Abstract. We present a new theoretical framework for the study of B → Xu ν decays, which includes all known perturbative and non-perturbative contributions and a description of leading and subleading Fermi motion effects. The perturbative and non-perturbative regimes are separated by a "hard" Wilsonian cutoff µ ∼ 1 GeV. We bring into focus some problems related to the high q 2 region and to Weak Annihilation effects. We provide estimates of the CKM parameter |V ub | using the described framework and discuss the related theoretical uncertainty.
IntroductionThe precise determination of the element |V ub | of the CKM matrix is an important test of the flavour structure of the Standard Model (SM) and is crucial in the indirect search for New Physics. The latest global fit to the Unitarity Triangle (UT) including all flavour changing observables but a direct determination of |V ub | predicts |V ub | = (3.44 ± 0.16) × 10 −3 [1]. This value agrees within errors with the exclusive determination, that relies on lattice QCD or light-cone sum rules [2,3] and that is still affected by somewhat large theoretical errors. An inclusive analysis is in principle the cleanest method to precisely determine |V ub |. This is based on the comparison between the decay rate of B → X u ν measured by experiments and the corresponding theoretical prediction. The latest HFAG world average [4] yields an inclusive |V ub | which is about 2.5σ higher than the value preferred by the global UT fit, calling for a deeper investigation of the process.The theoretical description of B → X u ν is based on a local Operator Product Expansion. Inclusive quantities are organized in a double series in α s (perturbative QCD corrections) and in 1/m b (Heavy Quark Expansion). The very same method was successfully applied to the b → c decay and led to a precise determination of |V cb |, within 2%. The description of charmless decays is more involved due to the dominant charmed background that needs to be rejected by experiments imposing very stringent cuts. These cuts can spoil the convergence of the OPE introducing sensitivity to nonlocal effects, such as the motion of the b quark inside the meson (Fermi motion), that can be parameterized in terms of a light-cone distribution function (or "shape function"). The lowest integer moments of the distribution function are constrained by the OPE [5] and they are expressed in terms of the b quark mass and of the same 5 and 6 dimensional operators that contribute to B → X c ν. Such expressions are universal, i.e. independent of the process, and shared by the radiative decay B → X s γ only as long as 1/m b corrections are neglected.