A Mini-Neptune from TESS and CHEOPS Around the 120 Myr Old AB Dor Member HIP 94235

Young planets offer a time-lapse view of the construction of the exoplanet demographics. Planetary systems are thought to undergo rapid evolution within the first hundreds of millions of years after their formation. Follow-up characterization of small young planets helps to test our models for the contraction and mass-loss processes that they undergo during this time frame.

Thousands of close-in Neptunes and super-Earths were discovered by the primary Kepler mission (e.g., Petigura et al. 2013; Burke et al. 2015; Zhu et al. 2018). The mechanisms that sculpted the period–radius distribution of these planets can shed light on the early formation and evolution of planetary systems. The evaporation of primordial hydrogen and helium envelopes, driven by UV and X-ray radiation from young stars, can reproduce the sub-Neptune desert and radius valley (e.g., Lopez et al. 2012; Owen & Wu 2013, 2017). These processes act on rapid timescales because the high-energy fluxes from young stars rapidly decline over time. Mass loss can also be driven by heat leaking out from a planet's deep interior, a process that can last hundreds of millions of years as the planets cool down after accretion (e.g., Lopez & Fortney 2013; Ginzburg et al. 2018; Gupta & Schlichting 2021). Giant impacts within compact super-Earth and Neptune systems may erode the envelopes of some planets, occurring after disk dissipation and before the systems dynamically cool within the first hundred million years (e.g., Marcus et al. 2009; Inamdar & Schlichting 2015). As their primordial gaseous envelopes are stripped away, it is possible that some planets lying in less energetic environments may be replenished by secondary atmospheres (e.g., Kite & Barnett 2020). Other mechanisms, such as in situ formation in gas-poor disks, may naturally carve out the current period–radius distribution of small planets (Lee & Connors 2021; Lee et al. 2022), and the radii of small planets may evolve much slower than predicted from runaway mass-loss models (David et al. 2021).

Young planets can help establish the timescales for the evolution of the period–radius distribution of the super-Earth and Neptune populations. For some planets, mass loss through photoevaporation occurs throughout their lifetime without significantly changing their radii. Extended atmospheres have been observed for Neptune-sized planets about older field stars in X-rays (Ehrenreich et al. 2012), Lyα (Kulow et al. 2014; Ehrenreich et al. 2015; Lavie et al. 2017; Bourrier et al. 2018), and He i (Spake et al. 2018; Allart et al. 2019; Kirk et al. 2020). However, for planets between 2 and 4 R_⊕, runaway evaporation may occur. This process strips away the outer primordial envelope of these Neptunes and super-Earths, leaving behind rocky cores. Real-time measurements of photoevaporation for younger systems can help establish the timescale for this process and its influence on the radius evolution of planets. Zhang et al. (2022b) detected the Lyα transit of the outer planet HD 63433 c in a 400 Myr old planetary system. They also reported the lack of a detectable escaping hydrogen atmosphere for the inner planet, suggesting that it may have already undergone runaway evaporation and lost its primordial atmosphere. Rockcliffe et al. (2021) reported a nondetection of escaping hydrogen for the 650 Myr old K2-25b, with one possibility for the nondetection being factors related to the star's youth. Zhang et al. (2022a) reported escaping He i for the 500 Myr old mini-Neptune HD 73583 b, showing an excess 0.68% absorption for the He i 10830 Å line. Tentative detections of atmospheric escape have been reported for planets in the V1298 Tau system in Ca ii, H-α (Feinstein et al. 2021), and He i (Gaidos et al. 2022).

The K2 mission (Howell et al. 2014) yielded some of the first young transiting planets (e.g., Mann et al. 2016a, 2016b; David et al. 2016; Rizzuto et al. 2018; Vanderburg et al. 2018; David et al. 2019a, 2019b; Livingston et al. 2019; Barragán et al. 2019), and provided insight into the radius distribution of these planets compared to those about older stars (e.g., Rizzuto et al. 2017; David et al. 2021). The Transiting Exoplanet Survey Satellite (TESS) mission (Ricker et al. 2015) is finding young planets that are more suitable for in-depth characterization: DS Tuc A (Newton et al. 2019), HD 63433 (Mann et al. 2020), and AU Mic (Plavchan et al. 2020) are among the brightest planet-hosting stars known. These are the best targets for in-depth follow-up studies with the suite of new astronomical observatories coming online this decade.

In this paper, we present the discovery of a mini-Neptune-sized planet transiting the V = 8.3 star HIP 94235, a member of the ∼120 Myr old AB Doradus moving group, one of the youngest and closest stellar associations. HIP 94235 was identified as a planet host through a search for planets around active stars (Zhou et al. 2021). The star's youth is confirmed from its spectroscopic and photometric characteristics, such as its rapid rotation and significant photometric modulation, strength of lithium absorption, and X-ray emission. The kinematics and independent age estimation of HIP 94235 agree with that of members of the AB Doradus moving group. The shallow 600 ppm transits of HIP 94235 b were identified in the single sector of observations obtained by TESS during the first sector of its ongoing Extended Mission. Subsequent observations via the CHaracterising ExOPlanet Satellite (CHEOPS) space telescope made it possible to confirm the existence of the transits and improved our ability to predict future transit times for follow-up studies.

2.1. Candidate Identification with TESS

TESS (Ricker et al. 2015) performed photometric measurements of HIP 94235 in its Sector 27 Camera 2 observations between 2020 July 04 and 30. HIP 94235 was observed at two-minute cadence via target pixel stamp observations, and was also included as a target of multiple TESS Guest Investigator Programs (G03251, D. Huber; G03272, J. Burt). We make use of Simple Aperture Photometry light curves (Twicken et al. 2010; Morris et al. 2020) from the Science Processing Observation Center (SPOC; Jenkins et al. 2016), extracted from the two-minute target pixel files for subsequent analyses. The field around HIP 94235 is shown in Figure 1 as seen from ground-based surveys, TESS, and subsequent follow-up observations.

HIP 94235 was identified as a planet candidate by a dedicated search for planets around young stars (Zhou et al. 2021). We first identified HIP 94235 as a possible young star via its rotationally modulated light curve. Three transits were identified using a Box-fitting Least Squares search (Kovács et al. 2002) on the light curve after detrending using a high-order spline-fitting procedure (Vanderburg & Johnson 2014). HIP 94235 was also identified as a threshold crossing event in both the SPOC pipeline (Jenkins et al. 2016) and the MIT QuickLook Pipeline (Huang et al. 2020a). The SPOC threshold crossing event diagnostic tests indicate the transit is on target to within 53 ± 25. However, it did not survive the TESS object of interest (TOI) vetting process initially because of the large amplitude of the residuals in the detrended light curves, which is due to the intrinsic variability of the star. The candidate was promoted to become a TOI after the confirmation of the photometric transit signals using the CHEOPS observations (Section 2.2).

Zoom In Zoom Out Reset image size

To account for the instrumental noise due to spacecraft motion and stellar variability in the light curves, we perform a simultaneous detrending using the spacecraft quarternions and basis splines following a similar process to that described in Vanderburg et al. (2019). We iteratively fit for linear coefficients of the mean, standard deviation, and skew terms of the three quarternions, together with a spline matrix created by the lightkurve package (Lightkurve Collaboration et al. 2018). The detrended light curve is adopted for our subsequent analyses in Section 4.

The corrected TESS light curve from Sector 27 is shown in Figure 2. Figure 3 shows close-ups of individual transits of HIP 94235 b in the raw and detrended TESS light curves. Figure 4 shows the phase-folded TESS transit light curve and its best-fit model from our global modeling (Section 4).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

2.2. Follow-up Photometry with CHEOPS

To confirm the existence and constrain the transit ephemeris of HIP 94235 b, we obtained space-based photometric observations using ESA's CHEOPS (Benz et al. 2021), through the CHEOPS Guest Observers Programme (AO2-005). CHEOPS is a 0.32 m Ritchey–Chrétien telescope in a nadir-locked 700 km low-Earth orbit, along the Earth's day/night terminator, with a period of 98.725 minutes. The telescope has a field of view of $17^{\prime} \times 17^{\prime}$ and is instrumentally defocused with a point-spread function of 16'' at a plate scale of 1'' pixel⁻¹.

A transit of HIP 94235 b was observed by CHEOPS between 2021 August 19 23:59 and 2021 August 20 07:33 UTC (visit ID 1568350). The visit consists of five orbits, with an exposure time of 17 s coadded on board to a cadence of 34 s. The visit had an observing efficiency of 61% (fraction of time on target), with interruptions primarily due to Earth stray light. A total of 498 exposures were obtained, of which, nine exposures were affected by stray light and Earth occultation, eight by South Atlantic Anomaly crossings. An example frame from CHEOPS is shown in Figure 1. The observations were reduced by the CHEOPS Data Reduction Pipeline v13.1.0 (Hoyer et al. 2020), accounting for bias, dark current, flat-fielding, bad pixel correction, smear contamination, and linearization. We adopt the optimal aperture light curve, with a circular aperture of 31 pixels, for our analysis. Aperture contamination by background stars is estimated to be at the 6 × 10⁻⁴ level and is accounted for via the simulations from the data reduction pipeline.

The CHEOPS light curve exhibits a smooth hour-long trend that can be attributed to the spot modulated rotational variations of HIP 94235. Shorter timescale instrumental variations associated with the spacecraft roll angle are seen on the orbital timescales (Maxted et al. 2021). We model the spot modulation over the five orbits of observations with a fourth-order polynomial after removal of the transit model. We also fit for a fifth-order correlation between the spacecraft roll angle and the resulting light-curve residuals. This modeling is performed simultaneous to the global fit of the TESS light curve and associated parameters, such that the uncertainties from this detrending process are fully accounted for in our results (Section 4). The raw and corrected CHEOPS light curves are shown in Figure 5, and the binned and corrected phase-folded data are compared with those from TESS in Figure 4.

Zoom In Zoom Out Reset image size

The CHEOPS observation was most crucial in refining the ephemeris of HIP 94235 b. With a transit depth of 600 ppm, detecting the photometric transit event with ground-based facilities is difficult. With only three transits available during the single sector of TESS observations over the entire primary and first extended mission, the TESS ephemeris would have quickly become stale. The ephemeris uncertainty using TESS observations alone would have been 5.2 hr after five years, making any transit follow-up studies more difficult to schedule. The single CHEOPS transit reduced the five-year transit timing uncertainty to 8 minutes. Figure 6 illustrates the reduction in transit ephemeris uncertainty enabled by the CHEOPS observation.

Zoom In Zoom Out Reset image size

2.3. Clearing of Nearby Eclipsing Binaries from Las Cumbres Observatory

2.4. Reconnaissance Spectroscopy

HIP 94235 shares space velocities with the AB Doradus moving group (Section 3.1.1). The age of the group has been estimated to be between 50 and 150 Myr. The group has also been linked to be coevolving with the Pleiades. We adopt an AB Doradus age of 120 Myr for the remainder of this discussion. In addition, a neighborhood search for tangentially comoving stars with HIP 94235 reveals a tentative population of stars, with an age of ∼120 Myr, determined by their rotation periods. The following sections examine the kinematic properties of HIP 94235 and the association, as well as the photometric and spectroscopic characteristics of HIP 94235 that identify its youth independent of the kinematics.

As accretion dwindles, young stars spin up as they conserve their angular momentum while they contract in radius. The spin up leads to increased magnetic activity, large photospheric spots, and increased chromospheric activity. Both rotation and activity then decay with time as angular momentum is lost through the stellar wind over the main-sequence lifetime of the star. The TESS light curve of HIP 94235 exhibits significant rotational spot modulation, which first drew our attention to its potential youth. The presence of X-ray emission and lithium absorption for the Sun-like host star confirm its youth. A flare event was also observed during the second orbit of the TESS sector, consistent with the behavior expected for a young star.

3.1. Kinematics

3.1.1. The AB Doradus Moving Group

Accurate estimation of the age of a single star is notoriously difficult (Soderblom 2010). Stars in clusters and associations have accurate age estimates as the population can be assessed as a whole; the stars can be seen to be coevolving based on their color–magnitude, rotation, and lithium abundance distributions. Planets found in coevolving populations offer much more stringent tests on the temporal evolution of planet properties.

Zuckerman et al. (2004) identified a set of stars within ∼50 pc comoving with AB Doradus. The star AB Doradus itself is among the closest (15 pc) and most-well-studied young stars. Membership of the group has been revised based on chemo-kinematic analyses of the homogeneity of the stars (da Silva et al. 2009) and updated kinematics from new missions (e.g., Malo et al. 2013; Gagné et al. 2018). Today, dozens of bona fide members define the extent and characteristics of the group, and its age has been estimated to range from 50 Myr (Zuckerman et al. 2004), ∼100–120 Myr (Luhman et al. 2005), >110 Myr (Barenfeld et al. 2013), to ∼150 Myr (Bell et al.2015).

Zoom In Zoom Out Reset image size

Figure 8 shows the color–magnitude and space motion of HIP 94235 alongside the AB Doradus moving group and the Pleiades cluster. HIP 94235 shares kinematic properties with the moving group, and has been classified as a bona fide member via Hipparcos (Malo et al. 2013) and Gaia (Gagné et al. 2018; Ujjwal et al. 2020) space motions.

Zoom In Zoom Out Reset image size

The group has long been linked to Pleiades cluster due to their shared kinematic velocities (Luhman et al. 2005; Ortega et al. 2007). Recent mapping of new low-density moving groups from Gaia (Kounkel & Covey 2019) suggest that the AB Doradus moving group, alongside newly identified Theia 301 and Theia 369 associations, form a long tidal tail streaming away from the Pleiades (Gagné et al. 2021).

Though more distant (130 pc), the Pleiades cluster contains ∼1000 members, and its age has been more thoroughly investigated than AB Doradus. Estimates converge to ∼120 Myr via nonrotating isochrones (e.g., Meynet et al. 1993), lithium depletion boundary (e.g., Stauffer et al. 1998; Barrado y Navascués et al. 2004), and 3D rotational isochrones (e.g., Brandt & Huang 2015). The spectroscopic and photometric characteristics of HIP 94235 agree with members of the Pleiades, reaffirming our adopted age of ∼120 Myr for the system.

3.1.2. An Independent Rotation Sequence for Comoving and Coevolving Stars

The sparsity and spread of the AB Doradus moving group makes it difficult to securely identify true members of the group. The canonical membership list has been evolving with each new chemo-kinematic data set as it was originally defined in Zuckerman et al. (2004). Upcoming Gaia releases with refined radial velocities for fainter distributions of cool stars may help redefine groups like AB Doradus and their links with other associations.

To ensure that our age estimation does not hinge on literature classifications of HIP 94235, we also independently search for a comoving population that may confirm its youth.

We follow Tofflemire et al. (2021) and search for stars within 50 pc of HIP 94235 that share its tangential velocity to within 5 km s⁻¹ using the comove package. ³¹ In this search, we assume the radial velocity of HIP 94235 for all neighborhood stars. The search returns a set of ∼2000 stars between 3 < G < 20 in magnitude. We then query for TESS MIT QLP light curves, returning 300 matches up to T_mag = 13.5. We apply a Lomb–Scargle period search for each light curve, with an upper limit on the rotation period of P_rot < 13. After manual examination, we find 140 stars that exhibit rotational modulation in their light curves with secure periods.

Figure 9 shows the rotation distribution of these 140 stars relative to the Pleiades. The vast majority of these stars are relatively bright and have well-measured radial velocities from Gaia. We find 21 stars with space motion velocities within 5 km s⁻¹ of HIP 94235; these are colored in blue for clarity. This population forms a rotation sequence that agrees with HIP 94235. This set of coevolving stars are presented in Table 2.

Zoom In Zoom Out Reset image size

Table 2. Candidate Comoving and Coevolving Stars with HIP 94235

TIC	R.A. (deg)	Decl. (deg)	U (km s−1)	V (km s−1)	W (km s−1)	B (mag)	V (mag)	T (mag)	K (mag)	Prot (d)	Li EW (Å)
71314712	330.63261	−47.67748	−4.0	−26.0	−4.0	9.3	8.8	8.2	7.4	2.4
91231096	310.66707	−46.67197	−1.9	−26.3	−1.9	11.0	10.0	9.1	7.6	4.6
93839949	176.15935	−49.41756	−7.3	−24.7	−7.3	9.9	8.9	8.0	6.5	1.9
101403239	300.79541	−52.96796	−1.7	−26.0	−1.7	12.8	11.9	10.9	9.4	12.7
108272865	285.27521	−28.71442	−5.2	−27.4	−5.2	9.0	8.5	8.0	7.2	2.4
142144969	99.73088	−74.42526	−5.2	−26.1	−5.2	10.5	9.7	9.0	7.9	7.0
197597944	327.20235	−39.48626	−8.5	−28.1	−8.5	10.3	9.7	9.0	8.0	4.1
206603521	21.80129	−57.29366	−7.9	−26.8	−7.9	14.5	13.0	11.1	8.8	3.8
234299476	358.66836	−60.85971	−8.3	−27.2	−8.3	11.1	10.0	9.0	7.5	9.0
270200832	331.81221	−74.08641	−3.0	−26.4	−3.0	9.5	9.0	8.5	7.8	2.0
270259954	359.04562	−39.05314	−7.5	−27.7	−7.5	9.2	8.2	7.3	5.9	7.8
270377865	345.08089	−26.15444	−3.1	−26.3	−3.1	8.1	7.5	6.9	5.9	3.6
278271178	112.29658	−82.20387	−4.4	−24.9	−4.4	14.3	13.1	11.1	8.9	10.9
280683734	342.73660	−79.16594	−7.5	−27.0	−7.5	9.6	9.0	8.4	7.6	2.6	0.136
290081380	319.52174	−29.52160	−7.0	−29.4	−7.0	15.1	13.9	11.5	9.3	10.3
341498715	318.52194	−63.70066	−6.9	−27.1	−6.9	10.9	10.1	9.4	8.2	5.9
357709300	189.06702	−79.52628	−7.3	−26.8	−7.3	12.2	11.0	10.2	8.7	7.0
369897885	171.32165	−84.95449	−8.4	−25.4	−9.6	8.1	7.6	7.2	6.5	1.5	0.030
389660501	321.33728	−43.51207	−3.5	−25.8	−3.5	13.4	12.1	11.0	9.1	7.1
405077613	112.74683	−84.32411	−8.6	−27.2	−8.6	10.8	10.0	9.2	7.9	4.9	0.200
409141582	312.08079	−28.02455	−5.7	−27.3	−5.7	13.7	13.6	11.4	9.1	2.7
409141582	312.08194	−28.02453	−9.2	−25.0	−9.2	13.7	13.6	11.4	9.1	2.7
439417806	301.15247	−35.21447	−3.7	−26.2	−3.7	9.4	8.9	12.2	7.7	1.4
464646604	287.74114	−60.27264	−5.7	−27.0	−5.7	8.9	8.3	7.8	6.9	2.2	0.141
466277708	305.70238	−65.25823	−6.9	−27.0	−6.9	10.3	9.6	9.0	8.1	3.8

Download table as:  ASCIITypeset image

To estimate the age of the distribution, we fit its color and rotation period with a rotation-age relationship. We select eight F, G, K stars (0.1 < B − V < 0.8) within our sample that lie on the slow sequence of the rotation sequence, and model their rotation periods with the age–color–rotation relationship from Mamajek & Hillenbrand (2008). The posterior age distribution for these eight stars, and their joint posterior, is shown in Figure 10. We find that the color–rotation distribution of this subsample can be described by the slow sequence with an age of ${118}_{-16}^{+18}$ Myr.

Zoom In Zoom Out Reset image size

We obtained a small number of spectra of this subsample using the Las Cumbres Observatory (Brown et al. 2013) Robotic Echelle Spectrographs (NRES) facilities. The stars that were observed exhibit lithium absorption as expected for their youth. The lithium 6708 Å equivalent width has been noted in Table 2 where available. Future works examining the lithium absorption of this subsample may yield a lithium depletion boundary to confirm this age estimate.

Of these 21 comoving stars, only one is a canonical member of AB Doradus (Gagné et al. 2018). It is likely that the AB Doradus moving group itself is dispersed and ill defined, and our subset of comoving and potentially coevolving stars form part of the extended AB Doradus group.

3.2. X-Rays

3.3. Lithium

Stars undergo lithium depletion during their main-sequence evolution. Lithium is depleted during proton collisions in the cores of stars. Convective mixing between the envelope and the core leads to a gradual depletion of lithium absorption in the observed spectra of Sun-like stars. As such, the lithium abundance of a star can be a tracer for its youth, though direct age estimates from the lithium absorption strength is only qualitative, as is the case for any age indicator when interpreting single stars. Our CHIRON spectra of HIP 94235 reveal significant absorption about the 6708 Å Li doublet. We model the doublet and the nearby Fe I line simultaneously, measuring a Li equivalent width of 0.1413 ± 0.0092 Åfor HIP 94235.

Figure 11 compares the lithium 6708 Å equivalent width against distributions from known AB Doradus members (da Silva et al. 2009), Pleiades, and Praesepe clusters (Quinn et al. 2012, 2014; Zhou et al. 2021). HIP 94235 is consistent with that of other known members of AB Doradus and Pleiades, and exhibits convincingly stronger lithium absorption than members of the 600 Myr Praesepe cluster, in agreement with a 120 Myr age estimate.

Zoom In Zoom Out Reset image size

3.4. Rotation

The TESS observations show that HIP 94235 exhibits significant photometric variability at the 2% level, consistent with the semi-periodic signature of spot modulation. Figure 12 shows the periodicity of HIP 94235 via a Lomb–Scargle periodogram, with a peak rotation period of 2.24 ± 0.11 days. We adopt the width of the rotational peak of the periodogram as the uncertainty on the measured period. The rotational modulation is clear in phase-folded light curve in the right panel of Figure 12, and the evolution of starspots over successive rotations can also be seen. Additionally, we modeled the light curve via a stochastically driven damped harmonic oscillator through a Gaussian process using the celerite package (Foreman-Mackey et al. 2017). Modeling the posterior via a Markov chain through emcee (Foreman-Mackey et al. 2013), we found a posterior distribution for the frequency term $\mathrm{log}{\omega }_{0}$ to be $-{0.791}_{-0.045}^{+0.047}$, corresponding to a period of 2.20 ± 0.10 days.

Zoom In Zoom Out Reset image size

Figure 11 shows the rotation periods of AB Doradus, as well as the Pleiades and Praesepe cluster members. We adopt the membership list for AB Doradus from Gagné et al. (2018), and derived rotation periods for stars with available TESS light curves from the MIT FFI QLP library (Huang et al. 2020b) for stars with available light curves. Rotation periods are measured via a Lomb–Scargle period search and down-selected by hand to remove stars that do not show unambiguous rotational signatures. By-hand corrections of period aliases were also applied. Of the 66 members listed in Gagné et al. (2018), 55 were bright enough and had accessible TESS QLP light curves, and 37 yielded convincing detections of a rotation signal. Rotation periods for Pleiades members were adopted from Rebull et al. (2016), and for Praesepe from Rebull et al. (2017). The rotation of HIP 94235 agrees with the Pleiades and AB Doradus distribution, in agreement with its kinematic age estimates.

To best estimate the stellar and planetary properties of the HIP 94235 system, we perform a global modeling incorporating the TESS and CHEOPS transits, radial velocities, and photometric and spectroscopic properties of HIP 94235.

The transit models were computed as per Mandel & Agol (2002), implemented via batman (Kreidberg 2015). The free parameters for this model are the stellar mass and radius, the time of transit center T₀, the orbital period P, the planetary radius ratio R_p /R_⋆, and the eccentricity parameters $\sqrt{e}\cos \omega$ and $\sqrt{e}\sin \omega$, where e is the eccentricity and ω is the longitude of periastron.

We model the CHEOPS transit and its associated stellar variability and instrumental characteristics simultaneous to the global modeling. The transit model is computed as per Mandel & Agol (2002). The hour-long rotational modulation signal is modeled via a fourth-degree polynomial with respect to time. The correlation between the light curve and the spacecraft motion is modeled via a fifth-degree polynomial against the roll angle. Figure 5 shows the CHEOPS light curve before and after the removal of the best-fit stellar variability and instrumental model.

The out-of-transit Keplerian radial velocity was modeled with additional parameters describing the systemic velocity γ, planetary mass M_p , and a jitter term for each instrument.

The stellar mass and radius were modeled using the Modules for Experiments in Stellar Astrophysics Isochrones and Stellar Tracks (MIST; Dotter 2016), and constrained by their photometric magnitudes and parallax priors from Gaia G, Bp, Rp (Gaia Collaboration et al. 2018), the Hipparcos TYCHO B and V bands (Perryman et al. 1997), the Two Micron All Sky Survey (2MASS) J, H, and Ks bands (Skrutskie et al. 2006). Additionally, the age was restricted to be 120 ± 50 Myr as per the age of the AB Dor moving group, and the limb darkening coefficients were fixed to theoretically interpolated values (Claret & Bloemen 2011; Claret 2017). All other parameters were assigned uniform priors with physically motivated boundaries.

The modeling was performed simultaneously for all parameters using Monte Carlo Markov Chain (MCMC) analysis, making use of the emcee package (Foreman-Mackey et al. 2013). Results are listed in Tables 3 and 4, the best-fit light-curve model is shown in Figure 2, and the best-fit radial velocity model is shown in Figure 7. The spectral energy distribution of HIP 94235 is shown in Figure 13, along with the template ATLAS9 model spectrum (Castelli & Kurucz 2004) computed at the best-fit stellar parameters of HIP 94235. We note that the mid-infrared Wide-field Infrared Survey Explorer (WISE) magnitudes show no excess that might be indicative of a remnant debris disk around the young star.

Zoom In Zoom Out Reset image size

Table 3. Properties of HIP 94235

Parameter	Value	Source
Astrometry
R.A. ...	19:10:57.87	Gaia Collaboration et al. (2021)
decl. ...	−60:16:21.49	Gaia Collaboration et al. (2021)
Parallax (mas) ...	17.061 ± 0.037	Gaia Collaboration et al. (2021)
Proper Motion
Gaia (2016.1) R.A. Proper Motion (mas yr−1) ...	11.632 ± 0.025	Gaia Collaboration et al. (2021)
Gaia (2016.1) decl. Proper Motion (mas yr−1) ...	−100.836 ± 0.025	Gaia Collaboration et al. (2021)
Hipparcos (1991.1) R.A. Proper Motion (mas yr−1) ...	12.755 ± 0.890	199 (1997)
Hipparcos (1991.4) decl. Proper Motion (mas yr−1) ...	−99.617 ± 0.796	199 (1997)
Hipparcos–Gaia Average R.A. Proper Motion (mas yr−1) ...	11.509 ± 0.025	Brandt (2021)
Hipparcos–Gaia decl. Average Proper Motion (mas yr−1) ...	−100.628 ± 0.019	Brandt (2021)
Photometry
TESS (mag) ...	7.758 ± 0.006	Stassun et al. (2018)
B (mag) ...	8.943 ± 0.027	Henden et al. (2016)
V (mag) ...	8.31 ± 0.03	Henden et al. (2016)
J (mag) ...	7.201 ± 0.023	Skrutskie et al. (2006)
H (mag) ...	6.966 ± 0.023	Skrutskie et al. (2006)
K (mag) ...	6.881 ± 0.027	Skrutskie et al. (2006)
Gaia (mag) ...	8.173207 ± 0.00031	Gaia Collaboration et al. (2021)
GaiaBP (mag) ...	8.4634 ± 0.0013	Gaia Collaboration et al. (2021)
GaiaRP (mag) ...	7.70637 ± 0.00087	Gaia Collaboration et al. (2021)
WISE W1 (mag) ...	6.841 ± 0.066	Cutri et al. (2012)
WISE W2 (mag) ...	6.823 ± 0.02	Cutri et al. (2012)
WISE W3 (mag) ...	6.836 ± 0.016	Cutri et al. (2012)
WISE W4 (mag) ...	6.728 ± 0.067	Cutri et al. (2012)
Kinematics and Position
U (km s−1) ...	−5.61 ± 0.34	Derived
V (km s−1) ...	−27.03 ± 0.17	Derived
W (km s−1) ...	−11.82 ± 0.18	Derived
Distance (pc) ...	${58.54}_{-0.07}^{+0.08}$	Derived
Physical Properties
M⋆ (M⊙) ...	${1.094}_{-0.007}^{+0.024}$	Fitted (Uniform Prior)
R⋆ (R⊙) ...	${1.08}_{-0.10}^{+0.11}$	Fitted (Uniform Prior)
Effective Temperature Teff (K) ...	5991 ± 50	This paper
Surface Gravity $\mathrm{log}{g}_{\star }$ (cgs) ...	4.460 ± 0.05	This paper
[m/H] ...	0.0	(Barenfeld et al. 2013)
$v\sin {I}_{\star }$ (km s−1) ...	24.4 ± 1.0	This paper
I⋆(°)...	>70 (3σ)	Calculated as per Masuda & Winn (2020)
Age (Myr) ...	50–150	Zuckerman et al. (2004)
		Luhman et al. (2005)
		Bell et al. (2015)
Limb darkening coefficients (TESS) ...	(0.17, 0.41)	Claret (2017)
Limb darkening coefficients (CHEOPS) ...	(0.27, 0.29)	Claret (2017)
Activity indicators Prot (days) ...	2.24 ± 0.11	This paper
Lithium 6708 Å Equivalent Width (Å) ...	0.1413 ± 0.0092	This paper
X-Ray luminosity $\mathrm{log}({L}_{X}/{L}_{\mathrm{bol}})$	−3.93 ± 0.13	(Boller et al. 2016)

Download table as:  ASCIITypeset image

Table 4. Derived Parameters for HIP 94235 b

Parameter	Joint model	Priors
Fitted Parameters
T0 (BJD) ...	${2459037.8704}_{-0.0022}^{+0.0011}$	Uniform
P (days) ...	${7.713057}_{-0.000021}^{+0.000021}$	Uniform
Rp /R⋆ ...	${0.0253}_{-0.00059}^{+0.00075}$	Uniform
i (deg) ...	${87.14}_{-0.17}^{+0.16}$	Uniform
$\sqrt{e}\cos \omega$ ...	${0.07}_{-0.54}^{+0.50}$	Uniform
$\sqrt{e}\sin \omega$ ...	${0.26}_{-0.27}^{+0.20}$	Uniform
Mp (M⊕) ...	<379 (3σ)	Uniform
γMINERVA−Australis 3 ...	${9013}_{-70}^{+58}$	Uniform
γMINERVA−Australis 5 ...	${8960}_{-126}^{+123}$	Uniform
γMINERVA−Australis 6 ...	${9050}_{-81}^{+88}$	Uniform
γCHIRON ...	${8264}_{-30}^{+31}$	Uniform
RV Jitter MINERVA-Australis 3 ...	${80}_{-56}^{+73}$	Uniform
RV Jitter MINERVA-Australis 5 ...	${154}_{-98}^{+210}$	Uniform
RV Jitter MINERVA-Australis 6 ...	${140}_{-79}^{+113}$	Uniform
RV Jitter CHIRON ...	${100}_{-28}^{+35}$	Uniform
Inferred parameters
e ...	${0.32}_{-0.20}^{+0.20}$	Derived
ω (deg) ...	${17}_{-92}^{+67}$	Derived
Rp (R⊕) ...	${3.00}_{-0.28}^{+0.32}$	Derived
a/R⋆...	${15.7}_{-1.5}^{+1.6}$	Derived
a (AU) ...	${0.07870}_{-0.00017}^{+0.00056}$	Derived
Transit duration (days) ...	${0.103}_{-0.018}^{+0.009}$	Derived
Krv (m s−1) ...	<150 (3σ)	Derived
Teq (K) ...	1060 ± 50	Derived

Download table as:  ASCIITypeset image

Diffraction limited observations revealed a bound M-dwarf companion to HIP 94235. The companion was identified in speckle imaging observations of HIP 94235 during the candidate vetting process, and also identified from archival adaptive optics observations searching for wide Jovian companions to young stars 11 yr prior. We describe the observations below and perform compute astrometric orbital solutions for HIP 94235 B.

We obtained high-contrast imaging at 562 nm and 832 nm with the Zorro speckle imager on the 8 m Gemini South Observatory on 2021 July 23 and 2021 October 22, which revealed a faint stellar companion to HIP 94235. Zorro is a dual-channel speckle imager with a pixel scale of 001pixel⁻¹ and an approximate FWHM of 002. Data reduction and analysis were performed as per Howell et al. (2011) and Howell et al. (2016). Figure 14 shows the Zorro image at 562 nm and 832 nm of HIP 94235 on 2021 October 22. The 832 nm observation achieved a contrast ratio of Δm = 6.68 at 05 separation. A companion of Δm = 5.3 was identified at a separation of 06 in the red band. No companions with contrasts brighter than Δm = 4.7 were detected in the blue arm. Observations from 2021 July 23 also identified the same companion in the 832 nm observations, though the blue arm was not functional at the time.

Zoom In Zoom Out Reset image size

Adaptive optics imaging of HIP 94235 was also carried out as part of a large program to characterize the occurrence rates of giant planets at large separations by Chauvin et al. (2015). These observations, using the NaCo high-contrast adaptive optics (AO) imager on VLT-UT4, were obtained on 2010-07-30. They revealed the stellar companion at a separation of 05 with a contrast of ΔH = 3.8 ± 0.3. The positional information from the diffraction limited imaging observations are listed in Table 5.

Based on the Gaia and Hipparcos proper motions measurements, the accumulated motion of HIP 94235 over the 11 yr interval is 11. In contrast, the relative motion between the imaged companion and the target star is 01, which strongly suggests the pair is a a bound stellar binary with a projected separation of 31 au. We henceforth refer to this companion as HIP 94235 B.

To determine the approximate properties of HIP 94235 B, we adopt the 100 Myr MIST isochrones (Choi et al. 2016) to model its 832 nm and H-band magnitudes. We approximate the 832 nm band with the I-band magnitudes from MIST, and adopt magnitudes for the companion of I = 10.76 ± 0.3 and H = 13.05 ± 0.4. We adopt a fixed metallicity of [M/H] = 0 as per that for the AB Doradus moving group (da Silva et al. 2009) and adopt a strong Gaussian prior for the distance to HIP 94235 B as per its Gaia parallax. We find the companion is an M-dwarf of mass 0.26 ± 0.04 M_⊙ and radius 0.31 ± 0.03 R_⊙. The companion is incapable of hosting the transiting companion responsible for the planetary transit signal detected in TESS. To check if this companion is capable of being the source of the transit signal, we deblend and refit for the transit about the M-dwarf companion. We find a best-fit radius ratio of R₂/R₁ = 0.46 ± 0.01 for such a system, with a V-shaped transit that is incompatible with the observed light curve. A model comparison yields a Bayesian information criterion difference of ΔBIC = 65.9 between the best-fit model of a transit about the companion M-dwarf and that of our nominal planetary transit scenario. We therefore rule out the companion being the source of the observed transit signal.

Although the available imaging detections cover only a small part of the orbital arc of HIP 94235 B, it is possible in principle that the binary orbit could be constrained. Several authors (e.g., Brandt 2018; Calissendorff & Janson 2018; Snellen & Brown 2018; Kervella et al. 2019) have demonstrated that it is possible to combine proper motion measurements from Gaia and Hipparcos to produce long-timescale astrometric data that can be used to detect the reflex orbital motion of orbiting companions. Brandt et al. (2019) extended this technique further by jointly fitting Hipparcos–Gaia astrometry with radial velocities and relative astrometry from direct imaging, and found that the orbits of massive companions can be precisely constrained even when the orbital periods are much longer than the observational duration. This method has been profitably applied to many further systems; Bowler et al. (2021), for example, were able to extract precise orbital parameters and masses for two white dwarfs with orbital periods in excess of ≳ 200 yr despite possessing no more than ∼30 yr of observational data in both cases.

Inspection the Hipparcos–Gaia Catalog of Accelerations (Brandt 2021) shows that HIP 94235 displays a statistically significant astrometric acceleration of 0.24 mas yr⁻¹ (χ² = 58.5) between the Gaia and Hipparcos–Gaia proper motions, equivalent to a drift in the stellar tangential velocity of ≈70 m s⁻¹, which can plausibly be attributed to HIP 94235 B. This motivates us to attempt an orbital fit for HIP 94235 B based on the available data.

To initialize our model we first used orbitize! (Blunt et al. 2020) to fit the relative astrometry of HIP 94235 B assuming a total system mass of 1.09 + 0.26 = 1.35 M_⊙. orbitize! makes use of the orbits for the impatient (OFTI) Bayesian sampling method (Blunt et al. 2017) that is well-suited to fitting the orbits of directly detected companions with orbital periods much longer than the observational span such as HIP 94235 B. Next we extracted the posteriors from the orbitize! fit and calculated the corresponding χ² value for a fit to the Hipparcos–Gaia astrometry for each set of orbital parameters assuming stellar masses of M_A = 1.09 M_⊙, M_B = 0.26 M_⊙. This allowed us to identify a tightly constrained initial parameter space for most orbital parameters.

To model the orbit of HIP 94235 B, we run a joint fit to the astrometry based on that of Brandt et al. (2019). For modeling the Hipparcos–Gaia astrometry we use the equations described in Venner et al. (2021b), while the corresponding expressions for the relative astrometry can be found in Pourbaix (1998). As in Venner et al. (2021a, 2021b), we resample the Hipparcos and Gaia proper motions using the observational epochs recorded in the Hipparcos Intermediate Astrometric Data for the former and the Gaia Observation Forecast Tool ³² for the latter. To explore the model parameter space we use the differential evolution MCMC sampler edmcmc ³³ (Vanderburg 2021).

A total of 11 parameters are used for the model: the system parallax ϖ, the primary mass M_A , the secondary mass M_B , the semimajor axis a, the eccentricity e, and argument of periastron ω parameterized as $\sqrt{e}\sin \omega$ and $\sqrt{e}\cos \omega$, the mean anomaly at an arbitrary reference epoch BJD = 2457000, the orbital inclination i, the longitude of node Ω, and finally two terms for the proper motion of the system barycenter.

The radial velocity trend generated by HIP 94235 B is too small to be detected in the available data so we do not make use of radial velocity (RV) data in the joint fit. The lack of radial velocity information in the fit results in the classical 180° degeneracy in the longitude of node Ω and argument of periastron ω; following convention we report the solution with Ω in the range [0, 180] deg. The argument of periastron used in our model is that of the primary's orbit rather than that of the companion.

Of the 11 parameters used in the joint model, the parallax ϖ was assigned a Gaussian prior of 17.061 ± 0.037 mas based on the Gaia EDR3 astrometric solution while the primary mass M_A was given a Gaussian prior of 1.094 ± 0.05 M_⊙, with an inflated uncertainty as compared to the value in Table 3 to avoid unduly biasing the model. Initial trials of the orbital fit were run without informed priors on the secondary mass M_B however these runs tended to produce results skewed toward implausibly small values for this parameter (M_B < 0.1 M_⊙), which in turn resulted in excessively broad distributions in the orbital parameters. It was therefore deemed prudent to adopt an informed prior of M_B = 0.26 ± 0.04 M_⊙ for the final model.

The results of our joint model for the orbit of HIP 94235 B are presented in Table 6. Despite the span of observations being much shorter than the orbital period of the binary, we are able to robustly constrain most orbital parameters. We measure a semimajor axis of $a={56}_{-7}^{+9}$ au for HIP 94235 B, approximately ≈80% larger than the projected separation, corresponding to an orbital period of ${365}_{-69}^{+92}$ yr. Our model shows a preference for relatively low orbital eccentricities (${0.25}_{-0.14}^{+0.22}$, e < 0.61 at 95% confidence), resulting in a periastron separation of ${43}_{-15}^{+11}$ au for the binary. We obtain a tightly constrained orbital inclination of $67\buildrel{\circ}\over{.} {8}_{-2.9}^{+2.7}$ and a precise longitude of node of ${20}_{-7}^{+11}$ degrees, while our measurements of ω and the time of periastron for HIP 94235 B are comparatively imprecise (${300}_{-80}^{+30}$ deg, ${2184}_{-74}^{+107}$ CE).

Table 6. Orbital Parameters of HIP 94235 B

Parameter	Median ± 1σ	Mode	Priors
Informed Priors
Parallax (mas) ...	17.061 ± 0.037	−	Gaussian
MA (M⊙) ...	1.094 ± 0.05	−	Gaussian
MB (M⊙) ...	0.26 ± 0.04	−	Gaussian
Fitted Parameters
a (AU) ...	${56}_{-7}^{+9}$	54	Log-uniform
$\sqrt{e}\sin \omega$ ...	${0.35}_{-0.19}^{+0.14}$	0.40	Uniform
$\sqrt{e}\cos \omega$ ...	${0.20}_{-0.34}^{+0.48}$	0.48	Uniform
Mean anomaly at BJD = 2457000 (deg) ...	${170}_{-80}^{+110}$	130	Uniform
i (deg) ...	${67.8}_{-2.9}^{+2.7}$	67.4	$\sin i$
Ω (deg) ...	${20}_{-7}^{+11}$	18	Uniform
Barycentric R.A. proper motion (mas yr−1) ...	10.50 ± 0.24	10.46	Uniform
Barycentric decl. proper motion (mas yr−1) ...	−102.99 ± 0.35	102.91	Uniform
Derived Parameters
P (years) ...	${365}_{-69}^{+92}$	362	Derived
e ...	${0.25}_{-0.14}^{+0.22}$	0.19	Derived
ω (deg) ...	${300}_{-80}^{+30}$	320	Derived
Periastron distance (AU) ...	${43}_{-15}^{+11}$	47	Derived
Time of periastron (years CE) ...	${2184}_{-74}^{+107}$	2164	Derived

Download table as:  ASCIITypeset image

Our fit to the astrometric data is shown in Figures 15 and 16. In the fit to the absolute astrometry it can be observed that the Hipparcos proper motion measurement is too imprecise to significantly detect the astrometric signal. Conversely, the Gaia and averaged Hipparcos–Gaia proper motions are so precise that their uncertainties are not visible in the figure, and it is these measurements that drive the detection of the astrometric reflex signal of HIP 94235 B. The importance of the >10 yr time span of relative astrometry is evident from Figures 15 and 16, as were it not for the detection of HIP 94235 B in an archival NaCo observation in 2010 it would not be possible to derive such strong constraints on the binary orbit.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

We note that the astrometric orbit solution is dependent on the priors we adopt for the system. For example, when the mass prior for HIP 94235 B is removed, the orbital semimajor axis becomes degenerate with eccentricity. As such, future diffraction limited observations remain important in validating and refining the orbital parameters of the system.

Based on our results, we predict the magnitude of the radial velocity trend on HIP 94235 generated by HIP 94235 B to be 9.6 ± 1.3 m s⁻¹ yr⁻¹. This trend is not detectable in our RV data, but it is possible that future high-precision radial velocity measurements will be able to detect this acceleration. If so, the radial velocity information could then be used to improve the constraints on the orbit of HIP 94235 B.

We report the discovery and statistical validation of a mini-Neptune around a bright young Sun-like star. HIP 94235 b is a ${3.00}_{-0.28}^{+0.32}\,{R}_{\oplus }$ planet in a 7.7 day period orbit around its V_mag = 8.31 host star. Based on its kinematics, HIP 94235 can be placed in the AB Doradus moving group, with an age of ∼120 Myr. The kinematics age is in agreement with that expected from the stellar rotation rate, lithium abundance, and X-ray emission intensity.

The 600 ppm transits of HIP 94235 b were identified at a signal-to-noise of 9.7 from TESS Extended Mission Section 27 observations. However, the two-day periodicity 2% stellar variability led to the planetary signal slipping through the TESS official planet selection processes. The detection of HIP 94235 b demonstrated the necessity of a dedicated search for planets around noisy, active young stars. Confirmation of transits came from the CHEOPS mission. We obtained five orbits of observations with CHEOPS to recover the transit of HIP 94235 b. Such follow-up would have been difficult to schedule with ground-based facilities due to the shallow transit.

Figure 6 demonstrates the importance of CHEOPS follow-up in preserving the transit timing ephemeris. Without such observations, the timing derived from the three transits observed in the single-sector TESS observation would have eroded by ∼1 hr per year. If a follow-up confirmation was not obtained within the first year after discovery, targeted transit observations would have been difficult to schedule and the recovery of such a small planet challenging (see Dragomir et al.2020).

HIP 94235 b is one of the smallest planets that has been found transiting a young star. Figure 17 shows the position of HIP 94235 b in the distribution of small planets around young stars. Most TESS planets that transit young stars, including HIP 94235 b, have radii that place them above the radius valley delineated by Fulton et al. (2017) and Owen & Wu (2017). HIP 94235 b is at the edge of detectability for the TESS light curve of HIP 94235, and selection biases likely shape the current TESS distribution of young planets. However, Rizzuto et al. (2017) found young clusters and associations surveyed by K2 hosted planets larger than the equivalent distribution about field stars. A similar holistic study of young stars surveyed by TESS may elucidate the radius evolution timescale for young planets.

Zoom In Zoom Out Reset image size

Interestingly, despite continuous monitoring from the K2 mission of ∼1000 Pleiades members, no confirmed planets have yet been found in the 125 Myr old cluster (e.g., Hartman et al. 2010; Rizzuto et al. 2017). Systems such as TOI-451 (Newton et al. 2021) in the Pisces–Eridanus stream (Meingast et al. 2019; Curtis et al. 2019), and HIP 94235 in AB Doradus, may help constrain the occurrence rates and radius properties of planets at the ∼100 Myr age range. These can help infer if the absence of planets in the Pleiades is due to detection biases, or if other astrophysical mechanisms may be at play.

Planets like HIP 94235 b, lying near the edge of the sub-Neptune valley, can provide key observational tests for the mechanisms of mass loss in young planets. HIP 94235 b and other recently discovered planets around young stars are subjected to significant high-energy radiation, which can be a dominant driver for rapid mass loss within the first hundreds of millions of years post formation (Owen & Jackson 2012). To estimate the mass evolution of HIP 94235 b, we adopt the analytical approach from Owen & Wu (2017). We find that we can replicate the current radius of HIP 94235 b via a planet model that has a high initial envelope mass fraction, and undergoes rapid mass loss with a timescale of ∼100–200 Myr (Figure 18). HIP 94235 b has a current energy-limited mass-loss rate of ∼5 M_⊕ Gyr⁻¹. At the end of this process, we expect the envelope mass fraction to reduce from 10% to ∼1% of the total planet mass. Such mass and radius evolution is expected for many close-in Neptunes and super-Earths around Sun-like stars.

Zoom In Zoom Out Reset image size

Figure 19 shows the X-ray irradiation experienced by HIP 94235 b compared to other systems about X-ray sources identified in the Second ROSAT Point-source Catalog (Boller et al. 2016). To convert ROSAT count rates to fluxes, we adopt the calibration in Fleming et al. (1995) and parallaxes made available from Gaia Collaboration et al. (2018). The majority of planetary systems around ROSAT X-ray sources are young, and the recent TESS discoveries of nearby planet-hosting young stars are the most suitable targets for follow-up X-ray and UV observations. au Mic (Plavchan et al. 2020) has the highest ROSAT count rate for any planet-hosting star. Similarly, other nearby systems such as the planets around V1298 Tau (David et al. 2019a, 2019b), the ultrashort period super-Earth TOI-1807b (Hedges et al. 2021), DS Tuc Ab (Newton et al. 2019), and HIP 94235 b mark the inner boundary of X-ray irradiation for small planets. The only planets around stars not identified as young in literature residing in more energetic environments are the hot Jupiter NGTS-16b (Tilbrook et al. 2021) and the Earth-sized inner planet in the Kepler-1514 system (Dalba et al. 2021). Neither have radii susceptible to significant modification by photoevaporation. Recent observations by Zhang et al. (2022b) detected the Lyα transit of HD63433 c. Similar observations in the X-ray for young active stars have the potential of anchoring the photoevaporation models.

Zoom In Zoom Out Reset image size

Other avenues of mass loss are also possible. Atmospheric erosion from giant impact events can occur even more quickly, before compact super-Earth systems settle dynamically, acting on the tens of millions of years timescale post disk dispersal (e.g., Izidoro et al. 2017). Core-powered mass loss may also reproduce the observed radius distribution independent of the host-star irradiation (Lopez & Fortney 2013; Ginzburg et al. 2018). By inferring the ages of Kepler systems through numerous indicators, David et al. (2021) found the radius gap may form at longer timescales. As such, it is unclear which mechanism dominates in shaping our current planet distribution.

At 50 au, HIP 94235 is one of the tightest stellar binaries (Su et al. 2021). Like the DS Tuc AB system (Newton et al. 2019), the orbit of the binary is aligned with that of the planetary orbit. This follows the trend from Christian et al. (2022) that wide transiting planet-hosting binaries with separations between 100–700 au are preferentially found in edge-on orbits. Christian et al. (2022) suggest that such trends may be due to the companions being formed from disk fragmentation, or the realignment of the inner disk by the perturbing outer companion. The efforts by TESS follow-up teams to provide diffraction limited imaging of a majority of planet candidates, such as HIP 94235, will help probe the continuation of this trend to <100 au separations.

The brightness of HIP 94235 makes the system suitable for follow-up atmospheric characterization with the next generation of space and ground-based facilities. Adopting the mass–radius relationship from Wolfgang et al. (2016), HIP 94235 b has a predicted mass of 11.2 ± 1.4 M_⊕, yielding a transmission spectroscopic metric of 96 ± 25. As such, HIP 94235 b ranks among the top dozen known planets between 1.6–4 R_⊕ in its suitability for follow-up transmission observations. In the era of JWST, we can compare the atmosphere of HIP 94235 b against planets of similar radii about older stars. We may find that young planets host primarily primordial atmospheres dominated by hydrogen and helium, with older planets hosting heavier water-rich atmospheres, or that some highly energetic environments may never allow secondary atmospheres to form. Orbital obliquities of young planets can help constrain the timescales of migration for planets that may have formed further out in their planetary systems. We expect a ∼10 m s⁻¹ Rossiter–McLaughlin signal if HIP 94235 b is in a well-aligned projected orbit. The stellar activity will be a limiting factor in achieving a secure detection of the spectroscopic transit, though past works have shown that this can be mitigated on transit timescales due to the smoothly varying nature of the rotational modulated velocity noise (e.g., Palle et al. 2020; Benatti et al.2021).

We respectfully acknowledge the traditional custodians of all lands throughout Australia, and recognize their continued cultural and spiritual connection to the land, waterways, cosmos, and community. We pay our deepest respects to all Elders, ancestors and descendants of the Giabal, Jarowair, and Kambuwal nations, upon whose lands the MINERVA-Australis facility at Mount Kent is situated. G.Z. thanks the support of the ARC DECRA program DE210101893. C.W. and G.Z. thank the support of the TESS Guest Investigator Program G03007. C.H. thanks the support of the ARC DECRA program DE200101840. E.G. gratefully acknowledges support from the David and Claudia Harding Foundation in the form of a Winton Exoplanet Fellowship. CHEOPS is an ESA mission in partnership with Switzerland with important contributions to the payload and the ground segment from Austria, Belgium, France, Germany, Hungary, Italy, Portugal, Spain, Sweden, and the United Kingdom. We thank support from the CHEOPS GO Programme and Science Operations Centre for help in the preparation and analysis of the CHEOPS observations. This research has used data from the CTIO/SMARTS 1.5 m telescope, which is operated as part of the SMARTS Consortium by RECONS. This study was based in part on observations made using the Las Cumbres Observatory global telescope network, using time allocated by the National Science Foundation's NOIRLab (NOIRLab Prop. ID NOAO2021A-009; principal investigator: J. Hartman). Some of the observations reported in this paper were obtained with the Southern African Large Telescope (SALT). Some of the observations in the paper made use of the high-resolution imaging instrument Zorro obtained under Gemini LLP Proposal Number: GN/S-2021A-LP-105. Zorro was funded by the NASA Exoplanet Exploration Program and built at the NASA Ames Research Center by Steve B. Howell, Nic Scott, Elliott P. Horch, and Emmett Quigley. Zorro was mounted on the Gemini North (and/or South) telescope of the international Gemini Observatory, a program of NSF's OIR Lab, which is managed by the Association of Universities for Research in Astronomy (AURA) under a cooperative agreement with the National Science Foundation on behalf of the Gemini partnership: the National Science Foundation (United States), National Research Council (Canada), Agencia Nacional de Investigación y Desarrollo (Chile), Ministerio de Ciencia, Tecnología e Innovación (Argentina), Ministério da Ciência, Tecnologia, Inovações e Comunicações (Brazil), and Korea Astronomy and Space Science Institute (Republic of Korea). 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. Funding for the TESS mission is provided by NASA's Science Mission directorate. We acknowledge the use of public TESS Alert data from pipelines at the TESS Science Office and at the TESS Science Processing Operations Center. This research has made use of the Exoplanet Follow-up Observation Program website, which is operated by the California Institute of Technology, under contract with the National Aeronautics and Space Administration under the Exoplanet Exploration Program. This paper includes data collected by the TESS mission, which are publicly available from the Mikulski Archive for Space Telescopes (MAST). Resources supporting this work were provided by the NASA High-End Computing (HEC) program through the NASA Advanced Supercomputing (NAS) Division at Ames Research Center for the production of the SPOC data products. MINERVA-Australis is supported by Australian Research Council LIEF Grant LE160100001, Discovery Grants DP180100972 and DP220100365, Mount Cuba Astronomical Foundation, and institutional partners University of Southern Queensland, UNSW Sydney, MIT, Nanjing University, George Mason University, University of Louisville, University of California Riverside, University of Florida, and The University of Texas at Austin. D.D. acknowledges support from the TESS Guest Investigator Program grants 80NSSC21K0108 and 80NSSC22K0185.

Facility: TESS, CHEOPS, Exoplanet Archive, CTIO 1.5 m, LCOGT, Gemini:Zorro, SALT, MINERVA-Australis.

Software: emcee (Foreman-Mackey et al. 2013), batman (Kreidberg 2015), astropy (Astropy Collaboration et al. 2018), PyAstronomy (Czesla et al. 2019), comove (Tofflemire et al.2021).