Peculiar Disk Substructures Associated with the Young Eruptive Star EX Lupi

TL;DR

This study reveals a multiscale, structured protoplanetary disk around the young eruptive star EX Lupi by combining ALMA Band 3/4 continuum observations with PRIISM-based super-resolution imaging and archival SPHERE near-infrared data. A newly detected crescent-shaped inner arc within $\lesssim 10$ au and a narrow outer ring at $r_{ring} \approx 28.6$ au indicate dynamically active dust rearrangement, with the outer ring showing marginal confinement ($\hat{w}_{d}/h_{p} \approx 0.95$) consistent with dust trapping in a gas pressure maximum. The disk is moderately flared with a surface height $H_{s}\approx8.2$ au at $r\approx77$ au and a small flaring angle $\varphi(r)\approx0.006( r/\mathrm{au})^{0.09}$, yielding a dust temperature profile of $T_{d}(r) \approx 75\pm6\, (L_{*/L_{\odot}})^{0.25}(r/\mathrm{au})^{-0.48}$ K and a disk-averaged temperature of $\langle T_{d}\rangle \approx 18\pm1$ K. The authors discuss massive companion-driven mechanisms, such as vortex formation and gap carving, as a unified explanation for both substructures and episodic accretion in EX Lupi, suggesting that such activity can occur in relatively mature disks and may regulate accretion through dynamical disk-planet interactions.

Abstract

Young eruptive stars such as EXors undergo dramatic accretion outbursts characterized by sudden optical brightenings, yet the underlying physical mechanism remains uncertain. We present high-resolution Atacama Large Millimeter/submillimeter Array (ALMA) Band 3 and 4 continuum observations of EX Lupi, the prototypical EXor-type variable, reconstructed using super-resolution imaging with sparse modeling. Our images reveal, for the first time, two distinct substructures: a compact, crescent-shaped inner arc within 10 au of the star, and a narrow outer ring at 30 au. The inner arc is strongly elongated and casts a shadow observed in the Very Large Telescope/SPHERE near-infrared scattered light. The outer ring exhibits a radial width comparable to the local pressure scale height, consistent with moderately efficient dust trapping. Geometric and thermal analysis of the disk surface, based on combined ALMA and SPHERE data, indicates that the disk is moderately flared with an average disk temperature consistent with that of classical T Tauri disks. The observed substructures suggest dynamical perturbations-plausibly induced by a massive companion companion-that may modulate accretion rates through gravitational interaction with the inner arc. These findings provide morphological evidence linking disk substructure to episodic accretion in the structurally mature disk.

Figures (10)

• Figure 1: Gallery of ALMA (PRIISM) continuum images and their residual maps at 3.2 mm / 93 GHz (Band 3: bottom row) and 2.0 mm / 153 GHz (Band 4: top row) of the EX Lupi disk. The same linear color scale is adopted. A white bar of $0\hbox{$.\!\!^{\prime\prime}$}1$ (= 16 au) is provided to reference the angular scales. The central stellar symbol ($\star$) indicates the stellar position measured in Gaia observations. Left: Model image reconstructed using PRIISM imaging. The image initially has a unit of janskys per pixel ($\rm Jy~pixel^{-1}$; Section \ref{['sec:imaging_priism']}), while it is converted to millijanskys per square arcsecond ($\rm mJy~arcsec^{-2}$) for comparison with the restored image (middle). The filled white ellipse denotes the effective spatial resolution $\theta_{\rm eff}$. The resolution is estimated from an artificial point-source injection method (Appendix \ref{['sec:effective_resolution']}). Middle: restored image after convolution with an elliptical Gaussian representing the main lobe of the synthesized beam, along with adding the dirty map of residuals (right). Each white contour corresponds to $[3, 4, 5]\times \sigma_{\rm noise}$, where $\sigma_{\rm noise}$ is the RMS noise value calculated from the emission-free area ($36~\mu \rm Jy~\rm beam^{-1}$ for Band 3 and $48~\mu \rm Jy~\rm beam^{-1}$ for Band 4). The restored image is expressed in a unit of $\rm Jy~beam^{-1}$, while being converted to $\rm mJy~arcsec^{-2}$. Right: dirty map of the residuals, obtained by subtracting the model from the observation in the visibility domain. The unit is expressed in image RMS noise. The black contours represent the same as the white ones on the restored image (middle).
• Figure 2: Left: the best-fit ellipse to the outer ring overlaid on the restored Band 4 image, with contour levels at $[3, 4, 5] \times \sigma^{\rm B4}_{\rm noise}$, where $\sigma^{\rm B4}_{\rm noise}$ is the RMS noise of $48~\mu \rm Jy~ beam^{-1}$. The purple curves represent the distribution of solutions considering the estimated $1\sigma$ error. The white dots along the outer ring indicate the radially averaged peak positions, spaced at $20^{\circ}$ intervals. The outer peak in the PA of $140^{\circ} \sim 160^{\circ}$ could not be identified due to insufficient angular resolution or low SNR and was therefore excluded from the fit. Right: the deprojected image with the inclination ($=42^{\circ}.6$) and PA ($=51^{\circ}.7$) derived from the best-fit ellipse. The center of the ellipse fit is used as the center position for deprojection. The purple circle indicates the best-fit model, while the white star symbol indicates the position of EX Lupi.
• Figure 3: Top panel: deprojected PA profile of the outer ring of EX Lupi, derived from the Band 4 restored image. The contour levels are $[3, 4, 5] \times \sigma_{\rm noise}$. Note that the emission from the inner arc ($r < 20$ au) is removed. The center coordinates are set based on an ellipse fit to the outer ring. Bottom panel: Radial intensity profile averaged over the full azimuthal angle, displayed on a linear color scale. The profile (solid purple curve) is linearly interpolated onto radial grid points spaced by 0.1 au using $\tt interpolate.interp1d$ from the $\tt SciPy$ module. The light purple ribbon represents the error of the mean at each radius. The dashed orange curve corresponds to the best-fit Gaussian profile to the observed data. The geometric mean of the spatial resolution is shown in the inset at the top-right corner.
• Figure 4: Close-up views of the inner arc of EX Lupi in ALMA Band 3 (left) and Band 4 (right) PRIISM model images. The same linear color scale is used. The white stellar symbol indicates the position of the star and the red one the position of the peak intensity of the inner arc. Contour lines at the $(40,50,60,70,80,90)\%$ value of the peak emission are overplotted. A white bar of $0\hbox{$.\!\!^{\prime\prime}$}05$ (= 8 au) is provided as a reference to the angular scales. The filled white ellipse indicates the effective spatial resolution $\theta_{\rm eff}$.
• Figure 5: Panels of (a) and (b): Deprojected position angle profiles for Band 3 and Band 4 model images. The profiles show the radial location of the peak (circle points) and its $50\%$ of the intensity distribution of the inner arc overlaid by the model images. The origin of the orbital radius is the position of the central star. Panel (c) shows the radial FWHM sizes of the intensity distribution. Panel (d) shows the peak intensity (circle points). In all panels, the circle symbols denote PA-binned means computed over $10^{\circ}$-wide PA bins; the shaded regions indicate the corresponding standard deviations. The vertical bars indicate the geometric mean of the effective spatial resolutions.