Early formation of planetary building blocks in the Solar System

Planets form in disks surrounding young stars, called protoplanetary or circumstellar disks. Planet formation models usually start with the protoplanetary disk fully formed (corresponding to the so-called Class II disks). We studied whether planet formation can already begin during the star+disk formation process, in the Class 0 and Class I stages.

We focused on modeling a Solar-type star. We connected models of protoplanetary disk formation, dust growth, planetesimal formation, and their internal evolution. Planetesimals are the first gravitationally-bound building blocks of planets, the precursors of asteroids and comets that we still observe in the Solar System. In our models, the first planetesimals form when the material is still being deposited onto the disk from the surrounding environment. Interestingly, our results are consistent with constraints derived from the meteoritic evidence. Our results were published in the Science paper (Lichtenberg, Drążkowska, et al. 2021).

We found a scenario of the early evolution of the solar disk that fits the constraints from accretion chronology, thermochemistry, and the mass divergence of the inner and outer Solar System. 

In our model, planetesimals form along the water snow line, which is first moving outwards as the protoplanetary disk becomes denser and hotter. Some planetesimals already form during the disk infall (Class I) stage, which we call Reservoir I planetesimals. Much more planetesimals form later, during the nominal protoplanetary disk (Class II) stage, which we call Reservoir II. 

All the planetesimals form outside of the water snow line, which means that they are initially water-rich. The Reservoir I planetesimals inherit a significant fraction of short-lived radionuclides, in particular Aluminium-26, in which the solar nebula was enriched. The decay of Aluminium-26 is the main source of internal heating. The left panel below shows the results of our models of the internal structure of planetesimals in terms of the internal temperature the planetesimals reach depending on their size and formation time. The streaming instability leads to top-heavy mass distribution, with most of the mass deposited in large planetesimals (~100 km). If such heavy planetesimals form early, they significantly heat up which leads to melting of the water ice, hydrous rock decomposition, and evaporation of water as well as formation of an iron core. These processes are illustrated on the right panel below. 

We use the models of internal evolution of planetesimals to compare the metal-silicate separation times (left panel below) and the hydrothermal activity (right panel below) in the simulated planetesimal population to the meteorite record. We find that the planetesimals from Reservoir I undergo the core formation much earlier than the planetesimals in Reservoir II. The metal-silicate separation times in Reservoir I generally fit well with the meteorites from the so-called non-carbonaceous reservoir (NC), while the Reservoir II planetesimals correspond to the meteorites from the carbonaceous reservoir (CC). 

Simulated planetesimals in Reservoir I experience a brief phase of hydrothermal activity between 0.25 and 0.7 Myr after CAI formation and then dehydrate rapidly at ~0.7 Myr. After that initial peak of water ice melting and hydrothermal activity, less than 1 vol% of water ice is retained. Reservoir II planetesimals, which form later with less Aluminum-26, experience protracted hydrothermal activity lasting for several Myr with a peak at 5 Myr after CAI formation. The Reservoir II planetesimals retained ~15 vol% water ice and ~75 vol% hydrous rock. The peak for hydrothermal activity in Reservoir II reproduces the clustering of aqueous alteration in the CC meteorite record. The single available age for the NC population (an ordinary chondrite, OC) does not coincide with the peak in the Reservoir I population, but this can be explained by most of the hydrous phases formed in the Reservoir I peak being subsequently destroyed by high-temperature internal processing.

All the meteorites we know fall into one of the two classes: the NC or the CC. As described above, the two-stage planetesimal formation scenario we found allows us to explain the difference in iron core formation ages and the content of water observed in these two classes. We find another confirmation of our scenario in the isotopic dichotomy found between the NC and CC meteorites (Warren 2011). Planetary materials from the NC and CC groups show a distinct split in the abundances of isotopes that are formed in stellar nucleosynthesis processes (see the left panel below). In our model, the two planetesimals reservoirs were formed at different times and orbital locations. During later disk stages, material from the outer disk parts was incorporated into outer Solar System planets but did not substantially contribute to the inner terrestrial planets, as the mixing was restricted by the formation of massive Reservoir II (see the right panel below). Our simulations of planetary migration during the disk phase indicate that migration of planetary cores formed from Reservoir I planetesimals would keep them separated from the material from outside of the water snow line (see the terrestrial planets migration paths indicated in the planetesimal formation history figure above). 

In the longer run, our scenario also explains why the Solar System has two quite different populations of planets: the terrestrial planets, Mercury, Venus, Earth, and Mars, which are all relatively small and dry, and the volatile-rich giant planets, Jupiter, Saturn, Uranus, Neptune. In our models, this split arises due to distinct paths of internal evolution related to the different amounts of radiogenic heating in the two reservoirs of pre-planetary bodies. The terrestrial planets are dry not because they formed closer to the Sun, but because their planetary building blocks formed earlier. Although the cores of giant planets started to form later, their growth was greatly accelerated by the high amount of planetesimals and pebbles available in the outer parts of the solar disk.

We used the simple rotating infalling molecular cloud model (Hueso & Guillot 2005). In this model, there are three components: the molecular cloud, the central star, and the circumstellar disk. We assume that the molecular cloud is initially spherically symmetric, homogenous, isothermal, and it has a solid rotation. A single central star forms, which is surrounded by an axisymmetric circumstellar disk. This model leads to an inside-out disk buildup.

Material infalling on the disk from the molecular cloud consists of 99% hydrogen/helium gas and 1% heavier elements, which condense as dust grains. In the disk, dust coagulation timescales become short enough for the dust to grow to pebble-sizes. We model dust coagulation using a simplified model similar to the one proposed by Birnstiel et al. (2012). We take into account the evolution of the water ice component of dust grains by including evaporation and condensation of water. We assume that the ice-rich dust aggregates are more sticky than the dry aggregates and thus grow to larger sizes. This leads to the traffic jam effect as the large aggregates outside of the water snow line drift faster than the small aggregates in the inner part of the disk (see panel b) of the figure on the right). What is more, the water vapor diffuses across the snow line and re-condenses on dust aggregates, which is called the cold finger effect (illustrated by panel c). Both the effects act together to increase the solids-to-gas ratio at the water snow line (Drążkowska & Alibert 2017).

Dust growth is hindered by fragmentation and radial drift barriers, making it virtually impossible to reach meter-sizes. We include planetesimal formation in the process of streaming instability (Johansen et al. 2007). The streaming instability is concentrating pebbles into dense filaments, some of which become gravitationally unstable and collapse to planetesimals with typical size of 100 km. Planetesimal formation via the streaming instability requires prior enhancement of solids-to-gas ratio. In our models, planetesimals form at the water snow line. In the early phases, during disk infall, the cold finger is the dominant effect enhancing the solids-to-gas ratio because it acts faster but it is less efficient and thus not so many planetesimals form comparing to Class II disk phase, when the traffic jam effect is dominant. Further details of this planetesimal formation scenario are described in Drążkowska & Dullemond (2018).

Once planetesimals are formed, we follow their internal evolution, which is mostly determined by the planetesimal size and its formation time (Lichtenberg et al. 2019). We take into account that the solar nebula was initially enriched with the short-lived radionuclides, in particular Aluminium-26. Aluminum-26 has a half-time of 0.7 Myrs, comparable to the timescale of the circumstellar disk formation. The formation time of planetesimals thus defines the amount of radiogenic heating of their interiors from Aluminium-26 decay. We performed a grid of models taking into account the various planetesimal sizes and initial abundance of Aluminium-26. We followed how the internal evolution of planetesimals lead to their dehydration and iron core formation. Then, we compared the results of our models to the evidence from the meteoritic record of the Solar System.

• The architecture of the Solar System as we know it today may be intimately connected to the earliest phases of the solar circumstellar disk existence. How the material is delivered to the disk may play a critical role in creating the observed diversity of exoplanetary systems.
  
  
• The initial inventory of radioactive isotopes in the molecular cloud and forming circumstellar disk is a decisive parameter to determine the properties of the planetary system as it drives the geophysical evolution of the protoplanets.
  
  
• The planetesimal formation is not a single burst as it was assumed in the classical theory of planet formation, but a continuous process spanning the whole lifetime of the protoplanetary disk. Future models should aim at connecting the star and disk formation and planet formation into one comprehensive framework. 
  
  
• Check the Science paper (Lichtenberg, Drążkowska, et al. 2021) for all the details of our model 😉