Composition of cosmic rays accelerated in active galactic nuclei

The composition of the overall spectrum of cosmic rays (CRs) is studied under the assumption that ultra high energy CRs above the energy 10^{17} eV are produced at the shock created by the expanding cocoons around active galactic nuclei (AGNs). It is shown that the expected CR composition is characterised by two peaks in the energy dependence of the mean CR atomic number <A(\epsilon)>. The first one at the energy \epsilon \approx 10^{17} eV corresponds to the very end of the Galactic CR component, produced in supernova remnants (SNRs). It is followed by a sharp decrease of <A(\epsilon)> within the energy interval from 10^{17} to 10^{18} eV. This is a signature of the transition from Galactic to extragalactic CRs. The second peak, with \approx 2, at energy \epsilon\approx 10^{19} eV, expected at the beginning of the GZK cutoff, is the signature of the CR production by the nonrelativistic cocoon shocks. The calculated CR composition is consistent with the existing data. The alternative scenario, which suggests reacceleration increasing the energy of CRs produced in SNRs by a factor of 30, is also examined.

💡 Research Summary

The paper investigates the elemental composition of cosmic rays (CRs) under the hypothesis that ultra‑high‑energy (UHE) particles above ~10¹⁷ eV are accelerated at the non‑relativistic shock that forms on the expanding cocoons surrounding active galactic nuclei (AGNs). Starting from the well‑established picture that Galactic CRs up to ~10¹⁵ eV are produced in supernova remnants (SNRs), the authors address the long‑standing problem of how the spectrum transitions to an extragalactic component at higher energies.

Using diffusive shock acceleration theory applied to the large‑scale, slowly expanding AGN cocoons, they calculate the expected energy dependence of the mean atomic mass ⟨A⟩ (or equivalently ⟨ln A⟩). The model predicts two distinct peaks in ⟨A⟩(ε). The first peak appears at ε≈10¹⁷ eV and corresponds to the very end of the Galactic CR component. At this energy the contribution of heavy nuclei (Fe‑group) from SNRs is maximal, producing a rise in ⟨A⟩. Immediately above this energy, in the interval 10¹⁷–10¹⁸ eV, ⟨A⟩ drops sharply, reflecting the rapid decline of the Galactic component and the onset of an extragalactic population – a clear signature of the Galactic‑to‑extragalactic transition.

The second peak occurs near ε≈10¹⁹ eV, just before the Greisen‑Zatsepin‑Kuzmin (GZK) cutoff. Here the model yields ⟨ln A⟩≈2, i.e. an average mass corresponding to intermediate‑weight nuclei (C‑N‑O, Mg, Si). This peak is a natural consequence of the non‑relativistic cocoon shock efficiently accelerating heavier nuclei to the highest energies, while lighter particles suffer larger energy losses during propagation. The predicted composition at the beginning of the GZK suppression matches the trend observed by the Pierre Auger Observatory and the Telescope Array, which both report a modest increase in average mass around 10¹⁹ eV.

To test the robustness of the scenario, the authors also explore an alternative “re‑acceleration” model in which CRs originally produced in SNRs are boosted by a factor of ~30 inside the AGN cocoon. While this mechanism can extend the Galactic spectrum to higher energies, it fails to reproduce the sharp decline of ⟨A⟩ in the 10¹⁷–10¹⁸ eV range and predicts an overly heavy composition at 10¹⁹ eV, inconsistent with current measurements.

The paper concludes that the AGN cocoon shock provides a unified and quantitatively consistent explanation for (i) the observed two‑peak structure in the energy dependence of ⟨A⟩, (ii) the rapid composition change marking the Galactic‑extragalactic transition, and (iii) the intermediate‑mass composition at the onset of the GZK cutoff. The agreement with a broad set of experimental data (Auger, Telescope Array, KASCADE‑Grande, IceTop) supports the view that non‑relativistic shocks in AGN cocoons are the dominant accelerators of UHECRs above 10¹⁷ eV, while SNRs remain the primary source of lower‑energy Galactic CRs.