Lunar and Planetary Science XXXIV (2003) 1781.pdf
HYDROTHERMAL ALTERATION OF THE MARTIAN CRUST: AN EXPERIMENTAL APPROACH
J. A. Hurowitz, S. M. McLennan, D.H. Lindsley and M.A.A. Schoonen. Department of Geosciences, Stony Brook
University, Stony Brook, NY 11794-2100, (firstname.lastname@example.org).
Introduction: On Mars, the geomorphology of Experimental Procedures: A unique and impor-
numerous surface features has been attributed to the tant component of this research project will involve
interaction of magmatic and impact heat sources with synthesis of the host-rock materials to be utilized dur-
an H2O-rich Martian crust/megaregolith [1-3]. In ad- ing hydrothermal alteration experiments. Studies of
dition, geochemical analyses of soil at the Pathfinder terrestrial hydrothermal systems indicate that at tem-
and Viking landing sites indicate high concentrations peratures ≤ 280 0C, host-rock composition exercises
of SO3 and Cl [4-6], consistent with interaction be- clear control over the secondary mineral assemblages
tween a basaltic protolith and fluids rich in acidic sul- formed as a result of hydrothermal alteration .
fate and chloride species. Such fluids are often found Griffith and Shock  concluded, on the basis of
in association with subsurface hydrothermal systems thermodynamic modeling of hydrothermal alteration
and crater lake environments in terrestrial settings [7, under conditions relevant to Mars, that the chemical
8]. These factors have led to the suggestion that and mineralogical characteristics of the modeled host-
hydrothermal alteration of the Martian crust may have rock is “the most important factor” controlling the fi-
been partially or wholly responsible for the production nal assemblage of secondary minerals produced in
of the chemically unique sedimentary deposits ob- their simulations.
served on the Martian surface [9-11]. However, the Basalt Synthesis: Two batches of synthetic crystal-
secondary mineral phases formed via alteration of the line basalt were prepared for use in the recently com-
Martian regolith under hydrothermal conditions rele- pleted and upcoming set of alteration experiments.
vant to Mars remain largely unconstrained [12-14]. Basalts were prepared utilizing a mixture of oxide
The few relevant experimental hydrothermal stud- components designed to match the chemical composi-
ies published to date have emphasized the conditions tion of the Los Angeles meteorite (Stone 1) . This
of formation of alteration minerals observed in the mixture of oxide components was heated above the
Shergottite-Nahklite-Chassigny class of meteorites liquidus (T≅1135oC) in Au80Pd20 tubing (housed in a
[15-17]. Given the exceedingly small volume of the vacuum-sealed silica tube), and subsequently cooled to
observed alteration mineralogy, and the shock-related a final temperature of 955oC before quenching in order
thermal history of these meteorites, there remains con- to produce a crystalline basaltic rock sample. Syn-
siderable doubt as to how much can be inferred about thetic basalt samples were characterized by petro-
bulk Martian soil properties from these unique samples graphic examination, electron microprobe, and X-ray
. Thermodynamic modeling of hydrothermal al- diffraction (XRD) analysis.
teration under conditions relevant to Mars has been Flow Through Experiments: In preparation for
performed utilizing only deionized water equilibrated loading into the flow-through reactor, synthetic basalt
with variable pCO2 as the simulated solvent [19, 20]. samples were lightly crushed, and the grain size rang-
To date, these types of studies have generally ignored ing from 125-710 µm (medium to fine sand) separated
other fluid compositions that may be of considerable by sieving. Ultrafine particles were subsequently re-
importance in Martian hydrothermal systems, such as moved by sonicating in acetone, and the samples dried
acid-sulfate and sulfate-chloride type fluids [9, 21]. overnight at 60oC. The 125-710 µm fraction was then
The results presented herein represent the first in a loaded into two flow-through lines housed in the reac-
series of experiments aimed at assessing the role that tor system, which was preheated to a temperature of
hydrothermal systems might play as a source of secon- 75oC. Sample Line #1 was loaded with 53.4 mg of
dary minerals at the Martian surface. Utilizing the basaltic sample, and Sample Line #2 with 64.3 mg.
flow through reactor described in Tosca et al. , two Fluid (T=75oC) was injected into the sample lines
successful experiments have recently been completed. by syringe pump at a flow rate of 3.5 µL/min. For
In addition, four batch-type experiments will be initi- Sample Line #1, a total of ~7 g of fluid was pumped
ated shortly. These experiments simulate the low- through the sample chamber, resulting in a water-rock
temperature (epithermal) alteration of a synthetic Mar- ratio (WR) ≅ 130. For Sample Line #2, a total of ~47 g
tian basalt by an acidic sulfate-chloride type fluid; an of fluid was pumped through the sample chamber,
analog to the near surface environment peripheral to a resulting in a WR ≅ 730. Effluent samples were col-
Lunar and Planetary Science XXXIV (2003) 1781.pdf
Hydrothermal Alteration of the Martian crust: J. A. Hurowitz et al.
lected in separate aliquots and placed in refrigerated ples at experimental temperatures (75oC); as well as a
storage for later analysis. characterization of the nature of basalt dissolution
The injected fluid was prepared by mixing deion- within the experimental hydrothermal system (i.e.:
ized water with reagent-grade H2SO4 and HCl, result- stoichiometric vs. non-stoichiometric dissolution).
ing in a solution with the composition 0.24M H2SO4, Further fluid geochemical modeling will involve
0.045M HCl. This composition is modeled after crater evaporation simulations that utilize the effluent com-
lake fluids found in terrestrial settings such as Mt. positions derived from the first set of flow-through
Ruapehu, NZ . The molar S: Cl ratio of this mix- experiments. Mineral phases precipitated as a result of
ture (5.3) is also nearly identical to the average molar evaporation of fluids found in Martian crater lakes or
S: Cl ratio of Martian soil (6.2). hydrothermal fluid seeps may form an important com-
Following completion of flow-through experimen- ponent of the Martian soil deposits, which are thought
tation, samples were removed from the flow-through to be enriched in salt components such as MgSO4,
reactor and washed with ethanol using a bottletop vac- MgCl2 and NaCl [26, 27].
uum filtration unit. This was done in order to remove Upcoming experiments will most likely be per-
residual acid from the solid sample, thereby preventing formed as batch type experiments in screw-top Tef-
further reaction from occurring after experiment com- lon beakers at 75oC, under a wider range of WR (10,
pletion. Approximately half of the reacted solid sam- 50, 100, and 500). These experiments will be per-
ple was then separated and ground using an agate mor- formed in attempt to promote precipitation of mineral
tar and pestle. The resultant sample powder was phases that would be expected to form in the near-
mounted on a petrographic slide using standard smear surface epithermal zone of a Martian hydrothermal
mounting techniques for XRD analysis. system. Such mineral phases could be incorporated
Preliminary Results: Synthesis resulted in basalt into Martian soil deposits by a variety of erosive
samples dominated by pyroxene (~46%), plagioclase mechanisms, e.g., see Refs [1, 2, 3].
(~38%), titanomagnetite (~7%), silica (~4.5%), and K- References:  Tanaka, K., et al. (2002) GRL 29,
rich residual glass (~4.5%). Clinopyroxene composi- 1-4.  Gulick, V. (1998) JGR 103, 19365-19387. 
tions range from Wo14-39En5-48Fs15-79. Plagioclase Hartmann, W. (2001) Space Sci. Rev. 96, 405-410. 
compositions average An58Ab41Or1.0. Titanomagnetite Bruckner, J., et al. (2001) LPS 32, Abstract # 1293. 
compositions range from Usp11-37Mt63-89. Co-existing Foley, C., et al. (2001) LPS 32, Abstract #1979. 
oxide, clinopyroxene and SiO2 equilibria suggest an Clark, B., et al. (1982) JGR 105, 9623-9642. 
ƒO2 of -0.75 to +2.1 ∆FMQ at 955oC. Varekamp, J., et al. (2000) Journal Volc. Geotherm.
These results compare to a Los Angeles meteorite Res. 97, 161-179.  Giggenbach, W. (1997) in Geo-
composition having pyroxene (43.7%), plagioclase chemistry of Hydrothermal Ore Deposits, 737-796. 
(44.8%), titanomagnetite (2.1%), silica (2.4%), and K- Newsom, H., et al. (1999) JGR 104, 8717-8728. 
rich felspathic glass (1.6%) . Los Angeles clino- Newsom, H. and J. Hagerty (1997) JGR 102, 19345-
pyroxene compositions range from Wo12-41En50-6Fs81- 19355.  Morris, R., et al. (1995) JGR 100, 5319-
33, plagioclase compositions range from An50-62Ab37- 5328.  Rathbun, J. (2002) Icarus 157, 362-372.
48Or1-4, titanomagnetite compositions range from  Morris, R., et al. (2000) JGR 105, 1757-1817.
Usp67-72Mt22-32, with a predicted ƒO2 of –1 to –2  Bell III, J., et al. (2000) JGR 105, 1721-1755.
∆FMQ at 900 – 950oC .  Baker, L. et al. (2000) Met. Planet. Sci. 35, 31-38.
Results of XRD analysis on reacted samples from  Kent, A., et al. (2001) Geochim. Cosmochim.
the flow-through experiments described above indicate Acta 65, 311-321.  Golden, D., et al. (2000) Met.
that no new mineral phases were precipitated during Planet. Sci. 35, 457-465.  Bridges, J., et al. (2001)
the course of experimentation. The primary mineral Space Sci. Rev. 96, 365-392.  Griffith, L. and E.
phases pyroxene, plagioclase, and titanomagnetite Shock (1995) Nature 377, 406-408.  Griffith, L.
were readily dissolved, leaving a residual solid com- and E. Shock (1997) JGR 102, 9135-9143.  New-
posed largely of primary silica, particularly at high som, H. (1980) Icarus 44, 207-216.  Tosca, N., et
WR. al. (2002) LPS 33, Abstract #1354.  Brown, P.
Future Work: Effluent samples collected during (1978) Annual Rev. Earth Planet. Sci. 6, 229-250. 
the course of experimentation will be analyzed for Rubin, A., et al. (2000) Geology 28, 1011-1015. 
dissolved cation and anion concentrations by emission Xirouchakis, D., et al. (2002) Geochim. Cosmochim.
spectroscopy (DCP-AES) and ion chromatography, Acta 66, 1867-1880.  McLennan, S. (2000) GRL
respectively. This will allow for the prediction of 27, 1335-1338.  McSween, H. and Keil, K. (2000)
mineral saturation states within the reacted fluid sam- Geochim. Cosmochim. Acta 64, 2155-2166.