⚛️ high-energy experiments

Measurements of lepton-flavour universality in semileptonic $B$ decays at Belle and Belle II

This paper presents new measurements of lepton-flavour universality ratios $R(D^{(*)})$ and angular observables in semileptonic $B$ decays from the Belle and Belle II experiments, which, while individually consistent with Standard Model expectations, collectively contribute to the ongoing tension between experimental world averages and theoretical predictions.

Original authors: Andreas Warburton

Published 2026-10-14

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Andreas Warburton

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the subatomic world, nature appears to follow a strict rule of fairness known as lepton-flavour universality. This principle suggests that the three types of charged particles called leptons—the electron, the muon, and the tau—should behave almost exactly the same way when they interact with other particles, differing only in their mass. Imagine these three particles as identical twins; while one might be slightly heavier than the others, they should all react to forces in the same manner. For decades, this rule has held up under intense scrutiny, forming a cornerstone of our understanding of physics. However, recent observations have hinted that this fairness might be broken. When heavy particles called B-mesons decay, they sometimes seem to favor the heavy tau particle over the lighter electron or muon more often than the standard laws of physics predict. This discrepancy has sparked a global hunt for new physics, a search for forces or particles that exist beyond our current models.

To investigate this mystery, scientists at the Belle and Belle II experiments in Japan have been meticulously counting these rare decays. These experiments take place in a particle accelerator where electrons and their antimatter counterparts, positrons, are smashed together. These collisions create pairs of B-mesons that fly apart and eventually decay. Because the initial conditions of the collision are so well understood, researchers can reconstruct the entire event, even when some particles, like neutrinos, escape detection. The team recently presented a suite of new measurements that compare how often B-mesons decay into a tau particle versus how often they decay into an electron or a muon. They also examined the angles at which these particles fly apart, looking for subtle differences in behavior that might reveal a hidden interaction.

The researchers focused on two main ways of identifying the decays. In one approach, they reconstructed the partner B-meson in the collision using a method called hadronic tagging, which looks for specific combinations of particles. This method is very clean and precise but captures fewer events. In the other approach, they used semileptonic tagging, which identifies the partner B-meson by looking for a lighter lepton in its decay. This method is less precise regarding the partner's details but captures a much larger number of events. By combining these strategies, the team measured the ratio of decays involving the heavy tau particle against those involving the lighter electrons or muons.

The results from the Belle II experiment, using the hadronic tagging method, found that the ratio of tau decays to light-lepton decays was approximately 0.439 for one type of decay and 0.242 for another. When using the semileptonic tagging method, the team found ratios of 0.418 and 0.306. While these numbers are slightly higher than what the standard model of physics predicts, the difference is not large enough to be considered a discovery. The measurements are consistent with the long-standing average of all previous experiments, which collectively show a tension with the standard model, but the new data alone does not confirm a violation of the rules. The statistical uncertainty in these measurements means that the observed differences could still be due to random chance rather than new physics.

Beyond just counting the decays, the scientists also looked at the direction in which the particles were emitted. They analyzed the angular distributions of the decay products to see if the electron and muon modes behaved differently from each other. If the laws of physics were treating these two light particles differently, it would show up as a distinct pattern in how they flew away from the collision point. The team measured various asymmetries and coefficients that describe these angles. In every case, the electron and muon behaved identically, with no evidence of a difference in their interactions. The data aligns perfectly with the standard model's expectation that electrons and muons are interchangeable in these processes.

The combined picture from these new measurements is one of cautious consistency. Individually, none of the four new results—the two ratios from the hadronic tag, the two from the semileptonic tag, and the angular tests—shows a statistically significant deviation from what the standard model predicts. The slight excesses in the tau decay rates remain within the margin of error. However, when these new results are added to the global collection of data from around the world, the overall tension with the standard model persists. The world average still suggests that tau particles might be favored more than expected, a hint that keeps the search for new physics alive.

The Belle II experiment has now collected more data than its predecessor, the original Belle experiment, and continues to grow. With this larger dataset, the researchers can refine their reconstruction methods and explore even more decay channels. The future of this field lies in combining these increasingly precise rate measurements with detailed studies of how the particles move and interact. If the current hints of a violation are real, the next generation of data will likely reveal the structure of the new forces at play. Until then, the universe continues to hold its secrets, with the lepton-flavour universality rule standing firm in the face of these new, precise tests, yet still leaving a small door open for discovery.

Technical Summary: Measurements of Lepton-Flavour Universality in Semileptonic $B$ Decays at Belle and Belle II

Problem and Motivation
In the Standard Model (SM), the charged-current couplings of the three charged lepton flavours ( $e, \mu, \tau$ ) to the $W$ boson are universal, with differences arising solely from mass effects. Semileptonic $B$ -meson decays serve as precise probes for testing Lepton-Flavour Universality (LFU) and searching for physics beyond the SM, particularly additional charged-current interactions. The primary observables of interest are the branching-fraction ratios $R(D^{(*)})$ , defined as:
$R(D^{(*)}) \equiv \frac{\mathcal{B}(\bar{B} \to D^{(*)} \tau^- \bar{\nu}_\tau)}{\mathcal{B}(\bar{B} \to D^{(*)} \ell^- \bar{\nu}_\ell)}, \quad \ell = e, \mu$
These ratios benefit from the cancellation of many experimental and theoretical uncertainties. While the underlying $b \to c$ quark transition is identical for both numerator and denominator, the larger $\tau$ mass allows sensitivity to non-SM scalar, vector, and tensor interactions. Previous combined measurements of $R(D)$ and $R(D^*)$ have exhibited tension with SM predictions, motivating improved rate measurements and complementary studies of differential and angular observables. Additionally, angular distributions in $B \to D^* \ell \nu$ decays provide information on the Lorentz structure of the interaction, offering direct tests of light-lepton ( $e-\mu$ ) universality.

Methodology
The analysis utilizes data from the Belle and Belle II experiments operating at the KEKB and SuperKEKB $e^+e^-$ colliders in Japan, respectively. Both experiments collect collisions at the $\Upsilon(4S)$ resonance, producing $B\bar{B}$ pairs. The clean environment, constrained initial-state kinematics, and excellent lepton identification make these datasets ideal for studying semileptonic decays involving neutrinos.

The analyses employ "tagging" strategies where one $B$ meson (the tag) is reconstructed to constrain the kinematics of the signal $B$ meson. The paper highlights the complementarity of two approaches:

Semileptonic Tagging: High efficiency but with relatively limited knowledge of the tag-side kinematics.
Hadronic Tagging: Higher purity but lower efficiency, providing fully constrained event kinematics.

Key Contributions and Results

New Belle II $R(D^{(*)})$ Measurements:
- Semileptonic-Tag Measurement: Using a multivariate classifier to separate signal ( $B \to D^{(*)} \tau \nu$ $B \to D^{(*)} τ ν$ ), normalization ( $B \to D^{(*)} \ell \nu$ $B \to D^{(*)} ℓ ν$ ), and background, the analysis fits four channels ( $D^+ e^-, D^+ \mu^-, D^{*+} e^-, D^{*+} \mu^-$ $D^{+} e^{-}, D^{+} μ^{-}, D^{*+} e^{-}, D^{*+} μ^{-}$ ) simultaneously.
  - Results: $R(D^+) = 0.418 \pm 0.075 (\text{stat}) \pm 0.056 (\text{syst})$ and $R(D^{*+}) = 0.306 \pm 0.035 (\text{stat}) \pm 0.018 (\text{syst})$ .
  - The result differs from the SM expectation by $1.7\sigma$ and is compatible with experimental world averages.
- Hadronic-Tag Measurement: This analysis addresses backgrounds from $B \to D^{**} \ell \nu$ $B \to D^{**} ℓ ν$ modes by requiring no additional charged tracks and utilizing the squared missing mass ( $M^2_{\text{miss}}$ $M_{miss}^{2}$ ) and residual energy ( $E_{\text{ECL}}$ $E_{ECL}$ ) in a two-dimensional extended maximum-likelihood fit.
  - Results: $R(D) = 0.439 \pm 0.055 (\text{stat}) \pm 0.046 (\text{syst})$ and $R(D^*) = 0.242 \pm 0.019 (\text{stat}) \pm 0.016 (\text{syst})$ .
  - This constitutes the most precise determination of $R(D^{(*)})$ using hadronic tagging to date. The combined measurement differs from the SM by $1.5\sigma$ and from the world average by $1.3\sigma$ .
Angular Tests of Light-Lepton Universality:
- Belle II ( $e-\mu$ differences): Using a $189 \text{ fb}^{-1}$ sample, angular asymmetries ( $A_{FB}, S_i$ ) were measured for $B^0 \to D^{*-} \ell^+ \nu_\ell$ in inclusive and recoil-binned samples. Joint tests of LFU-sensitive differences ( $\Delta A_{FB}, \Delta S_3, \Delta S_5$ ) yielded $p$ -values of $0.57$ (inclusive) and $0.13$ (subranges), providing no evidence for LFU violation.
- Belle (Angular Coefficients): Using the final $711 \text{ fb}^{-1}$ dataset, Belle measured twelve angular coefficients ( $\hat{J}_i$ ) describing the fully differential decay distribution for all $B \to D^* \ell \nu$ modes. A simultaneous fit across 576 phase-space regions compared $e$ and $\mu$ observables. The comparison with the SM yielded $\chi^2/\text{ndf} = 41/48$ ( $p=0.76$ ), showing no deviation from SM expectations.

Significance and Outlook
The paper reports that all four new measurements (two $R(D^{(*)})$ ratios from Belle II and two angular analyses from Belle II and Belle) are individually consistent with Standard Model expectations, though they are statistically limited. The new $R(D^{(*)})$ results are compatible with the experimental world averages, which continue to exhibit tension with SM predictions.

The significance of this work lies in the demonstration of complementary analysis strategies (semileptonic vs. hadronic tagging) and the provision of precise constraints on both total rates and angular observables. With Belle II having now accumulated an $\Upsilon(4S)$ data sample exceeding that of Belle, the paper notes that the field is entering a phase of heightened precision. Future combinations of increasingly precise $R(D^{(*)})$ values with differential observables will be critical for determining whether the existing tensions persist and, in the event of a confirmed deviation, for identifying the structure of the underlying physical interactions.

Technical Summary: Measurements of Lepton-Flavour Universality in Semileptonic BBB Decays at Belle and Belle II

More like this

Technical Summary: Measurements of Lepton-Flavour Universality in Semileptonic $B$ Decays at Belle and Belle II