According to the Global Hierarchical Collapse (GHC) scenario a molecular cloud should globally collapse and start individual small collapses at its density peaks. Thus the average density of the cloud should increase with time and its Jeans mass should decrease, naturally fragmenting into lower-mass fragments as the cloud evolves. There are hints that this is happening observationally but no clear systematic study has been performed so far. Thus, a great contribution to test the GHC scenario would be to systematically show, using submillimeter high angular resolution images of infrared-quiet clumps, that the less evolved clumps fragment into more massive fragments.

Theoretical models suggest that molecular cores with density power-law indices steeper than -2 should not exist. However, a non-negligible number of these cores has been found. The goal of this thesis is test whether the claimed cases of "steep cores" are actually so steep, by fitting a sophisticated model which does not adopt the typical assumptions made in the literature to infer density power-law indices.

In the context of cluster-forming clumps collapsing from large scales, one would expect that fragmentation takes place simultaneously with the global collapse of the clump. If this is actually happening, one would expect to find that typical separations between fragments decrease with evolutionary stage. A sample of cluster-forming clumps at different evolutionary stages and undergoing fragmentation will be analized to answer this question.

The answer to this question has direct implications on the structure and dynamics of molecular clouds. We aim at answering it from an observational point of view by observing the submillimeter polarized emission of a sample of massive dense cores at clump-scales. Since this sample has already been observed at core-scales, the project will directly answer the question.