betaFeedback

WorksPhysics and Astronomy

Expanding the Volatile Abundance Inventory for High-Priority K-Dwarf JWST Exoplanet Systems

Abstract

There are exceedingly few chemical abundance measurements available in the literature for K dwarf stars, especially for those with Teff<_{eff} < 4700 K due to an incompatibility of traditional analyses with their molecular-line-ridden optical spectra. To address the need for K dwarf abundance constraints, we present an elemental ([C/H], [O/H]) and isotopic (12^{12}C/13^{13}C, 16^{16}O/18^{18}O) abundance analysis for six K dwarf exoplanet host stars: K2-141, NGTS-10, WASP-43, WASP-69, WASP-80, and HATS-72. Each system hosts an exoplanet targeted by JWST characterization efforts, making host star abundances especially critical as baseline references. Using MARCS model atmospheres, the TurboSpectrum spectral synthesis code, and near-infrared IGRINS spectra, we derive novel [C/H] and [O/H] abundance ratios for these high-priority K dwarf host stars. We further investigate prospects for determining 12^{12}C/13^{13}C and 16^{16}O/18^{18}O ratios using 13^{13}C16^{16}O and 12^{12}C18^{18}O lines from the 2.3 micron CO bandhead. The observed spectra are too noisy to measure 16^{16}O/18^{18}O ratios for the sample, but we do find sub-solar 12^{12}C/13^{13}C ratios for five out of six targets. We also derive uniform gyrochronological ages for each host star to compare with galactic chemical evolution models and identify underlying abundance-age trends. NIR H+K band spectroscopy provides an alternative to optical, atomic line-derived carbon and oxygen abundances for cool dwarf stars. The growing catalog of precise host star elemental and isotopic abundances complements ongoing observations of their companions and provides a baseline for understanding their chemical composition, origin, and evolution.

Introduction

Stellar chemical composition plays a central role in understanding planet formation, as stars and their planets form from the same protoplanetary disk. Higher bulk stellar metallicities are strongly linked to increased occurrence rates of close-in Jupiter-sized planets ((J. A. Valenti & D. A. Fischer 2005; E. M. Johnson 2010; A. Mortier et al. 2013; T. Gan et al. 2025)), and to a lesser extent, sub-Neptunes and super-Earths ((J. Wang & D. A. Fischer 2015; E. A. Petigura et al. 2018; K. M. Boley et al. 2024)). Weaker trends have also been reported for individual refractory elements, with subsolar-metallicity stars hosting gas planets showing enhanced refractory-to-iron abundance ratios compared with stars without detected planets ([3]). The abundances of volatile elements such as C, O, N, and S in both stellar and planetary atmospheres provide insight into planet formation mechanisms and chemical evolution ((M. Bedell et al. 2018; J. M. Brewer & D. A. Fischer 2016; J. J. Fortney 2012; E. Delgado Mena et al. 2021; D. Turrini et al. 2021; I. J. M. Crossfield 2023)). Carbon and oxygen abundances, in particular, can be used to differentiate between gravitational instability which yields stellar or super-stellar C/O ratios in brown dwarfs (J. D. Ilee et al. 2017; K. Hawkins et al. 2020) while core accretion results in a more diverse range of C/O values in Jupiter-class exoplanets (J. W. Xuan et al. 2024). As an example, N. Hejazi et al. (2025) shows that the C/O ratio of the young M-type T Tauri star DH Tau A is fully consistent with that of its super-gas giant companion DH Tau Ab, suggesting that the companion likely formed through rapid gravitational collapse rather than a slower core accretion process. This formation scenario is further supported by the system’s young age and relatively large companion-to-star mass ratio.

The C/O ratio was once considered the ideal formation and migration tracer for exoplanets (K. I. Öberg et al. 2011; N. Madhusudhan 2012): planets forming within the H2OH_2O and CO snowlines retain stellar C/O ratios, while those forming farther out and migrating inward display a range of volatile enhancements for various reasons (N. Madhusudhan et al. 2014, 2017; A. D. Schneider & B. Bitsch 2021a,b; B. Bitsch et al. 2022). Today however, it is clear that a direct inversion from atmospheric abundances to planet formation models is quite difficult and such inferences rely heavily on our understanding of volatile delivery mechanisms and protoplanetary disk chemistry (C. Mordasini et al. 2016; P. Mollière et al. 2022). Planetesimal impacts and pebble accretion during planet formation greatly influence the final atmospheric composition of the gas giant exoplanets, as does the chemical evolution of the protoplanetary disk (C. Mordasini et al. 2016; P. Mollière et al. 2022). This demonstrates that chemical inheritance from the host star is only one area requiring careful consideration to connect planet formation mechanisms with observations of exoplanet atmospheres.

Recently, star-planet connection studies have been extended beyond the C/O ratio to include other volatile abundance ratios like C/N, N/O, and S/N which also offer diagnostics for a planet’s formation location and migration history (E. Pacetti et al. 2022; K. I. Öberg et al. 2023). In the directly imaged HR 8799 system, J. W. Xuan et al. (2026) finds a tentative increase of planetary S/H with increasing orbital distance and also implicates NH3NH_3 abundances as a formation location tracer relative to the CO and N2N_2 snowlines. Other studies suggest that N/S and C/O (K. Nakazawa & K. Ohno 2026) or C/S and O/S (I. J. M. Crossfield 2023; A. Baburaj et al. 2025) can disentangle various accretion scenarios (e.g. gas vs. solid, pebble vs. planetesimal accretion respectively) and are thus useful for testing planet formation models. This sort of testing requires complementary stellar abundance measurements to serve as a baseline for interpreting planetary chemical abundances (D. R. Coria et al. 2024; A. Baburaj et al. 2025).

With powerful instruments like the James Webb Space Telescope and ground-based high-resolution spectrographs, astronomers can now detect even faint isotopologue lines from 13C{}^{13}\mathrm{C} O, C 18O{}^{18}\mathrm{O}, H218OH_2{}^{18}\mathrm{O}, and 15NH3{}^{15}\mathrm{N}H_3 in stars, brown dwarfs, and exoplanets (M. R. Line et al. 2021; Y. Zhang et al. 2021b,a; R. B. Botelho et al. 2020; D. R. Coria et al. 2023; D. González Picos et al. 2025a; I. J. M. Crossfield et al. 2019; J. W. Xuan et al. 2024; J. C. Costes et al. 2024; C. E. Hood et al. 2024; S. de Regt et al. 2024; S. Gandhi et al. 2023; F. Finnerty et al. 2024; D. González Picos et al. 2026a). These isotopic measurements, especially for 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} and 16O/18O{}^{16}\mathrm{O}/{}^{18}\mathrm{O}, provide an additional tool for distinguishing between formation pathways and tracing the chemical evolution of planetary systems (P. M. Woods & K. Willacy 2009; S. K. Atreya et al. 2016; Y. Zhang et al. 2021b,a; D. Barrado et al. 2023).

Motivation

K Dwarf Stars and Their Companion Exoplanets

K dwarf host stars and their companion exoplanets are undeniably important to various aspects of exoplanet science, but they are not as widely studied as other dwarf stars. K dwarfs are more abundant than the best studied G-type stars, comprising ∼12%\sim12\% of the total stellar population compared with ∼8%\sim8\% for G stars ([10]). One study finds that 4.0%±2.3%4.0\% \pm2.3\% of late K dwarf stars (masses of 0.50.5-0.750.75 M⊙M_{\odot} and effective temperatures of 3900-4800 K) host Saturn-mass or larger planets with orbital periods <245< 245 days (E. Gaidos et al. 2013). Furthermore, the sensitivity of exoplanet detection techniques like the radial velocity method and the transit method scales inversely with stellar mass and radius, such that smaller planets can be detected around smaller, less massive stars (E. Gaidos et al. 2013). Radial velocity measurements rely on strong, numerous absorption lines that are prominent in the spectra of cool dwarf stars and scarce in hotter stars. The planet-to-star radius ratio (RP/R⋆R_P/R_\star) is also larger for planets orbiting smaller, cool dwarf stars which ultimately produces deeper, more detectable transits. K dwarf stars are also significantly less active than M dwarfs which not only makes these stars more amenable to RV and transit surveys (H. Isaacson & D. Fischer 2010), but also provides a more stable environment for companion planets (T. Richey-Yowell et al. 2019). These positive attributes clearly make K dwarf host stars ideal candidates for characterization. So this raises the question, “why are K dwarf stars so neglected?”К

Late K-type stars in particular are left out from stellar, planetary, and chemical abundance surveys because they are faint and not M dwarfs. M dwarfs tend to spark more general interest because they are preferable targets for the search and characterization of habitable worlds due to their greater numbers and close-in habitable zones (E. Gaidos et al. 2013; T. Richey-Yowell et al. 2019). Most stellar surveys are flux- or magnitude-limited at visible wavelengths, meaning brighter, late F to early K spectral type stars are the optimal targets over late-type K and M dwarfs. Furthermore, deriving chemical abundances for cooler stars using traditional optical spectroscopy methods becomes more complicated as these stars contain millions of optical molecular lines, typically metal hydrides for K dwarfs and metal oxides for M dwarfs, that blend and obscure atomic signatures of interesting species. This is why established radial-velocity surveys like the California Planet Search (CPS) and the HARPS survey demonstrate a notable lack of late K stars (H. Isaacson & D. Fischer 2010; F. Pepe et al. 2000; R. Cosentino et al. 2012) as do major optical chemical abundance surveys (J. M. Brewer et al. 2016; M. Bedell et al. 2018). Even state-of-the-art, optical EPRV spectrographs like ESPRESSO, EX-PRES, and NEID observe only a handful of the hotter, early-type K dwarfs. Near-infrared instruments like CARMENES, SPIRou, and NIRPS, on the other hand, have a preference for M dwarfs instead with few dedicated K dwarf surveys like the KOBE (K-dwarfs Orbited By habitable Exoplanets) experiment (J. Lillo-Box et al. 2022; O. Balsalobre-Ruza et al. 2025). A similar challenge persists in IR spectra of cool dwarfs, where millions of background molecular lines, dominated by H2OH_{2}O, blend with and obscure many potentially useful atomic and molecular abundance diagnostics. Even recent abundance surveys considering JWST host stars exclusively like J. R. Kolecki & J. Wang (2022) and A. S. Polanski et al. (2022) covering Cycle 1 targets and Q. Sun (2026) covering Cycle 2 targets do not include more than one target (18 Indi for J. R. Kolecki & J. Wang (2022); WASP-69 for A. S. Polanski et al. (2022); Q. Sun (2026)) with effective temperatures <4800< 4800 K in their sample.

Stars hosting giant exoplanets are also considered high-priority targets because of the influence that giant exoplanets have on their planetary siblings. Gas giant exoplanets are thought to play a crucial role in the formation of their planetary systems as they facilitate the formation of planetary cores, alter protoplanetary disk chemistry by delivering volatile-rich planetesimals to the inner disk, and modify the influx of asteroids and comets to young planets (K. Batygin & G. Laughlin 2015; N. Georgakarakos et al. 2018; M. L. Bryan & E. J. Lee 2024). Such investigations extend to multi-planet systems hosting warm-Jupiters as well. Warm Jupiters keep companions at rates comparable to smaller-planet systems, while hot Jupiters do not (C. Huang et al. 2016). However, these companions almost always lie interior to the warm Jupiter’s orbit, and the habitable zone to date is unprobed at terrestrial-planet sensitivity.

Furthermore, gas giant exoplanets to date are more observable than their rocky exoplanet companions. Several ground-based observatories are dedicating significant time to the characterization of exoplanet atmospheres using high-resolution cross-correlation spectroscopy techniques (M. Yakubu 2026). Observing campaigns like the VLT/ESPRESSO GTO program, ESO/NIRPS GTO program, the “Roasting Marshmallows” program with Gemini/IGRINS/IGRINS-2 (M. Brogi et al. 2023; P. C. B. Smith et al. 2024; V. Panwar et al. 2025) and the “ESO SupJup Survey” with VLT/CRIRES+ (S. Gandhi et al. 2025; Y. Zhang et al. 2024; D. González Picos et al. 2025b, 2026a) reveal detections of various molecules, minor isotopologues, and other atmospheric properties. Likewise, JWST observations have found a wealth of information on the composition, chemistry, structure, etc. of their atmospheres as discussed in the following section. Meanwhile, JWST surveys of sub-Neptunes and smaller rocky exoplanets (e.g. the JWST COMPASS and Rocky Worlds programs) continue to reveal flat, featureless spectra—in some cases expected for bare rock planets [1] (Q. Xue et al. 2025; J. Lustig-Yaeger et al. 2023; J. Kirk et al. 2024; N. Scarsdale et al. 2024; E. M. May et al. 2023; R. Luque et al. 2025) or otherwise flattened or muted by unocculted starspots, aerosol clouds/hazes or high mean molecular weight atmospheres [5, 9, 11, 5, 7] (S. E. Moran et al. 2023; A. Meech et al. 2026; N. L. Wallack et al. 2024; J. Teske et al. 2025; E. M.-R. Kempton et al. 2023; E. Schlawin et al. 2024a; K. Ohno et al. 2025; P.-A. Roy et al. 2026; C. Piaulett-Ghorayeb et al. 2024; S. Mukherjee et al. 2025). In these cases, the nature of the atmospheres and interiors of terrestrial exoplanets are best inferred from the properties of their host stars and gas giant companions.

Elemental and Isotopic Abundances of K Dwarf Stars

Despite significant progress over the past two decades in linking stellar chemical compositions to the properties of their orbiting planets, there remains a notable lack of such studies focused on low-mass stars—specifically latetype K and M dwarfs. These stars are intrinsically faint which means obtaining high-resolution, high signal-to-noise spectra suitable for abundance analyses requires substantial observing time on large, 8-meter-class telescopes. The challenges are compounded by the complex spectra of cool dwarf stars, which are dominated by molecular line blending, making precise abundance measurements especially difficult (N. Hejazi et al. 2023, 2024). As a result, only a limited number of studies report detailed elemental abundances for low-mass stars (D. Souto et al. 2017, 2018, 2022; N. Hejazi et al. 2023, 2024; E. Melo et al. 2024; F. Jahandar et al. 2025) in contrast to the extensive analyses and significantly larger samples available for hotter F, G, and early K-type stars (J. M. Brewer et al. 2016; M. Bedell et al. 2018; E. Delgado Mena et al. 2021; J. R. Kolecki & J. Wang 2022; A. S. Polanski et al. 2022).

Nevertheless, given the high occurrence rate of exoplanets around low-mass stars (E. Gaidos et al. 2013; K. Ment & D. Charbonneau 2023; Kaminski, A. et al. 2025), expanding high-resolution spectroscopic studies of these hosts is essential for advancing our understanding of planet formation and composition. While recent efforts have begun to explore the chemical properties of hotter stars hosting high-priority JWST exoplanet targets (J. R. Kolecki & J. Wang 2022; A. S. Polanski et al. 2022), a majority of high-priority host stars have little to no chemical abundance constraints in the literature. Nearly half of the JWST exoplanet host star sample consists of cooler stars with effective temperatures below 4700 K. Some of these targets have archival high-resolution spectra that could be useful for measuring chemical abundances, but many will have to rely on ongoing or future observations. As JWST and state-of-the-art ground-based observatories continue to deliver precise planetary abundance measurements, parallel spectroscopic analyses of cool dwarf host stars will provide critical context for understanding the processes that shape planetary atmospheres.

The paucity of elemental abundance constraints, even for high priority K dwarf exoplanet hosts, can be addressed moving forward using near-infrared H and K band spectroscopy covering wavelengths of 1.5 - 2.5 microns. There are significantly fewer chemical species accessible here; however, recent studies focused on deriving abundances for K and M dwarf stars have successfully constrained the abundances of up to 15 species (C, N, O, Na, Mg, Al, Si, K, Ca, Ti, V, Cr, Mn, Fe, Ni) for stars with effective temperatures << 4700 K (D. Souto et al. 2022; N. Hejazi et al. 2023, 2024). Notably, those abundance studies and earlier works like T. Tsuji et al. (2015) utilize molecular OH, CO, and H2_2O lines found throughout the H and K bands respectively to constrain carbon and oxygen abundance ratios crucial for contextualizing the yields from exoplanet atmosphere studies.

Besides elemental abundances, near-infrared spectroscopy may also be the key to determining 12{}^{12}C/13{}^{13}C and 16{}^{16}O/18{}^{18}O ratios for cool dwarf stars. Recent studies have used lines in the fundamental (v=1v = 1 to v=0v = 0 transitions) rovibrational CO band located in the M band at 4.5 - 5.0 microns to derive 12{}^{12}C/13{}^{13}C and 16{}^{16}O/18{}^{18}O ratios for M dwarfs (I. J. M. Crossfield et al. 2019; D. González Picos et al. 2026b) and solar twins (D. R. Corea et al. 2023) respectively. However, because K dwarfs are intrinsically faint, it is impractical to obtain sufficiently high signal-to-noise, high-resolution spectra at these wavelengths in a reasonable amount of telescope time, even on large, 8m-class telescopes. Another approach is to turn to the first overtone (v=2v = 2 to v=0v = 0 transitions) rovibrational band of CO located at 2.3 - 2.45 microns (T. Tsuji & T. Nakajima 2014, 2016; T. Tsuji 2016; D. González Picos et al. 2025a). Here the lines are weaker than in the M band, but K and M dwarfs are also brighter at these near-infrared wavelengths. Combined with the high-throughput of current near-infrared spectrographs like the Immersion GRating INfrared Spectrometer (IGRINS) (C. Park et al. 2014; G. Mace et al. 2018), it suddenly becomes more practical to obtain a sufficiently high signal-to-noise, high-resolution K band spectrum making this wavelength regime a viable option for measuring isotopic carbon and oxygen abundance ratios.

In this study, we use high-resolution (R∼45,000R \sim45{,}000), high signal-to-noise IGRINS spectra to first constrain elemental carbon and oxygen abundances for the selected sample of K dwarf stars. Then, using these new elemental carbon and oxygen constraints, we investigate whether the isotopic abundance ratios, 12{}^{12}C/13{}^{13}C and 16{}^{16}O/18{}^{18}O, can also be inferred for these targets.

In § 3, we describe the K dwarf exoplanet hosts selected for this abundance analysis including an overview of their giant exoplanet companions and other relevant studies. In § 4, we provide a brief description of the observations, spectral reduction, final stellar spectra used in this analysis. Then, in § 5 and § 6, we detail the methodology used to generate synthetic spectra and measure elemental and isotopic abundance ratios for these K dwarf host stars. Finally, in § 7, we derive gyrochrone-based ages for our stellar sample and, along with the measured abundances, contextualize them with several galactic chemical evolution models.

The K Dwarf Sample

The cool dwarf stars selected for this elemental and isotopic abundance analysis are classified as K spectral type, main sequence stars with effective temperatures 4000 K <Teff<< T_{\mathrm{eff}} < 4800 K. These stars also host exoplanets that are ideal candidates for atmospheric characterization via transmission spectroscopy, emission spectroscopy, or high-resolution cross-correlation spectroscopy (L. Kreidberg 2018; M. Brogi & M. R. Line 2019), and have the most to gain from measurements of their host star’s carbon and oxygen abundances. The majority are giant exoplanets with Rp>0.6RJR_{\mathrm{p}} > 0.6R_{\mathrm{J}}, but we also include a highly irradiated lava world to provide an alternative scenario for using host star abundances to contextualize exoplanet observations. Finally, the selected stars also have high-resolution, high signal-to-noise H+K band IGRINS spectra available through the Raw & Reduced IGRINS Spectral Archive (E. Sawczynec et al. 2025) and via the following observing programs: GS-2023B-LP-108 (PI: Crossfield), GS-2024A-LP-108 (PI: Crossfield), and GS-2023A-Q-203 (PI: Crossfield).

The sample consists of six exoplanet host stars: K2-141, NGTS-10, WASP-43, WASP-69, WASP-80, and HATS-72. This includes exoplanets discovered by the Wide Angle Search for Planets (WASP) collaboration (D. L. Pollacco et al. 2006) that have been extensively studied for years but also some newer discoveries like K2-141 and NGTS-10 that are still not well characterized. The properties of the selected K dwarf host stars and their planetary companions are discussed further in the following section. Refer to Table 1 for a list of stellar parameters.

ParameterK2-141NGTS-10WASP-43WASP-69WASP-80HATS-72
Spectral TypeK7VK5VK7VK5VK7VK4V
K Mag8.40114.2619.677.4598.3519.764
G Mag10.718714.314511.89369.491511.268512.0646
Parallax (mas)16.13033.871411.474019.885820.11417.8880
Radius (R⊙R_{\odot})0.68 ± 0.020.70 ± 0.040.65 ± 0.010.86 ± 0.030.59 ± 0.020.72 ± 0.01
∗Teff{}^{*}T_{\mathrm{eff}} (K)4571 ± 704600 ± 1504308 ± 764748 ± 404143 ± 944656 ± 9
∗log⁡g{}^{*}\log g4.62 ± 0.034.59 ± 0.024.55 ± 0.114.50 ± 0.074.66 ± 0.024.59 ± 0.01
∗[Fe/H]{}^{*}[\mathrm{Fe}/\mathrm{H}]−0.03 ± 0.08−0.02 ± 0.12+0.12 ± 0.12+0.24 ± 0.02−0.13 ± 0.17+0.10 ± 0.01
vsin⁡iv\sin i (km/s)3.0 ± 1.74.0 ± 0.62.26 ± 0.542.20 ± 0.401.27 ± 0.170.80 ± 1.30
References(1, 7)(2)(3)(4, 8, 9)(5)(6)

Table 1. Stellar Parameters. Parameter search conducted using the NASA Exoplanet Archive (J. L. Christiansen et al. 2025). Spectral types and KK-band magnitudes from 2MASS (R. M. Cutri et al. 2003). GG-band magnitudes and parallaxes from Gaia DR3 (Gaia Collaboration et al. 2016, 2023). (1) L. Malavolta et al. (2018); (2) J. McCormac et al. (2020); (3) M. Esposito et al. (2017); (4) K. G. Stassun et al. (2017); (5) A. H. M. J. Triaud et al. (2015); (6) J. D. Hartman et al. (2020); (7) O. Barragán et al. (2018); (8) A. S. Bonomo et al. (2017); (9) D. R. Anderson et al. [4]. A star* indicates values adopted in this abundance analysis. Refer to Table 4 for the original values and references.

The K2-141 System

K2-141 (EPIC 246393474) is a relatively bright (Kmag = 8.4), active K7V dwarf star about 68% the size of the Sun with an effective temperature of approximately 4571 K (L. Malavolta et al. 2018; O. Barragán et al. 2018). K2-141 is located ∼61\sim61 pc away in direction of the constellation Pisces. This K dwarf star is host to K2-141 b, an ultra-short period (USP) super-Earth, and an outer companion K2-141 c, a Neptune-like planet, both discovered in 2018 by two independent studies using observations of the Kepler telescope during its extended “second light” mission, K2 (L. Malavolta et al. 2018; O. Barragán et al. 2018).

K2-141 b orbits its host every 0.28 days (6.7 hours), while K2-141 c has an orbital period of 7.7 days (L. Malavolta et al. 2018; O. Barragán et al. 2018). Multiple transit and radial velocity studies studies have derived consistent parameters for this planet: a radius of approximately Rp=1.51±0.05 R⊕R_{\mathrm{p}} = 1.51 \pm0.05\ R_{\oplus} and mass of Mp=5.08±0.41 M⊕M_{\mathrm{p}} = 5.08 \pm0.41\ M_{\oplus}, leading to a high bulk density of around 8.2±1.1 g cm−38.2 \pm1.1\ \mathrm{g\,cm^{-3}}, indicative of a rocky, iron-silicate composition and the absence of a significant atmosphere (L. Malavolta et al. 2018; O. Barragán et al. 2018). K2-141 c’s size is not well constrained due to its grazing transit and RV non-detections, but it is more likely a Neptune-like planet rather than a rocky super-Earth or hot Jupiter (L. Malavolta et al. 2018).

Most Kepler and K2 ultra-short period planet candidates orbit stars too faint for precise radial velocity follow-up; however, K2-141’s relative brightness allows for a precisely determined mass and thus a reliable density estimate (L. Malavolta et al. 2018). K2-141 is also bright enough to conduct follow-up observations of the planet’s emission in the infrared as previously done with exoplanet 55 Cnc e, one of only a handful of other small USP planets accessible in both the optical and infrared (B.-O. Demory et al. 2016a,b). Although the larger planet K2-141 c is not a good candidate for characterization due to its grazing transits, K2-141 b’s high surface temperature >2000> 2000 K makes it an excellent target for thermal emission observations as in the Spitzer analysis by Zieba, S. et al. (2022).

The mineralogy of highly irradiated lava worlds is fundamentally shaped by both stellar carbon-to-oxygen and magnesium-to-silicon ratios, both in the planet’s deep interior and the observable atmosphere. Considering interior models of exoplanets, stellar C/O and Mg/Si ratios serve as the critical inputs used to break interior structure degeneracies, particularly silicate versus carbide-dominated interiors (C. Dorn et al. 2015, 2017). The redox state and mineral composition of a rocky planet’s mantle further control the chemistry of its outgassed atmosphere, meaning that the host star’s C/O ratio leaves a fingerprint on the very gases a lava world releases under extreme irradiation (G. Ortenzi et al. 2020). For example, magma ocean surfaces may produce chemically distinct magma compositions depending on the rock type present (L. Schaefer & B. Fegley 2009), and such mantle outgassing drives measurable differences in atmospheric species like SiO, CO2_2, and SO2_2 across different planetary compositions (E. S. Kite et al. 2016). K2-141 b, as a highly irradiated rocky planet, is a prime testbed for mantle dynamics, nightside volcanism, volatile outgassing models, and the search for signatures of vaporizing surface rocks and minerals. Using two-dimensional convection models with tracer-based volatile tracking, studies like T. G. Meier et al. (2026) find that magma ocean thickness varies from 200 to 300 km depending on the model parameters, corresponding to 2-3 % of the planet’s radius. Nightside volcanism gradually depletes the mantle of volatiles resulting in the outgassing of tens of bars of CO2\mathrm{CO}_{2} and H2O\mathrm{H}_{2}\mathrm{O} over the course of a billion years. Notably, they demonstrate that even large volcanic eruptions on the nightside produce thermal emission signals of no more than 1 ppm which is well below the current detectability limits in thermal phase curves (T. G. Meier et al. 2026). Nevertheless, the host star’s C/O ratio then provides crucial context for constraining the planet’s composition of the planet’s atmosphere and interior.

The JWST Cycle 1 proposal (GO #2347, L. Dang et al. (2021)) monitored K2-141 b for three complete orbits using MIRI/LRS, aiming to detect SiO in emission of this alleged lava planet, search for silicate clouds, and investigate the presence/structure of a global atmosphere. Analysis of the full-orbit spectroscopic MIRI/LRS (5-12 μ\mum) phase curve places constraints on the phase offset, Bond albedo, and heat recirculation efficiency from the phase curve, but ultimately inconclusive regarding the presence, or lack-of a global atmosphere on a lava planet (L. Dang et al. 2024). The full results are not yet published.

The NGTS-10 System

NGTS-10 is a 10.4 ± 2.5 Gyr old K5V star with effective temperature Teff=4400±100T_{\mathrm{eff}} = 4400 \pm100 K and solar metallicity showing moderate signs of stellar activity. The star is named after its companion planet NGTS-10 b, the tenth discovery by the Next Generation Transit Survey (NGTS) at the ESO Paranal Observatory. The NGTS

consists of an array of twelve 20 cm telescopes optimized for detecting small planets around K and early M stars. NGTS-10 b is an ultrashort period transiting hot Jupiter with a mass of 2.162−0.107+0.092 MJ2.162^{+0.092}_{-0.107}\ M_{\mathrm{J}}, radius of 1.205−0.083+0.117 RJ1.205^{+0.117}_{-0.083}\ R_{\mathrm{J}}, and orbital period of 0.767 days (18.4 hours) (J. McCormac et al. 2020). As of its discovery date in 2020, NGTS-10 b is the shortest period hot Jupiter yet discovered, and joins the list of rare, ultrashort period (USP) hot Jupiters which includes notable, well-studied exoplanets such as WASP-18 b, WASP-19 b, WASP-103 b, HATS-18 b, KELT-16 b and WASP-43 b (also selected for this elemental and isotopic abundance study).

USP hot Jupiters are the easiest transiting planets to observe from the ground due to their relatively large transit depths and geometrically increased transit probability making them ideal targets for studying star–planet interactions and atmospheric characterization through transmission and emission spectroscopy (J. McCormac et al. 2020). NGTS-10 b is likely to undergo strong tidal interactions with its host star due to its proximity at just 1.46 ± 0.18 Roche radii2, which could trigger orbital decay and ultimately lead to its in-spiraling.

NGTS-10 b was observed by JWST in Cycle 1 through the GO Program Proposal 2158 (V. Parmentier et al. 2021). This program measured a full 18-hour orbital phase curve of the hot Jupiter (0.6 - 5.3 μm\mu\mathrm{m}) using JWST/NIRSpec. The study detects the presence of CO and H2_2O with similar abundances on both dayside and nightside. Although the observations are consistent with a solar-composition atmosphere at chemical equilibrium for the dayside, the nightside favors disequilibrium chemistry due to its significant CH4_4 depletion relative to equilibrium chemistry predictions (V. Parmentier et al. 2026). The team argues that this is likely due to horizontal chemical quenching rather than non-solar elemental abundances or vertical mixing mechanisms. Most relevant to modeling the host star’s spectrum, target acquisition imaging from this program revealed a stellar mass binary companion, NGTS-10 B, separated from the primary by 0.33”. Although our target NGTS-10 A was initially modeled using the combined spectra from both components, the final effective temperature from this study (Teff=4567±36T_{eff} = 4567 \pm36) is consistent with the previous estimated adopted here: Teff=4400±100T_{eff} = 4400 \pm100, 4600±1504600 \pm150 from Double SED fitting and HARPS spectroscopy, respectively (J. McCormac et al. 2020).

The WASP-43 System

WASP-43 is a K7V dwarf star with effective temperature Teff=4500±100T_{eff} = 4500 \pm100 K and near-solar bulk metallicity and slightly super-solar metal abundances. The early discovery of hot Jupiter WASP-43 b, announced in 2011, makes WASP-43 one of the better characterized host stars in this sample. In fact, it is one of the targets of precursor spectroscopic studies for the ESO’s Atmospheric Remote-sensing Infrared Exoplanet Large (Ariel) survey (Magrini, L. et al. 2022; da Silva, R. et al. 2024; M. Tsantaki et al. 2025; H. Ramler et al. 2026).

The companion planet WASP-43 b has been extensively studied by both ground-based and space-based observatories. WASP-43 b, as the name suggests, was discovered by the ground-based WASP collaboration (D. L. Pollacco et al. 2006). The study found the planet transiting its host star every 0.81 days corresponding to a semi-major axis of only 0.014 AU or 2.1 stellar Roche radii (C. Hellier et al. 2011). The planet is about the size of Jupiter, but more massive with M=1.8 MJM = 1.8\ M_J, radius R=0.9 RJR = 0.9\ R_J, and a high bulk density (C. Hellier et al. 2011). Like NGTS-10, WASP-43 is believed to undergo strong star-planet tidal interactions and possible decay of its orbit.

Hot Jupiters are among the easiest exoplanets to characterize from the ground due to their frequent transits and large transit depths, and this is certainly the case for WASP-43 b. Early transmission spectroscopy observations of WASP-43 sought to characterize its atmosphere both in the optical (F. Murgas et al. 2014) and near-infrared (G. Chen et al. 2014) but struggled to distinguish between a cloudy, high mean molecular weight, or hazy atmosphere due to large observational uncertainties. Another study focused instead on high-precision radial velocity measurements to measure the Rossiter-McLaughlin effect and refine orbital parameters like the spin-orbit angle (M. Esposito et al. 2017). WASP-43 and WASP-80, both included in this elemental and isotopic abundance analysis, are among the coolest stars for which the Rossiter-McLaughlin effect has been observed so far. This study also found evidence for enhanced stellar activity possibly induced by the prominent star-planet interactions in the WASP-43 system. More recent ground-based studies focus on high-resolution, near-infrared cross-correlation spectroscopy of WASP-43 b. The hot Jupiter has been observed by both VLT/CRIRES+ and Gemini/IGRINS, the premier high-resolution, near-infrared spectrographs used for these types of analyses. The CRIRES+ retrieval reported the first detection of CO, provided evidence for the existence of H2_2O, and measured a super-solar C/O ratio of 0.78 in the atmosphere of WASP-43 b (F. Lesjak et al. 2023). Similarly, the IGRINS study detected H2_2O but did not detect the presence of CH4_4, CO, or CO2_2 in WASP-43 b’s atmosphere (D. Bartellet et al. 2025). Measurements of the C/O ratio of hot Jupiter atmospheres and their host stars are key to understanding both planet formation and migration but also cloud formation and atmospheric composition (H. Reggiani et al. 2022; F. Lesjak et al. 2023).

Space-based observations of WASP-43 b have revealed differences between the planet’s nightside and dayside. Thermal emission and transmission spectroscopy measurements obtained with instruments like Hubble allow for a precise determination of the water abundance in the atmospheres of hot Jupiters (L. Kreidberg et al. 2014). Photometric reflected light observations of WASP-43 b using with Hubble reveal a very dark dayside atmosphere (J. Fraine et al. 2021). Later studies using TESS and CHEOPS light curves of the planet argue this low dayside albedo is either due to a cloud-free atmosphere (G. Scandariato et al. 2022) or the presence of clouds containing strongly absorbing Fe-bearing aerosol species (K. L. Chubb et al. 2024). Full-orbit phase curves from Spitzer are used to determine that WASP-43 b likely has a cloudy nightside that transitions to a relatively cloudy-free dayside (M. M. Murphy et al. 2023). A recent study observes WASP-43 b with JWST’s MIRI instrument and finds a large day–night temperature contrast, evidence for water absorption, the presence of nightside clouds, and an absence of dayside clouds (T. J. Bell et al. 2024).

Because of WASP-43 b’s numerous atmospheric abundance constraints, this planet is one of only a handful where a direct comparison can be made with its host stars’ chemical abundances. This is essential to tracing a planet’s formation, migration, and accretion histories and also to better understanding an exoplanet’s atmospheric evolution.

The WASP-69 System

WASP-69 is an active, mid-K dwarf with effective temperature Teff=4748±40T_{\mathrm{eff}} = 4748 \pm40 K and super-solar bulk metallicity. WASP-69 is thought to be relatively young at an age of 0.8 - 1.1 Gyr estimated from its rotation rate, lithium abundance, and activity indicators [4]. Like WASP-43, WASP-69 is another well-characterized system due to the early detection of a hot Jupiter companion orbiting the host. WASP-69 b is an inflated Saturn-mass planet (M=0.26 MJM = 0.26\,M_{\mathrm{J}}, R=1.06 RJR = 1.06\,R_{\mathrm{J}}) on a 3.868 day orbit [4].

Due to its proximity to its active host star and extended atmosphere, WASP-69 b is a favorable target for studies of atmospheric loss. Several high-resolution, near-infrared studies have targeted WASP-69 b in search of the metastable helium triplet line located at 1083.3 nm which is one of the best tracers of atmospheric escape in transiting exoplanets. This atmospheric escape is largely driven by the host star’s x-ray and ultraviolet irradiation (J. Sanz-Forcada et al. 2025). Studies using data from the ESO/NIRPS, Calar Alto/CARMENES, CFHT/SPIRou, and Keck/NIRSPEC spectrographs have confirmed excess absorption in the helium triplet during the transit of WASP-69 b and provided constraints on the planet’s atmospheric mass-loss rate [2] (L. Nortmann et al. 2018; M. Lampón et al. 2023; A. Masson et al. 2024). A recent Keck/NIRSpec study of WASP-69 b’s atmospheric escape further finds that the atmospheric outflow extends at least 7.5 times the radius of the planet and corresponds to a mass-loss rate of 1 M⊕M_{\oplus} per Gyr (D. Tyler et al. 2024).

Other ground-based studies of WASP-69 b find evidence for the presence of several key molecular species in the hot Jupiter’s atmosphere. Near-infrared transmission spectroscopy reveals the presence of CH4_4, NH3_3, CO, C2_2H2_2, and H2_2O indicative of a high C/O ratio (G. Guilluy et al. 2022) while optical transmission spectroscopy suggests the presence of TiO and high-altitude aerosols (Q. Ouyang et al. 2023). Even low-resolution spectrophotometry using VLT/FORS2 demonstrates evidence for H2_2O and CH4_4 absorption and high-altitude cloud decks while accounting for stellar faculae (D. J. M. Petit dit de la Roche et al. 2024).

Recent observations of two WASP-69 b eclipses using JWST’s NIRCam and MIRI instruments provide an emission spectrum covering 2 - 12 microns (E. Schlawan et al. 2024b). The data reveals absorption features from water vapor, carbon dioxide, and carbon monoxide, but no strong evidence for methane. Furthermore, WASP-69 b’s emission spectrum seems inconsistent with cloud-free homogeneous models, instead favoring models with aerosols. The authors suggest possible model scenarios for the planet with the most plausible having higher metallicity and super-solar C/O: a cloud-layer model that includes high-altitude silicate aerosols to moderate long-wavelength emission; or a two-region model that includes significant dayside inhomogeneity and cloud opacity with two different temperature-pressure profiles (E. Schlawin et al. 2024b). New modules like NEXOTRANSNEXOTRANS leverage Bayesian and machine learning techniques for retrievals while also incorporating disequilibrium chemistry, which when applied to the same JWST/NIRCam + JWST/MIRI spectra, place robust constraints on the volume mixing ratios of H2_2O, CO2_2, CO, CH4_4, and potentially SO2_2 (T. Deka et al. 2026).

Like WASP-43, WASP-69 b’s numerous atmospheric abundance constraints make this planet one of only a handful where a direct comparison can be made with its host stars’ chemical abundances to reveal the planet’s formation, migration, accretion, and evolution.

The WASP-80 System

WASP-80, otherwise known as TYC 5165-481-1, is a relatively bright (Kmag = 8.4) late-type K dwarf star with sub-solar metallicity and effective temperature Teff=4143±94T_{\mathrm{eff}} = 4143 \pm94 K. The hot Jupiter WASP-80 b was discovered transiting this star in 2013 and has mass M=0.55±0.04 MJM = 0.55 \pm0.04\,M_{\mathrm{J}}, radius R=0.95±0.03 RJR = 0.95 \pm0.03\,R_{\mathrm{J}} and a circular, 3.07 day orbit (A. H. M. J. Triaud et al. 2013). Like the other planets whose host stars were selected for this elemental and isotopic abundance study, WASP-80 b is another worthwhile target for characterization via transmission and emission spectroscopy.

From the ground, WASP-80 b has been observed in transmission by various high- and low- resolution spectrographs. At low-resolution, WASP-80 b’s transmission spectrum is most consistent with a hazy atmosphere (J. Kirk et al. 2018). At high-resolution, WASP-80 b’s transmission spectra reveal a mostly flat spectrum that favors high-altitude cloud models over cloud-free models with stellar or sub-stellar metallicities (H. Parviainen et al. 2018). High-resolution, near-infrared transmission spectroscopy suggests an atmosphere consistent with solar composition and the presence of disequilibrium chemistry (I. Carleo et al. 2022). This study also detects the presence of several molecular species in WASP-80 b’s atmosphere including H2_2O, CH4_4, NH3_3, and HCN with high significance, tentative detection of CO2_2, and inconclusive results for C2_2H2_2 and CO (I. Carleo et al. 2022). Using the same data, another study looks for planetary absorption at He I triplet tracer of atmospheric escape but does not find any signature (L. Fossati et al. 2022).

From space, WASP-80 b has been studied both by Spitzer and JWST. Analysis of early Spitzer phase curves again show a flat transmission spectrum but do reveal WASP-80 b is as bright and as red as a T4-class brown dwarf, but that its temperature is much cooler (A. H. M. J. Triaud et al. 2015). JWST/NIRCam transmission and emission spectra spanning 2.4-4.0 microns show strong evidence for CH4_4 which are consistent with solar to sub-solar C/O and around five times solar metallicity for WASP-80 b’s atmosphere (T. J. Bell et al. 2023). A comparison of the shorter wavelength eclipse spectrum obtained with JWST’s NIRISS/SOSS instrument with aerosol models demonstrates that the spectrum favors cloud species with weak-to-moderate near-infrared reflectance, soots, and hazes over manganese sulfide and silicate clouds (K. Morel et al. 2025). More recently, a 2.4 - 10 μ\mum transmission spectrum obtained by JWST/NIRCam and MIRI reveals evidence for H2_2O, CH4_4, CO2_2, NH3_3, and CS2_2 in the atmosphere of WASP-80 b and place upper limits on CO and SO2_2 (A. Triantafillides et al. 2026). This study establishes CS2_2 as an observable tracer of sulfur disequilibrium chemistry that is also consistent with recent chemically validated networks that include efficient carbon-sulfur coupling through CH2_2S.

WASP-80 b is yet another of only a handful where a direct comparison can be made with its host stars’ chemical abundances. In this case, however, the limiting factor is the lack of chemical abundances reported for its cool, late-K-type host star WASP-80.

The HATS-72 System

HATS-72 is a main sequence K dwarf with super-solar metallicity and effective temperature Teff=4656±9T_{eff} = 4656 \pm9 K. HATS-72 b was discovered alongside three other giant planets transiting K dwarf stars– HATS-47 b, HATS-48 b, and HATS-49 b– by the HAT-South (HATS) collaboration (G. A. Bakos et al. 2013; J. D. Hartman et al. 2020). This hot Jupiter, also known as WASP-191 b or TOI 294.01, was detected independently by the WASP collaboration and NASA’s Transiting Exoplanet Survey Satellite (TESS; G. R. Ricker et al. 2015). HATS-72 b has a mass M = 0.1254±\pm0.0039 MJ_J, R = 0.7224±\pm0.0032 RJ_J, and an orbital period of 7.3279 days (J. D. Hartman et al. 2020). Unlike the companion planets to the other host stars selected for this abundance study, HATS-72 b is classified as a warm-Jupiter rather than a hot Jupiter due to its longer orbital period PP = 7.33 days. The expected equilibrium temperature of the companion planet is below 1000 K because of its host star’s relatively small size and luminosity (J. D. Hartman et al. 2020).

HATS-72 b was observed by JWST’s NIRSpec and NIRISS/SOSS instruments as part of the JWST Cycle 1 GO Program 2594 (J. Spake et al. 2021). The proposal aimed to test the “CH4_4 thermometer” identified by J. J. Fortney et al. (2020) which establishes that atmospheric CH4_4/H2_2O and CH4_4/CO2_2 abundance ratios are good probes of the interior temperatures of moderately-irradiated exoplanets (Teq∼750T_{eq} \sim750 K). HATS-72 b, as the older and denser planetary twin (same mass, host star type and equilibrium temperature) to WASP-107 b, is expected to have stronger CH4_4 features than its twin, implying WASP-107 b’s hotter interior is the reason for its inflated radius (J. Spake et al. 2021). The final analysis of these JWST observations has not yet been published. The host star HATS-72 also has no published carbon and oxygen abundances to date, so JWST’s observations of HATS-72 b will rely on this work for additional context.

Observations & Data

High-Resolution, NIR IGRINS Spectra

In this study, we use high-resolution R∼45,000R \sim45{,}000, near-infrared (NIR) spectra obtained from the Immersion Grating INfrared Spectrograph (IGRINS, C. Park et al. (2014); G. Mace et al. (2018)) at the Gemini-South and Lowell observatories. IGRINS is a compact, cross-dispersed spectrograph offering simultaneous coverage of the H and K bands located between 1.45–2.45 microns.

Data were retrieved from the Raw & Reduced IGRINS Spectral Archive (E. Sawczynec et al. 2025) including archival observations and data from our group’s observing programs: GS-2023B-LP-108 (PI: Crossfield), GS-2024A-LP-108 (PI: Crossfield), and GS-2023A-Q-203 (PI: Crossfield). All IGRINS spectra are reduced using the IGRINS Pipeline Package V3 (E. Sawczynec et al. 2025). This pipeline performs a number of reduction processes including flat-fielding, background removal, order extraction, distortion correction, wavelength calibration and telluric correction using A0V spectra to produce a final multi-order (23 orders in the H band, 21 orders in the K band) echelle spectrum per target. The reduced multi-order spectra are then combined using the combspec tool in the SpeXTool package ((M. Cushing et al. 2004)) and then flattened. These K dwarf spectra range in S/NS/N from S/N=171.5S/N = 171.5 in the H band and S/N=84.2S/N = 84.2 in the K band for NGTS-10 to S/N=408.0S/N = 408.0 in the H band and S/N=275.3S/N = 275.3 in the K band for WASP-69. We plot these final, flattened spectra in Figure 1. Additional observational parameters are provided in Table 2.

ParameterK2-141NGTS-10WASP-43WASP-69WASP-80HATS-72
FacilityGemini SouthGemini SouthGemini SouthLowell ObservatoryLowell ObservatoryGemini South
Program IDGS-2022B-Q-226GS-2023A-Q-203GS-2023A-FT-106——GS-2023A-Q-203
Date (UT)2022-09-062023-03-262023-04-122018-09-232017-09-122023-04-28
Time (UT)02:12–02:1500:43–02:1200:30–00:3304:04–04:0607:06–07:1108:52–08:57
Integration Time (s)15045015090300270
Mean H-band S/N259.8171.5223.5408.0253.0292.4
Mean K-band S/N181.784.2163.3275.3142.6173.7

Table 2. Observational parameters for the six K dwarfs observed with the IGRINS spectrograph ((C. Park et al. 2014; G. Mace et al. 2018)). Mean H-band S/N calculated over the 1.50–1.80 µm wavelength range. Mean K-band S/N calculated over the 2.20–2.45 µm wavelength range.

Figure 1. Plots of the observed, high signal-to-noise H (top panel) and K (bottom panel) band IGRINS spectra used in this abundance analysis. Oxygen abundances are determined from OH lines located in the H band (1.45–1.80 microns) while carbon, 12C/13C^{12}\mathrm{C}/^{13}\mathrm{C}, and 16O/18O^{16}\mathrm{O}/^{18}\mathrm{O} abundances are derived from various CO isotopologue lines located in the 2.3–2.45 micron region of the K band (1.95–2.45 microns).

Methods: Elemental Carbon and Oxygen Abundance Analysis

The main goal of this study is to measure 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} and 16O/18O{}^{16}\mathrm{O}/{}^{18}\mathrm{O} ratios in a sample of K dwarf stars hosting high-priority giant exoplanets. Previous studies have made these measurements in solar type stars assuming solar carbon and oxygen elemental abundances (D. R. Coria et al. 2023; D. R. Coria et al. 2024). This is acceptable considering the stars in those samples are of solar metallicity; however, the same assumption cannot be made for the selected K dwarfs as most have non-solar [Fe/H] and a handful of which have non-solar [C/H] or [O/H]. The remaining K dwarfs have no [C/H] or [O/H] available in the literature at all. So, prior to determining 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} and 16O/18O{}^{16}\mathrm{O}/{}^{18}\mathrm{O} ratios, elemental carbon and oxygen abundances must be constrained.

Model Atmospheres, Line Data, and Spectral Synthesis

The grids of model spectra for this study are generated following the methods described in N. Hejazi et al. (2023) and D. R. Coria et al. (2024). We begin by synthesizing a grid of continuum-normalized model spectra using the TurboSpectrum 1D LTE radiative transfer code (B. Plez 2012) together with 1D, hydrostatic, and plane-parallel MARCS atmosphere models, (Gustafsson et al. 2008) and a select set of atomic and molecular line lists presented in N. Hejazi et al. (2023). The line lists most crucial to a near-infrared carbon and oxygen abundance analysis are those for OH (A. Goldman 1982; A. Goldman et al. 1998) and CO (D. Goorvitch 1994).

First, we use the interpolation routine developed by T. Masseron to interpolate the MARCS model atmosphere associated with each star’s physical parameters. These adopted TeffT_{\mathrm{eff}}, log⁡g\log g, and [Fe/H] (used interchangeably with [M/H], the bulk stellar metallicity relative to solar) values are listed in Table 4. Then, we generate synthetic spectra using the interpolated model, while customizing the abundance of element X in Turbospectrum through an input quantity named “relative abundance” ABUND(X), which is related to the abundance [X/H], by

ParameterK2-141NGTS-10WASP-43WASP-69WASP-80HATS-72
TeffT_{\mathrm{eff}}4599 ±\pm 79(3), 4472 ±\pm 150(9)4600 ±\pm 150(4)4400 ±\pm 100(5), 4184 ±\pm 116(9)4700 ±\pm 50(6), 4876 ±\pm 77(2), 4693 ±\pm 150(9)4143 ±\pm 94(7)4656 ±\pm 9(8)
log gg4.62 ±\pm 0.03(3), 4.64 ±\pm 0.11(9)4.59 ±\pm 0.02(4)4.50 ±\pm 0.20(5), 4.61 ±\pm 0.14(9)4.50 ±\pm 0.15(6), 4.50 ±\pm 0.09(2), 4.46 ±\pm 0.15(9)4.66 ±\pm 0.02(7)4.59 ±\pm 0.01(8)
[Fe/H]−0.06±0.10-0.06 \pm0.10(3), +0.07±0.15+0.07 \pm0.15(9)−0.02±0.12-0.02 \pm0.12(4)−0.01±0.15-0.01 \pm0.15(5), +0.29±0.22+0.29 \pm0.22(9)+0.15±0.08+0.15 \pm0.08(6), +0.40±0.02+0.40 \pm0.02(2), +0.38±0.09+0.38 \pm0.09(9)−0.13±0.17-0.13 \pm0.17(7)+0.10±0.01+0.10 \pm0.01(8)
[C/H]+0.04±0.09+0.04 \pm0.09(1)——+0.40±0.06+0.40 \pm0.06(1), +0.27±0.05+0.27 \pm0.05(2)——
[O/H]−0.04±0.09-0.04 \pm0.09(1)—+0.30±0.09+0.30 \pm0.09(1)+0.14±0.02+0.14 \pm0.02(1), +0.18±0.07+0.18 \pm0.07(2)——
C/O∼0.66\sim0.66(1)——∼1.00\sim1.00(1), ∼0.68\sim0.68(2)——
Adopted (This Work)
TeffT_{\mathrm{eff}}4571 ±\pm 704600 ±\pm 150(4)4308 ±\pm 764748 ±\pm 404143 ±\pm 94(7)4656 ±\pm 9(8)
log gg4.62 ±\pm 0.034.59 ±\pm 0.02(4)4.55 ±\pm 0.114.50 ±\pm 0.074.66 ±\pm 0.02(7)4.59 ±\pm 0.01(8)
[Fe/H]−0.03±0.08-0.03 \pm0.08−0.02±0.12-0.02 \pm0.12(4)+0.12±0.12+0.12 \pm0.12+0.24±0.02+0.24 \pm0.02−0.13±0.17-0.13 \pm0.17(7)+0.10±0.01+0.10 \pm0.01(8)
Derived (This Work)
[C/H] (NlinesN_{\mathrm{lines}})−0.03±0.05-0.03 \pm0.05 (19)+0.05±0.06+0.05 \pm0.06 (12)+0.24±0.05+0.24 \pm0.05 (18)+0.26±0.06+0.26 \pm0.06 (10)+0.05±0.06+0.05 \pm0.06 (10)+0.18±0.05+0.18 \pm0.05 (18)
[O/H] (NlinesN_{\mathrm{lines}})+0.02±0.08+0.02 \pm0.08 (32)+0.11±0.08+0.11 \pm0.08 (37)+0.11±0.08+0.11 \pm0.08 (46)+0.35±0.09+0.35 \pm0.09 (18)−0.12±0.08-0.12 \pm0.08 (57)+0.20±0.08+0.20 \pm0.08 (29)
C/O0.53 ±\pm 0.130.51 ±\pm 0.140.79 ±\pm 0.200.48 ±\pm 0.130.87 ±\pm 0.230.56 ±\pm 0.14
12C/13C^{12}\mathrm{C}/^{13}\mathrm{C} (NlinesN_{\mathrm{lines}})51 ±\pm 12 (10)48 ±\pm 11 (3)50 ±\pm 12 (9)38 ±\pm 10 (4)—52 ±\pm 16 (3)

Table 4. Fundamental Stellar Parameters and Derived Abundances

[X/H]=ABUND(X)+[M/H],(1)[\mathrm{X}/\mathrm{H}] = \mathrm{ABUND}(\mathrm{X}) + [\mathrm{M}/\mathrm{H}], \tag*{(1)}

where [M/H] indicates the overall metallicity. For carbon and oxygen, the abundances are given by

[C/H]=ABUND(C)+[M/H],(2)[\mathrm{C}/\mathrm{H}] = \mathrm{ABUND}(\mathrm{C}) + [\mathrm{M}/\mathrm{H}], \tag*{(2)}

and

[O/H]=ABUND(O)+[M/H]+[α/Fe],(3)[\mathrm{O}/\mathrm{H}] = \mathrm{ABUND}(\mathrm{O}) + [\mathrm{M}/\mathrm{H}] + [\alpha/\mathrm{Fe}], \tag*{(3)}

respectively. Oxygen is an α\alpha element and is therefore subject to α\alpha-enhancement scaling relations, with [α/Fe]=−0.4×[M/H][\alpha/\mathrm{Fe}] = -0.4 \times[\mathrm{M}/\mathrm{H}] for −1.0≤[M/H]≤0.0-1.0 \le[\mathrm{M}/\mathrm{H}] \le0.0, and [α/Fe]=0.0[\alpha/\mathrm{Fe}] = 0.0 for [M/H]>0.0[\mathrm{M}/\mathrm{H}] > 0.0, as described in Gustafsson et al. (2008). These corrections are negligible for our sample of stars with near-solar to super-solar metallicities.

The “absolute abundance” of element X is defined as

A(X)=log⁡10(NXNH)+12,(4)A(\mathrm{X}) = \log_{10}\left(\frac{N_{\mathrm{X}}}{N_{\mathrm{H}}}\right) + 12, \tag*{(4)}

where NXN_{\mathrm{X}} and NHN_{\mathrm{H}} are the number densities of element X and hydrogen, respectively. Recalling that the abundance of element X is defined relative to the Sun,

[X/H]=log⁡10(NXNH)−log⁡10(NXNH)⊙,(5)[\mathrm{X}/\mathrm{H}] = \log_{10}\left(\frac{N_{\mathrm{X}}}{N_{\mathrm{H}}}\right) - \log_{10}\left(\frac{N_{\mathrm{X}}}{N_{\mathrm{H}}}\right)_{\odot}, \tag*{(5)}

the absolute abundance can be expressed as

A(X)=A(X)⊙+[X/H].(6)A(\mathrm{X}) = A(\mathrm{X})_{\odot} + [\mathrm{X}/\mathrm{H}]. \tag*{(6)}

The C/O ratio is then calculated from the absolute carbon and oxygen abundances as

C/O=NCNO=10A(C)−A(O),(7)\mathrm{C}/\mathrm{O} = \frac{N_{\mathrm{C}}}{N_{\mathrm{O}}} = 10^{A(\mathrm{C})-A(\mathrm{O})}, \tag*{(7)}

with its propagated uncertainty, assuming independent uncertainties, given by

σC/O=ln⁡(10) (C/O)σA(C)2+σA(O)2,(8)\sigma_{\mathrm{C}/\mathrm{O}} = \ln(10)\,(\mathrm{C}/\mathrm{O})\sqrt{\sigma_{A(\mathrm{C})}^{2}+\sigma_{A(\mathrm{O})}^{2}}, \tag*{(8)}

where σA(C)\sigma_{A(\mathrm{C})} and σA(O)\sigma_{A(\mathrm{O})} are the uncertainties in the absolute carbon and oxygen abundances, respectively, calculated as

σA(C)=σA(C)⊙2+σ[C/H]2,(9)\sigma_{A(\mathrm{C})} = \sqrt{\sigma_{A(\mathrm{C})_{\odot}}^{2}+\sigma_{[\mathrm{C}/\mathrm{H}]}^{2}}, \tag*{(9)}

and

σA(O)=σA(O)⊙2+σ[O/H]2.(10)\sigma_{A(\mathrm{O})} = \sqrt{\sigma_{A(\mathrm{O})_{\odot}}^{2}+\sigma_{[\mathrm{O}/\mathrm{H}]}^{2}}. \tag*{(10)}

For each K dwarf, we generate a grid of synthetic spectra using its respective adopted TeffT_{\mathrm{eff}}, log⁡g\log g, and [Fe/H] values with different relative abundances of carbon and oxygen: ABUND(X) = +0.40, +0.30, +0.20, +0.15, +0.10, +0.05, +0.00, −0.10, −0.20. The selected K dwarfs have slightly sub-solar to super-solar [Fe/H] so the grids center on ABUND(X) = +0.10.

Line Identification and Continuum/Pseudo-continuum Placement

We use a custom list of 27 CO lines and 78 OH lines to determine [C/H][\mathrm{C}/\mathrm{H}] and [O/H][\mathrm{O}/\mathrm{H}] abundances respectively. This list is adopted from N. Hejazi et al. (2024) and includes line locations and the most appropriate continuum/pseudo-continuum regions around each line to be used for normalization. The observed spectrum must be normalized relative to the grid of synthetic spectra. This is done by fitting a low-order polynomial to the residuals R=O/SR = O/S (where OO is the observed flux and SS is the interpolated synthetic flux at a given wavelength) of the data points within the selected continuum/pseudo-continuum regions. The normalized spectrum is obtained by dividing the observed flux by this final polynomial-fitted residual. The full continuum/pseudo-continuum placement procedure is described in N. Hejazi et al. (2024); N. Hejazi et al. (2025).

Deriving Elemental Carbon and Oxygen Abundances

Carbon [C/H][\mathrm{C}/\mathrm{H}] and oxygen [O/H][\mathrm{O}/\mathrm{H}] abundance values are determined line-by-line using a χ2\chi^2-minimization routine as in D. R. Coria et al. (2024). A χ2\chi^2 value is calculated between the normalized observed spectrum and each synthetic spectrum in the grid, where each synthetic spectrum corresponds to a different relative abundance value [ABUND(X)]=+0.40,+0.30,+0.20,+0.15,+0.10,+0.05,+0.00,−0.10,−0.20[\mathrm{ABUND}(\mathrm{X})] = +0.40, +0.30, +0.20, +0.15, +0.10, +0.05, +0.00, -0.10, -0.20 with the same physical stellar parameters. A cubic spline is fit to the set of χ2\chi^2 and relative abundance points, and then we calculate the relative abundance where χ2\chi^2 is minimized. This minimum in the χ2\chi^2 curve represents the best-fit relative abundance value for the given line. This process is repeated for each OH and CO line; however, due to poor line-profile modeling, bad pixels in the observed spectrum, or issues with either normalization or fitting, not all lines are included in the final abundance calculation. An example of this process is shown in Figure 2.

Figure 2. Deriving best-fit abundances for K2-141 using a χ2\chi^2-minimization routine. χ2\chi^2 is calculated between each synthetic spectrum corresponding to a different abundance value [X/H]. A cubic spline connects these χ2\chi^2 points as shown in the right plots and is used to interpolate the value of [X/H] such that χ2\chi^2 is minimized. This is the best-fit abundance value for a particular line. This routine is run only on the line core and includes only points inside the selected χ2\chi^2 fit window for each line. Uncertainties represent 1σ1\sigma confidence intervals using the region where Δχmin⁡2≤1\Delta\chi^2_{\min} \leq1 (Y. Avni 1976). An example of this calculation is shown for an OH, CO, and 13CO^{13}\mathrm{CO} line in the top, middle, and bottom rows respectively.

A weighted average3 of all the best-fit relative abundances for selected OH lines gives the relative oxygen abundance value for the star. Similarly, a weighted average of all the best-fit relative abundances for selected CO lines gives the relative carbon abundance value for the star. Final [O/H][\mathrm{O}/\mathrm{H}] and [C/H][\mathrm{C}/\mathrm{H}] abundances are obtained by adding the stellar metallicity values to the inferred relative abundances using Equations (2) and (3).

The statistical uncertainties of these abundance measurements are given as the standard error on the mean of the abundances measured for each line. This way, we account for the overall spread of calculated abundance values and the number of lines used to derive the final abundance. Statistical uncertainties on the [C/H][\mathrm{C}/\mathrm{H}] values range from ∼0.01−0.03\sim0.01 - 0.03 dex while those for [O/H][\mathrm{O}/\mathrm{H}] are ∼0.01\sim0.01 dex. N. Hejazi et al. (2024) found that systematic uncertainties on these abundance values arise mostly from the choice of effective temperature and metallicity for the model spectra resulting in systematic uncertainties of σT≤0.08\sigma_T \le0.08 and σZ≤0.03\sigma_Z \le0.03. We adopt these values here and report final uncertainties by adding σstat\sigma_{\mathrm{stat}}, σT\sigma_T, and σZ\sigma_Z in quadrature. Systematic uncertainties arising from the choice of surface gravity (log g) are negligible.

The final [C/H][\mathrm{C}/\mathrm{H}] and [O/H][\mathrm{O}/\mathrm{H}] values and uncertainties for each K dwarf in the sample, scaled according to Equations (2) and (3), are provided in the bottom panel of Table 4.

Methods: Isotopic Carbon Abundance Analysis

To measure the 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} ratio for the K dwarf sample, we follow the same methods as for deriving [C/H][\mathrm{C}/\mathrm{H}] and [O/H][\mathrm{O}/\mathrm{H}] described above. The MARCS model atmosphere grid is first interpolated to the adopted TeffT_{\mathrm{eff}}, log g, and [Fe/H][\mathrm{Fe}/\mathrm{H}] for each star as before. This time, the relative abundances for carbon and oxygen are fixed at the elemental abundance values derived above. This is done to ensure that the 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} model grids accurately reflect the carbon and oxygen ratios measured for each K dwarf rather than solar abundances.

Next, a grid of synthetic spectra is generated for each K dwarf using its respective adopted atmospheric parameters from Table 1, the fixed relative abundances of carbon and oxygen, and varying 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} ratios: 12C/13C=30,50,60,80,90,100,110,130,160{}^{12}\mathrm{C}/{}^{13}\mathrm{C} = 30, 50, 60, 80, 90, 100, 110, 130, 160. These 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} grids center around the solar value, 12C/13C=91.4±1.3{}^{12}\mathrm{C}/{}^{13}\mathrm{C} = 91.4 \pm1.3 (T. R. Ayres et al. 2013).

We employ the CO line list from D. Goorvitch (1994), which also contains information for minor isotopologues of CO, and TurboSpectrum's equivalent width calculation function (B. Plez 2012) to identify strong lines that may be used to derive the 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} ratio. We visually inspect the isotopologue model grid at any locations where Turbospectrum calculates EQW>1\mathrm{EQW} > 1 mÅ. This filtering leaves thirteen 13CO{}^{13}\mathrm{CO} regions for which we determine suitable continuum/pseudo-continuum regions for normalization and χ2\chi^2 fitting windows as in (N. Hejazi et al. 2023; D. R. Coria et al. 2024). These normalization and fitting regions are available in the accompanying machine-readable line list (see Table A1). It is impor-

BytesFormatUnitsLabelExplanations
1–9A9—SpeciesLine species (OH, CO, or 13CO^{13}\mathrm{CO})
11–19F9.3A˚\mathring{A}CenterLine center wavelength
21I1—FlagNumber of exclusion zones for continuum selection
23–31F9.3A˚\mathring{A}MinLamContMinimum wavelength of the continuum selection window
33–41F9.3A˚\mathring{A}MaxLamContMaximum wavelength of the continuum selection window
43–52F10.3A˚\mathring{A}MinLamExc1Minimum wavelength of exclusion window 1
54–63F10.3A˚\mathring{A}MaxLamExc1Maximum wavelength of exclusion window 1
65–74F10.3A˚\mathring{A}MinLamExc2Minimum wavelength of exclusion window 2; ’999999’ if Flag == ’1’
76–85F10.3A˚\mathring{A}MaxLamExc2Maximum wavelength of exclusion window 2; ’999999’ if Flag == ’1’
87–95F9.3A˚\mathring{A}MinLamFitMinimum wavelength of the χ2\chi^2 abundance fit window
97–105F9.3A˚\mathring{A}MaxLamFitMaximum wavelength of the χ2\chi^2 abundance fit window

Table A1. Column Descriptions for Supplementary datafileA1.txt

tant to note that the line centers and χ2\chi^2 fit windows provided for the minor isotopologue 13CO{}^{13}\mathrm{C}\mathrm{O} may, but do not necessarily refer to individual lines and can refer to regions containing multiple blended lines. Furthermore, the 12C/{}^{12}\mathrm{C}/{}^{13}\mathrm{C}$ ratios derived from these overtone rovibrational band lines are highly sensitive to the normalization parameters, requiring a careful continuum/pseudo-continuum selection, but also to spectral noise and the choice of fundamental stellar parameters for the model spectra. For this reason, we consider more conservative uncertainties that account for the systematics associated with these measurements in addition to the relatively small statistical uncertainties as in D. R. Coria et al. (2024). In the previous study, σsys=6−7\sigma_{sys}=6-7 for high quality VLT/ESPRESSO and Keck/HIRES spectra from a well-characterized star (WASP-77 A), but since our spectra are lower S/N and our uncertainties on the stellar effective temperatures are higher, we adopt σsys=10\sigma_{sys}=10. Final uncertainties are estimated by adding σstat\sigma_{stat} and σsys\sigma_{sys} in quadrature.

Although the noise in WASP-80’s spectrum prevents a reliable 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} measurement, we are able to derive 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} ratios for the remaining five targets using at least three 13CO{}^{13}\mathrm{C}\mathrm{O} lines. These values are all sub-solar ranging from 12C/13C=38−52{}^{12}\mathrm{C}/{}^{13}\mathrm{C}=38-52, indicating some 13C{}^{13}\mathrm{C} enrichment for all five hosts.

Even using the high signal-to-noise IGRINS spectra (84<S/N<27584<S/N<275) poses a challenge for deriving 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} ratios from weak 13CO{}^{13}\mathrm{C}\mathrm{O} isotopologue lines. The C18O{}^{\mathrm{C}18}\mathrm{O} isotopologue lines are significantly weaker, so the prospects for deriving reliable 16O/18O{}^{16}\mathrm{O}/{}^{18}\mathrm{O} ratios from K band spectra are poor as shown in D. González Picos et al. (2025a). This means that perhaps deriving 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} and 16O/18O{}^{16}\mathrm{O}/{}^{18}\mathrm{O} ratios is best done in the M band using fundamental rovibrational CO isotopologue lines as in I. J. M. Crossfield et al. (2019), D. R. Coria et al. (2023), and D. González Picos et al. (2026b).

Discussion

Stellar Activity Effects

Targets NGTS-10, WASP-43, WASP-69, and WASP-80 are known to have high activity indexes, but this does not significantly affect our abundance measurements. Magnetic effects are negligible for molecular OH and CO lines used in this work (A. Reiners 2012; P. I. Cristofari et al. 2023) as lines from the first overtone of CO, and similarly the NIR OH band, form deeper in the star and are significantly less affected by any chromospheric activity (T. R. Ayres & G. R. Wiedemann 1989; G. Wiedemann et al. 1994). WASP-69 is the most active star in our sample and known to exhibit starspots and faculae (D. J. M. Petit dit de la Roche et al. 2024). However, the effect is minimal: TESS light curve analysis yields a spot filling factor of 0.77 of the stellar surface (H. Chakraborty et al. 2024). For a photospheric temperature of approximately 4700 K and spots several hundred Kelvin cooler, this filling factor produces only a few-kelvin change in the flux-weighted effective temperature, suggesting that the effect of the observed spot activity on our derived abundances is small compared with the current uncertainties.

The Stellar Age Problem

While many fundamental stellar parameters like mass, radius, and chemical composition can be measured with relative ease, determining the age of an individual star remains notoriously difficult. This is a problem because so many areas of astronomy rely on this chronological context for testing stellar evolution models, understanding galactic archaeology and evolution, and exploring the formation and evolution of exoplanetary systems. Regarding stellar chemical abundances, stellar ages enable the search for abundance-age relationships for different groups of stars. Because solar twin stars are the most amenable to chemical abundance measurements, several relationships have been found between the isochrone ages of these stars and individual elemental abundance ratios, establishing abundance ratios as stellar age proxies. We refer the reader to Section 2.3 of D. R. Coria et al. (2023) and references therein for more details on chemical abundances as stellar chronometers. These abundance-age relationships break down, however, beyond solar twin parameters and remain largely unexplored due to the lack of chemical abundance measurements– and associated challenges in making them– for cooler K/M dwarfs as well as the difficulty in constraining the ages for these stars.

The standard approach for determining stellar ages is to apply isochrones (curves on a Hertzsprung–Russell diagram that connect stars of the same age but different initial masses) that relate the star’s physical properties (e.g. effective temperature, luminosity, chemical composition) to stellar evolution models. Notable isochrone packages in astronomy include Yonsei–Yale isochrones (S. Yi et al. 2001; Y.-C. Kim et al. 2002; S. K. Yi et al. 2003; P. Demarque et al. 2004), the PAdova and TRieste Stellar Evolution Code (PARSEC) (A. Bressan et al. 2012; C. T. Nguyen et al. 2022), and MESA Isochrones and Stellar Tracks (MIST) (J. Choi et al. 2016; A. Dotter et al. 2026). However, this approach is notoriously challenging for cool dwarfs which move imperceptibly across this parameter space while on the main sequence ([8] (M. J. Veyette & P. S. Muirhead 2018; J. Choi et al. 2016; A. Dotter et al. 2026)). Furthermore, isochrones are too closely spaced in effective temperature vs. luminosity space for field-age main-sequence cool dwarfs, meaning isochrone fitting alone often produces unrealistic and unreliable ages for cool dwarf stars.

One alternative is to instead use asteroseismology to measure surface oscillations and infer a star’s evolutionary stage based on its internal structure and composition. Stellar surface oscillations reveal internal structure allowing for precise determination of radii, masses, and ages (Y. Lebreton & J. Montalbán 2009; R. A. García & J. Ballot 2019). This again is challenging for cool dwarfs because the oscillations are so weak, and the technique has only been applied to a handful of bright, nearby K dwarfs like ε\varepsilon Indi, HD 219134, and σ\sigma Draconis (T. L. Campante et al. 2024; M. Hon et al. 2024; Y. Li et al. 2025).

Other approaches for determining stellar ages rely on lithium abundances or stellar activity indicators. Lithium is a powerful stellar age indicator because stars destroy their initial lithium supply as they age through mixing and burning processes. This results in an observable Li depletion by a factor of ∼50\sim50 as age increases from 1 to 9 Gyr (M. Carlos et al. 2016). However, this Li clock is most sensitive for younger stars with ages <1< 1 Gyr and provides the best age constraints for coeval groups of stars versus single stars (R. D. Jeffries 2014). Regarding stellar activity, there is a strong anticorrelation between chromospheric activity and age leading to the use of the Ca II H and K emission index as an age indicator for late F- to early K-type stars (E. E. Mamajek & L. A. Hillenbrand 2008). Other relations including the H-alpha emission index can only be used to roughly estimate ages of field stars with 4000 K <Teff<< T_{\rm eff} < 5500 K with an accuracy of about 40-60 percent (J. Zhang et al. 2019). In addition, these chromospheric activity age indicators are best suited to stars younger than 2 Gyr (G. Pace 2013).

The last ray of hope for obtaining cool dwarf ages is through gyrochronology which assigns ages based on empirical temperature-rotation-age relations observed in clusters. Gyrochronology, in contrast to the methods described above, is ideal for assigning ages to single main sequence stars, middle-aged/old stars, and cool K/M dwarfs (E. Gaidos et al. 2023; L. G. Bouma et al. 2024; Y. Lu et al. 2024). Stellar ages have been primarily estimated through gyrochronology for many planetary systems, particularly for those discovered by the WASP/SWASP surveys (A. S. Bonomo et al. 2017). In most cases, gyrochronology- and isochrone-based stellar ages agree closely (A. S. Bonomo et al. 2015), but gyrochronological ages may not be as reliable for exoplanet hosts [6] because of possible spin-up of the host stars by their close-in exoplanets (F. Pont et al. 2009; P. F. L. Maxted et al. 2015). However, they are often the easiest to calculate for cool dwarf stars as many often have reliable and readily accessible rotation periods from WASP, Kepler, or TESS light curves (C. Hellier et al. 2011; L. G. Bouma et al. 2024; A. W. Boyle et al. 2026) in addition to other required fundamental parameters like temperature, distance, and magnitude.

Literature Ages for the K Dwarf Sample

Literature ages for late K/M dwarfs are exceedingly rare, however, given that the planetary systems in this work are prime targets for ground- and space-based characterization, several efforts have been made to estimate the host star ages. Table 3 provides a comprehensive list of the stellar rotation periods and ages, both isochrone and gyrochrone, available for our K dwarf sample: K2-141, NGTS-10, WASP-43, WASP-69, WASP-80, and HATS-72. Most have at least one published rotation period, isochrone age, and gyrochrone age (HATS-72 lacks a gyrochrone age). The better studied WASP systems have multiple and sometimes conflicting reported values for each of these parameters. One example is the conflicting isochrone age estimates for WASP-43 Age=0.2±0.6Age = 0.2 \pm0.6 Gyr (A. Bonfanti et al. 2015) or Age=11.21±1.5Age = 11.21 \pm1.5 Gyr (Y. Huang et al. 2022). Similarly, WASP-80’s rotation period analyses show a concerning range of viable periods from 3 to 23 days (M. Salz et al. 2015; K. Penev et al. 2018; K. Triantafillidis et al. 2026). Some values in the table, like WASP-69’s 1.8 ±\pm 0.8 Gyr isochrone age [2], are actually composite ages derived from several of the numerous techniques presented in the previous section. R. Allart et al. [2] uses PARSEC isochrones, Li abundances, Ca II H&K emission, chromospheric age, gyrochronology relations, and metallicity-age relations to determine their final age value for WASP-69. Similarly, WASP-69’s composite gyrochrone age of 1.1 ±\pm 0.15 Gyr is derived using some older rotation-age relationships (S. A. Barnes 2007), Li abundance-age relations (P. Sestito & S. Randich 2005), and activity-age relationships (E. E. Mamajek & L. A. Hillenbrand 2008).

K2-141NGTS-10WASP-43WASP-69WASP-80HATS-72
(Ref) Prot(\mathrm{Ref})\ P_{\mathrm{rot}} (days)(1) 15.17 −2.36+3.63^{+3.63}_{-2.36}<br>(2) 14.03 ±\pm 0.09<br>(3) 7.88 ±\pm 0.29(4) 17.29 ±\pm 0.008<br>—<br>—(5) 15.60 ±\pm 0.4<br>(3) 15.25 ±\pm 0.94<br>—(6) 23.07 ±\pm 0.16<br>(3) 20.87 ±\pm 1.75<br>(7) 18.7 ±\pm 3.5(16) 23.5 ±\pm 3.0<br>(5) 8.1 ±\pm 0.8<br>(17) 3.15 ±\pm 0.01(9) 48.725 ±\pm 0.015<br>—<br>—
P‾rot\overline{P}_{\mathrm{rot}} (days)13.49 ±\pm 0.1517.29 ±\pm 0.00815.55 ±\pm 0.6423.04 ±\pm 0.28—48.725 ±\pm 0.015
(Ref)(\mathrm{Ref}) Isochrone Age (Gyr)(10) 6.30−4.70+6.60^{+6.60}_{-4.70}<br>—<br>—(4) 10.4 ±\pm 2.5<br>—<br>—(11) 0.2 ±\pm 0.6<br>(12) 2.40 ±\pm 2.30<br>(13) 11.12 ±\pm 3.8(15) 4.675 ±\pm 4.012<br>(13) 11.21 ±\pm 1.5<br>(14) 0.758(14) 12.5<br>—<br>—(9) 12.17−0.45+0.24^{+0.24}_{-0.45}<br>—<br>—
Isochrone Age (Gyr)<br>(∗∗^{**}This Work)13.65−0.55+0.26^{+0.26}_{-0.55}10.07−1.78+1.75^{+1.75}_{-1.78}13.93−0.13+0.05^{+0.05}_{-0.13}13.97−0.05+0.02^{+0.02}_{-0.05}13.97−0.06+0.03^{+0.03}_{-0.06}13.25−0.88+0.54^{+0.54}_{-0.88}
(Ref)(\mathrm{Ref}) Gyrochrone Age (Gyr)(2) 0.740 ±\pm 0.360<br>—(8) 1.068 ±\pm 0.025<br>—(5) 0.8<br>(8) 0.769 ±\pm 0.028(15) ∗^{*}1.8 ±\pm 0.8<br>(7) 1.1 ±\pm 0.15(5) 0.2<br>(8) 1.601 ±\pm 0.202—
Gyrochrone Age (Gyr)<br>(∗∗^{**}This Work)0.86 ±\pm 0.011.79 ±\pm 0.120.99 ±\pm 0.072.93 ±\pm 0.040.0691 ±\pm 0.000312.46 ±\pm 0.01

Table 3. Stellar Rotation Periods & Ages (Isochrone vs. Gyrochrone)

Determining Isochrone & Gyrochrone Ages for the K Dwarf Sample

Because there are significant differences in the methods and resulting literature ages for our targets, we decide to conduct a new isochrone fitting and gyrochrone age analysis.

We use the spectroscopic parameters listed in Table 1 along with the Gaia DR3 parallax and magnitudes (GG, BPBP, and RPRP), Two Micron All Sky Survey (2MASS) magnitudes (JJ, HH, and KSK_S), and Wide-field Infrared Survey Explorer (WISE) magnitudes (W1W_1, W2W_2, and W3W_3) to perform an isochrone fit in order to derive the physical parameters of the host star, including the age. The spectroscopic parameters, parallax, and magnitudes are used as priors to determine the goodness of fit. We use the isochrone package (T. D. Morton 2015) to generate the MIST (MESA Isochrones and Stellar Tracks) isochrone models (J. Choi et al. 2016) used to sample the stellar parameters and find the best-fit parameters by using a Markov Chain Monte Carlo (MCMC) routine using the emcee package (D. Foreman-Mackey et al. 2013). We obtain ages consistent with our upper bound of 14 Gyr for 5 of our 6 targets, with NGTS-10 being the exception with an age of 10.1−1.8+1.710.1^{+1.7}_{-1.8} Gyr. The issue of K-dwarf isochrone ages greater than the age of the universe is not new, particularly in the case of WASP-69 (P. F. L. Maxted et al. 2015). These isochrone ages are listed at the bottom of Table 3.

We use a simple code, GPGyro to infer gyrochronological ages of single, main-sequence stars between 0.67 and 14 Gyr from effective temperature, rotation period measurements, absolute magnitude and parallax (Y. Lu et al. 2024). Effective temperature, magnitude, and parallax values for the sample are found in Table 1. Rotation periods are listed in Table 3. The effective temperature, rotation period, and absolute magnitudes are passed to GPGyro which calculates a gyrochrone age. For the targets with multiple, consistent rotation periods, we perform the gyrochrone age calculation using a weighted average of the stellar rotation periods. WASP-80 notably has three conflicting rotation periods 8.1 ±\pm 0.8 (M. Salz et al. 2015), 3.15 ±\pm 0.01 (A. Triantafillides et al. 2026), and 23.5 ±\pm 3.0 (K. Penev et al. 2018) days derived from WASP light curves, TESS light curves, and stellar-spin models respectively. We consider only the rotation periods derived from WASP-80 light curve analyses for uniformity. Because the two light-curve-derived rotation periods are quite different, we run GPGyro separately for each rotation period for this target.

Following the methods in E. Gaidos et al. (2023), we perform Monte Carlo realizations (500 points) of these calculations to estimate the uncertainty in the age, incorporating errors in rotation period, effective temperature, and the intrinsic gyrochronology relations. Histograms for the resulting age distributions are shown in Figure 3 and the mean and standard deviation of the distributions are adopted as the final age and uncertainty as listed at the bottom of Table 3.

Comparison to Kepler Field gyrochrones and GPGyro Age Distributions

Figure 3. Comparison to Kepler Field gyrochrones and GPGyro Age Distributions. Top: This stellar rotation period vs. effective temperature plot shows gyrochrone curves determined for Kepler field stars (L. G. Bouma et al. 2024). The K dwarfs analyzed in this work are overplotted to demonstrate that the adopted effective temperatures and rotation periods mostly agree with existing gyrochrones in the literature except HATS-72 which lies outside the explored parameter space and WASP-80 which appears younger on this plot than the derived gyrochrone ages suggest. Bottom: Histograms representing Monte Carlo realizations of 500 GPGyro (Y. Lu et al. 2024) runs per target. The mean (solid lines) and standard deviation (dotted lines) of each distribution are adopted as the final gyrochrone age and uncertainty.

The new gyrochrone ages for our K dwarf sample are mostly consistent with previous gyrochrone age estimates within the uncertainties. Figure 3 (top) shows

the rotation period vs. effective temperature gyrochrone curves from L. G. Bouma et al. (2024) used to determine rotation-based ages for stars in the Kepler field younger than 4 Gyr. HATS-72’s slow rotation period is inconsistent with any of the gyrochrones, indicating an age >4> 4 Gyr. WASP-69 lies right along the 2.6 Gyr gyrochrone, and this is consistent with the 2.86 ±\pm 0.38 Gyr GPGyro age derived in this work. K2-141, NGTS-10, and WASP-43 lie between the 2.6 and 1 Gyr gyrochrones, again consistent with our new estimates. Finally, our new age estimates for WASP-80, 60 Myr and 470 Myr based on the A. Triantafillides et al. (2026) and M. Salz et al. (2015) rotation periods respectively, both agree and disagree with L. G. Bouma et al. (2024) gyrochrones. The former age lies well below the 120 Myr isochrone and is consistent with a <120< 120 Myr age. The latter age lies along the 120 Myr gyrochrone, but is several hundred Myr higher, in contention.

Compared to previous isochrone age estimates, the K dwarfs appear significantly younger according to rotation-age relationships with K2-141, NGTS-10, WASP-43, and WASP-80 having ages <2< 2 Gyr. The exceptions are WASP-69 which seems to be more of a middle-aged star compared to previous gyrochrone age estimates and HATS-72 which appears to be old ∼12\sim12 Gyr regardless of the isochrone or gyrochrone age estimate.

Galactic Chemical Evolution Context & Tentative 12^{12}C/13^{13}C Trends

Now with chemical abundances and stellar ages in hand, we may use Galactic chemical evolution (GCE) models to contextualize our results and search our K dwarf sample for abundance-metallicity or abundance-age trends. In this study, we consider various state-of-the-art GCE models (D. Romano et al. 2017; N. Prantzos et al. 2018; C. Kobayashi et al. 2020) which have previously issued a call for more observational constraints from cool dwarf stars. We refer the reader to D. R. Coria et al. (2023), § 5, for a summary of these GCE models and a detailed description of their different nucleosynthetic prescriptions.

In Figure 4, we plot several Galactic chemical evolution models (D. Romano et al. 2017; N. Prantzos et al. 2018; C. Kobayashi et al. 2020) against [C/H], [O/H], and 12^{12}C/13^{13}C ratios from several cool dwarf stars (I. J. M. Crossfield et al. 2019; R. B. Botelho et al. 2020; D. González Picos et al. 2025a) including the K dwarf sample from this work. Regarding elemental carbon and oxygen abundances (top two rows in Figure 4), the younger members of our K dwarf sample appear to follow trends previously observed in solar twin

Figure 4. Elemental and Isotopic Abundance Comparison to Galactic Chemical Evolution (GCE) Models. First Row: Elemental carbon and oxygen abundances vs. [Fe/H] from our K dwarf sample, solar twins (R. B. Botelho et al. (2020)), and M dwarfs (I. J. M. Crossfield et al. (2019)), plotted against the fiducial GCE model from C. Kobayashi et al. (2020). Second Row: Same as first row, but now using time on the x-axis instead of [Fe/H]. The time axis represents the age of the universe, meaning any stellar abundances are plotted at their expected formation times. Third Row: 12C/13C^{12}\mathrm{C}/^{13}\mathrm{C} ratios derived in this work and the R. B. Botelho et al. (2020) solar twin sample plotted against four GCE models from D. Romano et al. (2017), each with a different nucleosynthetic prescription. Fourth Row: Same as previous, but now using three GCE models from C. Kobayashi et al. (2020). Fifth Row: Similar to previous, but now using two GCE models from N. Prantzos et al. (2018) and [Fe/H] replacing time on the x-axis. The cool dwarfs from this work, R. B. Botelho et al. (2020); I. J. M. Crossfield et al. (2019); D. González Picos et al. (2025a) are added back in as they all have [Fe/H] measurements, but not ages. In all the plots above, solar values are marked by gray dotted lines. We refer the reader to D. R. Coria et al. (2023), § 5, for a summary of these GCE models and a detailed description of their different nucleosynthetic prescriptions.

stars from R. B. Botelho et al. (2020). When looking at the K20 fiducial model, [O/H] is better reproduced than [C/H]. While O and Fe which have single production mechanisms (core-collapse supernovae and Type Ia supernovae, respectively), carbon is produced via two major channels: core collapse supernovae and low- to intermediate-mass asymptotic giant branch (AGB) stars. This makes [C/H] more difficult to reproduce using first-principle GCE models like Kobayashi+ 2020 which do not consider carbon yields from fast-rotating massive stars. Moreover, the dominant carbon-producing channel is still heavily debated as discussed in D. A. Boyea et al. (2026).

The 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} for our one old K dwarf, HATS-72, helps us better understand chemical evolution in the Galaxy’s early history– favoring Models 1 and 2 over Models 3 and 4 from D. Romano et al. (2017) like the rest of our K dwarf sample as seen in Figure 4. In the C. Kobayashi et al. (2020) GCE model plot from Figure 4, Model 3 without nucleosynthetic contributions from AGB and super-AGB stars significantly underestimates the 13C{}^{13}\mathrm{C}-enrichment observed in several cool dwarf stars which instead favor Models 1 and 2. Similarly, most cool dwarf 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} measurements favor N. Prantzos et al. (2018) Model 2 with nucleosynthetic prescriptions from non-rotating massive stars and non-rotating low-to-intermediate mass (NR-LIM) stars rather than Model 1 which instead considers rotating massive stars and NR-LIM stars.

From Figure 4, it is clear that 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} GCE models still need more observational constraints from cool dwarf stars, but particularly from older, low metallicity stars.

In Figure 5, we search the observed 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} ratios for abundance-metallicity-age trends. Each sample of stars is fit using a linear orthogonal distance regression (ODR) as there are uncertainties on both the x- (i.e. 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C}) and y-values (i.e. [Fe/H], stellar ages, and effective temperatures) considered. First Row: While there is a tentative trend between 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} and [Fe/H], there is no clear trend between 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} and the isochrone ages for solar twin stars like those in R. B. Botelho et al. (2020). This is inconsistent with most GCE models which predict 13C{}^{13}\mathrm{C}-enrichment with increasing [Fe/H] and younger ages (N. Prantzos et al. 2018; C. Kobayashi et al. 2020). The exceptions are Models 3 and 4 from D. Romano et al. (2017) which do predict increasing 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} ratios for younger stars. Only the K dwarfs “12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} vs. Age (Gyr)” plot (Figure 5, second row, right panel) is consistent with these model trends as the 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} ratio increases towards the younger stars in the sample. Because our K dwarfs fit the trend but not the 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} values predicted by Models 3 and 4 (Figure 4, third row), this could be indicative of inaccurate nucleosynthetic yields at earlier times in the galaxy’s history. More observational constraints could help address this inconsistency, but the 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} ratio parameter space for middle-aged dwarf stars remains largely unexplored.

Figure 5. Searching for abundance-metallicity-age trends. First Row: 12C/13C^{12}\mathrm{C}/^{13}\mathrm{C} vs. [Fe/H] (left) and Stellar Age (right) for R. B. Botelho et al. (2020) solar twin sample. Second Row: Same as above, but now for K dwarfs analyzed in this work and in Y. Zhang et al. (2024); D. González Picos et al. (2025b). Third Row: Similar to above, but for M dwarfs that only have [Fe/H] measurements, not ages. Fourth Row: 12C/13C^{12}\mathrm{C}/^{13}\mathrm{C} vs. effective temperature for all cool dwarfs plotted above. In all plots, dotted lines represent solar values while dashed lines represent the corresponding linear or parabolic orthogonal distance regression (ODR) fits, each labeled with its corresponding equation of the 12C/13C^{12}\mathrm{C}/^{13}\mathrm{C} as a function of effective temperature.

Second Row: For the K dwarf from this work and the ESO SupJup Survey (Y. Zhang et al. 2024; D. González Picos et al. 2025b) there appears to again be a trend between 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} and [Fe/H] but now opposite to that observed for solar twins (left). There also appears to be a tentative trend with age (right), but the lack of constraints from older stars makes this unreliable. While they follow the general trends predicted by the GCE models considered here, the models somewhat underestimate the 13C{}^{13}\mathrm{C}-enrichment observed in these stars.

Third Row: This M dwarf sample from D. González Picos et al. (2025a) is much larger than the K dwarf sample in this work. While the observed 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} ratios hover around the solar value, the scatter is much more pronounced for these M dwarfs than for the solar twins from R. B. Botelho et al. (2020). Nevertheless, the observed trend appears consistent with that observed for K dwarfs and most GCE models considered here.

Fourth Row: We plot 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} vs. effective temperature for the stellar samples discussed above. A parabolic orthogonal distance regression fits the data well; however, the observed trend of K dwarfs leaning more 13C{}^{13}\mathrm{C}-rich than either the solar twins or M dwarfs may not be due to their unique range of effective temperatures but rather Galactic chemical evolution as the K dwarf sample is likely younger than either the hotter or cooler stars. A couple caveats to encourage further exploration: (1) the M dwarfs from D. González Picos et al. (2025a) do not have measured ages which could alter our interpretation; (2) our K dwarf sample is quite small and does not probe the middle-aged (4-10 Gyr) parameter space, meaning the tentative trend could change or disappear with further observations; (3) we do not investigate the reliability of the isochrone ages for the R. B. Botelho et al. (2020) solar twin sample, particularly for the middle-aged stars, which again could alter our interpretation of these abundance-age trends.

Conclusion

This work demonstrates the viability of using near-infrared CO and OH lines to derive elemental carbon and oxygen abundances in K dwarf exoplanet host stars, revealing sub-solar to super-solar C/O ratios across the sample. This chemical abundance analysis has yielded novel [C/H] and [O/H] abundance ratios for K dwarf host stars NGTS-10, WASP-80, and HATS-72 and ad-

References include: (1) A. S. Bonomo et al. (2023), (2) O. Barragán et al. (2018), (3) A. W. Boyle et al. (2026), (4) J. McCormac et al. (2020), (5) M. Salz et al. (2015), (6) A. S. Bonomo et al. (2017), (7) D. R. Anderson et al. [4], (8) F. Gallet (2020), (9) J. D. Hartman et al. (2020), (10) L. Malavolta et al. (2018), (11) A. Bonfanti et al. (2015), (12) A. Bonfanti et al. (2016), (13) Y. Huang et al. (2022), (14) J. H. Hamer & K. C. Schlaufman (2022), (15) R. Allart et al. [2], (16) K. Penev et al. (2018), (17) A. Triantafillides et al. (2026).

∗^{*}This is a composite age measurement using numerous techniques including PARSEC isochrones, lithium abundances, Ca II H&K emission, metallicity-age relations, and even rotation-age relationships (R. Allart et al. [2]).

∗∗^{**}Isochrone ages in this work are calculated using the isochrone package (T. D. Morton 2015) and the emcee (D. Foreman-Mackey et al. 2013) Markov Chain Monte Carlo (MCMC) package. Gyrochronological stellar ages in this work are calculated using the GPGyro code (Y. Lu et al. 2024). We list two such ages for WASP-80 using the stellar rotation periods from A. Triantafillides et al. (2026) and M. Salz et al. (2015), respectively.

Note— This table shows the fundamental stellar parameters and carbon and oxygen abundances for the selected K dwarf sample. The top panel includes literature values for the stellar effective temperature, surface gravity and metallicity [Fe/H]. This list is not exhaustive and compiles values from discovery papers the hosts’ companion planets or host star abundance surveys. K dwarfs NGTS-10, WASP-80, and HATS-72 have no published [C/H] or [O/H] abundance ratios according to the Hypatia Catalog (N. R. Hinkel et al. 2014). The middle panel lists adopted fundamental stellar parameters for each target. Where multiple values from different references are shown for a target’s effective temperature, surface gravity, or metallicity, we adopt a weighted average of the values. The bottom panel lists the derived abundances and C/O ratios from this work. C/O ratios are calculated from Equations (7) and (8) using the solar abundances from [12]. References: (1) da Silva, R. et al. (2024); (2) A. S. Polanski et al. (2022); (3) L. Malavolta et al. (2018); (4) J. McCormac et al. (2020); (5) …

ditional carbon and oxygen constraints for K2-141 and WASP-43.

While initial efforts to extract isotopic ratios (12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C}, 16O/18O{}^{16}\mathrm{O}/{}^{18}\mathrm{O}) from 13C16O{}^{13}\mathrm{C}^{16}\mathrm{O} and 12C18O{}^{12}\mathrm{C}^{18}\mathrm{O} lines in the K band showed promise, the weak signal-to-noise and challenging normalization of these features limit their immediate utility and are perhaps best measured from M-band CO isotopologue features. Nevertheless, high-resolution NIR spectroscopy will enable robust elemental and possibly isotopic abundance measurements of cool dwarf stars using new and archival data.

Expanding this approach to a larger sample of K dwarfs hosting high-priority exoplanets will establish a much-needed baseline of host-star composition for interpreting exoplanetary atmospheric data, particularly from current and upcoming JWST and ground-based campaigns (e.g., ESO SupJup Survey Collaboration; KPIC Collaboration; and other high-resolution, cross-correlation spectroscopy studies of exoplanets).

Beyond C/O Ratios

This work focuses primarily on the elemental and isotopic ratios of carbon and oxygen because carbon- and oxygen-bearing molecules are the most commonly detected in the atmospheres of giant exoplanets. Atmospheric retrievals on JWST NIRCam/MIRI/NIRSpec data from the exoplanet companions to our K dwarf host stars have detected and constrained the volume mixing ratios for different molecules including H2O\mathrm{H_2O}, CO, CO2\mathrm{CO_2}, CH4\mathrm{CH_4}, and most notably, nitrogen- and sulfur-bearing molecules like NH3\mathrm{NH_3}, SO2\mathrm{SO_2} and CS2\mathrm{CS_2} (NGTS-10 b: V. Parmentier et al. (2026); WASP-43 b: J. Yang et al. (2024); WASP-69 b: G. Guilluy et al. (2022); E. Schlawin et al. (2024b); WASP-80 b: A. Triantafillides et al. (2026)).

Multiple studies find that complementary stellar and exoplanetary C/O ratios alone provide non-unique formation and accretion scenarios ((H. Reggiani et al. 2022; D. R. Coria et al. 2024)), but C/O, C/N, N/O, S/N, and 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} ratios combined can break degeneracies in planet’s formation, migration, and accretion histories ((K. I. Öberg et al. 2011; B. Bitsch et al. 2022; E. Pacetti et al. 2022; I. J. M. Crossfield 2023; D. R. Coria et al. 2024; J. W. Xuan et al. 2026)).

Complementary host star nitrogen abundances are particularly lacking for K/M dwarf stars as both atomic nitrogen and molecular NH lines at optical wavelengths are inaccessible due to strong molecular absorption. A recent study ((N. Hejazi et al. 2023)) has shown that [N/H] can be derived for late K dwarf stars in the H band using CN lines, but it is yet unclear whether this technique applies to cooler M dwarfs. Stellar sulfur abundance measurements are challenging for the similar reasons; however, they may may be possible using Y-band atomic sulfur lines, H+K band H2S\mathrm{H_2S} lines or L+M band SO lines by addressing the difficult continuum/pseudocontinuum normalization or lack of high-quality near- to mid-infrared (Y, L and M band) spectra for these cool dwarf stars.

Regarding exoplanet abundance constraints, 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} ratios remain largely unexplored although retrievals on both space- and ground-based observations have demonstrated their capability of providing constraints ((M. R. Line et al. 2021; Y. Zhang et al. 2021b; D. González Picos et al. 2025a; J. C. Costes et al. 2024; C. E. Hood et al. 2024; S. de Regt et al. 2024; S. Gandhi et al. 2023; L. Finnerty et al. 2024; D. González Picos et al. 2026a)).

These complementary stellar elemental abundance measurements provide crucial context for interpreting the atmospheric compositions and formation pathways of their giant planet companions and ultimately may provide evidence for outer disk formation and accretion for these hot Jupiters.

ACKNOWLEDGMENTS

The authors thank the anonymous referee for their thorough review and invaluable discussion, which substantially improved this manuscript. The authors would also like to thank Chiaki Kobayashi and Darío González Picos for providing their GCE models and M dwarf 12C/13C{}^{12}\mathrm{C}/{}^{13}\mathrm{C} measurements, respectively. D. C. would like to acknowledge Yuxi (Lucy) Lu for their attentive help in using and troubleshooting the GPgyro code for determining gyrochronological ages.

D. C. and D. D. would like to thank the Robert E. Young Origins of the Universe Chair fund for its generous support. This work was supported in part by the National Science Foundation (NSF) under grant AST-2441920.

This work has made use of data from the European Space Agency (ESA) mission Gaia (https://www.cosmos.esa.int/gaia), processed by the Gaia Data Processing and Analysis Consortium (DPAC, https://www.cosmos.esa.int/web/gaia/dpac/consortium). Funding for the DPAC has been provided by national institutions, in particular the institutions participating in the Gaia Multilateral Agreement.

This research has made use of the Astrophysics Data System, funded by NASA under Cooperative Agreement 80NSSC21M00561. This research has made use of the NASA Exoplanet Archive, which is operated by the California Institute of Technology, under contract with the National Aeronautics and Space Administration under the Exoplanet Exploration Program. N.H. acknowledges support from the Rubin/Chile Joint Research Program, funded by the U.S. National Science Foundation (NSF), and thanks the University of Tarapacá for hosting this research.

This work used The Immersion Grating Infrared Spectrometer (IGRINS) was developed under a collaboration between the University of Texas at Austin and the Korea Astronomy and Space Science Institute (KASI) with the financial support of the US National Science Foundation under grants AST-1229522, AST-1702267 and AST-1908892, McDonald Observatory of the University of Texas at Austin, the Korean GMT Project of KASI, the Mt. Cuba Astronomical Foundation and Gemini Observatory. The RRISA is maintained by the IGRINS Team with support from McDonald Observatory of the University of Texas at Austin and the US National Science Foundation under grant AST-1908892.

Facilities: Gemini:South, LDT

OH, CO, and 13CO^{13}\mathrm{CO} Line List: Normalization and Fit Parameters

**Note—**Wavelength units are A˚\mathring{A} (= 0.1 nm). — denotes a dimensionless quantity. Flag values indicate the number of exclusion zones applied by the continuum selection routine (0, 1, or 2). The cell value ”999999” is a null value.

Line-by-Line Relative Abundances

Figure B1. Histograms for line-by-line abundance measurements for oxygen (top, blue), carbon (middle, red), and 12C/13C^{12}\mathrm{C}/^{13}\mathrm{C} (bottom, green) The solid black lines represent the weighted average of the abundance values, weighted by statistical uncertainty (Δχ2=1\Delta\chi^{2} = 1) for each measurement. The dotted lines represent the standard deviation (±1σ\pm1\sigma) bounds. We are unable to determine the 12C/13C^{12}\mathrm{C}/^{13}\mathrm{C} ratio for WASP-80. Final [O/H], [C/H], and 12C/13C^{12}\mathrm{C}/^{13}\mathrm{C} ratios are given in Table 4.

References

  1. [1]Adams Redai, J., Wogan, N., Wallack, N. L., et al. 2025, The Astronomical Journal, 170, 219, doi: 10.3847/1538-3881/adee92DOI
  2. [2]Allart, R., Carteret, Y., Bourrier, V., et al. 2025, A&A, 700, A7, doi: 10.1051/0004-6361/202452525DOI
  3. [3]Adibekyan, V., Sousa, S. G., Santos, N. C., et al. 2012, A&A, 545, A32, doi: 10.1051/0004-6361/201219401DOI
  4. [4]Anderson, D. R., Collier Cameron, A., Delrez, L., et al. 2014, MNRAS, 445, 1114, doi: 10.1093/mnras/stu1737DOI
  5. [5]Ahrer, E.-M., Radica, M., Piauleт-Ghorayeb, C., et al. 2025, ApJL, 985, L10, doi: 10.3847/2041-8213/add010DOI
  6. [6]Angus, R., Aigrain, S., Foreman-Mackey, D., & McQuillan, A. 2015, MNRAS, 450, 1787, doi: 10.1093/mnras/stv423DOI
  7. [7]Alam, M. K., Gao, P., Adams Redai, J., et al. 2025, AJ, 169, 15, doi: 10.3847/1538-3881/ad8eb5DOI
  8. [8]Angus, R., Morton, T. D., Foreman-Mackey, D., et al. 2019, AJ, 158, 173, doi: 10.3847/1538-3881/ab3c53DOI
  9. [9]Alderson, L., Batalha, N. E., Wakeford, H. R., et al. 2024, AJ, 167, 216, doi: 10.3847/1538-3881/ad32c9DOI
  10. [10]Arney, G. N. 2019, ApJL, 873, L7, doi: 10.3847/2041-8213/ab0651DOI
  11. [11]Alderson, L., Moran, S. E., Wallack, N. L., et al. 2025, AJ, 169, 142, doi: 10.3847/1538-3881/adad64DOI
  12. [12]Asplund, M., Amarsi, A. M., & Grevesse, N. 2021, A&A, 653, A141, doi: 10.1051/0004-6361/202140445DOI

Paper details

vlnc:W_27kf4jxw
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents