acidification of europa's subsurface ocean

7/29/2019 Acidification of Europa's Subsurface Ocean

1/9

Research Article

Acidification of Europas Subsurface Oceanas a Consequence of Oxidant Delivery

Matthew A. Pasek1 and Richard Greenberg2

Abstract

Oxidants are formed at the surface of Europa and may be delivered to the subsurface ocean, possibly in greatquantities. Whether these substances would be available for biological metabolism is uncertain, because theymay react with sulfides and other compounds to generate sulfuric and other acids. If this process has beenactive on Europa for much of its age, then not only would it rob the ocean of life-supporting oxidants but the

subsurface ocean could have a pH of*

2.6, which is so acidic as to present an environmental challenge forlife, unless organisms consume or sequester the oxidants fast enough to ameliorate the acidification. KeyWords: EuropaOcean chemistryOxidantsSulfatePeroxideOxygenSulfuric acid. Astrobiology 12,xxxxxx.

1. Introduction

Strong oxidants, compounds capable of receiving elec-trons from other compounds, are rare in the Solar System

due to the abundance of chemical reductants such as H and C,which rapidly react with oxidants to form oxides that includewater and carbon dioxide. Whether a compound is an oxidant

is dependent on the electron availability of the local envi-ronment, and most oxidants are formed by non-equilibriumprocesses. Some Solar System oxidants are likely produced byhigh-energy events that occur in atmospheres, for instance,the production of trace O2 on Mars (McElroy et al., 1977).However, for the three large Solar System bodies that havethus far been found to be rich in strong oxidants, Earth, Eu-ropa, and Ganymede, production occurs at the surface. Thesource of oxidants on Earth is oxygenic photosynthesis,whereas for Europa and Ganymede it is irradiation of the icycrusts by high-energy particles.

Europas oxidants include a thin molecular oxygen (O2)atmosphere (Hall et al., 1995), hydrogen peroxide (H2O2) andO2 as constituents of the icy crust (Carlson et al., 1999), and

various oxidized sulfur species (SO2

4 compounds) on thesurface (McCord et al., 1998: Carlson et al., 1999). Both O2 andH2O2 probably formed by the radiolysis of ice at Europassurface (Carlson et al., 2009; Paranicas et al., 2009) and arestrong oxidants in most environments. Sulfates, as sulfatesalts (McCord et al., 1998, 2010) and sulfuric acid (Carlsonet al., 1999, 2002), are distinctive components of Europassurface. Also present are weak oxidants such as SO2, S, and

CO2. Possible sources of sulfur compounds are substancesdelivered from Io or endogenic material, the latter of which issuggested by association with surface features that mayrepresent connections to the ocean. Although it is unclearwhether these sulfates are actually capable of oxidizing othercompounds on Europas surface or subsurface, they con-tribute to environmental conditions more oxidizing than

those common elsewhere in the Solar System.Surficial oxidants are likely transported to the subsurfaceocean, possibly in substantial quantities. There they en-counter reductants and modify the ocean chemistry. Here,we investigate those effects and show how reaction of theoxidants with sulfide may acidify Europas ocean, with po-tentially significant consequences for biological systems inthe ocean.

2. Fluxes of Oxidants and Reductants

to Europas Subsurface Ocean

The rate of delivery of oxidants to Europas ocean dependson the net rate of radiolytic production near the surface and

on geological processes in the crust that may then deliver theoxidants down through the ice to the ocean below. Handet al. (2007), who built on approaches by Chyba and Phillips(2001) and Cooper et al. (2001), estimated the amount ofoxygen and H2O2 that accumulates as radiolytic chemistryproceeds and impact gardening protects the products byburial within the top few meters of ice. The dominant oxi-dant is O2, with 310% H2O2 (Carlson et al., 1999). Hand et al.

1Department of Geology, University of South Florida, Tampa, Florida.2Department of Planetary Science, University of Arizona, Tucson, Arizona.

ASTROBIOLOGYVolume 12, Number 2, 2012 Mary Ann Liebert, Inc.DOI: 10.1089/ast.2011.0666

1


2/9

(2007) then assumed, in effect, that after the quantities in-creased over a certain characteristic period of time, the oxi-dants would be delivered to the ocean, after which the nearsurface would begin to be reloaded for the next delivery.Hand et al. suggested that a plausible delivery timescale isthe apparent age of the surface, *50 million years, a figurebased on constraints from the small number of craters onEuropa (Zahnle et al., 2003). However, the process of delivery

and, hence, the context of this timescale was not defined.With this assumption, they calculated that an average rate of*4 109 mol/yr of O2 and H2O2 would be delivered into theocean.

Hand et al. noted that shorter delivery intervals wouldresult in a greater average delivery rate; longer intervalswould yield slower delivery into the ocean. They also notedthat reaction with endogenic reductants would consume theoxidants, a process we consider in detail in this paper. Theseendogenic reductants are generated by fluid-rock interac-tions and include CH4, H2, H2S, and Fe

2+ . With productionrates roughly estimated as 1.5 108, 3.6 108, 1.0109, and9107 mol/yr, respectively (McCollom, 1999; Hand et al.,2007), Hand et al. calculated that a substantial fraction of

oxidants would remain in the ocean even after the reductantsare consumed. These calculations were expanded by Handet al. (2009) to include oxidized, reduced, and biologi-cal ocean compositions, as well as the effect of exogenousorganics.

The oxidant delivery process assumed by Hand et al.(2007) may be an oversimplification for various reasons. Forexample, the timescale for erasing craters may not be thesame as the oxidant delivery interval. Moreover, the geo-logical processes on Europa that modify the surface probablydo it on a continual basis, whereas the scenario by Hand et al.(2007) implicitly involves a gradual surface buildup of oxi-dants punctuated by infrequent sudden delivery to theocean, which seems unlikely. Without the substantial rates of

delivery assumed by Hand et al., oxidants that reach theocean would be quickly consumed by the endogenic reduc-tants as well as by accretionary organics (Zolotov and Shock,2004). Moreover, Zolotov and Shock suggested that O2 andH2O2 would escape from the ocean during melt-throughevents (Greenberg et al., 1999) that expose the ocean to thesurface.

More recently, Greenberg (2010) considered the variousgeological processes that might be operating to bury near-surface oxidants and work them down to the ocean, in-cluding ridge formation and surface burial. That analysissuggested that the oxidant-rich layer near the top of the icecrust gradually thickened and occupied an ever-greaterfraction of the crust until by *2 Gyr it occupied the entire

crust. Up to that point, little oxygen reached the ocean.Thereafter, O2 and H2O2 (mostly O2) were delivered at a rateof*13 1011 mol/yr, assuming the same radiolytic pro-duction rate of Hand et al. (2007). While the actual deliveryrate remains uncertain, the plausibility of such high ratesmotivates consideration of the consequences for the chem-istry of the ocean and for any life there.

3. Modeling the Ocean Chemistry

Oxidants that enter the subsurface ocean would be ex-pected to affect its pH. Zolotov and Kargel (2009) showed

that, with only a small amount of oxidant delivery, the pHwould be high. But with the substantial delivery flux thatmay be possible, oxidants that react with the endogenic re-ductants, especially sulfide or organic compounds, couldacidify the ocean.

Two reactions of special note include the oxidation ofmethane

CH4 2O2/

H2CO3H2O (1)

and the oxidation of hydrogen sulfide

H2S 2O2/SO2

4 2H (2)

The latter reaction is known to be important in acidificationof terrestrial geological environments. Experiments by Mill-ero et al. (1987) and by Jennings et al. (2000) showed that thereaction of O2 and H2O2 with H2S and sulfide minerals pyriteor pyrrhotite decreases the pH of solutions to values between2 and 3, and that these reactions occur rapidly [ t1/2*10100h for sub-mM concentrations of H2O2 or O2 and H2S(Satterfield et al., 1954; Zhang and Millero, 1993)]. In those

experiments, approximately two H

+

ions were formed forevery sulfur oxidized.Here, we evaluate the generation of acid by reaction of

oxidants with a wide range of plausibly relevant reductants,and we quantify the accompanying acidification of Europassubsurface ocean by calculating the equilibrium chemistry.Computations were carried out with the program HSC(version 7.0, Outokompu Research Oy)1. This code uses theGIBBS energy solver (White et al., 1958) to determine equi-librium concentrations and has been used previously toconstrain sulfur chemistry in the Solar System (Pasek et al.,2005). The behavior of aqueous species in water is approxi-mated by using the HSC Chem AQUA module that appliesthe Davies model (extended Debye-Huckel), the semi-

empirical Pitzer model (with binary interactions only), andHarvies modification of the Pitzer model (binary and ter-nary parameters). The code allows for the injection or re-moval of species and computes the resulting changes insolution chemistry with respect to time. Reaction kinetics areignored in this study, because the reactions of major note(oxidation of reduced S species) are rapid and thus quicklyreach equilibrium even at 273 K (Millero et al., 1987; Zhangand Millero, 1993; Jennings et al., 2000).

In our first model system, the oxidant flux into the oceanreacts with accumulating endogenic reductants dissolved inwater, but we assume negligible chemical reaction with theunderlying rock itself. Similarly, Kargel et al. (2000) consid-ered a scenario in which the interaction with rock was

minimal. As discussed in Section 5.1 below, it is quite plau-sible that physical conditions minimize the chemical inter-face at the ocean-rock boundary. Even if the rock is able toreact freely with ocean water, this first type of model systemmight apply in the upper layer of the ocean (Section 5.1).

In this system, the initial abundances of H and O were setequal to the mass of Europas ocean, which we set to 3 1021

kg and corresponds to a 100 km ocean depth (Anderson et al.,

1The HSC chemical code stands for enthalpy, entropy, and heatcapacity. More details on the code can be found at www.hsc-chemistry.net.

2 PASEK AND GREENBERG


3/9

1997). The salinity was set equivalent to the salinity ofEuropan Seawater in McCollom (1999) for Na, K, Cl, andMg. In this system, the dominant anion is Cl-. The abun-dances of C, S, Fe, H, and O were then modified to accountfor the flux of substances that enter the ocean. First, reduc-tants from the interior (H2, CH4, H2S, and Fe

2 + ) were addedaccording to rates from McCollom (1999) and Hand et al.(2007). Then, after 2 billion years, oxidants from the crust (O2

and H2O2) were added, according to the timing and ratecalculated by Greenberg (2010), which were based on theradiolytic production model of Hand et al. (2007) and on anassessment of the dominant geological processes. We alsotested cases with much slower delivery of oxidants into theocean, but always commencing 2 billion years after the startof the endogenic reductant flow. These calculations wereperformed at 0C and 120 bar (equivalent to 10 km of iceload) and solved for H + concentration with respect to water.Changing the pressure of the system to larger pressures (12kbar) was not found to change the results appreciably. Si-milarly, the solution chemistry is insensitive to the volume ofthe ocean (e.g., reducing the ocean depth from 100 to 50 kmdecreased the pH at 4.5 Gyr by only 0.04 units).

Our second type of model system includes chemical re-action with the rock on the seafloor. The rock would be ex-pected to neutralize the acid produced. Our computationalexperiments quantified this effect for various quantities ofreactive rock and showed the chemistry of the ocean aftersuch reactions. We considered cases with basaltic and withultramafic rock, using compositions given by Turekian andWedepohl (1961). For each rock type, cases were examinedwith reactive rock volumes set equal to the upper 1 m, 1 km,and 25 km of sub-oceanic rock, for a total of six model sys-tems. The interaction depth of 25 km corresponds to themaximum depth of porosity in the rock, according to Vanceand Goodman (2009). For these experiments, we exposed tothe rock seawater in which new compounds had first been

generated from the reaction of oxidants with reductants. Infact, as shown in Section 4, for the rock-free cases, the pHreached a near steady-state very soon after the oxidants be-gan to arrive in the already reductant-rich ocean. For ourcases with the reactive rock component, we started with theocean composition in this steady-state obtained from therock-free cases after 4.5 billion years (i.e., 2.5 billion yearsafter the oxidant flux begins entering the ocean). This choiceof initial ocean composition is conservative in the sense thatit maximizes the role of the rock exposure, because the totalquantity of oxidants is limited to the amount delivered overthe last 2.5 billion years, as opposed to 4.5 billion years ofoxidant flux. Note too that the final ocean composition andpH at 4.5 Gyr would be the same if the rock were introduced

earlier, because the total elemental budget at 4.5 Gyr wouldbe unchanged. The calculations were performed at 2 kbar(expected under 150 km of water) for a range of tempera-tures from 0C to 250C.

For all the cases considered here, once the abundances of theelements were supplied, the HSC program then calculated thedistribution of elements among the various chemical species(listed below) by solving the minimum thermodynamic energystate coupled to mass balance calculations. With the code, wetracked the quantities of the following compounds, which rep-resent contributions from basaltic or ultramafic silicates, as wellas from a suite of sulfate minerals that may form during oxi-

dation of sulfides and iron sulfide minerals, as well as ironoxides, phyllosilicates, and carbonates: Al2O3, Al2SiO5 (anda-lusite), Al2Si2O5(OH)4 (kaolinite), CaAl2Si2O8, CaCO3, CaMg-Si2O6, CaSO42H2O, Fe, Fe(OH)2, Fe(OH)3, Fe0.877S, Fe0.945O,Fe0.947O, Fe2(SO4)3, Fe2O3, Fe2O33H2O, Fe2O3$H2O, Fe3O4,FeCO3, FeO, FeS, FeS2, FeSiO3, FeSO4$4H2O, FeSO4$7H2O,FeSO4H2O, KAlSi3O8, MgCO3, MgO, Mg(OH)2, MgSO4H2O, MgSO42H2O, MgSO44H2O, MgSO45H2O, MgSO4

6H2O, MgSO47H2O, MgSiO3, Mg3Si4O10(OH)2 (talc),Mg7Si8O16H2 (anthophyllite), Mg3Si2O5(OH)4 (serpentine),Mg5Al2Si3O10(OH)8 (chlorite), Mg4Si6O21H12 (sepiolite), NaAl-Si3O8, NaCl, NaAl2(AlSi3O10)(OH)2 (paragonite), Ca2FeAl2Si3-O12OH (epidote), Ca2Mg5Si8O16H (tremolite), CaAl2Si3O10-(OH)2 (prehnite), Mg7Si8O22(OH)2 (cummingtonite), Mg3Al2-Si3O12 (pyrope), Fe3Al2Si3O12 (almandine), Ca3Al2Si3O12 (gros-sular), Ca3Fe2Si3O12 (andradite), S, and SiO2 (quartz).

We also considered the following aqueous species in orderto model the complex chemistry of iron in aqueous solution,mineral dissolution, and acid-base chemistry that may resultfrom oxidation of endogenic compounds: Al3+ , CH4, CO, CO2,CO23 , Ca

2+ , Cl- , Fe2+ , Fe3+ , Fe(OH)2 , Fe(OH)2, Fe(OH)3,Fe(OH)3 , Fe(OH)

4 , Fe2(CO3)3, FeCO3, FeO+ , FeO, FeO2 ,

FeOH2+

, FeOH+

, FeSO

4 , H+

, H2, H2CO3, H2O, H2O2, H2S,H2SO3, H2SO4, H4SiO4, HCO

3 , HFeO2, HFeO

2 , HO

2 , HS- ,

HS2 , HSO

3 , HSO

4 , HSO

5 , K+ , Mg2+ , Na+ , O - , O 22 , O2,

O2 , O3, OH, OH- , S2- , S 22 , SO2, SO

23 , SO3, and SO

24 .

4. Results

4.1. Cases without direct interaction with rock

In the absence of reactive rock, if the quantity of oxidantsdelivered to Europas ocean becomes stoichiometricallyequivalent to or exceeds the quantity of reductants, then theocean becomes strongly acidic. Given the rate of delivery ofreductants assumed in our calculations [based on McCollom(1999)], the early europan ocean would have been chemically

reduced. Only when the oxidants begin to flow after the 2-billion-year delay [based on the timing and flux estimate ofGreenberg (2010)] would the ocean begin to become oxi-dized. We found that the pH of the ocean drops below 3 inonly 30 million years once the oxidants start to arrive. ThepH then decreases slowly to 2.6 after another 2.5 billion years(i.e., at 4.5 Gyr from the start).

As shown in Fig. 1, similar low pH values would bereached even with much lower oxidant fluxes. In all thesecases, most of the drop in pH occurred shortly after the de-livery of oxidants begins (Fig. 2). The subsequent decreasewas slow, because no sulfide remained after the initial drop;it was all consumed to form sulfate and acid. Oxidants wouldstill be left over despite reaction with the reductants, so the

ocean would remain strongly oxidizing with free O2 availablefor other reactions, including metabolic processes. As long asthe quantity of oxidants in this scenario is greater than thequantity of reductants, the final pH is fairly independent ofhow much greater it is; the pH will always drop to near 2.6.

Because the low pH results from the complete oxidation ofsulfide, changes to this value are possible only with changesto the fluxes of reductants. The key reductant that leads tothe acidification of Europas ocean is sulfide as H2S. If theflux of H2S has changed significantly over time, or if the fluxis different from the estimates by McCollom (1999) and byHand et al. (2007), then the resulting pH of Europas ocean

EUROPAS ACIDIC OCEAN 3


4/9

could be affected. For example, an increase in the flux of H2Sby a factor of 100 to 1011 mol/yr yields a pH as low as 1.1.Changing the CH4 flux would also change the resulting pHslightly as a result of formation of H2CO3. However, H2Soxidation is the most important reaction in acidification,unless the total H2S flux is low, in which case the oxidationof CH4 can be important. The dependence of the final pH onthe H2S flux is shown in Fig. 3.

In all the cases described above, iron speciates as hematite(Fe2O3); carbon is a mixture of dissolved CO2 gas andH2CO3; sulfur is principally in sulfate (10:1 SO

2

4: HSO

4);

and Mg, Ca, K, and Na are dissolved ions.

4.2. Reaction with rock

The pH values resulting for the models that include rockinteractions are summarized in Table 1. If the volume ofinteracting rock is equivalent to the upper 1 m, the rock haslittle effect on acidification, independent of composition ortemperature; the pH becomes 2.6, just as for the case with norock. However, if rock equivalent to a thickness of a kilo-meter or more is involved, then dissolution of olivine canneutralize the acid:

Mg2SiO4 4H/2Mg2 H4SiO4 (3)

For example, with ultramafic rock composition and a lowtemperature (0C), the pH can become very high if kilome-ters of rock participate, although the effect is less pronouncedat higher temperatures. If the rock is basaltic, then the degreeof buffering of acid is decreased because basaltic rock isricher in Al and Si, which have poorer potential for neu-tralization of H2SO4 than Mg, Na, K, and Ca oxides. More-over, in the basalt model the ocean is already close to

saturation in Mg and Ca, which further limits neutralizationby preventing Reaction 3 from reaching completion.

In these calculations, basalt rock alters to prehnite, sepiolite,and paragonite, whereas ultramafic rocks alter to prehnite andbrucite. All iron is present as hematite in both systems. Nosulfate minerals are produced in these calculations.

5. Discussion

If the flux of oxidants from the crust exceeds the flux ofreductants from the interior (as seems plausible), the pH ofEuropas subsurface ocean is generally decreased significantly.

FIG. 1. pH as a function of O2 flux and H2O2 flux, after the flux of reductants has continued for 4.5 billion years. These valuesare fairly constant from shortly after the oxidants begin to enter the ocean, which is taken for these calculations to commence 2billion years after the reductant flux begins. Except for very small amounts of both H2O2 and O2, pH values are generally

significantly less than 3. For the maximum flux rates from Greenberg (2010), shown by the black dot, the resulting pH is 2.6.



5/9

The decrease in pH is primarily due to the formation of acidduring the oxidation of sulfide and is directly related to theflux of sulfide in the oceans. The acidification would have

occurred soon after the oxidants began to enter the ocean,even if the sulfide (and other endogenic reductants) hadbeen accumulating for 2 billion years beforehand. Evenwith an oxidant delivery rate much lower than that esti-mated by Greenberg (2010), the pH would have decreased

to near its low steady-state value in less than a 100 millionyears (Fig. 2). An exception to these conclusions would be ifthe ocean were in chemical equilibrium with a large volume

of ultramafic rock, which could neutralize the acid. We nextdiscuss why that condition seems unlikely to be relevant forEuropa. We then discuss the implications of our model as tothe chemical composition of the ocean and for the possi-bility of life there.

FIG. 2. The change in pH of Europas ocean over time. The pH drops quickly once oxidant delivery begins at 2 Gyr. Valuesat 4.5 Gyr are also shown in Fig. 1. Three cases are shown here: 100%, 10%, and 1% of the flux estimated by Greenberg (2010).In the latter case (dashed line), the changes in slope after the commencement of oxygenation are a result of buffering bycarbonate and other species.

FIG. 3. The relationship between the flux of H2Sand the resulting pH. Here, the flux of oxidants isset equal to Greenbergs (2010) maximum esti-mates, but the result would be similar for sloweroxidant delivery as long as the rate were adequateto consume the endogenic reductants.



6/9

5.1. Acid neutralization by rock

Neutralization of the acidic europan ocean by rock wouldprimarily occur through dissolution of silicates to release Na,K, Ca, and Mg. Neutralization of the ocean will occur only ifthe rock at the bottom can provide these elements in suffi-cient quantity. Kargel et al. (2000) proposed a scenario in

which the ocean bottom consists of non-reactive sulfates;hence, this system would not neutralize acid produced byreaction of sulfides with oxidants (though whether suchchemistry would arise in this ocean is another question). Ourresults indicate that basalts are similarly ineffective, unless ahuge volume, representing a layer tens of kilometers thick, isintimately reactive with the ocean. The basalts are relativelyineffective because they contain substantial amounts of SiO2,which does not neutralize acid. Ultramafic rocks could bemore effective, but the rocks must be reactive to a depth of*1 km, and temperatures should be less than 200C to keepneutralization efficient.

Although ocean water may penetrate kilometers deep intothe rock layer (Vance and Goodman, 2009), it is unlikely that

more than a small fraction of the rock in that layer is able tointeract with the ocean. For a substantial portion of the porouslayer of rock to be chemically exposed to seawater, the rockwould need to be porous on a very fine scale. For example, ifthe porous layer extended 25 km into the rock, it is unlikelythat more than 1 in 25,000 atoms in that layer (equivalent toour 1-meter-thick-rock case, which was inadequate for neu-tralization) would be exposed to the ocean intimately enoughto be in chemical equilibrium. In principle, more atoms couldbe exposed, but only with very small capillaries to carry thewater, in which case the flow could hardly be adequate tokeep the rock in equilibrium with the ocean. Moreover, itseems unlikely that the rock involved would be predomi-nantly ultramafic, which (according to Table 1) further re-

duces the likelihood that the ocean has been neutralized.Even so, it is conceivable that enough rock could be en-

gaged to neutralize some ocean water just above the rocksurface. More likely, vertical transport would keep the oceanwater mixed by convection or thermal plumes over hot spots,in which case the neutralizing effect of the rock would bediluted in the large volume of the ocean. As we have shown,the entire ocean could only be neutralized if a huge volume ofthe rock were in intimate chemical equilibrium with the ocean.

Thus, an important issue is the rate and extent of verticalmixing. With vertical mixing, it is unlikely that enough rockcan participate to neutralize the acidic ocean water. Such

neutralization would require interaction with a huge amountof rock material, even if it were ultramafic. On the otherhand, if vertical mixing is slow enough, or if it does notextend all the way up to the ice or down to the rock, theupper ocean would be acidic (as calculated in Section 4.1) atthe same time as the deepest part has a higher pH (as cal-culated in Section 4.2).

The transport timescale due to chemical diffusion of

components through the depth of the ocean is*

Z2

/k, whereZ is the depth of the europan ocean (from the water-ice to thewater rock boundary) and k is the vertical diffusion coeffi-cient. For a depth of 100 km, and a vertical diffusion coeffi-cient of 10 - 5 m2/s, set equivalent to Earths (Ledwell et al.,1993), the mixing timescale is hence of the order of*107

years. This timescale suggests that chemical diffusion alonewould be slow enough to allow a significant difference in pHbetween the upper ocean and the bottom of the ocean.However, vertical transport of components by heat plumesor convection in the ocean may mix the ocean on shortertimescales. Acid neutralization by rock will only be signifi-cant if the lower portion of the ocean is not involved in suchmixing, so that the neutralizing species are not overly diluted

or, if ultramafic rock is involved, in very large quantities.

5.2. Implications for ocean composition

In the primordial jovian nebular region, sulfur likelyaccreted into the jovian moons primarily as FeS and H2S(Pasek et al., 2005). Sulfates have been observed on Europassurface and are attributed to radiolysis that involved sub-stances delivered from Io (Carlson et al., 2002, 2005, 2009)and to direct delivery from the subsurface ocean (Zolotovand Shock, 2001; McCord et al., 2010). If the sulfates have anoceanic origin, they may have been a leachate of the chon-dritic accretionary material of Europa (Kargel et al., 2000;Fanale et al., 2001) or, more likely, products of reactions of

sulfides with oxidants (e.g., McKinnon and Zolensky, 2003).Estimates of the oceanic composition of Europa have been

based on extracts of carbonaceous chondritic meteorites.These estimates suggest a salty ocean initially comprisingoxidized components, including sulfate, which is a majoranion in carbonaceous chondrite extracts (Fanale et al., 2001).However, the sulfate extracted from carbonaceous chondritesmay have been formed by weathering on Earths surface inthe years following their fall. Sulfate veins bear O isotopecharacteristics indicative of terrestrial oxygen, as does someof the water-soluble sulfate (Gounelle and Zolensky, 2001;Airieau et al., 2005). Sulfates are not found in more pristinechondrites, such as those of Tagish Lake (McKinnon andZolensky, 2003). In most meteorites, sulfur is in sulfide min-

erals such as troilite, FeS, and pyrrhotite, Fe1-xS. Additionally,sulfur chemistry in the nebula near the location of Jupiteraccretion should have been dominated by reducing speciessuch as H2S (Grossman, 1972; Pasek et al., 2005). This suggeststhat sulfur around Jupiter was initially in a reduced state andpresent as sulfides. Moreover, sulfides are the dominantconstituent from black smoker deep sea vents on Earth(Von Damm, 1990) and are predicted to dominate hydro-thermal vent fluids on Europa (McCollom, 1999).

Our analysis suggests that sulfate on Europa may be anoxidation product of sulfide in Europas subsurface ocean,which is in agreement with Hand et al. (2007). So long as the

Table 1. Values of Oceanic pH for 12 Cases withReactive Rock: Two Rock Types, Two Temperatures,

and Three Assumed Volumes of Reactive Rock

Depth of reactive rock layer

Composition Temperature (C) 1 m 1 km 25 km

Ultramafic 0 2.6 9.96 10

250 2.6 4.8 5Basaltic 0 2.6 3.1 4.6250 2.6 3.4 3.4

Quantities of reductants are based on 4.5 billion years of deliveryat a constant rate; quantities of oxidants are based on 2.5 billion yearsof delivery at a constant rate (after the 2-billion-year delay).



7/9

flux of oxidants is adequate, as seems plausible, then allsulfide will be consumed rapidly (on the timescale of mixingin the subsurface ocean) to form sulfate. If a flux of 109 mol/yr of endogenic sulfide enters Europas ocean (as estimatedby McCollom, 1999 and Hand et al., 2007), then, over the ageof Europa, approximately 5 1018 mol of sulfate will haveformed from this reaction. Consequently, approximately 1019

mol of H + would also form during this oxidation process,

yielding the pH of the europan ocean of*

2.6. At a pH of 2.6,the SO 24 to HSO

4 ratio is approximately 10:1 (or less withhigher H2S fluxes); hence, sulfate salts from the subsurfaceocean should be a mixture of compounds with SO 24 andHSO4 anions. If these salts reach the surface of Europa, forexample through cracks or melt-through sites in the ice crust,then the radiolysis of a HSO4 salt to H2SO4 would be ex-pected (Carlson et al., 2005). The best fits to Europas hydrateappear to be mixtures of Na sulfates, Mg sulfates, and Hsulfates (Carlson et al., 2005; Orlando et al., 2005), though aprecise identification is not presently possible. Both sulfuricacid hydrate and sulfate salts are plausible products on Eu-ropas surface via reaction of O2 with sulfide in the subsur-face ocean, even without a contribution from Io.

The possibility of an acidic ocean was previously raised byKargel et al. (2000) as one of many possible scenarios for Eu-ropas subsurface ocean system. The acidification in this sys-tem was attributed to reaction of SO2likely delivered by Io toEuropas surfacewith water to form H2SO4, and required anon-reacting rock wall surface. Other scenarios suggest littleacidity for Europas subsurface ocean, especially if the oxidantflux to the ocean is limited, and the sub-oceanic rock has acomposition equivalent to carbonaceous chondrite material(e.g., Zolotov and Kargel, 2009). Many models of Europassubsurface ocean predict saturation with MgSO4 (McCordet al., 1999; Kargel et al., 2000; Hand and Chyba, 2007), based inpart on carbonaceous chondrite extracts. Hydrothermal reac-tions of acid in the ocean with rock may also dissolve Mg

silicates, which could be placed on the surface as Mg-sulfatesas Mg- and sulfate-rich water freezes. However, our modeldoes not predict saturation of an MgSO4 mineral in the ocean,as there is too little sulfur present from hydrothermal sourcesto saturate the ocean. If the amount of sulfur is changed sub-stantially, as indicated in Fig. 2, then the water may becomeincreasingly concentrated in sulfate, though saturation requiresabout 103 times the current estimated flux of sulfur.

5.3. Conditions for life

The implications of an oxygenated ocean have been dis-cussed by Hand et al. (2007) and by Greenberg (2010). Therates of delivery of oxidants into the ocean could be adequate

to support significant biota, with concentrations adequate tosupport large eukaryotes, even macrofauna. However, asshown here, the oxygenation of the ocean would likely beaccompanied by acidification. Eukaryotes, especially mac-rofauna reliant on biomineralization for structure, farepoorly in acidic water. Hence, if the acidification occurred,life in Europas subsurface ocean would likely consist pri-marily of organisms analogous to one-celled bacteria- orarchaea-like cells, unless more complex life-forms evolvedspecifically for this environment.

If the flux of oxidants to Europas subsurface ocean wasdelayed by 13 billion years in accord with the analysis by

Greenberg (2010), then prebiotic chemistry may have had am-ple time to generate organisms with genetic systems and pro-tective structures (e.g., cells). An active anaerobic ecosystemcould have developed prior to an increase in oxidant flux. Then,as the oxidant flux commenced, organisms would have hadstructures in place to protect them from damage due to oxi-dation, which would have given them a chance to evolve high-efficiency oxygen-consuming metabolisms (as happened on

Earth as oxygen built up in the atmosphere and ocean). How-ever, as the acidification process discussed here proceeded, thepH could have become so low that the continued existence oflife, especially of any macroorganisms, would be problematic.

On the other hand, if organisms developed oxygen- or acid-consuming metabolisms or chemically changed the environ-ment more quickly than the acidification occurred, then theymight have minimized the deleterious effects. For example,just as terrestrial microorganisms can cause mineral weather-ing (Uroz et al., 2009), a similar effect by europan organismscould enhance acid neutralization by the seafloor rock. Also, iforganisms could have metabolized the oxygen quickly enoughas it entered the ocean, they might have delayed or preventedthe acidification, but only if enough oxygen atoms remained

sequestered within the bodies or remains of the organisms.The ecosystem would need to evolve quickly to meet this

challenge. According to Fig. 2, even if the oxidant flux were10% of Greenbergs estimate, oxygen metabolisms and acidtolerance would need to evolve in *50 million years. It maybe that at first only organisms close to the ocean bottomwould be sufficiently protected from the acidic, oxic envi-ronment by the neutralizing effect of the rock. The ecosystemmight then gradually evolve so that organisms could exploitthe oxic environment and stop the acidification.

5.4. Biominerals

Formation of biominerals may be an important step in thediversification of macrocellular life (e.g., Ward and Brownlee,2000). The carbonate minerals that form shells of mollusksand other invertebrates on Earth are highly soluble in waterwith a pH of 2.6 [up to concentrations in excess of 1 mol/Lwater (hereafter M) of Ca2+ or carbonate]. Similarly, calciumphosphate minerals, such as those that form vertebrateskeletons, are also highly soluble (up to concentrations inexcess of 0.1 M of phosphate). However, if acidificationproceeded in the europan ocean, then production of two ofthe dominant biominerals (apatite and calcite/aragonite)used by modern terrestrial life would not be feasible. Alter-native biominerals are possible. For example, silica is an al-ternative used principally by terrestrial algal forms, andsiliceous biominerals are not affected by acidic solutions.However, they can only form in water saturated in Si (ap-proximately 1 mM concentration of Si as silicic acid). Giventhe importance of phosphorus to life (e.g., Pasek, 2008), analternative to calcium phosphate biominerals might be ironphosphate biominerals, as iron phosphate is poorly solublein acidic solutions (phosphate solubility of less than 10 -9M);hence, it is significantly more stable than calcium phosphatesunder these conditions.

6. Conclusion

Acidification of Europas ocean likely commenced as soonas oxidants began to be delivered to the ocean, especially in



8/9

the upper regions of the ocean. This effect may have beenprevented if organisms evolved a process to utilize andconsume oxidants and the ecosystem and rock-bufferingmodified the acidity over a relatively short timescale.Otherwise, the oxidant flux could have quickly over-whelmed the accumulated reductants and generated signif-icant acid during the process. Life at the bottom of the oceanwould have had a better chance to be protected from this

acidification than life higher up in the ocean, closer to thewater-ice boundary.

Unless europan ocean life evolved an effective biominer-alization process and the ability to consume and sequesterarriving oxidants, any organisms would have to tolerate ahigh quantity of oxidants and low pH. In that case, a sur-viving ecosystem in Europas ocean might be analogous tomicrobial communities that live in acid mine drainage (e.g.,Edwards et al., 2000). Thus, studies of organisms at Ro Tintomay be especially pertinent to putative europan life (e.g.,Fernandez-Remolar et al., 2005; Amils et al., 2007; Davila et al.,2008). These communities are dominated by acidophiles thatoxidize iron and sulfide as sources of metabolic energy. In-deed, in these environments, Fe3+ can be a more efficient

oxidant than even O2; hence, these microbes may promote theactive acidification of their own environment. If the dominantreductant in Europas ocean is indeed sulfide, then such mayhave been the fate of Europas ecosystem as well. It appearsthat only if life were able to evolve quickly enough as theoxidants arrived in the ocean could it have prevented devel-opment of such a hostile environment or come to live with it.

Acknowledgments

The authors thank Virginia Pasek for helpful edits andTom McCollom for helpful discussion. Insightful reviewsfrom Chris McKay and other, anonymous reviewers wereespecially helpful. M.A.P. and R.G. were supported by

grants from NASAs Exobiology and Evolutionary Biologyprogram (NNX10AT30G) and Planetary Protection program,respectively.

Disclosure Statement

The authors declare no competing interests.

References

Airieau, S.A., Farquhar, J., Thiemens, M.H., Leshin, L.A., Bao,H., and Young, E. (2005) Planetesimal sulfate and aqueousalteration in CM and CI carbonaceous chondrites. GeochimCosmochim Acta 69:41674172.

Amils, R., Gonzalez-Toril, E., Fernandez-Remolar, D., Gomez, F.,

Aguilera, A., Rodriguez, N., Malki, M., Garcia-Moyano, A.,Fairen, A.G., de la Fuente, V., and Sanz, J.L. (2007) Extremeenvironments as Mars terrestrial analogs: the Ro Tinto case.Planet Space Sci 55:370381.

Anderson, J.D., Lau, E.L., Sjogren, W.L., Schubert, G., and MooreW.B. (1997) Europas differentiated internal structure: infer-ences from two Galileo encounters. Science 276:12361239.

Carlson, R.W., Johnson, R.E., and Anderson, M.S. (1999) Sulfuricacid on Europa and the radiolytic sulfur cycle. Science 286:9799.

Carlson, R.W., Anderson, M.S., Johnson, R.E., Schulman, M.B.,and Yavrouian, A.H. (2002) Sulfuric acid production on Eu-ropa: the radiolysis of sulfur in water ice. Icarus 157:456463.

Carlson, R.W., Anderson, M.S., Mehlman, R., and Johnson, R.E.(2005) Distribution of hydrate on Europa: further evidence forsulfuric acid hydrate. Icarus 177:461471.

Carlson, R.W., Calvin, W.M., Dalton, J.B., Hansen, G.B., Hudson,R.L., Johnson, R.E., McCord, T.B., and Moore, M.H. (2009)Europas surface composition. In Europa, edited by R.T. Pap-palardo, W.B. McKinnon, and K. Khurana, University of Ar-izona Press, Tucson, AZ, pp 283327.

Chyba, C.F. and Phillips, C.B. (2001) Possible ecosystems and thesearch for life on Europa. Proc Natl Acad Sci USA 98:801804.

Cooper, J.F., Johnson, R.E., Mauk, B.H., Garrett, H.B., andGehrels, N. (2001) Energetic ion and electron irradiation of theicy Galilean satellites. Icarus 149:133159.

Davila, A.F., Fairen, A.G., Gago-Duport, L., Stoker, C., Amils, R.,Bonaccorsi, R., Zavaleta, J., Lim, D., Schulze-Makuch, D., andMcKay, C.P. (2008) Subsurface formation of oxidants on Marsand implications for the preservation of organic biosignatures.Earth Planet Sci Lett 272:456463.

Edwards, K.J., Bond, P.L., Gihring, T.M., and Banfield, J.F. (2000)An archaeal iron-oxidizing extreme acidophile important inacid mine drainage. Science 287:17961799.

Fanale, F.P., Li, Y.-H., De Carlo, E., Farley, C., Sharma, S.K.,Horton, K., and Granahan, J.C. (2001) An experimental esti-

mate of Europas ocean composition independent of Galileoorbital remote sensing. J Geophys Res 106:1459514600.

Fernandez-Remolar, D.C., Morris, R.V., Gruener, J.E., Amils, R.,and Knoll, A.H. (2005). The Ro Tinto basin, Spain: mineral-ogy, sedimentary geobiology, and implications for interpre-tation of outcrop rocks at Meridiani Planum, Mars. EarthPlanet Sci Lett 240:149167.

Gounelle, M. and Zolensky, M.E. (2001) A terrestrial origin forsulfate veins in CI1 chondrites. Meteorit Planet Sci 36:13211329.

Greenberg, R. (2010) Transport rates of radiolytic substances intoEuropas ocean: implications for the potential origin andmaintenance of life. Astrobiology 10:275283.

Greenberg, R., Hoppa, G.V., Tufts, B.R., Geissler, P., Riley, J.,and Kadel, S. (1999) Chaos on Europa. Icarus 141:263286.

Grossman, L. (1972) Condensation in the primitive solar nebula.Geochim Cosmochim Acta 36:597619.

Hall, D.T., Strobel, D.F., Feldman, P.D., McGrath, M.A., andWeaver, H.A. (1995) Detection of an oxygen atmosphere on

Jupiters moon Europa. Nature 373:676679.Hand, K.P. and Chyba, C.F. (2007) Empirical constraints on the

salinity of the europan ocean and implications for a thin iceshell. Icarus 189:424438.

Hand, K.P., Carlson, R.W., and Chyba, C.F. (2007) Energy,chemical disequilibrium, and geological constraints on Eu-ropa. Astrobiology 7:10061022.

Hand, K.P., Chyba, C.F., Priscu, J.C., Carlson, R.W., and Neal-son, K.H. (2009) Astrobiology and the potential for life onEuropa In Europa, edited by R.T. Pappalardo, W.B. McKinnon,

and K. Khurana, University of Arizona Press, Tucson, AZ, pp589629.Jennings, S.R., Dollhopf, D.J., and Inskeep, W.P. (2000) Acid

production from sulfide minerals using hydrogen peroxideweathering. Appl Geochem 15:235243.

Kargel, J.S., Kaye, J.Z., Head, J.W., Marion, G.M., Sassen, R.,Crowley, J.K., Ballesteros, O.P., Grant. S.A., and Hogenboom,D.L. (2000) Europas crust and ocean: origin, composition, andthe prospects for life. Icarus 148:226265.

Ledwell, J.R., Watson, A.J., and Law, C. (1993) Evidence for slowmixing across the pycnocline from an open-ocean tracer-release experiment. Nature 364:701703.



9/9

McCollom, T.M. (1999) Methanogenesis as a potential source ofchemical energy for primary biomas production by autotro-phic organisms in hydrothermal systems on Europa. J GeophysRes 104:3072930742.

McCord, T.B., Hansen, G.B., Fanale, F.P., Carlson, R.W., Matson,D.L., Johnson, T.V., Smythe, W.D., Crowley, J.K., Martin, P.D.,Ocampo, A., Hibbitts, C.A., Granahan, J.C., and the NIMSteam. (1998) Salts on Europas surface detected by GalileosNear Infrared Mapping Spectrometer. Science 280:12421245.

McCord, T.B., Hansen, G.B., Matson, D.L., Johnson, T.V.,Crowley, J.K., Fanale, F.P., Carlson, R.W., Smythe, W.D.,Martin, P.D., Hibbitts, C.A., Granahan, J.C., and Ocampo, A.(1999) Hydrated salt minerals on Europas surface from theGalileo near-infrared mapping spectrometer (NIMS) investi-gation. J Geophys Res 104:1182711851.

McCord, T.B., Hansen, G.B., Combe, J.-P., and Hayne, P. (2010)Hydrated minerals on Europas suface: An improved lookfrom the Galileo NIMS investigation. Icarus 209:639650.

McElroy, M.B., Kong, T.Y., and Yung, Y.L. (1977) Photo-chemistry and evolution of Mars atmosphere: a Viking per-spective. J Geophys Res 82:43794388.

McKinnon, W.B. and Zolensky, M.E. (2003) Sulfate content ofEuropas ocean and shell: evolutionary considerations and

some geological and astrobiological implications. Astrobiology3:879897.

Millero, F.J., Hubinger, S., Fernandez, M., and Garnett, S. (1987)Oxidation of H2S in seawater as a function of temperature,pH, and ionic strength. Environ Sci Technol 21:439443.

Orlando, T.M., McCord, T.B., and Grieves, G.A. (2005) Thechemical nature of Europa surface material and the relation toa subsurface ocean. Icarus 177:528533.

Paranicas, C., Cooper, J.F., Garrett, H.B., Johnson, R.E., andSturner, S.J. (2009) Europas radiation environment and itseffects on the surface. In Europa, edited by R.T. Pappalardo,W.B. McKinnon, and K. Khurana, University of Arizona Press,Tucson, AZ, pp 529544.

Pasek, M.A. (2008) Rethinking early Earth phosphorus geo-chemistry. Proc Natl Acad Sci USA 105:853858.

Pasek, M.A., Milsom, J.A., Ciesla, F.J., Lauretta, D.S., Sharp,C.M., and Lunine, J.I. (2005) Sulfur chemistry with time-varying oxygen abundance during Solar System formation.Icarus 175:114.

Satterfield, C.N., Reid, R.C., and Briggs, D.R. (1954) Rate ofoxidation of hydrogen sulfide by hydrogen peroxide. J AmChem Soc 76:39223923.

Turekian, K.K. and Wedepohl, K.H. (1961) Distribution of theelements in some major units of the Earths crust. Geol Soc AmBull 72:175192.

Uroz, S., Calvaruso, C., Turpault, M-P., and Frey-Klett, P. (2009)Mineral weathering by bacteria: ecology, actors, and mecha-nisms. Trends Microbiol 17:378387.

Vance, S. and Goodman, J. (2009) Oceanography of an ice-covered moon. In Europa, edited by R.T. Pappalardo, W.B.McKinnon, and K. Khurana, University of Arizona Press,Tucson, AZ, pp 459484.

Von Damm, K.L. (1990) Seafloor hydrothermal activity: blacksmoker chemistry and chimneys. Annu Rev Earth Planet Sci18:173204.

Ward, P.D. and Brownlee, D. (2000) Rare Earth: Why Complex Lifeis Uncommon in the Universe, Springer Verlag, New York.

White, W.B., Johnson, S.M., and Dantzig, G.B. (1958) Chemicalequilibrium in complex mixtures. J Chem Phys 28:751755.

Zahnle, K., Schenk, P., Levison, H., and Dones, L. (2003) Cra-tering rates in the outer Solar System. Icarus 163:263289.

Zhang, J.-Z. and Millero, F.J. (1993) The products from the oxi-dation of H2S in seawater. Geochim Cosmochim Acta 57:17051718.

Zolotov, M.Y. and Kargel, J.S. (2009) On the chemical composition

of Europas icy shell, ocean, and underlying rocks. In Europa,edited by R.T. Pappalardo, W.B. McKinnon, and K. Khurana,University of Arizona Press, Tucson, AZ, pp 431458.

Zolotov, M.Y. and Shock, E.L. (2001) Composition and stabilityof salts on the surface of Europa and their oceanic origin. JGeophys Res 106:3281532827.

Zolotov,M.Y. and Shock, E.L.(2004)A model for low-temperaturebiogeochemistry of sulfur, carbon, and iron on Europa. J Geo-phys Res 109, doi:10.1029/2003JE002194.

Address correspondence to:Matthew A. Pasek

University of South FloridaDepartment of Geology

4202 E Fowler AveSCA 528

Tampa, FL 33620

E-mail: [email protected]

Submitted 8 April 2011Accepted 20 November 2011


acidification of europa's subsurface ocean

Documents