Charge coupled devices for detection of coherent neutrino-nucleus scattering

In this article the feasibility of using charge coupled devices (CCD) to detect low-energy neutrinos through their coherent scattering with nuclei is analyzed. The detection of neutrinos through this standard model process has been elusive because of the small energy deposited in such interaction. Typical particle detectors have thresholds of a few keV, and most of the energy deposition expected from coherent scattering is well below this level. The CCD detectors discussed in this paper can operate at a threshold of approximately 30 eV, making them ideal for observing this signal. On a CCD array of 500 g located next to a power nuclear reactor the number of coherent scattering events expected is about 3000 events/year. Our results shows that a detection with a confidence level of 99% can be reached within 16 days of continuous operation; with the current 52 g detector prototype this time lapse extends to five months.

Figure 1

Cross section of a $250\mu \mathrm{m}$ thick CCD developed at Lawrence Berkeley National Laboratory [17]. (a) Layout of the three gates that form one pixel. (b) Electrostatic potential (V) generated through the three gated phases shown as function of depth (y axis) and one of the lateral directions (x axis).

Figure 2

Compendium of images from recent measurement of background at sea level in a CCD.

Figure 3

RMS pixel error (${\sigma }_{\mathrm{RMS}}$) caused by the output amplifier, as a function of pixel read-out time.

Figure 4

Histogram of single pixels of images taken with a CCD running at 140 K and pixel-time of $30\mu \mathrm{s}$.

Figure 5

Total neutrino-nucleus coherent cross section ${\sigma }_{\mathrm{T}}$ for silicon from Eq. (2) (blue curve, left), and weighted by the reactor antineutrino spectrum (red curve, right). The dashed curves correspond to a threshold energy of 28 eV, approximately $5{\sigma }_{\mathrm{RMS}}$.

Figure 6

Antineutrino spectrum for each process [15].

Figure 7

Total reactor antineutrino spectrum per fission in the reactor per MeV. The red solid line reflects the fissile isotopes, and the blue dashed line the sum of the fissile isotopes and the neutron capture by ${}^{238}\mathrm{U}$.

Figure 8

Energy spectra for expected events in silicon detectors: (– – blue) nuclear-recoil energy spectrum, (— red) spectrum for detectable events using the quenching factor from Lindhard’s theory and (- - - green) using the efficiency curve from detection algorithm in Sec. 5c; recoil spectrum using neutrino calculations by Mueller et al. ($\square$) [23] and Huber ($\times$) [24].

Figure 9

Total number of events as a function of the threshold energy for different constant quenching factors (blue curves), and Lindhard factor (orange curve). The red dotted curve shows the total number of events as a function of the maximum detectable recoil energy using $Q=1$.

Figure 10

Silicon quenching factor. Compendium of measurements of [10], and theoretical prediction by Lindhard [27, 28].

Figure 11

Measurement and simulation of the reconstructed lateral deviation of events from X-rays from the $K_{\alpha }$ peak of a ${}^{55}\mathrm{Fe}$ source when the CCD is exposed from the back (black curve) and from the front (red curve).

Figure 12

Two simulated neutrino events generated at different depth of the detector and at different relative positions in the pixel. (a) $\overline{{\nu }_e}$-event interacting close to the gates of the detector, with ${\sigma }_{\text{Diff}}=0.2$ pixels, and (b) $\overline{{\nu }_e}$-event interacting close to the back (large $y$) of the detector, where ${\sigma }_{\text{Diff}}=0.5$ pixels. The $x_i$ and $z_i$ values are the coordinates of the point of origin of the events in the array.

Figure 13

Detection efficiency of a simple threshold algorithm of 28 eV when the CCD is configured for a binning of $10\times 10$ pixels.