Results and Prospectives of Reactor Neutrino Experiments

Liangjian Wen

The 12th particle physics phenomenology workshop (PPP12) 16-19 May, NCTU, Hsinchu, Taiwan

Outline• Reactor Neutrinos• Oscillation measurement: θ13 and Δm2

ee– Daya Bay, Double Chooz, RENO

• Sterile Neutrino Search• Rate and spectrum anomaly• Future: determining Neutrino Mass Ordering

– JUNO, RENO-50

• Summary

2

What we have learned?

3

𝑉𝑉 =1 0 00 𝑐𝑐23 𝑠𝑠230 −𝑠𝑠23 𝑐𝑐23

𝑐𝑐13 0 𝑠𝑠13𝑒𝑒−𝑖𝑖𝛿𝛿0 1 0

−𝑠𝑠13𝑒𝑒𝑖𝑖𝛿𝛿 0 𝑐𝑐13

𝑐𝑐12 𝑠𝑠12 0−𝑠𝑠12 𝑐𝑐12 0

0 0 1

𝑒𝑒𝑖𝑖𝜌𝜌 0 00 𝑒𝑒𝑖𝑖𝜎𝜎 00 0 1

Standard Parametrization of the PMNS Matrix

𝜽𝜽𝟐𝟐𝟐𝟐 ~ 𝟒𝟒𝟒𝟒∘

Atmospheric,LBL accelerator

𝜽𝜽𝟏𝟏𝟐𝟐 ~ 𝟗𝟗∘

Reactor,LBL accelerator

𝜽𝜽𝟏𝟏𝟐𝟐 ~ 𝟐𝟐𝟒𝟒∘

Solar,KamLAND

0ν2β, LNV?

Quarks vs. Leptons: A big puzzle of fermion flavor mixings

|𝑈𝑈| = |𝑉𝑉| =

CKM PMNS

Hierarchy! Approximate μ-τ symmetry?

|𝚫𝚫𝒎𝒎𝟐𝟐𝟐𝟐𝟐𝟐 | ~ 𝟐𝟐.𝟒𝟒 × 𝟏𝟏𝟏𝟏−𝟐𝟐 eV𝟐𝟐 𝜹𝜹 ~ ? 𝚫𝚫𝒎𝒎𝟐𝟐𝟏𝟏

𝟐𝟐 ~ 𝟖𝟖× 𝟏𝟏𝟏𝟏−𝟒𝟒 eV𝟐𝟐

• Discovery of neutrino in 1956• Early search for oscillation 70’s-80’s• Small θ13 in 1990s• limit on neutrino magnetic moment (00’s)• Observation of reactor 𝝂𝝂𝒆𝒆 disappearance in 2003• Discovery of non-zero θ13 in 2012• Mass hierarchy and precision measurements• Sterile neutrinos, Magnetic moment, …

Reactor Neutrinos

4

5

Neutrino spectra of fission isotopes, ILL

Isotope evolution

fissio

n fra

ctio

n (%

)

X

• Neutrino flux of a commercial reactor with 3 GWth : ~6×1020 ν/s

• Distinguishing correlated and un-correlated errors is important

Reactor ν Flux

Capture on H

Capture on Gd

0.1% Gd

nepe +→+ +ν

10-40 keV

++ +−++≅ epnne mMMTTE )(ν

1.8 MeV: Threshold

• ν-e scattering• Inverse-β reaction (IBD)

Reactor ν Detection

6

A little history of νe disappearance …

7

PRD 62 (2000) 072002

PRL 90, 021802 (2003)

Precision of early Reactor experiments (3-6%):• Reactor power：~1%• ν spectrum：~0.3%• Fission rate： ~ 2%• Target mass：~1-2%• Backgrounds：~1-3%• Efficiency：~2-3%

Near-far relative measurement was proposed (Mikaelyan and Sinev, hep-ex/9908047) to reduce the uncertainties from reactor and detector

Daya Bay, Double Chooz, RENO

8

Daya Bay Double Chooz

RENO

Daya BayRENO

Double Chooz

40t

40t80t

365 m

490 m17.4 GWth

8.5 GWth

8t

8t290 m

1380 m

16t 16t

16.8 GWth

Daya Bay: Best Site for θ13

• Powerful reactor complex (Top 5)• Close to mountains enough shielding • Luminosity 5-20 times of DC and RENO• Featured design side-by-side

calibration (2-4 ADs at each site) actual relative det. error 0.13% /√N,

• Discovered an unexpectedly large θ13 in Mar. 2012. precision measurement afterwards Huber et al. JHEP 0911:044, 2009

9

Daya Bay

RENO

DC

Designs Luminosity(ton⋅GW)

Detector Systematics

Overburden (near/far, mwe)

Sensitivity(3y, 90%CL)

Daya Bay 1400 0.38%/√N 250 / 860 ~ 0.008Double Chooz 70 0.6% 120 / 300 ~ 0.03

RENO 260 0.5% 120 / 450 ~ 0.02

Daya Bay Experiment

10

Aug. 2011

Nov. 2011

Dec, 2012

Full detectors:Aug, 2012

Water pool

The Daya Bay Detectors

11

• Multiple AD modules at each site to check Uncorr. Syst. Err. – Far: 4 modules，near: 2 modules

• Multiple muon detectors to reduce veto eff. uncertainties– Water Cherenkov： 2 layers – RPC： 4 layers at the top + telescopes Redundancy !!!

40 t MO

20 t LS20 t Target

reflector

reflector

Automated Calibration Units (ACU)

Data Periods

12

EH1 EH2 EH3

2017/02, EH1-AD1 offline

θ13, ∆𝑚𝑚𝑒𝑒𝑒𝑒2 [PRD]θ13, ∆𝑚𝑚𝑒𝑒𝑒𝑒

2 [PRL]nH θ13 [PRD]sterile ν [PRL, PRL]reactor [CPC]wave packet [EPJC*]

θ13, ∆𝑚𝑚𝑒𝑒𝑒𝑒2 [PRL]

nH θ13 [PRD(R)]sterile ν [PRL]reactor [PRL]

θ13 [PRL]

θ13 [CPC]

*submitted

55 d

ays

139

days

217

days

621

days

1230

day

s

8-AD Data Taking6-AD 2012 Summer

reactor evolution [PRL*]

40 t MO

20 t LS20 t Target

reflector

reflector

Automated Calibration Units (ACU)

Energy Calibration

13

The relative energy scale uncertainty <0.2%

PMT gain calibration

(dark noise, LED)

Energy Rec.

• calibration Sources• spallation neutrons

Relative Energy Scale

• 68Ge, 60Co, 241Am-13C• Neutrons (IBD, spallation)• Special sources• Natural radioactivity

Detector energy response model

14

• Energy model: fit to various gamma lines and 12B beta-decay spectrum

• Validated with– 208Th, 212Bi, 214Bi beta-decay spectrum; Michel electron

– Bench tests of Compton scattering electrons in LS

• Non-linear energy response in Liquid scintillator– Quenching (known as Birks’ law) and Cerenkov (particle-, E- dep.)

– Electronics (E- dep, modeled based on MC and single channel FADC measurement)

gamma electron positron

Uncertainty <1%

Experience on DYB energy model is valuable to JUNO (similar LS).

Signal & Backgrounds

15nGd Background (Far hall)n-H Background (Far hall)

• Accidental background• 9Li/8He• Fast neutrons• Am-C neutron calib. sources• 13C(α,n)16O

AD

IWS

RPC Array

Telescope RPC

RPC-only taggedOWS tagged

Side-By-Side Comparison• Cross check with multiple detectors at the same site• The relative detection efficiency uncertainty down to 0.13%,

verified by comparing the rates of detectors

16

χ2/NDF = 232.6/263

Precision Oscillation Measurement

17

1230 days

Sin22θ13 = [8.41 ± 0.33] × 10-2

NH: Δm232 = [2.45 ± 0.08] × 10-3 eV2

IH: Δm232 = [-2.55 ± 0.08] × 10-3 eV2

1230 days• Independent sin22θ13 meas. from nH• run until 2020, achieve uncertainty ≤3%

Global Comparison

18

NH

Most precise measurement• Sin22θ13 uncertainty: 3.9%• |Δm2

32| uncertainty: 3.4%

Consistent results with reactor and accelerator experiments.

Independent n-H analysis

19

The n-H analysis, where the neutron captures on hydrogen, is statistically and (largely) systematically independent from the nGd one.One of the challenges: large accidental background

Phys. Rev. D 93, 072011 (2016)Rate analysis: sin22θ13 = 0.071±0.11, χ2/ndf = 6.3/6

Future Sensitivity• Daya Bay:

– Δ(sin22θ13) ~ 0.003 ~3%– Δ(Δm2

ee) ~ 0.07 ~ 3%

• RENO: ~5%. • Double Chooz: ~10%

20

Daya Bay: operation till 2020RENO: “operation funding secured until Feb. 2019”Double Chooz: “secured to Jan. 2018 (may change)”

by J. Zhao

Reactor Anomaly• ILL spectra agree with data• Mueller/Huber spectra higher than data• Sterile neutrino?

21

G. Mention et al. Phys.Rev. D83 (2011) 073006

fit with sterile 𝜈𝜈Δ𝑚𝑚2 ≈ 1 𝑒𝑒𝑉𝑉2

Absolute Reactor νe flux

22

Discrepancies to the Huber+Mueller model indicate: Over estimated flux and/or underestimated flux uncertainty Or the existence of a sterile neutrino

Consistent flux measurement with previous short baseline reactor experiments

PRL 116, 061801 (2016)

• Daya Bay measurement of absolute Flux– Data/(Huber+Mueller)：0.946±0.020– Data/(ILL+Vogel)：0.992 ± 0.021

621 days data

CPC Vol. 41, No. 1 (2017) 013002

Future Reactor Exp. for Sterile Neutrino

23

• Different technologies: (Gd, Li, B) (seg.)(movable)(2 det.)• Most have sensitivity 0.02~0.03 @∆m~1eV2 @90%CL

Talk by Nathaniel Bowden @NEUTRINO2016

Search for light sterile neutrinos

24

• An unique opportunity for sterile neutrino searches– Sterile neutrino would introduce additional oscillation mode– Relative meas. at multiple baselines: EH1 (~350m), EH2 (~500m), EH3 (~1600m)

• Oscillation analysis– No significant signal observed, consistent with 3-flavor neutrino oscillation. – Set most stringent limit at 10-3 eV2 < Δm2

41 < 0.1 eV2

PRL 113, 141802 (2014)

Excluded

MINOS, Daya Bay and Bugey-3

25

Parameter space allowed by LSND and MiniBooNE is excludedPRL. 117, 151801 (2016)

5 MeV Bump on Reactor Spectrum

26

• Events are reactor power related & time independent

• Events are IBD-like:– Disfavors unexpected backgrounds

• No effect to θ13 if near-far meas.• Possibly due to forbidden decays

(PRL112: 2021501; PRL114:012502)

RENO, Neutrino 2014

DC, Neutrino 2014

Jetter, Tau2014

Reactor Antineutrino Spectrum• Bump local significance ~ 4.4 σ• Unfolding the reactor neutrino spectrum

27Chin. Phys. C, 41(1) 13002-013002 (2017) PRL 116, 061801 (2016)

Evolution of reactor neutrino flux and spectrum

28

Probe the reactor anomaly

1230 days

Rescaled model (equal deficit in the IBD yields from four fission isotopes)

arXiv:1704.01082

Evolution of reactor neutrino flux and spectrum

29

• Combined fit for major fission isotopes 235U and 239Pu

• σ235 is 7.8% lower than Huber-Mueller model (2.7% meas. uncertainty)

• σ239 is consistent with the prediction (6% meas. uncertainty)

• 2.8σ disfavor equal deficit (H-M model & sterile hypothesis)

Exploring ν mass ordering with Reactors

30

Phys.Rev.D78:111103,2008

• Large θ13 open doors to MH– Exploit L/E spectrum with reactorsPrecision energy spectrum measurementLook for interference between solar- and atmospheric- oscillations relative measurement

S.T. Petcov et al., PLB533(2002)94S.Choubey et al., PRD68(2003)113006J. Learned et al., PRD78, 071302 (2008)L. Zhan, Y. Wang, J. Cao, L. Wen, PRD78:111103, 2008, PRD79:073007, 2009J. Learned et al., arXiv:0810.2580…

Independent on CP phase and θ23 (Acc. & Atm. do)Energy Resolution is the key

JUNO and RENO-50

31

Hong KongJUNO

Daya Bay

RENO-50

RENO

18 kton LS Detector ~47 km from YG reactors

Mt. Guemseong (450 m) ~900 m.w.e. overburden

20 kton LS Detector

~53 km from Taishan & Yangjiang reactors ~750 m rock overburden

J. Phys. G43:030401 (2016)(arXiv: 1507.05613)

∆M223

6 years,3σ

Determine NMO at JUNO

32

L. Zhan, Y. Wang, J. Cao, L. Wen, PRD78:111103, 2008, PRD79:073007, 2009

Detector size: 20kt Energy resolution: 3%/√EThermal power: 36 GWBaseline 58 km

3%/ 𝐸𝐸

17.4 GW18.4 GW

17.4 GW

33

Cosmic muons~ 250k/day

Atmospheric νseveral/day

Geo-neutrinos1-2/day

Solar ν(10s-1000s)/day

reactor ν, ~ 60/day

700 m

Supernova ν~ 5k in 10s for 10kpc

36 GW, 53 km

0.003 Hz/m2

215 GeV10% multiple-muon

Neutrino Rates

Δχ2 Relative Meas.

(a)Use absolute ∆mμμ

2

Ideal case 4 5(b)Realistic case 3 4

Sensitivity on NMO

Y.F Li et alPRD 88, 013008 (2013)

JUNO MH sensitivity with 6 years' data:

(a) If accelerator experiments, e.g NOvA, T2K, can measure ∆M2

µµ to ~1% level(b) Take into account multiple reactor cores, uncertainties from energy non-linearity, etc

Ideal Core distr. DYB & HZ Shape B/S (stat.) B/S (shape) |∆m2µµ|

Size 52.5 km Real Real 1% 6.3% 0.4% 1%

∆χ2ΜΗ +16 - 3 -1 - 1 - 0.6 - 0.1 + (4-12)

34

Precision Measurement

0.16%0.24% 0.39%0.54%0.16%0.27%E resolution

Correlation among parameters

Statistics+BG, +1% b2b+1% EScale , +1% EnonL

sin2 θ12 0.54% 0.67%

Δm221 0.24% 0.59%

Δm2ee 0.27% 0.44%

Probing the unitarity of UPMNS to ~1%, more precise than CKM matrix elements!

35

Current precision

J. Phys. G43:030401 (2016)

Matter Effects

36

• With six years of running, the Δχ2 of mass ordering measurements will reduce from 10.28 (vacuum) to 9.64 (matter).

• The reduction of Δχ2 is comparable with other systematic error.

Chinese Physics C 40(9) (2016) 091001

Supernova neutrinos at JUNO

37

no osc.complete

Measure energy spectra & fluxes of almost all types of neutrinos

Correlated events. Better detection in LSthan in Water

Typical galactic SN assumptions: 10 kpc galactic distance, 3×1053 erg, Lν the same for all types

SN researches at JUNO• NMO sensitivity• Absolute v mass• SN direction• flux/spectra meas. & recon.• time evolution• test various SN models• …

J. Phys. G43:030401 (2016)

Supernova neutrinos

38

• Test average-energy hierarchy of SN ν, and how total energy is partitioned among ν flavors

• ν mass: < 0.83±0.24 eV at 95% CL (JCAP 05 (2015) 044)• Locating the SN: ~9°

PRD 94, 023006 (2016)

Spectra

Precision of 1%

Precision of 10%

Supernova neutrinos• Reconstruct the flux & spectra of all flavor neutrinos from the

observed ν-e, ν-p and IBD interactions, by using the SVD-unfolding technique

• Toy MC including E-resolution, quenching, E-threshold, etc• Verify the method with different SN models model

independent

39H. Li et al, paper to appear soon

Supernova neutrinos

40

K-C. Lai et al, JCAP 1607 (2016) no.07, 039

Probing neutrino mass ordering with SN νs

Diffuse Supernova Neutrino Background

• DSNB: Past core-collapse events– Cosmic star-formation rate– Core-collapse neutrino spectrum– Rate of failed SNe

10 Years’ sensitivity

34J. Phys. G43:030401 (2016)

Other Physics Topics

42

• JUNO will have 1-2 σ NMO sensitivity with atmospheric neutrinos– Measure both lepton and hadron energy– Good tracking and energy resolution

• Geo-v measurement at JUNO– Huge reactor neutrino backgrounds– Need accurate reactor spectra

– 20 x statistics than previous meas.KamLAND: 30±7 TNU (PRD 88 (2013) 033001)

Borexino: 38.8±12.2 TNU (PLB 722 (2013) 295)

Best fit 3 y 5 y 10 y

U+Thfix ratio

0.96 10% 8% 6%

U (free) 1.03 19% 15% 11%

Th (free) 0.80 37% 30% 21%

Combined shape fit of geo-ν and reactor-ν

Others (sterile neutrinos, proton decay, neutrinos from dark matter… etc)

J. Phys. G43:030401 (2016)

Central detector

Water Cherenkov

Top Tracker

Calibration

Pool

Dep

th: 4

4m

Pool ID:43.5m

AS: ID35.4m

SSLS: ID40.1m

AS: Acrylic sphere; SSLS: stainless steel latticed shell

Filling +Overflow

Acrylic sphere(20Kt LS in it)

~18000 20” PMT+~25000 3’’ PMT

~2000 20’’ PMT

SS latticed shell

Electronics

Acrylic Sphere: ID: 35.4mThickness:120mm

SSLS:ID: 40.1mOD: 41.1m

Water poolID: 43.5mHeight: 44mWater Depth: 43.5m

JUNO Detectors

R&D of 20” MCP-PMT

44

MCP Principle

Project Team

• Advantages:– Higher QE: transmissive

photocathode at top + reflective photocathode at bottom

– High CE: less shadowing effect– Easy for production: less manual

operation and steps

5”(8”) Prototype

20” Prototype

Design Production

2009 2010~2013 2013~2015 2016~2019

MCP-PMT Performance

45Min:24.5%; Max:29%

Average:26.5%

QE & uniformity Dark rate

After pulse

PMT Purchasing of JUNO

46

15k MCP-PMT (75%) from NNVT5k Dynode(25%) from Hamamatzu

Dec.16, 2015Characteristics unit MCP-PMT

（NNVC）R12860

（Hamamatsu）Detection Efficiency

(QE*CE*area) % 27%, > 24% 27%, > 24%

P/V of SPE 3.5, > 2.8 3, > 2.5TTS on the top point ns ~12, < 15 2.7, < 3.5Rise time/ Fall time ns R~2, F~12 R~5,F~9 Anode Dark Count Hz 20K, < 30K 10K, < 50K

After Pulse Rate % 1, <2 10, < 15

Radioactivity of glass ppb238U:50232Th:5040K: 20

238U:400232Th:40040K: 40

By Scaling PMT Spec for LS quantity to reach 3σ@ 6year

Decision based on risk, price, performance merit for physics

47

Acrylic Sphere R&D

48

Acrylic divided into 200+ panels Acrylic nodes test:

Upto ~700 kN pulling forces when it breaks

Part of sphere are manufactured to test the techniques

Sphere sheet manufacturing: 3m X 8m X 0.12 mThe problems of shrinkage and shape variation were

resolved.

Progress of LS R&D

49

Al2O3column

LAB and Al2O3mixing tank

Pure LAB

• Test the overall design of purification system at Daya Bay. Replaced the target LS in one detector

• Quantify the effectivities of subsystems– Optical : > 20m A.L @430nm– Radio-purity: < 10-15 g/g (U, Th)

• Determine the choice of sub-systems– Al2O3 column, distillation, gas striping, water extraction

Distillation system

Steam stripping system

Distillation and steam stripping system.

JUNO-LS Pilot plant

50

Full system tested in Daya Bay LS hall.A new batch of purified LS was produced and filled into DYB-AD1.LS radioactivity being evaluated with data. Further tests to optimize the LS recipe

CLS GT

Calibration System

51

Guide Tube

JUNO Schedule

52

Schedule:• Civil preparation：2013-2014• Civil construction：2014-2018• Detector component production：2016-2017• Detector assembly & installation：2018-2019 • Filling & data taking：2020-2021

Future Plan• Run for 20-30 years• Likely, double beta decay experiment in 2030

vertical shaft sloped tunnel

Exploring nature of v’s mass• 0νββ is currently the viable and

sensitive probe to the Majorananature of ν

53

Z.Z. Xing, Y.L. Zhou, CPC39 (2015) 011001

Current Limit

Next generation𝑻𝑻𝟏𝟏/𝟐𝟐𝟏𝟏𝝊𝝊𝝊𝝊𝝊𝝊~1028 yr

ν is MajoranaYes

NO

Inverted NMO? Yes(by other exp.)

Observed 0νββ?

ν is Dirac(if NOT intro. new physics)

NO

Little hope for next generation 0νββexperiment to determine ν’s Majorana

The JUNO LS detector has good potential for DBD search• 3%/ 𝐸𝐸 energy resolution• 35.4 m diameter of LS• Clean balloon hold enrXe gas (>80% 136Xe) dissolved LS• LS purity: 10-15 g/g (U/Th) 10-17 g/g (U/Th)• Excellent muon track rec. reject cosmogenic

isotopes

54

Ultimate 0νββ search at JUNO

Chin.Phys.C 2017, 41(5): 53001-053001

Expected Background: 1.34/FWHM/ (ton 136Xe)/yr

Possibly in 2030

Summary• Significant improvement on Sin22θ13 precision from the Daya

Bay, Double Chooz and RENO experiments. Ultimate precision of Sin22θ13 will reach ~ 3%

• Precision measurement of the absolute neutrino flux and spectrum.

• Reactor anomaly may have a definite answer before 2020.• Future reactor neutrino experiments:

– Mass hierarchy (3-4 σ in 2026)– Precision measurement of 3/6 mixing parameters up to <

~1% level unitarity test of the mixing matrix– Sterile neutrinos– Rich physics topics with a detector like JUNO

55

Thanks!

56

Spectra of Isotopes

57

• Ab initio: Nuclear database, Σ fragments, Σ chains, Σ branches 10%uncertainty (e.g. Vogel et al., PRC24, 1543 (1981)).

• ILL measured the β-spectra convert to neutrino spectra– ILL spectra: Use spectra of 30 virtual branches, fit amplitude and

endpoints – Mueller spectra: 90% Ab initio, 10% fit rate anomaly– Huber spectra: fit w/ improved nuclear effects

K. Schreckenbach et al. PLB118, 162 (1985)A.A. Hahn et al. PLB160, 325 (1985) Shape verified by Bugey-3 data

Normalization by Bugey-4, 1.6%

Combination Prospectives• The 3 experiments started to discuss combination• To the end of 2016 (by J. Zhao)

– Daya Bay: 0.00307– Daya Bay + RENO: 0.00287 improve by 6.5%– Daya Bay + RENO + DC: 0.00282 improve by 8.2%

• Assumed uncorrelated– Reactor, detector, background– Cosmogenic bkg correlation worse– Cosmogenic bkg (50% systematic uncer.) reduction better

58

Zoom

Beyond Photo-statistics

• Generic form of E resolution– a: stochastic term– b: constant term– c: noise term

Impact to MH sensitivity

41

• Data validated Full MC (DYB&DC)• Noise term dominated by PMT dark noise• Constant term

– Residual non-uniformity– Flaws in readout electronics– Artifacts from resolution plotting

• No JUNO show stopper found in DYB model

NH IH

Dighe & Smirnov, hep-ph/9907423

L – Resonance Res. for neutrinos Always adiabatic

( )12221 θ ,Δm H – Resonance

Res. for neutrinos (NH) Res. for antineutrinos (IH)

( )13231 θ ,Δm

Adiabatic for a ‘large’ θ13

Supernova Neutrinos: Elementary Particle Physics

Flavor Conversion: MSW effects in the SN envelope

for

for

for

0eF0

eF0

xF

eν

eν

ττµµ νννν ,,,

00 )1( xee FpFpF −+=00 )1( xee FpFpF −+=

000

41

41

42

41

eexx FpFpFppF −+

−+

++=∑

Normal

Inverted

sin2(2θ13)

≳ 10−3

Mass ordering

sin2(θ12)

0 cos2(θ12)

0

Case

A

B

Survival probabilities

)for(p eν )for(p eν

Dighe & Smirnov, hep-ph/9907423 Dighe, Kachelriess, Raffelt & Tomàs, hep-ph/0311172

Primary fluxes (+ collective) Leaving the SN (+ Collective & MSW effects)

Supernova Neutrinos: Elementary Particle Physics

Flavor Conversion: MSW effects in the SN envelope

Short Baseline Exp. with Gas TPC• Gas TPC detector at ~20 m from a reactor

– ν-e scattering– High energy precision ( <3%/sqrt(E) )

• Major motivation: high precision reactor spectrum to 1%– Input for JUNO. Daya Bay energy resolution 8%, JUNO 3%

• Other motivations:– The weak mixing angle θw– Abnormal magnetic moment– Sterile neutrino

• Prototyping at IHEP. Prepare LOI in 2017MUNU exp:µν < 0.9 × 10−10 µBCF4 , T > 700 keVPLB 615(2005)153

Instrumental background: PMT Flasher

63

Flashers

Signal Flasher

Common phenomena in the past large neutrino experiments (KamLAND, Super-K, Borexino, SNO), caused by discharge on dynodes or bases. Not easy to efficiently reject at that time Turn off the problematic PMTA highly-efficient flasher cut developed for DYB: ~100% eff., 0.01% error

8He/9Li

8He/9Li background

64

Old approachNew approach

Classical method: time-since-last-muon fitDifficulty: large error when muon rate is high, impossible to estimate the Li9 production yield for muons that havelow energy deposition

Guess: 8He/9Li production accompanied with neutron reduce the rate of muon sampleResult: successful data-driven test and significantly reduced the error

L. J. Wen et al., NIMA 564 (2006) 471-474

β-n emitter: muon spallation on Carbon nuclei

13C μn?

8He/9Li dominates the background uncertainty

Optical Model for large LS detectors

65

L.J Wen, Ph.D thesisX.F. Ding, L.J. Wen,et al, CPC 39 (12) (2015) 126001.X.Y. Li, L.J. Wen, et al, paper in preparation

Developed a new optical model to address those issues. Key optical parameters:• Molar attn. coefficient of

each composition• Quantum yield of each fluor

LAB PPO

bis-SMB

• What’s the best LS recipe?– Detector size dependent, cannot rely on lab

experiment

• Predict the detector non-uniformity• Understand non-linearity from Cerenkov

PreliminaryTab-top meas.

By Wavelength shifter (WLS)?

By Primary fluor (PF)?

By Solvent ?Photon

Absorbed?

TransportationReemission?

Dead

Reemission?

Re-emission (by PF or WLS)

Dead

Yes

NO