Peter Pan disk

A Peter Pan disk is a circumstellar disk around a star or brown dwarf that appears to have retained enough gas to form a gas giant planet for much longer than the typically assumed gas dispersal timescale of approximately 5 million years. Several examples of such disks have been observed to orbit stars with spectral types of M or later. The presence of gas around these disks has generally been inferred from the total amount of radiation emitted from the disk at infrared wavelengths, and/or spectroscopic signatures of hydrogen accreting onto the star. To fit one specific definition of a Peter Pan disk, the source needs to have an infrared "color" of ${\displaystyle Ks-W4>2}$, an age of >20 Myr and spectroscopic evidence of accretion.^[1][2]

In 2016 volunteers of the Disk Detective project discovered WISE J080822.18-644357.3 (or J0808). This low-mass star showed signs of youth, for example a strong infrared excess and active accretion of gaseous material. It is part of the 45⁺¹¹
₋₇ Myr old Carina young moving group, older than expected for these characteristics of an M-dwarf.^[3][4] Other stars and brown dwarfs were discovered to be similar to J0808, with signs of youth while being in an older moving group.^[4][2] Together with J0808, these older low-mass accretors in nearby moving groups have been called Peter Pan disks in one scientific paper published in early 2020.^[5][2] Since then the term was used by other independent research groups.^[6][7][8]

Name

Peter Pan disks are named after the main character Peter Pan in the play and book Peter Pan; or, the Boy Who Wouldn't Grow Up, written by J.M. Barrie in 1904. The Peter Pan disks have a young appearance, while being old in years. In other words: The Peter Pan disks "refuse to grow up", a feature they share with the lost boys and titular character in Peter Pan.^[2][1]

Characteristics

The known Peter Pan disks have the H-alpha spectroscopic line as a sign of accretion. J0808 shows variations in the Paschen-β and Brackett-γ lines, which is a clear sign of accretion.^[1][2] It was also identified as lithium-rich, which is a sign of youth.^[4] Two peter pan disks (J0808 and J0632) show variation due to material from the disk blocking the light of the star.^[1][9] J0808 and J0501 also showed flares.^[1][2] Some of the Peter Pan disks (J0446, J0949, LDS 5606 and J1915) are binaries or suspected binaries.^[2][10][11] J0226 is a candidate brown dwarf^[2] and Delorme 1 (AB)b is a planetary-mass object in a circumbinary orbit.^[7][12][13] A detailed study of J0446B with JWST MIRI detected 9 hydrocarbons, two nitrogen-bearing species, two isotopes of CO₂, molecular hydrogen and two noble gases. Neon and molecular hydrogen strongly supports the idea that this disk is a long-lived primordial disk.^[14]

It was suggested that Peter Pan disks take longer to dissipate due to lower photoevaporation caused by lower far-ultraviolet and X-ray emission coming from the M-dwarf.^[2] Modelling has shown that disk can survive for 50 Myrs around stars with a mass less than 0.6 M_☉ and in low-radiation environments. At higher masses of 0.6 to 0.8 M_☉ the stars form an inner gap before 50 Myr, preventing accretion.^[15] Observations with the Chandra X-ray Observatory showed that Peter Pan Disks have a similar X-ray luminosity as field M-dwarfs, with properties similar to weak-lined T Tauri stars. The researchers of this study concluded that the current X-ray luminosity of Peter Pan disk cannot explain their old age. The old age of the disk could be the result of weaker far-ultraviolet flux incident on the disk, due to weaker accretion in the pre-main sequence stage.^[16] It was proposed that disks do form with a lifetime distribution, with some disks only existing for a few Myrs and others for dozens of Myrs. This would explain why some >20 Myr old M-dwarfs show accretion due to a disk, but not all M-dwarfs of this age. The research team found an initial disk fraction of 65% for M-dwarfs (M3.7-M6) and the disk lifetime distribution matches a Gaussian or Weibull distribution.^[17]

Known Peter Pan disks

Artist's Impression of a Peter Pan disk

SPHERE image of the disk around PDS 111, which is a higher-mass analogue of a Peter Pan disk

The prototype Peter Pan disk is WISE J080822.18-644357.3.^[2] It was discovered by the NASA-led citizen science project Disk Detective.^[18]

Murphy et al. found additional Peter Pan disks in the literature, which were identified as part of the Columba and Tucana-Horologium associations. The Disk Detective Collaboration identified two additional Peter Pan disks in Columba and Carina associations.^[2] The paper also mentions that members of NGC 2547 were previously identified to have 22 μm excess and could be similar to Peter Pan disks.^[2][19] 2MASS 08093547-4913033, which is one of the M-dwarfs with a debris disk in NGC 2547 was observed with the Spitzer Infrared Spectrograph. In this system the first detection of silicate was made from a debris disk around an M-type star. While the system shows the H-alpha line, it was interpreted to be devoid of gas and non-accreting.^[20]

In the following years additional objects were discovered.^{[7][9][10][11]} Some objects do not exactly fit the definition of Peter Pan disks, but are similar enough to be analogs: The object 2MASS J06195260-2903592 was found to be a 31+22
−10 Myr old analog to Peter Pan disks. This object does however not show accretion.^[21] The star PDS 111 is interpreted as a higher-mass analog of Peter Pan disks, with an age of 15.9+1.7
−3.7 Myrs, a mass of 1.2±0.1 M_☉, active accretion and a directly imaged disk.^[22] One team also found old accreting stars in the Large Magellanic Cloud in the Tarantula Nebula.^[23] This might be explained with a low metallicity in the LMC, which can lead to more massive disks that are less opaque.^[15]

List of Peter Pan disk candidates

Note: Wang et al. 2025^[24] lists 14 Peter Pan disks, here only 4 are listed that are older than 20 Myrs. Not included is US 3566 (Gaia DR3 155649614856576), which is a binary of a white dwarf and M-dwarf,^[25] which could be a cataclysmic variable. Not included are also 2MASS J04141188+2811535 and 2MASS J04091380+3136325, which could be Taurus members.^[24][26]

Name	Age (Myrs)	Association	spectral type	infrared excess	accretion	Reference
WISE J080822.18-644357.3	45+11 −7	Carina association	M5	yes	yes	[3][4]
2MASS J05010082-4337102	42+6 −4	Columba association	M4.5	yes	yes	[2][27]
2MASS J02265658-5327032	45±4	Tucana-Horologium association	L0δ	yes	yes	[2][27]
WISEA J044634.16-262756.1	42+6 −4	Columba association (but might be χ1 Fornacis member, which is 34 Myr old)	M6+M6	yes	likely	[2][28]
WISEA J094900.65-713803.1	45+11 −7	Carina association	M4+M5	yes	yes both	[2]
2MASS J15460752-6258042	~55	Argus association (but might be Beta Pictoris member)	M5	yes	yes	[10][28]
2MASS J05082729−2101444	30–44	Columba association (but could be Beta Pictoris member)	M5	yes	yes	[10]
LDS 5606	30–44	Columba association (but could be Beta Pictoris member)	M5+M5	yes	yes	[29][10]
Delorme 1 (AB)b	30–45	Tucana-Horologium association	L0 (very low gravity)	no	yes	[7][12][13]
2MASS J06320799-6810419	~45	Carina association	M4.5	yes	yes	[9]
2MASS J19150079-2847587	24±3	Beta Pictoris moving group	M4.8 (binary candidate)	yes	yes	[11]
StHα34	24.7+0.9 −0.6	Beta Pictoris moving group	M3+M3	yes	yes	[28][30][31]
Gaia DR3 2162887638405193216	>50		K9.4	yes	yes	[24]
CVSO 1241 (Gaia DR3 3223542525253775104)	25.1	Orion OB1? (would be 5–10 Myr)[32]	M3.8	yes	yes	[24]
2MASS J05171175+0702232 (Gaia DR3 3241216624914091136)	24.7	Lambda Orionis ring?[33]	M3.2	yes	yes	[24]
Gaia DR3 3319360599927089024	29.9		M3.6	yes	yes	[24]

2MASS J0041353-562112 was discarded as it belongs to the Beta Pictoris moving group and does not show excess.^[2]

Implications for planet formation around M-stars

There are different models to explain the existence of Peter Pan disks, such as disrupted planetesimals^[4] or recent collisions of planetary bodies.^[34] One explanation is that Peter Pan disks are long-lived primordial disks.^[6] This would follow the trend of lower-mass stars requiring more time to dissipate their disks. Exoplanets around M-stars would have more time to form, significantly affecting the atmospheres on these planets.^[1][2]

Peter Pan disks that form multiplanetary systems could force the planets in close-in, resonant orbits. The 7-planet system TRAPPIST-1 could be an end result of such a Peter Pan disk.^[9]

A Peter Pan disk could also help to explain the existence of Jovian planets around M-dwarfs, such as TOI-5205b. A longer lifetime for a disk would give more time for a solid core to form, which could initiate runaway core-accretion.^[35]

See also

• WISE J080822.18-644357.3
• Disk Detective
• Circumstellar disk
• Protoplanetary disk
• Debris disk
• T Tauri star
• HD 74389
• AU Microscopii
• List of nearby stellar associations and moving groups