Neutrino physics

A deliberate mix-up in flavour

Neutrinos come in three 'flavours', as do antineutrinos, and they all change flavour as they travel. New measurements of the mixing of different neutrinos may help to explain why our Universe is made of matter and not antimatter.

For decades, physicists have been intrigued as to why the visible Universe is made of matter, rather than antimatter. Intuitively, studying the smallest of the known fundamental particles, the neutrino, may reveal the answer to this question. As neutrinos travel through space and time, they can oscillate between three different types (flavours). If the way in which antineutrinos change flavour differs from the way neutrinos change flavour, these relatively small differences may have led to a matter-based Universe.

Experiments have been set up to study the amplitude and frequency of these changes to determine the measurable parameters associated with flavour oscillations, such as mass differences (neutrinos have three possible masses, denoted 1, 2 and 3) and mixing angles (how much mixing occurs between neutrinos of mass 1, 2 and 3). Writing in Physical Review D, the Daya Bay Collaboration1 reports that it has used a new detection method to make the most precise measurement yet of θ13, the mixing angle between neutrino mass 1 and neutrino mass 3. This will help in our quest to understand the origin of the dominance of matter in the Universe.

The Daya Bay experiment is located on the southern coast of China, approximately 55 kilometres northeast of Hong Kong. The site consists of eight identical antineutrino detectors, located between 360 m and 1.9 km from the Daya Bay and Ling Ao nuclear power plants. Approximately 3,500 billion billion electron antineutrinos (one flavour of antineutrino) are emitted from the nuclear reactors every second, and a small number of these interact with the 20 tonnes of gadolinium-doped liquid scintillator in each detector; the scintillator emits light when it absorbs an energetic particle. The electron antineutrinos interact with protons in the nuclei of the scintillator molecules to produce positrons (antielectrons) and neutrons in a process known as inverse β-decay. Within a few nanoseconds, the positrons produce light by scintillation, and this is detected by an array of photomultiplier tubes, which amplify the incident light.

When a neutron is captured by a gadolinium nucleus embedded in the scintillator, it forms an excited state, which subsequently decays to a more-stable state by emitting a series of γ-rays with a total energy of 8 megaelectronvolts. These γ-rays scatter and interact with the scintillator to produce another light pulse, typically tens to hundreds of microseconds after the first pulse that was generated by the positrons. Searching for two coincident light pulses that have specific properties can confirm the detection of electron antineutrinos.

Other reactor-based experiments2,3 use the same gadolinium capture process. But the Daya Bay Collaboration now reports an additional, independent method to identify neutrons produced through inverse β-decay. In the detectors, neutrons are preferentially captured on gadolinium, but capture on a hydrogen atom of the organic liquid scintillator is also possible. Neutron capture on hydrogen produces deuterium and releases a low-energy (2.2-MeV) γ-ray, but this is hard to detect because many similar γ-rays with energies below 4 MeV are emitted from decays of naturally occurring radioisotopes, such as potassium-40 (1.4 MeV) and thallium-208 (2.6 MeV). Such γ-rays are problematic because they can mimic those released by neutron capture on hydrogen, obscuring the antineutrino signals. Radioactivity in the detector must therefore be minimized and controlled by careful purification of the scintillator, selection of low-radioactivity materials and stringent cleanliness. Despite such precautions, radioactivity is never completely eliminated, and this ultimately limits the precision of the experiment.

Detecting inverse β-decay from antineutrino interactions requires the identification of two light signals occurring within 1 to 1,400 microseconds of each other. This timing requirement reduces the impact of coincidental light flashes from background radiation that are not time correlated. The Daya Bay Collaboration was able to estimate the rate of accidental coincidences by using data recorded in its detectors. This quantity, subtracted from the total number of observed coincidences, yields the number of time-correlated coincidences.

Further corrections were required to account for any time-correlated backgrounds, the most significant of which arises from the activation by cosmic rays of carbon-12 in the scintillator to produce lithium-9 and helium-8. Both of these isotopes decay by β-emission to neutron-unstable excited states, producing a signal of an electron followed by neutron capture that is indistinguishable from the signal from inverse β-decay.

When all the backgrounds were appropriately accounted for, the collaboration used the number of genuine time-correlated coincidences to measure the flux of electron antineutrinos. The measured number of electron antineutrinos was found to be less than the collaboration had predicted from reactor flux models. The missing electron antineutrinos demonstrate that they oscillate into other flavours that cannot be detected in this experiment.

This analysis from the Daya Bay experiment provides an independent measurement of the number of electron-antineutrino interactions in the detectors and represents an important cross-check of previous measurements of the mixing angle θ13. Moreover, the combination of the collaboration's latest hydrogen-capture results with those from its gadolinium analysis provides the most precise measurement of θ13 so far, where sin2 (2θ13) = 0.082 ± 0.004, an 8% improvement in precision.

This measurement compares favourably with those from other reactor studies2,3 and long-baseline experiments4,5,6,7, which provide long distances over which to observe changes in neutrino flavour. Closer examination of the results (Fig. 1) suggests slight discrepancies between current measurements of θ13. Results from Double Chooz3 and T2K (ref. 4) tend to favour higher values of θ13, although not significantly so.

Figure 1: Neutrino measurements.

The mixing angle θ13 is a measure of how much neutrinos that have masses 1 and 3 mix. The graph compares measurements of θ13 made by the Daya Bay1, Double Chooz3 and T2K (ref. 4) experiments; values are plotted as sin2 (2θ13). Two values are shown for the T2K experiment because the best-fit value of θ13 depends on the ordering of the neutrino masses: m3 > m2 > m1 is known as the normal hierarchy (NH), whereas m2 > m1 > m3 is the inverted hierarchy (IH). The Daya Bay measurement is the most precise so far. A non-zero θ13 would open the way to experiments that might explain why the Universe is dominated by matter rather than antimatter.

Measurements of θ13 are particularly valuable in our quest to understand the matter-dominated Universe through measurement of another neutrino oscillation parameter, δCP. Non-zero δCP would indicate that neutrinos violate charge–parity symmetry (under this symmetry, the laws of physics should be the same when a particle is replaced by its antiparticle and inverted through a mirror); this may explain why the Universe is dominated by matter. Combining results from reactor and long-baseline experiments provides a hint that δCP is indeed non-zero, but current experiments are not precise enough to reach the statistically significant '5 sigma' discovery threshold. The next generation of experiments8,9 will seek to make the first measurement of δCP and thus help us to understand the origin of our matter-dominated Universe.Footnote 1


  1. 1.

    See all news & views


  1. 1

    An, F.P. et al. (Daya Bay Collaboration) Phys. Rev. D 93, 072011 (2016).

    ADS  Article  Google Scholar 

  2. 2

    Ahn, J. K. et al. Phys. Rev. Lett. 108, 191802 (2012).

    CAS  ADS  Article  Google Scholar 

  3. 3

    Ishitsuka, M. (2016).

  4. 4

    Abe, K. et al. Phys. Rev. Lett. 112, 061802 (2014).

    CAS  ADS  Article  Google Scholar 

  5. 5

    Abe, K. et al. Phys. Rev. D 91, 072010 (2015).

    ADS  Article  Google Scholar 

  6. 6

    Adamson, P. et al. Phys. Rev. Lett. 112, 191801 (2014).

    CAS  ADS  Article  Google Scholar 

  7. 7

    Adamson, P. et al. Preprint at (2016).

  8. 8

    Abe, K. et al. Prog. Theor. Exp. Phys. 5, 053C02 (2015).

    Google Scholar 

  9. 9

    Acciarri, R. et al. Preprint at (2016).

Download references

Author information



Corresponding author

Correspondence to Helen O'Keeffe.

Related links

Related links

Related links in Nature Research

Neutrino physics: What makes the Sun shine

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

O'Keeffe, H. A deliberate mix-up in flavour. Nature 533, 327–328 (2016).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.