Simulating the spatial distributions of gas- and ice-phase molecules in galaxies: a new method and preliminary results

TL;DR

This work integrates interstellar chemistry with galaxy-scale simulations via a post-processing astrochemistry pipeline to predict gas- and ice-phase abundances for hundreds of species as functions of local ISM conditions, focusing on H2O, CO2, CH3OH ices in a Milky Way–like disk. Using a MW-type isolated disk with bar-driven structure, the study finds strong radial gradients in dust temperature and UV fields that drive ice-formation efficiencies, producing negative gradients and large dispersions in ice abundances, especially CO2 ice, while H2O ice remains relatively robust to $T_{\rm dust}$. The analysis shows that dispersions in $T_{\rm dust}$ and in gas-phase elemental abundances jointly amplify ice-abundance spreads, implying substantial cloud-to-cloud chemical diversity within a single galaxy and potential implications for planet formation and the Galactic Habitable Zone. The work highlights current limitations—ice feedback on dust growth, fixed dust size, and simplified CR treatment—and points to future directions including a fast astrochemical emulator to enable self-consistent, fully coupled galaxy-scale simulations with ice physics.

Abstract

Recent observations have revealed significant variations in the abundances of gas- and ice-phase molecules in galaxies with different luminosities and types. In order to discuss the physical origins of these variations, we incorporate gas- and dust-phase interstellar chemistry into galaxy-scale simulations with various baryonic physics including dust formation, evolution, and destruction, all of which are essential for the calculations of 400 interstellar molecule species. The new simulations can accordingly predict the abundances of gas- and ice-phase molecular species such as H_2O and CO_2 ice within individual molecular gas cloud of galaxies based on gas density and temperature, dust temperature (T_dust), elemental abundances (e.g., CHNOPS), UV radiation strength (F_UV), and cosmic ray ionisation rate (zeta_CR) within the clouds. Since this is the first of the series of papers, we describe the details of the new simulations and present the preliminary results focused on the spatial distributions of H_2O, CO, CO_2, and CH_3OH ice species in a disk galaxy similar to the Milky Way. We particularly discuss how T_dust and gas-phase elemental abundances can control the spatial distributions of the above molecules in galaxies. We briefly discuss the total amount of H_2O and CO_2 ices and radial distributions of PN and PO molecules in the Galaxy.

Figures (18)

• Figure 1: Illustration of key physical processes that are modelled in the present simulations with a new code. The journey of a dust grain within a galaxy is divided into the following environments depending on their physical properties: (i) dying stars such as AGB stars, core collapse supernovae (CCSNe), and novae (ii) diffuse interstellar medium, (iii) dense cores of molecular clouds, and (iv) forming stars and planets. In these environments, various dust formation and evolution processes can occur, including dust formation through nucleation within gaseous ejecta from CCSNe and AGB stars in (i), dust growth through accretion of metals in (ii), ice formation on the surface of dust grains in (iii), and complex organic matter formation due to UV irradiation from a baby star in (iv). Glycine is shown in (iv) as a possible example of organic matter formation, though it is yet unclear whether it is produced in ISM. Ice formation on dust can possibly suppress the dust growth, whereas dust growth can promote ice formation on dust surface. This mutual interaction between ice and dust is called "ice feedback" in this illustration. Tiny dust particles merge together to finally form a planet, and the complex organic matter can be delivered to the planet. The baby star finally dies away so that its matter can be again circulated to the ISM.
• Figure 2: Time evolution of the surface gas densities ($\Sigma_{\rm g}$) projected onto the $x$-$y$ plane in the fiducial model. The gas disk is divided into $100 \times 100$ mesh points (i.e., $3.5 \times 3.5$ kpc) to estimate $\Sigma_{\rm g}$ at each point.
• Figure 3: The same as Fig. 1 but for dust temperature, $T_{\rm dust}$.
• Figure 4: The same as Fig. 1 but for normalised UN radiation fields, $R_{\rm UV}$, where $R_{\rm UV}=F_{\rm UV}/F_{\rm UV,0}$.
• Figure 5: Radial profiles of $T_{\rm dust}$ (top), $\log R_{\rm UV}$ (second from the top), $f_{\rm dust}$ (second from the bottom), and $f_{\rm o,g}$ (bottom) in the fiducial model. The error bars at each radial bin represent $1\sigma$ dispersions of these physical properties. These properties for the central bulge ($R<2$ kpc) in the first bin are estimated using all particles within $R<2$ kpc.