Search For New Physics at BABAR

Using a full BABAR data sample of 426 $fb^{-1}$, we present improved measurements of the ratio ${\cal{R}}(D^{(*)}) = {\cal{B}} (\bar{B} \to D^{(*)}\tau^{-}\bar{\nu}_{\tau})/$ ${\cal{B}} (\bar{B} \to D^{(*)}\ell_{\ell}^{-}\bar{\nu}_{\ell})$, where $\ell$ is either electron or muon. We measure ${\cal{R}}(D) = 0.440 \pm 0.058 \pm 0.042$ and ${\cal{R}}(D^*) = 0.332 \pm 0.024 \pm 0.018$. These ratios exceed the Standard Model predictions by $2.0\sigma$ and $2.7\sigma$, respectively. The results disagree with the Standard Model predictions at the level of $3.4\sigma$. The ratios are sensitive to new physics contributions in the form of a charged Higgs boson. However, the access cannot be explained by a charged Higgs boson in the type II two-Higgs-doublet model.


INTRODUCTION
The semileptonic physics in B meson sector played a prominent role in investigating of new physics effect at low-energy region. Semileptonic transitions are the simplest process in B mesons decay. In the Standard Model (SM), the heavy b quark decays to either a c or an u quark and the virtual W boson [1][2][3]. Experimentally, semileptonic decays have the advantage of large branching fraction and are used to determine the weak couplings, the Cabibbo-Kobayashi-Maskawa (CKM) matrix elements, |V cb | and |V ub | [4].
The parton level diagram ofB → D ( * ) τ −ν τ decays where D ( * ) refers to either a D or a D * meson is shown in Fig. 1. The decayB → D ( * ) τ −ν τ is sensitive to charged Higgs contribution at the tree level. The three-body de- cayB → D ( * ) τ −ν τ permits the study of decay distribution which discriminate between W − and H − exchange [5,6].
The decay ofB → D ( * ) τ −ν τ with τ lepton in the final state offer possibilities of significant new physics contribua This work was supported by the U.S. Department of Energy under grant No. DE-FG02-96ER-40970 e-mail: godang@usouthal.edu tions that is not present in the process where light lepton such as electron and muon in the final state. The study of B → D ( * ) τ −ν τ have already shown the new physics contributions can be over-constrained [7][8][9][10]. The existing studies of theB → D ( * ) τ −ν τ based on two-Higgs-doublet model (2HDM) predict a substantial impact on the ratio R(D ( * ) ) and R(D) [7,8,10,12].

The BABAR DETECTOR AND DATA SET
The BABAR detector was operated at the PEP-II asymmetric-energy storage rings at the SLAC National Accelerator Laboratory. The data used in this analysis were collected with the BABAR detector. We analyze data recorded with the BABAR detector at a center of mass energy of 10.58 GeV. The data sample consist of an integrated luminosity of 426 f b −1 , corresponding to 471 ×10 6 BB pairs. An additional sample of 40 f b −1 , taken at energy 40 MeV below the Υ(4S ) resonance. This additional sample of data is used to study the continuum background from the decays of e + e − → qq(γ) pairs where q can be u, d, s, c, τ.
A detail description of the BABAR detector is presented elsewhere [11]. The momenta of the charged particles are measured in a tracking system consisting of a 5-layer double sided silicon vertex tracker (SVT) and a 40-layer drift chamber (DCH). The SVT and DCH operate within a 1.5 T solenoid field and have a combined solid angle coverage in the center of mass frame of 90.5%. A detector of internally reflected Cerenkov radiation (DIRC) is used for charged particle identifications of pions, kaons, and protons with likelihood ratios calculated from dE/dx measurements in the SVT and DCH. Photons and longlived neutral hadrons are detected and their energies are measured in a CsI(Tl) electromagnetic calorimeter (EMC). For electrons, energy lost due to bremsstrahlung is recovered from deposits in the EMC.

ANALYSIS
In this analysis, instead of measuring the absolute branching fraction ofB → D ( * ) τ −ν τ , we measure the ratios where ℓ is either electron or muon. In the standard model (SM), the relative rate R(D ( * ) ) have less than 6% uncertainty [12]. In the decay ofB → D ( * ) τ −ν τ , we construct the τ lepton only from the purely lepton decays: τ − → e −μ e ν τ and τ − → e −ν µ ν τ so that the signal events (B → D ( * ) τ −ν τ ) and the normalization events (B → D ( * ) ℓ −ν ℓ ) are identified by the same particles in the final state. When taking the ratio of R(D ( * ) ), the various sources of uncertainties will be canceled and reduced.
We reconstruct candidate events produced in Υ(4S ) → BB decays by selecting the hadron decay of one of the B meson (B tag ). The other candidate events are reconstructed semileptonically, specially a charm meson (either charged or neutral D or D * ) and a charged lepton (either e or µ). The signal events (B → D ( * ) τ −ν τ ) and the normalization events (B → D ( * ) ℓ −ν ℓ ) are extracted using unbinned maximum-likelihood fit to the two-dimensional distributions of the invariant mass of the undetected particles. Basically it is the invariant mass of the neutrinos.
where p is the four-momenta of the colliding beams, B tag , D ( * ) and charged lepton, respectively. The lepton threemomentum in the B rest frame is denoted by p * ℓ . The distribution of the lepton three-momentum of the signal events is softer than the distribution of the lepton threemomentum of the normalization events because the observed lepton in the signal events is a secondary particle originated from the τ decay, τ − → ℓ −ν ℓ ν τ . If all particles are properly reconstructed the invariant mass of the undetected particles (m 2 miss ) with a single missing neutrino peaks at zero, whereas the signal events which have three missing neutrinos have a wide m 2 miss distribution that extends from -1 GeV 2 to 10 GeV 2 . The two observable kinematic variables are used to select the B tag candidates: and where the p tag and E tag refer to the center-of-mass momentum and energy of the B tag . E beam is the center-of-mass of a single beam particle. If the B decays are correctly reconstructed, the distribution of the m ES is centered at the B meson mass with a resolution of 2.5 MeV. The distribution of ∆E is centered at zero with a resolution of 18 MeV. In this analysis we required m ES > 5.27 GeV and |∆E| < 0.072 GeV.
The main background contributions to the signal events are the following: to the charm resonances heavier than the D * meson such as D * 0 , D 1 , D ′ 1 , and D * 2 orbital excitations of the cq pairs. The decay ofB → D * * ℓ −ν ℓ where the D * meson decays to D ( * ) π 0 peaks in the m 2 miss distribution. These events are estimated using the Monte Carlo samples.
• Charge cross-feed events: these events come from the decay ofB → D * * (τ − /ℓ − )ν ℓ . These background events were reconstructed with the incorrect charge where one of the charges particles in the final state has been assigned to the different B meson.
• Other BB background: These events come from the decay of B → D ( * , * * ) D ( * , * * )+ s due to the large leptonic and semileptonic branching fractions of D + s mesons. We estimate these events using Monte Carlo sample and its contribution is fixed in the fitting process.   The results of the signal and normalization yields are extracted using unbinned maximum-likelihood fit to the two dimensional, m 2 miss -p * ℓ contributions. The fit is performed simultaneously to the four D ( * ) ℓ samples and four D ( * ) π 0 ℓ samples. The distributions of each D ( * ) ℓ and D ( * ) π 0 ℓ sample is fitted to the sum of either eight or six contributions, respectively. The fit relies on 8 × 4 + 6 × 4 = 56 probability density functions (PDFs). The two dimensional, m 2 miss -p * ℓ contributions for each of the 56 PDFs are Hadron Collider Physics symposium 2012 described in detail using smooth non-parametric kernel estimator [13]. The m 2 miss distributions of the signal events and the normalization events cane be easily distinguished due to the different number of neutrino as the undetected particles in its corresponding decays in the final state. However, the m 2 miss distributions of the backgrounds resemble those of the signal events, and therefore in the fitting procedure these contributions are either fixed fitted or constrained by the D ( * ) π 0 ℓ Monte Carlo sample. Figure 3   We extract the branching fraction ratios as define in the following where N sig and N norm are the number of signal and normalization events extracting from the fitting process, respectively. The ǫ sig /ǫ norm is the ratio of the efficiencies of the signal and the normalization events. We impose the isospin relations of R(D * ) ≡ R(D * + ) = R(D * 0 ) and R(D) ≡ R(D + ) = R(D 0 ). Table 1 shows the fit results of the yield of B → D * τν with the statistical uncertainties only. Figure 4 shows the yield of B → Dτν comparison of the m 2 miss (left) and p * ℓ (right) distributions of the B → Dτν (data points) with the projections of the results of the isospin-unconstrained fit.
The region above the dashed line is the background component corresponds to BB background. The region Mode 0.43 ± 0.08 0.47 ± 0.08 0.44 ± 0.06 B(Dτν) 0.99 ± 0.19 1.01 ± 0.18 1.02 ± 0.13 below the dashed line corresponds to the continuum background. In the p * ℓ distributions, we only include events with m 2 miss > 1 GeV 2 . Table 2 shows the fit results of the

SYSTEMATICS UNCERTAINTIES
The largest systematic uncertainties in this analysis is due to the poorly understood of the decay B → D * * (ℓ/τ)ν background. The systematic uncertainty due the PDF that describe these contributions including the uncertainty on the branching fractions of the four B → D * * ℓν decays, the branching fraction ratio of B → D * * τν to B → D * * ℓν, and its relative efficiency. We assign 2.1% on R(D) and 1.8% on R(D * ), respectively. We also assign a systematic uncertainty due to the observed variation of the decay of B → D * ηℓν, nonresonance B → D * π(π)ℓν, B → D * * (ℓ/τ)ν, and D * * → D ( * ) ππ. They are 2.1% for R(D) and 2.6% for R(D * ).
The other largest systematic uncertainties are due to the continuum and BB backgrounds. We assign 4.9% for R(D) and 2.7% for R(D * ). The systematic uncertainties due to the PDFs for the signal and normalization decays are 4.3% for R(D) and 2.1% for R(D * ). The systematic uncertainties due to the efficiency ratios ǫ sig and ǫ norm are 2.6% on R(D) and 1.6% on R(D * ), respectively.
By choosing the decay of τ lepton only from the purely lepton decays: τ − → e −ν e ν τ and τ − → e −ν µ ν τ , uncertainties due to the particle identification, final state radiation, soft-pion reconstruction, and other related detector performance are largely cancel in taking the ratios. They only contribute about 1% in the systematic uncertainty.
All systematic uncertainties are added in quadrature to assign the total systematic uncertainty. There is a positive correlation between some of the systematic uncertainties R(D) and R(D * ). As a result the correlation of the total uncertainties is reduced to -0.27 for R(D) and R(D * ).

CONCLUSIONS
We have measured the R(D) = 0.440 ± 0.058 ± 0.042 and R(D * ) = 0.332 ± 0.024 ± 0.018. These ratios exceed the Standard Model predictions by 2.0σ and 2.7σ, respectively. They are disagree with the SM prediction at the level of 3.4σ. The results are compatible with the results measured by the Belle Collaboration [14,15]. Together with the results measured by the Belle Collaboration, it could be an indication of new physics processes in B mesons decay.
The measured values of R(D) and R(D * ) match the predictions of the particular Higgs model where the ratio of neutral Higgs field vacuum expectation values, tanβ, and the mass of the physical charged Higgs boson, m H + is tanβ/m H + = 0.44 ± 0.02 GeV −1 and tanβ/m H + = 0.75 ± 0.04 GeV −1 , respectively. Figure 5 shows the comparison results of this paper with the charged Higgs boson of type II 2HDM predictions. The Standard Model predictions correspond to tanβ/m H + = 0. However, these results are not compatible with a charged Higgs boson in the type II 2HDM with a 99.8% confidence level for any value of tanβ/m H + . More general charged-Higgs models or other New Physics contributions may explain the access of R(D) and R(D * ).