Measurement of $B_d$ mixing using opposite-side flavor tagging

We report on a measurement of the $B_d^0$ mixing frequency and the calibration of an opposite-side flavor tagger in the D0 experiment. Various properties associated with the $b$ quark on the opposite side of the reconstructed $B$ meson are combined using a likelihood-ratio method into a single variable with enhanced tagging power. Its performance is tested with data, using a large sample of reconstructed semileptonic $B\to \mu D^{0}X$ and $B\to \mu D^{*}X$ decays, corresponding to an integrated luminosity of approximately $1{\mathrm{fb}}^{-1}$. The events are divided into groups depending on the value of the combined tagging variable, and an independent analysis is performed in each group. Combining the results of these analyses, the overall effective tagging power is found to be $\epsilon {\mathcal{D}}^2=(2.48\pm {0.21}_{-0.06}^{+0.08})\%$. The measured $B_d^0$ mixing frequency $\Delta m_d=0.506\pm 0.020(\mathrm{stat})\pm 0.016(\mathrm{syst}){\mathrm{ps}}^{-1}$ is in good agreement with the world average value.

Figure 1

Diagram of an event with a reconstructed $B^0$ candidate. PV indicates the primary vertex for the event.

Figure 2

The $M(K\pi \pi )-M(K\pi )$ invariant mass distribution for selected $\mu D^{*}$ candidates. The curve shows the result of the fit described in Sec. 7.

Figure 3

The $K\pi$ invariant mass distribution for selected $\mu D^0$ candidates. The curve shows the result of the fit described in Sec. 7.

Figure 4

The $K\pi$ invariant mass for selected $\mu D^{*}$ candidates. The curve shows the result of the fit described in Sec. 7.

Figure 5

(a) Distribution of the jet charge $Q_J^{\mu }$ for muons with hits in all three layers of the muon detector. (b) Distribution of the jet charge $Q_J^{\mu }$, for muons with fewer than three hits. (c) Distribution of the jet charge for electrons $Q_J^e$. Here $q(b^{\mathrm{rec}})$ is the charge of the muon from the reconstruction side.

Figure 6

(a) Distribution of the secondary vertex charge for events with an opposite-side muon. (b) Distribution of the secondary vertex charge for events without an opposite-side muon. (c) Distribution of the event jet charge. $q(b^{\mathrm{rec}})$ is the charge of the $b$ quark from the reconstruction side.

Figure 7

Normalized distributions of the combined tagging variable. $q(b^{\mathrm{rec}})$ is the charge of the $b$ quark from the reconstruction side.

Figure 8

Normalized distributions of the combined tagging variable for three multidimensional taggers for the simulated samples $B^{\pm }\to J/\psi K^{\pm }$. Here $q(b^{\mathrm{rec}})$ is the charge of the $b$ quark from the reconstructed side. (a) Distribution of $d$ for $\mathrm{Tag}(\mu +\mathrm{SV})$. (b) Distribution of $d$ for $\mathrm{Tag}(\mu -\mathrm{SV})$. (c) Distribution of $d$ for $\mathrm{Tag}(\mathrm{SV}-\mu )$.

Figure 9

The fit to $M(K\pi )$ mass for nonoscillating (left) and oscillating (right) $B\to {\mu }^{+}\nu D^{*-}$ events tagged by the muon tagger with $|d|>0.3$ in VPDL bins $0.0<L\le 0.025\mathrm{cm}$ (2a, 2b), $0.025<L\le 0.050\mathrm{cm}$ (3a, 3b), $0.075<L\le 0.100\mathrm{cm}$ (5a, 5b), and, $0.100<L\le 0.125\mathrm{cm}$ (6a, 6b).

Figure 10

Asymmetries obtained in the $D^{*}$ and $D^0$ sample with the combined tagger in $|d|$ bins. Circles are data, and the result of the fit is superimposed.