/ AbstractBorexino is a liquid scintillator detector located at the Laboratori Nazionale del Gran Sasso, Italy with the main goal to measure solar neutrinos. The experiment recently provided the first direct experimental evidence of CNOcycle neutrinos in the Sun, rejecting the no-CNO signal hypothesis with a significance greater than 5σ at 99%C.L. The intrinsic 210Bi is an important background for this analysis due to its similar spectral shape to that of CNO neutrinos. 210Bi can be measured through its daughter 210Po which can be distinguished through an event-by-event basis via pulse shape discrimination. However, this required reducing the convective motions in the scintillator that brought additional 210Po from peripheral sources. This was made possible through the thermal insulation and stabilization campaign performed between 2015 and 2016. This article will explain the strategy and the different methods performed to extract the 210Bi upper limit in Phase-III (Jul 2016Feb 2020) of the experiment through the analysis of 210Po in the cleanest region of the detector called the Low Polonium Field. 1. CNO neutrino detection with Borexino Solar neutrinos are elementary particles that are produced inside the Sun, by the same nuclear fusion processes that generate the heat. Neutrinos interact rarely after their production and are therefore a direct probe of these nuclear fusion processes. According to the Standard Solar Model (SSM), which represents the best knowledge available about the Sun, the heat in the Sun’s core is generated by two main series of processes, fusing protons to Helium: the primary proton–proton (pp) chain, responsible for about 99% of the solar energy production and the sub-dominant Carbon – Nitrogen – Oxygen (CNO) cycle. Neutrinos are detected via their elastic scattering on electrons in the liquid scintillator. Borexino [1] has already published a complete spectroscopy of pp chain neutrinos [2] and has recently provided the first direct experimental evidence of solar neutrinos produced in the rare CNO nuclear fusion cycle [3]. The main challenges of this analysis are the very low interaction rate of CNO neutrinos and the similarity of its spectral shape to that of pep solar neutrinos and the intrinsic 210Bi background. The pep neutrino rate can be independently determined with 1.4% precision using the constraint on solar luminosity, global analysis with all solar neutrino experiments excluding the latest Borexino data, exploiting theoretically precisely known ratio of pep and pp neutrino fluxes, and using the most recent values of the oscillation parameters. The 210Bi background—the short-living decay product of 210Pb, can be determined via the counting of α-decays of its daughter 210Po. This assumes secular equilibrium of the chain down from 210Pb, which is a long-living isotope contaminating the liquid scintillator. Alpha particles can be identified in Borexino through pulse shape discrimination techniques. Until mid-2016, additional 210Po was brought from peripheral sources to the fiducial volume through the convective motions of the scintillator, triggered by seasonal temperature changes. However, between 2015 and 2016, the detector was thermally insulated and an active temperature control system was installed. This has minimized the residual convection in the innermost parts of the detector, making it possible to measure 210Bi via 210Po and has helped in obtaining a 210Bi upper limit from the cleanest region of the detector called the Low Polonium Field. A multivariate fit was then performed, i.e. the energy spectra in the window between 320 keV and 2,640 keV and the radial distribution of the Phase-III data (July 2016 February 2020) was simultaneously fitted, after constraining the rates of pep and 210Bi. This study excluded the no-CNO signal scenario with a significance greater than 5.0σ at 99.0% CL. The total contribution of the systematics was evaluated as +0.6 −0.5 cpd/100t using 13.8 million pseudo-datasets with the same exposure as Phase-III. The systematic uncertainties included the 210Bi spectral shape, the energy scale and resolution of the Monte Carlo model, non-linearity and non-uniformity of the detector’s response, as well as variation in the absolute value of the scintillator light yield. A simple counting analysis, complementary to the multivariate fit, rejected the null CNO hypothesis at 3.5σ. 2. The Low Polonium Field The Low Polonium Field of Borexino developed just above the equator of the detector around mid-2016. The 210Po rate in this field is the sum of two components: a scintillator component that comes from the 210Pb in the scintillator and is assumed to be in secular equilibrium with 210Bi, and a vessel component which is due to the migration of 210Po from the 210Pb on the vessel. The migration process is driven by convective currents. However, the parent 210Pb and 210Bi isotopes of this component stay on the IV. The qualitative shape and approximate position of the LPoF have been