A subset of samples obtained from Hole 1188A were found to have 5-10 wt% total sulfur contents and were initially analyzed with the inductively coupled plasma-atomic emission spectroscopy (ICP-AES) using the Sulfur Method (see "Geochemistry" in the "Explanatory Notes" chapter). However, these samples proved mismatched to the standard reference materials for sulfides and were, therefore, reanalyzed using the method developed for igneous rocks with a dacite-rhyodacite composition (see "Geochemistry" in the "Explanatory Notes" chapter). Despite the high degree of alteration, there was, in general, a much better match of the samples with the matrix of the igneous rock standards. In most altered rocks from Hole 1188A, a major portion of the total measured sulfur is sulfate sulfur (dominantly anhydrite, with barite and possibly other minor sulfates). In addition, samples that were analyzed by ICP-AES were analyzed for their total water-soluble sulfate by a gravimetric method to better characterize the nature and abundance of the sulfates.
A subset of samples from Hole 1188A was used to further examine the issue of sulfate dissolution (see "Sulfate Analysis" in "Geochemistry" in the "Explanatory Notes" chapter). The gravimetric dissolution experiments were repeated, without the addition of BaCl2, on aliquots (0.5 g) from the cleaned and ground samples mixed with 25 mL of nanopure water. After 12 hr at 4°C, the samples were centrifuged, and the supernatant filtered through 0.45-µm Gelman polysufone disposable filters. The extracted solutions were diluted with nanopure water to a volume of 50 mL and stored in acid-washed plastic vials for analysis. They were then analyzed using the shipboard ion chromatograph (Dionex, DX-120) to check the sulfate measurements obtained by gravimetric analysis. The waters were analyzed for Ca, Mg, K, and SO4 following the procedures outlined by Gieskes et al. (1991) for dissolved solids. The International Association of Physical Sciences Organizations (IAPSO) standard seawater was used for calibrating the instrument. The reproducibility of these analyses, determined by repeated measurements of standards for Ca, Mg, K, and SO4 were to within 3% to 5% on 1/200 diluted aliquots in nanopure water. It was found that the ion chromatography (IC) measurements for total dissolved cations were in good agreement with the gravimetric measurements for sulfate. However, the cation analysis of the solutions showed that other cations in addition to the Ca were also present and that not all dissolved sulfate is originally from pure anhydrite.
Following the analysis of the filtered supernatant liquid by IC, the same solutions were analyzed for Si, Mn, Al, and Na by ICP-AES. For these ICP-AES measurements, single element standards from Spex Certiprep were used. Both accuracy and precision are generally better than 1%-2% for these analyses.
The abundance of pyrite and clay minerals in cores recovered from Hole 1188A, and the scarcity of base metal sulfides, raised the question of whether the subsurface PACMANUS hydrothermal system might be relatively enriched in gold. This analysis could not be run on board, but the opportunity arose to send samples ashore via helicopter on 25 November 2000. Samples were sent to the Commonwealth Science and Industry Research Organisation (CSIRO) Division of Exploration and Mining in Sydney, Australia, for analysis.
Three representative composite samples weighing ~150 g each were prepared. The samples from Hole 1188A were a combination of rock chips taken from bags of fine-grained "residue" that had been collected for certain cores and of an approximately equal quantity of randomly picked small fragments from the trays of cores where no residue had been retained. Because the purpose was to test the possibility of an auriferous character, these were considered to adequately represent average material, though small for typical bulk assay work. The samples covered the following intervals:
At CSIRO Exploration and Mining in Sydney, the samples were dried, ground to nominal 200 mesh in a tungsten carbide ring mill, and thoroughly homogenized. Subsamples weighing 10 g were irradiated at the Lucas Heights Nuclear Reactor (Australian Nuclear Science and Technology Organization), and the neutron activation spectra were collected for 7 days and analyzed using routine methods by Becquerel Laboratories Pty Ltd at Lucas Heights.
Separate 0.1-g subsample splits of the grab samples were retained at CSIRO Exploration and Mining in Sydney, where they were analyzed by ICP-AES. The samples were dissolved in an HF-HClO4-HNO3 acid mixture, evaporated to dryness, and redissolved in an HNO3 solution for analysis by ICP-AES using a Spectro Analytical Instruments SpectroFlame instrument. The procedure normally retains all the sulfur present in the solution for whole-rock samples. Intensities were corrected for spectral interferences and referenced to synthetic calibration standards. Instrument performance was monitored with a range of standard reference soils.
Surface water samples were collected off the railing of the ship by bucket grab for seawater chemistry. Two samples were collected over the drill sites. Owing to the close proximity of the drill sites, and the position of the JOIDES Resolution over the drill sites, the three samples were thought to represent the surface water chemistry over the drill site. These samples of surface water were used as a comparison for the samples collected by the WSTP. The sampling probe was filled with deionized water for the trip down the drill string to the sampling level. The valves were timed to open and release the deionized water to be replaced by the borehole fluids at the desired depth. After equilibration time the valves were then closed for the trip back up the drill string. Analysis of salinity, pH, and alkalinity were conducted as soon as the samples were brought to the laboratory; IC and ICP-AES measurements for anions and cations were then completed on acidified samples with a nitric acid spike.
The results of the shipboard ICP-AES, NCS, IC, and gravimetric analysis of representative rock samples from Hole 1188A are given in Tables T11, T12, and T13 with the samples listed by assigned lithologic units (see "Igneous Petrology") and in order of increasing depth. They are reported in Table T11 ordered with respect to major element oxides expressed as weight percent. The values for total sulfur and water (by NCS) are reported in weight percent and follow the whole-rock compositions. The volatile free concentration data were recalculated to account for the abundance of H2O, S, and SO3 and are reported in the lower half of Table T11. The data for the NCS analysis of the unignited (pre-loss on ignition [LOI]) water and sulfur contents and the ignited (post-LOI) sulfur content are found in Table T12. The results for the four supernatant analyses by IC and ICP-AES methods are given in Table T13. Tables containing the data from the midcruise shore-based neutron activation analysis (NAA) and ICP-AES for the bulk samples are found in Tables T14 and T15, respectively, and are presented with relevant detection limits.
Tables T16 and T17 present the rock data from Hole 1188F. As with the data from Hole 1188A, the samples are listed by assigned units and in order of increasing depth. They are reported in Table T16 ordered with respect to major element oxides expressed as weight percent. The values for total sulfur and water (by NCS) are reported in weight percent and follow the whole-rock compositions at the bottom of the table. In Table T17, the comparison between sulfur pre- and post-LOI is presented. It was found that a modification to the LOI procedure was required to remove the majority of the sulfur in the samples. Roasting time at 1050°C was increased from 1 to 4 hr. This increased roasting time significantly improved the efficiency of sulfur removal in the samples from Hole 1188F.
Table T18 contains the data of the water samples collected from the boreholes using the WSTP tool. A single sample was taken from Hole 1188B, and two samples were taken at different times from Hole 1188F. Temperatures recorded by the WSTP for the collected fluids range from 3° to 21°C.
The raw ICP-AES intensity for each element of the standard reference rocks used (see "Geochemistry" in the "Explanatory Notes" chapter and Murray et al., 2000) were drift and blank corrected and were then used to calculate concentrations for the ignited samples tabulated in the top half of Table T11. For all subsequent discussions, figures, and plots, the major element oxide and trace element concentrations were recalculated to the original rock composition by multiplying the ignited oxide value by 100 + LOI% then dividing by 100. Note that LOI represents the volatile content in the powdered and dried samples and consists mainly of H2O+ and sulfur compounds. Trace amounts of CO2 and other minor volatile components may also be present but are considered negligible.
Late during the Leg 193 cruise, results from a sulfate sulfur doping experiment of samples of fresh dacite and standard reference materials suggested that sulfate contents >1 wt% S affect the ICP-AES analyses, leading to an underestimation of the oxide percentages of Si, Al, Fe, Ca, and Na. Therefore, the results of the ICP-AES analyses in samples that have a reported total sulfur content >1 wt% should be considered only as approximate concentrations (see "Geochemistry" in the "Explanatory Notes" chapter for discussion of instrument precision and accuracy).
For the midcruise shore-based analyses, the NAA intensities were corrected for spectral interferences and referenced to synthetic calibration standards. The precision of the ICP-AES for a majority of the elements (i.e., K, Na, and Cr) duplicated from the NAA analysis is higher than that of NAA. Performance was monitored with a range of standard reference soils.
The variations of selected major oxide and trace element compositions are plotted against depth and are compared with the bulk mineralogy as determined by XRD in Figure F114. The observed chemical variation is generally consistent with the bulk mineralogy, including any primary minerals retained in the rock (i.e., plagioclase), as well as the alteration minerals (i.e., chlorite and illite) described in the "Site 1188 Alteration Log". For example, increases and decreases of MgO coincide with the appearance and disappearance of chlorite downhole, the increases of total Fe (as Fe2O3) correlate with the occurrences of magnetite and/or pyrite, and Na2O is elevated in intervals containing plagioclase. The majority (14 out of 20) of the rocks sampled from Hole 1188A contain >1% total sulfur after ignition; SiO2, Al2O3, FeO, CaO, and Na2O data cited for these rocks are therefore preliminary.
Only one sample analyzed from Hole 1188A represents a fresh volcanic rock (Unit 1; Sample 193-1188A-2R-1, 9-12 cm). This rock can be classified as transitional between dacite and rhyolite (i.e., rhyodacite) using the total alkalis vs. silica (TAS), or as dacite using the International Union of Geological Sciences (CIPW norm) classification. A full chemical review for this unit was deemed inappropriate based on a single sample.
Three major alteration types have been identified from the cores recovered from Hole 1188A: (1) green silica-clay, (2) bleached (Bl), and (3) silicified. For more detailed descriptions of these alteration types, see "Hydrothermal Alteration". In Table T11 the alteration types assigned to each sample are noted. In Figure F114A the variation in XRD mineralogy with depth is compared with the geochemical profiles for key major elements. In general, the changes in MgO and total Fe (plotted as Fe2O3) correspond to the formation of chlorite, clay minerals, magnetite, and pyrite. In general, the abundance of Na, Mg, and K in the altered rocks correspond broadly with the presence of plagioclase, chlorite, and illite, respectively.
The bleached rocks show a decrease in their CaO, Na2O, K2O, and MgO contents relative to the fresh rock and to the GSC altered and silicified rocks. Although macroscopic and microscopic observations indicate an increase in quartz content (e.g., vesicle fill), the silicified rocks do not show elevated SiO2 contents when compared to the fresh dacite. This apparent contradiction can be attributed to the multiple phases of alteration identified. Silica may have been lost by early GSC alteration (clay forming) and regained by later silicification. If the magnitude of the silica fluxes are similar, then no apparent net change would be observed. Based on the preliminary nature of the shipboard major element data, which is a result of high S contents, this hypothesis could not be fully tested.
Comparison of total sulfur (TS) in the unignited samples to that of the ignited samples showed that traditional ignition to 1050°C does not fully remove all sulfur components (i.e., 1.09 wt% TS in ignited Sample 193-1188-14R-1, 105-108 cm [117.05 mbsf] compared to 2.98 wt% TS in the unignited [pre-LOI] sample). However, the proportion of sulfur retained during ignition varies from sample to sample, possibly indicating variable proportions of sulfur-bearing compounds that do not break down during the standard ignition procedure. A significant portion of the total sulfur found in the samples analyzed from Hole 1188A consists of soluble-sulfate sulfur, and its concentration is variable with alteration type. The values obtained for the total sulfur and calculated for the water-soluble sulfate phase are found at the bottom of Table T11. Gravimetric measurements were used to determine the water-soluble sulfate concentrations. However, in many samples, there is an apparent deficiency in Ca as measured to account for all of the sulfate as anhydrite. Replicate samples measured by gravimetric analysis yielded reproducible results. Owing to this apparent disparity, the second aliquot of the liquid from soaking the powdered rock (the supernatant) was analyzed by ICP-AES and IC for Ca, Na, Si, Al, and Mn. The results are also presented in Table T13. Although the total cation concentrations are not large enough to account for all the sulfate found in the dissolution experiments, they indicate that in the supernatant solution there is a mixture of other sulfate-complexing cations, besides Ca. However, these ions do not seem to play a part in the gravimetric measurements, as the XRD spectra of the precipitate only show barite peaks.
Based on soluble-sulfate concentrations, the data suggest that many of the altered rock samples in Hole 1188A contain from 5% to 15% anhydrite. The mass of the powdered rock residues dried after soaking in comparison to their original masses indicates between 10% and 20% loss, which is in general agreement with the observed range of sulfate concentrations found in the supernatant. In contrast, no significant sulfate dissolution could be detected after soaking minicores in seawater (see "Physical Properties" in the "Explanatory Notes" chapter).
The sulfate-sulfur doping experiment quantified the amount of total sulfur necessary to push the error for the ICP-AES analyses outside the ±2% error range at 1 wt% total sulfur (pre-LOI). The presence of the sulfate was directly influencing the ICP-AES measurements by either suppressing the ionization or by forming colloids or precipitates in the solution, which were not properly transferred into an aerosol by the ICP nebulizer. Alternatively, the high S contents may cause an unusual matrix mismatch between samples and standards. In either case, the high sulfate sulfur remaining in the sample post-LOI means that measured concentrations of some of the major oxides (SiO2, Al2O3, Na2O, CaO, and Fe2O3) are markedly underestimated.
The midcruise shore-based NAA and ICP-AES analyses of samples from Hole 1188A established that the subsurface PACMANUS hydrothermal system is not enriched in gold or elements commonly associated with gold. The three NAA samples from Hole 1188A contain from 3 to 9 ppb gold (Tables T14, T15), covering the same range shown by fresh dacites and rhyodacites from Pual Ridge (Moss et al., 2001). Zinc and copper contents are low in all samples in Hole 1188A. Barium levels reported by the midcruise shore-based analyses are comparable to those obtained shipboard for Hole 1188A. In the upper 100 m of Hole 1188A, some of the Ba is associated with barite (as noted in the XRD traces). Below 100 m, the majority of the Ba is associated with silicates. This is also observed in the comparability between NAA and ICP-AES data, which imply again that the Ba is contained in silicates rather than barite. Geochemical indicator elements for an epithermal or acid sulfate style of mineralization, especially Sb and As, are exceptionally low. These data, and the low precious metal and base metal contents of the cored samples, constitute a marked contrast from the massive sulfide chimneys of the PACMANUS hydrothermal field (Parr et al., 1996).
A single sample of fluid from Hole 1188B was taken from 3 mbsf. The data are presented below along with the fluid sample data for Hole 1188F.
The variations of selected major oxide and trace element compositions are plotted against depth and are compared with the bulk mineralogy as determined by XRD in the lower half of Figure F114B. The observed chemical variation is generally consistent with the bulk mineralogy and the alteration log, as observed in Hole 1188A. Examples of this agreement include the increases and decreases of MgO with the appearance and disappearance of chlorite, subtle increases of total Fe with increased contents of magnetite and/or pyrite, and the elevated Na2O concentrations in intervals containing fresh plagioclase. The H2O+ and TS data are presented in Table T17 and illustrate the effectiveness of the longer roasting time for removal of sulfur.
Fresh volcanic rock was not encountered in Hole 1188F. The Zr/TiO2 ratios of the altered rocks (where Zr is in parts per million and TiO2 in weight percent) show moderate variability around 250, whereas the fresh sample from the top of Hole 1188A gives a ratio of 300 and may suggest two different parent materials, both dacitic.
Two major alteration types have been identified from the cores recovered from Hole 1188F: GSC alteration and silicification. For more detailed descriptions of these alteration styles, see "Hydrothermal Alteration". In Table T16, the alteration types assigned to each sample are noted. The variation in XRD mineralogy with depth is compared with the geochemical profiles for key major elements. Similar to the relationship between chemical and mineralogical compositions in Hole 1188A, Na2O, MgO, and Fe2O3 are low and K2O is high where alteration is complete and illite is the dominant phyllosilicate, whereas these trends are reversed where fresh plagioclase is preserved and chlorite is abundant.
The salinity of the surface water at Site 1188 has a value typical of that expected for Equatorial Pacific waters (35 g/kg) (Sverdrup et al., 1942). The temperature remained constant at 30°C, with a pH of 8.4. The composition of fluids collected from both boreholes is dominated by bottom seawater. The salinity of the Hole 1188B sample, at 31 g/kg, seems anomalously low. This may be caused by a dilution of the fluid by deionized water used to fill the WSTP collection coil for transfer from the rig deck to the borehole. However, the Mn is anomalously high. Concentrations of major dissolved ions in the borehole water from the WSTP from within Hole 1188F are also presented in Table T18.