Icarus 156, 76–106 (2002) doi:10.1006/icar.2001.6758, available online at http://www.idealibrary.com on Photochemistry of a Volcanically Driven Atmosphere on Io: Sulfur and Oxygen Species from a Pele-Type Eruption Julianne I. Moses Lunar and Planetary Institute, 3600 Bay Area Blvd., Houston, Texas 77058-1113 E-mail: [email protected] Mikhail Yu. Zolotov Department of Earth and Planetary Sciences, Washington University CB1169, One Brookings Drive, St. Louis, Missouri 63130-4899 and Bruce Fegley, Jr. Department of Earth and Planetary Sciences, Planetary Chemistry Laboratory, McDonnell Center for the Space Sciences, Washington University CB1169, One Brookings Drive, St. Louis, Missouri 63130-4899 Received November 15, 2000; revised September 17, 2001 polar regions might also affect the surface spectral properties. We To determine how active volcanism might affect the standard predict that the S/O ratio in the torus and neutral clouds might picture of sulfur dioxide photochemistry on Io, we have developed be correlated with volcanic activity—during periods when volcanic a one-dimensional atmospheric model in which a variety of sulfur-, outgassing of S2 (or other molecular sulfur vapors) is prevalent, we oxygen-, sodium-, potassium-, and chlorine-bearing volatiles are would expect the escape of sulfur to be enhanced relative to that of volcanically outgassed at Io’s surface and then evolve due to photol- oxygen, and the S/O ratio in the torus and neutral clouds could be ysis, chemical kinetics, and diffusion. Thermochemical equilibrium correspondingly increased. c 2002 Elsevier Science (USA) calculations in combination with recent observations of gases in the Key Words: Io; photochemistry; atmospheres, composition; vol- Pele plume are used to help constrain the composition and physi- canism; geochemistry. cal properties of the exsolved volcanic vapors. Both thermochemi- cal equilibrium calculations (Zolotov and Fegley 1999, Icarus 141, 40–52) and the Pele plume observations of Spencer et al. (2000; Sci- 1. INTRODUCTION ence 288, 1208–1210) suggest that S2 may be a common gas emitted in volcanic eruptions on Io. If so, our photochemical models indicate Jupiter’s satellite Io has one of the most unusual atmospheres that the composition of Io’s atmosphere could differ significantly in our solar system (see the reviews of Lellouch 1996, Spencer from the case of an atmosphere in equilibrium with SO2 frost. The major differences as they relate to oxygen and sulfur species are an and Schneider 1996), and our understanding of the composi- increased abundance of S, S2,S3,S4, SO, and S2O and a decreased tion and physical properties of Io’s atmosphere has evolved abundance of O and O2 in the Pele-type volcanic models as com- rapidly in the past few years. Recent observations of SO2,a pared with frost sublimation models. The high observed SO/SO2 molecule that has long been known to be present in the io- ratio on Io might reflect the importance of a contribution from vol- nian atmosphere (Pearl et al. 1979), have helped constrain the canic SO rather than indicate low eddy diffusion coefficients in Io’s surface pressure and the degree of atmospheric “patchiness” atmosphere or low SO “sticking” probabilities at Io’s surface; in on Io (Lellouch et al. 1990, 1992, Ballester et al. 1990, 1994, that case, the SO/SO2 ratio could be temporally and/or spatially Sartoretti et al. 1994, 1996, Lellouch 1996, Trafton et al. 1996, variable as volcanic activity fluctuates. Many of the interesting vol- Hendrix et al. 1999, McGrath et al. 2000). Other molecular con- canic species (e.g., S ,S,S, and S O) are short lived and will be 2 3 4 2 stituents such as SO and S have recently been identified on the rapidly destroyed once the volcanic plumes shut off; condensation 2 of these species near the source vent is also likely. The diffuse red satellite (Lellouch et al. 1996, McGrath et al. 2000, Spencer deposits associated with active volcanic centers on Io may be caused et al. 2000), and atomic and molecular emissions from neutral by S4 radicals that are created and temporarily preserved when sul- O, S, Na, Cl, H, SO2 (and/or SO) have been mapped on and about fur vapor (predominantly S2) condenses around the volcanic vent. Io’s disk from Galileo Orbiter, Hubble Space Telescope (HST), Condensation of SO across the surface and, in particular, in the and ground-based observations (e.g., Belton et al. 1996; Geissler 76 0019-1035/02 $35.00 c 2002 Elsevier Science (USA) All rights reserved. SULFUR AND OXYGEN PHOTOCHEMISTRY ON IO 77 et al. 1999b; Roesler et al. 1999; Ballester et al. 1999; Bouchez have concentrated on sublimation-generated or even sputtering- et al. 2000; McGrath et al. 2000; Feldman et al. 2000a; Trafton generated atmospheres (e.g., Kumar 1980, 1982, 1984, 1985, 2000; Retherford et al. 2000a, 2000b; see also the Voyager ob- Summers 1985, Summers and Strobel 1996, Wong and Johnson servations of Cook et al. 1981). Continuing Io plasma torus and 1996, Wong and Smyth 2000). The effects of active volcan- neutral cloud observations, including the recent detection of Cl+ ism on ionian photochemistry have remained relatively unex- and Cl++ in the torus (Kuppers¨ and Schneider 2000, Schneider plored. To determine how volcanic outgassing might affect the et al. 2000, Feldman et al. 2000b), confirm that Na, K, and Cl standard picture of SO2 photochemistry on Io, we have devel- in atomic and/or molecular form must be present in Io’s atmo- oped a simple one-dimensional (1-D) model in which a vari- sphere (e.g., Spencer and Schneider 1996). These observations ety of S-, O-, Na-, K-, Cl-, and H-bearing volatiles are vol- have refocused attention on the composition and chemistry of canically outgassed from the surface and then evolve due to Io’s atmosphere and on the importance of active volcanism in photolysis and subsequent chemical kinetics. Initial and bound- maintaining that atmosphere. ary conditions for the volcanic outgassing are specified (with The three main proposed mechanisms for generating an at- some modification) from the thermochemical equilibrium cal- mosphere on Io include frost sublimation, surface sputtering, culations of Zolotov and Fegley (2000a) and Fegley and Zolotov and active volcanism. Although the relative importance of each (2000). The former used recent HST observations (McGrath mechanism is not well understood, both frost sublimation and et al. 2000, Spencer et al. 2000) to constrain the temperature, active volcanism appear to play a role. Wong and Smyth (2000) pressure, and speciation of the exsolved sulfur- and oxygen- summarize the evidence for a sublimation-driven component. bearing gases in the vicinity of the Pele volcanic vent, and the The role of active volcanism is demonstrated by several aspects latter calculated the equilibrium chemistry of Na-, K-, and Cl- of the observations. First, the sulfur dioxide vapor appears to be bearing volcanic gases on Io. Eddy and molecular diffusion are nonuniformly distributed over Io’s surface; the observed patch- assumed to dominate the vertical transport of the atmospheric iness does not have a subsolar latitude dependence nor does it constituents, and the atmosphere is assumed to be in hydro- appear to follow the surface SO2 frost distribution as would be static equilibrium. Although our modeling is designed to rep- expected from a sublimation-driven atmosphere (e.g., Lellouch resent average atmospheric density conditions within a large- et al. 1992, 1994, 1996, Sartoretti et al. 1994, 1996, Ballester scale quasi-hydrostatic region surrounding an active Pele-type et al. 1994, Lellouch 1996, Trafton et al. 1996, Hendrix et al. eruption, we do not consider plume dynamics, and our “vol- 1999, McGrath et al. 2000). Second, the observed red shifts and canic” models are thus zeroth-order approximations to the actual the degree of broadening of the millimeter-wave SO2 and SO situation. lines suggest that energetic outgassing of gases from volcanic Dynamical studies by Ingersoll et al. (1985), Ingersoll (1989), plumes may be responsible for these aspects of the microwave Moreno et al. (1991), Wong and Johnson (1995, 1996), Wong emission (Lellouch et al. 1994; see also Ballester et al. 1994, and Smyth (2000), and Austin and Goldstein (1996, 2000), Strobel et al. 1994, Lellouch 1996). Third, visible-wavelength along with the observed heterogeneity of Io’s atmosphere (e.g., emissions from what is presumed to be neutral atomic oxygen, McGrath et al. 2000), illustrate that the ionian atmosphere is atomic sodium, and molecular SO2 and/or SO were observed on complex, highly dynamic, and not likely to be in hydrostatic Io when the satellite passed into Jupiter’s shadow (e.g., McEwen equilibrium. One-dimensional models will not be able to re- et al. 1998a, Geissler et al. 1999b, Bouchez et al. 2000; see also produce all the complexities. However, 1-D models are very the ultraviolet observations of Roesler et al. 1999). The corre- useful for first-order predictions of the relative abundances of lation of many of the emission features in these eclipse images different atmospheric species and for determining the impor- with known volcanic plumes or other centers of volcanic activity tance of photochemical processing of the volcanic gases. For the on Io suggests that volcanoes may be a major source of these first time ever, our photochemical models include sulfur com- neutral species (Geissler et al. 1999b). Finally, the H Lyman al- pounds such as S3,S4, and S2O that have been proposed as pha emission observed near high latitudes on Io might be caused surface coloring agents on Io (e.g., Nelson and Hapke 1978, by diffuse reflection of the solar Ly α line from Io’s surface when Hapke 1979, Smith et al.