First JUNO data put bi-large neutrino mixing theories to the test

Few experiments in particle physics have been awaited as eagerly as the Jiangmen Underground Neutrino Observatory, and the first results are already reshaping how theorists think about one of the deepest puzzles in nature: why neutrinos mix with one another in such a strange, lopsided way. Buried under a mountain in southern China, JUNO has now delivered its first precision measurements of the solar neutrino mixing angle and the associated mass-squared difference, and a new theoretical analysis published in The European Physical Journal C shows that these numbers are sharp enough to put decades-old ideas about the origin of lepton mixing under real, quantitative pressure.

The measurements themselves are striking. Using just 59.1 days of reactor antineutrino data, the JUNO collaboration reports a best-fit value of the solar mixing angle of sin-squared theta-12 equal to 0.3092, with an uncertainty of only 0.0087, and a solar mass-squared difference of (7.50 plus or minus 0.12) times ten to the minus five electron-volts squared, assuming normal mass ordering. When the new SNO+ reactor measurements are folded in with the Particle Data Group 2025 global dataset, the combined values shift only slightly, to 0.310 plus or minus 0.012 and (7.63 plus or minus 0.17) times ten to the minus five electron-volts squared. Remarkably, JUNO’s standalone determination already surpasses the precision of the existing worldwide oscillation global fits, which pool together decades of data from solar, atmospheric, reactor and accelerator experiments.

That leap in precision matters because the mixing angles are not arbitrary numbers. In the Standard Model extended to include neutrino masses, the leptonic mixing matrix, the analogue of the famous Cabibbo-Kobayashi-Maskawa matrix that governs quark mixing, is built from two pieces: a diagonalization matrix for the charged leptons and one for the neutrinos. The peculiar pattern of its elements has intrigued theorists ever since neutrino oscillations were discovered, and the interest only intensified after reactor experiments revealed that the third mixing angle theta-13 is not zero, while accelerator results from T2K hinted at a nearly maximal value of the CP-violating phase delta, the parameter that controls differences between neutrinos and antineutrinos.

One particularly elegant family of explanations goes by the name of bi-large mixing. The guiding observation is numerical coincidence: the smallest lepton mixing angle is close in size to the largest quark mixing angle, the Cabibbo angle, which has the value lambda equal to 0.22501 with a tiny uncertainty. This suggests that the Cabibbo angle might act as a universal seed for both quark and lepton mixings, hinting at a common origin in some grand unified theory where quarks and leptons are treated on the same footing. In bi-large schemes, two of the three neutrino mixing angles are large, of order one, while all three are expressed as simple multiples or combinations of lambda, producing sharp, testable predictions.

The new analysis, carried out by Gui-Jun Ding, Ranjeet Kumar, Newton Nath, Rahul Srivastava and José W. F. Valle, examines four such patterns, labelled T1 through T4. In the T1 pattern, the atmospheric angle satisfies sin theta-23 equal to one minus lambda, the solar angle sin theta-12 equals two lambda, and the reactor angle sin theta-13 equals lambda, with charged-lepton corrections of a CKM-like form motivated by SO(10) unification. The T2 pattern instead takes sin theta-23 equal to three lambda, with SU(5)-motivated charged-lepton corrections. The T3 and T4 patterns share the same neutrino matrix, with sin theta-12 and sin theta-23 both equal to psi times lambda, where psi is a free parameter near three, but differ in which charged-lepton correction is applied. In each case a phase phi remains as the single adjustable parameter for T1 and T2, making them one-parameter schemes whose predictions are tightly correlated.

Before JUNO, the verdict from global fits was already mixed. The T1 pattern emerged as the most compatible with oscillation data, successfully reproducing the best-fit values in the planes correlating the atmospheric angle with the CP phase, and theta-13 with theta-23. The T2 pattern survived only at the two-sigma level, preferring the lower octant of theta-23. The T3 pattern was excluded at three sigma, while T4 sat barely inside the allowed region and was expected to fall out as measurements of delta improved. In other words, the family of bi-large ideas was already cornered, and the question was whether JUNO’s solar measurement would finish the job or refine the survivors.

The answer is a dramatic tightening. JUNO is a twenty-kiloton liquid-scintillator detector located about 53 kilometres from the Yangjiang and Taishan nuclear power plants, viewed by roughly 18,000 twenty-inch photomultiplier tubes plus some 25,000 three-inch tubes, achieving an energy resolution of about three percent divided by the square root of the energy in MeV. That instrument performance translates into sub-percent goals on the solar parameters. When the JUNO constraint on sin-squared theta-12 is imposed, the T1 pattern is found to be compatible only with the higher atmospheric octant, with sin-squared theta-23 squeezed to a narrow range around 0.57, thereby excluding maximal atmospheric mixing. The T2 pattern, by contrast, still prefers nearly maximal mixing with sin-squared theta-23 near 0.5 at three sigma.

The CP predictions are equally distinctive. Imposing the JUNO solar constraint splits the T1 prediction for the Dirac CP phase into two branches, delta of approximately 0.7 pi and 1.3 pi at one sigma, which simultaneously disfavours maximal CP violation at delta equal to plus or minus pi over two and CP conservation at delta equal to pi, although the latter survives at three sigma. The T2 pattern is harsher still: it excludes CP conservation even at three sigma, because its predicted branch falls outside the JUNO band in that region. For the two-parameter patterns T3 and T4, both are excluded at two sigma in the CP-phase plane, and while one branch remains viable at three sigma, the JUNO one-sigma range pushes the allowed phase near 1.8 pi, excluding CP conservation at that level. Overall, the authors conclude that T3 is expected to be decisively tested, and potentially ruled out, by future precision measurements of the atmospheric angle and the CP phase.

The implications extend beyond oscillations into the hunt for neutrinoless double beta decay, the hypothetical process that would reveal whether neutrinos are their own antiparticles. The decay amplitude depends on the effective Majorana mass, a coherent sum over the three neutrino masses weighted by mixing matrix elements and two Majorana phases that oscillation experiments cannot access. Because the bi-large patterns do not predict the mass-squared differences or the Majorana phases, all four schemes yield similar double-beta predictions, and the analysis focuses on T1. JUNO’s precise theta-12 and solar mass-squared values slightly shrink the allowed region of the effective Majorana mass for normal mass ordering, and the T1 pattern occupies a smaller slice of that region than the JUNO-allowed band itself.

The experimental stakes are high. The current KamLAND-Zen limit corresponds to an effective mass window of 36 to 156 milli-electron-volts, while next-generation searches, including JUNO-beta-beta with 5 to 12 milli-electron-volts sensitivity, KamLAND2-Zen below 20, LEGEND-1000 at 8.5 to 19.4, nEXO at 5.7 to 17.7, and SNO+ Phase II at 19 to 46 milli-electron-volts, will probe much of the remaining parameter space, though a substantial portion of the normal-ordering region will stay out of reach. Cosmological constraints from Planck and the latest KATRIN direct-mass bound of 0.45 electron-volts further bracket the lightest neutrino mass. Taken together, the first JUNO results mark the moment when flavour model building enters a precision era: one-parameter schemes like T1 now make predictions sharp enough to be confirmed or killed within a decade, and the era of comfortable theoretical ambiguity in neutrino mixing is rapidly drawing to a close.

Subject of Research: Testing bi-large neutrino mixing patterns with the first JUNO oscillation measurements

Article Title: Zooming in on ‘bi-large’ neutrino mixing with the first JUNO results

Article References: Ding, G.-J., Kumar, R., Nath, N., Srivastava, R., & Valle, J. W. F. (2026). Zooming in on ‘bi-large’ neutrino mixing with the first JUNO results. The European Physical Journal C, 86(9), Article 1037. https://doi.org/10.1140/epjc/s10052-026-16285-9

Image Credits: AI Generated

DOI: 10.1140/epjc/s10052-026-16285-9

Keywords: neutrino oscillations, JUNO, bi-large mixing, Cabibbo angle, lepton mixing matrix, CP violation, theta-12, neutrinoless double beta decay, flavour models, grand unified theories, mass ordering, liquid scintillator