The onset of stellar multiplicity in massive star formation: A search for low-mass companions of massive young stellar objects with L-band adaptive optics imaging

Context. Given the high incidence of binaries among mature field massive stars, it is clear that multiplicity is an inevitable outcome of high-mass star formation. Understanding how massive multiples form requires the study of the birth environments of massive stars, covering the innermost to outermost regions.Aims. We aim to detect and characterise low-mass companions around massive young stellar objects (MYSOs) during and shortly after their formation phase. By the same means, we also probed the 3.8-mu m emission that surrounds these massive protostars, in order to link the multiplicity to their star-forming environment.Methods. To investigate large spatial scales, we carried out an L'-band high-contrast direct imaging survey seeking low-mass companions (down to L-bol approximate to 10(circle dot )or late A-type) around thirteen previously identified MYSOs using the VLT/NACO instrument. From those images, we looked for the presence of companions on a wide orbit, covering scales from 300 to 56 000 au. Detection limits were determined for all targets and we tested the gravitational binding to the central object based on chance projection probabilities.Results. We have discovered a total of thirty-nine potential companions around eight MYSOs, the large majority of which have never been reported to date. We derived a multiplicity frequency (MF) of 62 +/- 13% and a companion fraction (CF) of 3.0 +/- 0.5. The derived stellar multiplicity and companion occurrence are compared to other studies for similar separation ranges. The comparisons are effective for a fixed evolutionary stage spanning a wide range of masses and vice versa. We find an increased MF and CF compared to the previous studies targeting MYSOs, and our results match the multiplicity rates derived among more evolved populations of massive stars. For similar separation ranges, we however confirm a higher multiplicity than that of T Tauri stars (similar to 30%), showing that the statement in which multiplicity scales with primary mass also extends to younger evolutionary stages. The separations at which the companions are found and their location with relation to the primary star allow us to discuss the implications for the massive star formation theories.Conclusions. Our findings do not straightforwardly lift the uncertainty as to the formation process of massive stars as a whole but we rather examine the likely pathways for individual objects. However, the wide distance at which companions are detected rather supports core fragmentation or capture as the main mechanisms to produce wide multiples. We find hints of triggered star formation for one object and discuss the massive star against stellar cluster formation for other crowded fields.