Draft:Coherent Neutrino-Nucleus Interaction Experiment

Coherent Neutrino-Nucleus Interaction Experiment (CONNIE) is an experiment developed with the aim of detecting coherent elastic neutrino-nucleus scattering (CE𝜈NS) using silicon CCDs. The experiment is located at the Angra-2 nuclear power plant, in Angra dos Reis, Brazil. The CONNIE experiment aims to measure CE𝜈NS at very low nuclear recoil energies (approximately 0.2 keV ), which could be sensitive to new physics beyond the Standard Model. CONNIE is an international collaboration and is the leading experiment of its field of research in Latin America.

Physics goals

The primary goal of the CONNIE experiment is to detect the coherent scattering of reactor-produced antineutrinos interacting with silicon nuclei. This effort aims to probe physics beyond the Standard Model (BSM) by exploring potential new interactions and particles affecting neutrino behaviour, which could increase the interaction probability. The experiment specifically seeks to:

• Detect coherent scattering and measure its rates to improve understanding of neutrino properties and interactions.
• Search for new physics, including light vector and scalar mediators that could influence neutrino interactions.
• Investigate other exotic scenarios such as relativistic millicharged particles, which may offer insights into new physics phenomena.

CONNIE ultimately aims to enhance the sensitivity and precision of neutrino detection, contributing to advancements in fundamental particle physics.

Coherent elastic neutrino-nucleus scattering (CE𝜈NS)

Coherent elastic neutrino-nucleus scattering (CE𝜈NS) is a type of interaction in which neutrinos interact with the entire nucleus of an atom at low energies. Neutrinos are electrically neutral leptons that only interact through the weak nuclear force, making their detection challenging due to the extremely low probability of interaction with matter.

There are two primary ways neutrinos can interact with atoms: either with the electrons in the atomic shell or with the protons and neutrons in the nucleus. CE𝜈NS occurs when neutrinos interact with the nucleus as a whole rather than with individual nucleons (protons and neutrons). This interaction is coherent, meaning the neutrino “sees” the nucleus as a whole due to its low energy and long wavelength, which is comparable to the size of the nucleus. CE𝜈NS is a neutral-current weak interaction that proceeds via the exchange of a Z boson, and can occur for all neutrino flavours.

Illustration of the coherent elastic neutrino scattering off an atomic nucleus via the exchange of a neutral Z-boson

CE𝜈NS was first predicted in 1974 and occurs at energies up to few tens of MeV. In this energy range, the neutrino interacts coherently with the nucleus, resulting in an enhanced cross-section. Despite the larger cross-section compared to other neutrino interactions, detecting CE𝜈NS is difficult due to the small amount of energy transferred to the nucleus. This energy, known as the nuclear recoil energy, is typically below 1 keV, making it hard to observe with standard detection technologies.

Experiments investigating CE𝜈NS require intense neutrino sources and detectors with very low energy thresholds to measure these small recoils, as well as very low backgrounds from other processes. The first experiment to observe this phenomenon experimentally was COHERENT, by using an intense pulsed beam of neutrinos from a spallation neutron source. The experiment demonstrated that CE𝜈NS occurs and can be detected with sufficiently sensitive instruments.

CONNIE detector description

The CONNIE detector at the Angra-2 reactor is a solid-state system designed to detect coherent elastic neutrino-nucleus scattering of reactor neutrinos. Its core component is an array of thick, high-resistivity silicon Charge-Coupled Devices (CCDs), originally designed as memory devices but now valued for their high-resolution imaging capabilities. CCDs offer high detection efficiency, low noise, good spatial resolution, and low dark current, making them ideal for scientific applications such as ground and space-based astronomy and X-ray imaging. Thick CCDs with more active mass have also been used in several particle detection experiments such as the DAMIC dark matter search .

The CONNIE CCD sensors are enclosed within a copper box that is maintained in a vacuum vessel and operate at low temperatures of approximately 100 K. The detector is surrounded by passive shielding, which includes 30 cm thick outer and inner layers of high-density polyethylene to absorb neutrons, placed around a 15 cm layer of lead to absorb photons.

The detector is positioned in the ground-level neutrino laboratory Nu Lab, shared with the Neutrinos Angra experiment, located outside the reactor dome at a distance of 30 metres from the 3.95 GW_th Angra-2 nuclear reactor. Antineutrinos are produced in the reactor primarily when uranium-235 fission products undergo beta decay. Although emitted in large quantities, their weak interaction with matter makes detection difficult. CONNIE aims to detect these antineutrinos by using coherent elastic neutrino-nucleus scattering. The estimated antineutrino flux at that distance is 7.8x10⁺¹²  /cm²/s .

Experiment phases and measurements

The CONNIE experiment began deploying a CCD-based detector at the Angra-2 reactor site in 2013, utilising the laboratory of the Neutrinos-Angra project. The initial phase (2014–2016) used engineering-grade CCDs, which served as a proof of concept by measuring background levels at the reactor site and demonstrating the remote operation of the CCD array and the feasibility of CCD technology for neutrino detection at a nuclear reactor .

In 2016, the detector was updated with 12 scientific CCDs, each with 4120 × 4120 pixels of 15 × 15 µm², and a thickness of 675 µm, giving a total mass of 6.0 g per CCD. The improved setup included fully depleted thick CCDs, which significantly enhanced low-energy sensitivity compared to previous versions. This enabled CONNIE to complete its first search for coherent elastic neutrino-nucleus scattering (CEνNS) using reactor antineutrinos, achieving a detection threshold of 75 eV electron equivalent. In this phase of the CONNIE experiment, data collected from 2016 onward using the scientific-CCD detector were analysed, achieving an excellent sensitivity down to recoil energies of 1 keV. This new detection capability allowed CONNIE to set an upper limit at 95% confidence level on the CEνNS event rate, placing constraints approximately 40 times above the expected rate from the Standard Model .

During 2017 and 2018 an operating efficiency of more than 95% was achieved, thanks to a monitoring, alarms and interlock system developed to record and report the status of all the critical values of the experiment . The low-energy data from this run were also used to set the first constraints on physics beyond the Standard Model for simplified models with new light mediators. The data provided the best limits at the time for a light vector mediator with a mass below 10 MeV and a light scalar mediator below 30 MeV. These results constituted the first use of the CONNIE data as a probe for BSM physics .

In 2019-2020, the CONNIE experiment implemented a hardware binning readout mode that improved the signal-to-noise ratio by summing pixel charges before readout. This enhancement allowed the detector to reach a lower energy threshold of 50 eV. With an exposure of 2.2 kg-days and a fiducial mass of 36.2 g across eight stable CCDs, no CEνNS excess was observed, leading to limits at 95% confidence level on the neutrino interaction rates. The lowest observed limit in the 50–180 eV ionisation energy range was approximately 70 times the Standard Model prediction .

In July 2021, the experiment was upgraded with two Skipper-CCDs, each containing 682 × 1022 pixels and a thickness of 675 µm, with a mass of 0.247 g per sensor. The Skipper-CCDs allow non-destructive repeated measurements of pixel charges, reducing electronic noise and enabling single-electron sensitivity . This technology improvement lowered the detection threshold to 15 eV, providing enhanced sensitivity to low-energy events and making CONNIE the first experiment to employ Skipper-CCDs for reactor neutrino detection.

The 2021-2022 data collected by the Skipper-CCDs in the CONNIE experiment enabled a refined analysis in multiple areas. Despite not detecting a CEνNS signal, the experiment achieved an upper limit at 95% confidence level on the neutrino interaction rate , comparable to that from the previous standard CCDs even with a reduced sensor mass (a thousand times smaller exposure). The new low threshold allowed CONNIE to improve its constraints on BSM physics for light vector mediator models. Moreover, the CONNIE collaboration conducted its first dark matter search through diurnal modulation, establishing the most stringent surface-level limits on DM-electron scattering cross-sections, as of 2024, thereby proving the excellent capabilities of Skipper-CCDs in the search for rare neutrino interactions and dark matter signatures .

Additionally, recent analyses have focused on searching for millicharged particles, which can be candidates for dark matter, by combining Skipper-CCD data from the CONNIE and the Atucha-II experiments. These analyses yielded world-leading limits on millicharged particle coupling, constraining its parameter space at charge fractions near 1x10-6 for several orders of magnitude of low masses, thus narrowing the viable range for millicharged particles as dark matter candidates .

In 2024, the two Skipper-CCDs were replaced with a Multi-Chip Module (MCM), consisting of 16 Skipper-CCDs mounted on a single board, and designed by the Oscura collaboration . This upgrade is intended to increase the detector mass and enhance sensitivity in the search for rare particle interactions.

A view of the CONNIE detector and its shielding

16 Mpix standard CCD which took data in 2016-2020

0.9 Mpix Skipper CCD which took data in 2021-2023

Multi-Chip-Module installed in 2024

Collaborating institutions

The CONNIE collaboration includes researchers and students from institutionalised entities such as universities and national labs:

• Federal University of Rio de Janeiro (UFRJ);
• Brazilian Centre for Research in Physics (CBPF);
• CEFET Angra dos Reis;
• Universidade Federal do ABC;
• Instituto Tecnológico de Aeronáutica;
• Centro Atómico Bariloche;
• Universidad de Buenos Aires;
• Universidad del Sur / CONICET;
• Universidad Nacional Autónoma de México;
• Universidad Nacional de Asunción;
• Fermi National Accelerator Laboratory (Fermilab);
• University of Zurich.