 PRELIMINARY RESULTS FOR THE ISOTOPIC COMPOSITION OF MARTIAN ATMOSPHERIC CO2 AS DETERMINED WITH THE SAMPLE ANALYSIS AT MARS (SAM) QUADRUPOLE MASS SPECTROMETER.  H. B. Franz1,2, J. C. Stern1, E. Raaen1, M. G. Trainer1, M. H. Wong3, H. L. K. Manning4, S. K. Atreya3, G. J. Flesch5, L. A. Leshin6, P. R. Mahaffy1, T. C. Owen7, C. R. Webster5, and the MSL Science Team.  1NASA Goddard Space Flight Center, Code 699, Greenbelt, MD 20771, Heather.B.Franz@nasa.gov, 2University of Maryland Baltimore County, Baltimore, MD 21228, 3University of Michigan, Ann Arbor, MI 48109, 4Concordia College, Moorhead, MN 56562, 5Jet Propulsion Laboratory, Pasadena, CA 91009, 6Renssalaer Polytechnic Institute, Troy, NY 12180, 7University of Hawaii, Honolulu, HI 96822.  Introduction: Shortly after the Mars Science Laboratory (MSL) "Curiosity" rover landed at Gale Crater,  the Sample Analysis at Mars (SAM) instrument suite began sampling the martian atmosphere to determine its chemical and isotopic composition. Analyses of the martian atmosphere by the Viking landers  revealed enrichment in heavy isotopes of nitrogen and noble gases, suggesting selective loss of light isotopes due to escape processes [1-3]. Data collected by Viking 1 also allowed estimates of the isotopic composition of atmospheric CO2, yielding 13 C/ 12 C of 0.0115 ± 0.0003 based on m/z 12 and 13 and 0.0115 ± 0.004 based on m/z 44 and 45 [2]. These ratios suggest enrichment in 13 C of ~23‰ compared to the terrestrial standard (V-PDB), but with the large uncertainties the  13 C cannot be clearly resolved from zero. Subsequent measurements of atmospheric CO2 by the Phoenix lander produced  13 C of -2.5 ± 4.3‰ [4]. Isotope ratios are measured by SAM with the quadrupole mass spectrometer (QMS) and the tunable laser spectrometer (TLS). The TLS is designed to perform in situ measurements of the abundances and isotope ratios of carbon, oxygen, and hydrogen in CH4, CO2, and H2O at unprecedented precision [5-7]. Isotope ratios of other compounds must be determined with the QMS, which can scan continuously over m/z values of 1.5 to 535.5. Carbon isotope ratios may also be determined from QMS atmospheric data to complement the highprecision measurements performed with the TLS. Here we report preliminary results for measurements of  13 C from martian atmospheric CO2 using the SAM QMS. Experimental Methods: The data described here were obtained via SAM's direct atmospheric experiment mode, in which martian atmospheric gas is introduced directly into the manifold of SAM's gas processing system for sampling by the QMS and/or the TLS. The procedure is described in detail in [8]. Briefly, prior to measuring martian atmospheric gas, the gas manifold and transfer lines are heated and evacuated, and background measurements are obtained.  Martian atmospheric gas is then introduced directly into the manifold and leaked into the ion source of the QMS through a capillary tube. The QMS scans continually across m/z values from 1.5 to 149.9 for a specified duration and mass step size.   The QMS employs different three scanning modes through control of the quadrupole rod voltages, as described in [8].  All ratios for atmospheric experiments described here were computed from integrated fractional scan peak areas at each m/z ± 0.4, averaged over multiple scans. Data Corrections and Calibration: Corrections for detector dead time are critical to obtaining accurate estimates of isotope ratios from QMS data. The SAM QMS contains two single-channel electron multiplier detectors, operated in pulse counting mode. Two effects can cause pulse counting systems to detect fewer events than actually occur at high count rates. The first effect is known as the "dead time," the minimum length of time that must separate two events for them to be recorded as two separate pulses [9]. A second effect involves dynode gain loss at high count rates, which may manifest as an increase in apparent dead time at high count rates [10]. Both effects cause the measured signal to diverge from the true signal as count rate increases, leading to erroneously low measurements at high count rates.  Accurate quantitative analysis requires a correction to the raw QMS data to account for these phenomena. A standard form for dead time correction for paralyzable detector systems is o = ne -n where o = observed count rate, n = the true count rate, and  = the dead time [9]. The correction for loss of QMS signal with increasing count rate was determined empirically by analyzing gases under nominal operating conditions, based on the assumption that the ratio of two isotopes (or isotopologues) of a given gas should be constant at all count rates. By allowing the "dead time" to grow with count rate according to the relationship ae bo , where a and b are constants and o = observed count rate, and applying the standard form for dead time correction as described above, a correction was derived for the SAM detector that encompasses both dead time and gain loss effects due to the electron multiplier, as well as contributions to the dead time from other components in the system. For simplicity, this is commonly referred to as just the "dead time correction."  A preliminary step in the analysis of all QMS data  is to apply the dead time correction to the raw data based on the current best estimate of the coefficients a and b. For experiments during the first 100 sols, the dead time coefficients were based on pre-flight calibration with O2. The SAM team will monitor QMS detector performance throughout the mission by examination of martian atmospheric data, to determine when and if the coefficients require updating.  Carbon Isotope Ratios: Isotope ratios based on atmospheric QMS data are nominally computed from the average ratios of fractional scan peak areas. The carbon isotopic composition of CO2 may be computed directly from the ratio of signals at m/z 12 and 13, although those masses may see minor interferences from CO and hydrocarbons. Because the major molecular ion of CO2 at m/z 44 saturates the detector under nominal direct atmospheric experiment conditions, a second method for computing the carbon isotopic composition utilizes m/z 45 and 46, applying an assumed oxygen isotopic composition to obtain  13 C. The analyses reported here were performed with an oxygen isotopic composition determined from TLS measurements. Carbon isotope ratios were normalized to pre-flight calibration data, which included QMS analysis of CO2 for which   13 C and   18 O were determined independently at high-precision with a ThermoFinnigan MAT-253 mass spectrometer.  Peliminary Results: During Curiosity's first 100 sols, SAM analyzed Mars' atmosphere with the QMS three times. Table 1 gives results for  13 C determined from m/z 12 and 13 as well as from m/z 45 and 46. The latter calculations assumed an oxygen composition with  18 O = 48 ± 6‰ (V-SMOW), as measured by the TLS [7].   17 O was assumed to 0.32‰, the average value for martian silicates [11]. However, note that the calculation of  13 C is fairly insensitive to the value assumed for  17 O. For example, assuming  17 O of 1.04‰ instead, the highest value measured for carbonates in martian meteorites [12], yields  13 C values identical within SAM measurement precision to those obtained using  17 O of 0.32‰. Uncertainties for individual experiments include propagation of statistical errors through the normalization process, while the uncertainties given for the weighted mean values include additional sources of systematic error. The average values of   13 C reported in the table for both methods overlap with the average  13 C of 45 ± 5‰ determined from TLS data [7]. The weighted mean for all QMS observations, including all ratios computed by both analysis methods, is 40.2 ± 10.8‰. These results are consistent with enrichments in heavy isotopes of other elements in the martian atmosphere, including hydrogen, nitrogen, argon, and other noble gases, and support models invoking large-scale atmospheric loss processes on Mars [1-3, 13].  Further implications of these results are discussed in other abstracts at this meeting [14-15]. Refinement of this analysis through more sophisticated modeling of instrument behavior is currently underway.   Table 1. Preliminary QMS results for 13C of  martian atmospheric CO2 measured during sols 0-100.   Test ID Sol 13C from m/z 12 & 13 (‰) 13C from m/z 45 & 46 (‰) 25009 27 25.4 ± 14.7 48.7 ± 4.2 25012 45 55.5 ± 4.8 38.5 ± 3.5 25027 77 25.0 ± 4.0 38.3 ± 3.3 47.4 ± 4.3 38.0 ± 3.2 Weighted mean for each method 40.4 ± 15.5 40.1 ± 5.2 Weighted mean of all  measurements 40.2 ± 10.8    References: [1] Biemann et al. (1976) Science 194. [2] Nier et al. (1976) Science 194. [3] Owen et al. (1977) JGR 82. [4] Niles et al. (2010) Science 329. [5]  Mahaffy (2008) Space Sci. Rev. 135. [6] Mahaffy  et al. (2012) Space Sci. Rev. 170. [7] Webster et al. (2013) LPSC XLIV. [8] Trainer et al. (2013) LPSC XLIV. [9] Knoll (2000) Radiation Detection and Measurement, John F. Wiley and Sons. [10] EPA (2007) Method 6800, Rev. 0. [11] Franchi et al. (1999) MAPS 34. [12] Farquhar et al. (2000) MAPS 26. [13] Watson et al. (1994) Science 265. [14] Leshin et al. (2013) LPSC XLIV. [15] Jones et al. (2013) LPSC XLIV. 
