Saturn's largest satellite, Titan, has a dense atmosphere of nitrogen with a few per cent of methane1. At visible wavelengths its surface is hidden by dense orange-brown smog, which is produced in the stratosphere by photochemical reactions following the dissociation of methane by solar ultraviolet light. The most abundant of the products of these reactions is ethane, and enough of it should have been generated over the life of the Solar System to form a satellite-wide ocean one kilometre deep2. Radar observations3 have found specular reflections in 75 per cent of the surface spots observed, but optical searches for a sun-glint off an ocean have been negative4. Here I explain the mysterious absence or rarity of liquid ethane: it condenses onto the smog particles, instead of into liquid drops, at the cold temperatures in Titan's atmosphere. This dusty combination of smog and ethane, forming deposits several kilometres thick on the surface, including the observed dunes and dark areas, could be named ‘smust’. This satellite-wide deposit replaces the ocean long thought to be an important feature of Titan.
Smog particles and ethane molecules exist in Jupiter's atmosphere as well as Titan's, and their behaviour on Jupiter gives important clues for the understanding of Titan. On Jupiter they are less abundant than on Titan because of the large quantity of hydrogen present, which is responsible for some recycling of methane from its dissociation products5, but this small difference does not affect the important point: the vertical distribution of ethane in Jupiter's atmosphere requires that it should condense on the hydrocarbon smog particles. It is released only at temperatures above 300 K or, in the stratosphere, by photon-stimulated desorption. At the very low temperatures of Titan's lower atmosphere, the ethane (as well as other non-methane hydrocarbons) remains bound to the smog particles and is therefore not available for condensation into liquid ethane.
Ethane in Jupiter's troposphere was measured by the Galileo Probe Mass Spectrometer6,7, and its mixing ratio at 10–15 bar (6 × 10-6) is much greater than it is near 1 bar (4 × 10-8). A source is required in the deeper (higher-pressure) region, and it was suggested7 to be evaporation and dissociation, at the relatively warm temperatures of that level, of molecules from the smog particles. Here I note that the ethane mixing ratio of 10-5 measured in the stratosphere (at 10-3 bar) by its infrared emission8 is also much greater than at 1 bar, and that the small value of 4 × 10-8 near the latter level represents a deep minimum. The temperature is cold, about 150 K, but not cold enough to liquefy ethane. The minimum can be explained by the condensation of ethane molecules on (and in) the smog particles. They are released at and below Jupiter's 10 bar level, along with the original smog constituents, by the higher temperatures there: 350 K and hotter at deeper levels. Stratospheric temperatures, although warmer than those near 1 bar, are still not warm enough to release the observed ethane molecules; instead they are undoubtedly released by ultraviolet photons which do not penetrate to the 1 bar region on Jupiter, and the same must be true on Titan.
All these processes (except evaporation at deep levels) should operate in Titan's atmosphere with its much lower temperatures. Condensation of ethane into a liquid requires a sufficiently high concentration, which cannot be attained because ethane is more strongly bound to the smog particles. The lack of an ocean or other large bodies of liquid thus has a natural explanation.
The name I suggest for the material that accumulates on the surface is ‘smust’, a contraction of ‘smog + dust’. The basic point of this paper depends only on the observations of the vertical distribution of ethane in Jupiter's atmosphere and not on any other properties, which are discussed below on the basis of observations in Titan's atmosphere from the Huygens probe and on photochemical theory confirmed by remote observations of hydrocarbon molecules in the atmosphere.
The particles, observed in detail by the DISR (Descent Imager and Spectral Radiometer) instrument9 on Huygens, were found to have a radius of around 0.9 µm. Even if they form larger clumps once they have been deposited on the surface, they are more like ‘dust’ than ‘sand’. The amount generated over the life of the Solar System can be obtained from published theoretical calculations, the same ones that were used2 to find an amount for the ethane in the proposed ocean. Depending on the assumed density, the depth of a uniform layer could be several kilometres. The optical properties of a particle as observed from the Huygens probe9 are attributed to a loose aggregate of monomers with a radius of 0.05 µm. Each particle contains 256 or 512 monomers; the following description adopts the 512 value.
The radius of a sphere having the same surface area as an aggregate particle is 0.9 µm, but the loose, fractal structure of the particles implies a considerably greater physical dimension. For an estimate of the composition it is necessary to turn to photochemical computations5 of production rates. The basic composition is assumed to be of polyyne (polyacetylene) molecules ranging from C2H2 to C8H2; the production rate of longer carbon chains is believed to be sufficiently smaller that such chains can be neglected. Acetylene (C2H2) is assumed here to behave like the ethane. Following the evidence discussed above, ethane is included in the mix; it accounts for 48% of the mass while the four polyynes amount to 10, 14, 26 and 2%, respectively. (Ethylene, C2H4, is observed and computed to be much rarer, because it is readily photolysed.) The mean mass is 45.4 atomic mass units. The volume of an individual monomer, assumed spherical, is 5.24 × 10-16 cm3; if the density is assumed to be 0.1 g cm-3 the mass is 5.2 × 10-17 g and the number of molecules is 7 × 105. The corresponding numbers for a particle of 512 monomers are 3 × 10-14 g and 4 × 108 molecules.
The total flux down to the surface may be estimated from the values for the individual molecules5; starting with ethane, they are (in molecules cm-2 s-1): 5.8 × 109, 1.2 × 109, 1.7 × 109, 3.1 × 109 and 2.5 × 108. Weighting each of these by the number of carbon atoms in the molecule gives a total flux of 4.1 × 1010 carbon atoms cm-2 s-1; if this flux is assumed to have been constant for the life of the Solar System, the accumulated quantity is 5.9 × 1027 carbon atoms cm-2 and the surface density of carbon is 1.2 × 105 g cm-2; adding an allowance of one hydrogen atom per carbon atom increases the surface density of the deposit to 1.3 × 105 g cm-2. Undoubtedly most of the material is compressed by the mass lying above it; if the density is assumed to be 0.5 g cm-2 the average depth of the deposit is 2.6 km and its total volume 2 × 108 km3.
Widespread dunes have been observed by the Cassini radar. The paper reporting them10 proposes, as one option, that the origin of dunes lies in “Titan's atmospheric methane photochemistry” and gives an estimate of the volume of the material smaller than the one above by a factor of 20. Here I make the similar proposal that smust is the material of the dunes, and in addition that it contains the missing ethane and thus accounts for the rarity of liquid surfaces. The grain diameter for formation of dunes was suggested10 to be 180–250 µm, far larger than the 1.8 µm inferred for the grains in the atmosphere from the Huygens observations. However, if the structure of these aggregates is very loose the outside dimension could be considerably greater. The compressive forces in a deep layer may cause these original smust particles to aggregate into larger ones, which would naturally migrate to the surface as the small ones fill the voids in the material (the so-called ‘Brazil nut effect’). Only a small fraction of the particles in the smust deposit need be larger than those observed by Huygens, and would still be plenty to account for the dunes.
Alternatively, the particles in the dunes may be much smaller than the 180–250 µm, which is the optimum11 size for dune formation. Smaller particles can still be raised by the wind if the speeds are greater, especially if the interparticle cohesive forces are smaller than was assumed—it is these forces that inhibit the smallest particles from being lofted by the wind. The particles near the surface are found9 to be rather different from those at higher altitudes; they may include a component raised from the surface in addition to those falling from the stratosphere.
It would be desirable to verify in the laboratory the attachment of ethane molecules to smog particles, here deduced from their behaviour on Jupiter. Such particles, with their fluffy structure, have, however, not been produced in experiments, which instead generate a dense deposit on the walls of the vessel. It will be a challenge to reproduce this structure along with a realistic composition, and then to expose the particles to ethane molecules.
Note added in proof: A few days before this paper went to press, a report12 appeared suggesting the presence of a tenuous ethane cloud bordering the northern (winter) polar region and perhaps extending over the rest of the polar cap, which is in darkness. It occupies the height region of 30–50 km and contains a column of particles of radius 3 µm containing 60,000 particles cm-2: one particle per 33 cm in the 1 cm2 column. The evidence that the particles are composed of ethane is indirect but convincing. Their presence may be compatible with the smust particles discussed here, because they contain only a tiny fraction of the total ethane. The mixing ratio quoted12 for the stratosphere is 2.2 × 10-5. The number of ethane molecules in the proposed cloud is 7.6 × 1016 cm-2; the corresponding number of molecules in the ambient atmospheric column from 30 to 50 km is 4.1 × 1025 cm-2 and the ethane mixing ratio is therefore 1.8 × 10-9, only 8 × 10-5 of the available 2.2 × 10-5. It is entirely reasonable that these few molecules would not reside on the smust particles. A possible difficulty is that this small amount of ethane vapour would be unable to condense. On the other hand, if more ethane were available one would expect the cloud particles to grow larger; probably the attachment of most of the ethane to the smust particles is necessary to prevent this.
I thank A. L. Sprague, L. Doose and M. Tomasko for discussions. I also thank H. Niemann, the Principal Investigator of the Galileo Probe Mass Spectrometer.
About this article