The pollution from massive AGB stars favoured by strong hot bottom burning

Technical Summary: The Pollution from Massive AGB Stars Favoured by Strong Hot Bottom Burning

Problem Statement
The paper addresses significant uncertainties in the modeling of intermediate-mass stars (4–8 $M_{\odot}$) evolving through the Asymptotic Giant Branch (AGB) phase, specifically those undergoing "Hot Bottom Burning" (HBB). While HBB is a critical mechanism for the nucleosynthesis of light elements (N, Al, Mg) and the chemical evolution of globular clusters (GCs), current models yield drastically different predictions due to poorly constrained physical inputs. Key uncertainties include the description of over-adiabatic convection and the lack of a first-principles theory for mass loss. These uncertainties lead to elemental yield variations of factors greater than 10 between different model sets. Furthermore, recent observations of high-redshift protogalaxies (NOEGs) reveal unexpectedly high Nitrogen-to-Oxygen (N/O) ratios in gas near active galactic nuclei (AGN), a feature not predicted by standard chemical evolution models. The paper investigates whether specific AGB models, characterized by strong HBB and high mass-loss rates, can simultaneously explain the chemical anomalies of second-generation (2G) stars in GCs and the high N/O ratios observed in these distant galaxies.

Methodology
The authors employ a comparative analysis of stellar evolutionary models for the upper mass range of AGB stars (4–8 $M_{\odot}$), including "super-AGB" stars. The study focuses on two primary physical parameters:

1. Convection Modeling: Comparing the standard Mixing Length Theory (MLT) against the "Full Spectrum of Turbulence" (FST) model. The FST model is noted for producing higher temperatures at the base of the convective envelope ($T_{HBB}$), which is crucial for efficient proton-capture nucleosynthesis.
2. Mass Loss Formulations: Contrasting mass-loss prescriptions with different dependencies on stellar luminosity. The authors utilize a formulation with a high power dependence on luminosity (Ventura et al. [5]) which limits the total number of thermal pulses and third dredge-up (3DU) episodes, versus formulations that allow for more extensive 3DU (Vassiliadis and Wood [67]).

The study synthesizes these models to generate elemental yields (specifically N/O, C/O, and He) and compares them against:

• Observational constraints from 2G stars in globular clusters (helium limits, Lithium abundances, Mg depletion).
• Spectroscopic data from the high-redshift galaxy GS 3073 ($z=5.5$), particularly the N/O ratios in the dense central regions near the supermassive black hole.

Key Contributions and Results

• Model Sensitivity and Yield Discrepancies: The paper demonstrates that the choice of convection and mass-loss models fundamentally alters the predicted chemical yields. Models employing FST convection and high-luminosity-dependent mass loss (V13 models) achieve strong HBB, depleting Oxygen and Magnesium while producing primary Nitrogen. In contrast, models using MLT and standard mass loss (F14 models) experience extensive third dredge-up, resulting in Carbon-rich envelopes (C/O > 1) and significantly lower N/O ratios. The authors show that the F14 models fail to reproduce the high N/O ratios observed in specific high-redshift environments.
• Constraints on Globular Cluster Formation: The authors argue that the AGB scenario for 2G formation is strongly supported by specific observational constraints that alternative models (such as those involving Very Massive Stars or Extremely Massive Stars) struggle to satisfy simultaneously:
  • Helium Limit: Observations of 2G stars show a helium mass fraction limit of $\Delta Y \approx 0.1$ (max $Y \sim 0.38$). AGB models naturally predict this upper limit due to the specific timing of the second and third dredge-ups, whereas massive star models often predict unbounded helium enrichment.
  • Lithium Abundances: The AGB model uniquely allows for the production of fresh Lithium via the Cameron-Fowler mechanism during the HBB phase, potentially explaining high Lithium content in specific "extreme" giants (e.g., in NGC 2808) that would otherwise be destroyed in massive star scenarios.
  • Magnesium Depletion: The models correctly predict that Mg depletion is metallicity-dependent, occurring primarily in low-metallicity clusters where $T_{HBB}$ is highest, consistent with data from clusters like $\omega$ Cen and NGC 2419.
  • Type II Clusters: The paper offers a coherent explanation for "Type II" clusters (e.g., NGC 1851, M 22) which exhibit s-process and CNO enrichment alongside iron spreads. This is attributed to a delayed star formation burst where pristine gas is mixed with AGB ejecta enriched by effective third dredge-up, occurring after the initial 2G formation but before Type Ia supernovae terminate star formation.
• High-Redshift N/O Anomalies: The study connects the high N/O ratios observed in the central regions of NOEGs (like GS 3073) to the intermittent super-Eddington accretion phases of central black holes. The authors propose that during quiescent phases between accretion bursts, the central region is replenished by winds from a nearby Nuclear Star Cluster (NSC) dominated by massive AGB stars. These AGB winds, modeled with strong HBB, provide the pure, high-N/O ejecta required to match the observations. This scenario suggests the high N/O is a temporary feature of the AGN growth phase rather than a permanent signature of 2G formation in a standard GC.

Significance and Claims
The paper claims that the "strong HBB" AGB models (utilizing FST convection and high mass-loss rates) provide a consistent framework for two distinct astrophysical problems: the chemical composition of multiple populations in globular clusters and the anomalous N/O ratios in the centers of high-redshift galaxies.

The authors assert that the success of these models in explaining the helium limits, Lithium presence, and metallicity-dependent Mg depletion in GCs validates their use for interpreting high-redshift data. Conversely, they argue that models failing to produce strong HBB (and thus failing to deplete Oxygen and enhance Nitrogen sufficiently) cannot explain the high N/O ratios in NOEGs. The work suggests that the high N/O in these distant galaxies is likely a transient signature of AGB pollution during the growth of supermassive black holes, potentially leading to complex abundance patterns (including iron spreads) more akin to systems like $\omega$ Cen than to standard globular clusters. The paper concludes that experimental verification of which AGB models are most appropriate is critical for understanding both the early chemical evolution of the universe and the formation history of globular clusters.