW CENPA Annual Report 2000-2001 15 Figure 1.10-1. One of two data sets with fits. The fits are described in the text. The energy axis is the actual beam energy at the entrance to the gas cell. running periods the 4 He(α, γ)8 Be results, including the measured efficiencies compare very well, but in the other running period there is a discrepancy which we are attempting to resolve. For the self-consistent run period, we fit the data from all three detectors, with the measured efficiencies applied. The free parameters were the three relative widths and a single overall normal- ization. The results are shown in Fig. 1.10-1.


  • Page 25

    16 129 1.11 Search for a permanent electric dipole moment in liquid Xe M. Ledbetter,∗ M. Romalis and M. Skoda∗ We are presently working on the development of the liquid 129 Xe EDM experiment to search for new sources of CP violation beyond the Standard Model. In the last year we have constructed and tested the apparatus needed to produce polarized liquid 129 Xe and made measurements of the transverse spin relaxation time T2 . We investigated the effects of magnetic interactions between 129 Xe atoms and found new non-linear effects in the response of the 129 Xe magnetization to a magnetic field gradient.1 We have also constructed a magnetically shielded low-field system and observed the spin precession signal of 129 Xe atoms using a high-Tc SQUID. The relaxation studies were performed in a magnetic field of 32 G. To suppress spin dephas- ing due to residual external magnetic field gradients, we used refocusing π pulses in a standard spin-echo sequence. However, spin-echo techniques do not prevent spin dephasing due to gradients created by polarized 129 Xe itself, since these gradients are also reversed by π pulses. For a uniform 129 Xe polarization in a spherical cell the magnetic field due to other 129 Xe atoms averages to zero. However, in the presence of a small gradient of the external magnetic field the magnetization of 129 Xe will develop a helix which in turn produces a gradient of the magnetic field. We found that this positive feedback mechanism causes the gradients of the magnetic field and the magnetization to grow exponentially in time and results in highly non-exponential decays of the transverse mag- netization, as shown in Fig. 1.11-2. We developed a simple model for a spherical cell to calculate the rate of exponential growth of the gradients. We also found that imperfect π pulses suppress the exponential growth of the magnetization gradients. After the exponential growth of the gra- dients is suppressed, we measured the transverse spin relaxation time T2 = 1300 sec, close to the longitudinal spin relaxation time T1 = 1800 sec. The SQUID measurements were performed in a magnetic field of 10 mG inside five-layer mag- netic shields with a shielding factor of 106 . Unlike an NMR coil, whose signal is proportional to the derivative of the magnetic field flux, the SQUID detector allows us to detect 129 Xe spin precession in a very small magnetic field with high signal-to-noise ratio. We are now beginning to study the effects of magnetic field interactions between 129 Xe atoms in this new regime of very small magnetic fields. The results of these studies will allow us to finalize the design of the actual EDM experiment. ∗ Physics Department, University of Washington, Seattle, WA 98195. 1 M. V. Romalis and M. P. Ledbetter, submitted to Phys. Rev. Lett.


  • Page 26

    UW CENPA Annual Report 2000-2001 17 15 Transverse Magnetization (µG) 10 5 0 0 500 1000 1500 2000 Time (sec) Figure 1.11-2. Envelope of the 129 Xe NMR signal obtained with a CPMG spin-echo sequence. Non-exponential decay of the magnetization due to magnetic field interactions is clearly seen. The onset of the non-exponential decay can be suppressed by using imperfect π pulses in the CPMG sequence, as indicated by the broken line, for which the length of the π pulses is reduced by 3%. This behavior is consistent with a simple pertubative model of the magnetic field interactions. Exponential fit to the later parts of the decay curve gives a value of T2 = 1300 sec.


  • Page 27

    18 1.12 f t value of the 0+ → 0+ decay of 32 Ar: a measurement of isospin breaking in a superallowed decay E. G. Adelberger, M. Bhattacharya, B. A. Brown,∗ M. W. Cooper,† A. Garcia,‡ T. Glasmacher,∗ V. Guimaraes,‡ A. Komives,‡ P. F. Mantica,∗ A. Oros,∗ J. I. Prisciandaro,∗ H. E. Swanson, S. L. Tabor† and M. Wiedeking† The precisely measured ft-values of nine 0+ ;T = 1 → 0+ ;T = 1 β-decays from 10 C to 54 Co yield a result for Vud that implies a deviation of more than 2σ from the unitarity of the CKM matrix1 . However, in order to extract Vud from the measured ft value one needs accurate theoretical calcula- tions for the nucleus-dependent isospin-breaking and radiative corrections. Given the impact of Vud on the unitarity of the CKM matrix, it is important to check predicted corrections in systems where these corrections are particularly large. We have chosen to check the calculated isospin-breaking correction for the T = 2 → T = 2 superallowed decay of 32 Ar, as the predicted correction in this system, 2.0±0.4%, is over 3 times greater than the largest isospin-breaking correction in T = 1 → T = 1 decays. We recently performed an experiment at the National Superconducting Cyclotron Laboratory at Michigan State University with a goal of determining the superallowed branch of 32 Ar β-decay with a precision of 0.4%. We directly counted the 32 Ar parents and detected the delayed proton and gamma decays with high-efficiency detectors. Mass-separated 32 Ar ions from the A1200 spec- trometer were implanted in a 500-µm thick silicon surface barrier (SSB) detector which also served as our delayed-proton counter. The 32 Ar ions were produced in the fragmentation of a 3.6 GeV 36 Ar beam on a Be target. The implantation detector was sandwiched between two similar SSB detectors which served as our beta detectors. They also helped us determine the actual number of implanted 32 Ar ions and reject fast particles in the beam. A 500-µm thick PIN Silicon counter located upstream from our telescope provided us with energy loss and time-of-flight information to identify the incoming fragments. In addition 5 high efficiency (three segmented clover units and two large single crystal) high purity germanium detectors surrounded our particle detection setup to look for βγ and βpγ decays of 32 Ar. We are in the process of analyzing the data. A preliminary analysis from our initial run indicates that it should determine ft with a precision of better than 1%. In addition we detected a number of previously unobserved gamma decay branches and obtained an improved value for the 32 Cl mass. ∗ Michigan State University, East Lansing, MI 48824. † Florida State University, Tallahassee, FL 32306. ‡ University of Notre Dame, Notre Dame, IN 46556. 1 J. C. Hardy et al., Nucl. Phys. A 509, 429 (1990).


  • Page 28

    UW CENPA Annual Report 2000-2001 19 2 Neutrino Physics 2.1 SNO 2.1.1 Status of the SNO solar neutrino analysis Q. R. Ahmad,∗ C. A. Duba, S. R. Elliott, A. A. Hamian, R. Hazama, K. M. Heeger, J. L. Orrell, R. G. H. Robertson, K. K. Schaffer, M. W. E. Smith, T. D. Steiger,† J. F. Wilkerson and the SNO collaboration The Sudbury Neutrino Observatory (SNO) is a second generation solar neutrino detector designed to measure the spectral distribution and flavor composition of the 8 B neutrinos from the Sun. Using heavy water as its primary target SNO detects the charged-current interaction, neutral- current interaction, and elastic scattering of neutrinos. This gives SNO the unique capability to measure both the flux of electron neutrinos as well as the flux of other neutrino flavors. In the first phase of the experiment SNO runs with pure D2 O to make a measurement of the charged-current interaction of neutrinos with deuterium. SNO has been online and taking production data since November 1999 while calibrating the detector and measuring the background rates at the same time. With an electronics threshold of about 0.25 photo-electron per channel and a multiplicity trigger of 18 Nhit the trigger rate is about 15 Hz and the hardware threshold corresponds to roughly 2 MeV. A number of calibrations are performed on a regular basis in SNO to measure the optical response of the detector, its energy resolution, and background rates. The optical calibration, performed using a laser source at different wavelengths and a diffuse laser ball suspended in the D2 O, shows that the timing resolution of the phototubes is near the expected design goal of about 2 ns. Data cleaning cuts are applied before reconstruction to remove any unphysical events from the raw data set. In addition, a set of high-level parameters have been developed to describe the Cerenkov nature of events. They are based on the in-time light detected in the detector and the average angle between hit PMTs. The discrimination between single Cerenkov electrons and multiple vertex events, such as β-γ’s, is used to distinguish background events and select a candidate neutrino event set. The neutrino signal loss due to these high-level cuts is less than 2%. A number of triggered and untriggered sources are used to determine SNO’s energy response. These include 16 N (6.13 MeV), a (p,T) source (19.8 MeV), a 8 Li β spectrum and neutrons (6.25 MeV). In addition, the endpoint of the 8 B spectrum itself provides a source-independent calibration of the neutrino data. As the systematics of the neutrino flux measurement is coupled to the uncertainty in the detection threshold, the goal of the SNO calibration program is to reduce the energy uncertainty to <1%. Preliminary results of SNO’s energy calibration program are shown in Fig. 2.1.1-1. The 16 N, 8 Li, and (p,t) data are in good agreement with the Monte-Carlo prediction of roughly 8-9 hits/MeV, electron equivalent. The 16 N and 8 Li sources are also used to determine the systematics associated with the event reconstruction and the angular resolution of neutrino events. The reconstruction systematics are position dependent and the resulting systematic error on the fiducial volume less than 3%. ∗ Presently at Sapient Corporation, Cambridge, MA 02142. † Presently at Cymer, Inc. San Diego, CA 92127-1712.


  • Page 29

    20 Normalized Distributions 2 N16 Data 10 Li8 Data PT Data Monte Carlo CC Monte Carlo 10 1 40 60 80 100 120 140 160 180 200 220 Number of PMT Hits Figure 2.1.1-1. Comparison of SNO calibration data from 16 N, 8 Li, and (p,t) with Monte-Carlo predictions. The predicted energy scale is roughly roughly 8-9 hits/MeV, electron equivalent. The backgrounds in the D2 O and H2 O are determined through radioassays and in-situ Cerenkov measures. The target levels for the internal Th and U backgrounds are each 7% of the Standard Solar Model equivalent in neutrons. These goals have been met and all radioactive backgrounds are at or below this target level. The extraction of the neutrino signal is performed in two different ways for independent veri- fication. One approach utilizes an essentially background free, small fiducial volume in the center of the detector while the second method extracts the different neutrino signals and background contributions for variable fiducial volumes and energy thresholds. Fig. 2.1.1-2 shows Monte-Carlo simulations of the radial distribution of neutrino signals and the primary backgrounds in the acrylic vessel (AV) and the light water. The units of the abscissa are normalized to equal volume elements and the radius of the acrylic vessel. NHITS > 45 AV CC PSUP NHITS > 65 AV CC PSUP ES ES Neutrons Neutrons 6000 6000 208 208 Tl in AV Tl in AV 214 214 Bi in H2O Bi in H2O Number of Events Number of Events 4000 4000 2000 2000 SNO Preliminary SNO Preliminary 0 0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 (R/AV)^3 (R/AV)^3 Figure 2.1.1-2. Monte-Carlo simulations of the radial distribution of neutrino signals and the primary backgrounds in the acrylic vessel (AV) and the light water. The two figures illustrate the threshold dependence of the background contributions. All elements of the solar neutrino analysis have been developed on a subset of the neutrino data. Redundant analyses are in place to verify all components of the analysis. Once the analysis components have been finalized a blind data set will be used for statistical comparisons and to establish an unbiased result. First results from the SNO experiment are to be expected in the near future.


  • Page 30

    UW CENPA Annual Report 2000-2001 21 2.1.2 Verifying event building in SNO data J. L. Orrell, K. K. Schaffer and J. F. Wilkerson The SNO data acquisition system was designed to continuously read and record information from the detector’s 9688 photomultiplier tubes (PMTs). Immediately after it is read from the hardware, information from individual PMTs is passed to the builder/recorder software, where it is assembled into “events” and recorded to disk and tape. An important verification of the hardware and the data acquisition process is to verify that PMT information is being properly associated with triggered events. When the first set of the detector trigger conditions is satisfied, two sets of data are read from the hardware. The data containing information about the trigger type and time is read out from the Master Trigger Card. Data containing information from all individual PMT tubes that were hit is read out separately from the Front End Cards (FECs). A complete “event” is the combined information from the trigger (including the time of the event) and the information from the PMT tubes. Both types of data are tagged with a Global Trigger Identification number, or GTID. The builder uses this GTID to match together all of the data that constitutes a single complete event. Two possible problems could interfere with proper building of events. First, hardware errors could attach faulty GTIDs to parts of the data; and second, event rates higher than the system’s maximum throughput could force data to be recorded faster than it can be built into events. PMT information in these conditions will still be recorded, but it will not be associated with its “parent” triggered event. Therefore, such PMT data is called “orphan” data. Surveys of the neutrino data have shown that the typical number of PMT hits that are “or- phaned” (usually due to the hardware errors mentioned above) is on the order of only a few PMT hits per 10,000 recorded. Tools have been developed to monitor the rate of orphans, as an ongoing diagnostic. Additionally, software is being developed to ensure that data from rate conditions be- yond the system throughput can be systematically rebuilt by reuniting the orphan PMT hits with the proper events.


  • Page 31

    22 2.1.3 Channel status verification for SNO data A. A. Hamian, P. Harvey,∗ M. A. Howe, N. McCauley,† C. Okada,‡ J. L. Orrell, R. Tafirout§ and J. F. Wilkerson A project to verify the integrity of a crucial aspect of the SNO data was undertaken in 2000. The goal of the work was to check the reliability of the DQXX banks. These banks are created and written to titles files by the SHaRC program1 at the start of each run. They contain channel- by-channel status information for the entire detector. This information is used in Monte Carlo simulations of SNO data, and is critical to the reliability of the simulated data. It is important to the overall SNO analysis to demonstrate that the DQXX banks correctly reflect the detector configuration for every run. The first step in this project was to compare the detector snapshots saved by SHaRC for each run with the DQXX titles files. Since the titles files are created from these same snapshots, this checks that there are no coding errors in translating information from the snapshot database to the titles files format. The results indicate that the titles files accurately reflect the SHaRC detector configuration for all runs. The second step was to compare the titles files information with the SNO database, SNODB. Since the DQXX banks are eventually incorporated into SNODB, it was necessary to verify that the banks in SNODB are consistent with those in the titles files. Of the 3708 runs that were checked, it was found that 647 runs have some discrepancy. These were traced to two sources. One is that the start time of some runs is incorrect due to a problem with the clock used to determine this information. This problem will be addressed by recreating the correct time stamp using an alternate clock, and regenerating the affected titles files. The second source of errors was that some DQXX banks do not appear in the database, because the titles file was not copied to the appropriate directory in time to be included. This will be fixed by modifying the database update procedure to require that there be no gaps in the run number sequence. SNODB will be re-generated in the near future with these fixes, which should provide 100% agreement between the DQXX titles files and the database. The final step was to compare the DQXX information with channel occupancy histograms. This was to check if channels which are “off” according to the DQXX banks indeed have zero occupancy according to the actual phototube data, and conversely, if channels which are “on” according to the banks have normal occupancies. Of the 294 neutrino runs which were checked, 44 runs showed some discrepancy between the DQXX information and the occupancy data. All discrepancies were cases where the channel is deemed “on” but has low or zero occupancy. These discrepancies are most likely channels which developed a hardware problem that hadn’t yet been discovered and addressed by the detector operators. This should be resolved once a new set of analysis banks (ANXX) is implemented, which will flag such cases. ∗ Queen’s University, Kingston, Ontario, Canada. † Oxford University, Oxford, Great Britain. ‡ Lawrence Berkeley National Laboratory, Berkeley, CA 94720. § Laurentian University, Sudbury, Ontario, Canada. 1 SNO Hardware Acquisition Realtime Control program; see Nuclear Physics Laboratory Annual Report, University of Washington (2000) p. 19 and references therein.


  • Page 32

    UW CENPA Annual Report 2000-2001 23 2.1.4 Muons and muon induced spallation neutrons in SNO Q. R. Ahmad,∗ R. Hazama, J. L. Orrell and J. F. Wilkerson, The detection of the neutral-current (NC) reaction is one of the primary goals of the SNO experiment. In order to determine the NC reaction rate accurately, one needs to account for the various forms of background properly, especially the largest source of background which comes from neutrons produced by nuclear spallation of the 16 O and 2 H nuclei in water by muons. In addition to producing neutrons, nuclear spallation products emit electrons, positrons, and γ-rays with energies similar to the solar neutrino events. If muons can be accurately identified, one can eliminate muon- induced background events from the neutrino sample by placing a simple muon-tagged time cut based on a half-life of spallation products and a capture time of neutrons in a water. The depth of the SNO detector make such a time cut feasible. The measured muon rate at SNO is approximately 2.9 hour−1 , while that at the Super-Kamiokande (Super-K) is 7.9 ×103 hour−1 , which is about 3000 times larger. However, in order to sacrifice as little live-time as possible, one needs to do a through and systematic study of all muon-induced spallation products. Muons and muon-induced spallation products are identified by the following procedures. A list of candidate muons is obtained by running the SNOMAN process called MuonID on the selected SNO data files. This is a first pass filter which provides an initial best guess for the path length of an event, which is used to categorize the prospective muons into AV (Acrylic Vessel) and non-AV going events. Any event giving a track with a path length greater than 1179 cm (a muon that grazes the AV or is tangential to it will have a path length = 1179 cm) is categorized as an AV going muon. Since all muons that go through the detector(AV) can be responsible for producing spallation events within the fiducial volume of the detector, AV-though going muons are considered. The efficiency is evaluated by hand-scanning, and the MuonID is 99.1 % efficient at identifying true muon events and only contributes ∼ 0.4% contamination to the final muon sample. The muon correlated events were identified by first selecting a through going muon and then collecting events within a 20 second time window. The standard SNOMAN time fitter (FTT) was used to reconstruct the position and direction of these events. It was shown from Monte Carlo that free neutrons in the D2 O volume are typically captured by deuterium within 500 ms of their generation. Furthermore, only ∼ 1% of the spallation products generated within 500 ms are β − γ events. Therefore, only interactions following within 500 ms of a muon were considered. The spallation neutrons after thermalization are captured on deuterium to produce a 6.25 MeV gamma. These neutrons can be utilized to perform an independent measurement of the SNO γ-ray energy calibration. Unlike deployable devices such as the SNO 16 N γ-ray calibration source, muon- induced neutrons are being constantly produced at SNO and can be compared to the calibration source without any disruption to normal data taking. In this context, Super-K is using muons that stop in the detector and then decay to produce Michel electrons and corresponding neutrinos. However, the energy is much higher (∼ 37 MeV) than that of solar neutrino events and cannot be used for an absolute energy scale, just for systematic checks. The number of µ-decay electron events at SNO is about 500 times lower than that at Super-K, so these events are not useful, but SNO can utilize the AV though-going muons for the systematic checks in addition to the 6.25 MeV gammas. ∗ Presently at Sapient Corporation, Cambridge, MA 02142.


  • Page 33

    24 2.1.5 SNO operations Q. R. Ahmad,∗ C. A. Duba, S. R. Elliott, A. A. Hamian, R. Hazama, K. M. Heeger, J. L. Orrell, R. G. H. Robertson, K. K. Schaffer, M. W. E. Smith, T. D. Steiger,† J. F. Wilkerson and the SNO collaboration In its second calendar year of production data taking in the pure heavy water phase, the SNO detector continues to meet or exceed its target goals. The experiment has been in production mode since November 2, 1999, with a total integrated livetime of 90% including calibrations. A significant fraction of the non-live time has been due to circumstances beyond the collaboration’s control, such as power outages in the INCO mine where the detector is housed. Over 98.5% of the 9688 PMT channels are fully operational, with a modest PMT death rate of 0.5% per year. The average event rate showed a decline over the course of the year, going from approximately 15-20 Hz down to approximately 10-15 Hz, reflecting the decreasing radioactive background levels. In response to these lower rates, the hardware trigger threshold was lowered in December 2000 from 18 PMTs (roughly 2 MeV) to 15 PMTs above pedestal threshold. While the rates of instrumental backgrounds such as flashers, which are light-emitting breakdowns inside a PMT, remain constant, no new instrumental backgrounds have been observed. A complete set of data cleaning cuts has been developed and implemented to remove the instrumental backgrounds. In last year’s Annual Report, two calibration sources were described.1 They were a laser source with a diffuser ball for the optical calibration, and a triggered 16 N source for an energy calibration point at 6.1 MeV, near the analysis threshold. A number of new sources were commissioned and deployed in 2000. Two untriggered 252 Cf sources were used to measure the neutron capture efficiency at low and high rates in order to predict the measured event rate due to the neutral current interaction of neutrinos with deuterium. Uranium and Thorium chain sources encapsulated in acrylic were deployed in both the heavy and light water regions of the detector to measure the shape of the U and Th backgrounds. A triggered 8 Li source which produces electrons up to roughly 16 MeV was employed to investigate the sacrifice of the data cleaning cuts, and to verify the SNO energy scale and event reconstruction algorithms. The analysis threshold for neutrino data is determined by the backgrounds in SNO. Water assays indicate that background levels in the H2 O and D2 O are near or below the design goals, with continuing improvement in water cleanliness as the D2 O is recirculated. Overall, SNO has operated very smoothly in its first phase of data-taking, and the collabora- tion looks forward to implementing the neutral current measurement phases of the experimental program. The University of Washington group continues to be actively involved in the daily oper- ations of the detector, providing manpower for regular monitoring shifts as well as for maintenance activities and general detector support. ∗ Presently at Sapient Corporation, Cambridge, MA 02142. † Presently at Cymer, Inc. San Diego, CA 92127-1712. 1 Nuclear Physics Laboratory Annual Report, University of Washington (2000) p. 18.


  • Page 34

    UW CENPA Annual Report 2000-2001 25 2.1.6 The SNO data acquisition system Q. R. Ahmad,∗ C. A. Duba, A. A. Hamian, P. Harvey,† K. M. Heeger, M. A. Howe, R. Meijer Drees,‡ A. W. Myers, J. Roberts,§ P. M. Thornewell,¶ T. D. Van Wechel and J. F. Wilkerson The SNO data acquisition system has been described extensively in previous reports.1 Only a summary of operations and significant updates is given here. In its second calendar year of production data taking, the SNO data acquisition (DAQ) system continued to perform in a stable and reliable manner. Relative to previous years where various parts of the DAQ system required one or more major upgrades, there were very few changes to the system in 2000. The only significant new feature added to the SHaRC2 program was a series of system checks which compares key aspects of the detector configuration with a standard configuration at the start of each neutrino run and reports any anomalies. A modification to the event builder code, along with corresponding updates to the embedded processor code and the SHaRC program, has been shown to reduce the residency time of SNO events in the builder queue due to an improved building algorithm. This update has been tested extensively, and will be put into production use in late April, 2001. In addition to the above enhancements, there were a number of minor (mostly cosmetic) changes to SHaRC. Several memory leaks were found and fixed, which has improved the system’s reliability. The overall stability of the DAQ system has been remarkably good, with an average of 1 or less SHaRC crashes per month. The DAQ group has continued to provide continuous off-site support, and regular periods of on-site support during significant upgrades. ∗ Presently at Sapient Corporation, Cambridge, MA 02142. † Presently at Queen’s University, Kingston, Ontario, Canada. ‡ 8140 Lakefield Drive, Burnaby, British Columbia, Canada. § Presently at Sudbury Neutrino Observatory, Sudbury, Ontario, Canada. ¶ Presently at F5 Networks Inc., Seattle, WA. 1 Nuclear Physics Laboratory Annual Report, University of Washington (2000) p. 19 and references therein. 2 SNO Hardware Acquisition and Realtime Control.


  • Page 35

    26 2.1.7 Time stamp validation in the SNO experiment J. L. Orrell and J. F. Wilkerson The Sudbury Neutrino Observatory (SNO) is a “real time” detector. The SNO detector is capable of achieving 20 nanosecond resolution of event trigger times. The photo-multiplier tubes used can achieve approximately 100 picosecond resolution within an event. This nanosecond by nanosecond data collection allows for the reconstruction of physics events. The data acquisition and record keeping must deal with time records spanning 17 orders of magnitude. Of primary importance for proper neutrino analysis, the accurate determination of a physics event’s time of occurrence is needed. Additionally, the SNO scientists must be able to communicate accurate event times to the physics community at large. This is particularly germane to the SuperNova Early Warning System (SNEWS) project.1 The SNO detector uses three separate time keeping devices in conjunction to produce event time stamps. There are two clocks located at the detector site which are used for moment to moment data collection. These clocks are referred to as the 10 MHz and 50 MHz clocks because of the rate at which they “tick”. The experiment also utilizes the Global Positioning System (GPS) as an absolute time reference. The 10 MHz clock is periodically synchronized with GPS time and, thus, is the time standard used to determine absolute event times. The 50 MHz clock serves as the relative time basis for event-to-event timing. That is, detector global triggers are latched on the 20 nanosecond ticks of the 50 MHz clock. The 50 MHz clock can also act as a secondary time standard. A secondary time standard is needed in situations where the 10 MHz clock has failed, the GPS communication is down or unavailable, or there are other uncertainties in the validity of the time stamps supplied by the 10 MHz clock. However, since the 50 MHz clock is not synchronized to GPS time we need to ensure we know the actual rate at which the supposed 50 MHz clock is ticking. It is possible to use a long period of detector data taking to determine the actual rate of the 50 MHz clock. We select events during periods known to have valid GPS synchronizations. For a given period we compare the elapsed times of the 50 MHz clock to that of the 10 MHz (i.e. GPS verified) clock. For each period a linear fit through the origin is used to determine difference between 50 MHz and the actual rate of the 50 MHz clock. It is then trivial to convert to the actual rate. From 27 periods (of durations between 2 and 45 days) we have determined the 50 MHz clock’s actual rate: R50 = 49999462.68 ± 0.87(Hz). (1) A weighted mean and error are used in this calculation because the measurement intervals are not uniform. This information will help ensure that accurate timing information is being used in neutrino data analysis. 1 SNEWS: http://hep.bu.edu/∼snnet/


  • Page 36

    UW CENPA Annual Report 2000-2001 27 2.1.8 Variation of neutrino flux with (local) solar position S. R. Elliott and M. W. E. Smith The solar neutrino flux is expected to vary with solar position, as measured with respect to a local set of co-ordinates: θ⊙ (zenith angle), φ⊙ (azimuth), and r⊙ (Earth-Sun distance). These variables are being studied with the Sudbury Neutrino Observatory (SNO). The coordinate r⊙ varies seasonally, with the Earth making its closest approach to the Sun (perhelion) in early January and furthest (apehelion) in early July. With this comes a modulation in the solar neutrino flux that goes as 1/r 2 . This can be seen in the simulated SNO charge current (CC) event rate shown in Fig. 2.1.8-1. In addition to this 1/r 2 behavior, there may be a variation of the neutrino species due to neutrino oscillation. The CC reaction is sensitive only to electron neutrinos, so that this additional modulation may show up in this reaction. Finally, assuming a no-oscillation hypothesis, the month-to-month behavior of the SNO event rate will provide a systematic check of detector stability. The zenith angle is a measure of an object’s location with respect to the horizon; θ⊙ = 0 for objects directly overhead and θ⊙ = π for objects directly underfoot. For θ⊙ < 0, the solar neutrinos are coming from below the horizon, and hence are passing through substantial Earth material. Under certain oscillation scenarios, this can lead to an enhancement of the electron neutrino flux, known as the Earth-matter or day-night effect. The path length through the Earth goes as: L = −2R⊕ cosθ⊙ θ⊙ < 0 where R⊕ is the Earth radius. For many regions of Mikheyev-Smirnov-Wolfenstein (MSW) pa- rameter space, the amount of regeneration is proportional to L. For other regions, the amount of regeneration also depends on the composition of the material. Thus, a higher regeneration is possible when the sun is behind the Earth and the neutrinos pass through the Earth’s outer core, where the electron density is higher than the mantle. Fig. 2.1.8-2 shows the Earth density profile being used to model the Earth-matter effect. Under the no-oscillation hypothesis, the neutrino flux will be unaffected. 1.06 Event rate relative to 1 AU 1.04 1.02 1.00 0.98 0.96 Monte Carlo 0.94 Expected 2 4 6 8 10 12 Month Figure 2.1.8-1. Simulated CC reaction in SNO. The high statistics of this simulation shows the expected seasonal modulation caused by the eccentricity of the Earth’s orbit. The rate is shown relative to that at 1 AU.


  • Page 37

    28 14 max. depth reached 12 by SNO neutrinos inner core density (g/cm ) 10 3 outer core mantle + crust 8 6 4 2 0 0 1000 2000 3000 4000 5000 6000 7000 radius (km) Figure 2.1.8-2. Density profile as a function of radius from Earth center. Due to SNO’s latitude, only the region to the right of the dashed line is accessible.


  • Page 38

    UW CENPA Annual Report 2000-2001 29 2.1.9 Neutrino events with solar flares revisited R. Hazama and R. G. H. Robertson Now we are in the maximum solar activity during solar cycle 23, and intense solar flares occurred in July and November 2000, just 11 years after big events in October 1989. In particular, the proton flux (≥ 10 MeV) as measured by the GOES satellite reached a maximum of ∼ 24000 p.f.u. (= protons cm−2 s−1 sr−1 ) on July 15, 2000 1230UT. This flux is about one order higher than that in March 1989 (and April 1984), which is the search period of the Kamiokande report,1 and almost equivalent to flux on August 4, 1972 0617UT. That is when the largest energetic proton event occurred during solar cycle 20, with the maximum proton flux (≥ 10 MeV) of 25000 p.f.u., and this is just the time of the first reported excess neutrino capture rate in the Homestake (37 Cl) detector. No significant neutrino signal was found by Kamiokande for the period of July 1983 - July 1988 and in March 1989, giving a limit on the time-integrated solar-flare νe flux. The sensitivity of SNO is presented in Fig. 2.1.9-1. SNO has a good sensitivity for low energy neutrinos. One possible loophole the Kamiokande result cannot rule out is that the 37 Ar excess in the chlorine detector is caused by very-low-energy neutrinos, e.g., those from positron emitters produced by flare particles in the Earth’s atmosphere. Furthermore SNO is the most northerly underground neutrino detector in the world and the geomagnetic field guides cosmic rays into the earth’s atmosphere. Thus, the flux of low energy neutrinos will be larger at SNO than at more southerly sites (Super-Kamiokande). Now we are reanalyzing the production rate of low energy neutrinos via cosmogenic radioisotopes produced by the interplanetary protons. Especially β − emitters are important in SNO, because we can identify a ν¯e by utilizing detection of two neutrons in coincidence with a positron. Figure 2.1.9-1. The sensitivity of the SNO detector to the solar-flare νe flux plotted as a function of νe energy, corresponding to one neutrino event in the detector. The dashed line is for the Kamiokande detector. 1 Kamiokande Collaboration, K. S. Hirata et al., Phys. Rev. Lett. 61, 2653 (1988); ibid. 62, 694 (1989); Astrophys. J. 359, 574 (1990).


  • Page 39

    30 2.2 SNO/NCD 2.2.1 Neutral current detectors in SNO J. F. Amsbaugh, T. V. Bullard, T. H. Burritt, G. A. Cox, P. J. Doe, C. A. Duba, S. R. Elliott, R. M. Fardon,∗ J. E. Franklin,† R. Hazama, K. M. Heeger, K. M. Kazkaz, A. W. Myers, R. G. H. Robertson, K. K. Schaffer, M. W. E. Smith, T. D. Steiger,‡ T. D. Van Wechel and J. F. Wilkerson In the third phase of the SNO experiment, an array of 3 He proportional counters will be installed within the D2 O volume. These Neutral Current Detectors (NCDs) will allow direct, real-time detection of the neutrons produced in neutrino neutral current reactions with D2 O, facilitating a better measurement of the total active neutrino flux from the sun as well as a more precise measurement of the shape of the charged-current spectrum. When neutrons enter the NCDs, they capture on the 3 He via: n + 3 He −→ p + 3 H + 764 keV. The charged particles produced in this reaction ionize the gas within the NCD, resulting in a pulse signal that travels down a central anode wire to be read and processed by the NCD electronics. In order to minimize radioactive backgrounds, extremely high purity materials were required in the production of the NCDs. The detector bodies are made from chemical-vapor-deposited (CVD) nickel, with specially designed CVD endcaps and quartz high-voltage feedthroughs. The high purity materials were chosen to specifically minimize contamination by uranium and thorium. Cosmogenic 56 Co in the nickel detector walls could also cause backgrounds, so the NCDs will be stored underground before deployment, allowing the 56 Co to decay according to its 79 day halflife. This “cool down” period will also be used for measuring Uranium and Thorium backgrounds and characterizing the NCD array. As an additional measurement of backgrounds from the NCD materials, a device built from the same materials (the CHIME calibration source) was successfully deployed in SNO during August, 2000. There are currently 262 NCDs stored underground on site. Ten completed NCDs await shipment to site, and 21 more remain at UW, requiring repair or gas fill. The data acquisition and analysis tools for the NCD array continue to be developed and tested. The NCD electronics and deployment equipment are being completed, and the NCDs are expected to be installed in Spring or Summer of 2002. ∗ Physics Department, University of Washington, Seattle, WA 98195. † Retired. ‡ Presently at Cymer Inc. San Diego, CA 92127-1712.


  • Page 40

    UW CENPA Annual Report 2000-2001 31 2.2.2 NCD data taking and analysis T. V. Bullard, S. R. Elliott, K. M. Heeger, R. G. H. Robertson and M. W. E. Smith Over the last year, the analysis of NCD data has been ramping up in preparation for a larger influx of data. With the complete electronics system nearly ready for installation, and with 87% of the array in its “cooldown” phase at SNO, the amount of data that we will be taking is going to drastically increase. The main push in our analysis ramp up has been the development and verification of software tools to be used in analyzing the data. These tools include low and intermediate level filters, rate counters, an ion tail removal routine, and a high level pulse shape fitter. In addition, the tools already in place have been packaged into a software program called Analyst assembled from sequential modules. Furthermore, the chain of required analysis routines has been outlined in order to direct further developments. In addition to software development, there have also been two data analysis studies in progress this year. The first of these was the “Water Wall” study. In this experiment, data was taken from a few NCD strings that were set up in a water enclosure underground at SNO to measure and compare the thermal and fast components of the neutron flux. These data may also be used for the development of neutron identification and data analysis routines. A second project involved the assembly of two short counters with high levels of 210 Po con- tamination. One counter has the contamination in the end cap region, while the other has the contamination in the mid-body of the counter. The goal of this project is to characterize the risetime vs energy distributions of alphas originating from the end cap region where the electric field is weak. These events might mimic neutron events that would otherwise be distinct from the distributions of bulk and surface alphas. This data has recently been taken and analysis of the data will be carried out once the required tools are in place. Analysis effort has also gone into determining counter rates from recent, as well as past, cooldown data. The average gas event rate for counters that were in the water wall was about 11 events/hour, while that for counters outside the water wall was about 29 events/hour. The individual counter rates, along with other relevant information, have been stored in a database created for pre-deployment data. The backgrounds will provide part of the information needed to assess readiness for deployment.


  • Page 41

    32 2.2.3 NCD electronics C. A. Duba, A. W. Myers, R. G. H. Robertson, T. D. Van Wechel and J. F. Wilkerson The Sudbury Neutrino Observatory (SNO) provides a unique window into both the neutral and charge current neutrino flux through the use of heavy water. SNO needs to differentiate between neutrons and other events in order to distinguish potential neutral-current neutrino flux. The Neutral Current Detector (NCD) array is set to be placed within SNO in one year, and has the potential to provide extremely accurate neutron recognition. The low expected rate of neutral current events generated by solar neutrinos necessitates event-by-event recognition while the high neutron rate from galactic supernovae requires a fast data acquisition system. Pulser HP 33120A Multiplexer (12 NCDs per) Fan In/Out Function threshold Generator Current Preamp Log Amp Summing Tektronix Junction (1 of 8) TDS 754A (1 of 8) Oscilloscope HV (1 of 4) Distribution GPIB panel Tektronix Offset Voltage DACs Delay Box TDS 754A HV (8 16-ch (~300ns) Oscilloscope Supplies DACs (1 of 96) Analog Thresholds & ADCs) HV HV Relays Digital Control Interface MUX Alternating Scope Trigger SNO MTC-D Controller DACs PulseGT NCD Card HV (8 16-ch Controller Synclr16,24 Trigger DACs Card & ADCs) IP-408 Shaper/ADC NCD GTID IP-408 VME 32 Bit Board Counter 32 Bit VME GBIP digital (8 NCD inputs) (modified digital ECPU Controller I/O (1 of 12 boards) Shaper/ADC) I/O VME Bus SBS 618/620 Optical VME Controller NCD DAQ Figure 2.2.3-1. The NCD Electronics control diagram. The NCD electronics were designed with both of these goals in mind. Dual data paths provide the low-rate scenario with digitized data while allowing for high-rate data taking. Each string of NCD counters has a pair of independent thresholds, one level for low rate pulse digitization and one for high rate signal integration. The NCD electronics also provide 8 individually controllable high voltage power supplies for selective distribution amongst the NCDs 96 strings.


  • Page 42

    UW CENPA Annual Report 2000-2001 33 All signals that come from charge deposited on the NCD center anode are run through a preamplifier, which separates out the current pulse from the high DC voltage. The pulse is carried into one of eight multiplexing (MUX) boxes, where it is split into the fast and slow data paths. The fast data path carries the signal through a MUX box and a shaper into an ADC, which takes the integrated signal and converts it to a digital charge value. If the event has sufficient peak current to trigger the digitizing threshold in the MUX box, a delayed signal will be latched through a log amp and into both digitizers. The ADC information passes through a VME bus to a SBS 618 VME controller. The SBS 618 connects to the controlling computer with a fiber-optic line. In parallel, an event that triggers digitization results in the MUX controller placing the hit information onto a line of the 32bit differential I/O module while triggering the appropriate digitizer. The computer grabs the hit information through VME, and requests the digitized event from the digitizer through GPIB. When the system is complete and installed into SNO, the trigger pulse will be copied to the Global Trigger I.D. (GTID) counter. The GTID counter will request a SNO GTID from the SNO Master Trigger Card Digital (MTCD). The computer will grab the GTID information and tie this value to any events coming in at this time. In the final system, an independent Embedded CPU (ECPU) will be placed in the VME crate to control the electronics, so that ephemeral computer functions will not interfere with electronics performance. The ECPU control will also increase the total data acquisition speed of the system by sharing some of the event organization and packing duties with the control computer. The high voltage distribution system is controlled by a set of hardware very similar in design to the threshold control hardware in the MUX control. There are 8 high voltage Spellman 3kV power supplies that connect to a high voltage distribution panel. The power supplies can be selectively connected to differing numbers of strings, and each individual power supply has sufficient current and stability to power the entire NCD array. The digital to analog converter board (DACs) is used in the high voltage system to set the NCD anode voltages to within a few volts. The high voltage control board operates the HV DACs board and regulates the power to the Spellman supplies. The HV safety system prevents the hardware from damaging the NCDs through a number of failsafe mechanisms.


  • Page 43

    34 2.2.4 Overview and status of the NCD DAQ software G. A. Cox, M. A. Howe, F. McGirt∗ and J. F. Wilkerson The Neutral Current Detector data acquisition (NCDDAQ) software is taking shape as more and more of the finalized electronics become available. NCDDAQ will provide an easy-to-use graphical interface to the NCD system at the Sudbury Neutrino Observatory (SNO). It is written in C++ and is compiled with the MetroWerks compiler for running on Macintosh computers. NCDDAQ uses much of the code base of the successful SHaRC DAQ control software1 that has been in routine operation at SNO for several years. The main dialog of the NCDDAQ interface provides the operator with a complete overview of the system’s current status. All of the various parts of the detector are shown in a layout that matches the electronic schematic of the system. By clicking on the various elements of the picture, dialogs for controlling that part of the system are displayed. Most of the major elements that are needed for the final system are well developed, including system initialization, a database for storing electronics constants, run control, high-voltage control, and an alarm system. The following figure shows NCDDAQ in use. Figure 2.2.4-1. A screen dump of the NCD DAQ in use. One of the most important things that NCDDAQ provides is a means for conveniently initializing the system. For this purpose, there is a Hardware Wizard that allows an operator to program in a highly selectable set of hardware parameters. The Hardware Wizard is closely coupled to a database that holds all electronic constants. A well-developed run control system is in place for testing the current system. As well as Los Alamos National Laboratory, Los Alamos, NM 87545. ∗ 1 Nuclear Physics Laboratory Annual Report, University of Washington (1997) pp. 20-23; (1998) pp. 18-20; (1999) pp. 16-18.


  • Page 44

    UW CENPA Annual Report 2000-2001 35 providing a one-button start run capability, it also provides controls for setting the length of a run and whether a run is to be repeated. In this developmental phase of the hardware/software the run control also provides controls for taking data with sub-parts of the system, i.e. just data from the digital scopes. In addition, all actions are logged to a status file on a run-by-run basis. A high-voltage control and safety system is built into NCDDAQ. HV supplies can be ramped using custom ramping profiles. Each supply can be controlled separately or as part of a group. The HV supplies are constantly monitored to protect the NCD tubes at all times. The monitoring includes checks to verify that the set voltage matches the read-back voltage, the relays are set to the proper state, and whether the electronic racks are on and communicating. In the event of any problems or inconsistencies, alarms are posted to alert the user, and if necessary, the software enters a safe mode that may ramp the HV to zero in certain extreme conditions. Also the operator can panic ramp the high voltage back to zero at any time for a particular supply, or for the entire detector. The state of the HV system is displayed on all hardware-display dialogs. An alarm system is in place for alerting the operators to unusual events. Each alarm can provide detailed help on what the problem is and how to solve it. So far, most of the defined alarms are for the HV system. Other alarms are being added as needed. The final data path is now under development. In this phase, the actual data taking will be moved off of the Mac and into an embedded eCPU running in the VME crate. The eCPU will read the hardware as needed, do some initial event building, and put the data into a dual-port memory where it can be grabbed by the Mac and dispatched to the analysis and event-by-event monitoring tools.


  • Page 45

    36 2.2.5 First Deployment of a Neutral Current Detector in the Sudbury Neutrino Observatory: The CHIME Engineering Run J. F. Amsbaugh, T. H. Burritt, P. J. Doe, C. A. Duba, S. R. Elliott, G. C. Harper, K. M. Heeger, G. Miller,∗ A. W. P. Poon,† R. G. H. Robertson and M. W. E. Smith The use of ultra-low background 3 He proportional counters as neutron detectors in the SNO detector will provide a means of measuring the neutral-current interaction rate of solar neutrinos. Since the binding energy of the deuteron is only 2.2 MeV gamma rays from natural decay chains in the proportional counters photodisintegrate the deuteron and thus simulate the neutral current signal. This poses stringent limits on the radiopurity requirements for the counters. An in-situ background test was performed in the heavy water volume of the SNO detector using seven Neutral Current Detectors. The goal of this measurement was to look for any unexpected backgrounds from the production counters used in the Neutral Current Detector array. The Neutral Current Detector (NCD) array consists of 775 m of 3 He proportional counters arranged in 96 strings with 300 counters. The counter bodies are made of about 450 kg of ultra-pure CVD nickel which contains natural Uranium and Thorium in equilibrium. The photodisintegration gammas are produced in the decays of the 232 Th (208 Tl → 208 Pb, Eγ =2.615 MeV) and 238 U (214 Bi → 214 Po, Eγ =2.445 MeV) chains and in the decay of the relatively long-lived 56 Co (56 Co → 56 Fe, Eγ >2.224 MeV). A Construction Hardware In Situ Monitoring Experiment (CHIME) has been designed to mea- sure the NCD originated background in the presence of the D2 O contribution prior to the deploy- ment of the entire NCD array. It is an in-situ measurement of the construction materials used in the NCD array. Seven individual counters arranged in a close-packed configuration with a total mass of about 5000 g and an overall length of 112 cm form a background test source that can be deployed in the SNO detector. The construction materials and procedures for the CHIME counters are essentially identical to those in the real NCD array. To assure its cleanliness before deployment the CHIME unit was checked for leachable Radon and surface dust deposition. Extensive tape lift studies of the CHIME and the Neutral Current Detector array in underground storage were made to estimate the amount of surface contaminants that will be introduced into the heavy water during the installation of the NCD array. The CHIME has been stored underground since December 20, 1998 in order to allow the cos- mogenically produced 56 Co to decay. In September 2000 the CHIME was deployed into the inner volume of the SNO detector. The CHIME is negative buoyant and was deployed along the central axis of SNO using a specially designed deployment mechanism with an ultra-clean Vectran line. During a 71.8 hrs long commissioning run no sign of unexpected backgrounds was detected. In principle, this verifies the successful execution of the cleanliness procedures during NCD construc- tion. A detailed analysis of the CHIME engineering run is in progress. A longer CHIME run is required to determine the activity of 232 Th in the Neutral Current Detectors and thus determine an upper limit on the photodisintegration background generated by the radioactivity in the NCD array. ∗ Los Alamos National Laboratory, Los Alamos, NM 87545. † Lawrence Berkeley National Laboratory, 1 Cyclotron Road, Berkeley, CA 94720.


  • Page 46

    UW CENPA Annual Report 2000-2001 37 2.2.6 The NCD laser welding equipment J. F. Amsbaugh, T. A. Burritt, P. J. Doe, B. Morissette∗ and T. D. Van Wechel We have completed construction and testing of the laser welding equipment needed for neutral current detector (NCD) deployment1 in the Sudbury Neutrino Observatory (SNO) detector. The NCD welding will occur in two stages with the welding fixture (WF) and its controller used for both. During the first stage, pre-deployment, individual detectors are welded into the largest segments that can fit into the room above the SNO detector. For this operation, the WF is mounted horizontally on a bench supported rail, with several roller–wheel carriages supporting the NCDs for insertion into the WF. During the second stage, deployment, these segments are welded together as they are inserted into the SNO detector. Here the WF is mounted vertically on a linear rotary shaft so it can be raised, lowered and swung out of the way during the welding and insertion procedures. The WF uses inflating cuffs to grip two NCD tubes or one tube and an end component. One tube which can move opens a gap for center conductor connection. The WF either twists the NCD tubes or orbits the laser output head when welding in the vertical configuration. The output head mount rotates to relieve fiber cable stress of the orbiting mode and slides with a tube follower to maintain laser focus on an out-of-round NCD tube. A motor and linkage assembly provides the relative rotation and locks tube rotation end to end. A He leak checking station is at one Figure 2.2.6-1. The weld fixture in use during March 2001 safety measurements. Point of view is from lower side opposite the laser-head. The WF is configured for pre deployment welding and mounted on similar rail for tests. end. The WF opens up for cleaning weld dust and changing the laser output lens protective cover. ∗ Sudbury Neutrino Observatory, Lively, Ontario, Canada P3Y 1M3. 1 Nuclear Physics Laboratory Annual Report, University of Washington (2000) p. 27.


  • Page 47

    38 Microswitches on the WF indicate door closure, motor engagement, and slide holdout release for the interlock system. The welding control provides safety and process interlocks, manifolds for cuff and cover gases, motor controls, a leak test volume pressure gauge, status indicators and the laser remote control panel. The interlock signal voltages to and from the laser are isolated by a master interlock relay from the weld control interlock voltage system. The primary laser safety is a shutter at the input to the fibre optic output cable and all laser, weld control, and process interlocks must be met before it will open. The operator can be seen adjusting laser parameters on the weld control in Fig. 2.2.6-1. The WF access door is open and two short NCD sample tubes are inserted. A roller–wheel carriage is in the foreground near the leak check station. The right angle weld view camera installed in place of the alignment eyepiece on the laser fibre output housing is seen on the left in Fig. 2.2.6-2. This housing is the middle black cube with the stainless steel fibre termination coming up from below. The welding laser2 delivers 1 kW peak optical power in 8.5 ms long pulses at 3 Hz pulse repetition rate. The WF design prevents this laser emission from escaping and measurements indicate that the accessible exposure level is below the Class I laser device limit. Air samples of the welding dust were taken and are below the permissible exposure limit. Figure 2.2.6-2. The welding fixture during laser emission measurements. Point of view is looking down along the NCD axis. 2 Model LuxStar-50, Lumionics Ltd., Warwickshire CV21 1QN, England.


  • Page 48

    UW CENPA Annual Report 2000-2001 39 2.2.7 NCD deployment equipment progress J. F. Amsbaugh, M. Anaya,∗ T. A. Burritt, P. J. Doe, G. C. Harper, G. Miller,∗ J. Wilhelmy∗ and J. Wouters∗ Development continues on the equipment needed to deploy the neutral current detectors (NCDs) into the heavy water acrylic vessel (AV) of the Sudbury Neutrino Detector (SNO). Also many suggestions and problems indentified by the design and procedure review1 have been implemented. This progress includes a revised glove box glove mount, a laser welding fixture, a counter balanced fixture mount, a predeployment welding bench, a redesigned hauldown mechanism, and a gantry crane. During the NCD deployment, the calibration glove box is replaced with the deployment plate. The plate has several 8–inch ports for gloves, view ports, covers and cameras. The new glove mount allows easier exchange of gloves and cover ports on the deployment plate. One of these interchangable covers will be modified to mount the neck view video camera to be used to thread the NCD cables into the correct pre-amplifier bulkhead feed-through. All components have been recieved but the housing design is not yet finalized. The bench supporting the laser welding fixture and NCDs horizontally for predeployment welding was completed and tested in Nov. 2000. Several improvements and changes were added. The laser welding fixture is described in Section 2.2.6 of this report. The NCD anchor is engaged in a float that the hauldown mechanism manually pulls down with a 1/16-inch vectran line. The remotely operated vessel (ROV) can then take the NCD string from the float to its correct attachment point. The mechanism then pulls the float back up to the deployment plate for the next NCD. The minumum vectran length is 3 times the AV depth, about 175 feet in total, and before installation it must be stored on the takeup spools. The new hauldown mechanism has greater vectran line capacity and improved user controls. The friction brake which locked the elevation and impeded motion has been replaced with two ratchet assemblies. This prevents accidental NCD movement if the operator releases one of the cranks during haul up or haul down. With both engaged the float is locked in position. A full NCD string can be as much as 50 lbf bouyant so a gear train reduces the required force an operator needs to apply to haul it down. The new mechanism is finished and was shipped to the test pool at Los Alamos National Laboratory. The gantry crane is a commerical2 all aluminum construction A-frame unit with 2000-lb ca- pacity. The cross beam has been hard anodized and the fixed turning pulley for the 5/16-inch vectran rope is all stainless steel as is its mounting hardware. The vectran rope is wound on an all aluminum manual winch3 with automatic brake and enclosed gear train. Finally, the casters were replaced with clean room compatible ones. Preventing contamination of the heavy water and damage to the acrylic vessel are important considerations. To address this, radon and leaching tests on the ROV and other equipment that will enter the AV or heavy water are planned. Clean laboratory space to do such tests were damaged by the recent fire in the Los Alamos area delaying the tests. The leach tank calibration tests have begun and the equipment tests should be completed soon. ∗ Los Alamos National Laboratory, Los Alamos, NM 87545. 1 Nuclear Physics Laboratory Annual Report, University of Washington (2000) p. 26. 2 Model 1ALU1208B, Spanco, Morgantown, PA 19543. 3 Model CMA-1760, Jeamar Winches, Inc., Buffalo, NY 14206.


  • Page 49

    40 2.3 Lead as a neutrino detector 2.3.1 Lead perchlorate as a neutrino detection medium M. K. Bacrania, P. J. Doe, S. R. Elliott and L. C. Stonehill Due to its apparent transparency, large interaction cross section and relatively low cost, lead per- chlorate Pb(ClO4 )2 is an attractive candidate for a Cerenkov type neutrino detector. Neutrino interactions with lead may occur by either charged current (CC) or neutral current (NC) reactions: νe +208 Pb −→ 208 Bi∗ + e− (CC) ↓ 208−X Bi + Nγ + Xn ′ νx +208 Pb −→ 208 Pb∗ + νx (N C) ↓ 208−X Pb + Nγ + Xn At 30 MeV the CC cross section with lead is about 600 times that of carbon and the NC cross section is about 100 times that of carbon. The signature of a CC interaction consists of a prompt electron followed by gamma rays and neutrons. The signature for NC interactions consists only of gamma rays and neutrons with no prompt electron. The number of neutrons (0, 1, or 2) depends upon the energy of the interacting neutrino. Because lead perchlorate solutions contain an appreciable amount of hydrogen, the neutrons quickly thermalize in the solution and are captured by the 35 Cl which then emits an 8.6 MeV gamma ray. The gamma ray is detected by subsequent Compton scattered electrons. This is the same reaction that will be utilized by the SNO Cerenkov detector to identify neutrons with high efficiency. To determine if a lead perchlorate Cerenkov detector can be built we have investigated the optical properties of the solution. Studies using a spectrophotometer revealed that there were no obvious absorbtion lines between wavelengths of 250 to 600nm. We constructed a special apparatus to measure the attenuation of 460 nm light in various strength solutions. Initial measurements yielded attenuation lengths of less than half a meter, which is insufficient to build a large lead perchlorate Cerenkov neutrino detector. Diluting the solution further reduced the attenuation length, suggesting that light loss may be due to the formation of Pb salts which scatter the light. To remove scattering particles we filtered the lead perchlorate solution using a series of filter pore sizes from 5.0 microns down to 0.2 microns. This resulted in the attenuation curve shown in Fig. 2. 3.1-1. This improvement is sufficient for a reasonably sized Cerenkov detector and suggests that additional filtering might further increase the attenuation length. Currently we are characterizing the Cerenkov light production of various strengths of lead perchlorate solutions. Once the production and attenuation of light has been understood and quantified we will be in a position to make meaningful assessment of the physics opportunities presented by this exciting new detector.


  • Page 50

    UW CENPA Annual Report 2000-2001 41 15 PMT output (mV) 10 Filtered 70% Solution Measurements 5 Fit 0 0 10 20 30 40 50 60 Column Height (cm) Figure 2.3.1-1. Attenuation of 460nm light in a 70% lead perchlorate solution. The attenuation length is approximately 4.3m.

  • View More

Get the full picture and Receive alerts on lawsuits, news articles, publications and more!