230Th Excesses Research Articles

U-series disequilibria generated by partial melting of spinel lherzolite

We have experimentally investigated the effects of crystal chemistry on the partitioning of U and Th between clinopyroxene and anhydrous silicate melt in a wide range of synthetic systems from 1 to 8 GPa and 1330 to 1735°C. We have also generated limited additional data for garnet–melt and pigeonite–melt U–Th partitioning. We find that for sub-calcic clinopyroxenes, of the variety which characterise the mantle solidus, D U is greater than D Th, as previously predicted from theoretical considerations based on the size of the M2 site into which substitution occurs. Melt composition appears to have a minor effect on partitioning. The new partitioning data have been combined with existing experimental data to derive a thermodynamically based expression that predicts D Th and D U from crystal chemistry, pressure and temperature. This model has been combined with published experimental determinations of clinopyroxene composition along the mantle solidus to derive the variation in D Th and D U/ D Th during mantle melting. A simple, one-dimensional model dynamic mantle melting using the new data shows that 230Th excesses, expressed as the activity ratio ( 230Th/ 238U), of up to 1.34 can be generated within the spinel lherzolite stability field. This value is in excellent agreement with the maximum 230Th excess (1.35) observed in mid-ocean ridge basalt (MORB). The magnitude of excess correlates linearly with the depth to the onset of melting. These findings, which are relatively insensitive to plausible variations in mantle porosity and melt productivity, are consistent with observed global variations in MORB 230Th excesses with ridge axial depth. We conclude that significant 230Th excess can be generated by melting of spinel lherzolite, in the absence of garnet.

Melting of the Earth's lithospheric mantle inferred from protactinium-thorium-uranium isotopic data

The processes responsible for the generation of partial melt in the Earth's lithospheric mantle and the movement of this melt to the Earth's surface remain enigmatic, owing to the perceived difficulties in generating large-degree partial melts at depth and in transporting small-degree melts through a static lithosphere. Here we present a method of placing constraints on melting in the lithospheric mantle using 231Pa-235U data obtained from continental basalts in the southwestern United States and Mexico. Combined with 230Th-238U data, the 231Pa-235U data allow us to constrain the source mineralogy and thus the depth of melting of these basalts. Our analysis indicates that it is possible to transport small melt fractions--of the order of 0.1%--through the lithosphere, as might result from the coalescence of melt by compaction owing to melting-induced deformation. The large observed 231Pa excesses require that the timescale of melt generation and transport within the lithosphere is small compared to the half-life of 231Pa (approximately 32.7 kyr). The 231Pa-230Th data also constrain the thorium and uranium distribution coefficients for clinopyroxene in the source regions of these basalts to be within 2% of one another, indicating that in this setting 230Th excesses are not expected during melting at depths shallower than 85 km.

Assessing the presence of garnet‐pyroxenite in the mantle sources of basalts through combined hafnium‐neodymium‐thorium isotope systematics

The existence of an enriched component in the mantle with a pyroxenitic or eclogitic composition and its importance for basalt genesis has been discussed for over two decades. Inferences about the depth of melting as well as the dynamics of melting based on the presence of garnet and the location of the spinel‐garnet transition are different if garnet‐pyroxenite is present in the peridotitic mantle. Trace element partition coefficients are dependent on composition, and the differences between garnet‐pyroxenite and peridotite are large enough to produce significant differences in trace element fractionation between melts derived from these different lithologies. Melts derived from garnet‐pyroxenite or eclogite‐bearing sources will have small or no 230Th excesses, which are largely independent of melting and upwelling rate. Melts derived from garnet‐peridotite will have significant 230Th excesses, which are dependent on melting and upwelling rate. We show that the combined hafnium‐neodymium‐thorium (Hf‐Nd‐Th) isotope and trace element data can distinguish between melts derived from peridotitic and pyroxenitic or eclogitic sources. We also present new Hf isotope data for Hawaiian basalts and use the combined Hf‐Nd‐Th isotope and trace element systematics to argue against the existence of garnet‐pyroxenite or eclogite in the source of Hawaiian basalts. It is especially the large variation in degree of melting for relatively constant isotopic composition that allows us to rule out garnet‐pyroxenite in the source of the Hawaiian basalts.

Insights into mid‐ocean ridge basalt petrogenesis: U‐series disequilibria from the Siqueiros Transform, Lamont Seamounts, and East Pacific Rise

Parent‐daughter disequilibria between (230Th)/(238U), (231Pa)/(235U) and (226Ra)/(230Th) (parentheses refer to activities) have been measured by thermal ionization mass spectrometry and inductively coupled plasma‐mass spectrometry in basalts from three tectonomagmatic settings of the East Pacific Rise (EPR) at 8°20′‐10°N. Mid‐ocean ridge basalts (MORB) from the Siqueiros Transform, the Lamont Seamounts, and the EPR ridge crest span a large compositional range from primitive, high‐MgO basalts with strong incompatible element depletions (DMORB) to typical normal MORB (NMORB) to rare incompatible element enriched basalts (EMORB) derived from a more enriched source isotopically. Concentrations of U vary from <13 ppb in DMORB to >400 ppb in EMORB while Th/U ranges from 2 in DMORB up to 3 in EMORB. The young‐looking high‐MgO basalts have (226Ra)/(230Th) that ranges from 3.2 to 4.2, while EMORB appear old being near secular equilibrium. Initial (231Pa)(235U) are very high (>2.5) in all of the Siqueiros basalts. Three basalts from the Lamont Seamounts have low incompatible element concentrations and low Th/U and are in secular equilibrium for (226Ra)/(230Th) while the sample located closest to the ridge axis has significant 226Ra and 231Pa excesses and minor 230Th excess. DMORB lack 230Th excess, have high excesses of226Ra and 231Pa, and resemble experimentally determined melts of peridotite at 1 GPa, implying derivation from relatively shallow level melting of spinel lherzolite at low residual porosity. Disequilibria for all three parent‐daughter pairs are consistent with typical axial NMORB resulting from mixing of melts derived from heterogeneous sources, specifically 90–95% DMORB with 5–10% EMORB. The observation that all samples, regardless of tectonomagmatic setting, lie on the same mixing trend suggests that melting beneath seamounts and transforms is similar to melting beneath the ridge axis. Variations in 230Th excess over short spatial scales imply that garnet‐bearing mafic veins create all of the 230Th excess observed in typical NMORB.

The role of clinopyroxene in generating U-series disequilibrium during mantle melting

We develop recent models of crystal-liquid partitioning (Blundy and Wood, 1994; Wood and Blundy, 1997) to predict the effects of crystal chemistry on the partitioning of U-series elements between clinopyroxene and coexisting silicate melt. With increasing pressure along the mantle solidus, clinopyroxene becomes richer in Na 2O and Al 2O 3 and these components are predicted to reduce the size of the large M2 site. This means that, because U 4+ is smaller than Th 4+, clinopyroxene should progressively discriminate in favour of U 4+ with increasing pressure. We find that the effect is sufficiently great that D U/D Th should change from <1.0 at pressures up to 1.0 GPa to >1.0 above 1.5 GPa. Thus, melting in the stability field of clinopyroxene should produce excess 238U relative to 230Th at low pressure and excess 230Th above 1.5 GPa. Observed excesses of 230Th in MORB may not, therefore, require the presence of garnet in the source region. We performed experiments at 1.5–1.9 GPa on compositions doped with U and Th to test the calculations. Crystals and quenched melts were analysed by ion microprobe. On the mantle solidus at 1.5 GPa, D U/D Th for clinopyroxene was found to be 1.07 ± 0.08, while a value of 1.19 ± 0.04 was obtained at 1.9 GPa. These results confirm the prediction that mantle clinopyroxenes have D U/D Th > 1.0 at high pressure. Absolute values of D U for clinopyroxene are 0.029 at 1.5 GPa and 0.023 at 1.9 GPa in good agreement with earlier measurements on aluminous clinopyroxenes. Corresponding values for orthopyroxene at 1.5 GPa are 1.7 × 10 −3 for D U and 8 × 10 −4 for D Th while olivine gave D U of 1.8 × 10 −5 and D Th of 1.2 × 10 −5 under similar conditions. When the new data are used in dynamic melting calculations we find that 230Th/ 238U activity ratios of up to about 1.23 in the liquid are consistent with melting in the spinel lherzolite field provided porosity is approximately 10 −3 and melting rate ≤ 3 × 10 −5 kg. m −3 yr −1. This confirms our model prediction that residual garnet is not required to generate excess 230Th in mantle melts. Solid-liquid partition coefficients for other members of the U decay series, Ra and Ac may be calculated from the model. For clinopyroxene we obtain D Ra of about 10 −6 and D Ac of 5 × 10 −4. The value for Ac is large enough that 227Ac activities may provide information on melt extraction processes. The M2 site of orthopyroxene is found to be elastically similar to the corresponding clinopyroxene site, which enables us to calculate for this phase, D Ra of 4 × 10 −10 and D Ac of 2 × 10 −6.

Th–U fractionation and mantle structure

In the Rio Grande rift (RGR) asthenospheric mantle-derived (high ϵ Nd) alkali basalts are strongly enriched in 230Th over 238U (10–27%). In contrast, lithospheric mantle-derived alkali and tholeiitic lavas (high ϵ Nd) have secular equilibrium ( 230Th/ 238U) values, within analytical error. These lavas are young, 3–50 ka old. Asmerom et al. reported ( 231Pa/ 235U) for one of the tholeiites of 1.40 [Y. Asmeron, H. Cheng, R. Thomas, M. Hirschmann, R.L. Edwards, 231Pa– 235U and 230Th– 238U Isotope Systematics in Continental Basalts, Mineral. Mag. 62 (1998) 81–92]. Thus, the lack of 230Th enrichment is not due to magma transport or post-eruption related decay. The contrast in the ( 230Th/ 238U) between asthenospheric and lithospheric mantle lavas is exactly similar to the contrast preciously described by Asmerom and Edwards [Y. Asmerom, R.L. Edwards, U-series isotope evidence for the origin of continental basalts, Earth Planet. Sci. Lett. 134 (1995) 1-7], based on a study of lavas from the Colorado Plateau and the Pinacate volcanic field (PVF), Mexico. Collectively, there is a distinct separation between low ϵ Nd value lavas with no 230Th enrichment on one hand and the high ϵ Nd lavas with large 230Th excesses, 10–40% on the other, irrespective of whether they are tholeiites or alkali basalts. This observation, in corroboration with ( 231Pa/ 235U) data, show that source mineralogy of the point of initiation of melting is the deciding factor in Th–U fractionation during mantle melting. The lavas with no 230Th enrichment are postulated to have a spinel lherzolite source, while a garnet peridotite source for lavas with 230Th enrichment. It is shown that, degree of 230Th enrichment during dynamic melting reflects the length and depth of the melting column. The factors that modulate the depth and length of the melting column include, variations in lithospheric thickness and heat flow. During the transient thermal pulse associated with igneous activity in continental rifts, the solidus should mirror the topology of the lithosphere. This is due to the thermal inertial of the lithosphere. The second related factor is the physical thickness of the lithosphere, which determines the height to which the melting column can rise to melt. Thirdly, the mantle solidus deepens in areas of high heat flow. The combined effects modulate the structure of the mantle solidus, resulting in differences in the length of the melting column and depth of initiation of melting. Thus, during dynamic melting, the areas with the thinnest lithosphere and highest heat flow, i.e. the longest and deepest melting column, have the highest degree of 230Th enrichment, similar to MORB.

Trace element partitioning during the initial stages of melting beneath mid-ocean ridges

We have measured partition coefficients for U, Th, REE, and high field strength element for orthopyroxene–liquid (3) and garnet–liquid (9) pairs from 2.4 to 2.8 GPa and for clinopyroxene–liquid (6) from 1.2 to 2.8 GPa. These are the first measurements of partition coefficients from experiments that are close to multiple saturation with an anhydrous lherzolite assemblage, similar to what is believed to be the source of primary melts formed beneath mid-ocean ridges. In determining the phase compositions of the anhydrous lherzolite we found that the composition of clinopyroxene on the anhydrous lherzolite solidus changes from high-Ca to low-Ca with increasing pressure. Garnet and clinopyroxene partition coefficients for the elements we studied (U, Th, Nb, Hf, Y, Zr, Ce, Nd, Sm, Er, Yb and Lu) are dependent on pressure, temperature, and composition; although crystal composition appears to be the dominant control on pyroxene coefficients. The garnet partition coefficients most applicable to lherzolite melting are significantly higher than those from previous studies, whereas the clinopyroxene partition coefficients are significantly lower. These new partition coefficients in combination with the new phase equilibria relax the constraints from uranium–thorium disequilibria on mantle porosities in the melting region and reconcile the implications of excess 230Th with melting models based on other geochemical and geophysical parameters [Kelemen et al., A review of melt migration processes in the adiabatically upwelling mantle beneath oceanic spreading ridges, Philos. Trans. R. Soc. London Ser. A 355 (1997) 283-3182; Toomey et al., Tomographic constraints on upper mantle structure beneath the MELT region of the East pacific Rise, Trans. AGU 78 (1997) 705]. With these new partition coefficients significant 230Th excess can be created with residual porosities of up to 1%.

Implications for sedimentation changes on the Iberian margin over the last two glacial/interglacial transitions from ( 230Th excess) 0 systematics

The Portuguese margin is well-suited for studies of the contrasts in North Atlantic circulation during glacial and interglacial times because of its rapid sediment accumulation rate. This paper reports a ( 230Th excess) 0-based study of sediment accumulation over the past 140 ky, a period which includes the last two glacial/interglacial transitions, in two cores at 2.4 and 3.5 km water depth on a slope transect at 40°N. Although the independently-determined mean sediment accumulation fluxes over the past 140 ky are unequivocally high with means of 13.2 and 10.5 g cm −2 ky −1 in the two cores, conventional application of the ( 230Th excess) 0 method yields consistently lower fluxes with means of 3.5 and 3.8 g cm −2 ky −1. These ( 230Th excess) 0 estimates are interpreted as representations of the regional depositional flux through time. This ( 230Th excess) 0 regional flux is composed of a carbonate flux of <1 g cm −2 ky −1 and a larger and variable clay input which indicates the importance of the sea level control on the clay input to the basin. Clay flux dominates the regional sediment accumulation which has a total flux of ∼2 g cm −2 ky −1 and a CaCO 3 content up to 50% in interglacial times, and a total flux up to ∼5 g cm −2 ky −1 and a CaCO 3 content down to 10% in glacial times. This pattern of change of sediment composition through time is also typical of the NE Atlantic, and sediment focusing (contourite formation) appears responsible for the high actual fluxes observed in glacial compared with interglacial times on the Iberian margin. The inventories of ( 230Th excess) 0 exceed those which could have been supplied from the overlying water column alone over 140 ky by factors of ×3.2 and ×4.0 in the two cores. This is ascribed to preferential current focusing of fine sedimentary material in glacial times as a result of a systematic change in deep ocean circulation in response to climatic forcing. The presence of Heinrich events in the sediments is clearly evident, but at this southerly latitude they produce muted regional increases in accumulation flux (<2 g cm −2 ky −1), apart from the large Heinrich event 4 which introduced an additional 10–20 g cm −2 ky −1 over ambient background levels.

Ra–Th–Sr isotope systematics in Grande Comore Island: a case study of plume–lithosphere interaction

Here, we present new mass spectrometry measurements of U–Th–Ra disequilibria and Sr isotopes for historical volcanics from the Karthala and La Grille volcanoes in Grande Comore Island, Comores Archipelago. Alkali basalts from the Karthala are characterised by large 230Th (33–47%) and 226Ra excesses (21–53%), and radiogenic Sr compositions (0.7034–0.7041). In contrast, La Grille basanites have less radiogenic 87Sr/ 86Sr (∼0.7032) with less 226Ra excesses (21%) but similar 230Th– 238U disequilibria (45%). The Karthala samples display much more scatter in the Ra–Th isochron diagram than in the Th–U isochron diagram, suggesting a decoupling of the two parent–daughter systems. Correlations between the Sr isotopes and U–Th–Ra disequilibria suggest a mixing relationship between La Grille and Karthala sources. La Grille basanites are shown to result from partial melting of the old metasomatised oceanic lithosphere beneath the archipelago whereas Karthala lavas are derived from the Comore plume. The presence of amphibole in the lithosphere is responsible for Ra–Ba, Ba–Th and Ra–Th fractionation during melting such that D Ba > D Th > D Ra. The partitioning of these elements clearly differs from partial melting of garnet–lherzolite where D Th > D Ra > D Ba. Part of the variations in the 226Ra excesses is attributed to radioactive decay between the source and the surface. Short residence times (<1000 year) preclude the presence of a large magma chamber and may be an argument for rapid segregation of the melts through cracks or veins during melt transport.

Systematics of U-series nuclides in primitive lavas from the 1730–36 eruption on Lanzarote, Canary Islands, and implications for the role of garnet pyroxenites during oceanic basalt formations

Inferences on mantle melting and related parameters are often derived from U-series results on oceanic basalts. However, most of these basalts come from magma chambers in which magma mixing amongst other processes is likely to have reduced the true compositional spread created during partial melting. Here are presented U–Th–Ra disequilibrium results on historical lavas of near-primary composition from Lanzarote, Canary Islands, where magma chambers have not been detected. During the 1730–36 fissure eruption, the magma composition varied with time from basanites through alkali basalts to tholeiites. Approximately 3–5 km 3 of primitive and often ultramafic xenolith-bearing lavas were produced. Abundances of Ba, Th and U generally decreased towards the end of the eruption, and show more than four-fold variations (e.g. Th = 5.75–1.55 ppm). Ba/Th and Ba/U vary by more than 75%, whereas Th/U vary by less than 10%. Excesses of 230Th over 238U range from 18% to 33% and generally increase with alkalinity. ( 230Th/ 232Th) decreased significantly from 1.02 to 0.94 with time and is positively correlated with the alkalinity of the lavas. Isotope ratios of Sr and ( 226Ra/ 230Th) also correlate with SiO 2. ( 226Ra/ 230Th) increased with time from 1.10 to 1.56. The correlations between ( 230Th/ 232Th), 87Sr/ 86Sr and SiO 2 suggest that the lava compositions represent nearly unmodified primary melts from a lithologically and compositionally heterogeneous mantle source. Decreasing ( 230Th/ 238U) and increasing ( 226Ra/ 230Th) in basanites to tholeiites can be explained by partial melting of a mixed garnet pyroxenite–lherzolite source, during which the proportion of pyroxenite and extent of melting diminished. Alternatively, the large variations of excess 226Ra and Ba/Th may reflect fluid addition to such a heterogeneous mantle source. In order to conserve the systematic variations of ( 226Ra/ 230Th) with other geochemical variables, transfer time through the lithosphere is likely to have been short relative to the half-life of 226Ra. A magma velocity through the lithosphere on the order of 10 −5 m/s (0.5 km/year) is compatible with the data. Both the Th–U disequilibria and the Th-isotope ratios of the Lanzarote lavas appear to be related to the mantle source lithology and composition, where garnet pyroxenite melting produced the early alkaline basalts and increasing melting of garnet lherzolite produced the tholeiites. Furthermore, correlations between ( 230Th/ 238U) and ( 232Th/ 238U) for different volcanic provinces worldwide, strongly suggest that Th–U disequilibria in oceanic basalts are principally caused by partial melting of garnet pyroxenites.

Melting of a subducting oceanic crust from U–Th disequilibria in austral Andean lavas

Understanding crustal genesis at convergent plate boundaries is important for determining mass transfer between different geochemical reservoirs in the Earth's mantle, and for deciphering the long-term growth of the continental crust. Most arc magmas are thought to be generated from fluid-induced melting of the mantle wedge above slabs of subducting oceanic crust1. Such magmas frequently display 238U enrichments or radioactive equilibrium2,3 between 238U and its radiogenic product 230Th. But where a young and hot oceanic crust is being subducted it may itself partially melt and produce calc-alkaline andesites and dacites, termed adakites4. Here we report a uniform excess of 230Th over 238U, but variable Th isotope ratios, in young adakites from the Andean austral volcanic zone south of the triple junction where the Chile ridge subducts beneath South America. We show that these results are compatible with the adakites having been formed by approximately 20% equilibrium melting due to amphibole decomposition in a heterogeneous5 oceanic crust. Moreover, both the degree of melting of the oceanic crust and its thermal structure appear to be uniform under most of the Andean austral volcanic zone. Such partial melting of subducted oceanic slabs may have occurred throughout the Earth's history where young oceanic plates were subducted.

UTh isotope disequilibria and ocean island basalt generation in the Azores

A suite of 23 alkalic lavas from the Azores islands of Sao Jorge, Pico, Faial, Terceira and São Miguel have been analysed for their major and tram element and Sr, Nd, Pb and Th isotopic compositions. Previous studies have shown that lavas along the MAR show increases in La/Sm and 87Sr/ 86 Sr coupled with decreasing Fe towards the Azores platform inferred to reflect increasing dominance of a plume component over a depleted MORB-mantle component. The La/Sm and Sr, Nd and Pb isotope data of lavas from São Jorge, Pico, Faial and Terceira overlap the high La/Sm, 87Sr/ 86Sr end of the MAR lava array inferred to represent the deep seated plume and they have low Fe 8.0 ∗ compared with for example, Iceland basalts. When compared with the MAR array, higher Fe 8.0 ∗ and Na 8.0 and lower Si 8.0 probably reflect lower degrees of melting at higher average pressures beneath a lithospheric lid. The combination of enrichment in incompatible elements (including volatiles) with depletion in Fe in the regional plume component is a feature most likely to have developed within the lithospheric mantle. Lavas from São Miguel require an additional high 87Sr/ 86Sr, 207Pb/ 204Pb, low 143Nd/ 144Nd component with high Th/Nb which is suggested to result from incorporation of sediment in the source of these lavas. UTh isotope data indicate 9–29% 230Th excesses similar to those observed in alkalic basalts from Iceland or Hawaii. This, combined with the observation that lavas from Sao Miguel at the periphery of the inferred plume have smaller 230Th excesses (average 14%) than those nearer the centre, including the recent measurements from the MAR-ridge (average 23%), suggests that there is no simple correlation between buoyancy flux or lithospheric lid thickness and 230Th excesses. Long-term source Th/U ratios inferred from Pb isotopes are greater than those measured indicating U addition some time prior to partial melting. The large variations in ( 230Th/ 232Th) and ( 238U/ 232Th) are negatively correlated with 87Sr/ 86Sr indicating that much of the variation in U/Th reflects long-lived source heterogeneity. Moreover, the degree of 230Th excess is also negatively correlated with 87Sr/ 86Sr suggesting a link between melting process and source heterogeneity. The variations in 230Th excess are unlikely to result from thermally controlled variations in solidus pressure since this would require temperature differences (250°C) that are improbable at the length scale between São Miguel and Terceira (200 km). Instead, it is suggested that the incorporation of sediment and volatiles beneath São Miguel has led to increased rates of melting, resulting in smaller 230Th excesses, consistent with the higher volcanic output on Sao Miguel compared with Terceira.

Ridge-hotspot interaction along the Mid-Atlantic Ridge between 37°30′ and 40°30′N: the U [sbnd]Th disequilibrium evidence

This study reports U Th disequilibrium data obtained by mass spectrometry for basaltic glasses collected along the Azores platform portion of the Mid-Atlantic Ridge (37°30′–40°30′N), a region characterized by both a geochemical and bathymetric gradient. High Th and U concentrations, as well as Th/U ratios, document an enriched geochemical signature. ( 230Th 238U ) activity ratios range from 1.20 to 1.35 and are thus systematically larger than most EPR MORBs reported in the literature. ( 230Th 232Th ) activity ratios show remarkable homogeneity for multiple samples taken from single dredge hauls. Additionally, samples with the highest Th concentrations (2.4 ppm) have among the highest Th isotope ratios. Taken together, these observations rule out assimilation of 230Th-rich sediment as an explanation for the 230Th 238U systematics. The relatively large 230Th excesses in the erupted lavas may be related to the influence of the enriched Azores mantle plume source. The lack of observed correlations between 230Th excess and trace element and isotopic indices of source enrichment, however, rules out source composition as an explanation for the variations in ( 230Th 238U ). Excess 230Th is correlated with the axial depth of the ridge in the study area, with the shallowest portions showing the largest extents of disequilibrium. This may reflect more melting in the presence of garnet for the shallow segments, and suggests that melting begins well within the garnet peridotite stability field (∼ 35 kbar) in the mantle beneath the Azores segment of the MAR. At the ridge segment scale, 230Th excesses tend to be smaller near segment boundaries. This could reflect differences in the melting process or less frequent magmatism in these zones. These results demonstrate the potential for U Th systematics to constrain the depth and degree of melting as well as the rate of mantle upwelling, even in the presence of source chemical heterogeneity.

A geochemical criterion for the recognition of Heinrich events and estimation of their depositional fluxes by the ( 230Th excess) 0 profiling method

A core from the Armorican Seamount (46°23.8′N, 12°32.8′W; 3849 m), in the deep NE Atlantic southwest of the British Isles, exhibits unusual Mg/Al ratios in the sediments deposited during the last glacial period. High Mg/Al ratio values are restricted within layers of ≤ 14 cm thick, and mark the presence of ice-rafted detrital material deposited by Heinrich events H1, H2 and H4. These peaks in the ratio are caused by local increases in the dolomite, quartz and feldspar content and concomitant dilutions in the phyllosilicate content, which overall increase Mg and decrease Al. The known times of the Heinrich events provide a chronology for the core, and indicate a considerably slower mean sediment accumulation rate during and after the glacial/interglacial transition (2.2 cm ky −1) than in earlier glacial times (mean 5.6 cm ky −1). Within the glacial, however, ( 230Th excess) 0 profiling demonstrates that accumulation fluxes were at least 7 × higher than normal during the deposition of Heinrich events (up to 25 g cm −2 ky −1 versus a background glacial level of 3.7 g cm −2 ky −1). Because of post-depositional bioturbation, peak fluxes are likely to have been even higher. The ( 230Th excess) 0 method also reveals the presence of event H3, which bears minor dolomite and is the least well developed of the last four Heinrich events. Despite the marked compositional and flux variations, the total mass deposited at this site is estimated at only 2.4, 4.9 and > 7.9 g cm −2 for events H1, H2 and H4 respectively. These inventories were deposited over maximum periods of 2.0, 1.1 and > 0.7 ky respectively, but delivery could have been much more rapid (of the order of a few hundred years).

A comparison of sediment accumulation chronologies by the radiocarbon and 230Th excess methods

Four Northeast Atlantic cores which exhibited well-correlated linear progressions of conventional radiocarbon age with depth were analysed to provide 230Th excess profiling data. The current evaluation of marine radiocarbon data suggests that conventional radiocarbon ages consistently underestimate true time. This deficit progressively increases from 4 to 22 ky BP. As a result, steady-state sediment accumulation rate models based on conventional radiocarbon ages are expected to be in error by up to 17%, dependent on the period over which the estimates are made. When proxy timescales from the radiocarbon and 230Th excess data for the four cores are compared over periods between 0–11.5 and 0–18.2 ky BP, the trend of cumulative time derived from the 230Th excess profiling data is shown to be consistent with the expected trend of conventional radiocarbon age vs the best current estimate of true time. This agreement supports the use of 230Th excess as a tracer for detailed estimates of changing deep-sea sediment accumulation fluxes.

Experimental investigations of the partitioning of Nb, Mo, Ba, Ce, Pb, Ra, Th, Pa, and U between immiscible carbonate and silicate liquids

Liquid carbonate/ liquid silicate partition coefficients ( D) for a suite of incompatible elements (Nb, Mo, Ba, Ce, Pb, Th, U) have been determined experimentally. For one experiment, D's for Ra and Pa were also determined. The partitioning behavior is shown to be a regular function of the ionic field strength ( z r ) of the partitioned elements. At constant temperature, increasing pressure from 7 to 20 kbar caused little change in the measured D. At constant pressure, increasing temperature causes all D's to approach unity, as the temperature nears the critical temperature and the two liquids become compositionally similar. Addition of P or F to the system enlarges the stability field of carbonate but does not affect the partitioning behavior of our suite of elements. However, both P and F are strongly partitioned into the carbonate liquid. Omphacite/carbonate liquid D's at 50–60 kbar have also been determined. Our results for D U and D TH are similar to those found previously for augite/silicate liquid. Thorium is more compatible than U in pyroxene, indicating that excesses of 230Th/ 238U found in MORB cannot be explained by fractionation between pyroxene and melt. The Soret effect in carbonate liquids is quite small, in contrast to the effect in silicate liquids. In general, we see no evidence for Soret separation in our experiments, implying that endmember Ca-, Na-, and Mg-carbonate liquids have simple speciations and mix nearly ideally. This observation suggests that the results of our liquid/liquid partitioning studies do not strongly depend on the exact composition of the carbonate. Our experimentally determined liquid-liquid partition coefficients can explain recent fractionations of 210T/ 238U and 226Ra/ 238U in carbonatitec lavas from the Oldoinyo Lengai volcano, but cannot explain similar fractionations in 230Th/ 238U. 210Pb has the shortest half-life of these nuclides, and is, therefore, the most sensitive to the exact timing and mechanism of fractionation. Because we can explain 210Pb/ 238U in terms of recent, instantaneous liquid immiscibility, we favor this model for the origin of the deviations of 210Pb/ 238U and 226Ra/ 238U from those expected under secular equilibrium. If our interpretation is correct, some other fractionation process(es) must have fractionated Th (and perhaps Pa) from U, Ra, and Pb in the Oldoinyo Lengai carbonatites.

238U [sbnd]234U and 232Th [sbnd]230Th in the Baltic Sea and in river water

The concentration ( C) of dissolved 238U, 234U, 232Th and 230Th in fresh and brackish waters from the Baltic Sea were determined using TIMS. The brackish waters range in salinity from that of sea water (SW) to 2.5‰. C 238U in oxygen-saturated, surface waters is well correlated with salinity and shows quasi-conservative behavior, as does Sr. Samples from the redox water interface show depletion in C 238U, demonstrating that dissolved U is being removed by FeMn oxyhydroxides. From a simple mixing relationship for the brackish water,δ 234U * = 1000‰ was calculated for the fresh water source in the northern Baltic. A study of the Kalixälven River over an annual cycle yields highδ 234U during spring and summer discharge and lower values during fall and winter, showing that different sources contribute to the U load in the river during different seasons. C 232Th and C 230Th in river water are governed by the discharge, reflecting the importance of the increased abundance of small particles ( < 0.45 μm) for the 232Th 230Th load at high discharge. 232Th/ 238U in river water is about 40 times less than in detrital material. In the brackish water, C 232 Th drops 2 orders of magnitude in the low salinity region ( < 5‰), reaching a value close to that of sea water at a salinity of 7.5‰. Almost all of the riverine 232Th must be deposited in the low-salinity regions of the estuary. The 230Th/ 232Th in river waters is about twice the equilibrium value for 232Th/ 238U (3.8). In the brackish waters, 230Th/ 232Th is greater by a factor of 10–100 than both river water and SW. The big increase in 230Th/ 232Th in the Baltic Sea waters over the riverine input indicates that the Th isotopes enter the estuary as a mixture of two carrier phases. We infer that about 96% of 232Th in river water is carried by detrital particles, whereas the other phase (solution, colloidal) has a much higher 232Th/ 232Th. Entering the estuary, the detrital particles sediment out rapidly, whereas the non-detrital phase is removed more slowly, causing a marked increase in 230Th/ 232Th in the brackish water. In SW, 230Th/ 232Th is closer to river input and detrital material than in brackish water. We conclude that in the deep sea, 232Th is almost exclusively dominated by windblown dust and can be used to monitor dust flux. The 230Th excess in Baltic rivers is produced in U-rich, 232Th-poor peatlands and trapped in authigenic particles and transported with the particles. Time scales for producing the 230Th excess are ∼ 2000–8000 yr. This is younger than, but comparable to, the time of the latest deglaciation, which ended some 9000 yr ago when the mires were forming. These results have implications for the possible mobility of actinides stored in repositories.

Compositional controls on the partitioning of U, Th, Ba, Pb, Sr and Zr between clinopyroxene and haplobasaltic melts: implications for uranium series disequilibria in basalts

The partitioning of U, Th, Pb, Sr, Zr and Ba between coexisting chromian diopsides and haplobasaltic liquids at oxygen fugacities between the iron-wüstite buffer and air at 1285°C has been characterized using secondary ion mass spectrometry. The partition coefficients for Th, U and Zr show a strong dependence on the Al and Na content of the clinopyroxene. A good correlation between IVAl and D Th exists for all recent Th partitioning studies, providing a simple explanation for the two order of magnitude variation in D Th observed in this and previous studies [1,2]. Because mantle clinopyroxenes generally have greater than 5 wt% Al 2O 3, we suggest that the relevant partition coefficients for U and Th are between 0.01 and 0.02. While variations in Al and Na in clinopyroxene affect the absolute value of the Th and U partition coefficients, they have no effect on their ratio, D Th D U . Our results reinforce the inference that equilibrium partitioning of U and Th between clinopyroxene and melt cannot explain the observed 230Th excesses in basalts. Indeed, under the oxygen fugacities relevant to MORB petrogenesis, clinopyroxene has little ability to fractionate U from Th ( D Th D U < 2 ), implying that chemical disequilibrium between melt and wall rock during transport is not required to preserve 230Th excesses generated in the garnet stability field. If the Ba partition coefficient serves as an analog for Ra and the partition coefficient of U 5+ serves as an analog for Pa 5+, then 226Ra and 231Pa excesses can be generated by clinopyroxene-melt partitioning. Using compositionally dependent partition coefficients, a melting model is used to show that equilibrium porous flow can explain variations in uranium series activities from the East Pacific Rise by varying the depth of melting.

Consequences of melt transport for uranium series disequilibrium in young lavas

Radioactive disequilibrium of 238U nuclides is commonly observed in young lavas and has often been used to infer the rates of melting and melt migration. However, previous calculations do not actually include melt transport. Here we explore the behaviour of short-lived radionuclides in a new calculation that includes the fluid dynamics of melt segregation. We emphasize that series disequilibrium results from the differences in residence time of parent and daughter nuclides. Unlike previous models, contrasts in residence times are controlled by differences in transport velocities caused by melt separation and continued melt-solid interaction throughout the melting column. This “chromatographic” effect can produce larger excesses of both 230Th and 226Ra within the same physical regime compared to previous calculations which do not include melt transport. Using this effect to account for U-series excesses leads to radically different inferences about the rates of melt migration. Where previous models require rapid melt extraction, our calculation can produce larger excesses with slow melt extraction. Nevertheless, reproducing the large ( 226Ra/ 230Th) activity ratios observed in fresh mid-ocean ridge glasses is still problematic if the residence times are controlled solely by bulk equilibrium partitioning. While it still remains to be shown conclusively that the large 226Ra excesses are produced during melting, our calculation only requires differences in transport velocities to produce secular disequilibrium. Thus we speculate that other processes, such as crystal surface interaction, may also contribute to the production of the observed excesses.

The generation of uranium series disequilibria by partial melting of spinel peridotite: constraints from partitioning studies

Alkaline earth, Pb and actinide partition coefficients have been measured for olivine-, orthopyroxene- and clinopyroxene-melt pairs in experimental charges using secondary ion mass spectrometry. The bulk partition coefficients for spinel peridotites in equilibrium with basaltic melts are 1.1 × 10 −4 and 1.7 × 10 −4 for U and Th, respectively; both are dominated by the presence of clinopyroxene. These new data demonstrate that the bulk partition coefficient for uranium is lower than that for thorium in spinel peridotites, even if hydrous or accessory phases are stable at the solidus. U is similarly incompatible compared to Th for spinel peridotite in equilibrium with carbonate melts. These results demonstrate that the partial melting of a spinel peridotite at chemical equilibrium cannot account for the ( 230Th/ 238U) ratios in excess of unity observed in MORB. The generation of the observed ( 230Th) excesses by disequilibrium partial melting of spinel peridotite requires that the diffusivity of Th is greater than that of U, which is improbable, as the ionic radius of Th is greater than that of U. Similarly, ( 230Th) excesses can be produced on melt percolation only if the bulk partition coefficient of Th is smaller than that of U, and cannot therefore be generated in the spinel stability field. ( 230Th/ 238U) ratios greater than unity might be produced by slow, near fractional partial melting in the garnet stability field, if D U * gt/L is large compared to D Th * gt/L and D Th * cpx/L.