Detection of the emission from the secondary component in a binary system can be extremely challenging, but equally rewarding. In the case of intermediate to high-mass binaries, detection of close companions can inform formation theories. In the extreme low mass-ratio case, where the secondary component is in fact a planet, detection of the emission in high resolution spectroscopy can be used to determine the true planet mass. In this thesis, we describe a technique to detect the thermal emission from the secondary component of a low mass-ratio binary system. We apply this technique to archived observations of early B-type stars using VLT/CRIRES, and simulate future observations of planetary systems with IGRINS, a near-infrared spectrograph being built now. / text

Since high-mass stars form deeply embedded within dense molecular clumps, the evolution of young stars and of dense clumps is inextricably linked. Previous datasets, however, lack the information necessary to test the prevailing theories. Definitive tests require a sufficiently large sample of molecular clumps and maps of their gas temperatures, column densities, velocity dispersions, and velocities at a spatial resolution comparable to, or smaller than, the clump scale (~1 pc). The Radio Ammonia Mid-Plane Survey (RAMPS), a new molecular line survey of thermal NH3 and H2O masers, provides the necessary data. In this dissertation, I used RAMPS data and archival datasets to test several theories of high-mass star formation and to investigate the co-evolution of molecular clumps and high-mass stars. All theories of high-mass star formation make testable predictions regarding clump kinematics and gravitational stability. Analyses of RAMPS kinematic data revealed that the majority of molecular clumps, particularly those in early evolutionary stages, are unstable to gravitational collapse. Further, they display infall motions, a key prediction of the theory of competitive accretion. I also investigated the kinematics of molecular filaments by comparing their measured velocity gradients to those predicted by hydrodynamical simulations. The measured spatial distributions of velocity gradients are inconsistent with existing models. Feedback from protostars and stars is predicted to alter the properties of surrounding clumps. I investigated feedback size scales and found that high-mass protostellar and stellar feedback significantly changes the temperatures, chemical abundances, and velocity dispersions of clumps on scales of ~0.3 to 3 pc. Finally, I observed a massive molecular cloud filament undergoing an interaction with a supernova shock, which is accelerating, heating, and injecting turbulence into the filament's gas. Although the molecular cores hosted by the filament may remain gravitationally bound, the filament is gravitationally unbound and likely being dispersed. Given that the shock is removing a reservoir of gas that could have been accreted by the cores, these data suggest that the supernova is inhibiting star formation.