RIKEN modelers derive 3I/ATLAS's extreme heavy water from a metal-poor birth cloud

The deuterium was locked in at about 10 kelvin, in gas too cold for most reactions to proceed ^[12]. Cosmic rays ionize the gas into H3+, which occasionally collides with an HD molecule and comes away as H2D+, leaving H2 and a little excess energy ^[13]. The reaction normally runs both ways. At 10 K the gas cannot supply the energy to push the deuterium back into inert HD, so it stays in H2D+ until an electron breaks that molecule apart, and the freed deuterium atoms meet oxygen ^[14].

Low metallicity helps at two points in this chain. Carbon monoxide destroys H2D+, and a metal-poor cloud has less CO, so H2D+ survives longer and yields more atomic deuterium ^[16]. In metal-rich clouds, ultraviolet photons split water into OH and H, flooding the gas with ordinary hydrogen and diluting the D/H ratio; a metal-poor cloud produces no such flood ^[17]. The paper gives three effects of this kind, all consequences of the same low metallicity ^[15].

Dividing the measured 1% by the 0.03% top of the cometary range gives 33, and by the 0.015% bottom gives 67, so the comet's water is tens of times richer in deuterium than anything comparable here ^[9]. The carbon anomaly is milder, between 1.4 and 2.1 times the solar system value ^[10]. It also sits above the interstellar medium average, higher than anywhere the object has traveled through in the past few billion years ^[6]. That points to formation as the source of the signature.

The economy of the result is its strength. One property of the birth cloud accounts for both anomalies, and low metallicity was already the accepted explanation for a high C-12 to C-13 ratio, so the paper extends a familiar idea to water ^[18][3]. The simulation demonstrates compatibility and does not measure the metallicity of the nursery; the case rests on a single object, the third interstellar visitor known ^[1][11]. The paper has been submitted to The Astrophysical Journal Letters and posted as a preprint, so no referee has passed on it yet ^[2].

The phys.org account credits the isotope measurements to observatories that turned to 3I/ATLAS after its July 2025 discovery, and it does not name the instruments or the teams ^[19]. Uncertainties matter for a result built on ratios, and the deuterium figure is given as around 1%, with the carbon ratio arriving as a range from 123 to 191 ^[7][6].