IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.309312.201.2008

Microanalytical results

Compositions of the analyzed primary magmatic phases (clinopyroxene, pigeonite, plagioclase, and Fe-Ti oxides) occurring in the investigated basalts are shown in Table T2.

Pyroxenes

Analyzed clinopyroxenes show strong zonation with decreasing Mg# toward the rims. Their core compositions vary between En45Fs19Wo36 and En58Fs11Wo31 with Mg# varying from 70.4 to 84.1. The rims show marked iron enrichment (Fig. F2), with Mg# as low as 42.0, corresponding to a FeOtot concentration of 24.6 wt%. Phenocrysts are slightly more primitive in composition compared to groundmass crystals, as reflected by the slightly higher Mg# (Fig. F3A). As expected, the Mg# of the clinopyroxenes correlates with the Mg# of the corresponding bulk rock (Fig. F3A).

Relics of pigeonite as cores in augite

Pigeonite has been reported as a groundmass mineral occurring as discrete crystal and thin prismatic cores sandwiched by augite in some lavas (e.g., Crispini et al., 2006; Umino, 2007). Our analyses show that these pigeonites have CaO contents ranging from 4.4 to 5.2 wt%. Mg# of pigeonites in augite ranges from 67.3 to 82.2 and are generally slightly higher than those of the host. Two textural types of pigeonite were observed in the samples:

  1. Type I pigeonites (Figs. F4, F5A) correspond to those described in Crispini et al. (2006) and Umino (2007) and can be easily observed under the light microscope as elongated prismatic crystals embedded in augite with a sharp contact between the two. Both pigeonite and augite show strong variations in Mg#. This textural type occurs in the fine-grained samples of Unit 1256D-1, which has been interpreted as a ponded lava flow, as well as in one fine-grained sample (309-1256D-118R-1, 43–48 cm) of the flow–dike transition.
  2. Type II pigeonites (Figs. F5B, F5C) show different textural features and were first observed in cryptocrystalline and microcrystalline samples of the sheet flows via BSE images. In contrast to Type I pigeonites, they appear as strongly disrupted, often cloudy diffuse patches in intensely zoned groundmass clinopyroxenes, meaning that single pyroxenes bear one or multiple diffuse cores of pigeonitic composition with a more diffuse contact between each other. Because of the strong zonation of the host clinopyroxene, Type II pigeonite is difficult to detect with the light microscope.

The observation of pigeonite in Hole 1256D is significant. Although low-Ca pyroxene is present in many gabbroic rocks from the plutonic crust of fast-spreading ridges (e.g., Hole 1256D; see the “Site 1256” chapter and Gillis, Mével, Allan, et al., 1993), it is normally not observed as phenocrysts in the corresponding erupted lavas (“the orthopyroxene paradoxon”). Here, low-Ca pyroxene was detected, but only as relics with the composition of pigeonite and not of orthopyroxene, which is the typical low-Ca pyroxene occurring in gabbros. Future studies will shed light on this phenomenon.

Sector zoning

Figure F6 presents a BSE image of clinopyroxene showing two domains that can be distinguished by a small difference in their gray levels. The brighter zone is, relative to the darker zone, enriched in Al, Ti, Cr, and Ca and impoverished in Fe and Mg (Fig. F6). Such zoning is hard to detect during routine microprobe analysis and is probably the reason for the relatively large standard deviations related to some averages (Table T2).

Plagioclase

In general, groundmass plagioclase shows marked zoning, with a significant decrease in An content toward the rims. Although the An content of the cores varies between 55.5 and 69.3 mol%, the rims are much lower in An, ranging from 34.7 to 67.1 mol%. The standard deviation for the analyses are large, and some cores show exceptionally high An contents up to >80 mol%. Figure F7 shows some concentration profiles for An to illustrate zoning trends observed in groundmass plagioclases. An and FeOtot content in the plagioclase are negatively correlated, suggesting the bulk rock evolution trend to more iron-rich composition with increasing differentiation (Fig. F8). This is confirmed by a diagram of An content versus bulk Mg# of the host rocks that shows a positive correlation (Fig. F3B).

Analyzed phenocrysts are significantly enriched in An content relative to the groundmass plagioclase, reaching values up to 83.9 mol% (Figs. F3B, F8). This enrichment implies that these phenocrysts are out of equilibrium with the groundmass, representing obvious crystallization products of more primitive magmas, although rims have lower An (<70.9 mol%). Some plagioclase phenocrysts are completely altered to albite, probably because of secondary hydrothermal alteration.

Fe-Ti oxides

Coexisting magnetite and ilmenite were found in only one sample of the lava pond (Fig. F9). For this sample, equilibration temperature and oxygen fugacity was calculated using the “QUILF” software (Andersen, 1993). Equilibrium temperature was estimated to be 784° bration during cooling. Calculated oxygen fugacity corresponds to ΔFMQ of –0.4 (0.4 log units below the oxygen fugacity of the fayalite-magnetite-quartz buffer), which is within the range of oxygen fugacities observed in fresh mid-ocean-ridge basalt (MORB) glasses (Bezos and Humler, 2005).

All other samples contain only titanomagnetite. In the diagram for Fe-Ti oxides (Fig. F10), they plot on the ulvospinel-rich side of the ulvospinel-magnetite solid solution.

Mineral downhole variations

We observed slightly lower iron content and a slightly higher An in plagioclase, along with a higher clinopyroxene Mg# in basalt, in the sheeted dikes, compared to the immediately overlying flows (Fig. F11), but no overall trend directly correlated to depth.

Differentiation trend

Kvassnes et al. (2004) estimated typical differentiation trends for MORB by plotting Mg# of clinopyroxene versus An contents of plagioclase in gabbros from different tectonic settings (“dry” and “wet” fractionation). As expected, our data reveal a trend typical for “dry” fractionation (Fig. F12). Including the corresponding mineral data from the two gabbro screens in the future will help clarify the genetic relationship between gabbros and basalts from the drilled section.