Circumventing the challenges in the choice of the nonconformal coupling function in inflationary magnetogenesis

As is well known, in order to generate magnetic fields of observed amplitudes during inflation, the conformal invariance of the electromagnetic field has to be broken by coupling it either to the inflaton or to the scalar curvature. Couplings to scalar curvature pose certain challenges even in slow roll inflation and it seems desirable to consider couplings to the inflaton. It can be shown that, in slow roll inflation, to generate nearly scale invariant magnetic fields of adequate strengths, the nonconformal coupling to the inflaton has to be chosen specifically depending on the inflationary model at hand. In a recent work, we had found that, when there arise sharp departures from slow roll inflation leading to strong features in the scalar power spectra, there inevitably arise sharp features in the spectra of the electromagnetic fields, unless the nonconformal coupling functions are extremely fine tuned. In particular, we had found that, if there occurs an epoch of ultra slow roll inflation (that is often required either to lower scalar power on large scales or to enhance power on small scales), then the strength of the magnetic field over large scales can be severely suppressed. In this work, we examine whether these challenges can be circumvented in models of inflation involving two fields. We show that the presence of the additional scalar field allows us to construct coupling functions that lead to magnetic fields of required strengths even when there arise intermediate epochs of ultra slow roll inflation. However, we find that the features in the spectra of the magnetic fields that are induced due to the departures from slow roll inflation cannot be completely ironed out. We make use of the code MagCAMB to calculate the effects of the magnetic fields on the anisotropies in the cosmic microwave background and investigate if the spectra with features are broadly consistent with the current constraints.