As indicated by Lalou et al. (1993), oxidation of massive sulfides may cause mobilization of uranium isotopes and, consequently, such material may not be useful for age determinations by means of 230Th/234U and 231Pa/235U disequilibria. In our study most of the samples consist of pyrite, pyrrhotite, and marcasite with trace amounts of magnetite, chalcopyrite, quartz, and barite (Table T1), and from these mineralogical results there are no indications that oxidized sulfides were analyzed. It was only in Sample 9 that goethite was found apart from marcasite. This sample was obtained from the CORK, which was deployed during Leg 139 (1991) and recovered during Leg 169 (1996). Age dating of these minerals within such a short time scale is not possible using the methods applied. 232Th was detected in four samples (8, 13, 15, and 17). Whereas the uranium and thorium contents are relatively high in these samples, the effect of exogenous 230Th and 231Pa on the 230Th/234U and 231Pa/235U disequilibria is expected to be low.
Table T2
presents the data that were used to calculate the apparent 230Th/234U
and 231Pa/235U ages using Equations 1 and 2. Average
radionuclide activities were used for those samples where duplicate or
triplicate analyses were performed. The apparent 230Th/234U
and 231Pa/235U ages are between 8.2 and >300 ka and
between 8.2 and >120 ka, respectively. The relation between both radionuclide
ratios is shown in Figure F3.
In the case of age concordance, the measured 230Th/234U
and 231Pa/235U ratios of samples investigated should fall
on the concordance line (solid line in Fig. F3).
This concordance line was calculated for an initial 234U/238U
ratio of 1.14, which is the ratio of seawater (Chen et al., 1986). Only Samples
1, 2, and 7 (within a 2-
level) fall on this line and therefore their apparent 230Th/234U
and 231Pa/235U ages can be regarded as "true"
ages. This suggests a formation period between 8.5 and 16.0 ka (mean of 230Th
and 231Pa ages) for the clastic sulfides at the top of the BHMS and
the upper part of the massive sulfide zone (Fig. F4).
The large offset of the other ratios from the concordance curve may have several causes. One cause may be that the massive sulfides investigated are not a chemically closed system for 230Th and 231Pa and addition and/or loss of these isotopes may have occurred. Most investigations dealing with the behavior of thorium and protactinium in oceanic environments have shown that these elements are less soluble than uranium (Cochran, 1982). Therefore, discordant ratios are more likely to be caused by a chemically open system with respect to the gain and/or loss of uranium.
Most of the uranium in massive sulfides is believed to have a seawater origin (Lalou et al., 1993), because the 234U/238U ratio measured in most sulfides is similar to that of seawater (234U/238U = 1.14). Relative to the uranium content in seawater (3.32 ppb) the hydrothermal fluids are poor in uranium (0.06-0.18 ppb) (Chen et al., 1986). Laboratory experiments indicated that in H2S-bearing fluids the mobile form of uranium (U+6) may be stable and no reduction was observed (Anderson et al., 1989; Kochenov et al., 1977). In the presence of mineral surfaces, however, there is a reduction of U+6 to less soluble U+4 and the uptake of uranium from the solution was found to be very effective (Wersin et al., 1994). Uranium is sorbed as finely dispersed uraninite minerals on the surface of the sulfide minerals (Wersin et al., 1994), which have a low solubility under chemically reducing conditions (Langmuir, 1978).
Massive sulfides precipitate because of cooling during mixing of hydrothermal fluids with seawater. Lalou et al. (1996) have shown that uranium uptake by sulfides occurs during and/or very soon after their formation and that the resulting uranium content is a function of local environmental conditions and thus very variable. As long as a reducing environment persists, uranium loss of the sulfides is expected to be minimal. Recrystallization of the sulfide minerals was observed in the Bent Hill deposits (Zierenberg et al., 1998; Krasnov et al., 1994) and large parts of the deposit have undergone changes because of continued circulation of hydrothermal fluids and seawater through the sulfide mound after initial deposition of the sulfide (Marchig and Dietrich, 1996). Such processes are likely to affect the uranium concentrations of the sulfides.
The 234U/238U ratios in the samples investigated are within the analytical uncertainties in the same range as that of seawater (234U/238U = 1.14). If there was a closed system for uranium in the samples investigated, all ratios measured should fall on the solid line in Figure F5, which indicates the change of the seawater ratios with time (i.e., 230Th/234U). The effect of a recent loss of uranium (with 234U/238U = 1.14) is indicated by the broken line. As shown in Figure F5, it is not possible to decide within the analytical 234U/238U uncertainties whether individual samples have been a closed system for uranium or not.
There is remarkable change in uranium vs. depth distribution in the BHMS deposit (Fig. F6). Apart from two sulfide samples (15 and 17), which were derived from Hole 1035F, there is a decrease in the uranium concentrations of sulfides investigated with depth. All samples with concordant ages were sampled near the surface of the deposit and these samples also have high uranium concentrations. Therefore, it seems likely that a closed system of uranium exists only for sulfides deposited near the surface (e.g., Samples 1, 2, 6, and 7) and that in the deeper part recrystallization of the sulfide deposits resulted in an open system for uranium. In this case the true ages of the samples lying below the concordia curve (Fig. F3) are expected to be older than the calculated 230Th/234U and 231Pa/235U ages (231Pa age < 230Th age < true age), whereas for the two samples above the concordia curve the true ages should be younger than the calculated apparent 230Th ages (Cheng et al., 1998).
