Studying Neutrino Directionality with Double Chooz

Erica Caden for the Double Chooz Collaboration

Drexel University

5 October, 2012 Outline

Motivation Double Chooz Experiment Inverse Beta Decay Kinematics Neutrino Directionality Results Comparing with CHOOZ Conclusion and future work

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 2 / 35 Motivation

Measuring the direction of incoming anti-neutrinos could be very useful for locating a supernova or detecting geoneutrinos. It is also a good veriﬁcation of an experiment’s Monte Carlo simulation. CHOOZ is the only non-segmented reactor anti-neutrino detector to produce a neutrino directionality measurement1. It is then natural for Double Chooz to try and improve that measurement.

1PRD61, 012001 (1999) E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 3 / 35 The Double Chooz Collaboration

Brazil France Germany Japan Russia Spain USA CBPF APC EKU Tohoku U. INR RAS CIEMAT– U. Alabama UNICAMP CEA/DSM/ T¨ubingen Tokyo Inst. IPC RAS Madrid ANL UFABC IRFU: MPIK Tech. RRC U. Chicago SPP Heidelberg Tokyo Kurchatov Columbia U. SPhN RWTH Metro. U. UCDavis SEDI Aachen Niigata U. Drexel U. SIS TU M¨unchen Kobe U. IIT SENAC U. Hamburg Tohoku Kansas State CNRS/IN2P3: Gakuin U. LLNL Subatech Hiroshima MIT IPHC Inst. Tech. U. Notre Dame Spokesperson: H. de Kerret (IN2P3) Project Manager: Ch. Vessiere (CEA–Saclay) www.doublechooz.in2p3.fr

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 4 / 35 The Double Chooz Experiment

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 5 / 35 The Double Chooz Far Detector

Glove Box Deployment of Calibration sources Outer Veto Extruded scintillating strips with WLS ﬁbers connected to 64 anode PMTs Steel Shielding 250t, 150mm Inner Veto 78 PMTs, 90m3 liquid scintillator Buﬀer 390 PMTs, 114m3 non-scintillating mineral oil Gamma Catcher 23m3 liquid scintillator Target 10m3 Gd-doped liquid scintillator

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 6 / 35 Detecting Anti-Neutrinos with Double Chooz

Reactors are cleanν ¯e source, and backgrounds are suppressed by a two-part coincidence signal Prompt signal: positron + annihilation γs = 1 ∼ 12 MeV Delayed signal: γs from neutron capture on Gd = 8 MeV Time interval: ∆t ∼ 30µsec

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 7 / 35 Anti-Neutrino Candidate Selection

Muon Veto

I No Coincidence in the IV I ∆t > 1ms Prompt Event

I 0.7 < Evis < 12.2 MeV I Light noise cuts Delayed Event

I 6 < Evis < 12 MeV I Light noise cuts Delayed coincidence

I 2 < ∆t < 100µs Multiplicity

I No extra events around signal Further Background Rejection

I ∆tµ >500ms (Eµ > 600MeV) I No coincidence signal in OV

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 8 / 35 Prompt and Delayed events for ν candidates in DC

Prompt: Delayed:

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 9 / 35 Reactorν ¯e Inverse Beta Decay Event

θ e ν θ e p n1 n θ n2

θ n3

+ ν¯e + p → e + n

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 10 / 35 IBD: Positron Angular Distribution

From Vogel and Beacom2, the positron angular distribution is given by dσ ' 1 + a(Eν)ve cos θe . d cos θe

where θe is the positron recoil angle and ve is its velocity in c = 1 units.

Inﬁnitely Massive nucleon (1/Mn ≈ 0)

Asymmetry coeﬃcient, a(Eν), reduces to: a(0) = −0.102.

2PRD60, 053003 (1999) E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 11 / 35 IBD: Positron Angular Distribution

The average cosine, weighted by the diﬀerential cross section, is

1 (0) hcos θe i = a ve = −0.034ve . 3 The positron is relativistic, so therefore the average positron angle is constant, independent of Eν, and slightly backwards. CHOOZ showed that the average positron displacement is -0.05 cm. The small displacement and negligible preferred scattering angle mean that the positron movement can be neglected. We assume the positron interaction vertex to be the anti-neutrino interaction vertex.

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 12 / 35 IBD: Neutron Angular Distribution

In the Laboratory frame, conservation of momentum gives:

~pν = ~pe + ~pn.

And conservation of energy yields: q 2 2 |pe | ≤ (Eν − ∆) − me < Eν.

where ∆ = mn − mp = 1.293 MeV. So the neutron is always emitted in the forward direction. The maximum angle between the anti-neutrino and initial neutron directions has been calculated to be: p 2 2 2Eν∆ − (∆ − me ) cos(θn)max = . Eν

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 13 / 35 IBD: Neutron Angular Distribution

Initital Neutron Distribution ) n

θ 1.0 cos( 0.8

0.6

0.4

0.2 Eν = 1.806 MeV

0.00 2 4 6 8 10 12 Eν [MeV]

p 2 2 2Eν∆ − (∆ − me ) cos(θn)max = Eν

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 14 / 35 IBD: Neutron Angular Distribution

Accounting for the small kinetic energy of the neutron and ﬁnite neutron mass, 1/mn terms arise in the neutron’s kinetic energy:

2 2 Eν (Eν − ∆) ∆ − me En = (1 − cos θe ) + mn 2mn

For a typical neutrino of 3.5 MeV and assuming cos(θe ) = 0, the resulting neutron initially has just 10 keV.

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 15 / 35 IBD: Neutron Moderation

At reactor energies, neutron thermalization is a non-isotropic process in which the neutron’s initial direction is preserved. With each scatter, the average cosine with respect to the incoming direction is: 2 hcos θni = , 3A where A is the mass number of the scattering nucleon. After thermalization, the neutron diﬀuses until it is captured. Diﬀusion is isotropic, but the average initial displacement is still away from the anti-neutrino source.

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 16 / 35 IBD: Neutron Moderation & Diffusion 5000 neutrons with 10keV were simulated using GEANT4 at the center of the detector with momentum in the initial y direction. The figure on the left shows the moderation time needed to thermal energy of 0.025eV. The diffusion time is exponential. The figure on the right shows the number of collisions needed for moderation and diffusion.

Langbrandtner, (2011) E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 17 / 35 IBD: Neutron Moderation & Diﬀusion The red dot indicates the initial neutron position. After ∼ 10 scatterings, the average y-displacement is 1.72 cm.

Langbrandtner, (2011) E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 18 / 35 IBD Kinematics - Summary

The positron is emitted slightly backwards, minimally scattered, and quickly captures on an electron. We assume the positron interaction point to be the neutrino interaction point. The neutron’s kinetic energy is small, and it is always emitted along the anti-neutrino’s direction. The neutron’s momentum is still in the forward direction through the moderation phase, and its displacement is measurable.

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 19 / 35 Principles of ν Direction Reconstruction

We ﬁrst deﬁne the reactor-detector vector: ~ ~ ~ XRD = XDetector − XReactor

For each neutrino candidate i, we have the reconstructed delayed and prompt vertices: ~ i ~ i Xn and Xe . We can construct the neutrino candidate displacement vector:

~ i ~ i ~ i Xen = Xn − Xe

and its unit vector: ˆ i ~ i i Xen = Xen/ Xen

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 20 / 35 Double Chooz Layout

(82.5,0,120) (-82.5,0,120) B1 B2 +x'

ν e

+x +y

+y' (-747.1,744.1,84)

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 21 / 35 Measuring ν Directionality cos(θ) between the two unit vectors should tend more towards +1 − than 1: ˆ i ˆ cos(θ) = Xen · XRD

1000 Entries 800

600

Data: = 0.05846 400 MC: = 0.05280

200 PRELIMINARY

-1.00 -0.8 -0.6 -0.4 -0.2 -0.0 0.2 0.4 0.6 0.8 1.0 cos(θ) The χ2/dof between the Data and a ﬂat distribution is 91.9/9, compared to the χ2/dof for a linear ﬁt is 4.83/8. (stat. errors only) E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 22 / 35 Measuring ν Directionality

We then deﬁne ~p as the average anti-neutrino candidate displacement unit vector: N 1 X ~p = Xˆ i N en i=1 The average anti-neutrino candidate’s direction, in the detector’s coordinate system, is

Data : ~p = (−0.0058, −0.0584, 0.0049) mm

MC: ~p = (−0.0066, −0.0524, 0.00044) mm

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 23 / 35 ~ Distribution of Xen Components

1200 1200

1000 1000 1000 µ µ Data: µ = -0.206 MC: µ = -0.174 Data: µ = -1.640 MC: µ = -1.366 Data: = 0.180 MC: = 0.025 800 800 σ σ σ = 15.98 σ = 15.10 800 σ = 15.51 σ = 15.02 = 15.51 = 15.34 PRELIMINARY PRELIMINARY PRELIMINARY 600 600 600

400 400 400

200 200 200

-1500 -100 -50 0 50 100 150 -1500 -100 -50 0 50 100 150 -1500 -100 -50 0 50 100 150 x:n-e+ [cm] y:n-e+ [cm] z:n-e+ [cm]

The distribution of X~en components are Gaussian and have width P. The average distance between the prompt and delayed candidates and the width of their distributions indicate how well we can locate the anti-neutrino candidate source.

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 24 / 35 ~ Distribution of Xen Components

1200 1200

400 400 400

200 200 200

-1500 -100 -50 0 50 100 150 -1500 -100 -50 0 50 100 150 -1500 -100 -50 0 50 100 150 x:n-e+ [cm] y:n-e+ [cm] z:n-e+ [cm] The neutron’s displacement is measured distance between the average delayed candidate interaction point and the average prompt candidate q 2 2 2 interaction point, ` = µx + µy + µz .

Data : ` = 1.66 ± 0.31 cm

MC : ` = 1.38 ± 0.29 cm This agrees well with the original simulated result of 1.72 cm.

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 25 / 35 ~ Distribution of Xen Components

1200 1200

400 400 400

200 200 200

-1500 -100 -50 0 50 100 150 -1500 -100 -50 0 50 100 150 -1500 -100 -50 0 50 100 150 x:n-e+ [cm] y:n-e+ [cm] z:n-e+ [cm] The position resolution is the average width of these distributions:

P = 1/3(σx + σy + σz ).

Data : P = 15.67 cm MC : P = 15.15 cm Due to the symmetry of the Double Chooz detector’s cylindrical geometry, P is expected to be the same for all directions.

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 26 / 35 Preliminary Directionality Results

The average anti-neutrino candidate’s direction, in the detector’s coordinate system, is

Data : ~p = (−0.0058, −0.0584, 0.0049) mm

MC: ~p = (−0.0066, −0.0524, 0.00044) mm Using this, we calculate the azimuthal angle of the neutrino wind : py Data : φ = tan−1 = −96◦. px py MC : φ = tan−1 = −97◦. px The anti-neutrino source is 180◦ from the anti-neutrino wind.

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 27 / 35 Preliminary Directionality Results

The possibility of reconstructing the direction of the anti-neutrino candidate source depends on the error on the average angle, ∆φ. The 1σ uncertainty for the angular location of a single source is: h √ i ∆φ(1σ) = tan−1 (P/`)/ N

We measured the average anti-neutrino candidate displacement to be ` = 1.66 cm. The total width of the prompt and delayed candidate vertex reconstruction has been measured to be P = 15.67 cm. (15.67/1.66) ∆φ = tan−1 √ = 6◦ 8249

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 28 / 35 Preliminary Directionality Results

With N = 8249 events, we can locate the anti-neutrino candidate source at 84◦ to within ∆φ = 6◦ half-aperture cone.

Our Monte Carlo is in agreement, indicating the anti-neutrino source to be at 83◦ within a ∆φ = 7◦ half-aperture cone.

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 29 / 35 Comparison to CHOOZ:

CHOOZ’s result for 2500 candidates was an 18◦ half-aperture cone.3

CHOOZ: |~p| φ θ Uncertainty Data 0.055 -70◦ 103◦ 18◦ MC 0.052 -56◦ 100◦ 19◦

The detector coordinate system used in CHOOZ is not the same as used in Double Chooz: “The average direction of the two rectors in the CHOOZ polar coordinate system ... [is] φ = (−50.3 ± 0.5)◦ and θ = (91.5 ± 0.5)◦ where θ is the zenith angle and the origin of the azimuthal (φ) coordinate is arbitrarily ﬁxed.” Double Chooz’s detector coordinate system is ﬁxed to the orientation of the hall. 3PRD61, 012001 (1999) E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 30 / 35 Conclusions & Future Work

With N = 8249 events, Double Chooz can locate the anti-neutrino candidate source at to within ∆φ = 6◦ half-aperture cone.

We are investigating cuts on the prompt-delayed candidate separation, and work is still ongoing in improving our position reconstruction algorithm using the calibration data from the Z-axis and Guide-Tube systems. The ratio of forward to backward scattered events can be used in estimating the signal to background ratio, as done by Palo Verde4. This would be very useful for the near detector, which has shallower overburden than the far detector.

4PRD62, 072002 (1999) E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 31 / 35 Back Up Slides

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 32 / 35 Directionality in CHOOZ CHOOZ was the ﬁrst non-segmented scintillator detector to measure reactor neutrino directionality. With ∼ 2500 events, they located the reactors within an 18◦ cone.

|~p| φ θ Uncertainty Data 0.055 -70◦ 103◦ 18◦ MC 0.052 -56◦ 100◦ 19◦ {PRD.61.012001(1999)} E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 33 / 35 Segmented Detector Directionality Results - Palo Verde

“Deﬁning the asymmetry

AData = (R − L)/(R + L)

in terms of the number of neutrons captured one column away from the source R and one column toward the source L, a slight asymmetry 0.050 ± 0.017 is found in the data, at 2.9 σ signiﬁcance. Using the Monte Carlo simulation which gives AMC = 0.134 to estimate the portion of the data consisting ofν ¯e signal and assuming the background to be symmetric in this variable, an eﬀective signal to noise ratio

AData +0.4 S/N = = 0.6−0.3 AMC − AData is found. This value agrees well with the ratio of 0.81 ± 0.03 found with the swap analysis method described below.”5

5PRD62, 072002 (1999) E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 34 / 35 Then a(Eν) is given by the competition of the non-spin-ﬂip (Fermi) and spin-ﬂip (Gamow-Teller) contributions.

f 2 − g 2 a(0) = ' −0.10 f 2 + 3g 2 and thus the angular distribution of the positrons is weakly backward. We have deﬁned the vector and axial-vector coupling constants by f = 1, g = 1.26.

E. Caden (Drexel U.) AAP 2012 - Honolulu 5 October, 2012 35 / 35