Beschreibung: Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the Massachusetts Institute of Technology and the Woods Hole Oceanographic Institution September 1983

Beschreibung: Novel methods were developed for the determination of 12 of the 14 Rare Earth Elements (REE) in seawater. Initial extractions of the REE by chelating ion exchange chromatography is followed by cation exchange for removal of co-extracted U and remaining traces of major ions. Finally traces of U are removed by anion exchange before irradiation for 8 hours at a flux of 5 x 1013 neutrons.cm-2.sec-l. After post-irradiation separation of 24 Na, the gamma spectra are recorded over four different time intervals with a Ge(Li) detector. An internal standard (144Ce) is carried all along the procedure for improved precision by avoidance of counting geometry errors. Vertical profiles are reported for three stations in respectively the Northwest Atlantic Ocean, the Eastern Equatorial Pacific Ocean and the Cariaco Trench, an anoxic basin. This data set represents the first detailed profiles of Pr, Tb, Ho, Tm and Lu in seawater, together with profiles of La, Ce, Nd, Sm, Eu, Gd and Yb. The first observations of positive Ce anomalies in seawater are ascribed to regeneration of Ce under reducing conditions. The first reported positive Gd anomalies are ascribed to the unique chemical properties of the Gd(III)-cation, which has an exactly half-filled 4f electron shell. Concentrations of the REE range from 0.3 pmol.kg-l (Lu) to 86 pmol.kg-l (Ce) and are among the lowest reported so far for trace elements in seawater. The REE as a group typically exhibit a quasi-linear increase with depth. In the deep water there appears to be some degree of correlation with silicate. Concentration levels in the deep Pacific Ocean are 2-4 times those in deep Atlantic waters. Ce has an opposite behaviour, with very strong depletions in deep Pacific waters. In the Cariaco Trench all REE, but especially Ce, are strongly affected by the chemical changes across the oxic/anoxic interface. The REE distributions normalized versus shales (crustal abundance) exhibit four major features: i) a gradual enrichment of the heavy REE, most strongly developed in the deep Pacific Ocean. This is compatible with the stabilization of heavy REE by stronger inorganic complexation in seawater as predicted by the TURNER- WHITFIELD-DICKSON speciation model. ii) the first description of positive Gd anomalies, in agreement with the anomalously strong complexation of the Gd(III)-cation predicted by the same speciation model. iii) most commonly negative, but sometimes positive, Ce anomalies. iv) a linear Eu/Sm relation for all samples. Distributions of the dissolved REE in ocean waters seem to be dominated by their internal cycling within the ocean basins. With a few notable exceptions, the ultimate external sources (riverine, aeolian, hydrothermal) and sinks (authigenic minerals) appear to have little impact on the spatial distribution of the REE in oceanic water masses. Analogies with distributions of other properties within the oceans suggest that the REE as a group are controlled by two simultaneous processes: A) cycling like or identical to opal and calcium-carbonate, with circumstantial evidence in support of the latter as a possible carrier. B) adsorptive scavenging, possibly by manganese-oxide phases on settling particles. The latter mechanism is strongly supported by the parallels between REE(III) speciation in seawater and the 'typical 1 seawater REE pattern. This general correspondence is highlighted by the very distinct excursions of Gd in both Gd(III) speciation and the observed seawater REE patterns. Combination of both apparent mechanisms, for instance scavenging of REE by adsorptive coatings (Mn oxides) on settling skeletal material, is very well conceivable. Upon dissolution of the shells at or near the seafloor the adsorbed REE fraction would be released into the bottom waters. The observations of positive Ce anomalies in Northwest Atlantic surface waters, enhanced Ce anomalies and Mn levels in the OZ-minimum zone of the Eastern Equatorial Pacific Ocean, and enhanced Ce concentrations in anoxic waters all support the contention that a vigorous cycling driven by oxidation and reduction reactions dominates both Ce and Mn in the ocean basins. Under conditions of thermodynamic equilibrium, Ce tends to become depleted in well-oxygenated open ocean waters, and normal or enriched in waters below a pOZ threshold of about 0.001-0.010 atm partial pressure. The latter threshold level generally lies below the sediment/water interface. However, the kinetics of oxidation (and reduction) of Ce appears to be slow relative to various transport processes. This leads to disequilibria, i.e. a major uncoupling of the pOZ threshold level and the Ce anomaly distribution. The REE are definitely non-conservative in seawater and in general the REE pattern or 143Nd/144Nd isotopic ratio cannot be treated as ideal water mass tracers. The continuous redistribution of Ce within the modern ocean, combined with the likelihood of active diagenesis, precludes the use of Ce anomalies as indicators of oxic versus anoxic conditions in ancient oceans. On the other hand, the Eu/Sm ratio, possibly combined with 143Nd/144Nd , would have potential as a tracer for understanding modern and ancient processes of hydrothermal circulation.

Beschreibung: This research was supported by Department of Energy contract DE-AS02-76EV03566 and Office of Naval Research Contract NOOOl 4-82-C-00l 9 NR 083-004.