🔭 astrophysics

Planet or brown dwarf? Constraints on the formation of H-type objects in IC348

By demonstrating that the spatial distribution of newly discovered H-type substellar objects in IC348 is indistinguishable from that of stars and brown dwarfs but inconsistent with the dispersed pattern expected from dynamically ejected planets, this study concludes that these objects likely formed via a star-like mechanism rather than a planetary one.

Original authors: Richard J. Parker (Sheffield, UK), Catarina Alves de Oliveira (ESA, Spain)

Published 2026-06-10

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Richard J. Parker (Sheffield, UK), Catarina Alves de Oliveira (ESA, Spain)

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a cosmic nursery called IC 348, a bustling cloud of gas and dust where new stars are being born. Recently, astronomers discovered nine tiny, dim objects floating in this nursery. They are too small to be full-blown stars, but too big to be typical planets. They are called "H-type" objects because their atmospheres have a unique chemical fingerprint (a hydrocarbon smell, if you will) that had never been seen in space before.

The big question is: What are these things, and how did they get there?

There are two main theories, like two different stories about how a child ends up in a new town:

The "Star" Story: They were born right there in the cloud, collapsing from gas just like stars do, but they just didn't grow up to be very big.
The "Planet" Story: They were born orbiting a star (like a planet), but a violent encounter with another star kicked them out of their home, leaving them to wander the nursery alone.

This paper acts like a cosmic detective, trying to figure out which story is true by looking at where these objects are standing in the nursery.

The Detective Work: Looking at the Crowd

The authors, Richard Parker and Catarina Alves de Oliveira, decided to look at the spatial distribution of these H-type objects. Think of this as checking the seating chart at a crowded party.

If they are "Star-like": They should be sitting right in the thick of the crowd, mixed in with the other stars and brown dwarfs (failed stars).
If they are "Planet-like" (ejected): They should be scattered around the edges of the room, far away from the main group, because they were kicked out and are slowly drifting away.

The Finding: The authors measured the distance between these H-type objects and their neighbors. The result? They are sitting right in the middle of the crowd. Their location is statistically identical to the stars and brown dwarfs. They aren't drifting away at the edges; they are part of the main group.

The Simulation: The "Cosmic Pinball" Experiment

To be absolutely sure, the team ran computer simulations. They built a virtual version of IC 348 and played a game of cosmic pinball.

The Setup: They created a virtual nursery with 495 stars and brown dwarfs.
The Planets: They added a "Jupiter-sized" planet to every star in the simulation.
The Action: They let the simulation run for millions of years, letting the stars bump into each other. Sometimes, a star would fly close enough to a planet to knock it out of orbit, turning it into a "free-floating planet."

They tested three scenarios for where the planets started:

Close to the star (1 AU): Like Earth.
Medium distance (5 AU): Like Jupiter.
Far away (30 AU): Like Neptune.

The Result:

When planets started far away, the simulation produced too many free-floating planets (20–40), and they were scattered all over the place.
When planets started close, they rarely got kicked out.
When planets started at 5 AU (Jupiter's distance), the simulation produced about 9 free-floating planets, which matches the number of H-type objects found in real life.

However, there was a catch. Even though the number matched, the location did not. In the simulation, the kicked-out planets were much more spread out than the stars. They were like guests who had been thrown out of the party and were wandering the parking lot, while the H-type objects in the real universe were still dancing on the dance floor with the stars.

The Conclusion: Born as Stars, Not Planets

The authors conclude that the "Planet Story" doesn't fit the evidence.

If the H-type objects were planets that got kicked out, they would be scattered far and wide, like seeds blown by the wind. Instead, they are huddled together with the stars, just like they were born right there in the cloud.

The Verdict: These H-type objects are likely not ejected planets. They are probably a new type of brown dwarf that formed directly from the collapse of gas clouds, just like stars do, but they just happen to have a unique chemical signature (that hydrocarbon smell) that makes them stand out.

In short: They aren't runaway planets; they are just the "little siblings" of the stars, born in the same cradle.

Technical Summary: Constraints on the Formation of H-type Objects in IC 348

Problem Statement
The formation mechanism of substellar objects, specifically distinguishing between brown dwarfs (formed via star-like collapse and fragmentation) and free-floating planets (formed in circumstellar discs and subsequently ejected), remains a significant challenge in stellar and planetary physics. This distinction is complicated by the overlap in mass regimes between gas giant planets and brown dwarfs (below the hydrogen-burning limit of $\sim 0.08 M_\odot$ ). Recent JWST observations of the IC 348 star-forming region have identified nine "H-type" objects (masses $0.002 – 0.012 M_\odot$ ) characterized by a unique 3.4 $\mu$ m spectral absorption feature from an unidentified aliphatic hydrocarbon. It is unclear whether these objects represent a distinct formation channel (e.g., planetary formation followed by ejection) or a variation of standard star-like formation.

Methodology
The authors employ a two-pronged approach combining observational data analysis with N-body simulations:

Observational Analysis:
- The study utilizes the updated census of IC 348 (Luhman et al. 2016, 2024; Luhman & Alves de Oliveira 2025), comprising 495 stars and brown dwarfs, including the nine H-type objects.
- Spatial distributions are quantified using three metrics:
  - Local Surface Density ( $\Sigma$ ): Calculated using the distance to the $N$ -th nearest neighbor ( $N=10$ ).
  - Q-parameter: Measures the degree of substructure vs. smoothness in the cluster.
  - $\Lambda_{MSR}$ Mass Segregation Ratio: Compares the Minimum Spanning Tree (MST) length of a specific subset (e.g., H-types) against randomly drawn subsets to determine if they are more clustered or dispersed than the general population.
- Statistical significance is assessed using Kolmogorov-Smirnov (KS) tests and specific thresholds for $\Lambda_{MSR}$ (requiring values $<0.5$ or $>2$ to be significant in small-N samples).
N-body Simulations:
- Simulations model the dynamical evolution of IC 348 using the kira integrator within the Starlab environment over 10 Myr.
- Initial Conditions: Regions are generated using a box-fractal method with varying fractal dimensions ( $D=1.6, 2.0, 3.0$ ) and radii ( $r_F$ ) to match the observed substructure and density. The Initial Mass Function (IMF) follows Maschberger (2013).
- Planet Injection: To test the ejection hypothesis, a $1 M_{Jup}$ planet is placed around every star in the $0.1 – 3 M_\odot$ range. Three scenarios are tested with initial semi-major axes ( $a_p$ ) of 1, 5, and 30 au.
- Constraints: Simulations are run to match the observed median surface density ( $\tilde{\Sigma} \approx 95$ stars pc $^{-2}$ ) and Q-parameter ( $Q \approx 0.98$ ) of IC 348 at ages of 1–5 Myr.

Key Results

Observational Spatial Distribution:
- The nine H-type objects in IC 348 exhibit a spatial distribution indistinguishable from the surrounding stars and brown dwarfs.
- KS tests on local surface densities yield p-values of 0.16 (H-types vs. all objects) and 0.12 (brown dwarfs vs. all objects), failing to reject the hypothesis that they share the same parent distribution.
- The $\Lambda_{MSR}$ ratio for the H-type objects is slightly above unity but not statistically significant, indicating no mass segregation (neither clustered nor dispersed) relative to the general population.
Simulation Constraints on Formation:
- The best-fitting initial conditions for IC 348 involve an initial radius $r_F = 1.3$ pc and fractal dimension $D = 1.6$ .
- Planet Ejection Counts:
  - Planets initially at 1 au produce too few free-floating planets (usually $<4$ ) to account for the nine H-types.
  - Planets initially at 30 au produce too many (20–40).
  - Planets initially at 5 au produce a number of free-floating planets (5–15) consistent with the observed count of H-types in 8 out of 10 simulation realizations.
- Spatial Distribution of Ejected Planets: In the simulations that produce the correct number of free-floating planets (initial $a_p = 5$ $a_{p} = 5$ au), these planets exhibit a significantly different spatial distribution compared to stars and brown dwarfs:
  - They are much more dispersed (inverse mass segregation, $\Lambda_{MSR} \ll 1$ ).
  - They reside in regions of significantly lower surface density (median $\sim 5.5$ stars pc $^{-2}$ for planets vs. $\sim 134$ stars pc $^{-2}$ for the whole region).
- This discrepancy persists for at least 3 Myr. While dynamical relaxation can make distributions indistinguishable after $\sim 5$ Myr in some simulations, the observed H-types at 3 Myr do not show this dispersion.

Significance and Conclusions
The paper concludes that the H-type objects in IC 348 are unlikely to have a planetary origin involving formation in circumstellar discs followed by dynamical ejection. While it is theoretically possible to produce the correct number of free-floating planets via stellar fly-bys (specifically if planets formed at $\sim 5$ au), the resulting spatial distribution of such ejected objects would be significantly more dispersed and lower in surface density than what is observed for the H-type objects.

Since the H-type objects share the same spatial and kinematic distribution as the stars and brown dwarfs in the cluster, the authors argue they likely formed via a star-like mechanism (e.g., collapse and fragmentation of the molecular cloud, or photoerosion of cores) rather than a planetary-like mechanism. The study suggests that the unique hydrocarbon spectral feature of H-type objects does not necessarily imply a different formation channel regarding their spatial origin, but rather distinguishes them as a specific subclass of brown dwarfs.

Caveats Noted by Authors
The authors acknowledge several limitations:

The age of IC 348 is uncertain (1–6 Myr); if the region is older ( $\sim 5$ Myr), dynamical relaxation could erase the spatial differences between ejected planets and the cluster.
Simulations do not include primordial binary populations, which could increase the rate of planet ejection.
The assumption of a single $1 M_{Jup}$ planet per star at fixed semi-major axes is a simplification of real planetary system architectures.
The Q-parameter in simulations is slightly higher than observed, potentially due to incompleteness in observational data.

The Detective Work: Looking at the Crowd

The Simulation: The "Cosmic Pinball" Experiment

The Conclusion: Born as Stars, Not Planets

More like this