Supernova remnants interacting with molecular clouds are ideal laboratories to study the acceleration of particles at shock waves and their transport and interactions in the surrounding interstellar medium. In this paper, we focus on the supernova remnant W28, which over the years has been observed in all energy domains from radio waves to very-high-energy gamma rays. The bright gamma-ray emission detected from molecular clouds located in its vicinity revealed the presence of accelerated GeV and TeV particles in the region. An enhanced ionization rate has also been measured by means of millimetre observations, but such observations alone cannot tell us whether the enhancement is due to low energy (MeV) cosmic rays (either protons or electrons) or the X-ray photons emitted by the shocked gas. The goal of this study is to determine the origin of the enhanced ionization rate and to infer from multiwavelength observations the spectrum of cosmic rays accelerated at the supernova remnant shock in the unprecedented range spanning from MeV to multi-TeV particle energies. We developed a model to describe the transport of X-ray photons into the molecular cloud, and we fitted the radio, millimeter, and gamma-ray data to derive the spectrum of the radiating particles. The contribution from X-ray photons to the enhanced ionization rate is negligible, and therefore the ionization must be due to cosmic rays. Even though we cannot exclude a contribution to the ionization rate coming from cosmic ray electrons, we show that a scenario where cosmic ray protons explain both the gamma-ray flux and the enhanced ionization rate provides the most natural fit to multiwavelength data. This strongly suggests that the intensity of CR protons is enhanced in the region for particle energies in a very broad range covering almost 6 orders of magnitude: from $\lesssim 100$ MeV up to several tens of TeV.

Low energy cosmic rays are the major ionization agents of molecular clouds. However, it has been shown that, if the cosmic ray spectrum measured by Voyager 1 is representative of the whole Galaxy, the predicted ionization rate in diffuse clouds fails to reproduce data by 1-2 orders of magnitude, implying that an additional source of ionization must exist. One of the solutions proposed to explain this discrepancy is based on the existence of an unknown low energy (in the range 1 keV-1 MeV, not probed by Voyager) cosmic ray component, called carrot when first hypothesized by Reeves and collaborators in the seventies. Here we investigate the energetic required by such scenario. We show that the power needed to maintain such low energy component is comparable of even larger than that needed to explain the entire observed cosmic ray spectrum. Moreover, if the interstellar turbulent magnetic field has to sustain a carrot, through second-order Fermi acceleration, the required turbulence level would be definitely too large compared to the one expected at the scale resonant with such low energy particles. Our study basically rules out all the plausible sources of a cosmic ray carrot, thus making such hidden component unlikely to be an appealing and viable source of ionization in molecular clouds.

Cosmic rays are usually assumed to be the main ionization agent for the interior of molecular clouds where UV and X-ray photons cannot penetrate. Here we test this hypothesis by limiting ourselves to the case of diffuse clouds and assuming that the average cosmic ray spectrum inside the Galaxy is equal to the one at the position of the Sun as measured by Voyager 1 and AMS-02. To calculate the cosmic ray spectrum inside the clouds, we solve the one-dimensional transport equation taking into account advection, diffusion and energy losses. While outside the cloud particles diffuse, in its interior they are assumed to gyrate along magnetic field lines because ion-neutral friction is effective in damping all the magnetic turbulence. We show that ionization losses effectively reduce the CR flux in the cloud interior for energies below $\approx 100$ MeV, especially for electrons, in such a way that the ionization rate decreases by roughly 2 order of magnitude with respect to the case where losses are neglected. As a consequence, the predicted ionization rate is more than 10 times smaller than the one inferred from the detection of molecular lines. We discuss the implication of our finding in terms of spatial fluctuation of the Galactic cosmic ray spectra and possible additional sources of low energy cosmic rays.