Water vapor has no business existing near a B[e] supergiant. The star is too hot, the radiation too intense, the environment too harsh. Water molecules should dissociate instantly.
Researchers report the first detection of hot water vapor emission in a B[e] supergiant — LHA 115-S 18 in the Small Magellanic Cloud. High-resolution near-infrared spectra reveal rotationally broadened CO emission from a Keplerian molecular ring, strong hydrogen wind features, and — the surprise — H₂O emission lines that shouldn't be there.
Note (2026-11-10): The arXiv identifier in this essay's front matter, 2603.19946, is a different paper (an order-reversing embedding of Turing degrees into Arthur-Nimue-Merlin degrees). The paper the essay describes is arXiv:2603.22334 (Arias, Torres, Kraus and Cidale, high-resolution near-infrared spectroscopy of the B[e] supergiant LHA 115-S 18 with a discovery of hot water vapor emission). I checked its abstract and searched its text: it reports rotationally broadened CO emission from a Keplerian molecular ring, strong hydrogen wind features, the first detection of hot water vapor in a B[e] supergiant, and a radial velocity offset between molecular and Pfund line emission that supports a binary system with possibly circumbinary molecular gas. The essay's closing account of how tidal interactions and orbital motion protect the water is its own.
B[e] supergiants are massive stars surrounded by cool, dense circumstellar disks. The disks host complex chemistry: CO molecules, dust, metallic emission. These are understood. The disk provides shielding, the density provides stability, and the Keplerian rotation keeps material in place long enough for chemistry to proceed. But water requires conditions even more specific — the right temperature range, the right density, the right shielding from UV. The conventional wisdom says the disk can support CO but water is too fragile.
The detection overturns this. Water vapor survives in regions that are simultaneously cool enough for molecular bonding and dense enough for self-shielding, even in the immediate environment of a star that should destroy it. The key is that the disk is denser and more extended than models predicted — cool, dense pockets persist deeper into the circumstellar environment.
A radial velocity offset between the molecular emission and Pfund hydrogen lines suggests a binary system. The molecular gas may be circumbinary — orbiting both stars, not just one. If so, the binary dynamics create the geometry that protects the water: tidal interactions maintain the disk structure, orbital motion creates shielded zones, and the binary potential well traps material long enough for water chemistry to proceed.
The classical models said water couldn't survive here. The water disagrees.