Constraints on low-mass, relic dark matter candidates from a surface-operated SuperCDMS single-charge sensitive detector

This article presents an analysis and the resulting limits on light dark matter inelastically scattering off of electrons, and on dark photon and axionlike particle absorption, using a second-generation SuperCDMS high-voltage eV-resolution detector. The 0.93 g Si detector achieved a 3 eV phonon energy resolution; for a detector bias of 100 V, this corresponds to a charge resolution of 3% of a single electron-hole pair. The energy spectrum is reported from a blind analysis with 1.2 g-days of exposure acquired in an above-ground laboratory. With charge carrier trapping and impact ionization effects incorporated into the dark matter signal models, the dark matter-electron cross section ${\overline{\sigma }}_e$ is constrained for dark matter masses from 0.5 to ${10}^4\mathrm{MeV}/c^2$; in the mass range from 1.2 to $50\mathrm{eV}/c^2$ the dark photon kinetic mixing parameter $ϵ$ and the axioelectric coupling constant $g_{ae}$ are constrained. The minimum 90% confidence-level upper limits within the above-mentioned mass ranges are ${\overline{\sigma }}_e=8.7\times {10}^{-34}{\mathrm{cm}}^2$, $ϵ=3.3\times {10}^{-14}$, and $g_{ae}=1.0\times {10}^{-9}$.

Figure 1

The top panel shows the DM-search spectrum (red) in units of event rate per 3 eV bin (left y axis) and the laser-calibration spectrum (blue) in units of events per 3 eV bin (right y axis). Both spectra show the data measured with a detector bias of 100 V after applying the live-time and data-quality cuts. The peak seen at $\sim 50\mathrm{eV}$ in the DM-search data is due to nonquantized events restricted to the outer QET channel [14]. Light gray-shaded regions on the left- and right-hand sides mark the energy ranges outside the region of interest; vertical lines correspond to the phonon energy $E_n$ of the $n$-photon absorption peak [Eq. (1)]. The black curve is an example of a signal produced by electron-recoiling dark matter particles with a mass of $1\mathrm{GeV}/c^2$ and form factor $F_{\mathrm{DM}}\propto 1/q^2$. This model assumes a Fano factor of $F=0.155$, an impact ionization (II) probability of 2%, and a charge trapping (CT) probability that varies from 0% to 15% shown by the hatched region. The curve is scaled to the dark matter-electron cross section ${\overline{\sigma }}_e$ that sets the limit at the second $e^{-}{h}^{+}$-pair peak. The bottom panel shows the binned efficiency data $\epsilon (E_i)$ (gray solid line), where the corresponding shaded region indicates the $1\sigma$ statistical uncertainty in each bin. The red dashed curve is the efficiency curve, and the corresponding shaded region is the conservative efficiency uncertainty envelope, which accounts for the statistical and systematic uncertainties.

Figure 2

90% C.L. limits on the effective dark matter-electron scattering cross section with form factor $F_{\mathrm{DM}}=1$ (top) and $F_{\mathrm{DM}}\propto 1/q^2$ (bottom) and with Fano factor of 0.155 (solid blue curve). The light blue band represents our estimate of the systematic uncertainty, which is dominated by varying the Fano factor assumption in the ionization model from $F={10}^{-4}$ to 0.3. Other direct detection constraints shown include SuperCDMS HVeV R1 [12] (red), DAMIC [58] (green), SENSEI [23] (orange), EDELWEISS [59] (gray), XENON10 [60, 61] (teal), and XENON1T [62] (pink).

Figure 3

90% C.L. limits on the dark photon ($V$) kinetic mixing parameter $ϵ$ (top) and axioelectric coupling constant $g_{ae}$ (bottom) with Fano factor of 0.155 (solid blue curve). The light blue band represents our estimate of the systematic uncertainty, which for masses $\gtrsim 4\times {10}^{-3}\mathrm{keV}/c^2$ is dominated by varying the Fano factor assumption in the ionization model from $F={10}^{-4}$ to 0.3; for masses $\lesssim 4\times {10}^{-3}\mathrm{keV}/c^2$, the uncertainty is dominated by the discrepancy in the photoelectric absorption cross section. Other direct detection constraints shown for $V$ and ALPs include SuperCDMS Soudan [34] (maroon), XENON10 (teal), and XENON100 (purple) [63]; additional constraints on $V$ include SuperCDMS HVeV R1 [12] (red), DAMIC [58] (green), SENSEI [23] (orange), EDELWEISS [59] (salmon), and anomalous energy loss mechanisms in the Sun [24]. For the axioelectric coupling, the entire region shown is disfavored by the observed cooling of red giant [25, 26] and white dwarf stars [26, 27].