cenpa
play

CENPA Center for Experimental Nuclear Physics and Astrophysics ! 1 - PowerPoint PPT Presentation

May 22, 2023 •242 likes •421 views

Dark matter search results from DAMIC at SNOLAB Alvaro E. Chavarria University of Washington CENPA Center for Experimental Nuclear Physics and Astrophysics ! 1 Outline Charge-coupled devices to search for dark matter. Response of

  1. Dark matter search results from DAMIC at SNOLAB Alvaro E. Chavarria University of Washington CENPA Center for Experimental Nuclear Physics and Astrophysics ! 1

  2. Outline ‣ Charge-coupled devices to search for dark matter. ‣ Response of DAMIC CCDs to signal and backgrounds. ‣ DAMIC at SNOLAB. ‣ DM-e scattering search ( results ). ‣ WIMP search ( status ). ! 2

  3. Charge coupled device y Pixel array σ xy σ xy ~ z x z y Free 675 µm charge ± x x carriers Ionizing particle z Fully depleted Device is “exposed,” collecting charge until substrate user commands readout Silicon band-gap: 1.2 eV Mean energy for 1 e-h pair: 3.8 eV 15 µm Standard fabrication in semiconductor industry and easy cryogenics (~100 K) ! 3

  4. Perfomance Pixel charge distribution 1330 3 Electron 10 Low-energy 1320 candidates 2 10 1310 ! = 5.9 eV 50 pixels = 1.6 e - 10 α 1300 1 Muon 1290 15x15 µm 2 26 13 0 13 26 − − pixels Energy / eV Very low noise and dark current 1280 4180 4190 4200 4210 4220 5 10 15 20 25 30 5 10 15 20 25 30 Energy measured by pixel [keV] lowest dark current ever measured particle identification and in a silicon detector: background characterization 5x10 -22 A/cm 2 (at 140 K) ! 4

  5. Detector response Mn K from front and back α 7 Energy / keV 1615 200 6.8 1610 180 0 2 4 >6 y [pix] 1605 Ionization [keV ee ] 6.6 160 1600 6.4 1595 140 6.2 1590 120 1585 6 100 1580 1380 1400 1420 1440 1460 5.8 80 x [pix] 5.6 60 5.4 40 Front Back 5.2 z reconstruction with X rays 20 5 0 and cosmic rays 0 0.2 0.4 0.6 0.8 1 1.2 1.4 σ / pixels xy 1.08 ) ee X-rays k(E) / k(5.9 keV 1.06 Optical photons CCD linearity down 1.04 1.02 to 40 eV ee with 1 optical photons 0.98 0.96 1 − 10 1 10 Ionization signal [keV ] ee ! 5

  6. Nuclear recoil response ‣ Detector response calibrated with 24 ] 10 ee Dougherty (1992) [keV Gerbier et al. (1990) keV neutrons from 9 Be( γ ,n) reaction. Zecher et al. (1990) 124 9 Sb- Be (2016) e Antonella (2016) E ‣ By comparing data and Monte Carlo Lindhard, k=0.15 1 spectra, ionization efficiency was PRD94 082007 measured to be lower than predicted Calibration by Lindhard model. down to 60 eV ee 1 − 10 ‣ Also validates diffusion model at low energies. 1 10 E [keV ] r nr ] 1000 Number of events per bin -1 Data - full BeO ) Monte Carlo ee 0.025 Number of nuclear recoils [(10 eV Simulation Best-fit with Monte Carlo spectrum reproduces σ xy 800 SbAl + FullBe 0.02 Single-recoil spectrum distribution at 2 / ndf � 142 / 154 600 very similar to signal low energies Prob 0.74 0.015 -1 f (0.06) 0.63 � 0.01 from 3 GeV WIMP. -1 f (0.3) 1.94 � 0.02 400 End-point = 3.2 keV r f(3.2) 0.61 � 0.02 0.01 y offset 1.4 � 1.0 200 Data <0.15 keV ee 0.005 Simulation 0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 σ / pixels E [keV ] xy e ee ! 6

  7. Flexibility in readout Pixels can be readout in “groups” and the total charge estimated in a single measurement. Less pixels but same noise per pixel! 1x10 3x3 Loss of x, y and z information 55 Fe from back: 1x1 Data shows clear improvement in 1640 1660 1680 1700 1720 1740 5510 5515 5520 5525 5530 5535 5540 5545 energy resolution 1x1 α - β coincidence 5 6 7 8 Energy [keV] ! 7

  8. SNOLAB Installation 16 Mpix CCD 5.8 g Poly- 6 cm VIB ethylene Lead Kapton Copper Lead block Kapton signal cable module signal cable Cu box with CCDs J. Zhou Cu vacuum vessel ! 8

  9. Current status ‣ 7 CCDs in stable data taking since 2017 (1 CCD sandwiched in ancient lead). ‣ 40 g target mass. ‣ Operating temperature of ~140K. ‣ Exposure for image: 8h and 24h 
 (each image acquisition is followed by a “blank” exposure) . ‣ 7.6 kg-day of data for background characterization in 1x1 format. ‣ 13 kg-day of data collected for DM search in 1x100 format. ‣ Since Jan 2019, resumed background run and detector studies (e.g., 125 K operation for lower leakage current) in preparation for DAMIC-M . ! 9

  10. Leakage current analysis CCD 1 CCD 2 CCD 3 CCD 4 CCD 5 CCD 6 CCD 7 Pixel distribution of 200 g-d of ‣ Select CCDs with constant data in 100 ks exposures leakage current. ‣ Compare pixel distribution to Bulk leakage current at the level leakage-only hypothesis + of 2 e - mm -2 d -1 at ~140 K signal from DM-e interactions. (Before 4 e - mm -2 d -1 at 105 K) ! 10

  11. DM-e results arXiv:1907.12628 Best exclusion limit for the absorption of hidden photons with masses 1-10 eV/ c 2 Best exclusion limits for the scattering of dark matter particles with masses <5 MeV/ c 2 ! 11

  12. WIMP Search ‣ Remove pedestal and subtract correlated noise. ‣ Mask defects: repeating patterns in images. ‣ Select images with expected noise profile. ‣ Perform a log-likelihood fit for a signal in a moving window across the image. Δ LL = ℒ n - ℒ s Example of one event Gauss signal flat noise + flat noise E = 0.14 keV, σ = 0.5 Δ LL = -130 For every event we have its statistical significance Δ LL above noise , its amplitude ( E , energy) and its spread ( σ x proportional to z ) ! 12

  13. Noise rejection ‣ We introduce leakage current on the blank (zero-exposure) images using a simple Poisson model. ‣ We run the full cluster extraction to obtain the Δ LL profile for “noise” clusters. ‣ Select a Δ LL value that removes all noise and calculate the event selection efficiency. dLL dLL 10% efficiency at 50 eV ee analysis threshold ! 13

  14. Background model ‣ Background model constructed from full particle tracking + detector response Monte Carlo. Two-D (E, σ x ) fit to data above 6 keV ee with constraints from known radioactive contaminants. D. Baxter ’s presentation from yesterday! ‣ Dominant systematic uncertainty are radioactive contaminants on the back of the active region, e.g., implanted 210 Pb or 3 H migration. Reconstructed depth allows to distinguish from WIMP signal. Background model Comparison of Back Exponential and WIMP Signal 1.2 4 [pixels] 1.2 [pixels] Back 3.5 1 1 3 x x σ σ 0.8 0.8 2.5 ~5 d.r.u. 
 0.6 0.6 2 22 Na Bulk Back Surface Exponential, α =0.5 keV 1.5 0.4 0.4 1 -2 WIMP Signal, M=2 GeV c 0.2 0.2 0.5 Front 0 0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0.2 0.4 0.6 0.8 1 1.2 1.4 Energy [keV ] Energy [keV ] ee ee ! 14

  15. Expected sensitivity ‣ Independent 2D unbinned likelihood fit with background model + WIMP signal to search for dark matter. ‣ Free parameters included in background model to account for systematic uncertainties. ‣ Analysis in its final stages. Results soon! ‣ We use latest background model and full analysis to generate expected sensitivity. ‣ Potential for discovery of WIMPs with masses 1–2 GeV/ c 2 . ‣ Result can exclude a significant fraction of CDMS II-Si. ! 15

  16. Conclusions ‣ DAMIC at SNOLAB has demonstrated CCDs as an excellent technology for dark matter direct detection. ‣ Extensive understanding of CCD response and backgrounds for an experiment with potential for discovery. ‣ Best results for DM scattering with masses <5 MeV/ c 2 . ‣ WIMP search data campaign complete. Exposure of 13 kg-d under analysis. Expect results soon. ‣ Particularly good sensitivity for WIMPs with 1-2 GeV/ c 2 . ‣ Next step in the program: DAMIC-M. See P. Privitera talk later today. ! 16

  17. DAMIC Collaboration Thank you! ! 17

More recommend