Oxidant Enhancement in Martian Dust Devils and Storms: Implications for Life and Habitability

ASTROBIOLOGYVolume 6, Number 3, 2006© Mary Ann Liebert, Inc.

Special Paper

Oxidant Enhancement in Martian Dust Devils andStorms: Storm Electric Fields and Electron

Dissociative Attachment

GREGORY T. DELORY,1 WILLIAM M. FARRELL,2 SUSHIL K. ATREYA,3NILTON O. RENNO,3 AH-SAN WONG,3 STEVEN A. CUMMER,4 DAVIS D. SENTMAN,5

JOHN R. MARSHALL,6 SCOT C.R. RAFKIN,7 and DAVID C. CATLING8,9

ABSTRACT

Laboratory studies, numerical simulations, and desert field tests indicate that aeolian dusttransport can generate atmospheric electricity via contact electrification or “triboelectricity.”In convective structures such as dust devils and dust storms, grain stratification leads to macro-scopic charge separations and gives rise to an overall electric dipole moment in the aeolianfeature, similar in nature to the dipolar electric field generated in terrestrial thunderstorms.Previous numerical simulations indicate that these storm electric fields on Mars can approachthe ambient breakdown field strength of �25 kV/m. In terrestrial dust phenomena, poten-tials ranging from �20 to 160 kV/m have been directly measured. The large electrostatic fieldspredicted in martian dust devils and storms can energize electrons in the low pressure mar-tian atmosphere to values exceeding the electron dissociative attachment energy of both CO2and H2O, which results in the formation of the new chemical products CO/O� and OH/H�,respectively. Using a collisional plasma physics model, we present calculations of the CO/O�

and OH/H� reaction and production rates. We demonstrate that these rates vary geometricallywith the ambient electric field, with substantial production of dissociative products whenfields approach the breakdown value of �25 kV/m. The dissociation of H2O into OH/H� pro-vides a key ingredient for the generation of oxidants; thus electrically charged dust may sig-nificantly impact the habitability of Mars. Key Words: Mars—Dust storm—Dust devil—Elec-tric field—Oxidant—Habitability. Astrobiology 6, 451–462.

451

1Space Sciences Laboratory, University of California, Berkeley, Berkeley, California.2Laboratory for Extraterrestrial Physics, NASA Goddard Space Flight Center, Greenbelt, Maryland.3Department of Atmospheric, Oceanic, and Space Sciences, University of Michigan, Ann Arbor, Michigan.4Department of Electrical and Computer Engineering, Duke University, Durham, North Carolina.5Geophysical Institute, University of Alaska, Fairbanks, Alaska.6SETI Institute, Mountain View, California.7Southwest Research Institute, Boulder, Colorado.8Department of Atmospheric Sciences/Astrobiology Program, University of Washington, Seattle, Washington.9Department of Earth Sciences, University of Bristol, Bristol, United Kingdom.

6278_06_p451-462 6/15/06 3:40 PM Page 451

INTRODUCTION

ACENTRAL QUESTION relevant to the habitabil-ity of Mars relates to the origin and presence

of oxidants in the atmosphere and soil (Klein,1998), which may have sterilized the surface andhence led to the failure of the Viking life sciencesexperiments to detect organics (Oyama et al.,1977). Photochemical processes have been in-voked to explain the presence of oxidants onMars, with hydrogen peroxide (H2O2) the mostlikely product (Krasnopolsky, 1993, 1995; Atreyaand Gu, 1994; Nair et al., 1994; Clancy and Nair,1996). While the recent discovery of H2O2 at 20–40parts per billion volume on Mars (Encrenaz et al.,2004) is consistent with production by photo-chemical processes in the atmosphere, the soil re-activity implied by the Viking results indicate lev-els ranging from at least 1 part per million (Zentand McKay, 1994) up to �250 parts per million(Mancinelli, 1989). The short lifetime of H2O2 inthe atmosphere relative to its rate of diffusion intothe soil makes the production of the inferred lev-els of oxidants difficult to explain from photo-chemical processes alone, and thus additionalsources should be explored. Mills (1977) was oneof the first to consider an alternative productionmechanism when he suggested that the electrifi-cation in dust storms may be an added physio-chemical energy source that has the ability to create a number of new species, including the ox-idant H2O2. Oyama and Berdahl (1979) describedthe possible creation of an oxygen plasma fromdust electrification on Mars, while Ballou et al.(1978) showed that oxygen plasmas can create ox-idants when exposed to basalts.

Here we investigate the implications of theubiquitous presence of strong electric fields onMars for atmospheric chemical processes relevantto oxidant production, which result from the ion-ization and dissociation of H2O by energized elec-trons in the martian atmosphere. Large electro-static fields created by dust devils and storms orother aeolian processes can lead to the creationand energization of electrons as an ambient corepopulation is accelerated and the ionization ofCO2 occurs. Using a detailed numerically basedplasma physics model, we calculate the electronenergy distribution for electric fields rangingfrom small values (�5 kV/m) to levels near thebreakdown potential (�25 kV/m) that would bepresent in a given dust storm event. We thenstudy the impact of this process on the local at-

mospheric chemistry with an emphasis on prod-ucts relevant to subsequent oxidant formation.Under these conditions we find that the dissoci-ation of H2O via electron collisions produces neg-ative ions at rates that vary strongly with the ap-plied electric field and become greater thanphotochemical rates by several orders of magni-tude. The generation of OH/H� will lead to thesubsequent production of H2O2 at rates greaterthan photochemical processes (Atreya et al.,2006). This process would be common in thelower atmosphere down to the surface and pre-dominant when high atmospheric dust loads at-tenuate photochemical processes. This ubiqui-tous, rapid source of oxidants would providestrong support for the interpretation of the Vikingresults as indicative of a chemically reactive, ox-idant-rich soil. This mechanism also has implica-tions for the lifetime and spatial distribution ofmethane, recently measured in the atmosphere ofMars (Formisano et al., 2004; Krasnopolsky et al.,2004; Mumma et al., 2004).

THE ELECTRIC DUST STORM

The dynamic martian atmosphere is character-ized by ubiquitous aeolian activity, with signifi-cant dust lofting and transport occurring over awide dynamic range of spatial and temporalscales. Global dust storms can envelope a signif-icant fraction of the planet, and appear to be sea-sonally dependent in that a large number of thesestorms occur between southern spring and sum-mer, around perihelion (Martin and Zurek, 1993).Over 780 local to regional (�102–106 km2) duststorms during the 1999 storm season have beencataloged by Cantor et al. (2001), who studied thegeographic and seasonal dependencies of theseevents. The most common dust activity on Marsoccurs at the smallest scale; warm-cored, convec-tive vortices such as dust devils are likely presentto varying extents during most seasons (Newmanet al., 2002; Fisher et al., 2005). Dust devils withdiameters between 100 m and 1 km, and heightsof up to 5–10 km, are frequently observed onMars (Thomas and Gierasch, 1985; Fisher et al.,2005). Based on their Mars Global Surveyor-ob-served tracks, Mars’ dust devils are found atnearly all locations on the planet, and because oftheir dynamic nature, they are continuously re-moving dust from the surface and maintainingthe bulk atmospheric dust opacity during non-

DELORY ET AL.452

6278_06_p451-462 6/15/06 3:40 PM Page 452

FIG. 1. a: Terrestrial electric field mea-surements of dust devils such as this eventnear Eloy, AZ. Maximum electric fields oc-cur near the core of the events, in manycases saturating the instrument with read-ings below �20 kV/m. b: Models for bothmacro- and microelectrification in a dustdevil. Larger dust storm electrification isenvisioned to occur in a similar manner.

b

a

6278_06_p451-462 6/15/06 3:40 PM Page 453

storm seasons (Ryan and Lucich, 1983; Smith andLemmon, 1999; Ferri et al., 2003).

Experimental and theoretical investigations offrictional charging mechanisms in both small-and large-scale meteorological phenomena sug-gest that Mars very likely possesses an electricallyactive atmosphere as a result of dust-liftingprocesses of all scales, including dust devils anddust storms. Naturally occurring dust activity isnearly always associated with significant electri-fication via the process of triboelectricity—thefrictional charging of dust grains in contact withone another or the surface as they are transportedby wind or convective circulations. In terrestrialsystems, early studies clearly demonstrated thepresence of electric fields in the kilovolt/meterrange within about 100 m of dust devils (Freier,1960; Crozier, 1964, 1970). More recently, electricfields ranging from �3 kV/m to greater than 20kV/m have been measured within dust devils, asshown in Fig. 1a (Farrell et al., 2003, 2004; Rennoet al., 2004). Surface processes can also generatesignificant electrification. In saltating sand, inwhich impacts from sand particles on approxi-mately centimeter-scale ballistic trajectories gen-erate lofted dust, fields in excess of 160 kV/mhave been measured within the first few cen-timeters of the surface (Schmidt et al., 1998). Thelargest terrestrial dust events, volcanic plumes,can generate lightning, and thus serve as ademonstration that electrified dust can indeedreach breakdown potentials (Anderson, 1965).Contact electrification can lead to differentialcharging of dust grains via a variety of mecha-nisms; in events with grains of similar composi-tion, smaller particles typically obtain a net neg-ative charge, while larger particles becomepositive (Ette, 1971; Melnik and Parrot, 1998; Far-rell et al., 2003; Renno et al., 2003). Thus a large-scale electric dipole moment can be generated bynearly any process with a vertical lifting compo-nent, as the smaller, negatively charged grains aretransported to higher altitudes than the heavier,positively charged grains. In dust devils and duststorms, the vertical stratification of grains basedon size and mass will create a stratification ofcharge, which creates an electric dipole momentwith a spatial scale on the order of the storm size(Fig. 1b). Based on the results of terrestrial ex-periments and their implications for the presenceof electrification processes on Mars, Melnick andParrot (1998) used a particle-in-cell numericalmodel to show that electric fields up to the break-

down potential of 25 kV/m can easily occur nearthe martian surface.

DUST STORM-DRIVEN ELECTRON ENERGIZATION

The martian troposphere may be considered asa tenuous electron plasma due to the presence ofa high-density neutral background dominated byelectron-neutral collisional processes. Based onthe penetration of ionizing radiation from cosmicrays and radioactive elements in the martiancrust, a representative core population electrondensity near the surface is ne � 5 � 106/m3

(Whitten et al., 1971). These electrons interact withthe ambient CO2 via a multitude of collisionalprocesses with cross sections that vary stronglywith energy. At low energies (�1 eV), elastic scat-tering and momentum transfer determine theelectron drift velocities through the medium. Inthe range of 1–10 eV, CO2 possesses large crosssections for vibrational and electronic excitationthrough electron collisions.

An important process in this energy range isdissociative attachment, in which an electron at-taches to a CO2 molecule, which then rapidly dis-sociates into a negative ion and a neutral mole-cule. Electron attachment processes are animportant part of the chemistry of the terrestrialatmosphere, where they are the dominant meansby which negative ions are produced from O2(Viggiano and Arnold, 1995). In dissociative at-tachment, the following reaction occurs:

e � AB � A� � B (1)

A second process, associative electron attach-ment, involves a third body:

e � AB � M � AB� � M (2)

In the lower terrestrial atmosphere, the three-body process dominates. However, at higher al-titudes and under an appreciable electric field,electron dissociative attachment is the dominatemechanism for free electron removal and nega-tive ion formation, particularly near dischargeevents in the vicinity of thunderstorm activity.

The behavior of tenuous electron plasmas inthe presence of a neutral CO2 background wasinvestigated numerically by Nighan (1970), whoconsidered a range of electron–CO2 collisional

DELORY ET AL.454

6278_06_p451-462 6/15/06 3:40 PM Page 454

processes in the presence of an electric field forparameters relevant to Mars. In this approach theelectron distribution is described by the Boltz-mann equation:

� ��v f(v�) � � � c(3)

where e is the electron charge, me is the electronmass, E� is the applied electric field, and f(v�) is theelectron distribution as a function of the three-di-

∂f(v�)

∂teE�me

mensional velocity vector v�. The subscript c onthe right-hand side of Eq. 3 denotes effects dueto electron collisions and is a sum over all rele-vant collisional cross sections describing elec-tron–CO2 interactions, including momentumtransfer, vibrational and electronic excitations,dissociative attachment, and impact ionization.Figure 2 shows graphically some of these inter-actions, i.e., electron impact ionization and disso-ciation of both CO2 and H2O, including the rele-vant electron energies. In Eq. 3, the acceleration

MARS DUST STORM CHARGING AND OXIDANTS 455

FIG. 2. Electron–CO2 and –H2O impact processes. Cross-section data shown for CO2 ionization and dissociationwere compiled by Itikawa (2002); H2O dissociation cross-section data were provided by Itikawa and Mason (2005).

6278_06_p451-462 6/15/06 3:40 PM Page 455

of the electrons via the electric force is offset byenergy-depleting interactions with the CO2 mol-ecule, which occur on spatial scales of the meanfree path; the combination of these effects can cre-ate a statistical electron energy distribution thatvaries substantially from a Maxwellian distribu-tion. The average electron energy increases withelectric field strength, which results in the devel-opment of a high-energy tail in the electron en-ergy distribution.

Numerical solutions to Eq. 3 can be foundbased on the Pitchford, O’Neil, and Rumble(POR) technique (Pitchford et al., 1981). For theelectron–CO2 interactions on the right-hand sideof Eq. 3, we consider the vibrational, excitational,attachment, and ionization processes shown inTable 1. The values of electric field E we use rangefrom �5 kV/m to near the theoretical breakdownpotential of �25 kV/m for Mars. Solutions to Eq.3 are greatly facilitated by the transformation tothe new variable u � mev2/2e, the electron energyin eV; we then define a new normalization for thedistribution function f(u) such that �u1/2f(u)du �1. Figure 3 shows the results of solving Eq. 3 forf(u), where the effect of an increasing electric fieldon the electron distribution is dramatic and pro-duces significant high-energy tails for fields be-tween 8 and 25 kV/m compared with the lowerfield cases. It is important to note that solutionsto Eq. 3 do not represent an actual large-scale dis-charge process (i.e., martian lightning); instead,we have estimated the energization of a thermalcore of electrons in the presence of a chargingfield E that interacts with the ambient CO2 at-mosphere prior to a catastrophic release of elec-tron current in a full breakdown scenario. By farthe most important implication of Eq. 3 is that thesolutions are remarkably non-Maxwellian. In

most cases, a Maxwellian distribution will not bea good approximation to the electron distributionfor the range of electric field values we considerhere (Nighan, 1970). The distribution f(u) is in factheavily modified by its interaction with CO2 andthus must be explicitly calculated based on therelevant physical parameters for Mars, which areprimarily governed by E/N, the ratio of the elec-tric field to the neutral density.

PRODUCTION OF NEGATIVE IONS AND NEUTRALS

Once produced, the enhanced populations ofelectrons at higher energies will then have amarkedly increased probability of impacts withother trace atmospheric constituents through avariety of interaction cross sections. While ion-ization processes are important, many additionalreactions will occur at energies less than the ion-ization potentials of ambient molecules, particu-larly in the 4–12 eV range where dissociative at-tachment processes come into play. Motivated bythe expected correlation between the presence ofH2O and H2O2 (Encrenaz et al., 2002), we now di-rect our attention to the impact of the electron dis-tribution f(u) on H2O. For the enhanced electronpopulation we see in f(u) above 4 eV, the break-down of H2O via dissociative electron attachmentwill be a significant possibility. These reactionsare outlined in Table 2. In this process, three neg-ative ion species are possible: H�, OH�, and O�.The cross section for the production of H� isgreater than those of the other negative ions byone or more orders of magnitude (Itikawa andMason, 2005) and will be the dominant productconsidered here. The production rate of a prod-

DELORY ET AL.456

TABLE 1. CO2–ELECTRON INTERACTIONS

Cross sections andProcess Mode energy range

CO2–electron Momentum transfer �4–180 � 10�16 cm2

collisions below 1 eVCO2 vibrational Symmetric, asymmetric, and bending modes �3 � 10�16 cm2, 0.1–20 eV

excitation (000 → 010,000 → 020 � 100,000 → 001, 0n0 → n00)CO2–electronic Various electronic transitions (3�

u, 1�u, 1�u . . .) 0.01–6 � 10�16 cm2,

excitation 0.1–100 eVCO2 → CO2

� Impact ionization �4 � 10�16 cm2, 13.7–100 eVCO2 → CO/O� Electron dissociative attachment 1.5 � 10�19 cm2 at 4.3 eV

dissociation 4.28 � 10�19 cm2 at 8.1 eV

After Itikawa (2002).

6278_06_p451-462 6/15/06 3:40 PM Page 456

uct n� from the electron distribution acting on amaterial of number density n is given by

� knne (4)

where k is the chemical rate constant and ne is theelectron density. The presence of energized electronsin the CO2 atmosphere will augment the productionrate of any species n� in two ways: first, via the in-teraction cross sections for specific reactions thatform the rate constant k, and also through increas-ing the degree of ionization of the atmosphere lead-ing to larger values of ne. Using the electron distri-butions obtained numerically, we can calculate thechemical rate constant based on the interaction crosssections for dissociative attachment interactions:

k � ��v� � � 1/2 �(u)uf(u)du (5)2e

me

dn�dt

where f(u) is the electron distribution function wehave determined using Eq. 3, u is the electron en-ergy in eV, and �(u) is the energy-dependentcross section for electron–H2O dissociative at-tachment.

The calculation of ne, the bulk electron densityin the presence of an electric field, follows theTownsend formalism (Llewellyn-Jones, 1981), inwhich ne is given by

ne � noe�x (6)

where no is the initial seed electron density (dueto steady-state conditions at the surface of Marsoutside the influence of electric fields), � is theTownsend first ionization coefficient, and x is aphysical length scale over which the electric fieldspersist. The Townsend coefficient � is determinedsimilarly as with the rate constant k

� � � 1/2

13.8 eV

�i(u)uf(u)du (7)

where nCO2 is the density of CO2, �i(u) is the ion-ization cross section for CO2 above �13.8 eV, andvd is the electron drift velocity:

vd � � � 12

uu du (8)

The constant � then describes the growth of elec-trons via electron–CO2 ionization as suggested byEq. 6. The choice of x in Eq. 6 is usually given bythe electrode spacings in laboratory experimentsdesigned to study ionization processes. In ourcase, we consider x to be the length scale of co-herent electric fields occurring in dust events.Though x may be considerably larger, we madea conservative estimate of x � 1 m, based on mea-

1�(u)

∂f∂u

2eme

E3N

2eme

nCO2

vd

MARS DUST STORM CHARGING AND OXIDANTS 457

FIG. 3. The electron density distribution in the atmos-phere at the surface of Mars for varying values of a duststorm-generated electric field. A neutral density of n � 2 �1023/m3 and an initial core electron density of no � 5 �106/m3 is assumed. The variable u is the electron energy ineV, and f(u) defined such that �u1/2f(u)du � 1. For electricfields �8 kV/m, a substantial high-energy tail develops.

TABLE 2. H2O DISSOCIATIVE ATTACHMENT PROCESSES

Dissociation product Energy (eV)

O(3P) � H2(X) 5.03OH(X) � H(n � 1) 5.10O*(1D) � H2(X) 7.00OH*(A) � H(n � 1) 9.15O*(1S) � H2(X) 9.22O(3P) � 2H 9.51O*(3s3SO) � H2(X) 14.56OH(X) � H*(n � 2) 15.30OH(X) � H*(n � 3) 17.19

Compiled by Itikawa and Mason (2005).

6278_06_p451-462 6/15/06 3:40 PM Page 457

surements taken in terrestrial dust devils (Deloryet al., 2002; Farrell et al., 2003) and consistent withthe desire to keep ne �� nCO2 such that the en-ergetic electrons remain a perturbation againstthe neutral background, with a degree of ioniza-tion �10�5.

The cross sections for dissociative attachmentof H2O were compiled by Itikawa and Mason(2005) and based on measurements conducted byMelton (1972) and Compton and Christophorou(1967). The highest cross sections for negative ionproduction from dissociative electron attachmentof H2O are for H�, peaking at �6.37 � 10�18 cm2

near �6.4 eV, and also �1.16 � 10�18 cm2 near�8.2 eV, while cross sections for OH� and O�

production peak at values �1 � 10�19 and�5.8 � 10�19 cm2, respectively. Using these crosssections and an H2O column density of �20 pr-�m (where pr-�m represents “precipitable mi-cron,“ and 1 pr-�m � 10�4 g/cm2 of H2O �3.35 � 1018/cm2 of H2O molecules, correspond-ing to �0.03% of the neutral number density atthe surface), we calculate the production rates ofthese products as a function of electric field inTable 3. Results of similar calculations for disso-ciation of CO2 into CO and O� are also listed inTable 3. Figure 4 shows the corresponding pa-rameters for the electron energetics and CO2 ion-ization, where ne, νd, �, and the rates of CO/O�

and OH/H� production are plotted as a functionof electric field E in kV/m.

DISCUSSION AND SIGNIFICANCE

Using a rigorous plasma physics approach, wearrive at a physical model, as discussed below,that reveals the impact of dust devil- and storm-generated electric fields on the atmosphericchemistry of Mars.

During virtually any dust activity arising fromthe ubiquitous saltation processes in dust devils

or the larger dust storms, significant triboelectriccharging of dust grains occurs. Subsequent loft-ing followed by vertical stratification of thesecharge carriers results in large-scale electricfields. Acting under the influence of the electricfield, a population of preexisting ambient elec-trons becomes energized and impacts atmo-spheric CO2 via a multitude of elastic, vibrational,electronic excitational, ionization, and dissocia-tive attachment processes. The electron density negrows substantially as ionization of the CO2 at-mosphere occurs. The mediation between energygained through the electric field versus the en-ergy lost due to CO2 interactions is expressed bythe final solution to the electron distribution func-tion f(u) in Fig. 3. These distributions are re-markably non-Maxwellian, with increasing high-energy tails becoming prominent with increasingelectric field. Once derived, the electron distribu-tion f(u) enables the calculation of the productionrates of other chemical species. In this case, be-cause of the importance of H2O in the formationof H2O2, we examine the impact of f(u) on the sta-bility of atmospheric H2O and find that dissocia-tive attachment from the energized electronsleads primarily to the production of OH/H�, fol-lowed by O� (and either 2H or H2) and OH/H�.

It is important to make the distinction betweena full discharge process and the electron ener-gization that we model here. In particular, wenote that the electron density expression in Eq. 6also possesses a denominator of the form 1 �(�/�)(e�x � 1) (Uman, 1969; Llewellyn-Jones,1981), which is ignored in our current calculation.The variable � is the growth rate of secondaryelectrons produced by the gas (electrons pro-duced by processes other than electron/moleculeimpact ionization), with �/� often referred to asTownsend’s second coefficient, and x is the elec-trode/anode distance. If a substantial number ofsecondary electrons are produced in the plasma,this denominator can go to zero, thereby making

DELORY ET AL.458

TABLE 3. CO2 AND H2O DISSOCIATION RATES

Rate (m�3s�1)

Electric field (kV/m) CO2 → CO/O� H2O → OH/H� H2O → O� H2O → OH�/H

5 4.17 � 109 1.31 � 107 5.30 � 105 2.77 � 105

10 7.75 � 1011 3.17 � 109 1.83 � 108 7.63 � 107

16 7.53 � 1012 3.27 � 1010 2.85 � 109 9.30 � 108

20 1.45 � 1014 6.24 � 1011 6.54 � 1010 1.93 � 1010

25 3.38 � 1017 1.42 � 1015 1.75 � 1014 4.75 � 1013

6278_06_p451-462 6/15/06 3:40 PM Page 458

the electron density tend toward infinity (allbound electrons are free). This situation of complete gas breakdown defines a “filamentary”discharge like the one that occurs in terrestrialcloud-to-ground lightning. Secondary electrongeneration processes are difficult to quantify, andtheir behavior is usually a strong function of theboundary conditions imposed on the plasma.While filamentary discharge and complete gasbreakdown is possible in a low-density CO2 gaswith secondaries (Llewellyn-Jones, 1981), wechoose to ignore secondary processes and dis-charges for our martian dust storm model giventhe high degree of uncertainty that would thenaffect the reliability of the derived chemical ratesfor OH/H� and other negative ion production.

Hence, by not including secondary electrongeneration processes (� � 0), we rule out chem-istry from filamentary, ionized discharges. How-ever, our calculations show that in pre-break-down conditions, significant chemical reactionrates are generated even in the case where the

plasma electron density is relatively small com-pared with the ambient density and the electrondrift velocities �d are �106 m/s. These drift ve-locities describe a slow-moving, mildly ionizedplasma that is very different from a typical dis-charge event, such as those that occur in terres-trial cloud-to-ground lightning stroke where ion-ization is nearly 100% and the electron speed is0.1–0.3c (Uman, 1969). The plasma in the martiandust storms may be more reminiscent of that interrestrial sprites: transient mesosphere luminousemissions (Rowland, 1998; Sentman et al., 1995;Sentman and Wescott, 1995) that are also mildlyionized (a few percent) and relatively slow mov-ing (Pasko et al., 1995, 1997). Further, the 30–50-km region of the terrestrial atmosphere has sim-ilar pressures as the lower martian atmosphere,which suggests similar atmospheric environ-ments for the two plasmas. Sprites result whenlightning-related post-discharge electrostaticfields between the cloudtops and ionosphere be-come large, and create electric field-driven elec-

MARS DUST STORM CHARGING AND OXIDANTS 459

Electron Drift Velocity

0 5 10 15 20 25 30E (kV/m)

105

v d m

/s

Townsend Coefficient

0 5 10 15 20 25 30E (kV/m)

10-12

10-10

10-8

10-6

10-4

10-2

100

102

α cm

-1

Electron Density

0 5 10 15 20 25 30E (kV/m)

106

107

108

109

1010

1011

1012

n e m

-3

CO/O-, OH/H- Rates

0 5 10 15 20 25 30E (kV/m)

106

108

1010

1012

1014

1016

1018

dn’ /d

t m-3 s

-1

CO/O-OH/H-

FIG. 4. The electron drift velocity (�d), Townsend first ionization coefficient (�), electron density (ne), and CO2 andH2O dissociation rates (dn’/dt) as a function of dust storm-driven electric fields.

6278_06_p451-462 6/15/06 3:40 PM Page 459

tron energization that acts to initiate excitation ofmesospheric gas. In the martian dust devil/stormcase, we are considering electrostatic fields thatform between the analogous ground and duststorm top that also give rise to electric field-driven electron energization. Certainly for thecase of filamentary discharges and large electrondensities, Eqs. 4 and 5 indicate greater produc-tion of CO, OH, and the negative ions. However,substantial reaction rates still occur in the low-density, slow-moving corona-like plasma wehave modeled here (see Fig. 4).

The rates calculated for the production ofOH/H� are particularly significant in terms ofthe subsequent generation of H2O2 (Atreya et al.,2006). OH/H� production from photochemicalprocesses alone is estimated to be �1010 m�3 s�1,while our calculations indicate that electric field-driven OH/H� rates become equal to photo-chemical rates in the presence of �10 kV/m fieldsand grow exponentially to orders of magnitudegreater values by 20–25 kV/m. As shown byAtreya et al. (2006), the rates of OH/H� produc-tion listed in Table 3 will enable the creation ofH2O2 at rates of up to �200 times typical photo-chemical rates. While the impact of electrifieddust on oxidant production is potentially sub-stantial, this process may also be limited by sev-eral key factors. First and foremost, the produc-tion of OH/H� is clearly dependent on theamount of water vapor present, which could bethe limiting factor for the subsequent generationof H2O2. Thus, we expect a wide variability in theefficiency of our electrochemical model, depend-ing on the distribution and amount of H2O avail-able during dust events. The spatial extent anddegree of electrification within martian duststorms and devils will also ultimately define lim-its. While discharges are common in terrestrialstorm systems, macroscopic electric fields rarelyexceed 400 kV/m (Winn et al., 1974), a fraction ofthe atmospheric breakdown potential of �3,000kV/m. This discrepancy represents an unre-solved issue in terrestrial thunderstorm electrifi-cation and may indicate that breakdown poten-tials are highly spatially localized within a givenstorm and difficult to measure in situ. Thus, if theelectrification of martian dust storms is analo-gous to terrestrial thunderstorms, these eventsmay only reach some fraction of the martianbreakdown potential over most of the storm vol-ume. However, details of the charging process arelikely different for dust storms on Mars compared

with analogous terrestrial processes. While large-scale, macroscopic electric fields may be limitedon Mars as in the terrestrial case, numerous mod-els do admit the possibility for higher fields thatapproach breakdown levels on large scales (Mel-nik and Parrot, 1998; Farrell et al., 2003). Pendinga verification of these electric fields using mea-surements from a future instrument on Mars, weinclude the full range of likely potentials that maybe encountered within a martian dust storm suchthat all likely production rates can be considered.

Despite these potential limitations, there is rea-son to believe that dust storm-driven electro-chemistry on Mars could be a compelling expla-nation for the degree of chemical reactivity andresulting sterility of martian soil as implied by theViking lander results. Even if larger-scale fieldsare limited on Mars, reaction rates on the orderof photochemical contributions will still occur forfields of �10 kV/m, less than half of the theoret-ical breakdown. We have also modeled predis-charge conditions only; subsequent electricalbreakdown and the possibility of runaway elec-trons in analogy with terrestrial sprites may in-troduce new and more energetic electrochemicalprocesses, albeit over short temporal and spatialscales. Additional aspects of our model relevantto enhanced oxidant production include the factthat these processes occur close to the surface,which results in a higher probability for any H2O2condensate to enter the regolith before significantre-vaporization. The near-surface location of elec-tric field-driven ion production also translatesinto the availability of a larger amount of ambi-ent H2O compared with what is available for pho-tochemical processes at higher altitudes, whichresults in enhanced production of H2O2.

CONCLUSION

Theory and experiment provide ample evi-dence that dynamic dust events are nearly alwayssignificantly electrified. Using a collisionalplasma physics model, we have shown that elec-trons acting under the influence of these electricfields in the martian atmosphere can significantlyenhance the creation of chemical products thatare important precursors to the production of theoxidant H2O2. Measurements of electric fields onMars, obtained simultaneously with the charac-terization of any trace species production using agas analyzer, would provide some observational

DELORY ET AL.460

6278_06_p451-462 6/15/06 3:40 PM Page 460

constraints on the importance of the electro-chemical process we have outlined. Assumingthat electrochemical processes in martian duststorms remain a possibility, future work will in-clude an examination of the direct impact of en-ergized electrons on other trace atmospheric con-stituents on Mars, such as methane and anyorganic compounds. Additionally, there is thequestion of the global and historical impact of at-mospheric electricity on atmospheric chemistry,which depends on the spatial and temporal cov-erage of these events as well as their degree oflarge-scale electrification.

ACKNOWLEDGMENTS

Partial support for this research includes fund-ing from NASA grant NAG5-9523, the Mars Fun-damental Research Program, and the PhoenixMars Scout Mission.

REFERENCES

Anderson, R. (1965) Electricity in volcanic clouds. Science48, 1179–1189.

Atreya, S.K. and Gu, Z.G. (1994) Stability of the Martianatmosphere: is heterogeneous catalysis essential? J. Geo-phys. Res. 99, 13133–13145.

Atreya, S.K., Wong, A.-S., Renno, N.O., Farrell, W.M., De-lory, G.T., Sentman, D.D., Cummer, S.A., Marshall, J.,Rafkin, S.C.R., and Catling, D. (2006) Oxidant en-hancement in martian dust devils and storms: implica-tions for life and habitability. Astrobiology 6, 439–450.

Ballou, E., Wood, P.C., Wydeven, T., Lehwalt, M.E., andMack, R.E. (1978) Chemical interpretation of VikingLander 1 life detection experiment. Nature 271, 644–645.

Cantor, B.A., James, P.B., Caplinger, M., and Wolff, M.J.(2001) Martian dust storms: 1999 Mars Orbiter Cameraobservations. J. Geophys. Res. 106, 23653–23687.

Clancy, R.T. and Nair, H. (1996) Annual (perihelion-aphe-lion) cycles in the photochemical behavior of the globalMars atmosphere. J. Geophys. Res. 101, 12785–12790.

Compton, R.N. and Christophorou, L.G. (1967) Negative-ion formation in H2O and D2O. Phys. Rev. 154, 110–116.

Crozier, W.D. (1964) The electric field of a New Mexicodust devil. J. Geophys. Res. 69, 5427–5429.

Crozier, W.D. (1970) Dust devil properties. J. Geophys. Res.75, 4583–4585.

Delory, G.T., Farrell, W.M., Hillard, B., Renno, N.O.,Smith, P., Marshall, J.R., and Eatchel, A. (2002) The elec-trical structure of terrestrial dust devils: implications ofmultiple vertical measurements of the electric field [ab-stract 0335]. In AGU Fall Meeting Abstracts, Vol. 51,American Geophysical Union, Washington, DC.

Encrenaz, T., Greathouse, T.K., Bezard, B., Atreya, S.K.,Wong, A.S., Richter, M.J., and Lacy, J.H. (2002) A strin-gent upper limit of the H2O2 abundance in the Martianatmosphere. Astron. Astrophys. 396, 1037–1044.

Encrenaz, T., Bezard, B., Greathouse, T.K., Richter, M.J.,Lacy, J.H., Atreya, S.K., Wong, A.S., Lebonnois, S.,Lefevre, F., and Forget, F. (2004) Hydrogen peroxide onMars: evidence for spatial and seasonal variations.Icarus 170, 424–429.

Ette, A.I.I. (1971) The effect of the harmattan dust on at-mospheric electric parameters. J. Atmospheric TerrestrialPhys. 33, 295–300.

Farrell, W.M., Delory, G.T., Cummer, S.A., and Marshall,J.R. (2003) A simple electrodynamic model of a dustdevil. Geophys. Res. Lett. 30, 2050–2053.

Farrell, W.M., Smith, P.H., Delory, G.T., Hillard, G.B.,Marshall, J.R., Catling, D., Hecht, M., Tratt, D.M.,Renno, N., Desch, M.D., Cummer, S.A., Houser, J.G.,and Johnson, B. (2004) Electric and magnetic signaturesof dust devils from the 2000–2001 MATADOR deserttests. J. Geophys. Res. 109, E03004.

Ferri, F., Smith, P.H., Lemmon, M., and Renno, N.O.(2003) Dust devils as observed by Mars Pathfinder. J.Geophys. Res. 108, doi:10.1029/2000JE001421.

Fisher, J.A., Richardson, M.I., Newman, C.E., Szwast, M.A.,Graf, C., Basu, S., Ewald, S.P., Toigo, A.D., and Wilson,R.J. (2005) A survey of Martian dust devil activity usingMars Global Surveyor Mars Orbiter Camera images. J.Geophys. Res. 110, E03004, doi:10.1029/2003JE002165.

Formisano, V., Atreya, S., Encrenaz, T., Ignatiev, N., andGiuranna, M. (2004) Detection of methane in the at-mosphere of Mars. Science 306, 1758–1761.

Freier, G.D. (1960) The electric field of a large dust devil.J. Geophys. Res. 65, 3504.

Itikawa, Y. (2002) Cross sections for electron collisionswith carbon dioxide. J. Phys. Chem. Ref. Data 31, 749–767.

Itikawa, Y. and Mason, N. (2005) Cross sections for elec-tron collisions with water molecules. J. Phys. Chem. Ref.Data 34, 1–22.

Klein, H.P. (1998) The search for life on Mars: what welearned from Viking. J. Geophys. Res. 103, 28463–28466.

Krasnopolsky, V.A. (1993) Photochemistry of the Martianatmosphere (mean conditions). Icarus 101, 313–332.

Krasnopolsky, V.A. (1995) Uniqueness of a solution of asteady state photochemical problem: applications toMars. J. Geophys. Res. 100, 3263–3276.

Krasnopolsky, V.A., Maillard, J.P., and Owen, T.C. (2004)Detection of methane in the martian atmosphere: evi-dence for life? Icarus 172, 537–547.

Llewellyn-Jones, F. (1981) The development of theories ofthe electrical breakdown of gases. In Electrical Break-down and Discharges in Gases. Part A: FundamentalProcesses and Breakdown, edited by E.E. Kunhardt andL.H. Luessen, Plenum Press, New York, pp. 1–71.

Mancinelli, R.L. (1989) Peroxides and the survivability ofmicroorganisms on the surface of Mars. Adv. Space Res.9, 191–195.

Martin, L.J., and Zurek, R.W. (1993) An analysis of thehistory of dust activity on Mars. J. Geophys. Res. 98,3221–3246.

MARS DUST STORM CHARGING AND OXIDANTS 461

6278_06_p451-462 6/15/06 3:40 PM Page 461

Melnik, O. and Parrot, M. (1998) Electrostatic dischargein Martian dust storms. J. Geophys. Res. 103, 29107–29118.

Melton, C.E. (1972) Cross sections and interpretation ofdissociative attachment reactions producing OH�,O/sup�, and H� in H2O. J. Chem. Phys. 57, 4218–4225.

Mills, A.A. (1977) Dust clouds and frictional generationof glow discharges on Mars. Nature 268, 614.

Mumma, M.J., Novak, R.E., DiSanti, M.A., Bonev, B.P.,and Dello Russo, N. (2004) Detection and mapping ofmethane and water on Mars [abstract 26.02]. In DPSMeeting 36, American Astronomical Society, Washing-ton, DC.

Nair, H., Allen, M., Anbar, A.D., Yung, Y.L., and Clancy,R.T. (1994) A photochemical model of the Martian at-mosphere. Icarus 111, 124–150.

Newman, C.E., Lewis, S.R., Read, P.L., and Forget, F.(2002) Modeling the Martian dust cycle. 1. Representa-tions of dust transport processes. J. Geophys. Res. 107,5123, doi:10.1029/2002JE001910.

Nighan, W.L. (1970) Electron energy distributions andcollision rates in electrically excited N2, CO, and CO2.Phys. Rev. A 2, 1989–2000.

Oyama, V.I. and Berdahl, B.J. (1979) A model of Martiansurface chemistry. J. Mol. Evol. 14, 199–210.

Oyama, V.I., Berdahl, B.J., and Carle, G.C. (1977) Prelim-inary findings of the Viking gas exchange experimentand a model for Martian surface chemistry. Nature 265,110–114.

Pasko, V.P., Inan, U.S., Taranenko, Y.N., and Bell, T.F.(1995) Heating, ionization and upward discharges inthe mesosphere due to intense quasi-electrostatic thun-dercloud fields. Geophys. Res. Lett. 22, 365–368.

Pasko, V.P., Inan, U.S., Bell, T.F., and Taranenko, Y.N.(1997) Sprites produced by quasi-electrostatic heatingand ionization in the lower ionosphere. J. Geophys. Res.102, 4529–4561.

Pitchford, L.C., O’Neil, S.V., and Rumble, J.R., Jr. (1981)Extended Boltzmann analysis of electron swarm ex-periments. Phys. Rev. A 23, 294–304.

Renno, N.O., Wong, A.-S., Atreya, S.K., de Pater, I., andRoos-Serote, M. (2003) Electrical discharges and broad-band radio emission by Martian dust devils and duststorms. Geophys. Res. Lett. 30(22), 2140, doi:10.1029/2003GL017879.

Renno, N.O., Abreu, V.J., Koch, J., Smith, P.H., Harto-gensis, O.K., De Bruin, H.A.R., Burose, D., Delory, G.T.,Farrell, W.M., Watts, C.J., Garatuza, J., Parker, M., and

Carswell, A. (2004) MATADOR 2002: a pilot field ex-periment on convective plumes and dust devils. J. Geo-phys. Res. 109, doi:10.1029/2003JE002219.

Rowland, H.L. (1998) Theories and simulations of elves,sprites and blue jets. J. Atmospheric Solar-Terrestrial Phys.60, 831–844.

Ryan, J.A. and Lucich, R.D. (1983) Possible dust devils—vortices on Mars. J. Geophys. Res. 88, 11005–11011.

Schmidt, D.S., Schmidt, R.A., and Dent, J.D. (1998) Elec-trostatic force on saltating sand. J. Geophys. Res. 103,8997–9001.

Sentman, D.D. and Wescott, E.M. (1995) Red sprites andblue jets: thunderstorm-excited optical emissions in thestratosphere, mesosphere, and ionosphere. Phys. Plas-mas 2, 2514–2522.

Sentman, D.D., Wescott, E.M., Osborne, D.L., Hampton,D.L., and Heavner, M.J. (1995) Preliminary results fromthe Sprites94 aircraft campaign. 1. Red sprites. Geophys.Res. Lett. 22, 1205–1208.

Smith, P.H. and Lemmon, M. (1999) Opacity of the Mar-tian atmosphere measured by the Imager for MarsPathfinder. J. Geophys. Res. 104, 8975–8986.

Thomas, P. and Gierasch, P.J. (1985) Dust devils on Mars.Science 230, 175–177.

Uman, M.A. (1969) Lightning, McGraw-Hill Book Co.,New York.

Viggiano, A.A. and Arnold, F. (1995) Ion chemistry andcomposition of the atmosphere. In Handbook of Atmo-spheric Electrodynamics, edited by H. Volland, CRCPress, Boca Raton, FL, pp. 1–26.

Whitten, R.C., Poppoff, I.G., and Sims, J.S. (1971) Theionosphere of Mars below 80 km altitude—I. Quiescentconditions. Planet. Space Sci. 19, 243–250.

Winn, W.P., Schwede, G.W., and Moore, C.B. (1974) Mea-surements of electric fields in thunderclouds. J. Geophys.Res. 79, 1761–1767.

Zent, A.P. and McKay, C.P. (1994) The chemical reactiv-ity of the martian soil and implications for future mis-sions. Icarus 108, 146–147.

Address reprint requests to:Gregory T. Delory

Space Sciences Laboratory, MS 7450University of California, Berkeley

Berkeley, CA 94720

E-mail: [email protected]

DELORY ET AL.462

6278_06_p451-462 6/15/06 3:40 PM Page 462