Letter of Intent for KEK Super B Factory Part I: Physics

Table 1.3: B → K ∗ γ direct CP -asymmetry
CLEO (9.1 fb −1 ) (8 ± 13 ± 3) × 10 −2
BaBar (20.7 fb −1 ) (−4.4 ± 7.6 ± 1.2) × 10 −2
Belle (78 fb −1 ) (−0.1 ± 4.4 ± 0.8) × 10 −2
Table 1.4: 90% confidence level upper limits on the B → ργ and ωγ branching fractions.
ρ + γ ρ 0 γ ωγ
CLEO (9.1 fb −1 ) 13 × 10 −6 17 × 10 −6 9.2 × 10 −6
Belle (78 fb −1 ) 2.7 × 10 −6 2.6 × 10 −6 4.4 × 10 −6
BaBar (78 fb −1 ) 2.1 × 10 −6 1.2 × 10 −6 1.0 × 10 −6
matrix element Vts with Vtd. A measurement of the b → dγ process will therefore provide the
ratio |Vtd/Vts| without large model dependent uncertainties. This mode is also the place where
a large direct CP -asymmetry is predicted within and beyond the Standard Model.
The search for the exclusive decay B → ργ is as straightforward as the measurement of
B → K ∗ γ, except for its small branching fraction, the enormous combinatorial background from
copious ρ mesons and random pions, and the huge B → K ∗ γ background that overlaps with
the B → ργ signal window. B → ωγ is not affected by B → K ∗ γ background, but it is still
unobserved. The upper limits obtained by BaBar [43], Belle [44] and CLEO [32] are summarized
in Table 1.4. The upper limits are still about twice as large as the Standard Model predictions [?]
(9.0 ± 3.4) × 10 −7 for ρ + γ, and (4.9 ± 1.8) × 10 −7 for ρ 0 γ and ωγ.
From these upper limits the bound |Vtd/Vts| < 0.34 can be obtained, which is still weaker
than the corresponding bound derived from ∆ms/∆md.
1.2.4 Electroweak Rare B Decays
Observation of B → K ∗ ℓ + ℓ −
The first signal for B → Kℓ + ℓ − was observed by Belle [17] using 29 fb −1 of data and confirmed
by BaBar [45] with 78 fb −1 , while the B → K ∗ ℓ + ℓ − signal, whose branching fraction is expected
to be larger, was not significant with those data samples.
The B → K (∗) ℓ + ℓ − signal is identified using Mbc, ∆E (and M(Kπ) for K ∗ ℓ + ℓ − ). Belle
has updated the analysis using a 140 fb −1 data sample, with a number of improvements in
the analysis procedure [18]. The most significant improvement is a lower minimum lepton
momentum of 0.7 (0.4) GeV for muons (electrons) from 1.0 (0.5) GeV to gain 12% (7%) in the
total efficiency. In addition, a K ∗ ℓ + ℓ − combinations are removed if there can be an unobserved
photon along with one of the leptons that can form a B → J/ψK → ℓ + ℓ − γK decay. As a result,
the first B → K ∗ ℓ + ℓ − signal was observed with a statistical significance of 5.7 from a fit to Mbc,
as shown in Figure 1.9, together with an improved B → Kℓ + ℓ − signal with a significance of 7.4.
The branching fractions obtained are summarized in Table 1.5, together with the BaBar
results [46]. For the combined B → K ∗ ℓ + ℓ − results, B(B → K ∗ ℓ + ℓ − ) = B(B → K ∗ µ + µ − ) =
0.75B(B → K ∗ e + e − ) is assumed which compensates for the enhancement at the q 2 = 0 pole
that appears more significantly in K ∗ e + e − , using the expected Standard Model ratio [?]. The
measured branching fractions are in agreement with the Standard Model, for example [47, 48]
(3.5 ± 1.2) × 10 −7 for B → Kℓ + ℓ − and (11.9 ± 3.9) × 10 −7 for B → K ∗ ℓ + ℓ − . We note that the
19