Abstract

We study the current estimates of B-c -> B-s pi to extract the fragmentation fraction f(c)/f(s) at the LHCb. A rather robust estimate of Br(B-c -> B-s pi) based on factorization and lattice results for the form factor gives f(c)/f(s) similar to 0.056 with a 16% error. We also revisit the extraction of f(s)/f(d) using B -> D pi instead of the theoretical cleaner but more suppressed channel B -> DK. We also find a tension on the predictions of Br(B-c -> J/psi pi) and Br(B-c -> B-s pi) considering the measurements of these modes at LHCb, and find that, within a 23% uncertainty, only the lower end of the current prediction range Br(B-c -> J/psi) similar to 0.4% - 1.7% would be consistent with the LHCb measurements.