The decoupling of Nitrogen from the Noble Gases in specific regions

The decoupling of Nitrogen from the Noble Gases in specific regions of Earth’s mantle. S. Mikhail1 and D. A.
Sverjensky2. 1The University of St. Andrews, The Department of Earth and Environmental Sciences, St. Andrews,
UK, email: [email protected], 2Johns Hopkins University, The Department of Earth and Planetary Sciences,
Baltimore, M.D., USA
Introduction:
One of the biggest puzzles in science today regards
the origin and uniqueness of life on Earth, and to address this requires an understanding of planetary habitability. An important parameter in this endeavour is the
composition and evolution of the atmosphere. A major
process in forming planetary atmospheres is mantle
degassing via volcanism. Where the primary control on
what elements are degassed by this process is their
compatibility in mantle minerals relative to melts and
fluids. The geochemistry of all telluric planetary atmospheres are dominated by carbon (CO2), nitrogen
(N2), and the noble gases. For elements like carbon and
nitrogen, the redox state of the environment is paramount in governing their speciation, which in turn controls their partitioning behaviour. However the noble
gases are insensitive to redox reactions, and therefore
provide a unique insight. In combination, the geochemistry of C, N, and the noble gases can illuminate the
processes and sources responsible for atmosphere formation and evolution.
Earth’s nitrogen abundance
Molecular nitrogen is a neutrally charged and highly-volatile molecule, and therefore, N2 behaves like a
noble gas. In the absence of speciation data for nitrogen at high temperatures, it has been commonly assumed that nitrogen in the mantle behaves like a noble
gas. Therefore, ratios such as N2/He and N2/Ar have
been measured, and the data modeled to constrain the
volatile fluxes during subduction [1] the nitrogen
budget of the mantle [2], and the evolution of the atmosphere through time [3]. Based on this assumption,
the empirical data show that Earth has a missing nitrogen conundrum (Fig.1). This is derived by summing up
the volatile abundances for the bulk silicate Earth
(BSE) and comparing those to chondritic meteorites.
This approach establishes that nitrogen is the most
depleted of the volatile elements [4].
Fig. 1. The relative abundances of the volatile elements in the BSE relative to chondrites, from [4].
This missing nitrogen is either stored in the mantle,
core, lost during accretion, or a combination of all
three [4-5]. However, there are also the possibilities of
(a) no nitrogen depletion, or (b) the magnitude of the
nitrogen depletion is overestimated. These bold alternatives are based on what is now known about the speciation of nitrogen in the mantle (Fig.2).
The speciation of nitrogen in the silicate mantle
Nitrogen can behave like an atmophile, lithophile,
or siderophile element depending upon its redox state
(i.e. N20 vs. NH4+). Recent theoretically- [6] and experimentally- [7] determined equilibrium constants have
shown that ammonic nitrogen should dominate over
molecular nitrogen in the uppermost mantle where the
redox state (fO2) is buffered below the quartz-fayalitemagnetite reaction, which is in fact the redox state of
Earth’s convecting upper mantle (Fig.2).
Fig. 2. The speciation of nitrogen in fluids in the
upper mantle [6].
The data in Fig.2 provide a significant insight, because nitrogen compatibility is directly related to speciation: molecular nitrogen is highly incompatible
whereas ammonic nitrogen is moderately compatible
in K-bearing phases [8]. Additionally, ammonium can
dissolve as a trace component in K-absent phases [9].
Therefore, the predicted stability of NH4+ (Fig.2)
alongside the the stability of K-bearing phases thoughout the entire mantle pressure and depth range [10]
means the inaccessible (i.e. unsampled) parts of the
mantle (>98% by volume) have the potential to store
significant quantities of ammonic nitrogen.
Evidence from planetary atmpsheres
By definition, there are no data from the inaccessibble portions of the mantle. However, there are data
for the atmospheres of planets with and without plate
tectonics. The data shown in Fig.2 make the following
prediction: A planet with Earth-like subduction zones
should have degassed more nitrogen relative to the
primordial noble gases compared with a planet devoid
of Earth-like subduction zones. This is because volcanism sourced from the ambient mantle (MORB- and
OIB source) should sample nitrogen dominantly present as NH4+. Conversely, volcanism sourced from the
mantle wedge (arc magmatism) should sample nitrogen dominantly present as N20. Note that, regardless of
the mantle redox state, the very low partition coefficients of the noble gases will not vary (e.g. they remain
highly incompatible). We compiled a dataset for the
atmospheric chemistries of Earth, Mars, and Venus and
find a striking match to the prediction, Earth’s atmosphere shows a large enrichment of N/primordial noble
gas ratios relative to the atmospheres of Mars and Venus (Fig.3).
Fig. 3. The relative abundances for nitrogen relative to the primordial noble gases in the atmopsheres
of Earth, Mars, and Venus. Data from [11].
Owing to the lack of data for volcanic degassing
over geologic time, we are unable to quantify the relative importance of the N2/noble gas mantle wedge
mechanism to that of other parts of the deep N-cycle.
However, we can make a simple calculation of how
long it would take to produce the Earth's atmospheric
N2/primordial noble gas enrichment relative to the Venusian and Martian atmospheres. Assuming all volcanism being equal in nature between the planets (for example, rifting and hotspots), and the Earth being exceptional in having arc systems, we use an empirical
nitrogen flux from the Central American volcanic arc
system of 0.63x1010 mol yr-1 [1]. To apply this flux
globally over 1-2 Ga, it must be amplified to represent
the global flux (factor of 20 or 10, respectively), to
produce the Earth's N2 enrichment.
Overall, our results suggest the uniqueness of
Earth's habitable outer layer has been strongly influenced by subduction-zone plate tectonics. In contrast,
the atmospheres, and potentially the habitabilities, of
Venus and Mars have evolved along different paths
through geologic time because they lacked plate tectonics.
Broader Implications
First, the storage capacities and total abundance for
nitrogen in the BSE (Fig.1), and other telluric planets
require revision. For example, the data in Fig.3 imply
that the relative abundance for nitrogen in the Venusian and Martian mantles should exceed that of Earth's
mantle.
Second, on the basis of the diagrams in Figs. 2 & 3,
the behaviour of nitrogen and the noble gases are clearly not equivalent in aqueous upper mantle fluids within
the telluric planets [6]. Therefore, estimations of volatile abundances and fluxes based on N2/noble gas systematics may produce inaccurate flux estimates [1] and
may underestimate planetary volatile budgets, e.g.
Earth’s missing nitrogen conundrum (shown in Fig.1).
Finally, the same caveat applies to the use of C/N
ratio as a tool to constrain the chondritic or degassed
nature of melts from Earth, and other extra-terrestrrial
planets and moons [e.g. 12]. This is because the nitrogen can behave like a lithophile element (similar to K+
& Rb+) in the mantle where the redox conditions are
between the QFM and IW buffers. Conversely, during
melting under these redox conditions carbon will be
stable as graphite or diamond (pressure dependant
[12]), and as such, will not partition like a lithophile
element.
References Cited:
[1] Fischer et al., (2002), Science, 297, 1154-1157; [2]
Marty, (1995), Nature, 377, 326 – 329; [3] Marty et al.,
(2013), Science, 342, 101- 104; [4] Marty, (2012),
EPSL, 314, 1-114; [5] Halliday, (2013), GCA,
105,146-171; [6] Mikhail & Sverjensky, 2014, Nature
Geoscience, 7, 816–819; [7] Li & Keppler, 2014,
GCA, 129, 13-32; [8] Watenpaul et al., (2010),
Chem.Geol, 270, 240-248; [9] Li et al., (2013), EPSL,
378, 311–323; [10] Harlow & Davies, (2004) Lithos,
77, 647-653 [11] Porcelli & Pepin, (2002), Treatise on
Geochemistry; [12] Stagno et al., (2013), Nature, 493,
84–88