Introduction

The discovery of the first unambiguous evolved brown dwarf Gliese 229B , the confirmation of the existence of young brown dwarfs in the Pleiades open cluster , and the detection of dozens of others from photometric  and proper motion surveys has restricted the intriguing gap between stars and planets. In fact, brown dwarfs are bright enough to be easily detected in standard bandpasses (ZJHK) from ground-based facilities. This is now understood as the natural consequence of strong absorption bands of H₂O and H₂ that depress the infrared flux in favor of the near-infrared bandpasses, far from any black body distributions. This effect was already apparent from the results of our model atmospheres , which were later extended deeper into the temperature regime of evolved brown dwarfs by . But as brown dwarf discoveries unfold, several questions arise. Why do brown dwarfs appear to form two distinct subgroups: 1) hotter red objects just below the stellar main sequence, and 2) much cooler and blue methane dwarfs? Is the apparent gap between these groups real? Are different physical processes involved in their atmospheres beyond the change of effective temperature?

From previous model atmospheres, it immediately appeared more difficult to explain the behavior of the hotter brown dwarfs, than that of the methane dwarfs in which water vapor and methane bands naturally matched the predictions of models ( and ) models within the accuracy of available molecular absorption profiles. Even the most detailed model atmospheres had failed to reproduce accurately the spectroscopic and photometric properties of red dwarfs later then about M6: all models were systematically too blue by as much as a magnitude in standard infrared colors (V-K, I-K, J-K). has reviewed this situation and the atmosphere modeling at the bottom of the main sequence. The reasons for these discrepancies was the onset of dust grain formation. As mass decreases along the main sequence, the latest-type red dwarfs bear outer atmospheric layers which reach temperatures well below 1800K, favoring the formation of dust grains. While it has been long suspected that grains could form under these conditions , the inclusion of such computation to non-grey model atmosphere calculations had to wait until -- using Gibbs free energies of formation and a simple sphere approximation for the Mie opacity of the grains -- published their exploratory non-grey dusty brown dwarf models. They explored the formation of three dust grain species: Al₂O₃, Fe, and MgSiO₃, and found corundum (Al₂O₃) to be a very abundant and powerful continuous absorber in red dwarfs with spectral type later than M8, while cooler methane brown dwarfs appeared comparatively grainless. But their models were based on band models for molecular opacities and could not reproduce the optical spectral distributions and several photometric properties of brown dwarfs.

Recently, and have explored models with finite radial extention of silicate clouds to address the systematic difference between early L and late T type brown dwarfs. While it is pertinent to explore such processes, several parameters must inevitably be used to characterize them, and moreover, these are likely time-dependant processes . So while we do believe that dust diffusion (referred to herafter as gravitational settling) contributes in reducing the radial extention of clouds in these atmospheres, we feel that, as in all unsolved physical problems, it is important to explore carefully the limiting cases. In this paper, we present therefore our model calculations for two opposite limiting cases of dust content in brown dwarfs atmospheres: 1) inefficient gravitational settling i.e. the dust is distributed according to chemical equilibrium predictions, which provides the maximum impact of dust upon the brown dwarf properties, and 2) efficient gravitational settling i.e. the dust forms and depletes refractory elements from the gas, but their opacity does not affect the thermal structure, hence a minimal effect upon brown dwarfs properties.

These models bring substantial improvements upon the previous similar work of in the number of grains included both in the chemical equilibrium and in the opacity database, as well as with regards to the molecular opacities. In Section we describe how hundreds of grain species are now included self-consistently to the chemical equilibrium calculation to allow us to identify the hot condensates that are most abundant in the atmospheres of late-type M dwarfs and brown dwarfs. In Section 3 we describe our treatment of the grain opacities. And Section 4 we describe the detailed opacity sampling model atmospheres to which these grain formation and opacities are incorporated. Section 6 is reserved for the discussion of the atmospheric structures, while the effects of grains on the spectroscopic and photometric properties of brown dwarfs are discussed in Section 7 and 8.