FIG. 1: DEAP-1 detector configuration. Liquid argon is contained in a stainless steel target chamber, which includes an inner acrylic cylinder and a diffuse reflector. Scintillation light from the liquid argon is wavelength-shifted by TPB on the inner detector surface and is transmitted to PMTs op- erating at room temperature through PMMA acrylic lightguides. The PMTs (5-inch ETL 9390B) are held in place by spring-loaded polyethylene supports, which maintain contact between the PMTs and lightguides.
FIG. 2: Calibration source geometry. The decay of 22 Na is tagged by gammas in the back NaI PMT, the annulus detector, and in DEAP-1, which allows a very low non-γ background for calibration of the pulse-shape discrimination.
FIG. 3: A sample event from a Na PSD calibration run. Shown are the recorded traces for each of the two PMTs, labeled A and B. The shaded area shows the prompt region which begins 50ns before the leading edge of thepulse and ends 150ns after the leading edge.
FIG. 4: The timing distributions for a typical run. The vertical lines indicate the cut values. For the top plot the cuts are always imposed at +/- 20ns. For the bottom plot the mean value is determined by a Gaussian fit and the cuts are imposed at the mean +/- 30ns.
FIG. 5: Single photoelectron spectra of the DEAP-1 PMTs, used to calibrate the absolute light yield and number of detected photons.
Fprompt versus energydistributionfor the triple-coincidence gamma-ray events. The region between 120-240 photoelectrons for Fprompt>0.7 contains no events.
FIG.7: Fprompt versus energy distribution for neutrons and gamma rays from an Am-Be calibration source. The upper band is from neutron-induced nuclear recoils in argon, the lower- band is frombackground gamma-ray interactions.
FIG. 8: Fprompt distribution for 16.7 million tagged gamma-ray events from the 22Na calibration, and nuclear recoil events from the Am-Be calibration, between 120 and 240 photoelectrons(approximately 43–86keVee). No gamma-ray events are seen in the nuclear recoil region.
FIG. 9: Comparison of Zfit distribution for gamma-rays from the PSD data, and for high-Fprompt backgrounds during the run (labeled Surface backgrounds). Also shown, for reference, is the distribution of high-Fprompt background events with the detector operating underground at SNOLAB.
FIG. 10: High-Fprompt background-event rate versus time. The average background rate is 4.6+/-0.2mHz.
FIG. 11: Light yield stability during the run. The average light yield is 2.8 photoelectrons/keVee. Shown are statistical uncertainties only. The scatter of the data points is within the systematic uncertainty of the measurement.
FIG.12: Measured lifetime of the triplet component in liquid argon during the run. Uncertainties shown are statistical only. The scatter in the data points is consistent with the systematic uncertainty of themeasurement. No significant change in the lifetime was seen.
FIG. 13: The noise level sigmaP in the prompt time window is shown as a function of the total number of photoelectrons in the pulse. The broken lines show how the four sources of noise and uncertainty that are taken into account in our model contribute to the total noise level.
FIG. 14: The noise level sigmaL in the late time window is shown as a function of the total number of photoelectrons in the pulse. The broken lines show how the four sources of noise and uncertainty that are taken into account in our model contribute to the total noise level.
FIG. 15: Comparison of the data and the analytic model in the region 60-120 photoelectrons (approximately 21 to 43 keVee). The gray shaded area represents the uncertainty in the noise parameters and in the energy calibration.
FIG. 16: Comparison of the data and the analytic model in the region 120-240 photoelectrons (approximately 43-86 keVee). The gray shaded area represents the uncertainty in the noise parameters and in the energy calibration.
FIG. 17: Pleak distribution from 22Na calibration data from DEAP-1, and analytic models with and without additional noise parameters for 120-240 photoelectrons. The lower curve shows the expected backgrounds in the measurements from high-Fprompt events.
FIG. 18: Pleak distribution expected for 20–40keVee with 6 photoelectrons/keVee yield. Shown are the distributions with and without systematic noise terms, and the Fprompt valuesfor 50 and 90 percent nuclear-recoil acceptance.
FIG. 19: Dark matter sensitivity of liquid argon. Shown are the current experimental limits fromthe CDMS and XENON-10 collaborations, and the expected sensitivity for 1000kg of liquid argon with a 20keVee threshold, and with a 19keVee threshold for argon that has been depleted in 39Ar by a factor of 20.