Abundant atmospheric methane from volcanism on terrestrial planets is unlikely and strengthens the case for methane as a biosignature
AAbundant Atmospheric Methane from Volcanism on Terrestrial Planets Is Unlikely andStrengthens the Case for Methane as a Biosignature
Nicholas Wogan , Joshua Krissansen-Totton , and David C. Catling Dept. Earth and Space Sciences, University of Washington, Seattle, WA, USA; [email protected] Virtual Planetary Laboratory, University of Washington, Seattle, WA, USA Dept. of Astronomy and Astrophysics, University of California, Santa Cruz, CA, USA
Received 2020 June 10; revised 2020 August 14; accepted 2020 September 15; published 2020 October 29
Abstract
The disequilibrium combination of abundant methane and carbon dioxide has been proposed as a promisingexoplanet biosignature that is readily detectable with upcoming telescopes such as the James Webb SpaceTelescope. However, few studies have explored the possibility of nonbiological CH and CO and relatedcontextual clues. Here we investigate whether magmatic volcanic outgassing on terrestrial planets can produceatmospheric CH and CO with a thermodynamic model. Our model suggests that volcanoes are unlikely toproduce CH fl uxes comparable to biological fl uxes. Improbable cases where volcanoes produce biologicalamounts of CH also produce ample carbon monoxide. We show, using a photochemical model, that high abioticCH abundances produced by volcanoes would be accompanied by high CO abundances, which could be adetectable false-positive diagnostic. Overall, when considering known mechanisms for generating abiotic CH onterrestrial planets, we conclude that observations of atmospheric CH with CO are dif fi cult to explain without thepresence of biology when the CH abundance implies a surface fl ux comparable to modern Earth ’ s biological CH fl ux. A small or negligible CO abundance strengthens the CH + CO biosignature because life readily consumesatmospheric CO, while reducing volcanic gases likely cause CO to build up in a planet ’ s atmosphere. Furthermore,the dif fi culty of volcanically generated CH -rich atmospheres suitable for an origin of life may favor alternativessuch as impact-induced reducing atmospheres. Uni fi ed Astronomy Thesaurus concepts: Biosignatures ( ) ; Astrobiology ( ) ; Planetary atmospheres ( ) ;Volcanoes ( )
1. Introduction
Large telescopes will soon be used to search for biogenicwaste gases in exoplanet atmospheres. Oxygen is the mostextensively studied biosignature gas ( Meadows 2017; Meadowset al. 2018 ) . Although many studies have proposed ways ofidentifying scenarios where nonliving processes might mimic lifeby producing oxygen ( i.e., false positives; Domagal-Goldmanet al. 2014; Tian et al. 2014; Wordsworth & Pierrehumbert 2014;Harman et al. 2015; Luger & Barnes 2015; Schwieterman et al.2019 ) , the circumstances are unusual and contextual clues candistinguish abiotic scenarios ( Meadows et al. 2018 ) .However, even when life is present, oxygen biosignaturesmay be uncommon. Oxygenic photosynthesis is a complexmetabolism that only evolved once on Earth ( Fischer et al.2016 ) . Additionally, oxygen was slow to accumulate in theEarth ’ s atmosphere ( Lyons et al. 2014 ) , and other planets mayhave low O concentrations for billions of years despite havingoxygenic photosynthetic life if there are large oxygen sinks ( Claire et al. 2006 ) . Accumulation of oxygen may be especiallychallenging on planets orbiting M-dwarf stars due to their lowvisible photon fl ux, which potentially limits primary production ( Lehmer et al. 2018 ) .One alternative to detecting oxygen-rich planets like themodern Earth is to look for methane on planets like theArchean Earth. Before the rise of oxygen, methanogenic lifecould have sustained a methane-rich atmosphere, which could be detected with remote spectroscopy ( Schindler & Kasting2000; Kasting & Catling 2003 ) .Recently, Krissansen-Totton et al. ( ) proposed acriterion for methane biosignatures: fi nding abundant CH inthe presence of CO ( abbreviated CH + CO ) . This combina-tion is compelling if the CH mixing ratio is greater than 0.1%because it is dif fi cult to explain such an abundance with theshort atmospheric lifetime of CH in terrestrial atmospheresand nonbiological methane sources such as serpentinization ( Krissansen-Totton et al. 2018b ) . This 0.1% threshold value isfor planets that orbit stars like the Sun and must be adjusted fordifferent stellar types. For example, planets orbiting M-starstypically receive less near-UV radiation than planets orbitingSun-like stars, resulting in different photochemistry thatpromotes the buildup of CH ( Segura et al. 2005; Grenfellet al. 2007, 2014; Rugheimer et al. 2015; Rugheimer &Kaltenegger 2018 ) . Krissansen-Totton et al. ( ) arguedthat the CH biosignature is strengthened by a low COabundance because volcanoes that produce CH should alsolikely generate CO. Additionally, living planets might have lowCO because microbes consume CO ( Kharecha et al. 2005 ) ;coupled ecosystem-planetary models of the early Earth suggestatmospheric CO / CH ratios declined dramatically with theemergence of chemoautotrophic ecosystems ( Sauterey et al.2020 ) .Exploring false positives for methane biosignatures istimely. Biogenic O or O detections with upcomingtelescopes, such as the James Webb Space Telescope ( JWST ) ,will be extremely dif fi cult ( Barstow & Irwin 2016; Krissan-sen-Totton et al. 2018a; Fauchez et al. 2019; Lustig-Yaegeret al. 2019; Wunderlich et al. 2020 ) , whereas CH + CO The Planetary Science Journal, ( ))
Large telescopes will soon be used to search for biogenicwaste gases in exoplanet atmospheres. Oxygen is the mostextensively studied biosignature gas ( Meadows 2017; Meadowset al. 2018 ) . Although many studies have proposed ways ofidentifying scenarios where nonliving processes might mimic lifeby producing oxygen ( i.e., false positives; Domagal-Goldmanet al. 2014; Tian et al. 2014; Wordsworth & Pierrehumbert 2014;Harman et al. 2015; Luger & Barnes 2015; Schwieterman et al.2019 ) , the circumstances are unusual and contextual clues candistinguish abiotic scenarios ( Meadows et al. 2018 ) .However, even when life is present, oxygen biosignaturesmay be uncommon. Oxygenic photosynthesis is a complexmetabolism that only evolved once on Earth ( Fischer et al.2016 ) . Additionally, oxygen was slow to accumulate in theEarth ’ s atmosphere ( Lyons et al. 2014 ) , and other planets mayhave low O concentrations for billions of years despite havingoxygenic photosynthetic life if there are large oxygen sinks ( Claire et al. 2006 ) . Accumulation of oxygen may be especiallychallenging on planets orbiting M-dwarf stars due to their lowvisible photon fl ux, which potentially limits primary production ( Lehmer et al. 2018 ) .One alternative to detecting oxygen-rich planets like themodern Earth is to look for methane on planets like theArchean Earth. Before the rise of oxygen, methanogenic lifecould have sustained a methane-rich atmosphere, which could be detected with remote spectroscopy ( Schindler & Kasting2000; Kasting & Catling 2003 ) .Recently, Krissansen-Totton et al. ( ) proposed acriterion for methane biosignatures: fi nding abundant CH inthe presence of CO ( abbreviated CH + CO ) . This combina-tion is compelling if the CH mixing ratio is greater than 0.1%because it is dif fi cult to explain such an abundance with theshort atmospheric lifetime of CH in terrestrial atmospheresand nonbiological methane sources such as serpentinization ( Krissansen-Totton et al. 2018b ) . This 0.1% threshold value isfor planets that orbit stars like the Sun and must be adjusted fordifferent stellar types. For example, planets orbiting M-starstypically receive less near-UV radiation than planets orbitingSun-like stars, resulting in different photochemistry thatpromotes the buildup of CH ( Segura et al. 2005; Grenfellet al. 2007, 2014; Rugheimer et al. 2015; Rugheimer &Kaltenegger 2018 ) . Krissansen-Totton et al. ( ) arguedthat the CH biosignature is strengthened by a low COabundance because volcanoes that produce CH should alsolikely generate CO. Additionally, living planets might have lowCO because microbes consume CO ( Kharecha et al. 2005 ) ;coupled ecosystem-planetary models of the early Earth suggestatmospheric CO / CH ratios declined dramatically with theemergence of chemoautotrophic ecosystems ( Sauterey et al.2020 ) .Exploring false positives for methane biosignatures istimely. Biogenic O or O detections with upcomingtelescopes, such as the James Webb Space Telescope ( JWST ) ,will be extremely dif fi cult ( Barstow & Irwin 2016; Krissan-sen-Totton et al. 2018a; Fauchez et al. 2019; Lustig-Yaegeret al. 2019; Wunderlich et al. 2020 ) , whereas CH + CO The Planetary Science Journal, ( )) , 2020 December https: // doi.org / / PSJ / abb99e © 2020. The Author ( s ) . Published by the American Astronomical Society. Original content from this work may be used under the termsof the Creative Commons Attribution 4.0 licence. Any furtherdistribution of this work must maintain attribution to the author ( s ) and the titleof the work, journal citation and DOI. + CO biosignature is potentiallydetectable on the planet TRAPPIST-1e, with just 10 transits ( Krissansen-Totton et al. 2018a ) . Thus, exploration ofpotential methane biosignature false positives and theircontextual discriminants is needed.The literature exploring false positives for methane bio-signatures has primarily focused on CH generation indeep-sea serpentinizing hydrothermal vents. Guzmán-Marmo-lejo et al. ( ) estimated a maximum CH surface fl ux of0.18 Tmol yr − ( ´ molecules cm − - s ) from hydro-thermal vents for planets with the same mass as Earth.Additionally, Krissansen-Totton et al. ( ) used MonteCarlo simulations to estimate a probability distribution formaximum abiotic CH production from this process. Theysuggest that >
10 Tmol CH4 yr − is highly unlikely. Theseestimated maximum fl uxes are small compared to modernEarth ’ s biological CH fl ux of 30 Tmol yr − .However, investigations of abiotic CH on Earth suggest thatthese estimates of abiotic CH from hydrothermal vents arepotentially unrealistically large. Serpentinization reactions invol-ving water and ultrama fi c oceanic crust generate H ; then,purportedly, H might react with inorganic carbon in hydro-thermal systems to generate CH . Krissansen-Totton et al. ( ) and Guzmán-Marmolejo et al. ( ) both estimatedabiotic CH fl uxes, assuming ef fi cient reactions between H andinorganic carbon. However, laboratory experiments have shownthat, uncatalyzed, this reaction is extremely slow at hydrothermalvent temperatures and pressures preventing chemical equilibriumon timescales of at least months ( Reeves & Fiebig 2020 ) .Additionally, various lines of evidence suggest that much of theCH observed in deep-sea hydrothermal vent waters is ultimatelyfrom biology ( Reeves & Fiebig 2020 ) . Furthermore, lifelessplanets without silica-secreting organisms should have highocean-water SiO concentrations, which suppresses the H andtherefore abiotic CH produced from serpentinization ( Tutoloet al. 2020 ) .Impacts can likely generate abiotic CH ( Zahnle et al. 2020 ) ,although impact-generated CH is only probable early in a solarsystem ’ s lifetime. The cratering record on the Moon shows thatEarth ’ s impact fl ux decreased dramatically by 3.5 Ga ( Marchiet al. 2014 ) . Thus, extra-solar systems that are several billionyears old are probably unlikely to have abiotic CH from thissource.Here we investigate another potential false-positive for theCH + CO biosignature: magma-sourced volcanic outgassing ( i.e., not metamorphic ) . Negligible CH has been observed ingases emitted by magmatic volcanoes on Earth ( Catling &Kasting 2017; Reeves & Fiebig 2020 ) , although it has not beeninvestigated whether substantial CH is feasible for volcanoesin vastly different thermodynamic regimes. We simulateoutgassing speciation for a range of magma temperatures,outgassing pressures, oxygen fugacities, volatile composition,and variable partitioning between subaerial and submarinevolcanism. We examine whether volcanoes can produce CH fl uxes comparable to biological fl uxes. Using a photochemicalmodel, we also investigate the atmospheric composition ofhypothetical planets by reducing volcanic gases to see whethervolcanic CH coincides with large atmospheric CO, whichcould be a detectable false-positive marker.
2. Methods
Below, we describe our model for predicting the gasesproduced by an erupting mantle-sourced volcano. We followGaillard & Scaillet ( ) and solve for the gas – gas and gas – melt equilibrium in a C – O – H system. Our model differs fromGaillard & Scaillet ( ) because we do not consider nitrogenor sulfur species. Despite these differences, we obtain similarresults to calculations made in Gaillard & Scaillet ( ) . Wehave also validated our code against the work of Liggins et al. ( ) and Ortenzi et al. ( ) , which have independentlyconstructed similar outgassing models. Our Python code ispublished as open-source software on the GitHub page https: // github.com / Nicholaswogan / VolcGases.Figure 1 shows a highly schematic conceptualization ofvolcanic degassing typical of low-viscosity magma. Gasbubbles form in the magma when molecules like H O andCO are exsolved. Within the gas bubbles, reactions drive thesystem to chemical equilibrium. The oxygen fugacity ( f O ) ofthe gas bubble is controlled by equilibrium with the oxygenfugacity of the magma ( e.g., Kadoya et al. 2020 ) . Gasesbubbles are released from the magma and enter the overlyingatmosphere or ocean.A mathematical model describes the volatiles in gas bubblesand magma. The amount of carbon and hydrogen that areexsolved by the magma into bubbles is governed by thesolubility of CO and H O, which we calculate with thesolubility relations for ma fi c magmas described in Iacono-Marziano et al. ( ) : ( ) ( ) ( ) = + + x x d a P S ln ln , 1 CO H O H O CO CO 1 ( ) ( ) ( ) = + x a P S ln ln . 2
H O H O H O 1
Here x CO and x H O are mol fractions of CO and H O in themagma, respectfully. Additionally, P CO and P H O are the partialpressure of CO and H O in gas bubbles suspended in themagma. The other terms in Equations ( ) and ( ) are solubilityparameters with values shown in Table 1, except S and S ,which are further described in Appendix A.1. We use solubilityrelations appropriate for ma fi c magmas because rocky planetsand moons in our solar system usually have basaltic crusts,suggesting that ma fi c magma is common to most terrestrialbodies.Volatile mol fractions ( e.g., x H O ) can be converted to massfractions with the formula ( ) mm = m x i i i magma Here m i is the mass fraction, m i is the volatile ’ s molar mass, and i can be either H O or CO . Table 1 gives the units ofeach term.We assume that after the hot gas exsolves from the magmainto bubbles, it achieves thermodynamic equilibrium from thereactions ( ) « + H O H 12 O , 4 ( ) « +
CO CO 12 O , 5 ( ) + « +
CO 2H O CH 2O . 6 The Planetary Science Journal, ( ))
CO 2H O CH 2O . 6 The Planetary Science Journal, ( )) , 2020 December Wogan, Krissansen-Totton, & Catling t thermodynamic equilibrium, the ratios of the fugacities ofvolatile species ( denoted f i ) are related to the equilibriumconstant corresponding to each chemical reaction. We assumethat we can replace fugacities with partial pressures ( denoted P i ) . This approximation is reasonable for the temperatures andpressures involved in volcanic outgassing ( Holland 1984 ) .Thus, ( ) = » K f ff P fP , 7 ( ) = »
K f ff P fP , 8
22 22 ( ) = »
K f ff f P fP P . 9 CH O CO H O2 CH O CO H O2
We calculate equilibrium constants ( e.g., K ) using the NASAthermodynamic database ( Burcat & Ruscic 2005 ) . We assumethat the gas is thermally and chemically coupled to the magmaso that the oxygen fugacity ( f O ) of the gas is set by the oxygen fugacity of magma, as observed ( Symonds et al. 1994 ) . So far,we have seven unknowns ( x CO , x H O , P CO , P H O , P CO , P H , P CH ) and only fi ve equations. To close the system, we addthree more equations and one more unknown. The fi rstequation requires that the partial pressures sum to the totalpressure: ( ) + + + + = P P P P P P . 10
H H O CO CO CH
The fi nal two equations are atom conservation equations forcarbon and hydrogen: ( ) ( ) mm aa = + ++ - m P P PP x COtot magmaCO CO CO CH gasgas CO ( ) ( ) mm aa = + ++ - m P P PPx
21 . 12
H Otot magmaH O H O H CH gasgas H O
Equations ( ) and ( ) state that the total moles of eithercarbon or hydrogen should be equal to the moles of either Figure 1.
Qualitative sketch of degassing typical of low-viscosity magma ( e.g., Hawaiian volcanoes ) . Here a gas bubble reaches thermal and chemical equilibriumwith a melt ( no crystals are present ) . Note, degassing can occur in many different ways depending on magma viscosity and volatile content ( Gonnermann &Manga 2013 ) . The Planetary Science Journal, ( ))
Qualitative sketch of degassing typical of low-viscosity magma ( e.g., Hawaiian volcanoes ) . Here a gas bubble reaches thermal and chemical equilibriumwith a melt ( no crystals are present ) . Note, degassing can occur in many different ways depending on magma viscosity and volatile content ( Gonnermann &Manga 2013 ) . The Planetary Science Journal, ( )) , 2020 December Wogan, Krissansen-Totton, & Catling lement in the gas phase plus the moles in the magma. Here a gas is the fi nal unknown. It is the total moles in the gas phasedivided by the total moles in the gas and magma combined.See Appendix A.2 for a full derivation of Equations ( ) and ( ) .Given a gas and magma temperature ( T ) , pressure ( P ) ,oxygen fugacity ( f O ) , and the total mass fraction ( or molfraction ) of CO and H O in the magma ( m COtot , and m H Otot ) ,Equations ( ) , ( ) , ( ) – ( ) are a system of eight equations andeight unknowns ( x CO , x H O , P CO , P H O , P CO , P H , P CH , a gas ) .We solve this system of equations numerically with the ScipyPython package.The solution to this system of equilibrium equationsprovides an estimate of the amount of each volatile species ingas bubbles in magma immediately before the gas leaves themagma. We assume bubbles remain in thermodynamicequilibrium with the surrounding melt until they are releasedinto the overlying atmosphere or ocean, and volatile speciationdoes not continue to evolve upon release. This does not exactlyre fl ect real degassing. Observed outgassing chemistry suggeststhat volcanic gas re-equilibrates to temperatures slightly lowerthan the magma as the gas leaves the magma and is no longerchemically buffered by it ( Oppenheimer et al. 2018; Moussal-lam et al. 2019; Kadoya et al. 2020 ) . We do not capture thiscomplexity in the main text, although in Appendix A.4 we investigate the closed system re-equilibration of volcanic gasesand show that this process does not change our conclusions.Once the unknowns are solved for, they can be used tocalculate the gas production ( i.e., the moles of gas produced perkilogram of magma erupted ) : ⎛⎝⎜⎜ ⎞⎠⎟⎟ ( ) ( ) am a = - q PP
10 1 . 13 i i Here q i is the gas production of species i in mol gas kg − magma. Calculating q i is useful because it is related to the fl ux F i of gas i to the atmosphere by the magma production rate: ( ) = F q Q . 14 i i m
Here Q m is the magma production rate in kg magma yr − and F i is in mol yr − .Several authors have shown that degassing can be affectedby graphite saturation of magma ( Hirschmann & Withers 2008 ) or by the solubility of CO, CH , and H in magma ( Hirschmannet al. 2012; Ardia et al. 2013; Wetzel et al. 2013 ) . The gasspeciation model described previously does not account forthese processes. However, in Appendix A.3, we introduce amore complex model that accounts for graphite saturation andCO, CH , and H solubility, and show that this model producesvery similar results to the simpli fi ed model described here. Table 1
Model Constants and VariablesConstant or variable Value Units De fi nitionConstants d H O L Solubility constant a a CO L Solubility constant a a H O L Solubility constant a S L L
Solubility constant a S L L
Solubility constant a m magma g magmamol magma Molar mass of magma b m H O g H Omol H O Molar mass of H O m CO g COmol CO Molar mass of CO K - + e T bar Equilibrium constant c K - + e T bar Equilibrium constant c K - + e T L Equilibrium constant c Input P L bar Total pressure of degassing T L K Temperature of magma and gas f O L bar Oxygen fugacity of the magma m COtot L gg COgas and magma Mass fraction CO in magma before degassing m H Otot L gg H Ogas and magma Mass fraction H O in magma before degassingOutput x H O L mol H Omol magma Mol fraction of H O in the magma after degassing x CO L mol COmol magma Mol fraction of CO in the magma after degassing P H O L bar Partial pressure of H O P CO L bar Partial pressure of CO P H L bar Partial pressure of H P CO L bar Partial pressure of CO P CH L bar Partial pressure of CH a gas L mol gasmol gas and magma Mol fraction in gas phase
Notes. a From Iacono-Marziano et al. ( ) . See Appendix A.1 to calculate S and S . b Molar mass of Mount Etna magma. c Calculated from the NASA thermodynamic database ( Burcat & Ruscic 2005 ) . The Planetary Science Journal, ( ))
Notes. a From Iacono-Marziano et al. ( ) . See Appendix A.1 to calculate S and S . b Molar mass of Mount Etna magma. c Calculated from the NASA thermodynamic database ( Burcat & Ruscic 2005 ) . The Planetary Science Journal, ( )) , 2020 December Wogan, Krissansen-Totton, & Catling .2. Monte Carlo Simulations We investigate volcanic false positives to the CH + CO biosignature on two types of worlds: an Earth-like world withsubaerial and submarine outgassing ( Figure 2 ) and an oceanworld with only submarine outgassing. For each type of planet,we search for false-positive scenarios by calculating volcanicoutgassing speciation with a wide range of input parameters.To explore volcanism on Earth-like planets, we calculateoutgassing speciation 10,000 times. For each calculation,we sample either uniform or log -uniform distributions ( see Table 2 ) of 10 parameters: T submarine , P submarine , m CO , submarinetot , m H O, submarinetot , T subaerial , P subaerial , m CO , subaerialtot , m H O, subaerialtot , f O , and X . The width of each uniform samplingdistribution is given and explained in Table 2. We use inputs with subscripts “ subaerial ” to calculate subaerial volcanicspeciation and inputs with subscripts “ submarine ” to calculatesubmarine volcanic speciation, and then we combine the resultsof each calculation with the formula ( ) ( ) = + - n PP X PP X i i i , subaerialsubaerial , submarinesubmarine Here n i is the mixing ratio of averaged outgassed volatiles ofspecies i produced by the combination of subaerial andsubmarine volcanoes and X is the fraction of subaerialvolcanism ( < < X ) . Also, P i ,subaerial and P i ,submarine are thepartial pressure of species i in subaerial and submarineoutgassing, respectively. Figure 2.
Illustration of the parameters considered in the Monte Carlo simulations.
Table 2
Monte Carlo Sampling DistributionsVariable Low High Sampling method Justi fi cation T submarine
873 K 1973 K Linear uniform Range of submarine magma temperatures observed on Earth a T subaerial
873 K 1973 K Linear uniform Range of subaerial magma temperatures observed on Earth a P submarine
100 bar 1000 bar Linear uniform Degassing pressure at 1 km to 10 km ocean depth b P subaerial uniform Rough range of subaerial degassing pressure in solar system m CO , submarinetot − − log uniform Approx. CO mass fraction range in Earth magma ( Wallace 2005; Wallace et al. 2015; Anderson &Poland 2017; le Voyer et al. 2019 ) m CO ,subaerialtot − − log uniform Approx. CO mass fraction range in Earth magma ( Wallace 2005; Wallace et al. 2015; Anderson &Poland 2017; le Voyer et al. 2019 ) m H O, submarinetot − − log uniform H O mass fraction range for Earth submarine outgassing ( Wallace et al. 2015 ) m H O, subaerialtot − − log uniform H O mass fraction range for Earth subaerial outgassing ( Wallace et al. 2015 ) f O FMQ-4 FMQ + uniform Oxygen fugacity of most reducing Martian meteorite ( Catling & Kasting 2017 ) to most oxidizedmagma on Earth ( Stamper et al. 2014 ) c X Notes. a Coldest rhyolite magma and hottest komatiites magmas ( Huppert et al. 1984 ) . b Assumes Earth ’ s gravity. The solubility of H O in magma does not allow for signi fi cant CH degassing at pressures greater than 1000 bar, equivalent to a depth of10 km. c FMQ is the fayalite-magnetite-quartz mineral redox buffer. See Chapter 7 in Catling & Kasting ( ) for a description of mineral redox buffers. We use theparameterization for the FMQ buffer de fi ned by Wones & Gilbert ( ) . This parameterization has only been experimentally validated to 1400 K ( O ’ Neill 1987 ) , butwe extrapolate using the parameterization to 1973 K. The Planetary Science Journal, ( ))
100 bar 1000 bar Linear uniform Degassing pressure at 1 km to 10 km ocean depth b P subaerial uniform Rough range of subaerial degassing pressure in solar system m CO , submarinetot − − log uniform Approx. CO mass fraction range in Earth magma ( Wallace 2005; Wallace et al. 2015; Anderson &Poland 2017; le Voyer et al. 2019 ) m CO ,subaerialtot − − log uniform Approx. CO mass fraction range in Earth magma ( Wallace 2005; Wallace et al. 2015; Anderson &Poland 2017; le Voyer et al. 2019 ) m H O, submarinetot − − log uniform H O mass fraction range for Earth submarine outgassing ( Wallace et al. 2015 ) m H O, subaerialtot − − log uniform H O mass fraction range for Earth subaerial outgassing ( Wallace et al. 2015 ) f O FMQ-4 FMQ + uniform Oxygen fugacity of most reducing Martian meteorite ( Catling & Kasting 2017 ) to most oxidizedmagma on Earth ( Stamper et al. 2014 ) c X Notes. a Coldest rhyolite magma and hottest komatiites magmas ( Huppert et al. 1984 ) . b Assumes Earth ’ s gravity. The solubility of H O in magma does not allow for signi fi cant CH degassing at pressures greater than 1000 bar, equivalent to a depth of10 km. c FMQ is the fayalite-magnetite-quartz mineral redox buffer. See Chapter 7 in Catling & Kasting ( ) for a description of mineral redox buffers. We use theparameterization for the FMQ buffer de fi ned by Wones & Gilbert ( ) . This parameterization has only been experimentally validated to 1400 K ( O ’ Neill 1987 ) , butwe extrapolate using the parameterization to 1973 K. The Planetary Science Journal, ( )) , 2020 December Wogan, Krissansen-Totton, & Catling o investigate volcanism on an ocean world, we also calculateoutgassing speciation 10,000 times. For each calculation, wesample either uniform or log -uniform distributions of inputs T submarine , P submarine , m CO , submarinetot , m H O, submarinetot , and f O , withranges de fi ned and justi fi ed in Table 2. We further investigate the CH + CO biosignature bymodeling the atmospheric composition of hypothetical unin-habited ocean worlds with reducing volcanic gases. We considerplanets orbiting the Sun and a late M star — the latter becauseplanets orbiting M-dwarfs are the most feasible targets for near-term telescopes like JWST ( Barstow & Irwin 2016 ) . Addition-ally, we simulate ocean worlds because ocean-bottom degassingis most thermodynamically prone to produce CH , as revealed byour Monte Carlo simulations and previous studies ( French 1966;Kasting & Brown 1998; see Section 4.1.1 for further discussion ) .To simulate atmospheres on uninhabited planets, we use the1-D photochemical model contained within the open-sourcesoftware package Atmos. Atmos is derived from a modeloriginally developed by the Kasting group ( Pavlov et al. 2001 ) ,and versions of this code have been used to simulate theArchean and Proterozoic Earth atmosphere ( Zahnle et al.2006 ) , Mars ( Zahnle et al. 2008; Smith et al. 2014; Sholes et al.2019 ) , and exoplanet atmospheres ( Harman et al. 2015;Schwieterman et al. 2019 ) .
3. Results
Figure 3 shows joint distributions of gas ratios CH / CO andCO / CO from the Monte Carlo simulation described inSection 2.2. These results suggest that for most combinationsof parameters, volcanoes are most likely to produce more CO than CO, and negligible CH , which is the case for the modernEarth ( Catling & Kasting 2017 ) . About 7% and 2% ofcalculations produce more CH than CO for ocean worlds andEarth-like worlds, respectfully. In the vast majority of cases,either CO or CO is the dominant carbon-bearing species. Figures 4 ( a ) and ( b ) show CH production from the MonteCarlo simulations in terms of mol CH kg − magma. To give asense of the gas fl uxes implied by these CH productions, wemultiply the distributions in Figures 4 ( a ) and ( b ) by the magmaproduction rate of modern Earth of 9 × kg yr − ( Crisp1984 ) , which gives the gas fl uxes shown in Figures 4 ( c ) and ( d ) , respectively. About 0.1% of calculations predict more than10 Tmol CH yr − for both Earth-like worlds and oceanworlds. This small fraction suggests that for modern Earthmagma production rates, volcanoes are unlikely to produceCH fl uxes comparable to modern Earth ’ s biological fl ux of30 Tmol yr − ( Hauglustaine et al. 2007 ) .Magma production rates larger than modern Earth ’ s increasethe probability that volcanic fl uxes of CH become comparableto biological CH fl uxes. For example, the early Archean Earthcould have had magma production rates up to about 25 timesmodern Earth ’ s ( Sleep & Zahnle 2001 ) . Such a magmaproduction rate would shift the distributions in Figures 4 ( c ) and ( d ) to larger values by a factor of 25 ( or in log -space, by afactor of 1.4 ) . In this case, ∼
2% of calculations ( for eitherEarth-like world or ocean world ) would predict more than 10Tmol CH yr − .Crucially, large CH fl uxes should almost always coincidewith even larger CO fl uxes ( horizontal axis in Figure 3 ) .Therefore, the unlikely cases where volcanoes mimic biologicalCH fl uxes can be identi fi ed by detecting abundant CO in aplanet ’ s atmosphere. We further investigate CO as a CH + CO biosignature discriminant using a photochemical model in thefollowing section. We use the
Atmos photochemical model to simulate thepotential observable gas abundances of uninhabited Earth-sizedocean worlds with reducing volcanic gases. We consider suchplanets because they are the most prone to mimic biology byproducing volcanic CH ( see Section 4.1.1 for more details ) .Our hypothetical planets have 1 bar N dominated atmospheres,400 bars of ocean water, magma degassing at 1473 K,and mantle redox states of FMQ-4. Here FMQ is the Figure 3.
Results of the Monte Carlo simulation described in Section 2.2. ( a ) and ( b ) show normalized count as a function of ( ) log CH CO and ( ) log CO CO for anocean world and Earth-like world, respectively. The white dotted lines indicate where CH / CO = / CO =
1. For almost all calculated gas speciations, CO and CO are much more abundant than CH . The Planetary Science Journal, ( ))
1. For almost all calculated gas speciations, CO and CO are much more abundant than CH . The Planetary Science Journal, ( )) , 2020 December Wogan, Krissansen-Totton, & Catling ayalite-magnetite-quartz buffer, which is a synthetic reference f O value at fi xed temperature-pressure conditions. Additionally,we assume that the magma contains 0.1 wt% CO and 1 wt%H O. Our assumed H O concentration is comparable to thoseobserved in submarine hot-spot magmas ( ) ; however, the CO concentration weassume is slightly lower ( Anderson & Poland 2017 ) . Giventhese inputs, our speciation model ( Section 2.1 ) predicts gasproduction from erupted magma of = ´ - q H 2 molgas / kg magma, = ´ - q CO 2 mol gas / kg magma, and = ´ - q CH 3 mol gas / kg magma.The magnitude of gas fl uxes to the atmosphere resultingfrom chemically reducing volcanism depends on the magmaproduction rate ( Equation ( )) . We consider magma produc-tion rates between about 10 − and 10 Earth ’ s modern magmaproduction rate of 9 × kg magma yr − ( Crisp 1984 ) .For each magma production rate, we calculate the outgassing fl ux of CH , H , and CO and set these fl uxes as lower boundaryconditions to the Atmos photochemical model. ( The outgassingmodel also gives CO and H O fl uxes, but we do not use themin our photochemical modeling. ) Atmos only allows fi xed CO mixing ratios and not CO fl uxes, so we consider cases withlow and high CO (
100 ppm and 10% ) . Additionally, we set thedeposition velocity of CO to 10 − cm s − to re fl ect the abioticuptake of CO by the ocean ( Kharecha et al. 2005 ) . All other boundary conditions are speci fi ed in Appendix B. Givenvolcanic outgassing fl uxes and other boundary conditions, Atmos calculates the mixing ratios of all species when theatmosphere is at photochemical equilibrium.Figure 5 shows the photochemical modeling results ofreducing volcanic gases on an uninhabited Earth-sized oceanworld orbiting the Sun. Figure 5 ( a ) assumes that theatmosphere has 100 ppmv CO , while Figure 5 ( b ) assumesthat atmospheric CO is 10%. Carbon monoxide and methaneare more abundant in the model with more CO because CO shields the lower atmosphere from hydoxyl ( OH ) productionfrom water photolysis. In anoxic atmospheres, OH is asigni fi cant sink for both CO and CH through the reactions + + CO OH CO H and + +
CH OH CH H O .OH is generated primarily from H O photolysis ( [ ] n l + < + h H O 200 nm OH H ) , but CO shieldsH O from photolysis in model runs with 10% CO , thuslimiting the CH and CO destruction from OH. Also, CH ismore abundant in atmospheres with more CO because CO shields CH from direct photolysis in cases when CO is > . This factor of ∼
200 comes fromcomparing Ly α ( λ = ) CO and CH cross sections.Ly α is the portion of the UV spectrum primarily responsiblefor photolyzing CH .Figure 5 suggests that reducing volcanic gases on an oceanworld orbiting a Sun-like star will only mimic biological CH Figure 4.
Normalized count of methane production ( mol gas / kg magma ) for ( a ) ocean worlds and ( b ) Earth-like worlds. Distributions were calculated by sampling theranges in Table 2. Multiplying Earth ’ s magma production rate of 9 × kg magma yr − by ( a ) and ( b ) gives the methane fl uxes in ( c ) and ( d ) , respectively. Formodern Earth ’ s magma production rate, volcanoes are likely to produce negligible CH . The Planetary Science Journal, ( ))
Normalized count of methane production ( mol gas / kg magma ) for ( a ) ocean worlds and ( b ) Earth-like worlds. Distributions were calculated by sampling theranges in Table 2. Multiplying Earth ’ s magma production rate of 9 × kg magma yr − by ( a ) and ( b ) gives the methane fl uxes in ( c ) and ( d ) , respectively. Formodern Earth ’ s magma production rate, volcanoes are likely to produce negligible CH . The Planetary Science Journal, ( )) , 2020 December Wogan, Krissansen-Totton, & Catling uxes and abundances for large magma production rates.Volcanism can generate Earth ’ s modern biological CH fl uxwhen the magma production rate is ∼
50 times modern Earth ’ s ( Figure 5 ) . In this case, the photochemical model predicts anatmospheric CH abundance between 0.01% and 0.3%,depending on the CO mixing ratio. Such CH abundancesare similar to the 0.01% to 1% expected in the early ArcheanEarth atmosphere ( Catling & Zahnle 2020 ) . In contrast, magmaproduction rates comparable to the modern Earth ’ s result in aCH fl ux of ´ molecules cm − s − ( − ) and CH abundances <
30 ppm , which are likely to beconsidered abiotic levels in an anoxic atmosphere.Figure 6 shows the CO and CH mixing ratios on an Earth-sized ocean world with reducing volcanic gases orbiting a coldM star. CO and CH are more abundant on the ocean worldorbiting the M star compared to the ocean world orbiting a Sun-like star ( Figure 5 ) . This is because M8V stars have a low fl ux of near-ultraviolet radiation compared to Sun-like stars. The lownear-ultraviolet fl ux reduces OH produce from H O photolysis,thus allowing for relatively high CO and CH concentrations.One consequence of M-dwarf photochemistry is a higherlikelihood of Archean Earth-like CH abundances on unin-habited planets with reducing gases from volcanism. Figure 6shows that modern Earth magma production rates can result inCH abundances up to 0.01%, which is comparable to what isexpected in the Archean atmosphere.Potential CH biosignature false positives from reducingvolcanic gases might be discriminated from inhabited worldsusing observations of CO. For planets orbiting Sun-like stars ( Figure 5 ) or M stars ( Figure 6 ) , the CO abundance is higherthan the CH abundance in every case that is a potentialoutgassing false-positive. Some authors have argued that a largeCO abundance is unlikely on an inhabited planet, becauseatmospheric CO should be readily consumed by biology Figure 5.
Atmospheric mixing ratios of CO and CH as a function of magma production rate relative to modern Earth ’ s ( or CH fl ux ) on an anoxic ocean world withreducing volcanic gases orbiting a Sun-like star. ( a ) and ( b ) are identical model runs, except ( a ) assumes a constant atmospheric CO mixing ratio of 0.0001, and ( b ) assumes a constant atmospheric CO mixing ratio of 0.1. Modern Earth ’ s biological CH fl ux is indicated on the horizontal axes. Archean Earth-like CH fl uxes andabundances are only mimicked by volcanoes for magma production rates >
10 times modern Earth ’ s. Such false-positive cases can be distinguished from biologybecause the CO abundance exceeds the CH abundance, which would likely not be the case for an inhabited planet. Figure 6.
Identical to Figure 5, except for a planet that orbits an M8V star instead of a Sun-like star. The Planetary Science Journal, ( ))
Identical to Figure 5, except for a planet that orbits an M8V star instead of a Sun-like star. The Planetary Science Journal, ( )) , 2020 December Wogan, Krissansen-Totton, & Catling Krissansen-Totton et al. 2018a ) . Conversely, Schwietermanet al. ( ) has demonstrated hypothetical cases where large COcan coincide with biology in an anoxic atmosphere. We furtherdiscuss CO as a false-positive discriminant in Section 4.2.
4. Discussion Our modeling results show that for modern Earth magmaproduction rates, volcanic fl uxes of reducing gases are unlikelyto produce more than 1 Tmol CH yr − , even in an extreme case ( Figure 4 ) . This fl ux is relatively small compared to the fl ux ofother volcanic gases on modern Earth. For example, Earth ’ smodern volcanoes produce about yr − and O yr − ( Catling & Kasting 2017, p. 203 ) . There arethree main reasons why the outgassing model predicts little CH ,which we explore further in the following discussion. because of Water Solubility inMagma One reason for small CH outgassing is the high solubility ofwater in magma at high pressures. Consider Equation ( ) ,which can be re-arranged as follows: ( ) = PP K Pf . 16
CHCO 3 H O2O2
42 22
The ratio
P P
CH CO in a gas bubble in magma is directlyproportional to P H O2 within that bubble. Generally speaking, P H O increases as the total pressure of degassing increasesbecause all partial pressures must sum to the total pressure ( Equation ( )) . For example, subaerial degassing at ∼ P H O and thus a small P P
CH CO ratio. On the other hand, submarine degassing at ∼
400 barshould have a larger H O partial pressure and thus a larger
P P
CH CO ratio. Here the equilibrium constant and oxygenfugacity have extremely weak pressure dependencies ( i.e., theyare effectively constant as degassing pressure changes ) . Figure 7 ( a ) shows modeled gas speciation for highlyreducing volcanism ( = f O FMQ-4 ) as a function of pressure.For small pressures ( <
100 bar ) , CH increases with increasingpressure and then asymptotes for pressures >
100 bar.CH asymptotes because of the high solubility of water inmagma at high pressure. High pressures dissolve a large fractionof the total available hydrogen as H O into the magma, which isshown in Figure 7 ( b ) . Dissolving a large amount of H O into themagma limits the amount of hydrogen available in the gas phasefor making H-bearing species, like CH , H O, and H .In summary, high pressure is in some ways thermodynamicallyfavorable for making methane because µ P P P
CH CO H O2 , but itis also unfavorable because high pressure dissolves a largefraction of the available hydrogen in the magma as H O. Limitedamounts of hydrogen in gas bubbles result in small amounts ofCH produced.Kasting & Brown ( ) used Equation ( ) to argue that ∼
1% of the carbon outgassed by submarine volcanoes shouldbe CH for magma with = f FMQ O . They assumed that » P P
H O , the total pressure. This assumption is valid foroxidized subaerial volcanoes because ∼
90% of the gasexsolved by Earth ’ s subaerial volcanoes is H O ( Catling &Kasting 2017, p. 203 ) . However, < P P
H O for submarinevolcanoes because of the high-water solubility in magma athigh pressure. Our outgassing model, which accounts forwater ’ s solubility in magma, produces negligible methane.Li & Lee ( ) also predict abundant CH produced bysubaerial and submarine volcanoes ( their Figure 5 ) . However,they calculated equilibrium constants in units of bars and thenused units of Pascals for equilibrium chemistry calculations. Theresult was that they calculated speciation for pressures a factor10,000 times greater than reported. For example, we were able toreproduce their subaerial outgassing case ( their Figure 5 ( a )) byassuming P = P = ( ) , theydid not account for the high solubility of H O in magma at highpressure. Their methods assume the total hydrogen outgassed forsubmarine volcanoes is the same as the total hydrogen outgassedby subaerial volcanoes. This should not be the case, because at
Figure 7. ( a ) Modeled gas speciation as a function of pressure. ( b ) Mole fraction of total hydrogen dissolved in the magma as a function of pressure. Model assumes = f O FMQ-4, T = = m H Otot wt%, and = m COtot wt%. Methane becomes more prevalent in volcanic gases at higher pressures but asymptotes becausehydrogen dissolves into the magma, reducing the total amount of H-bearing volatiles released from the magma. The Planetary Science Journal, ( ))
Figure 7. ( a ) Modeled gas speciation as a function of pressure. ( b ) Mole fraction of total hydrogen dissolved in the magma as a function of pressure. Model assumes = f O FMQ-4, T = = m H Otot wt%, and = m COtot wt%. Methane becomes more prevalent in volcanic gases at higher pressures but asymptotes becausehydrogen dissolves into the magma, reducing the total amount of H-bearing volatiles released from the magma. The Planetary Science Journal, ( )) , 2020 December Wogan, Krissansen-Totton, & Catling igh pressure water dissolves in magma and is unavailable formaking H-bearing gas species ( Figure 7 ( b )) .The pressure dependence of volcanic outgassing has implica-tions for planetary atmospheres generally ( Gaillard & Scail-let 2014 ) . Thin atmospheres will allow substantial degassing ofboth carbon and hydrogen bearing species. However, planets withthick atmospheres or large global oceans will have volcanicdegassing dominated by CO and CO, and almost no hydrogenbearing species. The overburden pressure where C-bearingspecies dominate depends primarily on the un-degassed concen-trations of H O and CO in the magma. In Figure 7, CO and COoverwhelm H-bearing species at ∼ = m COtot and = m H Otot . In contrast,Figure 8 in Gaillard & Scaillet ( ) illustrates a case with lessvolatiles ( = m COtot % and = m H Otot ) where C-bearingspecies eclipse H-bearing species at ∼ because Magma Is Hot Relatively little CH is produced by volcanoes because CH isgenerally not thermodynamically favorable at typical magmadegassing temperatures. Figure 8 shows gas speciation as afunction of temperature for a submarine outgassing case. Forthese chosen inputs, CH is the dominant carbon-bearing speciesfor < T K. Mid-ocean ridge basalts ( MORB ) are about 2 / ( Crisp 1984 ) . MORB magmaerupts at temperatures between 1473 and 1650 K ( Scheideg-ger 1973 ) and are thus in a temperature regime where CH isunfavorable, even from more reducing volcanism.On the other hand, magma from arc volcanoes is generallymuch colder than MORB magma. Moussallam et al. ( ) report magma temperatures for many arc volcanoes ( their TableS3 ) , the coldest of which are 1123 K. Thus, it does seempossible for magma to be cold enough for CH to be thedominant carbon-bearing outgassed species from an extremelyreducing volcano with = f O FMQ-4.Recall that large magma production rates ( ∼
30x modern ) arerequired for volcanoes to produce CH fl uxes compared tobiological ones ( Figure 5 ) . It seems unlikely that planets withlarge magma production rates will have magma temperaturescold enough to produce plentiful CH . For example, the ArcheanEarth may have had a larger magma production rate than the modern Earth because the Earth ’ s mantle was hotter in the distantpast ( Sleep & Zahnle 2001 ) . The hotter Archean mantle resultedin the eruption of ∼ ( Huppert et al.1984 ) or possibly only ∼ ( McKenzie 2020 ) . Such hotmagma degassing is unfavorable for methane ( Figure 8 ) . because Very Low OxygenFugacity Is Required The fi nal reason why volcanic CH is unlikely on terrestrialplanets is because very low f O is required to make abundantmethane. Figure 9 shows gas speciation as a function of oxygenfugacity for submarine volcanism. For these assumed inputs,methane is a substantial fraction of outgassed species for < f O FMQ-3, and at FMQ-5 ( roughly equivalent to the quartz-fayalite-iron buffer ) , half the carbon is converted to CH , while the otherhalf is CO. Most degassing on Earth occurs at approximately = f FMQ O ( Catling & Kasting 2017, p. 208 ) , but magma spansFMQ-4 to FMQ + ( Stamper et al. 2014 ) . Additionally, theoxygen fugacity of Martian meteorites ranges between FMQ andFMQ-3.7 ( Catling & Kasting 2017, p. 363 ) . Therefore, the < f O FMQ-3 required for plentiful CH outgassing is at the extremesof the oxygen fugacities observed for Earth and Mars.Astronomical observations and geochemical experimentssuggest Earth-sized planets should generally have relativelyoxidized magmas. Doyle et al. ( ) spectroscopically measuredthe oxygen fugacity of material polluting the surface of severalwhite dwarfs. Their observations suggest that rocky exoplanetsare likely to have similar oxygen fugacities to Earth and Mars.Additionally, high pressure experiments suggest that the uppermantles of Earth-sized planets should self-oxidize by iron oxidedisproportionation to roughly FMQ during the magma-oceanphase, early in a planet ’ s life ( Armstrong et al. 2019 ) . CO-consuming life evolved very early on Earth ( Adam et al.2018 ) and is a relatively simple metabolism. Therefore, it seemspossible that life on other planets will evolve to consume CO.Planets with atmospheric CH + CO produced by life might alsohave relatively small amounts of atmospheric CO because of CO Figure 8.
Modeled volcanic outgassing speciation as a function of temperature.Model assumes = f O FMQ-4, P =
400 bar, = m H Otot wt%, and = m COtot wt%. CH is more thermodynamically favorable at lower degassingtemperatures. Figure 9.
Modeled volcanic outgassing speciation as a function of oxygenfugacity. Model assumes P =
400 bar, T = = m H Otot wt%, and = m COtot wt%. Methane is most favorable at low oxygen fugacity. The Planetary Science Journal, ( ))
400 bar, T = = m H Otot wt%, and = m COtot wt%. Methane is most favorable at low oxygen fugacity. The Planetary Science Journal, ( )) , 2020 December Wogan, Krissansen-Totton, & Catling onsumers. Consequentially, the presence of abundant CO alongwith CH can discriminate abiotic situations.Monte Carlo simulations show that volcanoes should almostalways produce more CO than CH ( Figure 3 ) . Additionally,photochemical modeling ( Figures 5 and 6 ) suggests that COshould build up in the atmospheres of uninhabited planets withreducing submarine volcanic gases. Thus atmospheric CO + CH produced by volcanoes is likely accompanied by a large COconcentration. This is distinct from an inhabited world, which canhave lower CO concentrations due to CO-consuming life.However, the mere presence of large atmospheric CO is nota de fi nitive sign of an uninhabited planet with reducingvolcanic gases ( Schwieterman et al. 2019 ) . This is becausethere are limits to how quickly gases can be transported fromthe atmosphere into the ocean where they can be consumed bylife ( Kharecha et al. 2005 ) . For example, consider a planet witha very large volcanic CO fl ux ( e.g., 100x modern ) . CO couldbuild up in this planet ’ s atmosphere even if CO consumerswere present in an ocean because CO transport from theatmosphere to the ocean would not be suf fi cient to maintainlow atmospheric CO.In summary, the CH + CO biosignature is most compellingwhen the CO abundance is low or negligible because a lack ofCO potentially implies the presence of CO-consuming biology.In comparison, atmospheric CH + CO and large CO isambiguous, and can either be explained by reducing volcanicgases or by an inhabited world that is unable to sequesteratmospheric CO.JWST might be able to put a tentative upper limit onatmospheric CO. Krissansen-Totton et al. ( ) simulatedJWST retrievals of TRAPPIST-1e with an atmosphericcomposition similar to the Archean Earth containing 10 ppbvCO. Their synthetic retrieval suggested CO was below 652ppmv with 90% con fi dence after 10 transits. CO constraintscould be improved by co-adding more transits and positive COdetections may also be possible with JWST ( Wunderlich et al.2020 ) .However, even if observational CO constraints are poor, it maystill be possible to say something about the abiotic or biotic originof atmospheric CH . Reducing gases from volcanism is unlikelyto mimic the modern biological CH fl ux of 30 Tmol yr − ( Section 4.1 ) . Additionally, serpentinization is unlikely to produce30 Tmol CH yr − , and impact-generated CH might bedistinguished with system age ( Krissansen-Totton et al. 2018b ) .Therefore, JWST observations of atmospheric CH + CO wouldbe challenging to explain without the presence of biologyregardless of atmospheric CO, as long as the CH abundanceimplies a surface fl ux similar to the modern Earth ʼ s. Levels and Implications for the Origin on Life
Much current origin of life research revolves around the “ RNA world ” hypothesis ( Gilbert 1986; Joyce & Szostak 2018;Sasselov et al. 2020 ) . This hypothesis proposes an interval of timewhen primitive life consisted of self-replicating, evolving RNAmolecules, which, at some point, were encapsulated in cells. On arocky world, “ RNA world ” requires that RNA is synthesized fromearly raw materials. Laboratory experiments that have successfullysynthesized nucleobases, which are building blocks of RNA,require the following nitriles: hydrogen cyanide ( HCN ) , cyanoa-cetylene ( HCCCN ) , and cyanogen ( NCCN; Sutherland 2016;Ritson et al. 2018; Benner et al. 2019 ) . In addition, nitriles have also been used to synthesize amino acids ( Miller & Urey 1959;Sutherland 2016 ) .The known natural source of nitriles is photochemistry in achemically reducing atmosphere containing H , CH and N orperhaps NH . For example, Titan ’ s photochemistry producesall the aforementioned nitriles ( Strobel et al. 2009 ) . Impor-tantly, to make the simplest nitrile, HCN, requires abundantCH because HCN is formed from photochemical products ofCH and nitrogen ( Zahnle 1986; Tian et al. 2011 ) .Our results show that volcanic gases generally are unlikely tocause high atmospheric CH abundances in prebiotic atmo-spheres. Consequently, the results lend credence to alternativeproposals for creating early CH -rich, reducing atmospheres,such as impacts ( Zahnle et al. 2020 ) . Impacts can create areducing atmosphere when reactions between iron-rich impactejecta and shock-heated water vapor from an ocean generatecopious H , CH , and NH . Subsequent photochemistry wouldgenerate HCN and other prebiotic nitriles over thousands tomillions of years ( Zahnle et al. 2020 ) .
5. Conclusions
Our modeling of volcanic outgassing speciation suggests thatchemically reducing volcanism on terrestrial planets is unlikelyto mimic biological CH fl uxes. The improbable cases wherevolcanoes do produce biological CH fl uxes also often produceCO. Volcanoes are not prone to produce CH for several reasons.First, the high solubility of H O in magma limits the amount oftotal hydrogen outgassed, thus preventing the production ofH-bearing molecules like CH . Second, CH outgassing requiresrelatively low magma temperatures compared to the majority ofmagma erupted on Earth. Finally, CH outgassing requires a verylow magma oxygen fugacity, unlike that of most terrestrialplanets inferred from astronomical data ( Doyle et al. 2019 ) .We use a photochemical model to calculate the atmosphericcomposition of planets with volcanoes that produce CH . We fi nd that atmospheric CH should coincide with abundant CO.On the other hand, biogenic CH can coincide with a low COabundance if CO-consuming microbial life is present.Therefore, the CH – CO biosignature is most compellingwhen little or no atmospheric CO is detected. AtmosphericCH -CO and large CO is ambiguous and can be explained byan uninhabited planet with highly reducing volcanic gases, oran inhabited planet where biology is unable to sequesteratmospheric CO ( Schwieterman et al. 2019 ) .However, observations of CO are not required to makeconclusions about the abiotic or biotic origin of observedatmospheric CH . Atmospheric CH and CO alone would havea reasonable probability of being biological if the observed CH abundance implies a surface fl ux similar to modern Earth ’ sbiological CH fl ux (
30 Tmol yr − ) . Such a large CH fl ux isdif fi cult to explain with reducing volcanic gases or other abioticprocesses that generate CH , such as serpentinization.These conclusions should be taken with caution because theyare based on what is understood about processes occurring onthe Earth and our Solar System, which may be a very sparsesampling of what is possible.We thank Lena Noack, Michael McIntire, and the twoanonymous reviewers for very knowledgeable and constructivecomments and conversations. We also thank Max Galloway forpointing out a mistake in our calculations in an early draft of thisarticle. N.W. and D.C.C. were supported by the Simons11 The Planetary Science Journal, ( ))
30 Tmol yr − ) . Such a large CH fl ux isdif fi cult to explain with reducing volcanic gases or other abioticprocesses that generate CH , such as serpentinization.These conclusions should be taken with caution because theyare based on what is understood about processes occurring onthe Earth and our Solar System, which may be a very sparsesampling of what is possible.We thank Lena Noack, Michael McIntire, and the twoanonymous reviewers for very knowledgeable and constructivecomments and conversations. We also thank Max Galloway forpointing out a mistake in our calculations in an early draft of thisarticle. N.W. and D.C.C. were supported by the Simons11 The Planetary Science Journal, ( )) , 2020 December Wogan, Krissansen-Totton, & Catling ollaboration on the Origin of Life grant 511570 ( to D.C.C. ) , aswell as the NASA Astrobiology Program grant No.80NSSC18K0829, and bene fi ted from participation in the NASANexus for Exoplanet Systems Science research coordinationnetwork. J.K.-T. was supported by the NASA Sagan Fellowshipand through the NASA Hubble Fellowship grant HF2-51437awarded by the Space Telescope Science Institute, which isoperated by the Association of Universities for Research inAstronomy, Inc., for NASA, under contract NAS5-26555. Appendix ADetails of Outgassing Speciation Model
A.1. Solubility Constants for H O and CO Our outgassing model uses solubility equations for H O andCO in ma fi c magmas from Iacono-Marziano et al. ( ( ) and ( )) . The parameters S and S in the solubilityequations depend on the chemical make-up of the magma. Wefound that different ma fi c magma compositions did notsigni fi cantly affect the outputs of our outgassing speciationmodel ( Section 2.1 ) ; therefore, for the purposes of calculatingmelt solubility, we fi xed the chemical make-up of the magma tothe magma erupting at Mount Etna, Italy, reported by Iacono-Marziano et al. ( ) . This reduced the complexity of the modelwithout sacri fi cing any signi fi cant amount of accuracy.Table 3 shows the chemical make-up of the magma at MountEtna, and Table 4 shows several solubility constants fromIacono-Marziano et al. ( ) . Together, these values de fi nethe solubility parameters S and S : ⎛⎝⎜⎜ ⎞⎠⎟⎟ ⎡⎣⎢ ⎤⎦⎥⎛⎝⎜ ⎞⎠⎟ ( ) ( ) ( ) ( ) mm = + + ++ + ++ + + + + ++ + A1 S C PT B bxx x x dx x d x x d ln 10 NBOO , ⎛⎝⎜⎜ ⎞⎠⎟⎟ ⎡⎣⎢ ⎤⎦⎥ ( ) mm = + + + S C PT B b ln 10 NBOO , A2
Here T is magma temperature, P is the total pressure ofdegassing, and ⎡⎣ ⎤⎦ NBOO is the amount of nonbridging oxygen peroxygen in the melt.
A.2. Derivation of Equations ( ) and ( ) The following is a derivation for the atom conservationequation for carbon used in our outgassing model ( Equation ( )) . The derivation for the atom conservationequation for hydrogen follows the exact same procedure, so wedo not include it.Consider some volume of magma with gas bubbles in it thatcontains a total number of moles g tot . The total moles is the sumof the moles of magma ( g magma ) , and the moles of gas inbubbles suspended in that magma ( g gas ) : ( ) g g g = + . A4 tot gas magma Within this same volume of magma, the total moles of carbon ( g Ctot ) is equal to the moles of carbon in the gas phase ( g Cgas ) andthe moles of carbon dissolved in the magma ( g Cmagma ) combined: ( ) g g g = + . A5 Ctot Cgas Cmagma
We assume that the only carbon-bearing molecule that candissolve in the magma is CO ; therefore, g g = Cmagma COmagma .Dividing by g tot and expanding gives ( ) gg gg gg g g gg = + . A6 Ctottot gastot Cgasgas magmatot COmagmamagma Table 3
Mount Etna Magma CompositionMagma component Mole fraction x SiO x TiO x Al O x FeO x MgO x CaO x Na2O x K O x P O
Note.
Taken from Iacono-Marziano et al. ( ) . Table 4
Solubility ConstantsConstant Value C CO B CO − b CO B H O − b H O ( ) + + d Al O CaO K O Na O + d FeO MgO − + d Na O K O
Note. “ Anhydrous ” case from Iacono-Marziano et al. ( ) . ⎡⎣⎢ ⎤⎦⎥ ( ) ( ) = + + + + -+ + + + + + + x x x x x xx x x x x x x x NBOO 22 2 3 . A3
K O Na O CaO MgO FeO Al OSiO TiO Al O MgO FeO CaO Na O K O The Planetary Science Journal, ( ))
K O Na O CaO MgO FeO Al OSiO TiO Al O MgO FeO CaO Na O K O The Planetary Science Journal, ( )) , 2020 December Wogan, Krissansen-Totton, & Catling e can replace g g magmatot with - gg gastot using Equation ( A4 ) . Thisleaves us with ⎛⎝⎜ ⎞⎠⎟ ( ) gg gg gg gg gg = + - Ctottot gastot Cgasgas gastot COmagmamagma Here gg CO2magmamagma is just x CO ( the mol fraction of CO in the magma;see Table 1 ) . Also, we assume that CO , CO, and CH are theonly carbon-bearing gas species, so g g g g = + + Cgas COgas COgas CHgas .Making substitutions gives ⎛⎝⎜ ⎞⎠⎟ ( ) gg gg g g gg gg = + + + - x Ctottot gastot COgas COgas CHgasgas gastot CO
Assuming the ideal gas law, g g = P P i i gas gas . Also, to makethe equation more manageable, we substitute a = gg gas gastot , whichis the total mols in the gas phase divided by the moles in thegas phase and magma combined: ( ) ( ) gg a a = + + + - P P PP x
Ctottot CO CO CH gas gas CO Magma sometimes freezes deep in the Earth as a glass before itreleases any volatiles. Measurements of volatiles like CO insuch glasses are reported in terms of mass fractions ( Wallaceet al. 2015 ) . To stay consistent with these unit conventions, weindicate the total carbon in un-degassed magma as a massfraction of CO ( m CO2tot ) . We can convert the mass fraction to amole fraction using Equation ( ) : ( ) mm gg gg = = = m x . A10 CO2tot magmaCO CO2tot COtottot Ctottot
Substituting Equation ( A10 ) into Equation ( A9 ) gives ( ) ( ) mm aa = + ++ - m P P PP x COtot magmaCO CO CO CH gasgas CO
Equation ( A11 ) is identical to Equation ( ) . A.3. Graphite Saturation and the Solubility of CO, CH , and H Several studies have shown that degassing can be affected bygraphite saturation of magma ( Hirschmann & Withers 2008 ) orby the solubility of CO, CH , and H in magma ( Hirschmannet al. 2012; Ardia et al. 2013; Wetzel et al. 2013 ) . Our modelfor outgassing speciation used throughout the main text doesnot account for these complications. Here we show that ourassumption is valid because it does not signi fi cantly change ourresults.Consider the following equilibrium: ( ) + « C O CO , A12 ( ) = »
K fa f Pa f . A13 COC O COC O
22 22
Here K is the equilibrium constant given by ( ) + T exp 47457 0.136 , and a C is the activity of carbon. Toincorporate graphite saturation into our model, we fi rst calculateoutgassing speciation using the model described in the main text ( Section 2.1 ) . Next, we check for graphite saturate by calculating the activity of carbon using Equation ( A13 ) . If < a C , then weassume the melt is not graphite saturated and that the calculation isvalid. If > a C , then we assume graphite is saturated andrecalculate outgassing speciation by replacing the carbonconservation equation ( Equation ( )) , with the graphite saturationequation with = a C ( Equations ( A13 )) . Here we are consideringgraphite saturation in the magma just before degassing occurs. Ourtreatment is different from, for example, the methods of Ortenziet al. ( ) because they are accounting for graphite saturationmuch deeper in a planet during partial melting of the mantle.Figure 10 is identical to Figure 3, except Figure 10 accountsfor graphite saturation. Graphite saturation appears to have asmall effect on the results; therefore, it is justi fi ed to ignore it.To incorporate the solubility of H , CH , and CO into ourmodel, we add the following solubility relationships to orsystem of original outgassing equations ( Section 2.1 ) : ( ) ( ) - - = = » P K xf xP exp 11.403 0.000076 , A14
22 22 ( ) ( ) - - = = »
P K xf xP exp 7.63 0.000193 , A15
44 44 ( ) ( ) ( ) ( ) - - = = »
P K xa f xa P exp 41.02 0.00056 .A16
Here pressure-dependent equilibrium constants K , K , and K are from Hirschmann et al. ( ) , Ardia et al. ( ) ,and Wetzel et al. ( ) , respectively. For Equation ( A16 ) ,we take the activity of iron to be = a Fe , based on theexperiments in Wetzel et al. ( ) . Also, we only includethe Equation ( A16 ) when < f O IW-0.55 ( IW is the iron-wustite mineral buffer ) because Wetzel et al. ( ) onlyobserved CO dissolved in magma for these low oxygenfugacities.We also alter the carbon and hydrogen atom conservationequations to accommodate for new molecules in the melt. ( )( ) ( ) mm aa = + ++ - + + m P P PP x x x COtot magmaCO CO CO CH gasgas CO CO CH ( )( ) ( ) mm aa = + ++ - + + m P P PP x x x
21 2 . A18
H Otot magmaH O H O H CH gasgas H O H CH
Here x i is the mol fraction of species i in the melt.Figure 11 is identical to Figure 3, except Figure 11 accountsfor H , CH , and CO solubility in magma. The solubility ofthese three molecules has a small effect on the results; thereforethey can be ignored. A.4. Closed System Cooling and Chemical Kinetics
Our model for volcanic outgassing is a thermodynamicequilibrium model. We assume that during magma eruptions,gas bubbles chemically and thermally equilibrate with magma,and then they are released to the atmosphere unaltered ( Figure 1 ) . This does not exactly re fl ect real degassing.In reality, the chemical composition of gas bubbles changesas bubbles leave the magma and enter the atmosphere13 The Planetary Science Journal, ( ))
Our model for volcanic outgassing is a thermodynamicequilibrium model. We assume that during magma eruptions,gas bubbles chemically and thermally equilibrate with magma,and then they are released to the atmosphere unaltered ( Figure 1 ) . This does not exactly re fl ect real degassing.In reality, the chemical composition of gas bubbles changesas bubbles leave the magma and enter the atmosphere13 The Planetary Science Journal, ( )) , 2020 December Wogan, Krissansen-Totton, & Catling Moussallam et al. 2019; Kadoya et al. 2020 ) . As a bubbleleaves magma, it cools down and new chemical equilibria arepreferred. When a gas bubble fi rst begins cooling, it is still veryhot, so chemical reactions keep the bubble near chemicalequilibrium. Once the bubble is cold enough, chemicalreactions slow, and ultimately cease, quenching or freezingthe chemical composition of the gas bubble. Therefore, thecooling process alters the chemistry of the gas.Gas re-equilibration to lower temperatures explains theobserved chemistry of volcanic gases globally ( Moussallamet al. 2019 ) , and Oppenheimer et al. ( ) provides a speci fi cexample of this phenomenon at in the Kilauea volcano inHawaii. During eruptions at Kilauea, gas bubbles in the magmawould rise to the surface. As the bubbles rose in the magma,they adiabatically expanded, which cooled the gas below thetemperature of the magma. Chemical reactions during adiabaticexpansion changed the chemical make-up of the bubble.For the purposes of understanding potential CH biosigna-ture false positives from volcanoes, we need to know if bubble cooling might generate a substantial amount of CH . Here we fi rst consider the kinetics of methane generation and show thatreactions are likely too slow to generate substantial CH duringgas cooling. Next, we show that our Monte Carlo simulationresults ( Figure 4 ) remain qualitatively unchanged, even if ourkinetics calculations are wrong, and CH can be generatedduring gas cooling.CO or CO is converted to CH through either of the netreactions ( Schaefer & Fegley 2010 ) : ( ) + « + CO 3H CH H O, A19 ( ) + « +
CO 4H CH 2H O. A20
The rate-limiting step to either CO or CO conversion to CH isdebated in the literature ( Zahnle & Marley 2014 ) , but thefollowing are two solid candidates and their corresponding rateconstants: ( ) + + H H CO CH OH, A21 ( ) ( ) = ´ - - k T
10 10
Figure 10.
Identical to Figure 3, except here we account for graphite saturation in the melt. As in Figure 3, ( a ) is for ocean worlds and ( b ) is for Earth-like worlds.Graphite saturation has a small effect on the results. Figure 11.
Identical to Figure 3, except here we account for the solubility of H , CH , and CO in the melt. As in Figure 3, ( a ) is for ocean worlds and ( b ) is for Earth-like worlds. H , CH , and CO solubility have a small effect on the results. The Planetary Science Journal, ( ))
Identical to Figure 3, except here we account for the solubility of H , CH , and CO in the melt. As in Figure 3, ( a ) is for ocean worlds and ( b ) is for Earth-like worlds. H , CH , and CO solubility have a small effect on the results. The Planetary Science Journal, ( )) , 2020 December Wogan, Krissansen-Totton, & Catling ) + H H CO CH , A23 ( ) ( ) = ´ - - k T
12 11
Here k and k are rate constants ( cm s − ) . The lifetime ofCO or CO conversion to CH is thus one of the following: ( ) ( ) t = Nk N N
CO , A25 CO
10 H H CO ( ) ( ) t = Nk N N
CO , A26
12 CO12 H H CO ( ) ( ) t = Nk N N
CO , A27
10 2 CO10 H H CO
22 2 ( ) ( ) t = Nk N N
CO . A28
12 2 CO12 H H CO
Here τ is the chemical lifetime in seconds, and N i is the numberdensity of species i in molecules cm − .Figure 12 shows timescales of CH generation ( Equations ( A25 ) – ( A28 )) during the closed system cooling ofsubmarine volcanic gas. To determine gas chemistry just before abubble is released from magma, we use our speciation model ( Section 2.1 ) . At 1473 K, we calculate gas speciation assuming P =
400 bar, = f O FMQ-4, = m COtot , and = m H Otot .We then calculate new chemical equilibrium as the gas cools,assuming it is a closed system ( i.e., we assume the gas is Figure 12. ( a ) Equilibrium composition as a function of temperature for a submarine volcanic gas that is cooled as a closed system and ( b ) timescales of CH formation during closed system cooling. Timescales of volcanic gas cooling are not shown or calculated. Figure 13.
The blue histograms in ( a ) and ( b ) are identical to Figures 4 ( c ) and ( d ) , and orange histograms are identical Monte Carlo simulations, except they accountfor the closed system cooling of volcanic gases to equilibrium temperatures observed on Earth (
800 to 1500 K ) . To calculate CH fl uxes, we used modern Earth ’ smagma production rate. The Planetary Science Journal, ( ))
800 to 1500 K ) . To calculate CH fl uxes, we used modern Earth ’ smagma production rate. The Planetary Science Journal, ( )) , 2020 December Wogan, Krissansen-Totton, & Catling hermally and chemically decoupled from the magma;Figure 12 ( a )) . Figure 12 ( b ) shows the corresponding timescale ofCH generation ( Equations ( A25 ) – ( A28 )) at each temperature.The quench temperature ( i.e., the temperature where out-gassing chemistry is frozen-in due to slow kinetics ) of CH depends on the cooling timescale of volcanic gases ( not shownin Figure 12 ) . CH should quench where the cooling timescaleis about the same as the timescale of CH generation. Aftergases are released from a submarine volcano, we suspect theycool from magma temperatures to ocean temperatures on theorder of seconds. If this is the case, then the CH quenchtemperature is probably > content of the gas ( Figure 12 ( a )) .Suppose that the CH quench temperature was instead1000 K. In this case, the CH content of the gas would beincreased by about a factor of fi ve ( Figure 12 ( a )) . There are twoways that a ∼ quench is possible. First, gas coolingcould occur on timescales of months rather than seconds.According to Figure 12 ( b ) , month-long gas cooling shouldquench CH by 1000 K. Second, catalysts could dramaticallyspeed up the reactions creating CH , which might allow forquench temperatures near 1000 K for even gas coolingtimescales of seconds. In the following two paragraphs, weshow that either of these scenarios would not signi fi cantlychange our results.To demonstrate that re-equilibration of gases to feasiblelower temperatures does not change our conclusions, assuminglow CH quench temperatures can be achieved, we performanother Monte Carlo simulation identical to the one describedin Section 2.2, except we account for closed system cooling ofvolcanic gases. In the Monte Carlo simulation, we fi rstcalculate gas composition using our outgassing model ( Section 2.1 ) ; then we re-equilibrate this gas mixture to theuniformly sampled gas equilibrium temperature between 800and 1500 K. This range of gas equilibrium temperatures is therange observed in Earth ’ s volcanic gases ( Moussallam et al.2019 ) . In cases where the randomly drawn gas equilibriumtemperature is higher than the magma temperature, we assumeno closed system cooling occurs.Figure 13 is identical to Figures 4 ( c ) and ( d ) , exceptFigure 13 accounts for closed system cooling of gases. Closedsystem cooling allows more CH production on average, butstill only 0.3% and 0.1% of calculations for ocean worlds orearth-like worlds, respectively, produce more than 10 TmolCH yr − . The probability of volcanic CH fl uxes beingcomparable to modern Earth ’ s biological fl ux (
30 Tmol yr − ) is still low.In summary, changes in gas chemistry during cooling mightcause our speciation model to under-predict the CH producedby an amount that does not change our conclusionssigni fi cantly. Further consideration of the kinetics of CH generation in volcanic gases is beyond the scope of this paper. Appendix BPhotochemical Model Boundary Conditions
Table 5 shows boundary conditions used for the
Atmos photochemical model. We used the same H O and temperaturepro fi le as Kharecha et al. ( ) for all simulations. Theversion of Atmos that we used has updated rate constants andH O cross sections following Ranjan et al. ( ) .Every simulation for planets orbiting the Sun uses a solarspectrum at 2.7 Ga, calculated via the methods described in Claire et al. ( ) , although our results are not sensitive to theage of the Sun. For planets orbiting an M8V star, we useestimates of TRAPPIST-1 ʼ s spectrum derived by Lincowskiet al. ( ) , scaled so that the solar constant of the planet is0.822 relative to modern Earth ’ s. We use this solar constantbecause it places the simulated planet at the same relativedistance from the inner edge of the habitable zone as Earthtoday ( Kopparapu et al. 2013 ) .All of our models include the modern production rate of NOfrom lightning. ORCID iDs
Nicholas Wogan https: // orcid.org / References
Adam, P. S., Borrel, G., & Gribaldo, S. 2018, PNAS, 115, E1166Anderson, K. R., & Poland, M. P. 2017, NatGe, 10, 704Ardia, P., Hirschmann, M. M., Withers, A., & Stanley, B. 2013, GeCoA,114, 52
Table 5
Boundary Conditions for Photochemical ModelingChemical species Deposition velo-city ( cm s − ) Mixing ratio Flux ( moleculescm − s − ) O L L O ´ - L L
H O L L H L L OH L L HO L L
H O × − L L H L F CH CO × − L F CH HCO L L
H CO × − L L CH L variable CH L L
C H L L NO × − L L NO × − L L
HNO L L O × − L L
HNO × − L L
H S × − L F CH SO L L S L L
HSO L L
H SO L L SO L F CH SO L L SO aerosol 1 × − L L S aerosol 1 × − L L
Hydrocarbonaerosol 1 × − L L CO L Variable L N L L Note.
Species included in the photochemical scheme with a deposition velocityand fl ux of 0 include N, C H , C H , CH C H, CH CCH , C H , C H CHO,C H , C H , C H , C H OH, C H OH, C H , C H , CH, CH O , CH O,CH CO, CH CO, CH CHO, C H , ( CH ) , C H, C , C H , HCS, CS , CS,OCS, S, and HS. Here deposition velocities follow those used by Schwieter-man et al. ( ) . The Planetary Science Journal, ( ))
Species included in the photochemical scheme with a deposition velocityand fl ux of 0 include N, C H , C H , CH C H, CH CCH , C H , C H CHO,C H , C H , C H , C H OH, C H OH, C H , C H , CH, CH O , CH O,CH CO, CH CO, CH CHO, C H , ( CH ) , C H, C , C H , HCS, CS , CS,OCS, S, and HS. Here deposition velocities follow those used by Schwieter-man et al. ( ) . The Planetary Science Journal, ( )) , 2020 December Wogan, Krissansen-Totton, & Catlingrmstrong, K., Frost, D. J., McCammon, C. A., Rubie, D. C., & Ballaran, T. B.2019, Sci, 365, 903Barstow, J. K., & Irwin, P. G. 2016, MNRAS: Letters, 461, L92Benner, S. A., Bell, E. A., Biondi, E., et al. 2019, ChemSystemsChem, 2,e1900035Burcat, A., & Ruscic, B. 2005, Third Millenium Ideal Gas and CondensedPhase Thermochemical Database for Combustion ( with update from activethermochemical tables ) Tech Rep. ANL-05 / ( Argonne National Lab,Argonne, IL ) , doi:10.2172 / ( Cambridge: Cambridge Univ. Press ) Catling, D. C., & Zahnle, K. J. 2020, SciA, 6, eaax1420Claire, M. W., Catling, D. C., & Zahnle, K. J. 2006, Geobiology, 4, 239Claire, M. W., Sheets, J., Cohen, M., et al. 2012, ApJ, 757, 95Crisp, J. A. 1984, JVGR, 20, 177Domagal-Goldman, S. D., Segura, A., Claire, M. W., Robinson, T. D., &Meadows, V. S. 2014, ApJ, 792, 90Doyle, A. E., Young, E. D., Klein, B., Zuckerman, B., & Schlichting, H. E.2019, Sci, 366, 356Fauchez, T. J., Turbet, M., Villanueva, G. L., et al. 2019, ApJ, 887, 194Fischer, W. W., Hemp, J., & Johnson, J. E. 2016, AREPS, 44, 647French, B. M. 1966, RvGeo, 4, 223Gaillard, F., & Scaillet, B. 2014, E&PSL, 403, 307Gilbert, W. 1986, Natur, 319, 618Gonnermann, H. M., & Manga, M. 2013, in Modeling Volcanic Processes: ThePhysics and Mathematics of Volcanism, ed. R. M. C. Lopes,S. A. Fagents, & T. K. P. Gregg ( Cambridge: Cambridge Univ. Press ) , 55Grenfell, J. L., Gebauer, S., Paris, P. v., Godolt, M., & Rauer, H. 2014, P&SS,98, 66Grenfell, J. L., Grießmeier, J.-M., Patzer, B., et al. 2007, AsBio, 7, 208Guzmán-Marmolejo, A., Segura, A., & Escobar-Briones, E. 2013, AsBio,13, 550Harman, C. E., Schwieterman, E. W., Schottelkotte, J. C., & Kasting, J. F.2015, ApJ, 812, 137Hauglustaine, D., Heinze, C., Holland, E., et al. 2007, in Climate Change 2007:The Physical Science Basis. Contribution of Working Group I to the FourthAssessment Report of the Intergovernmental Panel on Climate Change, ed.S. Solomon et al. ( Cambridge: Cambridge Univ. Press ) Hirschmann, M. M., Withers, A., Ardia, P., & Foley, N. 2012, E&PSL, 345, 38Hirschmann, M. M., & Withers, A. C. 2008, E&PSL, 270, 147Holland, H. D. 1984, The Chemical Evolution of the Atmosphere and Oceans ( Princeton, NJ: Princeton Univ. Press ) Huppert, H. E., Sparks, R. S. J., Turner, J. S., & Arndt, N. T. 1984, Natur,309, 19Iacono-Marziano, G., Morizet, Y., le Trong, E., & Gaillard, F. 2012, GeCoA,97, 1Joyce, G. F., & Szostak, J. W. 2018, Cold Spring Harbor Perspectives inBiology, 10, a034801Kadoya, S., Catling, D. C., Nicklas, R. W., Puchtel, I. S., & Anbar, A. D. 2020,Geochemical Perspectives Letters, 13, 25Kasting, J. F., & Brown, L. L. 1998, in The Molecular Origins of Life, ed.A. Brack ( Cambridge: Cambridge Univ. Press ) , 35Kasting, J. F., & Catling, D. 2003, ARA&A, 41, 429Kharecha, P., Kasting, J., & Siefert, J. 2005, Geobiology, 3, 53Kopparapu, R. K., Ramirez, R., Kasting, J. F., et al. 2013, ApJ, 765, 131Krissansen-Totton, J., Garland, R., Irwin, P., & Catling, D. C. 2018a, AJ,156, 114Krissansen-Totton, J., Olson, S., & Catling, D. C. 2018b, SciA, 4, eaao5747le Voyer, M., Hauri, E. H., Cottrell, E., et al. 2019, GGG, 20, 1387Lehmer, O. R., Catling, D. C., Parenteau, M. N., & Hoehler, T. M. 2018, ApJ,859, 171Li, Z.-X. A., & Lee, C.-T. A. 2004, E&PSL, 228, 483 Liggins, P., Shorttle, O., & Rimmer, P. B. 2020, E&PSL, 550, 116546Lincowski, A. P., Meadows, V. S., Crisp, D., et al. 2018, ApJ, 867, 76Luger, R., & Barnes, R. 2015, AsBio, 15, 119Lustig-Yaeger, J., Meadows, V. S., & Lincowski, A. P. 2019, AJ, 158, 27Lyons, T. W., Reinhard, C. T., & Planavsky, N. J. 2014, Natur, 506, 307Marchi, S., Bottke, W., Elkins-Tanton, L., et al. 2014, Natur, 511, 578McKenzie, D. 2020, JPet, egaa061Meadows, V. S. 2017, AsBio, 17, 1022Meadows, V. S., Reinhard, C. T., Arney, G. N., et al. 2018, AsBio, 18, 630Miller, S. L., & Urey, H. C. 1959, Sci, 130, 245Moussallam, Y., Oppenheimer, C., & Scaillet, B. 2019, E&PSL, 520, 260O ’ Neill, H. S. 1987, AmMin, 72, 67, http: // / ammin / AM72 / AM72_67.pdfOppenheimer, C., Scaillet, B., Woods, A., et al. 2018, NatGe, 11, 678Ortenzi, G., Noack, L., Sohl, F., et al. 2020, NatSR, 10, 10907Pavlov, A. A., Brown, L. L., & Kasting, J. F. 2001, JGR, 106, 23267Ranjan, S., Schwieterman, E. W., Harman, C., et al. 2020, ApJ, 896, 148Reeves, E. P., & Fiebig, J. 2020, Elements: An International Magazine ofMineralogy, Geochemistry, and Petrology, 16, 25Ritson, D. J., Battilocchio, C., Ley, S. V., & Sutherland, J. D. 2018, NatCo,9, 1Rugheimer, S., & Kaltenegger, L. 2018, ApJ, 854, 19Rugheimer, S., Kaltenegger, L., Segura, A., Linsky, J., & Mohanty, S. 2015,ApJ, 809, 57Sasselov, D. D., Grotzinger, J. P., & Sutherland, J. D. 2020, SciA, 6, eaax3419Sauterey, B., Charnay, B., Affholder, A., Mazevet, S., & Ferrière, R. 2020,NatCo, 11, 1Schaefer, L., & Fegley, B., Jr 2010, Icar, 208, 438Scheidegger, K. 1973, JGR, 78, 3340Schindler, T. L., & Kasting, J. F. 2000, Icar, 145, 262Schwieterman, E. W., Reinhard, C. T., Olson, S. L., et al. 2019, ApJ, 874, 9Segura, A., Kasting, J. F., Meadows, V., et al. 2005, AsBio, 5, 706Sholes, S. F., Krissansen-Totton, J., & Catling, D. C. 2019, AsBio, 19, 655Sleep, N. H., & Zahnle, K. 2001, JGR, 106, 1373Smith, M. L., Claire, M. W., Catling, D. C., & Zahnle, K. J. 2014, Icar, 231,51Stamper, C., Melekhova, E., Blundy, J., et al. 2014, CoMP, 167, 954Strobel, D. F., Atreya, S. K., Bézard, B., et al. 2009, in Titan from Cassini-Huygens, ed. R. H. Brown et al. ( Berlin: Springer ) , 235Sutherland, J. D. 2016, Angewandte Chemie International Edition, 55, 104Symonds, R. B., Rose, W. I., Bluth, G. J. S., & Gerlach, T. M. 1994, inVolatiles in Magma, Reviews in Mineralogy, Vol. 30, ed. M. R. Carrol &J. R. Halloway ( Chantilly, VA: Mineralogical Society of America ) , 1Tian, F., France, K., Linsky, J. L., Mauas, P. J., & Vieytes, M. C. 2014,E&PSL, 385, 22Tian, F., Kasting, J., & Zahnle, K. 2011, E&PSL, 308, 417Tutolo, B. M., Seyfried, W. E., & Tosca, N. J. 2020, PNAS, 117, 14756Wallace, P. J. 2005, JVGR, 140, 217Wallace, P. J., Plank, T., Edmonds, M., & Hauri, E. H. 2015, Volatiles inMagmas ( Amsterdam: Elsevier ) , 163, doi:10.1016 / B978-0-12-385938-9.00007-9Wetzel, D. T., Rutherford, M. J., Jacobsen, S. D., Hauri, E. H., & Saal, A. E.2013, PNAS, 110, 8010Wones, D. R., & Gilbert, M. C. 1969, AmJS, 267, 480Wordsworth, R., & Pierrehumbert, R. 2014, ApJL, 785, L20Wunderlich, F., Scheucher, M., Godolt, M., et al. 2020, ApJ, 901, 126Zahnle, K., Claire, M., & Catling, D. 2006, Geobiology, 4, 271Zahnle, K., Haberle, R. M., Catling, D. C., & Kasting, J. F. 2008, JGRE, 113,E11004Zahnle, K. J. 1986, JGR, 91, 2819Zahnle, K. J., Lupu, R., Catling, D. C., & Wogan, N. 2020, PSJ, 1, 11Zahnle, K. J., & Marley, M. S. 2014, ApJ, 797, 41 The Planetary Science Journal, ( ))