Abstract

Context. The close binary system NN Serpentis must have gone through a common envelope phase before the formation of its white dwarf. During this phase, a substantial amount of mass was lost from the envelope. The recently detected orbits of circumbinary planets are probably inconsistent with planet formation before the mass loss. Aims: We explore whether new planets may have formed from the ejecta of the common envelope and derive the expected planetary mass as a function of radius. Methods: We employed the Kashi {\amp} Soker model to estimate the amount of mass that is retained during the ejection event and inferred the properties of the resulting disk from the conservation of mass and angular momentum. The resulting planetary masses were estimated from models with and without radiative feedback. Results: We show that the observed planetary masses can be reproduced for appropriate model parameters. Photoheating can stabilize the disks in the interior, potentially explaining the observed planetary orbits on scales of a few AU. We compare the expected mass scale of planets for 11 additional systems with observational results and find hints of two populations, one consistent with planet formation from the ejecta of common envelopes and the other a separate population that may have formed earlier. Conclusions: The formation of the observed planets from the ejecta of common envelopes seems feasible. The model proposed here can be tested through refined observations of additional post-common envelope systems. While it appears observationally challenging to distinguish between the accretion on pre-existing planets and their growth from new fragments, it may be possible to further constrain the properties of the protoplanetary disk through additional observations of current planetary candidates and post-common envelope binary systems.