The origin of Titan s atmosphere: some recent advances

The origin of Titan s atmosphere: some recent advances By Tobias Owen 1 & H. B. Niemann 2 1 University of Hawaii, Institute for Astronomy, 2680 Woodlawn Drive, Honolulu, HI 96822, USA 2 Laboratory for Atmospheres, Goddard Space Fight Center, Greenbelt, MD 20771, USA It is possible to make a consistent story for the origin of Titan s atmosphere starting with the birth of Titan in the Saturn subnebula. If we use comet nuclei as a model, Titan s nitrogen and methane could easily have been delivered by the ice that makes up 50% of its mass. If Titan s atmospheric hydrogen is derived from that ice, it is possible that Titan and comet nuclei are in fact made of the same protosolar ice. The noble gas abundances are consistent with relative abundances found in the atmospheres of Mars and Earth, the sun, and the meteorites. Keywords: Origin, atmosphere, composition, noble gases, deuterium 1. Introduction In this note, we will assume that Titan originated in Saturn s subnebula as a result of the accretion of icy planetesimals: particles and larger lumps made of ice and rock. Alibert & Mousis (2007) reached this same point of view using an evolutionary, turbulent model of Saturn s subnebula. They found that planetesimals made in the solar nebula according to their model led to a huge overabundance of CO on Titan. We obviously have no direct measurements of the composition of these planetesimals. We can use comets as a guide, always remembering that comets formed in the solar nebula where conditions must have been different from those in Saturn s subnebula, e.g., much colder. We can deduce the composition of the solar nebula at Saturn s distance from the sun by studying the atmosphere of Saturn itself, as this giant planet presumably captured both particles and gas during its formation (Pollack et al. 1996). We therefore begin with a brief discussion of Saturn before moving on to Titan. 2. Saturn The atmosphere of Saturn is >99% hydrogen and helium. In Saturn s H 2, we find D/H = 1.9 10 5 (Lellouch et al. 2001) compared to the value of 1.5 ± 0.1 10 5 in atomic hydrogen in the local interstellar medium (Linsky 2003). Except for trace constituents (see Atreya et al. 1999), the only other gas detected with certainty in Saturn s atmosphere is methane, with a mixing ratio of CH 4 /H 2 + He = 7 10 4 (Flasar et al. 2005). The value of D/H in Saturn s methane is the same as the value of this ratio in the H 2 (Lellouch et al. 2001). These numbers are all consistent with TEX Paper

2 T. Owen & H. B. Niemann models for Saturn s formation in which the methane or at least the carbon in it was delivered by icy planetesimals that formed at temperatures below 37 K (Owen et al. 1999; Gautier et al. 2001a, b; Owen & Encrenaz 2006; Gautier & Hersant 2005). There is no indication from accurate observations of Saturn s atmosphere so far that there was anything different about the solar nebula from which this planet formed compared with conditions at Jupiter as deduced from the Galileo Probe measurements in Jupiter s atmosphere (Niemann et al. 1996). However, it is essential to understand that we have only the abundance of carbon to use in the evaluation of different possibilities. Thus we are unable to evaluate models that suggest strongly nonsolar abundances relative to carbon in the local solar nebula and hence in Saturn. In particular, we cannot test the interesting suggestion by Gautier & Hersant (2005) that nitrogen may be deficient and xenon enhanced in Saturn s atmosphere. These discrepancies would arise from a hypothetical depletion of H 2 O in the solar nebula at Saturn s distance from the sun. Absent accurate observations of additional atmospheric abundances, we will simply assume that the composition of the solar nebula at Saturn was essentially identical with its composition at Jupiter. This then furnishes our reference for studying Titan. 3. Titan s atmosphere We are immediately struck by the dominance of nitrogen in Titan s atmosphere. Earth is the only other body in the solar system with an atmospheric pressure greater than a few µbars that exhibits this characteristic. Unlike Earth, we find a highly reducing atmosphere on Titan with methane instead of CO 2 and free H 2 instead of O 2. This mixture of gases resembles the putative composition of the outer solar nebula although not in these proportions. However, it is clear that Titan s atmosphere is secondary, produced by degassing of the icy planetesimals that formed the satellite. If Titan had collected its atmosphere directly from the solar nebula or from Saturn s subnebula, its atmosphere would contain slightly more Ne than N, while Ar/N would equal 1/30 according to modern solar abundances (Grevesse et al. 2005). Instead, on Titan today we find Ne/N < 10 8 and Ar/N = 1.5 10 7 (Niemann et al. 2005). (a) Methane The primitive reducing conditions in Titan s atmosphere today are a direct result of the 94 K surface temperature (Fulchignoni et al. 2005). The vapor pressure of water ice at 94 K is so low (<10 2 µbars; List 1958) that we expect no oxidation of methane from this source. Thus any original methane or methane made on the planet will not be converted to CO 2 as it must be on any warm inner planet in any planetary system. We expect the ratio of carbon to nitrogen to be 20 ± 10, as observed in the atmosphere of Venus and in the reconstituted atmosphere of Earth (Owen & Bar- Nun 1995, 2000). This expectation is consistent with the observed depletion of nitrogen in comets (Geiss 1988), attributed to the difficulty in trapping N 2 in ice (Owen & Bar-Nun 1995). N 2 is assumed to be the dominant carrier of nitrogen in the outer solar nebula following conditions in the interstellar medium where atomic N is also prevalent (van Dishoeck et al. 1993). Delivery of volatiles to Earth and

The origin of Titan s atmosphere 3 Venus by icy planetesimals apparently preserved this nitrogen depletion, resulting in the observed value of C/N = 20 ± 10 compared with the solar value of C/N = 4 (Grevesse et al. 2005). Interstellar abundances predict NH 3 /N 2 10 with NH 3 assumed to be the dominant condensed nitrogen compound. These values again lead to C/N 20, in good agreement with abundances found on comets, Venus and Earth (Owen & Bar-Nun 1995). Thus we might predict 4 C/N 20 ± 10 on Titan. Instead we observe C/N < 0.04 (Niemann et al. 2005), well outside this domain. On Earth, the carbon missing from the atmosphere is primarily bound up in carbonate rocks with some contribution from buried organic carbon (e.g., coal and petroleum). Where is the missing carbon on Titan? Another forceful argument that leads to this same question is the relatively short lifetime of methane in Titan s atmosphere. Photochemistry is destroying methane in Titan s upper atmosphere with subsequent reactions ultimately leading to escape or producing aerosols that are continually precipitating to the satellite s surface. The rate of this conversion is such that the present complement of atmospheric methane will be gone in 10 20 10 6 years (Strobel 1982; Yung et al. 1984). Thus the methane in the atmosphere requires a source. That source is not in the famous lakes and seas; it must be inside Titan. A detailed model for episodic renewal of atmospheric methane by release from a clathrate cap on a putative subsurface ocean has been developed in detail by Tobie et al. (2006). In this model, the methane in the subsurface ocean was originally brought to the satellite as clathrate hydrate, becoming a major component of the primitive atmosphere. It then formed a layer of clathrate at the surface of the ocean, from which it can escape into the atmosphere (Tobie et al. 2006). Manufacture of some of the methane on Titan could have occurred in a two-step process: water-rock reactions such as serpentinization produce hydrogen followed by Fischer-Tropsch type (FTT) reactions with CO and CO 2 to produce methane (Owen et al. 2005; Niemann et al. 2005). In the coma of the Oort cloud comet Hale-Bopp, CO and CO 2 were respectively 20 and 6 times as abundant as CH 4 (Bockelée-Morvan et al. 2004). The total carbon in comets is larger still, owing to the carbon-containing grains in their nuclei (Geiss 1988). Thus there should be plenty of carbon for the hypothesized FTT reactions. However, these reactions require catalysis, whereas the first step in this scenario serpentinization does not. Hence, it is more likely that this scenario will simply produce hydrogen. Furthermore, if it was preserved in the planetesimals that formed Titan, the 1% methane in Hale- Bopp s coma would provide far more of this gas than is necessary to account for an adequate reservoir to supply Titan s atmosphere (cp., NH 3 in 3b). Thus it seems more likely that direct delivery was responsible for the methane we see today. In principle, it might be possible to determine the relative importance of endogenous and exogenous methane on Titan by comparing isotope ratios with those in protosolar = cometary methane. However, we already know that the carbon in Titan s methane is fractionated: 12 C/ 13 C = 85 = ± 1 vs. 89.0 (VPDB), presumably by preferential escape of the light isotope from the atmosphere (Niemann et al. 2005). Hence D/H in the methane must be higher than the starting value (presumably the D/H in cometary methane) just as we anticipate that cometary methane has a higher value of 12 C/ 13 C than the value we find in atmospheric methane on Titan today. Thus we can conclude that today s methane is not protosolar, unlike

4 T. Owen & H. B. Niemann the methane we find in comets. The same problem would confront an effort to compare D/H in Titan s ice with D/H in contemporary atmospheric methane: the ice may well be protosolar, the methane certainly isn t. (b) Nitrogen As we have seen, nitrogen must have arrived on Titan as an easily trapped condensed compound, most likely NH 3. This is apparently what happened in meteorites, comets and the inner planets (Owen & Bar-Nun 1995; Owen et al. 2001; Meibom et al. 2007). In the case of the meteorites and Earth, this conclusion is supported by studies of 15 N/ 14 N (Owen et al. 2001; Meibom et al. 2007). In contrast, N 2, the dominant form of nitrogen in the outer solar nebula, was most probably included in the hydrodynamic collapse of surrounding nebula gases that brought H 2 and He to Saturn. Alternatively, Gautier & Hersant (2005) have suggested that NH 3 was the dominant form of nitrogen captured by Saturn, in which case the ratio of 15 N/ 14 N in Saturn s nitrogen should be similar to the one on Earth. The comets tell us that icy planestimals formed at T > 32 K will not contain N 2. This again supports our scenario for the origin of Titan s nitrogen: The solar nebula contains nitrogen primarily in two forms: N 2 plus 10% NH 3. Saturn acquires all forms of nitrogen in the solar nebula. The icy planetesimals that accreted to form Titan predominantly captured NH 3. Cometary comae contain about 1% nitrogen from NH 3 (Bockelée-Morvan et al. 2004). If this holds true for comet nuclei as well, and if comets indeed have compositions closely similar to the planetesimals that formed Titan, we can see that Titan ice ( 0.5 of the total mass of the satellite) could hold several times the amount of ammonia degassed to form the original atmosphere. The NH 3 in that atmosphere would have been photochemically converted to N 2 (Atreya et al. 1978). On Jupiter, and thus most likely in the sun, 15 N/ 14 N = 2.7 10 3 This nitrogen comes predominantly from N 2. On Earth, and in all varieties of meteorites, 15 N/ 14 N = 3.6 10 3. This nitrogen comes from some condensed nitrogen compound, probably NH 3. On Titan, 15 N/ 14 N = 6.0 10 3. The most straightforward interpretation of this high value for Titan is the loss of 14 N through atmospheric escape and upper atmosphere chemistry. The fractionation of the isotopes observed in Titan s nitrogen suggests that the original amount of atmospheric N 2 was considerably larger than the present one (Lunine et al. 1999). Assuming that 36 Ar degassed to the same extent as NH 3, we can then calculate the value of 36 Ar/N in the original atmosphere. Today, 36 Ar/N 2 is 2.8 10 7. Correcting for a factor 1.5 to obtain the original N 2 abundance produced from NH 3, we find N/ 36 Ar 10 7. On Earth N/ 36 Ar = 6 10 4. Some part of this large difference with Titan could be attributed to the warmer temperature at which Titan s ices formed in the Saturn subnebula, compared with the comets that brought 36 Ar to Earth (if they did!).

The origin of Titan s atmosphere 5 (c) Hydrogen The important ratio of D/H has been measured in methane (CH 3 D/CH 4 ) as D/H = 1.32 +0.15 0.11 10 4 by Bezard et al. (2007), while D/H in hydrogen (H 2 ) determined by the Huygens Gas Chromatograph Mass Spectrometer (GCMS) is approximately twice as large (the exact value has not yet been established). The atmospheric H 2 may be a product of the photolysis of methane. The difference in the values of D/H for CH 4 and H 2 would then be ascribed to a reactiondriven enrichment of D in the H 2 at the expense of the D in methane (Coustenis et al. 2007). Alternatively, the hydrogen may be leaking out from Titan s interior, the excess gas from reactions that are the first step in the possible production of methane described above. The classic example of such a reaction is serpentinization that would produce H 2 from water melted from Titan s ice interacting with rocks (Owen et al. 2005; Niemann et al. 2005). In this case the higher value of D/H in atmospheric H 2 would require a higher value in Titan s H 2 O. The problem is to know how, when and where this process occurred. This ambiguity in the explanation for the difference between D/H in methane and in H 2 could be resolved if we could measure D/H in Titan s ice directly. It would be especially interesting if the value of D/H in Titan s ice is indeed the value measured in atmospheric H 2 as that value may well be within the error bars of ground-based observations of D/H in cometary H 2 O (Meier & Owen 1999). (For example, the measurement of D/H in Comet Hyakutake was 2.9 ± 1.0 10 4 ; Bockelée-Morvan et al. 1998). In this case there would be a strong presumption that the ice making up 50% of the mass of Titan is identical with the ice in comet nuclei. Cometary ice is most likely protosolar, in which case so is its value of D/H. The only way to avoid this would be through a sequence in which the ice is warmed until it vaporizes, allowing cometary H 2 O to exchange isotopes with solar nebula H 2, and then the vapor re-freezes, forming ice with a lower value of D/H. For this to happen, the temperature in the outer solar nebula would have to be about 200 K or more. However, this warming would also lead to complete loss of CO from comets, which is not observed (e.g., Bockelée-Morvan et al. 2004). Thus a coincidence between the value of D/H in Titan s atmospheric H 2 with the value measured in cometary H 2 O could mean that Titan s ice is also protosolar that it also never became warm enough to go through this exchange phase. This possibility makes the determination of D/H in both the H 2 and in the ice especially interesting. (d) Noble gases Ne has not been detected on Titan, as mentioned previously. It is not expected to be present owing to the extreme difficulty in trapping this highly volatile gas. The production of radiogenic 40 Ar on Titan has already been discussed in Niemann et al. (2005). Its main relevance here is that it demonstrates there are pathways from the rocky interior of Titan through the thick ice mantle to the atmosphere. These pathways could be followed by the primordial noble gases and by methane as well (cf. 3a). If Titan s atmosphere follows the relative abundances of 36 Ar and the isotopes of krypton and xenon in the sun, or even in the atmospheres of Earth, Mars, Venus

6 T. Owen & H. B. Niemann and in the meteorites, it is not surprising that the GCMS did not detect Kr and Xe. The upper limits on both of these gases in Titan s atmosphere were 10 8 (Niemann et al. 2005). That means that 36 Ar/Kr > (2.8±0.3 10 7 )/10 8 ), = >28±3. On both the sun and Venus, 36 Ar/ 84 Kr 2 10 3. This means that if Titan had captured the noble gases in solar abundances, our upper limit was 140 times too high to detect krypton. Owen & Bar-Nun (1995, 2000) have suggested that the heavy noble gases on Earth and Mars were brought to those planets by icy planetesimals. On Earth 36 Ar/ 84 Kr = 28; on Mars it is 20.5±2.5 (Bogard et al. 1998). Given the uncertainties in these determinations, it might be possible that the Martian value overlaps the Titan upper limit. However, there is no trace of krypton in the GCMS mass spectra (Niemann et al. 2005). Xenon is approximately 10 times less abundant than Kr in the sun (Grevesse et al. 2005) and in the atmospheres of Earth and Mars (Bogard et al. 1998). Xenon has not yet been measured on Venus and is nearly equal to krypton in the meteorites. Without detecting Kr, it is therefore unlikely that Xe would be detectable if Titan noble gas abundances follow those of the sun or Earth, Mars, and the meteorites. Of course, there are other possibilities. If 36 Ar were strongly depleted on Titan by atmospheric escape or by some difficulty in trapping this gas in icy planetesimals compared with Kr and Xe, (e.g., Gautier & Hersant 2005), then 36 Ar/(Kr, or Xe) could be even smaller than on Mars. In this case, because we did detect 36 Ar, we should have detected Kr. Since we didn t achieve this detection, we can conclude with some confidence that there has been no selective depletion of 36 Ar relative to Kr greater than that on Mars. But suppose 36 Ar diffuses into the atmosphere more readily than Kr (and Xe). Or suppose one or both Kr and Xe are trapped as clathrates at the bottom of the putative methane ocean (Tobie et al. 2006). Then atmospheric 36 Ar would appear highly enriched relative to the other two elements, again giving us no chance to detect them in the GCMS spectra. The conclusion of this discussion is that it is not surprising that the GCMS did not detect krypton and xenon unless there was a unique depletion of 36 Ar in the atmosphere. Thomas et al. (2007) have proposed the formation of clathrate hydrates on Titan s surface as a sink for krypton and xenon, while Osegovic & Max (2005) have calculated that compound clathrates could explain the absence of xenon and presumably krypton as well. The arguments presented above show that the special conditions required for clathrate formation are not required to explain the upper limits on these two gases given the detection threshold of the GCMS. Jacovi & Bar-Nun (2008) have demonstrated in laboratory experiments that aerosols made from C 2 H 2 and HCN can trap the heavy noble gases in proportions consistent with the GCMS results. This process is also not required. 4. Future work It will be many years before we return to Titan with still more capability than Cassini-Huygens was able to achieve. Meanwhile, there are many more things to do. There must be a laboratory investigation of the hypothesis that photolysis of

The origin of Titan s atmosphere 7 CH 4 can lead to enrichment of D in the H 2 that is produced. This can be done in the next few years. We may hope that future studies of comets will improve our understanding of the starting mix of volatiles on Titan. Measuring the ratio of D/H in cometary CH 4 and other H-containing molecules, searching for the heavy noble gases and measuring their abundances and isotope ratios in comets are all critical goals. The value of N/ 36 Ar and the xenon isotope ratios in comets will be especially interesting. All of these measurements, along with many others, will be made by the Rosetta mission when it reaches comet Churyumov-Gerasimenko in 2012. We will need missions to other types of comets as well especially pristine comets entering the inner solar system for the first time before we know true cometary abundances. We await all these results with great interest. We thank Ingo Mueller-Wodarg for the invitation to this conference. We are grateful to D. Gautier for many helpful discussions and to O. Mousis for his comments and references. References Alibert, Y. & Mousis, O. 2007 Formation of Titan in Saturn s subnebula: constraints from Huygens probe measurements. Astron. Astrophys. 465, 1051 1060. (DOI 10.1051/0004-6361:20066402.) Atreya, S. K., Donahue, T. M. & Kuhn, W. R. 1978 Evolution of a nitrogen atmosphere on Titan. Science 201, 611 613. Atreya, S. K., Wong, M. H., Owen, T. C., Mahaffy, P. R., Niemann, H. B., de Pater, I., Drossart, P. & Encrenaz, T. 1999 A comparison of the atmospheres of Jupiter and Saturn: deep atmosphere composition, cloud structure, vertical mixing, and origin. Planet. Space Sci. 47, 1243 1262. Bezard, B., Nixon, C., Kleiner, I. & Jennings, D. 2007 Detection of 13 CH 3D on Titan. Icarus 191, 397 400. (DOI 10.1016/j.icarus.2007.06.004.) Bockelée-Morvan, D., et al. 1998 Deuterated water in comet C/1996 B2 (Hyakutake) and its implications for the origin of comets. Icarus 133, 147 162 (DOI 10.1006/icar.1998.5916.) Bockelée-Morvan, D., Crovisier, M. J., Mumma, M. J. & Weaver, H. J. 2004. The composition of cometary volatiles. In Comets II (ed. M. C. Festou, H. U. Keller & H. A. Weaver), pp. 391 423. Tucson: University of Arizona Press. Bogard, D. D. & Garrison, D. H. 1998 Relative abundances of argon, krypton, and xenon in the Martian atmosphere as measured in Martian meteorites. Geochim. Et Cosmochim. Acta 62, 1829 1835. Coustenis, A., et al. 2007 The composition of Titan s stratosphere from Cassini/CIRS mid-infrared spectra. Icarus 189, 35 62. (DOI 10.1016/j.icarus.2006.12.022.) Flasar, F. M., et al. 2005 Temperatures, winds, and composition in the Saturnian system. Science 307, 1247 1251. (DOI 10.1126/science.1105806.) Fulchignoni, M., et al. 2005 In situ measurements of the physical characteristics of Titan s environment. Nature 438, 785 791. (DOI 10.1038/nature04314.) Gautier, D., & Hersant, F. 2005 Formation and composition of planetesimals. Space Sci. Rev. 116, 25 52. (DOI 10.1007/s11214-005-1946-2.) Gautier, D., Hersant, F., Mousis, O. & Lunine, J. I. 2001a Enrichments in volatiles in Jupiter: a new interpretation of the Galileo measurements. Astrophys. J. 550, L227 L230. (DOI 10.1086/319648.) Gautier, D., Hersant, F., Mousis, O. & Lunine, J. I. 2001b Erratum: Enrichments in volatiles in Jupiter: a new interpretation of the Galileo measurements. Astrophys. J. 559, L183.

8 T. Owen & H. B. Niemann Geiss, J. 1988 Composition in Halley s Comet: clues to origin and history of cometary matter. Rev. Mod. Astron. 1, 1 27. Grevesse, N., Asplund, M. & Sauval, A. J. 2005 The new solar chemical composition. In Element stratification in stars: 40 years of atomic diffusion (ed. G. Alecian, O. Richard & S. Vauclair), EAS Publications Series, no. 17, pp. 21 30. (DOI 10.1051/eas:2005095.) Jacovi, R. & Bar-Nun, A. 2008 Removal of noble gases from Titan s atmosphere by trapping in its haze. Icarus 196, 302 304. (DOI 10.1016/j.icarus.2008.02.014.) Lellouch, E., Bezard, B., Fouchet, T., Feuchtgruber, H., Encrenaz, T. & de Graauw, T. 2001 The deuterium abundance in Jupiter and Saturn from ISO-SWS observations. Astron. Astrophys. 370, 610 622. (DOI 10.1051/0004-6361:20010259.) Linsky, J. L. 2003 Atomic deuterium/hydrogen in the Galaxy. Space Sci. Rev. 106, 49 60. (DOI 10.1023/A:1024673217736.) List, R. J. 1958. Smithsonian meteorological tables, 6th edn., p. 360. Washington, D.C.: Smithsonian Institution. Lunine, J. I., Yung, Y. L. & Lorenz, R. D. 1999 On the volatile inventory of Titan from isotopic abundances in nitrogen and methane. Planet. Space Sci. 47, 1291 1303. Meibom, A., Krot, A. N., Robert, F., Mostefaoui, S., Russell, S. S., Petaev, M. I. & Gounelle, M. 2007 Nitrogen and carbon isotopic composition of the sun inferred from a high-temperature solar nebular condensate. Astrophys. J. 656, L33 L36. (DOI 10.1086/512052.) Meier, R. & Owen, T. C. Cometary deuterium. 1999 Space Sci. Rev. 90, 33 43. (DOI 10.1023/A:1005269208310.) Niemann, H. B., et al. 1996 The Galileo Probe Mass Spectrometer: composition of Jupiter s atmosphere. Science 272, 846 849. Niemann, H. B., et al. 2005 The abundances of constituents of Titan s atmosphere from the GCMS instrument on the Huygens probe. Nature 438, 779 784. (DOI 10.1038/nature04122.) Osegovic, J. P. & Max, M. P. 2005 Compound clathrate hydrate on Titan s surface. J. Geophys. Res. 110, E8, E08004 Owen, T. & Bar-Nun, A. 1995 Comets, impacts and atmospheres. Icarus 116, 215 226. (DOI 10.1006/icar.1995.1122.) Owen, T. & Bar-Nun, A. 2000 Volatile contributions from icy planetesimals. In Origin of the Earth and Moon (ed. R. M. Canup & K. Righter), 459 471. Tucson: University of Arizona Press. Owen, T. & Encrenaz, Th. 2006 Compositional constraints on giant planet formation. Planet. Space Sci. 54, 1188 1196. (DOI 10.1016/j.pss.2006.05.030.) Owen, T., Mahaffy, P. R., Niemann, H. B., Atreya, S., Donahue, T., Bar-Nun, A. & de Pater, I. 1999 A low-temperature origin for the planetesimals that formed Jupiter. Nature 402, 269269. Owen, T., Mahaffy, P. R., Niemann, H. B., Atreya, S. & Wong, M. 2001 Protosolar nitrogen. Astrophys. J. 553, L77 L79. (DOI 10.1086/320501.) Owen, T., Niemann, H. B. & Atreya, S. 2005 Preliminary results from Huygens GCMS. Phys. Usp. 635, 6 38. Pollack J. B., Hubickyj O., Bodenheimer P., Lissauer J. J., Podolak M., Greenzweig Y. 1996 Formation of the giant planets by concurrent accretion of solids and gas. Icarus 124, 62 85. (DOI 10.1006/icar.1996.0190.) Strobel, D. F. 1982 Chemistry and evolution of Titan s atmosphere. Planet. Space Sci. 30, 839 848. (DOI 10.1016/0032-0633(82)90116-7.) Thomas, C., Mousis, O., Ballnegger, V. & Picard, S. 2007 Clathrate hydrates as a sink of noble gases in Titan s atmosphere. Astron. Astrophys 474, L17 L20. (DOI 10.1051/0004-6361:20078072.)

The origin of Titan s atmosphere 9 Tobie, G., Lunine, J. L. & Sotin, C. 2006 Episodic outgassing as the origin of atmospheric methane on Titan. Nature 440, 61 64. (DOI 10.1038/nature04497.) van Dishoeck, E. F., Blake, G. A., Draine B. T. & Lunine J. I. 1993. The chemical evolution of protostellar and protoplanetary matter. In Protostars and planets III (ed. E. H. Levy & J. I. Lunine), pp. 163 241. Tucson: University of Arizona Press. Yung, Y. L., Allen, M. & Pinto, J. P. 1984 Photochemistry of the atmosphere of Titan: comparison between model and observations. Astrophys. J. Suppl. Ser. 55, 465 506. (DOI 10.1086/190963.)