The fragments of comet Shoemaker-Levy 9 struck Jupiter in July 1994. Each impact produced strong shocks both at the impact site and again when the ejecta plume fell back on the atmosphere. The resulting chemistry was distinctive. Reported products ranged from mildly oxidized (H₂O) to neutral (CO, S₂, OCS) to reduced (CS, CS₂, HCN, C₂H₄). The impacts also produced a great deal of dust, in part silicate (mildly oxidized) and in part carbonaceous (reduced). Evidently the impacts sampled several kinds of gas. Some gases were jovian, very dry to begin with, and by shock chemistry converted to HCN, perhaps CS and CS₂ (if H₂S were present), and tholins and/or soot (both words used in broad sense). Other parcels were derived more from comets. These supplied the water and the oxygen for CO, and were probably the source of sulfur for S₂. But apparently there were no very strongly oxidized parcels, which would have produced the unobserved compounds SO, SO₂, and CO₂.

Reentry gas temperatures were directly measured via hot CO. Gas temperatures rose quadratically with time, reaching 5000 K towards the end of the event. The dust was much cooler, roughly 600 K, and dust temperatures were sensibly constant over the ten minutes of strong IR radiation. Such temperatures are nearly ideal for surface-catalyzed chemistry, especially for the Fischer-Tropsch reactions that convert CO to hydrocarbons.

Although chemistry on Jupiter does not directly relate to chemistries of early Earth, it is clear that atmospheric chemistry of imaginary terrestrial planets could be greatly perturbed by impacts. Shocks at the impact site, shocks as ejecta reenter the atmosphere, and catalytic action by warm grains all offer means by which an atmosphere could be put into transient thermochemically disequilibrated states resembling somewhat the chemistry of relatively high temperatures. Very big impacts would process the whole atmosphere. It may take a long time to return to the "steady state".

To first approximation, one expects to thermally process some 10 impactor masses of air in an impact. An event on the scale of that made the lunar Orientale impact basin could process a 1 bar atmosphere. Under favorable conditions, as might be produced by a relatively reduced impactor, one might expect unexpected species (e.g., CH₄, say) at the percent level, which would then take thousands or perhaps even millions of years of atmospheric photochemistry to remove. Earth received hundreds of impacts on this scale in the hundreds of millions of years before 3.8 Ga; thus it is possible that a fair fraction of EarthÕs youth was spent in impact-perturbed states rather than in geochemically and photochemically mediated steady states. Thus otherwise similar young planets may exhibit a wide range of early atmospheres, any of which might have characterized Earth at a particular moment in time, and some of which would seem especially conducive to the origin of life.