Performance results from the ZEPLIN-III Second Science Run

Pawel Majewski STFC/Rutherford Appleton Laboratory

On behalf of ZEPLIN-III Collaboration: Edinburgh University (UK), Imperial College London (UK) , ITEP-Moscow (Russia) LIP-Coimbra (Portugal) STFC Rutherford Appleton Laboratory (UK)

PSD9, September 2011, Aberystwyth

Friday, 16 September 2011 1 WIMP detection techniques and DM experiments around the world

Heat and ionisation Bubbles and Droplets: bolometers : CDMS CUOPP EDELWEISS PICASSO Light and heat Bolometers: CRESST ROSEBUD Phonons

dE/dx

Charge Light Scintillators : DAMA Scintillation and ionisation charge detectors: LIBRA XENON XMASS WARP CLEAN ArDM ANAIS ZEPLIN KIMS Ionisation detectors: DMTPC LUX DRIFT, GENIUS, NEWAGE, HDMS, IGEX

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 2 ZEPLIN III at the Boulby mine

Peterlee

Hartlepool Middlesborough

Newton Aycliffe Redcar

Billingham Inverness Stockton Middlesbrough

Darlington Staithes

Whitby Edinburgh

Newcastle Belfast

York Dublin Liverpool

Birmingham Location: Boulby, North Yorkshire (UK)

London Vertical Depth: 1070 m (2850 mwe) Muon flux: ~4*10-8cm-1s-1 Plymouth

Other experiments: ZEPLIN II DRIFT II SKY

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 3 ZonEd Proportional scintillation in LIquid Noble gases programme

ZEPLIN I: Single phase, 3 PMTs, 5/3.1 kg LXe run 2001/04 ZEPLIN III: Double phase, Limit: 1.1*10-6 pb 31 PMTs, high electric field, 12/6.5 kg LXe First Science Run II-V 2008 (83 days) Limit: 8.1*10-8 pb Second Science Run 2010/11 (319 days) Limit: ongoing analysis

ZEPLIN II: Double phase, 7 PMTs, moderate E field, 31/7.2 kg LXe run 2005/06 Limit: 6.6*10-7 pb

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 4 ZEPLIN III - double phase liquid Xenon detector

• Scintillation light collection ‐ 31 2” PMTs immersed in liquid ‐ 36 mm drift height ‐ 400 mm diameter ‐ 4 mm gas gap with 1.6 atm operating pressure ‐ no PTFE reflector

• Electric field operation ‐ 4 kV/cm in liquid and 8 kV/cm in gas ‐ no extraction grids

• Light yield at 4 kV/cm ‐ S1 ~ 1.8 pe/keV ‐ S2 ~ 11 pe/keV

• Event 3D position reconstruction @ 122 keV (x,y) 2 mm, (z) micrometers

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 5 ZEPLIN III : operation principle

+7 kV Gas E S2

Liquid E S1 –10 kV

Single scatter

Waveforms from PMTs are acquired using 8‐bit dual‐range Acqiris system (64 channels) with 2ns sampling rate. Maximum acquisition speed : 100 events/s.

Example of double scatter event registered fully with low sensitive readout channel.

Double scatter

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 6 Detector calibration Energy resolution at 122 keV & E=4 kV/cm: S1 (16.3 %), S2 (8.8 %), correlated 5.4 %.

Combined data from runs with 137Cs (red) and AmBe (blue) sources. 40 keV line from inelastic scattering is clearly visible on the ER band.

Simulated energy deposition

Smeared simulated energy deposition Difference for E>160 keV due to: no background substraction and single scatter evet selection in data

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 7 ZEPLIN III SSR upgrade: photomultipliers

- PMT gamma-rays limited sensitivity of first run by a large factor (>1,000 mBq/PMT) - Custom-built low-background tubes, pin-by-pin compatible (from ETEL) - Assembly onto ZEPLIN-III array in early 2010 (delivery 18 months late!) - Aiming for 30x reduction in radioactivity to <50 mBq/PMT - Achieved – but with optical performances very poor! FSR: QE MEAN=30.0%, STDEV=3.4(%); GAIN VARIATION (MAX/MIN)=6.5 SSR: QE MEAN=26.2%, STDEV=6.2(%); GAIN VARIATION (MAX/MIN)=100!

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 8 ZEPLIN III SSR upgrade: anti-coincidence veto detector

Active Veto System (> 3π coverage): 20 (roof) and 32 (barrel) plastic slabs equipped with 52 PMTs Principle of operation: surrounding Gd‐loaded polypropylene - multiple scattering in hydrocarbon - radiative capture (mean capture time: 11 µs) by 157Gd accompanied Single plastic slab with MeV gammas (polystyrene based UPS-932A)

Single Gd-loaded polypropylene slab

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 9 ZEPLIN III SSR upgrade: anti-coincidence veto detector assembly

Inside the lead castle

Gd-loaded slabs around ZIII detector

Veto barrel section fully assembled

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 10 Second Science Run: Veto Performance

- 52 readout channels with CAEN-1724 ADCs

- Waveform duration -20 to +300 µs, 100 ns sampling

- Neutron tagging efficiency: 60 %

- Gamma tagging efficiency: 28 %

- Mean neutron capture time: Pulse time distribution when triggered by ZEP-III: Peak at 17.7 µs - prompt coincidence (PTAG) delayed coincidence pulses appeared later (GTAG)

Pulse time in the veto relative Delayed neutron tagging efficiency to S1 from nuclear elastic scatters in LXe PTAG efficiency as a function of energy deposited in LXe

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 11 Energy and position reconstruction

Mercury: least squares technique based on PMT response functions calculated from gamma calibration data

Input: PMT signal and error: {Ai,σi} Model: for an event of energy e at point {x,y} the signal from PMT is:

ai = eCiηi(x, y)

where:

e - energy, Ci - C x gain x QE, ηi -light collection efficiency

2 2 (ai Ai) χ (x, y, e)=Σ −2 σi Minimization

Output : x0, y0, e0

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 12 Energy and position reconstruction

How it works:

1. Reconstruct Co-57 calibration data with centroid algorithm 2. Plot S2 for a given PMT vs distance from its axis (R) for events in the central region 3. Convert it to profile histograms and fit to get the first approximation for the PMT response functions 4. Run the least squares reconstruction with the new set of response functions 5. Update S2 vs R plot 6. Find second approximation for response functions 7. Repeat (4)...(6) until response functions don't change

1. Energy spectrum for Co-57 (resolution is ~8% FWHM for 122 keV peak)

2. y(S1) vs y(S2) for Co-57

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 13 Position reconstruction

X and Y position resolution 2 mm FWHM

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 14 Second Science Run Performance

First science run (FSR) at Boulby: 83 days in 2008 Strong constraints on WIMP-nucleon scattering XS Phase-II upgrades commissioned in 2009/10 New photomultiplier array (ultra-low background) New anti-coincidence veto (background reduction, diagnostic) New calibration hardware (reduction of systematics) System automation (improved stability, underground effort) Second science run (SSR) - data-taking now completed 319 days of continuous data taking run Effective fiducial exposure ~560 kg*days (4x FSR)

One month’s worth of daily LN2 refills

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 15 Second Science Run Backgrounds

Main sources of background radiation: Radial - gamma: U,Th and 40K

- neutrons: muon induced showers, spontaneous fission of 238U and (α,n) reaction from U and Th decay chains

Depth

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 16 Analysis of Single Electron events

SE area distribution Clusters of single photoelectron pulses 1e = 30 phe

Excellent for S2 signal calibration and xenon purity assessment !

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 17 Second Science Run Performance: ongoing analysis - Veto prompt tags (green) not blinded in signal region (28% tagging fraction) - Veto delayed tags (red) consistent with predicted 0.5% accidentals rate and confirming negligible neutron background (e.g. no tags below NR median) however : - Optical performance of new PMTs is very challenging – still being addressed - Gamma background 20x lower but distribution is wider and threshold is higher

FSR SSR

Signal Region

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 18 Conclusions

Second Science Run

- Upgraded instrument successfully with new PMTs and new veto detector

- Achieved required background in both neutrons and gamma-rays

- Significant effort into vertex reconstruction algorithms delivered excellent spatial and energy resolutions

- Second science run accumulated 1,750 kg·days of raw exposure collected over 319 days of continuous operation

- SSR WIMP result coming soon

FSR and SSR in numbers

- 2500 person·day underground effort - 100 TB of raw data - 12 xenon liquefactions - 15000 litres of Liquid Nitrogen used - SSR - longest continuous operation of any two phase detector in the world !

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 19 Second Science Run Sensitivity

PSD9, September 2011, Aberystwyth Pawel Majewski

Friday, 16 September 2011 20