avatar University of Washington Inc Services
  • Location: Washington 
  • Founded:
  • Website:

Pages

  • Page 1

    Center for Experimental Nuclear Physics ref 0.3 ∆ρ / ρ and 0.2 Astrophysics 0.1 0 University of Washington φ 4 ∆ 3 2 1 1 1.5 2 0 0.5 η∆ −1 −1−0.5 0 −2 −1.5 Annual Report 2009


  • Page 2

    ANNUAL REPORT Center for Experimental Nuclear Physics and Astrophysics University of Washington May, 2009 Sponsored in part by the United States Department of Energy under Grant #DE-FG02-97ER41020.


  • Page 3

    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, makes 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. The upper, right cover figure is from an analysis of semi-peripheral 200 GeV Au-Au collisions measured in the STAR detector at RHIC. It shows the two-particle correlation strength for particles separated in azimuth by φ∆ and in psuedo-rapidity by η∆ . There is a prominent same-side peak centered at (η∆, φ∆ ) = (0, 0) and an away-side ridge at φ∆ = π. Those two features represent pairs of fragments (jets) from large-angle scattering of partons. There is also a cos(2φ∆ ) structure which has a small amplitude for this centrality and is convention- ally associated with elliptic flow. A close look also shows a small ridge at η∆ = 0 due to longitudinal fragmentation of projectile nucleons. Figure provided by Duncan Prindle. The left cover photo shows Seth Hoedl and Frank Fleischer next to the axion-search torsion balance. The experiment is designed to look for axions by trying to measure the macroscopic force mediated by them. The right, center photo is of Alexis Schubert installing the first functional detector into a mechanical, electrical, and thermal test cryostat for MAJORANA at LANL. The right, bottom photo shows Carin Schlimmer assembling her experiment to measure gravity gradients. Photos by Vic Gehman and Greg Harper.


  • Page 4

    UW CENPA Annual Report 2008-2009 May 2009 i INTRODUCTION The Center for Experimental Nuclear Physics and Astrophysics, CENPA, was established in 1998 at the University of Washington as the institutional home for a broad program of research in nuclear physics, astrophysics and related fields. Research activities are conducted locally and at remote sites. CENPA has been a major participant in the Sudbury Neutrino Observatory (SNO) and is presently a major participant in the KATRIN tritium beta de- cay experiment and the MAJORANA double beta decay experiment. Most recently, CENPA physicists have joined the SNO+ double beta decay experiment in Canada. The current program includes “in-house” research on nuclear astrophysics and fundamental interactions using the local tandem Van de Graaff, as well as local and remote non-accelerator research on fundamental interactions and user-mode research on relativistic heavy ions at the Relativistic Heavy Ion Collider at Brookhaven. The 2008-9 period has seen dramatic change at CENPA. Assistant Professor Nikolai Tolich, who joined the faculty from Lawrence Berkeley Lab in the fall of 2007, has started a research program with SNO+. Professor John Wilkerson, a faculty member since 1994, accepted a position at the University of North Carolina and departed in January 2009. A search is underway for a faculty member to replace John. Our first CENPA Fellow, Research Assistant Professor Michael Miller, joined us from MIT in November and is leading our MAJORANA project. Professor Eric Adelberger, and Research Professors Kurt Snover and Derek Storm all retired within the past two years, but they remain active in research as Emeritus faculty. Derek served as the Director of the Nuclear Physics Laboratory (CENPA’s predecessor) and Executive Director of CENPA for more than two decades. We take this opportunity to thank him for his exceptional and dedicated leadership. Professor Hamish Robertson has taken over as Director of CENPA, and Greg Harper, Senior Research Engineer for many years, has accepted the position of Associate Director. This marks a change in management organization, with the positions of Executive Director and Scientific Director being eliminated in favor of the two described. Hamish was also appointed to the Boeing Distinguished Professorship last summer. The DOE Office of Nuclear Physics, which provides our operating grant, reviewed our program in September 2008 and subsequently approved funding for a further three years (FY09-11), contingent on successful yearly continuation proposals. We thank our external advisory committee, Baha Balantekin, Stuart Freedman, and Bill Zajc, for their valuable recommendations and advice. The committee reviewed our program in July, 2008 in advance of the DOE panel review. Results from the third phase of the Sudbury Neutrino Observatory (SNO) with our 3 He- filled proportional counters deployed to detect neutrons from the neutral-current disintegra- tion of deuterium were published and reported at Neutrinos 2008 in Christchurch, NZ. The measured fluxes in the three reactions registered by SNO agreed with previous phases, lead- ing to improved precision in the determination of the mixing angle θ12 , as well as providing a test of systematic uncertainties given the very different method. For the SNO+ project we are in the process of replacing the SNO data acquisition software with an ORCA based system. So far we have coded the communication and readout of all


  • Page 5

    ii the major electronics components used in SNO, and plan to have the entire DAQ system completed later this year. We have also measured the quenching factor for alpha particles in the liquid scintillator. The construction of the detector system, the US contribution to KATRIN, has continued to make progress during the past year, albeit with some delays as vendors have encountered problems. One of the two superconducting magnets, the magnet support system, the multi- pixel Si PIN diode array, most of the vacuum system, and calibration equipment, have now been completed with support from both DOE and the University. The electronics has been completed and delivered by our colleagues at Forschungszentrum Karlsruhe. On the MAJORANA R&D project, we have completed the design for a pulsed-reset pream- plifier for the front-end of the detector amplifier chain. Fabrication of a prototype is begin- ning. New concepts for a mechanical design of the detector holders that lends itself to ultra-clean construction have been developed and prototyped. Our efforts on target developments for an implanted 22 Na target paid off and we were able to halve the damage by the high intensity beams on the targets. We have now taken most of the data that we need to determine the consumption of 22 Na in explosive stellar environments. Although our data analysis is in progress we have already concluded that a resonance at Ep ∼ 198 keV that was proposed based on indirect measurements and expected to dominate the consumption rate, actually does not contribute at all. We published our work on the determination of the electron-capture branch from 100 Tc and started working on a similar experiment on 116 In that will take place later in 2009. Both these numbers will provide benchmarks for nuclear structure calculations for double-beta decays. The UCNA collaboration published the first determination of the beta asymmetry using ultracold neutrons. Although the uncertainty in that publication was large at almost 5%, we now have in hand data that should determine the beta asymmetry to < 1% and we expect to take data within 2009 that could reduce the uncertainty down to < 0.3%. At that level this measurement will be an important check on extraction of Vud from nuclear beta decay and will likely be the best determination of the ratio of the axial to vector coupling constants. A new upper limit on the permanent electric dipole moment of atomic mercury recently has been reported. The electric dipole moment is generated by interactions that violate time reversal symmetry. The UW result is the most sensitive test of time reversal symmetry violation on ordinary matter. The test of the possibility of observer-to-observer communication with quantum nonlo- cality has continued. A new type of nonlinear crystal with vastly improved entangled-photon production is being designed into the experiment. Work on the quantum mechanical DWEF model describing RHIC collisions has contin- ued. It had been discovered that an additional term must be added to the formalism when imaginary potentials are used. We have extended our measurements of 2D angular autocorrelations to determine the pt


  • Page 6

    UW CENPA Annual Report 2008-2009 May 2009 iii dependence of the azimuth quadrupole (v2) free of jet contributions and the A dependence of minijet angular correlations. We have also established a quantitative connection between minimum-bias jets in spectra and correlations and pQCD calculations of fragment distribu- tions which indicates that almost all scattered partons survive to the final state in the form of correlated hadrons. Three CENPA graduate students, Adam Cox, Erik Mohrmann, and Sky Sjue, obtained their Ph.D. degrees during the period of this report. As always, we encourage outside applications for the use of our facilities. As a con- venient reference for potential users, the table on the following page lists the capabili- ties of our accelerators. For further information, please contact Greg Harper, Associate Director, CENPA, Box 354290, University of Washington, Seattle, WA 98195; (206) 543- 4080, or gharper@u.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 underlined, to whom inquiries should be addressed. Hamish Robertson, Director Greg Harper, Associate Director and Editor Victoria Clarkson, Assistant Editor


  • Page 7

    iv 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. Methods 122, 65 (1974). Recently adapted to an (optional) terminal ion source and a non-inclined tube #3, which enables the accelerator to produce high intensity beams of helium and hydrogen isotopes at energies from 100 keV to 7.5 MeV. Some Available Energy Analyzed Beams Ion Max. Current Max. Energy Ion Source (particle µA) (MeV) 1H or 2 H 50 18 DEIS or 860 3 He or 4 He 2 27 Double Charge-Exchange Source 3 He or 4 He 30 7.5 Tandem Terminal Source 6 Li or 7 Li 1 36 860 11 B 5 54 860 12 C or 13 C 10 63 860 ∗14 N 1 63 DEIS or 860 16 O or 18 O 10 72 DEIS or 860 F 10 72 DEIS or 860 ∗ Ca 0.5 99 860 Ni 0.2 99 860 I 0.001 108 860 *Negative ion is the hydride, dihydride, or trihydride. Additional ion species available including the following: Mg, Al, Si, P, S, Cl, Fe, Cu, Ge, Se, Br and Ag. Less common isotopes are generated from enriched material. In addition, we have produced a separated beam of 15-MeV 8 B at 6 particles/second. BOOSTER ACCELERATOR See “Status of and Operating Experience with the University of Washington Superconducting Booster Linac,” D. W. Storm et al., Nucl. Instrum. Methods A 287, 247 (1990). The Booster is presently in a “mothballed” state.


  • Page 8

    UW CENPA Annual Report 2008-2009 May 2009 v Contents INTRODUCTION i 1 Neutrino Research 1 SNO 1 1.1 Status of the SNO project . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 NCD pulse fitting with simulated-pulse libraries . . . . . . . . . . . . . . . . 3 1.3 Development of a background-free NCD pulse-shape analysis . . . . . . . . . 4 SNO+ 5 1.4 Status of the SNO+ experiment . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.5 ORCA DAQ system for SNO+ . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.6 Quenching and energy resolution of alpha lines in SNO+ . . . . . . . . . . . 7 1.7 Understanding light propagation in the SNO+ detector . . . . . . . . . . . . 8 KATRIN 9 1.8 Status of the CENPA contribution to the KATRIN experiment . . . . . . . . 9 1.9 Estimation of the environmental radioactivity in the KATRIN spectrometer hall 11 1.10 Monte Carlo Studies of low energy electrons incident on silicon . . . . . . . . 12 1.11 Status of the vacuum system for the KATRIN detector . . . . . . . . . . . . . 13 1.12 Absolute efficiency calibration of KATRIN Si multipixel focal-plane detector 14 1.13 Status of the superconducting magnets for KATRIN . . . . . . . . . . . . . . 15 1.14 Preparations for KATRIN FPD electronics commissioning . . . . . . . . . . 16 1.15 Preparations for KATRIN FPD commissioning . . . . . . . . . . . . . . . . . 18 MAJORANA 20 1.16 MAJORANA R&D activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 1.17 Material screening with germanium detectors . . . . . . . . . . . . . . . . . . 21


  • Page 9

    vi 1.18 Surface-alpha measurements on HPGe detectors . . . . . . . . . . . . . . . . . 22 1.19 Methods for deploying ultra-clean detectors . . . . . . . . . . . . . . . . . . . 23 2 Fundamental Symmetries and Weak Interactions 24 Torsion Balance Experiments 24 2.1 Charge measurement for gravitational wave observatories . . . . . . . . . . . 24 2.2 Progress on improved equivalence principle limits for gravitational self-energy 26 2.3 Continued progress toward a new sub-millimeter test of the gravitational inverse square law . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.4 A cryogenic torsion balance for gravitational experiments . . . . . . . . . . . 28 2.5 Wedge pendulum progress report: testing the gravitational inverse-square law 29 2.6 Progress on a torsion balance test of new spin coupled forces . . . . . . . . . 30 2.7 Designing an equivalence principle pendulum with hydrogen rich test bodies . 31 Weak Interactions 32 2.8 Production of 6 He to determine the e-ν e correlation . . . . . . . . . . . . . . 32 2.9 Measurement of the neutron beta asymmetry with ultracold neutrons . . . . 34 2.10 Characterization of ultracold neutron detectors for use in the UCNA experi- ment at LANL: conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 2.11 Development of a 114mIn calibration source for the UCNA experiment . . . . 36 2.12 Parity non-conserving neutron spin rotation experiment . . . . . . . . . . . . 37 2.13 Permanent electric dipole moment of atomic mercury . . . . . . . . . . . . . . 38 2.14 The APOLLO lunar laser ranging project progress report . . . . . . . . . . . 40 Quantum Optics 41 2.15 Progress on a test of quantum nonlocal communication . . . . . . . . . . . . . 41 3 Axion Searches 42 3.1 First results from the “axion” torsion-balance experiment . . . . . . . . . . . 42


  • Page 10

    UW CENPA Annual Report 2008-2009 May 2009 vii 3.2 The Axion Dark-Matter Experiment at CENPA: the phase II upgrade . . . . 43 4 Nuclear Astrophysics 46 4.1 Target development for 22 Na(p,γ) measurements . . . . . . . . . . . . . . . . 46 4.2 Determining the reaction rate for 22 Na(p,γ): preliminary results . . . . . . . 48 4.3 Studies of explosive nucleosynthesis in proton-rich environments . . . . . . . 50 4.4 Investigation of claims about correlations between the decay rate of 54 Mn and solar activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 5 Nuclear Structure 54 5.1 Electron capture branch for 100 Tc and 116 In and nuclear structure relevant for double-beta decays . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 5.2 Precision mass measurements of 20 Na, 24 Al, 28 P, 31 S, and 32 Cl . . . . . . . . 55 6 Relativistic Heavy Ions 57 6.1 Summary of event structure research . . . . . . . . . . . . . . . . . . . . . . . 57 6.2 Pileup rejection for angular correlations in high-luminosity A-A data . . . . 58 6.3 Cu-Cu angular correlations and the sharp transition . . . . . . . . . . . . . . 59 6.4 Azimuth quadrupole marginal pt dependence . . . . . . . . . . . . . . . . . . 60 6.5 Azimuth quadrupole joint (pt1 , pt2 ) structure . . . . . . . . . . . . . . . . . . 61 6.6 Online quality assurance for STAR data acquisition . . . . . . . . . . . . . . . 62 6.7 The blast-wave model and the myth of radial flow . . . . . . . . . . . . . . . 63 6.8 The spectrum soft component as a unversal feature of all fragmentation processes 64 6.9 Dramatic differences between p-p̄ and e+ -e− fragmentation functions: Does the hard Pomeron break FF universality? . . . . . . . . . . . . . . . . . . . . 65 6.10 The minimum-bias parton spectrum and saturation-scale arguments . . . . . 66 6.11 pQCD calculations of minimum-bias fragment distributions . . . . . . . . . . 67 6.12 Comparisons of calculated pQCD fragment distributions and measured spec- trum hard components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 6.13 Good, bad and ugly ratio measures of fragmentation . . . . . . . . . . . . . . 69


  • Page 11

    viii 6.14 Centrality evolution of fragment distributions and the sharp transition in jet angular correlations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 6.15 Periodicity effects and jet-structure modeling on azimuth . . . . . . . . . . . 71 6.16 The ZYAM prescription and underestimation of jet yields at RHIC . . . . . . 72 6.17 Recovering valid jet structure – two RHIC case studies . . . . . . . . . . . . . 73 6.18 Relativistic heavy ion physics-analysis of pionic interferometry: the DWEF model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 7 Electronics, Computing, and Detector Infrastructure 75 7.1 Electronic equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 7.2 Pulsed reset preamp for MAJORANA . . . . . . . . . . . . . . . . . . . . . . . 76 7.3 Data acquisition development for the MAJORANA experiment . . . . . . . . . 78 7.4 Laboratory computer systems . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 7.5 Athena: a high end computing deployment for scientific computing . . . . . 80 7.6 Linux driver development for VME-based single-board computers . . . . . . 81 7.7 Drift filter for difference measurements - understanding the errors . . . . . . 82 7.8 Characterization of the version 3 IPE crate and optimization of the onboard trapezoidal filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 8 Accelerator and ion sources 85 8.1 Van de Graaff accelerator and ion source operations and development . . . . 85 8.2 Modification of the DEIS and injector deck for noble gas ion implantation . 86 9 Status of the Career Development Organization 87 10 CENPA Personnel 88 10.1 Faculty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 10.2 CENPA External Advisory Committee . . . . . . . . . . . . . . . . . . . . . . 88 10.3 Postdoctoral Research Associates . . . . . . . . . . . . . . . . . . . . . . . . 89 10.4 Predoctoral Research Associates . . . . . . . . . . . . . . . . . . . . . . . . . 89


  • Page 12

    UW CENPA Annual Report 2008-2009 May 2009 ix 10.5 Non-RA graduate students taking research credit . . . . . . . . . . . . . . . . 89 10.6 NSF Research Experience for Undergraduates participants . . . . . . . . . . . 89 10.7 University of Washington undergraduates taking research credit . . . . . . . . 90 10.8 Professional staff . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 10.9 Technical staff . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 10.10 Administrative staff . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 10.11 Part time staff and student helpers . . . . . . . . . . . . . . . . . . . . . . . 91 11 Publications 92 11.1 Published papers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 11.2 Papers submitted or to be published 2009 . . . . . . . . . . . . . . . . . . . . 96 11.3 Book Publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 11.4 Invited talks, abstracts and other conference presentations . . . . . . . . . . . 98 11.5 Ph.D. degrees granted . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102


  • Page 13


  • Page 14

    UW CENPA Annual Report 2008-2009 May 2009 1 1 Neutrino Research SNO 1.1 Status of the SNO project N. S. Oblath, R. G. H. Robertson, N. R. Tolich, B. VanDevender, H. S. Wan Chan Tseung, J. F. Wilkerson∗ , plus the SNO Collaboration. The primary scientific goal of the Sudbury Neutrino Observatory (SNO) experiment was to measure the flux of the different neutrino flavors coming from the sun in order to understand the solar neutrino anomaly. Data taking ended in November, 2006, and since then all efforts have focussed on analysis of the recorded data. The three distinct phases of the experiment each measured the total flux for all three light neutrino flavors in a different way. Data from the first two phases1,2 confirmed that the solar neutrino anomaly was due to the nature of the neutrino, with neutrinos changing from electron neutrinos at production in the sun to an admixture of all three light neutrino flavors by the time they reached detectors on the earth. However, these two phases had a strong correlation between the measured electron neutrino flux and the total neutrino flux. For the third phase, proportional counters filled with 3 He were deployed within the heavy water in the SNO detector, allowing for an independent measurement of the total neutrino flux. In June, 2008, the SNO collaboration published3 an analysis of the third phase data that agreed with the result from the previous two phases and reduced the uncertainty on the neutrino mixing angle θ12 . Because this data was collected with a new detector system, there were a number of systematic error estimates and new analyses that had to be completed for this paper. People at CENPA were involved in many aspects of this analysis, and much of this has been described in last years annual report4 . In the last year Noah Oblath led the group developing a simulation of the pulse shapes caused by various sources of charged particles in the proportional counters. This resulted in probability density functions versus energy for alpha particles from various sources that included a full accounting of the systematic errors due to inaccuracies in their simulation model. The largest systematic error came from not knowing the depth from the inner surface for alphas from 210 Po decays. Although these were assumed to come from the surface, due to a surface layer of 210 Pb from 210 Rn decay, there was evidence for migration with a depth of approximately 1 µm. Nikolai Tolich led the group responsible for extracting the neutrino flux from the data. In order to handle systematic errors more rigorously than previous SNO analyses, this group decided to use a new method based on a Markov Chain Monte Carlo method and Bayesian statistics. After the release of the first results from the third phase, Noah Oblath has worked on developing a method to distinguish neutrons and alphas in the proportional counter data, in an effort to remove a significant background, reducing both statistical and systematic errors. ∗ Currently at the University of North Carolina 1 B. Aharmim, et al., Phys. Rev. C 75, 045502 (2007). 2 B. Aharmim, et al., Phys. Rev. C 72, 055502 (2005). 3 B. Aharmim, et al., Phys. Rev. Lett. 101, 111301 (2008). 4 CENPA Annual Report, University of Washington (2008) p. 1.


  • Page 15

    2 The SNO experiment is unique, and there is a commitment to optimize the analysis of these data as this will likely provide the best constraint on the neutrino mixing angle θ12 for the foreseeable future. In order to achieve this there has been an effort underway for a number of years to lower the energy threshold for the previous analyses of the first two phases from 5 MeV to 3.5 MeV. This process is reaching a conclusion with blindness removed, and the final analysis currently underway. As analysis co-ordinator Nikolai Tolich has ensured that this analysis proceeded with a thorough internal review. Once this analysis is completed, there is a plan to combine it with an analysis of the final phase, providing the most complete analysis of the SNO data. As analysis co-ordinator Nikolai Tolich has also overseen the completion of a paper5 measuring the flux of both muons and neutrinos produced in cosmic ray showers. The depth of SNO uniquely allowed for a measurement of the atmospheric neutrino flux coming from above the horizon where the neutrinos were not expected to have undergone neutrino oscillations. There is also an on-going analysis searching for so-called hep neutrinos from the sun. The predicted flux for these neutrinos has a large uncertainty, and SNO has previously published limits based on the first phase data. However, there is a prospect that with data from all three phases we would observe a signal if it were at the upper limit of expectations. Finally, there are also searches underway for “exotic” physics, such as fractionally charged particles, neutron anti-neutron oscillations, and astronomical sources. 5 B. Aharmim, et al., submitted to Phys. Rev. D (2009).


  • Page 16

    UW CENPA Annual Report 2008-2009 May 2009 3 1.2 NCD pulse fitting with simulated-pulse libraries N. S. Oblath We have previously developed a unique and detailed pulse simulation for the SNO Neutral Current Detection system1 . With the goal of using the simulation to separate neutron-capture signal events from the alpha-decay backgrounds, simulated neutron-capture and alpha pulses are used to fit the NCD data. Due to the speed of the NCD pulse simulation it is impractical to simulate pulses during the fitting process. Instead libraries of pulses are created before fitting. Three libraries are used: neutron captures, wall alphas (alphas originating from decays in or on the NCD walls), and wire alphas (from decays in or on the NCD anode wires). The variety of neutron and alpha pulse shapes, real and simulated, is a result of the initial energy and the location and direction of the ionization track inside the NCD. Four coordinates define the parameter spaces of the neutron and alpha simulations, and the pulses in each library are simulated on a grid in each space. For neutrons, the parameters are the track direction, the position of the capture along the counter, and the capture radius from the anode wire. For alphas the radius is constant but the initial energy varies. Each library is composed of approximately 3,500 pulses. This particular library size was optimized for each library to cover the various pulse shapes sufficiently without making the fits take too long. Each pulse in a data set is fit with all pulses in a library. The fit region of each data pulse extends from the leading edge (10% of the pulse amplitude) to beginning of the ion tail. A χ2 is calculated for the comparison of each library pulse to a data pulse. An additional term is added to the χ2 to account for the energy difference between data and library pulses. The result is a fit that reliably finds a simulated pulse shape to match the events in the NCD data set. The quality of the fits was verified by using neutron calibration pulses and alpha pulses from the 4 He strings. Fig. 1.2-1 shows a neutron pulse fit with the neutron library (left), and an alpha pulse fit with the alpha library (right). This fitting algorithm is being used to determine the number of neutrons detected by the NCDs (see Sec. 1.3). Figure 1.2-1. Examples of a neutron (left) and alpha (right) pulse fit. 1 CENPA Annual Report, University of Washington (2008) p. 5.


  • Page 17

    4 1.3 Development of a background-free NCD pulse-shape analysis N. S. Oblath We have developed a method for determining the number of neutron captures in the SNO NCD data set based on pulse shapes. We fit the NCD pulses with libraries of signal and background events (see Sec. 1.2) to identify a subset of neutron-capture pulses that are unique from the alpha backgrounds. A data pulse is fit with the neutron library and the alpha (wall & wire) library, producing two fit results, χ2n and χ2α . In Fig. 1.3-1 the left plot shows neutron and alpha events in the two-dimensional parameter space, log(χ2α ) versus log(χ2n ). There is a distinct region of the parameter space that is almost entirely free of alpha events. A linear combination of the two χ2 values can be used to simplify the event discrimination: ∆ log(χ2 ) ≡ log(χ2α ) − log(χ2n ). The right plot in Fig. 1.3-1 shows the ∆ log(χ2 ) distribution for neutron captures, alpha events, and the full NCD data set. The alphas fall in a single peak near zero. The neutrons also exhibit a peak around 0, which is composed of pulses that resemble alpha events. Additionally, however, there is a shoulder of neutron capture events that is composed of pulses that look uniquely like neutron-captures. Alpha pulses are created by a single ionizing alpha particle, whereas neutron pulses are a result of a back-to-back proton/triton pair. When the proton and triton travel parallel to the anode wire they create a pulse that resembles an alpha event. When they travel perpendicular to the anode wire the two-particle structure is revealed. The pulses created by a perpendicular ionization track are easily distinguishable from alpha pulses. Since there is limited information about the shape of the alpha ∆ log(χ2 ) distribution (≈ 1, 000 events on the 4 He strings, versus ≈ 80, 000 neutron events from the 24 Na calibra- tions), systematic errors resulting from the alpha backgrounds can be minimized by using a background-“free” cut. The cut fraction for neutron pulses is well-determined by the 24 Na calibrations. A preliminary cut removes ≈ 99% of the alpha events, while keeping ≈ 38% of the neutron-capture events. By cutting so many signal events the statistics become the dominant source of error. Optimization of the cut is currently underway to minimize the total error. Figure 1.3-1. The two-dimensional (left) and one-dimensional (right) χ2 space.


  • Page 18

    UW CENPA Annual Report 2008-2009 May 2009 5 SNO+ 1.4 Status of the SNO+ experiment M. A. Howe∗ , J. Kaspar, J. K. Nance, N. R. Tolich, H. S. Wan Chan Tseung, and J. F. Wilkerson∗ The Sudbury Neutrino Observatory (SNO) project ended data taking in November, 2006. Filling the detector with a liquid scintillator will transform the SNO experiment into the SNO+ project. The scintillator will provide 25 times more photoelectrons per electron than the Cherenkov process in SNO, which will improve the energy resolution and allow the energy threshold to be decreased. The SNO+ project is being built upon existing infrastructure of the SNO experiment allowing for a shorter time to the first data, which is expected in 2011. The diverse scientific goals address fundamental questions in particle physics, astro- physics, and geosciences. In particular, adding 150 Nd to the scintillator will enable a search for neutrino-less double-beta-decay with the neutrino mass sensitivity of 50 meV (5σ c.l.). SNO+ will detect geoneutrinos, i.e., electron antineutrinos from natural radioactivity in the earth, to understand the radiogenic component of the earth’s heat flow, and to test compet- ing models of the composition of the continents. The measurement is possible thanks to the position of the detector in a well-studied continental crust in a region with low background from nuclear reactors. However, the reactor signal still allows SNO+ to detect antineutrinos from distant reactors and to observe spectral distortions due to neutrino oscillations. Before Nd is added to the scintillator, SNO+ will measure the low-energy solar neutrinos (pep and CNO reactions) to report on the sterile neutrino admixture, non-standard interactions, and sun’s metallicity. Finally, the SNO+ will maintain excellent supernova neutrino capabilities. CENPA has been active in the Monte Carlo group working on energy and position fitters. New fitters are under development to handle scintillation light which is uniform in direction. In the SNO project the information on position came from the Cherenkov cone which is now strongly suppressed by the scintillation light. This work has been accompanied by a detailed understanding of the optical properties of the SNO+ setup. A dedicated measurement of the SNO+ scintillator was performed in summer 2008. A bucket filled with the scintillator was placed into the center of the detector filled with water. We have contributed to the analysis of the bucket data focusing on energy resolution and quenching of the alpha lines coming from Rn dissolved in the scintillator. This analysis is crit- ical for determining the detector response to different ionizing particles and for determining sensitivity to neutrino mass, since it depends directly on the energy resolution achieved. The UW group has been leading the DAQ task to deliver a new ORCA based system to read out SNO+ data. ORCA was used to read out data from the proportional counters in the final SNO phase. The present system to read out the PMT data in SNO could not handle two orders of magnitude higher data rate from the scintillator. The new system is under development to be delivered in fall 2009. ∗ University of North Carolina, Chapel Hill, NC


  • Page 19

    6 1.5 ORCA DAQ system for SNO+ M. A. Howe∗ , J. Kaspar, N. R. Tolich, and J. F. Wilkerson∗ The scintillator filling the SNO+ detector will provide 25 times more photoelectrons per elec- tron than the Cherenkov process in SNO, which will improve the energy resolution and allow the energy threshold to be decreased. However, the present data acquisition system, based on OS9 and SUN computers mastering VME crates controlled by a Motorola embedded CPU, cannot handle such a data rate, and is not scalable enough to be upgraded in a reasonable and reliable way. Therefore, it is necessary to develop a new DAQ system. The new DAQ system is based on ORCA, which was developed in CENPA led by Mark Howe. ORCA is a general purpose, highly modular, object-oriented, acquisition and control system that is easy to use, develop, and maintain. It has been well received in the final SNO phase to read out the proportional counters. Its general-purpose design enables moving from a master-slave paradigm towards more independent standalone units. The former SNO and future SNO+ topologies are compared in Fig. 1.5-1. A test stand located in Sudbury has been used for our DAQ development. It allows us to run both the former SHaRC and the future ORCA system in parallel to cross-check the outputs. We have managed to initialize, control, and read out all the hardware devices in the SNO+ setup with ORCA. We have developed a control code for the single board computer mastering the interface crate running Linux (see Sec. 7.6). Now we are proceeding with the run control, hardware calibration and validation. The first underground tests are scheduled for May, 2009. The full DAQ system will be running in fall 2009. Besides the DAQ system itself we have also started to code the DAQ part of the Monte Carlo model of the SNO+ detector. Figure 1.5-1. The left figure shows a DAQ scheme for the PMT readout in SNO. The right figure shows the planned DAQ scheme for the SNO+ . Only one out of nineteen PMT data crates is shown. ∗ University of North Carolina, Chapel Hill, NC


  • Page 20

    UW CENPA Annual Report 2008-2009 May 2009 7 1.6 Quenching and energy resolution of alpha lines in SNO+ N. R. Tolich and H. S. Wan Chan Tseung During October 2008, several calibrations were carried out to study the optical properties of the SNO+ scintillator. A cylindrical acrylic container containing about 1 L of scintillator was deployed within the SNO acrylic vessel, which was filled with water. The scintillation light yield L from electrons (in units of phototube hits per MeV) was obtained by positioning an AmBe neutron source near the bucket, and locating the Compton edges of the 2.2 MeV neutron-capture γ and the 4.4 MeV γ from the first excited state of 12 C. Corrections due to Cherenkov light contributions, multi-photon hits and random noise triggers were applied1 . We measured the quenching of alphas by using the decays of dissolved 222 Rn and its daugh- ters, which produced two noticeable peaks below 1 MeV equivalent electron energy in the spectrum. The higher-energy peak was found to be predominantly composed of 214 Po α de- cays via the delayed coincidence method. The number of 214 Bi–214 Po coincidences decreased over time with a half life of about 4 days, confirming the presence of 222 Rn. The low-energy peak consisted mainly of 218 Po and 222 Rn alphas. The positions of these three alpha lines were obtained through Gaussian fits to the peaks, after subtracting out the underlying β–γ background, which is dominated by 214 Pb and 214 Bi decays. Monte Carlo studies indicated that the shape of the β–γ background spectrum can be approximated by a 6th order poly- nomial. An example fit to a group of 8 runs, after β–γ removal, is shown in Fig. 1.6-1 (left), where the green-dashed, red-dashed and blue curves are the 222 Rn, 218 Po and 214 Po alphas, respectively. For Nd-doped scintillator, the corresponding quenching factors √ were 10.2 ± 0.8, 9.6 ± 0.8 and 8.1 ± 0.5, while the energy resolution was found to be 6.5%/ E (MeV). The quenching effect is well-described by Birks’ law2. By using L and the three alpha peak positions, a value for Birks’ constant can be extracted ( Fig. 1.6-1, right). This was found to be 73.2 ± 3.6 µm/MeV. ×103 number of counts phototube hits Runs 70771 - 70779 400 1 Birks’ law (kB=73.2 microns/MeV) Data 300 0.8 0.6 200 0.4 100 0.2 0 0 200 300 400 500 600 2 4 6 8 10 12 14 phototube hits Energy (Mev) Figure 1.6-1. Left: alpha peak components (see text). Right: fit to alpha peak positions using Birks’ law. 1 Helen O’Keeffe (Queen’s University), private communication. 2 J. B. Birks, The theory and practice of scintillation counting, Pergamon Press, 1964.


  • Page 21

    8 1.7 Understanding light propagation in the SNO+ detector J. K. Nance and N. R. Tolich A significant source of backgrounds in any large spherical detector such as the SNO+ detector is external radiation either penetrating from the surrounding materials or present on the inner surface of the detector itself. In order to reduce the influence of those backgrounds, the SNO+ detector will implement a fiducial volume cut such that only events occurring well within the scintillation volume are accepted. Event reconstruction is then performed using timing and charge collection information from the surrounding photomultiplier tubes. However, the nature of excitations in organic scintillating cocktails such as that proposed for use in SNO+ complicates the problem of event reconstruction. Because the scintillator can absorb and re-emit photons generated in an interaction, the original vertex of a physics event can become obscured and more difficult to understand. A GEANT4 based C++ Monte Carlo simulation is used to model the response of the detector to physics events in the scintillation volume, but information about the physics processes that take place in between the vertex and the photomultiplier tubes is lost. Work is therefore currently being done to recover this lost information and make use of it in the event reconstruction process. It is hoped that the effect of position on the timing distributions can be determined and thus improve the PDFs used for reconstruction. Below is a figure which helps motivate this type of analysis. Shown in Fig. 1.7-1 is the photomultiplier tube hit time distribution for photons which reflected from surfaces outside the acrylic vessel in the SNO+ detector. At this level of the analysis, it demonstrates that there is a significant contribution to the spectrum from these reflections. Importantly, if the reflections occur outside the acrylic vessel, they should be independent of the position of the event inside the vessel, which is important information from the reconstruction perspective. Large R Multi›Reflections TD Entries 27331 Mean 83.07 RMS 46.07 0.16 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0 0 50 100 150 200 250 300 350 400 450 Hit Time(ns) Figure 1.7-1. Simulated time spectrum for PMT hits caused by photons reflected from surfaces outside the SNO+ AV.


  • Page 22

    UW CENPA Annual Report 2008-2009 May 2009 9 KATRIN 1.8 Status of the CENPA contribution to the KATRIN experiment J. F. Amsbaugh, L. Bodine, T. H. Burritt, P. J. Doe , G. C. Harper, M. L. Leber, E. L. Martin, A. W. Myers, R. G. H. Robertson, K. Tolich, B. VanDevender, T. D. Van Wechel, B. Wall. The KATRIN collaboration aims to set a limit on the neutrino mass to a sensitivity of 0.2 eV via a study of tritium beta decay. The main components of the experiment are a windowless, gaseous tritium source, a transport section that carries the beta electrons to a pair of electrostatic spectrometers and, finally, a detector system that measures the flux of electrons passing through the spectrometers. The US institutes, consisting of the University of Washington (UW), the Massachusetts Institute of Technology (MIT) and the University of North Carolina (UNC) are responsible for providing the detector system and the data acquisition software. The University of California, Santa Barbara (UCSB) has recently joined the US contingent and will be participating in diagnostics, simulations and analysis. The KATRIN project is currently at the height of its construction activities at the Forschungzentrum, Karlsruhe (FZK). The US is scheduled to complete installation of the detector at the FZK by December 2010, and data taking with tritium is expected to begin in September 2012. Significant progress has been made in acquiring the main KATRIN components although challenges remain. Principal among these is the windowless gaseous tritium source (WGTS). The company originally contracted to supply this component was acquired, and the division responsible for supplying the WGTS was split off. A solution to this is in hand, and the WGTS is expected to be commissioned in early 2012. Tritium gas emerging from the windowless source is returned via a differential pumping system (DPS). The DPS has passed its cooling tests and field mapping. Delivery is expected in May 2009. Tritium escaping the DPS is trapped by the cryogenic pumping system (CPS) in order to prevent tritium contaminating the spectrometers. Trapping is achieved by coating the walls of the beam pipe with argon snow. Delivery of the CPS is expected in September 2010. The spectrometer system consists of a pre-spectrometer that is used to filter out the uninteresting, low energy parts of the spectrum and a main spectrometer. The prespectrom- eter was delivered in 2004 and has been used to optimize the electrode design for the main spectrometer and study sources of backgrounds such as Penning traps. The main spectrom- eter, with a resolution of 0.93 eV, enables the precision measurement of the beta spectrum. This vessel, delivered in late 2006, has passed its vacuum commissioning and is currently having a field-shaping, background-suppressing electrode installed. Completion is expected in October 2009 after which the spectrometer will undergo electrostatic commissioning be- ginning in March 2010. Commissioning and background studies conducted with the pre and main spectrometers are being carried out with the data acquisition software supplied by the US. Assembly and commissioning of the detector system is taking place at UW with MIT


  • Page 23

    10 providing the calibration, and background control hardware, consisting of inert shielding and a scintillator veto. UNC is providing the DAQ software for the system. Assembly has been slowed by delays in supplying certain critical components; the magnets, a custom, low background feedthrough, and the post acceleration electrode system, although resolution of these difficulties is underway. The problem that prevented the pair of superconducting magnets from reaching their design field of 6T was traced to mismatched coefficients of expansion in the material chosen to wrap the magnet coil. This has now been rectified, and delivery is anticipated in early May 2009. The electroformed copper horn of the post acceleration electrode suffered blistering during the assembly brazing process. This component has now been replaced with a spun copper part, and delivery is expected in early May 2009, allowing completion of the vacuum system. Finally, the low background sapphire signal feedthough continues to experience manufacturing problems. While these are being resolved, two feedthroughs using standard, less radioactively clean, glass insulators have been ordered. This will allow testing of the electronics and evaluation of the detector wafers. The detector wafers were received in July. They have met their dimensional tolerances, and room temperature leakage currents are within specifications. Further evaluation will take place upon receiving the custom signal feedthrough on which the detectors are to be mounted. The detector electronics have been received from FZK and are currently being assembled and commissioned. The vacuum system and pumping station have been assembled, with final commissioning awaiting the delivery of the post acceleration electrode and baking system. The calibration system, provided by MIT, has been delivered to UW and installed on the vacuum system. A method of measuring the absolute efficiency of the detector by recording the photo current from the electron calibration source has been prototyped by UW and is undergoing testing. The copper and lead shield and handling device has been completed at MIT and is ready to be shipped to UW. Measurements of the laboratory background as a function of shielding have begun and will be used to better understand the measured efficiency of the shield when tested in the laboratory. The veto is currently under construction at MIT with delivery expected in May 2009. The background model continues to be improved and tested. Incorporating the results of the assays of detector materials gives us confidence that we will meet the background goal of 1 mHz. Testing of the Mk3 DAQ crate hardware has been carried out at UW, and the Mk4 version is expected in Fall 2009. The Slow Control system was delivered to UW in Fall 2008 with final commissioning expected in summer 2009. A revised schedule accounting for the delays in equipment delivery has been produced. The fully commissioned detector system will be shipped to the FZK on July 2010 and instal- lation is expected to be complete in December 2010.


  • Page 24

    UW CENPA Annual Report 2008-2009 May 2009 11 1.9 Estimation of the environmental radioactivity in the KATRIN spec- trometer hall M. L. Leber, P. Renschler∗ , and S. Kage∗ The detector-related backgrounds in the KATRIN experiment need to be under 1 mHz in the region of interest. By radio-assay of construction materials and using known cosmic ray fluxes, the background from these sources can be estimated with the existing KATRIN Geant4 simulation. A challenging background to estimate is the environmental radioactivity which originates in materials outside of the detector section. By measuring this background with a Germanium detector and comparing to an estimate from the simulation, we can make a projection about the rate in the KATRIN silicon detector. Three measurements were made in the KATRIN prespectrometer hall with a Germanium detector: an unshielded measurement, a measurement inside 15 cm of lead shielding, and a measurement inside the 15 cm shield with a twelve-degree opening that matches the KATRIN shield. This combination of measurements allowed an estimate of the total flux of photons from radioactive decay and an estimate of the flux coming from the direction of the shield opening. To simulate the environmental radioactivity, we assumed the initial unstable isotopes are embedded in concrete, so the emitted photons can Compton scatter within the source. This ensures that the spectrum incident on the germanium detector contains both full energy peaks and the photons which have already Compton scattered within the concrete. Uranium, Thorium, and Potassium are spread throughout the concrete surrounding the Germanium detector. The full-energy peaks were fit in the simulation and measurement to determine the area, or total number of counts. The simulation was then scaled to match the measurement and determine the total flux. To estimate the angular dependence of the incident photons, the measurement with the completely closed shield was compared to the measurement with a twelve-degree opening in the shield. The full-energy peaks of each measurement were fit to determine the area. Photons entering through the opening were simulated, and the peak areas in the simulation were matched to the difference of the two measurements. Finally, to determine the background in the KATRIN silicon detector, the initial energy spectra of photons determined by the unshielded measurements was used. The angular dis- tribution of the photons was isotropic, except the rates were different through the shield opening and otherwise, as determined from the two shielded measurements. The background rate without post-acceleration is expected to be 0.92 mHz from the environmental radioac- tivity alone. Therefore, to reach our background goal of 1 mHz total, post-acceleration will be necessary. ∗ Forschungszentrum Karlsruhe, Institut für Experimentelle Kernphysik, Postfach 3640, 76021 Karlsruhe, Germany.


  • Page 25

    12 1.10 Monte Carlo Studies of low energy electrons incident on silicon H. Bichsel, Z. Chaoui∗ , P. Renschler† and R. G. H. Robertson The purpose of this simulation is to determine the energy loss of low-energy electrons (e.g. E = 18.6 keV ) in silicon for the KATRIN focal plane detector. Electrons backscattered from the detector surface will return to the detector due to magnetic reflection in the detector pinch magnet or electrostatic reflection at the main spectrometer. This leads to multiple passages through the detector surface layer (d≈100 nm) in which deposited energy does not contribute to the detector readout. Both effects are included in the simulation results shown in Fig. 1.10-1 and Fig. 1.10-2. In Fig. 1.10-2 the effect of different inelastic cross sections on the simulation results is shown. One is obtained with a model dielectric function given by Penn1 , the other was calculated using the Bethe-Fano approach2 . Assuming a cut at 1% of F(T), the observable energy windows differ by ≈16%. In the current Monte Carlo program all energy lost by primary electrons is assumed to be deposited locally. Further studies3 will show if there is a need for a more detailed simulation that also takes production and tracking of secondary electrons by atomic ionization and relaxation into account. ✬ ✭ ✮ ✯ ✰ ✱ ✲ ✳ ✴ ✵ ✭ ✶ ✳ ✷ ✯ ✸ ✹ ✺ ✻ ✼ ✽ ✾ ✿ ✽ ✻ ✾ ✾ ❀ ✧ ★ ✫ ✧ ★ ✫ ❉ ❊ ❋ P ❍ ▼ ▼ ◗ ◆ ❘ ❍ ❙ ❖ ☛ ☞ ✌ ✍ ✎ ✏ ✑ ✒ ✓ ❉ ❊ ❋ ● ❍ ■ ❏ ❍ ❑ ▲ ▼ ◆ ❖ ✏ ✕ ✖ 4 ✗ ☞ ✔ ✌ ✕ ✕ ✘ ✙ ✦ ❬ ❭ ✜ ❪ ❫ ❴ ❵ ❛ ❜ ❝ ❞ ❡ ❢ ❣ ❤ ✧ ★ ✪ ✧ ★ ✪ ❡ ❥ ❦ 4 ❧ ❛ ✐ ❜ ❜ ❥ ♠ ♥ ❩ ❳ ❨ ✄ ✂ ❲ ✄ ✂ ✁ ❯ ❱ r ✁ ❚ ☎ q ♣ ✕ ✖ 4 ♦ ✔ ✚ ✧ ★ ✩ ✧ ★ ✩ ❁ ❂ ❃ ❄ ❅ ❆ ❇ ❈ ✦ ✦ ❭ ✜ ❪ ❫ ❴ ✆ ✝ ✞ ✟ ✠ ✡ Figure 1.10-1. Spectrum f(T) shows en- ✛ ✜ ✢ ✜ ✣ ✤ ✥ ✦ ✢ ✦ ✜ ✦ ✣ ✦ ✤ ✦ ✥ ✜ ✢ s t ✉ ✈ ✇ ① ergy lost in the sensitive volume for incident Figure 1.10-2. Comparison between two in- electrons with E = 18.6 keV and θ=60◦ elastic cross section models. Shown is F(T) (solid line). The dotted line shows the in- obtained with the Penn and the Bethe-Fano tegral F(T). The chained line shows F(T) approach for E = 18.6 keV and θ=60◦ . for θ=30◦ . ∗ University of Setif, Algeria † Karlsruhe Institute of Technology, Germany 1 D.R. Penn, Phys. Rev. B 35 482 (1987). 2 H. Bichsel, Rev. Mod. Phys. 60 663 (1988). 3 Z. Chaoui et al., Phys. Lett. A 373 1679 (2009).


  • Page 26

    UW CENPA Annual Report 2008-2009 May 2009 13 1.11 Status of the vacuum system for the KATRIN detector J. F. Amsbaugh, T. H. Burritt, G. C. Harper, and K. Tolich The sensitivity of KATRIN critically depends on minimizing background by avoiding inter- actions among particles along the path of electrons from the tritium source to the focal plane detector (FPD). Therefore the FPD will be placed in an extreme high vacuum (XHV) cham- ber whose design pressure is less than 10−10 mbar. The XHV chamber is separated from the main spectrometer by a DN250 gate valve. The XHV chamber provides mounts for its vacuum pumps, vacuum measuring equipment, the post-acceleration electrode, and the FPD calibration devices. Signals from the FPD exit the XHV chamber through a feedthrough flange and are fed into preamplifiers placed in a high vacuum (HV) chamber whose design pressure is less than 10−6 mbar. The HV chamber provides mounts for its vacuum pumps, vacuum gauges, a pulse tube cooler, and another signal feedthrough flange, through which signals exit the HV chamber. The initial rough pumping of each chamber is performed and monitored by its designated combination of roughing pumps and vacuum gauges, all of which reside on a mobile cart (roughing station). After rough pumping, the design pressure is achieved and maintained by a cryopump mounted on each chamber. The chambers and their accessories are supported by a stand that rolls on rails allowing adjustment of the vacuum system position. Since the previous report1 , the majority of components for the vacuum system has been either delivered or manufactured onsite and assembled. The XHV chamber has been mounted on the stand, and the DN250 gate valve, the vacuum measuring devices, cryopump, and the FPD calibration devices are mounted on it. The HV chamber with its cryopump and gauges has been mounted on the XHV chamber. A prototype roughing station cart has been manufactured, and all the components housed by the cart are assembled. Fig. 1.11-1 shows a photograph of the vacuum system assembled as of April 30, 2009. Figure 1.11-1. KATRIN FPD vacuum system assembly 1 CENPA Annual Report, University of Washington (2008) p. 11.


  • Page 27

    14 1.12 Absolute efficiency calibration of KATRIN Si multipixel focal-plane detector E. L. Martin, A. W. Myers, R. G. H. Robertson, and T. D. Van Wechel In order to determine the absolute efficiency of the KATRIN focal-plane detector we intend to measure the electron current from a photoemissive electron gun and compare it to the count rate detected by the detector. This requires a precise current measurement of a femtoamp at a voltage of 20 kV. PULCINELLA (Precision Ultra-Low Current Integrating Normalization Electrometer for Low-Level Analysis) is a current meter capable of measuring fA scale current. The low current measurement was accomplished by use of a current integrator and ADC from Texas Instruments, the DDC-114. For the scale used, full scale charge is 12 pC with 20 bit resolution. The chip contains reset circuitry and twin ADCs that alternate charge collection and readout switched by a supplied clock. The ADC was placed on a single integrator board with a voltage regulator, oscillator, and FPGA used to control the ADC and convert the digital output to serial format and transmit it to a receiver board. To run the 5 V integrator board floating on 20 kV the power and output are optically isolated. Optical power isolation is accomplished with an array of LEDs and a solar panel while data are transmitted over a fiber optic cable. The electron emitter to calibrate the focal-plane detector is a copper disc attached to an isolated shaft passing through a bellows to allow moving it in and out of the beam path from the KATRIN main spectrometer. UV light will pass through a sapphire window to illuminate the photoelectrode. To remove zero current offsets the UV light will be pulsed and status of the light will also be sent to the receiver board and packaged with the charge measurement data. The receiver board also adds an incremental counter for data loss detection and converts the data to an Ethernet signal that can be downloaded to a computer. PULCINELLA was calibrated using a square wave generator with a voltage divider and a 5 GΩ resistor to generate a pulsed 9.24 pA signal. Due to uncertainty on the measurement of the voltage and the value of the 5 GΩ resistor systematic uncertainty was 1.6%. Full scale charge was found to be 11.8 pC with negligible statistical uncertainty. Each charge measurement cycle takes from 10 to 2550 ms. The major noise contribution to individual charge measurements was independent of integration time, favoring long mea- surement cycles for accurate results. Noise was 0.87 fC rms at 20 ms (44 fA) and .98 fC at 1 s (.98 fA).


  • Page 28

    UW CENPA Annual Report 2008-2009 May 2009 15 1.13 Status of the superconducting magnets for KATRIN L. I. Bodine, R. G. H. Robertson, and D. I. Will The two superconducting solenoidal magnets for KATRIN1 are being manufactured by Cry- omagnetics, Inc., Oak Ridge, TN. Construction of the pinch magnet system has been com- pleted and preliminary tests of the system have been performed. The pinch system can safely maintain the full 6 T central field and survive a heater induced quench. The detector mag- net system is nearing completion. Installation in the cryostat (see Fig. 1.13-1) is the only remaining construction task. Our lab in the Physics Astronomy Building (B037) has been prepared for the magnets. A layout that does not conflict with neighboring, field-sensitive experiments has been found and the magnet stands as well as the necessary electrical and water connections have been installed. The necessary cryogen transfer equipment is available and a cryogen supplier has been identified. A preliminary integration of the magnet electronics with the KATRIN Slow Control sys- tem has been completed. The temperature sensors, cryogen level sensors and the power supplies can be computer controlled and the final test of the integration will be completed upon the magnets’ arrival at UW. D. I. Will has been designated as the safety officer and has designed a mandatory super- conducting magnet safety training. The training will take place following the testing of the magnets at Cryomagnetics, Inc. Pinch Magnet preliminary Field Scan @ 1T, 1.6cm is cryostat flange reference 12000 10000 8000 Gauss 6000 4000 2000 0 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 Centimeters Figure 1.13-1. The left panel shows a picture of the completed pinch magnet system. The right panel shows an axial scan of the magnetic field of the pinch coil system with a 1 T central value 1 CENPA Annual Report, University of Washington (2008) p. 15.


  • Page 29

    16 1.14 Preparations for KATRIN FPD electronics commissioning B. VanDevender, B. L. Wall The Institut fur Prozessdatenverarbeitung und Elektronik at the Forschungszentrum Karl- sruhe has designed and built a commissioning set of electronics. The electronics for the KATRIN Focal Plane Detector (FPD) consists of 4 major parts and works as follows. The Power And Control (PAC) board provides power regulation to the preamplifier boards. It controls the preamp module and channel multiplexing. It measures leakage currents, temper- atures and bias voltages, and sets the variable gains on the Optical Sender board. The PAC board is connected to 24 preamp modules via an interface board that distributes the power and multiplexing signals and returns the temperature and leakage current read out voltages. Each preamp module has 6 or 7 channels providing the 148 channels needed for the FPD. The signals from the 24 modules are connected to the Optical Sender board which provides a secondary amplification stage and converts the charge signal to optical. The optical signals are received by the IPE crate which is equipped with Optical Receiver boards. Each signal is then converted from an analog to digital signal in the IPE crate where ORCA1 does the data read out. Figure 1.14-1. The electronics data and control diagram. The red arrows show the data path and the blue arrows show the control path. The IPE group has delivered a commissioning set of electronics to CENPA. The delivery 1 CENPA Annual Report, University of Washington (2008) p. 81.


  • Page 30

    UW CENPA Annual Report 2008-2009 May 2009 17 included 24 preamp modules for a total of 148 channels, the PAC board, a single optical sender board, and two optical receiver boards. An IPE version 3 crate already resides at CENPA. This will allow us to test one quadrant of the detector at a time. After receiving the electronics Lars Petzold of the IPE group visited CENPA and helped establish operation of the electronics. Also during this time Mark Howe visited and established ORCA control of the PAC board and the programmable power supplies that are going to be used to power the detector electronics. We are currently able to select individual preamp channels for pulsing or leakage current measurement, to read the temperature of individual modules, and to select the gains on the optical sender board. Figure 1.14-2. 241 Am and 109 Cd spectrum taken with IPE preamp module. The most energetic lines are the 59 keV and 88 keV γs from those sources Preliminary tests of the preamplifier performance have also begun. We connected a Hamamatsu S3096 to the input of an IPE preamp module and connected the output to the IPE v3 crate for analog to digital conversion. The detector was illuminated with an 10 µC 241 Am and a 100 µC 109 Cd. The width of the 59.5 keV line of the 241 Am is 1.91 keV.


  • Page 31

    18 1.15 Preparations for KATRIN FPD commissioning B. A. VanDevender, M. Steidl, B. L. Wall Three KATRIN Focal-Plane Detectors (FPDs) were received at CENPA in August, 2008. The manufacturer, Canberra Belgium, verified specifications for reverse-bias leakage currents less than 100 nA/cm2 at 20 ◦ C before delivery (Fig. 1.15-1). All three devices were examined at UW under a microscope with a high-precision translating stage (Fig. 1.15-2). The software interface to the scope allowed mapping of all mechanical dimensions of the device. The coordinates of all 24 wafer corners, 148 pixels, guard ring and bias contact region of all three devices meet specifications well within allowed tolerances provided to Canberra. Other specifications are listed in Table 1.15-1. The devices are currently stored in a dry nitrogen environment awaiting full characteriza- tion of their properties. They will be installed in CENPA’s electron gun1 with the front-end electronics described in this report (see Sec. 1.14). The full commissioning program will de- termine the optimum reverse-bias potential, leakage currents, and responses to gamma rays and electrons with energies in KATRINs signal region at the optimum bias at both room temperature and -100 ◦ C. 2 10 leakage current (nA) 10 1 0 20 40 60 80 100 120 140 pixel number 102 leakage current (nA) 10 1 0 20 40 60 80 100 120 140 pixel number 2 10 leakage current (nA) 10 1 0 20 40 60 80 100 120 140 pixel number Figure 1.15-1. Room-temperature leakage current measurements on the three delivered KATRIN FPDs. Closed circles represent measurements at the recommended 120 V bias potential. Open circles represent measurements at the maximum 150 V bias potential. 1 CENPA Annual Report, University of Washington (2008) p. 17.


  • Page 32

    UW CENPA Annual Report 2008-2009 May 2009 19 Figure 1.15-2. On the left is the back side of a KATRIN FPD. The wafer diameter is 114 mm from corner to corner. The dark gray color is due to a TiN layer intended to provide ohmic contact between the p++ doping and front-end electronics. The central dartboard pattern contains 148 44.1 mm2 active pixels. A continuous guard ring surrounds the pixels. The outermost ring of TiN extends around the edge of the device so that the bias potential on the front of the detector can be supplied with contacts on the back. The picture on the right shows the inspection of an FPD. The image on the monitor is the intersection of three pixels (which appear red on the monitor). The gap between pixels (white) is 0.05 mm. Table 1.15-1. Specifications for KATRIN FPDs Geometry: bulk material silicon shape regular 24-sided polygon, 114 mm corner-to-corner thickness 500 µm sensitive area 90 mm diameter circle number of pixels 148 pixel area 44.1 mm2 pixel capacitance 9.45 pF guard ring 90–94 mm diameter circle Doping: substrate n front (entrance) side n++ unsegmented, no metallization back side p++ segmented, TiN metallization dead layer < 150 nm Operating Conditions: temperature -100–30 ◦ C pressure < 10−10 mbar magnetic field < 6T recommended bias 120 V maximum bias 150 V


  • Page 33

    20 MAJORANA 1.16 MAJORANA R&D activities J.F. Amsbaugh, T.H. Burritt, P.J. Doe, A. Garcı́a, M. A. Howe∗ R.A. Johnson, M.G. Marino, M.L. Miller, A.W. Myers, R.G.H. Robertson, A.G. Schubert, T.D. Van Wechel, B.A. VanDevender, J.F. Wilkerson∗, and D.I. Will The MAJORANA Collaboration is performing R&D to demonstrate the feasibility of building a 1 tonne modular array of 86% enriched 76 Ge capable of searching for neutrinoless double- beta (0νββ) decay in the inverted mass hierarchy region (≈30 meV). A cornerstone of the plan is the development and construction of the MAJORANA Demonstrator, a R&D module comprised of 30 kg of 86% enriched 76 Ge and 30 kg of non-enriched Ge detectors. The use of a mixture of both enriched and natural or depleted Ge has the advantages of lowering the costs in the R&D phase, accelerating the deployment schedule, and also giving MAJORANA an opportunity to verify that any observed peak in the 0νββ region of interest is directly associated with the presence of 76 Ge. The goals for the Demonstrator are: • Show that backgrounds, at or below 1 count/tonne/year in the 0νββ decay peak 4 keV region of interest, can be achieved, a necessary condition for scaling to a 1 tonne or larger mass detector. • Demonstrate sensitivity by testing the validity of the Klapdor-Kleingrothaus reported 76 Ge 0νββ observation1 . • Show successful long-term operation of crystals in their respective environments. • Demonstrate a cost-effective and scalable approach. The proposed method uses the well-established technique of searching for 0νββ in high- purity Ge (HPGe) diode radiation detectors that play both roles of source and detector. These detec- tors will be located in specially fabricated ultra low-background, electroformed Cu cryostats. The technique is augmented with recent improvements in signal processing, detector design, and advances in controlling intrinsic and external backgrounds. Major advances in detector R&D (both laboratory and industrial) have led to the choice of p-type point-contact (PPC) detectors for the demonstrator module. These detectors offer low capacitance and excellent resolution, providing the ability to (1) tag and veto two-site background interactions via pulse shape analysis and (2) probe ultra low-energy events (E>≈300 eV) to competitively search for dark matter candidates (WIMPs, axions, etc)2 . Now at University of North Carolina ∗ 1 H. V. Klapdor-Kleingrothaus, I. V. Krivosheina, A. Dietz, and O. Chkvorets, Physics Letters B 586 198 (2004). 2 C.E. Aalseth, P.S. Barbeau, D.G. Cerdeno, Phys. Rev. Letters 101, 251301 (2008).


  • Page 34

    UW CENPA Annual Report 2008-2009 May 2009 21 1.17 Material screening with germanium detectors J. F. Amsbaugh, K. Boddy, R. A. Johnson, M. G. Marino, A. G. Schubert, H. Simons, J. F. Wilkerson∗, and D. I. Will The MAJORANA neutrinoless double-beta decay experiment will require extremely low back- ground rates. The MAJORANA Collaboration has a background goal of 1 count per tonne-year in the 4-keV region of interest surrounding the 76 Ge double-beta decay endpoint. To achieve MAJORANA background goals, materials used in the experiment must meet high radiopurity standards. Assaying raw materials at the sensitivity required for MAJORANA is costly and time consuming. A surface-level material-screening facility has been established at CENPA to prescreen materials for MAJORANA. The CENPA material-screening facility consists of a high-purity germanium detector en- closed in a six-inch-thick lead shield. A scintillator panel was constructed and added to the system. The scintillator tags events in the germanium detectors associated with inci- dent cosmic rays. The original lead shielding was dismantled and reconstructed from lead known to be radiologically clean. The background spectrum in the detector with and without the cosmic-ray veto is shown in Fig. 1.17-1. The facility allows materials not clean enough for MAJORANA to be identified without sending the materials to higher-sensitivity facilities. The minimum sensitivity of the facility, calculated for a massless point source 1 mm from the front face of the detector after seven days of counting, is ≈20 mBq for 232 Th and 238 U in equilibrium, and ≈30 mBq for 40 K. all counts unvetoed vetoed 104 103 counts 102 10 0 500 1000 1500 2000 2500 3000 energy [keV] Figure 1.17-1. Background spectrum inside lead shielding in the MAJORANA lab. The full spectrum, counts tagged by the veto, and counts not tagged by the veto are shown. Data were collected for 5.6 days. ∗ University of North Carolina, Chapel Hill, NC


  • Page 35

    22 1.18 Surface-alpha measurements on HPGe detectors T. H. Burritt, S. R. Elliott∗, V. M. Gehman∗ , V. E. Guiseppe∗ , R. A. Johnson, and J. F. Wilkerson† A test stand for the study of surface-alpha measurements on HPGe detectors has been built and data have been collected. The goal is to understand the response of a detector to alpha particles of different incidence angles and build a more realistic background model for surface alpha decays on HPGe detectors in double-beta decay experiments. The test stand allows an alpha source to shine on the surface of an N-type HPGe detector through various collimation holes of different angles. An 241 Am source attached to a rotational feedthrough allows alignment of the source with a particular collimation hole via manipulation of the feedthrough. In this way, several data runs of different collimation angles can be obtained without opening up the cryostat. Data were taken (Fig. 1.18-1) with alphas of incidence angles 0◦ , 30◦ , 45◦ , and 60◦ , with the angle defined with respect to the normal of the surface. The 241 Am source consists of three main alpha peaks (5388, 5443, and 5486 keV) as well as a 59.5 keV gamma that accompanies the 5486 keV alpha and results in a 4th peak. The width and offset of the peaks is due to energy losses within the dead region of the detector; larger incidence angles result in wider spread and greater offset. The data are being fit using a PDF that matches the width and offset of the peaks to the dead layer of the detector. This information will be used to validate surface-alpha decay simulations and infer purity requirements. 0o 30o Counts Counts 800 1500 600 1000 400 200 500 0 0 5100 5200 5300 5400 5500 5600 5100 5200 5300 5400 5500 5600 Energy [keV] Energy [keV] 45o 60o Counts Counts 600 600 400 400 200 200 0 0 5100 5200 5300 5400 5500 5600 5100 5200 5300 5400 5500 5600 Energy [keV] Energy [keV] Figure 1.18-1. Alpha spectra from 241 Am at four different incidence angles. Alphas at larger incidence angles must travel through more dead regions of the crystal, and therefore lose more energy and are subject to more energy straggling. ∗ Los Alamos National Laboratory, Los Alamos, NM † University of North Carolina, Chapel Hill, NC


  • Page 36

    UW CENPA Annual Report 2008-2009 May 2009 23 1.19 Methods for deploying ultra-clean detectors T. H. Burritt, S. R. Elliott*, V. M. Gehman*, V. E. Guiseppe∗ A. G. Schubert, and J. F Wilkerson† The MAJORANA Collaboration will use an array of germanium crystals enriched in 76 Ge to search for neutrinoless double-beta decay. The germanium crystals will be housed in radiologically-pure copper cryostats. Several vertical strings of four or five Ge crystals will be deployed in each cryostat. Strings for the MAJORANA experiment must be constructed from ultra-pure materials and maintain proper thermal and electrical operating conditions for Ge detectors. A preliminary design for the building block of a detector string, a detector mount, is shown in Fig. 1.19-1. The intent is to minimize the amount of material near the detector and to minimize handling of the parts during fabrication. A test cryostat at Los Alamos National Laboratory has been used to test electrical and mechanical properties of string designs for MAJORANA. Several temperature sensors within the cryostat are used to evaluate the thermal performance. Temperature data taken while cooling the system with liquid nitrogen has been used to study the flow of heat from detector blanks and has provided information about the conduction of heat through the detector string. Figure 1.19-1. A design for a detector mount to hold a 70 mm diameter Ge detector. The detector is shown in grey; the detector mount is created from ultra-pure copper and small amounts of electrically-insulating material. ∗ Los Alamos National Laboratory, Los Alamos, NM † University of North Carolina, Chapel Hill, NC


  • Page 37

    24 2 Fundamental Symmetries and Weak Interactions Torsion Balance Experiments 2.1 Charge measurement for gravitational wave observatories J.H. Gundlach, C.A. Hagedorn, J.L. McIver, S.E. Pollack∗, S. Schlamminger, M. Turner Electrical charges on the end masses of interferometric gravitational wave observatories are an important noise source for the planned space based gravitational wave antenna LISA, as well as for existing ground based observatories like LIGO. In both systems an interferometer compares the light travel time between two pairs of end masses. The end masses are inertial along the direction of the light and an incoming gravitational wave will alter the light travel time in one arm with respect to a second arm. For LISA and LIGO, the end masses are electrically isolated. For LISA, the isolation occurs because the end mass is freely falling inside a spacecraft which is servoed such that the mass is essentially force free. For Advanced LIGO, the end masses are suspended from insulating fused silica fibers, which have been chosen for their low mechanical loss. In both cases, fluctuating charge on the end masses can couple to conductive surfaces nearby and produce small spurious accelerations, limiting the sensitivity of the gravitational wave antenna. We have been using our torsion balance exper- iment to investigate such charge fluctuations. Fig. 2.1-1 shows the geometry of the experi- ment. A silicon pendulum is suspended from a quartz fiber. Two copper plates are placed in the vicinity of the pendulum, and the gap between plates and pendulum can be varied from 0.05 mm to 10 mm. The angular excur- sion of the pendulum is measured by an au- tocollimator and processed by a digital feed- back loop. The feedback loop creates a con- trol voltage that is applied to either one of the larger electrodes such that the pendulum re- mains parallel to the copper plates. The sign of the control voltage can be selected by the user and the measured control voltage is recorded as the science signal. In order to avoid large Figure 2.1-1. The geometry of the torsion contact potentials, all surfaces have been gold balance experiment. plated. In the past year we have established an electrostatic model of the geometry described above. This model has been tested with measurements and a finite element analysis. The ∗ Presently at Rice University, Houston, Tx.


  • Page 38

    UW CENPA Annual Report 2008-2009 May 2009 25 0 10 -1 10 Charge [pC/rtHz] -2 10 10-3 -4 10 Measured charge fluctuations LISA requirement 10-5 0.1 1 10 Frequency (mHz) Figure 2.1-2. A preliminary measurement of the power spectral amplitude of the charge on the pendulum. The thick solid line is the LISA requirement. model, experiment, and simulation agree within their uncertainties. We have developed a procedure to infer the charge on the pendulum by holding the pendulum parallel to the plate with a positive control voltage V+ and then with a negative voltage V− . The charge on the pendulum is given by Q = C(V+ + V− ), where C is the capacitance between the pendulum and one copper plate. The data acquisition software has been altered to automatically change the polarity of the feedback voltage every 20 s. The mean feedback voltage and the known capacitance C is used to calculate the charge fluctuation. A preliminary measurement of the power spectral amplitude of the charge fluctuation is shown in Fig. 2.1-2. In the measured √ frequency range, the charge fluctuations appear to be white with a value of 0.2 pC/ Hz. There are several noise sources that may be contributing to the measured level of charge fluctuations; therefore, this measurement is an upper limit of the level of charge noise which may be expected for LISA. We are currently characterizing our systematics, including surface potential fluctuations on the nearby conducting surfaces, thermally activated torque noise in the quartz fiber, and seismic couplings. In addition we are preparing schemes to actively discharge the pendulum and hence reduce the noise in the system.


  • Page 39

    26 2.2 Progress on improved equivalence principle limits for gravitational self-energy E.G. Adelberger, J.H. Gundlach, B.R. Heckel, S. Schlamminger, T.A. Wagner Lunar laser ranging now makes range measurements with 1 mm uncertainty1 . This improve- ment will result in better constraints on equivalence principle violations. However, an ambi- guity exists in testing equivalence principle violations with the Earth-Moon system because the Earth and Moon have significant gravitational self-energy as well as different composi- tions. In the laboratory we test for equivalence principle violation based only on composition differences since the test bodies have insignificant gravitational self-energy. Combining the lunar laser ranging and laboratory measurements allows for separate limits on equivalence principle violations due to gravitational self-energy and composition differences. We test for composition dependent equivalence principle vi- olation using a rotating torsion balance. The pendulum is con- Fiber Positioning Stage figured with a composition dipole based upon the differences in the Earth’s and Moon’s compositions. In previous measure- ments, our torsion balances have operated near the thermal Support limit imposed by the torsion fiber. However, the pendulum Plate used for the Earth-Moon test experienced noise about a factor Upper Support Wings Optical of 3 greater than this. The dominant noise source was traced to Readout extremely small pressure bursts in our vacuum chamber. These Vacuum Pump pressure bursts occurred roughly once an hour. The pressure increased to about two times the base pressure of 5 × 10−5 Pa, but was pumped back down within 20 s. Extensive testing of our vacuum system led to a redesign of some of the O-ring seals. During the redesign, we took the opportunity to stiffen Vacuum the central column as well, as shown in Fig. 2.2-1. Chamber The new vacuum chamber was designed to minimize changes to our apparatus, while improving the O-ring seals and avoiding new weldments. With the redesign we improved mechanical Tilt Sensor Enclosure stiffness and immunity to temperature gradients. Two sliding Figure 2.2-1. New design for O-ring seals on the central column were replaced with more our vacuum chamber column. reliable compression O-ring seals. The column wall is a factor of six thicker, and the thickness of the rocket-wing like support fins was doubled. These changes resulted in more than an order of magnitude increase in the stiffness of the vacuum chamber column. The seal between the column and the fiber positioning stage was separated from structural support of the column. The fiber support stage rests on top of the column which extends through the fiber support plate. There is a gap between the column and fiber support stage to reduce thermal coupling. We previously determined that the top support plate was the location of the apparatus’ greatest sensitivity to temperature gradients. We expect to resume taking data with the Earth-Moon test bodies in May. 1 Battat, J. B. R., et al., PASP, 121, 29 (2008)


  • Page 40

    UW CENPA Annual Report 2008-2009 May 2009 27 2.3 Continued progress toward a new sub-millimeter test of the gravitational inverse square law J.H. Gundlach, C.A. Hagedorn, S. Schlamminger, M. D. Turner We have continued work on a new torsion balance for a parallel plate test of the gravitational inverse square law1,2 . Our experiment is within reach of new limits, but further character- ization of systematic effects is required. We have undertaken many improvements to the apparatus and our data analysis software to increase the experiment’s intrinsic sensitivity, close systematic loopholes, and more cleanly extract signals from the data stream. Important improvements include: (1) replacement of the aluminum isolating foil with beryllium copper; (2) fabrication and installation of an improved attractor plate with sig- nificantly smaller gravitational field inhomogeneities; (3) improvements in alignment and cleaning techniques; (4) the installation of a commercial in-vacuum encoder on the attractor to supplement and, recently, replace our simple optical readout; (5) acquisition and instal- lation of an electrically isolated data acquisition and control system for attractor-related signals; (6) installation of additional feedthroughs and cabling to further segregate signaling; (7) improved internal grounding and added optical isolation for critical feedback voltages; and (8) the fabrication and installation of new thermal insulation for the bell jar. Furthermore, intensive gravitational simulation work to both optimize our attractor design and predict expected Newtonian torques has reached improved levels of sophistication and detail. These improvements have brought us closer to a successful p measurement. The torque noise in our region of interest (≈ 2-10 mHz) is < 2 × 10 −14 N-m/ (Hz) at ≈ 50 µm pendulum-foil separation. We are able to achieve pendulum-attractor separations of 60 µm while maintain- ing feedback lock at an increased noise level. We anticipate releasing our first results in the coming year. Calibration signal -12 Torque Spectral Amplitude (N-m/√(Hz)) 10 Science frequency -13 10 -14 10 Approximate Thermal Limit -15 10 10-4 10-3 10-2 Frequency (Hz) Figure 2.3-1. Torque noise performance with 50 micron pendulum-foil separation. 1 CENPA Annual Report, University of Washington (2007) p. 40. 2 CENPA Annual Report, University of Washington (2008) p. 33.


  • Page 41

    28 2.4 A cryogenic torsion balance for gravitational experiments E. G. Adelberger, F. Fleischer and B. R. Heckel A fundamental limiting factor for our torsion balance experiments is thermal noise. To overcome this limitation, we have started a new project aimed at the development of a torsion balance experiment operating at cryogenic temperatures. Because of its relative simplicity, its first application will be a non-rotating equivalence principle test. We have developed a setup which allows to cool down a torsion balance to temperatures near liquid helium temperature (4.2 K). Due to the extreme sensitivity of these experiments, special consideration was given to maintaining a low level of vibrations. A two-stage pulse tube cooler made by Sumitomo (model RP-062B) has been chosen for its low-vibration char- acteristics, and the setup has been designed to provide vibrational isolation between the cryo-cooler and the vacuum chamber containing the torsion balance. Fig. 2.4-1 shows a schematic cross section of the experiment. The vacuum vessel contains two thermal shields, an outer one at an intermediate temperature of ≈ 90 K and an inner one at ≈ 4 K. The inner shield will hold the torsion balance, consisting of a composition dipole pendulum hanging from a thin conductive fiber. Both shields are connected to the corresponding stages of the pulse tube cooler via flexible heat links made from copper braids. Furthermore, to insure good vibrational decoupling, the cold head is supported by a separate wall mount featuring a set of three air springs with a low resonance frequency of ≈ 3 Hz. A laser autocollimator will be used for monitoring the torsion balance’s movement. The construction of the actual experimental setup started in the fall of 2008. By the end of the report period, the structural parts, the vacuum system, the flexible heat links and the thermal shields as well as a system for montoring the temperatures of different parts using silicon diode temperature sensors were completed and assembled. The vacuum system has been successfully evacuated to pressures in the upper 10−7 mbar range at room temperature. The first cool-down test for setup was a success. ➇ ➇ ➇ ➇ ➉ ➉ ➉ ➉ air springs ➈ ➈ ➈ ➈ ➊ ➊ ➊ ➊ ➇ ➇ ➇ ➇ ➉ ➉ ➉ ➉ ➈ ➈ ➈ ➈ ➊ ➊ ➊ ➊ ➇ ➇ ➇ ➇ ➉ ➉ ➉ ➉ ➈ ➈ ➈ ➈ ➊ ➊ ➊ ➊ ➇ ➇ ➇ ➇ ➉ ➉ ➉ ➉ ➈ ➈ ➈ ➈ ➊ ➊ ➊ ➊ independent cold ➇ ➇ ➇ ➇ ➉ ➉ ➉ ➉ ➈ ➈ ➈ ➈ ➊ ➊ ➊ ➊ ➇ ➇ ➇ ➇ ➉ ➉ ➉ ➉ ➈ ➈ ➈ ➈ ➊ ➊ ➊ ➊ edge−welded ❽ ❽ ❽ ❾ ❾ ❿ ➀ ❿ ➀ ⑩ ⑩ ⑩ ⑩ ⑩ ⑩ ❷ ❷ ❷ ❷ ❷ ❷ ❷ ❷ ❷ ❷ ❷ ❷ ❷ ❷ head support bellows ⑧ ⑨ ⑧ ⑧ ⑨ ⑧ ⑧ ⑨ ⑧ ⑧ ⑨ ⑧ ❶⑩ ❶⑩ ❶⑩ ❸❷ ❷❸ ❷❸ ❸❷ ❸❷ ❸❷ ❸❷ motorized pendulum ⑨ ⑨ ⑨ ⑨ ❶ ❶ ❶ ❸ ❸ ❸ ❸ ❸ ❸ ❸ ➁⑧ ➁⑧ ➁⑧ ➁⑧ ➁ ➁ ➁ ➂⑨ ➂⑨ ➂⑨ ➂⑨ ➂ ➂ ❶ ➂ ❶ ❶ ❸ ❸ ❸ ❸ ❸ ➆ ❸ ➆ ❸ ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ levelling ➁ ➂ ➁ ➂ ⑥ ➁ ➂ ➁ ➂ ⑥ ⑦ ➁ ➂ ➁ ➂ ⑥ ⑦ ➁ ➂ ➁ ➂ ⑥ ⑦ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ④ ⑤ ➁ ➂ ➁ ➂ ④ ⑤ ④ ⑤ ② ③ ② ③ ② ③ ② ③ ➅ ➆ ➅ ➆ ➅ ➆ ➅ ➆ suspension mounts ⑥ ⑥ ⑥ ⑥ ④ ④ ④ ② ② ② ② ⑦ ⑦ ⑦ ⑤ ⑤ ⑤ ③ ③ ③ ③ ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ wall mount ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ flexible ➁ ➂ ➁ ➂ ➁ ➁ ➂ ➁ ➂ ➁ ➁ ➂ ➁ ➂ ➁ ❻ ❼ ❻ ❼ ➁ ➂ ➁ ➂ ➁ ❻ ❼ ❻ ❼ ➁ ➂ ➁ ➂ ➁ ❻ ❼ ❻ ❼ ➁ ➂ ➁ ➂ ➁ ❻ ❼ ❻ ❼ ➁ ➂ ➁ ➂ ➁ ❻ ❼ ❻ ❼ ❻ ❼ ❻ ❼ ❹ ❺ ❹ ❺ ❹ ❺ ❹ ❺ ❹ ❺ ❹ ❺ ➅ ➆ ➅ ➆ ➅ ➅ ➆ ➅ ➆ ➅ pulse tube cooler ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ Cu heat links ➁ ➂ ➁ ➁ ➂ ➁ ➁ ➂ ➁ ➁ ➂ ➁ ➁ ➂ ➁ ➁ ➂ ➁ ➁ ➂ ➁ ➅ ➆ ➅ ➅ ➆ ➅ magnetic damper ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ G10 spacers ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ vacuum chamber ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ multi−layer ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➅ ➆ ➅ ➆ ➅ ➆ ➅ ➆ insulation ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➅ ➆ ➅ ➆ ➅ ➆ ➅ ➆ ➅ ➆ ➅ ➆ fiber ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ outer ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ autocollimator ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➅ ➅ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➆ ➆ ➃ ➃ ➃ ➃ ➃ ➃ ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➄ ➄ ➄ ➄ ➄ ➅ ➅ thermal shield ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➃ ➃ ➃ ➃ ➃ ➃ viewport ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➄ ➄ ➄ ➄ ➄ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➃ ➃ ➃ ➃ ➃ ➃ ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➄ ➄ ➄ ➄ ➄ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➃ ➃ ➃ ➃ ➃ ➃ ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➄ ➄ ➄ ➄ ➄ inner ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➃ ➃ ➃ ➃ ➃ ➃ ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➄ ➄ ➄ ➄ ➄ ➂ ➂ ➂ ➂ ➂ ➂ ➂ ➃ ➃ ➃ ➃ ➃ ➃ ➁ ➁ ➁ ➁ ➁ ➁ ➁ ➄ ➄ ➄ ➄ ➄ ➂ ➂ ➂ ➂ ➂ ➂ ➂ thermal shield ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➁ ➂ ➃ ➃ ➃ ➄ ➃ ➄ ➃ ➄ ➃ ➄ ➃ ➄ ➃ ➄ ➃ ➄ ➃ ➄ ➃ ➄ ➃ ➄ pendulum pumping port (to turbo pump) Figure 2.4-1. A schematic cross sectional view of the cryostat setup for the cryogenic torsion balance.


  • Page 42

    UW CENPA Annual Report 2008-2009 May 2009 29 2.5 Wedge pendulum progress report: testing the gravitational inverse- square law E. G. Adelberger, T. S. Cook and H. E. Swanson Last year we reported solving our technology hurdles and were preparing to take data1 . As it turns out we were only partially correct in our assessment. While our data collection is now nearly complete, we were forced to abandon one piece of technology due to its failure to work reliably. This past year has also been marked by the development of code to calculate torques for non-aligned geometry. We began taking data with the Wedge pendulum in July of 2008. About a month into the process our Nanomotion motor began to exhibit erratic behavior showing large swings in temperature and ultimately losing the ability to turn in one direction. We decided it was no longer worth our time to pursue this technology and returned to our previous method of rotation, a stepper motor. Fortunately, with only minor modifications we were able to use the feedback system designed for the Nanomotion motor to run the stepper motor in lock-step with our angle encoder. While not as precise as the Nanomotion motor (by approximately a factor of 2), the fed-back stepper motor still represents an improvement in angle error of over two orders of magnitude when compared to the stepper motor run with a clock pulse (the method used to run the stepper in previous incarnation of the short range experiment). One feature of the wedge geometry is that it allows for a nearly analytic solution for calculating torques. This solution involves a single integral over some Bessel functions which result from writing down the Green’s function for a gravity-type or Yukawa-type potential in cylindrical coordinates. When the pendulum is not exactly aligned with the attractor, however, the cylindrical coordinate system is broken and the solution now requires more complicated integrals over the geometry. To calculate torques for such alignments (which is the more accurate physical description of our experiment), we developed code in C++ that runs on the Athena Cluster (see Sec. 7.5), a highly parallel computer located at CENPA, to perform the Monte Carlo integrations. Two versions of the code were developed: (1) a simple yet time consuming point-by-point integration over both attractor and pendulum and (2) an extension of the Bessel function solution that integrates the analytic solution of the attractor for each point in the offset pendulum. The Athena Cluster allows for extreme parallelization of the calculation to quickly build statistical error bars within our desired accuracy. Current testing shows perfect agreement between methods (1) and (2) (both of which agree with the analytic solution on center), with the Bessel method converging about 8 times faster. We are now taking data, exploring changes both in separation distance and in radial alignment. We anticipate ending data taking and performing a complete analysis of the data shortly. 1 CENPA Annual Report, University of Washington (2008) p. 28.


  • Page 43

    30 2.6 Progress on a torsion balance test of new spin coupled forces E. G. Adelberger, B. R. Heckel and W. A. Terrano The torsion balance experiment to measure spin dependent forces1 is nearly ready for com- missioning. All that remains is assembling it in the vacuum system with the torsion fiber attachments. Since our pendulum will have large numbers of aligned spins in order to be sensitive to spin coupled forces, an essential component to the experiment is demonstrating that its magnetic couplings can be controlled. To this end, after assembling the ring we tune the magnetizations of each individual segment. This reduces the measured peak to peak magnetic field 70 thousandths of an inch above the rings from 600 G to 30 G (Fig. 2.6-1). Housing the magnets under two layers of magnetic shielding, as in the experimental design, further reduces the measured magnetic field to ≈ 1 mG (Fig. 2.6-2). Since we are only interested in torques at 10 times the rotation frequency of the attractor, we can take the Fourier transform of these traces to get an idea of the shielding at the science signal. The torque on the pendulum depends on the product of the magnetic fields of the pendulum and attractor, and so we can get a very rough estimate of the shielding factor by comparing 5 · 103 G of spin contrast every 18 ◦ to a magnetic field variation of 4 · 10−5 G at the same frequency outside the ring. This means we have room to see spin coupled interactions down to 10−16 compared to magnetic interactions. This is compatible with our experimental goal. Figure 2.6-1. Left: Before tuning the Alnico magnetization. Right : After tuning Figure 2.6-2. With magnetic shielding 1 CENPA Annual Report, University of Washington (2007) p. 38.


  • Page 44

    UW CENPA Annual Report 2008-2009 May 2009 31 2.7 Designing an equivalence principle pendulum with hydrogen rich test bodies E. G. Adelberger, J. H. Gundlach, B. R. Heckel, S. Schlamminger, W. A. Terrano One of the deepest puzzles of contemporary physics is the nature of Dark Matter. Despite the fact that Dark Matter appears to be around five times more prevalent than ordinary matter we have very little experimental information about its behaviour and interactions. There are limits on weak and electromagnetic point contact interactions from WIMP and axion detection experiments. Galactic rotation curves, large scale structure and CMB/BBN con- siderations indicate that Dark Matter interacts gravitationally with hydrogen. An interesting question then is the nature of long range interactions between dark matter and other ma- terials. Such an additional, non-gravitational, force would appear similar to an Equivalence Principle violation, pointed in the direction of the galactic dark matter halo. To investigate such interactions, we are gearing up for a version of the classic torsion balance tests of the Equivalence Principle with test bodies that are rich in Hydrogen. The major difficulty at this point is working with materials that have a high concentration of Hydrogen. At this time we have designed and are building a prototype that uses Ultra-High- Molecular-Weight Polyethylene. A cross section of the prototype is shown below. Figure 2.7-1. The Polyethylene is in dark grey, and the Aluminum test masses are in light grey. They each make up one half of a cylindrical shell. The unshaded regions are an aluminum shell that surrounds the entire pendulum for electrostatic shielding and the rods that attach the pendulum to the torsion fiber.


  • Page 45

    32 Weak Interactions 2.8 Production of 6 He to determine the e-ν e correlation G. C. Harper, A. Garcı́a, Z. T. Lu∗ , P. Muller∗ , R. G. H. Robertson, C. Wrede, and D. W. Zumwalt We are planning to develop a laser trap for 6 He with the aim of determining the e-ν e cor- relation with unprecedented precision. We are presently building an apparatus to serve as a production target to produce an intense source (109 s−1 ) of 6 He. We plan to use the 7 Li(d,3 He)6 He reaction on molten Li from a 12 MeV beam of deuterons at a current of order 10 µA from the CENPA Van de Graaff. This should allow sufficient 6 He production, but it is unclear how fast the 6 He will be transmitted to the trapping site, so we plan to test this system. Figure 2.8-1. Lithium cup for production of 6 He. We are using a cylindrical cup made of stainless steel (volume roughly 70 cm3 ) to contain the molten Li. Lithium does not have any corrosive reaction with stainless steel, so it is a good material for a holding cell. The cylinder will then be partially milled flat on two sides. On one side, we will drill a hole into the cup and attach over the hole a stainless steel foil through which the deuteron beam can pass (see Fig. 2.8-1). This attachment will be made by clamping the foil onto the flat face of the cylinder using a custom Stainless Steel flange with a semi-circular ridge machined around the face of the flange. On the other flat face, a ∗ Physics Division, Argonne National Laboratory, Argonne, IL


  • Page 46

    UW CENPA Annual Report 2008-2009 May 2009 33 block of copper will be attached for use in regulating the temperature to melt or solidify the lithium. Copper is used here for its good thermal conductivity properties. An electrical cartridge heater will be placed inside the copper block and will be controlled by a temperature regulator attached to the cup. We will also have two cooling-gas lines attached to the copper block, an inlet and an outlet. In the event of overheating, compressed air will be used to cool the copper block, and the power to the cartridge heater will be shut off. This is important because our beam current can be of order 10 µA, and so at 12 MeV we would be introducing power of order 100 W. To isolate our apparatus thermally and electrically, we have used ceramic breaks. There are three isolators used for our design. The first is a ceramic electrical isolator used on the electrical feed-through port, which prevents electrical noise from interfering with the reading on the temperature controller. The other two are for both thermal and electrical isolation. The primary insulator is attached to the main outlet of the cup, which will then transfer the 6 He into a turbopump and into an experimental area. The other two insulators form a break in the two cooling lines. Initial testing will address concerns regarding the integrity of the stainless steel foil after the molten lithium cools. If the foil breaks, we have alternative (more complicated) designs that we will consider.


  • Page 47

    34 2.9 Measurement of the neutron beta asymmetry with ultracold neutrons A. Garcı́a, S. Hoedl, A. L. Sallaska, S. K. L. Sjue∗ , C. Wrede, and UCNA collaboration† Precision measurements of the neutron half life and β asymmetry parameter A0 together could provide a value for the CKM quark-mixing matrix-element Vud that is competitive with, and complementary to, the most precise value that is determined from superallowed 0+ → 0+ β decays. Traditionally, measurements of A0 have been made using polarized cold neutrons (ECN < 25 meV) that include substantial systematic uncertainties related to their polarization and self-induced background via activation of the apparatus. To reduce these uncertainties the UCNA collaboration made the first measurement of A0 using polarized ultra-cold neutrons (EU CN < 340 neV) at the Los Alamos Neutron Science Center in 2006 and 2007. Results of these limited post-commissioning measurements were published in 2009 and yielded A0 = −0.1138(0.0046)stat(0.0021)syst, in agreement with the world average of A0 = −0.1173(0.0013) from the Particle Data Group that is based entirely on cold-neutron measurements. In 2008 the efforts of the UCNA collaboration were focused on reducing both the statistical and systematic uncertainties to achieve a 1% measurement of A0 . Statistical uncertainties have been reduced by a factor of ≈ 5 by increasing the proton production-beam current, improving neutron transport from the UCN source to the UCNA spectrometer, and running for a much longer period of time. Systematic uncertainties have been improved by a factor of ≈ 4 by carrying out several dedicated systematic studies. For example, Monte Carlo studies of electron backscattering and angle-dependent energy loss have been tested by running in different geometries, depolarization studies have been refined, and the energy calibration has been improved by using new calibration sources (section 2.11). Based on these improvements we expect to report a sub-1% measurement of A0 for 2008 (see Table 2.9-1), with a goal of 0.3% for the future. 2007 2008 Statistics 4.0% < 0.8% Polarization 1.3% < 0.4% Energy calibration 1.5% 0.3% Angle effects 0.5% 0.2% Backscattering 0.4% 0.2% Total 4.5% < 1.0% Table 2.9-1. Uncertainties contributing to the measurement of A0 by the UCNA collabora- tion in 2007 and 2008. The 2008 numbers are preliminary. ∗ Present address: TRIUMF, Vancouver, BC, Canada † A. Saunders, A. Young spokespersons; the collaboration is formed by approximately 30 scientists from Caltech, Duke Univ., Idaho State Univ., Univ. of Kentucky, Los Alamos National Lab., North Carolina State Univ., Texas A&M Univ., Virginia Tech, Univ. of Washington and Univ. of Winnipeg.


  • Page 48

    UW CENPA Annual Report 2008-2009 May 2009 35 2.10 Characterization of ultracold neutron detectors for use in the UCNA experiment at LANL: conclusions A. Garcı́a, S. A. Hoedl, and A. L. Sallaska We have fabricated ultracold neutron (UCN) detectors that consist of silicon charged particle detectors coupled with thin nickel foils coated with either natural LiF or 10 B implanted into vanadium. The foils convert neutrons into energetic, charged particles which readily produce measurable signals in silicon detectors. The detectors were tested with a gravitational spectrometer at the Institut Laue-Langevin, and both detectors and experimental setup have previously been described in detail1 . The central problem is that neutrons will only be detected if they penetrate the surface (or surface window) of the detector. This will occur if the velocity of the UCN is greater than a “cutoff velocity,” defined by the effective potential barrier of the foil. The analysis to determine the values of the cutoff velocities for the foils has concluded and has been accepted for publication2 . Because of the availability of the Athena cluster (see Sec. 7.5), the Monte Carlo simulation portion of the analysis was extended to include a more accurate treatment of the finite implantation distribution of 10 B implanted into vanadium, as well as widening the scope of systematic error checks in order to fix simulation parameters accurately. The vast extent of the parameter space probed would not have been possible without access to the cluster. The minimum detection cutoff velocities (effective potentials) were determined to be 309±17 cm/s (49.8±2.7 neV) for LiF and 367±39 cm/s (70.3±7.5 neV) for 10 B in vanadium. Although the result for LiF is consistent with expectations, the result for 10 B in vanadium is significantly higher. We interpret this discrepancy as due to contam- ination. We also show that while a thicker foil is more efficient for UCN detection, a thinner foil is more suitable for determining the cutoff velocity. 1 CENPA Annual Report, University of Washington (2007) p. 50 2 10.1016/j.nima.2009.02.014


  • Page 49

    36 114m 2.11 Development of a In calibration source for the UCNA experiment A. Garcı́a, G. C. Harper, A. Palmer, D. I. Will, C. Wrede, and UCNA collaboration Uncertainties in detector response and nonlinearity contributed the largest systematic uncer- tainty (1.5%) to the 2007 measurement of A0 by the UCNA collaboration (section 2.9). A goal of the 2008 measurements was to reduce the total uncertainty in A0 to 1%, which required a substantial improvement in the energy calibration. The 2007 calibration relied on conversion electrons from 113 Sn and 207 Bi sources, which provided good calibration points for the high- energy portion of the neutron β-decay spectra (Eβ > 400 keV), but a low-energy source was not available. The scarce availability of commercial calibration sources that produce intense low-energy conversion-electron lines suitable for detector calibrations of this sort prompted us to develop a 114m In calibration source (Eβ < 200 keV) at CENPA in collaboration with North Carolina State University (NCSU). 45000 40000 35000 30000 25000 20000 15000 10000 5000 0 0 200 400 600 800 1000 1200 1400 1600 1800 Energy (arbitrary units) Figure 2.11-1. Raw energy spectrum of 114mIn source measured with UCNA scintillator detector. The peak corresponds to conversion electrons from the decay of 114mIn to 114g In and the continuum is from the subsequent β decay of 114g In to 114 Sn. The general procedure was to prepare samples containing 113 In at CENPA and ship them to NCSU for irradiation in the PULSTAR research reactor to produce 114mIn (T1/2 = 50 days) via the (n, γ) reaction. A sample of 113 In was initially prepared by vacuum evaporation of natural indium (4.3% 113 In, 95.7% 115 In) onto a thin Kapton substrate and irradiated to produce 114m In successfully. However, evidence for the oxidation and migration of the exposed indium during irradiation was observed afterwards. Attempts to encapsulate the evaporated In between evaporated layers of other elements in the next iteration of source development improved the situation, but showed that migration was still possible. This prompted us to prepare 113 In samples via ion implantation on the low-energy beam line of CENPA’s tandem Van de Graaff instead. A 100 nA, 113 In16 O− molecular ion beam was mass selected with a 90◦ magnet, accelerated to 45 keV, and magnetically rastered over a 3 mm collimator to implant a 3 mm diameter, 5 µg/cm2 sample of 113 In in a thin aluminized-mylar substrate. This sample was irradiated and used in December of 2008 to calibrate the UCNA scintillators. Subsequent analysis of the acquired spectra has shown that the source fabrication was successful. The 114mIn spectra will be used with spectra from other sources to improve the energy-calibration uncertainties in A0 to ≈ 0.3% for the 2008 runs.


  • Page 50

    UW CENPA Annual Report 2008-2009 May 2009 37 2.12 Parity non-conserving neutron spin rotation experiment C. Bass∗ , K. Gan†, B. R. Heckel, D. Luo‡ , D. Markoff§, H. P. Mumm∗ , J. Nico∗ , A. Opper† , W. M. Snow‡, and H. E. Swanson The parity non-conserving (PNC) neutron spin rotation experiment1 was completed in June, 2008. The apparatus was subsequently removed from the NIST beam line and shipped to Indiana University. The experiment ran for 3 reactor cycles or about 106 calendar days. Near the end of the last cycle a coupling on the pump shaft broke and we were no longer able to transfer liquid Helium between target chambers. We decided against making a repair and instead continued to investigate systematics until the end of the cycle using cold Helium and Nitrogen gases. The hold time for liquid Helium in the targets was 8 hours which determined the length of each run. We started the analysis phase by reviewing the entire data set and discarded data where target chambers failed to completely fill, the ion chamber showed signs of sparking, or there were abrupt changes in residual magnetic fields. The surviving data set amounted to 21 days of live running. The weak interaction produces small rotations of the neutron spins as they pass through liquid Helium targets. The approximately 100 µGauss residual field produces a much larger precession but the design of the experiment makes the measured PNC angle relatively in- sensitive to this field. Periodically throughout the run set a known angle was applied to the polarized neutrons and the response of the apparatus was measured. This calibration turned out to be important in maintaining the precision angle measuring capabilities of the appara- tus as well as its insensitivity to various systematic effects. Slow drifting of the residual field however can be a significant source of systematic error. To reduce the effects of varying fields we employ two independent strategies. The first of these uses linear regression on the data in each run where the angle is then determined from the residuals. This analysis is being carried out at NIST. The second uses a sliding filter (see Sec. 7.7) applied to sequential data that removes any linear or quadratic dependence prior to obtaining the angle. This later method has been part of the analysis effort at CENPA. Both of these should give similar results and comparing them will contribute to our understanding of the drift systematic. We searched the surviving data set for any significant correlations between PNC precession angle and target Helium levels, residual magnetic fields, detector segments, or neutron velocity and none have been found. Preliminary results show no resolved angle and that we are limited by statistical errors consistent with neutron shot noise. ∗ NIST Center for Neutron Research, 100 Bureau Drive, Stop 8461, Gaithersburg, MD 20899-8461 † The George Washington University, Department of Physics, Corcoran 105, 725 21st St, NW, Washington, DC 47408 ‡ Department of Physics, Indiana University Cyclotron facility, 2401 Milo B Sampson Land, Bloomington, IN 47408 § North Carolina State University, TUNL, Raleigh, NC 27695 1 CENPA Annual Report, University of Washington (2003) p. 5.

  • View More

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