PROSPECT: Precision Reactor Oscillation and SPECTrum Experiment

PROSPECT: Precision Reactor Oscillation and SPECTrum Experiment Penn Neutrino Seminar November 20, 2015 1

Outline 1. Introduction to neutrinos and neutrino anomalies 2. PROSPECT short baseline reactor experiment 3. PROSPECT Phase I program 4. Concluding remarks L/E Osc./Unosc. 1.02 1.00 0.98 0.96 0.94 0.92 0.90 0.88 0.86 2 Mass Splitting: 1.78 ev ; Osc. Amplitude: 0.09 Phase I (1 yr) Phase I + II (3+3 yr) 0 1 2 3 4 5 6 7 8 9 (m/mev) 2

Outline 1. Introduction to neutrinos and neutrino anomalies 2. PROSPECT short baseline reactor experiment 3. PROSPECT Phase I program 4. Concluding remarks L/E Osc./Unosc. 1.02 1.00 0.98 0.96 0.94 0.92 0.90 0.88 0.86 2 Mass Splitting: 1.78 ev ; Osc. Amplitude: 0.09 Phase I (1 yr) Phase I + II (3+3 yr) 0 1 2 3 4 5 6 7 8 9 (m/mev) 3

Beginning of neutrino physics neutrinos postulated by W. Pauli in 1930 to explain energy conservation in β decay in 1956, Cowan and Reines observed (anti-)neutrinos from a reactor added to the Standard Model as chargeless, massless, weakly interacting leptons that come in 3-flavors study neutrinos from different sources (e.g. reactors, solar, atmospheric, accelerators) 4

Neutrino anomalies solar neutrino anomaly fusion νe atmospheric neutrino anomaly T. Kajita, New J. Phys. 6 (2004) 194 5

lead to discoveries Super-Kamiokande 1998: solved atmospheric anomaly SNO 2001: solved solar anomaly water Cherenkov νμ ντ oscillation through the earth heavy-water Cherenkov sum of all ν matched solar prediction for the discovery of neutrino oscillations which shows that neutrinos have mass 6

Current picture of neutrino oscillation physics Neutrinos undergo quantum mechanical oscillations between flavor and mass states, implying they are massive (although very light) particles. weak states 0 1 e @ A = U µ interact 0 @ mass states 1 2 3 1 A where U = propagate 0 1 U e1 U e2 U e3 @ U 1 U 2 U A 3 U 1 U 2 U 3 PMNS mixing matrix U = atmospheric/accelerator reactor solar 0 @ 1 0 0 1 0 0 c 23 s 23 A @ c 13 0 s 13 e i 0 1 0 0 s 23 c 23 s 13 e i 0 c 13 1 A 0 @ c 1 12 s 12 0 s 12 c 12 0A 0 0 1 where c ij = cos( ij ) s ij = sin( ij ) = CP phase 7

Oscillations - two neutrino approximation P!, 6= = sin 2 (2 )sin 2 m 2 L 4E where, = e, µ, source (Eν) detector L Parameters θ (mixing angle - amplitude) and Δm 2 (mass splitting - frequency) are defined by nature. We can target specific Δm 2 measurements by designing our experiments to have a certain L/E. 8

Precision reactor antineutrino experiments Kilometer baseline θ13 precision experiments: Daya Bay, Double Chooz, RENO Daya Bay antineutrino detectors Daya Bay site m P ( e! e )=1 sin 2 (2 )sin 2 2 L 4E survival probability fraction anti-νe detect inverse beta decay: anti-νe + p e + + n reactor core near detector far detector Flux measured between detectors for relative oscillation measurement. Can also be compared to predictions from reactor models. 9

Reactor flux anomaly relative counting measurement between near and far shows deficit in anti-νe flux indicates oscillations to other SM neutrino flavors over kilometer baselines consistent with solar, atmospheric, and accelerator data Data / Prediction 1.2 1 0.8 0.6 10 anti-νe disappearance 2 10 Distance (m) arxiv:1508.04233 (Daya Bay) Previous data Daya Bay Global average 1-σ Experiments Unc. 1-σ Model Unc. 3 10 near and far detectors observe overall flux deficit when compared to 2011 prediction new measurements agree with old reactor experiments with better precision reactor global flux deficit ~5% 10

Reactor flux anomaly relative measurement between near and far shows deficit in anti-νe flux indicates oscillations to other SM neutrino flavors over kilometer baselines consistent with solar, atmospheric, and accelerator data LEP Z boson width measurements only 3 active neutrinos allowed Data / Prediction 1.2 1 0.8 0.6 10 anti-νe disappearance 2 10 Distance (m) arxiv:1508.04233 (Daya Bay) Previous data Daya Bay Global average 1-σ Experiments Unc. 1-σ Model Unc. 3 10 near and far detectors observe overall flux deficit when compared to 2011 prediction new measurements agree with old reactor experiments with better precision reactor global flux deficit ~5% 11

Flux hypothesis - sterile neutrino oscillations Δm14 2 > 1 ev 2 Δm23 2 =2.3x10-3 ev 2 Δm12 2 =8x10-5 ev 2 Blue shows fit to reactor data with a 3+1 neutrino model sin 2 2θ14 > 10-3 arxiv:11204.5379 (Light sterile neutrinos white paper) 12

Flux hypothesis - sterile neutrino oscillations high frequency oscillations Δm14 2 > 1 ev 2 Δm23 2 =2.3x10-3 ev 2 Δm12 2 =8x10-5 ev 2 Blue shows fit to reactor data with a 3+1 neutrino model need short baseline data!! sin 2 2θ14 > 10-3 arxiv:11204.5379 (Light sterile neutrinos white paper) 13

ev-scale sterile neutrinos and other anomalies LSND decay at rest MiniBooNE short baseline accelerator GALLEX/SAGE Ga source calibration anti-νe appearance low energy νe appearance νe disappearance 14

Impacts of sterile neutrinos Of course, the addition of a each sterile state adds a set of new parameters to the mixing framework. What is the practical effect? DUNE response to steriles An ev-scale sterile would impact: long-baseline experiments measuring CP violation neutrinoless double beta decay observing Majorana neutrinos 0νββ response to steriles Giunti & Zavanin, JHEP 1507 (2015) Steriles indicate new physics and will have a profound effect on future experiments. We need definitive short-baseline experiments that don t rely on predictions! 15

Reactor spectrum anomaly Entries / 250 kev Ratio to Prediction Entries / 250 kev (Huber + Mueller) Ratio χ 2 contribution to Prediction (Huber ( χ + ) Mueller) contribution ( χ ) χ 2 20000 15000 10000 20000 5000 15000 1.2 10000 1.1 1 5000 0.9 i i 0.8 1.2 4 1.1 2 1 0 0.9 2 0.8 4 4 2 0 Data Full uncertainty Antineutrino energy Reactor spectra uncertainty from Daya Bay and Double Chooz. ILL+Vogel Data Full uncertainty Integrated Reactor uncertainty ILL+Vogel 2 Prompt Positron 4 Energy (MeV) 6 8 Integrated 2 Prompt 4Energy (MeV) 6 8 arxiv:1508.04233 (Daya Bay) 2 4 6 8 Prompt Energy (MeV) 10 1 10 10 10 6 2 3 4 Local p-value (1 MeV windows) Local p-value (1 MeV windows) 2 10 5 4 10% excess of events in the 4-6 MeV 10 region 6 when compared to reactor models, 2 4 6 8 10 Prompt Energy (MeV) also known as the bump. 1 1 10 10 10 10 10 2 3 4 5 16

Spectrum hypothesis - deficiencies in models power reactor fuels composed of 235 U, 238 U, 239 Pu and 241 Pu arxiv:1503.01059 total emitted spectrum is an admixture, thousands beta branches Two major approaches to calculate spectrum: 1. Ab-initio calculate spectrum branch-by branch using beta branch databases 2. Beta conversion measure beta spectrum of main isotopes fit with virtual branches and converted to antineutrino spectra Dwyer & Langford, PRL 114 012502 (2015) Both methods have complications and difficulties. 17

Outline 1. Introduction to neutrinos and neutrino anomalies 2. PROSPECT short baseline reactor experiment 3. PROSPECT Phase I program 4. Concluding remarks L/E Osc./Unosc. 1.02 1.00 0.98 0.96 0.94 0.92 0.90 0.88 0.86 2 Mass Splitting: 1.78 ev ; Osc. Amplitude: 0.09 Phase I (1 yr) Phase I + II (3+3 yr) 0 1 2 3 4 5 6 7 8 9 (m/mev) 18

Precision Reactor Oscillation and SPECTrum experiment Physics objectives: 1. Search for short-baseline oscillation at distances <10 m 2. Perform a precision measurement of 235 U reactor anti-νe spectrum Phase near detector I ~ 7m Pb shield wall Phase I 7-11m near detector ~ 7m reactor core baselines 7-11m far detector ~ 18m reactor core HFIR core Phase II 16-20 m HFIR core phase 1 phase 1+ phase 2 Münster, April 10, 2014 HFIR, ORNL Requirements: energy resolution of 4.5%/ E (σ/e) for spectral measurement good position resolution for comparing spectra between baselines excellent background rejection capabilities at near-surface, reactor site 19

Detection mechanism - lithium liquid scintillator prompt (t0 = 0): 1-10MeV annihilation β + typical reactor spectrum νe p n 6 Li α Ev Eβ + 1.8MeV inverse beta decay (IBD) t d ~100μm delay (tc = 40μs): 0.6MeVee (n, 6 Li) capture coincidence of these two signals indicates an IBD event 20

PROSPECT site - High Flux Isotope Reactor (HFIR) HFIR Core power: 85MW (research) fuel: highly enriched uranium ( 235 U) core shape: cylindrical size: h=0.5m r=0.2m (compact) duty-cycle: 41% baselines: 7-11m, 16-20m 5m PHASE I PROSPECT AD PROSPECT baselines Phase II PROSPECT AD 21

Background challenges and measurements Experimental set-up: near-surface, short-baseline liquid scintillator (LS) detectors Correlated Backgrounds: cosmogenic fast neutrons multiple neutron captures Cosmogenic fast neutrons NIM A A806 (2016) Unshielded gammas (near) Uncorrelated Backgrounds (accidentals): reactor-related gammas gammas from internal backgrounds ( 232 Th, 40 K) NIM A A806 (2016) 22

Outline 1. Introduction to neutrinos and neutrino anomalies 2. PROSPECT short baseline reactor experiment 3. PROSPECT Phase I program 4. Concluding remarks L/E Osc./Unosc. 1.02 1.00 0.98 0.96 0.94 0.92 0.90 0.88 0.86 2 Mass Splitting: 1.78 ev ; Osc. Amplitude: 0.09 Phase I (1 yr) Phase I + II (3+3 yr) 0 1 2 3 4 5 6 7 8 9 (m/mev) 23

Short-baseline, near-surface detection techniques 1. Segmented detector relative oscillations between segments several pure L/E measurements independent of predictions 2. Novel 6 Li liquid scintillator distinct (α, triton) signature highly localized, time and space separate from reactor γ energy good energy resolution excellent pulse-shape discrimination safe, non-flammable 3. Shielding design localized Pb reactor wall neutron shielding package 4. Compact, HEU reactor point source of neutrinos one fuel isotope Osc./Unosc. 1.02 1.00 0.98 0.96 0.94 0.92 0.90 0.88 0.86 2 Mass Splitting: 1.78 ev ; Osc. Amplitude: 0.09 Phase I (1 yr) Phase I + II (3+3 yr) 0 1 2 3 4 5 6 7 8 9 L/E (m/mev) (n, Li) event 24

PROSPECT Phase I detector High Flux Isotope Reactor (ORNL) Phase 1 near detector: 3 tons 6 Li-loaded EJ-309 liquid scintillator 10x12 segmented array low-mass optical separators segment size: 120x15x15 cm 3 double-ended PMT readout movable, baseline coverage 7-11m Phase I lead wall reactor core Physics goals: 1. probe sterile ν parameter space at 3σ in 1 calendar year 2. precision measurement of 235 U neutrino spectrum ~100cm active length 6 Li-doped LS Light Guide 5 PMT 25

Phased detector development approach PROSPECT-0.1 Characterize LS Aug 2014 Spring 2015 5cm 0.1liter LS, 6 LiLS PROSPECT-2 Background studies Winter 2014-15 Aug 2015 PROSPECT-20 Characterize segment Spring-Summer 2015 PROSPECT-60 Mechanical prototype Late 2015* 12.5cm 1.7 liter 6 LiLS 1m 23 liter LS, 6 LiLS 1m 60 liter (1x2) 6 LiLS segments PROSPECT-2k Physics measurement Late 2016* 1m 3 tons 6 LiLS segments *technically driven schedule 26

Lithium-loaded liquid scintillator (LiLS) development Novel scintillator cocktail: PSD LiLS that is non-toxic, non-flammable extensive studies with LAB, Ultima Gold EJ-309 gave best light yield, PSD Scintillator specs (PROSPECT-0.1): Light YieldEJ-309 = 11500 ph/mev Light YieldLiLS, measured = 8200 ph/mev prominent neutron capture peak in LiLS PSD FOM at (n, Li) is 1.79 energy resolution (σ/e) of 5.2% at 0.6MeVee (n,li) Cf-252 Co-60 developed novel LiLS with excellent light yield, PSD, and neutron capture capabilities 27

Pulse-shape discrimination Can take advantage of how different particles deposit energy in scintillator using pulse-shape discrimination (PSD). Gives particle identification information. (n,li) Cf-252 PSD = Q tail Q full n Qfull Qtail γ arxiv1309.7647 PROSPECT-2 (LiLS) particle classification: light particles = gamma-like, heavy charged = neutron-like 28

Pulse-shape discrimination Can take advantage of how different particles deposit energy in scintillator using pulse-shape discrimination (PSD). Gives particle identification information. energy+psd cuts for prompt and delay signals PSD = Q tail Q full Qfull Qtail 10MeV PROSPECT-2 (LiLS) arxiv1309.7647 particle classification: light particles = gamma-like, heavy charged = neutron-like 29

Tackling backgrounds with PSD prompt signal: 1-10 MeV positron from inverse beta decay delay signal: 0.6 MeV signal from neutron capture on 6 Li IBD-like PROSPECT-2 (LiLS) fast neutron Signal Background PSD signatures inverse beta decay γ-like prompt, n-like delay fast neutron n-like prompt, n-like delay accidental gamma γ-like prompt, γ-like delay accidental gammas particle ID strongly suppresses cosmogenic correlated and reactor-induced uncorrelated backgrounds 30

Tackling backgrounds with detector design the neutron capture on 6 Li allows for event localization, and combined with the localized e + gives a spatial correlation in addition to the IBD temporal correlation easy fiducialization to control gamma backgrounds designed localized shielding to suppress cosmogenic and reactor correlated backgrounds cosmogenic neutron backgrounds can be subtracted with reactor-off data sets IBD in 6 Li e + n effect of increasing lead wall on gammas (1)! (2)! Counts/MeV/s 3 10 10 2 Config. (1) Config. (2) Config. (3) Config. (4) neutron shielding (3)! 10 (4)! 1 NIM A A806 (2016) 0 1 2 3 4 5 6 7 8 Energy (MeV) detector structure and passive shielding designed for near-surface backgrounds 31

Compatibility and design of low-mass separators Compatibility: extensive material compatibility testing required to ensure long-term LS performance focus on materials proven in recent experiments - PTFE, acrylic, polypropylene, long-term mechanical stability verified Separators: physics goals demand low inactive mass, high reflectivity, and long-term compatibility developed multi-layer system meeting all requirements fabrication procedures for full-scale system under validation Events/IBD/15 p.e. 10 6 Low-mass reflector prototypes No inactive mass 5 4 3 3 e + wall study PROSPECT AD Bugey 3 2 1mm thick 1 0 0 500 1000 1500 2000 2500 Photoelectrons produced robust low-mass separators from LS-compatible materials 32

PROSPECT-20 segment studies 522 + 16 PE/MeV (n, Li) n-like γ-like low energy (0.5-0.7MeV) : 99.99% rejection of γ 99% acceptance n events LiLS with realistic geometry, above target light collection goal with excellent PSD 33

Simulation to benchmark prototype data Simulations have been developed to meet distinct needs: Phase 1 cosmic neutron event 1. Segment design simulations further used to optimize light transport, shielding etc in context of science goals 2. Background mitigation develop single flexible Monte Carlo, benchmark against prototypes, enable extrapolation to full detector 3. Detector response detailed model of detector response ensures PROSPECT has precision spectral measurement capability Representative 500 MeV primary Winston cones simulation of PROSPECT-20 with optical transport prototyping program has enabled validation of mechanics, detector response, and simulation models 34

Simulation to benchmark prototype data Simulations have been developed to meet distinct needs: power of fiducialization 1. Segment design simulations further used to optimize light transport, shielding etc in context of science goals 2. Background mitigation develop single flexible Monte Carlo, benchmark against prototypes, enable extrapolation to full detector 3. Detector response detailed model of detector response ensures PROSPECT has precision spectral measurement capability Representative 500 MeV primary Winston cones simulation of PROSPECT-20 with optical transport prototyping program has enabled validation of mechanics, detector response, and simulation models 35

Validation of Monte Carlo from HFIR data PROSPECT-20 IBD-like prompt cosmics PROSPECT-20 at HFIR Phase I simulated signal + cosmics after cuts PROSPECT-20 measured cosmic backgrounds during reactor-off PROSPECT-20 simple Monte Carlo agrees reasonably with data confident in extrapolating MC to Phase I detector after series of effective cuts, can reach S:B > 3:1 surpasses physics goals target will measure these backgrounds during reactor-off time in Phase I prototype deployment validated background MC to project Phase I S:B 36

PROSPECT Phase I physics reach Sterile neutrino search Sensitivity parameters: fiducialized volume: 8x10 segments segment fiducialization: 90x15x15 cm 3 S:B: 3:1 for nearest position + Phase II energy resolution: 4.5%/ E position resolution: 15cm Spectral measurement 37

What we are currently working on 1. PROSPECT-60 1x2 LiLS array Mechanical prototype to validate detector components and test operation of subsystems. corner rod piece separators 1 segment with corner rods 2. LS nonlinearity measurements PMT housing + light guide HPGe with Co-60 and Bi-207 14000 Co 1.173 6us Study non-linearities in scintillator at low energies using mono-energetic electrons using Compton coincidence spectrometer. counts 12000 10000 8000 6000 4000 2000 Bi.570 Bi 1.064 Co 1.333 Bi 1.77 0 0 500 1000 1500 2000 2500 3000 3500 integral (ADC channel) 38

Outline 1. Introduction to neutrinos and neutrino anomalies 2. PROSPECT short baseline reactor experiment 3. PROSPECT Phase I program 4. Concluding remarks L/E Osc./Unosc. 1.02 1.00 0.98 0.96 0.94 0.92 0.90 0.88 0.86 2 Mass Splitting: 1.78 ev ; Osc. Amplitude: 0.09 Phase I (1 yr) Phase I + II (3+3 yr) 0 1 2 3 4 5 6 7 8 9 (m/mev) 39

Concluding statements 1. Anomalies have led to great discoveries in the history of neutrino physics. 2. The reactor anomaly entails both a flux deficit hinting at sterile neutrinos and a spectral deviation, the PROSPECT experiment can address both. 3. The PROSPECT R&D program has: successfully deployed multiple prototype detectors developed a detailed understanding of near-surface backgrounds at HFIR developed technology required for the Phase I detector produced simulation models validated against prototype data 4. PROSPECT Phase I is ready to proceed with precision 235 U spectrum measurement and cover the sterile global best fit in 1 calendar year at 3σ 40

The PROSPECT Collaboration 4 national laboratories 9 universities 63 collaborators prospect.yale.edu 41

Anomalies lead to discoveries solar anomaly SNO 2001 detected all flavors total matches solar prediction atmospheric anomaly Super-Kamiokande 1998 directional info νμ ντ oscillation T. Kajita, New J. Phys. 6 (2004) 194 43

Where to look for sterile neutrinos? [ev 2 ] 2 14 m 10 Reactor anomaly Allowed Regions: SBL Anomaly (Kopp), 95% CL All ν e Disappearance Exps (Kopp), 95% CL Gallium Anomaly (Kopp), 95% CL SBL + Gallium Anomaly (RAA), 95% CL Daya Bay Exclusion, 95% CL 1 10-1 sterile global fit νe, anti-νe disappearance -2-1 10 10 sin 2 1 2θ 14 44

PROSPECT: Hands-on science DPF August 2015: 5 August 2015 45 45

Back-up: PROSPECT-2 at HFIR Detector geometry: 1.7L cylinder Scintillator: Li-loaded EJ-309 PMTs: 5 flat ET9823 Shielding: poly, Pb, Bpoly Reflectors: diffuse Gore DAQ: CAEN 1720 (12bit) Purpose: background reduction method DPF August 2015: 5 August 2015 46

PROSPECT-2 at HFIR (n,li) arxiv1506.03547 Coincidence analysis: cosmogenic fast neutrons (real) cosmogenic showers (multiple captures) reactor-related gammas (accidental) PSD cuts on prompt and delayed signals rejects many of these backgrounds. DPF August 2015: 5 August 2015 47

Back-up: PROSPECT-20 at HFIR Detector geometry: 23L 1-meter rectangle Scintillator: Li-loaded EJ-309 PMTs: 5 flat ET9823 Shielding: poly, Pb, Bpoly, water bricks Reflectors: 3M SolarMirror DAQ: CAEN 1720 (12bit) Purpose: Operate full PROSPECT segment DPF August 2015: 5 August 2015 48

PROSPECT-20 at HFIR (n,li) Accidentals reduced significantly with energy and PSD cuts. DPF August 2015: 5 August 2015 49

Back-up: PROSPECT-20 at Yale Optics optimization studies: Reflector type Reflector coupling PMT read-out Compare to simulation Soon to come: Optical coupler geometries Li-loaded EJ-309 PMT Optical coupler 100cm EJ-309 15.2cm Reflectors Detector geometry: 23L 1-meter rectangle Scintillator: EJ-309 PMT(s): 5 spherical Hamamatsu R6594 Shielding: Pb Reflectors: variable DAQ: CAEN 1730 (14bit) Purpose: optimize optics of full segment DPF August 2015: 5 August 2015 50

Segment response: light collection and PSD studies Bi-207 PMT Optical coupler 100cm JINST 10 P11004 (2015) EJ-309 15.2cm Reflectors PROSPECT-20atYale (EJ-309): optimize collection, PSD with air-coupled specular external separators average light collection: 527+ 10 photoelectrons/mev low energy PSD (0.5-0.7MeV) allows for 99.99% rejection of γ, 99% acceptance n events can improve geometry using internal separators Cf-252 excellent PSD is obtained in realistic geometry at target light collection of 500pe/MeV 51

Segment response: double-ended readout light collection measurements Bi-207 position reconstruction (timing) pulse-shape discrimination 5cm resolution Cf-252 JINST 10 P11004 (2015) double-ended readout allows for uniform optical collection, enhanced PSD, and axial position resolution 52

Lithium dopant in liquid scintillator 1. Small detectors that do not have full calorimetry information. But, neutron capture on 6 Li allows for single-site topology. α (n, Li) event n 6 Li t d ~100μm 2. PROSPECT will be in a high gamma environment, with energies ranging from 1-10MeV. This background will not interfere with neutron captures since (n, Li) events fall in the n-like pulse shape discriminate (PSD) band. Can contain (n, Li) events in segments and extract from backgrounds. DPF August 2015: 5 August 2015 53

Validation of MC from prototypes at HFIR site PROSPECT-2 at HFIR combined exposure PROSPECT-2 PROSPECT-20 PROSPECT-20 at HFIR 54