Could nova explosions explain the chemistry of phosphorus-rich stars?

A chemical puzzle in ancient star clusters

Phosphorus-rich stars have been identified both in the Milky Way’s general stellar population and in globular clusters, dense stellar systems whose formation histories remain debated. The preprint cites three confirmed cluster cases. Their unusual chemistry raises a basic question: what supplied so much phosphorus to the gas from which they formed?

Astronomers express these abundances using notation such as [P/Fe], which compares a star’s phosphorus-to-iron ratio with the Sun’s on a logarithmic scale. A value of 1 means ten times the solar ratio. The study defines phosphorus-rich stars as having [P/Fe] of at least 1.

Bekki and Tsujimoto investigate enrichment by oxygen–neon novae: outbursts in binary systems containing oxygen–neon white dwarfs. Their proposed setting is a giant molecular cloud forming a globular cluster, surrounded by an older stellar population. Nova ejecta enter the cloud, mix with some of its cold gas, and supply material for new stars.

The work was posted as an arXiv preprint on November 6, 2026; its landing record states that it has been accepted for publication in Monthly Notices of the Royal Astronomical Society.

Testing a mixture, not simulating the whole process

The central calculations are idealized analytical mixing models. The authors combine the starting composition of cloud gas with published predictions of the elements expelled by seven nova types. They examine both single-type enrichment and mixtures of different nova types, with randomly assigned contributions in the composite models.

Their main control is the dilution factor: the mass of initially unpolluted gas divided by the mass of nova ejecta mixed into it. A smaller factor means more concentrated enrichment. Most calculations begin with gas containing about one-tenth the solar iron-to-hydrogen ratio and explore dilution factors from 10 to 10,000.

The benchmark nova yields were calculated for solar metallicity, rather than the lower metallicity typical of the stars being investigated. Earlier calculations suggest phosphorus production may depend only moderately on metallicity, but suitable yield tables are missing. The authors therefore also test phosphorus yields reduced to 30% and 10% of the benchmark values.

Promising patterns, incomplete matches

Many models predict that stronger phosphorus enrichment accompanies higher nitrogen, oxygen, aluminium and silicon abundances relative to iron. Silicon rises more strongly than sulfur, producing a positive relationship between phosphorus enrichment and the silicon-to-sulfur ratio—one feature seen in the observations used for comparison.

Among the single-type calculations, the model labelled ONe5 provides the best overall match. However, no one phosphorus-yield choice fits every element equally well. Composite models cover more of the observed chemical diversity, but discrepancies remain: some very oxygen-rich stars are not reproduced, the predicted sulfur spread is too narrow, and certain mixtures produce phosphorus abundances higher than those observed.

Nova enrichment alone also cannot explain enhanced heavy elements such as cerium and barium. The authors add a second proposed source: winds from lower-mass asymptotic giant branch stars. With suitable mixing proportions, these calculations can reproduce substantial cerium enrichment. Whether both sources can enrich the same star-forming gas in the required proportions remains untested.

The challenge of catching fast ejecta

The scenario faces a physical tension. Too much cloud gas dilutes the phosphorus below the required abundance. Too little may fail to slow the ejecta enough for the cloud’s gravity to retain them.

In a simplified momentum-conservation calculation for a cloud of three million solar masses, the ONe5 case requires a dilution factor of at least about 196 for retention, but no more than about 286 for phosphorus-rich star formation. That narrow overlap illustrates the constraint, rather than establishing that real clouds achieve it.

The retention model does not follow shock cooling or detailed gas dynamics. Earlier simulations of dwarf galaxies help estimate how often clouds encounter potential nova systems, but do not simulate the full enrichment process. Those estimates favour massive clouds in compact, gas-rich dwarf galaxies.

An isotope test

The models predict unusually low nitrogen-14-to-nitrogen-15 and oxygen-16-to-oxygen-17 ratios—below about 20 and 400, respectively, in the highlighted cases. These are predictions, not measurements of phosphorus-rich stars. Detecting such signatures in main-sequence stars would support the scenario, although the measurements would be exceptionally difficult.

The proposed origin of phosphorus-rich field stars in disrupted globular clusters is likewise an assumption, not an established result. The preprint offers a chemically promising explanation and observational tests; detailed simulations must still show whether the necessary capture, cooling and localized mixing can actually occur.