eROSITA X-ray analysis of the PeVatron candidate Westerlund 1

Context. It is still unclear which fraction of cosmic rays above an energy of 1 PeV is accelerated by the observed Galactic PeVatron population. These sources of unknown physical origin are detected through their gamma-ray emission, which also identifies them as particle accelerators. However, their gamma-ray data are typically degenerate between hadronic and leptonic emission scenarios, which hinders their straightforward association with the mainly hadronic cosmic ray population. Aims. In this study, we aimed to distinguish between leptonic and hadronic particle acceleration scenarios for the PeVatron candidate HESS J1646-458, which is associated with the star cluster Westerlund 1 (Wd 1). To this end, we first studied the diffuse X-ray emission from Wd 1 to better understand if its origin is of thermal or nonthermal nature. In addition, we searched for X-ray synchrotron emission from the associated PeVatron candidate HESS J1646-458 to put new constraints on the magnetic field strength and the leptonic particle population of this source. Methods. We used data from the all-sky surveys 1 to 4 of the extended Roentgen Survey with an Imaging Telescope Array (eROSITA) on board the Spectrum-Roentgen-Gamma orbital platform to spectrally analyze the diffuse emission from Wd 1 and HESS J1646-458. For Wd 1, we fitted and compared a purely thermal model and a model with a thermal and a nonthermal component. Next, we analyzed the spectra of four annuli around Wd 1 that coincide with HESS J1646-458 to search for synchrotron radiation. Results. We find that eROSITA data cannot distinguish between thermal and nonthermal source scenarios for the diffuse emission from Wd 1 itself. For a thermal source scenario, the observed X-ray flux can be explained in large part by unresolved pre-main sequence stars or by thermalized stellar wind shocks. In the case of the PeVatron candidate HESS J1646-458, we find no evidence of synchrotron emission. We estimated an upper confidence bound of the synchrotron flux up to 40 ' around Wd 1 of 1.9 & sdot; 10(-3) keV(-1) cm(-2) s(-1). We used this result to study the spectral energy distribution of the source. From that, we obtained an upper 1 sigma confidence bound on the magnetic field strength of HESS J1646-458 of 7 mu G. Conclusions. Our upper bound on the magnetic field strength in HESS J1646-458 is compatible with a previous estimate in the literature for a fully leptonic source scenario. Therefore, a purely leptonic emission scenario is compatible with our results. The same is the case for hadronic and hybrid scenarios, for which even less synchrotron flux is expected compared to the leptonic scenario.