Imaging the LkCa 15 system in polarimetry and total intensity without self-subtraction artefacts

TL;DR

The paper tackles the challenge of interpreting small-scale disk substructures in protoplanetary systems without self-subtraction artefacts by employing star-hopping RDI with SPHERE to image LkCa 15 in Ks-band, complemented by ALMA continuum data. Through radiative-transfer modeling with RADMC-3D, the authors test dust configurations and find that porous CAHP aggregates (CAHP-128-100 nm) in the disk surface layer best reproduce both the total and polarized near-infrared scattering, as well as the millimeter morphology when paired with compact mm grains. The study places upper limits on unseen planets (e.g., ≈$1.5$–$3.6$ $M_J$ depending on location) and reveals a radially segregated dust distribution, with micron grains inside ≈50 au and mm grains extending from ≈58 to 127 au, coupled with a shallow grain-size distribution slope of $ ilde{\ ext{ζ}} \,\approx -2.31$. Overall, combining star-hopping imaging with phase-function diagnostics provides strong constraints on dust grain properties and disk structure, informing theories of planet formation in gas-rich environments.

Abstract

Studying young protoplanetary disks is essential for understanding planet formation, but traditional angular differential imaging can introduce self-subtraction artefacts that hinder interpretation of small-scale structures. We present high-resolution total- and polarized-intensity Ks-band images of the LkCa~15 system obtained with SPHERE using near-simultaneous reference-star differential imaging (star-hopping), yielding self-subtraction-free images beyond 0.1 arcsec. LkCa~15 hosts a ~160 au protoplanetary disk and has previously been reported to harbour candidate protoplanets at separations of 15--18 au. We analyse the disk morphology and dust properties and search for super-Jupiter planets beyond 20 au. We first model the near-infrared scattered-light images together with ALMA submillimetre continuum data using RADMC-3D and a two grain-size (micron and millimetre) compact olivine model. While this model broadly reproduces the disk geometry, it overpredicts the degree of forward scattering in the near-infrared. To investigate this discrepancy, we extract the scattering phase function S(theta) and polarized fraction P(theta) from the SPHERE data and compare them with aggregate-scattering models. The observed phase functions disfavour compact Mie spheres and are better matched by porous aggregates (CAHP). Recomputing the scattered-light models with porous CAHP grains in the disk surface layer significantly improves agreement with the observed Ks-band morphology and polarization, while retaining compact millimetre grains to reproduce the ALMA continuum. No new planetary companions are detected; we place upper mass limits of ~1.5 MJ beyond 200 au and ~3.6 MJ in the inner disk. Our results demonstrate that combining star-hopping imaging with phase-function diagnostics provides strong constraints on dust grain properties in protoplanetary disks.

Figures (19)

• Figure 1: Left to right: IRDIS total intensity $K$-band, Stokes $Q-K$-band and Stokes $U-K$-band images as observed from VLT-SPHERE.
• Figure 2: Inverted $K$-band polarization fraction (${I_{\text{tot}}}/{\sqrt{Q^2 + U^2}})$ map of LkCa 15 system. Planets with low polarization, if exist, overlay on a backdrop of highly polarized light coming from the disk, resulting in a robust discrimination caused by a high signal-to-noise ratio (SNR). However, there is no unambiguous planetary signature in this image.A map value of 14 corresponds to a polarization of roughly $1/14\approx7\%$.
• Figure 3: Schematic diagram showing the axis orientation for the E-field chosen in this paper. The +Q is aligned along the X-direction and -Q is aligned along the Y-direction. The blue points represent the dust grains, while the arrows indicate the scattered photons towards observer, an incomparably large distance away.
• Figure 4: The signs of Stokes parameters taken in this paper indicating the positive and negative regions in Q and U images.
• Figure 5: Comparison of RADMC-3D model with Olivine grains and observed data. The sequence from left to right shows the data, the simulated model, and the residuals. The top row is the $K$-band total intensity image, followed by the $K$-band Stokes Q, and the Stokes U. The fourth and fifth rows at the bottom correspond to ALMA $880~\mu$m and $1300~\mu$m images, respectively.