arX

iv:1

309.

6016

v1 [

astr

o-ph

.EP]

24

Sep

2013

Draft version September 25, 2013Preprint typeset using LATEX style emulateapj v. 5/2/11

BIOSIGNATURE GASES IN H2-DOMINATED ATMOSPHERES ON ROCKY EXOPLANETS

S. Seager,, W. Bains1,, R. Hu1,

Draft version September 25, 2013

ABSTRACT

Super Earth exoplanets are being discovered with increasing frequency and some will be able toretain stable H2-dominated atmospheres. We study biosignature gases on exoplanets with thin H2atmospheres and habitable surface temperatures, by using a model atmosphere with photochemistry,and biomass estimate framework for evaluating the plausibilty of a range of biosignature gas candi-dates. We find that photochemically produced H atoms are the most abundant reactive species in H2atmospheres. In atmospheres with high CO2 levels, atomic O is the major destructive species for somemolecules. In sun-Earth-like UV radiation environments, H (and in some cases O) will rapidly destroynearly all biosignature gases of interest. The lower UV fluxes from UV quiet M stars would produce alower concentration of H (or O) for the same scenario, enabling some biosignature gases to accumulate.The favorability of low-UV radiation environments to accumulation of detectable biosignature gasesin an H2 atmosphere is closely analogous to the case of oxidized atmospheres, where photochemicallyproduced OH is the major destructive species. Most potential biosignature gases, such as DMS andCH3Cl are therefore more favorable in low UV, as compared to solar-like UV, environments. A fewpromising biosignature gas candidates, including NH3 and N2O, are favorable even in solar-like UVenvironments, as these gases are destroyed directly by photolysis and not by H (or O). A more subtlefinding is that most gases produced by life that are fully hydrogenated forms of an element, such asCH4, H2S, are not effective signs of life in an H2-rich atmosphere, because the dominant atmosphericchemistry will generate such gases abiologically, through photochemistry or geochemistry. Suitablebiosignature gases in H2-rich atmospheres for super Earth exoplanets transiting M stars could poten-tially be detected in transmission spectra with the James Webb Space Telescope.

1. INTRODUCTION

The detection of exoplanet atmospheric biosignaturegases by remote sensing spectroscopy is usually taken asinevitable for the future of exoplanets. This sentimentis being borne out with the discovery of increasing num-bers of smaller and lower mass planets each year. Inaddition, the natural evolution to development of largerand more sophisticated telescopes (such as the JamesWebb Space Telescope (JWST) slated for launch in 2018,Gardner et al. (2006)) and the giant 20- to 40-meter classground-based telescopes1 continues to fuel the conceptthat the eventual detection and study of biosignaturegases is a near certainty.The topic of biosignature gases, however, may remain

a futuristic one unless a number of extreme challengescan be overcome. The biggest near-term challenge is tofind a large enough pool of potentially habitable exo-planets accessible for followup atmosphere study2. Bypotentially habitable we mean rocky planets with sur-face liquid water, and not those with massive envelopesmaking any planet surface too hot for complex moleculesrequired for life. The large pool of such planets is neededbecause there could be a large difference in the num-bers of seemingly potentially habitable planets (basedon their measured host star type, orbit, and mass or size

1 The Extremely Large Telescope(http://www.eso.org/public/teles-instr/e-elt.html), the GiantMagellan Telescope (http://www.gmto.org/), and the ThirtyMeter Telescope (http://www.tmt.org/).

2 For example the all-sky space-based TESS mission (TranstingExoplanet Planet Survey Satellite, PI George Ricker) has beenselected under NASAs Astrophysics Explorer Program for launchin 2017.

and inferred surface temperature) and those that are in-habited by life that produces useful biosignature gases(which will be inferred from measured atmospheric spec-tra). Useful biosignature gases means those that can ac-cumulate in the planet atmosphere, are spectroscopicallyactive, and are not overly contaminated by geophysicalfalse positives. A contemporary, related point on iden-tifying a large enough pool of planets is that even thefraction of small or low-mass planets that are potentiallyhabitablethat is with surface conditions suitable forliquid wateris not yet known. The reason is that thefactors controlling a given planets surface temperaturesare themselves not yet observed or known, including theatmosphere mass (and surface pressure), the atmosphericcomposition, and hence the concomitant greenhouse gaspotency (see the review by Seager 2013).A second major challenge for the study of biosigna-

ture gases is the capability of telescopes to robustly de-tect molecules in terrestrial exoplanet atmospheres. Thischallenge is continously faced in todays hot Jupiter at-mosphere studies (e.g., Seager 2010), where many at-mospheric molecular detections based on data from theHubble Space Telescope or the Spitzer Space Telescoperemain controversial (see Deming et al. 2013, and ref-erences therein). For transiting planets, the ability toidentify and remove systematics to a highly precise levelwhile adding together numerous transit events from dif-ferent epochs is a necessity to reach the small signalsof terrestrial planet atmospheres. For directly imagedplanets, the ability to reach down to Earth-sized plan-ets in Earth-like orbits is one of the most substantialtechnological challenges to ever face astronomers. While

http://arxiv.org/abs/1309.6016v1http://www.eso.org/public/teles-instr/e-elt.htmlhttp://www.gmto.org/http://www.tmt.org/

2

technology development is ongoing, there are as yet nosolid plans to launch a space telescope capable of directlyimaging terrestrial-size planets.A third major challenge in the study of biosignature

gases has to do with the geological false positive signa-tures. These false positives are gases that are producedgeologically and emitted by volcanoes or vents in thecrust or ocean. Geochemistry has the same chemicals towork with that life does, and therefore false positives areinevitable. While early theoretical studies favored detec-tion of redox disequilibrium (such as O2 and CH4) thatshould not both exist in an atmosphere in photochemicalsteady state, often one molecule of the set is too weakspectroscopically for potential detection. The conven-tionally adopted approach (at least in theoretical studies)is therefore to identify a biosignature gas that is manyorders of magnitude out of thermodynamic equilibriumwith the expected gas composition of the atmosphere andto study the gas in the context of the planet atmosphereenvironment via atmospheric spectra that cover a widewavelength range. A more likely outcome to the field ofbiosignature gases will be to develop probabilistic assess-ment of the likelihood a molecule in a given atmospherecan be attributed to life, because spectroscopic data andthe information for a complete assessment of the plane-tary environment will be limited.To increase the chances of detecting exoplanet atmo-

spheric biosignature gases we are motivated to widen theparameter space of types of planets where biosignaturegases can accumulate and should be sought out obser-vationally. We here describe, for the first time to ourknowledge, the case for and against biosignature gasesin hydrogen-rich atmospheres. Some massive enough orcold enough super Earths (loosely defined as planets withup to 10 Earth masses) will be able to retain hydrogenin their atmospheres (see the discussion in 5.2). In gen-eral, planets are expected to outgas or capture hydro-gen from the nebula during planet formation. Here weare concerned with super Earths with relatively thin hy-drogen atmospheres and not planets with massive atmo-spheres or envelopes (as in mini-Neptunes) which willhave surfaces too hot for liquid water (Rogers and Sea-ger, in prep.) or may not even have a surface. A thinhydrogen atmosphere does not add much to either themass or the size of the planet (Adams et al. 2008), sothat an H2-rich atmosphere itself does not aid in planetdiscovery or detection.Super Earths with H2-rich atmospheres are nonetheless

in some ways more favorable for detection and study thantheir terrestrial planet counterparts with N2- or CO2-dominated atmospheres. A more massive planet thanEarth (i.e., more likely to retain atmospheric H2 thanEarth) is easier to discover than an Earth-mass planetvia the radial velocity technique. A more massive planetthan Earth is also larger and so easier to discover or de-tect by the transit technique than a lower-mass planet.For example a 10 M planet of Earth-like compositionwould have a radius 1.75 times larger than Earth (e.g.,Seager et al. 2007). The larger planet area is more fa-vorable for atmosphere study in reflected or thermallyemitted radiation than an Earth-size planet. For transittransmission spectra, planets with H2-rich atmosphereshave a much larger signal compared to H-poor atmo-spheres because of the larger scale height H , based on

the mean molecular weight (e.g., Seager 2010),

H =kT

mHg, (1)

where k is Boltzmanns constant, T is temperature, mHis the mass of the hydrogen atom, and g is the surfacegravity. The point is that when H2 dominates the atmo-spheric composition over terrestrial planet atmospheregases CO2 and N2, the mean molecular weight is 20times smaller and hence the scale height is 20 timeslarger. The observational imprint of an atmosphere isusually taken as about 5H .Turning back to biosignature gases, they have been

studied theoretically as indicators of life on planets withoxidized atmospheres for over half a century, beginningwith Lederberg (1965) and Lovelock (1965). One high-light from the last decade is the realization that low UVradiation environments compared to solar lead to a muchhigher concentration of biosignature gases, as studied forEarth-like planet atmospheres. This is because the stel-lar UV creates the radical OH (in some cases O) whichdestroys many gases in the atmosphere and thus reducesthe gas lifetime. A low UV radiation environment istaken to be that of a planet orbiting a UV quiet M dwarfstar (see Figure 1 and discussion in 5.6).A second highlight in biosignature gas research in the

last decade is the theoretical exploration of potentialbiosignature gases beyond the conventionally considereddominant Earth or early Earth based ones of O2, O3,N2O, and CH4. The variations studied include dimethyl-sulfide (DMS; Pilcher 2003), methyl chloride (CH3Cl;Segura et al. 2005), and other sulfur compounds includ-ing CS2 and COS (Domagal-Goldman et al. 2011). SeeSeager et al. (2012) for a review, Seager et al. (2013) fora biosignature gas classification scheme, and Seager et al.(2013) for a biomass model estimate intended as a plaus-ability check to consider biosignature gas surface fluxesdifferent from Earth values.We begin with a description of our atmosphere and

biomass estimate model in 2. We present general resultsin 3 and specific results for a number of potential andunlikely biosignature gases in 4. A discussion in 5 isfollowed by a summary and conclusion in 6.

2. MODEL

The model goal is to computationally generate atmo-spheric spectra for exoplanets with H2-rich atmosphereswith biosignature gases. The model consistes of a photo-chemistry code which takes biosignature gases as surfacefluxes, an approximate temperature profile calculation,and a line-by-line spectral calculation (Seager et al. 2013;Hu et al. 2012). The model also uses a biomass model es-timate, to check whether or not a biosignature gas couldbe the result of a plausible surface ecology.

2.1. Model Atmosphere

Photochemistry Model The focus on chemistry is criti-cal for biosignature gases, because sinks control a biosig-nature gas lifetime and hence the gas potential to accu-mulate in the planetary atmosphere. A model for atmo-spheric chemistry is required to connect the amount ofbiosignature gas in the atmosphere (as required for de-tection) to the biosignature source flux at the planetarysurface.

3

102

103

104

1020

1015

1010

105

100

Wavelength [nm]

Flux in the Habitable Zone [W m

2 nm1]

Sun

M star with Teff = 3000 K without chromospheric emission

GJ 1214

Figure 1. Comparison of stellar fluxes. The radiative flux received by a planet in the habitable zone of a solar-like star, a weakly active Mdwarf star (like GJ 1214), and a theoretically simulated quiet M dwarf star with an effective temperature of 3000 K with no chromosphericemission. The flux is scaled so that the planet has a surface temperature of 290 K. The spectrum of the sun-like star is from the AirMass Zero reference spectrum during a solar quiet period (http://rredc.nrel.gov/solar/spectra/am0/). The spectrum of GJ 1214 containstwo parts: for wavelengths shorter than 300 nm we take the most recent HST measurement (France et al. 2013), for wavelengths longerthan 300 nm we take the NextGen simulated spectrum for an M dwarf star having parameters closest to those of GJ 1214 (i.e., effectivetemperature of 3000 K, surface gravity log(g) = 5.0, and metallicity = 0.5). The spectrum with no chromospheric emission is also from theNextGen model (Allard et al. 1997). Under the common definition of weakly active, or relatively quiet M dwarf, the UV environment inits habitable zone can differ by more than six orders of magnitude.

Our photochemical model is presented in Hu et al.(2012). The photochemical code computes the steady-state chemical composition of an exoplanetary atmo-sphere. The system can be described by a set of time-dependent continuity equations, one equation for speciesat each altitude. Each equation describes: chemical pro-duction; chemical loss; eddy diffusion and molecular dif-fusion (contributing to production or loss); sedimenta-tion (for aerosols only); emission and dry deposition atthe lower boundary; and diffusion-limited atmosphericescape for light species at the upper boundary. The codeincludes 111 species, 824 chemical reactions, and 71 pho-tochemical reactions. Starting from an initial state, thesystem is numerically evolved to the steady state in whichthe number densities no longer change.The generic model computes chemical and photochem-

ical reactions among all O, H, N, C, S species, and forma-tion of sulfur and sulfate aerosols. The numerical codeis designed to have the flexibility of choosing a subsetof species and reactions in the computation. The codetherefore has the ability to treat both oxidized and re-duced conditions, by allowing selection of fast species.For the chemical and photochemical reactions, we usethe most up-to-date reaction rates data from both theNIST database (http://kinetics.nist.gov) and the JPLpublication (Sander et al. 2011). Ultraviolet and visi-ble radiation in the atmosphere is computed by the -Eddington two-stream method with molecular absorp-

tion, Rayleigh scattering and aerosol Mie scattering con-tributing to the optical depth. The model was developedfrom the ground-up and has been tested and validated byreproducing the atmospheric composition of Earth andMars (Hu et al. 2012; Hu 2013).For biosignature gases that are minor chemical per-

turbers in the atmosphere, the biosignature lifetime canbe estimated based on the abundance of the major chem-ical sink. For this paper, the values of NH3 and N2Osurface source fluxes are calculated from the full pho-tochemistry model, whereas the calculations of surfacesource fluxes for other biosignature gases are simplifiedestimates. One more point to note is that photochem-istry is relevant high in the atmosphere typically abovembar levels to which stellar UV radiation can penetratefrom above.Temperature-Pressure Profile The precise temperature-

pressure structure of the atmosphere is less impor-tant than photochemistry for a first-order description ofbiosignatures in H2-rich atmospheres. The reason is thatmost biosignature gases of interest have sources and sinksnot signficantly affected by minor deviations in the tem-perature pressure profile. Morever the biosignature gasesthemselves are secondary players in governing the heat-ing structure of the atmosphere.We therefore justify using the photochemistry model in

a stand-alone mode, with a pre-calculated temperature-pressure profile. The calculated temperature-pressure

http://rredc.nrel.gov/solar/spectra/am0/http://kinetics.nist.gov

4

profile is approximate and is one which assumes a sur-face temperature (e.g., 290 K), an appropriate adiabaticlapse rate for H2-rich compositions, and a constant tem-perature above the convective layer of the atmosphere.Such assumed temperature profiles are consistent withgreenhouse warming in the troposphere and lack of ul-traviolet absorber in the stratosphere. The semi-majoraxis of the planet is then derived based on the assumedtemperature profile by balancing the energy flux of in-coming stellar radiation and outgoing planetary thermalemission. The details of this procedure are described inHu et al. (2012).Synthetic Spectra To generate exoplanet transmis-

sion and thermal emission spectra, we use a line-by-line radiative transfer code (Seager et al. 2000;Madhusudhan & Seager 2009; Hu et al. 2012). Opaci-ties are based on molecular absorption with cross sec-tions computed based on data from the HITRAN 2008database (Rothman et al. 2009), molecular collision-induced absorption when necessary (e.g., Borysow 2002),Rayleigh scattering, and aerosol extinction computedbased on Mie theory. The atmospheric transmission iscomputed for each wavelength by integrating the op-tical depth along the limb path (as outlined in, e.g.,Seager & Sasselov 2000; Miller-Ricci et al. 2009). Theplanetary thermal emission is computed by integratingthe radiative transfer equation without scattering foreach wavelength (e.g., Seager 2010).We consider clouds in the emergent spectra for thermal

emission by considering 50% cloud coverage by averag-ing a cloudy and cloud-free spectra. We omit clouds forthe transmission spectra model because the clouds are atlow altitudes whereas the spectral features form at highaltitudes.

2.2. Biomass Model Estimates

A biomass model estimate has been developed bySeager et al. (2013) that ties biomass surface density toa given biosignature gas surface source flux. The moti-vating rationale is that with a biomass estimate, biosig-nature gas source fluxes can be free parameters in modelpredictions, by giving a physical plausibility check interms of reasonable biomass. The approach aims to en-able consideration of a wide variety of both gas speciesand their atmospheric concentration to be considered inbiosignature model predictions. The biomass model es-timates are valid to one or two orders of magnitude. Weprovide a summary of the biomass model here with thefull details available in Seager et al. (2013).The biomass model is used in the following algorithm.

First we calculate the amount of biosignature gas re-quired to be present at detectable levels in an exo-planet atmosphere from a theoretical spectrum (we de-fine a detection metric in 2.3). Second, we determinethe gas source flux necessary to produce the atmosphericbiosignature gas in the required atmospheric concentra-tion. The biosignature gas atmospheric concentration isa function not only of the gas surface source flux, butalso of other atmospheric and surface sources and sinks.Third, we estimate the biomass that could produce thenecessary biosignature gas source flux. Fourth, we con-sider whether the estimated biomass surface density isphysically plausible, by comparison to maximum terres-trial biomass surface density values and total plausible

surface biofluxes.Based on life on Earth, a summary overview is that a

biomass surface density of 10 g m2 is sensible, 100 g m2

is plausible, and 5000 g m2 is possible. In real sit-uations, the total biomass is nearly always limited byenergy, bulk nutrients (carbon, nitrogen), trace nutri-ents (iron, etc.) or all three. Regarding global surfacebiofluxes we provide values and references where neededin our results and discussion.The biomass model estimates are tied to the type

of biosignature gas and so we briefly summarize ourbiosignature classification scheme before discussion eachbiomass model estimate.Type I biosignature gases are generated as byproduct

gases from microbial energy extraction. For example,on Earth many microbes extract energy from chemicalenergy gradients using the abundant atmospheric O2 foraerobic oxidation,

X + O2 oxidized X. (2)

For example: H2O is generated from H2; CO2 from or-ganics; SO2 or SO

24 from H2S; rust from iron sulfide

(FeS); NO2 and NO

3 from NH3; etc.On an exoplanet with an H2-rich atmosphere, the

abundant reductant would now atmospheric H2 such that

H2 +X reduced X. (3)

The oxidant must come from the interior.In other words, for chemical potential energy gradi-

ents to exist on a planet with an H2-rich atmosphere,the planetary crust must (in part) be oxidized in orderto enable a redox couple with the reduced atmosphere.The byproduct is always a reduced gas, because in a re-ducing environment H2-rich compounds are the availablereductants. To be more specific, oxidants would includegases such as CO2 and SO2.The Type I biosignature gas biomass model is based

on thermodynamics and is derived from conservation ofenergy and discussed in detail in Seager et al. (2013).The biomass model estimate is

B G

[

FsourcePme

]

. (4)

Here B is the biomass surface density in g m2, G

is the Gibbs Free energy of the chemical redox reac-tion from which energy is extracted (e.g., equation (2)).G depends on the standard free energy of reaction(G0), and the concentration of the reactants and prod-ucts. Reactant and product concentrations can includeocean pH (concentration of H+) in reactions that gen-erate or consume protons. G0 values are taken fromAmend & Shock (2001).The term Pme is an empirically determined microbial

maintenance energy consumption rate, that is, the mini-mum amount of energy an organism needs per unit timeto survive in an active state (i.e., a state in which theorganism is ready to grow). An empirical relation hasbeen identified by Tijhuis et al. (1993) that follows anArrhenius law

Pme = A exp

[

EART

]

. (5)

Here EA = 6.94 104 kJ mol1 is the activation energy,

5

R = 8.314 kJ mol1 K1 is the universal gas constant,and T in units of K is the temperature. The constantA is 4.3 107 kJ g1 s1 for aerobic growth and 2.5 107 kJ g1 s1 for anaerobic growth (Tijhuis et al. 1993).Here per g refers to per g of wet weight of the organism.Pme is in units of kJ g

1 s1.The free parameter in this biomass model estimate

(equation (4)) is the biosignature gas source flux Fsource(in units of mole m2 s1). Fsource is the flux of themetabolic byproduct and is also the surface bioflux re-quired to generate a given biosignature gas concentrationin the atmosphere.Type II biosignature gases are byproduct gases pro-

duced by the metabolic reactions for biomass building,and require energy. On Earth these are reactions thatcapture environmental carbon (and to a lesser extentother elements) in biomass. Type II biosignature re-actions are energy-consuming, and on Earth the energycomes from sunlight via photosynthesis.There is no useful biomass model for Type II biosig-

nature gases because once the biomass is built a Type IIbiosignature gas is no longer generated.Type III biosignature gases are produced by life but not

as byproducts of their central chemical functions. TypeIII biosignature gases appear to be special to particu-lar species or groups of organisms, and require energyfor their production. Because the chemical nature andamount released for Type III biosignature gases are notlinked to the local chemistry and thermodynamics, theType III biosignature gas biomass model is an estimatebased on lab culture production rates.We estimate the biomass surface density by taking the

biosignature gas source flux Fsource (in units of mole m2

s1) divided by the mean gas production rate in the labRlab (in units of mole g

1 s1),

B FsourceRlab

. (6)

We take the maximum observed for the Type III Rlabrates Ffield values from different studies (Seager et al.2013). The caveat of the Type III biomass estimate ex-plicitly assumes that the range of R for life on exoplan-ets is similar to that for life in Earths lab environment.Nonetheless we have showed based on Earths values thatthe Type III biomass model is valid to one or two ordersof magnitude. The goal, again, is to use the biomass esti-mate to argue for or against plausibilty of a biosignaturegas based on Earths biomass surface density values andnot for any prediction of quantitative values.Bioindicators are defined as the end product of chem-

ical reactions of a biosignature gas.Model caveats are related to the order of magnitude

nature of the biomass estimates, the possible terracen-tricity of the biomass model estimates, and the lack ofecosystem context (see Seager et al. (2013) 6.1, 6.2, 6.3for a detailed discussion). Here we provide a summaryoverview.The order of magnitude nature of the Type I biomass

estimate derives from the dependency of the estimateon Pme , itself very sensitive to temperature. The pos-sible terracentricity of our estimates is related to use ofPme , which is derived from observations of terrestrial mi-croorganisms, but we argue the dependency is largely

on thermodynamics (Seager et al. 2013). The order ofmagnitude nature of the Type III biosignatures derivesthe reliance on laboratory rates for microbial productionrates, and this is also possibly a terracentricity issue.The lack of ecosystem context is a major limitation

for the biomass estimate. An ecosystem contains notonly the producers (i.e., the biomass estimate derivedultimately from the bioflux Fsource) but also consumers,whereas the biomass model estimate considers only theproducers. In this sense the biomass estimate in equa-tion (4) is a minimum. We can fairly say that in the caseof a very small or very large biomass estimate, the as-sessment of biosignature gas plausibility is valid: a smallbiomass estimate gives room for consumers even as aminimum biomass and a large biomass estimate as a min-imum will remain large regardless of the consumers. Forthe intermediate case where a large but not unreason-able biomass is needed to generate a detectable biosig-nature, the decision on whether the gas is a plausiblebiosignature is more complicated, and will depend onthe planetary context: geochemistry, surface conditions,atmospheric composition and other factors.Again, we do not argue the biosignature biomass model

estimates are an accurate prediction of an extraterrestrialecology, rather we emphasize the goal of the biomassmodel estimates is the order of magnitude nature fora first order asssessment of the plausability of a givenbiosignature gas candidate.

2.3. Detection Metric

We now describe our metric for a detection that leadsto a required biosignature gas concentration. For now,the detection has to be a theoretical exercise using syn-thetic data. We determine the required biosignature gasconcentration based on a spectral feature detection witha SNR=10. Specifically, we describe the SNR of the spec-tral feature as the difference between the flux in the ab-sorption feature and the flux in the surrounding contin-uum (on either side of the feature) taking into accountthe uncertainties on the data,

SNR =|Fout Fin|

2Fout

+ 2Fin

, (7)

where Fin Fin is the flux density inside the absorp-tion feature and Fout Fout is the flux density in thesurrounding continuum, and is the uncertainty on themeasurement.The uncertainties of the in-feature flux and continuum

flux are calculated for limiting scenarios. For thermalemission we consider a futuristic space telescope able toblock out the light of the host star. The uncertaintiesof the in-feature flux and continuum flux are calculatedfor a limiting scenario: an 1.75 times Earth-sized planetorbiting a star3 at 10 pc observed (via direct imaging)with a 6 m-diameter telescope mirror operating within50% of the shot noise limit and a quantum efficiency of20%. The integration time is assumed to be 20 hours.We note that collecting area, observational integrationtime, and source distance are interchangeable dependingon the time-dependent observational systematics. This

3 Assuming perfect removal of starlight.

6

telecope scenario is based on a TPF-I type telescope(Lawson et al. 2008).For transit transmission spectra, we use the same equa-

tion as above, but with the denominator replaced bythe noise in the stellar flux (F), as in

(42F), be-

cause transmission observations measure the differencebetween the in-transit stellar flux and out-of-transit stel-lar flux. For transmission spectra we consider a 6.5-mspace telescope, having quantum efficiency of 0.25 ob-serving with 50% photon noise limit, with integrationtime of 60 hours in-transit and 60 hours out-of-transit(assuming observing of multiple transits). Again, wenote that collecting area, observational integration time,and source distance are interchangeable depending on thetime-dependent observational systematics. This scenariois based on the JWST.

3. PHOTOCHEMISTRY RESULTS: H IS THE DOMINANTPHOTOCHEMICALLY-PRODUCED REACTIVE SPECIES

IN H2-RICH ATMOSPHERES

In an H2-rich terrestrial exoplanet atmosphere, atomicH is the largest sink for most atmospheric molecules in-cluding biosignature gases. This is in contrast to oxi-dizing atmospheres (atmospheres with substantial O2 orCO2 and H2O and without H2) where the OH radical(and in some cases O) plays the role of the dominantsink. We note that for H2-rich atmospheres with highCO2 levels, atomic O will be abundant (Figure 2) andfor some molecules will dominate the removal chemistry.To explain the high H concentration we review the pro-

duction of H, OH and O in H2-rich atmospheres. Toqualitatively outline the main points we use a simplifieddescription of the main chemical pathways. This dis-cussion serves for illustration only, and is later backedup with a more detailed computational photochemistrymodel.To derive atmospheric concentrations of a species [A]

we take photochemical equilibrium,

d[A]

dt= P [A]L = 0, (8)

[A] =P

L. (9)

where [A] is the mixing ratio of species A, and P and Lare the production and loss rates respectively of speciesA. Below, the K are reaction constants and J is thephotodestruction rate associated with a stated reaction.We consider an H2 atmosphere with some H2O.

H2O+ h H+OH J, (10)

OH+ H2 H2O+H K, (11)

H + H +M H2 +M Km. (12)

Combining the above two equations we have

[H] =

K[OH][H2]

Km[M]=

J [H2O]

Km, (13)

and

[OH] =J [H2O]

K[H]. (14)

The simplified atmosphere reveals a number of relevantpoints. The first major point is that the role of OH in

103

102

101

100

101

102

103

1018

1017

1016

1015

1014

1013

1012

1011

1010

109

108

107

106

105

104

103

102

101

100

UV Flux [Solar UV Flux]M

ixin

g R

atio

CO2

H

OH

O

Figure 2. Mixing ratio dependence of the reactive species (H,OH, and O) on UV flux in a H2-rich atmosphere with some CO2.Shown are different CO2 levels. The curves correspond to a CO2surface emission flux of Earths volcanic emission rate (31015 m2

s1; solid lines), CO2 emission rate 100 times higher than Earthsrate (dashed lines) and CO2 emission rate 100 times lower thanEarths rate (dotted lines). The planet has 10 M and 1.75 Rand is in a 1.6-AU orbit of a sun-like star (with UV adjusted). Thefiducial atmosphere is 90% H2 and 10% N2 by volume, in a 1 baratmosphere with a 290 K surface temperature. The main point isthat the H concentration does not depend on the amount of CO2in the atmosphere, whereas the amount of O is critically controlledby the level of CO2 in the atmosphere. Compared to H, OH isalways a minor constituent in the atmosphere (by a few orders ofmagnitude). As the UV flux increases, more of the destructive,reactive species are generated.

forming H from H2 (equation (11)) illustrates the impor-tance of water vapor. Water is needed to form H in thefirst place, in this case.The second major point is that the reason H can ac-

cumulate to high concentrations is because the H + H +M reaction rate that removes the H atoms is relativelyslow. The rates are

K = 2.8 1018 exp(1800.0/T ) [m3s1], (15)

Km = 6.64 1039(T/298.0)1 N [m3s1], (16)

J 106 [s1], (17)

where N is the number density of species M in units ofmolecules m3. The rates are from Sander et al. (2011).The third major point is that in the H2-rich atmo-

sphere the OH concentration is low, because [OH] reactswith H2 to recombine to H2O.H is produced by photodissociation of water vapor and

not predominantly by the direct photodissociation of H2.The reason is that the photons with high enough energy( < 85 nm; Mentall & Gentieu 1970) to photodissociateH2 are not available. The high-energy photons that coulddissociate H2 directly are absorbed at the pressure lev-els of nanobars by H2 itself, and the photons that could

7

penetrate down to pressure levels relevant to observations(0.1 mbar to 1 bar) are those that can dissociate water.Photodissociation of H2O, in comparison, is caused byphotons of lower energy ( 240 nm; Banks & Kockarts1973), photons which can penetrate more deeply in theatmosphere than the ones that photodissociate H2. Wenote that like other photochemical products, H is formedprimarily above the mbar level, before all of the photodis-sociating stellar photons are absorbed.The H concentration is dependent on the stellar UV

levels and presence of H2O. Low-UV environments arefavorable for biosignature build up, since the initial pho-tolysis that starts the OH formation chain will be weaker.A similar situation is described for oxidized atmospheresin Segura et al. (2005).We must beware that for some molecules, in some

situations, atomic O will be the dominant destructivespecies. There is no simple model (as in the above equa-tions) but with our full simulation we find in atmosphereswith high CO2 atomic O will abundant (see Figure 2).The key point is that reaction rates with O are fasterthan reaction rates with H for some molecules (see Ta-ble 1).There is a very important point of comparison between

the dominant reactive species, H in H2-rich atmospheresand OH in oxidized atmospheres. The concentrations ofH and OH in the two different types of atmospheres vary(see Figure 3 and Table 1, as well as more generally Table4 in Hu et al. (2012)). This can be understood qualita-tively because OH is much more reactive than H. OHwill react faster with any atmospheric component thanH, and so, for the same impinging stellar UV flux, OHwill build up to a lower atmospheric concentration thanH. The rate of removal of a biosignature gas by H or OHis a product of the concentration and the reactivity. OH,with lower concentration but greater reactivity will re-move a biosignature gas at a similar rate to H, which hasa greater concentration but a lower reactivity. In otherwords, while the mechanism of chemistry clearance andthe end products are different, the loss rate is fairly sim-ilar. For more details of the formation and destructionof the reactive species H, OH, and O in reduced and ox-idized atmospheres, see Hu et al. (2012).

4. RESULTS: POTENTIAL AND UNLIKELYBIOSIGNATURE GASES

We now turn to describe the potential and unlikelybiosignature gases in an H2 atmosphere by their biosig-nature category. The biosignature categories developedin Seager et al. (2013) and summarized in 2.2 are an es-sential aide for calculations because of the common for-mation pathways that belong to each biosignature class.We consider a planet with 10M, 1.75 R, an atmo-

sphere with 90% H2 and 10% N2 by volume. The atmo-sphere scenario is the hydrogen-rich case among the exo-planet benchmark scenarios detailed in Hu et al. (2012),and we here outline the key specifics. The planet surfacepressure is 1 bar and the planet surface temperature is290 K. The temperature drops with increasing altitudeaccording to an adiabatic lapse rate, until reaching 160K and is prescribed as constant above. The semi-majoraxis of the planets orbit is 1.4 AU if orbiting a sun-likestar, 0.037 AU if orbiting an M5V dwarf star, this is theconsistent planet-star separation given the atmospheric

Mix

ing

Ra

tio

1018

1017

1016

1015

1014

1013

1012

1011

1010

109

108

107

106

H

OHO

UV Flux [Solar UV Flux]

Ch

em

ica

l R

em

ova

l R

ate

fo

r C

H3C

l [m

s

1]

103

102

101

100

101

102

10310

1210

1110

1010

910

810

710

610

510

410

310

210

110

010

1

H

OHO

Figure 3. Destructive power of reactive species (H, OH, and O) ina reduced atmosphere. The atmosphere considered has 90% H2 and10% N2 by volume, with CO2, CH4, SO2, and H2S emission fromits surface, for a 1 bar atmosphere on a planet with 10 M and1.75 R. Shown for comparison, are cases for an N2-dominatedatmosphere (diamond markers) and Earths current atmosphere(circular markers). Top panel: Mixing ratios of H, OH, and O asa function of UV flux. The mixing ratio of H exceeds the otherreactive species OH and O. Bottom panel: The column-integratedchemical removal rates as a result of reactions with H, O, and OH,for which we have used CH3Cl as an example. The removal ratesare scaled by the steady-state mixing ratio of CH3Cl to have adimension of velocity. This panel shows that removal by H is thedominant loss rate, and that the loss rates scale approximatelylinearly with UV flux incoming to the exoplanet atmosphere.

composition and the required surface temperature. Theeddy diffusion coefficients are scaled up by a factor of 6.3from those measured in Earths atmosphere, in order toaccount for the difference in the mean molecular mass.Important minor gases considered are H2O (evaporatedfrom a liquid water ocean), CO2 (about 100 ppm), andCH4 and H2S (emitted from surface). Deposition veloc-ities of H2, CH4 are assumed to be zero, the depositionvelocity of CO is 1010 m s1, and deposition velocities ofoxidants including O2, O3, and H2O2 and sulfur speciesare assumed to be the values as on Earth. See Hu et al.(2012) for the rationale for these specifics and for the de-scription of carbon, oxygen, and sulfur chemistry in suchan H2-dominated atmosphere.The amount of UV flux on the planet from the star

is critical to destroying biosignature gases and so we

8

Reaction A n E T = 270 K T = 370 K T = 470 KDMS + H CH3SH + CH3 4.81 1018 1.70 9.00 2.63 1024 2.11 1022 2.91 1021

CH3Cl + H CH3 + HCl 1.83 1017 0 19.29 1.97 1024 1.46 1022 1.92 1021

CH3Br + H CH3 + HBr 8.49 1017 0 24.44 1.59 1021 3.01 1020 1.63 1019

CH3I + H CH3 + HI 2.74 1017 1.66 2.49 7.67 1018 1.75 1017 3.09 1017

DMS + OH CH3SCH2 + H2O 1.13 1017 0 2.10 4.43 1018 5.71 1018 6.60 1018

CH3Cl + OH CH2Cl + H2O 1.40 1018 1.60 8.65 2.54 1020 1.89 1019 3.17 1019

CH3Br + OH CH2Br + H2O 2.08 1019 1.30 4.16 2.87 1020 7.13 1020 1.30 1019

CH3I + OH CH2I + H2O 3.10 1018 0 9.31 4.90 1020 1.50 1019 2.86 1019

DMS + O CH3SO + CH3 1.30 1017 0 -3.40 5.91 1017 3.93 1017 3.10 1017

CH3Cl + O CH2Cl + OH 1.74 1017 0 28.68 1.77 1023 8.77 1022 8.26 1021

CH3Br + O CH2Br + OH 2.21 1017 0 30.76 1.77 1023 8.77 1022 8.26 1021

CH3I + O CH3 + IO 6.19 1018 0 -2.84 2.19 1017 1.56 1017 1.28 1017

Table 1Reaction rates with H, OH, and O of select Type III biosignature gases. Second order reaction rates in units of m3 molecule1 s1 are

computed from the formula k(T ) = A(T/298)n exp(E/RT ) where T is the temperature in K and R is the gas constant(R = 8.314472 103 kJ mole1). The reactions rate are compiled from the NIST Chemical Kinetics Database.

consider the same planet orbiting three different startypes. The first star type is a sun-like star. The sec-ond star type is a weakly active 0.2 R M5V dwarfstar, with EUV taken as that expected for GJ 1214b(France et al. 2013). The third star type is a quiescentM star with no chromospheric and only photospheric ra-diation, again an 0.2 R M5V dwarf star (see 5.6). UVradiation received by the planet is scaled according to thesemi-major axis, and the stellar UV spectra are from:the Air Mass Zero reference spectrum for the sun-likestar (http://rredc.nrel.gov/solar/spectra/am0/); the UVfluxes are from France et al. (2013) for the weakly activeM dwarf star (using the values for GJ 1214b), and fromsimulated spectra of cool stars (Allard et al. 1997) forthe UV quiet M dwarf star (see Figure 1).Whether or not a biosignature gas is detectable can be

technique and spectral feature dependent. The requiredatmospheric concentration depends on the strength ofa given absorption feature, and different techniques aresensitive to different wavelength ranges. For example,thermal emission detection sensitivity follows the plane-tary thermal emission flux (approximately a black bodypeaking in the mid-IR), whereas the transmission spectrasensitivity in the infrared follows the thermal emissionflux of the star (approximately a black body). An illus-trative example is NH3 with a strong absorption featureat 10.3-10.7 m suitable for planetary thermal emissionbut for transmission a weaker absorption feature at 2.8-3.2 m is more easily detected than the 10 m featurebecause of overall photon fluxes of the star. For trans-mission spectra we avoid consideration of sun-like starsbecause the observational signal (the planet atmosphere-star area ratio) is too low (e.g., Kaltenegger & Traub2009).Biosignature gas results are summarized for thermal

emission detectability (for sun-like and M dwarf stars)in Table 2 and for transmission spectra detectability (forM dwarf stars only) in Table 3. Select promising biosig-nature gases are shown via their thermal emission spectrafor a variety of atmospheres for intercomparison: CH3Cl(Figure 4); DMS (Figure 5); N2O (Figure 6); NH3 (seeSeager et al. (2013) Figure 2), and via their transmissionspectra for H2-dominated atmospheres (Figure 7).

4.1. Type I Biosignature Gases: Fully Reduced Forms

We start by focusing on the Type I biosignature gases,gases generated by reactions that extract energy from ex-ternal, environmental redox gradients. The most likelyType I metabolic product in an H2-rich atmospherewould be those in which non-hydrogen elements are intheir most hydrogenated form4. In a reducing environ-ment, life captures chemical energy by reducing envi-ronmental chemicals. In the presence of excess hydro-gen, the most energy life could extract from chemicalpotential energy gradients would be from converting ele-ments from relatively oxidized compounds to their fullyreduced form. An additional reason for focusing on TypeI biosignature gas products that are in their most hydro-genated from is that they are likely long lived in an H2-dominated atmosphere, because molecules in their mosthydrogenated form cannot undergo any further reactionswith H.

4.1.1. Type I Biosignature Gas Overview

The most reduced form of the most abundant non-metal elements, C, N, O, P, S, H, Si, F, and Cl areCH4, NH3, H2O, PH3, H2S, H2, SiH4, HCl, and HF. Themost promising Type I biosignature gas is NH3, whichis further described below (4.1.2). The other reducedmolecules are unlikely biosignature gases for a variety ofreasons. Some (PH3, SiH4), require energy input to makethe reduced product from geologically available materi-als, and so would not be produced by Type I biosignaturegas reactions. Some are always present in their most re-duced form and so life cannot reduce them further (F,Cl). H2S and CH4 are not viable for reasons discussedbelow, largely because geological and biologically sourcescannot be discriminated between. In the case of H2 andH2O, they are naturally present in an H2-dominated at-mosphere at relevant potentially-habitable planet tem-peratures.An aside about PH3. We note that trace amounts

of phosphine is produced by some anaerobic ecologieson Earth (Glindemann et al. 2005). It is controver-sial whether the microorganisms in these environmentsare making PH3, or whether the bacteria are makingacid which is attacking environmental iron that con-

4 Life might produce molecules with elements in intermediateredox states as life does on Earth. In an H2 atmosphere suchmolecules are likely to be photochemically hydrogenated.

http://rredc.nrel.gov/solar/spectra/am0/

9

Molecule Mixing Wave- Surf. Flux Biomass Surf. Flux Biomass Surf. Flux Biomass DominantRatio length Sun-Like Estimate Active M Estimate Quiet M Estimate Removal[ppm] [m] [m2 s1] [g m2] [m2 s1] [g m2] [m2 s1] [g m2] Path

Type I

NH3 0.10 10.3-10.8 2.41015 4.0104 5.11014 8.0106 8.2105 9.5106 photolysis

Type III

CH3Cl 9.0 13.0-14.2 1.01017 2.8103 2.91015 77 4.71011 0.013 HDMS 0.10 2.2-2.8 4.21019 190 1.81019 82 2.41013 1.1104 OCS2 0.59 6.3-6.9 8.71017 5.5107 3.61017 2.3107 5.91011 37 OOCS 0.10 4.7-5.1 2.51015 1.3105 1.01014 5.5103 1.31010 0.67 HN2O 0.38 7.5-9.0 3.81015 5.41014 1.31011 photolysis

Table 2Results for thermal emission spectra. Potential biosignature gas required concentrations, related required biosignature gas surface fluxes(in units of molecules m2 s1), estimated biomass surface densities, and the dominant removal path or destructive species. Results aregiven for three cases: for a planet orbiting a sun-like star, a weakly active M5V dwarf star (denotedActive M) and a quiescent M5V

dwarf star (denoted Quiet M). The planet considered has 10 M, 1.75 R, an atmosphere with 90% H2 and 10% N2 by volume, with asurface temperature of 290 K and a surface pressure of 1 bar. Note that compounds with the removal path dominated by O, the required

surface flux sensitively depends on the CO2 emission/deposition.

Molecule Mixing Wave- Surf. Flux Biomass Surf. Flux Biomass DominantRatio length Active M Estimate Quiet M Estimate Removal[ppm] [m] [m2 s1] [g m2] [m2 s1] [g m2] Path

Type I

NH3 11 2.8-3.2 5.51016 1.1 8.8107 1.8109 photolysis

Type III

CH3Cl 10 3.2-3.4 3.21015 8.6102 5.21011 1.4102 HDMS 0.32 3.1-3.6 5.81019 2.6102 7.81013 3.6104 OCS2 0.38 6.4-6.9 2.31017 1.5107 3.81011 24 OOCS 1.8 4.7-5.1 1.91015 9.9104 2.31011 12 HN2O 11 3.8-4.1 4.81015 3.71012 photolysis

Table 3Results for transmission spectra. Potential biosignature gas required concentrations, related required biosignature gas surface fluxes (inunits of molecules m2 s1), estimated biomass surface densities, and the dominant removal path or destructive species. Results aregiven both for two cases, a planet orbiting a weakly active M5V dwarf star (denoted Active M) and a quiescent M5V dwarf star

(denoted Quiet M). The planet considered has 10 M, 1.75 R, an atmosphere with 90% H2 and 10% N2 by volume, with a surfacetemperature of 290 K and a surface pressure of 1 bar. Note that a planet orbiting a sun-like star is not considered for transmission

spectra because the overall detection signal is too low because of the small planet atmosphere annulus area to sun-like star area. Notethat compounds with the removal path dominated by O, the required surface flux sensitively depends on the CO2 emission/deposition.

tains traces of phosphide, and this attack is making thephosphine gas (Roels & Verstraete 2001). Phosphine isa potential biosignature in other highly reduced envi-ronments. Phosphine is reactive and thermodynami-cally disfavored over elemental phosphorus and hydrogenat Earth surface pressure and temperature. Phosphinemight be a Type I biosignature gas under conditions ofvery high H2 pressure, which would favour productionof PH3 over elemental phosphorus. Phosphine could bealso be produced as a Type III biosignature gas, analo-gous to reactive signaling molecules such as NO or C2H4on Earth.

4.1.2. NH3 as the Strongest Candidate Biosignature Gas inan H2 Atmosphere

NH3 is the strongest candidate biosignature gas in athin, H2 atmosphere because, like O2 in Earths atmo-sphere, there is no plausible geological or photochemicalmechanism for producing high concentrations on rockyplanets with thin atmospheres (but c.f. the false posi-tive discussion below). NH3 is readily photolyzed in theupper atmosphere to yield N2 and in volcanic gases isthermally broken down at high temperatures. The triplebond of N2 makes it extremely kinetically stable and soany N in the atmosphere ends up being trapped as N2.

We have therefore proposed NH3 as a biosignature gasin an H2-rich atmosphere (Seager et al. 2013). NH3 isa good biosignature gas candidate for any thin H2-richexoplanet atmosphere because of its short lifetime andlack of geological production sources. NH3 as a biosig-nature gas is a new idea, and one that is specific to anon-Earth-like planet. On Earth, NH3 is not a usefulbiosignature gas because, as a highly valuable moleculefor life that is produced in only small quantities, it israpidly depleted by life and is unable to accumulate inthe atmosphere. NH3 is also a very poor biosignature gason Earth because it is very soluble, so the trace amountsproduced will stay dissolved in water and not escape tothe atmosphere.The summary of the biosignature gas idea is that NH3

would be produced from hydrogen and nitrogen, in anatmosphere rich in both,

3H2 +N2 2NH3. (18)

This is an exothermic reaction which could be used tocapture energy. The industrial version of this reaction iscalled the Haber process for ammonia production at hightemperatures; hence we call such a planet a cold Haberworld. We proposed that in an H2-rich atmosphere, lifecould find a way to catalyze the breaking of the N2 triple

10

bond and the H2 bond to produce NH3, and capture theenergy released. In contrast, life on Earth solely fixesnitrogen in an energy-requiring process. Energy capturewould yield an excess of NH3 over that needed by lifeto build biomass, and so the excess would accumulate inthe atmosphere. Is a cold Haber World possible? We be-lieve yes, based on synthetic chemistry on Earth that cancatalyze the breakage of each of H2 (Nishibayashi et al.1998) and N2 bonds (Yandulov & Schrock 2003; Schrock2011) at Earths surface pressure and temperature; whatis not yet known is a catalytic system that can breakboth at once.We showed in Seager et al. (2013) that for an Earth-

size, Earth-mass planet with a 1 bar atmosphere of 75%by volume N2 and 25% by volume H2 (including carbonspecies via a CO2 emission flux), a potentially detectableNH3 atmosphere concentration of 0.1 ppm is sustainableby a very reasonable biomass surface density of 9 102

g m2. This modest surface density corresponds to alayer less than one bacterial cell thick. For compari-son, the phytoplankton that are the major contributor toEarths oxygen atmosphere are present in Earths oceansat around 10 g m2. For interest, we note that standardprinter paper is between 80 and 100 g m2.For an H2-dominated atmosphere with 90% H2 and

10% N2 on a planet with 10 M, 1.75 R orbiting a sun-like star, but all other parameters the same as the above,the viability of NH3 as a biosignature gas in a thermalemission spectrum still holds based on a physically rea-sonable biomass surface density. We now describe the es-timate for the biomass surface density, using the Type Ibiomass equation (equation (4)). We use the NH3 sourceflux of 2.41015 molecule m2 s1 (see Table 2). To com-pute G we used T = 290 K, and reactant and productconcentrations at the surface in terms of partial pres-sures of N2 = 0.1, H2 = 0.9, NH3 = 1.4 10

7, givingG = 85.6 kJ mole1. Pme = 7.0 10

6 kJ g1 s1,we find a biomass surface density of 4.9 102 g m2.Based on a reasonable biomass surface density, we there-fore consider the NH3 production flux to be viable inour Haber World scenario. The global annual biogenicNH3 surface emission in the Haber World would be about1100 Tg yr1. This is much higher than the Earthsnatural NH3 emission at 10 Tg yr

1 (Seinfeld & Pandis2000). Comparing NH3 production on the Haber worldand on Earth, however, is not valid. We are postulatingthat production of NH3 on the Haber world is a majorsource of metabolic energy for life. A better emission ratecomparison is to the biosignature gas O2 from Earthsprinciple energy metabolism, photosynthesis. Earthsglobal oxygen flux is 200 times larger than the HaberWorlds NH3 surface emission, at about 2 10

5 Tg yr1

(Friend et al. 2009).Turning to a weakly active M5V dwarf star, for the

same fiducial planet, the NH3 surface flux required tosustain a detectable level of atmospheric NH3 in a ther-mal emission spectrum is 5.1 1014 molecule m2 s1

(see Table 2). This value is about 5 times lower than thesun-like star example above and therefore converts intoa biomass estimate of about 5 times smaller than thesun-like star example above, or about 1 102 g m2,due to the linear scaling of the problem. For a transmis-sion spectra measurement for the same planet orbiting

the same weakly active M dwarf star, the optimal wave-length range for detection is 2.83.2 m, the requiredconcentration is 11 ppm, and the required surface sourceflux is 5.5 1016 molecule m2 s1, resulting in a sur-face biomass of 1 g m2. For this particular example, NH3in transmission vs. thermal emission, it is more difficultto detect NH3 and hence a higher biomass is actuallyrequired for the same hypothetical type of life.We emphasize that the NH3 biosignature gas concept is

not changed for a planet with a massive (yet still thin)atmosphere with high surface pressure. As long as thesurface conditions are suitable for liquid water, NH3 willnot be created by uncatalyzed chemical reactions.NH3 is not immune to false postives. Although a rocky

planet with a thin H2-dominated atmosphere is unlikelyto have an NH3 false positive, the challenge is in iden-tifying the planetary (and stellar) characteristics. Wedescribe three scenarios that could lead to the nonbio-logical production of NH3.A rocky world with a hot surface of 820 K could

generated NH3 by the conventional Haber process if thereis surface iron. Such a hot surface temperature couldpresumably be ruled out from other observations.A second scenario where NH3 is naturally occuring is in

the atmospheres of gas giant planets or the so-called miniNeptunes. The deep atmosphere may reach conditionswhere NH3 can be formed kinetically at the extremelyhigh pressures necessary for NH3 formation to be pos-sible thermodynamically. On Jupiter, for example, theH2 +N2 NH3 reaction becomes significant in compar-ison with vertical transport at about 1500 K, 1400 bar(Prinn & Olaguer 1981). The only way we can discrim-inate between planets with a massive envelope and arocky planet with a thin atmosphere where the pres-sures for the thermodynamic formation of NH3 are notreached, is with high-resolution spectra to assess the sur-face pressure (Benneke & Seager 2012, 2013).A third scenario for an NH3 false positive is for planets

with outgassed NH3 during evolution. The importance ofammonia for the atmospheric evolution of Titan relatesto primordial ammonia which accreted with the ices ofthe moon and has not subsequently been broken downeither by internal heat (likely on a rocky planet) or byexternal UV photolysis (which will rapidly break downany NH3 in the atmosphere) (Shin et al. 2012). In thiscase, ammonia is therefore present as ice in the interior.This would be a challenging case to ascertain, and illus-trates of how an assignment of any gas as a biosignaturegas candidate has to be given a detailed probabilistic as-sessment based on what we know about the planet con-cerned.For any case, a quiescent M star with no chromospheric

UV emissionhence a planet with little to no destruc-tive UV fluxNH3 can easily accumulate in the planetatmosphere and act as a signficant false positive. NH3is destroyed by photolysis and is very sensitive to theamount of UV radiation.

4.1.3. CH4 and H2S as Unlikely Biosignature Gases

CH4 has been described at length as a possible biosig-nature gas on early Earth and on exoplanets in ox-idized atmospheres (e.g., Hitchcock & Lovelock 1967;Des Marais et al. 2002). This is despite the risk of a geo-logically derived false positive, because it is believed that

11

in an oxidized environment geological production of CH4will be small, and so if enough CH4 is produced it maybe attributed to life. This is the case on Earth where atleast 99% of the atmospheric CH4 derives directly fromlife or from industrial destruction of fossil hydrocarbonsformed from past life (Wang et al. 2004). However, the1775 ppb concentration of CH4 in Earths atmosphere(Solomon et al. 2007) is not enough to be detected re-motely with envisioned space telescope capabilities.CH4 is a poor biosignature gas in an H2-rich atmo-

sphere because it is both produced volcancially and is anend product of CO2 photochemistry in the atmosphere.Terrestrial volcanic emission rates of CH4 and CO2 wouldlead to substantial build-up of CH4 in H2 dominated at-mospheres. Even small amounts of outgassed CO2 willlead to an accumulation of CH4 in the atmosphere, be-cause CH4 has a very long lifetime in an H2-rich atmo-sphere and CH4 would be produced by

4H2 +CO2 CH4 + 2H2O. (19)

More specifically, considering Earths volcanic emissionrates of CH4 and CO2, and with deposition velocities of106 m s1 for CO2 and zero m

2 s1 for CH4, CH4 willaccumulate up to 10 ppm in a 1 bar 90% H2, 10% N2atmosphere with a temperature profile similar to Earths(Hu et al. 2012). Even in the case of no surface CH4emission, CO2 emission into the same atmosphere wouldlead to the atmospheric production and accumulation ofCH4 up to 5 ppm. This example is intended to show thatthe false positive risk of CH4 is so high in a H2-dominatedatmosphere as to make CH4 an implausible biosignaturegas.H2S is even more unfavorable than CH4 as a biosig-

nature gas in an H2-rich atmosphere because of thesame geological false positive issues as with carbon gases.An added problem is the generation of aerosols whichmay blanket any spectral features and the fact that theH2S spectral features are heavily contaminated by atmo-spheric water vapor making them potentially difficult todetect (Hu et al. 2013).

4.2. Type II Biosignature Gases: No ViableBiosignature Gases

Type II biosignature gases are those produced bymetabolic reactions for biomass building. Biomass build-ing on Earth primarily occurs by photosynthesis, whichhas the dual goal of harvesting light energy to use formetabolism and also for capturing carbon for biomassbuilding.We have not identified any useful biosignature gases of

Type II in an H2-rich, 1 bar atmosphere. Photosynthesisin a reduced environment such as an H2-dominated atmo-sphere would generate reduced byproduct gases, whichare not useful as biosignature gases because those speciesare already expected to be present in their most reducedforms in the H2-dominated atmosphere.The concept of photosynthesis on a planet with an

H2-dominated atmosphere is nonetheless worth some dis-cussion5, starting with a brief review of photosynthesis

5 see AbSciCon 2008 abstract by N. Sleephttp://online.liebertpub.com/doi/pdf/10.1089/ast.2008.1246,and Pierrehumbert & Gaidos (2011) for a discussion of photosyn-thetic active radiation that reaches the surface under thick H2

in the familiar Earth environment. Photosynthesis mustconvert carbon from its environmental form, which is theform most thermodynamically stable at surface tempera-tures and pressures, into biomass. Biomass is of interme-diate redox state (Bains & Seager 2012). The key pointtherefore is that in an oxidized environment like Earth,photosynthesis must reduce oxidized carbon (CO2) andwill generate an oxidized byproduct. On Earth environ-mental carbon is captured in photosynthesis, producingO2 as a byproduct,

H2O+CO2 CH2O+O2. (20)

Here CH2O represents biomass.Photosynthesis, by definition, will have the same goal

in an H2-rich environment as in an oxidized environ-ment; to harvest light energy and to build carbon-basedbiomass. Because CH4 is the most thermodynamicallystable gaseous form of carbon in this environment, pho-tosynthesis would oxidize the carbon in CH4 and producea reduced byproduct. The lowest energy route is to di-rectly split the CH4, as,

CH4 +H2O+X CH2O+XH, (21)

where X is an atom that is oxidized in the environment,and has been reduced to XH, consuming energy in theprocess, and CH2O again represents biomass. (We notethat the oxidation state of the oxygen is not changed inthis process, unlike oxygenic photosynthesis on Earth, soformally this is not splitting water even though water isinvolved.)The null result for biosignature photosynthetic biosig-

natures on an H2-dominated atmosphere is based on thepoint that most non-metals (C, O, S, the halogens) arelikely to be in their most reduced state already on thesurface of this world, and so cannot play the role of X inthe above described photosynthesis process.One exception might have been hydrogen, which is ox-

idized in water and methane, and so a possible photo-synthetic reaction is

CH4 +H2O CH2O+ 2H2, (22)

but again H2 is not a useful biosignature gas because itis already present in the H2-dominated atmosphere.For completeness, we describe some other unlikely but

interesting possibilites for X and XH. Silicon, phosphorusand boron are likely to be present as the oxidized min-erals silicates, phosphates and borates respectively, butusing these as a sink for the electrons in photosynthesis,for example in the reaction with silica to generate silane,

CH4 +1

2SiO2 CH2O+

1

2SiH4, (23)

requires more energy than the reaction in equation (22)under a range of conditions, and so would represent avery inefficient way of generating biomass.Reduction of a metal with a positive electrochemical

potential would be more energetically efficient, as for ex-ample in the reduction of copper oxide to copper,

CH4 + 2Cu(O) CH2O+ 2Cu + H2O, (24)

but produces no volatile product, and is dependent ona supply of oxidized metal. (There are clear parallels

atmospheres.

http://online.liebertpub.com/doi/pdf/10.1089/ast.2008.1246

12

with anoxygenic photosynthesis on Earth for this type ofreaction.) By contrast, the reaction in equation (22) islimited only by the supply of methane, as life in wateris not limited by the chemical availability of water. Insummary, photosynthesis in the reducing environmentwill either generate H2, which will not be detectable ina hydrogen-dominated atmosphere, or will produce non-volatile products, i.e. not products not in gas form whichby definition will not be detectable as atmospheric gases.

4.3. Type III Biosignature Gases are Most Viable inLow-UV Environments

Life produces many molecules for reasons that are notrelated to the generation of energy, which we refer to asType III biosignature gases. The gases are produced forreasons such as stress, signaling, and other physiologicalfunctions, and some of these have already been discussedquantitatively in detail as biosignature gases in oxidizedatmospheres, (e.g., CH3Cl (Segura et al. 2005) and DMSand other sulfur compounds (Domagal-Goldman et al.2011)).The fate of Type III biosignature gas molecules de-

pends on the level of relevant reactive species in the at-mosphere, and hence on stellar UV flux. In low UV en-vironments, some Type III gases can accumulate to de-tectable levels. In the relatively high UV environments ofsun-like stars, in an H2-rich atmosphere, many Type IIIgases could be rapidly driven to their most hydrogenatedform, and in some cases will not accumulate to detectablelevels unless we assume unrealistic production rates. Inthese extreme cases in a high UV environment, we wouldonly be able to infer the presence of the biosignature gasby detecting the end-product of photochemical attack,which we call a bioindicator. Only in a few cases mightbioindicators be useful, because many are not spectro-scopically active (and hence not detectable) and othersare indistinguishable from geological cases as well (e.g.,DMS will end up as CH4 and H2S, and N2O will end upas N2 and H2O.)We now show that Type III biosignature gas survival

and hence plausibility depends highly on the UV fluxlevel of the host star. We consider the three fiducialstar types that differ in UV radiation levels: the sun-likestar; the weakly active M5V dwarf star; and the quies-cent M5V dwarf star (Figure 1). We consider the samemodel planet as above, a 10 M, 1.75 R planet with anatmosphere with 90% H2 and 10% N2 by volume, witha surface temperature of 290 K and a surface pressureof 1 bar. Results for the cases we modeled are listed forthermal emission spectra in Table 2 and for transmissionspectra in Table 3.Our first example of a Type III biosignature gas is

methyl chloride (CH3Cl). CH3Cl is produced in traceamounts by many microorganisms on Earth. The de-tectability of CH3Cl in Earth-like atmospheres in thelow UV environment of UV quiet M stars has alreadybeen studied by Segura et al. (2005) and later as a po-tential biosignature gas in more generalized oxidized at-mospheres by Seager et al. (2013).Here for the first time we study CH3Cl as a poten-

tial biosignature gas in a thin H2-rich atmosphere. Forthis, we go beyond previous work not only by consid-ering an H2 atmosphere but also by using our biomassestimate framework so as not to be constrained by terres-

trial bioflux production rates. We now show why CH3Clis a potential biosignature gas in H2 rich atmospheres inlow UV environmentsbecause the amount of biomassto generate a detectable concentration of CH3Cl is phys-ically plausible. We use our biomass estimate framework(2.2 and Seager et al. (2013)).Considering the thermal emission spectrum for our

fiducial planet with a 1 bar atmosphere of 90% H2 and10% N2, a spectral signature of 9 ppm is required forspectral detection using our detection metric. This state-ment is for a spectral band feature in absorption at 13.0-14.2 m (see Figure 4); this is the band accessible inan H2 atmosphere, weaker than the 6.6-7.6 m bandthat would be masked by H2-H2 collision-induced absorp-tion. In order to sustain an atmospheric concentrationof 9 ppm of CH3Cl on our model planet in the habitablezone for a sun-like, weakly active M5V dwarf star, andUV quiet M5V dwarf star, the surface bioflux produc-tion rate would need to be 1.0 1017 molecule m2 s1

(1.7 107 mole m2 s1), 2.9 1015 molecule m2 s1

(4.8 109 mole m2 s1), 4.7 1011 molecule m2 s1

(7.8 1013 mole m2 s1), respectively. Estimatingthe biomass with equation (6) and with the lab rateat 6.17 1011 mole g1 s1 (see Seager et al. 2013),the biomass surface density would need to be about3000 g m2, 80 g m2, 0.001 g m2 for each star-type re-spectively. A globally averaged density of 3000 g m2 islikely too high, one of 80 g m2 is high but not impossi-ble, according to terrestrial biodensities (see Seager et al.2013).The results show that CH3Cl is a more viable biosig-

nature gas in low-UV as compared to high-UV environ-ments. We emphasize that although our estimates ofbiomass surface density for Type III biosignature pro-duction are approximate, the resulting trend is robust.For a spectral detection in transmission for our fidu-

cial Earth transiting an M5V star, the required con-centration is about 10 ppm in the wavelength range3.2-3.4 m. The surface bioflux and biomass esti-mates for a weakly active and quiet star respectivelyare 3.2 1015 molecule m2 s1 and 900 g m2 and5.2 1011 molecule m2 s1 and 0.001 g m2. The re-quired biomass surface density for the weakly active Mstar is higher than the average surface biomass in Earthsoceans and the biomass surface density for the quiet M5Vdwarf star is much lower than Earths and very plausible,again emphasizing the trend that low-UV radiation en-vironments are more favorable for Type III biosignaturegas accumulation.The different values for biosignature gas surface flux

and for the biomass estimates for transmission spectraas compared to thermal emission spectra are in generaldue to either or both of longer atmospheric pathlengthsand different favorable wavelengths (depending on themolecule of interest).The fate of CH3Cl in its destruction by H, is to end

up in its fully hydrogenated form, HCl, with the overallreaction as,

CH3Cl + H2 CH4 +HCl. (25)

HCl could be a bioindicator. The HCl molecule is stableto further photochemistry, because if it is photolyzed,the Cl atoms generated will be predominately react with

13

H2 Atmospheres

180

200

220

240

260

280

300

Brig

htn

ess T

em

pe

ratu

re [K

]

N2 Atmospheres

2 3 4 5 8 10 20 30 50 80 100

160

180

200

220

240

260

280

300

Wavelength [microns]

Brig

htn

ess T

em

pe

ratu

re [K

]

CO2 Atmospheres

CO2

CO2

CO2

CO2

CO2

H2O

H2O

H2O

CO2

CO2

CH4

CH4

CH4

H2 - H

2

H2O

H2O

H2O

CH3Cl

CH3Cl

CH3Cl

CH3Cl

CH3Cl

0 ppm

0.5 ppm

5 ppm

50 ppm

500 ppm

150

200

250

300B

rig

htn

ess T

em

pe

ratu

re [K

]

Figure 4. Theoretical infrared thermal emission spectra of a superEarth exoplanet with various levels of atmospheric CH3Cl in a1 bar atmosphere with a surface temperature of 290 K for a planetwith 10 M and 1.75 R. From top to bottom, the panels show thespectra of H2-, N2-, and CO2-dominated atmospheres, respectively,and the detailed compositions of these reference atmospheres aredescribed in 4 for the H2-dominated planet and in Hu et al. (2012)for the N2 and CO2 dominated cases. We find that over 5 ppm ofCH3Cl is required for detection via thermal emission for H2-, N2-,and CO2-dominated atmospheres.

H to reform HCl. HCl would not be expected to bepresent in significant levels at atmospheric altitudes forspectral detection without life taking non-volatile formsof Cl and putting it into the atmosphere, because all ge-ological sources are non-volatile chlorides (such as NaCl)and any HCl that is volcanically released would be effi-ciently rained out of the troposphere. The limiting prob-lem is that the HCl spectral features are too weak to bedetectable and are likely to be contaminated by CH4 inthe 3 to 4 m range.As a second Type III biosignature gas exam-

ple we consider dimethyl sulfide (DMS). DMS hasbeen studied before in oxidizing atmospheres byDomagal-Goldman et al. (2011) who concluded thatDMS itself is not a potentially detectable biosignaturegas in oxidized atmospheres under sun-like UV radiation,but one of its photolytic breakdown products ethane isdetectable (we call this a bioindicator gas). Using thesame atmosphere and framework as the above CH3Clexample, for thermal emission spectra we find a mix-ing ratio required for detection of 0.1 ppm in the 2.2-

2.8 m band (see Figure 5). Via photochemistry, thismixing ratio corresponds to a surface flux in our fidu-cial H2-dominated atmosphere for a sun-like star, weaklyactive M5V dwarf star, and quiet M5V dwarf star as4.21019 molecules m2 s1 (6.9105 moles m2 s1),1.81019 molecules m2 s1 (3.0105 moles m2 s1),2.41013 molecules m2 s1 (4.11011 moles m2 s1),respectively. Using a DMS lab production rate of 3.64107 moles g1 s1 (Seager et al. 2013), we come up witham implied biomass surface density estimate for the threestar types of about 200 g m2, 100 g m2, 104 g m2,respectively. The first two values are high, but physicallyplausible as compared to Earth biomass surface densityranges. For transmission spectra the numbers are abouta factor of two higher for the weakly active and quietM5V dwarf star (see Figure 7 and Table 3).

H2 Atmospheres

180

200

220

240

260

280

300

Brig

htn

ess T

em

pe

ratu

re [K

]

N2 Atmospheres

2 3 4 5 8 10 20 30 50 80 100

160

180

200

220

240

260

280

300

Wavelength [microns]

Brig

htn

ess T

em

pe

ratu

re [K

]

CO2 Atmospheres

H2 - H

2

CO2

CO2

CO2

CO2

CO2

CO2

CO2

H2O

H2O

H2O

CH4

CH4

CH4

H2O

H2O

H2O

DMS DMS DMS

DMS DMS DMS

DMS DMS

0 ppm

0.5 ppm

5 ppm

50 ppm

500 ppm

150

200

250

300

Brig

htn

ess T

em

pe

ratu

re [K

]

Figure 5. Theoretical infrared thermal emission spectra of a superEarth exoplanet with various levels of atmospheric DMS in a 1 baratmosphere with a surface temperature of 290 K for a planet with10 M and 1.75 R. From top to bottom, the panels show thespectra of H2-, N2-, and CO2-dominated atmospheres, respectively,and the detailed compositions of these reference atmospheres aredescribed in 4 for the H2-dominated planet and in Hu et al. (2012)for the N2 and CO2 dominated cases. We find that 0.1 ppm of DMSis required for future detection via thermal emission for H2, N2,and CO2-dominated atmospheres.

The DMS results show again that the lowest UV en-vironment is most favorable. There are two other otherrelevant points related to DMS appearing to be a favor-able biosignature gas in each of the three UV radiation

14

environments studied. The first point is that gases de-stroyed by reaction with O (as opposed to gases destroyedby reactions with H) show a similar surface flux require-ment between the sun-like and weakly active M dwarfstar. This is because the release of O from CO2 photoly-sis is largely driven by Lyman-alpha emission which aresimilar at the habitable zones for the sun-like and weaklyactive M dwarf star used in this study (Figure 1).The second point is that the high Rlab values used for

DMS, and hence the low biomass surface density esti-mates, are a result of the biology of DMS production.On Earth, DMS is the waste product of consumption ofDMSP by marine organisms consuming marine plankton.DMSP is accumulated in large amounts by some marinespecies. Thus organisms that generate DMS do not haveto invest their own resources to make DMS, and so arenot limited to how much they can make. Maximal pro-duction rates are therefore very high. This is discussedfurther in Seager et al. (2013).In terms of a bioindicator, DMS will react with H2 to

generate CH4 and H2S. Neither is a useful bioindicatoras CH4 and H2S are expected to be present in the atmo-sphere naturally. This is in contrast to oxidized atmo-spheres, where ethane may be a bioindicator gas, as theexpected to be the end product of DMS photodestruc-tion by combination of methyl radicals generated fromattack of O on DMS (Domagal-Goldman et al. 2011).As a third and fourth Type III biosignature gas ex-

ample, we used CS2 and OCS. For these two gases wefind the same trend as the other Type III biosignaturegases, that is in a low UV environment the biosignaturegases can accumulate (see Tables 2 and 3). The biomassestimates (as a plausibility check) are too high for thesun-like and weakly active M dwarf star environmentsto be plausible as compared to terrestrial biomass sur-face density values. OCS in the UV environment of aweakly active M dwarf star may be an exception with anestimate at the upper limit of plausibility.As a fifth example we describe N2O. On Earth N2O is a

Type I biosignature gas produced by nitrifying bacteria.N2O is not likely to be produced in a thin H2-rich atmo-sphere because there is unlikely to be much nitrate avail-able. Here we explain further. N2O has been suggestedas a biosignature gas in Earths atmosphere (Segura et al.2005). N2O is a Type I biosignature gas formed by twoprocesses on Earththe oxidation of ammonia by at-mospheric oxygen and the reduction of nitrate in anoxicenvironments,

2NH3 + 2O2 N2O+ 3H2O, (26)

NO3 +H N2O+H2O. (27)

Analogous reactions on a hydrogen-dominated worldwould be the reduction of nitrate by atmospheric hy-drogen

NO3 +H2 N2O+H2O+OH, (28)

or the oxidation of ammonia by a geologically derivedoxidant.Nitrate is formed on Earth by oxidation of NO gener-

ated by lightning in Earths oxygen-rich atmosphere, orby biological processesneither are likely in an H2-richenvironment, so it is not clear whether nitrate reductionis a useful energy source in a world with an atmosphere

rich in H2. Ammonia oxidation requires a strong oxidiz-ing agent, which again is likely to be missing from theenvironment.N2O as a Type I biosignature gas, therefore, seems

unlikely, although not impossible, from very rare envi-ronments in which there are oxidized nitrogen speciesgenerated geochemically.N2O, however, could be a Type III biosignature gas

as NO is for some organisms on Earth. We have calcu-lated the surface fluxes for a detectable amount of N2Oin a thin, H2-dominated atmosphere and find relativelylow required surface fluxes. The reason is that N2O isdestroyed by photodissociation, a slower rate than byreaction with H. N2O may therfore be a plausible biosig-nature gas candidate, even in an atmosphere subject tostrong UV radiation (see Figures 6 and 7 and Tables 2)and 3). A biomass estimate (as a plausibility check) isnot possible for N2O, as it is only known as a Type Ibiosignature on Earth (and so therefore Type III Rlabrates are not available).

5. DISCUSSION

5.1. What Constitutes an H2-Dominated Atmosphere?

We have calculated biosignature gas accumulation inan atmosphere with 90% H2 and 10% N2 by volume. Asuper Earth exoplanet atmosphere can have many othergas species. The concentration of the major destructivespecies, H, O, and OH will depend on the amounts ofthese other gas species.As an example, we explore the changing effect of the

reactive species in an H2-dominated atmosphere for dif-ferent UV flux levels, based on the surface flux levels ofCO2 (Figure 2). A few key points are as follows. TheH abundance is almost not affected by the CO2 mixingratios ranging from 108 to 102. The O abundance de-pends on both CO2 and UV levels, such that both a highCO2 level and a high UV flux lead to high atmospheric O.Only in extreme cases (e.g., H2-dominated atmosphereswith >1% CO2, shown by dashed lines), the abundanceof O may be very close to the abundance of H. The OHabundance depends on a complex source-sink network,ultimately driven by H2O and CO2 photolysis. Notably,the amount of H is always at least 4 orders of magnitudehigher than the amount of OH.The effect of changing the H2 mixing ratio and the

addition and variation of other active gases on the Hconcentration will need to be considered on a case-by-case basis as they will react not only with H and OH butalso with other gas species.

5.2. Can Super Earths Retain H2-DominatedAtmospheres?

Whether or not a super Earth planet can retain H2stably from atmospheric escape is not known. Althoughmany models and studies for exoplanet atmospheric es-cape exist (see e.g., Lammer et al. 2012, and referencestherein), the permanent limitation is that there are toomany unknowns to provide a definitive and quantitativestatement on which planets will retain H2. One of thechallenges is the unknown history and present state ofthe host stars EUV flux. Another major challenge isthe defining the mechanism for atmospheric escape for agiven exoplanet, for example whether or not the regime

15

CO2

CO2

H2O

H2 - H

2

H2O

N2O

N2O N

2O N2O

180

200

220

240

260

280

300

Brig

htn

ess T

em

pe

ratu

re [K

]

N2 AtmospheresCO

2 CO2

H2O

H2O

N2ON

2ON2O

N2O

2 3 4 5 8 10 20 30 50 80 100

160

180

200

220

240

260

280

300

Wavelength [microns]

Brig

htn

ess T

em

pe

ratu

re [K

]

CO2 Atmospheres

CO2H2O

CO2

CO2

CH4

H2O

N2O

N2O

0 ppm

0.5 ppm

5 ppm

50 ppm

500 ppm

0 ppm

0.5 ppm

5 ppm

50 ppm

500 ppm

0 ppm

0.5 ppm

5 ppm

50 ppm

500 ppm

H2 Atmospheres

150

200

250

300B

rig

htn

ess T

em

pe

ratu

re [K

]

Figure 6. Theoretical infrared thermal emission spectra of a superEarth exoplanet with various levels of atmospheric N2O in a 1 baratmosphere with a surface temperature of 290 K for a planet with10 M and 1.75 R. From top to bottom, the panels show thespectra of H2-, N2-, and CO2-dominated atmospheres, respectively,and the detailed compositions of these reference atmospheres aredescribed in 4 for the H2-dominated planet and in Hu et al. (2012)for the N2 and CO2 dominated cases. We find that about 0.4 ppmof N2O is required for future detection via thermal emission forH2, N2, and CO2-dominated atmospheres.

of rapid hydrodynamic escape was reached in a planetshistory or which non-thermal mechanism, if any, cameinto a dominant role (see Table 4.1 and references thereinin Seager 2010). With an unknown initial atmosphericreservoir and an unknown present atmospheric composi-tion, the regime and type of atmospheric escape is diffi-cult to impossible to identify.Some super Earths will have been formed with at-

mospheres with H2, based on both theoretical and ob-servational evidence. Theoretically, planetary buildingblocks containing water-rich minerals that can release H(Elkins-Tanton & Seager 2008; Schaefer & Fegley 2010).Observationally, a large number and variety in radius ofKepler mini-Neptunes that must have H or an H/He en-velope to explain their radii. So either from outgassingor nebular capture of gases, some super Earths shouldhave started out with H2-rich atmospheres and thosewith high enough gravity and low enough temperaturesand/or with magnetic fields should be able to retainthe H2 (e.g., Pierrehumbert & Gaidos 2011). Observa-

tional detection of H2-rich atmospheres will ultimatelybe needed to confirm the scenario of thin H2-dominatedatmospheres on super Earths.

5.3. Upper Temperatures for Life

Super Earths with H2-dominated atmospheres canhave surface temperatures hotter than Earth due toan H2 greenhouse effect from H2-H2 collision-inducedopacities (Borysow 2002; Pierrehumbert & Gaidos 2011).While the hypothetical planets we have described in thispaper were constructed to have 1 bar atmospheres withEarth-type surface temperatures, many H2-dominatedplanet atmospheres are likely to have hotter surface tem-peratures than Earth, even for planets orbiting beyond1 AU of their host stars.An important question for understanding the poten-

tial of biosignature gases on a planet with an H2-dominated atmospheres is therefore, how hot a planetcan be and still sustain life? On Earth, organismsthat grow at 395 K are known (Lovley & Kashefi 2003;Takai et al. 2008) and have been cultured in the labat elevated pressures equal to in situ pressures. Fur-thermore, proteins can function at 410 K to 420 K(Tanaka et al. 2006; Sawano et al. 2007; Unsworth et al.2007) motivating a consensus that life at 420 K is plau-sible (Deming & Baross 1993; Cowan 2004).Life might exist at temperatures even higher than

420 K. The main argument for a maximum tempera-ture for life involves the temperature at which the basicbuilding blocks of life (DNA, proteins, carbohydrates,and lipids) break down. Many of the component chemi-cals of life, including DNA, many of the amino acids thatmake up proteins, and many of the key metabolites thatallow lifes biochemistry to function are rapidly chemi-cally broken down above 470 K (e.g., Cowan 2004) Themaximum temperature at which life could exist thereforemay lie between 420 K and 470K.

5.4. What Surface Pressure is too High?

Many super Earth atmospheres will be much moremassive than the 1 bar atmosphere on Earth. For tem-peratures suitable for the existence of liquid water (see5.3), the surface pressure could be as high as 1000 baror higher (Wagner & Pru 2002). There are three keypoints to show that the high surface pressures does notdestroy the biosignature gases before they can reach thehigh atmosphere.Can life generate potentially detectable biogsignature

gases under a massive atmosphere? The answer is yes,provided the surface temperature is compatible with life,then in principle life can survive and generate biosigna-ture gases. The chemistry described in this paper stillholds under a massive atmosphere, because the photo-chemical destruction occurs above 1 mbar. Furthermore,we showed in Seager et al. (2012) that the biomass sur-face density estimates are unchanged under a massiveatmosphere as long as the photochemical loss rate dom-inates. For biosignature gases whose loss is dominatedby deposition at the surface (i.e. are absorbed by thesurface), then the biosignature source flux and hencebiomass surface density will scale linearly with planetaryatmosphere mass.The second key question is, Can the high density and

pressures on the surface under a massive atmosphere gen-

16

1.75

1.755

1.76

1.765

1.77

Pla

ne

t R

ad

ius [E

art

h R

ad

ius]

N2O

N2O

CO2

H2O

H2 - H

2

H2 - H

2

H2 - H

2

CO2

CO2

H2O

H2O

CO2

CO2

H2O

H2O

H2O

CH4

CH4

CH4

0 ppm

0.5 ppm

5 ppm

50 ppm

500 ppm

0 ppm

0.5 ppm

5 ppm

50 ppm

500 ppm

0 ppm

0.5 ppm

5 ppm

50 ppm

500 ppm

Pla

ne

t R

ad

ius / S

tella

r R

ad

ius

0.016

0.0161

0.0161

0.0162

0.0802

0.0804

0.0806

0.0808

0.081

Pla

ne

t R

ad

ius / S

tella

r R

ad

ius

0.016

0.0161

0.0161

0.0162

0.0802

0.0804

0.0806

0.0808

0.081

Pla

ne

t R

ad

ius / S

tella

r R

ad

ius

0.016

0.0161

0.0161

0.0162

0.0802

0.0804

0.0806

0.0808

0.081

Pla

ne

t R

ad

ius / S

tella

r R

ad

ius

0.016

0.0161

0.0161

0.0162

0.0802

0.0804

0.0806

0.0808

0.081

H2 - H

2

CO2

CO2

H2O

CH4

NH3

NH3

0 ppm

0.5 ppm

5 ppm

50 ppm

500 ppm

1.75

1.755

1.76

1.765

1.77

Pla

ne

t R

ad

ius [E

art

h R

ad

ius]

1.75

1.755

1.76

1.765

1.77

Pla

ne

t R

ad

ius [E

art

h R

ad

ius]

0.2 0.5 1 2 5 10 20 50 1001.75

1.755

1.76

1.765

1.77

Wavelength [microns]

Pla

ne

t R

ad

ius [E

art

h R

ad

ius]

CH3Cl

CH3Cl

DMS DMS

Figure 7. Theoretical transmis