Evidence for the Decay $B_s^0\to D_s^{(*)}{D}_s^{(*)}$ and a Measurement of $\Delta {\Gamma }_s^{CP}/{\Gamma }_s$

We search for the semi-inclusive process $B_s^0\to D_s^{(*)}{D}_s^{(*)}$ using $2.8{\mathrm{fb}}^{-1}$ of $p\overline{p}$ collisions at $\sqrt{s}=1.96\mathrm{TeV}$ recorded by the D0 detector operating at the Fermilab Tevatron Collider. We observe $26.6\pm 8.4$ signal events with a significance above background of 3.2 standard deviations yielding a branching ratio of $\mathcal{B}(B_s^0\to D_s^{(*)}{D}_s^{(*)})=0.035\pm 0.010(\mathrm{stat}.)\pm 0.011(\mathrm{syst}.)$. Under certain theoretical assumptions, these double-charm final states saturate $CP$-even eigenstates in the $B_s^0$ decays resulting in a width difference of $\Delta {\Gamma }_s^{CP}/{\Gamma }_s=0.072\pm 0.021(\mathrm{stat}.)\pm 0.022(\mathrm{syst}.)$.

Figure 1

Invariant mass distribution of the $\varphi \pi$ system for the $B_s^0\to D_s^{(*)}\mu \nu$ sample. The two peaks correspond to the $D^{\pm }$ candidates (lower masses) and $D_s$ candidates (higher masses).

Figure 2

Projections of the two-dimensional maximum likelihood fit onto invariant mass spectra of the (a) $\varphi \pi$ system from hadronic $D_s$ decays and (b) $KK$ system from semileptonic $D_s$ decays. The peaks in both distributions are explored to search for the correlation between the two systems.

Figure 3

Constraints in the $\Delta {\Gamma }_s-{\varphi }_s$ plane. The solid line represents our measurement under the theoretical assumptions stated in the text and with $x_f=0$. Two pairs of lines are 68% (dashed) and 90% (dotted) C.L. intervals of $\Delta {\Gamma }_s$ for a given assumed value of ${\varphi }_s$. Contours from the $B_s^0\to J/\psi \varphi$ decay are the equivalent C.L. regions of ($\Delta {\Gamma }_s$, ${\varphi }_s$) when measuring simultaneously both parameters. No theoretical uncertainties are reflected in the plot. The SM prediction is represented by the thick vertical line.