Search for lepton-flavor-violating τ decays into a lepton and a vector meson using the full Belle data sample

N. Tsuzuki, K. Inami, I. Adachi, H. Aihara, D. M. Asner, H. Atmacan, T. Aushev, R. Ayad, V. Babu, Sw. Banerjee, P. K. Behera, K. Belous, J. V. Bennett, M. Bessner, B. Bhuyan, T. Bilka, D. Biswas, D. Bodrov, J. Borah, A. Bożek, M. Bračko, P. Branchini, T. E. Browder, A. Budano, M. Campajola, D. Červenkov, M.-c. Chang, B. G. Cheon, K. Chilikin, H. E. Cho, K. Cho, S.-j. Cho, S.-k. Choi, Y. Choi, S. Choudhury, D. Cinabro, J. Cochran, S. Das, N. Dash, G. De Nardo, G. De Pietro, R. Dhamija, F. Di Capua, Z. Doležal, T. V. Dong, D. Dossett, S. Dubey, D. Epifanov, T. Ferber, D. Ferlewicz, B. G. Fulsom, V. Gaur, A. Giri, P. Goldenzweig, Y. Guan, K. Gudkova, X. Han, T. Hara, K. Hayasaka, H. Hayashii, M. T. Hedges, D. Herrmann, W.-s. Hou, C.-l. Hsu, T. Iijima, N. Ipsita, A. Ishikawa, R. Itoh, M. Iwasaki, W. W. Jacobs, E.-j. Jang, S. Jia, Y. Jin, T. Kawasaki, C. Kiesling, C. H. Kim, D. Y. Kim, K.-h. Kim, Y.-k. Kim, K. Kinoshita, P. Kodyš, T. Konno, A. Korobov, S. Korpar, E. Kovalenko, P. Križan, P. Krokovny, M. Kumar, K. Kumara, A. Kuzmin, Y.-j. Kwon, S. C. Lee, J. Li, L. K. Li, Y. Li, J. Libby, K. Lieret, Y.-r. Lin, D. Liventsev, Y. MaJournal of High Energy Physics 2023 · DOI

Abstract

Charged-lepton-flavor-violation (CLFV) is predicted in several new physics scenarios. We update the analysis of $\tau$ lepton decays into a light charged lepton ($\ell = e^\pm$ or $\mu^\pm$) and a vector meson ($V^0 = \rho^0, \phi, \omega, K^{*0}$, or $\bar{K}^{*0}$) using 980 fb$^{-1}$ of data collected with the Belle detector at the KEKB collider. No significant excess of such signal events is observed, and thus 90% credibility level upper limits are set on the $\tau \to \ell V^0$ branching fractions in the range of $(1.7–4.3) \times 10^{-8}$. These limits are improved by 30% on average from the previous results.

Introduction

In the Standard Model (SM) of particle physics, lepton flavor is a conserved quantity. While neutrino oscillations demonstrate flavor violation in the neutral sector, charged-lepton-flavor-violation is so heavily suppressed in the SM that any observation at current experimental sensitivities would be an unambiguous signal of New Physics.

The $\tau$ lepton, as the heaviest known lepton, is theoretically more sensitive to new, heavy particles—such as leptoquarks or heavy neutral leptons—that might mediate these forbidden decays. This study focuses on the search for $\tau \to \ell V^0$, where $\ell$ is an electron or muon and $V^0$ is a neutral vector meson. This search is particularly timely given recent "B-anomalies" in flavor physics, which suggest potential new physics interactions involving the third generation of particles.

Experimental Methodology

The analysis utilizes the full 980 fb$^{-1}$ dataset collected by the Belle detector at the KEKB $e^+e^-$ collider. The core strategy employs a "tag-and-probe" approach where each $e^+e^- \to \tau^+\tau^-$ event is divided into two hemispheres based on the thrust vector. One hemisphere is searched for the LFV candidate ($\ell V^0$), while the other is used to "tag" the event by identifying a standard $\tau$ decay.

Distribution of $\cos\theta_{\text{miss-tag}}^{\text{c.m.}}$ for $\tau \to \mu\rho^0$ with an electron tag, showing data versus signal MC.

To enhance sensitivity, this study implements Boosted Decision Trees (BDTs) via the LightGBM library. The BDTs are trained to distinguish signal from background using variables such as invariant mass, missing energy, and thrust magnitude. This multivariate approach allowed for a nearly 10% increase in signal efficiency compared to previous rectangular cut-based analyses.

Data-Driven Background Estimation

To prevent bias, a blind analysis was performed. Signal modeling relied on large-scale Monte Carlo simulations. Crucially, a unique "hadron-enhanced" sideband dataset—created by intentionally failing lepton identification requirements—allowed for direct data-driven modeling of hadronic backgrounds.

Distribution of $M_\nu^2$ for $\tau \to \mu\rho^0$ with a hadronic tag, comparing data to the signal expectation.

The BDT effectively learns the correlation between kinematic variables like $\Delta E$ (energy difference from the beam) and $M_{\ell V^0}$ (invariant mass). In a signal event, $\Delta E \approx 0$ and $M_{\ell V^0} \approx M_\tau$, allowing the BDT to filter out continuum $e^+e^- \to q\bar{q}$ events that would otherwise mimic the signal.

Results and Discussion

No significant excess of signal events was observed across any of the 20 searched decay modes. Consequently, 90% credibility level (C.L.) upper limits were set on the branching fractions ($\mathcal{B}$), ranging from $1.7 \times 10^{-8}$ to $4.3 \times 10^{-8}$. These results represent a 30% average improvement over previous best limits. For instance, the limit for $\tau \to \mu \rho^0$ was pushed down to $< 1.7 \times 10^{-8}$.

Conclusion

While no discovery was made, these results set the world's most stringent limits on these LFV processes. The main limitation remains statistical. Future experiments like Belle II, which aims to collect 50 times more data, will utilize these refined MVA techniques to push limits into the $10^{-9}$ or $10^{-10}$ regime, potentially uncovering the first signs of New Physics in the lepton sector.

Summary updated

Abstract

A bstract Charged-lepton-flavor-violation is predicted in several new physics scenarios. We update the analysis of τ lepton decays into a light charged lepton ( ℓ = e ± or μ ± ) and a vector meson ( V 0 = ρ 0 , ϕ , ω , K *0 , or $$ \overline{K} $$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mover> <mml:mi>K</mml:mi> <mml:mo>¯</mml:mo> </mml:mover> </mml:math> *0 ) using 980 fb − 1 of data collected with the Belle detector at the KEKB collider. No significant excess of such signal events is observed, and thus 90% credibility level upper limits are set on the τ → ℓV 0 branching fractions in the range of (1.7–4 . 3) × 10 − 8 . These limits are improved by 30% on average from the previous results.
BibTeX
@article{tsuzuki2023search,
  author = {N. Tsuzuki and K. Inami and I. Adachi and H. Aihara and D. M. Asner and H. Atmacan and T. Aushev and R. Ayad and V. Babu and Sw. Banerjee and P. K. Behera and K. Belous and J. V. Bennett and M. Bessner and B. Bhuyan and T. Bilka and D. Biswas and D. Bodrov and J. Borah and A. Bożek and M. Bračko and P. Branchini and T. E. Browder and A. Budano and M. Campajola and D. Červenkov and M.-C. Chang and B. G. Cheon and K. Chilikin and H. E. Cho and K. Cho and S.-J. Cho and S.-K. Choi and Y. Choi and S. Choudhury and D. Cinabro and J. Cochran and S. Das and N. Dash and G. De Nardo and G. De Pietro and R. Dhamija and F. Di Capua and Z. Doležal and T. V. Dong and D. Dossett and S. Dubey and D. Epifanov and T. Ferber and D. Ferlewicz and B. G. Fulsom and V. Gaur and A. Giri and P. Goldenzweig and Y. Guan and K. Gudkova and X. Han and T. Hara and K. Hayasaka and H. Hayashii and M. T. Hedges and D. Herrmann and W.-S. Hou and C.-L. Hsu and T. Iijima and N. Ipsita and A. Ishikawa and R. Itoh and M. Iwasaki and W. W. Jacobs and E.-J. Jang and S. Jia and Y. Jin and T. Kawasaki and C. Kiesling and C. H. Kim and D. Y. Kim and K.-H. Kim and Y.-K. Kim and K. Kinoshita and P. Kodyš and T. Konno and A. Korobov and S. Korpar and E. Kovalenko and P. Križan and P. Krokovny and M. Kumar and K. Kumara and A. Kuzmin and Y.-J. Kwon and S. C. Lee and J. Li and L. K. Li and Y. Li and J. Libby and K. Lieret and Y.-R. Lin and D. Liventsev and Y. Ma},
  title = {{Search for lepton-flavor-violating τ decays into a lepton and a vector meson using the full Belle data sample}},
  journal = {Journal of High Energy Physics},
  year = {2023},
  doi = {10.1007/jhep06(2023)118},
}
Figure