HD 72659 c: Brown Dwarf Desert Companion

• HD 72659 c is an outer substellar companion to a solar analog, residing in the brown dwarf desert with a mass of ~19 MJ and a semimajor axis of ~21.5 au.
• It was confirmed through a joint multi-technique analysis combining long-term radial velocity data, Gaia DR3 astrometry, and SPHERE direct-imaging constraints.
• Dynamical studies indicate significant Kozai–Lidov oscillations, revealing complex secular interactions with an inner gas giant that refine our understanding of multi-body systems.

HD 72659 c is an outer substellar companion to the solar analog HD 72659, confirmed through a joint analysis of radial velocities, Gaia DR3 high-precision astrometry, and direct-imaging constraints in "The GAPS Programme at TNG LXVIII. Characterization of the outer substellar companion around HD 72659 with a multi-technique approach" (Ruggieri et al., 25 Jul 2025). The system was already known to host a gas giant on a $\sim 10$-yr orbit, while the outer object had been identified as a massive companion whose true nature depended on breaking the radial-velocity mass–inclination degeneracy. The combined solution yields a companion with $M_{\rm c}=19.4^{+0.8}_{-0.5}\,M_{\rm J}$, $a_{\rm c}=21.5^{+0.5}_{-0.4}$ au, and $P_{\rm c}=97.1^{+3.4}_{-2.5}$ yr, placing it in the Brown Dwarf desert and making the system a case of particular interest for secular dynamics, including possible Kozai–Lidov forcing (Ruggieri et al., 25 Jul 2025).

1. Identification and astrophysical setting

HD 72659 is described as a solar analog with two known companions: an inner gas giant, conventionally designated planet b, and the outer substellar companion HD 72659 c (Ruggieri et al., 25 Jul 2025). The outer object was the subject of confirmation and characterization using a multi-technique approach that combined long-baseline RV monitoring, proper-motion anomaly measurements from Gaia DR3 and Hipparcos, and high-contrast imaging constraints.

The principal astrophysical significance of HD 72659 c is its placement in the Brown Dwarf desert. In the reported joint solution, the companion mass is well above the planetary minimum-mass regime yet below the hydrogen-burning stellar domain, with a semimajor axis in the outer-system range. This combination makes the system relevant both to substellar-companion demographics and to dynamical studies of hierarchical architectures around Sun-like stars (Ruggieri et al., 25 Jul 2025).

The study explicitly aimed to confirm HD 72659 c, which had been recently announced from HIRES and HARPS data combined with Gaia astrometry. The new work confirms the literature mass scale for the object while revising its orbital period upward by a factor of about two relative to the earlier report (Ruggieri et al., 25 Jul 2025). This revision is central to the present understanding of the system.

2. Observational basis

The characterization of HD 72659 c relies on four RV datasets, astrometric proper-motion anomalies, and SPHERE direct imaging (Ruggieri et al., 25 Jul 2025). The RV coverage is unusually extended in time and heterogeneous in instrumental origin, which is essential for constraining a companion on a multidecadal orbit.

Data type	Instrument or source	Reported coverage or property
RVs	HIRES/Keck	66 observations from 1998-01-25 to 2019-10-24 (21.74 yr), median RV uncertainty 1.44 m s$^{-1}$
RVs	HARPS/ESO	65 spectra from 2004-02-23 to 2021-06-05 (17.28 yr), median RV uncertainty 1.86 m s$^{-1}$
RVs	HARPS-N/TNG	91 spectra from 2013-01-03 to 2023-04-11 (10.27 yr), median RV uncertainty 0.62 m s$^{-1}$
RVs	HRS/HET	32 high-uncertainty RVs from 2004-12-03 to 2007-11-27 (2.96 yr), median uncertainty 9.14 m s$^{-1}$
Astrometry	Gaia DR3 and Hipparcos PMa from Brandt (2021)	Gaia epoch 2016.0, Hipparcos epoch 1991.25
Direct imaging	SPHERE/VLT IRDIFS and IRDIFS_EXT	Observations on 2015-12-31 and 2024-01-21

The HARPS dataset is split into pre-upgrade and post-upgrade subsets because a fiber-upgrade in May 2015 altered the instrumental configuration (Ruggieri et al., 25 Jul 2025). The HARPS-N sequence, obtained under the GAPS project, is especially precise, with a reported median RV uncertainty of 0.62 m s$^{-1}$; one outlier at 4.37 m s$^{-1}$ was removed (Ruggieri et al., 25 Jul 2025).

The astrometric component consists of proper-motion anomalies from Gaia DR3 and Hipparcos. The reported values are $M_{\rm c}=19.4^{+0.8}_{-0.5}\,M_{\rm J}$0 mas yr$M_{\rm c}=19.4^{+0.8}_{-0.5}\,M_{\rm J}$1, $M_{\rm c}=19.4^{+0.8}_{-0.5}\,M_{\rm J}$2 mas yr$M_{\rm c}=19.4^{+0.8}_{-0.5}\,M_{\rm J}$3, $M_{\rm c}=19.4^{+0.8}_{-0.5}\,M_{\rm J}$4 mas yr$M_{\rm c}=19.4^{+0.8}_{-0.5}\,M_{\rm J}$5, and $M_{\rm c}=19.4^{+0.8}_{-0.5}\,M_{\rm J}$6 mas yr$M_{\rm c}=19.4^{+0.8}_{-0.5}\,M_{\rm J}$7 (Ruggieri et al., 25 Jul 2025). These measurements provide the astrometric lever arm needed to infer the true companion mass.

The direct-imaging observations were obtained with SPHERE/VLT in two epochs. On 2015-12-31, a SHINE filler observation used IRDIFS with YJH IFS plus H-band IRDIS, total integration 1024 s, field rotation 9.6°, and seeing 0.72″. On 2024-01-21, a GAPS follow-up used IRDIFS_EXT with YJH IFS plus K1K2 IRDIS, 48×64 s = 3072 s, rotation 28.49°, and seeing 0.72″ (Ruggieri et al., 25 Jul 2025).

3. Radial-velocity and astrometric inference

The RV analysis is framed in the standard Keplerian form for the semi-amplitude $M_{\rm c}=19.4^{+0.8}_{-0.5}\,M_{\rm J}$8 of a companion of mass $M_{\rm c}=19.4^{+0.8}_{-0.5}\,M_{\rm J}$9 orbiting a star of mass $a_{\rm c}=21.5^{+0.5}_{-0.4}$0 with period $a_{\rm c}=21.5^{+0.5}_{-0.4}$1 and eccentricity $a_{\rm c}=21.5^{+0.5}_{-0.4}$2:

$a_{\rm c}=21.5^{+0.5}_{-0.4}$3

The corresponding minimum-mass relation is written through the mass function:

$a_{\rm c}=21.5^{+0.5}_{-0.4}$4

In this formulation, RVs alone determine only $a_{\rm c}=21.5^{+0.5}_{-0.4}$5, not the true mass (Ruggieri et al., 25 Jul 2025).

The mass–inclination degeneracy is broken by astrometry. The reported relation for the sky-projected semimajor axis of the stellar reflex motion is

$a_{\rm c}=21.5^{+0.5}_{-0.4}$6

where $a_{\rm c}=21.5^{+0.5}_{-0.4}$7 is in au and $a_{\rm c}=21.5^{+0.5}_{-0.4}$8 in pc. Measuring $a_{\rm c}=21.5^{+0.5}_{-0.4}$9 in mas yields

$P_{\rm c}=97.1^{+3.4}_{-2.5}$0

while the inclination is fixed by

$P_{\rm c}=97.1^{+3.4}_{-2.5}$1

This is the methodological core of the HD 72659 c confirmation: RVs constrain the orbital signal and astrometry supplies the missing geometric information required for the true mass (Ruggieri et al., 25 Jul 2025).

The final orbital solution was obtained from a joint fit of RVs and Gaia DR3 PMa using DE-MCMC, with methodological references given to Sozzetti (2023) and Ruggieri et al. (2024b) (Ruggieri et al., 25 Jul 2025). This places the inference in a standard Bayesian orbital-fitting framework appropriate for sparse-orbit, long-period companions.

4. Derived orbital and physical parameters

The joint RV + Gaia DR3 PMa fit yields the following posterior medians and $P_{\rm c}=97.1^{+3.4}_{-2.5}$2 intervals for HD 72659 c (Ruggieri et al., 25 Jul 2025).

Parameter	Posterior (median $P_{\rm c}=97.1^{+3.4}_{-2.5}$3)	Prior
$P_{\rm c}=97.1^{+3.4}_{-2.5}$4 [yr]	$P_{\rm c}=97.1^{+3.4}_{-2.5}$5	U(30,150)
$P_{\rm c}=97.1^{+3.4}_{-2.5}$6 [au]	$P_{\rm c}=97.1^{+3.4}_{-2.5}$7	U(10,30)
$P_{\rm c}=97.1^{+3.4}_{-2.5}$8	$P_{\rm c}=97.1^{+3.4}_{-2.5}$9	U(0,1)
$^{-1}$0 [°]	$^{-1}$1	U(0,360)
$^{-1}$2 [°]	$^{-1}$3 (retrograde)	U(cos)
$^{-1}$4 [°]	$^{-1}$5	U(0,360)
$^{-1}$6 [$^{-1}$7]	$^{-1}$8	derived

The reported epoch parameter is $^{-1}$9 BJD, with prior $^{-1}$0 (Ruggieri et al., 25 Jul 2025). The orbit is moderately eccentric, and the inclination is reported as retrograde. The mass is consistent with a brown-dwarf classification in the substellar sense used by the study.

A central result is the discrepancy in period relative to Feng et al. (2022). That earlier work reported $^{-1}$1 and $^{-1}$2 d ($^{-1}$3 yr) from Gaia-Hipparcos PMa plus HIRES and HARPS data. The newer analysis finds $^{-1}$4 yr, described as twice longer and a $^{-1}$5 discrepancy, while the mass agrees within $^{-1}$6 (Ruggieri et al., 25 Jul 2025). The paper attributes the revised period to the extended RV baseline of 25 yr and refined RV + DR3 astrometry.

The study also states that the parameters of planet b are refined with reduced uncertainties compared to previous works (Ruggieri et al., 25 Jul 2025). No numerical values for the revised planet-b fit are provided in the supplied data, so the significance of this refinement is primarily contextual: improved knowledge of the inner orbit is necessary for any robust secular analysis of the coupled system.

5. Direct-imaging non-detection and mass limits

SPHERE imaging did not detect HD 72659 c directly (Ruggieri et al., 25 Jul 2025). The 2024 SPHERE IRDIFS_EXT dataset was converted into 5$^{-1}$7 contrast-based mass limits using AMES-COND at 8 Gyr, yielding the following thresholds: at 0.3″ (6.5 au), $^{-1}$8; at 0.6″ (13 au), $^{-1}$9; at 1″ (21 au, the projected location), $^{-1}$0; and beyond 1.5″, $^{-1}$1 (Ruggieri et al., 25 Jul 2025).

These limits are entirely consistent with the non-detection because the dynamically inferred mass is $^{-1}$2, well below the imaging sensitivity floor at the expected projected separation (Ruggieri et al., 25 Jul 2025). The imaging data therefore do not independently measure the companion mass, but they do exclude a much more massive and brighter object at the relevant separations.

A common misconception in long-period companion studies is that a direct-imaging non-detection argues against the companion’s existence. In this case, the published limits indicate the opposite: the non-detection is expected under the favored dynamical solution because the companion is too faint, given the assumed age and model mapping, to exceed the SPHERE detection threshold (Ruggieri et al., 25 Jul 2025). This suggests that the imaging data function chiefly as an upper-bound consistency check rather than a discovery channel.

6. Dynamical configuration and secular evolution

The dynamical analysis was performed by numerically integrating the system with the RADAU integrator, using the nominal masses and orbits reported in the study (Ruggieri et al., 25 Jul 2025). Under the assumption of a mutual inclination $^{-1}$3 between b and c, the integrations show classical Kozai–Lidov oscillations. Specifically, the eccentricity of the inner planet cycles between $^{-1}$4 and $^{-1}$5, while its inclination oscillates from prograde ($^{-1}$6) through $^{-1}$7 into retrograde ($^{-1}$8) (Ruggieri et al., 25 Jul 2025).

The study gives the critical Kozai angle as

$^{-1}$9

Because $^{-1}$0, large eccentricity–inclination exchange is enabled (Ruggieri et al., 25 Jul 2025). No chaotic diffusion is seen over 100 Myr, and secular theory is said to reproduce these cycles qualitatively.

This dynamical architecture is important because it provides a mechanism for explaining the moderate eccentricity of the inner planet. The paper explicitly states that Kozai–Lidov cycles may explain the moderate $^{-1}$1, while a past planet–planet scattering event or a stellar fly-by could have excited the mutual inclination (Ruggieri et al., 25 Jul 2025). These latter scenarios are not presented as direct detections; they are possible pathways for producing the current secular configuration.

A plausible implication is that HD 72659 is a dynamically mature hierarchical system in which the present-day orbital architecture preserves information about earlier excitation processes. That inference remains conditional, because the true mass and orbital plane of planet b are not yet directly constrained in the supplied analysis.

7. Literature context and prospective constraints

The paper places HD 72659 c within a line of earlier work that progressively narrowed the allowed parameter space for the outer companion (Ruggieri et al., 25 Jul 2025). Bryan et al. (2016) had reported lower limits of $^{-1}$2 and $^{-1}$3–35 au from RV trends; the current solution lies in the upper part of that semimajor-axis range and is described as firmly in the brown-dwarf desert (Ruggieri et al., 25 Jul 2025). Feng et al. (2022) then reported a mass close to the current value but with a substantially shorter period. The new analysis revises the orbital timescale while preserving mass consistency at the $^{-1}$4 level.

The broader interest of the system derives from the coexistence of two gas-giant/substellar companions around a Sun-like star with $^{-1}$5 (Ruggieri et al., 25 Jul 2025). This makes HD 72659 a useful target for studies of long-term secular coupling, hierarchical stability, and the boundary region between giant planets and brown dwarfs.

The study identifies several future observational paths. Continued HARPS-N monitoring will cover a larger phase of HD 72659 c’s orbit, and future Gaia releases are expected to detect HD 72659 b astrometrically, which would constrain its true mass and the mutual inclination (Ruggieri et al., 25 Jul 2025). The paper further notes that JWST and ELT prospects are limited by age and brightness but could probe $^{-1}$6 (Ruggieri et al., 25 Jul 2025). This suggests that the decisive near-term gains are more likely to come from longer-baseline precision astrometry and RVs than from direct imaging alone.