The complete set of the recognized secondary phases together with their distribution and relative abundance in the pillow basalts from the different sites is reported in Figure 4. Secondary products are found mainly in four distinct modes of occurrence: (1) vesicle or cavity linings or fillings; (2) coatings, fracture fillings, and veins; (3) pseudomorphic replacement of mafic phenocrysts and microphenocrysts (mainly olivine and subordinately clinopyroxene); and (4) patches within mesostasis.
Clay minerals are ubiquitous and are the most abundant secondary products, present in all four modes discussed above. Clay minerals were determined optically, following the criteria of Honnorez et al. (1983) and Laverne et al. (1996), by XRD data, including those of air-dried and glycolated powder mounts, and by microprobe quantitative and semiquantitative analyses.
XRD data of clay minerals indicate that trioctahedral smectite is the predominant phyllosilicate (d001 = 13-15 Å expanding after glycolation to d001 = 17.1-17.8 Å and d060 = 1.53 Å; Fig. 5). Subordinate types of mixed-layer clays, or mechanical mixtures of two or more clays, have been recognized: celadonite-smectite (Sites 1024-1029 and 1031-1032), celadonite-smectite-nontronite (Sites 1026-1027, 1029, 1032), smectite-chlorite (Sites 1027 and 1032), and smectite-montmorillonite (Sites 1027 and 1032).
Only one sample (Sample 168-1029A-25X-3, 77-82 cm) gave an X-ray spectrum unequivocally referable to a dioctahedral mica (celadonite type, Fig. 6), although there is evidence that trioctahedral smectite and possible smectite-celadonite mixed-layer minerals are present (as evidenced by the d060 = 1.532 Å reflection of smectite together with the d060 = 1.509 Å reflection of celadonite; Fig. 6).
On the basis of mineral chemistry and optical features we have distinguished five main different types of clay minerals or clay mixtures (as schematized in Table 2; Types 1-5), which can be classified into two main groups: saponite group and celadonite or celadonite mixtures group.
Saponite is the most common clay, and it has been identified at all sites. On the basis of more than 140 microprobe analyses, two subgroups may be divided (Types 1 and 2; Table 2): (1) Fe-rich saponites (MgO = 13-19 wt%; FeOt = 13-21 wt%); and (2) Mg-rich saponites (MgO = 18-23 wt%; FeOt = 8-12 wt%). Whereas Mg-rich saponite occurs often associated with celadonite and subordinately to Fe-oxyhydroxides in fractures or adjoining oxidation halos, Fe-rich saponite is restricted to the reducing assemblages together with carbonates and sulfides. Both saponite types are present as vesicle fillings, patches within mesostasis, fracture coatings and vein fillings, and pseudomorphic replacement of mafic minerals (mainly olivine).
Analyses were calculated on the basis of 20 oxygens and four hydroxyls (Table 3). Structural formulas were calculated considering that (Si + Al)IV = 8 atoms per formula unit (a.p.f.u.); the excess Al was assigned to octahedral sites together with Fet and Mg. When present, the slight excess of Mg in octahedral sites was placed in interlayer positions as an exchange cation. The compositional variations of the two saponite types (Table 3) agree well with the chemical constraints suggested by Andrews (1980) and reveal a considerable range of Fe-Mg substitution in octahedral coordination (Fig. 7), with the higher Fe contents occurring in samples from Site 1025 (i.e., in the fractionated ferrobasalt). Interlayer sites are mainly occupied by Ca and subordinately by K and Mg, whereas Na is generally below the detection limits.
Material identified as celadonite during Leg 168 shipboard analyses (Davis, Fisher, Firth, et al., 1997) appears in hand specimen as green to very dark green with resinous luster and waxy consistency, and in thin section as bright green to yellow-green cryptocrystalline or fibrous aggregates (Types 3-5; Table 2). Nevertheless (as discussed above), only one sample from Site 1029 shows an XRD pattern corresponding to a dioctahedral mica of the celadonite type. Celadonite or celadonite-bearing mixtures mainly occur within fractures and veins or in adjacent greenish and subordinately reddish brown halos (as vesicle fillings, irregular patches or pseudomorphic replacements).
Microprobe analyses were calculated on the basis of 20 oxygens and four hydroxyls per unit cell (Table 4). Aluminum was distributed between tetrahedral and octahedral coordination so that the tetrahedral occupancy (Si + Al) always sums to 8 a.p.f.u. No normalizations were performed for the octahedral sites to verify the possible excess of occupancy. Following these assumptions, all celadonite analyses invariably exhibit a large excess of cations in octahedral coordination (from 4.70 to 5.43 a.p.f.u.) compared to a value of 4.00 a.p.f.u. required by the dioctahedral structure. As evidenced in Figure 8, the octahedral occupancy excess shows a significant positive correlation with the total Fe content (a.p.f.u.). This evidence suggests that the high octahedral total is mainly related to the presence of a finely intermixed Fe-oxide contaminant extraneous to the celadonite structure and/or to the presence of interlayered Fe-rich smectite (as discussed by Andrews [1980] and Li et al. [1997]). TEM observations, performed on celadonites by Peacor (1992) and Li et al., (1997) demonstrate that submicrometer mixed layering is a common feature in celadonites formed at low temperatures.
Fe-oxyhydroxides and mixtures of Fe-oxyhydroxides and clay minerals strictly occur within fractures and veins or in reddish brown halos immediately around them (as vesicle filling, irregular patches within mesostasis, or pseudomorphic replacements). They commonly occur alone or associated with celadonite or celadonitic mixtures. Most of the Fe-rich compounds actually correspond to mixtures of Fe-oxyhydroxides and clay minerals (iddingsite), although optical examinations under microscope (reflected light) and microprobe analyses suggest the presence of pure and finely disseminated goethite and/or hematite lamellae in some microveins from Sites 1026 to 1027 and 1032.
Mixtures have been divided into four main types (Types 6-9; Table 2) on the basis both of the optical features (mainly the color in plane-polarized light) and chemical composition. Types 6 and 7 mixtures appear in thin section as massive aggregates of cryptocrystalline material varying in color respectively from bright orange yellow to orange red. As evidenced in SEM microphotographs, they clearly represent heterogeneous mixtures of Fe-oxyhydroxides or iddingsitic material and celadonite with the celadonite content progressively decreasing from Type 6 to 7 (Pl. 1; Fig. 1, Fig 2).
Types 8 and 9 correspond to iddingsitic mixtures that appear, both in hand specimen and thin section, as dense compounds varying in color respectively from dark brownish red to dark bright red. The mixture between Fe-oxyhydroxides and undifferentiated clay minerals is never discernible even at the resolution limits of the SEM, and consequently the actual end member cannot be recognized. The chemical composition of the two types is comparable (FeOt = 65-70 wt%; SiO2 = 9-12 wt%; and minor amounts [always <1 wt%] of CaO, K2O, and MgO), and no major variations correspond to the observed change in color.
Calcite and aragonite are widely distributed in pillow basalts, massive basalts, and diabase, although these minerals are not ubiquitous. Calcium carbonate occurs alone or associated with saponite in veins and vesicles, and also forms as a replacement of olivine, pyroxene, and plagioclase. The vein widths vary from about 0.5 to 2 mm. Crosscutting relationships with respect to clay veins and haloes surrounding clay veins indicate that the carbonate minerals generally occur late in the alteration sequence. Similarly, textures commonly indicate that aragonite formed earlier than calcite, and in many cases was actually replaced by calcite. The temperatures of present-day water-rock interaction at sites with evidence of carbonate mineralization range from Site 1025 at 35°C to Sites 1026-1027 at 63°C.
Chemical analyses of calcite and aragonite exhibit systematic differences (Table 5; Yatabe et al., Chap. 11, this volume). Calcites contain appreciable MgCO3, MnCO3, and FeCO3, as well as Ni, Cu, Zn, La, Ce, and Pb, whereas aragonites are relatively poor in these constituents. Relative to calcites, however, aragonites are enriched in Sr and Rb. These element distributions partly reflect crystal structural effects (Yatabe et al., Chap. 11, this volume).
Secondary sulfides occur as minor or trace constituents at all sites. They are present as fine-grained euhedral to subhedral crystals disseminated in open fracture surfaces, fine grains within altered mesostasis, and as veinlets and vesicle linings. Sulfides are commonly associated with Fe-rich saponite and, when present, with carbonates. Most of the sulfide occurrences correspond to pyrite and subordinately to pyrrhotite; to date, the very subordinate occurrences of chalcopyrite and millerite are of unknown origin (e.g., secondary alteration, igneous, deuteric?).
The presence of talc has been revealed by XRD analyses in samples from Sites 1026-1028 and 1032. It occurs almost exclusively as colorless to pale tan fibrous aggregates within pseudomorphic replacement of olivine together with Fe-rich saponites, or as the last product in the sequential fillings of fractures.
Quartz has been revealed by XRD analyses of two clay-bearing veins from Sites 1027 and 1032, but its presence was not confirmed by optical examinations under microscope.
Trace amounts of undifferentiated zeolites and phillipsite occur as colorless fibrous radiating aggregates in veinlets, mainly crosscutting the glassy rims of pillows, from Sites 1025 to 1032. Their presence has been confirmed by single crystal or separated aggregates X-ray analyses (with a Gandolfi Camera). Undifferentiated zeolites are present in significant amounts only in the hyaloclastite breccia from Hole 1026B (Section 168-1026B-3R-1) in the saponite-carbonates-bearing matrix.