Abstract
Context . In the early phases of star formation, dense and cold structures such as filaments and prestellar cores are commonly probed and studied via observations of thermal dust emission. The inferred core properties from these observations are critically dependent on the adopted dust model, which must be characterized carefully. However, analysis of dust properties in individual prestellar cores is often limited by the wavelength range and spatial resolution of the observations. Aims . Our target in this work, L1506C, is part of the filament L1506 in the Taurus molecular cloud ( D ~ 140 pc). Previous studies using dust and gas tracers suggest that it is a prestellar core in the making and have revealed signatures of dust evolution. Our goal was to bring further constraints on both the cloud structures and dust properties. To do so, we used observations of dust emission covering, for the first time, a broad wavelength range, from the far-infrared (FIR) to millimeter range, as well as 3D radiative transfer modeling while also taking into account constraints from extinction data. Methods . We used both Herschel FIR observations and complementary millimeter wavelength observations with NIKA2 at the IRAM 30m telescope. Analysis of the spectral energy distributions over the whole spectral range was first performed with modified blackbody (MBB) fits. The data were then analyzed with 3D modeling using the radiative transfer code SOC and the latest THEMIS 2 dust model. As additional constraints we used extinction observations from WIRCam at CFHT and from Spitzer . Results . The MBB modeling over the wavelength range from 160 μm to 2 mm revealed that L1506C is fragmented into two low density cores with n H 2 < 5 × 10 4 cm −3 , independently of the assumed dust opacity, with masses smaller than their Jeans masses. The dust color temperature ( T ) drops from 16 K in the filament to 11 K in the cores, and the spectral index β increases from 1.4 to 1.9, showing a clear T − β anticorrelation and a change in grain properties. To model the source with 3D radiative transfer modeling, grains more evolved than the diffuse interstellar medium (ISM) are required in the densest part. Using grains with sizes up to 0.7 μm, 50% porosity, and 50% ice content, we were able to reproduce the observed millimeter emission within a 10% residual in the central core region and the near-infrared scattering (coreshine) within 1 σ . Extinction data strongly constrain the dust properties, but the best results we obtained still overestimate the extinction by a factor of more than 2.5 with the current THEMIS 2 dust model. The model N H 2 has a maximum value of 1.3 × 10 22 cm −2 , 1.7 times larger than the one calculated with MBB modeling, and the model central dust temperature is down to 8 K. Good fits to the observations were obtained with a transition between diffuse and evolved grains at low density, within the range 1500 ≤ n H ≤ 4500 cm −3 . When adopting a model with a transition threshold of n H = 3000 cm −3 , evolved grains account for 50% of the dust content along the line of sight, where A V ≈ 4 on the observed extinction map. Conclusions . Grain growth can already happen at a very early stage of star formation, even before the onset of gravitational collapse. It is difficult to precisely constrain the cloud properties with 3D radiative transfer modeling when only using dust emission observations because some of the parameters are degenerate. Observations of near-infrared extinction and scattering and independent measurements from molecular line data are crucial complements to these data, providing independent constraints and breaking degeneracies in the analysis.