1:
2: \noindent
3: We fully reconstruct one \B\ meson~($\B_{\rm REC}$) in hadronic
4: (\Bz $\to D^{(*)-} \pi^+$, $D^{(*)-} \rho^+$,
5: $D^{(*)-} a_1^+$ and $\jpsi \Kstarz$) or
6: semileptonic ($\bztodstarlnu$) decay mode \footnote{Throughout this paper,
7: charge conjugate modes are implied.}. A total of 2577 neutral \B\ candidates is
8: reconstructed in hadronic decay modes, with an average purity close to $90\%$.
9: The main background for these modes is combinatorial. 7517 \Bz\ candidates
10: are reconstructed in the semileptonic mode, with an average purity close to
11: $84\%$. Backgrounds to the semileptonic mode are due to combinatorial \Dstar\,
12: fake leptons, uncorrelated \Dstar\ $l$ combinations, $c\bar c$ events,
13: and charged \B\ decays from $B^-\rightarrow D^{*+}(n\pi) l^- \nu$.
14:
15: The other two important ingredients for this analysis are the vertex
16: reconstruction
17: and the identification of the flavor of the other \B\ meson~($\B_{\rm
18: TAG}$) in the event. The flavor of the $\B_{\rm TAG}$ is
19: determined from the correlation between the particle types and the
20: charge of its decay products \cite{BabarPub0008}.
21: If there is an identified lepton its charge
22: is used; otherwise the summed charge of identified kaons provides the tag.
23: An event with no tagging leptons or kaons can still be tagged by the
24: use of a neural network that exploits the flavor information carried by
25: other decay products, such as soft leptons from charm semileptonic decays and soft pions
26: from \Dstar\ decays.
27:
28: At \pep2\, the \B\ meson pairs produced in the decay of the
29: \FourS\ resonance are moving in the lab frame along the
30: beam axis ($z$ direction) with a Lorentz boost of $\beta_z \gamma = 0.56$.
31: The separation between the two \B\ vertices along the boost direction,
32: $\Delta z = z_{\rm REC} - z_{\rm TAG}$, is measured and used to
33: estimate the decay time difference, $\Delta t \approx \Delta z/\beta_z
34: \gamma c$.
35: The $\B_{\rm TAG}$ vertex is determined via
36: an inclusive procedure applied to all tracks
37: not associated with the $\B_{\rm REC}$
38: meson \cite{BabarPub0001}.
39: The typical separation between the two vertices is
40: $\deltaz = \beta_z \gamma c \tau_B \approx 260$\mum, to be compared to the
41: experimental resolution $\sim 100$\mum.
42: The $\Delta t$ resolution is limited by the precision on the
43: $\B_{\rm TAG}$ vertex, and has little dependence on the decay mode of
44: the $\B_{\rm REC}$. The $\Delta t$ resolution function is well described
45: by three Gaussians: core, tail and outlier. We calculate the uncertainty
46: on $\Delta t$ by using a globally-fitted rescaling of the event-by-event
47: vertex separation errors. Most of the events, $\sim 70\%$, are in the
48: core Gaussian, with $\sigma \sim 0.6$ ps.
49: