White dwarf mergers and the origin of R Coronae Borealis stars

We present a nucleosynthesis study of the merger of a 0.4 solar masses helium white dwarf with a 0.8 solar masses carbon-oxygen white dwarf, coupling the thermodynamic history of Smoothed Particle Hydrodynamics particles with a post-processing code. The resulting chemical abundance pattern, particularly for oxygen and fluorine, is in qualitative agreement with the observed abundances in R Coronae Borealis stars.

💡 Research Summary

The paper investigates the origin of hydrogen‑deficient, carbon‑rich stars—specifically R Coronae Borealis (RCB), hydrogen‑deficient carbon (HdC), and extreme helium (EHe) objects—by focusing on the double‑degenerate (DD) merger scenario in which a helium white dwarf (He WD) merges with a carbon‑oxygen white dwarf (CO WD). The authors perform three‑dimensional Smoothed Particle Hydrodynamics (SPH) simulations of a 0.4 M⊙ He WD merging with a 0.8 M⊙ CO WD, following the thermodynamic histories of 10 000 tracer particles that sample the hot corona surrounding the merger remnant (radial range 0.005–0.05 R⊙). These histories are post‑processed with a 327‑isotope nuclear network (H through Ga) using reaction rates from the REACLIB library, supplemented by recent experimental updates (e.g., Iliadis et al. 2010).

The remnant structure consists of a central hot core (essentially the primary CO WD), a hot corona composed mainly of disrupted secondary material with a small admixture of primary matter, and an outer, rapidly rotating accretion disk. The authors explore two mixing prescriptions for the corona: “deep” mixing (homogeneous mixing of the entire 0.005–0.05 R⊙ region) and “shallow” mixing (mixing only the outer 0.014–0.05 R⊙). Initial compositional stratification includes CO‑rich, He‑rich, and H‑rich shells in the CO WD (M_CO:CO = 0.78 M⊙, M_He:CO = 0.019 M⊙, M_H:CO = 0.001 M⊙) and analogous shells in the He WD (M_He:He = 0.399 M⊙, M_H:He = 0.001 M⊙). Both solar metallicity and a low‑metallicity case (Z = 10⁻⁵) are examined.

Key nucleosynthetic outcomes are as follows:

• Carbon – The final carbon abundance is high in all models and strongly dependent on the mixing depth. The trend with metallicity matches observed RCB data, but the absolute carbon mass fraction is higher than typical observations, suggesting that the model may over‑produce carbon or that additional dilution processes are missing.

• Carbon isotopes – The ¹³C/¹²C ratio for the solar‑metallicity case is ≈2 × 10⁻⁵, consistent with material that has undergone He‑burning, yet it exceeds the observed ratios in RCB stars, indicating possible limitations in the nuclear network or mixing assumptions.

• Oxygen – The ¹⁶O/¹⁸O ratio is 370 for deep mixing and 19 for shallow mixing. The shallow‑mixing value approaches the low ratios (high ¹⁸O) observed in some HdC and RCB stars, supporting the idea that ¹⁸O can be synthesized in the hot corona during the merger.

• Nitrogen – Nitrogen remains largely unchanged from its pre‑merger CNO‑processed level, as the short timescales of the merger do not allow significant destruction via ¹⁴N(α,γ)¹⁸F.

• Fluorine – Fluorine is produced throughout the corona, relatively insensitive to mixing depth, and reaches abundances higher than in the cold‑merger scenario. However, the modeled fluorine still falls short of the extreme enhancements measured in some RCB stars, implying that additional production channels or more extreme temperature conditions may be required.

• Neon – Neon shows no significant variation with mixing depth or metallicity, indicating that it is not processed during the merger and retains its original abundance set by the progenitor evolution.

The authors conclude that a “hot” merger, in which nuclear burning occurs during the coalescence, can qualitatively reproduce several hallmark abundance anomalies of RCB and HdC stars—most notably the fluorine enrichment and the presence of ¹⁸O. Nevertheless, discrepancies remain: carbon is over‑produced, fluorine is still under‑produced relative to observations, and neon cannot be raised to the observed levels. These gaps point to uncertainties in the initial shell masses, metallicity, reaction rate data, and the treatment of convective mixing in the post‑merger remnant.

Future work should incorporate higher‑resolution 3‑D hydrodynamics, an expanded nuclear network (including more neutron‑capture pathways), and a systematic exploration of progenitor metallicities and shell thicknesses. Such refinements will help determine whether the hot DD merger scenario can fully account for the complex chemical signatures of hydrogen‑deficient, carbon‑rich stars, or whether additional astrophysical processes (e.g., late thermal pulses, binary interaction histories) must be invoked.