research

first stars#

Born in the pristine hydrogen gas of the early universe, the first generation of stars (also known as Population III, or Pop III, stars) are believed to be very different from the stars we see today. In order to condense to high densities and form stars, clumps of gas need to radiate away their thermal energy (such that gravitational collapse can beat out the pressure support). In the local universe (for basically every star we’ve observed), this ‘cooling’ of the gas is governed by the “heavy” elements produced by previous generations of stars. Pop III stars, however, are not so fortunate. In the early universe, the gas is too cold to excite hydrogen and cooling can therefore only occur inefficiently through molecular hydrogen, or H2. This results in stars thought to be tens to hundreds of times the mass of our Sun. At such large masses, they lived short, few million year lives, which, in some cases, ended in superluminous supernova explosions (SNe), releasing heavy elements and enriching their interstellar environments, setting the stage for more complex structures and later generations of stars to form.

Artist’s rendition of a Pop III star (source: National Astronomical Observatory of Japan)

The formation of these stars crucially depends on the presence of molecular hydrogen in a dark matter (DM) halo. Therefore, any processes that affect the H2 content of a DM halo will in turn affect the ability of that halo to form stars. There are several key processes believed to significantly affect the H2 fraction in a halo—the relic velocity between baryons and DM in the early universe, the buildup of a photodissociating UV background, and the development of a photo-ionizing and -heating X-ray background. The first of these effects, known as the ‘stream velocity’, supresses the gas fraction in a DM halo and delays the formation of stars. Once the first Pop III stars form, they emit UV photons that can dissociate molecular hydrogen. Therefore, as more stars form, a metagalactic UV background—known as the ‘Lyman-Werner (LW) background’—builds up and future generations of star formation are suppressed. These stars also contribute to a growing X-ray background, which can both positively and negatively feed back into the star formation process. First, the X-ray photons can ionize hydrogen, boosting the electron fraction in a gas cloud, which in turn promotes the formation of molecular hydrogen (thereby supporting star formation). On the other hand, these hot electrons can also distribute their thermal energy throughout the gas, making the cooling process more challenging. The competition between all of these effects sets the star formation rate in a given DM halo over time.

These processes, while all relating to the star formation process in different ways, can all be parameterized in terms of a single quantity: the minimum dark matter halo mass for stars to form. I am using a series of simple models for these various effects to generate an analytic estimate for this mass scale. I can then fold this estimate into a semi-analytic model for star formation to understand how these processes will affect global, potentially observable quantities, such as the star formation rate density.

galaxies#

I have also worked on galaxy simulations—specifically searching for so-called ‘prolate rotators’ in the IllustrisTNG (TNG) cosmological simulation suite. Prolate rotation is an observational phenomenon wherein massive, cigar-shaped, elliptical galaxies are observed to rotate around their long axes (i.e., the spin and shape major axes are coincident). Using the TNG simulations, we analyzed the evolution of massive galaxies to (1) find a robust way to identify prolate rotators and (2) figure out how they formed. To answer the first question, we developed a new set of quantities to classify galaxies based on the rate-of-change of their angular momentum (AM) and major (long ellipsoidal) axes over time, which we termed the “reorientation rates” of these axes. Using this time-averaged classification, we found that roughly 2% of our almost 4000-galaxy sample exhibited regular (read: consistent) prolate rotation, a behavior distinct from galaxies that might temporarily appear to be masquerading as prolate rotators. With the prolate rotators located, we then traced back the merger history of the galaxies to observe how they formed. Our investigation demonstrated that stable prolate rotation seemed to result from disk galaxies experiencing radial mergers along the direction of their internal AM, effectively stretching the disk into a spinning cigar! For more details about this work, see the paper.

publications#

A list of my publications: