Search for the B+ -> K+vv Decay Using Semi-Leptonic Tags

B. Aubert, Y. Karyotakis, J. P. Lees, V. Poireau, E. Prencipe, X. Prudent, V. Tisserand, J. Garra Ticó, E. Graugés, M. Martinelli, A. Palano, M. Pappagallo, G. Eigen, B. Stugu, Luyan Sun, M. Battaglia, D. N. Brown, B. H. Hooberman, L. T. Kerth, Yu. G. Kolomensky, G. Lynch, I. L. Osipenkov, K. Tackmann, T. Tanabe, C. M. Hawkes, N. Soni, A. T. Watson, Henrik Koch, T. Vazquez Schroeder, D. J. Asgeirsson, C. Hearty, T. S. Mattison, J. A. McKenna, M. Barrett, Ayaz H. Khan, A. Randle-Conde, V. E. Blinov, A. D. Bukin, А. Р. Бузыкаев, В. П. Дружинин, V. B. Golubev, A. P. Onuchin, S. I. Serednyakov, Yu. I. Skovpen, E. P. Solodov, K. Yu. Todyshev, M. Bondioli, S. Curry, I. Eschrich, D. Kirkby, A. J. Lankford, Per Lund, M. Mandelkern, E. C. Martin, D. P. Stoker, H. Atmacan, J. W. Gary, Franklin Liu, O. R. Long, G. M. Vitug, Z. Yasin, V. Sharma, C. Campagnari, T. M. Hong, D. Kovalskyi, M. A. Mazur, And J. D. Richman, T. W. Beck, A. M. Eisner, C. A. Heusch, J. Kroseberg, W. S. Lockman, A. J. Martínez, T. Schalk, B. A. Schumm, A. Seiden, L. Wang, L. O. Winstrom, C. H. Cheng, D. Doll, B. Echenard, F. Fang, D. G. Hitlin, I. Narsky, P. Ongmongkolkul, T. Piatenko, F. C. Porter, R. Andreassen, G. Mancinelli, B. Meadows, K. Mishra, M. D. Sokoloff, P. C. Bloom, W. T. Ford, A. Gaz, J. F. Hirschauer, M. Nagel, U. Nauenberg, J. G. Smith, S. R. WagnerarXiv 2009 · arXiv

Abstract

This study presents a sensitive search for the rare decay $B^+ \to K^+\nu\bar{\nu}$, a process involving flavor-changing neutral currents (FCNC) that are highly suppressed in the Standard Model. Using $319 \text{ fb}^{-1}$ of data collected by the BABAR detector at the $\Upsilon(4S)$ resonance, we employ a semi-leptonic tagging technique to reconstruct one $B$ meson and search for the signal in the recoil. A multivariate Random Forest classifier was used to enhance signal-to-background separation. We find 38 candidate events, consistent with the expected background of $31 \pm 12$. We set an upper limit on the branching fraction of $\mathcal{B}(B^+ \to K^+\nu\bar{\nu}) < 4.5 \times 10^{-5}$ at the 90% confidence level.

Introduction

In the Standard Model (SM), the decay $B^+ \to K^+\nu\bar{\nu}$ is a flavor-changing neutral current process that cannot occur at the tree level. Instead, it proceeds through higher-order quantum loops, such as the Z-penguin and box diagrams. Because these processes are extremely rare—with an SM predicted branching fraction of approximately $3.8 \times 10^{-6}$—they serve as sensitive probes for New Physics. Contributions from Supersymmetry, unparticles, or light dark matter candidates could significantly enhance this rate.

Feynman diagrams showing the loop-level processes (Z-penguin and box diagrams) responsible for the $B^+ \to K^+\nu\bar{\nu}$ decay in the Standard Model.

Detecting this decay is exceptionally challenging because the two neutrinos escape the detector unseen, leaving only a single charged kaon and "missing" energy as experimental signatures.

Experimental Methodology

Dataset and Tagging

The analysis utilized data from the BABAR detector at the SLAC National Accelerator Laboratory, comprising 351 million $B\bar{B}$ pairs. To identify the signal, the researchers used a Semi-Leptonic Tagging method. Since $B$ mesons are produced in pairs at the PEP-II collider, one $B$ meson (the "tag") is fully reconstructed through the decay $B^- \to D^0 \ell^- \bar{\nu} X$. Once the tag is identified, the remaining particles in the event are analyzed for the signal $B^+ \to K^+\nu\bar{\nu}$.

Analysis Strategy

To distinguish the rare signal from the large background of other $B$ decays and continuum events, the team employed a Random Forest (RF) classifier. This multivariate machine learning approach utilized 19 different kinematic and topological variables. Key variables included the number of remaining charged tracks, unmatched neutral energy in the calorimeter, the momentum of the signal kaon candidate, and the total missing energy in the event.

Distributions of four key variables used in the Random Forest classifier: (a) remaining charged tracks, (b) unmatched neutral energy, (c) signal kaon momentum, and (d) missing energy. The thick black line represents the signal, while shaded regions represent various background sources.

The RF classifier provided a 34% improvement in sensitivity over traditional rectangular selection cuts by capturing complex correlations between these variables. To ensure objectivity, the analysis was conducted "blind," with the signal region hidden until the selection criteria were finalized.

Results and Discussion

After unblinding the data, 38 candidate events were observed in the signal region. This is consistent with the predicted background of $31 \pm 12$ events, which was estimated using Monte Carlo simulations and validated in data sidebands.

Given the lack of a statistically significant excess, an upper limit was calculated. The resulting limit of $\mathcal{B}(B^+ \to K^+\nu\bar{\nu}) < 4.5 \times 10^{-5}$ at the 90% confidence level is an improvement over previous BABAR results. While this result is consistent with the Standard Model, it further constrains the parameter space for potential New Physics models.

Conclusion

This search represents one of the most sensitive probes of the $B^+ \to K^+\nu\bar{\nu}$ decay to date. The current sensitivity is primarily limited by the size of the available dataset. Future experiments, such as Belle II, are expected to collect significantly more data, which will be necessary to either observe the decay at its predicted SM rate or uncover definitive evidence of physics beyond the Standard Model.

Summary updated

Abstract

11 pages, 6 figures, submitted to Phys. Rev. D
BibTeX
@misc{aubert2009search,
  author = {B. Aubert and Y. Karyotakis and J. P. Lees and V. Poireau and E. Prencipe and X. Prudent and V. Tisserand and J. Garra Ticó and E. Graugés and M. Martinelli and A. Palano and M. Pappagallo and G. Eigen and B. Stugu and Luyan Sun and M. Battaglia and D. N. Brown and B. H. Hooberman and L. T. Kerth and Yu. G. Kolomensky and G. Lynch and I. L. Osipenkov and K. Tackmann and T. Tanabe and C. M. Hawkes and N. Soni and A. T. Watson and Henrik Koch and T. Vazquez Schroeder and D. J. Asgeirsson and C. Hearty and T. S. Mattison and J. A. McKenna and M. Barrett and Ayaz H. Khan and A. Randle-Conde and V. E. Blinov and A. D. Bukin and А. Р. Бузыкаев and В. П. Дружинин and V. B. Golubev and A. P. Onuchin and S. I. Serednyakov and Yu. I. Skovpen and E. P. Solodov and K. Yu. Todyshev and M. Bondioli and S. Curry and I. Eschrich and D. Kirkby and A. J. Lankford and Per Lund and M. Mandelkern and E. C. Martin and D. P. Stoker and H. Atmacan and J. W. Gary and Franklin Liu and O. R. Long and G. M. Vitug and Z. Yasin and V. Sharma and C. Campagnari and T. M. Hong and D. Kovalskyi and M. A. Mazur and And J. D. Richman and T. W. Beck and A. M. Eisner and C. A. Heusch and J. Kroseberg and W. S. Lockman and A. J. Martínez and T. Schalk and B. A. Schumm and A. Seiden and L. Wang and L. O. Winstrom and C. H. Cheng and D. Doll and B. Echenard and F. Fang and D. G. Hitlin and I. Narsky and P. Ongmongkolkul and T. Piatenko and F. C. Porter and R. Andreassen and G. Mancinelli and B. Meadows and K. Mishra and M. D. Sokoloff and P. C. Bloom and W. T. Ford and A. Gaz and J. F. Hirschauer and M. Nagel and U. Nauenberg and J. G. Smith and S. R. Wagner},
  title = {{Search for the B+ -> K+vv Decay Using Semi-Leptonic Tags}},
  year = {2009},
  eprint = {0911.1988},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/0911.1988},
}
Figure