Stable Isotope Measurements of Martian Atmospheric CO2 at the Phoenix Landing Site

See allHide authors and affiliations

Science  10 Sep 2010:
Vol. 329, Issue 5997, pp. 1334-1337
DOI: 10.1126/science.1192863


Carbon dioxide is a primary component of the martian atmosphere and reacts readily with water and silicate rocks. Thus, the stable isotopic composition of CO2 can reveal much about the history of volatiles on the planet. The Mars Phoenix spacecraft measurements of carbon isotopes [referenced to the Vienna Pee Dee belemnite (VPDB)] [δ13CVPDB = –2.5 ± 4.3 per mil (‰)] and oxygen isotopes [referenced to the Vienna standard mean ocean water (VSMOW)] (δ18OVSMOW = 31.0 ± 5.7‰), reported here, indicate that CO2 is heavily influenced by modern volcanic degassing and equilibration with liquid water. When combined with data from the martian meteorites, a general model can be constructed that constrains the history of water, volcanism, atmospheric evolution, and weathering on Mars. This suggests that low-temperature water-rock interaction has been dominant throughout martian history, carbonate formation is active and ongoing, and recent volcanic degassing has played a substantial role in the composition of the modern atmosphere.

Carbon dioxide (CO2) is the dominant gas in the martian atmosphere, and its stable isotopic composition can provide constraints on Mars volatile reservoirs, carbon cycle, and climate history. CO2 is involved in many major planetary processes that affect volatile reservoirs of the planet’s surface, including volcanic degassing, atmospheric loss, and rock weathering and carbonate formation. Additionally, CO2 rapidly equilibrates with water, and the oxygen isotopic composition of CO2 may provide insight into the action of liquid water at the planet’s surface. Stable isotopes are effective tracers of the history of CO2 because they are fractionated in predictable ways by each geological process. Thus, the stable isotopic composition of the atmospheric CO2 of Mars will record the sum of the geologic processes in which it has participated, making it possible to quantifiably constrain them.

Previous measurements on the surface of Mars of the isotopic composition of atmospheric CO2 have lacked sufficient precision to be useful for meaningful interpretation (Table 1). Measurements of trapped CO2 gas in martian meteorite EETA 79001 impact glass claimed higher precision and showed elevated δ13C values (Table 1), supporting a hypothesis suggesting that the modern martian atmosphere was enriched in 13C (1). However, more recent Earth-based spectroscopic measurements of the martian atmosphere have measured the martian CO2 to be depleted in 13C relative to CO2 in the terrestrial atmosphere (2).

Table 1

All published measurements of the isotopic composition of martian atmospheric CO2, including results from this study (1, 2, 2832). Uncertainties for the Phoenix measurements are calculated as SE from the 2σ SD of all of the measurements. GC, gas chromatograph; MS, mass spectrometer.

View this table:

Here, we report high-precision measurements, made at the surface of Mars, of atmospheric CO2. The Thermal and Evolved Gas Analyzer (TEGA) instrument on the Mars Phoenix Lander (3) included a magnetic-sector mass spectrometer (EGA) (4), which had a goal of measuring the isotopic composition of martian atmospheric CO2 to within 0.5%. Its counting rates were averaged over ~5-min increments for masses 44, 45, and 46. These masses include all of the major stable isotopes of carbon (12C and 13C) and oxygen (16O, 17O and 18O) and are universally used to make isotopic measurements of CO2 in terrestrial laboratories. Isotope ratios were calculated from these averages and were corrected for the deadtime in the preamplifier discriminator circuit (fig. S1), the background measured in the instrument before ingesting gas, and the 12C17O16O contribution to mass 45. The correction for 17O contribution is relatively small as compared with our uncertainties because of the extremely low abundance of 17O in the solar system. Therefore, only unrealistically large uncertainties in the martian δ17O value [>60 per mil (‰)] could affect the magnitude of our uncertainties. Finally, the corrected isotope ratios were calibrated by referencing analyses of a CO2 gas of known composition carried to Mars on the lander (Fig. 1) (5).

Fig. 1

Corrected but uncalibrated isotope ratios of measurements performed at the martian surface by the TEGA instrument. Small triangles are individual ~5-min measurements of martian atmospheric CO2 that have not yet been calibrated by results from calibration gas measurements. Small circles are corrected measurements of calibration gas. Larger triangle and circle are averages showing 2σ SD. These are also plotted with results from the laboratory measurement of the TEGA calibration gas before flight. Data are listed in Table 1 and in (5).

The Phoenix results (Table 1) show that the carbon isotopic composition of martian atmospheric CO2 is not enriched in 13C, which is in agreement with (2), and are within the uncertainty of most previous isotopic measurements (Fig. 2). The δ18O referenced to the Vienna standard mean ocean water (VSMOW) (δ18OVSMOW) value of martian atmospheric CO2 (31.0‰) is lighter than terrestrial atmospheric CO2 (41‰), whereas the δ13C referenced to the Vienna Pee Dee belemnite (VPDB) (δ13CVPDB) (–2.5‰) is similar to terrestrial atmospheric CO2 (–7‰). However, this result for carbon is in contradiction with values reported from trapped gases in impact glass from EETA 79001 (1), which measured CO2 that was more enriched in 13C. It is possible that this trapped CO2 was derived from the impact volatilization of carbonates that were enriched in 13C. This is indicated by the presence of martian carbonates elsewhere in EETA 79001, and the suggestion that the impact glass from which the gas measurements were made may include a martian regolith component (6), which probably contained 2 to 3% carbonate (3).

Fig. 2

Corrected and calibrated results of martian atmospheric CO2 measurement plotted next to results from Viking lander (28, 29), Earth-based spectroscopy (2), and measurements of calibration gas on Earth and Mars. The result for calibration gas on Mars is the uncalibrated value to show the magnitude of the calibration correction. Error bars on TEGA measurements are 2σ SE.

Carbonates in martian meteorites preserve the carbon and oxygen isotopic composition of the CO2 from which they formed and provide an opportunity to look back into time to evaluate the history of carbon and H2O on the planet. On Earth, carbonates are typically fractionated in both carbon and oxygen from both CO2 and water. The δ13C of terrestrial CO2 averages around –7‰, which is very similar to magmatic CO2 (7). The δ18O averages ~41‰ and is strongly influenced by low-temperature equilibration with seawater (δ18O = 0‰), which has a fractionation factor of ~40‰ at 25°C (7). Terrestrial carbonates in equilibrium with these reservoirs at 25°C would have δ13C ~ 3‰ and δ18O ~ 29‰ (7). Variations in these isotopic values can yield valuable information regarding the influence of biological processes, temperatures of formation, and other details of the carbonate formation environment.

Unfortunately, δ13C measurements from acid extractions and thermal analyses of martian meteorites range from 65‰ to –20‰, revealing substantial variability that has been very difficult to understand (815). Much of these data come from three martian meteorites that each sample alteration events from different periods of martian history: ALH 84001 (<4 billion years ago) (16), Nakhla (<600 million years ago) (17), and EETA 79001 (<200 million years ago) (18). A large hindrance for interpreting these data has been a lack of information about the isotopic composition of larger reservoirs on Mars.

ALH 84001 contains carbonates that have been dated to 3.9 billion years ago and are generally enriched in 13C (14, 16). This signature is likely to be originally derived from the ancient martian atmospheric CO2 that had been enriched in 13C by early atmospheric loss processes (19). In situ ion microprobe measurements of the δ13CVPDB of ALH 84001 carbonates range from 30‰ to 65‰, with δ18OVSMOW ion microprobe measurements ranging from –10‰ to 25‰ (20, 21). Isotope measurements of CO2 extracted by phosphoric acid generally agree with these measurements but also indicate the presence of a lighter carbon component with δ13C near 10‰ (Fig. 3).

Fig. 3

Carbon and oxygen isotope measurements of martian meteorite carbonates. All measurements (except where noted) are results from phosphoric acid treatment of meteorites at a variety of temperatures (8, 1015). The ALH 84001 SIMS data are a combination of in situ carbon (20) and oxygen (21) data by use of Mg content to correlate the two data sets. Arrows indicate probable sources for isotope variation within the diagram, as discussed in the text. The dashed oval represents the equilibrium composition of carbonates that could form from the modern martian atmosphere measured in this study.

Nakhla carbonates are in general less abundant than ALH 84001 carbonates, and the aqueous alteration event has been dated to be ~600 million years old (17). Phosphoric acid measurements show a similar pattern to ALH 84001 measurements, with some measurements showing very enriched δ13C values near 40‰ and other measurements clustering around 10‰ (10, 14).

EETA 79001 is a Shergottite that features a “white druse” material that has been suggested to be martian in origin on the basis of petrologic arguments (22); isotopic measurements of this material result in a δ13CVPDB near 8‰ and a δ18OVSMOW near 22‰ (12, 15). These isotopic values have an oxygen isotope composition that is too enriched for Antarctic weathering, which features waters extremely depleted in 18O (15, 23).

When all of the published carbonate analyses are plotted next to the Phoenix results, they reveal a pattern that suggests a potential 4-billion-year history of H2O and CO2 on Mars (Fig. 3). This history is dominated by three components: modern martian atmosphere, early martian atmosphere, and magmatic CO2.

The modern atmosphere is recorded by carbonates in the 200-million-year-old martian meteorite EETA 79001. The isotopic composition of these carbonates is consistent with formation at low temperatures (~0 to 25°C) from the modern martian atmosphere measured by Phoenix (Fig. 3). In addition, some carbonate measurements in Nakhla, Governor Valadares, and ALH 84001 are also in this range, suggesting that these meteorites also contain a modern carbonate component.

The ancient atmosphere of Mars is likewise recorded in carbonates from the oldest martian meteorite ALH 84001. The average carbon and oxygen isotopic composition of the carbonates in ALH 84001 suggest an ancient atmospheric composition enriched in δ13C but of similar δ18O content to the modern atmosphere (24). These isotope values are more widely scattered because of large heterogeneities within the meteorite itself (20, 21). Careful ion microprobe analyses have shown the values to follow a covariant trend, possibly reflecting kinetic fractionation processes that were active during their precipitation (20).

Magmatic CO2 appears to be preserved in the martian meteorite Zagami. Phosphoric acid CO2 extractions of Zagami (10) show a mixing line between modern martian CO2 and a component with δ13C of –20 to –30‰ and δ18O of ~10‰ (Fig. 3). This is consistent with thermal analyses of martian meteorites (25) and theoretical high-temperature equilibrium between silicate rocks and CO2 (15).

All of these measurements can be synthesized into a model that broadly constrains the history of water, volcanism, atmospheric development, and water-rock interaction on Mars. In this model, early atmospheric CO2 was enriched in 13C through early sputtering and hydrodynamic escape (25, 26). This early enrichment has since been removed over the course of history (Fig. 3) through carbonate formation and dilution from mantle degassing that outweighed ongoing atmospheric loss. This suggests that atmospheric loss processes were important early in martian history but have been less important since that time, whereas volcanic degassing still plays a substantial role. Thus, on the basis of the carbon isotopes it is unlikely that the somewhat enriched δ18O values for the modern CO2 are due to fractionation stemming from atmospheric loss processes. Carbonate formation is an important factor in removing 13C from the atmosphere, and indications of modern carbonate formation in many of the martian meteorites (Fig. 3) suggests that carbonate formation is a substantial ongoing process in the current martian environment, which conflicts with the widely held belief that aqueous environments on modern Mars are largely acidic.

The model also calls for the oxygen isotope composition of CO2 to be buffered by surficial reservoirs of water or ice (23), which in turn were buffered by substantial early water-rock interaction. This composition has remained similar through time, as indicated by the carbonate record (Fig. 3). The average oxygen isotope composition (δ18O ≈ 19‰) of the ancient ALH 84001 carbonates (14) is similar to the oxygen isotopic composition of the modern carbonates in EETA 79001 (δ18O ≈ 21‰) (12, 15). This could be the result of substantial ongoing water-rock interaction, indicating an active low-temperature hydrology on Mars throughout its history (19) and would contradict the commonly held view that aqueous activity has been very limited on Mars for the past 2 to 3 billion years. Another possibility to explain the oxygen isotope record is that the current water reservoir on Mars is very large as compared with magmatic degassing and atmospheric loss since the end of the early Noachian. This extensive reservoir could then buffer the oxygen isotopic composition of water and CO2 through time (19), even if the total amount of water-rock interaction has been limited through much of the planet’s history.

The oxygen isotope composition of martian atmospheric CO2 is enriched in 18O (δ18O = 31.0 ± 5.7‰) when compared with that of the silicates on Mars (δ18O = 4.2). This is substantially different from magmatic CO2, which would be in high-temperature equilibrium with the rocks and have a δ18O near 8‰ (15). Similar to terrestrial CO2, the most likely explanation for this isotopic enrichment is through low-temperature isotopic exchange with liquid water in which the fractionation of oxygen isotopes strongly enriches CO2 in 18O. If this equilibration took place at an average temperature of 0°C, the δ18O of liquid water would be close to –15‰ (15). CO2 and H2O may also exchange oxygen via photochemical reactions in the atmosphere; however, fractionations attributable to this mechanism are unknown. Therefore, if photochemical reactions dominate over aqueous reactions then the δ18O of the water on Mars may be very different. However, aqueous reactions probably dominate, given that both martian atmospheric CO2 and carbonates in EETA 79001 have δ18O values that are consistent with H2O, with a δ18O near –15‰.

Previous estimates of the δ18O of water on Mars have relied on a model described by Clayton and Mayeda (15), in which water and CO2 equilibrate with silicate rocks at high temperature during initial outgassing and then equilibrate at low temperature with no further interaction with the silicate crust. In this model, δ18O values of both water and CO2 are strongly dependent on the ratio of H2O/CO2. The recent discoveries by Mars Express and the Mars Reconnaissance Orbiter that the ancient crust of Mars contains substantial weathering products suggest that the assumption of no further interaction with the silicate crust after initial outgassing is invalid. Thus, the δ18O of water on Mars was and may continue to be buffered by exchange with the silicate crust.

The average δ18O of water present on early Mars can be independently calculated assuming it is buffered by low-temperature (0 to 25°C) water-rock interaction with a fractionation of 20‰ (27). The δ18O of the water would be ~–16‰, which is nearly identical to δ18O of water calculated from the Phoenix results. If this water-rock interaction had primarily occurred at higher temperatures, then it would probably have resulted in water with δ18O close to that of terrestrial oceans (27), which have a δ18O near 0‰ and are strongly influenced by high-temperature water-rock interaction at mid-ocean ridges. If the water in equilibrium with martian atmospheric CO2 does represent average water on the martian surface, then the preponderance of water-rock interaction on Mars took place at low temperatures, indicating that the extensive phyllosilicate exposures in the ancient crust were predominately formed at low temperatures. This is consistent with a lack of plate tectonics and a lack of substantial hydrothermal systems on early Mars.

Biological processes strongly fractionate carbon isotopes, typically incorporating a larger proportion of 12C than is present in the CO2 reservoir in which they live. On a global scale, biological processes should preferentially remove 12C and store it in organic matter, increasing the δ13C of the remaining CO2 reservoir. This process is one of many that may be operating on Mars; thus, the atmospheric measurement reported here does not provide a means for unambiguous detection of biological processes. Instead, this measurement clarifies the characteristics of major carbon and oxygen isotope reservoirs on Mars.

Supporting Online Material

Materials and Methods

Figs. S1 and S2

Tables S1 to S3


References and Notes

  1. Materials and methods are available as supporting material on Science Online.
  2. If the δ13C-enriched carbonates in Nakhla formed in equilibrium with the atmosphere, it would be difficult to explain how the atmospheric composition evolved from close to 40‰ to the modern value of –2‰ in only 600 million years. It might instead be the case that the 13C-enriched carbon is derived from a more ancient carbonate deposit that was incorporated into the younger nakhlites through dissolution and reprecipitation at the time of the Nakhla alteration event.
  3. We thank the members of the Mars Phoenix Project who enabled daily science operations at the Mars Phoenix landing site. We especially thank the Thermal and Evolved Gas Analyzer Science and Engineering Team, including H. Enos, G. Droege, C. Fellows, M. Finch, M. Fitzgibbon, and M. Williams. The Phoenix Mission was lead by the University of Arizona, Tucson, AZ, and was managed by NASA’s Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA. The spacecraft was developed by Lockheed Martin Space Systems, Denver, CO. This paper benefited from two anonymous reviews and helpful discussions with J. Jones, E. Gibson, and R. Socki.
View Abstract

Navigate This Article