The binary fraction of planetary nebula central stars I. A high-precision, I-band excess search

TL;DR

This study measures the binary fraction of planetary nebula central stars using high-precision I-band and J-band photometry, revealing a higher binary fraction than the progenitor main sequence stars, suggesting binary interactions are common in planetary nebulae formation.

Contribution

The paper introduces a high-precision photometric method to determine the binary fraction of planetary nebula central stars and provides a de-biased estimate, comparing it with main sequence binary fractions.

Findings

De-biased binary fraction of 67-78% for central stars.

Binary fraction of 54% in J band for brighter companions.

Estimated total binary fraction possibly exceeds 100%, indicating high binary interaction prevalence.

Abstract

In an attempt to determine how many planetary nebulae derive from binary interactions, we have started a project to measure their unbiased binary fraction. This number, when compared to the binary fraction of the presumed parent population can give a first handle on the origin of planetary nebulae. By detecting 27 bona fide central stars in the I band we have found that 30% of our sample have an I band excess between one and a few sigmas, possibly denoting companions brighter than M3-4V and with separations smaller than approximately 1000 AU. By accounting for the undetectable companions, we determine a de-biased binary fraction of 67-78% for all companions at all separations. We compare this number to a main sequence binary fraction of (50+/-4)% determined for spectral types F6V-G2V, appropriate if the progenitors of today's PN central star population is indeed the F6V-G2V stars. The error on our estimate could be between 10 and 30%. We conclude that the central star binary fraction may be larger than expected from the putative parent population. Using the more sensitive J band of a subset of 11 central stars, the binary fraction is 54% for companions brighter than approximately M5-6V and with separations smaller than about 900 AU. De-biassing this number we obtain a binary fraction of 100-107%. The two numbers should be the same and the discrepancy is likely due to small number statistics. We also present an accurately vetted compilation of observed main sequence star magnitudes, colours and masses, which can serve as a reference for future studies. We also present synthetic colours of hot stars as a function of temperature (20-170kK) and gravity (log g= 6-8) for Solar and PG1159 compositions.