2011A&A...531A..33C

Current surveys of nearby star-forming regions and young aggregates that produce samples of objects spanning the entire mass range down to masses of a few Jupiter masses can address questions about differences in the substellar mass function among different regions and the lowest masses of objects that can form in isolation. Deep imaging of a selected area in the Lupus 3 star-forming region, characterized by a large number of known young stellar objects, has been carried out with the goal of producing a complete sampling of the mass function down to sub-Jupiter masses. I_CJHK_S imaging complemented with intermediate-band imaging sensitive to methane absorption has been obtained with the VLT. The observed area measures 7'3x7'4 (0.42x0.43pc² at 200pc) and the background is obscured by A_V<5mag in approximately 70% of it. Detection limits (3σ) are I_C=25.6, J=23.1, H=22.3, and K_S=21.3. Cool objects with temperatures below 3000K are identified by means of reddening-free indices. Objects cooler than ∼1300K should also be detectable by comparing the flux in the H and methane-continuum bands. The luminosities of the 17 identified cool objects are too low in all cases to be consistent with membership in the star-forming region. Their number agrees with the statistical expectation for the background population of very low-mass stars. A reexamination of known candidate young stellar objects leads to excluding eight of them, reducing the number of bona fide members in the area to 15, of which ten lie in low extinction regions. The expectation based on the stars-to-brown dwarf ratios derived in other young aggregates is to find two or three new brown dwarfs, whereas none have been found. The low-mass young stellar object Par-Lup3-3 is found to consist of two close components of similar brightness separated by 0.3". A very faint object is also detected 1.2" from the jet-driving very low-mass star Par-Lup3-4 that, if confirmed as a physical companion, would have an estimated mass of 1-2M_Jup. The absence of brown dwarfs in one of the most crowded areas of the Lupus 3 clouds, although limited by small-number statistics, argues against a mass function rich in very low-mass objects, as might have been expected if the apparent overabundance of mid-M spectral types in Lupus extended toward later types. The results also stress the importance of critically examining the actual nature of previously identified candidate young stellar objects at the time of drawing statistical conclusions about the substellar population.