CoCoNUTS: Wide-Orbit Companion Survey

Updated 3 July 2026

CoCoNUTS is a large-scale, volume-limited survey that identifies wide-orbit planetary and substellar companions using precise Gaia and infrared data to establish well-defined benchmark systems.
The program employs multi-wavelength spectrophotometry and advanced spectral retrieval techniques (e.g., using JWST and Gemini) to robustly determine atmospheric properties, isotopic ratios, and evolutionary parameters.
Its discoveries, including planetary-mass companions at extreme separations, challenge conventional formation models and refine our understanding of cloud physics, chemical disequilibrium, and substellar evolution.

CoCoNUTS refers to the COol Companions ON Ultrawide orbiTS survey and its resulting discoveries: a systematically constructed population of wide-orbit planetary and substellar companions, currently serving both as key benchmark systems for low-temperature astrophysics and as a crucial resource for planet formation theory. Since its inception, the COCONUTS survey has led to the direct imaging and rigorous physical characterization of benchmark systems spanning the late-M, L, T, and Y spectral types at projected separations up to ∼10,000 astronomical units (au), with emphasis on planetary-mass companions in the T/Y regime such as COCONUTS-2b. COCONUTS has enabled precise constraints on atmospheric properties, evolutionary models, isotopic abundances, and population statistics by leveraging multi-wavelength spectrophotometry and host-star determinations of age and metallicity (Zhang et al., 2021, Zhang et al., 2020, Zhang et al., 2022).

1. Survey Motivation and Design

COCONUTS was designed as a large-scale, volume-limited survey targeting wide-orbit planetary-mass and substellar companions (≳500 au) to nearby stars (<100 pc) (Zhang et al., 2020, Zhang et al., 2022, Zhang et al., 2021). Its primary goals are:

To establish a well-defined sample of ultracool benchmarks (mass 5–70 MJup) with precisely determined ages and metallicities from their host stars.
To provide ground truth for substellar atmosphere and evolution models by enabling direct spectral and photometric comparisons across a broad temperature, gravity, and metallicity regime.
To probe the occurrence rate and orbital architecture of wide companions, illuminating formation channels inaccessible to classical core accretion models.

Target selection uses cross-matching of primary stars from Gaia (DR2/EDR3) within 100 pc against deep infrared sky surveys (Pan-STARRS1, AllWISE, CatWISE2020, UKIDSS, UHS, VHS), seeking co-moving ultracool candidates within 10,000 au and requiring proper motion and parallax consistency to within 3–5σ. Initial selection is based on color-magnitude cuts, followed by spectroscopic confirmation and full kinematic association (Zhang et al., 2020, Zhang et al., 2022).

2. Key Discoveries: Systems and Physical Properties

COCONUTS has led to the direct imaging of benchmark companions across multiple systems:

COCONUTS-1: White dwarf (5115 K) + T4 brown dwarf at 1280 au. COCONUTS-1B characterized at Teff = 1255 K, log g = 5.44 (evolution-based), M = 69 MJup, showing photometric/spectral anomalies indicative of condensate clouds and CO non-equilibrium chemistry (Zhang et al., 2020).
COCONUTS-2: M3V (10.9 pc) + T9/T9.5 planetary-mass companion at 6471 au, Teff ≈ 480–495 K, M ≈ 7.3–8.0 MJup, log g ≈ 4.2–4.3, R ≈ 1.03–1.11 RJup, [M/H] ≈ –0.12–0.0. COCONUTS-2b is among the coldest and widest-separation directly imaged planets, providing reference points for atmospheric and formation models (Zhang et al., 2021, Zhang et al., 2024, Ravet et al., 8 Apr 2026, Kühnle et al., 29 Apr 2026).
COCONUTS-3: M5.5 + very red L6 INT-G companion at 1891 au, with the companion inferred at Teff = 1362 K, log g = 4.96, M = 39 MJup, exhibiting an extremely red J–K color (2.11 mag) and evidence for abundant condensate clouds (Zhang et al., 2022).

Each benchmark system provides empirical anchors for age, metallicity, and evolutionary status, enabling detailed physical and atmospheric property estimation.

3. Atmospheric Characterization and Modeling

The COCONUTS sample, particularly COCONUTS-2b, is characterized using multi-instrument panchromatic coverage (1–15 μm or wider) with Gemini/FLAMINGOS-2, JWST/NIRSpec, JWST/MIRI-LRS/MRS, WISE, and Spitzer (Zhang et al., 2021, Zhang et al., 2024, Ravet et al., 8 Apr 2026, Kühnle et al., 29 Apr 2026). Key analysis steps include:

Forward modeling and retrieval using state-of-the-art grids: ATMO2020(++), Sonora Diamondback, Sonora Elf Owl, Exo-REM, with self-consistent treatment of equilibrium/disequilibrium chemistry, vertical mixing (Kzz ≈ 10⁶–10⁷ cm²/s), diabatic temperature-pressure profiles, and cloud physics (Zhang et al., 2024, Ravet et al., 8 Apr 2026).
Robust molecular identification (H2O, CH4, NH3) via high S/N JWST spectra and cross-correlation with opacity templates, achieving >4σ detections of key bands (e.g., H2O at 6.3 μm, CH4 ν3 at 3.3 μm and ν4 at 7.7 μm, NH3 at 10.5 μm) (Ravet et al., 8 Apr 2026).
Spectral retrieval frameworks using nested sampling (ForMoSA, PyMultiNest) and, critically, incorporating Gaussian Process covariance models to account for wavelength-correlated residuals arising from instrumental systematics and model deficiencies (Ravet et al., 8 Apr 2026).
All leading models require non-equilibrium chemistry and/or diabatic structure and clouds to explain observed spectra, particularly the strong CH4 bands and mid-IR flux redistribution. Cloudless models systematically underpredict observed Y- and J-band fluxes, indicating deficiencies in line profiles and alkali condensation/rainout treatments (Zhang et al., 2024, Zhang et al., 2020, Zhang et al., 2022).

Best-fit physical/chemical parameters for COCONUTS-2b fall within Teff = 483–496 K, log g = 4.19–4.30 dex, R ≈ 1.03–1.11 RJup, [M/H] ≈ –0.12–0.0, C/O ≈ 0.4–0.5, and M ≈ 7–8 MJup (Zhang et al., 2024, Ravet et al., 8 Apr 2026).

4. Mass, Age, and Evolutionary Status

Total system ages are established via stellar activity indicators, rotation period, and comparison to moving group populations (e.g., Ursa Major Corona, t = 414 ± 23 Myr for COCONUTS-2 system) (Kiman et al., 25 Nov 2025). Companion mass is derived by interpolating measured luminosity against evolutionary cooling tracks (ATMO2020, Sonora Bobcat/Diamondback grids, hot/cold start assumptions).

In all recent COCONUTS-2b work, independent methods converge to M = 7.3 ± 0.3 MJup, logL/L⊙ ≈ –6.17 dex, reinforcing planetary-mass status and establishing these as among the coldest objects with precisely determined mass, radius, age, metallicity, and atmospheric properties (Ravet et al., 8 Apr 2026, Zhang et al., 2024). The inferred mass ratio (q ≈ 0.02) places COCONUTS-2b well within the planetary regime and disfavours formation via classical core accretion at such extreme separations; formation via cloud fragmentation or disk gravitational instability is favored (Zhang et al., 2024, Zhang et al., 2020).

5. Isotopic and Compositional Diagnostics

COCONUTS-2b is the first planetary-mass companion with robust measurements of minor isotopologue abundances, achieved through full spectral resolution retrieval on JWST/MIRI/MRS data (Kühnle et al., 29 Apr 2026):

Clear detections of H₂¹⁸O (log ℬ=72.5), H₂¹⁷O (25.8), and ¹⁵NH₃ (24.7) via Bayesian nested sampling and leave-one-out Bayes factor analysis.
Derived oxygen and nitrogen isotope ratios: ¹⁶O/¹⁸O = 256⁺²⁹₋₂₅, ¹⁶O/¹⁷O = 934⁺¹⁷⁴₋¹³⁹, ¹⁴N/¹⁵N = 324⁺⁴⁶₋⁴⁰. The O isotopic ratios indicate heavy-isotope enrichment relative to solar and mean ISM; the N ratio is ISM-like.
These isotopologue diagnostics provide a new pathway to trace formation scenarios, comparing disk or cloud inheritance and possible fractionation/snowline processing effects.

The derived global metallicity and C/O ratios ([M/H] = –0.12⁺⁰.⁰¹₋⁰.⁰², C/O = 0.45±0.02) for COCONUTS-2b, together with its lack of heavy-element enrichment, support a stellar-like formation channel and limited core erosion or planetesimal pollution.

6. Population Context, Evolution, and Formation Pathways

The COCONUTS sample covers a wide range of system architectures (see table below), extending benchmarks for atmospheric and evolutionary model calibration beyond what was previously accessible.

System	Primary	Companion	Sep (au)	SpT (BD/comp)	Teff (K)	log g (dex)	Mass (MJup)	[M/H]
COCONUTS-1	White dwarf	T4 BD	1280	T4	1255	5.44	69	[not stated]
COCONUTS-2	M3V	T9/T9.5 exoplanet	6471	T9.5	483–496	4.19–4.30	7–8	–0.12–0.0
COCONUTS-3	M5.5	L6 INT-G BD	1891	L6	1362	4.96	39	0.21 ± 0.07

In COCONUTS-2, the extreme separation (6471 au) makes in-situ core accretion untenable; current data and chemical signatures favor formation via gravitational fragmentation followed by outward migration or scattering, or direct collapse analogous to binary star formation. These wide, low-mass, low-metallicity companions challenge traditional planet formation models and serve as critical probes of late-stage disk and cloud evolution (Zhang et al., 2024, Zhang et al., 2020, Kiman et al., 25 Nov 2025).

7. Impact on Benchmarking and Model Development

The COCONUTS program has produced the most tightly age- and metallicity-anchored benchmarks for late-M, L, T, and Y dwarfs at planetary and brown dwarf masses, providing an empirical basis to evaluate and falsify atmospheric and evolutionary models (Zhang et al., 2020, Zhang et al., 2022). Key outcomes include:

Identification and quantification of systematic errors in cloud prescriptions and alkali line opacities, especially at low gravity and low temperature (Zhang et al., 2020, Zhang et al., 2022).
Demonstration of gravity-dependent color and luminosity effects across the L/T transition, with larger amplitude J- and H-band brightening and cooler effective temperatures for young, low-gravity objects (Zhang et al., 2020).
Panchromatic JWST+ground-based spectroscopy sets a new standard for exoplanet atmosphere retrieval, including robust treatment of instrument/model systematics (Gaussian Process likelihoods), and enables isotopologue abundance measurements previously beyond reach (Ravet et al., 8 Apr 2026, Kühnle et al., 29 Apr 2026).
COCONUTS systems form the reference set for next-generation atmospheric, population, and planet formation studies, anchoring evolutionary tracks, cloud microphysics, chemical disequilibrium, and isotopic evolution for planetary-mass companions in extreme environments.

Subsequent COCONUTS discoveries, expanded metallicity/age coverage, and continued JWST+ELT follow-up will deepen constraints on cold planet and brown dwarf demographics and formation channel diversity.