GEOCHEMISTRY AND ORIGIN OF A DEEP-SEA PACIFIC PALYGORSKITE DEPOSIT
THOMAS
M. C H U R C H
and B R U C E V E L D E
College of Marine Studies, University of Delaware, Newark, D E 19 711 (U.S.A.) Laboratoire de PJtrographie, Universitdde Paris VI, Paris (France) (Received March 16, 1978; revised and accepted September 8, 1978)
ABSTRACT Church, T.M. and Velde, B., 1979. Geochemistry and origin of a deep-sea Pacific palygorskite deposit. Chem. Geol., 25: 31--39. Palygorskite was isolated as the major fraction from one horizon of a mid-Pacific core and analyzed for major cations (Na, K, Rb, Ca, St, and Ba), the exchange characteristics of Na, K, and Ca, and the stable isotopes of O and Sr. These results are compared with the geochemistry of the major coexisting phase, montmorillonite. The data show that the palygorskite, nominally a magnesium alumino-silicate, also contains significant structural K and Ca. It probably formed diagenetically from detrital AI and Sr-bearing precursor phases at temperatures higher than those of the normal deep-sea. Sr-isotope ratios demonstrate that the palygorskite and montmorillonite neither formed from nor equilibrated later with ocean water.
INTRODUCTION The occurrence of authigenic minerals in the deep sea attracts substantial interest because their formation could indicate equilibrium conditions which partially control the composition of ocean water. Deposits of such minerals could provide important sinks for the dissolved components of seawater. Also, the occurrence of authigenic minerals could allow one to determine the chemical conditions and composition of formation waters whether they be deep sea or hydrothermal in origin. The occurrence of palygorskite (and sepiolite) in the deep sea is of particular interest because the formation of authigenic magnesium silicates could well be a major equilibrium factor controlling the abundance of Mg in seawater. Both minerals have been frequently reported in deep marine sediments, but evaluations of the conditions of their formation are widely varying. For instance, the source of Mg for the two minerals is crucial. Hathaway and Sachs (1965} propose stripping Mg from seawater by silica while Bonatti and Joensuu (1968) propose alteration of volcanic ash precursors by Mg-rich hydrothermal brines first to smectite phases and then conversion of these authigenic phases to palygorskite and clinoptilolite.
32 Bowles et al. (1971) state that Mg-rich hydrothermal solutions emanating from tectonic fissures mix with seawater and can alter pre-existing parent material to form deep-sea palygorskite and sepiolite. Unfortunately, these a~.ithors and others (e.g., Bartholomd, 1966) equate the conditions of formation, and more importantly, the stability for both minerals. Sepiotite and palygorskite have important distinctions. Palygorskite contains more A1 than sepiolite, and they occur in different mineral assemblages. Commonly, calcite, dolomite, and quartz are found with sepiolite and clinoptilolite, quartz, and smectites (montmorillonite) are more commonly found with palygorskite (Hathaway and Sachs, 1965; Bonatti and Joensuu, 1968; Wollastetal, 1968; Bowles et al., 1971). Arguments by Millot (1964) and Velde (1977) as to the origin of palygorskite suggest that nearly mono-mineralic layers of the mineral indicate authigenic formation. Most recently, Couture (1977) reports in some detail the chemical and mineralogical compositions of zeolite-containing, palygorskite- and montmoriUonite-rich deposits from the Pacific Ocean Sites 164 and 196 of the Deep-Sea Drilling Project (DSDP). It would be useful to examine in greater detail the geochemistry of deepsea magnesium silicates which could better reveal the chemical conditions under which they formed. The purpose of this paper is to present some analyses of trace elements and stable isotopes plus major cation-exchange characteristics for one Pacific palygorskite deposit which reveal some particular conditions for the formation of this magnesium silicate. THE SAMPLE A core was raised by the Proa expedition of the Scripps Institution of Oceanography (Proa 101-P) from abyssal depths in the central southwestern Pacific (7 ° N,117 ° W). It has remained largely a unique curiosity of the Scripps collection since the coring appears to have penetrated a doublelayered Mn crust sandwich which contained 240 cm of mostly palygorskite, clinoptilolite and montmorillonite. Some minor amounts of forams, quartz grains, striking single crystal panes of barite (100 × 50 × 30 pm), and micromanganese and phosphate nodules were also identified. Qualitative mineralogical abundance of several sections is presented in Table I. Quartz, calcite, and zeolite tend to be concentrated in the coarse (>2 um) fraction, while the zeolite clinoptitolite is present in important quantities (about 10%) in the finer fraction. The coexistence of palygorskite with the silica-rich authigenic minerals chert and clinoptilolite appear common (Couture, 1977), although no chert was detected in these deposits. MINERAL SEPARATION There appears to be complementary abundances of palygorskite and montmorillonite; thus an attempt was made to purify two portions of the core, 43--58 and 99--100 cm, in order to concentrate palygorskite and mont-
33 TABLE I Mineralogical analyses of bulk sections from the Pacific core Proa 101-P Section
Palygorskite
Clinoptilolite
Montmorillonite--illite
Quartz
Barite
--
(cm) 20--24
++
--
++
++
++
35--37
+++
--
+
+
--
+
43--58
+++
--
+
--
++
--
65--66
++
++
+++
+
--
+
99--100
++
++
+++
+
--
+
+++ = > 50%;++ = > 10%; + = > 5%;--: not detected morillonite, respectively. This was done by successive ultrasonic agitations in alcohol and acetone of the distilled-water washed material which passed through a 32-pm sieve. Although the montmorillonite remained largely in suspension after this treatment, pure separations were not possible; and the fractions were only enriched to about 90%. The barite crystals were separated on a small 28-pm nylon sieve and purified by settling in Clerici solutions. ANALYSIS M E T H O D S
The palygorskite and montmorillonite samples were analyzed for Na, K, and Ca by flame p h o t o m e t r y following complete acid digestion with HF. The Sr and Rb concentrations were determined by isotope-dilution mass spectrometry on the complete digests after separation on cation exchange columns with carrier-free radiotracers. The 87Sr/S6Sr isotope ratios were determined to 100 ppm by normalizing about two dozen sets of data to a ~6Sr/aSSr ratio of 0.1192. The oxygen isotopes were run by the normal BF3 digestion against the PeeDee belomnite standard. Exchangeable cations were determined by thoroughly dispersing 10--30 mg of material in 10 ml of a 0.02 m NH4C1 solution for a few days (Peech, 1945) and analyzing the Na, K, Ca by flame p h o t o m e t r y . RESULTS
: ' ~ ~
The mineralogical, chemical, and stable isotope compositions are presented in Table II and discussed in the following sections. MINERALOGY
All mineral phases detected in Proa 101 could a priori be of authigenic origin. One exception is the rounded quartz grains in the 45--58 cm section which is evidence for aeolian input during this period (Rex and Goldberg, 1958). Each mineral phase is discussed separately.
34 TABLE II Mineralogical, chemical, and isotopic contents of the palygorskite- and montmorilloniteenriched fractions of the Pacific core Proa-101 P
T h e o c c u r r e n c e o f p a l y g o r s k i t e in d e e p - o c e a n s e d i m e n t s has r e c e n t l y b e e n g e o c h e m i c a l l y c h a r a c t e r i z e d b y C o u t u r e ( 1 9 7 7 ) in t e r m s o f p h a s e equilibria or m i n e r a l c h e m i c a l c o m p o s i t i o n in t w o D S D P sites. H o w e v e r , f r e q u e n t c i t a t i o n s o f p a l y g o r s k i t e in m i x e d m i n e r a l a s s e m b l a g e s o f t h e d e e p e r sedim e n t s in s o m e earlier D S D P r e p o r t s s e e m t o be largely u n f o u n d e d , a p p a r e n t ly d u e to c o m p u t e r i z a t i o n o f d i f f r a c t i o n d a t a w h i c h d o e s n o t distinguish
mixed-layered minerals. Such reports of palygorskite in mixed mineral assemblages other than those considered by Couture to be typical (i.e.zeolite + chert + palygorskite) should thus be considered with caution. If one can assimilatepalygorskite into a compositional series contiguous with sepiolite,one can delineate certain general chemical requirements for its precipitation (Siffert,1962; Wollast et al.,1968; Christ et al.,1973). Most likely a p H above 7.5 is required, and SiO2 concentrations ) 10 30 p p m are necessary to form palygorskite. The simple combination of M g 2÷ ions with silicahas been proposed for the chemical precipitation of magnesian sepiolife observed to occur in the Santa Cruz basin, California (Fleischer, 1972). In the samples studied, the mineral is found forming on the surface and in cracks of diatomite rocks lining the SE wall of the Santa Cruz basin (605-1590 m depth); and chemical analysis of the sample shows an essentially pure MgO--SiO2 --H20 composition. Such compositions, however, seem to be rare for sepioliteand nonexistent for palygorskites. Analyses available for samples from other environments (weathering profiles,hydrothermal veins,
35
and closed-basin sediments) form a continuous compositional series with increasing CaO, A1203, and K20 contents (Velde, 1977). Ca can either be isomorphically substituted for Mg in the mineral structure or can be considered as a zeolite-type exchange ion. Both cations are relatively abundant in seawaters as well as in pore waters. K can be m u c h less easily accomodated (crystaUographically) in the proposed sepiolite and palygorskite structures and should normally be considered as an exchangeable ion (Brown, 1961). However, alumina is more difficult to account for by deep-sea aqueous precipitation because of its low solubility and low concentration in seawater. Of the two mineral species, species, palygorskite always contains more A1203,6--18 wt.%, compared to 0--6 wt.% A12O3 found in sepiolites. It is quite possible, therefore, that the palygorskite sample discussed here could have been formed by a reaction more complex than the simple precipitation from seawater of its comp o n e n t elements. The K content of the palygorskite studied here is rather high (3.46 wt.% K20). McLean et al. (1972) report a lacustrine palygorskite with 1.8% K20. K as well as Na (1.09 wt.% NaO2 in this sample) is present only to a minor e x t e n t as zeolite exchange ions as evidenced by the small a m o u n t exchanged from the structure (Table III). These ions are c o m m o n l y present in seawater, although K has a lower relative concentration. The Na/K ratio in this palygorskite is 0.61, whereas the Na/K ratio in seawater is 47. Moreover, the exchange capacity calculated from the Na, Ca, and K content exceeds by a factor of ten the measured exchangeable ions. Thus, it appears that the K is, in fact, mostly structural or non-exchangeable. The total positive ionic charges due to alkali and Ca 2÷ ions present for this palygorskite are 258 mequiv, per 100 g o f sample. This falls well above the exchangeable cations which are 20.42 mequiv./100 g for this palygorskite or the 20--45 mequiv./100 g of exchangeTABLE III Exchangeable ions as replaced by excess ammonium, and the total exchange capacity of the palygorskite and montmorillonite fractions of the Pacific core Proa 101-P Palygorskite fraction Exchangeable cation
Ca K Na
exchange capacity (mequiv./100 g) 16.4 1.96 2.06
Total exchangeable c a t i o n s (mequiv. Ca, K and Na) 20.42 Total cations (mequiv. Ca, K and Na) 258
Montmorillonite fraction percent of exchange capacity total cation (mequiv./100 g)
percent of total cation
10.8 5.3 11.8
52.9 22.7 57.2
60.5 4.25 7.95
72.7 177
36 able ions measured on other sepiolites and palygorskites (Brown, 1961). This suggests, even though the sample is n o t totally pure, that a significant portion, at least of the Ca, is found in the octahedrally coordinated Mg 2÷ ion positions of the structure. Even for the N a 2 0 and K20, which together reresent 107 mequiv., only 17.1 are actually exchanged by NH~. This exchange capacity also is much lower than most natural zeolites. Montmorillonite
The montmorillonite of this core appears to be dioctahedral by X-ray diffraction and the chemical composition of this phase is quite normal; the relatively high K 2 0 content (1.76%) is similar to that reported by other workers for detrital montmorillonite with 23.7% as an exchangeable cation. The total alkali and Ca contents indicate the presence of 183 mequiv, of charge, a b o u t 50% that of a muscovite-type structure with O~0(OH)2, a high b u t reasonable value for dioctahedral montmorillonites (Brown, 1961). The exchangeable ion contents in Table III indicate that the montmorillonite has not come to equilibrium with seawater with respect to its Na/K ratio in that the Na/K value of 47 (seawater) is not all approached. It can be seen in Table II that only 23% of the montmorillonite K was exchangeable, whereas 52% of the Na was exchangeable. In summary, the alkali and alkaline earth contents of the palygorskite and montmorillonite show that: (1) K and Na are high in palygorskite and lower in montmorillonite. Further, the Na/K ratios for palygorskite and montmorillonite indicate that these minerals apparently did not equilibrate with seawater. (2) The exchangeable Ca is 5a% of total CaO in montmorillonite and only 10% in the palygorskite, suggesting structural placement o f alkaline earths in the deep-sea palygorskite. CHEMISTRY
Minor elements
The Rb concentrations are 470 and 230 p p m in the palygorskite and montmorillonite phase, respectively, related perhaps to the relatively higher K content of the palygorskite phase. Most of the K is nonexchangeable, and this also may be true for R b when one considers that the K / R b ratio in the palygorskite is 67.0 compared with a ratio in seawater which is 7290. The Sr concentration shows an inverse correlation with Ca being more impoverished in the palygorskite than the montmorillonite, This could suggest that the two alkaline earths occupy different sites in palygorskite; more from structural components of detrital precursors in palygorskite and more from exchange with formation solutions for montmoriltonite. If one inspects the measured exchangeable cations in Table III, it is evident that, unlike the
37 palygorskite, the montmorillonite exchanges most of its Ca and probably also its Sr.
Isotope ratios The O-isotope abundances in both palygorskite and montmorillonite fractions are similar, yielding a fractionation coefficient, u- OXYpa_Mo,(Pa = palygorskite, Mo = montmorillonite) of 1.023 (SMOW). This is significantly lower than the fractionation factors reported by Savin and Epstein (1970a) for most marine clay minerals including montmorillonite ,ta oxy Mo = 1.027). The only montmorillonite samples showing these light values in the deep sea were deduced by Savin and Epstein (1970b) from iron oxide-rich cores near the East Pacific rise whose detrital clays appear to be terrigenous mixed assemblages, a characteristic of this palygorskite deposit also. Further, weathered soil profiles containing montmorillonite reported by Lawrence and Taylor (1972) are of the heavier variety ( ~dXoY_H O = 1.025--1.028), even though the aqueous solutions m whmh they crystalhzed were lighter. The authigenic montmorillonites richest in ~sO apparently form in isotopic equilibrium with seawater at deep-sea temperatures, and those of ~sO poor samples form by exchange with waters of greater temperature than the deep sea (Savin and Epstein, 1970a). The light O isotopes of this palygorskite deposit could reflect at least equilibration, if not formation, at temperatures elevated above the sea floor. Another possibility is that the light O isotopes were inherited from precursor detrital phases which did not exchange during alteration. There is only negligible exchange (< 10%) of detrital clay minerals such as illite and chlorite with water even when heated to 100°C over two years (O'Neil and Kharaka, 1976; Yeh and Savin, 1976). It seems that deep-sea silicates probably exchange isotopic O readily at low temperature only during formation or recrystallization. If the fractionation in palygorskite is assumed comparable to fractionation in carbonates and quartz, the temperature of formation for this palygorskite, although elevated above the sea floor, was probably not in excess of 50°C. Even if the detrital quartz present ( - 10%) is ~sO fractionated to 1.018, there is not the 50% necessary to depress the palygorskite deposit to its low ~sO value. An alternative hypothesis is that the formation waters from which the palygorskite deposit formed are meteoric or lighter than seawater, but this seems unlikely. It seems most likely that the O isotopes fractionated at higher hydrothermal temperatures. The 87Sr/S6Sr isotope ratio is 0.7192 + 0.0011 and 0.7154 + 0.0001 in the palygorskite and montmorillonite fractions, respectively. These Sr ratios are unusual since they are higher than either seawater (0.7092) or the basic volcanic rocks which may have provided precursor phases which might have altered to these authigenic minerals. On the other hand, the Sr ratio for palygorskite is further from seawater than montmorillonite, indicating perhaps a more labile Sr c o m p o n e n t in montmorillonite, in agreement with its much higher Ca exchange capacity.
38
The approximate age of the sample is roughly Pliocene--Miocene from either associated sparse nannofossils or extrapolated 23°Th/232Th deposition rates of roughly 1 mm/Ma in that area (Goldberg and Koide, 1962). The age may even be less than a million years considering the associated barites which have been He dated at 0.04--0.4 Ma with retention and excess U--Th nuclide corrections (T.M. Church, unpublished data, 1977). Using about 1 Ma as the age, the original S6Sr/87Sr complement for the palygorskite would be roughly 0.717, which according to Dasch (1969), could reflect a terrestrial inheritance. If one considers terrestrial detritus as possible precursors, a likely origin, as for the associated quartz, is aeolian material from the Asian subcontinent (Rex and Goldberg, 1958). Also, this palygorskite Sr-isotope ratio is significantly higher than the weak acid-leachable values reported by Daseh et al. (1971) for an Fe-rieh palagonite--smectite metalliferrous sediment, or isolated phfllipsites (Bernat, 1973). Both these other authigenie deposits from t h e Pacific yielded Sr-isotope ratios identical to modern seawater. In summary, the following conclusions can be drawn from the minor and isotopic chemistry of the palygorskite--montmoriUonite fractions: (1) 180/t60 ratios are lower than those expected for mineral formation in equilibrium with deep seawater, or those of detrital minerals formed under conditions of the earth's surface, and thus probably reflect a hydrothermal origin for this palygorskite. (2) STSr/86Sr ratios are higher than seawater, or that which can be explained as derived from seawater considering the age of the sediment and using common seawater Sr as a starting component, and thus suggest some detrital precursors for the origin of this palygorskite. CONCLUSIONS
It appears that the palygorskite and montmorillonite of the Proa 101-P Pacific deposit were formed, at least in part, from materials such as precursor detrital minerals rather than just dissolved seawater ionic species. Also, it appears that these minerals were formed by alteration reactions at temperatures higher than the sea floor, This agrees with the conclusion of Couture (1977) on the basis of silicaenrichment of palygorskite and coexistence with the silica-richzeolite,clinoptflolite.The associated clays and quartz minerals which occur in the core along with montmorillonite and palygorskite indicate that some of the precursor materials could be terrigenous in origin, which m a y not just be true for this palygorskite, in particular,but perhaps important for the origin of other deep-sea authigenic silicatesas well. ACKNOWLEDGEMENTS
T. Church gratefully acknowledges initial support of a C N R S - - N S F Post Doctoral Award under the sponsorship of Prof. C.J. All~gre of the Institute du Physique du Globe, University of Paris VI, Mr. Serge Fourcade of the
39
Institute graciously analyzed the oxygen isotopes. Subsequent support of the University of Delaware Research Foundation is also acknowledged. REFERENCES Bartholom~, P., 1966. Sur l'abondance de la dolomite et de la s~piolite dans les s~ries s~dimentaries. Chem. Geol., 1: 33--48. Bernat, M., 1973. Chronom~trie g ~ l o g i q u e h l'aide des isotopes a vie moyenne de l'uranium et du strontium. Thesis, University of Paris VI, Paris, 137 pp. Bonatti, E. and Joensuu, O., 1968. Palygorskite from Atlantic deep-sea sediments. Am. Mineral., 53: 975--983. Bowles, F.A., Angino, E.A., Hostermon, J.W. and Galls, O.K., 1971. Precipitation of deep-sea palygorskite and sepiolite. Earth Planet. Sci. Lett., 11: 324--332. Brown, G. (Editor), 1961. The X-ray identification and crystal structures of clay minerals. In: Minerals, Mineral. Soc., London. Christ, C.L., Hostetler, P.B. and Siebert, R.M., 1973. Studies in the system MgO--SiO 2 CO 2 - H 2 0 , III. The activity constant of sepiolite. Am. J. Sci., 273: 65--83. Couture, R.A., 1977. Composition and origin of palygorskite-rich and montmorilloniterich zeolite-containing sediments from the Pacific Ocean. Chem. Geol. 19: 113--130. Dasch, E.J., 1969. Strontium isotopes in weathering profiles, deep-sea sediments, and sedimentary rocks. Geochim. Cosmochim. Acta, 33: 1521. Dasch, E.J., Dymond, J.R. and Heath, G.R., 1971. Isotopic analysis of metalliferous sediment from the East Pacific rise. Earth Planet. Sci. Lett., 13: 175--180. Fleischer, P., 1972. Sepiolite associated with Miocene diatomite, Santa Cruz Basin. Am. Mineral. 57: 903--913. Goldberg, E.D. and Koide, M., 1962. Geochronological studies of deep-sea sediments by the ionium/thorium method. Geochim. Cosmochim. Acta, 26: 417--450. Hathaway, J.C. and Sachs, P.L., 1965. Sepiolite and clinoptilolite from the mid-Atlantic ridge. Am. Mineral., 50: 852--867. Lawrence, J.R. and Taylor, H.P., Jr.,1972. Deuterium and oxygen-18 correlations; clay minerals and hydroxides in Quaternary soils compared to meteoric waters. Geochim. Cosmochim. Acta, 36: 1377--1393. McLean, S.A., Allen, B.L. and Craig, J.R., 1972. The occurrence of sepiolite and attapulgite on the Southern High Plains. Clays Clay Mineral., 20: 143--151. Millot, G., 1964. G~ologie des argiles. Masson, Paris, 500 pp. O'Neil, J.R. and Kharaka, Y.K., 1976.Hydrogen and oxygen isotope exchange reactions clay minerals and water. Geochim. Cosmochim. Acta, 40: 241--246. Peech, M., 1945. Determination of exchangeable cations and exchange capacity of soils-rapid micro-methods utilizing centrifuge and spectrophotometer. Soil Sci., 59: 25--38. Rex, R. and E.D Goldberg, 1958. Quartz contents of pelagic sediments of the Pacific Ocean. Tellus, 10: 153--159. Savin, S.M. and Epstein, S., 1970a. The oxygen and hydrogen isotope geochemistry of clay minerals. Geochim. Cosmochim. Acta, 34: 25--42. Savin, S.M. and Epstein, S., 1970b. The oxygen and hydrogen isotope geochemistry of ocean sediments and shales. Geochim. Cosmochim. Acta 34: 43--63. Siffert, B., 1962. Quelques reactions de la silice en solution: la formation des argiles. M~m. S~r. Carte Geol. Alsace Lorraine, 21. Velde, B., 1977. Clays and Clay Minerals in Natural and Synthetic Systems. Elsevier, Amsterdam 218 pp. (see especially pp. 140--156). Wollast, R., MacKenzie, F.T. and Bricker, O.P., 1968. Experimental precipitation and genesis of sepiolite at earth surface conditions. Am. Mineral., 53: 1645--1662. Yeh, H.W. and Savin, S.M., 1976. The extent of oxygen isotope exchange between clay minerals and sea water. Geochim. Cosmochim. Acta, 40: 743--748.
Comments
Report "Geochemistry and origin of a deep-sea Pacific palygorskite deposit "