Neutrinos produced in the nuclear reaction that triggers solar-energy generation have been detected. This milestone in the search for solar neutrinos required a deep underground detector of exceptional sensitivity. See Article p.383
A remarkable detector of solar neutrinos called Borexino has operated for the past seven years in Italy's Gran Sasso Laboratory, shielded by more than a kilometre of rock from the cosmic rays that bombard Earth's surface. A prolonged effort has reduced background signals from radioactive elements present in the detector that would otherwise obscure the neutrino signal. On page 383 of this issue, the Borexino Collaboration1 reports how this background suppression has enabled direct detection of the low-energy neutrinos produced in the nuclear reaction that initiates solar-energy generation.
Attempts to identify the source of solar energy have a long history2. In the nineteenth century, gravitational contraction was the only known mechanism that could plausibly account for the Sun's luminosity and extended lifetime. Because this limited the Sun's age to about 30 million years, physicist Willliam Thomson (Lord Kelvin) concluded that Charles Darwin's geological estimate of 300 million years for the age of the Earth was incorrect.
The beginning of the twentieth century marked the discoveries of the equivalence of mass and energy, embodied in Albert Einstein's famous equation E = mc2, and of radioactive decays in which a nucleus changes its charge through weak interactions. Following Francis Aston's measurements in 1920 of the mass difference between four protons and a helium-4 (4He) nucleus, Arthur Eddington proposed that the source of solar energy is the fusion of four protons to form 4He. Later, responding to criticism that the Sun is not sufficiently hot to sustain nuclear fusion, Eddington invited his critic to “go and find a hotter place”.
This dispute was resolved by George Gamow, who showed that quantum tunnelling would allow two solar protons to approach one another within the range required for nuclear fusion to occur. The detailed reactions leading to the synthesis of 4He were then deduced3: the proton–proton (pp) chain (Fig. 1) in the case of small, slowly evolving stars such as the Sun, and the carbon–nitrogen (CN) cycle in more-massive, rapidly evolving stars. Steady nuclear-energy release in the solar core keeps the temperatures high, ionizing hydrogen (H) and 4He and producing a plasma in which the electrons act as a gas. The Sun burns in a long-lasting hydrostatic equilibrium, with the effects of gravity counterbalanced by the electron-gas pressure. The rate of energy generation must compensate for the radiative losses from the Sun's surface. The Sun has maintained this balance for 4.6 billion years by consuming about half the hydrogen fuel in its core.
This description of the Sun also applies to about 90% of the stars in the Milky Way. How can our understanding of energy generation in such 'main sequence' stars be tested? Conventional observations are limited to the Sun's surface: they do not directly probe the deep interior, where the temperatures of 107 kelvins necessary for fusion are found. Yet just as the Sun's surface shines in photons, its core shines in neutrinos. Each synthesized 4He nucleus requires the conversion of two protons to two neutrons, with each neutron being accompanied by an electron-type neutrino. The neutrinos emerge directly from the core and reach Earth in eight minutes, carrying in their flux and energy distribution a detailed account of the Sun's fusion processes. The neutrino core-shine at Earth is intense (1011 neutrinos per square centimetre per second) but difficult to detect, because neutrinos rarely interact with matter as they pass through it. For this reason, the bulk of these neutrinos have evaded direct detection — until now.
All but 1% of solar 4He synthesis takes place through the pp chain, which releases two electron neutrinos and 26.73 million electronvolts of energy. The three cycles of this chain (ppI, ppII and ppIII) are each associated with a characteristic neutrino source. All three cycles begin with the fusion of two protons in the solar plasma to form deuterium (2H), through the 'pp' and 'pep' reactions (Fig. 1). The former accounts for 99.76% of deuterium synthesis, and thus determines the rate of solar-energy generation.
In early experiments to detect neutrinos — the historic chlorine experiment4, Kamiokande5 and GALLEX/SAGE6,7 — the number of neutrinos recorded was about one-half to one-third of that predicted by theory, indicating that a basic flaw existed in our understanding of either the Sun or the physics of neutrinos. This solar-neutrino problem motivated the building of a new generation of massive detectors, namely, the Sudbury Neutrino Observatory8, Super-Kamiokande9 and Borexino (Fig. 2). The Sudbury Neutrino Observatory and Super-Kamiokande, which detected the high-energy neutrinos produced in the β-decay of boron-8 (8B) in the ppIII cycle (Fig. 1), traced the problem to new particle physics: neutrinos have a mass and can change 'flavour' during their transit from the solar core to Earth. Two-thirds of solar 8B electron neutrinos oscillate into other types (muon and tau neutrinos) before reaching Earth.
The 8B neutrinos comprise about 0.01% of the total flux of neutrinos coming from the Sun. The Borexino detector was designed to detect the remaining, lower-energy solar neutrinos through their scattering off electrons in liquid scintillator: the recoiling electrons emit light that is recorded in the detector. Measurements of the low-energy neutrinos produced in the pep reaction and the ppII cycle have been announced previously10,11. Now, after an extended detector-purification campaign, the Borexino Collaboration has measured the pp neutrinos — the lowest-energy neutrino branch, accounting for 90% of the total flux12.
This result provides an important test of how matter affects neutrino oscillations — oscillations occurring within the Sun differ from those occurring in a vacuum. Such 'matter effects' can be exploited to determine whether a given neutrino is heavier or lighter than another neutrino. Previous experiments8,9 fixed the relative masses of two of the three neutrinos in this way. The theory used predicts12 that oscillation probabilities will be lower for pp neutrinos than for higher-energy 8B neutrinos: a larger fraction of the pp neutrinos will arrive at Earth as electron neutrinos. The fraction found by Borexino, 0.64 ± 0.12, is nearly twice that found for 8B neutrinos. This verification of theory is important because future planned 'long-baseline' neutrino-beam experiments will exploit matter effects to determine the ordering among all three neutrino masses.
The Borexino results also provide new tests of the Sun. Most solar-neutrino analyses assume that the total flux of neutrinos is consistent with the solar luminosity. But the connection between neutrino emission, which measures the rate of energy generation in the solar core today, and luminosity is valid only in a steady-state Sun: because photons take about 100,000 years to diffuse out of the core, this connection would not hold if the temperature of the solar core varies on times less than 100,000 years. Tests of this connection thus constrain solar variability and well as certain new-physics phenomena, such as solar emission of undetected 'sterile' neutrinos. Because the solar luminosity has been measured to a precision of 0.01%, checks on this relationship are limited by neutrino-flux uncertainties. The Borexino Collaboration notes that the 10% uncertainty of its pp neutrino-flux determination could be reduced to 1% in an improved experiment.
Now that all four principal neutrino sources from the pp chain have been directly measured, one task remains: about 1% of solar 4He synthesis takes place through the CN cycle. Neutrinos produced through this channel have not yet been detected. Because the CN cycle is catalysed by reactions on C and N, its rate is proportional to the solar core's metallicity (the fraction of elements other than H or 4He). Consequently, the core metallicity can be deduced from the CN neutrino flux. A measurement of these neutrinos could directly confirm the solar-abundance problem12: helioseismic data suggest that the solar interior is metal-rich relative to the solar surface, contradicting the standard-solar-model assumption of a homogeneous primordial Sun, formed from gas that had mixed convectively.
Measuring CN neutrinos is a final challenge for Borexino, requiring further progress in background suppression. One of the candidate explanations12 for the solar-abundance problem is the effect of planetary formation on the early Sun: this process swept a great deal of metal from gas that, if then deposited on the Sun's surface, could have diluted the star's outer convective zone. The demonstration of a connection between a host star's metallicity and the presence of its planets would have significant implications for exoplanet searches.