Formation of stars is the principal driver of galaxy evolution and chemical enrichment in the Universe, which in turn affects the process of star birth. On the other hand, it also provides the sites of planet formation and development of life. Therefore, understanding how stars and planets form has long been one of the central problems in astrophysics. In fact, significant progress has been seen especially in the last decades, but the detailed processes are still quite elusive. ALMA, JWST, and next generation instruments on ground-based telescopes will yield further scientific advances prior to the TMT’s first light. However, we definitely need TMT to have unprecedented angular resolution with very high sensitivity in near- to mid-infrared, which is critical for this science topic. TMT, with the great complementarity with those telescopes and instruments, will provide a big step forward to more complete understanding of star and planet formation.
Star formation is a fundamental astrophysical process; and yet we still lack a quantitative and predictive theory for how stars and clusters form (Krumholz 2014). Amongst the key ingredients for building such a theory are observations of young star clusters over a wide range of environments that can be used as benchmarks to test star formation models. Young clusters serve as laboratories with controlled conditions (same age, metallicity, distance, initial conditions) and many stars to sample the distributions of stellar masses, kinematics, and multiplicities that arise during the star formation process. The requisite range of initial conditions in metallicity, external pressure, cloud mass/density, and environment can only be found outside of our own Galaxy where studying clusters in detail has been extremely challenging. Fortunately, TMT’s spatial resolution and sensitivity will allow individual young stars to be spatially resolved, even in the largest star clusters, within a wide range of environments throughout the Local Group galaxies and beyond. TMT observations of young clusters, which are the output of the star formation process, will complement ALMA studies of the gas clumps and cores that are the input to the star-formation process.
In the section below, we highlight TMT science cases for measuring the initial mass function (IMF), internal kinematics, and multiplicity in young clusters, followed by the sections focusing on high- and low-mass ends of the IMF. These observations will help constrain star formation theories such that reliable predictions can be incorporated into models of planet formation, stellar evolution, galaxy formation and evolution, and cosmology.
Sketch of different probes of molecules in disks. The mid-infrared predominantly traces the inner few AU, corresponding to the terrestrial planet-forming region.
Protoplanetary disks are flattened rotating structures of gas and dust, formed as a natural outcome of star formation process. They are the birth places of planets, and hence provide the initial condition of planet building including density and temperature distributions of disk material which are the ingredients for planets and possibly life. Disks also interact with newly-born planets, resulting in re-distribution of gas and dust or orbital migration of planets. After most gas dissipates, evolution of a planetary system is expected to further proceed, such as with violent orbital re-configuration and growth of rocky bodies in debris disks around young main-sequence stars (see Section 10.2 of the TMT DSC). Therefore, observational understanding of young disks is essential to address the questions of when, where, and how planets form and evolve?
What is most critical for disk observations is high angular resolution. It is required to spatially resolve the local physical and chemical conditions in a disk to finally constrain planet formation theories. In this respect, there is no doubt that TMT will play a major role in this science field, providing a resolution of ~1 AU at 1 µm for disks in the nearest star-forming regions at 140 pc. The high sensitivity of TMT allows us to significantly increase the sample size for resolved observations and to discuss any dependence in planet formation as a function of stellar properties. The list of disk-bearing objects has been growing thanks to Spitzer, AKARI, Hershel, and WISE (e.g., Evans et al. 2009), thus it is not hard at all to extract hundreds of targets for TMT.
The inner (鈮�10 AU) regions of protoplanetary disks are particularly interesting. Given the abundance of exoplanets, we know that protoplanets must be common in the 0.5—10 AU range (Cassan et al. 2012), while current direct detection surveys show that they are rare beyond ~30 AU (e.g., Chauvin et al. 2010). These inner regions also intersect the habitable zone of their parent stars. Such inner disks can be resolved either directly, or kinematically where the (nearly) Keplerian rotation of disks can be used to separate disk regions in velocity (and hence radius). Spectro-astrometry is another advanced, promising technique to probe spatial scales much smaller than the nominal spatial resolution. Given the expected temperatures of these disks, atomic and molecular lines of interest will lie between 1 and 25 µm. The classical infrared “finger-print” region contains transitions from many molecular species, including important bulk tracers, such as water, CO and CO2, as well as a potentially long list of organic species (HCN, C2H2, CH4, etc.).
Disk observations, including detection of footprints of planets, can be extensively done from the early phase of TMT without coronagraphy. As high contrast 2nd generation instruments become available on TMT, lots of detections of exoplanets will be expected. This may be the most compelling and exciting period since the connection between disks and planets can be established, which represents substantial progress toward the ultimate goal of understanding planet formation.
TMT will be highly complementary to JWST and ALMA. TMT and JWST will probe the inner disk region while ALMA more typically traces the outer disk. Because of its higher spatial and spectral resolution, TMT will be much superior to JWST in constraining the spatial origin of disk emission features and their dynamics.
In the following, we describe some intriguing science topics where a 30-m class telescope can uniquely contribute; conditions of planet formation, disk-planet dynamical interaction including growth of gaseous planets and circumplanetary disks, and distribution of ice, water, and organic material in young plant-forming disks.
Model images in H band for gaps and spirals in a disk caused by a 0.1 MJup planet around a solar-mass star, obtained with 2D hydro-dynamical simulations. Disk aspect ratio of 0.05, and local isothermal, hydrostatic balance was assumed for the vertical structure. The color scale denotes surface brightness multiplied by r2 (r is distance from central star). Top: The models at the raw resolution with a planet at r = 10 AU (left) and r = 30 AU (right). Middle: Images expected with an 8 m telescope. Bottom: Images expected with TMT, models were convolved with Gaussian functions of diffraction-limited FWHMs assuming a distance of 140 pc. Dark spots are "planet shadows", indicating the location of the planet (Jang-Condell 2009). Note: 3D thermal balance calculations are necessary to predict planet shadows; the sizes of dark spots are tentative in these images.
Planets can dynamically interact with their parent disk, leaving footprints which in most cases are more readily detected than planets themselves. In a gaseous protoplanetary disk, the simplest form of the footprints is a radial gap along the orbit of a giant planet. Once formed, the gap can excite instabilities due to the pressure gradient at its outer edge, resulting in dust traps which can be favorable locations of planetesimal formation and growth, providing the formation site of additional members of a young planetary system (e.g., Lyra et al. 2009). Dynamical processes can also excite small-scale structures such as spiral waves either in association with instabilities or companions even for those insufficiently massive to carve a gap. In fact, via direct imaging with 8-10 m class telescopes, gaps have been uncovered in transitional disks of T Tauri and Herbig Fe/Ae/Be stars, while spirals were detected preferentially toward warmer disks around earlier type stars in scattered light (e.g., Muto et al. 2012). Yet, with the current instruments, the primary focus is still onto the disk regions of several tens of AU. New high-contrast instruments are now being delivered, but if their extreme AO systems are not equipped with laser guide stars, most T Tauri stars, which are relatively faint at optical wavelengths, will not be observable.
Higher angular resolution and contrast, afforded by a 30 m class telescope with an advanced AO system, are the key to detect planetary signatures in the inner disk region and to resolve smaller-scale structures. For instance, the typical requirement for the detection of spirals is to resolve the spatial scale comparable to the disk vertical thickness. In the case of nearby star-forming regions at 140 pc, the resolution of about 0.01 arcsec is needed to distinguish tightly-wound spirals at 30 AU in a colder T Tauri disk than Herbig systems, with the aspect ratio (the ratio of the scale height to the radius) of ~0.1. This can be confirmed in the simulations of disk scattered light which demonstrate that TMT has an ability to reveal signatures of an embedded planet of 鈮�0.1 MJup at 鈮�10 AU from the central star at 140 pc. A gap and spirals by a planet at 10 AU can be detected in the mode of direct imaging, preferably with polarimetry. Structures within 10 AU can be studied with the smaller inner working angle provided by PFI or SEIT. Furthermore, the innermost region lying beneath the coronagraphic mask or the bright stellar halo in NIR can be explored in thermal emission in MIR. The diffraction-limited resolution in N band is ~0.07 arcsec for TMT, which allows us to study distributions of warm dust with a spatial scale of about 0.02 arcsec, corresponding to 3 AU at 140 pc.
It is worth noting that multi-epoch observations are quite useful to put a strong constraint on planet location through disk dynamics. A spiral arm caused by a planet in a disk is expected to co-rotate with the planet, resulting in the pattern rotation velocity slightly different from the local Keplerian velocity. With the spatial resolution of TMT, the rotation of the spiral can be detected by observations several years apart if a planet is orbiting at 30 AU around a solar-mass star.
Near- and mid-infrared observations are sensitive to the distribution of small, micron-sized grains at the surface of an optically thick protoplanetary disk, while (sub-)mm observations probe large, mm-sized grains in the mid-plane in the continuum as well as the gaseous component at various heights through line emission. The synergy of ALMA with TMT is thus of great importance to understand the 3D distribution of different components that make up the disk. Needless to say, such comprehensive understanding has a strong impact on theory of planet formation, and is critical to understanding the origin of the observed structures; a dust trap (Birnstiel et al. 2013) or a planet (Zhu et al. 2014).
Giant planets form during the gas-rich protoplanetary disk phase with the first few million years after the formation of the central star. Beyond a certain mass, the forming protoplanet will interact with its natal disk in multiple ways that may lead to direct or indirect detections of the planet itself. A giant planet may open a gap in the gas disk, inducing local non-Keplerian velocity fields with characteristic structure and time variability. This has been proposed to create observable spectroscopic signatures in the 4.7 µm rovibrational CO lines (e.g., Regály et al. 2014). The planet may form a relatively massive, moon-forming, circumplanetary disk (Lubow et al. 1999; Machida et al. 2008; Brittain et al. 2013), and it may even produce strong emission lines from accretion flows (Zhou et al. 2014). All of these tracers will be accessible to TMT. Indeed, they are already being pursued with existing 8-10 m class telescopes as well as with ALMA, and initial results are promising.
Therefore, while TMT may detect protoplanets through broad-band imaging of their thermal continuum emission, there is also the prospect of using kinematic gas tracers to detect the growth of giant protoplanets. TMT is able to detect gas tracers in the optical and infrared at very high spatial resolution; 40 mas for the four main CO isotopologues around 4.7 µm that trace the velocity field of the protoplanetary material, as well as a host of tracers of accretion onto the planet itself, including hydrogen recombination lines such as Brγ (2.16 µm).
It should be empathized that the detection of a gap-opening, accreting planet will have a strong impact on the theory of planet formation and migration. Measurements of the gas accretion rate onto a planet along with the depth and width of the gap are critical to know the angular momentum transfer rate through the gap, and therefore the final mass and type-II migration of the planet (Tanaka et al. 2002; Crida et al. 2006; Fung et al. 2014).
Planet and/or moon formation in action can be caught even after the gas dissipation. There have been recent detections of warm dust belts in a number of debris disks (e.g., Morales et al. 2011). This is the dust originating in the inner region at ~1—10 AU where the habitable zone resides, and can be interpreted to be an evidence for ongoing, oligarchic growth of terrestrial bodies. The high spatial resolution of TMT will be critical in characterizing this terrestrial material to gain new unique insight into the building blocks and mechanisms of terrestrial planet formation. In nearby debris disks, imaging from K to N band will be capable of resolving the 1 AU region. The precise structure such as azimuthal asymmetries and sharp boundaries of the disk edges can inform us of the presence of planets and help establish the connection between the dust belt and planet formation/evolution activity.
Scattered light image of the disk around the Herbig Fe star HD 142527 at 3.08 μm (left) and the scattered light spectra (right). The ~3.1 μm absorption feature, probably due to the scattering by water ice grains, is clearly seen in the spectra. As we go closer to the central star inside of the snow line, this absorption band will disappear due to the absence of water ice, which will directly show the snow line location at the disk surface.
The snow line is a condensation/sublimation front of (water) ice in a protoplanetary disk; water present in the form of vapor (gas) within the snow line and as ice (solid) beyond it. The formation of giant planet cores is believed to be enhanced because the solid surface density increases and the inward radial drift of solids may be slowed-down across the snow line (Brauer et al. 2008). It helps to explain why Jupiter-like gas giants are not the closest to the Sun in our Solar System. While the snow line is located at a radial distance of ~3 AU in the current Solar System, it may have been as close as ~0.7 AU in the early, optically thick phase (Garaud & Lin 2007), and therefore abundant water ice may have been present in the terrestrial planet region. Thus understanding the location of the snow line, at various evolutionary stages, is important given its relation to both planet formation and the origin of the water in terrestrial planets.
The snow line in protoplanetary disks can be predicted through observations of water vapor. Water emission lines in NIR and MIR can be used to determine the radial extent of water vapor in the warm disk atmosphere (e.g., Pontoppidan et al. 2010a). For instance, this approach was successful to show the snow line location of ~4 AU in the disk surrounding the nearby (54 pc) young star, TW Hya (Zhang et al. 2013). With the high-spectral resolution (R~100,000) of MICHI and MODHIS, the number of sources studied in this way can dramatically increase, see Figure 8.5, enabling us to investigate the evolution of snow line as a function of stellar age and mass.
The distribution of water ice can also be traced via spatially resolved spectroscopy of disk scattered light. Water ice has a strong absorption band at ~3.1 µm which should be imprinted in the scattered light spectra (Honda et al. 2009). Disk models predict that the surface snow line in protoplanetary disks around intermediate-mass young stars (Lstar = 10 L蕵) is located at ~20 AU (Oka et al. 2012), which corresponds to ~0.14 arcsec at 140 pc. This is expected to be within a reach of TMT without coronagraphy (IRIS, MODHIS), and the observations will surely be feasible with PFI owing to the much better inner working angle.
The outer region, well beyond the snow line, can be the formation site of icy planets. Comets can also form in such a cold region and could be ingredients of icy planets. Although comets are known to contain pristine frozen material, they also have incompatible, high-temperature product such as crystallized silicate grains, which infers dynamical mixing between the hot inner and the cold outer regions. It is still a matter of debate how the mixing occurs and how the high-temperature material is incorporated into comets. A key to understand such a process is obtaining the spatial distribution of various kinds of grains in young planet-forming disks. With the high-spatial resolution of TMT, MICHI will offer a unique opportunity to uncover both mineralogical evolution of solids and transport of material within a disk by spatially-resolved MIR spectroscopy (e.g., Okamoto et al. 2004).
