Proton Stability in Leptoquark Models

We show that in generic leptoquark (LQ) extensions of the standard model lepton and baryon numbers are broken at the level of renormalizable operators. In particular, this may cause fast proton decay unless the leptoquarks are heavy enough. We derive stringent bounds for the 1st generation LQ masses and couplings from the proton stability constraints. Leptoquark models [1, 2] remain an attractive possibility for new physics beyond the Standard Model (SM) admitting non-SM particles, leptoquarks (LQ), with masses at the electroweak scale. LQs are vector or scalar particles carrying both lepton and baryon numbers and, therefore, having a distinct experimental signature. For these reasons searching for LQs is a promising subject for present and near future experiments [3, 4]. Theoretical motivation for LQ models usually refers to the low energy limit of some more fundamental theory associated with an energy scale much higher than the electroweak scale. In the literature it has been argued that the origin of LQ models may reside in grand unified theories (GUT) [6], [7], models of extended technicolour [8]-[9], composite models [10] and some other high-energy scale model. However the arguments within this framework in favor of light LQs with masses of the order of the electroweak scale are quite vague. Nevertheless the generic structure of LQ models can be completely determined by the symmetry with respect to the SM gauge group [2, 11]. In this way LQ models with light LQs can be studied without referring to their high energy scale origin. Adopting this approach we examine the question of Lepton (L) and Baryon (B) number conservation in LQ models. We show that in the LQ Lagrangian there exist renormalizable ∆L = 1 and ∆B = 1 terms which may affect the conventional LQ phenomenology. In particular, the previously overlooked in the literature ∆B = 1 terms may cause fast proton decay unless the LQs are heavy enough. Thus the proton stability constraint casts lower limits on the LQ masses. We show that these limits are more stringent than those existing in the literature [3, 4]. e-mail: [email protected] e-mail: [email protected]