Overview of Scintillation Systems

Overview of Scintillation Systems ANT 11 Daniel Bick October 11, 2011 Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 1 / 21

Outline 1 Motivation 2 Scintillator Requirements for LENA 3 Different Scintillators for Neutrino Detection 4 Scintillator Properties 5 Conclusions Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 2 / 21

Motivation Past and present experiments have shown that liquid scintillator detectors are an excellent choice for detecting neutrinos KamLAND and Borexino recently explored neutrino fluxes below 5 MeV in the reactor, solar and geoneutrino sector Future experiments will require larger target masses and thus larger volumes Highly transparent scintillators needed Liquid scintillator is well suited for the detection of low energy neutrinos Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 3 / 21

Requirements for a Next Generation LSD What should the next liquid scintillator detector offer? Good energy resolution More than 200 p.e. per MeV sufficient light yield needed High transparency Large volume requires long attenuation lengths O(-20 m) Low detection threshold ν e : 1.8 MeV (threshold for inverse β-decay) ν x : 200 kev (Intrinsic 14 C) Radiopurity Excellent background discrimination Coincidence signal from inverse β-decay Pulse shape discrimination (scintillator type dependent) Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 4 / 21

Liquid Scintillator An organic liquid scintillator in general consists of a solvent and one or more solutes. Solvent Hydrocarbon molecules containing benzene-ring structures Luminescent when excited, deexcitation produces UV light Spectrum of a single component scintillator has a significant overlap with its own absorption spectrum Solute(s) One or two solutes added as wavelength-shifter or fluor Energy is transferred e.g. via dipole-dipole interactions or collisions Shifts emission spectrum to longer wavelengths where the scintillator is more transparent Optimization of emission spectrum to PMT sensitivity Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 5 / 21

Characteristics of Liquid Scintillators Emission spectrum UV or violet light, light transmission wavelength dependent Light Yield Affects energy resolution and detection threshold Fluorescence time profile Particle-type dependent particle ID Attenuation length Absorption affects energy resolution Scattering affects signal shape Purity Radioactive contaminations mimic neutrino signals Safety Environmental and health hazards Often low flash points Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 6 / 21

Solvents: PC PC C 9 H 12 Dodecane C 12 H 26 PC Successfully used in KamLAND and Borexino KamLAND: mixture of 80% dodecane, 20% PC Attenuation length 8 m @ 430 nm Low flash point (48 C) Needs to be purified +Dodecane High transparency can compensate lower light yield Offers many free protons Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 7 / 21

Solvents: PXE and LAB Attenuation length 12 m (after purification High flash point (167 C) Fast (τ 1 = 2.63 ns) High density (0.986 kg/l) Attenuation length 20 m High flash point (140 C) Many free protons (6.6 28 per m 3 ) PXE C 16 H 18 LAB C 16 19 H 26 32 Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 8 / 21

Solutes PPO C 15 H 11 NO primary fluor absorption: 280-325 nm emission 350-400 nm PMP C 18 H 20 N 2 bismsb C 24 H 22 secondary fluor absorption: 320-370 nm emission: 384-450 nm large stoke shift absorption maximum: 294 nm emission maximum 415 nm Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 9 / 21

n spectra show four main peaks. For example ence of the fluorescence spectra on the solvent in figure7.5), the peaks are at approximately ashow of Emission the the solute energy PPOspacing have been Spectra between measured thewhen firstit is dissolved in the tab. vibrational Figure8.2levels showsof a comparison the groundof state, boths0i spectra, exhibiting two the molecule can not be resolved at room principle gives the relative intensities LAB + 2g/l PPO of the 1 PXE + 2g/l PPO the vibrational lines in the spectra of PPO, e literature 0.8 results [94]. The energy spacing cited triplet state is, in general, smaller than 0.6 avelengths. As above 500 nm no significant 0.4 that there is no evidence for phosphorescence n unstructured 0.2 fluorescence spectrum which 0]. Relative light output 0 ra of various concentrations of PPO in the 250 300 350 400 450 500 550 7.2 Spectra obtained by excitation Wavelength in with nm UV-light ison of the PPO spectra for the solvents PXE + 0.1 g/l LAB PPO and PXE. With PXE 1 PXE + 0.5 g/l MSB f about (right), nm respectively. to longer When wavelengths the UV-light PXE + 1.0 can g/l from PPObe the observed deuteriumand lamp a more illator, 0.8 mostly PXE molecules are excited. PXE + Rapidly g/l PPO the energy is transainly by nonradiative processes. Part of the energy in the PPO molecule onal level structure. PXE + 50 g/l PPO urther0.6on to bismsb molecules in the same manner (compare figure7.8 op, left). As the concentration of bismsb is low (in the order of several First 0.4 of all, a shift of about nm on all the PPO vibrational peaks the PPO molecules find no bismsb partner to transfer energy. As a emore, light is the emitted resolution by theofppo the molecules vibrational themselves. levels varies: After the peaks in the Solvent propagating 0.2 e, are thebetter light isresolved. absorbed These very efficiently effects by could bismsb: be explained although the byconcen- considering SB solvent has molecules. The surroundings of the benzene rings produce 0 been changed by a factor of four, the difference between both is hich figure changes 7.8 is small. the orbital In figure spacing 7.9, theof spectra the π-bonds. of PXE with A consequence 2 g/l PPO of 300 320 340 360 380 400 420 440 460 480 500 Wavelength in nm fference in the emission of the solvents PXE and LAB, namely 290 ively [5][8]. PXE consists of 2g/l two PPO + 20mg/l benzene bismsb rings and CH3 groups ypxe 1 one and benzene PPO. ring Left: andemission a long hydrocarbon-chain spectra where Fluor 2g/l PPO (see table5.1 in 20mg/l bismsb observed (position A, see figure7.4). Right: 0.8 ation after 1.4 cm propagation 98 (position B). Relative light output Relative light output 0.6 0.4 tinctive examples are plotted. The complete 0.2 Emission spectra depend on Solvent Type of fluor Amount of fluor Typically around 290 nm Desired to emit above 400 nm Relative light output Relative light output Relative light output PPO spectrum 1 0.8 0.6 0.4 0.2 0 250 300 350 400 450 500 550 600 Wavelength in nm bismsb spectrum 1 0.8 0.6 0.4 0.2 0 250 300 350 400 450 500 550 600 Wavelength in nm PMP spectrum 1 0.8 0.6 0.4 0.2 0 300 350 400 450 500 550 Wavelength in nm 0 250 300 350 400 450 500 550 600 Wavelength in nm T. Marrodán Undagoitia, T. Marrodán Undagoitia, PhD thesis om into the spectra of 2g/l PPO and 20mg/l bismsb (violet circles), arxiv:0904.4602 n 2008 empty circles) and 20mg/l bismsb (blue stars). All mixtures dissolved XE. The three-component scintillator shows the sum of the spectra of Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 / 21

Light Yield The amount of light emitted depends not only on the particle s energy: Type of scintillator Type of particle (quenching) Ionization and radiationless deexcitation processes lead to a loss of fluorescence efficiency Effect higher for heavier particles (protons, α) Affects also pulse shape Typically O(000) photons per MeV Energy resolution is correlated to the number of detected p.e. per MeV Number of photo electrons Primary light output Absorption losses PMT coverage PMT quantum efficiency Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 11 / 21

Light Output Measurements for LENA 6.6 Relative light output compared to PXE + 6 g/l PPO (maximum) Scintillator excited by 54 Mn source (γ 834 kev) Maximum of pulse height spectrum used for comparison Light output increases up to 6 g/l ChapterPPO 6. Fluorescence decay-time measurements of organic liquids Relative light output 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 PXE 1 2 3 4 5 6 7 8 9 Concentration of PPO in g/l Fluorescence decay time in ns 3 2 2 2 2 MaxA 900 Entries 0000 800 700 600 500 400 300 200 0 0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 Volts in V 1 0.9 0.8 0.7 0.6 0.5 0.4 LAB 0.3 1 2 3 4 5 6 7 8 9 Concentration of PPO in g/l T. Marrodán Undagoitia, PhD Thesis 2008 16: Spectrum Daniel Bick of(universität the pulse height Hamburg) in the close PMT. A Gaussian ANT 11 curve is fitted 1 October 11, 2011 12 / 21 Relative light output tput Fluorescence decay time in ns in ns 4

Florescence Times Contributions from decay of different exited states and other processes 4.6 Light propagatio Number of excited states n(t) = i n i e t τ i Log (amplitude) Decay constants τ i typical for each scintillator First (fast) component usually large amplitude time response Quenching: amplitudes n i of time components depend on de α particles energy ofdxthe excitation to fulfill the condition for resonance absorption S S 1 state can be overpopulated and the fast time constant can be measure i.e. without being affected by other transitions. When the excitation of Scintillation pulse shape depends on fluorescence decay times is performed by particle radiation, all processes discussed in section 4.3 ca Daniel Bick (Universität Hamburg) to the pulse ANT 11 shape. In chapter 6, measurements October of the 11, decay 2011 constants 13 / 21 of Protons Electrons Time (ns) Figure 4.4: Comparison T. of Marrodán the pulseundagoitia, shapes for PhD alpha-particles, Thesis 2008 protons a in a liquid scintillator. The amplitude of the pulses in logarithmic scale is function of time. particle identification by pulse shape analysis

Measurement of Fluorescence Times for LENA 2 2 1 1 Measurements 0 50 0 150 performed 200 250 300 350using 400 450 single photon sampling technique Time in ns 0 50 0 150 200 250 300 350 400 450 Time in ns Events / ( 1 ) 5 LAB + 2g/l PPO 4 3 2 Events / ( 1 ) 5 τ s,ppo 3.2 PXE + 1.845 2g/l PMP ± 0.02878 k h,pxe 0.7217 ± 0.07904 in ns Fluorescence decay time τ 1 3 4 2.8 3 2.6 2 2.4 2.2 2 1 0 50 0 150 200 250 300 350 400 450 Time in ns Fluorescence decay times decrease with fluor concentration Contribution of fast component between 75 and 95% Variations of τ 1 between 2 and 8 ns PXE significantly faster 1 1 2 3 4 5 6 7 8 9 Concentration of PPO in g/l 0 50 8 0 150 200 250 300 350 τ 400 450 s,ppo Time 1.186 in ± ns 0.1511 k h,lab 0.1241 ± 0.01516 7 in ns Fluorescence decay time τ 1 in ns me τ 1 6 5 4 3 2 1 2 3 4 5 6 7 8 9 Concentration of PPO in g/l T. 4.6 Marrodán Undagoitia, arxiv:0904.4602 τ s,pmp 3.297 ± 0.04734 k h,pxe 0.5714 ± 0.08766 4.4 Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 14 / 21

α/β dicrimination in LAB LAB + 2 g/l PPO (SNO+ Measurements) Deoxygenating the scintillator reduces quenching Increase of relative amount of longer time components Better particle discrimination Number of events per bin 4 3 Oxygenated alpha Oxygenated electron Deoxygenated alpha Deoxygenated electron Number of events per bin 3000 2500 2000 Oxygenated alpha Oxygenated electron Deoxygenated electron Deoxygenated alpha 1500 2 00 500 0 50 0 150 200 250 300 350 400 450 Time (ns) 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Peak/Total ratio H. M. O Keeffee, arxiv:12.0797 Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 15 / 21

Neutron Gamma Discrimination Discriminate recoil protons from gammas through pulse shape analysis Am-Be source as neutron + γ source Figure 4.6: Energy dependence of the peak distance parameter gamma and neutron peak (see equation(4.7)) in PXE (plotted in LAB (plotted in red). The distance µ n µ γ between the pe as σn 2 + σγ 2 decreases with the visible energy, which leads to t 2 MeV visible energy in PXE. In contrast to that, D increas visible energy in LAB, because σn 2 + σγ 2 decreases faster than µ Neutron rejection efficiency 97.8% γ acceptance Visible Neutron rejection efficiency Energy [MeV] in PXE [%] in LAB [%] 0.5-1.1 99.80 93.28 1.1-1.7 99.97 98.27 1.7-2.3 99.99 99.05 2.3-2.9 99.99 99.33 2.9-3.5 99.98 99.36 R. Table Möllenberg, 4.1: Energy Dipl. Thesis dependence 2009 of the neutron rejection efficiency i LAB with 97.8% gamma acceptance. Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 16 / 21

MC of Neutron Gamma Discrimination in LENA MC: Works also on large scales T.O.F. corrections necessary Neutron and positron events generated PXE LAB R. Möllenberg, Dipl. Thesis 2009 Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 17 / 21

Light Attenuation: Absorption and Scattering Optical transparency is one of the key parameters for large volumes Absorption prevents some photons from reaching the PMTs Scattering redirects the propagation direction of photons lengthens trajectory smears arrival time and hit patterns Different contributions to scattering Rayleigh scattering on electrons natural limit 40 m Absorption and reemission Mie scattering by impurities influences energy, time and spatial resolution Wavelength-dependent: Transparency increases with the wavelength. Low absorption and scattering lengths are required Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 18 / 21

Understanding 3D Light Transport 4. MUON-TRACK RECONSTRUCTION MC of point like e event in LENA Anisotropic (Rayleigh) scattering D. Hellgartner, Diploma Thesis, 2011 M. Wurm, arxiv:04.0811 alculated Fig. 4.25: value Sum l of the TOF-corrected hit time ray =45 m. Sample l is [m] l an [m] l S [m] χ 2 /ndf l ray ; spectra this corresponds of 50000 to 1MeV the pointlike PXE u electron 22.8±1.0 events 33.6±4.0 13.6±0.7±1.0 1.39 32 ination thewith center organic of impurianes which are not supposed PXE p 40.0±3.9 51±13 22.3±2.7±1.6 3.71 32 the detector. C12Additionally, sa 258±54 40.9±3.9 the 35.3±3.0±2.2 0.92 contributions of the directc12 and acscattered 132±16 photons aredifference shown. can be LAB p550 60.5±3.7 40.5±5.2 24.3±1.9±1.5 1.29 45 48.5±5.6 35.4±3.1±2.3 0.77 ompare the results for C12). LAB p500 75.3±5.3 40.2±4.4 26.2±1.9±1.6 1.23 45 significant erent brands. LAB 550q 66.3±5.7 40.0±4.6 25.0±1.9±1.6 0.80 45 PC i) cane was is assumed. the most transparoth the samples from Sigma Note that h 13.0±0.9 19.3±3.3 7.8±0.6±0.6 1.52 21 CX ˆn i = > 0 3 for45.0±4.5 a 44.9±4.5±2.9 0.74 44 thedaniel exponential Bick (Universität as λhamburg) h. ANT 11 October 11, 2011 19 / 21 6

Purity Issues Impurity threats Dust particles containing K, U, Th 222 Rn emanating form materials of construction 85 Kr and 39 Ar from air leaks 2 Pb and 2 Po on metal surfaces Impurities are a source for both background and attenuation 14 C is intrinsic to the scintillator and cannot be removed (long storage, old carbon) Distillation Removes impurities that are less volatile Does not remove noble gas impurities Gas Stripping Nitrogen flushing Removes remaining noble gases very efficiently Water extraction Washing the scintillator thus removing polar impurities Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 20 / 21

Conclusions State of the art liquid scintillators Large attenuation lengths up to O(20 m) Combination of PPO and bismsb provides high stokes shift and good light yield for both PXE and LAB PXE: faster than other scintillators LAB: better effective light yield (due to high attenuation length) PMP makes second fluor needless but has a reduced light yield LAB + PPO + bismsb currently favored for LENA Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 21 / 21

End Daniel Bick (Universität Hamburg) ANT 11 October 11, 2011 22

Single Photon Smapling Technique PMT Signal far PMT 000000 111111 000000 111111 000000 111111 000000 111111 000000 111111 000000 111111 000000 111111 111111 000000 Signal close PMT Scintillator container 0000000 1111111 0000000 1111111 0000000 1111111 0000000 1111111 0000000 1111111 0000000 1111111 0000000 1111111 0000000 1111111 1111111 0000000 23 October 11, 2011 ANT 11 Daniel Bick (Universita t Hamburg) 54 Mn Source PMT base