Borexino: Every Branch of the Sun

The Sudbury Neutrino Observatory closed the solar neutrino problem in 2001 by showing that electron neutrinos from the Sun transform into other flavors on their way to Earth. That resolution was a landmark in physics, but it was not the end of solar neutrino science — it was the beginning of a precision era. What the SNO result established was that oscillation happens. What came next was the quantitative programme of measuring exactly which oscillation parameters were at work, at exactly which energies, and with exactly which deviations from the Standard Solar Model.

That precision programme was, for fifteen years, almost entirely a Borexino story. Between 2007 and 2021, a single 300-ton liquid scintillator detector at the Gran Sasso National Laboratory in central Italy recorded the most precisely characterised solar neutrino signal ever measured. It resolved every component of the pp chain and the CNO cycle, individually and with per-cent-level uncertainty. It observed the MSW transition from vacuum-dominated to matter-dominated oscillation across the 1–5 MeV energy range. And in 2020 — three years before shutdown — it made the first direct detection of neutrinos from the CNO fusion cycle, the last unobserved branch of solar nuclear burning.

This is the story of how the experiment was designed, what made it uniquely capable, what it measured, and why its legacy persists even after decommissioning.

The radiopurity problem

Solar neutrinos at the Earth’s surface arrive at a flux of about $6.5\times 10^{10}$ per square centimetre per second, dominated by low-energy (below 0.42 MeV) pp-chain neutrinos. The interaction cross-section at those energies is approximately $10^{-46}$ cm² per target electron — in practical terms, roughly one event per kiloton per day for the pp component. The detector technology must therefore be enormously sensitive, and — crucially — it must be radioactively pure to a degree that seems, at first glance, impossible.

Natural background radioactivity in any ordinary material includes traces of ${}^{238}$U and ${}^{232}$Th at concentrations of about $10^{-9}$ grams per gram — parts per billion. Those trace contaminants decay, and their decay products produce signals at the same energies as low-energy solar neutrinos. A scintillator detector built from ordinary materials would see background dwarfing the neutrino signal by five or six orders of magnitude.

The Borexino approach was to build a scintillator of extraordinary radiopurity — approximately $10^{-18}$ grams of U and Th per gram of scintillator, nine orders of magnitude below natural background — by multiply distilling the solvent (pseudocumene), sparging with high-purity nitrogen to remove dissolved gases, and exposing to water for extractive cleanup. The process was developed over years at the Counting Test Facility, a small precursor detector used as a proof-of-concept.

The final detector used a layered geometry to isolate the fiducial volume from the external world:

• Inner vessel (130-micron nylon sphere): 300 tons of the ultrapure scintillator as the target
• Outer vessel: a second nylon sphere at slightly larger radius, filled with pure buffer fluid that absorbed external gamma rays
• Stainless steel sphere (13.7 m diameter): surrounding the buffer, lined with 2212 photomultipliers
• Water tank (18.0 m diameter): an outer bath of pure water serving as a Cherenkov muon veto

The entire assembly was installed 1400 metres underground in Hall C of the Gran Sasso laboratory, where rock overburden suppressed cosmic-ray muons by a factor of $10^6$ relative to the surface. First data taking began in May 2007.

Observation of ⁷Be — the first milestone

Borexino’s primary initial target was the monoenergetic ${}^7$Be solar neutrino at 0.862 MeV. Before Borexino, the ${}^7$Be component had been measured only indirectly through Gallium-based radiochemical experiments (SAGE, GALLEX) that integrated the spectrum without distinguishing individual branches. A direct measurement of the ⁷Be flux would test the predicted fraction of proton-proton chain branching and provide a clean MSW oscillation probe at sub-MeV energies.

The 2008 Borexino result reported: $R_{\text{Be}}=(49\pm 3\pm 4)\text{ events per day per 100 tons}$ matching the MSW-LMA prediction and the Standard Solar Model within uncertainties. This was the first direct detection of a specific solar fusion branch through neutrinos.

The full spectrum, branch by branch

Over the following decade, Borexino progressively measured every component of the pp chain:

Branch	Year of direct detection	Flux (cm⁻²s⁻¹)
⁷Be	2007–2008	4.99 × 10⁹
⁸B	2008	2.4 × 10⁶
pep	2012	1.27 × 10⁸
pp	2014	6.6 × 10¹⁰
CNO	2020	6.6 × 10⁸

Each measurement required different analysis techniques: the ⁷Be line required spectral fitting, pep and CNO required three-fold coincidence tagging to suppress ${}^{11}$C backgrounds, and pp required the lowest-threshold analysis ever done.

The pp result was particularly notable. The pp fusion reaction produces 99% of the Sun’s luminosity and the pp neutrinos carry half the energy release. A direct measurement of their flux was, in a sense, the direct measurement of the Sun’s fusion rate. Borexino found: ${\Phi }_{pp}=(6.6\pm 0.7)\times 10^{10}{\text{ cm}}^{-2}{\text{s}}^{-1}$ which matched the Standard Solar Model prediction. This agreement is now one of the strongest quantitative tests of solar physics available.

The MSW survival probability, mapped

Beyond individual fluxes, Borexino’s suite of measurements at different energies traces out the electron-neutrino survival probability as a function of energy. At low energies (pp, ⁷Be), the survival probability is close to the vacuum-oscillation value of $\sim 0.55$. At high energies (⁸B), it drops to the MSW-matter-dominated value of $\sim 0.3$. The transition occurs around 2 MeV and has a specific shape predicted by the Mikheyev-Smirnov-Wolfenstein mechanism.

Before Borexino, the transition had only been inferred from the energy-integrated rates at different experiments (Homestake, Kamiokande, Super-K). Borexino was the first to resolve the transition within a single experiment, at its predicted shape. The result agreed with MSW-LMA at better than 1σ significance and ruled out several alternative oscillation scenarios — including non-standard interactions that would predict different spectral shapes.

The CNO detection

The Sun’s nuclear burning is overwhelmingly the pp chain, with the CNO cycle contributing only ~1% of total luminosity. Measuring the CNO component directly was the holy grail of solar neutrino physics because it carries information about the solar core metallicity — the abundance of elements heavier than helium — that has been a 15-year-long open question.

Helioseismology (the study of solar pressure waves) suggests a low-metallicity Sun, in tension with 3D-atmosphere-model spectroscopy of the photosphere, which suggests a high-metallicity Sun. A direct measurement of the CNO flux, which scales linearly with core metallicity, would settle the question.

Borexino achieved the CNO detection by developing an analysis technique called “thermally stabilised low-background phase” from 2016 onward. The detector was carefully thermally insulated and stabilised to minimise convection that was distributing the ${}^{210}$Bi contaminant throughout the fiducial volume. After three years of running, the CNO signal was extracted with a statistical significance of 5σ, published in Nature in 2022: ${\Phi }_{\text{CNO}}=6.6_{-0.9}^{+2.0}\times 10^8{\text{ cm}}^{-2}{\text{s}}^{-1}$ The result is compatible with the high-metallicity Standard Solar Model at 2σ and with the low-metallicity one at 1σ — not definitive, but the first direct constraint from neutrinos.

The decommissioning and legacy

Borexino was decommissioned in 2021. The scintillator sphere was flushed with nitrogen and sealed; the infrastructure remains available at Gran Sasso for potential future experiments. Plans for next-generation solar neutrino precision include:

• SNO+ in Canada, which uses tellurium-loaded scintillator primarily for a neutrinoless double-beta-decay search but retains solar neutrino sensitivity
• Jinping Neutrino Experiment (JUNO precursor, China) — aims for ~10-ton scintillator at ~2400-metre depth for precision solar + geoneutrino measurements
• DUNE with improvements — potentially sensitive to low-energy solar neutrinos after significant detector upgrades

None of these currently match Borexino’s radiopurity. Whether any ever will is an open question; ultra-low-background scintillator development is an active but slow technological frontier.

Why the experiment mattered

Borexino’s cumulative achievement is usually summarised as “measurement of solar neutrinos”. That understates it. What Borexino actually did was convert solar neutrino physics from an inference-based discipline — where fluxes were deduced from energy-integrated rates of a few experiments — to a direct-observation discipline, where every fusion branch is individually resolved and its flux precisely known.

That conversion matters beyond solar physics. Every future experiment that uses solar neutrinos as a calibration source, a background constraint, or a physics probe depends on the Borexino flux values. The neutrino floor in dark-matter detection, the MSW-parameter determination, the CNO metallicity probe — all now rest on Borexino’s precision measurements.

The experiment is also, in retrospect, a lesson in what sustained radiopurity engineering can achieve. The scintillator purification techniques developed for Borexino have propagated into numerous downstream experiments, raising the radiopurity state of the art by orders of magnitude. The intellectual property of Borexino was, arguably, as valuable as its physics results.

When historians write the story of twenty-first-century neutrino physics, the solar precision chapter will be dominated by this one 300-ton sphere of extraordinarily clean oil in a mountain in central Italy.