 OXYGEN AND CARBON ISOTOPE SYSTEMATICS IN CO2-H2O-(Ca,Mg,Na)ClO4 BRINE SYSTEM BELOW 0 °C: IMPLICATIONS FOR OXYGEN ISOTOPE COMPOSITION OF WATER AND CARBON SEQUESTRATION ON MARS. M. I. El-Shenawy1 and P. B. Niles2, 1USRA, NASA Johnson Space Center, Houston, TX 77058 (mohammed.i.elshenawy@nasa.gov), 2NASA Johnson Space Center, Houston, TX 77058.  Introduction:  Perchlorates have been detected on Mars at locations across the surface including Phoenix landing site at 68°N, Gale Crater at 4.5°S, and two Viking landing sites at 22.5°N and 48.3°N [1-3]. This wide occurrence of perchlorates ( ̴ 6.5 % of the Martian surface area) suggests the possible widespread existence of perchlorate brines on Mars at temperatures down to 198 K [4,5]. Evidence of liquid water on Mars was inferred from the spectral analysis of modern flow structures on surface (RSLs) and the radar imaging of subsurface large water body, which are potentially associated with the occurrence of Mg, Ca, Na - perchlorates [6,7].  Under modern Martian climactic conditions, cryogenic perchlorate brines would allow oxygen isotope exchange between H2O and atmospheric CO2. As a result, the oxygen isotope reservoir in water (minor oxygen reservoir) would be closely related to the reservoir in atmospheric CO2 (major oxygen reservoir). The relation between the oxygen isotope composition in both oxygen reservoirs, oxygen isotope fractionation between CO2 and water (1000 ln18αCO2-H2O), would be governed by ambient temperature [e.g., 8]. This relation has not been defined at temperatures below 273.15 K. This relationship is crucial to better understand the oxygen isotope composition of Martian surface water and the history of aqueous processes on Mars. In addition, cryogenic perchlorate brines could potentially have a high solubility of atmospheric CO2 [e.g., 5], trapping carbon either in the form of dissolved inorganic carbon (DIC) or carbonates (i.e., carbon sequestration). The latter would result from the combination of CO2 and cations (Mg, Ca, Na) in brines.  In this study, we attempt to determine the oxygen isotope fractionation (ln18αCO2-H2O - temperature relation) and to investigate the solubility of CO2 in perchlorate brines between +30 and -33 °C.           Methods:  Na, Ca, Mg-perchlorate solutions of eutectic concentration (i.e., 9.25, 4 and 3.4 mol.kg-1, respectively) were prepared in a dry glove box under continuous nitrogen gas flow at room temperature. To prepare these solutions, 18 MΩ deionized water was boiled to release any dissolved CO2, and was measured for its oxygen isotope composition (δ18OH2O) against the international water standards (i.e., VSMOW and SLAP). Then dried anhydrous perchlorate salts were dissolved in this carbon free-water. The solutions were poured into 20 mL vials sealed with aluminum seal and septum. Meanwhile, Exetainer® vials (10 ml) were automatically flushed and filled with a 0.4 % CO2 and 99.6 % He mixture for 10 mins using the Gas Bench II system to assure that similar volume of CO2-He mixture were added to the vials and no other gases were in the vials. Carbon and oxygen isotope compositions (δ13CCO2 and δ 18OCO2) of the starting CO2 gas in the mixture were -4 ‰ relative to VPDB and +26.12 ‰ relative to VSMOW, respectively. All vials containing solution and gas mixture were stored at either 30, 4, 15 or -33 °C, depending on the target experimental temperature for a day to establish the thermal equilibria. Subsequently, 0.4 mL of the perchlorate solution was drawn with a 1 mL syringe through the septum of the solution vial and injected into the Exetainer® vial containing the gas mixture. Then the Exetainer® vial was aged for a certain time at the storage temperature until the measurement of oxygen isotope composition of the CO2 in the headspace was performed. The δ13CCO2 and δ 18OCO2 were measured using the Gas Bench II system coupled with a Finnigan MAT 253 continuous flow isotope ratio mass spectrometer (CFIRMS) at JSC.   Results and Discussion:   Determination of oxygen isotope fractionation between CO2 and water (1000 ln18αCO2-H2O) in brines. The 18αCO2-H2O values were determined by the following expression:  1000 ln18αCO2-H2O = 1000*ln [(1000 + δ 18OCO2)/(1000 +  δ18OH2O)]  The 1000 ln18αCO2-H2O values increase exponentially with time from the initial 1000 ln18αCO2-H2O value at aging time zero and then level off at a certain timeinvariant 1000 ln18αCO2-H2O value (Fig. 1). In general, this time-invariant 1000 ln18αCO2-H2O value decreases with increasing temperature. Furthermore, it also depends on the cation in the perchlorate brines within a certain temperature (i.e., the time-invariant 1000 ln18αCO2-H2O value in Na > Ca > Mg) (Fig. 1). The latter observation is explained by the salt isotope effect which has been reported by other researchers [e.g., 9]. However, the time-invariant 1000 ln18αCO2-H2O value in NaClO4 brines is statistically indistinguishable from that determined in pure water at a same temperature (i.e., +30 and +4 °C). In this study, the time-invariant 1000 ln18αCO2-H2O values determined in pure water and NaClO4 brines at +30 and +4 °C are in a good agree ment with the previously reported 1000 ln18αCO2-H2O value at isotopic equilibrium in Bottinga and Craig [8]. This suggested that our CO2-brine system has reached isotopic equilibrium and can be used to calibrate the ln18αCO2-H2O - temperature relation at below 0 °C. At present time, our calibration of this relation is: 1000 ln18αCO2-H2O = 17.24 (1000/T) - 17.80; and  it is based upon the 1000 ln18αCO2-H2O values obtained from three temperatures (i.e.,  +30, +4 and -15 °C) because the experiments at -33 °C are still ongoing.    Fig. 1: 1000 ln18αCO2-H2O evolution versus time in CO2 - perchlorate brines. (a) at 30 °C and (b) at -15 °C. Full data will be presented in the poster due to the page limitation.   Carbon isotope systematics in CO2 - perchlorate brines. The δ13CCO2 value of CO2 in the headspace above the perchlorate brines quickly (compared to the 1000ln18αCO2-H2O) evolved from the starting δ 13CCO2 at zero time to a lower constant value which decreases with decreasing temperature (Fig. 2). At all temperatures, the CO2 above the MgCl2O8 brines possess the lowest δ13CCO2 value. The decrease in the  δ 13CCO2 value was also associated with a decrease in the CO2 fraction in the headspace suggesting that CO2 was trapped in the brines. If there was a CO2 leakage out of the Exetainer® vial, the δ13CCO2  value would be higher than the starting CO2 at zero time because the 12C in CO2 escapes first enriching the remaining CO2 fraction in 13C. We interpret the decrease in the δ13CCO2 value as an indication to the carbonate precipitation in the brines, especially in the MgCl2O8 brines. Time (h) 0 50 100 150 200 250 300 350 d1 3 C CO 2-7.5 -7.0 -6.5 -6.0 -5.5 -5.0 -4.5 -4.0 -3.5 H2O_30 o CNaClO4_9.25M_30 o CCaCl2O8_4M_30 o CMgCl2O8_3.4M_30 o C Time (h) 0 500 1000 1500 2000 2500 3000 3500 d1 3 C CO 2-10 -9 -8 -7 -6 -5 -4 -3 NaClO4_9.25M_-15 o CCaCl2O8_3.4M_-15 o CMgCl2O8_3.4M_-15 o C Fig. 2: The δ13CCO2 evolution versus time in CO2 - perchlorate brines. (a) at 30 °C and (b) at -15 °C. Full data will be presented in the poster due to the page limitation.  Conclusions: Based upon our 1000 ln18αCO2-H2O - temperature calibration, the δ18OH2O on Mars surface (average temperature = -63 °C and the δ18OCO2 of Mars atmosphere = 48 ± 5‰) would be -16 ± 5 ‰ VSMOW which is consistent with low temperature equilibrium with the Martian silicate crust. The potential carbonate precipitation inferred from the δ13CCO2 in our experiments suggests that carbon sequestration on Mars may have occurred via large subsurface carbonate deposition in subglacial salt lakes [e.g., 7]. This would preferentially remove the 13C driving the remaining atmospheric CO2 to lower δ 13C values. This could reduce 13C enrichment from atmospheric loss [10].    References:  [1] Glavin D.P. et al. (2013) JGR, 118, 1955-1973. [2] Hecht M.H. et al. (2009) Science, 325, 64-67. [3] Navarro-González, R. et al. (2010) JGR, 115, E12010. [4] Pestova  O. N. et al. (2005) Russ. J. Appl. Chem., 78, 409 -413. [5] Stamenković V. et al. (2018) Nat. Geosci., 11, 905-909. [6] Ojha L. et al. (2015) Nat. Geosci., 8, 929932. [7] Orosei et al. (2018) Science, 361, 1093 -1096. [8] Bottinga Y. and Craig H. (1968) Earth Planet. Sci. Lett., 5, 285 -295. [9]  Horita J. et al. (1993a) Geochim. Cosmochim. Acta, 57, 2797-2817. [10] Jakosky B. M. et al. (2017) Science, 355, 1408 -1410.   (a) (b) (a) (b) 
