The Selena Neutrino Experiment

The Selena Neutrino Experiment

CENPA

Center for Experimental

Nuclear Physics and Astrophysics

Alvaro E. Chavarria

University of Washington

Outline

• Selena detector concept.

• R&D progress and status.

• Prospects for neutrinoless !! decay.

• Prospects for a zero-background solar " experiment.

• Solar " spectroscopy with Selena.

Selena detector concept

Pixel charge

Bragg peak

Bragg peak

Vertex Simulation

JINST12(2017)P03022

+HV Digital

output

Integrated

electronics CMOS pixel array 100 cm² x 5 μm

Electrode (30 nm) Amorphous ⁸²Se (200

μm thick)

E

e^-

e^-

Signals

100s of modules

Spatial correlations

5330 5332 5334 5336 5338 5340 5342 5344 5346 5348 3152

3154 3156 3158 3160 3162 3164 3166 3168

x [pixel]

y [pixel]

4 keV e

5330 5332 5334 5336 5338 5340 5342 5344 5346 5348 3152

3154 3156 3158 3160 3162 3164 3166 3168

x [pixel]

y [pixel]

41 keV e

5330 5332 5334 5336 5338 5340 5342 5344 5346 5348 3152

3154 3156 3158 3160 3162 3164 3166 3168

x [pixel]

4323 keV α

y [pixel]

101

1 10 102

Energy [keV]

∆t = 4 days ∆t = 10 days

•

Pb surface background:

210

Pb

Bi

Po

• Cosmogenic

Si:

Si (T

= 150 y, β) ➔

P (T

= 14 days, β)

Search for spatially

correlated beta decays.

Sensitivity with current data is few Bq/kg.

JINST16(2021)P06019

Data from DAMIC: UW’s other imaging solid state detector.

Detector goals

• Excellent energy resolution to perform precision

spectroscopy, i.e., distinguish 0" signal from 2" !! decay background.

• Excellent spatial resolution to suppress backgrounds by event topology, e.g., number of Bragg peaks, decay

sequences, etc.

• Easy to build: room temperature operation, not very

stringent radio purity requirements or fiducialization, etc.

• Scalable: standard CMOS foundry process (300 mm diameter wafers) and industrial aSe deposition.

2021 2024 2029 2040

Small prototypes

Large detector module (0.25 kg)

100 kg (400 modules) demonstrator

10-ton detector

underground

E [keV]

0 500 1000 1500 2000 2500 3000

RMS / Mean

0 0.01 0.02 0.03 0.04 0.05

/mean, All Events σ

/mean, All Events, with Dead Layer σ

/mean, Cut on Track Length σ

/mean, Cut on Track Length, with Dead Layer σ

Invers Square Root

/mean, Previous Assumption σ

JINST16(2021)P06018

57

Co


122 keV

Source from above

R&D: Single pixel

R&D: Topmetal-II -

3

− −2 −1 0 1 2 3

X [mm]

3

−

−2

−1 0 1 2 3

Y [mm]

0 50 100 150 200 250 300 350 400 ]_-

Charge [e

image 36504

600

− −400−200 0 200 400 600 800 1000

-] charge [e 1

10 102

count

Pixel charge in image 36504

−3 −2 −1 0 1 2 3

X [mm]

−3 2

−

−1 0 1 2 3

Y [mm]

0 50 100 150 200 250 300 350 400 ]_-

Charge [e

image 30548

−600−400−200 0 200 400 600 800 1000

-] charge [e 1

10 102

count

Pixel charge in image 30548

Before aSe After aSe Test board

Electron tracks from

Sr-Y!

‣ From Y. Mei at LBNL.


‣ CMOS pixel array with

exposed metal electrodes.

‣ (83 #m)

pixels

‣ 15 e

pixel noise.

NIMA810(2016)144

Status

• We measured the energy response of aSe.

• We coupled aSe to a CMOS pixel array.

Milestones achieved:

We are already characterizing the response of our first aSe/CMOS image sensors!

• Many opportunities to improve energy and spatial resolution:

thinner dead layer, smarter energy reconstruction, smaller pixel size…

We present prospects for 0"!! decay search and solar "

spectroscopy assuming experimentally informed expected detector performance for an exposure of 100 ton-year

Publication in the coming months.

x [mm]

6.0 6.3 6.6 6.9 7.2 7.5

y [mm]

7.5 7.8 8.1 8.4 8.7 9.0

1 10 102

103

104

EVENT= 36 EDep

x [pixel]

400 420 440 460 480 500

y [pixel]

500 520 540 560 580 600

18 19 20 21 22 23 24 25 26

EVENT= 36 Layer

2800 2850 2900 2950 3000 3050 3100 3150 3200 E [keV]

0 0.05 0.1 0.15 0.2 0.25

5keV]yrEvents /[t

27 yr

1/2=10 , T 0

20 yr

1/2=10 , T 2

x [mm]

6.60 6.75 6.90 7.05 7.20 7.35 7.50 7.65

y [mm]

6.0 6.3 6.6 6.9 7.2 7.5

1 10 102

103

104

EVENT= 15 EDep

x [pixel]

440 450 460 470 480 490 500 510

y [pixel]

400 420 440 460 480 500

17 18 19 20 21 22 23 24 25 26

EVENT= 15 Layer

0 "!! decay

Single e

‣ By identification of Bragg peak we can achieve 10

suppression of single electron background, with 50% signal acceptance.

‣ Bulk backgrounds suppressed by α/β particle ID, spatial correlations.

Background rate <6 x 10

/keV/ton/year!

JINST12(2017)P03022

T

> 10

years limit on

Se 0"!!

‣ Pixel size would need to be improved to (15 #m)

and dead layers to ~10 #m.

Double e

Solar "

• Solar "s have been identified as a background for searches for 0" !! in

Se.

!10

PHYSICAL REVIEW C95, 055501 (2017)

Solar neutrino interactions with the double-β decay nuclei ⁸²Se, ¹⁰⁰Mo, and ¹⁵⁰Nd

H. Ejiri¹ and S. R. Elliott²

1Research Center for Nuclear Physics, Osaka University, Ibaraki, Osaka 567-0047, Japan

2Los Alamos National Laboratory, Los Alamos, New Mexico, USA

(Received 8 February 2017; published 2 May 2017)

Solar neutrinos interact within double-beta decay (ββ) detectors and contribute to backgrounds for ββ experiments. Background contributions due to solar neutrino interactions with ββ nuclei of ⁸²Se, ¹⁰⁰Mo, and

150Nd are evaluated. They are shown to be significant for future high-sensitivityββ experiments that may search for Majorana neutrino masses in the inverted-hierarchy mass region. The impact of solar neutrino backgrounds and their reduction are discussed for futureββ experiments.

DOI:10.1103/PhysRevC.95.055501

I. INTRODUCTION

Neutrino-less double beta decay [ββ(0ν)] is a unique and realistic probe for studies of neutrino (ν) properties and especially the Majorana mass character of the neutrino and the absolute mass scale.ββ studies and ν masses are discussed in recent reviews and their references [1–4].

The rate of ββ(0ν), if it exists, would be extremely small becauseββis a second-order weak process that requires lepton number conservation violation and Majorana fermions. There are numerous possible mechanisms that can give rise toββ(0ν) as is discussed in the reviews [1–4]. It is known, however, that light neutrinos exist and therefore it is convenient to consider light-neutrino exchange as the default process by which to benchmark ββ(0ν) rates. For light-neutrino exchange, the rate depends on the effective Majorana mass squared and the typical mass regions to be explored are about 45–15 meV and 4–1.5 meV in cases of the inverted-hierarchy (IH) and normal- hierarchy (NH) mass regions, respectively. The ββ(0ν) half- lives expected for these regions are near or greater than 10²⁷yr, depending significantly on the nuclear matrix element (NME), including the effective axial weak couplingg_A. As a result, the ββ(0ν) signal rate (S_ββ) is near or less than a few counts per ton of ββ isotope per year (t y). Accordingly the background rate necessarily has to be around or less than one count per t y.

Solar-νs are omnipresent and cannot be shielded, and thus their charged current (CC) and neutral current (NC) interac- tions are potential background sources for high sensitivityββ experiments as discussed in [5,6] and references therein. In fact, it has been shown that solar-ν CC interactions with ββ isotopes like ¹⁰⁰Mo [7], ¹¹⁶Cd [8], and ¹⁵⁰Nd [9] can be used for real-time studies of the low-energy solar-νs.

The ββ isotopes most often used or considered for high- sensitivity experiments are ⁷⁶Ge, ⁸²Se, ¹⁰⁰Mo, ¹³⁰Te, ¹³⁶Xe, and ¹⁵⁰Nd. Here, we classify them into two groups. Group A consisting of ⁸²Se, ¹⁰⁰Mo, and ¹⁵⁰Nd has a large solar-ν CC rate, whereas Group B consisting of ⁷⁶Ge, ¹³⁰Te, and ¹³⁶Xe has a rather small solar-ν CC rate. In the previous paper [5], background contributions from CC solar-ν interactions were discussed for the three ββ nuclei in Group B. Solar-ν interactions with atomic electrons in ββ isotopes and liquid scintillators used for ββ experiments were considered in

The present paper aims to evaluate the background contri- butions for the three nuclei in Group A and discuss the impact on high sensitivityββexperiments using them. The mechanics of these calculations are identical to those in Ref. [5], and we do not repeat all the details here.

II. SOLAR NEUTRINO BACKGROUNDS

The process of ββ decay from ^Z−1A to ^Z+1A via the intermediate nucleus ^ZA is shown in Eq. (1). Solar νs can produce background to this signal primarily through CC interactions with ββ nuclei. The CC interaction produces background in two ways. First the CC interaction itself can produce a signal (B_CC) given by the promptly emittede⁻ and, if the resulting nucleus is in an excited state, a number of γ rays may be emitted as the nucleus relaxes to its ground state.

Second, the resulting nucleus^ZA can then β⁻ decay to ^Z+1A by emitting a single β⁻ ray and possibly also γ ray(s) if the residual state is an excited state (B_SB). The interaction and decay schemes are shown in Fig. 1. The three processes are expressed as

ββ : ^Z−1A → ^Z+1A+β⁻ +β⁻ +Q_ββ,

CC : ^Z−1A+ν → ^ZA+e⁻ +γ(s)+Q_ν, (1) SB : ^ZA → ^Z+1A+β⁻ +γ(s)+Q_β,

where ββ, CC, and SB denote the ββ decay, the solar-ν CC interaction, and the singleβdecay processes, respectively. The Qvalues for each are given as Q_ββ,Q_ν, and Q_β, respectively, as shown in Fig. 1.

We consider ββ detectors where the sum energy of the β andγ rays is measured. Theββ(0ν) signature is a peak within the region of interest (ROI) at the ββ Q value (E = Q_ββ).

In contrast, the sum of the electron energy (E_e) and any successive γ ray energy (E_γ) is a continuum spectrum for both the CC and SB processes. The backgrounds of interest are estimated by their yields within the ROI and therefore are sensitive to the detector’s energy resolution. The energy width of the ROI is given by the FWHM resolution ($E) at the energy of the Q value. This ratio, defined as δ, is the fractional energy resolution at E = Q_ββ. Certainly the

• This is not a problem for our 0"!! search because

the capture product

Br is a ! emitter, which leads

to a decay sequence that we can tag and discard.

!11

D. FREKERS et al. PHYSICAL REVIEW C 94, 014614 (2016)

82Se 82Br

82Kr

5− 0+

0+

β−β−

Q β−β− = 2997.9(3)

Qβ− = 3093.0(10) QEC = 96.6(17)

35.3h 2− 45.95 6.13min 1+ 75.06

1+ states

{

FIG. 1. Sketch of the 2νββ decay process in ⁸²Se. The transition paths through the intermediate 1⁺ states in ⁸²Br are indicated.

The various Q values and excitation energies are taken from Refs. [1,30,31]. All energies are given in keV units.

T_1/2^2νββ(¹³⁰Te) = [6.9 ± 1.3] × 10²⁰yr. Nonetheless, the mod- eling of the time of mineralization, as well as of the retention times for the noble gases, and the assumptions made about alternative productions through fission, fission induced neutrons or cosmic rays over geological time scales, remain the main sources of systematic uncertainties in geochemical analyses. On the other hand, by today’s good knowledge of many 2νββ-decay half-lives one is now in the position to constrain many of these assumptions. This is indicated in detail in Ref. [14].

In this article the ⁸²Se(³He,t)⁸²Br reaction has been used to provide a detailed and high-resolution insight into the GT strength distribution in ⁸²Br, as this ties in to some of the aforementioned subjects. One may recall that hadronic charge-exchange reactions at intermediate ener- gies (i.e., 100–300 MeV/A) and low momentum transfers q_{t r} ≈ 0 fm⁻¹ are a well-established tool to selectively induce GT transitions [15]. This follows from the dominant σ τ component of the effective nucleon-nucleon (N N) interaction in this energy/momentum region [16–20], where, in addition, the nucleus exhibits a high nuclear transparency.

The present experiment is within the spirit of previous high-resolution (³He,t) charge-exchange experiment on light and medium-weight nuclei reported in Refs. [15,21–28], and many details of the reactions and analyses can be found there.

Charge-exchange reactions for ββ decay and astrophysical neutrino studies are, for instance, reviewed in Ref. [29].

II. EXPERIMENT

The experiment was performed at Research Center for Nuclear Physics (RCNP), Osaka University. A 420 MeV

3He⁺⁺ beam was accelerated using the Azimuthally Varying Field (AVF) Cyclotron in combination with the Ring Cyclotron and transported to the scattering chamber of the Grand Raiden Spectrometer [32]. The West-South (WS) beam line [33]

provided the dispersion of the beam necessary for obtaining high-resolution (³He,t) spectra. Several tuning techniques for dispersion matching between beam line and spectrometer were

employed to optimize energy and angular resolution. These are described in Refs. [32–36].

Outgoing reaction tritons were momentum analyzed in the Grand Raiden Spectrometer within its full acceptance of

±20 mrad in horizontal and ±40 mrad in vertical direction.

The detection system consisted of a set of two mutliwire drift chambers, which allowed precise track reconstruction on the focal plane [37]. They were followed by two thin (3 and 10 mm) plastic scintillators used for particle identification and for providing the event trigger.

A thin 1.79(5) mg/cm² Se target evaporated on a 150 µg carbon backing was employed. The target thickness was determined by performing an energy-loss measurement of α particles traversing the target foil in a specially designed setup. The thickness calculation was done with the computer code SRIM [38]. The Se material was isotopically enriched and specified at 97.43(2)% ⁸²Se.

After applying various off-line spectrometer aberration corrections, a final-state energy resolution of 38 keV was obtained, which was partly due to the energy-loss differences between ³He and triton ions in the target.

An energy calibration was performed using a ²⁶Mg and a

natSi target. These targets provide numerous levels at well- known excitation energies distributed over a large momentum in the focal plane. In the energy region up to the isobaric analog state (IAS) the accuracy is at a level of ±2keV.

Two spectrometer-angle settings, i.e., 0^◦ and 2.5^◦, allowed generating center-of-mass (c.m.) angular distributions ranging from θ_c.m. ≈ 0^◦ to θ_c.m. ≈ 4.0^◦.

III. ANALYSIS

Excitation-energy spectra of the ⁸²Se(³He,t)⁸²Br reaction are shown in Fig. 2. Three angular cuts have been overlaid to visualize the behavior of the angular distribution.

The spectra are dominated by the strongly excited IAS, which is located at an excitation energy of E_x = 9.576 MeV.

This energy agrees well with the value of E_x = 9.58 MeV reported in Ref. [13] from a ⁸²Se(p,n) reaction at 134.4 MeV and with an energy resolution of about 300 keV. The broad Gamow-Teller resonance (GTR) appears above the IAS at an excitation energy of E_x ≈ 12.1 MeV and has a width of

≈5 MeV.

The low excitation energy region is dominated by a GT transition to the 75 keV (1⁺) state. The lowest excited state at 45.9 keV (2⁻) (cf. Fig. 1) was too weakly populated at these angles to be identified against the strong 1⁺ state at 75 keV.

It appears that the spectra below ≈2.1 MeV excitation energy show a remarkably low level of fragmentation, contrary, for instance, to its close-by neighbor ⁷⁶Ge [22]. Only three strongly excited J^π = 1⁺ states, at 75, 1484, and 2087 keV appear in the spectra. On the other hand, the fragmentation suddenly increases dramatically above ≈2.1 MeV. In total, more than 60 states below 6 MeV were identified above a general background dominated by the tail of the GTR. This situation is more reminiscent of the ⁷⁶Ge case shown in Ref. [22], where this high level of fragmentation was attributed to the softness of the nuclear shape near A = 76.

"

+

Se

Br* + e

+ C.E. (29 keV)

82

Br*

Br + C.E. (46 keV)

“prompt”

82

Br

Kr + $ + !

%

= 6.1 min

%

= 35 hours 445 keV end-point

E

– 172 keV

Triple sequence:

We can also use the prompt events

to perform solar "

spectroscopy!

‣ Expected number of three accidental events in a (100

#m)

pixel (22 µg) is <10

in 100 ton year.

‣ Other &, p, or n reactions that make

Br* have a different prompt event topology.

‣ No cosmogenic isotope starts a decay chain that

mimics the triple coincidence.

Backgrounds

0 0.5 1 1.5 2 2.5 3

Energy / MeV 0

5000 10000 15000 20000 25000 30000 35000 40000

Events / keV / ton / year

Background spectrum for large scale Selena

!! decay spectrum 0.1 d.r.u

6.5 x 10

ton

year

‣ Some neutron captures on Se isotopes can give triple coincidences but their event topologies are also very different.

No identified background to mimic the triple coincidence.

Possibility of zero background ' spectroscopy!

Solar " measurements

Solar neutrinos

Solar fluxes Survival probability

‣ Nuclear reactions in the sun.

‣ pp + CNO cycles.

‣ Predicted by solar models.

‣ Vacuum oscillations, matter effects.

‣ Precise prediction by MSW-LMA

(parameters from PMNS matrix).

0 0.2 0.4 0.6 0.8 1 1.2 1.4 Energy / MeV

4

10−

−3

10

−2

10

−1

10

-1-1-1 keV yearRate / ton 1

Prompt spectrum (low energy)

Prompt spectrum

"

+

Se

Br* + e

+ C.E. (29 keV)

E

– 172 keV

pp

Be

pep

CNO CNO

Capture ((E) from

PRC 94, 014614 (2016)

Species E range

(keV)

N 1/√N

pp

6170 1.3%

7

Be

1850 2.3%

pep

151 8.1%

CNO

63 12.6%

8

B

209 6.9%

100 ton-year

Solar Physics

‣ Solar luminosity: pp rate unc. improved to 1.3% (currently

~10% by BX). Flux measurement uncertainty depends on

uncertainty on "

capture (. If calibrated to "-e scattering ( by comparing measured

Be rate with BX, sys. unc. = 3.5%.

‣ Solar metallicity: CNO flux measured to 13%. Difference

between high- and low-metallicity solar model predictions: 28%.

‣ Solar core temperature:

Species RMS width (keV)

Mean unc.

(keV)

Line diff.

(keV) Fraction

7

Be 14.8 0.34 1.29 26%

pep 19.8 1.60 7.59 21%

PRD 49, 3923 (1994)

Neutrino physics

‣ Onset of matter effects in " oscillations: pep/pp rate ratio measured to 8%. Flux ratio very well predicted by SSM, MSW

effect should suppress pep "

flux by 7%. NSI could increase this.

Sensitive probe for neutrino transport in the Sun

Conclusion

• A large Selena detector has the potential for background-free search for 0"!! decay and solar " spectroscopy.

• Could place a limit on T

> 10

years on

Se 0"!! 


(m

< 10 meV).

• Measure basic properties of the Sun to validate our solar / stellar models.

• We already fabricated and operated the lowest noise single- pixel aSe sensor ever!

• We demonstrated that aSe can be coupled to a CMOS pixel array to image electron tracks!

• Our R&D program continues to characterize and optimize the

CMOS/aSe sensor toward a large-area module for Selena.

Thank you!

• University of Washington: Alvaro E. Chavarria, Alex Piers, Harry Ni.

• Berkeley Lab: Yuan Mei, Xinran Li.

• Princeton University: Cristiano Galbiati.

• Hologic Coporation: Brad Polischuk, Snezana Bogdanovich.

CENPA

Center for Experimental

Nuclear Physics and Astrophysics