Limits on light WIMPs from the CDEX-1 experiment with a p-type point-contact germanium detector at the China Jinping Underground Laboratory
We report results of a search for light Dark Matter WIMPs with CDEX-1 experiment at the China Jinping Underground Laboratory, based on 53.9 kg-days of data from a p-type point-contact germanium detector enclosed by a NaI(Tl) crystal scintillator as anti-Compton detector. The event rate and spectrum above the analysis threshold of 475 eVee are consistent with the understood background model. Part of the allowed regions for WIMP-nucleus coherent elastic scattering at WIMP mass of 6-20 GeV are probed and excluded. Independent of interaction channels, this result contradicts the interpretation that the anomalous excesses of the CoGeNT experiment are induced by Dark Matter, since identical detector techniques are used in both experiments.
pacs:95.35.+d, 29.40.-n, 98.70.Vc
Corresponding author: ]firstname.lastname@example.org Corresponding author: ]email@example.com
Current direct-detection dark matter experiments aim at searching of the Weakly Interacting Massive Particles (WIMPs, denoted by ) via elastic scattering of nuclei in terrestrial detectors: cdmpdg12 (). Of particular interest are the potential positive signatures implied by data from the DAMA dama (), CoGeNT cogent (), CRESST-II cresst2 () and CDMS-II(Si) cdms2si () experiments. Such interpretations, however, are in conflict with the null results from other experiments texono13 (); lux (); xenon100 (); cdmslite (); supercdms (). Additional experiments with improved sensitivities to probe this parameter space are crucial.
Germanium detectors with sub-keV sensitivities were identified ulege () as effective means to probe the light WIMP regions, motivating development of point-contact germanium detectors () ppcge () and various experimental searches texono09 (); texono13 (); cogent (); cdex1 (); malbek (). In particular, the CoGeNT experiment cogent () with a 443 g detector reported possible WIMP-induced events as well as annual modulation signatures.
The China Dark Matter Experiment(CDEX) pursues direct searches of light WIMPs towards the goal of a ton-scale germanium detector array at the China Jinping Underground Laboratory(CJPL) cjplnews (); cdexcosmic () located in Sichuan, China, with about 2400 m of rock overburden. The combined cosmic-ray direct and induced rates at CJPL has been measured cdexcosmic () to be 61.7 m yr, consistent with expectations. Studies from a prototype CDEX-0 detector array with 20 g target mass at CJPL were reported cdex0 (). Our earlier measurement at CJPL from the first phase of CDEX experiment (CDEX-1) cdex1 () is with a of target mass 994 g and analysis threshold of 400 eVee (“ee” denotes electron-equivalent energy), but in the absence of Anti-Compton (AC) detector and prior to surface event suppression. We report new CDEX-1 results in this article with these two crucial features incorporated. A cylindrical NaI(Tl) crystal scintillator with a well-shaped cavity enclosing the target serves as the AC-detector. Identification of surface background and derivation of efficiency factors follow the procedures of Ref. texono14 (). Details of the hardware setup and shielding configurations can be referred to Refs. cdex1 (); cdexshielding ().
Signals from the point-contact electrode are processed by a pulsed reset preamplifier with three identical outputs. Two of them are distributed to the shaping amplifiers at 12 s and 6 s shaping time which also provides the trigger for data acquisition (DAQ), as well as the energy measurement (T). The remaining one is loaded to a timing amplifier (TA) which provides the fast rise-time information. The outputs are digitized by flash analog-to-digital converters at 100 MHz. A total of 58.7 days of data was recorded at a trigger rate of 5 Hz. The DAQ dead time is 0.1%, as measured by events due to random triggers (RT) generated by a precision pulser. Energy calibration was achieved by the cosmogenic X-ray peaks and the zero-energy was defined by the pedestals of RT events. Deviations from linearity is less than 0.8%. The trigger efficiency was unity above 320 eVee, as verified by in situ physics events via an extrapolation of the amplitude distributions to sub-noise edge energy texono09 (). The selection of candidate events based on timing correlation and basic pulse shape discrimination (“Basic Cut”, denoted by BC) as well as the derivation of their efficiencies were discussed in our earlier report cdex1 (). The microphonics effects and electronic events induced by the preamplifier reset timing were completely suppressed and the combined efficiencies of 86.3% were accurately evaluated. Events in anti-coincidence (coincidence) with the AC-detector are denoted by AC(). The AC selection discriminate -ray induced background at a signal efficiency of 100%, as measured by RT events.
|Energy Bin||0.475-0.575 keVee||1.975-2.075 keVee|
|ACB and Errors (kgkeVeeday)|
|I) Statistical Uncertainties (Combined) :||1.47||0.97|
|(i)Uncertainties on Calibration (,) :||0.32||0.08|
|(ii)Derivation of (,)-corrected Bulk Rates :||1.43||0.97|
|II) Systematic Uncertainties (Combined) :||0.87||0.27|
|(i) Rise-time Cut-Value||0.27||0.12|
|(ii) Fiducial Mass||0.05||0.05|
|(iii) Normalization Range (3-5 keVee)||0.07||0.01|
|(iv) (,) = (B,S) at Normalization||0.10||0.10|
|(v) Choice of Discard Region||0.30||0.06|
|(vi) Source Location||0.28||0.19|
|(vii) Source Energy Range and Spectra||0.72||0.12|
The surface electrode of is fabricated by diffusion of lithium ions, having a finite thickness and producing events with anomalous charge collection and rise time () surface (); texono14 (). The thickness of the Surface layer, including both the dead and inactive layers, was derived to be (0.970.15) mm, via the comparison of measured and simulated intensity ratios of various -peaks from a Ba source, with the thickness of the copper cryostat being the leading contribution to the uncertainties ppcmj12 (). This gives rise to a fiducial mass of 919 g and data strength of 53.9 kg-days for this analysis. Surface events have anomalous charge collection and risetime () distributions. The values are measured by fitting to a hyperbolic tangent function to the TA signals texono14 (). The -distribution of AC events is depicted in Figure 1a, showing the two-band structure characterizing bulk (B) and surface (S) events. Typical B and S events at 500 eVee, together with their fitted-profiles, are displayed in the inset. The distributions of AC and AC events at two energy-bands are depicted in of Figure 1b, together with those from Co -source. The three samples match well, indicating that the in situ background are dominated by ambient high energy (MeV range) ’s. Events due to low energy ’s from Am -source, also superimposed, show differences in the -distributions in S.
The -cut() for differentiating the observed B and S events is set at 0.8 s. Two factors are necessary to translate the measured rates (B,S) to the actual rates (B,S)- the B-signal retaining () and S-background rejection () efficiencies. These are related by the coupled equations:
The normalization of (B,S)=(B,S) is set at the high energy range of 35 keVee where the separation of the two bands is larger than the -measurement resolution.
Calibration data with Am, Co, Cs and Co are adopted to evaluate (,). Figures 2a shows the measured B spectra and their corresponding reference B derived from simulation. The allowed bands in (,) derived from the calibration data at 475-575 eVee and 2.8-3.0 keVee are illustrated in Figure 2b. The bands overlap at a common region indicating the results are valid over the entire energy range of interest. The measured (,) as a function of energy are depicted in Figure 2c. An additional consistency measurement is provided by the ratio of the Ga-L X-rays at 1.3 keVee after selection to its original intensity predicted by the Ga-K X-ray at 10.37 keVee.
|(Discard Am Data)||[sys.]||[sys.]||[sys.]|
The raw spectrum and those at various stages of selection procedures are depicted in Figure 3a. The peak at 600 eVee is due to induced electronic noise from preamplifier resets and is completely rejected by timing correlation in BC cdex1 (). The (,)-corrected spectra of the candidate events, defined as ACB and shown in the Figure 3b, can be derived via the solution of Eq. 1:
The peaks correspond to known K-shell X rays from the cosmogenically-activated isotopes. The analysis threshold is placed at 475 eVee, below which the sensitivity is constrained by the noise edge. The AC spectra, depicted in the inset of Figure 3a, correspond to events from ambient -rays which are in coincidence with the NaI(Tl) detector. An expected flat ACB spectrum down to threshold is obtained, demonstrating that the (,)-correction is valid.
The various components which contribute to the errors of ACB at threshold and at a typical high energy bin are summarized in Table 1. Uncertainty values of (,) and B at threshold are listed in Table 2. Systematic uncertainties originate from (i) parameter choices of the analysis procedures and (ii) possible differences in the locations and energy spectra between the calibration sources and background events. The combined errors of ACB are dominated by the statistical uncertainties in the measurement of (B,S), boosted by the factor of [1/(+-1)] from Eq. 2 as (,) deviates from unity below 1.5 keVee. Contributions of systematic uncertainties are minor (increasing the total error of ACB from 1.47 to 1.71 kgkeVeeday at threshold) and are taken into account in the analysis. The variations of the key parameters over changes of within the -scan range in Figure 1a are studied. The B spectra are stable, robust and independent of , as indicated by the small variations relative to the uncertainties.
The high statistics of Am surface events produce the narrow vertical bands in Figures 2b, which in turn drives the small statistical error of in Figures 2c. An additional stress-test was performed. The Am measurements are discarded altogether, and (,) are derived with the Co, Cs and Co data. The results are summarized in Table 2. The B at the threshold energy bin would be shifted only by 4.2% from 4.09 to 3.92 kgkeVeeday. The effects on the subsequent physics analysis are therefore negligible.
High energy -rays from ambient radioactivity produce flat electron-recoil background at low energy, as predicted by simulations and is verified by the Am, Cs and Co spectra of Figure 2a. The L-shell X-ray lines are predicted by the K-shell peaks. Both background are subtracted from the -corrected ACB spectrum as shown in Figure 3b. A minimum- analysis is applied to the residual spectrum within 0.475 and 2.25 keVee, adopting two free and positive definite parameters which characterize the flat ambient -background and the possible -N spin-independent cross-section (), respectively. Conventional astrophysical models cdmpdg12 () are adopted to describe WIMP-induced interactions, using the local WIMP density of 0.3 GeV/cm, the Maxwellian velocity distribution with =220 km/s and the galactic escape velocity of =544 km/s. The quenching function (QF) in Ge is evaluated with the TRIM software package trim (). The derived QF is depicted in Figure 4 with measured data geqf () showing good agreement over a large range of nuclear recoil energy. A systematic uncertainty of 10% is taken, corresponding to the spread of individual data points, as well as the deviations with the alternative Lindhard model lindhard (). Analysis is performed by scanning QF within 10% of their nominal value, and the most conservative constraints are adopted as the limits.
As illustration, the best-fit spectrum at =8 GeV, where =(-1.809.28)x10 cm at /dof=8.11/16 (p-value0.95), is depicted in Figure 3b, with the band representing the 2 uncertainties. Exclusion plot of versus at 90% confidence level is displayed in Figure 5. The bounds from other benchmark experiments are superimposed cogent (); cdmslite (); lux (); supercdms (). As comparison, different QFs (parametrization of CoGeNT cogent () and Lindhard theory with k=0.157 of CDMSlite cdmslite ()) are used to derive alternative exclusion curves, also displayed in Figure 5. It can be seen that the analysis procedures adopted in this work provide the most conservative constraints.
An order of magnitude improvement in the sensitivities of has been achieved over our previous results cdex1 (). Part of the light WIMP regions within 6 and 20 GeV implied by earlier experiments are probed and rejected. In particular, the CoGeNT anomalous events at sub-keV energy cogent () are not reproduced in these results based on identical detector techniques. This strongly disfavors the excess is induced by dark matter, independent of interaction channels. For instance, electromagnetic final states are not constrained by experiments like SuperCDMS supercdms () which measure nuclear recoil events. They can, however, be probed by the CDEX-1 data.
The CDEX-1 experiment continues to accumulate data at CJPL. Research programs are pursued to further reduce the physics threshold via hardware and software efforts. Time modulation of the data will be studied. A PCGe array of 10 kg target mass range enclosed in an active liquid argon anti-Compton detector is being constructed. Feasibility studies towards scale-up to ton-scale experiment cdexshielding () are being pursued.
This work was supported by the National Natural Science Foundation of China (contract numbers: 10935005, 10945002, 11275107, 11175099) and National Basic Research program of China (973 Program) (contract number: 2010CB833006) and NSC 99-2112-M-001-017-MY3 and Academia Sinica Principle Investigator Award 2011- 2015 from Taiwan.
- (1) M. Drees and G. Gerbier, Review of Particle Physics Phys. Rev. D 86, 289 (2012), and references therein.
- (2) R. Bernabei et al., Eur. Phys. J. C 56, 333 (2008); R. Bernabei et al., Eur. Phys. J. C 67, 39 (2010).
- (3) C.E. Aalseth et al., Phys. Rev. D 88, 012002 (2013). C.E. Aalseth et al., arXiv: 1401.3295 (2014).
- (4) G. Angloher et al., Eur. Phys. J. C 72, 1971 (2012).
- (5) R. Agnese et al., Phys. Rev. Lett. 111, 251301 (2013).
- (6) E. Aprile et al., Phys. Rev. Lett. 109, 181301 (2012).
- (7) D.S. Akerib et al., Phys. Rev. Lett. 112, 091303 (2014).
- (8) H.B. Li et al., Phys. Rev. Lett. 110, 261301 (2013).
- (9) R. Agnese, et al. Phys. Rev. Lett. 112, 041302 (2014).
- (10) R. Agnese et al., Phys. Rev. Lett. 112, 241302 (2014).
- (11) Q. Yue et al., High Energy Phys. and Nucl. Phys. 28, 877 (2004); H.T. Wong et al., J. Phys. Conf. Ser. 39, 266 (2006).
- (12) P.N. Luke et al., IEEE Trans Nucl. Sci. 36 926 (1989); P.A. Barbeau, J.I. Collar and O. Tench, JCAP 09 009 (2007).
- (13) H.T. Wong, Mod. Phys. Lett.A 23 1431 (2008); S.T. Lin et al., Phys. Rev.D 79, 061101(2009).
- (14) W. Zhao et al., Phys. Rev. D 88, 052004 (2013); K.J. Kang et al., Chinese Phys. C 37, 126002 (2013).
- (15) G.K. Giovanetti et al., arXiv:1407.2238 (2014).
- (16) K.J. Kang et al., J. Phys. Conf. Ser. 203, 012028 (2010).
- (17) Y. C. Wu et al., Chinese Phys. C 37, 086001 (2013).
- (18) S.K. Liu et al., Phys. Rev. D 90, 032003 (2014).
- (19) H.B. Li et al., Astropart. Phys. 56, 1 (2014).
- (20) K.J. Kang et al., Front. Phys. 8, 412 (2013).
- (21) U. Tamm, W. Michaelis, and P. Coussieu, Nucl. Instrum. Meth. 48, 301 (1967); M.G. Strauss and R.N. Larsen, Nucl. Instrum. Meth. 56, 80 (1967); E. Sakai, IEEE Trans. Nucl. Sci. 18, 208 (1971).
- (22) E. Aguayo et al., Nucl. Instrum. Meth. A 701, 176 (2013).
- (23) J.F. Ziegler, Transport of Ions in Matter, http://www.srim.org (1998).
- (24) S.T. Lin et al., arXiv:0712.1645v4 (2007), and references therein.
- (25) J. Lindhard et al., Dan. Vid. Mat.-Fys, Medd. 33, 10 (1963).