Search

Filling out the dearth of detections between direct imaging and radial velocity surveys will test theories of planet formation and (sub)stellar binarity across the full range of semi-major axes, connecting formation of close to wide separation gas giants and sub-stellar companions. Direct detection of close-in companions is notoriously difficult as classical techniques fail near the diffraction limit. In this dissertation, I present a new pipeline which searches for faint companions using kernel-phase interferometry (KPI), a technique utilizing interferometric analysis of the full unobscured telescope aperture. I demonstrate the pipeline, and the power of KPI, by performing two surveys and accompanying demographic analyses. First, I search for companions in the entire HST/NICMOS F110W and F170M image archive of nearby brown dwarfs (BDs), demonstrating significant sensitivity to companions at half the diffraction limit. I discover no new companions but recover and refine astrometry of 19 previous imaging companions and confirm two previous kernel-phase detections. I also present contrast curves to enable a population study and to demonstrate the strength of this technique at separations inaccessible to classical imaging techniques. Second, after estimating physical properties of a subset of this sample, I use a Bayesian framework to compare these detections and detection limits to a model companion population. When correcting for Malmquist bias, I find a smaller population of companions with tighter separations than seen in previous studies. This is due to our use of KPI, which enables us to resolve the peak of the semimajor axis distribution with significant sensitivity to low-mass companions. I confirm the previously-seen trends of decreasing binary fraction with decreasing mass and a strong preference for tight and equal-mass systems in the field-age sub-stellar regime; only ∼ 1% of systems are wider than 20 au or have a mass ratio q < 0.6. Third, I present the results of a second KPI based survey of sub-stellar objects in Taurus and Upper Scorpius using archival HST/ACS. I present 6 new candidate detections at extremely tight separations, in addition to 4 previously known companions. This is the first application of KPI to visible-wavelength observations. I find an observed companion frequency of F [subscript obs] = 0.24 [superscript +0.10][subscript −0.08] which is slightly higher than previous studies, consistent with our new detections. Combining these two surveys and demographic analyses, I find a significant excess of young wide (ρ > 10 au) companions compared to the field. I also find that the field population of wide companions is consistent with my population of systems formed in low density regions if it were to be diluted by single systems formed in high density regions. This is consistent with dynamical evolution preferentially dissolving systems born in high-density star-forming regions over those in low-density regions. I attribute the characteristics of the BD binary population to turbulent fragmentation setting the initial conditions followed by a brief period of dynamical evolution, removing the widest and lowest-mass companions, before the birth cluster dissolves.

Our understanding of how exoplanets form and evolve relies on analyses of both the mineralogy of protoplanetary disks and their detailed structures; however, these key complementary aspects of disks are usually studied separately. We present initial results from a hybrid model that combines the empirical characterization of the mineralogy of a disk, as determined from its mid-infrared spectral features, with the well-tested MCFOST radiative transfer disk model that takes into account realistic disk density and temperature structures, a combination we call the EaRTH Disk Model. With the results of the mineralogy detection serving as input to the radiative transfer model, we generate mid-infrared spectral energy distributions that reflect both the mineralogical and structural parameters of the corresponding disk. Initial fits of the SED output by the resulting integrated model to Spitzer Space Telescope mid-infrared (IRS) spectra of the protoplanetary disk orbiting the nearby T Tauri star MP Mus demonstrates the potential advantages of this approach by revealing details like the dominance of warm small olivine and cool small pyroxene in the dusty disk of MP Mus. The methodology proposed here provides insight into disk composition and structure, but requires fine-tuning in the radiative transfer stage to reproduce specific spectral features. However, it should be directly applicable to the interpretation of mid-infrared spectra of protoplanetary disks that will be produced by the James Webb Space Telescope. Additionally, while the model can successfully match known observations assuming a typical fiducial protoplanetary disk, it is necessary to alter the model to match data when structural differences are observed, such as a gap or rings; adjusting the model to match infrared observations while remaining consistent with millimeter observations demonstrates the model’s adaptability to different disk structures, which we do using the Spitzer IRS spectra and Atacama Large Millimeter/sub-Millimeter Array (ALMA) observations of MP Mus, as well as the transition disks of GM Aur and LkCa 15. The results illustrate the large differences that occur in protoplanetary disk development despite similar ages, likely the result of planetary formation.

It is now widely accepted that supermassive black holes (SMBHs) lie at the center of most massive galaxies. While SMBHs in these hosts have been well documented and studied, those in lower mass hosts, such as dwarf and spiral galaxies without a central bulge component (i.e. bulgeless galaxies), have remained quite elusive and insufficiently studied. Identifying and examining exactly how active galactic nuclei (AGN) co-evolve and influence bulgeless and dwarf host galaxies is the foundation of my dissertation research. These galaxies can be used as local analogs of the infant Universe (gas rich, low metallicity, rapidly growing) and their BH number density and occupation fraction can provide important constraints on the BH seed population. However, these AGN are difficult to study since they are energetically weak or heavily obscured. The IR provides a unique and compelling avenue to study these AGN since it is not as severely affected by dust obscuration as optical observations. As such, my research utilizes IR selection techniques and observations to identify a new population of obscured AGN that have been overlooked by large optical surveys. Much of my work has focused on coronal lines (CLs), outflows, and analyzing the impacts of AGN on their hosts. In my publications, we confirm AGN activity in a sample of dwarf and bulgeless galaxies through the presence of CL and hidden broad-line emission. In many of these, we find fast, kpc outflows that are likely driven by the AGN. We also find that the outflowing gas have sufficient escape velocities to allow the gas to enrich the circumgalactic medium. As a result, these outflows have the potential to suppress star formation and could play a key role in dwarf and bulgeless galaxy evolution. Additionally, these observations can help deliver key constraints to improve the theoretical feedback models of dwarf and bulgeless galaxies.

The ensemble emission from all sources outside of the Milky Way is known as the extragalactic background light (EBL). At optical and near-infrared (NIR) wavelengths, the EBL is primarily stellar emission tracing back to the Epoch of Reionization (EOR) at redshifts z > 6 when the first luminous structures formed. Given the large uncertainties in our understanding of the EOR, measurements of the EBL provide an important probe of the galaxies that were responsible for reionization. Direct observations of the EBL are challenging due to contamination from bright local foregrounds. In recent years, intensity mapping has emerged as a successful technique in which EBL fluctuations are measured on large angular scales where the known foreground contributions are minimal or well-modeled. To isolate the signals from EOR structures, intensity mapping can be applied to optical and NIR data to probe rest-frame UV emission in galaxies at z > 6. Multiple intensity mapping studies including the first Cosmic Infrared Background ExpeRiment (CIBER-1) have found that the optical/NIR EBL intensity and its large-scale fluctuations exceed predictions from galaxy models. The excess is above EOR level and also persists at wavelengths < 1 μm where we do not expect to see reionization signals. To explain the excess, a number of astrophysical sources have been proposed including intra-halo light (IHL) from low-mass stars at the outskirts of galaxies. Observations at 1.1 and 1.8 μm from CIBER-1 second and third flights suggest that the excess can be best described by a level of IHL comparable to the integrated light from known galaxy populations. While this result is intriguing, given CIBER-1 spectral coverage, the IHL and EOR components could not be distinguished, prompting interest in a new mission, CIBER-2. CIBER-2 is designed to disentangle the IHL and EOR signals using broader spectral coverage from 0.5 - 2.0 μm in six wavebands and larger light-gathering power. The wavebands are selected to provide 21 auto- and cross-spectra to probe the Lyman break that can distinguish EOR contributions from low-redshift foregrounds. CIBER-2 comprises a 28.5-cm telescope cooled to < 100K using liquid nitrogen, and three HAWAII-2RG detectors coupled with dual- band filters to obtain data in six wavebands simultaneously. CIBER-2 is planned for four flights on the Black Brant IX sounding rocket, with the first flight in mid-2021. In this dissertation, I present my work on the CIBER-2 design, characterization and payload integration, as well as constructing the EBL fluctuation power spectra from the data taken from CIBER-1’s final flight. This analysis relies on a previously developed pipeline for use with earlier flights, but has been revised to capture the hardware changes in the final flight and the corresponding systematic uncertainties. I will also outline the expected development of CIBER-2 post-flight analysis and highlight the advantage of CIBER-2 data for EOR studies.

As the birth places of super star clusters (SSCs), starbursts provide a testbed of astrophysics on a vast range of scales, from the microphysics of star formation to the macrophysics describing the evolution of galaxies and ultimately the Universe as a whole. SSCs represent the dominant mode of star formation in the early Universe. In the local Universe, these ~1-10 Myr old systems host dense concentrations of thousands of massive stars, offering an unprecedented view of the likely progenitors of globular clusters. This dissertation investigates the influence of SSCs on the gas that they ionize and subsequent effect of their starbursts on the galaxies in which they form. In particular, I explore the following questions: how does massive star feedback clear natal gas from SSCs? What conditions allow a cluster to remain bound through this gas dispersal phase or form multiple generations of stars? Through what mechanisms do starbursts regulate the central mass evolution of disk galaxies? How do SSCs drive chemically enriched, galaxy-scale winds from starbursts? To study the launching of winds, we need tracers that can penetrate the dusty clouds around young SSCs, and both high spatial and spectral resolution. The Brackett α line (4.05μm) is ideal, but not available to most observers. Using NIRSPEC (Keck II) we can map Br α and probe the dynamics of heavily embedded HII regions near SSCs. In this thesis, we present NIRSPEC observations in a variety of starburst environments, with the aim of providing an empirical basis for the emerging paradigm in the formation of the most massive stellar systems.

The center of the Milky Way Galaxy is a rather exotic place. It is home to a supermassive black hole (Sgr A*), 3 massive stellar clusters (the Arches, Central and Quintuplet clusters), and several isolated massive stars which are spread throughout the inner ~100 pc. In this dissertation, I present mid-infrared imaging data of three regions in the Galactic center including the Arches and Quintuplet clusters, and the H HII regions. Observations of these regions were taken with the Faint Object infraRed CAmera for the SOFIA Telescope (FORCAST) at 19.7, 25.2, 31.5, and 37.1 um. These data trace warm dust (~100 K) in the surrounding environment of stars and provide constrains on physical properties of the sources such as luminosity, mass, and temperature. The Arches cluster is surrounded by a grouping of molecular clouds which are heated and ionized by the cluster. Both the clouds (known as the Arched Filaments) and the cluster are named for the arc-like appearance of the clouds which are one of the most striking large scale features in the Galactic center. We examine the distribution of dust temperatures across the clouds in the region and find that they are relatively uniform (70-100 K) even though the filamentary clouds have a large physical size (~25 pc). Equilibrium heating analysis of the dust indicates that the characteristic grain size is smaller (~0.01 um) than typical values for the interstellar medium (~0.1 um). We also find evidence for depleted abundances of polycyclic aromatic hydrocarbons (PAHs). Both the grain size and the PAH depletion indicate processing of material in the clouds and may be explained by shocks or other processes occurring in the region. Adjacent to the Arched Filaments there is a collection of HII regions associated with isolated massive stars known as the H HII regions. Studying these sources provides important information on the prevalence and distribution of massive stars in the region and understanding their origin may provide insights on other field stars which are observed throughout the Galactic center. Observations with SOFIA/FORCAST provide high-angular resolution maps (~4") which allows us to study the detailed morphology of four of the HII regions. Analysis of dust temperatures show that three of the spatially extended objects require multiple heating sources, and we identify potential stellar counterparts. We also compare the size and morphology of the HII regions and argue from this that the properties of H2 are consistent with in situ star formation. Last, we identify 8 new sources that may be part of the H complex and provide initial characterizations of their infrared emission. The Quintuplet Proper Members (QPMs) are a collection of five bright infrared sources for which the Quintuplet cluster is named. These sources are thought to be binary systems with a carbon-rich Wolf-Rayet (WC) star and O/B companion. Dust production by these types of systems are poorly understood and constraining their mass-loss history can provide insight on this process. We produce models of the dust emission associated with the four detected sources using the radiative transfer code DUSTY and show that Q2 and Q3 are best fit by radial density profiles which are consistent with constant mass-loss rates. In contrast, Q1 and Q9 show shallower radial density profiles indicating more vigorous mass loss in the stars' recent past. Q9 is particularly interesting because it has a massive dust reservoir which is significantly larger than the other QPMs. N-band imaging obtained with the Gemini Telescope reveal an extended dusty nebulae associated with the source which suggests the material may have been produced by a recent episode of enhanced mass loss from the star. Such events are atypical of dusty WC stars, and we suggest that the dust may have been produced in an earlier phase of the star.

A central question in the study of star formation is: what sets the mass distribution of nascent stars? This distribution appears to vary little (especially at the low mass end) throughout our galaxy and others, despite variations in the heavy element abundance, interstellar radiation field, and amount of gas available. This work is a component of the Herschel Orion Protostar Survey (HOPS), a program to characterize hundreds of protostars in the Orion Molecular Clouds, largest site of on-going star formation with 0.5 kiloparsec. As part of the HOPS program, we have spectral energy distributions with a wavelength coverage of 1.2-870 µm, making this one of the most well characterized samples of nearby star formation. We extend this work with a large survey imaging 283 protostars in Orion with the Hubble Space Telescope, enabling us to observe with 80 AU resolution at 1.6 µm. At this wavelength, light from the central photosphere is scattered off the dust ubiquitous throughout the gas, resolving the structure of the protostellar envelope. We use the high angular resolution to perform a morphological study of the nearby protostellar gas (the envelope), finding that our morphologies are indicative of evolutionary trends. The high resolution of this data makes it uniquely suited for finding unambiguously edge-on protostars (of which we report thirteen at high confidence) and for tracing the structure of cavities cleared by protostellar outflows. We report on one of these unambiguously edge-on protostars, HOPS 171, which we've studied in detail with the Atacama Large Millimeter/submillimeter Array to find a complex structure. We use an edge detection routine to perform the analysis on tracing cavities caused by outflows, and report the first application of such a routine to study the role of feedback on the envelope. Using evolutionary diagnostics, we find no evidence for the growth of these outflow cavities as protostars evolve through the first 100000 years. This contrasts with previous authors who have invoked the growth of the outflow cavities as a mechanism for clearing the gas infalling into forming stars. We validate our approach by confirming that the number of protostars without observed outflow cavities is consistent with the distribution of cavity sizes inferred from our observation; i.e., that we are not appreciably affected by observation bias. We conclude that the reduction/halting of infall and the low star formation efficiency inferred from molecular clouds cannot be explained with envelope clearing by outflows.

This thesis presents Adaptive Optics (AO) imaging and spectroscopic measurements of infrared excess sources in the Galactic Center to determine if they are similar to G2, the first example of a spatially-resolved object interacting with the supermassive black hole at the center of the Galaxy, Sgr A$^{*}$. Our goal is to understand the debated nature of these infrared excess sources and to understand their relationship with the supermassive black hole. Our objects have been monitored with AO from he past decade (2004 - 2015) and are within the inner 1\farcs75 of the supermassive black hole. We initially focus on one source, G1, which gets comparably close to the supermassive black hole compared to G2 ($a_{min}\sim200-300$ AU) and lies on a very eccentric orbit ($e_{G1} = 0.99$). While G2 has been tracked before and during periapse passage ($T_{0} \sim$ 2014.2), G1 has been followed since soon after emerging from periapse ($T_{0} \sim$ 2001.3). Our observations of G1 double the previously reported observational time baseline, which improves its orbital parameter determinations. G1's orbital trajectory appears to be in the same plane as that of G2, but with a significantly different argument of periapse ($\Delta\omega$ = 21$\pm$4 degrees). This suggests that G1 is an independent object and not part of a gas stream containing G2 as has been proposed. Furthermore, we show for the first time that: (1) G1 is extended in the early epochs (those closest to periapse) along the direction of orbital motion and (2) G1 becomes significantly smaller over time, changing from 450 AU in 2004 to less than 170 AU in 2009 and thereafter. Based on these observations, G1 appears to be the second example of an object tidally interacting with a supermassive black hole. G1's continued survival 14 years after periapse, along with its compactness in epochs further from the time of periapse, suggests that this source is stellar in nature and may be consistent with a black-hole driven stellar binary merger induced by the Kozai mechanism.We then observationally characterize several other infrared excess sources to see whether or not they share similar properties. We find another source, G3, has nearly identical observation properties to G1 and G2 but lies further away from Sgr A${*}$. G3 does not lie on the same orbital plane to G1 or G2 at all, suggesting that these objects do not necessarily all come from the same region. G3 does not lie on a highly eccentric orbit at all ($e\sim0.4$) and its periapse passage distance is significantly further away than G1 or G2 ($a_{min}\sim5000$ AU). Several other sources also exhibit Br-$\gamma$ emission, but also do not lie on the same orbital plane as G1 or G2. G3 and other infrared excess sources supports the hypothesis that there is a population of objects that share similar observational qualities to G1 and G2 -- high infrared luminosities, cold dust temperatures, and Br-$\gamma$ emission line features. These new objects might therefore exist as a new population of sources in the Galactic center, and might not necessarily have the same physical manifestation.

We present a multi-wavelength study of the nearby supernova remnant Cassiopeia A (Cas A). Easily resolvable supernova remnants such as Cas A provide a unique opportunity to test supernova explosion models. Additionally, we can observe key processes in the interstellar medium as the ejecta from the initial explosion encounter Cas A’s powerful shocks. In order to accomplish these science goals, we used the Spitzer Space Telescope’s Infrared Spectrograph to create a high resolution spectral map of select regions of Cas A, allowing us to make a Doppler reconstruction of its 3-dimensional structure structure. In the center of the remnant, we find relatively pristine ejecta that have not yet reached Cas A’s reverse shock or interacted with the circumstellar environment. We observe O, Si, and S emission. These ejecta can form both sheet-like structures as well as filaments. Si and O, which come from different nucleosynthetic layers of the star, are observed to be coincident in some regions, and separated by >500 km s−1 in others. Observed ejecta traveling toward us are, on average, #24;800 km s−1 slower than the material traveling away from us. We compare our observations to recent supernova explosion models and find that no single model can simultaneously reproduce all the observed features. However, models of different supernova explosions can collectively produce the observed geometries and structures of the emission interior to Cas A’s reverse shock. We use the results from the models to address the conditions during the supernova explosion, concentrating on asymmetries in the shock structure. We also predict that the back surface of Cassiopeia A will begin brightening in #24;30 years, and the front surface in #24;100 years. We then used similar observations from 3 regions on Cas A’s reverse shock in order to create more 3-dimensional maps. In these regions, we observe supernova ejecta both immediately before and during the shock-ejecta interaction. We determine that the reverse shock of the remnant is spherical to within 7%, although the center of this sphere is offset from the geometric center of the remnant by 810 km s−1. We determine that the velocity width of the nucleosynthetic layers is #24;1000 km s−1 in a given region, although the velocity width of a layer along any given line of sight is

This thesis presents a new multi-component spectral energy distribution (SED) decomposition method for analyzing the ultraviolet to millimeter wavelength SEDs of dusty infrared-luminous galaxies. SEDs are constructed from spectroscopic and photometric data obtained with the Spitzer Space Telescope in conjunction with photometry from the literature. Each SED is decomposed into emission from populations of stars, an AGN accretion disk, PAH molecules, atomic and molecular lines, and distributions of graphite and silicate grains. Decompositions of the SEDs of ~20 template starburst galaxies and ~10 template AGNs provide baseline properties to aid in quantifying the star-formation and accretion contributions in composite sources. As case studies, we show that obscured radiation from stars is capable of powering the total dust emission from the composite galaxy NGC6240, while the decomposition of Mrk1014 is consistent with ~65% of its power emerging from an AGN and ~35% from star-formation. The properties of our template starburst galaxies and AGNs are used to estimate the AGN-strengths for some of our ~75 ULIRGs, although many are so obscured that uncovering their principal energy source is not possible in the infrared, and instead will require dust penetrating X-ray observations. We suggest that many of the variations in the SEDs of starburst galaxies may be explained in terms of the different mean optical depths through the clouds of dust surrounding the young stars within each. We also suggest that the divergent far-IR properties of AGNs may result from differences in the relative orientation of their host galaxy and AGN accretion disks, resulting in different amounts of AGN-heated cold dust emission emerging from their host galaxies. We estimate that 30-50% of the far-IR and PAH emission from Mrk1014 may originate from such AGN-heated material in its host galaxy disk, demonstrating that caution must be used when estimating the strength of star-formation based solely upon the far-IR or PAH properties of a source.

The dense, hot cores of 24 molecular clouds were examined at infrared wavelengths between 2.2 and 25 µm in search of embedded, luminous objects thought to massive 0 and B stars at an early stage of their evolution. 19 of the clouds were examined at 20 µm for high sensitivity to sources of low color temperature between 80 and 200 K. The clouds IC 1848 A, S 235, S 255, S 269, IR 12.4 + 0.5, S 140 and Cepheus A contain a total of 12 embedded sources of radiation more luminous than 6 x 10^3 L⊙ of which 7 were discovered during the course of this research. The 20 µm searches were particularly fruitful, yielding discoveries of 4 objects too red to be discovered in previous searches at 2.2 and 10 µm. In all cases the infrared sources are found projected against the densest part of the molecular cloud where the 13CO column density exceeds 2 x 1016 cm-2 On the other hand, no compact infrared sources were found in clouds such as Cepheus B which has high gas temperature, about 30 K, but low gas density less than 3 x 10^3 cm^-3 . The absence of embedded sources in some cases suggests that heating by stars external to the cloud can result in elevated temperatures in rarified clouds. The success of the 20 µm searches points out two biases inherent in studies made at shorter wavelengths; First, the effect of intra-cloud extinction is to obscure deeply embedded objects at wavelengths shorter than 20 µm. A reliance on counts of sources discovered at 2.2 and 10 µm will result in a serious underestimate of the rate of formation of massive stars. Second, short wavelength studies favor the discovery of objects which lie close to the front of a cloud where the intra-cloud extinction is low, thereby creating an artificial need to explain the preferential location of these objects close to the cloud boundaries. The importance of these prejudices to theories of star formation is discussed. This work has led to the realization that infrared sources in molecular clouds tend to form in clusters whose characteristic size is very similar to that found in trapezium-like groups of 0 and B stars. It seems likely that molecular clouds fragment as their densities reach 10^4 cm^-3 into independently collapsing nuclei separated by about 0.1 pc and each capable of producing one (or more) star(s). Some embedded infrared sources are surrounded by a small amount of ionized material. The strength of the radio emission seen toward objects brighter than 10^4 L⊙ such as Cepheus A and S 140 is consistent with these objects being pre-main sequence 0 and B stars.