Early planet formation in embedded protostellar disks: Setting the stage for the first generation of planetesimals

(Abridged) Recent surveys of young star formation regions have shown that the average Class II object does not have enough dust mass to make the cores of giant planets. Younger Class 0/I objects have enough dust in their embedded disk, which begs the questions: can the first steps of planet formation occur in these younger systems? The first step is building the first planetesimals, generally believed to be the product of the streaming instability. Hence the question can be restated: are the physical conditions of embedded disks conducive to the growth of the streaming instability? Here we model the collapse of a `dusty' proto-stellar cloud to show that if there is sufficient drift between the falling gas and dust, regions of the embedded disk can become sufficiently enhanced in dust to drive the streaming instability. We include four models, three with different dust grain sizes and one with a different initial cloud angular momentum to test a variety of collapse trajectories. We find a `sweet spot' for planetesimal formation for grain sizes of a few 10s of micron since they fall sufficiently fast relative to the gas to build a high dust-to-gas ratio along the disk midplane, but have slow enough radial drift speeds in the embedded disk to maintain the high dust-to-gas ratio. Unlike the gas, which is held in hydrostatic equilibrium for a time due to gas pressure, the dust can begin collapsing from all radii at a much earlier time. The streaming instability can produce at least between 7-35 M$_\oplus$ of planetesimals in the Class 0/I phase of our smooth embedded disks, depending on the size of the falling dust grains. This first generation of planetesimals could represent the first step in planet formation, and occurs earlier in the lifetime of the young star than is traditionally thought.Comment: Accepted for publication in A&A, 22 pages, 19 figure