Scintillation Light from Cosmic-Ray Muons in Liquid Argon 5 - - PowerPoint PPT Presentation
Scintillation Light from Cosmic-Ray Muons in Liquid Argon 5 November, 2015 Denver Whittington Stuart Mufson Bruce Howard Indiana University Outline Goal: Measure the time structure of the scintillation signal from liquid argon after
5 November 2015
2
➢ Goal: Measure the time structure of the scintillation signal from
➢
➢ Experiment
➢
➢
➢
➢ Scintillation structure analysis ➢ Physical model of signal ➢ Comparison of Models ➢ Results
5 November 2015
3
➢ TallBo at Fermilab (PAB)
➢
84” LAr dewar
➢
Data collected Nov./Dec. 2014
➢ Ultra-high purity liquid argon
➢
Vacuum to remove residual atmosphere
➢
Condenser to maintain closed system
➢
Active N2, O2, and H2O monitoring
➢ O2 ~40 ppb (negligible) ➢ N2 < 200 ppb (negligible) ➢ H2O ~8ppb (negligible)
➢ Multiple light guide designs
➢
Dip-coated acrylic bars
➢
Cast acrylic and polystyrene bars
➢ Hodoscope (cosmic ray) trigger
➢
2 8x8 Arrays of PMTs + BaF crystals
➢ CREST cosmic-ray balloon experiment
➢
2 scintillator paddle planes
➢ Allows shower rejection,
reconstruction of single tracks
5 November 2015
4
➢ Large active area UV-collecting light guides
➢
➢ 20 inch prototypes tested in this experiment
➢
➢
➢ Four light guides analyzed
5 November 2015
5
➢ Biased at 24.5 V (low noise, high gain) ➢ Excellent single-pixel resolution → ➢ Characteristics measured in LN2
➢
6
➢
➢
➢
all waveforms average waveform
5 November 2015
6
➢ Example waveform from a hodoscope-selected track
➢
➢
➢
➢ Superposition of
all waveforms average waveform
5 November 2015
7
➢ Average signal is convolution of illumination with SiPM response ➢ Use Gold deconvolution algorithm (in ROOT TSpectrum) to
➢
➢
➢
➢
average cosmic ray waveform measured by SiPM k
average time-dependent signal of scintillation photons “illumination function” average single-pe SiPM response function
(inset from slide 6) (inset from slide 5)
5 November 2015
8
➢ Fundamental signal expected to be an exponential probability
➢ “Exponentially-modified Gaussian” (EMG) function
➢ Gaussian-like rise with exponential tail
➢ Multi-Component Fit
➢ Two components insufficient ➢ Best fit with found using
➢ Early-light component ➢ Intermediate component
➢ Frequently reported
➢ Late-light component ➢ Fourth component
➢ Describes behavior
at > 6 μs
5 November 2015
9
➢ Features of note
➢
➢ Compatible with other measurements
➢
➢ Compatible with other measurements
using beta and gamma sources
➢
➢ Likely due to additional fast scintillation
from polystrene (reported elsewhere)
➢ Don't have sensitivity to resolve
this substructure here
➢
➢ Present in acrylic light guides only
5 November 2015
10
➢ The intermediate and fourth components appear instrumental
➢
➢ Data refit using a physical model description for the illumination
➢
➢ Exponential probability distribution functions
➢
➢ Exponential probability distribution functions
➢
➢ All convolved with a Gaussian function
➢ Result is again a sum of EMG functions, reparameterized to
➢
5 November 2015
11
➢ Same quality fit, amplitudes easier to interpret
➢
➢
5 November 2015
12
➢ Agreement in results
➢
➢
➢
➢ Similar delayed emission recently reported in Phys. Rev. C (E. Segretto) ➢ Agrees with “early light” as the 70% of singlet light converted promptly
5 November 2015
13
➢ Additional cross-check: Calculate “prompt fraction”
➢
➢
➢
➢
➢ t* = 130 ns, tf = 10 μs
➢ Same result of Fprompt ≈ 0.3
5 November 2015
14
➢ Triplet state Ar2* eximer lifetime measured as 1.52 μs ➢ Physical model indicates that ~30% of scintillation light is
➢ Calculation of “prompt fraction” agrees with results for electrons
5 November 2015
15
➢
Recovered time-dependent structure of scintillation signal detected by DUNE light guides with SiPMs by deconvolving the average SiPM single-pe response from the average cosmic-ray signal.
➢
Phenomenological model
➢
Physical Model
➢
Measured scintillation parameters associated with cosmic-ray muons in LAr
➢
τT = 1.52 μs
➢
Early light fraction ~25%
➢
Delayed emission from WLS
➢ ~30% effect
➢
Singlet LAr fraction ~36%
➢
Prompt signal compatible with various other electron signal measurements
5 November 2015
16
5 November 2015
17
➢ Injected xenon into the liquid argon
➢ GXe mixed with GAr, heated, and injected into the liquid at ~150 psi ➢ Increments of 20 ppm (by volume) ➢ Time structure determined using same deconvolution procedure
➢ 1.52 μs tail replaced by broad signal at ~200 ns (20 ppmv) ➢ Broad signal becomes more prompt as concentration increases ➢ Further analysis to be done
➢ Prompt signal possibly diminished ➢ Hodoscope-triggered data hints at ~50% more light from Xe-doped LAr
Time-dependent structure of the LAr+Xe signal Cumulative scintillation signal from LAr+Xe
(area normalized)
5 November 2015
18
➢ Charged particles create diatomic Ar-Ar eximers (Ar2*) ➢ Result is a prompt singlet signal and a long-lived triplet signal
➢
➢
(singlet)
(triplet)
e- μ-
e-
50% 50% 35% 65% 1.5 μs 5 ns 128 nm
5 November 2015
19
➢ Similar results on all four light guides
➢
5 November 2015
20
➢ Cross check: additional light guides
➢
➢
➢ Wide range of
lifetimes for intermediate component
➢ Clear separation
between acrylic and polystyrene