Abstract
While it now known that most stars are born in relatively massive clusters, i.e., with masses greater than 1000 solar mass, several fundamental questions remain open, including ``what is the timescale of the process?'' and ``what sets the star formation efficiency?'' Furthermore, there is no consensus on a general theory that predicts the outcome of star formation, e.g., the stellar initial mass function (IMF) or multiplicity/clustering properties, from given interstellar medium conditions. Development of such a theory requires testing by observations of forming clusters that span a wide range of environments and evolutionary stages. Now with the unprecedented sensitivity and resolution provided by facilities like the Atacama Large Mm/sub-mm Array (ALMA) and the forthcoming James Webb Space Telescope (JWST), it is possible to study star cluster formation in detail --- star by star --- in relatively distant regions throughout our Galaxy.
In this dissertation, I first present a case study of massive star cluster formation in G286.17+0.14 (hereafter G286), which is located at a distance of 2.5 +- 0.3 kpc in the Carina spiral arm. We have conducted a multi-wavelength survey with facilities including ALMA (millimeter) and Hubble Space Telescope (HST)/ Very Large Telescope (VLT) / Gemini (near-infrared [NIR]). This allows us to trace both the gas/dust component and the young stellar object (YSO) population. From the mm continuum we identified about 100 cores and derived the Core Mass Function (CMF). For M >1Msun, the fiducial dendrogram-identified CMF can be fit with a power law of the form dN/dlogM ~ M^-alpha with \alpha = 1.24 +- 0.17, which is slightly shallower than, but still consistent with, the index of the Salpeter stellar IMF of 1.35. This further strengthens the case of a correspondence between CMF and IMF that has been seen previously in local regions, but now is found in a more distant, massive protocluster. The kinematics and dynamics of the gas in G286 was studied using spectral lines, including C18O 2-1 and those from deuterated species like N2D+, DCO+ and DCN. The 0.02-pc-scale dense cores, and pc-scale filamentary structures in G286 show internal kinematics that are consistent with being in virial equilibrium. However, the velocity distribution of the whole cloud appears to be composed of two spatially distinct velocity groups, indicating that the dense molecular gas has not yet relaxed to virial equilibrium, perhaps due to there being recent or continuous infall into the system. With multi-epoch HST J/H band data we also characterize the stellar variability of the young stars. By comparing the NIR photometry for data taken in 2014 and 2017 for around 6000 stars, we found significant variability in about 7% of the sample. This percentage is higher (14%) for objects that show NIR color signatures of having a protostellar disk. An object with extreme variability was also found, with a K band brightening of 3.5 magnitudes. Follow-up observations indicate this object is a very low mass (<0.12Msun) example of an FU Ori type (accretion burst) source, which would be the lowest mass example of this class.
In the next part of the thesis I explore the environmental dependence of star formation by extending the developed analysis methods to other regions. One example is the Center Ridge Clump (CRC) of the Vela C giant molecular cloud (GMC). This dense clump was selected as showing the lowest level of sub-mm polarization angle dispersion in BLASTPOL mapping of the region and so is expected to be strongly magnetized. We have characterized the dense cores in the CRC with ALMA band 6 (1.3~mm) and 7 (0.87~mm) observations. We identified 11 dense cores from their continuum emission, with masses ranging from 0.1 to 4.5 Msun. Their deuteration ratios, determined from N2H+3-2 and N2D+ 3-2, span from 0.09 to 1.28, with the latter being one of the highest values yet measured. These ratios appear to be a good tracer of core evolution. Overall this region has a relatively low dense gas fraction compared with other typical clouds with similar column densities, which may be a result of its strong magnetic field.
As an attempt to extend the analysis to different environments and also study the detailed star formation process on protostellar disk scales, we also present ALMA band 6 and 7 and VLA Ka band (9~mm) observations toward NGC 2071 IR, an intermediate-mass star formation region in the L1630 cloud of Orion B. We characterize the continuum and associated molecular line emission towards the most luminous protostars, i.e., IRS1 and IRS3, as well as other protostellar objects, on ~ 40 au scales. IRS1 is partly resolved in millimeter
and centimeter continuum and shows a potential disk. IRS3 has a clear disk appearance in millimeter continuum and is further resolved into a binary system in our 9 mm map. Both sources exhibit clear velocity gradients across their protostellar disks in multiple spectral lines. We use an analytic method to fit the Keplerian rotational motion of the disks, and derive constraints on physical parameters, such as the dynamical mass of the central object. For both IRS1 and IRS3, the inferred ejection directions from different tracers, including radio jets, water masers, molecular outflows and H2 emission, are not always consistent and can be misaligned by up to ~50 degrees. IRS3 is better explained by a single precessing jet with its axis wiggling over a range of position angles. A similar mechanism may be present in IRS1, but unresolved multiplicity is also a possibility.
We conclude with a discussion of the prospects for extending such studies of star formation, where individual stars and disks are characterized across the full range of the mass spectrum, to other regions in the Galaxy.
