A tale of serendipity: quasi-periodic oscillations in pulsating ultraluminous X-ray sources and a search for new candidate X-ray pulsators

In the past 10 years, the discovery of pulsating ultraluminous X-ray sources (PULXs) has revealed that accreting neutron stars (NSs) can shine at extreme luminosities, well above their Eddington limit. This finding has caused a shift in the ULX paradigm. Previously thought to consist solely of intermediate-mass black holes, the current view is that the ULX population is more heterogeneous, potentially dominated by NSs accreting at super-Eddington rates, posing significant challenges to our understanding of the physics of accretion onto compact objects. Many questions about this class of sources remain open: Are the extreme luminosities of these NSs really super-Eddington, or is geometrical beaming playing a significant role? If they are indeed accreting at super-Eddington rates, how can they sustain this regime over extended periods of time? What is the fraction of ULXs powered by accreting NSs? Are they an exception, or do they represent a common X-ray binary phase? Each new (observational and theoretical) insight into their complex phenomenology can bring us closer to a deeper understanding of these fascinating sources. Given their rarity (only 6 PULXs are known, 11 taking into account transient PULXs too, out of almost 2000 ULXs), every new identified PULX can carry vital information on the whole population. A possible way to achieve this goal is by searching the archives of X-ray missions with good imaging and timing capabilities. As these archives become larger and larger, a data mining approach, with a systematic and automated search for sources showing coherent pulsations, provides the most efficient method to perform such an analysis in a reasonable time. NGC7793 P13 and NGC5907 ULX-1 are two examples of PULXs discovered thanks to data mining projects in the XMM-Newton archive. In the process, serendipitous sources (not only PULXs) are always behind the corner, sources that otherwise would have been lost in the archives. Both the study of PULXs and the search for new X-ray pulsators through data mining have defined the course of my PhD and are the subject of this thesis. In Chapter 1, I first introduce the class of compact objects and the basics of accretion physics, and then describe the main properties and the open questions about (P)ULXs. In Chapter 2, I report my discovery of mHz-QPOs in the flux of two PULXs at super-Eddington luminosities. These mHz-QPOs could represent a signature of super-Eddington accretion and a turning point in our comprehension of PULXs. I describe the data mining project thanks to which NGC7793 P13 pulsations have been discovered in Chapter 3. I am currently involved in the continuation of this project aimed at searching for previously unidentified X-ray pulsators in the newly published XMM-Newton observations. During one of these searches, I discovered a new pulsar in the Large Magellanic Cloud. This source is likely a new candidate magnetar, only the third known outside our Galaxy. I report the details of the discovery in Chapter 4. Finally, I summarise and draw the conclusions of the work of these three years in Chapter 5, where I also outline additional works in progress and possible new directions for my research. A few examples of these and other works I have been involved in are further discussed in Appendix A and Appendix B.