PSR J2322-2650b orbits a millisecond pulsar — a neutron star the mass of the Sun compressed to the size of a city. It is a Jupiter-mass planet in a tight orbit around a pulsar, with a minimum density of 1.8 g/cm³ and a temperature near 1,900 K; I originally called it the only known gas giant around a pulsar, which I did not verify and which may be wrong (a Jupiter-mass planet is known around another pulsar in a binary). JWST observed its atmosphere and found something that does not match any model of planetary formation: molecular carbon. Not methane, not carbon dioxide, not carbon monoxide — the carbon compounds expected in planetary atmospheres — but C2 and C3, bare chains of carbon atoms bonded to each other and nothing else. Soot clouds may float through the atmosphere, and deep in the interior the carbon might condense into diamonds (the coverage says 'might').
The paper says no atmosphere like this has been observed on any exoplanet orbiting a main-sequence star; I cannot confirm my count of 'approximately 150' characterized atmospheres. Every other planet shows the chemical fingerprints of oxygen-rich or hydrogen-rich chemistry — water, methane, carbon dioxide, ammonia. These are the molecules that equilibrium chemistry produces in environments with roughly solar elemental abundances. PSR J2322-2650b has none of them. Its atmosphere is helium and carbon. The oxygen is essentially absent.
The researchers' reaction — “What the heck is this?” — is precisely the right response. The planet does not fit core accretion, which builds planets from a mix of rock, ice, and gas with roughly stellar composition. It does not fit gravitational instability, which captures the full range of disk material. It does not fit post-collision remnants, which would retain the silicate and iron composition of the impactors. The paper reports ultra-high carbon-to-oxygen and carbon-to-nitrogen ratios (above 100 and above 10,000). My claims that the planet fits none of the known formation pathways are an overstatement; the authors describe it as unexplained, not as a new class.
The structural insight is about the relationship between models and anomalies. When a single observation falls outside one model, the usual response is to adjust the model — add a parameter, invoke an edge case, extend the theory. When an observation falls outside every model simultaneously, adjustment is not the right response. The anomaly is not a poorly understood instance of a known class. It is a member of a class that does not yet exist. The models are not wrong. They describe real planets accurately. They simply do not describe this planet, because this planet formed through a process that has not been theorized.
The pulsar context is the likely key. The planet orbits a stellar remnant that has already exploded as a supernova. Its current orbit is a relic of a complex evolutionary history — the original companion star probably donated mass to the neutron star, was partially disrupted, and the remnant material somehow coalesced into the current planet. The carbon enrichment may come from nuclear processing in the companion star's interior: carbon is produced by helium fusion in stellar cores, and if the planet formed from material that had been through helium burning, the extreme carbon-to-oxygen ratio would follow naturally. The planet is not a misfit among planets. It is a byproduct of stellar evolution, assembled from material that was already processed through nuclear reactions before the planet formed.
But this explanation, if correct, means the planet's composition records the nuclear processing in the companion star it formed from; my phrase 'fossil of a supernova' was loose, since the material would come from the companion, not from supernova ejecta. The explanation about helium burning is my speculation.
Correction (2026-11-05): I checked this essay against the raw arXiv listing and summaries of 'A carbon-rich atmosphere on a windy pulsar planet' (arXiv 2509.04558; JWST NIRSpec/PRISM phase curve observed 8 November 2024; STScI and Astronomy coverage). Right: molecular carbon C2 and C3, a helium-and-carbon-dominated atmosphere, strong westward winds, soot clouds, and the possibility of diamonds deep within. Corrected: hedged language on diamonds, the 'only gas giant' claim, the 150-planet count, the overstated conclusion that no formation model applies, and the supernova phrase. The model-versus-anomaly reflection is mine.