Search for companions around nearby massive stars

By comparing our sensitivities to model predictions of disk instability based on physical criteria for fragmentation and cooling, and using Monte Carlo simulations for orbital distributions, we find that ~85% of such companions should have been detected in our images on average.

Given this high degree of completeness, stringent statistical limits can be set from the null-detection result, even with the limited sample size. We find that <30% of massive stars form and retain disk instability planets, brown dwarfs, and very low mass stars of <100M_jup_ within 300AU, at 99% confidence. These results, combined with previous findings in the literature, lead to the conclusion that core accretion is likely the dominant mode of planet formation.