Updates and Status of the Noble Element Simulation Technique Jon Balajthy CPAD Workshop December 10, 2019 1
2
Noble Element Simulation Technique (NEST) Simulation package for modelling response of noble element detectors ● Interpolates between existing calibration datasets to provide predictions for arbitrary energies and fields ● LXe is most fully developed model, LAr is currently in development ● Collaborators from wide range of experiments and institutions ○ LUX/LZ, XENON, (n)EXO, DUNE Main webpage: http://nest.physics.ucdavis.edu/ Github page: https://github.com/NESTCollaboration/nest Several implementation platforms: ● C++ based executable ● Geant4 integration ● Python wrapper ● Online calculator 3
The NEST Physics model NEST is a set of mostly empirical models that simulates S1 and S2 size, timing, and location The inputs required: ● Event energy and/or interaction type ○ Beta/Compton (spectra for H-3 and C-14 included) ○ NR (AmBe, B-8, WIMP, neutrons) ○ Kr-83m ○ Heavy Ions ○ Alphas ○ Photoabsorption ● Drift field profile (can be a constant or a detailed map) ● Detector parameters (g1, g1_gas, density, extraction efficiency, etc.) 4
Current Status Very brief overview- more detailed information available at NEST website (http://nest.physics.ucdavis.edu/) Also see Greg Rischbieter’s talk from DANCE conference (https://indico.cern.ch/event/824917/contributions/3571627/att achments/1934381/3205056/NEST_DANCE_GR.pdf) 5
Nuclear Recoil Yields Same yields model used for WIMPs, D-D, B-8, Cf, AmBe Recent data from LLNL improves model at low energy Need Ly data in <1keV region M. Szydagis. (NEST Collaboration) A Comprehensive, Exhaustive, Complete Analysis of World LXe NR Data With a 6 Final Model. 2019.
Electron Recoil Yields Currently divided into two(ish) parts: beta/Compton ● Compton/beta yields ● Photo-absorption ● Also includes special case of Kr-83m Ly and Qy are complementary, (Ly = 1/W - Qy) photoabsorption 7
Ions and alphas and krypton, oh my! alpha-NR Heavy Ions Kr-83m 8
Resolution Model NEST doesn’t just do mean yields! Fluctuations from recombination + detector resolution, tuned to match band means+widths LUX C-14 LUX D-D Improved Modeling of β Electronic Recoils in Liquid Xenon Using LUX Calibration Data LUX Collaboration (D.S. Akerib et al. ). Oct 9, 2019. 17 pp. e-Print: arXiv:1910.04211 [physics.ins-det] 9
Recent/Ongoing Updates 1. LArNEST 2. Low energy NR model 3. ER model overhaul a. Inner shell interactions (nu-e, EC) b. Compton vs beta 4. Work Function 10
LAr ER Model Coming soon! Models based on LXe equations and fit to LAr calibration data 11
LAr NR Model NR, too! Low energy behavior modeled after LXe measurements 12
Low Energy NR Model New data from Livermore improves sub-keV Qy, but Ty and Ly are still unknown Need total yields measurements in sub-keV region ● Low energy behavior could trend high or low Lenardo, Brian, et al. "Measurement of the ionization yield from ● Difficult measurement to make due to nuclear recoils in liquid xenon between 0.3--6 keV with small LY at low E NR single-ionization-electron sensitivity." arXiv preprint arXiv:1908.00518 (2019). 13
ER-Model Overhaul ● Re-fit beta model using C-14 from LUX and H-3 data from XENON and XeLDA ● (?)Separate beta and Compton models ● Develop new electron capture/ nu-e scattering model 14
Beta vs EC XeLDA detector at FNAL observed offset of H-3 vs Xe-127 L-shell capture Primary decay mode: 5.1keV is divided ~evenly between 6 electrons -> Higher dE/dx means more recombination Slide from Dylan Temple’s TAUP talk 15
Beta vs EC This effect was also observed in LUX, but was within systematic error. Is not observed in K, M, or N Preliminary plot shells from Dylan Temples ER->NR leakage fraction for L-shell captures increases from Identical ~0.5% to ~3% offset in LUX and XeLDA Leads to ~20% increase in neutrino burden in LZ G3 experiments will have Need to develop L-shell NEST model O(10,000) 𝜉 -e scatters (~7% are 𝜉 -L, or ~1,000 events) ● Measurements limited to < 300 V/cm ● Direct 𝜉 -e measurements could be XeLDA result predicts ~30 leaked informative events vs. ~5 for current NEST 16
Beta vs Compton Possibly a similar effect in Compton scattering vs beta decay Tritium data is very consistent, but NEST tends to overshoot at high field -> Comes from previous lack of beta data above 180 V/cm 150-180 V/cm ~275 V/cm 330-366 V/cm ~100 V/cm 17
Beta vs Compton Compton data is also consistent, but disagrees with all available beta data No strong physical explanation for disparity ~2000 V/cm ~410 V/cm ~500 V/cm ~180 V/cm Need to investigate possible split of CS/beta models ● Compton/beta measurements from the same detector would cancel any possible systematic effects (possibly forthcoming from XeLDA?) 18
W-value: New EXO result The new EXO yields calibration includes a measurement of the average work function (W) Disagrees with accepted results by ~15% W (eV/quantum) Existing measurements of W agree with 13.7 +/-~0.2 eV Note: mass density shown, but # density is important. EXO arXiv:1908.04128v1 effective density is 2.9 g/cc. Density (g/cc) Most commonly cited: Absolute measurement from E. Dahl 19 Not listed: 13.4 +/-0.4 eV from neriX. W floated in MC fit.
W-value: Excess light yield? Points include CS and photoabsorption, so expect to fall between beta and gamma Qy seems ~correct, while Ly is high -> Additional IR photons detected by APDs? -> Add 35% to Ly for silicon PDs -> Adding ~15% to both Ly & Qy doesn’t work (Green is modified NEST model ) Least kludgy method- assumes everyone is right IR scintillation is expected (but not this much) Need more measurements ● Want to verify our explanation ● If confirmed, IR scintillation will need to be measured and modeled. A structured 35% effect might change the field/energy dependence, or even break the combined energy model (no more 1:1 recombination?) 20 arXiv:1908.04128v1
Conclusions ● NEST is a robust set of empirical models that currently matches all world data to ~10% (most places much better) ● There are still several open questions caused by lack of data/ contradictory data a. Total yield of low energy NR- Data does not exist b. L-shell interactions- Known to differ, but only measured at 2 fields in Xe-127 EC c. Beta v. Compton- All available data shows ~10-20% discrepancy. d. W value- Few absolute measurements available. Those that exist disagree by ~15%. Might be explained by detector effects 21
Extra Slides 22
IR Scintillation in LXe Significant IR scintillation is observed in SD 630-70-75-500 APDs GXe (~= VUV) Few direct measurements in LXe (none at wavelengths <1um ) Only measurements show LXe << GXe for IR scintillation PMT cutoff Need ~17 ph/keV NIR to explain EXO result ● GAr NIR yield is 17.5 ph/keV (Xe is higher) ● LAr yield is ~0.5 ph/keV (times ~30% QE) ● LXe NIR yield needs to be ~50x yield in LAr ( A Bondar et al 2012 JINST 7 P06014 ) DOI:10.1016/S0168-9002(00)01249-3 23
The NEST Physics model 1. a. Draws number of quanta given event type/energy/field b. Divides into N e , N γ according to E Extraction Gas mean yields model + recombination fluctuations 2. Propagate photons and calculate S1 E Drif size/timing e- e- t e- e- e- ○ Pulse shape model is included Liquid ○ Detailed hit-mapping is in - development + + * - + - * 3. Propagate electrons calculate drift time and S2 size 24
Recombination Model Recombination is integral to the observed of charge and light yield Note that recombination does not affect the total number of quanta. 25
Recommend
More recommend
