The highly collimated bipolar outflow of OH 231.8+4.2

We present high spatial resolution observations of the CO molecular emission (J=1-0 and J=2-1 lines) in the post-AGB bipolar nebula OH 231.8ler. High-quality NIR images (J, H, K' bands) of light scattered by grains were also obtained. Our observations probe the bulk of the nebular material, providing maps with a resolution ~ 1 arcsec of the mass distribution, both CO and NIR images being very closely coincident. The combination of the two ¹²CO lines has been used to measure the distribution of the kinetic temperature in the nebula, which is found to be very low, ranging between 8 K, in the outer southern clumps, and 35 K, in the central region. A relative temperature increase is found in the northernmost condensation, probably associated to a strong bow-like shock. Since velocities are also measured in CO, the dynamic parameters (kinetic momentum and energy) are also measured with high resolution. Most of the nebular mass ( ~ 0.64 M_sun) is located in the central condensation and flows at expansion velocities <=40 km s^-1. The rest of the gas, ~ 0.3 M_sun almost equally distributed in the two lobes, flows along the nebular axis at high velocities, that increase proportionally to the distance to the central star reaching values as large as 430 km s^-1, as a result of a sudden acceleration happened about 770 yr ago. The general mass distribution in OH 231.8ler\ is found to be clumpy and very elongated, with a length/width ratio reaching a factor 20 in the southern tail. In the center, however, we find a double hollow-lobe structure, similar to those found in other well studied protoplanetary nebulae. We stress the enormous kinetic linear momentum carried by the molecular nebula, about 27 M_sun,km s^-1 (5.5 x 10³⁹ g cm s^-1). The kinetic energy is also very high, ~ 1700 M_sun,km² s^-4 ~ 3.4 x 10⁴⁶ erg. Given the short time during which the acceleration of the molecular outflow took place, we conclude that the linear momentum carried by the stellar photons is about a factor 100 smaller than that carried by the outflow, even if the effects of multiple scattering are taken into account. We independently argue that radiation pressure directly acting onto grains (the mechanism thought to be responsible for the mass ejection in AGB envelopes) cannot explain the observed bipolar flow, since this would produce a significant shift between the dust and gas features that is not observed. Finally, we review the uncertain nature and evolutionary status of this unique object. Based on observations carried out with the IRAM Plateau de Bure Interferometer. IRAM is supported by INSU/CNRS (France), MPG (Germany) and IGN (Spain).