Letter

# The Charon-forming giant impact as a source of Pluto’s dark equatorial regions

• Nature Astronomy 1, Article number: 0031 (2017)
• doi:10.1038/s41550-016-0031
Accepted:
Published online:

## Abstract

Pluto exhibits complex regional diversity in its surface materials 1,2 . One of the most striking features is the dark reddish material, possibly organic matter, along Pluto’s equator coexisting with the H2O-rich crust 2 . Little is known, however, about the surface process responsible for the dark equatorial regions. Here, we propose that Pluto’s dark regions were formed through reactions in elongated pools of liquid water near the equator, generated by the giant impact that formed Charon 3,​4,​5 . Our laboratory experiments show that dark reddish organic matter, comparable to Pluto’s dark materials, is produced through polymerization of simple organic compounds 6,7 that would have been present in proto-Pluto (for example, formaldehyde) by prolonged heating at temperatures ≥50 °C. Through hydrodynamic impact simulations, we demonstrate that an impactor, one-third the mass of Pluto, colliding with proto-Pluto—with an interior potential temperature of 150–200 K—could have generated both a Charon-sized satellite and high-temperature regions around Pluto’s equator. We also propose that high-velocity giant impacts result in global or hemispherical darkening and reddening, suggesting that the colour variety of large Kuiper belt objects 8,​9,​10,​11,​12 could have been caused by frequent, stochastic giant impacts in a massive outer protoplanetary disk in the early Solar System 13,​14,​15,​16 .

The informally named Cthulhu Regio is the most striking example of reddish H2O-rich crusts present on Pluto’s equator 1,2 . The dark reddish organic matter on Pluto’s equator could have been formed by radiolysis and photolysis of the surface ices 17 and/or might have been delivered from comets 18 . However, these processes would have resulted in global darkening, which is inconsistent with the presence of bright H2O-rich bedrocks, for example, Pulfrich crater in the northeast of Tombaugh Regio 2 . Furthermore, these processes occur only in the uppermost surface (for example, a few metres). Although further geological investigations are required, Pluto’s dark materials may even be present deep in the bedrock, since they appear on the floors, walls and ejecta of craters and on the high scarps of extensional fractures 19 . Given how heavily cratered Cthulhu Regio is 19 , the reddening process of the surface and deep crust would have occurred in ancient times.

We propose that the dark reddish material near the equator was formed by the giant impact that created Charon 3,​4,​5 . The Charon-forming impact, if it occurred, must have been one of the most energetic events in Pluto’s history, possibly melting a significant part of the surface and subsurface ice. Considering that large Kuiper belt objects (KBOs) accreted from comet-like icy planetesimals, the inventory of cometary volatiles would be fundamentally similar to that of KBO volatiles 20 . Thus, both proto-Pluto and the impactor are highly likely to have contained simple organic compounds that are typically found in comets, such as formaldehyde and ammonia 21 . In the warm liquid H2O pools formed by the Charon-forming impact, such organic compounds can be polymerized, forming complex organic matter 6,7 .

Here, we performed laboratory experiments on organic syntheses from formaldehyde, glycolaldehyde and ammonia in solutions, using concentrations comparable to those in comets 21 (see ‘Methods’). Figure 1 shows the transmittance spectra and the visual appearance of the solutions after heating at various temperatures and for various reaction times. At the beginning of the experiments, the solutions became darker and redder over time. This was caused by formation of high-molecular-weight, complex organic matter that included olefinic and aromatic molecules (Supplementary Figs 1 and 2), largely derived from polymerization of formaldehyde 6,7 . Darkening and reddening of the solutions stopped after ~1,100, ~800 and ~20 h for reactions at 25, 50 and 100 °C, respectively (Fig. 1). At 100 °C, refractory organic solids precipitated when reaction times were ≥200 h (Fig. 1c and Supplementary Fig. 3). The reason the darkening and reddening stopped may have been that the reactions reached a steady state. As the organic synthesis proceeded, the building materials for complex organic matter would have been gradually depleted until the polymerization and hydrolysis rates were balanced (Supplementary Fig. 1).

After prolonged heating (for example, >103 h), the solutions at 50 and 100 °C exhibited steep red slopes in visible wavelengths (Fig. 1) and relatively flat spectra in the near infrared (Supplementary Fig. 2). The refractory organic solids formed at 100 °C produced similar red slopes in the visible range, but were slightly blue between 1.0 and 2.5 μm (Supplementary Fig. 3). The visible and near infrared spectra of these solutions and refractory organic solids resemble those of Pluto’s equatorial dark regions 2,22 , although the comparison is mostly qualitative. In fact, the present study does not aim to reproduce exactly the reddish organic matter observed on Pluto by New Horizons. This is because the organic material on Pluto would have been altered by radiolytic and photolytic reactions after its formation, possibly leading to dehydrogenation, oxygenation and/or carbonitization 23,24 . As a result, the exact current composition and optical properties of the material may be partly due to such processes. Although the detailed chemistry of these processes is beyond the scope of the present study, our results suggest that prolonged heating at ≥50 °C could have facilitated the conversion of simple organic molecules into the complex dark reddish organic matter seen on Pluto.

We then carried out hydrodynamics simulations to see whether a Charon-forming giant impact could produce a large, heated (≥50 °C), elongated area near Pluto’s equator using the smoothed-particle hydrodynamics (SPH) method (see ‘Methods’). We performed 180 simulations of a potential Charon-forming giant impact between an undifferentiated impactor and an undifferentiated target, varying the values of five parameters related to the planetary and impact conditions (see ‘Methods’). Of the 180 runs, 51 resulted in formation of Charon-sized satellites with a satellite-to-planet mass ratio of 0.12 ± 0.03 (see Supplementary Table 1). The unsuccessful giant impacts were classified into four categories: small-satellite producing collisions, huge-satellite producing collisions, hit-and-run collisions, and others (Supplementary Fig. 4).

Figure 2 shows the results of a typical successful giant impact between an impactor and its target, where both objects had rock mass ratios (f rock) of 0.8 and initial potential temperatures (T 0) of 150 K. The impact conditions for Fig. 2 are similar to those for successful Charon-forming scenarios in previous simulation studies 4,5 . We found that elongated high-temperature regions surrounding the equator can be formed (Fig. 2d). The length and width of the major heated region on the surface at ≥50 °C reached ~3,000 km and ~800 km, respectively, which is comparable to Cthulhu Regio. The impact mainly embedded heat up to ~200 km deep (Fig. 2f), forming dark mantles beneath the equator. In addition to the major heated region, there were several patchy heated regions near the equator (Fig. 2e), which may explain the presence of Krun Macula and Balrog Macula 1 . Our simulation results also showed the remnants of unheated areas near the equator, which can explain the presence of neutral-coloured H2O-rich regions (for example, Pulfrich crater).

Assuming radiative cooling of the H2O pool, temperatures in the heated volume can be sustained at ~50 °C for >>103 h (see ‘Methods’), which is sufficient to form dark reddish materials. On the other hand, the satellite produced by the impact remained largely unheated (Fig. 2d), which is consistent with previous studies 4,5 . This happens because the surface materials of the impactor, which are heated by the first contact with the target (Fig. 2a,b), get tidally peeled off during the second grazing contact (Fig. 2c). Then, the unheated inner part of the impactor becomes a satellite. For almost all of the successful giant impacts, we found that an unheated Charon-sized satellite was formed by a similar mechanism.

A series of our simulations statistically suggest that a Charon-forming giant impact can generate elongated heated regions comparable to Pluto’s dark areas. The other eight successful giant impacts whose initial conditions for the colliding bodies were the same (f rock = 0.8 and T 0 = 150 K) also generated heated regions (≥50 °C) mostly at low latitudes (0 ± 30°; Fig. 3a). The latitudinal distribution of heated regions by the nine successful impacts (Fig. 3a) shows good agreement with that of the dark areas observed on Pluto (Fig. 3d). In the dark areas formed at low latitudes, radiolytic and photolytic reactions would have altered organic materials 23,24 after their formation, which may have further darkened the surfaces and made it more difficult to condense volatiles from the atmosphere. In contrast, less-heated high-latitude regions absorb less sunlight and thereby make volatiles easier to condense, which further reflect sunlight 2 . This positive feedback mechanism 2 may have enhanced the colour and material heterogeneity between the equatorial and polar regions.

We found that the area of the heated region depends rather strongly on f rock and T 0: lower values for these lead to a narrower heated area (Fig. 3b,c) because the specific heat of typical rock (~1 kJ kg−1 K−1) is less than that of ice (~2 kg−1 K−1) and liquid water (~4 kg−1 K−1). Moreover, the very high melting heat of ice (~300 kJ kg−1) prevents temperature increases. Therefore, our results suggest that a Charon-forming giant impact on a rock-rich target (f rock ≈ 0.8) with a high initial potential temperature (T 0 ≥ 150 K) is required to form highly elongated, dark reddish areas. A rock-rich target is consistent with the observed high bulk density of Pluto (1.860 g cm−3) 1 because the estimated f rock of Pluto ranges from 0.65 to 0.77, assuming the typical densities of ice (1 g cm−3) and rock (2.5–3.5 g cm−3). Previous studies 4,5 have shown that a highly differentiated impactor and target cannot produce a Charon-like satellite, suggesting that the interior of proto-Pluto would not have greatly exceeded 200 K. Thus, we propose that T 0 needs to reach 150–200 K at the time of impact.

Although the surface temperature is equilibrated with the solar flux, the potential temperature of proto-Pluto’s interior depends on the timing of its formation relative to that of the formation of the Solar System, and also to the accretion rate. According to calculations of interior temperatures during the accretion (see ‘Methods’), proto-Pluto would not have formed within ~5 × 106 years after the formation of the Solar System (Supplementary Fig. 10a). On the other hand, an interior potential temperature of 150–200 K can be achieved if its accretion from planetesimals occurred in ≥~106 years (see ‘Methods’ and Supplementary Fig. 10b). This timescale may be consistent with theoretical predictions of 107–108 years 25 for the formation of Pluto-sized objects at 35–50 au from the Sun.

Heat from small impacts, which have occurred throughout Pluto’s history, does not penetrate far below the surface, where temperatures are close to those at the surface (40–50 K). Given that typical temperature increases for such impacts would be 150–200 K, global darkening would not have resulted. On the other hand, ancient large impacts (diameter > 100 km) could have darkened the surface. Although we suggest that Charon was left unheated by the giant impact that formed it, dark reddish materials on Mordor Macula, in its northern terrain 1,2 , might have been formed by another large impact (diameter, 230 km) that created the Dorothy Gale basin. Contrary to a hypothesis that Mordor Macula formed by photolysis of polar CH4 ice 2 , we predict the absence of dark materials on Charon’s south pole.

In addition to potential Charon-forming giant impacts (v imp/v esc = 1.0–1.2; v imp, impact velocity; v esc, two-body escape velocity), we also simulated higher-velocity giant impacts on Pluto-sized objects, where v imp/v esc was 2.0. For a near head-on collision with an impact angle of 15°, the target was globally heated, resulting in global darkening and reddening (Fig. 4a). The 45° impact did not result in global heating, but the impacted hemisphere was widely heated up (Fig. 4b).

The impact-dependent variation in surface darkening may explain the colour diversity observed among large KBOs 8,​9,​10,​11,​12 . This colour diversity might not be explained by a single phenomenon common to these objects; for example, photolysis 17 . A possible explanation is frequent giant impacts of large KBOs 26,27 during formation in a massive outer disk 13 and their subsequent outward migration on the giant planets’ migration 14,​15,​16,28 . This may be testable by observations of protoplanetary disks around young solar-mass stars 29 and variability in surface colour on KBOs, including spacecraft observations in the extended New Horizons mission.

## Organic synthesis and analyses

The methodology of the organic synthesis experiments was based on previous studies simulating aqueous chemistry in carbonaceous chondrites 6,7 , but the concentrations of formaldehyde (CH2O), glycolaldehyde (C2H4O2) and ammonia (NH3) in the present study were 4–40 times lower than those of the previous study, given the concentrations of these molecules observed in comets and molecular clouds 21,30,31 . In our experiments, the 2 ml starting solution was 97.5% deionized distilled water, 1% CH2O as paraformaldehyde, 1% NH3, and 0.5% C2H4O2. The solution was flame-sealed into Pyrex glass ampoules while purging the gas phase with pure Ar gas. The sealed ampoule was heated isothermally at 25, 50 and 100 °C in an oven for up to 2,880 h. After heating, the solution was introduced into a 1-mm-thick quartz cell and analysed using a ultraviolet-visible spectrometer (Lambda 650; PerkinElmer) and a Fourier-transform infrared (FTIR) spectrometer (Frontier IR/NIR; PerkinElmer). A sample of the solution (~0.2 ml)—formed at 50 °C for a reaction time of 2,880 h—was dried at ~40 °C under Ar gas flow, and the residue was mixed with KBr powder and analysed via the FTIR spectrometer using the diffuse reflectance spectroscopy technique (DiffusIR; PIKE Technologies). In the experiments at 100 °C for reaction times ≥200 h, organic solids precipitated in the sample solutions. For these, the solutions were separated by centrifugation at 2,500 r.p.m. for 30 min (max. G-force was ~700). The centrifuged organic solids were dried and analysed with the ultraviolet-visible and FTIR spectrometers.

We did not measure the optical constant, for example, the k value, of the organic solids for comparison with the dark materials on Pluto. This was because the optical properties of Pluto’s dark materials would have changed from their initial values as a result of radiolytic and photolytic reactions 23,24 over geological time (as discussed in the main text). In addition, the optical properties of the organic solids might depend on the composition of the starting solution. In reality, Pluto’s interior is likely to contain nitriles and alcohols 21,30 , such as HCN and CH3OH, which were not used in the present study. Further investigations of the effects of both radiolytic reactions and starting material composition are important for quantitative comparison with the spacecraft observations. In addition, characterization of the dissolved organic matter formed in the experiments would be useful to constrain the reaction mechanisms. Although beyond the scope of the present study, these analyses are important and will be performed in a follow-up paper.

## SPH simulations

To investigate increases in temperature by a potential Charon-forming giant impact, we perform three-dimensional impact simulations using the SPH method 32 . Here, we used the code developed by Genda et al. 33,34 , which includes mutual self-gravity, but does not include material strength or frictional heating.

A fluid description is valid when stresses associated with self-gravity and compressional forces exerted by a collision greatly exceed the material strength; that is, the Hugoniot elastic limit (HEL). The HEL for H2O ice is around 300 MPa (ref. 35 ), which is achieved by collisions at velocities of ~300 m s−1. Since a typical impact velocity for a Charon-forming giant impact is more than 1 km s−1, the vicinity of the impact site, whose size is around the impactor’s size, should behave fluidly. Although the behaviour of material strength becomes more important in regions distant from the impact site, energy deposition (that is, temperature increase) mainly takes place near the impact site (Fig. 2 and Supplementary Fig. 5). Since the present study discusses the formation of a warm liquid-H2O ocean near the impact point and its subsequent movement and distribution on Pluto, the elastic behaviour of the distant regions would not affect our conclusions significantly.

Frictional heating, which is not included in our SPH code, may occur during a giant impact and would increase the temperatures of the target and impactor. However, it occurs effectively for frozen water but not for liquid water (owing to its fluidity). As the synthesis of complex organic materials in the warm ocean requires higher temperatures ≥~50 °C (Fig. 1), frictional heating would not affect our results for formation of complex organic materials after a giant impact. Thus, although our SPH code does not include material strength or frictional heating, they would not change our conclusions significantly.

Regarding the equations of state (EOSs), we used 5-Phase EOS 36 for H2O materials and M-ANEOS 37,38 for rock materials. These EOSs are state-of-the-art for impact simulations, and they give a thermodynamically consistent set of state variables such as pressure, density, internal energy and temperature. The 5-Phase H2O EOS includes five phases for ice Ih, ice VI, ice VII, liquid and vapour. For both EOSs, a set of state variables is given in tabular form. For rock material, we used the M-ANEOS table for basalt. Here, we consider a collision between an undifferentiated impactor and an undifferentiated target. Collisions between differentiated impactors and targets tend to result in formation of a disk composed mainly of ice materials 4,5 . From this icy disk, a satellite composed mainly of ice is formed, which is not consistent with the currently observed bulk density of Charon (1.702 ± 0.021 g cm−3; ref. 1 ). We made a table of the ice and rock mixture using the 5-Phase H2O EOS and basalt M-ANEOS on the basis of pressure and temperature equilibrium. We applied this mixture table for undifferentiated bodies, and as a result, the ice temperature was always equal to that of rock. This assumption is valid for well-mixed undifferentiated bodies. We can obtain the temperature from this mixture table by using the density and internal energy that are calculated by the SPH code.

We carried out 180 simulations for giant impacts on proto-Pluto. The total mass of an impactor and target (M tot) was set to 1.46 × 1022 kg, which is the combined mass of Pluto (M P = 1.303 × 1022 kg) and Charon (M C = 1.587 × 1021 kg) (ref. 1 ). Two f rock values were considered: 0.5 and 0.8. For the initial potential temperature (T 0) of a target and impactor, we consider three values of T 0: 50, 150 and 200 K. Their interiors had an adiabatic temperature profile. Two impactor-to-target mass ratios were considered: M imp/M tot = 0.30 and 0.35, where M imp is the impactor’s mass. Here, the impactor and target were assumed to have no pre-impact spin. For the impact velocity (v imp), the following five ratios were considered: v imp/v esc = 1.00, 1.05, 1.10, 1.15 and 1.20, where v esc is the two-body escape velocity 39 . For the total angular momentum of the collision (L imp), three ratios were considered: L imp/L PC = 0.95, 1.00 and 1.05. where L PC is the current angular momentum of the Pluto–Charon system (6.00 × 1030 kg m2 s−1; ref. 40 ).

In total, we performed 180 impact simulations, comprising 2 (rock mass fraction) × 3 (initial temperature) × 2 (mass ratio) × 5 (impact velocity) × 3 (angular momentum) simulations. The total number of SPH particles used in each simulation was fixed at 100,000, and all SPH particles in the impactor and target had equal mass. Typical impact simulations took ~2 × 105 s (~2 d), while some took longer to finish the collisional evolution (~2 × 106 s).

Using a particle method such as SPH, it is not straightforward to determine the temperature at an arbitrary location (x). In the SPH method, a physical value (A(x)) should be determined by summing up the contributions of the neighbouring SPH particles 41 : $(1) A ( x ) = ∑ j m j A j ρ j W ( x − x j , h j )$ where m is the mass of the SPH particle, ρ is the density, h is the smoothing length and W is the kernel function. The subscript j represents neighbouring SPH particles, whose number was set to 64. For the kernel function W, we use the spherically symmetric spline kernel function 42 . Impact simulations give the values of m j , A j , ρ j , x j and h j . We apply Equation (1) to estimate the temperature in the planet using T instead of A.

It is also not straightforward to determine the surface boundary of the planet because the radial density distribution near the surface is not sharp in the SPH method. Here, we define the surface as the radius at which the density is 1,000 kg m−3. Density at an arbitrary location can be calculated from Equation (1) by using ρ instead of A. Since the planet is oblate due to its spin after formation of the satellite, the distance between the surface and the centre of mass depends on the latitude and longitude. The SPH method’s surface effects are examined by comparing the results of giant impact simulations for different spatial resolutions using 100,000 and 30,000 particles for the same impact conditions (Supplementary Fig. 6). We found that the major results for the formation of warm liquid oceans at low latitudes did not change, although the results using 100,000 particles showed wider areas suitable for organic synthesis (Supplementary Fig. 6). The results show that the surface effects would not affect our conclusions significantly.

## Timescale for cooling of a warm liquid-water pool

To estimate the timescale for the cooling of the main warm pool formed after a giant impact, we assumed that convection in the water pool occurs effectively, and considered first the radiative cooling from the surface.

We estimated the timescale for the cooling of a warm pool from 50 °C to 40 °C, that is, a temperature decrease of 10 °C, as our experimental results showed that darkening of the solution is sensitive to temperature, and that dark reddish materials are not generated at a temperature of 25 °C. According to our SPH calculations, the length and width of the main water pool at temperatures ≥50 °C would be ~3,000 km and ~800 km, respectively, on surface area of ~2.4 × 106 km2 (Fig. 2). Assuming a black body for the water pool, radiation from the surface at 50 °C would be ~1015 W.

We also estimated a timescale for cooling of the liquid-water pool through conduction into the surrounding ice. In the pool, rocky materials would sink and deposit readily at the bottom. Thus, heat would be lost by conduction through the rock layer at the bottom. The heat flux through the rock layer was calculated as C × T × S, where C is the thermal conductivity of rock (~3 W m−1 K−1) and S is the surface area (~2.4 × 1012 m2). For the boundary conditions, we considered the temperature of the ocean–rock boundary to be 277 K, or 4 °C, where the density of water becomes maximal. The temperature of the bottom of the rock layer was considered to be 273 K, or 0 °C, because the rock materials initially deposit at the water–ice boundary. T for the 10-m-thick rock layer was 0.4 K m−1, and the heat flux obtained was ~1012 W (even for a 0.1-m-thick rock layer, the heat flux reached only ~1014 W). This means that the deposition of rock materials works as a thermostat for the pool, preventing effective cooling. In reality, the rock layer has porosity, which significantly decreases the thermal conductivity. Although this is a first order estimate, the results suggest that heat flux due to the conductance would be lower than that of the radiation from the surface.

If we consider 200 km as the depth and 4.2 kJ kg−1 K−1 as the specific heat capacity of the water pool, the timescale for cooling can be estimated at ~106 h. Even when considering the cooling of a 100-m-thick surface layer of the water pool, it would take ~103 h for cooling, which is comparable to the time our organic synthesis at 50 °C took to reach a steady state (Fig. 1). In reality, the cooling of a water pool must be more complex than our model; however, the resulting estimates allow us to suggest that, even for the very surface of a water pool, temperatures sufficient to produce dark reddish materials would be sustained for a sufficient time.

## Interior temperature of pre-impact Pluto

We calculated the interior temperature of proto-Pluto based on a previous study 43 . We assumed that proto-Pluto had an accretional temperature profile, which was determined from the balance of energy: heating on the surface due to the impact and release of potential energy, and radiation from the surface 43 . The ambient temperature was set to 40 K, the density of accreting material to 1,860 kg m−3 (ref. 1 ), and the specific heat to 1.13 kJ kg−1 K−1. A constant increase in Pluto’s radius was assumed, as was a gravitational focusing factor (controls impact velocity) of five 43 . The temperature increase due to radiogenic heating of short-lived radionuclides was also taken into account; the parameters adopted to calculate this heating rate have been reported previously 43 . In addition to these assumptions, two parameters were required to calculate the accretional temperature profile: the time at which the planet began accreting relative to formation of the Solar System, t start, and the length of time for accretion, τ acc. The central temperature depends only on t start. This model assumes that all the impact energy was deposited within accreting Pluto where it is subject to radiation cooling (that is, the limit of small impactors) 43 . Thus, the obtained temperature profile for accreting Pluto provides a lower limit for a given accretion time. Large impacts would deposit heat deep below the surface, leading to a higher interior temperature 43 . This means that the obtained τ acc would be the shortest accretion time to achieve the final temperature profile of accreting Pluto.

## Image analysis

We calculated the fraction of dark areas on the surface of Pluto for different latitudinal bands using a panchromatic global map based on New Horizons’ images, which were acquired via the Long Range Reconnaissance Imager (LORRI). We first removed the pixels of a uniformly dark region in the southern hemisphere where data is not available yet. We then normalized the pixel value, D (darkness); D = 1 for the darkest and D = 0 for the brightest. The number of pixels that exceeded a threshold (D P) divided by the total number of pixels gave the dark-area fraction. This fraction was calculated for each latitudinal band, which had widths of 10°. We analysed data between 90° N and 30° S of Pluto where the number of pixels is sufficiently large. We used different values of D P, ranging from 0.0 to 0.9. The software package, R v.3.3.1 (R Foundation for Statistical Computing, 2016; https://www.r-project.org/), was used for this analysis, and the image that was analysed is available at http://pluto.jhuapl.edu/Multimedia/Science-Photos/image.php?page=2&gallery_id=2&image_id=432.

## Data availability

The data that support the plots within this paper and other findings of this study are available from the corresponding author on reasonable request.

How to cite this article: Sekine, Y., Genda, H., Kamata, S. & Funatsu, T. The Charon-forming giant impact as a source of Pluto’s dark equatorial regions. Nat. Astron. 1, 0031 (2017).

## References

1. 1.

et al. The Pluto system: initial results from its exploration by New Horizons. Science 350, aad1815 (2015).

2. 2.

et al. Surface compositions across Pluto and Charon. Science 351, aad9189 (2016).

3. 3.

On the origin of the Pluto–Charon binary. Astrophys. J. Lett. 344, L41–L44 (1989).

4. 4.

A giant impact origin of Pluto–Charon. Science 307, 546–550 (2005).

5. 5.

On a giant impact origin of Charon, Nix, and Hydra. Astron. J. 141, 35 (2011).

6. 6.

et al. Establishing a molecular relationship between chondritic and cometary organic solids. Proc. Natl Acad. Sci. USA 108, 19171–19176 (2011).

7. 7.

, & Exploring the potential formation of organic solids in chondrites and comets through polymerization of interstellar formaldehyde. Astrophys. J. 771, 19 (2013).

8. 8.

, & Discovery of a planetary-sized object in the scattered Kuiper belt. Astrophys. J. Lett. 635, L97–L100 (2005).

9. 9.

et al. (50000) Quaoar: surface composition variability. Astron. Astrophys. 584, A107 (2015).

10. 10.

et al. (90377) Senda: investigation of surface compositional variation. Astrophys. J. 140, 2095–2100 (2010).

11. 11.

, & Rotationally resolved spectroscopy of dwarf planet (136472) Makemake. Astron. Astrophys. 577, A86 (2015).

12. 12.

et al. The surface of the transneptunian object 90482 Orcus. Astron. Astrophys. 437, 1115–1120 (2005).

13. 13.

On the number of planets in the outer solar system: evidence of a substantial population of 1000-km bodies. Icarus 90, 271–281 (1991).

14. 14.

, , & Origin of the orbital architecture of the giant planets in the solar system. Nature 435, 459–461 (2005).

15. 15.

, , & Origin of the cataclysmic Late Heavy Bombardment period of the terrestrial planets. Nature 435, 466–469 (2005).

16. 16.

, , , & Origin of the structure of the Kuiper belt during a dynamical instability in the orbits of Uranus and Neptune. Icarus 196, 258–273 (2008).

17. 17.

, , , & Ice chemistry on outer solar system bodies: electron radiolysis of N2-, CH4-, and CO- containing ices. Astrophys. J. 812, 150 (2015).

18. 18.

, , & Black rain: the burial of the Galilean satellites in irregular satellite debris. Icarus 223, 775–795 (2013).

19. 19.

et al. The geology of Pluto and Charon through the eyes of New Horizons. Science 351, 1284–1293 (2016).

20. 20.

, , & In The Solar System Beyond Neptune (eds. Barucci, M. A., Boehnhardt, H., Cruikshank, D. P., Morbidelli, A. & Dotson, R. ) 213–241 (Univ. Arizona Press, 2008).

21. 21.

& The chemical composition of comets—emerging taxonomies and natal heritage. Ann. Rev. Astron. Astrophys. 49, 471–524 (2011).

22. 22.

et al. Pluto’s global surface composition through pixel-by-pixel Hapke modeling of New Horizons Ralph/LEISA data. Preprint at (2016).

23. 23.

, , , & IR spectra of irradiated organic materials. Infrared Phys. 28, 183–188 (1988).

24. 24.

, , & The alteration of organic matter in response to ionizing irradiation: chemical trends and implications for extraterrestrial sample analysis. Geochim. Cosmochim. Acta 70, 1020–1039 (2006).

25. 25.

Planet formation in the outer solar system. Publ. Astron. Soc. Pacific 114, 265–283 (2002).

26. 26.

et al. Satellites of the largest Kuiper belt objects. Astrophys. J. Lett. 639, L43 (2006).

27. 27.

& Interpreting the densities of the Kuiper belt’s dwarf planets. Mon. Not. R. Astron. Soc. 460, 1542–1548 (2016).

28. 28.

, & The evolution of a Pluto-like system during the migration of the ice giants. Icarus 246, 330–338 (2015).

29. 29.

et al. Direct imaging and spectroscopy of a young extrasolar Kuiper belt in the nearest OB association. Astrophys. J. Lett. 807, L7 (2015).

30. 30.

, , & In Comets II (eds Festou, M. C., Keller, H. U. & Weaver, H. A. ) 391–423 (Univ. Arizona Press, 2004).

31. 31.

et al. Organic compounds on comet 67P/Churyumov-Gerasimenko revealed by COSAC mass spectrometry. Science 349, aab0689 (2015).

32. 32.

A numerical approach to the testing of the fission hypothesis. Astron. J. 82, 1013–1024 (1977).

33. 33.

, , , & Resolution dependence of disruptive collisions between planetesimals in the gravity regime. Icarus 262, 58–66 (2015).

34. 34.

, & Warm debris disks produced by giant impacts during terrestrial planet formation. Astrophys. J. 810, 136 (2015).

35. 35.

& Shock Hugoniot of H2O ice. Geophys. Res. Lett. 30, 1332 (2003).

36. 36.

& Impact crater formation in icy layered terrains on Mars. Meteorit. Planet. Sci. 43, 1993–2013 (2008).

37. 37.

& Improvements in the CHART D Radiation-Hydrodynamic Code III: Revised Analytic Equations of State (Sandia National Laboratories , 1972).

38. 38.

A hydrocode equation of state for SiO2 . Meteorit. Planet. Sci. 42, 2079–2098 (2007).

39. 39.

, & Merging criteria for giant impacts of protoplanets. Astrophys. J. 744, 137 (2012).

40. 40.

, & Complete tidal evolution of Pluto–Charon. Icarus 233, 242–258 (2014).

41. 41.

Smoothed particle hydrodynamics. Annu. Rev. Astron. Astrophys. 30, 543–574 (1992).

42. 42.

& A refined particle method for astrophysical problems. Astron. Astrophys. 149, 135–143 (1985).

43. 43.

& Constraints on gas giant satellite formation from the interior states of partially differentiated satellites. Icarus 198, 163–177 (2008).

## Acknowledgements

Y.S. thanks S. Tachibana for providing the methods to produce the organic matter from formaldehyde solution. This study was supported by Grant-in-Aids for Scientific Research from the Japan Society for Promotion of Science (26707024, 16001111, 16K13873 and 15K13562), from the JGC-S Scholarship Foundation, and from the Astrobiolgy Center of the National Institutes of Natural Sciences (NINS).

## Author notes

• Yasuhito Sekine
•  & Hidenori Genda

These authors contributed equally to this work.

## Affiliations

1. ### Department of Earth and Planetary Science, The University of Tokyo, 7-3-1 Hongo, Bunkyo 113-0033, Japan

• Yasuhito Sekine
•  & Taro Funatsu
2. ### Earth-Life Science Institute, Tokyo Institute of Technology, 2-12-1 Ookayama, Meguro 152-8550, Japan

• Hidenori Genda
3. ### Creative Research Institution, Hokkaido University, Kita-10, Nishi-8, Sapporo 060-0810, Japan

• Shunichi Kamata

## Contributions

Y.S. developed the idea for the study, and Y.S. and T.F. performed the experiments. H.G. performed hydrodynamic simulations. S.K. performed interior temperature calculation and image analysis. All the authors contributed to the interpretation of the results.

## Competing interests

The authors declare no competing financial interests.

## Corresponding author

Correspondence to Yasuhito Sekine.

## PDF files

1. 1.

### Supplementary Information

Supplementary Table 1, Supplementary Figures 1–10.