Scientists working with the PandaX‑4T detector report a new measurement of the flux of solar proton–proton (pp) neutrinos. Using Run 2 data taken from 2024 to 2026, with an exposure of 1.9 tonne·yr, they find a pp neutrino flux of (8.5 ± 3.5) × 10^10 cm^−2 s^−1. The result is consistent with the Standard Solar Model and gives a 2.2σ statistical excess over backgrounds, which the authors call the first positive indication of pp neutrino–electron scattering below an electron recoil energy of 165 keV.

The experiment looks for pp neutrinos by observing neutrino–electron elastic scattering in a liquid xenon time projection chamber. This detector records two signals from each interaction: a prompt flash of light (S1) and a delayed charge signal turned into light (S2). Measuring both signals lets the team reconstruct the event position and energy in three dimensions, which helps reject background events that come from the detector edges.

Before Run 2 the PandaX‑4T team upgraded many parts of the detector and its readout. They replaced the cathode, improved photomultiplier tube electronics, installed more reflective field panels, and upgraded the external water shield. The readout electronics were changed to custom 14‑bit digitizers running at 500 million samples per second and to a triggerless mode that records low‑threshold signals (about one‑third of a photoelectron). They also kept the xenon clean with continuous circulation, achieving an average electron lifetime of 1.08 ± 0.05 ms for reliable charge collection.

Controlling radioactive backgrounds at low energy was a major focus. The team reduced radon (222Rn) in the xenon to a minimum of 2.9 ± 0.1 µBq/kg by running their distillation system in a reverse mode. They also found elevated argon‑39 (39Ar) and ran two argon distillation campaigns. Run 2 was therefore split into a High‑Ar period (233.8 live days, 39Ar activity 0.74 µBq/kg) and a Low‑Ar period (204.2 live days, 0.10 µBq/kg). Time variations of noble‑gas impurities were constrained using the physics data themselves and by gas assays. The data processing pipeline was unified and improved so it can pick out very small S1s and better separate overlapping S2 signals.