1. The accretion process in neutron-star low-mass X-ray binaries
     2. Discovery of the tidal disruption event candidate 2XMMi J184725.1-631724

1. The accretion process in neutron-star low-mass X-ray binaries
Lin, D., Remillard, R. A., & Homan, J. 2007, ApJ, 667, 1073:
          Evaluating Spectral Models and the X-ray States of Neutron-Star X-ray Transients
Lin, D., Remillard, R. A., & Homan, J. 2009, ApJ, 696, 1257:
          Spectral States of XTE J1701-462: Link between Z and Atoll Sources
Lin, D., Remillard, R. A., & Homan, J. 2010, ApJ, 719, 1350:
          Suzaku and BeppoSAX X-ray Spectra of the Persistently Accreting Neutron-Star Binary 4U 1705-44
Lin, D., Remillard, R. A., Homan, J., & Barret, D. 2012, ApJ, 756, 34:
          The Spectral Evolution along the Z track of the Bright Neutron Star X-ray Binary GX 17+2

One of my main research areas is the accretion process in low-mass X-ray binaries (LMXBs), in which a neutron star (NS) or a stellar-mass black hole (BH) accretes matter from a Roche-lobe filling low-mass companion star through an accretion disk (Figure 1.1). Intense X-ray emission is released from the inner accretion disk and/or the boundary layer produced by the impact of the accretion flow on the NS surface. LMXBs are unique targets for study of strong gravity and dense matter. Spectral modeling of NS LMXBs is critical for understanding of the spectral evolution and physical connection of various types of NS LMXBs (mainly atoll and Z sources) but had been impeded for decades by the problem of model degeneracy (i.e., many models can fit the spectra statistically equally well, ref). X-ray spectral studies of BH X-ray binaries have been relatively successful, e.g., spin parameters have been derived from relativistic accretion disk models for several BHs, and the spectral modeling difficulty for NS LMXBs is probably due to their additional emission component from the boundary layer. Since my PhD program, I have developed a new way to model the X-ray spectra of accreting NSs (Figure 1.2), found the connections between subtypes (Figures 1.4 and 1.5), provided new interpretations of their X-ray spectral evolution (Figures 1.3 and 1.6), and obtained hints on the origins of some quasi-periodic oscillations (QPOs).

Fig 1.1: An illustration of a LMXB

Fig 1.2: Sample spectral fits to broad-band spectra of 4U 1705-44 with our model (ref). The soft state is fitted with a model consisting of a thermal multicolor disk (MCD), a single-temperature blackbody (BB, used to describe the boundary layer) and a weak powerlaw (PL). The hard state is dominated by a cutoff powerlaw (CPL) plus a small BB.

Fig 1.3: Spectral fit results Aql X-1 using our model(ref). The blue squares, green triangles, and red crosses are for the hard, transitional, soft states, respectively.

Fig 1.4: The light curves of XTE J1701-462 in its 2006-2007 outburst(ref). See Figure 1.5 for the meaning of each symbol.

Fig 1.5: The hardness-intensity diagram of XTE J1701-462 in its 2006-2007 outburst(ref)

Fig 1.6: The spectral fit results of XTE J1701-462 (only the atoll stage and lower vertices are shown, ref). See Figure 1.5 for the meaning of each symbol.

2. Discovery of the tidal disruption event candidate 2XMMi J184725.1-631724
Lin, D., Carrasco, E. R., Grupe, D., Webb, N., Barret, D., Farrell, S. A. 2011, ApJ, 738, 52
Press Release

When unlucky stars get close to a SMBH, they can be tidally disrupted and subsequently accreted before they realize this is the end of their world. Such events are called tidal disruption events (TDEs). They provide a unique way to find and study dormant SMBHs believed to reside in the center of most massive galaxies. People have been searching for them for decades, but only find about a dozen of candidates thus far.
  During my large project of source type classification of the 2XMM catalog, I discovered the TDE candidate 2XMMi J184725.1-631724. This source is coincident with the center of the galaxy IC 4765-f01-1504 at a redshift of 0.0353 (about 150 Mpc away). Its four X-ray observations are shown in Figure 2.1, and the corresponding luminosity curve in Figure 2.2. The transient behavior of the source is obvious. It reached a luminosity about 1044 erg/s, the Eddington luminosity of a 106 solar-mass BH.
   Its two XMM-Newton X-ray spectra, of unprecedented quality among candidate TDEs, are very soft and can be fitted with a model of a thermal multicolor disk as the dorminating component plus a weak powerlaw (Figure 2.3). The disk temperatures are low (kT~0.06-0.1 keV) but expected for accretion onto a SMBH. The disk luminosities from the two XMM-Newton observations are consistent to be proportional to kT4, expected if the accretion disk was truncated at the innermost stable circular orbit. Thus the behavior of the accretion disk here is very similar to the thermal state of BH X-ray binaries, but now just for accretion of a tidally disrupted star onto a SMBH. Large variability on timescales of hours are seen in the two XMM-Newton light curves (Figure 2.4). Based on the fits to spectra at different flux levels, we find that such large fast variability can be explained as due to fast variations of the mass accretion rate probably caused by shocks during violent tidal disruption of the star.
Fig. 2.1: The images of four X-ray observations of 2XMMi J184725.1-631724 (red square).

Fig. 2.2: The luminosity curve from the above four X-ray observations.

Fig. 2.3: The spectral fits of the two XMM-Newton X-ray spectra using a multicolor disk plus a power-law.

Fig. 2.4: The pn light curves of the two XMM-Newton observations.