Dark Matter Searches in LUX Cláudio Pascoal Da Silva (LIP, Universidade De Coimbra) (On Behalf of the LIP-Coimbra Dark Matter Group) Coimbra, January 11Th, 2017

Dark Matter Searches in LUX Cláudio Pascoal da Silva (LIP, Universidade de Coimbra) (on behalf of the LIP-Coimbra Dark Matter Group) Coimbra, January 11th, 2017

mail: [email protected] 2

Outline

•Evidence for Cold Dark Matter •The LUX detector - two phase Liquid/ Gas time projection chamber o Direct dark matter detection o How LUX detector works •The LUX calibrations (Krypton, DD and Tritium) •332 live-days second science run (2014-2016) results o LUX backgrounds o Main dark matter search analysis o LUX dark matter search limits o LUX axion searches •The LZ experiment

Image: LUX inside the water tank (September 2012) Evidences for Cold Dark Matter 3

Rotation curves of galaxies2.2. Evidence for dark matter Cosmic Microwave Background

Planck

NGC-3198 Planck

Figure 2.2: Rotation curve of spiral galaxy NGC-3198. The sum of disk and expected dark matter halo contributions match the observations [14]. Galaxies Surveys Dwarf Galaxies between baryonic or dark origin. In fact, gravitational lensing was ﬁrst suggested by Adams, 2016 Zwicky as a viable technique to measure the mass distribution in our Universe [15]. We distinguish between three di↵erent classes of gravitational lensing: strong, weak and micro lensing. Strong lensing distorts the images of the lensed objects to great extent, resulting in clearly visible arcs and multiple images of the same source.Gravitational On the lensing contrary, micro-lensing imposes no visible distortion on the shape, but the amount of light detected from a background source changes over time. Bullet Cluster The weak lensing technique is based on the statistical analysis of numerous weakly lensed sources and is most commonly used for large sky surveys. When observing a preferred direction in the distortion of the intrinsic shape of captured galaxies, mass distributions inRed: the areaChandra may be X-Ray reconstructed. telescope. Recent advances in this technique, utilising the redshiftWhite: dependence Optical Hubble (higher redshiftTelescp galaxies experience stronger shear distortion), enableLight the blue: recovery Lensing of the Map full three-dimensional gravitational potential of the matter density, resolving large scale structures in both angle and time. This was achieved, for example, by studying the weak lensing data from the Hubble Space Telescope (HST)/Space Telescope A901/902 Galaxy Evolution Survey (STAGES) [16]. A very prominent example, demonstrating the presence of dark matter using the technique of weak gravitational lensing, is the observation of the Bullet cluster [17], a merger of two galaxy clusters. When the two clusters collided, the ﬂuid-like x-ray emitting hot gas or ICM was spatially separated from the visible stellar components, which simply passed through each other. However, the gravitational potential does not trace the ICM, the dominant baryonic mass fraction, but, rather approximately,

11 Evidences for Cold Dark Matter 4

Rotation curves of galaxies2.2. Evidence for dark matter Cosmic Microwave Background

Planck

NGC-3198 Planck Vera Rubin 1928/2016

11 PHYSICAL REVIEW LETTERS week ending 5 PRL 117, 201101Dark (2016) Matter or Modified Gravity?11 NOVEMBER 2016

] PRL 117, 201101 (2016) -2 TABLE I. Scatter budget for acceleration residuals. 2693 points in 153 galaxies Source Residual Rotation velocity errors 0.03 dex Disk inclination errors 0.05 dex MOND Galaxy distance errors 0.08 dex Variation in mass-to-light ratios 0.06 dex HI flux calibration errors 0.01 dex Total 0.12 dex

argued [41–43] to be the cause of the apparent discrepancy [44] with the predictions of numericalThe simulations Astrophysical[45] Journal. Kepler This cannot be the case here. If we had neglectedLetters, 648(2), some L109 important source of uncertainty,Yellow: we Chandra would X-Ray erroneously telescope. -2 infer a large intrinsic scatter, notBlue: a Optical small Hubble one. Telescope For the log(observed centripetal acceleration) [ms centripetal acceleration) log(observed log(predicted cent. acceleration) [ms ] intrinsic scatter to be non-negligible,White curves: the errorsLensing Map must be overestimated rather than underestimated. If there were no •All the evidences that we have for the existenceobservational uncertaintyof dark atmatter all, the intrinsic is gravitational scatter would still be limited by the small observed rms of 0.13 dex. •Modified Newtonian Dynamics (MOND) Regardless of whether the intrinsic scatter is zero or merely very small, the radial acceleration relation is an o Gravity switches from an inverse square law to importantone that empirical describes facet the of the rotation mass discrepancy curves problem.of galaxies. FIG.o The 3. analysis The centripetal of galaxy acceleration clusters observed (such in rotation as curves,the bulletWhen cluster)gbar is and observed, CMBgobs is follows,usually and used vice as versa. proof This g V2=R, is plotted against that predicted for the observed must be explained by any successful theory. obsagainst¼ MOND. distribution of baryons, gbar Φbar= R , in the upper panel. Discussion.—We find a strong relation between the o ¼ j∂ ∂ j NearlySymmetry 2700 individual argument: data points forgas 153 is SPARCat center, galaxies but are potentialobserved has radial two acceleration wells gobs and that due to the shown in a density map. The mean uncertainty on individual baryons, gbar. This radial acceleration relation is completely points is illustrated in the lower left-hand corner. Large squares empirical. It follows from a minimum of assumptions. The show the mean of binnedThey data. sit in Dashed caves, linesdeep show underground. the width of Surrounded the only by inputs lead, protected are the data,from thenoise, Poisson shielded equation, from the andwarmth the of ridge as measured bythe the Sun, rms they in each wait… bin.Maybe The dotted in 100 line years is the they’ll still sit in caves, deep underground. And wait. simplest possible conversion of starlight to stellar mass. line of unity. The solid line is the fit of Eq. (4) to the unbinned data using an orthogonal-distance-regression algorithm that We have not considered any particular halo model for considers errors on both variables. The inset shows the histogram the dark matter. Indeed, such models are unnecessary. The of all residuals and a Gaussian of width σ 0.11 dex. The distribution of dark matter follows directly from the ¼ residuals are shown as a function of gbar in the lower panel. The relation, and can be written entirely in terms of the baryons: error bars on the binned data are smaller than the size of the points. The solid lines show the scatter expected from gbar gDM gobs − gbar : 5 g =g observational uncertainties and galaxy-to-galaxy variation in ¼ ¼ ep bar † − 1 ð Þ the stellar mass-to-light ratio. This extrinsic scatter closely follows the observed rms scatter (dashed lines): the data are The dark and baryonic mass are stronglyffiffiffiffiffiffiffiffiffiffi coupled [13,14]. consistent with negligible intrinsic scatter. Possible interpretations for the radial acceleration relation fall into three broad categories: (1) it represents the end scatter in the relation must be smaller still once scatter due product of galaxy formation; (2) it represents new dark to errors is accounted for. sector physics that leads to the observed coupling; (3) it is There are two types of extrinsic scatter in the radial the result of new dynamical laws rather than dark matter. acceleration relation: measurement uncertainties and None of these options are entirely satisfactory. galaxy-to-galaxy variation in ϒ . Measurement uncertain- In the standard cosmological paradigm, galaxies form ⋆ ties in gobs follow from the error in the rotation velocities, within dark matter halos. Simulations of this process do not disk inclinations, and galaxy distances. The mean contri- naturally lead to realistic galaxies [44,46]. Complicated bution of each is given in Table I. Intrinsic scatter about the accessory effects (“feedback”) must be invoked to remodel mean mass-to-light ratio is anticipated to be 0.11 dex at simulated galaxies into something more akin to observa- 3.6 μm [24]. This propagates to a net residual of 0.06 dex in tions. Whether such processes can satisfactorily explain the gbar after accounting for the variable slope of the relation. radial acceleration relation and its small scatter remains to The total expected scatter is 0.12 dex (Table I), leaving little be demonstrated [47,48]. room for intrinsic scatter. Another possibility is new “dark sector” physics. The Astronomical data often suffer from unrecognized sys- dark matter needs to respond to the distribution of baryons tematics. In the case of rotation curves, this is frequently (or vice versa) in order to give the observed relation. This is 201101-4 The Astrophysical Journal,749:90(10pp),2012April10 Hlozek et al. Cold DM or 10Hot5 DM 6

104 ] Cold Dark Matter 3 Hot Dark Matter 103 •Masses > 100 GeV and low termal vel. •Masses ~ keV and high termal vel. (k,z=0) [Mpc

P 2 •Clumping of DM even at small scales •10No DM clumpingLyA (McDonald et al. 2006)even at small scales BCG Weak lensing (Tinker et al. 2011) •Dwarf galaxiesCCCP II (Vikhlinin and et al. 2009) satellite galaxies •Abundance of dwarf galaxies ACT Clusters (Sehgal et al. 2011) 10suppressed1 SDSS DR7 (Reid et al. 2010) ACT CMB Lensing (Das et al. 2011) ACT+WMAP spectrum (this work)

3 2 1 0 10− 10− 10− 10 1 Simulations k [Mpc− ]

101 Large Galaxies Cold Warm Hot Galaxy Clusters 100 Cold Dark Matter (250 eV) 1

M 10− / M ∆ 2 10− Warm Dark 3 10− Matter (1 eV)

Mass Variance LyA (McDonald et al. 2006) 10 4 Mass Variance − BCG Weak lensing (Tinker et al. 2011) CCCP II (Vikhlinin et al. 2009) ACT Clusters (Sehgal et al. 2011) 5

10− SDSS DR7 (Reid et al. 2010) Horizon Hubble ACT CMB Lensing (Das et al. 2011) ACT+WMAP spectrum (this work) 10 6 − 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 Mass scale M [Msolar] The Astrophysical Journal, 556:93-107 Figure 5 Reconstructed matter power spectrum: the stars showMass the power Scale spectrum from (Solar combining Mass) ACT and WMAP data (top panel). The solid and dashed lines show the nonlinear and linear power spectra, respectively, from the best-ﬁt ACT ΛCDM model with spectral index of ns 0.96 computed using CAMB and 1 1 = HALOFIT (Smith et al. 2003). The data points between 0.02 Mpc−

the matter power spectrum are in fact averages over a range measurements are shown by the leftmost and rightmost points of wavenumbers. For example, the ACT lensing constraint of the slanted error bars in the top panel of Figure 5. The is sensitive to a broad range of comoving distances between cluster constraints will vary according to the range of masses 2000 and 8000 Mpc (see Figure 2 of Sherwin et al. 2011), probed by each study (as the wavenumber kc is related to the with the lensing kernel spread peaked around ℓ 100. This mass Mc through Equation (5)). For the CCCP survey, the ≈ 14 15 translates into a range of wave numbers between 0.01 ! k ! range of masses is 1.35 10 M " Mc " 1.35 10 M 1 × ⊙ 1 × ⊙ 0.05. The upper and lower k bounds of the SDSS and Lyα or 0.24 Mpc−

8 Is DM a Standard Model Particle? 7 Is DM a Standard Model Particle? 8

Light particle - no mass

Couples with plasma (disproved by CMB and Big-bang nucleosynthesis) Unstable

Hot Dark Matter

No known particles from the SM model can be a Dark-Matter particle! Cold Dark Matter Particle Candidates 9

o WIMP’s (weakly interactive massive particles): o Stable and neutral in most scenarios; o Solves the Gauge Hierarchy Problem o WIMP Miracle (density from freeze out) o Physics beyond the standard model: o Super-symmetry - neutralino o Axions (solution to the strong CP problem - Peccei-Quinn solution) o nearly collisionless, neutral, non baryonic, provide the DM density; o Axions coming from the sun (light particle) XENON-100 o Axion Like Particles (ALPs) slowly moving in the galaxy o And many, many others… o superWIMPs, light gravitinos, branons, Sterile Neutrinos, Kaluza-Klein bosons … ΛCDM Empedocles World

Dark Energy

Cold Dark Matter

ΛCDM World 11 Indirect Detection

•Observation of the SM decay products of DM particles

•Neutrinos: DM sinks in Earth/Sun and AMS transform into neutrinos IceCUBE o IceCube, ANTARES •Cosmic Rays: DM decays into cosmic rays and gamma rays; o PAMELA, AMS, HESS, VERITAS, Fermi- LAT, HAWK, CTA •Astronomical uncertainties are signiﬁcant. LHC Production o How do we know a speciﬁc emission line is from dark matter and not from an unknown cosmological background?

Hess LHC12 Production

•DM particles or their “cousins” are produced in Collider Experiments •LHC o ATLAS, CMS ATLAS •May detect new particles, but is it Dark Matter?

Production CMS COUPP13 Direct Detection

•DM scatters off nuclei/electrons in a detector

Chalenges LNGS (Gran Sasso) •Isothermal model: expect recoil <10 keV requiring detectors with a very low threshold. •Challenge backgrounds - (Ambient radioactivity: ~100 evts/kg/s )

Solutions SNOLab •Careful selection of the target material DarkSide CMS •Go underground (cosmic protection) •Active and passive shield •Careful selection of the materials inside the detector

Boulay COUPP14 Direct Detection

•Possible detection of charge, Light, Heat, Phonons or combination of these signals •Different Target Materials:

o Germanium, Argon, Xenon, C3F8, C4F10, NaI(Tl)), etc ZEPLIN-III

Production DarkSide CMS Direct Detection 15

Our Signals

•WIMPs elastic scattering with the nucleus - contact operators from effective ﬁeld theory o Spin Independent (M) o Proportional to A2 o Spin Dependent (Σ’ and Σ’’) o Other effective ﬁeld interactions dependent on the linear and angular moment… Xe is good for most EFT •WIMPs inelastic scattering interactions with protons •Axions and ALPs couple with electrons through the axion electric effect; Direct Detection 16

Our Signals

•Annual and Diurnal Modulations o Observed in DAMA/Libra but not in xenon experiment

3σ 6 https://arxiv.org/pdf/1701.00769v1.pdf 7 † Wallenberg Academy Fellow ) 6 l Xenon (4 years2σ data) S 5 4 ‡ E-mail: [email protected] (T 3 §

∆ E-mail: [email protected]

- 2 1σ 1 XENON-100 ¶ Also with Coimbra Engineering Institute, Coimbra, Por- 40 60 80 100 120 140 160 180 200 220 )] tugal 12 12 Expected 1 2 3 σ σ σ day ⋅ ⇤⇤ E-mail: [email protected] DAMA/LIBRA 10 10 [1] R. Bernabei et al. (DAMA/LIBRA), Eur. Phys. J. C73, XENON100 tonne ⋅ expected 2648 (2013). 8 Run I-III 8 phase from [2] R. Agnese et al. (SuperCDMS), Phys. Rev. Lett. 112, Run II 241302 (2014). 6 DM halo 6 [3] D. S. Akerib et al. (LUX), arXiv:1608.07648.

4 4 [4] A. Tan et al.(PandaX-II),Phys.Rev.Lett.117,121303 (2016). 2 2 [5] E. Aprile et al.(XENON),Phys.Rev.D94,122001 (2016). 0 DAMA/Libra 9.2 σ C.L. Amplitude [events/(keV 0 40 60 80 100 120 140 160 180 200 220 2 4 6 8 1012 [6] C. Amole et al.(PICO),Phys.Rev.D93,061101(2016). Phase [Days] -∆(TSl) [7] P. Agnes et al. (DarkSide), Phys. Rev. D93,081101 FIG. 5. The XENON100 best-fit black dot, and 68% (light (2016). red shaded region) and 90% (green shaded region) confidence [8] J. Kopp, V. Niro, T. Schwetz and J. Zupan, level contours as a function of amplitude and phase relative Phys. Rev. D80,083502(2009). to January 1, 2011 for one year period. The corresponding [9] E. Aprile et al.(XENON),Science349,no.6250,851 Run II-only results [10] are overlaid with a black square and (2015) dotted lines. The phase is less constrained than in Run II [10] E. Aprile et al.(XENON),Phys.Rev.Lett.115,091302 due to the smaller amplitude. The expected DAMA/LIBRA (2015). signal (cross, statistical uncertainty only) and the phase ex- [11] K. Abe et al. (XMASS), Phys. Lett. B759,272(2016). pected from a standard DM halo (vertical dotted line) are [12] E. Aprile et al. (XENON100), Astropart. Phys. 35,573 (2012). shown for comparison. Top and side panels show (TSl)as a function of phase and amplitude, respectively, along with [13] E. Aprile et al.(XENON100),Phys.Rev.Lett.107, two-sided significance levels. 131302 (2011). [14] E. Aprile et al.(XENON100),Phys.Rev.Lett.109, 181301 (2012). [15] E. Aprile et al.(XENON100),Phys.Rev.Lett.111, Forschungsgemeinschaft, Max Planck Gesellschaft, Foun- 021301 (2013). dation for Fundamental Research on Matter, Weizmann [16] E. Aprile et al.(XENON100),Phys.Rev.D90,062009 Institute of Science, I-CORE, Initial Training Network (2014). Invisibles (Marie Curie Actions, PITNGA-2011-289442), [17] E. Aprile et al. (XENON100), Astropart. Phys. 54,11 Fundacao para a Ciencia e a Tecnologia, Region des Pays (2014). de la Loire, Knut and Alice Wallenberg Foundation, Is- [18] E. Aprile et al.(XENON100),Phys.Rev.D83,082001 (2011). tituto Nazionale di Fisica Nucleare, and the Laboratori [19] M. Weber, PhD Dissertation, University of Heidelberg Nazionali del Gran Sasso for hosting and supporting the (2013). www.ub.uni-heidelberg.de/archiv/15155. XENON project. [20] S. Lindemann and H. Simgen, Eur. Phys. J. C74,2746 (2014). [21] S. Lindemann, PhD Dissertation, University of Heidel- berg (2013). www.ub.uni-heidelberg.de/archiv/15725. [22] B.M. Roberts, V.A. Dzuba, V.V. Flambaum, ⇤ Also with Albert Einstein Center for Fundamental M. Pospelov and Y.V. Stadnik, Phys. Rev. D93, Physics, University of Bern, Switzerland 115037 (2016). Liquid Xenon TPC 17

•LUX is a direct detection experiment using a gas/liquid xenon time-projecting chamber •Energy depositions produce light and charge o Prompt scintillation (S1) o Proportional scintillation (S2): Measurement of the electrons extracted from the liquid to the gas •3D Position Reconstruction o Depth obtained from the time difference between S1 and S2 - called here drift time o XY reconstructed from the S2 light pattern Why a Liquid Xenon TPC? 18 LUX is a dual phase liquid/gas xenon time projecting chamber (TPC). Why?

•Xenon has a high atomic mass (A=131)High density (2.9 g/cm3) and High light yield •No intrinsic backgrounds (85Kr removed through chromatography) •TPCs scalable to multi-ton size (see LZ). •Ratio of charge to light (S2/S1) is a discriminator against backgrounds (>99%): o Nuclear Recoil (NR): WIMPs and neutrons interact with nuclei - short, dense tracks o Electronic Recoil (ER): axions, �s and e- interact with the electrons - longer, less dense tracks •Odd-neutron isotopes (129Xe, 131Xe) enable spin-dependent sensitivity studies. The LUX Experiment

•370 kg Liquid Xenon Detector (59 cm height, 49 cm diameter) o 250 kg in the active region (with ﬁeld) Construction materials chosen for low radioactivity (Ti, Cu, PTFE)

~4759 cm (cath.-surface) 49 cm

122 ultra low-background PMTs (61 on top, 61 on bottom) Active region defined by PTFE observe both S1 and S2 19 reflectors (high reflectivity >97%) - high light collection) Typical LUX Pulses

We analyze data from single-site energy deposits that look like this…

Top

S1 S2 phe/

Bottom PMTs Time (samples; 1 sample = 10 ns)

Waveform GUI and plotting: Daniel Typical LUX Pulses

S1 - Prompt scintillation •Sharp rise with a exponential decay o Pulse FWHM: ~100 ns o S1 area: 1-50 phd for WIMP search •~60-90% of light in bottom PMTs o Ratio depends on the depth of the event •Threshold of 2 detected photons 1.5 μs

S2 - Electroluminescence •Near-gaussian pulse shape o pulse width depends on the depth •~25 phd per extracted electron •~57/43% (top/bot.) light in PMTs •Threshold of 150-200 phd 5.5 μs LUX AT SURF 22 (Sanford Underground Research Facility)

•Sanford Underground Research Facility Lead, South Dakota, USA. •Former Home of the Homestake Solar Neutrino Experiment 1970-1994 •1478 m deep (4300 m.w.e.) • μ ﬂux reduced x10-7 compared to Raymond Davis sea level) (Nobelpriset i fysik 2002) The LUX Collaboration 23 Timeline 2006-2016 24

Two main WIMP search runs First Science Run (FSR): 2013/04-2013/09, 95 live-days Second Science Run (SSR): 2014/09-2016/09, 332 live-days

2008: LUX 2013 (Apr): 2014 (Jan): 2016 (May): funded First Science Run Grid SSR ﬁnished - (DOE+NSF) (FSR) starts conditioning 332 live-days 2014 (Sep): 2016 (Sep): 2010: Second Science Run LUX LIP Joined (SSR) Starts decommission

2006: 2012 (Jul): 2013 (Nov): 2015 (Dec): 2016 (Jul): week ending LUX collab. LUX moves First Science Run PRL 116, 1613013-month (2016) runPHYSICAL REVIEWSSR LETTERS results released22 APRIL 2016 105 week ending 10−40 formed underground PRL 112,results091303 (2014) reportedPHYSICAL REVIEW LETTERSreanalysis7 MARCH posted 2014 Brown University332 thank thelive-days UK Royal Society for travel

) funds under the International Exchange Scheme (IE120804). 2 LUX will10 continue4 operations at SURF during 2014 10−41 The UK groups acknowledge institutional support from and 2015. Further engineering and calibrationCDMSlite 2015 studies will

) Imperial College London, University College London and 2 establish the optimal3 parameters for detector operations, −42 10 10 with potential improvements in applied electric fields, Edinburgh University, and from the Science and Technology Facilities Council for PhD Studentships No. ST/K502042/1 −44 increased calibration statistics, decayingSuperCDMS backgrounds 2014 10 2 −43 and an instrumented10 water tank veto further enhancing 10 (AB), No. ST/K502406/1 (SS), and No. ST/M503538/1 PandaX 2015 the sensitivity of the experiment. Subsequently, we will −50 2015 (KY). The University of Edinburgh is a charitable body, 6 8 10 12 complete the10 ultimate1 goal of conducting a blindedDarkSide 300 10−44 registered in Scotland, with Registration No. SC005336. This nucleon cross section (zb)

XENON100 2012 nucleon cross section (cm − −40 live-day WIMP search8 further improving sensitivity to − 10 B LUX 2014 research was conducted using computational resources and nucleon cross section (cm − explore significant0 new regions of WIMP parameterThis Result −45 services at the Center for Computation and Visualization, −45 10 10

10 space. WIMP −42 WIMP

WIMP 10 Brown University. We gratefully acknowledge the logistical This work10−1 was partially supported by the U.S. and technical support and the access to laboratory infra- 1 2 3 −44 Department of Energy (DOE) under Awards No. DE- 10 10 2 10 10 structure provided to us by the Sanford Underground 1 2 3 m (GeV/c ) 10 10 10 FG02-08ER41549, No. DE-FG02-91ER40688,WIMP No. DE- Research Facility (SURF) and its personnel at Lead, m (GeV/c2) FG02-95ER40917, No. DE-FG02-91ER40674, WIMP South Dakota. SURF was developed by the South Dakota No. DE-NA0000979,FIG. 3. Upper No. DE-FG02-11ER41738, limits on the spin-independent No. DE- elastic WIMP- FIG. 5 (color online). The LUX 90% confidence limit on the SC0006605,nucleon No. cross DE-AC02-05CH11231, section at 90% C.L. No. Observed DE-AC52- limit in black, with Science and Technology Authority, with an important spin-independent elastic WIMP-nucleon cross section (blue), 07NA27344,the 1- and and No.2-σ ranges DE-FG01-91ER40618; of background-only the trials U.S. shaded green and philanthropic donation from T. Denny Sanford, and is PRL,together 112, with the 1σ variation091303 from repeated trials,2014 where trials PRL, 116, 161301 2016 arXiv 1608.07648 Æ Nationalyellow. Science Also Foundation shown are limits under from Awards the first No. LUX analysis [6] operated by Lawrence Berkeley National Laboratory for fluctuating below the expected number of events for zero BG are PHYS-0750671, No. PHY-0801536, No. PHY-1004661, forced to 2.3 (blue shaded). We also show Edelweiss II [45] (dark (gray), SuperCDMS [40] (green), CDMSlite [41] (light blue), the Department of Energy, Office of High Energy Physics. yellow line), CDMS II [46] (green line), ZEPLIN-III [47] No. PHY-1102470,XENON100 No.[42] PHY-1003660,(red), DarkSide-50 No. PHY-1312561,[43] (orange), and PandaX We thank Felix Kahlhoefer and Sebastien Wild for uncov- (magenta line), CDMSlite [48] (dark green line), XENON10 No. PHY-1347449;[44] (purple). the The Research expected Corporation spectrum of Grant coherent neutrino- ering a mistake in a preprint of this work. No. RA0350; the Center for Ultra-low Background S2-only [20] (brown line), SIMPLE [49] (light blue line), and nucleus scattering by 8B solar neutrinos can be fit by a WIMP Experiments in the Dakotas (CUBED); and the South XENON100 225 live-day [50] (red line) results. The inset (same model as in [45], plotted here as a black dot. axis units) also shows the regions measured from annual Dakota School of Mines and Technology (SDSMT). modulation in CoGeNT [51] (light red, shaded), along with LIP-Coimbra acknowledges funding from Fundação para *Corresponding author. exclusion limits from low threshold re-analysis of CDMS II data a Ciência e Tecnologia (FCT) through the Project-Grant [email protected] [52] (upper green line), 95% allowed region from CDMS II Letter but are reviewed in, e.g., [38]. Limits on spin- No. CERN/FP/123610/2011. Imperial College and Brown [1] J. I. Read, The local dark matter density, J. Phys. G 41, silicon detectors [53] (green shaded) and centroid (green x), 90% University thank the UK Royal Society for travel funds dependent cross sections are presented elsewhere [39]. 063101 (2014). allowed region from CRESST II [54] (yellow shaded) and under the International Exchange Scheme (IE120804). The In conclusion, reanalysis of the 2013 LUX data has [2] D. Harvey, R. Massey, T. Kitching, A. Taylor, and E. Tittley, DAMA/LIBRA allowed region [55] interpreted by [56] (grey UK groups acknowledge institutional support from shaded). (results sourced from DMTools [57]). excluded new WIMP parameter space. The added fiducial The nongravitational interactions of dark matter in colliding Imperial College London, University College London, mass and live time, and better resolution of light and charge and Edinburgh University, and from the Science & galaxy clusters, Science 347, 1462 (2015). WIMP masses, from 2.4 to 5.3. A variation of one standard Technologyyield Facilities a 23% Council improvement for Ph. in D. sensitivity studentship at high WIMP [3] Ade et al. (Planck Collaboration), Planck 2015 results. XIII. deviation in detection efficiency shifts the limit by an No. ST/K502042/1masses over (AB). the The first University LUX result. of Edinburgh The reduced, 1.1 keV Cosmological parameters, arXiv:1502.01589v2. average of only 5%. The systematic uncertainty in the is a charitablecutoff body, in registeredthe signal in modelScotland, improves with registration sensitivity by 2% at [4] M. W. Goodman and E. Witten, Detectability of certain position of the NR band was estimated by averaging the No. SC005336.high masses This research but is thewas dominant conducted effectusing com- below 20 GeV c−2, dark-matter candidates, Phys. Rev. D 31, 3059 (1985). difference between the centroids of simulated and observed putationaland resources the range and 5.2 services to 3.3 GeV at thec−2 Centeris newly for demonstrated to [5] J. L. Feng, Dark matter candidates from particle physics and AmBe data in log S2b=S1 . This yielded an uncertainty of Computation and Visualization, Brown University. We methods of detection, Annu. Rev. Astron. Astrophys. 48, ð Þ be detectable in xenon. These techniques further enhance 0.044 in the centroid, which propagates to a maximum acknowledge the work of the following engineers who 495 (2010). uncertainty of 25% in the high mass limit. played importantthe prospects rolesduring for discovery the design, in the construction, ongoing 300-day LUX [6] D. S. Akerib et al. (LUX Collaboration), First Results from The 90% upper C.L. cross sections for spin-independent commissioning,search and and operation the future phases LUX-ZEPLIN of LUX: S. Dardin[46] experiment. the LUX Dark Matter Experiment at the Sanford Under- WIMP models are thus shown in Fig. 5 with a minimum from Berkeley, B. Holbrook, R. Gerhard, and J. Thomson ground Research Facility, Phys. Rev. Lett. 112, 091303 46 2 This work was partially supported by the U.S. Department cross section of 7.6 × 10− cm for a WIMP mass of from University of California, Davis; and G. Mok, J. Bauer, (2014). 33 GeV=c2. This represents a significant improvement over and D.of Carr Energy from (DOE) Lawrence under Awards Livermore No. DE-FG02-08ER41549,National [7] D. S. Akerib et al. (LUX Collaboration), The large under- the sensitivities of earlier searches [46,47,50,51]. The low Laboratory.No. We DE-FG02-91ER40688, gratefully acknowledge the No. logistical DE-FG02-95ER40917, and ground xenon (LUX) experiment, Nucl. Instrum. Methods energy threshold of LUX permits direct testing of low technicalNo. support DE-FG02-91ER40674, and the access to laboratory No. infrastructure DE-NA0000979, No. DE- Phys. Res., Sect. A 704, 111 (2013). mass WIMP hypotheses where there are potential providedFG02-11ER41738, to us by the Sanford No. Underground DE-SC0006605, Research No. DE-AC02- [8] C. H. Faham, V. M. Gehman, A. Currie, A. Dobi, P. hints of signal [46,51,54,55]. These results do not Facility05CH11231, (SURF) and itsNo. personnel DE-AC52-07NA27344, at Lead, South and No. DE- Sorensen, and R. J. Gaitskell, Measurements of wave- support such hypotheses based on spin-independent iso- Dakota.FG01-91ER40618; SURF was developed the by U.S. the National South Dakota Science Foundation length-dependent double photoelectron emission from spin-invariant WIMP-nucleon couplings and conventional Science and Technology authority, with an important single photons in VUV-sensitive photomultiplier tubes, astrophysical assumptions for the WIMP halo, even philanthropicunder donation Grant from No. T. PHYS-0750671, Denny Sanford, and No. is PHY-0801536, J. Instrum. 10, P09010 (2015). when using a conservative interpretation of the existing operatedNo. by Lawrence PHY-1004661, Berkeley No. National PHY-1102470, Laboratory No. for PHY-1003660, [9] C. H. Faham, Ph.D. thesis, Brown University, 2014. low-energy nuclear recoil calibration data for xenon the DepartmentNo. PHY-1312561, of Energy, Office No. of PHY-1347449; High Energy the Research [10] V. Solovov et al. (ZEPLIN-III Collaboration), Position detectors. Physics. Corporation Grant No. RA0350; the Center for Ultra-low reconstruction in a dual phase xenon scintillation detector, Background Experiments in the Dakotas (CUBED); and the IEEE Trans. Nucl. Sci. 59, 3286 (2012). 091303-6 South Dakota School of Mines and Technology (SDSMT). [11] Y. Mei, Ph.D. thesis, Rice University, 2011. LIP-Coimbra acknowledges funding from Fundação para a [12] D. S. Akerib et al. (LUX Collaboration), Calibration, event Ciência e a Tecnologia (FCT) through the Project-Grant reconstruction, data analysis and limits calculation for the No. PTDC/FIS-NUC/1525/2014. Imperial College and LUX dark matter experiment (to be published).

161301-6 Krypton Calibrations 25 Krypton data 83mKr •83mKr is an internal source. It is injected (Drift Time 4 - 8 !s) in the gas system and decays uniformly Second Science Run inside the detector 83m o Kr emits two gamma rays Eγ,1 = 32.2 keV (T1/2 = 1.83 h) and Eγ,2 = 9.4 keV (T1/2 = 154 ns) o 1 to 2 times a week •83mKr used to o Develop S1 and S2 position corrections: both S1 and S2 pulses depend on the location of the event due to geometrical light collection and electronegative impurities. o Map variations of the electric ﬁeld in the detector o Develop and test the position reconstruction: krypton data is used to get the light response functions (LRFs)of the PMTs. These functions are found by iteratively ﬁtting the distribution of S2 signal for each PMT.

The large difference between the drift field (180 V/cm) and the extraction field (2.8 kV/cm in liquid) causes the the drift field lines to be compressed as they pass through the gate plane; any electrons leaving the drift volume appear only in narrow strips between each pair of gate wires. ER Calibrations 26 •Tritium is an ideal source for determination of the detector’s electron recoil band and low energy threshold o E(max) - 18.6 keV, - 5.9 keV

o β decay with T(1/2) = 12.6 a - Long Lifetime •Tritiated methane was injected in the gas system and removed by the getter. Charge •ER calibrations performed every three months

Light Light Yield (ph/keV) Charge Yield (e-/keV) Yield (ph/keV) Charge Yield Light

Phys. Rev. D 93, 072009 (2016) NR Calibration 27 •Deuterium-Deuterium neutron Generator installed outside LUX water tank Charge •The 2.45 MeV emitted neutrons are collimated to the level of ~1 degree •Two analysis are performed

o Double-scatters - ionization yield Qy (0.7 to 74 keVnr)

o Single-scatters - scintillation yield Ly and NR band calibration (1.1 to 74 keVnr) •Calibrations every three months and at different depths during the SSR

Light

Efﬁciency

https://arxiv.org/abs/1608.05381 ER/NR calibrations 28 4 ER Calibration 10.2 7.8 9.0 6.6 From FSR! 3.5

(S2) 3 10 log 5.4 4.2 2.5 2.9 1.7 0.5 keV ee S2=165 phd

2 0 5 10 15 20 25 30 35 40 45 50 S1 detected photons 4 NR Calibration 51.0 39.0 45.0 33.0 3.5 27.0 21.0

(S2) 3 10 log

2.5 15.0 9.0 S2=165 phd 3.0 keV nr 2 0 5 10 15 20 25 30 35 40 45 50 S1 detected photons Wall-surface backgrounds 29 •238U late chain plate-out on PTFE surfaces survives as 210Pb and its daughters (mainly 210Bi and 210Po). •Betas and 206Pb recoils travel negligible distance, but they can be reconstructed some distance from the wall as a result of position resolution (especially for small S2s). •These sources define the position of the wall in measured coordinates, for the 4 data bins and any combination of drift-time & ϕ. •The boundary of the fiducial volume is defined at 3 cm from the observed wall in S2 space.

Fiducial volume

Measured wall position Fiducial boundary Data Wall Model Wall Surface Model

1000 S2 ∈ [200, 500] phd S1 ∈ [ 55, 500] phd 100 How to predict the number of wall 10 Events 1 1000 events leaking into ﬁducial volume? S2 ∈ [500, 900] phd S1 ∈ [ 55, 500] phd 100 10 Events 1 Pwall N S ,S ,φ,z,DB ∗ PPrr SS ,rawφ,z,DB,dwall 1000 = ( 1 2 ) (( 22 )) S2 ∈ [900, 1300] phd S1 ∈ [ 55, 500] phd 100 10 Events Number of Events Radial Model 1

1000 S2 ∈ [1300, 1800] phd S1 ∈ [ 55, 500] phd 100 •Leakage depends on the number of events, N, 10 Events 1 and how they distribute around the wall, Pr. 1000 S2 ∈ [1800, 3000] phd S1 ∈ [ 55, 500] phd 100 •To get N we used our WIMP search data, but 10 Events with the radial position of the events larger 1 −5 −4 −3 −2 −1 0 1 2 3 4 5 than the radial position of the wall - events Distance to the wall (cm) leaking towards the wall. Into the detector Out the detector •To get Pr our strategy is: Wall PDFS o Describe the events as function of the ratio Wall PDF Proj log(S2) vs S1 Wall PDF Proj drift time vs r )

c 4 50 10-1 (S2

between the distance to the wall and the 10 Rel Prob 10-2 Rel Prob log 10-5 drift time [us] 100 uncertainty called pulls g 3.5 10-3

10-4 150 10-5 3 -6 10 d -6 Wall 10 200 g = -7 10

2.5 -8 r 250 10 10-7 10-9

2 300 10-10 o 10 20 30 40 50 0 5 10 15 20 Use the data with S1>50 phd to get systematics. S1c in detected photons r [cm] Estimation of Backgrounds 31

Expected number Background source below NR median Bulk volume, but leakage External Gamma Rays 1.51±0.19 at all energies Internal Betas 1.20±0.06 Low-energy, but conﬁned to 238U late chain wall back. 8.7±3.5 the edge of our ﬁducial volume Accidental S1-S2 0.34±0.10 In the bulk volume, low- energy, in the Solar 8B neutrinos 0.15±0.02 NR band

•These ﬁgures are ﬁgure of merit only. In our analysis we use a likelihood analysis. o + ~ 0.3 single scatter neutrons, e.g. from (α, n), not included in PLR S2 Quality Cuts and Efficiency 32

•After single-scatter event selection (1 S1 + 1 S2), fiducializing, and cutting out periods of detector instability, few pathologies remain. These pathologies are targeted applying cuts to the S2: Efficiencies calculated with o Position reconstruction χ2 cut analysis the PMT hit calibration data! pattern. It removes single-electron pile-up, additional multiple scatters (x,y separated) and PMT after-pulsing. o S2 waveform cuts to remove misclassified gas events and merged S2s from multiple scatters (z separated).

Position Reconstruction χ2 Cut S2 Gaussian Fit Width S2 Gaussian Sigma Cut LUX Likelihood Analysis 33

30 25

Events 20 •A proﬁle-likelihood test (PRL) was 15 10 5 implemented to compare the models 0 5 10 15 20 25 30 35 40 45 50

S1c (phd)

with the observed data 100 80 Events 60 •5 un-binned PLR dimensions 40 20 0 o 2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4 z/drift time, r, ϕ, S1 and log10(S2) log(S2 ) c 100 •1 binned PLR dimension: 80

Events 60 o 40 Event date 20 00 2 4 6 8 10 12 14 16 18 20 •Data in the upper-half of the ER band r (cm)

140 were compared to the model (plot at 120

Events 100 80 60 right) to assess goodness of ﬁt. 40 20 0 50 100 150 200 250 300 •For each detector S1 and S2 are modeled drift time (µs)

with NEST (Noble Element Simulation 250 200 Events 150 Technique, NEST, http://www.albany.edu/ 100 50 0 physics/NEST.shtml) −3 −2 −1 0 1 2 3 phi (radian) •Good agreement with background-only model, p-value >0.6 for each projection. Observed data Background only model ER/NR Bands - SSR (2014/2016) 34 4

3.8

3.6

3.4

3.2

[S2 (phd)] 3 10 log 2.8 ER Band Mean 2.6 NR Band Mean ER Band 80% contours 2.4 NR Band 80% contours

2.2 0 10 20 30 40 50 S1 (phd) LUX 2014/2016 Detector’s Response 35 4

3.8

3.6

3.4

3.2

[S2 (phd)] 3 10 log 2.8 ER Band Mean 2.6 NR Band Mean ER Band 80% contours 2.4 NR Band 80% contours

2.2 0 10 20 30 40 50 S1 (phd) LUX 2014/2016 Detector’s Response 36

4 9.8 keVee 8.7 7.5 3.8 6.3 5.2 4 3.6 2.9

3.4 1.7 3.2

[S2 (phd)] 3 10 log 2.8 33 keVnr 27 2.6 21

2.4 15 Energy contours (constant number of quanta) 2.2 3 9 0 10 20 30 40 50 S1 (phd) LUX Blind Analysis 37

4 9.8 keVee 8.7 7.5 3.8 6.3 5.2 4 3.6 2.9

3.4 1.7 A common approach is to blind oneself to 3.2 events in the signal regions but it often blinds us to rare backgrounds and [S2 (phd)] 3 10 pathologies log 2.8 33 keVnr 27 2.6 21 Instead of traditional blinding, we 2.4 15 employ a technique where fake signal events (“salt”) are injected into data stream. NOT SIM!! 2.2 3 9 0 10 20 30 40 50 S1 (phd) WIMP-search data from 332 live days 38

4 9.8 keVee 8.7 7.5 3.8 6.3 5.2 4 3.6 2.9

3.4 1.7 3.2

[S2 (phd)] 3 10 log 2.8 33 keVnr 27 2.6 21

15 2.4 Black Dots: Second Science Run Data (Salt Included) Open circles: High radius events 2.2 3 9 0 10 20 30 40 50 S1 (phd) WIMP-search data - Salt Identified 39

4 9.8 keVee 8.7 7.5 3.8 6.3 5.2 4 3.6 2.9

3.4 1.7 3.2

[S2 (phd)] 3 10 log 2.8 33 keVnr 27 2.6 21

2.4 15 Salt - blue dots 2.2 3 9 0 10 20 30 40 50 S1 (phd) WIMP-search data - S1 Pathologies 40 4 9.8 keVee 8.7 7.5 3.8 6.3 5.2 4 3.6 2.9

3.4 1.7 B 3.2 C

[S2 (phd)] 3 10 log 2.8 A 33 keVnr 27

2.6 red dots: 2 populations of rare 21 pathological events were identified 2.4 contributing 3 sub-NR-band15 events

2.2 3 9 0 10 20 30 40 50 S1 (phd) WIMP-search data - S1 Pathologies 41 top PMT array A phe / sample

B phe / sample

C phe / sample

phe samples (10 ns) S1 Quality Cuts 42 •Two additional cuts on the S1 pulse were implemented. •Flat signal acceptance of 98.5% when both cuts are applied to the DD and Tritium data

Density map: CH3T calibration data x : WS data passing S1 cuts : WS data cut by S1 Max. PMT Area cut x : WS data cut by S1 Prompt Fraction cut

Removes events with S1 that has gas- Removes events with S1 light overly event-like time structure concentrated in a single PMT WIMP-search Final Data 43 4 9.8 keVee 8.7 7.5 3.8 6.3 5.2 4 3.6 2.9

3.4 1.7 3.2

[S2 (phd)] 3 10 log 2.8 33 keVnr 27 2.6 21 Pathological events removed p-value = 40% consistent with 2.4 15 background-only hypothesis (includes also information on r, z, �) 2.2 3 9 0 10 20 30 40 50 S1 (phd) First Science Run - 95 Live Daysweek ending 44 PRL 116, 161301 (2016) PHYSICAL REVIEW LETTERS 22 APRIL 2016

5 4 was tuned to the S1-S2 distribution of 1.8 × 10 fiducial- 10.2 keVee 7.8 9.0 volume electron recoils from the internal tritium source. 6.6 Good agreement was obtained from threshold to the 5.4 3.5 4.2 18.6 keV end point, well above the WIMP signal in both 2.9 light and charge, and the reconstructed β spectrum validates 1.7 the g1 and g2 values measured with line sources [16]. 3 33 keVnr

Simulated waveforms, processed with the same data- [S2(phd)] reduction software and event selection as applied to the 10 log search data, are used to model the ER backgrounds in S1 and S2. 2.5

Events due to detector component radioactivity, both 3915 21 27 within and above the energy region of interest, were 2 simulated with LUXSim. The high-energy spectral agree- 0 5 10 15 20 25 30 35 40 45 50 ment between data and simulation based on γ screening is S1 (phd) generally good [20,28]; however, we observe an excess of ER events with 500–1500 keV energy concentrated in the FIG. 2. Observed events in the 2013 LUX exposure of 95 live lowest 10 cm of the active region. Its precise origin is days and 145 kg fiducial mass. Points at <18 cm radius are black; those at 18–20 cm are gray. Distributions of uniform-in-energy unknown but the spectrum can be reproduced by simulating −2 238 232 electron recoils (blue) and an example 50 GeV c WIMP signal additional, heavily downscattered U chain, Th chain, (red) are indicated by 50th (solid), 10th, and 90th (dashed) 60 and Co γ rays in the center of a large copper block below percentiles of S2 at given S1. Gray lines, with ER scale of keVee the PMTs. This implies an extra 105 low-energy Compton- at top and Lindhard-model NR scale of keVnr at bottom, are scatter events, included in the background model. The γ-ray contours of the linear combined S1-and-S2 energy estimator [19]. population is subdivided into two spatial distributions with floating normalization: one generated by the bottom PMT array, its support structure, and the bottom γ-ray shield; and aftermath of photoionization and delayed electron emission one from the rest of the detector. following large S2s. The threshold is set for >99% tritium A final source of background, newly modeled here, is the acceptance and removes 1% of gross live time [34].We tail in reconstructed r of events on the PTFE sidewalls. The report on 95.0 live days. Figure 2 shows the measured light S1-S2 distribution of background events on the walls and charge of the 591 surviving events in the fiducial differs from that in the liquid bulk. Charge collection is volume. incomplete, so the ER population extends to lower values A double-sided, profile-likelihood-ratio (PLR) statistic of S2. There are, in addition, true nuclear recoils from the [35] is employed to test signal hypotheses. For each WIMP daughter 206Pb nuclei of α decay by 210Po plated on the mass, we scan over cross section to construct a 90% con- wall. The leakage of wall events towards smaller r depends fidence interval, with test statistic distributions evaluated by strongly, via position resolution, on S2 size. The wall Monte Carlo sampling using the RooStats package [36].At population in the fiducial volume thus appears close to the all masses, the maximum-likelihood value of σn is found to S2 threshold, largely below the signal population in S2 be zero. The background-only model gives a good fit to the at given S1. It is modeled empirically using high-r and data, with KS test p values of 0.05, 0.07, 0.34, and 0.64 for low-S2 sidebands in the search data [33]. the projected distributions in S1, S2, r, and z respectively. Systematic uncertainties in background rates are treated Upper limits on cross section for WIMP masses from via nuisance parameters in the likelihood: their constraints 4 to 1000 GeV c−2 are shown in Fig. 3; above, the limit are listed with other fit parameters in Table I. S1, S2, z, and increases in proportion to mass until 108GeV c−2, 106 zb, r are each useful discriminants against backgrounds, and where the Earth begins to attenuate≳ the WIMP flux. The cross sections are tested via the likelihood of the search raw PLR result lies between one and two Gaussian σ below events in these four observables. the expected limit from background trials. We apply a Search data were acquired between April 24th and power constraint [37] at the median so as not to exclude September 1st, 2013. Two classes of cuts based on cross sections for which sensitivity is low through chance prevailing detector conditions assure well-measured events background fluctuation. We include systematic uncertain- in both low-energy calibration and WIMP-search samples. ties in the nuclear recoil response in the PLR, which has a Firstly, data taken during excursions in macroscopic modest effect on the limit with respect to assuming the best- detector properties, such as xenon circulation outages or fit model exactly: less than 20% at all masses. Limits instability of applied high voltage, are removed, constitut- calculated with the alternate, Bezrukov parametrization ing 0.8% of gross live time. Secondly, an upper threshold is would be 0.48, 1.02, and 1.05 times the reported ones at 4, imposed on summed pulse area during the event window 33, and 1000 GeV c−2, respectively. Uncertainties in the but outside S1 and S2. It removes triggers during the assumed dark matter halo are beyond the scope of this

161301-5 FSR Reanalysis - 95 Live Days 45 Old Limits from the 2015 Spin Dependent reanalysis of 2013 data (FSR) week ending PRL 116, 161301Spin (2016) IndependentPHYSICAL REVIEW LETTERS 22 APRIL 2016

5 −40 10 10 Brown University thank the UK Royal Society for travel

) funds under the International Exchange Scheme (IE120804). 2 104 10−41 The UK groups acknowledge institutional support from CDMSlite 2015 Imperial College London, University College London and 103 10−42 Edinburgh University, and from the Science and Technology

SuperCDMS 2014 FacilitiesWIMP-Neutron Council for PhD Studentships No. ST/K502042/1 102 10−43 (AB), No. ST/K502406/1 (SS), and No. ST/M503538/1 PandaX 2015 −50 2015 (KY). The University of Edinburgh is a charitable body, 101 DarkSide 10−44 registered in Scotland, with Registration No. SC005336. This nucleon cross section (zb)

XENON100 2012 nucleon cross section (cm − 8B LUX 2014 − research was conducted using computational resources and This Result 100 10−45 services at the Center for Computation and Visualization, WIMP WIMP Brown University. We gratefully acknowledge the logistical 10−1 and technical support and the access to laboratory infra- 101 102 103 structure provided to us by the Sanford Underground m (GeV/c 2 ) WIMP Research Facility (SURF) and its personnel at Lead, South Dakota. SURF was developed by the South Dakota FIG. 3. Upper limits on the spin-independent elastic WIMP- Science and Technology Authority, with an important nucleon cross sectionPRL, at 90%116, C.L. 161301 Observed (2016) limit in black, with WIMP-Proton the 1- and 2-σ ranges of background-only trials shaded green and philanthropic donation from T. Denny Sanford, and is yellow. Also shown0.60 are limits zeptobarns from the first LUX analysis [6] operated by Lawrence Berkeley National Laboratory for (gray), SuperCDMS (at[40] 33(green), GeV/c CDMSlite2) [41] (light blue), the Department of Energy,PRL, Office116, of16130 High2 Energy (2016) Physics. XENON100 [42] (red), DarkSide-50 [43] (orange), and PandaX We thank Felix Kahlhoefer and Sebastien Wild for uncov- [44] (purple). The expected spectrum of coherent neutrino- ering a mistake in a preprint of this work. nucleus scattering by 8B solar neutrinos can be fit by a WIMP model as in [45], plotted here as a black dot. *Corresponding author. [email protected] Letter but are reviewed in, e.g., [38]. Limits on spin- [1] J. I. Read, The local dark matter density, J. Phys. G 41, dependent cross sections are presented elsewhere [39]. 063101 (2014). In conclusion, reanalysis of the 2013 LUX data has [2] D. Harvey, R. Massey, T. Kitching, A. Taylor, and E. Tittley, excluded new WIMP parameter space. The added fiducial The nongravitational interactions of dark matter in colliding mass and live time, and better resolution of light and charge galaxy clusters, Science 347, 1462 (2015). yield a 23% improvement in sensitivity at high WIMP [3] Ade et al. (Planck Collaboration), Planck 2015 results. XIII. masses over the first LUX result. The reduced, 1.1 keV Cosmological parameters, arXiv:1502.01589v2. cutoff in the signal model improves sensitivity by 2% at [4] M. W. Goodman and E. Witten, Detectability of certain high masses but is the dominant effect below 20 GeV c−2, dark-matter candidates, Phys. Rev. D 31, 3059 (1985). and the range 5.2 to 3.3 GeV c−2 is newly demonstrated to [5] J. L. Feng, Dark matter candidates from particle physics and methods of detection, Annu. Rev. Astron. Astrophys. 48, be detectable in xenon. These techniques further enhance 495 (2010). the prospects for discovery in the ongoing 300-day LUX [6] D. S. Akerib et al. (LUX Collaboration), First Results from search and the future LUX-ZEPLIN [46] experiment. the LUX Dark Matter Experiment at the Sanford Under- ground Research Facility, Phys. Rev. Lett. 112, 091303 This work was partially supported by the U.S. Department (2014). of Energy (DOE) under Awards No. DE-FG02-08ER41549, [7] D. S. Akerib et al. (LUX Collaboration), The large under- No. DE-FG02-91ER40688, No. DE-FG02-95ER40917, ground xenon (LUX) experiment, Nucl. Instrum. Methods No. DE-FG02-91ER40674, No. DE-NA0000979, No. DE- Phys. Res., Sect. A 704, 111 (2013). FG02-11ER41738, No. DE-SC0006605, No. DE-AC02- [8] C. H. Faham, V. M. Gehman, A. Currie, A. Dobi, P. 05CH11231, No. DE-AC52-07NA27344, and No. DE- Sorensen, and R. J. Gaitskell, Measurements of wave- FG01-91ER40618; the U.S. National Science Foundation length-dependent double photoelectron emission from under Grant No. PHYS-0750671, No. PHY-0801536, single photons in VUV-sensitive photomultiplier tubes, No. PHY-1004661, No. PHY-1102470, No. PHY-1003660, J. Instrum. 10, P09010 (2015). No. PHY-1312561, No. PHY-1347449; the Research [9] C. H. Faham, Ph.D. thesis, Brown University, 2014. [10] V. Solovov et al. (ZEPLIN-III Collaboration), Position Corporation Grant No. RA0350; the Center for Ultra-low reconstruction in a dual phase xenon scintillation detector, Background Experiments in the Dakotas (CUBED); and the IEEE Trans. Nucl. Sci. 59, 3286 (2012). South Dakota School of Mines and Technology (SDSMT). [11] Y. Mei, Ph.D. thesis, Rice University, 2011. LIP-Coimbra acknowledges funding from Fundação para a [12] D. S. Akerib et al. (LUX Collaboration), Calibration, event Ciência e a Tecnologia (FCT) through the Project-Grant reconstruction, data analysis and limits calculation for the No. PTDC/FIS-NUC/1525/2014. Imperial College and LUX dark matter experiment (to be published).

161301-6 WIMP-nucleon SI Exclusion - SSR

Brazil bands show our 1 and 2 σ regions 0.22 zeptobarns (at 50 GeV/c2)

•We observed an improvement of a factor of four compared with the results from the ﬁrst science run. SI Exclusion - FSR+SSR 47

SUSY-cMSSM (1 σ)

0.11 zeptobarns (at 50 GeV/c2) •Both LUX Runs Combined o https://arxiv.org/abs/1608.07648 •LUX now excludes signiﬁcant portions of the 1-sigma regions for WIMPs favored by certain supersymmetric models. LUX Proposal VS Main Result 48

Blue: DOE/NSF LUX Proposal, Projected (2007) Black: Observed 332-liveday limit Brazil band: 1- and 2-sigma 332-LD expected sensitivities Axion Searches (Preliminary) 49

•Measure the coupling between axions/ALPs and electrons (gAe) •Axion signal: electronic recoil (ER) scattering - NR/ER discrim. does not help •The majority of the background for LUX sits in the ER band •Solar Axions: relativistic particles - mass assumed to be zero •ALPs: not relativistic particles •A Proﬁle Likelihood Ratio approach used as statistical strategy

Solar Axions Axion-like particles (ALPs) Neutrino Physics 50 •Neutrinoless double beta decay (0νββ) o Neutrinos are their own antiparticles (Majorana particles) o 0νββ decay rate is related to the square of an effective Majorana neutrino mass o 136Xe (Q=2.5 MeV), 134Xe (low Q) and 124Xe (low Q) •Double electron capture (2ν2EC or 2νß+EC) or neutrinoless double electron capture (0νDEC) o 2ν2EC could be observed for 124Xe, 126Xe o Extremely long lifetime •Solar neutrinos - background for DM search o Elastic scattering νe→νe which produce - ER recoils o Coherent neutrino-nucleus scattering - NR recoils • Most from 8B fusion, still unobserved •Other neutrino sources: o Atmospheric neutrinos - from cosmic rays o Geo-neutrinos - from the radioactive decays o Reactor neutrinos o Galactic neutrinos (from supernovas) LUX has a low mass, not very sensitive •Detection of sterile neutrinos to these interactions, but LZ, being •Measurement of the neutrino magnetic moment bigger (~28x), will be competitive. The LUX-ZEPLIN Experiment 51

•Turning on by 2020 with 1,000 initial live-days plan Instrumentation •10 tons total, 7 tons active, ~5.6 ton ﬁducial conduits •Unique triple veto •GOALS: < 2 x 10-48 cm2, at Existing 40 GeV ~100 times better water tank than LUX Gd-loaded liquid LXe heat scintillator exchanger tower

120 outer detector PMTs (Main Detector) n tubes 2-phase XeTPC 494 (131) TPC (Xe skin) PMTs The LUX-ZEPLIN Experiment

(LZ 5.6 Tonnes, 1000 live days, 6 keVnr analysis threshold) LIP-CoimbraLIP-Coimbra Dark in LUX Matter and Group LZ

Alexandre Vladimir Lindote Solovov

Paulo Cláudio Isabel Francisco João Brás Silva Lopes Neves Rodrigues

Searching for Dark Matter since 2003 LIP-Coimbra in LUX and LZ

•Main hardware/software contributions o Slow control development (L3) •Slow controlL managementU o Online monitor system •LN handling system •Data analysis tools in LZ o Position reconstruction •Data management and processing o Pulse finding •Major data analysis projects o S1 and S2 pulse identification o WIMP search final analysis •Physics beyond dark-matter o ER/NR discrimination o Neutrino-less double beta decay o Wall background radial model o Solar neutrinos •Data Analysis Tools in LUX •Background model and simulations (L3) o Position reconstruction o Position corrections •Research and development program o Photon counting o Characterization of charging effects in PTFE when immersed in liquid xenon •Research and development program o PTFE reflectance in Liquid xenon o PTFE reflectance in gas

o PMT uniformity studies https://arxiv.org/pdf/1612.07965.pdf •~700 days onsite shifts in South Dakota Conclusions 55 •LUX has since 2013 the world-leading result in the dark-matter research. •LUX had signiﬁcant improvements in the calibration of xenon detectors - essential to improve detector’s sensitivity. •The LUX’s 332 live-day search, cutting into un-probed parameter space. Excluding SI WIMPs down to 0.22 zeptobarns (2.2x10-46 cm2) at 50 GeV/c2. •When both runs are combined SI WIMPs are excluded down to 0.11 zeptobarns at 50 GeV/c2. •Results available on: o https://arxiv.org/pdf/1608.07648v2.pdf •More analysis forthcoming o Spin-dependent, update to axion searches/ALP, effective ﬁeld theory, neutrino less double beta decay, additional calibrations etc. •Onwards and downwards: LUX-ZEPLIN (LZ) experiment under construction, 7 tonne active mass (2020).

Postdoctoral position and Research position in LZ There are two open positions in our group ([email protected]) o http://www.lip.pt/ﬁles/rhumanos/1-345_10398375862febd4d0f48bb72d478a09d25ddf9120161219.pdf Backup Slides

LUX in the Water Tank 2012 Planck Collaboration: Cosmological parameters

Posteriors for the A reaction rate parameter for vari- Fig. 37. 2 As in Fig. 38, but now allowing YBBN and N to vary as ous data combinations. The vertical dashed line shows the value Fig. 39. P e↵ parameter extensions to base ⇤CDM. A2 = 1 that corresponds to the current experimental estimate of 57 the d(pEvidence, )3He rate used in the PArthENoPE forBBN code. non-baryonic cold DM abundance measurements and with the predictions of standard BBN.

0.35 https://arxiv.org/pdf/1502.01589v3.pdf There is a well-known parameter degeneracy between YP Excluded by andThe the radiation baryon-to-photon density (see the discussion inratioPCP13). Heliumhas been the Serenelli & Basu (2010) abundance predictions from the CMB are therefore particularly sensitivesame to thesince addition ~1 of the minute parameter N e↵afterto base ⇤theCDM. Big Bang. 0.30 BBN Allowing both YP and Ne↵ to vary we ﬁnd the following con- straintsBaryons (at 95 % CL): are ≲15.7% of the mass in

BBN Aver et al. (2013) P

0.25 matter. +0.058 Y 0.252 0.065 Planck TT+lowP ; . +0.058 + + Standard BBN BBN 8 0 251 0.064 Planck TT lowP BAO ; Y = P > +0.034 > 0.263 . Planck TT,TE,EE+lowP ; 0.20 Planck TT+lowP > 0 037

He massHe fraction > +0.035 4 Planck TT+lowP+BAO < . Planck + + > 0 262 0.037 TT,TE,EE lowP BAO . > Planck TT,TE,EE+lowP > (79) > BBN Contours in: the YP –Ne↵ plane are shown in Fig. 39. Here 0.15 0.020 0.022 0.024 0.026 again, the impact of Planck polarization data is important, and b helps to substantially reduce the degeneracy between these two Baryon Density parameters. The Planck TT,TE,EE+lowP contours are in very BBN Fig. 38. Constraints in the !b–YP plane from Planck and good agreement with standard BBN and Ne↵ = 3.046. However, Planck+BAO, compared to helium abundance measurements. even if we relax the assumption of standard BBN, the CMB does Here 68 % and 95 % contours are plotted for the CMB(+BAO) not allow high values of Ne↵. It is therefore dicult to accommo- BBN data combinations when YP is allowed to vary as an additional date an extra thermalized relativistic species, even if the standard parameter to base ⇤CDM. The horizontal band shows observa- BBN prior on the helium fraction is relaxed. tional bounds on 4He compiled by Aver et al. (2013) with 68 % and 95 % errors, while the dashed line at the top of the ﬁgure delineates the conservative 95 % upper bound inferred from the 6.6. Dark matter annihilation Solar helium abundance by Serenelli & Basu (2010). The green Energy injection from dark matter (DM) annihilation can stripe shows the predictions of standard BBN for the primordial alter the recombination history, leading to changes in the 4 abundance of He as a function of the baryon density. Both BBN temperature and polarization power spectra of the CMB predictions and CMB results assume Ne↵ = 3.046 and no signif- (e.g., Chen & Kamionkowski 2004; Padmanabhan & Finkbeiner icant lepton asymmetry. 2005). As demonstrated in several papers (e.g., Galli et al. 2009; Slatyer et al. 2009; Finkbeiner et al. 2012), CMB anisotropies o↵er an opportunity to constrain the nature of DM. Furthermore, CMB experiments such as Planck can achieve limits on the an- paper, the systematic e↵ects in the Planck polarization spectra, nihilation cross-section that are relevant for the interpretation of although at low levels, have not been accurately characterized the rise in the cosmic-ray positron fraction at energies > 10 GeV at this time. Readers should therefore treat the polarization con- observed by PAMELA, Fermi, and AMS (Adriani et⇠ al. 2009; straints with some caution. Nevertheless, as shown in Fig. 38, Ackermann et al. 2012; Aguilar et al. 2014). The CMB con- all three data combinations agree well with the observed helium straints are complementary to those determined from other astro-

50 Axion Searches Spectra 58 •Solar Axions: relativistic particles - mass assumed to be zero •The theoretical spectrum is the product of the solar axion ﬂux and the axio-electric cross section (solar axion assumed to be massless) •ALPs: not relativistic particles •The theoretical spectrum is a spike at the ALPs mass, because they are not relativistic particles; •The spike acquires then a width because of the experimental resolution;

Solar Axions Axion-like particles (ALPs) Neutrino Backgrounds/Signals in 59

LUX/LX - Coherent Scattering 3

8 1012 10 6GeVWIMP 100 GeV WIMP pp pp 10 pep ] pep

1 10 hep 6 hep

− 10 7Be [384.3 keV] 7Be [384.3 keV]

1] 7 8 7 Be [861.3 keV] 10 Be [861.3 keV] − 8

MeV 8 B B 4 13 N 1 13 10 15

− 6 N O

year 17 s 10 15 O F 1

2 DSNB 17 F −

− Atm DSNB 102 104 Atm

0 102 10

0 10 10−2 Event rate [ton

10−2 Neutrino Flux [cm 10−4 10−4 10−1 100 101 102 103 10−3 10−2 10−1 100 101 102 Neutrino Energy [MeV] Recoil energy [keV]

FIG. 1: Left: Neutrino energy spectra that are backgrounds to direct detection experiments: Solar, atmospheric, and the •Neutrinodi↵use supernova energy spectra background. due Theto coherent Solar neutrino neutrino-nucleus fluxes are normalised scattering to the that high metallicityare backgrounds standard Solarto direct model. The detectionatmospheric experiments: neutrino spectrum Solar, isatmospheric, the sum of contributions and the diffuse from electrons, supernova anti-electrons, background. muons The and anti-muons.Solar neutrino The di fluxes↵use supernova background is the sum of three di↵erent neutrino temperatures, 3, 5 and 8 MeV. Right: Xenon scattering event rate 45 2 areas normalised a function of to recoil the energy high formetallicity each type ofstandard neutrino Solar as well model. as a 6 GeV The WIMP atmospheric with n neutrino=5 10 spectrumcm (solid is black the line)sum of 49 2 ⇥ 8 contributionsand a 100 GeV from WIMP electrons, with nanti-electrons,=2.5 10 cm muons(dashed and black anti-muons. line) to show The how diffuse they overlap supernova with Bbackground and atmospheric is the ⇥ sumneutrino of three induced different recoils neutrino respectively. temperatures, 3, 5 and 8 MeV. Right: Xenon scattering event rate as a function of recoil energy for each type of neutrino as well as a 6 GeV WIMP with σχ−n = 5 × 10−45 cm2 (solid black line) and a 100 GeV WIMP with σχ−n = 2.5 × 10−49 cm2 (dashed black line) to show how they overlap with 8Bbackground and atmospheric for WIMP neutrino searches induced are displayed recoils in respectively. the left forming a line of sight integral of the spectrum of neu- hand of Fig. 1 with uncertainties listed in Table I. In fact trinos from a single supernova with the rate density of with advances in technology currently underway [45] it core-collapse supernovae as a function of redshift. See will be possible for direct detection experiments to make Ref. [52] for the full calculation of the predicted DSNB. competitive measurements of these neutrino fluxes [46] The total flux of the DSNB is considerably smaller than 2 1 and even constrain new physics such as the existence of for Solar neutrinos, around 86 cm s ,howeveritis sterile neutrinos [47] or new interactions between neutri- an important background to consider as it extends to a nos and nuclei or electrons [48]. higher energy range not occupied by Solar neutrinos. The calculated spectra have a Fermi-Dirac form with temper- Neutrinos from various fusion reactions in the interior atures in the range 3 to 8 MeV. In this study we use a of the Sun dominate the low energy-high flux regime and DSNB flux which is the sum of 3 temperatures: 3 and are the dominant background for direct detection with a 5 MeV for electron and anti-electron neutrinos respec- total flux at Earth of around 6.5 1011 cm 2 s 1 [49, 50]. tively, and 8 MeV for the sum of the remaining neutrino Neutrinos from the initial pp reaction⇥ make up 86% of flavours. There are considerable theoretical uncertainties all Solar neutrinos and have been detected most recently in this calculation, hence we will take a large systematic by the Borexino experiment, determining the flux with uncertainty of 50% on the total flux of DSNB neutri- an uncertainty of 1% [51]. However for the remain- nos [52]. ing Solar neutrinos⇠ the theoretical uncertainties in the fluxes are as large as or larger than the measurement The final type of neutrino remaining to be dis- uncertainty and rely on an assumption of a Solar model cussed are those from the atmosphere which provide for their calculation. In this work we assume the high the main neutrino background for WIMP masses above metallicity Standard Solar Model (SSM) [49] which is 100 GeV. These neutrinos occupy the high energy and the model most consistent with existing Solar data. Due low flux regime and will limit the sensitivity of experi- to their relatively low energies, Solar neutrinos will influ- ments to spin-independent cross sections below around ence the detection of WIMPs with masses less than 10 10 48cm2 [13, 15, 19]. The flux of atmospheric neutrinos GeV. with energies less than 100 MeV is diculat to measure For WIMP masses between 10 and 30 GeV, the neu- as well as predict theoretically [53–55] although the ex- 2 1 trino floor is caused by the sub-dominant di↵use super- pected flux is around 11 cm s . In this work we use a nova neutrino background (DSNB), the sum total of all calculation that is a sum of the contributions from elec- neutrinos emitted from supernovae over the history of tron, anti-electron, muon and anti-muon neutrinos and the Universe. The background flux is calculated by per- place a 20% uncertainty on the total flux [54]. ⇠ Grid Conditioning 60 •Results from the first science run featured a 48.9% electron extraction efficiency. •During the first half 2014 the voltage of the grids was raised for an extended period of time until significant current is drawn. The main objective was to burn any dust or asperities present in the grids. •After the grid conditioning the electron extraction efficiency increased to >70%. •…but upon refilling we observed a large radial component in the drift field. •Moreover the effect of the radial field is time dependent increasing along the run.

250 kg still in here - -2 kV e Krypton data Measured position reconstructed of wall

positions -4 kV

-6 kV z [cm] [cm] z

-8 kV

-10 kV

r [cm] Modeling the Electric Field 61 •A Fully 3-D model is constructed in the COMSOL Multiphysics® FEM software to compute the electric field in the active region of LUX •The observed radial field is consistent with a build up of negative charge (0 to -10 μC/m2) on the PTFE walls. •Charges are added to the walls to produce the radial field that best produces the observed distribution of 83mKr decays.

Charges in the PTFE 10x variation in the Wall field amplitude Dealing with the Fields 62 (How to deal with a field that is varying in space and in time?) •Detector’s volume sliced in M time bins and N z slices •In each of the MxN segments, we assume a uniform detector model for both ER and NR response. •83mKr is used to compute the fiducial volume in each segment •We found that 4 date bins and 4 z-slices captured the variation with sufficient calibration statistics. The data bins used were: o DB#1 (2014.09.09-2014.12.31): 46.8 live-days 105.4±5.3 kg fiducial mass o DB#2 (2015.01.01-2015.03.31): 46.7 live-days 107.2±5.4 kg o DB#3 (2015.04.01-2015.09.30): 91.6 live-days 99.2±5.0 kg o DB#4 (2015.10.01-2016.05.03): 146.9 live-days 98.4±4.9 kg •We effectively have 16 independent detectors •For each detector S1 and S2 are modeled with NEST (Noble Element Simulation Technique, NEST, http://www.albany.edu/physics/NEST.shtml) •NEST is “tuned” to each of the 16 detectors by varying the applied field until we see a match between model and calibration data.

M. Szydagis 2013 JINST 8 C10003 and J. Mock 2014, JINST 9 T04002 Dealing with the Fields 63

Gray density: CH3T calibration (ER)

Orange density: DD calibration (NR)

Solid lines: NEST model, ER, NR band mean

Dashed lines: NEST model, 10-90 percentile. S2/S1 Position Corrections 64

•Size of the S1 depends on the location of the event (due to P M T geometrical light collection), and S2 (due electronegative impurities) Anode •On the FSR the correction factors for both S1 and S2 were obtained by flat fielding a mono-energetic Gate source 83mKr. S1 larger for •However, a spatially varying E-field events lower in ALSO affects S1 and S2 sizes, but the detector differently for every particle type and energy. S2 larger for events higher in the detector

Cathode

P M T 073303-5 Manalaysay et al. Rev. Sci. Instrum. 81,073303͑2010͒

TABLE I. The measured zero-field LY, peak resolution ͑Res.͒, and field dependence fit parameters ͑ai͒.Therow following 41.5 keV gives the charge collection of the summed signal. Uncertainties shown in LY are statistical only because these two peaks are taken from identical events, their systematic uncertainties are highly corre- lated, and hence do not affect the significance of the relative rise in LY.

E LY Res. a2 −4 ͑keV͒ ͑pe/keV͒ ͑␴/␮͒ a1 ͑10 cm/V͒ a3

9.4 6.74Ϯ0.06 20.0% −0.35Ϯ0.06 6.3Ϯ3.0 1 32.1 6.43Ϯ0.04 14.4% −0.55Ϯ0.03 8.9Ϯ1.6 1 41.5 ¯¯0.406Ϯ0.006 17Ϯ2 0.074Ϯ0.012 123.6 6.38Ϯ0.05 11.5% −0.679Ϯ0.007 12.6Ϯ0.5 1

with the two signals exhibiting anticorrelation. It is then cru- ting the S1 peak positions versus the S2 peak positions from cial that the scintillation dependence on the applied field, data taken at various applied fields, shown in Fig. 8. As S1 called field quenching, be known quantitatively for any cali- and S2 are anticorrelated, these data lie along a line having bration sources. Figure 7 shows the LY as a function of the negative slope, with the line’s intercepts representing the to- applied field, normalized to the zero field value, of the two tal number of quanta, ions plus excitons ͑Nion+Nex͒. For 83m 57 Kr transitions and the Co line. The uncertainty in the electronic recoils, the ratio of excitons to ions, Nex/Nion, is 30 LY is dominated by a 2% systematic uncertainty taken from taken to be 0.06, and hence Q0 is 94.3% the value of the S2 the measured fluctuations in the PMT gain over the duration intercept. The horizontal positions and error bars are deter- of the run. The horizontal positions are determined by elec- mined in the same manner as those of the scintillation yield trostatic field simulations of the detector in each voltage con- measurements, while the vertical error bars are instead domi- figuration used; horizontal uncertainties are the 1−␴ varia- nated by the statistical errors in the peak fits and the uncer- tion in the field over the active volume. The simulations65 were tainty in Q0. S2/S1 Positioncarried Corrections out using the COMSOL simulation package ͑commer- The data are fit with a three-parameter function based on cially available ,29 and verified with software written in the Thomas–Imel box model for electron-ion 073303-4 Manalaysay͒ et al. Rev. Sci. Instrum. 81,073303͑2010͒ house. recombination,32 given by The time scale of the ionization signal 1–2 ␮s does not 60 S͑E͒ Q͑E͒ 1 permit the two 83mKr transitions to be resolved separately, , = a a E ln 1+ + a , ͑1͒ •Our strategy is: 100 1 2 3 and instead40 the S2 signal contains the combination of charge S͑0͒ Q0 ͩ a2Eͪ o Disentangle position effects from field emitted from both decays. This 41.5 keV summed-signal ion- where E is the electric field strength, and S and Q are the 20 effects; ization yield is also shown in Fig. 7 normalized to Q , the scintillation50 and ionization yields, respectively. This model is

0 Counts p.e./sample o Apply a correction to account for theoretical0 total amount of initial charge produced prior to used only to provide a convenient parametrization of the electron-ion0 recombination.500 This1000 value is determined1500 by plot-2000 data, and not to infer fundamental LXe physical properties position effects only. 0 t [ns] 0 200 400 600 800 1000 1200 1400 1600 from the results of the fits. The ai are the parameters of the 83m 120 ∆t [ns] • Kr has two decays close in Background 83mKr S1 1001 4 x 10 time. The ratio of the first-to- 5 2 80 10 83mKr (9.4 keV) second S1 pulse area 0 0.860 IS2 4 40 ounts depends on field alone. This 83m Kr (32 keV) C 57 Second S1 [p.e.] 0.620 Co

allows us to measure the ), Q(E)/Q 0 0 3 2nd S1 1st S1 0 100 200 300 400 0 100 200 300 400 1 component of variation due 0.4 10 First S1 [p.e.] [p.e.] 0 50 100 Higher150 200 250 300

S(E)/S( 83m S1Field [p.e.] to applied field alone. Kr (41.5 keV,83m Charge) S2 2 FIG. 5. ͑Top͒ PMT output from a Kr decay. In this double pulse of 0.2 primary scintillation light ͑S1͒, the first pulse corresponds to the 32.1 keV FIG. 6. ͑Color online͒͑Top͒ The distribution ofLower delay times between first transition with the second pulse resulting from the 9.4 keV transition. ͑Bot- and second S1 pulses for events in the 83mKr acceptanceField window. An expo- tom͒ The area of the first S1 pulse vs the area of the second, for events nential1 fit to the distribution gives a half-life of 156Ϯ5 ns,IS1 consistent with 0 83m showing0 this characteristic500 two-pulse1000 structure. Shown1500 are distributions2000 the published value of 154 ns. ͑Bottom͒ Spectra from the two Kr transi- taken before Rb exposure ͑A“background”pplied Field͒ and [V/cm] during Rb exposure ͑83mKr͒, tions, summed over all runs taken at zero field. demonstrating that the population of 83mKr decays is clearly separated from background events. The box represents the energy cuts used as the 83mKr 0 FIG. 7. ͑Color online͒ Field quenching, defined as the LY of a spectral line 0 50 100 150 200 250 300 350 400 acceptance window. divided by the LY obtained at zero field, or S͑E͒/S͑0͒. Data collected from S1 [p.e.] 21 57Co ͑open black squares͒ are consistent with those reported in Ref. 31 tent with the published value of 154.4Ϯ1.1 ns. This excel- lent agreement validates the claim that these events are in- ͑solid83Rb gray introduced diamonds͒. in Dashed the system lines correspond are discussed to a fit parametrization in Secs. III de- and FIG. 8. ͑Color online͒ The peak position in S2 vs S1 space for the 41.5 keV scribed in the text. Also shown is the field-dependent charge collection of emissiondeed ofcaused83mKr. by The83m dataKr are decays. taken from applied fields ranging from 100 IV. 83m the combination of both Kr transitions, Q͑E͒/Q0; the two transitions V/cm toDue 1 kV/cm. to the The shaping line is fit to of the the data PMT having signals vertical byand thehorizontal various occur too close in time for their ionization signals to be individually re- intercepts IS2 and IS1, respectively; these intercepts indicate the location of data acquisition ͑DAQ͒ components, multiple S1 pulses that solved.III. ANALYSIS AND RESULTS Nion +Nex. are delayed by less than ϳ100 ns cannot be separately re- Once the 83mKr has entered the LXe, a 32.1 keV transi- solved. Additionally, the signal is required to be “clean” i.e., Author complimentary copy. Redistribution subject to AIP license or copyright, see http://rsi.aip.org/rsi/copyright.jsp ͑ tion might occur in the active region, which will then be flat baseline͒ two samples before and after the pulse, in order followed by the 9.4 keV transition. A 83mKr decay is, there- to register as a positive S1 identification during the offline fore, indicated by two S1 pulses whose separation in time is processing of the data. This makes the efficiency for detect- characterized by a decaying exponential with t1/2 =154 ns. ing multiple S1 pulses less than unity for ⌬tS1Ͻ250 ns, as is Some of these transitions will occur too close in time to be obvious from Fig. 6 ͑top͒. Therefore, the double S1 selection resolved separately, giving a single 41.5 keV pulse; however, cut detects 83mKr decays with an efficiency of approximately the strength of this signal is well below the background level 32% under these conditions. in the Xürich detector. On the other hand, many of the 83mKr The spectra, in pe, obtained at zero field from the two decays have a double S1 structure, while only a small frac- transitions of 83mKr are displayed in Fig. 6 ͑bottom͒.A tion of non-83mKr decay events share this feature. An ex- Gaussian function is fit to each spectrum that is used to de- ample of the PMT response from a 83mKr decay is seen in termine the LY and energy resolution, shown in Table I. As Fig. 5 ͑top͒. mentioned in Sec. I, 57Co emits primarily 122 keV ␥ rays. The events with such a double S1 structure are shown However, there is a small additional contribution from 136 from one data set in Fig. 5 ͑bottom͒, with the area of the first keV. The two lines, however, are not distinguishable from pulse plotted versus the area of the second pulse. In this one another due to the detector’s energy resolution and in- space, it is evident that the 83mKr decays form a population stead give a single peak, whose weighted average energy is of events that is clearly separated from background. The box 123.6 keV. The measurements suggest a rise in the LY at indicates the energy cuts for first and second S1 pulses used lower energies, consistent with behavior previously observed to identify 83mKr decays; before opening the Rb valve, back- in LXe ͑Ref. 27͒ and also in the XENON10 detector.28 The ground data show no events within this box. After the Rb peak resolutions ͑␴/␮͒ are also shown at zero field. valve has been opened, the rate of 83mKr decays in the total As mentioned in Sec. II, most LXe detectors use an ap- LXe volume increases to the 20 Bq level in roughly 10 h. In plied external electric field in order to collect electrons emit- order to further check that these are indeed 83mKr decays, the ted from the interaction site. As the applied field is increased, distribution of S1 delay times ͑i.e., the time between the first more and more electrons leave the interaction, suppressing and second S1 pulses͒, ⌬tS1, of events within the box of Fig. the recombination process that contributes photons to the 5 ͑bottom͒ is fit with a decaying exponential. The result of scintillation signal. The result is that both the scintillation the fit, shown in Fig. 6 ͑top͒, gives t1/2 =156Ϯ5 ns, consis- and ionization responses vary strongly with applied field,

Author complimentary copy. Redistribution subject to AIP license or copyright, see http://rsi.aip.org/rsi/copyright.jsp