Soft lepton-flavor violation in a multi-Higgs-doublet seesaw model

We consider the Standard Model with an arbitrary number n_H of Higgs doublets and enlarge the lepton sector by adding to each lepton family \ell a right-handed neutrino singlet \nu_{\ell R}. We assume that all Yukawa-coupling matrices are diagonal, but the Majorana mass matrix M_R of the right-handed neutrino singlets is an arbitrary symmetric matrix, thereby introducing an explicit but soft violation of all lepton numbers. We investigate lepton-flavor-violating processes within this model. We pay particular attention to the large-m_R behavior of the amplitudes for these processes, where m_R is the order of magnitude of the matrix elements of M_R. While the amplitudes for processes like tau^- --> mu^- gamma and Z --> tau^+ mu^- drop as 1/m_R^2 for arbitrary n_H, processes like tau^- --> mu^- e^+ e^- and mu^- --> e^- e^+ e^- obey this power law only for n_H = 1. For n_H \geq 2, on the contrary, those amplitudes do not fall off when m_R increases, rather they converge towards constants. This non-decoupling of the right-handed scale occurs because of the sub-process ell^- --> ell'^- {S_b^0}^*, where S_b^0 is a neutral scalar which subsequently decays to e^+ e^-. That sub-process has a contribution from charged-scalar exchange which, for n_H \geq 2, does not decrease when m_R tends to infinity. We also perform a general study of the non-decoupling and argue that, after performing the limit m_R --> \infty and removing the \nu_R from the Lagrangian, our model becomes a multi-Higgs-doublet Standard Model with suppressed flavor-changing Yukawa couplings. Finally, we show that, with the usual assumptions about the mass scales in the seesaw mechanism, the branching ratios of all lepton-flavor-changing processes are several orders of magnitude smaller than present experimental limits.