Status and Results from the EXO Collaboration

The Enriched Xenon Observatory (EXO) is an experimental program searching for neutrinoless double-beta decay using 136Xe. Such a search can shed light on the Majorana nature of the neutrino (whether the neutrino is its own anti-particle), the absolute mass scale of neutrinos, and beyond standard model processes that violate lepton number conservation. The first phase of the experiment, EXO-200, uses 200 kg of xenon with 80% enrichment in 136Xe in a single-phase liquid xenon time projection chamber (TPC). The double-beta decay of xenon is detected in the ultra-low background TPC by collecting both the scintillation light and the ionization charge. The detector has been taking low background physics data with enriched xenon at the Waste Isolation Pilot Plant (WIPP) in New Mexico since early May 2011. The results produced from the collaboration include the first observation of two-neutrino double-beta decay of 136Xe, and a neutrinoless double-beta decay search result that places one of the most stringent limits on the effective Majorana neutrino mass. Building on the success of EXO-200, the collaboration is performing feasibility studies and R&D work for a future multi-tonne scale experiment named nEXO. During the talk, I will discuss the latest results from EXO-200 and prospects of neutrinoless double-beta decay search with both EXO-200 and nEXO.


Introduction
Neutrinos are some of the most difficult known particles to study.Their mass is presently unknown (although we know they do have mass from oscillation experiments), as is their potential to undergo CP-violating transformations.Their extremely low mass is one of the challenges to the standard model.Possibly even more fundamentally, we do not know whether neutrinos are Dirac or Majorana particles.Dirac particles (such as electrons and quarks) are distinct from their anti-particles (positrons and anti-quarks).Majorana particles are not distinct from their anti-particles.Photons and pi-zero particles are like this, though neither is a fundamental fermion.It is currently unknown whether any Majorana fermions exist.Neutrinos, with their lack of electric charge, could be Majorana in nature, in which case the only real difference between a neutrino and an anti-neutrino (as we have observed them) would be the particle spin.
The most promising avenue to get an answer to this question is through a search for neutrinoless double-beta decay (0νββ).In certain nuclei with an even number of both protons and neutrons, ordinary beta decay may be energetically forbidden, (2) This second-order weak interaction is rare, but has been observed for some isotopes, including 136 Xe.
If neutrinos are Majorana particles, it should be possible for one of the two anti-neutrinos in the 2νββ decay to act as a neutrino, and cancel out the other anti-neutrino.This interaction (Figure 1) is neutrinoless double-beta decay.The signature to observe this interaction is that the two emitted electrons will have the full Q-value energy, as none of it is lost to the anti-neutrinos, as it is in 2νββ.
where G 0ν is a phase space factor, M 0ν is a nuclear matrix element, and m ββ is given by and is the effective mass of the electron neutrino.Thus, observation of 0νββ and measurement of its rate would also be a measurement of the neutrino mass.

EXO-200 Detector
The Enriched Xenon Observatory (EXO) collaboration is searching for 0νββ in 136 Xe.This isotope was chosen for several reasons: • It is relatively easy to enrich, as it is the heaviest long-lived isotope of xenon.
• It can be both the source and detection medium (based on scintillation light and ionization of liquid xenon), yielding a monolithic detector.

EPJ Web of Conferences
08001-p.2 • It can be continuously circulated and purified.
• It has a reasonably high Q-value (2457.8keV), well above the energies of most background gammas.
• It may be possible to tag the resulting 136 Ba ++ daughter.The EXO-200 detector is a liquid xenon time projection chamber (TPC), enriched to 80.6% 136 Xe.A voltage gradient of 376 V/cm between the central cathode and the anode planes near the endcaps serve to drift charge from ionization to the induction and collection wire-planes.Avalanche photodiodes (APDs) on the endcaps of the cylindrical TPC collect scintillation light.The combined information of the crossed induction and charge collection wire-planes, along with the scintillation light, and the timing difference between scintillation and charge collection, allows 3-D reconstruction of events in the TPC.More information on the EXO-200 TPC can be found in [1].A photograph of the inside of the TPC (during assembly) is shown in Figure 2. The EXO-200 detector is installed underground (1600 meters water equivalent depth) at the Waste Isolation Pilot Plant (WIPP), near Carlsbad, NM.The TPC is enclosed in a cryostat (for temperature regulation as well as shielding), which is surrounded by further lead shielding.A drawing of the detector and shielding setup is in Figure 3.
A big feature of the TPC is that the scintillation and ionization channels can be combined to get an improved energy resolution [2].This effect is demonstrated in the thorium source calibration data, shown in Figure 4.The calibration sources are fed into the cryostat through a copper source tube that passes through the shielding and enters the cryostat, winding around the TPC.The gamma lines from 228 Th, 137 Cs, and 60 Co sources are used to map out detector response.
Another useful feature of the TPC is that gamma backgrounds can largely be rejected by a multiplicity cut.Gammas tend to Compton scatter and deposit charge at multiple positions in the TPC (known as multi-site, MS, events).Sometimes the clusters are too close together to resolve, resulting in a single-site (SS) event, but we can reduce gamma backgrounds at the Q-value by near a factor of 5 by rejecting MS events.Double-beta decays at the Q-value are expected to be 82.5% SS.

Design sensitivity and estimated backgrounds
While the measured performance of EXO-200 utilizing substantial low-background and calibration data sets will be the subject of a future paper, here we provide sensitivity figures assuming design parameters for the detector performance and background.Initial data taking roughly confirms the validity of such parameters.Using the expected energy resolution of s E /E = 1.6% at the 136 Xe end point, EXO-200 was designed to reach a sensitivity of T

EXO-200 Results
In 2011, the EXO collaboration published the first results from EXO-200, measuring 2νββ in 136 Xe for the first time [4].With just 31 live-days of data, the decay was measured with a half-life of (2.11 ± 0.04 stat ± 0.21 sys) × 10 21 years.This is the longest directly measured half-life of any radioisotope.
In 2012, the EXO collaboration published the first results for their 0νββ search [5].With 120.7 days livetime and a 98.5 kg fiducial volume, a maximum likelihood fit was performed simultaneously on the SS and MS spectra.The fitted spectra are shown in Figure 5, with the right plot zoomed into the region of interest.In the 1σ (2σ) region of interest (ROI), 1 (5) events were observed, with a fitted expected background of 4.1 ± 0.3 (7.5 ± 0.5) events.This corresponds to a 90% C.L. limit of T 0νββ 1/2 > 1.6 × 10 25 years.The fitted background corresponds to ∼ 60 counts/ ± 2σ ROI/140 kg/2 yr.This is not far from the design goal of 40 counts.The limit set on the 0νββ half-life corresponds to a Majorana neutrino mass of m ββ < 140 − 380 meV, depending on which nuclear matrix element is assumed.
U-dimension and 6 mm in z (drift time).The event sharing between SS and MS energy spectra is demonstrated for a 228 Th source in Fig. 3.The SS and MS spectra are compared to probability density functions (PDFs) generated by GEANT4 [15] simulations (MC).Simulated charge deposi-tions are transported to the wires erated using a model of the reado added and signals are reconstructe for the data.After events are design the total collected energy is con resolution function shown in Fig.
duces the shape of the 60 Co and 22 Fig. 3 for 228 Th).The MC reprod events, defined as N SS =ðN SS þ N M tion, for 228 Th (0) events th 70% (71%) efficiency for the requ fully reconstructed in 3D.The absolute, NIST-traceable, activity AE9:4%.These variations are use systematic uncertainties in the fi estimate for the 0 efficiency a broad energy range by compar ciency with low-background dat spectrum has vanishing statistical ciency is found to be a smooth fun agrees with the simulated efficien overall scale uncertainty mentione The fiducial volume used in this of 136 Xe (3:52 Â 10 26 atoms), corr active enr LXe.The trigger is fully The cut represented by the dashed eliminates a population of events d enr LXe region for which the charg FIG. 2. Top: uncertainty bands on the energy calibration residuals, using the full energy reconstruction described for the three sources.For both SS (solid line) and MS (dashed line) the position of the four lines is consistent with the calibration model within 0:1%.Bottom: energy resolution for various sources along with a fit to the empirical model (see text).

FIG. (color online). MS (top) and
tra from a 228 Th calibration run.T simulation (line) is fit to the data flo the MS and SS spectra independently with the rendered spectral shape.The l to reproduce the absolute source rate i is not illustrated by the figure. 1 (color).Correlation between ionization and scintillation for SS events from a 228 Th source.The energy resolution is considerably improved by forming the linear combination of both measurements.Events in the top-left quadrant are due to incomplete charge collection and are rejected by the cut (dashed line), removing only 0.5% of the total.The main peak at 2614 keV from 208 Tl decay can be seen, and it makes the anticorrelation between scintillation and ionization clear.We choose a "rotated" energy, a linear combination of the measured scintillation and ionization energies, which gives an improved energy resolution.The ionization-only and scintillation-only energy resolutions (σ E /E) at the 208 Tl peak are (for SS events) 3.5% and 6.0%, respectively [3], but this is improved to better than 1.8% with the "rotated" energy.The scintillation/ionization ratio can also be used to reject backgrounds such as alphas, which have a much larger scintillation fraction (the alpha events would be in the cut-off region above and to the left of this plot).
the necessary infrastructure for a barium tagging option will be included, even if barium tagging isn't ready when the data taking begins.
Figure 7 shows the projected sensitivity for EXO-200 and nEXO, as a function of the lightest neutrino mass.The present EXO-200 limit band is based on the published limit [5], and the "ultimate" sensitivity assumes radon removal (from the lead-cryostat air gap), improved analysis, and no signal in 4 years livetime.The "initial nEXO" band refers to a detector directly scaled from EXO-200, including its measured background and 10 year livetime.The "final nEXO" band refers to the same detector, but with no backgrounds other than 2νββ, as would be the case with barium tagging in use.The width of the horizontal bands is due to the uncertainty in nuclear matrix elements.The blue bands are 68% CL intervals from oscillation experiments for the normal and inverted hierarchies.Note that nEXO, with barium tagging, should be able to completely exclude the Majorana neutrino case for the inverted hierarchy.
Fig. 4. Primarily due to bremsstrahlung, a fraction of events are MS.The MC simulation predicts that 82.5% of 0 events are SS.Using a maximum likelihood estimator, the SS and MS spectra are simultaneously fit with PDFs of the 2 and 0 of 136 Xe along with PDFs of various backgrounds.Background models were developed for various components of the detector.Results of the material screen campaign, conducted during construction, such deviation measured with the source calibration spectra.The SS fractions forand -like events are also constrained in the fit to within AE8:5% of the MC predicted value.As a cross-check, the constraint on the 2 SS fraction is released in a separate fit of the low-background data.The SS fraction is found to agree within 5.8% of the value predicted by the MC simulation.
The energy scale is a free parameter in the fit, so that it is constrained by the 2 spectrum.The fit reports a scale factor of 0:995 AE 0:004.The uncertainty is inflated to AE0:006 as a result of an independent study of the possible

032505-4
tor, the SS and MS spectra are simultaneously fit with PDFs of the 2 and 0 of 136 Xe along with PDFs of various backgrounds.Background models were developed for various components of the detector.Results of the material screen campaign, conducted during construction, value.As a cross-check, the constraint on the 2 SS fraction is released in a separate fit of the low-background data.The SS fraction is found to agree within 5.8% of the value predicted by the MC simulation.
The energy scale is a free parameter in the fit, so that it is constrained by the 2 spectrum.The fit reports a scale factor of 0:995 AE 0:004.The uncertainty is inflated to AE0:006 as a result of an independent study of the possible a e-mail: joalbert@indiana.eduDOI: 10.1051/ C Owned by the authors, published by EDP Sciences, but double-beta decay (2νββ) may be allowed, 2n → 2p + 2e − + 2ν e .

2 beta
no-Less Double-Beta Decay only if neutrinos are (yes) and Majorana (?). a particles: are their own cles.s lepton number violation new physics).as peak at Q-value 3 Elliot, S. et al., Annu.Rev. Nucl.Part.Sci.2002.52:115-51 (t 1/2 ) can be relate to the effective Majorana mass (m ) that is a function of the light neutrino masses (m i ) ved by EXO-200.This was st NME ever directly (t 1/2 ) can be relate to the effective Majorana mass (m ) that is a function of the light neutrino masses (m i ) d by EXO-200.This was ME ever directly (simple 0νββ mechanism) (t 1/2 ) can be relate to the effective Majorana mass (m ) tha function of the light neutrino mas First observed by EXO-200.This was the smallest NME ever directly observed.(t 1/2 ) can be relat effective Majorana mass function of the light neutr First observed by EXO-200.This was the smallest NME ever directly observed.

Figure 1 .
Diagram of a simple mechanism for neutrinoless double-beta decay.Note that the neutrino can only be produced if it is equivalent to the anti-neutrino, that is, a Majorana fermion.If observed, 0νββ would be a lepton-number violating process, and require physics beyond the standard model.Observation would clearly indicate that neutrinos are Majorana particles, and, if the mass of the neutrino were known and sufficiently large, non-observation of 0νββ would clearly indicate that neutrinos are Dirac particles.The lifetime for 0νββ is related to the neutrino mass by

Figure 2 .
Figure 2. Photograph of the inside of the EXO-200 TPC, taken during assembly.This is one half of the TPC.The crossed induction wire plane and collection wire plane are visible above the APD plane.The copper field rings surround the cylindrical TPC, with white teflon reflector just inside of their radius.

Figure 2 .
Figure 2. Cutaway view of the EXO-200 setup, with the primary subassemblies identified.The outermost shielding layer, outside the outer vessel of the cryostat, consists of 25 cm of lead.The low-noise front end electronics are located outside of the lead shielding and are connected to the detector through thin polyimide cables.This choice trades some increased noise for the simplicity and accessibility of room temperature, conventional construction electronics.A cosmic-ray veto counter made of plastic scintillators surrounds the cleanroom housing the rest of the detector.EXO-200 is located at a depth of 1585 m water equivalent [21] in the Waste Isolation Pilot Plant (WIPP) near Carlsbad, New Mexico (32 22'30"N 103 47'34"W).

PRL 109 ,
032505 (2012) P H Y S I CA L R E V I E W L E T T E R S 032505-3

Figure 4 .
Figure 4. Plot of Thorium calibration data.The main peak at 2614 keV from208 Tl decay can be seen, and it makes the anticorrelation between scintillation and ionization clear.We choose a "rotated" energy, a linear combination of the measured scintillation and ionization energies, which gives an improved energy resolution.The ionization-only and scintillation-only energy resolutions (σ E /E) at the 208 Tl peak are (for SS events) 3.5% and 6.0%, respectively[3], but this is improved to better than 1.8% with the "rotated" energy.The scintillation/ionization ratio can also be used to reject backgrounds such as alphas, which have a much larger scintillation fraction (the alpha events would be in the cut-off region above and to the left of this plot).

FIG. 4 (
FIG. 4 (color).MS (top) and SS (bottom) energy spectra.The best-fit line (solid blue) is shown.The background components are 2 (grey region), 40 K (dotted orange), 60 Co (dotted dark blue), 222 Rn in the cryostat-lead air-gap (long-dashed green), 238 U in the TPC vessel (dotted black), 232 Th in the TPC vessel (dotted magenta), 214 Bi on the cathode (long-dashed cyan), 222 Rn outside of the field cage (dotted dark cyan), 222 Rn in active xenon (long-dashed brown), 135 Xe (long-dashed blue) and 54 Mn (dotted brown).The last bin on the right includes overflows (none in the SS spectrum).FIG. 5 (color).Energy spectra in the 136 Xe Q region for MS (top) and SS (bottom) events.The 1 ð2Þ regions around Q are shown by solid (dashed) vertical lines.The 0 PDF from the fit is not visible.The fit results have the same meaning as in Fig. 4.

FIG. 4 (
FIG. 4 (color).MS (top) and SS (bottom) energy spectra.The best-fit line (solid blue) is shown.The background components are 2 (grey region), 40 K (dotted orange), 60 Co (dotted dark blue), 222 Rn in the cryostat-lead air-gap (long-dashed green), 238 U in the TPC vessel (dotted black), 232 Th in the TPC vessel (dotted magenta), 214 Bi on the cathode (long-dashed cyan), 222 Rn outside of the field cage (dotted dark cyan), 222 Rn in active xenon (long-dashed brown), 135 Xe (long-dashed blue) and 54 Mn (dotted brown).The last bin on the right includes overflows (none in the SS spectrum).FIG. 5 (color).Energy spectra in the 136 Xe Q region for MS (top) and SS (bottom) events.The 1 ð2Þ regions around Q are shown by solid (dashed) vertical lines.The 0 PDF from the fit is not visible.The fit results have the same meaning as in Fig. 4.

Figure 5 .
Figure 5. Fits to the low background data from 120.7 days livetime on EXO-200.The left plot shows the full fitted spectrum on a logarithmic scale, and the right plot shows the spectra zoomed in near the region of interest.The 1σ (2σ) region of interest around the Q-value are marked by vertical solid (dashed) red lines.The MS (top) and SS (bottom) spectra are fit to simultaneously.Background components are 2νββ (grey region), 40 K (dotted orange), 60 Co (dotted dark blue), 222 Rn in the cryostat-lead air-gap (long-dashed green), 238 U in the TPC vessel (dotted black), 232 Th in the TPC vessel (dotted magenta), 214 Bi on the cathode (long-dashed cyan), 222 Rn outside of the field cage (dotted dark cyan), 222 Rn in active xenon (long-dashed brown), 135 Xe (long-dashed blue) and 54 Mn (dotted brown).

Figure 6 .Figure 7 .
Figure6.Drawing of a nEXO detector design.A water tank serves both as passive shielding from gamma rays, and as an active water Cherenkov muon veto.An area for barium tagging equipment is maintained above the water shield.