Six billion years from now, while evolving on the asymptotic giant branch (AGB), the Sun will metamorphose from a red giant into a beautiful planetary nebula. This spectacular evolution will impact the solar system planets, but observational confirmations of the predictions of evolution models are still elusive as no planet orbiting an AGB star has yet been discovered. The nearby AGB red giant L 2 Puppis (d = 64 pc) is surrounded by an almost edge-on circumstellar dust disk. We report new observations with ALMA at very high angular resolution (18 × 15 mas) in band 7 (ν ≈ 350 GHz) that allow us to resolve the velocity profile of the molecular disk. We establish that the gas velocity profile is Keplerian within the central cavity of the dust disk, allowing us to derive the mass of the central star L 2 Pup A, m A = 0.659 ± 0.011 ± 0.041 M (±6.6%). From evolutionary models, we determine that L 2 Pup A had a near-solar main-sequence mass, and is therefore a close analog of the future Sun in 5 to 6 Gyr. The continuum map reveals a secondary source (B) at a radius of 2 AU contributing f B / f A = 1.3 ± 0.1% of the flux of the AGB star. L 2 Pup B is also detected in CO emission lines at a radial velocity of v B = 12.2 ± 1.0 km s −1 . The close coincidence of the center of rotation of the gaseous disk with the position of the continuum emission from the AGB star allows us to constrain the mass of the companion to m B = 12 ± 16 M Jup . L 2 Pup B is most likely a planet or low-mass brown dwarf with an orbital period of about five years. Its continuum brightness and molecular emission suggest that it may be surrounded by an extended molecular atmosphere or an accretion disk. L 2 Pup therefore emerges as a promising vantage point on the distant future of our solar system.

Context. Evolved low-mass stars lose a significant fraction of their mass through stellar winds. While the overall morphology of the stellar wind structure during the asymptotic giant branch (AGB) phase is thought to be roughly spherically symmetric, the morphology changes dramatically during the post-AGB and planetary nebula phase, during which bipolar and multi-polar structures are often observed. Aims. We aim to study the inner wind structure of the closest well-known AGB star CW Leo. Different diagnostics probing different geometrical scales have implied a non-homogeneous mass-loss process for this star: dust clumps are observed at milli-arcsec scale, a bipolar structure is seen at arcsecond-scale, and multi-concentric shells are detected beyond 1 . Methods. We present the first ALMA Cycle 0 band 9 data around 650 GHz (450 μm) tracing the inner wind of CW Leo. The fullresolution data have a spatial resolution of 0. 42 × 0. 24, allowing us to study the morpho-kinematical structure of CW Leo within ∼6 . Results. We have detected 25 molecular emission lines in four spectral windows. The emission of all but one line is spatially resolved. The dust and molecular lines are centered around the continuum peak position, which is assumed to be dominated by stellar emission. The dust emission has an asymmetric distribution with a central peak flux density of ∼2 Jy. The molecular emission lines trace different regions in the wind acceleration region and imply that the wind velocity increases rapidly from about 5 R , almost reaching the terminal velocity at ∼11 R . The images prove that vibrational lines are excited close to the stellar surface and that SiO is a parent molecule. The channel maps for the brighter lines show a complex structure; specifically, for the 13 CO J = 6-5 line, different arcs are detected within the first few arcseconds. The curved structure in the position-velocity (PV) map of the 13 CO J = 6-5 line can be explained by a spiral structure in the inner wind of CW Leo, probably induced by a binary companion. From modelling the ALMA data, we deduce that the potential orbital axis for the binary system lies at a position angle of ∼10-20• to the north-east and that the spiral structure is seen almost edge-on. We infer an orbital period of 55 yr and a binary separation of 25 au (or ∼8.2 R ). We tentatively estimate that the companion is an unevolved low-mass main-sequence star. Conclusions. A scenario of a binary-induced spiral shell can explain the correlated structure seen in the ALMA PV images of CW Leo. Moreover, this scenario can also explain many other observational signatures seen at different spatial scales and in different wavelength regions, such as the bipolar structure and the almost concentric shells. ALMA data hence for the first time provide the crucial kinematical link between the dust clumps seen at milli-arcsecond scale and the almost concentric arcs seen at arcsecond scale.

Simultaneous MERLIN observations of the OH 1665‐ and 1667‐MHz maser lines in the circumstellar envelope of the semiregular star W Hya have been taken in all Stokes parameters. The 1665‐MHz emission comes from two elongated clusters located 80 au from the star. The 1667‐MHz emission arises in an incomplete shell of radius 130 au, with the blueshifted features located in the northern part of the envelope and the redshifted components clustered south of the centre. The circularly polarized maser components exhibit spatial separation along the north–south direction. The linearly polarized components were found from the near side of the envelope. Their polarization position angles indicate that the projected axis of the magnetic field at PA ≃ −20° is consistent with spatial segregation of circular polarization. The intensity of the magnetic field, estimated from a tentative measurement of Zeeman splitting, is about 0.6 mG at the location of the 1667‐MHz emission, with the field pointing away from the observer. A small change of position angles of linear polarization observed in both maser lines is interpreted as a weak Faraday effect in the maser regions with an electron density of about 2 cm−3. The overall polarization structure of the envelope suggests an ellipsoidal or weak bipolar geometry. In such a configuration, the circumstellar magnetic field may exert a non‐negligible influence on mass loss. The velocity field in the circumstellar envelope recovered from observations of SiO, H2O, OH and CO lines at five radial distances reveals a logarithmic velocity gradient of 0.25 and 0.21 in the 1665‐ and 1667‐MHz maser regions respectively. The acceleration within tens of stellar radii cannot be explained by the classical model of radiation pressure on dust.

Context. The condensation of inorganic dust grains in the winds of evolved stars is poorly understood. As of today, it is not yet known which molecular clusters form the first dust grains in oxygen-rich (C/O < 1) asymptotic giant branch (AGB) winds. Aluminium oxides and iron-free silicates are often put forward as promising candidates for the first dust seeds. Aims. We aim to constrain the dust formation histories in the winds of oxygen-rich AGB stars. Methods. We obtained Atacama Large Millimeter/sub-millimeter array (ALMA) observations with a spatial resolution of 120 × 150 mas tracing the dust formation region of the low mass-loss rate AGB star, R Dor, and the high mass-loss rate AGB star, IK Tau. We detected emission line profiles of AlO, AlOH, and AlCl in the ALMA data and used these line profiles to derive a lower limit of atomic aluminium incorporated in molecules. This constrains the aluminium budget that can condense into grains. Results. Radiative transfer models constrain the fractional abundances of AlO, AlOH, and AlCl in IK Tau and R Dor. We show that the gas-phase aluminium chemistry is completely different in both stars with a remarkable difference in the AlO and AlOH abundance stratification. The amount of aluminium locked up in these three molecules is small, ≤1.1 × 10 −7 w.r.t. H 2 , for both stars, i.e. only ≤2% of the total aluminium budget. An important result is that AlO and AlOH, which are the direct precursors of alumina (Al 2 O 3 ) grains, are detected well beyond the onset of the dust condensation, which proves that the aluminium oxide condensation cycle is not fully efficient. The ALMA observations allow us to quantitatively assess the current generation of theoretical dynamical-chemical models for AGB winds. We discuss how the current proposed scenario of aluminium dust condensation for low mass-loss rate AGB stars within a few stellar radii from the star, in particular for R Dor and W Hya, poses a challenge if one wishes to explain both the dust spectral features in the spectral energy distribution (SED) in interferometric data and in the polarized light signal. In particular, the estimated grain temperature of Al 2 O 3 is too high for the grains to retain their amorphous structure. We advocate that large gas-phase (Al 2 O 3 ) n clusters (n > 34) can be the potential agents of the broad 11 µm feature in the SED and in the interferometric data and we propose potential formation mechanisms for these large clusters.