M dwarfs
M dwarfs are three-quarters of the stars in the Milky Way, live for tens of billions of years, and lock in the composition of the gas from which they formed. Their cool atmospheres are packed with molecular lines — inconvenient for some abundance work, but ideal for CO isotopologues. If we want to know how $^{13}$C and $^{18}$O built up in the solar neighbourhood over time, these stars are the natural census.
That programme rests on careful stellar characterisation. High-resolution infrared spectrographs such as SPIRou now deliver the S/N and wavelength coverage needed for detailed atmospheric fits. Cristofari et al. (2022) showed how SPIRou spectra of 44 M dwarfs can yield $T_{\mathrm{eff}}$, $\log g$, $[\mathrm{M/H}]$, and even $[\alpha/\mathrm{Fe}]$ when compared to MARCS models — refining line lists and demonstrating that α-enhancement couples into the other parameters. Earlier isotope work on M dwarfs (Crossfield et al. 2019) already hinted that rare C and O isotopes were measurable from the ground. What was still missing was a metallicity baseline large enough to test Galactic chemical evolution (GCE) models: primary $^{12}$C and $^{16}$O from helium burning versus secondary $^{13}$C from the CNO cycle, with novae as a debated late-time $^{13}$C source.
Rare isotopes across the solar neighbourhood
In González Picos, Snellen & de Regt (2025, Nature Astronomy) we analysed archival SPIRou K-band spectra of 32 nearby M dwarfs (within $\sim15\,\mathrm{pc}$), spanning $[\mathrm{M/H}]$ from about $-0.4$ to $+0.4$. We detect $^{13}$CO at $>3\sigma$ in 29 of 32 stars, and $\mathrm{C}^{18}\mathrm{O}$ in a subset. The headline result is simple: more metal-rich stars have lower $^{12}\mathrm{C}/^{13}\mathrm{C}$, tracking progressive $^{13}$C enrichment of the ISM. Super-solar stars approach present-day ISM-like ratios ($\sim68$), while $^{16}\mathrm{O}/^{18}\mathrm{O}$ in the most metal-rich objects can fall well below solar — in line with theoretical yields. Fits with petitRADTRANS and FastChem reach $\sim1$–$2\%$ relative precision across the CO band.
This is the stellar counterpart to the planetary isotope work. When a super-Jupiter shows $^{12}\mathrm{C}/^{13}\mathrm{C}\sim50$–$80$, that number only becomes a formation diagnostic once we know what the local Galactic mixture looked like at the relevant metallicity. The same logic applies inside systems: GQ Lup A’s isotope ratio matches its companion (SupJup IV), tying stellar and substellar chemistry to a shared reservoir.
Looking ahead, extending this census to lower metallicity with ELT/METIS and JWST would test GCE where secondary enrichment is expected to be weakest — the next natural step beyond the SPIRou solar-neighbourhood sample.