R. A. Schmitt

Cerium: A chemical tracer for paleo-oceanic redox conditions☆

Abstract The Ce anomalies found in marine carbonates provide a sensitive indicator for the study of paleo-oceanic redox conditions. The dependency of CCe (the observed Ce concentration) and αCePO04 (the activity of the CePO4 complex) to the PO2, pH and PCO2 of seawater can be used for deciphering some aspects of the paleo-ocean chemistry, and is: log C Ce

Chemical evidence for the genesis of the ureilites, the achondrite Chassigny and the nakhlites

log α CePO 0 4 = 13.2 − 0.25 log P O 2 − 3.0 pH = −1.95 − 0.25 log P O 2 + 1.50 log P CO 2 + 1.50 log [Ca ++ ] The pH, which is related to PCO2, and PO2 are major factors for controlling the Ce concentration in seawater. Therefore Ce anomalies, CeAs (Ce observed/Ce interpolated between La and Pr or Nd) serve as indicators for pH (or PCO2) Or PO2 changes. Nearly uniform Ce abundances found in Pacific ~ 5000 m water columns ( Piepgras and Jacobsen , 1988) and the CeAs in ichthyoliths can be explained using this model. A general CeA shift in sediments from D.S.D.P. Holes 525A, 530A & B, 516, 516F suggest that the entire South Atlantic water redox condition changed from reducing to oxidizing at 56 My. In the North Atlantic, the Galicia Margin samples show a redox change between 132–148 My, indicating active surface currents between the Atlantic ocean and the shallow sea off the Iberian Peninsula. The carbonate samples from D.S.D.P. Hole 316 in the Central Pacific yielded an average CeA of 0.15 ± 0.01, which is 2−3× the CeAs observed in deep Pacific seawater by Piepgras and Jacobsen (1988). The relative enrichment of Ce in these samples is ascribed to coprecipitation of Ce(OH)4 onto Fe-Mn-O coatings. The striking similarities in the trace element profiles of REE + Sc + Hf + Th and CeAs and the low U abundances (~0.08 ppm) in many marine carbonates indicate insignificant Ce redox and trace elemental concentration changes for the Pacific ocean over the past 95 My: no significant Ce redox change occurred in the Pacific ocean within ± K T boundary. One 240 My China limestone, deposited under oxic sedimentary conditions, yielded a CeA of 0.53. Calculations indicate a PCO2 of ~1.9× present value prevailed over the ~240 My Tethyan Ocean. U Th ratios and low U concentrations are diagnostic indicators for obtaining reliable CeAs in marine carbonates, and perhaps for biogenic apatites and conodonts as well.

Elemental abundances in chondrules from unequilibrated chondrites: Evidence for chondrule origin by melting of pre-existing materials

Abstract Major element and REE, Cr, Sc, V, Ni, Co, Ir, Au, Sr, Ba abundances were determined in three ureilites and the unique achondrite, Chassigny. Chondritic-normalized REE abundance patterns for the ureilites are v-shaped, similar to pallasites, indicating a possible deep-seated origin. The lithophile trace element abundances and v-shaped REE patterns of the ureilites are consistent with a two-stage formation process, the first of which is an extensive partial melting of chondrite-like matter to yield ureilite precursors in the residual solid, which is enriched in Lu relative to La. The second step consists of an admixture of small and variable amounts of material enriched in the light REE. Such contaminating material may be magmas derived from the first formed melt of partial melting of chondrite-like matter. In contrast to the ureilites, Chassigny has a chondritic-normalized REE pattern which decreases smoothly from La(1.8 × ) to Lu(0.4 × ) and is parallel to and ~0.25 × the REE pattern in the nakhlitic achondrites. The composition of the magma from which Chassigny crystallized was highly enriched in the light REE; e.g. chondritic normalized La/Lu ~ 7. The similarity in the fractionated REE patterns (no Eu anomalies) for the olivine-pyroxene Chassigny and for the nakhlites suggests a genetic relationship. Siderophile trace element relationships in ureilites can be interpreted by three components: (1) ultramafic silicates enriched in Co relative to Ni, (2) an indigenous metal phase remaining after the partial melting event, and (3) a component of the carbon-rich vein material added after the partial melting.

Correlation of volcanic ash deposits by activation analysis of glass separates

Abstract Bulk abundances of Na, Mg, Al, Ca, Sc, V, Cr, Mn, Fe, Co, Ni, Zn, La, Sm, Eu, Yb, Lu, Ir, and Au were determined by neutron activation analysis of chondrules separated from unequilibrated H-, L-, and LL-chondrites (Tieschitz, Hallingeberg, Chainpur, Semarkona) and correlated with chondrule petrographic properties. Despite wellknown compositional differences among the whole-rock chondrites, the geometric mean compositions of their respective chondrule suites are nearly indistinguishable from each other for many elements. Relative to the condensible bulk solar system (approximated by the Cl chondrite Orgueil), chondrules are enriched in lithophile and depleted in siderophile elements in a pattern consistent with chondrule formation by melting of pre-existing materials, preceded or attended by silicate/metal fractionation. Relative to nonporphyritic chondrules, porphyritic chondrules are enriched in refractory and siderophile elements, suggesting that these two chondrule groups may have formed from different precursor materials.

Petrogenetic modeling of Hawaiian tholeiitic basalts: A geochemical approach

Abstract Volcanic ash deposits whose source is the Cascade Mountains area were correlated on the basis of 19 elemental abundances obtained by instrumental neutron activation analysis (INAA). After activation of glassy separates in a TRIGA reactor, gammaray spectra were obtained and analyzed with computer programs. The elements Na, Sm, Sc, Fe, Ce, Hf, and Th were determined with relative standard deviations less than 5%; the precision for La, Co, Eu, Yb, Cs, Ba, and Lu was less than 17%; larger errors were obtained for Rb, Ta, Nd, Tb, and Cr. A statistical method was developed for correlation on the basis of relative elemental compositions unique to the ash deposits. Elemental abundances of Mazama glassy separates were independent of distance from the source. The site to site chemical variability of crystal rich Glacier Peak and St. Helens ash layers was greater than for Mazama and Newberry ashes. The Rb, Yb, Lu, Th, and Ta contents in Newberry glass were more than twice those in Mazama glass. The concentrations of trace elements in Glacier Peak and St. Helens ashes generally were less than one-half those in Mazama glass. The presence of Mazama ash has been confirmed at sites in Oregon, Washington, Alberta, and in sediments of the Pacific Ocean.

Rare earth element geochemistry of South Atlantic deep sea sediments: Ce anomaly change at ~54 My

Abstract The abundances of 29 elements, including nine REE (La, Ce, Nd, Sm, Eu, Tb, Dy, Yb and Lu), have been determined by instrumental neutron activation analysis in 33 tholeiitic basalts from Mauna Loa, Kohala, Mauna Kea, Lanai and Koolau. These data have been combined with data provided by A. V. Murali et al. on seven basalts from Kilauea for geochemical evaluation. Partial melting models based on the partitioning of the REE and Sc suggest that these basalts can be produced from three distinct source compositions. Thus, the basalts from Mauna Kea, Kohala and Kilauea are generated by 2–10% partial melting of similar source materials in the compositional range 82 ± 4% olivine plus orthopyroxene (ol + opx), 14 ± 3% clinopyroxene (cpx) and 4 ± 1% garnet (gar). Similar amounts of melting of source materials having 74 ± 6% ol + opx, 21 ± 5% cpx and 5 ± 1% gar produce the basalts of Mauna Loa and Lanai. A source material composed of 86 ± 4% ol + opx, 11 ± 2% cpx and 3 ± 1% gar is proposed for the generation of basalts from Koolau. The olivine-crystallization model suggested here requires absolute REE abundances in the preferred Mauna Kea, Kohala and Kilauea source to range from 1.1× (times chondrite) for the LREE (La) to about 1.3× for the HREE; in the Mauna Loa and Lanai source, La is about 1.0× and the HREE are about 1.6×; and in the Koolau source, La is ~0.7× and the HREE are ~0.9×.

Trace element partitioning between volcanic plagioclase and dacitic pyroclastic matrix

The geochemistry of the REE (rare earth elements) in oceanic sediments is discussed, based mainly on samples from DSDP Holes 530A and 530B, Leg 75, and Hole 525A, Leg 74. The proposed mechanisms for incorporation of the REE into the marine carbonate phases are adsorption, chiefly onto the carbonate minerals and on Sc, Hf, and Ta-rich Fe-Mn hydroxide flocs as carbonate coatings. The Ce anomaly of marine carbonate was used as an indicator of paleo-ocean water redox conditions: the bottom water of the Angola Basin was in a reducing condition in the Cretaceous. At ca. 54 My, the South Atlantic water condition became oxidizing, similar to the present seawater redox condition. This change was related to the improvement of circulation due to the widening of South Atlantic and the subsidence of water circulation barriers such as the Walvis Ridge and perhaps the Romanche Fracture Zone. The younger (Eocene-Recent) and older (Albian-Santonian) argillaceous sedimentary rocks from 530A (denoted as YSAB and OSAB respectively) show different degrees of Eu depletion with a transition period in between. The REE patterns of OSAB suggest a basaltic origin. The possible sources are Kaoko basalt in Southwest Africa or Namibia and the basaltic Walvis Ridge itself. The decrease in the area covered by Kaoko basalt due to erosion, the subsidence of the Walvis Ridge, and the improvement of water circulation led to changes in the Eu anomaly from Campanian to Paleocene, and resulted in the YSAB REE pattern. Changes in the Sm/Eu, La/Th, Th/Yb, Ti/Al2O3, FeO/Al2O3, and Hf/Al2O3 ratios suggest changes of average source rock composition from and esite to granodiorite. The REE abundances and patterns of younger sediments in the Angola Basin (YSAB) are very similar to those observed in NASC, PAAS, and ES sediments. The YSAB REE abundances and patterns may represent the average REE distribution of the exposed African continental crust. The strong resemblance of REE distributions of YSAB, NASC, PAAS and ES suggests thorough REE mixing from different sources and the uniformity of the average crustal compositions of different continents: Africa, North America, Australia, and Europe

Alteration of basaltic glasses from north-central British Columbia, Canada

Abstract Abundances of several trace elements in plagioclase and its host pyroclastic matrix have been determined by neutron activation analysis and atomic absorption spectroscopy. Partition coefficients of the rare earth elements and Sr, Rb, Ba, and K show a regular pattern for zoned plagioclase and seem to be controlled by the crystallo-chemistry of the volcanic phenocrysts.

Apollo 15 yellow-brown volcanic glass: Chemistry and petrogenetic relations to green volcanic glass and olivine-normative mare basalts

Abstract Evidence of palagonitization (glass replacement by poorly crystallized, clay-like material) is seen on all glasses studied from three Pleistocene subglacial volcanoes in north-central British Columbia, Canada. Samples from foreset breccias of Tuya Butte are more highly palagonitized than those from the tephra cones of Ash Mountain and Southern Tuya. Extensive palagonitization is generally associated with authigenic mineralization (clays, zeolites). Palagonite composition varies widely relative to glass composition, and palagonite can be broadly categorized as either high-Al or low-Al, depending on whether Al was retained or lost to aqueous solutions during palagonitization. The lowest Al palagonites are those from the foreset breccias of Tuya Butte. Loss of Al during palagonitization is related to closed-system alteration, including precipitation of aluminosilicate authigenic cements. Low initial pH is suggested for Al depletion, consistent with the behavior of Ni, Co, and Cr, which are retained in high-Al and depleted in low-Al palagonite. Microenvironment appears to be more influential than macroenvironment in determining the composition of palagonite. Palagonite rinds are compositionally zoned, generally becoming progressively higher in Al and Ca, and lower in Fe and Mg, towards the innermost (later-formed) portions of the rinds. This apparently reflects changing solution composition (increasing pH) with time. Compositional zoning does not change the overall stoichiometry of palagonite which resembles smectite clay. Mineral paragenesis is related to the Ca content of the palagonite, with partial replacement of palagonite by smectite (Fe-saponite) occurring when Ca is retained in the rind. This replacement phenomenon occurs prior to zeolitization. No such replacement clay occurs with low-Ca palagonite, but a late-stage nontronite film overgrows zeolite. Phillipsite is the first zeolite formed, followed by chabazite. Analcime and calcite occur in the most highly palagonitized samples. Mass balance considerations indicate higher mass loss where palagonitization has not proceeded to the point where zeolite solubility limits were attained in the local solution. Zeolites occur in closed-system conditions (low flow rates), where little net system mass loss or gain has occurred. The colloidal nature of palagonite allows the effective adsorption of Rb, Cs, Sr, Ba, and REEs.

Radiochemical neutron activation analysis of Indium, Cadmium, Yttrium and the 14 rare earth elements in rocks

Abstract Apollo 15 yellow-brown glass is one of twenty-five, high-Mg, primary magmas emplaced on the lunar surface in pyroclastic eruptions. Forty spherules of this glass were individually analyzed by electron microprobe and INAA for major- and trace-elements ( e.g. Sc, V, Cr, Co, La, Ce, Nd, Sm, Eu, Ho, Tm, Yb, Lu, Hf, Ta, and U). The abundances demonstrate that this primary magma was produced by partial melting of differentiated cumulates in the lunar mantle. Models are developed to explain the possible source-regions of several Apollo 15 and Apollo 12 low-Ti mare magmas as being products of hybridization involving three ancient differentiated components of a primordial lunar magma ocean: (a) early olivine ± orthopyroxene cumulates; (b) late-stage clinopyroxene + pigeonite + ilmenite + plagioclase cumulates; and (c) late-stage inter-cumulus liquid. Isotopic constraints on Sm-Nd fractionation and least-squares tests of potential mixing models require models of Apollo 15 yellow-brown glass, and Apollo 12 and 15 olivine mare basalts, involving low F ~3% cumulate remelting of sources in which olivine and orthopyroxene are residual phases. Models of the Apollo 15 green glass composition indicate higher (~4–7%) degrees of melting, but with only a slight preference for orthopyroxene-rich relative to orthopyroxene -free sources. The modeled dependence on residual orthopyroxene in mare source regions places source depths at ~400 km or greater in order to maintain olivine and orthopyroxene on the liquidus. Hybridization at such depths would likely result from convective overturn of the primordial magma ocean and suggests that separate fractionating masses were positioned at various depths. Our models for developing hybridized source regions of primary low-Ti mare magmas do not require selective contamination of primary liquids. The preferred model for Apollo 15 yellow-brown glass is presented as the commingling of 95% early cumulate olivine + orthopyroxene, 1.9% clinopyroxene, 0.27% pigeonite, 1.7% plagioclase, 0.02% ilmenite, and 0.56% trapped liquid. Respective proportions for the Apollo 15 olivine mare basalt source model are 98.5%, 0.37%, 0%, 0.79%, 0.16%, and 0.13%; and the proportions used for the preferred Apollo 12 olivine mare basalt model are 97.6%, 0.73%, 0.24%, 0.77%, 0.42%, and 0.21%. Our preferred model for the green glass source composition ( F = 0.07) requires the commingling of 97.4% early olivine + orthopyroxene, 0.18% clinopyroxene, 1.32% pigeonite, 0.85% plagioclase, and 0.28% trapped liquid.