PATHWAYS to FORM KIESERITE from EPSOMITE at MID to LOW TEMPERATURES, with RELEVANCE to MARS, John J

Lunar and Planetary Science XXXVIII (2007) 1298.pdf

PATHWAYS TO FORM KIESERITE FROM EPSOMITE AT MID TO LOW TEMPERATURES, WITH RELEVANCE TO MARS, John J. Freeman, Alian Wang, Bradley L. Jolliff, Department of Earth & Planetary Sciences and the McDonnell Center for the Space Sciences, Washington University, St. Louis, MO, 63130 ([email protected])

Kieserite on Mars: The hydrous Mg-sulfate that as the major analytical tool, accompanied by mass-loss has been identified definitively on Mars by OMEGA measurements, XRD, and IR spectroscopy. We have (Mars Express orbiter) is kieserite (MgSO4·H2O) [1]. done 126 experiments between 800-2500 hrs duration An additional class of OMEGA spectra has been at- at three temperatures (50˚C, 21˚C, and 5˚C) and using tributed to “polyhydrated sulfate.” This class has spec- ten relative humidity (RH) buffers covering a RH tral features that match Mg-sulfates of higher hydra- range from 5.5% to 100%. The powdered pure sam- tion or multi-cation sulfates [2,3]. Ca-sulfates, espe- ples epsomite, starkeyite, kieserite and amorphous Mg- cially gypsum (CaSO4·2H2O), are also identified in sulfate (with two structural waters) were used as start- polar regions on Mars [4], but anhydrite (CaSO4) does ing phases. Six samples of powdered epsomite mixed not possess an absorption in the OMEGA spectral re- with powdered anhydrite, bassanite, and gypsum with gion. Of interest is an apparent systematic trend in the molar ratios of Mg:Ca at ~8:2 and ~4:6 were studied at geomorphic siting of different types of sulfates at re- 30% RH and 50˚C. gional and global scales on Mars [5,6,7]. Kieserite has Can kieserite form by dehydration of higher been found mostly on steep slopes or on plateaus, hydrates? Although kieserite has been identified at whereas “polyhydrated sulfates” occur on shallow many locations on Mars by orbital remote sensing, it slopes or on valley floors. This trend indicates a poten- cannot be formed in our experiments from the direct tial commonality in the geological processes that is dehydration of epsomite, hexahydrite, or starkeyite at responsible for transitions (hydration state, etc.) be- 5˚C ≤ T≤ 50˚C (kieserite can be produced from dehy- tween these two phases, and also between these and dration of epsomite at T ≥75˚C). Figure 1a & b shows other co-existing phases that would be invisible to the hydrous Mg-sulfate phases identified in the latest OMEGA, e.g., anhydrite. The new findings from mis- reaction products of our experiments based on their sions demonstrate that the study of potential martian Raman spectra. Except for a few cases that will be sulfates, their composition, hydration states, crystallin- discussed in later section, starkeyite is the apparent ity, and phase transition pathways are very important stable phase of Mg-sulfate at 5˚C ≤ T≤ 50˚C and mid- to understanding the early history, hydrologic evolu- low relative humidity 5.5% ≤ RH ≤ 50% (or 35% de- tion, and present-day surface of Mars. pending on T). Laboratory study of Figure 1a. Dehydration from Epsomite Starkeyite has a very stable four- Mg-sulfate: Laboratory ex- 1w 2w Am 4w 6w 7w Aq member-ring substructure [17], thus

periments [8-16] on hydrous 80 additional activation energy is sulfates under well-controlled needed to break the ring and to form conditions provide fundamen- a tighter framework in the kieserite 60 1979 hrs tal knowledge of phase structure where all SO4 tetrahedra

boundaries, reaction path- 40 and MgO5Ow octahedra share their ways, and conditions of phase coordinating oxygen. It appears 1979 hrs Temperature (C) transitions, which will help to 20 unlikely that temperatures high interpret surface and orbital 2157 hrs enough to provide these activation

observations. The limitations 0 energies exist at the surface of Mars of laboratory experiments 020406080100today. Therefore if the primary hy- Relative Humidity (%) include difficulties in simulat- drous Mg-sulfates precipitated from Figure 1b. Hydration and Dehydration from Starkeyite ing real, complex geologic aqueous solution on Mars were ep- 80 processes and the extremely somite (or MgSO4·11H2O at low T [18,19]), are there dehydration path- long durations needed for 972 hrs low-temperature experiments. 60 ways that could produce abundant We have studied the sta- kieserite detected by OMEGA today? bility field and phase- 40 1st Pathway for kieserite for- 909 hrs

transition pathways of hy- Temperature (C) mation at mid to low T: Amorphous 20 drous Mg-sulfates using hu- 2157 hrs Mg-sulfate is the first pathway to midity-buffer techniques and form kieserite from epsomite in our 0 vacuum desiccation. Laser 0 20 40 60 80 100 experiments (5˚C ≤ T≤ 50˚C). Dur- Raman spectroscopy is used Relative Humidity (%) ing the two experiments shown in Lunar and Planetary Science XXXVIII (2007) 1298.pdf

1w 2w Fig. 1a at 50˚C and an RH range when using pure Mg-sulfates as starting Fig. 2a 7w at 5.5%RH & 50˚C of 5.5-11.1%, the epsomite sam- 6w phases. In our experiments, the forma- 4w ples first convert to amorphous 1979 hrs tion of kieserite and sanderite was not Mg-sulfate, then a mixture of kie- observed when epsomite and anhydrite serite and sanderite appeared, and 1143 hrs powder samples in separate vials were 686 hrs finally the entire sample converted Am placed within the same humidity buffer, 224 hrs to crystalline kieserite and sande- 179 hrs or when epsomite was mixed with bas- 134 hrs rite. Fig. 2a shows a set of Raman 67 hrs sanite or with gypsum powder samples. spectra obtained during the entire 43 hrs It appears that the physical contact of 24 hrs 7 hrs experiment. A similar set of hy- 1 hrs 7w epsomite grains (<75 µm in our ex-

dration experiments using amor- 1120 1100 1080 1060 1040 1020 1000 980 960 periments) with those of anhydrous Ca- Counts / Raman Shift (cm-1) phous Mg-sulfate (MgSO4·2H2O) sulfate is required for this reaction, i.e., as the starting phase shows the 4w the surrounding anhydrite grains may same phenomenon: at 50˚C and Fig. 2b 7w + Anhy at provide a favorable microenvironment 31%RH & 50˚C 30.5% RH after 20 hrs with sam- Anhy 4w+5w for the dehydration reaction. ples in the humidity buffer, the 2531 hrs Implications for remote sensing

amorphous Mg-sulfate gradually 707 hrs observations: Amorphous Mg-sulfates 6w converted to a mixture of kieserite 587 hrs can be formed from the direct dehydra-

and sanderite (not shown). We did 184 hrs tion of epsomite and hexahydrite, 2w not observe a similar conversion 1w readily at T > 0˚C or at a much slower in the experiments done at 70 hrs rate when T < 0˚C [20]. The co-

T<50˚C, possibly owing to the 47 hrs existence of Ca- and Mg-sulfates is 2+ slow reaction rates at lower tem- 27 hrs anticipated on Mars because (1) Mg 1 hr 2+ perature. For example at 5˚C, we 1080 1060 1040 1020 1000 980 960 and Ca are released from igneous see a trace amorphous phase Counts / Raman Shift (cm-1) minerals (especially clinopyroxene) as formed from epsomite at low RH (7.4%) after 2157 a result of chemical weathering e.g., through reaction hrs, whereas it appeared quickly after 7 hrs in the with acidic aqueous fluid; (2) Mg- and Ca-sulfates buffer at 50˚C. With experiments done at 5˚C and likely coexist in evaporite sequences, and their coexis- starting with amorphous Mg-sulfate, the samples con- tence on Mars is implied by modal mineral analyses verted to a mixture of starkeyite and hexahydrite at based on MER results [21,22]. Thus the essential 33.6% RH or remained as the amorphous phase at phases for both pathways would be available on Mars, lower RH. We consider that the irregular structure of and could be involved in the origin of kieserite as a an amorphous Mg-sulfate must have provided a lower dehydration product of epsomite and MgSO4·11H2O. It energy threshold for the formation of kieserite. is also possible that martian kieserite was formed as 2nd Pathway for kieserite formation at mid to part of primary evaporite deposits similar to those that low T: A second potential pathway to form kieserite occur on Earth. from epsomite occurs in physical mixtures with anhydrite (CaSO4). A set of Raman spectra in Figure 2b Acknowledgment: This work is supported by NASA funding shows that epsomite first converts to hexahydrite, then NNG05GM95G, NAG5-12684, and for the MER Athena Science team. We thank Dr. I-Ming Chou for providing the phase boundary starkeyite, sanderite, and kieserite appear together after data and instructive discussions 27.3 hrs in RH buffered samples. The increase of Ra- man peak intensities of kieserite and sanderite contin- References: [1] Arvidson et al. (2005) Science 307, 1591-1593. ues until 91 hrs, then decreases. Because we used a [2] Bibring et al. (2005) Science 307, 1576-1581. [3] Gendrin et al. (2005), Science 307, 1587-1591. [4] Langevin et al. (2005) Science mid-RH (30.5%) for these experiments, starkeyite be- 307, 1584-1586. [5] Arvidson (2006) LPI Contribution No. 1331, p8. came the major Mg-sulfate in the mixture after 2531 [6] Bibring (2006) LPI Contribution No. 1331, p12. [7] Mangold et hrs, and only trace amounts of kieserite, sanderite, and al (2006) LPI Contribution No. 1331, p53. [8] Chou et al. (2002) Am. hexahydrite are still present. We postulate that kieser- Miner 87,108-114. [9] Chou et al. (2003) Astrobiology 3, 619-629. [10] Vaniman et al. (2004), Nature 431, 663-665. [11] Chou et al. ite and sanderite could become the major stable Mg- (2005) GSA. [12] Wang et al. (2006) LPSC, abs#2168. [13] Wang et sulfates in this type of mixture if the experiments were al. (2006) LPSC, abs#2191. [14] Vaniman & Chipera (2006) Am. done in a lower RH range (e.g., 5.5-11.1%, similar Miner. [15] Wang et al. (2006) GCA 70, 6118-6135. [16] Wang et conditions as the two points in Fig. 1a). No Raman al (2006) LPI Contribution No. 1331, p75. [17] Bauer, Acta Cryst. 17, 863-869. [18] Peterson & Wang (2006) Geology, 957-960. [19] peak position shifts in both Mg- and Ca-sulfates, as Wang et al (2007), this volume. [20] Wang et al. (2007) this Volume. would indicate cation substitutions (at least to a detect- [21] Clark et al. (2005) EPSL 240, 73-94. [22] Wang et al. (2006) JGR, 111, able level), were observed. The conditions to form E02S16. stable starkeyite from epsomite are 30.5% RH at 50˚C