Quentin Williams

Seismic Evidence for Partial Melt at the Base of Earth's Mantle

The presence of an intermittent layer at the base of Earths mantle with a maximum thickness near 40 kilometers and a compressional wave velocity depressed by ∼10 percent compared with that of the overlying mantle is most simply explained as the result of partial melt at this depth. Both the sharp upper boundary of this layer (<10 kilometers wide) and the apparent correlation with deep mantle upwelling are consistent with the presence of liquid in the lowermost mantle, implying that the bottom of the thermal boundary layer at the base of the mantle may lie above its eutectic temperature. Such a partially molten zone would be expected to have enhanced thermal and chemical transport properties and may provide constraints on the geotherm and lateral variations in lowermost mantle temperature or mineralogy.

The core-mantle boundary layer and deep Earth dynamics

Recent seismological work has revealed new structures in the boundary layer between the Earths core and mantle that are altering and expanding perspectives of the role this region plays in both core and mantle dynamics. Clear challenges for future research in seismological, experimental, theoretical and computational geophysics have emerged, holding the key to understanding both this dynamic system and geological phenomena observed at the Earths surface.

Seismic Detection of the Lunar Core

Reinterpreted Apollo-era seismic data from the Moon reveal a solid inner core and a fluid outer core. Despite recent insight regarding the history and current state of the Moon from satellite sensing and analyses of limited Apollo-era seismic data, deficiencies remain in our understanding of the deep lunar interior. We reanalyzed Apollo lunar seismograms using array-processing methods to search for the presence of reflected and converted seismic energy from the core. Our results suggest the presence of a solid inner and fluid outer core, overlain by a partially molten boundary layer. The relative sizes of the inner and outer core suggest that the core is ~60% liquid by volume. Based on phase diagrams of iron alloys and the presence of partial melt, the core probably contains less than 6 weight % of lighter alloying components, which is consistent with a volatile-depleted interior.

Hit-and-run planetary collisions

Terrestrial planet formation is believed to have concluded in our Solar System with about 10 million to 100 million years of giant impacts, where hundreds of Moon- to Mars-sized planetary embryos acquired random velocities through gravitational encounters and resonances with one another and with Jupiter. This led to planet-crossing orbits and collisions that produced the four terrestrial planets, the Moon and asteroids. But here we show that colliding planets do not simply merge, as is commonly assumed. In many cases, the smaller planet escapes from the collision highly deformed, spun up, depressurized from equilibrium, stripped of its outer layers, and sometimes pulled apart into a chain of diverse objects. Remnants of these ‘hit-and-run’ collisions are predicted to be common among remnant planet-forming populations, and thus to be relevant to asteroid formation and meteorite petrogenesis.

Mantle Melting and Basalt Extraction by Equilibrium Porous Flow

The chemical composition of mid-ocean ridge basalt, the most prevalent magma type on the planet, reflects the melts continuous reequilibration with the surrounding mantle during porous flow. Models of basalt extraction that account for the observed uranium-series disequilibria on the Juan de Fuca ridge constrain both the abundance of melt beneath ridges (0.1 to 0.2 percent) and the style of mantle melting. Unlike models that incorporate near-fractional melts (dynamic melting), mixing of equilibrium porous flow melts derived from heterogeneous source materials quantitatively explains the uranium-series observations.

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-wustite 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 DTh exists for all recent Th partitioning studies, providing a simple explanation for the two order of magnitude variation in DTh observed in this and previous studies [1,2]. Because mantle clinopyroxenes generally have greater than 5 wt% Al2O3, 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, DThDU. 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 (DThDU < 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 U5+ serves as an analog for Pa5+, 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.

Seismological constraints on a possible plume root at the core–mantle boundary

Recent seismological discoveries have indicated that the Earths core–mantle boundary is far more complex than a simple boundary between the molten outer core and the silicate mantle. Instead, its structural complexities probably rival those of the Earths crust. Some regions of the lowermost mantle have been observed to have seismic wave speed reductions of at least 10 per cent, which appear not to be global in extent. Here we present robust evidence for an 8.5-km-thick and ∼50-km-wide pocket of dense, partially molten material at the core–mantle boundary east of Australia. Array analyses of an anomalous precursor to the reflected seismic wave ScP reveal compressional and shear-wave velocity reductions of 8 and 25 per cent, respectively, and a 10 per cent increase in density of the partially molten aggregate. Seismological data are incompatible with a basal layer composed of pure melt, and thus require a mechanism to prevent downward percolation of dense melt within the layer. This may be possible by trapping of melt by cumulus crystal growth following melt drainage from an anomalously hot overlying region of the lowermost mantle. This magmatic evolution and the resulting cumulate structure seem to be associated with overlying thermal instabilities, and thus may mark a root zone of an upwelling plume.

Crystal chemical control of clinopyroxene-melt partitioning in the Di-Ab-An system: implications for elemental fractionations in the depleted mantle

Abstract The partitioning of fifteen trace elements (Rb, Sr, Zr, Nb, Ba, La, Ce, Nd, Sm, Gd, Yb, Hf, Ta, Pb, and Th) between clinopyroxene and synthetic melt has been studied in two compositions along an isotherm in the diopside-albite-anorthite ternary at 1 bar pressure. The two compositions correspond to ∼Di65An35 and ∼Di55Ab45 and produce clinopyroxenes distinct in chemistry while melt compositions range from 49 wt% SiO2 to 61 wt% SiO2. The partition coefficients of high field strength elements (HFSE) increase by factors of 2–8 in Di-An experiments relative to Di-Ab experiments while other elements show very little change (±20%) between compositions. The change in HFSE partitioning correlates with increases in tetrahedral Al203 (IVAl) content of clinopyroxenes in the anorthite-bearing experiments. Changes in DTa/DNb also correlate with IVAl based on a survey of previously published determinations. Tests of models of trace element substitution energetics produce values for Young’s modulus (E) and optimum D (Do) consistent with previous results for clinopyroxene for mono-, di-, and trivalent cations. The wide variations in partitioning behavior for tetra- and pentavalent cations are also consistent with these models because the high values for E make partition coefficients and relative HFSE partitioning sensitive to small changes in composition. The overall increase in HFSE partitioning and Do for the DiAn composition is consistent with Do increasing as a function of IVAl, consistent with the role of IVAl in charge balancing HFSE. However, IVAl must also cause lattice changes that affect the ability of the clinopyroxene to discriminate between Nb and Ta. There are two important implications to the observed dependence of HFSE partition coefficients on clinopyroxene aluminum content. First, HFSE will be fractionated from their adjacent REE within ultramafic samples during melting of spinel lherzolite as clinopyroxene Al2O3 content decreases. Second, fractionations between Nb and Ta and Zr and Hf observed in mantle-derived magmas are consistent with extraction of melt in equilibrium with spinel lherzolite having clinopyroxenes with ∼5 wt% Al2O3. The fractionations among HFSE (Nb/Ta, Hf/Zr) and between HFSE and REE observed in both arc magmas and upper mantle peridotites may simply reflect prior depletion by major melting events (F > 10%) which left clinopyroxene as a residual phase. We speculate that the peridotitic sources for MORB and arc lavas are similar in composition with both having significant HFSE anomalies. However, MORB do not typically record HFSE anomalies because of the complementary contribution of HFSE from enriched mafic veins interspersed within the peridotite.

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 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.

A geochemically consistent hypothesis for MORB generation

Abstract Geochemical observations of MORB including U-series disequilibria are used to examine the processes and timescales of MORB melt generation. Incompatible elements in MORB suggest that the MORB source region consists of a depleted lherzolite matrix interspersed with chemically enriched mafic veins. Wide variations in Th/U distinguish these source variations in MORB better than 87 Sr / 86 Sr and document that the relative chemical homogeneity of normal MORB reflects efficient melt mixing rather than a homogeneous source. Spinel compositional variations in MORB and in mantle solids (abyssal peridotites and dunites) reflect reactive flow of melts having significant compositional variations. High Cr# spinels result from reactive flow of chemically enriched melts derived from the mafic vein source ascending through the lherzolite of the upper melting column. High Cr# and TiO2 contents in dunite spinels indicate that dunites form by reactive flow of enriched melts through the upper melting column. Once formed, dunites act as high permeability pathways for melt from surrounding lherzolite and are responsible for the “fractional signatures” observed in the major element chemistry, melt inclusions, abyssal peridotites and Lu–Hf systematics of MORB. Based on the recognition that there are two sources melting beneath ridges that have different porosity characteristics, a melting model consistent with evidence for both fractional and equilibrium porous flow melting is proposed. In this model, the presence of dunite channels affect melt generation and transport in the lherzolite matrix, suggesting that mantle heterogeneity may be critical to the physical aspects of melting and melt transport in the mantle beneath mid-ocean ridges. U-series disequilibria provide information on how melting occurs in the two endmember sources and suggest that melt porosities in the lherzolite may be as low as 0.1%. Melt within lherzolite maintains equilibrium with the coexisting solid while it ascends porously. Primitive MORB with high Mg# consistently have low 230 Th excesses or deficits with major element chemical signatures of equilibration near 1.0 GPa suggesting that the depleted endmember melt maintains chemical equilibrium with lherzolite until shallow mantle depths (∼30 km). Melt porosities in enriched heterogeneities remain below 1% for perhaps 10s of km before losing chemical equilibrium with the solid during transport in the upper melting column. Because the porosities required by the observed disequilibria are small, the transition to porosities large enough to form “veins” of melt must occur over a timescale which is very long in comparison to the 226 Ra half-life and significantly long for 231 Pa . Thus, instantaneous transport dynamic melting models appear incompatible with the observed disequilibria even when initial melt productivities as low as 0.05%/km are used.