avatar University of Washington Inc Services

Pages

  • Page 1

    AER CENPA Center for Experimental Nuclear Physics and Astrophysics EEEEEEES SES University of Washington


  • Page 2

    ANNUAL REPORT Center for Experimental Nuclear Physics and Astrophysics University of Washington April, 2015 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. Cover design by Gary Holman. The background design on the front and back cover is a rendering of a g ´ 2 silicon photo-multiplier (SiPM) (Sec. 4.7). In the top left photo, Arnaud Leredde inspects part of the detection system for the 6 He experiment (Sec. 3.1). The top right photo depicts graduate student John Lee repairing the apparatus for the short range spin-coupled force experiment (Sec. 2.5). In the bottom left photo, graduate student Rachel Ryan is assembling the MuSun time-projection chamber in preparation of the 2014 data run at the Paul Scherrer Institute (Sec. 4.20). In the bottom right photo, postdoc Luiz de Viveiros (UC Santa Barbara) working on the Project 8 (Sec. 1.8) superconducting magnet in the basement of the UW Physics-Astronomy Building.


  • Page 4

    UW CENPA Annual Report 2014-2015 April 2015 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 and related fields. Research activities are conducted locally and at remote sites. The research program emphasis is fundamental symmetries and neutrinos. In neutrino physics, CENPA is the lead US institution in the KATRIN tritium beta decay experiment, the site for experimental work on Project 8, and a collaborating institution in the MAJORANA 76 Ge and the SNO+ 130 Te double beta decay experiments. The Muon Physics group is running the MuSun experiment to measure muon capture in deuterium at the Paul Scherrer Institute in Switzerland. The group has a leadership role in the new project to measure the anomalous magnetic moment of the muon at Fermilab to even higher precision than it is presently known from the collaboration’s previous work at Brookhaven. The fun- damental symmetries program also includes “in-house” research using the local tandem Van de Graaff accelerator with an experiment to measure the electron-neutrino correlation and Fierz interference in 6 He decay. In addition to the research directly supported by DOE’s Office of Nuclear Physics through the CENPA core grant, other important programs are located in CENPA, forming a broader intellectual center with valuable synergies. The ‘Gravity’ group, as it is known, carries out with both DOE and NSF support studies of the weak and strong Equivalence Principles, fundamental precepts of General Relativity, as well as searches for non-Newtonian weak forces such as are predicted by theories with extra dimensions. The DOE Office of High Energy Physics supports a unique experiment, the ADMX axion search, now in an extended data-taking phase. These unique CENPA activities have generated an unusual spinoff, a successful program on nanopore DNA sequencing, led by Jens Gundlach and supported by NIH. The past year has marked another exciting one in accomplishment and recognition. As- sistant Professor Jason Detwiler received the Physics Department Mentoring Award. Senior Research Scientist Erik Swanson received a Fermilab Intensity Frontier Fellowship, given to outstanding researchers in neutrino and muon physics. Todd Wagner successfully defended his thesis in December. Dmitry Lyapustin and Jared Kofron defended theirs, both on St. Patrick’s day in 2015, and we congratulate them all. Postdoctoral Fellow Diana Parno was appointed Acting Assistant Professor and Associate Director of CENPA, and has in a few short months made many improvements as well as taking on much of the day-to-day operation of CENPA. Diana takes over from Associate Director and technical guru Greg Harper, who retired in August, but lives in Seattle and has not disconnected his phone. Nomie Torres joined us in December from the University of Alaska in Fairbanks. Nomie is a Fiscal Specialist 2 with our Administrator Victoria Clarkson in the front office, and has contributed greatly to the operations side of CENPA. Hannah LeTourneau signed on in August as a Research Engineer to manage the new ADMX helium liquefier. It has also been a time of transition. Tom Trainor retired in December and is now Research Professor Emeritus. This brings to a close our long and noteworthy program on relativistic heavy-ion physics. Assistant Professor Nikolai Tolich resigned his position this year and


  • Page 5

    ii Postdoctoral Fellow Joulien Tatar, who joined us in March 2014 to work on SNO+ with Nikolai, is now at UC Berkeley pursuing his outstanding research on Cherenkov radiation from ice in Antarctica. Muon group postdoc Jason Crnkovic departed in April and Pete Alonzi departed in September. Jason is a Research Associate at Brookhaven, and Pete is now Senior Research Data Scientist in the University of Virginia Library System. ADMX group postdoc Michael Hotz departed in December to join Nion, Inc., and Andrew Wagner departed in January for Raytheon, Inc. Ana Malagon arrived in September to join the ADMX Group as a postdoc. In March, 2015, Kim Siang Khaw arrived to join the Muon Group as a postdoc. Kim Siang follows in the footsteps of Andreas Knecht and Martin Fertl, the third graduate of Klaus Kirch’s group at ETH to come to CENPA continuing a wonderful ‘pipeline’ of exceptional physicists for us. Our aging buildings decided they weren’t going to take it anymore and showered much of the Gravity Group’s electronics with rainwater on two stormy weekends. This was evidently collateral damage from a seemingly endless University re-roofing project that turned CENPA into a construction zone for a year. A steam line (fortunately low pressure) blew on a Tuesday evening turning much of the Cyclotron wing into a steam bath, soaking carpets, walls, keyboards, and papers with warm sticky water. Old PCB-filled 2400-volt transformers were extracted in a carefully choreographed super-crane operation that required several one- day shutdowns of electrical power. The State of Washington required significant upgrades to our pressurized gas storage for the tandem, to meet new regulations. Our unflappable and capable Senior Engineer Doug Will led us through this gauntlet with his good humor. In September, DOE convened a review panel to come to Seattle and review the CENPA proposal for the next 3-year renewal. The review went well, and the DOE Office of Nuclear Physics renewed our grant DE-FG02-97ER41020 for FY15-17. The continued support of CENPA by the Office of Nuclear Physics is greatly appreciated. In the following we record some of the highlights of our past year in research. • Construction and commissioning continues apace for the KATRIN neutrino mass exper- iment. The detector system, in conjunction with the suite of analysis tools, successfully completed a 135-day spectrometer commissioning campaign. The arrival of the tritium source and beam-line components in 2015 will further exercise the detector system as tritium operations approach in late 2016. • Following the first round of extensive KATRIN Detector-Spectrometer commissioning measurements in 2013, the second round of measurements was carried out from Oc- tober 2014 to March 2015, with improved detector alignment, an upgraded electron gun with pitch-angle control, fully functional Rn-trapping baffles, newly implemented active background removal methods using magnetic pulses and electric-dipole pulses, and a new detector wafer with 100% working pixels. This time the backgrounds from cosmogenic secondaries and stored electrons from Rn decays were fully understood and also confirmed to be under control, which allowed the collaboration to learn of new background source(s). The data are currently being analyzed extensively to plan the next round of measurements.


  • Page 6

    UW CENPA Annual Report 2014-2015 April 2015 iii • The KATRIN data-analysis tools provided by the UW group were used for the second- round measurements as KATRIN’s standard analysis platform. The tools were designed to be universal for offline, real-time and near-time analyses, and owing to that, a num- ber of sophisticated offline analysis programs developed after the first-round measure- ments were used for the second round in the real-time and near-time regimes, providing immediate feedback for measurement planning. The tools were extended for simulation- driven analysis in order to facilitate scenario testing on measured phenomena. • In July the Department of Energy, Office of Nuclear Physics undertook a comprehensive assessment of the KATRIN project’s plans. The five-member external peer panel found the physics to be compelling and construction progress impressive but that the path to tritium will be challenging. • A long paper on the tritium molecular final states by Laura Bodine, Diana Parno, and Hamish Robertson is published in Phys. Rev. C as an Editors’ Suggestion. The paper describing the KATRIN detector system was completed under the leadership of Diana Parno and published in NIM. • Project 8, an experiment that uses cyclotron radiation emitted by mildly relativistic electrons spiraling in a magnetic field to measure their energy, successfully demonstrated the principle of the method in June 2014, and is now moving toward a micro-scale tritium experiment. The paper reporting the success is in press in Phys. Rev. Lett. as an Editors’ Highlight. • The Majorana Demonstrator neutrinoless double-beta decay experiment commissioned and took data with a prototype array of natural HPGe detectors and began commis- sioning the first module of enriched germanium detectors. UW personnel participated in detector construction activities and are leading the data analysis effort. First results are expected in 2016. • The g-2 Storage Ring has been rebuilt in its newly constructed custom building on the Fermilab campus. It is expected to be cooled down and powered on in May, 2015. The field shimming operation commences soon afterwards, with new NMR probes, electronics, and data acquisition provided by the UW team. • The g-2 calorimeter development continues to progress on track. More than 1300 PbF2 crystals and silicon photomultipliers have been ordered. This follows a highly successful SLAC test-beam run, which proved the technical concept of a prototype detector. The results are published in NIM. • The MuSun experiment at PSI has completed 40% of its data taking on muon capture on the deuteron. The new UW-designed TPC and its critical electronics worked flawlessly, permitting the efficient collection of a very high-quality data set in 2014. Two papers


  • Page 7

    iv about muon capture on the proton (MuCap experiment) and another about the MuSun cryogenic preamplifiers developed at UW are published. • A new torsion-balance/rotating-attractor instrument with 20-pole magnet test-bodies has achieved attoNewton-meter sensitivity to the spin-dependent interactions mediated by Goldstone boson exchange with masses up to 0.1 meV. The null result is sensitive to new hidden symmetries broken at energy scales up to 100 TeV, about 5 times higher than the best prior constraint from the ALPS experiment at DESY. • ADMX received the nod from DOE, following a very positive recommendation from the P5 Panel, to go ahead with the ‘Gen-2’ version of the experiment. ADMX will explore higher frequencies and do it much faster thanks to low-noise SQUID electronics cooled with a new dilution refrigerator. • The relativistic heavy-ion group found that the nonjet cylindrical quadrupole compo- nent of 2D angular correlations from high-energy heavy ion collisions, conventionally measured by parameter v2 and interpreted to represent elliptic flow of a quark-gluon plasma or “perfect liquid,” exhibits several characteristics incompatible with a hydrody- namic interpretation. We have extracted a quadrupole-source boost distribution and mt spectrum that reveal the quadrupole source as an expanding thin shell with fixed boost (radial speed) value independent of the collision system and a cold spectrum quite differ- ent from that manifested by most final-state hadrons. We have also measured nonzero quadrupole amplitudes in 200 GeV p-p collisions increasing approximately as the cube of the charge multiplicity (whereas dijet production in p-p collisions increases as the square of the multiplicity). Those and other aspects of nonjet v2 data suggest that the quadrupole component is a novel QCD phenomenon unrelated to hydrodynamic flows in a dense medium. • We have continued to apply a two-component (soft + hard) model (TCM) of hadron production in high-energy nuclear collisions to data from the LHC. Previously we de- scribed hadron yields and mean-transverse momentum trends from several LHC colli- sion systems accurately with a TCM featuring a strong dijet component. We have now accomplished the same for event-wise mean-pt fluctuation data from the same systems. A dominant role for dijet production is clearly established, whereas Monte Carlo mod- els relying on extensive particle rescattering (expected for thermalization and hydro phenomena) are falsified by the LHC data. • Our setup for trapping 6 He and searching for tensor currents has come together and we have gotten our first data set. Over about 1 day of data taking yielded statistics for a determination of “little a” at the 4% level. We are still working on many improvements and expect to triple our trapping efficiency and get a determination to „ 1% before the end of 2015.


  • Page 8

    UW CENPA Annual Report 2014-2015 April 2015 v • The project testing for the possibility of nonlocal signaling using interference-pattern switching with momentum entangled photons, which has been pursued since 2007, has concluded with a publication that identifies the mechanism within quantum mechanics that blocks such signals. As always, we encourage outside applications for the use of our facilities. As a conve- nient reference for potential users, the table on the following page lists the capabilities of our accelerators. For further information, please contact Diana Parno, Associate Director (dparno@uw.edu) or Eric Smith, Research Engineer (esmith66@u.washington.edu), CENPA, Box 354290, University of Washington, Seattle, WA 98195; (206) 543 4080. Further informa- tion 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 nor to be quoted without permission of the authors. In each article the names of the investigators are listed alphabetically, with the primary author underlined in the case of multiple authors, to whom inquiries should be addressed. Hamish Robertson, Director Diana Parno, Associate Director, Editor Gary Holman, Technical Editor Victoria Clarkson, Assistant Editor


  • Page 9

    vi TANDEM VAN DE GRAAFF ACCELERATOR Our tandem accelerator facility is centered around a High Voltage Engineering Corporation Model FN purchased in 1966 with NSF funds, with operation funded primarily by the U.S. Department of Energy. See W. G. Weitkamp and F. H. Schmidt, “The University of Washing- ton Three Stage Van de Graaff Accelerator,” Nucl. Instrum. Methods 122, 65 (1974). The tandem was adapted in 1996 to an (optional) terminal ion source and a non-inclined tube #3, which enables the accelerator to produce high intensity beams of hydrogen and helium 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. Several additional ion species are 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. We recently have been producing the positive ion beams of the noble gases He, Ne, Ar, and Kr at ion source energies from 10 keV to 100 keV for implantation, in particular the rare isotopes 21 Ne and 36 Ar. We have also produced a separated beam of 15-MeV 8 B at 6 particles/second.


  • Page 10

    Contents INTRODUCTION i 1 Neutrino Research 1 KATRIN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1 KATRIN status and plans . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Wafer characterization and noise investigation . . . . . . . . . . . . . . . . . 6 1.3 Progress on veto upgrade . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.4 Analysis tools: status and development plan . . . . . . . . . . . . . . . . . . 9 1.5 Molecular effects and the KATRIN experiment . . . . . . . . . . . . . . . . . 10 1.6 Update on the Tritium Recoil-Ion Mass Spectrometer . . . . . . . . . . . . . 11 1.7 Single-electron detection for KATRIN time-of-flight operation . . . . . . . . 12 Project 8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 1.8 Status of the Project 8 neutrino mass experiment . . . . . . . . . . . . . . . 14 Ho-163 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 1.9 Searching for neutrino mass with 163 Ho: spectrum shape . . . . . . . . . . . 15 SNO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 1.10 Search for hep neutrinos in SNO data . . . . . . . . . . . . . . . . . . . . . . 16 MAJORANA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 1.11 Overview of the MAJORANA DEMONSTRATOR . . . . . . . . . . . . . . . . . 17 1.12 String building for Module 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 1.13 High-voltage cable and feedthrough characterization . . . . . . . . . . . . . 19 1.14 Low-background signal connector production and testing . . . . . . . . . . . 21 1.15 Simulation and analysis activities for the MAJORANA DEMONSTRATOR . . . 23 1.16 Low-noise forward-biased preamplifier tested with a mini-PPC Ge detector . 25 2 Fundamental symmetries and non-accelerator-based weak interactions 27


  • Page 11

    viii Torsion-balance experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.1 Progress on the upgrade of the wedge-pendulum experiment for testing short- range gravity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.2 Progress on a high-precision ground-rotation sensor for Advanced LIGO . . . 28 2.3 Silica fibers for increased torsion-balance sensitivity . . . . . . . . . . . . . . 30 2.4 Parallel-plate inverse-square-law test . . . . . . . . . . . . . . . . . . . . . . . 31 2.5 Improved short range spin-coupled force test . . . . . . . . . . . . . . . . . . 32 Non-accelerator-based weak interactions . . . . . . . . . . . . . . . . . . . . 35 2.6 The 199 Hg electric-dipole-moment experiment . . . . . . . . . . . . . . . . . 35 3 Accelerator-based physics 36 Accelerator-based weak interactions . . . . . . . . . . . . . . . . . . . . . . . 36 3.1 Overview of the 6 He experiments at CENPA . . . . . . . . . . . . . . . . . . 36 3.2 6 He source developments for the β-ν angular correlation experiment . . . . . 37 3.3 Systematics and calibrations of the array geometry and electric field for the 6 Heexperiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 3.4 β-detector calibration for the 6 He experiment . . . . . . . . . . . . . . . . . . 41 3.5 Time-of-flight spectrum and kinematics reconstruction using the β-recoil ion coincidence measurement for 6 He decay . . . . . . . . . . . . . . . . . . . . . 44 3.6 Cyclotron radiation emission spectroscopy: an alternate approach to measur- ing the Fierz interference coefficient . . . . . . . . . . . . . . . . . . . . . . . 46 3.7 Recent upgrades and achievements of the 6 He laser setup . . . . . . . . . . . 48 4 Precision muon physics 52 4.1 Overview of the muon physics program . . . . . . . . . . . . . . . . . . . . . 52 g´2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 4.2 Overview of the g ´ 2 experiment . . . . . . . . . . . . . . . . . . . . . . . . 53 4.3 NMR system for precision magnetic field measurement . . . . . . . . . . . . 56 4.4 PbF2 calorimeter with SiPM readout . . . . . . . . . . . . . . . . . . . . . . 57


  • Page 12

    UW CENPA Annual Report 2014-2015 April 2015 ix 4.5 Inflector beam-monitoring system . . . . . . . . . . . . . . . . . . . . . . . . 59 4.6 Test beam studies of the g ´ 2 PbF2 calorimeter . . . . . . . . . . . . . . . . 62 4.7 SiPM pulse shape studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 4.8 Calorimeter simulation studies . . . . . . . . . . . . . . . . . . . . . . . . . . 67 4.9 Q method for anomalous precession frequency . . . . . . . . . . . . . . . . . . 68 4.10 Collimator optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 4.11 Injection optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 4.12 Measuring the effect of external, ppm-strength, periodic perturbations on a 1.45T magnetic field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 4.13 Production of 400 pulsed proton NMR probes for the g ´ 2 Experiment . . . 72 4.14 Measuring the temperature dependence of the magnetic properties of petroleum jelly for pulsed proton NMR . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 4.15 NMR DAQ for ring-magnet shimming . . . . . . . . . . . . . . . . . . . . . . 76 4.16 Applying time-frequency analysis techniques to nuclear free induction decay (FID) signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 4.17 Quality-control database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 4.18 Quality-control procedures for lead fluoride (PbF2 ) crystals . . . . . . . . . . 80 4.19 Quality control of magnetic components . . . . . . . . . . . . . . . . . . . . . 81 MuSun . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 4.20 Overview of the MuSun experiment . . . . . . . . . . . . . . . . . . . . . . . 83 4.21 2014 production run . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 4.22 Determining the target gas purity . . . . . . . . . . . . . . . . . . . . . . . . . 87 4.23 Muon tracking in the TPC . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 AlCap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 4.24 Nuclear physics input in the search for charged lepton flavor violation, the AlCap experiment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91


  • Page 13

    x 5 Axion searches 94 ADMX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 5.1 Status of ADMX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 6 Relativistic Heavy Ions 98 6.1 UW URHI program overview . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 6.2 Evidence against “elliptic flow” derived from v2 ppt q data . . . . . . . . . . . 99 6.3 No centrality dependence for azimuth-quadrupole source boost . . . . . . . . 100 6.4 p-p event-wise mean-pt fluctuations at LHC energies . . . . . . . . . . . . . . 101 6.5 LHC Pb-Pb event-wise mean-pt fluctuations at 2.76 TeV . . . . . . . . . . . 102 6.6 LHC Pb-Pb pt fluctuations vs Monte Carlo models . . . . . . . . . . . . . . . 103 6.7 Bayesian inference and 200 GeV Au-Au azimuth-correlation models . . . . . 104 6.8 Geometric interpretation of Bayesian inference . . . . . . . . . . . . . . . . . 106 7 Other research 108 7.1 Final report on the nonlocal quantum communication test . . . . . . . . . . 108 8 Education 109 8.1 Use of CENPA facilities in education and course work at UW . . . . . . . . 109 8.2 Student training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 8.3 Accelerator-based lab class in nuclear physics . . . . . . . . . . . . . . . . . . 110 9 Facilities 111 9.1 Facilities overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 9.2 Van de Graaff accelerator and ion-source operations and development . . . . 112 9.3 Laboratory computer systems . . . . . . . . . . . . . . . . . . . . . . . . . . 113 9.4 Electronic equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 9.5 CENPA instrument shops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115


  • Page 14

    UW CENPA Annual Report 2014-2015 April 2015 xi 9.6 Building maintenance and upgrades . . . . . . . . . . . . . . . . . . . . . . . 115 10 CENPA Personnel 117 10.1 Faculty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 10.2 CENPA External Advisory Committee . . . . . . . . . . . . . . . . . . . . . . 117 10.3 Postdoctoral Research Associates . . . . . . . . . . . . . . . . . . . . . . . . 118 10.4 Predoctoral Research Associates . . . . . . . . . . . . . . . . . . . . . . . . . 118 10.5 NSF Research Experience for Undergraduates participants . . . . . . . . . . . 119 10.6 University of Washington graduates taking research credit . . . . . . . . . . . 119 10.7 University of Washington undergraduates taking research credit . . . . . . . . 119 10.8 Visiting students taking research credit . . . . . . . . . . . . . . . . . . . . . . 119 10.9 Professional staff . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 10.10 Technical staff . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 10.11 Administrative staff . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 10.12 Part-time staff and student helpers . . . . . . . . . . . . . . . . . . . . . . . 120 11 Publications 121 11.1 Published papers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 11.2 Invited talks at conferences . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 11.3 Abstracts and contributed talks . . . . . . . . . . . . . . . . . . . . . . . . . . 129 11.4 Papers submitted or to be published . . . . . . . . . . . . . . . . . . . . . . . 130 11.5 Reports and white papers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 11.6 Ph.D. degrees granted . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133


  • Page 15


  • Page 16

    UW CENPA Annual Report 2014-2015 April 2015 1 1 Neutrino Research KATRIN 1.1 KATRIN status and plans J. F. Amsbaugh, J. Barrett∗ , A. Beglarian† , T. Bergmann† , L. I. Bodine, T. H. Burritt, P. J. Doe, S. Enomoto, J. A. Formaggio∗ , F. M. Fränkle† , F. Harms† , A. Kopmann† , E. L. Martin, A. Müller† , N. S. Oblath∗ , D. S. Parno, D. A. Peterson, L. Petzold† , A. W. P. Poon‡ , R. G. H. Robertson, M. Steidl† , J. Schwarz† , D. Tcherniakhovski† , T. D. Van Wechel, K. J. Wierman§ , J. F. Wilkerson§ , and S. Wüstling† The layout of the KATRIN hardware is given in Fig. 1.1-1. Following the formation of the KATRIN collaboration in 2001, the first hardware for the experiment, the prespectrometer, appeared in 2005. This proved to be a workhorse in understanding and refining the design features of the main spectrometer, which was delivered in 2006. In 2011 the US institutes supplied the Focal Plane Detector (FPD) system and the data acquisition software. These, in conjunction with the suite of near-time/real-time analysis tools, have proved vital to the commissioning of the main spectrometer and understanding of the background sources. SOURCE   TRANSPORT   SPECTROMETERS   DETECTOR   νe$ 3H$ 3He$ e%$ Differen=al   Rear   Pre-­‐   Main   pumping   Detector   system   spectrometer   spectrometer   system   Windowless   Cryogenic   Field   gaseous   pumping   compensa=on   tri=um   system   coils   source   70  meters   Figure 1.1-1. The layout of the main components of the KATRIN experiment. The next two years will be exciting, but very challenging, times for KATRIN. The final components of the experiment, the Windowless Gaseous Tritium Source (WGTS), the Dif- ferential Pumping System (DPS) and the Cryogenic Pumping System (CPS), will all appear at KIT in 2015. These components must be installed, commissioned and integrated into the ∗ Massachusetts Institute of Technology, Cambridge, MA. † Karlsruhe Institute of Technology, Karlsruhe, Germany. ‡ Lawrence Berkeley National Laboratory, Berkeley, CA. § University of North Carolina, Chapel Hill, NC.


  • Page 17

    2 Tritium Laboratory Karlsruhe (TLK) before tritium operation can begin as scheduled in late 2016. Commissioning of this hardware will begin in early 2016 and will put great demands on the FPD system. A goal of the US institutes will be to ensure that the FPD system and the analysis tools are in optimal shape to meet the commissioning demands. The status of this hardware and the plans for the FPD are outlined below, starting at the source and continuing down the beam line to the detector. Two additional UW-based activities are also overviewed, TRIMS and Time of Flight, both of which may impact future KATRIN activities. • Hardware and software status: – The tritium source: At 2 ˆ 1011 Bq, the WGTS is a sufficiently bright source to satisfy KATRIN’s sta- tistical requirements. In order to meet the systematic requirements, the tritium gas must be maintained at a well-defined and stable density. This is achieved by maintaining the source at 30 ˘ 0.01 K. These demanding conditions were success- fully met in the WGTS demonstrator, a 10-meter-long by 90-mm-diameter tube whose temperature is maintained by boiling neon. Research Instruments Inc. is installing the demonstrator into the train of seven superconducting magnets nec- essary to constrain the electrons resulting from the tritium decay. The work is proceeding well and delivery to KIT is expected in August 2015. Preparation to receive this hardware is one of the major logistical challenges fac- ing the KATRIN collaboration. The auxiliary systems include cryogenics sup- ply/recovery lines, many hundreds of channels of slow controls instrumentation, and magnet safety systems. The installation is expected to be complete and com- missioning to begin in late December 2015. The WGTS is expected to be opera- tionally ready in May 2016. Monitoring the physical parameters of the WGTS is critical to the control and understanding of sources of systematic error. This is achieved in part by the Rear System that is attached to the upstream (rear) end of the WGTS and includes a steerable electron gun used to probe the gas column. This gun and associated vacuum assembly was delivered to KIT by UCSB in February 2015. The super- conducting magnet that will link the Rear System to the WGTS passed its site acceptance tests in March 2015. The entire rear system is expected to achieve operational readiness in April 2016, in time for completing the commissioning of the WGTS. – The transport system: Once the electrons have been produced in the WGTS it is necessary to adiabati- cally transport them to the main spectrometer while at the same time excluding the tritium molecules from entering and contaminating the main spectrometer. This is a technical challenge, since the source is at a pressure of 10´3 mbar and


  • Page 18

    UW CENPA Annual Report 2014-2015 April 2015 3 the spectrometer is at a pressure of 10´11 mbar, and the two regions are connected by an open-ended beam pipe. The Differential Pumping System (DPS) and the Cryogenic Pumping System (CPS) provide the solution, together reducing the tritium level by a factor of 1014 while allowing the free passage of electrons. The DPS consists of a chicane of five superconducting magnets separated by turbo- molecular pumps. The chicane prevents line-of-sight passage of the neutral tritium molecules while the magnets guide the electrons through the chicane. The turbo- molecular pumps return the tritium gas back to the source, resulting in a 107 tri- tium reduction factor at the output of the DPS. The original DPS was abandoned soon after delivery to KIT due to unresolvable problems with the magnet pro- tection diodes. The system was subsequently redesigned using five, free-standing magnets. These magnets are installed on the support structure and have passed their site acceptance tests in March 2015. The chicane beam pipe and tritium secondary enclosure is nearing completion. Manufacture of the turbomolecular pump ports is complete and site acceptance testing of the beam tube assembly will be complete in July 2015. The CPS traps the tritium molecules that escape the DPS. This system again consists of a chicane beam pipe, but instead of turbomolecular pumps, the tritium is captured on argon frost that lines the walls of the chicane, resulting in a further factor-of-107 reduction in the tritium entering the spectrometers. Between data- taking cycles the argon is sublimated and any captured tritium is removed. The CPS is being built by ASG Superconductors in Genoa, Italy. Assembly is almost complete and delivery to KIT is expected in June 2015. After installation in the summer and commissioning with argon in the fall, the CPS will be ready for tritium in June 2016. – The spectrometers: The KATRIN experiment includes three spectrometers: the pre-spectrometer, used to reduce the flux of electrons entering the main spectrometer; the main spectrometer, where the electron energy is analyzed with a resolution of 0.94 eV; and the monitor spectrometer, used as an independent monitor of the spectrometer voltage stability. The prespectrometer was used to guide the design of the internal electrodes of the main spectrometer and to understand sources of background. In mid-2015 it will be located in its final position at the entrance of the main spectrometer where it will reduce the flux of electrons by a factor of „ 107 , allowing only high-energy electrons to enter the main spectrometer. The main spectrometer was delivered to KIT in 2006, and installation of the field- shaping and background-reducing inner wire electrodes was completed in 2013. The spectrometer has since undergone two „ 4-month commissioning periods, which have demonstrated that the spectrometer meets the design energy resolu- tion. Unexpected sources of background have also been identified. 210 Pb from radon in the air is plated onto the wall of the spectrometer. This background may be reduced by use of the wire planes. An additional source of Rn is the 3 km of


  • Page 19

    4 NEG strips used to pump hydrogen emanating from the steel spectrometer walls. It has been demonstrated that the liquid-nitrogen-cooled baffles capture the Rn, effectively reducing the Rn entering the spectrometer to an insignificant amount. The background associated with the spectrometer is currently „ 400 mHz, com- pared to the goal of 10 mHz. The dominant background is uniformly distributed throughout the volume of the spectrometer. It was measured during a commis- sioning run with a high residual gas pressure, since the spectrometer was not fully baked out beforehand. A brief data-taking run is scheduled for June 2015 to measure the background with a properly baked system. The monitor spectrometer, consisting of the refurbished Mainz spectrometer, shares the same high-voltage retarding potential as the main spectrometer and, by moni- toring the lines from a 83 Kr source, monitors the stability of the retarding potential of the main spectrometer. The monitor spectrometer is fully operational and ready for KATRIN tritium operation. – The detector system: The detector system passed its operational readiness review in March 2013 and has been used extensively for commissioning the spectrometer. Detector operation has revealed areas in which improvements can be made. An example is the replacement of the detector cosmic ray veto system that is currently underway (Sec. 1.3). The new veto is expected to provide excellent signal-to-noise and will be simple and robust, requiring little attention from the user. Installation and commissioning of the new veto is expected to be complete in September 2015. Another example is the development of wafer characterization methods for the focal-plane detector itself (Sec. 1.2). In 2013 the pinch magnet began to fail, becoming unable to reach the nominal 6T field without quenching. The DOE and KIT rapidly responded by making funding available to secure a replacement pinch magnet. The new magnet was delivered to KIT in March 2015 and is currently undergoing site acceptance testing. The new pinch magnet incorporates many of the design features and lessons learned in the production of the DPS magnets and is expected to be robust and reliable. The detector system is expected to be fully operational by the time of the June background studies, and the final upgrades to the veto and electronics should be completed in time for the third round of commissioning in mid-2016. – Data acquisition and analysis tools: Data acquisition is controlled by the Object-oriented Real-time Acquisition and Control system (ORCA). This system is developed and maintained by UNC and is also used for the high-speed DAQ requirements of the monitor spectrometer and rear system. ORCA supports automation of data-taking sequences, enabling remote operation of routine activities such as calibration and repetitive data ac- quisition cycles. In 2010 development began of a suite of real-time/near-time analysis tools1 . Pri- 1 CENPA Annual Report, University of Washington (2014) p. 5.


  • Page 20

    UW CENPA Annual Report 2014-2015 April 2015 5 marily for use in commissioning the detector system, it includes tools to simulate and compare the detector performance. This has now evolved into the de facto analysis system for KATRIN (Sec. 1.4). Analysis routines are simply constructed by drag-and-drop from an extensive library of analysis tools, allowing the physicist to focus in the physics rather than the coding. Analysis routines can also be auto- mated allowing a real-time check on the quality of the data. The system provides a robust foundation that may be built upon to accommodate the evolving needs of the KATRIN experiment. – Other activities: In probing the neutrino mass, KATRIN examines the last 20 eV of the spectrum in order to avoid the electronic excited states of the T2 decay. Within the 20 eV window there exist vibrational and rotational excited states which are thought to be calculable to the necessary precision (Sec. 1.5). However, there is significant disagreement between theory and experiment. At the UW, the Tritium Recoil Ion Measurement Spectrometer (TRIMS) will provide data that makes fundamental tests of the theoretical model, either supporting or refuting its use by KATRIN (Sec. 1.6). The spectrometer operates in the retarding-potential mode, integrating all events above the retarding potential of the spectrometer. The spectral shape is deter- mined by operating the spectrometer at a number of different retarding potentials. It is also possible to operate the spectrometer in the Time-of-Flight (ToF) mode, in which the energies of all electrons above the retarding potential can be measured with a comparable resolution. Operating in the ToF mode will offer additional ways of suppressing backgrounds while drastically reducing the time to maximum neutrino mass sensitivity (Sec. 1.7). • Future plans: Exercising the detector system has provided insights into desirable improvements. As a result, upgrades are being implemented to the veto (Sec. 1.3) and electronics systems. The study of spectrometer backgrounds will be briefly revisited in June 2015 but the bulk of the commissioning activity will begin in mid-2016 when the source and transport systems are commissioned and brought online. By mid-2016, the detector system and its support team must be functioning like a well-oiled machine in order to meet the demands of commissioning and data-taking that will extended into 2021. A new Memorandum of Understanding has been drawn up and approved by the DOE Office of Nuclear Science. The MoU runs until 2019, well into the tritium operations phase. The primary US responsibility remains the maintenance and operations over- sight of the detector system. The detector system is complex and experience has taught us that a permanent on-site presence is required to maintain the knowledge base, ensure regular maintenance and respond to the unexpected. To this end we have identified a senior KIT scientist to fill the role. Regular tritium operations will likely consist of three, three-month run periods, in- terspaced by one-month maintenance periods, per year. Since the experiment can be


  • Page 21

    6 monitored and controlled remotely, the US institutes are positioned to remotely monitor data taking operations outside of the normal European working day. 24/7 operation of the tritium source requires a constant presence at the TLK. Support for the necessary staff is being sought by the KIT management. The US institutes will continue to support the activities for which they are primarily responsible, namely the maintenance and development of the DAQ (UNC) and sim- ulations software (MIT, UNC). The UW will continue to play a leading role in the maintenance, development and integration of the real-time analysis tools into the over- all analysis framework. A result from the TRIMS experiment is expected within a year and will be of interest outside the neutrino-mass community as well. Should the principle be proven for single- electron detection for ToF (Sec. 1.7), we will actively seek ways of incorporating it into the KATRIN experimental program to speed progress to the 200-meV neutrino mass sensitivity. 1.2 Wafer characterization and noise investigation S. Enomoto, E. L. Martin, D. A. Peterson, R. G. H. Robertson, and T. D. Van Wechel When a wafer from the 2014 batch of KATRIN focal plane detector wafers was installed as the KATRIN detector, the detector resolution went from 1.5 keV FWHM to 2.0 keV FWHM. Another wafer from the same batch was tested in an attempt to determine what might be the cause of the decreased performance. Three wafer properties were measured: inter-pixel resistance, sheet resistance, and bulk resistivity. Each wafer is a PIN diode detector with a heavily doped entrance window to apply bias to each pixel, and a downstream readout side segmented into 148 pixels surrounded by a guard ring surrounded by a bias ring, each coated in titanium nitride. The bias-ring coating wraps around the edges of the wafer, connecting the upstream side of the wafer to bias. Wafer inter-pixel resistance was measured by applying a fixed bias to a single pixel, and adjusting the bias on all surrounding pixels. The current on the single-pixel connection was measured. The measured current vs the inter-pixel voltage difference was fit to a model of two back-to-back diodes in parallel with an ohmic resistance and a fixed leakage current. The measured resistance was over 100 GΩ, too high to explain the noise. The measured leakage current was thought to be dominated by light leaks, yet still measured only 3.7 pA with a 100 V bias at -11 ˝ C, too low to explain the noise. Sheet resistance was determined by applying a current on opposite pins of the guard ring or bias ring, and measuring the voltage on the other pins. The measurements were compared to a numerical model to determine the sheet resistance. For the bias ring the relative sheet resistance of the TiN-coated region and uncoated region was determined from the relative voltage distribution around the bias pins, and then the absolute sheet resistance was determined. The sheet resistance of the pixels was assumed to be the same as that of the guard ring.


  • Page 22

    UW CENPA Annual Report 2014-2015 April 2015 7 The guard-ring sheet resistance for the new wafer measured 0.79 Ω/sq, compared to 48.7 Ω/sq for a wafer from the previous batch. The bias sheet resistance measured 4.0 Ω/sq for the TiN-coated region and 28.2 Ω/sq for the uncoated region, compared to 1580 Ω/sq and 35 Ω/sq for the wafer from the previous batch. The lower sheet resistance should result in less noise for the new wafer. Bulk resistivity was measured by applying a forward bias to a pixel, and measuring the current. This was fit to a model of a diode in series with an ohmic resistance and the contribution from sheet resistance and pin contact resistance was accounted for. Bulk resistivity came out to 93 Ω¨cm for the new wafer and 41 Ω¨cm for the wafer from the previous batch. The higher bulk resistivity for the new wafer indicates lower impurity concentration, which would be expected to result in improved resolution. As of yet no explanation for the reduction in detector resolution has been identified. All properties measured so far indicate the new wafer should actually be superior to the wafer from a previous batch. Further investigation is planned. 1.3 Progress on veto upgrade T. H. Burritt, P. J. Doe, S. Enomoto, D. A. Peterson, and T. D. Van Wechel The cosmic-muon veto for the detector wafer is being upgraded to take advantages of new silicon photomultiplier (SiPM) models and recent technical advancements developed by other experiments. The current KATRIN veto yields two to three detected photons per cosmic muon per SiPM, while other experiments typically report much higher numbers such as 20 detected photons per muon per SiPM, even with similar detector construction. Due to the high dark rate of SiPM’s at „100 kHz, 99% of our recorded data is currently dominated by accidental coincidences of the dark noise, even with precise control of SiPM temperature with Peltier coolers in a box with carefully regulated dry nitrogen flow. An upgrade to a high light- yield system will not only reduce the data size by 99% but also will improve the robustness and stability of the system, ease operation, and reduce the systematic effects associated with the instability. Cosmic-muon detectors of this type typically consist of plastic scintillator panels, wave- length-shifting (WLS) fibers inserted into the panels to collect photons and to bring them out of the panels, and SiPM photon detectors attached to the WLS fiber in some way. The scintillator panels are often wrapped with reflective material to improve photon collection efficiency. Sometimes SiPMs are attached to both ends of a WLS fiber, sometimes one end is treated to be reflective without a SiPM, and sometimes one end is just left unconnected and untreated, presumably due to the limited benefits of using both ends because of the short light attenuation length of the WLS fiber which is typically only a few meters. After some literature review, the system developed by the T2K experiment seemed to have the best overall performance1 . With considerable assistance from the T2K group, we utilized 1 P.-A. Amaudruz et al., The T2K Fine-Grained Detectors, Nucl. Inst. Meth. A 696, 1 (2012).


  • Page 23

    8 the T2K SiPM holders to directly attach the SiPMs to the WLS fibers, which eliminate our lossy clear-fiber connections between the WLS fibers and SiPMs. The new Hamamatsu SiPMs (product name MPPC) with 1.3 mm ˆ 1.3 mm sensitive area are used instead of our old 1.0 mm ˆ 1.0 mm MPPCs. Also copying the T2K design, the WLS fibers are replaced with 1-mm-diameter Kuraray Y11 from Bicron BCF-91A. The scintillator panel thickness is doubled to 20 mm to further increase the photon yield. With these improvements, the photon yield was expected to be large enough to allow the use of part of the “photon budget” to improve robustness, as done by T2K. The WLS fibers are replaceable if damaged (damage has already occurred to our current non-replaceable system) by making the fiber grooves sufficiently loose, the preamplifiers are placed away from the SiPMs using coaxial cables to an accessible location, and the SiPM coolers and associated dry nitrogen flow are all eliminated. The readout electronics were redesigned and a prototype board was produced by CENPA. All the potentiometers are replaced with digital-to-analog converters (DAC) for full auto- mated control. The bias supply circuitry was modified to reduce power consumption which is necessary for use without cooling fans under strong magnetic fields. A control board for DAC programming with temperature readout, based on a field-programmable gate array (FPGA), was designed and produced by CENPA. A set of software tools to operate the electronics was also developed. SiPM Ch 0 (Coincidence with Ch 1) counts/bin 2 pe gaus fitting (muons) chi2/ndf: 786.32 / 666 5 3 pe Constant: 714.5 ± 1.6 10 Mean: 2389.1 ± 0.6 4 pe Sigma: 232.55 ± 0.74 5 pe Dark, Crosstalk and 104 Low Energy Events Cosmic Muons 103 Electronics Saturation 102 101 100 0 500 1000 1500 2000 2500 3000 3500 4000 ADC 20 Apr 2015 23:52, results/NewVeto-00174.root/Adc0 Figure 1.3-1. Performance of prototype veto system. The ADC scale is calibrated using the separated pe peaks at ADC „ 100, resulting in 32 ADC/photo-electron. The photon yield for cosmic muons is then calculated to be 75 photons per muon per SiPM. The electronics are saturated at the higher end of the muon peak (ADC ą„ 2200) due to unexpectedly high light yield. The figure was created with the Koffein software tool (Sec. 1.4). The performance of the new design was evaluated with a small test scintillator panel of 275 mm x 210 mm dimension. Fig. 1.3-1 shows measured photon yield with a test DAQ system using NIM trigger logic and VME readout, controlled with KiNOKO software. The acquired data is automatically processed and managed by KAFFEE (Sec. 1.4). Although the short WLS fiber length makes the photon yield higher than in the final design, the muon


  • Page 24

    UW CENPA Annual Report 2014-2015 April 2015 9 signals are far above the SiPM dark rate and crosstalk. The peak position is estimated to be „75 photons per muon per SiPM, which is to be compared with the current veto system light yield of 2 to 3 photons per muon per SiPM. Based on this prototype test, two scintillator panels were produced and are currently being tested. The new veto will be installed on the detector system at KIT in September 2015. 1.4 Analysis tools: status and development plan S. Enomoto For the first round of commissioning measurements of the KATRIN Spectrometer - Detec- tor Section (SDS) carried out in 2013, UW provided a suite of analysis tools which is based on the toolkit originally developed for the detector commissioning analysis in 2011-2012. The analysis tools were also used for the second round of SDS commissioning measurements (SDS-II) in 2014 with extensive improvements as described below: The tools were originally developed for real-time and near-time applications, and were extended for used for offline analysis in the first round of the SDS commissioning measure- ments (SDS-I). Owing to the toolkit’s feature of online and offline unification, the set of sophisticated offline analysis programs developed for SDS-I data after SDS-I concluded, were used for SDS-II real-time and near-time analyses without modifications, providing immediate final results during and/or just after every run. For offline analysis use, the core tool for analysis logic (BEANS) was integrated with the KATRIN simulation tool, Kassiopeia. BEANS had been designed to be flexible for various analysis needs without having any pre-defined event data structure, and its dynamically- developing data model was found to be a good fit to the simulation’s trait of attaching various pieces of information arbitrarily. The analysis automation and run catalog tool (KAFFEE) was also extended for offline analysis and simulation. With a new MongoDB backend, KAFFEE maintains every analysis details as well as run conditions and simulation configurations, and enables to searches for specific runs and analysis results. In order to promote analysis sharing over remote locations, KAFFEE equips a convenient file uploader, Dripbox, allowing any files to be managed by KAFFEE. To facilitate interactive quick analysis on the data catalog, KAFFEE now contains a new web-based tool, Koffein, which can interactively draw histograms and graphs, read coordinates, count entries, compare histograms, fit any functions to histograms, and produce presentation-grade figures (see example in Fig. 1.3-1). The UW-provided tools are now used as the defacto standard analysis tools for KATRIN SDS. It is expected that the tools will be further extended for use in the KATRIN Source and Transport Section (STS), where the data basically consists of time series of slow-control readings, very different from the event-based data from the detector system. Some extensive modifications to the tools are being investigated towards this end.


  • Page 25

    10 1.5 Molecular effects and the KATRIN experiment L. I. Bodine, D. S. Parno, and R. G. H. Robertson The ongoing construction of the Karlsruhe Tritium Neutrino experiment (KATRIN) has renewed interest in the molecular final-state distribution (FSD) populated by T2 beta decay1 . An understanding of the molecular effects at the percent level is necessary for KATRIN to reach the ultimate sensitivity of 0.2 eV. While modern calculations2,3 claim accuracy on that level the interplay between experimental uncertainty and uncertainty on the theoretical FSD for KATRIN has largely been ignored. We re-examined the impact of molecular effects on KATRIN, quantifying the impact of molecular effects on KATRIN4 , as summarized below. We showed that the general features of the FSD can be reproduced quantitatively from considerations of kinematics and zero-point motion, developing a semiclassical model of the molecule as a simple harmonic oscillator. The variance of the FSD arising from zero-point motion agrees well with the variances from the full quantum mechanical calculations. From this expression for the variances we were able to derive the impact on the systematic errors on the neutrino mass-squared. Table 1.5-1 shows the resulting systematic uncertainty on the neutrino mass-squared owing to the identified molecular effects. Source of systematic shift Target accuracy σsyst pm2ν qr10´3 eV2 s FSD theoretical calculations |∆σFSD {σFSD | ď1% 6 temperature calibration |∆T {T | ď 0.005 - translational 0.05 - FSD 0.06 temperature fluctuations |∆T {T | ď 0.001 - translational 0.009 - FSD 0.01 ortho-para ratio |∆λ{λ| ď 0.03 0.44 isotopic impurities - tritium purity |∆ǫT {ǫT | ď 0.03 2.9 - ratio of HT to DT |∆κ{κ| ď 0.1 0.03 higher rotational states |∆T {T |rotational ď 0.1 1 Table 1.5-1. Summary of molecular-related sources of systematic shift in extracted neutrino mass-squared, the projected accuracy on the experimental parameters and the individual effect on m2ν for the nominal KATRIN parameters. In addition to the impact on KATRIN we found that when the LANL and LLNL neutrino- mass experiments performed in the 1980s with gaseous tritium are re-evaluated using the modern FSD calculations2 , the extracted neutrino mass-squared values are consistent with 1 E. W. Otten and C. Weinheimer, Rep. Prog. Phys. 71, 086201 (2008). 2 A. Saenz, S. Jonsell and P. Froelich, Phys. Rev. Lett. 84, 242 (2000). 3 N. Doss, J. Tennyson, A. Saenz and S. Jonsell, Phys. Rev. C 73, 025502 (2006). 4 L. I. Bodine, D. S Parno and R. G. H. Robertson, Phys. Rev. C 91 035505 (2015).


  • Page 26

    UW CENPA Annual Report 2014-2015 April 2015 11 zero instead of being significantly negative. This result highlights the importance of using the correct FSD to analyze molecular-tritium beta-decay neutrino-mass experiments. Fig. 1.5-1 shows the FSD calculations of Saenz1 compared to those of Fackler2 , which were available at the time of the original LANL and LLNL analyses. 100 Relative probability (%/eV) 10 1 0.1 0.01 0.001 -250 -200 -150 -100 -50 0 Binding energy (eV) Figure 1.5-1. Molecular spectrum excited in the beta decay of T2 (J “ 0) as calculated by Saenz (solid red curve) and by Fackler et al.2 (dotted blue curve). For the purposes of display and comparison, discrete states in the latter spectrum have been given a Gaussian profile with a standard deviation of 3 eV. 1.6 Update on the Tritium Recoil-Ion Mass Spectrometer L. I. Bodine, T. Lin, D. S. Parno, and R. G. H. Robertson The Tritium Recoil-Ion Mass Spectrometer, TRIMS3 , is an experiment designed to mea- sure the probability that beta decay within a T2 molecule will result in a bound 3 HeT` molecule, testing an observable related to the molecular final-state distribution relevant to KATRIN (Sec. 1.5). The beta electron will be detected in coincidence with a 3 He` , T` , or 3 HeT` ion, using a pair of silicon detectors at either end of a glass decay chamber. The charged particles will be accelerated toward the detectors using a 60-kV potential difference, and will be guided by a 2-kG magnetic field. The experiment is presently under construction with the goal of commissioning in late spring and summer of 2015. 1 A. Saenz, S. Jonsell and P. Froelich, Phys. Rev. Lett. 84, 242 (2000). 2 O. Fackler, B. Jeziorski, W. Kolos, H. J. Monkhorst and K. Szalewicz, Phys. Rev. Lett. 55, 1388 (1985). 3 CENPA Annual Report, University of Washington (2013) p. 15.


  • Page 27

    12 The assembly of the VCR and 2.75-inch CF portions of the ultra-high vacuum system, comprising the gas-handling, pressure-measurement, and pumping sections, is essentially complete. A partial redesign has been implemented to ensure electrical isolation between the front-end electronics and the vacuum gauges and pumps. Upgrades to the support struc- ture, which will simplify the final installation of the decay chamber and the final positioning of the vacuum system relative to the magnets, are underway. The beta detector will be held at 60 kV relative to the ion detector. In a test setup, the decay chamber, including detector blanks and the internal metal plates that will define the potential near the detectors, has been placed in the 2-kG magnetic field and successfully maintained at voltages up to 70 kV. The test has been performed both under a vacuum at 3 ˆ 10´8 Torr and with injected H2 gas at 1 ˆ 10´7 Torr, which matches the anticipated T2 pressure during eventual data-taking. Data will be acquired with a 250-MS/s digitizer from CAEN, which has now been inte- grated with ORCA by Mark Howe at the University of North Carolina, Chapel Hill. We have written OrcaRoot code, including an offline trapezoidal filter, to translate the raw data into an analysis-ready format. Initial tests of the detectors, digitizer, and front-end electronics, using 241 Am calibration data, give a preliminary energy resolution of „ 1.1 keV (σ). Once these tests are complete, the detectors and decay chamber will be installed on the main vacuum system. Pumps, pressure gauges, and the high-voltage supply will be continuously monitored via a LabJack ADC read out by ORCA. We have tested this functionality and are working to integrate the readout with a PostgreSQL monitoring database. A gaseous 83m Kr source1 has been produced at the CENPA tandem van de Graaff accelera- tor. The TRIMS apparatus will be commissioned with a measurement of the charge spectrum of the krypton ion following conversion. The resulting data may suggest modifications or up- grades before tritium data-taking commences. In preparation for the commissioning period, an upgraded simulation using Geant4 is in development. 1.7 Single-electron detection for KATRIN time-of-flight operation E. L. Martin and R. G. H. Robertson Implementing a time-of-flight capability in the KATRIN experiment could allow measurement of the energy spectrum above each retarding potential, instead of simply the rate of electrons above each retarding potential. This could significantly reduce the data collection time required to attain the same statistical uncertainty2 . In time-of-flight mode the energy of each electron is determined from the time it takes to pass through the main spectrometer. As flight time is mostly determined by the slow 1 CENPA Annual Report, University of Washington (2014) p. 13. 2 N. Steinbrink, V.‘Hannen, E L. Martin, R. G. H. Robertson, M. Zacher and C. Weinheimer, New J. Phys. 15, 113020 (2013).


  • Page 28

    UW CENPA Annual Report 2014-2015 April 2015 13 movement through the analyzing plane, where the remaining transverse momentum on the electron results in reduced energy resolution similar to high pass filter mode, the energy resolution in time-of-flight mode can be nearly as good as high-pass-filter mode. Energy resolution will be further reduced by timing resolution. The stop signal for time of flight is already available from the focal plane detector with around 100 ns timing resolution, but a start signal is still required. The best location for generating the start signal is between the pre-spectrometer and main spectrometer. Electrons trapped between the main spectrometer and pre-spectrometer would cause a background in the tens of MHz. A removal method using a small rod inserted in the beam line has been proposed. Most electrons that pass the pre-spectrometer will be reflected by the main spectrometer and pass the tagger twice. This background can be reduced by using a higher pre-spectrometer retarding potential, but will still be in the range of tens of kHz. Time of flight is tens of µs, so random triggers are expected to be nearly as frequent as signal events. Even so, reduced measurement time is expected. An electron tagger test setup using a resonant cavity is under construction. It uses a cylindrical cavity with a small hole through the center for electrons to pass through. A cylindrical cavity designed for the width of the KATRIN beam line would be too large to place between the pre-spectrometer and main spectrometer, so a different cavity with a helical design is also under development. The cavity is excited by a loop inserted through the cavity wall, and excitation is measured from another loop. As an electron passes through it will exchange a small amount of energy with the cavity. How much depends on one cavity excitation amplitude and phase. A change in cavity excitation would indicate the passing of an electron, and how much it changed measures the change in electron energy. The energy change of the electron is proportional to the cavity excitation voltage, but the energy stored in the cavity is proportional to the square of the excitation voltage. Excitation needs to be sufficient to overcome thermal noise, but not so large that the fractional change from a passing electron is too small. Based on component specifications the signal power is expected to be around 28 times the noise power for optimal excitation.


  • Page 29

    14 Project 8 1.8 Status of the Project 8 neutrino mass experiment P. J. Doe, J. A. Formaggio∗ , M. Fertl, D. Furse∗ , M. A. Jones† , J. Kofron, B. H. LaRoque‡ , E. L. McBride, B. Monreal‡ , N. S. Oblath∗ , R. Patterson§ , R. G. H. Robertson, L. J Rosenberg, G. Rybka, M. G. Sternberg, J. Tedeschi† , and B. A. VanDevender† Project 8 marked completion of phase 1 of its experimental effort with the publication of a highly resolved 83m Kr conversion electron spectrum, shown below in Fig. 1.8-1, which includes conversion from the K, L, M , and N shells of krypton. This is the first demonstration of the technique, now called Cyclotron Radiation Emission Spectroscopy (CRES). CRES combines the rich history of single-electron trapping with high resolution-analysis techniques to provide a unique avenue to electron energy measurement and spectroscopy. Figure 1.8-1. The conversion electron spectrum of 83m Kr as measured by Project 8. The inset spectrum shows high resolution data from conversion of L shell electrons, with a characteristic FWHM of 15 eV. With phase 1 completed, development efforts are ongoing for phase 2 of the prototype. In expectation of future challenges, initial designs for a version of the prototype experiment which contains tritium have been drawn. The fundamental difference in the design is an ex- tended magnetic bottle formed from a pair of solenoidal coils. Small magnetic inhomogeneities can easily spoil such a confining potential, and therefore extensive magnetic surveying of the components of the prototype has been performed. Data taking which is scheduled for later this year should probe both the design of the extended magnetic bottle as well as its performance, and lay the groundwork for a CRES measurement of the full β-decay spectrum of tritium. ∗ Massachusetts Institute of Technology, Cambridge, MA. † Pacific Northwest National Laboratory, Richland, WA. ‡ University of California Santa Barbara, Santa Barbara, CA. § California Institute of Technology, Pasadena, CA.


  • Page 30

    UW CENPA Annual Report 2014-2015 April 2015 15 Ho-163 163 1.9 Searching for neutrino mass with Ho: spectrum shape R. G. H. Robertson In our search for new and more sensitive methods to measure the mass of the neutrino, we have given some consideration1 to an idea that has its origins more than 30 years ago. It was noted that electron capture decay could be used to measure the mass because, although capture leaves a vacancy in an atomic orbital and ejects a nominally monoenergetic neutrino, in fact atomic vacancies have short lifetimes and therefore non-negligible widths. The decay process is then formally the same as a radiative decay, A Z Ñ A pZ ´ 1q ` νe ` γi ` Qi (1) with a 3-body phase space. The tails of the lines extend to the energy limit imposed by the ground-state Q-value, and at that limit are sensitive to the modification of phase space caused by neutrino mass. This method looks particularly attractive in the case of 163 Ho, which has a low Q-value in the vicinity of 2.5 keV. A number of experimental groups (ECHo, HoLMES, NuMECS) are developing high-resolution calorimeters to record the ‘visible’ energy, i.e. that not carried away by the neutrino. With such a detector, one observes sharp lines from the 7 occupied orbitals from which capture can occur: (3s), (3p1/2), (4s), (4p1/2), (5s), (5p1/2), and (6s). The lines have Breit-Wigner shapes extending to the energy cutoff at the Q-value, where neutrino mass information would be found. The determination of the mass requires knowledge of the spectral shape in the absence of neutrino mass, which can only be obtained from theory. Unfortunately the simplified theory just described neglects an important complication, the population of multivacancy states in the daughter 163 Dy atom. Fig. 1.9-1 shows the complex spectrum of satellite peaks that results. Their lineshapes are not Breit-Wigner and cannot easily be calculated, which makes the theoretical spectrum needed for extraction of neutrino mass problematic. 4 10 3 10 2 Relative Amplitude 10 1 10 0 10 -1 10 -2 10 -3 10 -4 10 0 500 1000 1500 2000 2500 3000 Visible energy (eV) Figure 1.9-1. The visible energy in a calorimeter following electron capture in 163 Ho. The simpler spectrum (blue) is calculated in the customary single-vacancy approximation. The more complex spectrum (red) includes configurations with 2 vacancies. 1 R. G. H. Robertson, Phys. Rev. C 91, 035504 (2015).


  • Page 31

    16 SNO 1.10 Search for hep neutrinos in SNO data T. J. Major and N. R. Tolich The SNO experiment collected six years’ worth of solar neutrino data. Most of the events were interactions of neutrinos originating from the decay of 8 B in the Sun, but a small fraction of the events correspond to events from the hep reaction, 3 He + p Ñ 4 He + ν + e` . Hep neutrinos are interesting because the hep reaction is less dependent on solar composition than are other fusion processes, and because calculation of the nuclear matrix element is prohibitively difficult. Determination of the rate of the hep reaction in the Sun would serve as a test of the Standard Solar Model in a new way. Hep neutrinos are distinguishable in principle from 8 B neutrinos because they have a different energy spectrum that extends to a higher energy regime than 8 B neutrinos. To improve our knowledge of the rate of the hep reaction, we have developed a new event fitter that uses the spatial distribution of detected photons in addition to the time distribution. This fitter has improved the spatial resolution of event locations by around 10%. This fitter is also more reliable than previous methods at identifying a class of instrumental backgrounds that previously required a considerable fiducial volume cut. Previous SNO analyses have used a rather conservative data set, but many runs that were excluded from this set may contain useful data. Some calibration runs that used a low-energy source may also contain useful higher-energy 8 B and hep neutrino events. We have developed a technique to test whether these runs contain data that are sufficiently similar to the conservative data set to be included in our analysis. Our technique uses a multi-dimensional modification of a Wald-Wolfowitz test, as discussed in Friedman and Rafsky1 . We begin by treating each event as a vector in a phase space composed of energy and spatial distribution information. We then combine the test data set and the accepted data set, and calculate the Mahalanobis distance between every pair of events. From these distances, we construct a graph composed of the union of an appropriate number of minimum spanning trees. An analysis of the interconnectivity of this graph produces a p-value that can be interpreted as a measure of similarity between the data sets. In this way, we can systematically (and blindly) assess individual events, runs, and sets of runs to decide which data to add to our analysis. 1 J. H. Friedman and L. C. Rafsky, Ann. Statist. 7, 697 (1979).


  • Page 32

    UW CENPA Annual Report 2014-2015 April 2015 17 MAJORANA 1.11 Overview of the MAJORANA DEMONSTRATOR T. H. Burritt, M. Buuck, C. Cuesta, J. A. Detwiler, J. Gruszko, I. Guinn, J. Leon, A. Li, D. A. Peterson, R. G. H. Robertson, A. H. Thompson, and T. D. Van Wechel The MAJORANA DEMONSTRATOR is a „40-kg array of high-purity germanium (HPGe) de- tectors enriched in 76 Ge that is under construction at the Sanford Underground Research Facility (SURF) in Lead, SD. The detectors are arranged in strings of 4-5 on low-background copper supports which are then placed in two cryogenic lead- and copper-shielded modules (see Fig. 1.11-1). The primary technical goal of the DEMONSTRATOR is the achievement of a radioactive background of 3.1 counts/ton/year within a 4-keV region of interest surrounding the 2039-keV Q-value for 76 Ge neutrinoless double-beta (0νββ) decay. Such a low background level would justify deeper investment in a much larger ton-scale experiment with sufficient sensitivity to definitively search for 0νββ decay for inverted hierarchical neutrino masses. In the process, the DEMONSTRATOR will simultaneously perform a sensitive test of a claimed observation of 0νββ decay1 , and will also perform searches for low-mass WIMP dark matter, solar axions, and other physics signals. Figure 1.11-1. MAJORANA DEMONSTRATOR schematic. The UW MAJORANA group leads the simulations and analysis effort within the collabora- tion, and contributes significantly to efforts in software development, data analysis techniques, background modeling, geometrical models for Monte Carlo simulations, data handling and storage issues, and feedback to design teams on background impact of proposed modifica- tions. We are also responsible for the design, production, and testing of the low-mass signal cable connectors to be used in the experiment, and for the testing of the high-voltage (HV) feedthrough flanges and Picocoax R cables that deliver the bias voltage to the detectors. In addition to these responsibilities, our group at the University of Washington is also contribut- ing to the ongoing construction efforts at SURF of the MAJORANA detector arrays. 1 H. V. Klapdor-Kleingrothaus et al., Phys. Lett. B 586, 198 (2004).


  • Page 33

    18 1.12 String building for Module 1 T. H. Burritt, M. Buuck, C. Cuesta, J. A. Detwiler, J. Gruszko, I. Guinn, J. Leon, D. A. Peterson, R. G. H. Robertson, and T. D. Van Wechel CENPA MAJORANA collaboration members have worked extensively building detector strings for the MAJORANA DEMONSTRATOR, contributing over 2000 on-site hours in the past year. The strings, which are tower assemblies holding four to five P-type point contact HPGe detectors and the low-mass front end boards that provide the first stage of signal amplification (seen in Fig. 1.12-1), are built entirely at the Sanford Underground Research Facility (SURF) Davis Campus on the 4850 ft. level. They are assembled in a class-10 glovebox to mitigate radon backgrounds, and then tested in individual string test cryostats. Following validation, they are installed in one of two electroformed copper cryostats, or the commercial copper prototype cryostat. Seven strings are installed into each of the modules that make up the MAJORANA DEMONSTRATOR. Figure 1.12-1. Left: The MAJORANA detector unit holds a P-type point contact HPGe detector and low-mass front end. Right: Detector assemblies are stacked to create a string. Over the past year, the MAJORANA collaboration has reached many milestones in string and module operations. Following improvements to the prototype cryostat, which was used to develop the design and procedures needed for the MAJORANA DEMONSTRATOR, the pro- totype module was commissioned in May and June of 2014. Since July of 2014, it has been stably taking data with nine natural-abundance detectors. The veto system was integrated into its operation in September of 2014. The procedures needed for the string-building of Module 1 were finalized via the con- struction of another natural-abundance detector string, and seven strings containing enriched detectors were built and tested for use in Module 1. This work continued over several months,


  • Page 34

    UW CENPA Annual Report 2014-2015 April 2015 19 Figure 1.12-2. Seven strings are installed in each module of the MAJORANA DEMON- STRATOR. during which CENPA members contributed to detector unit assembly, string assembly, and string characterization. After Cryostat 1 was commissioned, the 7 strings were installed, as in Fig. 1.12-2. Commissioning of the 29 detectors of Module 1 began in early April 2015, and will continue in the coming months. 1.13 High-voltage cable and feedthrough characterization T. H. Burritt, M. Buuck, C. Cuesta, J. A. Detwiler, J. Gruszko, I. Guinn, J. Leon, A. Li, D. A. Peterson, R. G. H. Robertson, A. H. Thompson, and T. D. Van Wechel The high-voltage (HV) cables and feedthrough flanges to be used at the MAJORANA DEMON- STRATOR have been characterized looking for micro-discharge events (MD) at the University of Washington. In total, 129 HV cables were tested: 42 for Module 1, 38 for Module 2 and 49 for the string test cryostats; and 280 HV feedthroughs since there are five 8” flanges with 40 feedthroughs each for Modules 1 and 2, and eight 6” flanges with 10 feedthroughs each for the string test cryostats. First, the leakage current of the flange feedthroughs was measured up to 5 kV, which is the maximum operating voltage of the detectors. The feedthroughs failing this test, i.e. with leakage current ą2 µA, were not further tested and used as ground at the MAJORANA DEMONSTRATOR. Then, each cable was tested with a feedthrough for at least 3 h, with an average time of 14 h. If a set did not pass the test, the cables were tested individually and the one creating MD events identified. The criteria that each cable and feedthrough have to fulfil to be accepted are: Holding 5 kV without current fluctuations, no excessive leakage current, and ă5 MD/h. The leakage current results are shown in Table 1.13-1, where it can be seen that 98% of the pins showed typical leakage current values at 5 kV. However, some pins were not


  • Page 35

    20 recommended to be used at the MAJORANA DEMONSTRATOR as high-voltage. Flange model Total High-leakage current Average feedthroughs feedthroughs leakage current 6 inch 80 5 1648 ˘ 61 nA 8 inch 200 2 1653 ˘ 57 nA Table 1.13-1. Leakage-current measurements results of the feedthroughs to be used at the Majorana Demonstrator. The MD event rate per cable results are shown in Fig. 1.13-1. The rate should be consid- ered as an upper limit since a contribution to the rate from the feedthrough or other parts of the electronic chain is expected. No issues were found for any of the cables, and a rate ă5 MD/h was measured for all cables, with an average of 0.60 ˘ 0.11 MD/h. One feedthrough was found to create a rate ą5 MD/h and was not recommended to be used at the MAJORANA DEMONSTRATOR as high-voltage. Figure 1.13-1. Micro discharge event rate per cable results for the cables to be used at the Majorana Demonstrator. Figure 1.13-2. Left: Picture of the MAJORANA Module 1 cross-arm with the signal and high-voltage cables installed. Right: Picture of Module 1 already installed with a closer view of the detectors and cables. The cables have already been installed in the string test cryostats and in Module 1. Fig. 1.13-2 shows the high-voltage and signal cables being installed into Module 1 and a close view of Module 1 with the cables and detectors installed.


  • Page 36

    UW CENPA Annual Report 2014-2015 April 2015 21 1.14 Low-background signal connector production and testing T. H. Burritt, M. Buuck, C. Cuesta, J. A. Detwiler, J. Gruszko, I. Guinn, J. Leon, A. Li, D. A. Peterson, R. G. H. Roberston, A. H. Thompson, and T. D. Van Wechel In order to reach the background goal of ă3 counts/ROI-ton-yr (ROI is the 4 keV region of interest around the Q-value of the 0νββ), the MAJORANA DEMONSTRATOR requires low- background signal connectors. Each signal connector must connect the 4 coaxial cables that run between the front-end charge-collecting amplifier placed near the point contact of each detector to a feedthrough flange running out of the cryostat. The connectors are placed above the cold plate directly above the detector array. Commercially available connectors use electrical contact springs made out of beryllium copper alloy (BeCu), which is difficult to get with a low enough uranium and thorium content to meet the background goal. For this reason, UW has developed signal connectors for the experiment. These connectors must have a low mass and use low-activity materials. They must also be able to survive multiple cycles to vacuum pressure and liquid-nitrogen temperatures, and must be easy to manipulate inside a glovebox without breaking. UW has developed a design consisting of electrical contact pins and sockets housed in Vespel R plugs, shown in Fig. 1.14-1. The connectors use gold-plated brass pins and sockets manufactured by Mill-Max R , with the BeCu contact springs removed. The pins and sockets are deliberately misaligned so that when plugged in, the pin is forced to bend against the socket. The spring force from this bending plays the role of the contact springs, providing reliable electrical continuity. The pins and sockets are attached to the signal cables using a low background Sn-Ag eutectic that was developed for the SNO experiment. Strain relief is provided by FEP heat shrink. Figure 1.14-1. Left: A signal plug pair, with a BNC connector for comparison. Right: A signal cable soldered to a signal plug. Production of signal plugs for the MAJORANA DEMONSTRATOR began in April 2014. The Vespel housing is machined at the 4850’ level of SURF. The housing is then leached in nitric acid, sonicated in ethanol and deionized water, and pumped and baked. The pins are


  • Page 37

    22 sonicated in ethanol and deionized water before being inserted into the housings underground. These plugs are then sent to UW to be soldered to signal cables in a clean room. Quality control (QC) tests are then performed to reject plugs that do not form reliable connections. The plugs are then sent to LBNL and SURF to be incorporated into the strings of detectors. So far, 159 female plugs and 109 male plugs have been produced. Of these, 112 females and 58 males have been soldered for use in Module 1 and the string test cryostats. Approximately 20-25% of plugs are rejected during the QC stage. In preparation for the production of signal plugs for Module 2, a batch of female plugs with different dimensions is being tested to improve on this figure. Assays of the materials in the signal plugs and a full-body assay of 2 signal plugs have been performed to estimate the activity of the connectors. The assays were combined with the results of the background model to estimate the background contribution in the region of interest. Table 1.14-1 lists all background contributions. The results are consistent with a background dominated by the pins. Notice that the BeCu contact springs alone would surpass the MAJORANA DEMONSTRATOR’s background goals by more than a factor of three. The current estimate for the contribution from signal connectors is 0.284 cts/ROI-t-yr. Material Assay Mass [g per Isotope Activity MJD BG Method conn. pair] [µBq/kg] [c/ROI/t/y] Pins (w/BeCu) ICP-MS 0.112 238 U 795000 ˘ 12000 8.8 ˘ 0.1 232 Th 41000 ˘ 1000 2.3 ˘ 0.1 Pins (no BeCu) ICP-MS 0.112 238 U 4600 ˘ 1500 0.05 ˘ 0.02 232 Th 5800 ˘ 100 0.32 ˘ 0.01 Vespel SP-1 NAA 0.95 238 U ă1000 ă0.20 232 Th ă12 ă0.01 Solder GDMS 0.04 238 U 5600 ˘ 1000 0.02 ˘ 0.004 232 Th ă12 ă0.0002 Solder flux GDMS 0.04 238 U 1200 ˘ 200 0.005 ˘ 0.001 232 Th ă400 ă0.007 FEP (shrink tube) NAA 0.1 238 U ă1250 ă0.012 232 Th ă138 ă0.007 Total (sum of 1.25 238 U ă1490 ă0.29 materials) 232 Th ă550 ă0.35 Total (full ICP-MS 0.95 238 U 1160 ˘ 20 0.110 ˘ 0.001 body assay) 232 Th 365 ˘ 6 0.174 ˘ 0.003 Table 1.14-1. Summary of all background contributions from signal plugs. ICP-MS is inductively coupled plasma mass spectrometry. GDMS is glow discharge mass spectrometry. NAA is neutron activation analysis. Material component masses are all approximate, with a bias towards larger masses.


  • Page 38

    UW CENPA Annual Report 2014-2015 April 2015 23 1.15 Simulation and analysis activities for the MAJORANA DEMONSTRATOR T. H. Burritt, M. Buuck, C. Cuesta, J. A. Detwiler, J. Gruszko, I. Guinn, J. Leon, D. A. Peterson, R. G. H. Robertson, and T. D. Van Wechel The UW MAJORANA group leads the simulations and analysis effort within the MAJORANA collaboration. Much of the low-level software for data I/O, event building, data processing, and simulation were written by CENPA personnel. Members of our group have played a central role in the building and validation of the background model for the MAJORANA DEMONSTRATOR, which informs the radiopurity criteria upon which the experimental design is evaluated. We also participate in the development and implementation of data analysis techniques, geometrical models for Monte Carlo simulations, and data handling, storage, and database technologies. Our efforts over the past year have focused on workflow management, improvements to data processiong and cleaning, the development and use of analysis tools for data from the MAJORANA DEMONSTRATOR Prototype Module, and preparation for data taking with enriched Ge detectors. As the Simulations and Analysis task leader for MAJORANA, Jason Detwiler oversees the development of the software tools necessary for full data taking. This past year, Detwiler implemented a working group structure to better organize the members of the collaboration around the software tasks at hand. This restructuring has been a great success, and has significantly accelerated the production of the MAJORANA software tools. Clara Cuesta took the role of head of the Data Cleaning and Run Selection working group. The group has implemented a Data Cleaning Framework to identify different types of events and to specify cuts to remove non-physics events from data-sets. External variables, such as environmental effects, are being integrated with the Ge data to be studied. Run and channel selection tools are being developed to automatically determine which runs will be used in neutrinoless double-beta decay analysis. Good runs will be integrated into data sets and a database will handle all the information, including the blind data. Finally, the total exposure and efficiency of the experiment will be determined. Run evaluation has been done with the prototype module data and will be done for Module 1 as soon as it starts taking data. Micah Buuck has primarily been focused on implementing a pulse-shape-based analysis technique for rejection of background multi-site events. In this capacity, he is a part of the Pulse Shape working group. The technique requires the generation of a “basis library” of single-site event pulses, which is then used to accept or rejcet incoming pulses based on a χ2 fit. He has successfully implemented the software necessary to generate the basis, and is now working on streamlining the process and applying it to physics data. Ian Guinn has significantly improved the event builder for the MAJORANA DEMONSTRAT- OR. The event builder has three main responsibilities. First, it converts the raw data files produced by the data acquisition system (ORCA) into ROOT files, known as built files, that are compatible with the MAJORANA software. Second, it combines waveforms and muon veto events that occur at closely separated times into a single event, in order to detect coincidences. Finally, it filters out corrupted data that may confuse the main analysis software, a process known as garbage collection. Guinn has made several improvements to the event builder in


  • Page 39

    24 the last year, including but not limited to: the addition of tools to read the ORCA XML header file, which contains important information about the hardware configuration, the ability to read out events from the muon veto, the addition of the garbage collector, and a 6-fold reduction in file size. Julieta Gruszko has successfully implemented and begun to use new frequency-domain analysis techniques. She will use the techniques to identify and eliminate noise sources and for data cleaning. The resulting noise curves of detector baselines will be used to characterize the detector, to accurately simulate pulses, and for optimum filtering. The tools she developed this year allow researchers to view noise curves over time, which can be used to identify intermittent noise sources and therefore tag events for data cleaning. Average noise curves taken with the Prototype Cryostat system are being used to identify the optimal electronics setup and laboratory conditions for data taking. For an example, see Fig. 1.15-1. 106 raw spectrum 105 with all cuts 104 counts/keV/tonne/year 103 102 10 1 10-1 0 1 2 3 4 5 6 7 8 9 10 energy [MeV] Figure 1.15-1. A comparison of the white Figure 1.15-2. Full-spectrum background noise levels showed that the use of fluores- model for the MAJORANA DEMONSTRA- cent lights in the lab was introducing noise. TOR, with and without analysis cuts. Following this study, the shielding of module electronics box was improved. Our attention is now turning toward completing preparations for the first data to come from the enriched Ge detectors. New assay results have improved the predicted background rate in the 4-keV region of interest surrounding the 2039 keV Q-value for double-beta de- cay of 76 Ge to 3.1 counts/ton-year. Major simulation campaigns are underway to provide up-to-date predictions for the full spectrum we expect to see with enriched detector data. Fig. 1.15-2 shows the full simulated spectrum, including the effect of analysis cuts. Other major activities include software quality assurance tests, database implementation of run in- formation recording and automatic data workflow management, refinement of event building routines, optimization of energy estimation and pulse-shape parameter extraction algorithms, and data monitoring and cleaning routines.


  • Page 40

    UW CENPA Annual Report 2014-2015 April 2015 25 1.16 Low-noise forward-biased preamplifier tested with a mini-PPC Ge detector J. A. Detwiler, J. Leon, D. A. Peterson, R. G. H. Robertson, and T. D. Van Wechel P-type point contact (PPC) germanium detectors are in principle sensitive to very low-energy nuclear recoils, such as those produced by coherent scattering of dark matter particles or neu- trinos. In order to realize this sensitivity, low-noise electronics capable of achieving sub-keV energy thresholds are required. We report on the development of a low-noise charge-sensitive preamplifier which is continuously reset by the forward-biased gate-to-source junction of an input tetrode JFET. In contrast to commonly used reset mechanisms, this design avoids the noise associated with a feedback resistor while also providing continuous operation. Refer to previous reports for more information on the preamplifier’s design principle and operation1 . Whereas previous protoypes have only been tested with Si PIN diode detectors, the most recent prototypes have been designed to work with a Ge detector. Given the small signals pro- duced by PPC Ge detectors, the first-stage (or front-end) amplifier must be connected near the detector. Otherwise, capacitive loading from long cables would reduce the signal-to-noise ratio. For this reason, it is crucial to choose materials that can be made or found to be low in radiactive backgrounds. Our front ends were fabricated out of gold on a fused-silica wafer2 . This substrate material has a low loss tangent (10´4 ) for low dielectric dissipation noise, and good thermal conductivity (41.9 W/mK) which makes it possible to control and optimize JFET temperature. Fabrication of the front ends was performed at the UW Nanofabrication Facility using standard photolithographic techniques. A thin adhesion layer of Ti was de- posited followed by 4000Å of Au to form the circuit traces. Bare-die Moxtek MX-30 JFETs were epoxied on the front end and wirebonded to the gold traces (see Fig. 1.16-1). Figure 1.16-1. Left: A front end board (20 mm X 10 mm) made from 500-mm-thick fused silica, shown here clamped in aluminum mount. Right: Pulse-height spectrum collected with Amptek MCA8000D using 241 Am and injected pulser. Semi-gaussian shaping time set to 24 µs and pulser frequency at 500 Hz. Refer to text for more information. 1 CENPA Annual Report, University of Washington (2013) p. 24. 2 From MarkOptics, Corning 7980


  • Page 41

    26 A miniature PPC Ge detector (mini-PPC) has been used to test our preamplifier. The mini-PPC1 has low leakage current (ă1 pA) and low capacitance („0.5 pF), which means it has the intrinsic low-noise performance necessary to test our preamplifier. A tensioned pin between the preamplifier and detector makes the electrical connection to the input. This system is cooled down to 87 K inside a cryostat via a copper coldfinger. Signals generated by the detector are amplified by the front end and passed to a second-stage amplifier. We can subsequently pass these amplified signals through a semi-gaussian shaping amplifier and measure the pulse-height spectrum with a multi-channel analyzer (MCA). Fig. 1.16-1 contains an energy spectrum collected in such a way using an 241 Am source. The peak on the left is from the source’s 59.5 keV gamma, which is used for energy calibration. The energy of the pulser peak (63.5 keV) on the right corresponds to an injected amount of charge, which in turn is used to calibrate the feedback capacitor („0.16 pF). The electronic noise level can also be inferred from the full-width-half-maximum (FWHM) of this pulser peak, 73 eV FWHM. This demonstrates one of the lowest electronic noise levels observed with a PPC Ge detector, even when the mini-PPC’s low capacitance is taken into account. Figure 1.16-2. Noise level of preamplifier connected to mini-PPC at 150 V bias and 87 K, in terms of squared electrons-RMS versus full-width-half-maximum time of a semi-gaussian shaper (blue dots). Also shown are a three-component fit to data (red solid) and individual noise component fits (red dashsed). Alternatively, the noise level can be measured from the RMS of baseline noise. In this method, the output from the shaping amplifier was passed on to a digital oscilloscope that was then used to average the RMS of ten baseline noise traces. Fig. 1.16-2 shows the electronic noise measured in this way versus shaping time. The minimum of this curve corresponds to about 65 eV FWHM, which is consist with the previous measurement method if one considers potential pulser instabilities and the effects of pileup. Additional tests with a 60 Co source show that this system has a linear energy response out to at least 1.33 MeV. 1 2 cm X 1 cm cylinder fabricated at Lawrence Berkeley National Lab.


  • Page 42

    UW CENPA Annual Report 2014-2015 April 2015 27 2 Fundamental symmetries and non-accelerator-based weak interactions Torsion-balance experiments 2.1 Progress on the upgrade of the wedge-pendulum experiment for test- ing short-range gravity E. G. Adelberger, S. Fleischer, C. D. Hoyle∗ , J. G. Lee, and H. E. Swanson The plans for a comprehensive upgrade of the Fourier-Bessel short-range experiment have been described in the previous annual report1 . In short, the upgrade consists of a new set of pendulum and attractor disks, an improved electrostatic shield with a motorized in-situ alignment system for the attractor-shield gap, a new vacuum chamber and pump line, a new calibration turntable, and new in-vacuum wiring and optical readout alignment. With the exception of the new pendulum and attractor, all parts of the upgrade have been manufactured, tested, and are ready to be deployed. After some further improvements to the alignment system, we have demonstrated working separations of less than 5 µm between the attractor and the electrostatic shield. Regarding the production of the new pendulum and attractor, we have developed a better process to glue the patterned platinum foils to the supporting glass substrate. It allows for significant improvements in terms of flatness, and it completely avoids the problem of surface irregularities as seen in the old process. Unfortunately an attractor disk assembled in this way showed clearly measurable permanent magnetization. The magnetic contamination probably comes from a layer of stainless-steel foils which — due to a miscommunication — surrounded the stack of platinum foils during the electric-discharge machining (EDM) of the foil patterns. In the meantime, we have investigated the potential of our new UV-laser prototyping setup for a radically simplified pattern-manufacturing process. While some issues related to burr formation in the laser-cutting process still need to be addressed, it looks as though this might be a feasible way to produce the pattern for the Fourier-Bessel experiment. It would be dramatically faster and cheaper than our previous EDM-based method. Additionally, as it would allow cutting the active pattern into a metal foil that has already been glued to a glass substrate, it would solve several problems in the gluing and centering process. As a further safeguard against the magnetic contamination issue, we will screen the raw materials for magnetic impurities before assembling the next pieces. In summary, almost all parts of the short-range setup upgrade have been manufactured and tested, and are ready to be deployed. They can be installed as soon as the data-taking ∗ Humboldt State University, Arcata, CA. 1 CENPA Annual Report, University of Washington (2014) p. 33.


  • Page 43

    28 for the spin-spin experiment has been completed. The only remaining question concerns the best process to create and mount the active patterns for the pendulum and the attractor. Significant progress has been made in this regard, and we may have a much cheaper and faster method available now. We are looking at mitigating the problems related to burr formation. Additionally, the extensively researched EDM-based process is available as a fallback solution. 2.2 Progress on a high-precision ground-rotation sensor for Advanced LIGO J. H. Gundlach, C. A. Hagedorn, M. D. Turner, and K. Venkateswara Advanced LIGO (aLIGO) is a next-generation gravitational-wave-detector system expected to make first detections in the next few years. Key among the improvements in the new detector is a better seismic isolation system to cancel the effect of ground motion on the suspended optics. This is done through an active-control system which measures the ground translation through seismometers and applies a corrective force on the optics platform. However, at low frequencies (10-500 mHz) conventional seismometers and tiltmeters are unable to separate ground rotation (tilt) and horizontal acceleration, which limits their ability to correct for ground translation. Thus, aLIGO needs a ground-rotation sensor to accurately distinguish between horizontal-ground acceleration and rotation. Such an instrument is not commercially available and the required sensitivity of the rotation sensor is challenging to reach. Over the last four years, we have developed two prototype instruments that meet a sig- nificant part of the requirement specified by the seismic-isolation team for aLIGO in the frequency range of 40 to 400 mHz. The device consists of a low-frequency flexure-beam bal- ance. We measure its angle using a multi-slit autocollimator mounted rigidly to the ground. Above its resonance frequency, the balance remains inertial, thus the autocollimator mea- sures ground rotation. Torque from horizontal acceleration is rejected by locating the center of mass at the pivot point of the flexure. The prototype beam-balance consists of a 0.75-m aluminum tube with 1.8 kg-brass weights attached at each end. It is suspended by two, 15-µm-thick, copper-beryllium flexures. The entire balance is surrounded by an aluminum heat shield and is placed in high vacuum to minimize thermal effects. One rotation sensor was installed at one of the end-stations at the LIGO Hanford Obser- vatory. It is now referred to as the Beam Rotation Sensor (BRS). Under windy conditions, tilt as measured by the BRS has been shown to be very coherent with a seismometer output at frequencies below 0.1 Hz, and has been used to remove tilt-induced noise from the seismome- ter output. Fig. 2.2-1 shows an example of a measurement taken when wind speeds were between 15-25 mph. In the top plot, the red trace shows the tilt measured by BRS, the blue curve shows the seismometer output (converted to angle units), and the cyan curve shows the tilt-subtracted ground translation signal. At frequencies greater than 0.1 Hz, the cyan curve follows the blue, but below that frequency, tilt dominates the seismometer output. Also shown as the purple trace is the BRS reference signal, which is indicative of the instrument


  • Page 44

    UW CENPA Annual Report 2014-2015 April 2015 29 noise, and the output of a seismometer mounted on the optics platform is shown in green. The bottom plot shows the coherence between the two seismometers and the BRS, indicating that tilt constitutes a significant part of the low-frequency signal in both seismometers. GND T240 and BRS motion compared to ST1 T240X −5 10 2 GND STSX ( * w /g) 2 −6 ST1 T240X ( * w /g) 10 BRS RY out GND X Super−sensor (tilt subtracted) ASD (rad/rt(Hz)) −7 10 BRS ref −8 10 −9 10 −10 10 −2 −1 0 10 10 10 frequency (Hz) created by BRSanalyze2 on 13−Oct−2014 J. Kissel, K. Venkateswara 1 0.8 coherence 0.6 0.4 GND STS and ST1T240X 0.2 ST1 T240X and BRS RY GND STSX and BRS RY 0 −2 −1 0 10 10 10 frequency (Hz) Figure 2.2-1. Seismic data measured at LHO. Top plot shows the ground rotation recorded by our sensor (BRS, red) in comparison to a seismometer (T240, blue) measured at LIGO Hanford Observatory. Also shown are the tilt-subtracted translation signal (cyan), sensor noise (purple) and the optics platform motion (green). Bottom plot shows the coherence between the two seismometers and the BRS. Measuring the tilt-free translation of the ground has been demonstrated to improve the isolation performance of the optics platform by reducing its low frequency motion, especially in windy conditions. Doing so enables more consistent locking of the interferometer and can help reduce up-conversion effects and hence improve its sensitivity. Further testing and characterization of this instrument is currently underway at LHO. At our lab, we are building a new compact version of the BRS with several improvements. A new compact readout using two fiber-optic interferometers is being developed to improve sensitivity. A mechanism for repeatable locking of the beam balance is being tested to enable the instrument to be transported safely. Further, the balance will have a cross shape to reduce sensitivity to gravity gradient noise.


  • Page 45

    30 2.3 Silica fibers for increased torsion-balance sensitivity J. H. Gundlach, C. A. Hagedorn, J. G. Lee, E. A. Shaw, W. A. Terrano, M. D. Turner, and K. Venkateswara We are exploring fabrication and use of fused silica as a superior torsion fiber material for several torsion-balance experiments, with a focus on our equivalence-principle test. Many of our torsion balances are limited by Brownian motion intrinsic toa the torsion fiber. The thermal torque noise power, for an internally-damped fiber, is τ pf q “ 2kB T κ{pπQf q, where kB is Boltzmann’s constant, T the fiber temperature, κ the fiber torsion constant, and Q the mechanical quality factor. To minimize noise we must minimize the ratio of the torsion constant κ to the Q factor1 . We have made silica fibers with torsion constants comparable to those of tungsten fibers with equal tensile strength. In addition, we have measured Q „ 270, 000 in one of these fibers compared to the best Q „ 6000 for tungsten fibers. It is likely that the Q of this fiber is still limited by non-fiber losses. In past work, flame-pulled fibers reached high Q at modest tensile strength, but fabrication was difficult and inconsistent1 . Our recent improvements were enabled by the CO2 laser at CENPA2 . Last year, we built an apparatus that used the laser to pull fibers, but not of the right diameter nor length. The main improvement was to switch from fiber pulling powered by a stepper motor to using the laser cutter’s integrated servo rotary stage. With smoother operation and integrated timing between the laser control and rotary stage, we were able to reliably explore the fiber-pulling and laser-heating parameter space. We have improved the characterization of our fibers. After pulling, fibers are checked with a reference pendulum for satisfactory κ and minimum breaking strength. At fixed κ the tensile strength is limited by thin spots in the fiber. Fiber uniformity is critical to optimal performance. We imaged our fibers along their entire length using the CENPA SmartScope and used custom Python software to determine fiber diameter curves. These characterizations allowed us to improve on uniformity and reproducibility. A fiber measurement is shown in Fig. 2.3-1 (left). To measure fiber Qs and torque noise, we restored the LISA apparatus3 and attached one of our new autocollimators4 . Our first laser-pulled fiber initially displayed Q „ 28, 000, low for a silica fiber. After excluding gas damping, we found that the dominant loss was magnetic. Canceling ambient fields with Helmholtz coils gave a Q of 60,000-70,000. Inserting a refurbished µ-metal shield gave us Q “ 271, 000 ˘ 1, 000, shown in Fig. 2.3-1 (right). We 1 C. A. Hagedorn, S. Schlamminger, and J. H. Gundlach, 2006 Proc. 6th Int. LISA Symp. AIP Conf. Proc. 873, 189 (2006). 2 CENPA Annual Report 2013, pp 118. 3 S. E. Pollack, M. D. Turner, S. Schlamminger, C. A. Hagedorn, and J. H. Gundlach, Phys. Rev. D 81, 021101(R) (2010). 4 CENPA Annual Report 2013, pp 47.


  • Page 46

    UW CENPA Annual Report 2014-2015 April 2015 31 Figure 2.3-1. Left: Diameter-vs-length characterization of fiber currently being used for Q measurements and investigation of noise. Right: Q measurement at large amplitude done by measuring the time of zero crossings of the pendulum mirrors and using that information to fit a sine curve for each period. are installing a purpose-built magnetic shield to minimize magnetic damping further. We will continue our investigation of silica fibers until they reliably surpass tungsten as our fiber of choice. 2.4 Parallel-plate inverse-square-law test J. H. Gundlach, C. A. Hagedorn, M. D. Turner, and K. Venkateswara This has been a year of documentation and data analysis. We will unblind our parallel-plate test of gravity1 at sub-millimeter scales on May 5, 2015. Charlie Hagedorn’s thesis is nigh complete, pending a post-unblinding conclusion. Our bootstrapped analysis method allows easy accommodation of difficult-to-model sys- tematic uncertainties, including the important effects of isolating-foil displacement and sys- tematic uncertainty in the pendulum-attractor distance (“horizontal error bars”). While bootstrapping removes analytic intuition from error propagation, its simplicity is an advan- tage. A small, important, insight is that, in the Yukawa ` parametrization, in which Newtonian gravity is modified to the form V prq “ ´G m1rm2 1 ` αe´r{λ , α and λ share no a priori ˘ relationship. Thus, the traditional ‘pick a λ, fit for α’ approach to constraining inverse- square-law violations is more appropriate than sophisticated approaches like kernel-density estimators. Our bootstrapped method is quite traditional, determining α exclusion limits at 1 CENPA Annual Report, University of Washington (2014) p. 32.


  • Page 47

    32 a given λ from best-fit points with nearby λ. If there is a signal to find, our method moves smoothly from exclusion to detection. 7 6 5 Stanford 08 4 Experimentally Excluded 3 log10(|α|) 2 1 Huazhong 07 0 Washington 04 -1 -2 Washington 12 -3 Gravitational Strength Washington 06 Huazhong 12 -4 -5 -4 -3 -2 log10(λ) (meters) Figure 2.4-1. Expected limits for this iteration of the parallel-plate experiment, including systematic uncertainties. Our 2σ limit should fall in the green or yellow (optimistic) regions. Useful upgrades for the experiment are clear: Re-enabling our pendulum/foil voltage control system will yield an immediate improvement in signal-to-noise ratio and decreased systematic coupling. This change, along with minor improvements to the autocollimator, alignment, and data-taking procedure, should yield an improved measurement within six months of running time. An upgraded pendulum may allow greater sensitivity through improved metrology and electrostatic properties. Our automated analysis software is ready for a second round of data. 2.5 Improved short range spin-coupled force test E. G. Adelberger, B. R. Heckel, J. G. Lee, and W. A. Terrano We have upgraded the torsion pendulum designed to look for short-range spin-coupled forces1 . Residual magnetic couplings and thick magnetic shielding limited the previous version. Up- grades consisted of multiple improvements to magnetic shielding and thinner components to decrease separations, increasing the sensitivity of the balance and increasing the signal respectively. Multiple layers of thin magnetic shielding proved to be more effective than a single thick layer. The 0.030” mu-metal cans on the 20-pole magnet of the pendulum and attractor were each replaced with two layers of 0.010” nested mu-metal cans with a layer of 0.001” aluminum foil between each can. The new shielding was tested on a turntable under a 1 CENPA Annual Report, University of Washington (2014) p. 36.


  • Page 48

    UW CENPA Annual Report 2014-2015 April 2015 33 GMR probe showing a factor of about 14 improvement in the magnetic field peak-to-peak (Fig. 2.5-1). Alternating 76-µm tungsten and titanium shims glued directly to the magnet rings eliminated a 0.050” alignment disk and gravitationally compensated both the pendulum and attractor. Figure 2.5-1. Magnetic field from enclosed 20-pole magnets. Black is with a single 0.030” mu-metal can; red is with two nested 0.010” mu-metal cans. Figure 2.5-2. Improved torque sensitivity in the 2015 vs. the 2014 gravitational data. The improved noise floor comes from moving from a 30-µm fiber to a 20-µm fiber, a higher Q from improved shielding, and a faster rotation rate. The smaller 10ω peak comes from improved gravitational compensation. The improved shielding decreased the loss of the pendulum oscillation, 1/Q. We also replaced the 30-µm fiber of the previous test with a 20-µm fiber. Additionally, the electrical contact on the turntable was replaced with a design that caused less friction and allowed a faster rotation rate. The increase in Q (1.6x), thinner fiber (2x), and faster attractor rotation rate (1.4x) improved the torque sensitivity of the balance by a factor of 4.5, as shown in Fig. 2.5-2. Finally, we replaced the shielding screen with 10 layers of 0.002” mu- metal and 1 layer of 0.010” mu-metal, all with 0.001” aluminum foil spacing between each layer.


  • Page 49

    34 Figure 2.5-3. New limits on a monopole-dipole force. The shaded region is excluded at 1σ Data-taking with the gravitational attractor is complete and has led to improved limits on CP -violating monopole-dipole couplings (Fig. 2.5-3) for bosons of mass between 1 and 400 µeV. Data-taking with the spin attractor is complete, having reached a torque sensitivity of 2 atto-Nm and an a attractor to pendulum separation of 4.1 mm, down from 6.1 mm. This data sets new limits on dipole-dipole interactions for all mass ranges (Fig. 2.5-4 left). Interpreting these results as excluding new Goldstone bosons implies that a new hidden symmetry must be broken at a scale F ą 100TeV for bosons with mass less than 100µeV (Fig. 2.5-4 right). Figure 2.5-4. Left: New limits on dipole-dipole interactions. The shaded region is excluded at 1σ. Right: New limits on a hidden high-energy symmetries. The shaded region is excluded at 1σ.


  • Page 50

    UW CENPA Annual Report 2014-2015 April 2015 35 Non-accelerator-based weak interactions 199 2.6 The Hg electric-dipole-moment experiment Y. Chen, B. Graner, B. Heckel, and E. Lindahl The 199 Hg EDM experiment has been taking data since the summer of 2013. After a several- month delay due to the failure of the UV laser system, data acquisition resumed. In February, 2015, a complete data set was completed. An initial analysis of the data revealed a new systematic error associated with small movements of the stack of 199 Hg vapor cells due to forces induced by the applied electric field. We are currently taking additional auxiliary data to better understand the impact of the vapor cell movement onto the EDM signal. The 199 Hg EDM apparatus uses UV laser light to polarize and measure the spin precession frequency of 199 Hg atoms in a stack of 4 vapor cells, 2 of which (the inner pair) have oppositely directed electric fields. A pump-probe sequence is employed, with the electric fields reversed between each pump-probe cycle. A blind frequency offset is added to the inner-cell frequency difference to mask the EDM signal until the analysis of the entire data set is complete. EDM data collection is divided into “sequences”, groups of 16 days of data distinguished by the ordering and orientation of the 4 vapor cells. Each sequence includes equal numbers of runs with electric field magnitudes of 6 kV/cm and 10 kV/cm, magnetic field normal and reversed, and slow and fast electric field ramp rates. We have completed 12 sequences of EDM data, a complete set of data for 3 vapor cells cycling through the EDM-sensitive positions in the apparatus. Fig. 2.6-1 shows the EDM measurements taken at 10 kV/cm for the 12 data sequences. Each run represents one day of data accumulation. A similar data set exists for measurements taken at 6 kV/cm. Figure 2.6-1. Angular frequency (rad/s) Hg EDM results for the 10 kV/cm data sequences. Preliminary analysis of the data gives a statistical error on the 199 Hg EDM of 2.7 ˆ 10´30 e cm. This represents an improvement by a factor of 5 over our previous 199 Hg EDM result from 2009. There is good agreement between the data taken at 10 kV/cm and 6 kV/cm, and between data taken for the two magnetic field directions. The systematic error analysis is in progress and we expect to complete the analysis and unblind the data in the summer of 2015. This project is supported primarily by the NSF (P. I. Heckel).

  • View More

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