Carina High-contrast Imaging Project for massive Stars (CHIPS) II. O stars in Trumpler 14

Context. Most massive stars belong to multiple systems, yet the formation process leading to such high multiplicity remains insufficiently understood. To help constrain the different formation scenarios that exist, insights into the low-mass end of the companion mass function of such stars is crucial. However, this is a challenging endeavour as (sub-)solar mass companions at angular separations (rho) below 1 '' (corresponding to 1000-3000 au in nearby young open clusters and OB associations) are difficult to detect due to the large brightness contrast with the central star. Aims. With the Carina High-contrast Imaging Project of massive Stars (CHIPS), we aim to obtain statistically significant constraints on the presence and properties of low-mass companions around massive stars in a previously unreachable observing window (Delta mag greater than or similar to 10 at rho less than or similar to 1 ''). In the second paper of the series, we focus on the Trumpler 14 cluster, which harbours some of the youngest and most massive O-type stars in the Milky Way. Methods. We obtained VET-SPHERE observations of seven O-type objects in Trumpler 14 using IRDIFS_EXT mode. These provide us with a 12 '' x 12 '' field of view (approximately ((3 x 3) x 10(4) au) centred on each O star and allow us to search for companions at separations larger than 0 ''.15 (approx. 360 au) and down to magnitude contrast >10 mag in the near-infrared. We used angular and spectral differential imaging along with Point Spread Function (PSF) fitting to detect sources and measure their flux relative to that of the central object. We then used grids of ATLAS9 and PHOENIX Local Thermodynamic Equilibrium (LTE) atmosphere models combined with (pre-)main-sequence evolutionary tracks to estimate the mass of the detected candidate companions. Results. We detected 211 sources with near-infrared magnitude contrast in the range of 2-12. Given the large surface number density of stars in Trumpler 14, one cannot reliably distinguish between cluster members and genuine companions for most of the detected sources. The closest companion, at only 0 ''.26, is characterised as a 1.4 M-circle dot star with an age of 0.6 Myr, in excellent agreement with previous age estimates for Tr 14. The mass function peaks at about 0.4 M-circle dot and presents a dearth of stars in the 0.5-0.8 M-circle dot mass range compared to previous estimates of the initial mass function in Tr 14. While statistically significant, part of these differences may result from contamination of the K-band fluxes by circumstellar material. Conclusions. SPHERE is clearly suitable to probe the low-mass end of the mass function in the vicinity of massive stars. Follow-up SPHERE observations to obtain the full Y to K spectral energy distribution would allow for better constraints on the masses of the detected sources, and to confirm (or invalidate) the curious mass function that we derived for low-mass stars in the vicinity of the O-type objects in Trumpler 14.