Measuring Dust Masses of Protoplanetary Disks in Serpens and L1641/L1647 with ALMA

TL;DR

This work leverages ANN-assisted SED modeling, grounded in the DIAD disk framework, to derive dust masses for 178 protoplanetary disks in Serpens and L1641/L1647, using ALMA 1.3 mm data and multiwavelength photometry. It reveals that many disks are optically thick at 1.3 mm, causing traditional flux-based methods to underestimate $M_{dust}$ by about a factor of 2 on average. The analysis shows that dust masses in these young regions exceed those in older regions and follow an age-related decline roughly as $M_{dust} \propto t^{-1}$, but even the increased SED-derived masses do not fully reconcile with exoplanet mass budgets, highlighting the persistent missing-mass problem. The study emphasizes the importance of observing at longer millimeter wavelengths (e.g., 7 mm / ALMA Band 1) to attain optically thin constraints and more accurate disk mass measurements, which are crucial for understanding disk evolution and planet formation timescales.

Abstract

Protoplanetary disks are an essential component of the planet-formation process. The amount of dust and gas in the disk constrains the number and size of planets that can form in a system. We analyze 178 T-Tauri stars, 18 in Serpens and 160 in L1641/L1647, and measure their disk dust masses using spectral energy distribution (SED) modeling and multiwavelength data, including 1.3 mm (ALMA band 6) fluxes from the literature. The disk masses calculated in this work are up to $\sim$2 times higher than those previously reported. We conclude that this is because disks may be partially optically thick at millimeter wavelengths while most calculations of the disk mass assume that the disk is optically thin at 1.3 mm. We calculate optical depths at 1.3 and 7 mm for a subset of the Serpens and L1641/L1647 disk sample and show that the vast majority of disks become optically thin at longer millimeter wavelengths; thus, observations at 7 mm (i.e., ALMA band 1) are vital to better characterize disk dust masses.

Figures (15)

• Figure 1: The observed SEDs and models for the 18 successfully fitted protoplanetary disks in our Serpens sample. The blue lines show 1000 randomly selected models from the posterior distributions. The models are reddened using 2009McClure extinction law and the extinction values gathered from the MCMC fitting process (see Section \ref{['subsubsec:sed modeling MCMC']}). Photometry labels are detailed in Table \ref{['table:photometry']}. Spitzer data is labeled as c2d. The complete figure set (18 images) is available in the online journal.
• Figure 2: The observed SEDs and successfully fitted models for the first 10 of the 160 protoplanetary disks in our L1641/L1647 sample. The blue lines show 1000 randomly selected models from the posterior distributions. The models are reddened using 2009McClure extinction law and the extinction values gathered from the MCMC fitting process (see Section \ref{['subsubsec:sed modeling MCMC']}). Photometry labels are detailed in Table \ref{['table:photometry']}. Spitzer data is labeled as c2d. The complete figure set (119 images) is available in the online journal. The additional SEDs for the 41 L1641 objects can be consulted in 2023Rilinger.
• Figure 3: Posterior distributions of the individual 18 successfully modeled sources of the Serpens sample (top panels) and ensemble distribution (bottom panels) for the following parameters (from top left to bottom right): $\alpha$ (disk viscosity), $\dot{M}$ (mass accretion rate), $R_{disk}$ (outer disk radius), $\epsilon$ (dust settling), $a_{max,midplane}$ (maximum grain size in the midplane), $a_{max,upper}$ (maximum grain size in the upper layers of the disk). The error bars denote the 16th and 84th percentiles; the bottom panels' histograms use 20 bins.
• Figure 4: (Continued). Posterior distributions of the individual 18 successfully modeled sources of the Serpens sample (top panels) and ensemble distribution (bottom panels) for the following parameters (from left to right): $M_{disk}/M_*$ (disk mass fraction to host star mass), $M_*$ (host star mass).
• Figure 5: Posterior distributions of the individual 160 successfully modeled sources of the L1641/L1647 sample (top panels) and ensemble distribution (bottom panels) for the following parameters (from top left to bottom right): $\alpha$ (disk viscosity), $\dot{M}$ (mass accretion rate), $R_{disk}$ (outer disk radius), $\epsilon$ (dust settling), $a_{max,midplane}$ (maximum grain size in the midplane), $a_{max,upper}$ (maximum grain size in the upper layers of the disk). The error bars denote the 16th and 84th percentiles; the bottom panels' histograms use 20 bins.