What can we learn from Bto a₁(1260)(b₁(1235))π(K) decays?

We investigate the $B\to a_1(1260)(b_1(1235))\pi(K)$ decays under the factorization scheme and find many discrepancies between theoretical predictions and the experimental data. In the tree dominated processes, large contributions from color-suppressed tree diagrams are required in order to accommodate with the large decay rates of $B^-\to a_1^0\pi^-$ and $B^-\to a_1^-\pi^0$. For $\bar B^0\to (a_1^+, b_1^+)K^-$ decays which are both induced by $b\to s$ transition, theoretical predictions on their decay rates are larger than the data by a factor of 2.8 and 5.5, respectively. Large electro-weak penguins or some new mechanism are expected to explain the branching ratios of $B^-\to b_1^0K^-$ and $B^-\to a_1^-\bar K^0$. The soft-collinear-effective-theory has the potential to explain large decay rates of $B^-\to a_1^0\pi^-$ and $B^-\to a_1^-\pi^0$ via a large hard-scattering form factor $\zeta_J^{B\to a_1}$. We will also show that, with proper charming penguins, predictions on the branching ratios of $\bar B^0\to (a_1^+, b_1^+)K^-$ can also be consistent with the data.