The cosmic journey of dust grains -- from nucleation to planetary system

TL;DR

This study investigates how dust grains released from stars survive their journey through a multi-phase Milky Way–like ISM and become incorporated into new stars, with a focus on distinguishing contributions from AGB stars versus core-collapse supernovae (SNe) to presolar grains. Using RAMSES-based hydrodynamics for an isolated MW-like disc and millions of Lagrangian dust tracers, the authors model thermal sputtering in the ISM and test variations in SNR environments to assess dust destruction, astration timing, and the relative stellar sources of presolar grains, including the $^{26}$Al budget. They find that roughly half of grains lose less than $10\%$ of their initial mass, SN grains are more eroded than AGB grains by about $10\%$ on average, and the AGB/SN astration ratio approaches $\sim0.8$ after a few hundred Myr, implying a SN-dominated presolar-dust fraction of around $55\%$; the results also indicate that SN-origin grains carry most of the short-lived $^{26}$Al. The study highlights the importance of the SN bubble environment in the first Myr after explosion and discusses limitations due to resolution and neglected processes, while offering a framework to connect galactic dust evolution to the origin of presolar grains and planetary system formation.

Abstract

Dust is essential to the evolution of galaxies and drives the formation of planetary systems. The challenge of inferring the origin of different presolar dust grains from meteoritic samples motivates forward modelling to understand the contributions of low- and high-mass stars to dust in our Solar System. In this work we follow the evolution of dust with tracer particles within a hydrodynamical simulation of a Milky Way-like isolated disc galaxy. We find that nearly half of the grains released from stars lose less than $10\%$ of their initial mass due to thermal sputtering in the interstellar medium (ISM), with an average degree of atomisation $\sim$$10\%$ higher for dust grains released by supernovae relative to asymptotic giant branch (AGB) star grains. We show through supernova remnant model variations that supernova (SN) dust survival is primarily shaped by the supernova bubble environment in the first million years (Myr) after the explosion rather than by its evolution during $10^2-10^3$ Myr in the ISM. The AGB/SN ratio of dust grains incorporated into newly formed stars approaches $0.8$ after a few hundred Myr of galactic evolution. Our analysis also shows that star-forming particles with short ($<$$10$ Myr) free-floating time-scales in the ISM are predominantly released from supernovae rather than AGB stars. This implies that the Solar System budget of short-lived radioactive isotopes such as $^{26}$Al, whose decay contributed to melting and differentiating planetesimals, should have been provided by massive stars with masses $M \gtrsim 8$ M$_{\odot}$.

Figures (8)

• Figure 1: Top-row: Projected gas density (left), temperature (centre), and star tracer density (right) of the simulated galaxy seen face-on at a time of $214.2 \, \mathrm{Myr}$. Hot, under-dense pockets are carved out by stellar feedback and SN explosions. The tracers attached to star particles sample the population of star particles in the simulation. Not all star particles host tracers, but the top right panel shows tracers which are in a star state in this simulation snapshot and may be released (i.e. transition to gas tracers) later on in the simulation through feedback processes. Bottom-row: number density (left), temperature (centre) and position (right) evolution for two tracer particles as a function of time $t$ in the simulation during $450 \, \mathrm{Myr}$ of galactic evolution. These illustrate that even though the particles travel on nearly circular orbits around the galaxy, they are exposed to abrupt changes in phases and traverse both dense and cold regions of the ISM, as well as the hot and under-dense ionized medium. The various ISM phases are shown as density and temperature intervals for reference. The cold neutral medium is associated with typical densities of $n_{\text{H}} \gtrsim 10 \, \mathrm{cm^{-3}}$, but our label "Dense" refers to star-forming regions (see Section \ref{['sec:SF']}).
• Figure 2: Amount of time $\tau$ between release from a star and new astration for dust particles in each category of atomisation fraction $f_{\text{a}}$. The histograms are binned in intervals of $10 \, \mathrm{Myr}$ and are normalized to the total number of astrating particles $N_{\star}$ (including all atomisation fraction categories). The solid lines represent the cumulative distributions and are also normalized to $N_{\star}$. The median free-floating time $\tau$ within each category is shown as a dashed blue line.
• Figure 3: Dust evolution as a function of time after release from a star $\Delta t$, for dust grains losing less than $1\%$ ($1^{\text{st}}$ column), between $1\%$ and $10\%$ ($2^{\text{nd}}$ column), between $10\%$ and $50\%$ ($3^{\text{rd}}$ column) and between $50\%$ and $100\%$ ($4^{\text{th}}$ column) of their initial mass by the end of the simulation or prior to being incorporated into a star again. The first two rows depict the temperature $T$ and the number density $n_{\text{H}}$ of the ambient gas. The two bottom rows represent grain properties: the absolute value of the sputtering rate ${\text{d}}a/{\text{d}}t$, and the grain size reduction $\Delta a = a_0 - a(\Delta t)$ relative to the initial grain size $a_0$ after release. As we do not consider grain growth, all grains can only decrease in size and thus $|{\text{d}}a/{\text{d}}t| = - {\text{d}}a/{\text{d}}t$. The two-dimensional histograms are represented on a $250 \times 250$ grid, with $N$ the number of particles per bin. Marginal distributions are shown in the side panels to the right of each main panel, and are normalized such that the area under the curve is of unity.
• Figure 4: Evolution of the mean atomisation fractions of AGB (blue) and SN (red) particles over the duration of the simulation. The delay in the release of AGB dust particles, and thus also their atomisation, is due to the relatively slower evolution of low-mass stars. The mean values of $f_{\text{a}}$ remaining well below $50\%$ for both AGB and SN particles suggests that the grains preserve a significant proportion of their initial mass, on average, throughout the simulation.
• Figure 5: Cumulative distribution of mass lost by AGB and SN dust grains before undergoing astration. The left and right panels represent the same data but on logarithmic and linear scales, respectively. The fiducial models are represented by the solid lines, whereas the dashed/dotted lines display the model variations for the SNR evolution from Table \ref{['tab:models_table']}. For readability, we do not to show the CM$_0$ and CM$_2$ lines as they are nearly identical and overlap very closely with with CM$_1$. The non-differentiable regions of the HM curves result from the transition in environment between the additional heating introduced in post-processing and the nominal temperature and density conditions encountered by SN particles in the simulation.