Substellar objects (mass M < 0.075 Msol) have core temperatures too low to produce stable hydrogen fusion. Therefore, they progressively contract and cool down.  Hundreds of brown dwarfs (M=0.013-0.075 Msol) and fewer hundreds of planetary-mass objects (M<0.013 M) or candidates have been discovered in the last fifteen years. Brown dwarfs have been characterized by ultra-cool spectral types M, L, and T. Due to their progressive dimming - especially rapid at ages <0.1 Gyr, substellar objects can be better detected when they are young. Nearby open clusters of well determined age and metallicity are idoneous regions to find such objects and test the predictions of theoretical models. In this PhD thesis we present the search, identification, and characterization of new substellar objects in the Pleiades and sigma~Orionis open clusters, and with theoretical masses of ~35-25 and ~4 times the mass of Jupiter (Mjup), respectively.

The Pleiades open cluster has receaved great attention in the litterature, due to its richness in objects, its proximity to the Sun (~120 pc), youth (~120 Myr), notable proper motion and small interstellar extinction. Because its population of fainter, L-type brown dwarfs was poorly known, we performed a search for these objects. We analyzed optical and near infrared imaging data (RIJ bands) in an area of 1.8 deg2 and we found 18 candidates. Subsequently, we performed a follow-up of these candidates in the near infrared (HKs bands), obtaining photometric and spectroscopic measurements to verify membership to the cluster  and determine  spectral characteristics. We confirm nine cluster members by proper motion, which allow us to improve the determination of the Pleiades substellar mass spectrum dN/dM (number of objects per unit of mass). In the range ~0.5-0.026 Msol, the mass spectrum can be fit by a power law M^-alpha with an index alpha=0.5+-0.2. This index is similar to those of open clusters about thirty times younger, suggesting that the differential evaporation of the low mass brown dwarfs relative to the more massive objects is insignificant. This result points to the existence of substellar objects of even lower mass in the Pleiades. The spectra obtained for six cluster members show  early- to mid L spectral types. At least two of them have spectral features which can be related to low surface gravity, dust, and youth. The Pleiades L-type brown dwarfs appear to be intrinsically as faint as field L-type dwarfs of known trigonometric parallax, which disagrees with luminosity predictions of theoretical models. A possible explanation could be that the Pleiades L-type brown dwarfs may have contracted faster than what the models indicate.

In the ~3 Myr old, extinction free, and relatively nearby (~360-440 pc) sigma Orionis open cluster, we searched for additional T-type planetary-mass candidates to S Ori 69 and S Ori 70. We collected and analyzed photometric data from the far red to the mid infrared in an area of ~840 arcmin2. We discovered three new planetary-mass candidates whose spectral types based on their colours would be L, transition L/T, and mid T, respectively. The L-type candidate is located near to a solar-type star belonging to the cluster (angular separation of 11.8 arcsec). The estimated theoretical masses of these candidates are in the range 2-7 Mjup, accounting for age, distance, and photometric uncertainties. If confirmed as cluster members, they would be some of the least massive planetary-mass objects identified by direct imaging outside the Solar System. Nevertheless, some of them are probably contaminating field dwarfs (that are much older and nearby) or galaxies, in which case the cluster mass spectrum would have a strong change in the slope, which could be evidence of the so-called opacity limit for the
fragmentation of molecular clouds.