From circumstellar disks to planetary systemsobservation and modeling of protoplanetary disks

Resumen

The existence of exoplanetary systems was first predicted after the discovery of accretion disks around young stars. Nowadays, with nearly 3500 exoplanets discovered, and almost 5000 more candidates identified by the Kepler space mission, planetary systems are now known to be ubiquitous around low-mass stars. The formation of these systems takes place during the stellar formation itself, from the dust and gas orbiting around the star in the protoplanetary disks. However, the process that leads to this formation is still not well understood. Studying the physical and chemical conditions of circumstellar disks is, thus, key to understand the planetary formation process. Planets can interact with the disk, creating structures such as spiral density waves, gaps, cavities or lopsided asymmetries. Studying disks showing these features could provide us with critical information about the planetary formation process itself. In particular, transitional disks, which are protoplanetary disks with central gaps or cavities in the dust distribution, appear as excellent candidates to study the first stages of planetary formation. Cavities in transitional disks were first identified through modeling of their spectral energy distributions (SEDs). This modeling also lead to the discovery of a subfamily of transitional disks, the so-called pre-transitional disks, which are thought to present a residual inner disk inside the cavity. (Sub-)mm and polarimetric IR observations have been able to image several of these disks and confirm the presence of the central cavities or gaps. Several mechanisms have been proposed to explain the clearing of these dust cavities: grain growth could reduce the emissivity of the dust grains, photoevaporative winds due to the high energy radiation from the star could remove material from the inner regions of the disk, and dynamical interactions with low-mass companions or planets could clear a gap or cavity. Even though some transitional disks have cavities that might be consistent with a photoevaporation origin, observations seem to indicate that most cavities in transitional disks are created by dynamical interactions with orbiting substellar or planetary companions. Nevertheless, the sample of observed transitional disks is not complete, and it is possible that, due to an observational bias, most of the transitional disks that have been studied belong to a family of denser disks that can form giant gas planets responsible for opening their cavities. More observations are needed to completely understand the impact of all the physical processes that take place during the last stages of the evolution of protoplanetary disks. Motivation and methodology: The main goal of this thesis is to study the initial conditions and overall process that leads to the formation of a planetary system. In order to do this we have observed three different transitional disks (HD 169142, XZ Tau B, and GM Aur) at mm and cm wavelengths with very high sensitivity and angular resolution. At mm wavelengths we have mainly studied the dust thermal emission of the disks. The emission at these long wavelengths is dominated by the large dust grains in the disk. Therefore, these observations have allowed us to study the locations in the disks where large dust grains may be trapped and, therefore, grow up to planetesimal sizes. On the other hand, the observations at cm wavelengths can trace the combination of dust thermal emission and thermal free-free emission from ionized gas. This ionization can be produced by high energy radiation from the star or by shocks in an accretion-driven jet. Thus, these observations have allowed us to study the impact of accretion-driven collimated jets and photoevaporative winds in the observed transitional disks. Furthermore, in order to obtain additional properties of the disks we have modeled the dust emission of the disk using selfconsistent radiative transfer codes. These codes calculate the temperature and density distributions of irradiated accretion disks, taking into account the settling of large dust grains, the viscous heating and the irradiation from the star. Finally, the codes can produce the expected SED of the disk and the directly irradiated walls of the gaps, as well as images at specific wavelengths. Slight modifications have been performed to the code to improve the calculation of the emission of the wall at mm wavelengths, where the disk can be optically thin, and to improve the calculation of the disk properties at the innermost regions of the disk. Results: In the first place, we present a study of the disk around the intermediate-mass star HD 169142. We report VLA observations at 7 mm, 9 mm, and 3 cm that trace the thermal emission of large dust grains (mm-cm sized) in the protoplanetary disk and the free-free emission of ionized gas. Our images have revealed the presence of a system of at least three gaps in the disk of dust, each ~0.20'' in width (~30 au at 145 pc), with outer radii of ~0.20'', ~0.48'', and ~0.83'' (~30,~70, and ~120 au, respectively). A bright and narrow clumpy ring of enhanced emission is observed at a radius of ~32 au, with a width of ~15 au. We interpret this ring to be tracing the rim of the inner gap. Unlike other sources, the radii of this structure of rings and gaps detected at 7 and 9 mm coincide with those obtained from previous near-infrared polarimetric images, which trace scattered light from small (micron sized) dust grains. We model the broad-band spectral energy distribution and the 7 mm images to constrain the disk physical structure. From this modeling we infer the presence of a small residual disk (~0.6 au in radius) inside the central cavity, indicating that the HD 169142 disk is a pre-transitional disk. The emission ring of the HD 169142 disk presents a high degree of substructure, which could be produced by dynamical interactions between the disk and multiple forming planets. We speculate that the ring could have become gravitationally unstable due to a pile-up of material produced by this dynamical interaction. As a consequence, the ring might be fragmenting into clumps that would end up forming new planets. Based on our observational and modeling results, we suggest that the two closest and most prominent gaps in the disk of HD 169142 could be created by dynamical interactions between the disk and forming planets, while the outer and shallower gap might be associated with dust growth and trapping close to the CO molecule snowline. Our highest angular resolution observations toward HD 169142 have also allowed us to detect a component of emission at 7 mm and 9 mm inside the innermost gap, with its peak of emission located at a projected distance of ~4 au toward the west from the central star. Its flux density and spectral index indicate that most of its emission is produced by ionized gas, which could be associated with an inhomogeneous photoionization of the inner disk, with an independent object, or with an (asymmetric) ionized jet. We favor the latter scenario and speculate that this jet might be significantly photoionized by the high energy radiation emitted from the central star. On the other hand, we report the discovery of a dwarf protoplanetary disk around the star XZ Tau B that shows all the features of a classical transitional disk but on a much smaller scale. The disk has been imaged with ALMA with extremely high angular resolution at mm wavelengths. While classical transitional disks have radii of~100 au, the disk around XZ Tau B presents a radius of only ~3.4 au, with a central cavity of ~1.3 au in radius. We have modeled the dust emission of the disk, showing that the mass of the disk could be large enough to form a compact planetary system like the ones detected in the exoplanet surveys. Due to the very small size of the disk, the dynamical evolution of the system is expected to be much faster than in classical larger disks. This opens up the possibility of monitoring the evolution of the system in very short (a few months) timescales. Finally, we performed multi-configuration VLA observations at 7 mm, 3 cm, and 5 cm toward the transitional disk of GM Aur. Our radio continuum observations have allowed us to image and spatially resolve, for the first time, the three main components at work in this stage of the disk evolution: the disk of dust, the ionized radio jet perpendicular to it, and the photoevaporative wind arising from the disk. The mass loss rate inferred from the flux density of the radio jet at 3 cm is consistent with the ratio between ejection and accretion rates typically found in younger objects (~10%), suggesting that transitional disks can power collimated ejections of material apparently following the same physical mechanisms as much younger protostars. Our results indicate that extreme-UV (EUV; 13.6-100 eV) radiation is the main ionizing mechanism of the photoevaporative wind traced by the free-free emission observed at cm wavelengths. The EUV photon luminosity inferred from this free-free emission is too low to produce the mass loss rates needed to disperse the disk in the timescales imposed by observations (<10 Myr). Therefore, we conclude that other additional mechanisms, such as X-ray-driven photoevaporation (which can launch a denser but only partially ionized wind), are required.