The miniTimeCube A Portable Directional AntiNeutrino Detector Steve

The mini-Time-Cube A Portable Directional Anti-Neutrino Detector Steve Dye, John Learned, Shigenobu Matsuno, Marc Rosen, Michinari Sakai, Stefanie Smith, Gary Varner, and students Univ. of Hawaii: Plus some other colleagues elsewhere Presentation at ANT 11, Philadelphia, 10 October 2011 1

m. TC Idea Do imaging with (100 ps) fast timing, not optics (time reversal imaging). Small portable 2. 2 liter scintillating cube, Boron doped plastic. 13 cm 4 x 6 MCP (x 64 pixels each) fast pixel detectors on surrounding faces Get neutrino directionality. Reject noise on the fly. 2. 2 liter 10 October 2011 ~10/day anti-neutrino interactions (inverse beta decay signature) from power reactor (San Onofre). John Learned at ANT 11 in Philadelphia 2

m. TC Virtues • Small size avoids positron annihilation gammas which smear resolution (Xo ~42 cm). . gammas mostly escape, permitting precise positron creation point location. • Fast pixel timing (<100 ps) and fast pipeline processing of waveforms rejects background in real time. • Having many pixels plus use of first-in light permits mm precision in vertex locations. • Neutrino directionality via precision positron production and neutron absorption locations. • No need for shielding (unlike other detectors). • Feasible even in high noise environment, near reactor vessel, at surface (eg. in a truck). 10 October 2011 John Learned at ANT 11 in Philadelphia 3

10 October 2011 Snapshot of the Fermat Surface for a Single Muon-like. Track Huygens wavelets John Learned at ANT 11 in Philadelphia Incoherent sum coincident with Cherenkov surface: Not polarized! J. Learned ar. Xiv: 0902. 4009 v 1 4

10 October 2011 Time Reversal Image Reconstruction John Learned at ANT 11 in Philadelphia Figure by Mich Sakai 5

10 October 2011 Directional Measurement John Learned at ANT 11 in Philadelphia Hiroko Watanabe (Workshop Towards Neutrino Technologies 2009) 6

Reactions in the Liquid Scintillator 2. α & triton stop ~immediately (mm). 2. 40 cm γ radiation length. 10 October 2011 John Learned at ANT 11 in Philadelphia Hiroko Watanabe (Workshop Towards Neutrino Technologies 2009) 7

10 October 2011 Neutrino Capture Nucleus Choices John Learned at ANT 11 in Philadelphia Hiroko Watanabe (Workshop Towards Neutrino Technologies 2009) 8

10 October 2011 Kam. LAND Calculations Kam. LAND resolution John Learned at ANT 11 in Philadelphia Hiroko Watanabe (Workshop Towards Neutrino Technologies 2009) 9

10 October 2011 Study using Kam. LAND LS and Resolution Hiroko Watanabe (Workshop Towards Neutrino Technologies 2009) We can, in principle, do much better with m. TC. Much further gain possible if can sense first neutron elastic scattering location. John Learned at ANT 11 in Philadelphia 10

Gammas Usually Leave m. TC Neutrino interaction point Mini-Time. Cube Gamma from neutron inelastic capture 2 m Gamma from positron annihilation Conclusion: Gammas from positron annihilation leave m. TC detector without interaction, thus not confusing vertex resolution. 10 October 2011 John Learned at ANT 11 in Philadelphia 11

10 October 2011 Average over reactor spectum Neutron only gets 15. 75 ke. V mean, most to positron. Inverse Beta Kinematics Trickier than you may realize. . . The positron gets the kinetic energy and the neutron gets the momentum. Plots by Stefanie Smith John Learned at ANT 11 in Philadelphia 12

We can select an event sample with good pointing Early neutrino captures point better 10 October 2011 Lower energies a little better Most early captures out ~ cm distance Early n captures: strong correlation between neutron and positron scattering angles. John Learned at ANT 11 in Philadelphia 13

Some tentative conclusions on neutrino direction determination in m. TC Get interaction vertex to ~1 mm and positron directions using fast timing and first hit reconstruction. Proton scatters by neutron would be most useful, but probably not enough light to reconstruct. Neutron capture location determined best with 6 Lithium by detecting alpha and triton. Can select golden events for best pointing. How good still TBD. Maximally utilize full information, employing full Max. L, yet to be done. We are exploring use of a neural network. 10 October 2011 John Learned at ANT 11 in Philadelphia 14

10 October 2011 John Learned at ANT 11 in Philadelphia 15

10 October 2011 John Learned at ANT 11 in Philadelphia 16

10 October 2011 Fitting the Positron Streak John Learned at ANT 11 in Philadelphia 17

First m. TC version using B loaded plastic scintillator First casting with bubble from Eljen 10 October 2011 John Learned at ANT 11 in Philadelphia 18

Mini-Time. Cube + PMTs and Readout Electronics (Portable) Volume ~ (2 ft)3 Weight < 30 kg ~43 cm (<1’ 6”) Plus separate processing electronics box. 10 October 2011 John Learned at ANT 11 in Philadelphia 19

Examples of PMT Read-outs Developed by IDL, Hawaii (Gary Varner) Fast waveform digitizer for the Photonis MCP is currently under development evolving from existing technology used in BELLE, BESS, ANITA. Length beyond photo-sensor will be ~125 mm. One module per MCP. 10 October 2011 John Learned at ANT 11 in Philadelphia 20

Data Acquisition System (DAQ) Based on c. PCI Format MCP PMT and Digitizer Internet x 1 c. PCI CPU Data processing card (x 3) 3 Gbs fiber link x 8 10 October 2011 x 3 (= Ts) M P 24 c. PCI crate John Learned at ANT 11 in Philadelphia 21

Mini Time Cube: 13 cm 3 Boron Loaded Plastic Scintillator 38 cm MTC with read-out electronics on one face MTC fully populated with read-out electronics Computer and DAQ fits upper case Detector and power supplies in lower case MTC within 2 ft 3 anodized Al enclosure 10 October 2011 Stackable transport cases John Learned at ANT 11 in Philadelphia 22

10 October 2011 The m. TC Being Assembled Gas and RF tight box Rosen design UV laser illumination John Learned at ANT 11 in Philadelphia 23

10 October 2011 Picosecond Calibration Laser System Matsuno/Rosen John Learned at ANT 11 in Philadelphia 24

Estimate Real World Unshielded Noise Rate from Italian CORMORAD Experiment ”A proposal for a high segmented power reactor antineutrino detector”, Marco Battaglieri, July 13~17, 2009, Workshop Towards Neutrino Technologies” 10 October 2011 Refer to CORMORAD talk given by Marco Battaglieri (Genoa) at Trieste 2009. Prototype segmented detector of square logs of NE 110 plastic scintillator, 3 inch PMTs on ends, 40 x 30 cm^3 total volume. No shielding => big background, outside containment building. CORMORAD noise rate near Romanian reactor: => R = ~120 Hz (single) => 2 x R^2 x τ = ~10 Hz (for two hits in time window τ = 330μs) m. TC noise rate implications: ≤ 1/30 vol. x 1/10 time resol. x CORM. rate = few Hz in m. TC Similar numbers from 2 expts at San Onofre (info: LLNL/Sandia group and Juan Collar) » good enough for real-time background analysis & rejection, with no shielding Software rejection needed ~10^5 (< Kam. LAND) John Learned at ANT 11 in Philadelphia 25

Mini-Time. Cube Sensitivity (13 cm^3 cube with 24 MCP's) • Photo sensitive Area 75% coverage of 1, 014 cm^2 • Pixel count 1536 (= 64/pmt * 4 pmt/side * 6 sides) • Typical reactor anti-neutrino several PE/pixel • 100 ps MCP time resolution 2 mm spatial resolution/hit • 2 ns scint decay constant 120 PE/Me. V in first 200 ps. • Vertex recon. with 1 st hits ~1 mm level. • Rate ~10 anti-neutrino events/day (25 m from 3. 3 GW reactor) • Rough cost ~$300 K (includes electronics, mostly MCP tubes) 10 October 2011 much less with future LAPPDs John Learned at ANT 11 in Philadelphia 26

Summary of what we can measure • Neutrino energy, via positron energy • Positron vertex and direction. • Neutron capture point & direction from vertex to capture point. • Enough information to get neutrino direction along a cone and maybe better using full constraints. • Possibly more: total neutron kinetic energy. • Probably too difficult: location of first n scatter. • Conclusion: We can do better than anyone has in the past. How much better remains to be demonstrated. 10 October 2011 John Learned at ANT 11 in Philadelphia 27

m. TC Study and Plans Backgrounds stopping muon, decay processes, random internal/external gamma (from reactor), thermal neutron. . . Solid scintillators boron loaded plastic from Eljen for initial build, not ideal Liquid scintillators short n capture time, best n absorb vertex: 6 Li loading. . . soon Pulse shape discrimination for neutron capture? Can we do anything with neutron elastic scattering? First scatter important but tough. GEANT Simulation of mini-Time Cube in progress…. Understanding tricks of simple kinematics and determining if golden sample identifiable. Careful evaluation of angular resolution with full analysis of waveforms. Time to activation waiting on electronics construction. m. TC operating in Lab early 2012, and take to reactor ASAP thereafter. 10 October 2011 John Learned at ANT 11 in Philadelphia 28

On the Road to Short and Long Range Nuclear Reactor Monitoring and Other Physics 2 liters m. TC • Start with demonstrations of detection of reactors, first close up, then further away. • Importance of directionality understood. • Focus upon resolving positron production and neutron absorption locations. 1 m^3 TC • Start with very small demonstration…. m. TC. • Build ~1 m^3 detector, range to ~200 m • Possible detailed search for short range oscillations • Place a number of modules in shipping container and monitor out to few km 10 October 2011 Truck size 200 m John Learned at ANT 11 in Philadelphia 29

10 October 2011 John Learned at ANT 11 in Philadelphia 30

10 October 2011 Photonis 8 x 8 Multi-channel plate PMT John Learned at ANT 11 in Philadelphia 31

Evaluation MCP Signal 408 nm laser, 100 Photo-Electrons Conclusion: Gain: 40 m. V/100 PE ~ 0. 4 m. V/PE (25 m) at 2100 V 5 m. V/100 PE ~ 50 V/PE (10 m) at 2500 V 10 m longer trailing edge Seems that rise time does NOT depend upon the amplitude => We choose 25μm pore size 10 October 2011 John Learned at ANT 11 in Philadelphia Jean-Francois Genat, ANT Workshop, August 13, 2009 Under development 32

The Photo-Sensor: Photonis XP 85012 (64 channel MCP) 10 October 2011 John Learned at ANT 11 in Philadelphia 33

Impulse Dark Noise vs HV Conclusion: 10 October 2011 At optimum efficiency (25 m 2000 V, 10 m 2400 V), dark counts rates are: 25 Hz (25 m), 20 Hz (10 m) per pixel John Learned at ANT 11 in Philadelphia 34

10 October 2011 Neutron Capture Cross Section John Learned at ANT 11 in Philadelphia 35

Neutrino Vertex Resolution Kam. LAND example: center of ionization for e+ and n capture >10 cm… not useful for present concerns. 120 PE/Me. V on Fermat surface => ~20 mm/sqrt(120*E/Me. V) = 1. 3 mm (2 Me. V anti-neutrino). Neutrino Vertex Resolution => Several mm -> directionality Problem: exponential decay of scintillator. Solution: employ first hits In actuality do full liklihood analysis. Initial promising results from NGA collaborators employing medical imaging like algorithms. 10 October 2011 John Learned at ANT 11 in Philadelphia 36