G. J. Wasserburg

Sm-Nd isotopic evolution of chondrites

The ^(143)Nd/^(144)Nd and ^(147)Sm/^(144)Nd ratios have been measured in five chondrites and the Juvinas achondrite. The range in ^(143)Nd/^(144)Nd for the analyzed meteorite samples is 5.3 e-units (0.511673–0.511944) normalized to ^(150)Nd/^(142)Nd= 0.2096. This is correlated with the variation of 4.2% in ^(147)Sm/^(144)Nd (0.1920–0.2000). Much of this spread is due to small-scale heterogeneities in the chondrites and does not appear to reflect the large-scale volumetric averages. It is shown that none of the samples deviate more than 0.5 e-units from a 4.6-AE reference isochron and define an initial ^(143)Nd/^(144)Nd ratio at 4.6 AE of 0.505828 ± 9. Insofar as there is a range of values of ^(147)Sm/^(144)Nd there is no unique way of picking solar or average chondritic values. From these data we have selected a new set of self-consistent present-day reference values for CHUR (“chondritic uniform reservoir”) of (^(143)Nd/^(144)Nd)_(CHUR)^0 = 0.511836 and (^(147)Sm/^(144)Nd)_(CHUR)^0 = 0.1967. The new ^(147)Sm/^(144)Nd value is 1.6% higher than the previous value assigned to CHUR using the Juvinas data of Lugmair. This will cause a small but significant change in the CHUR evolution curve. Some terrestrial samples of Archean age show clear deviations from the new CHUR curve. If the CHUR curve is representative of undifferentiated mantle then it demonstrates that depleted sources were also tapped early in the Archean. Such a depleted layer may represent the early evolution of the source of present-day mid-ocean ridge basalts. There exists a variety of discrepancies with most earlier meteorite data which includes determination of all Nd isotopes and Sm/Nd ratios. These discrepancies require clarification in order to permit reliable interlaboratory comparisons. The new CHUR curve implies substantial changes in model ages for lunar rocks and thus also in the interpretation of early lunar chronology.

238U234U230Th232Th systematics and the precise measurement of time over the past 500,000 years

We have developed techniques to measure the ^(230)Th abundance in corals by isotope dilution mass spectrometry. This, coupled with our previous development of mass spectrometric techniques for ^(234)U and ^(232)Th measurement, has allowed us to reduce significantly the analytical errors in ^(238)U-^(234)-^(230)Th dating and greatly reduce the sample size. We show that 6 × 10^8 atoms of ^(230)Th can be measured to ±30‰ (2σ) and 2 × 10^(10) atoms of ^(230)Th to ± 2‰. The time over which useful age data on corals can be obtained ranges from a few years to ∼ 500 ky. The uncertainty in age, based on analytical errors, is ± 5 y (2σ) for a 180 year old coral (3 g), ± 44 y at 8294 years and ± 1.1 ky at 123.1 ky (250 mg of coral). We also report ^(232)Th concentrations in corals (0.083–1.57 pmol/g) that are more than two orders of magnitude lower than previous values. Ages with high analytical precision were determined for several corals that grew during high sea level stands ∼ 120 ky ago. These ages lie specifically within or slightly postdate the Milankovitch insolation high at 128 ky and support the idea that the dominant cause of Pleistocene climate change is Milankovitch forcing.

238U,234U and232Th in seawater

We have developed techniques to determine ^(238)U,^(234)U and ^(232)Th concentrations in seawater by isotope dilution mass spectrometry. U measurements are made using a ^(233)U-^(236)U double spike to correct for instrumental fractionation. Measurements on uranium standards demonstrate that ^(234)U/^(238)U ratios can be measured accurately and reproducibly. ^(234)U/^(238)U can be measured routinely to ± 5‰ (2σ) for a sample of 5 × 10^9 atoms of ^(234)U (3 × 10^(−8) g of total U, 10 ml of seawater). Data acquisition time is ∼ 1 hour. The small sample size, high precision and short data acquisition time are superior toα-counting techniques. ^(238)U is measured to ± 2‰ (2σ) for a sample of 8 × 10^(12) atoms of ^(238)U (∼ 3 × 10^(−9) g of U, 1 ml of seawater). ^(232)Th is measured to ± 20‰ with 3 × 10^(11) ^(232)Th atoms (10^(−10) g ^(232)Th, 1 1 of seawater). This small sample size will greatly facilitate investigation of the ^(232)Th concentration in the oceans. Using these techniques, we have measured ^(238)U, ^(234)U and ^(232)Th in vertical profiles of unfiltered, acidified seawater from the Atlantic and ^(238)U and ^(234)U in vertical profiles from the Pacific. Determinations of ^(234)U/^(238)U at depths ranging from 0 to 4900 m in the Atlantic (7°44′N, 40°43′W) and the Pacific (14°41′N, 160°01′W) Oceans are the same within experimental error (± 5‰,2σ). The average of these ^(234)U/^(238)U measurements is 144 ± 2‰ (2σ) higher than the equilibrium ratio of 5.472 × 10−5. U concentrations, normalized to 35‰ salinity, range from 3.162 to 3.281 ng/g, a range of 3.8%. The average concentration of the Pacific samples (31°4′N, 159°1′W) is ∼ 1% higher than that of the Atlantic (7°44′N, 40°43′W and 31°49′N, 64°6′W). ^(232)Th concentrations from an Atlantic profile range from 0.092 to 0.145 pg/g. The observed constancy of the ^(234)U/^(238)U ratio is consistent with the predicted range of ^(234)U/^(238)U using a simple two-box model and the residence time of deep water in the ocean determined from ^(14)C. The variation in salinity-normalized U concentrations suggests that U may be much more reactive in the marine environment than previously thought.

Precise determination of SmNd ratios, Sm and Nd isotopic abundances in standard solutions☆

The methods used for precise calibrations of Sm/Nd ratios and the average isotopic abundances obtained for normal Sm and Nd are given. A mixed Sm-Nd normal solution with a precisely known ^(147)Sm/^(144)Nd ratio close to the nominal average chondritic value is described and the calibration discussed. Aliquots of this standard solution are available on request and may be useful for precise interlaboratory calibration of Sm and Nd.

U-Th-Pb systematics in three Apollo 14 basalts and the problem of initial Pb in lunar rocks

The isotopic composition of Pb and the elemental concentration of U, Th and Pb were measured on ‘total’ rock samples 14053, 14073 and 14310 and on mineral separates of 14310 and 14053. Sample #73 appears to be quite similar to #310. Sample #310 yields total rockmodel ages ofT(206Pb/238U) = 4.24AE,T(207Pb/235U) = 4.27AE, andT(208Pb/232Th) = 4.13AE. These arenearly concordant and distinct from the Rb-Sr and K-Ar crystallization ages of 3.88 AE. Mineral separates from 14310 show a wide spread in207Pb/206Pb ranging from 0.483 to 0.995. The data points define a reasonable linear array on the coupled Pb-U evolution diagram. Similar analyses of 14053 give high, discordant total rockmodel ages ofT(206Pb/238U) = 5.60AE,T(207Pb/235U) = 5.18AE, andT(208Pb/232Th) = 5.48AE. Mineral separates show a range of207Pb/206Pb from 0.716 to 1.209. These data also define a reasonable linear array on the coupled Pb-U evolution diagram. These are the first Pb-U isochrons obtained for lunar basalts and indicate a reasonable solution to the previous discrepancy between the different methods of ‘absolute’ age determination. The resulting U-Pb isochron ages are compatible with the Rb-Sr and K-Ar ages on the same rocks. However, it is not possible to establish aprecise time of ‘crystallization’ from the Pb-U data because of the small angle of intersection between the linear arrays and the concordia curve. These data show that total rock model ages do not in general yield crystallization ages. The data on #310 and #053 show that these rocks were formed containing a highly radiogenic initial lead (207Pb/206Pb)I0 ≈ 1.46 which accounts for the excessively high total rock model ages by the U-Th-Pb method. The only significant discrepancy in the data is the apparent variability of (208Pb/206Pb)I0in #053 which remains to be resolved. The (207Pb/206Pb)I0in these rocks corresponds to the radiogenic lead evolved between 4.51 and 3.88 AE in a U-rich environment. Such data from initial Pb may provide a new chronometer for early lunar evolution. The high207Pb/206Pb ages in some total lunar soils as well as in treated fractions may be partly explained as a consequence of the contribution of lunar basalts with radiogenic initial Pb. The data prove that at the time of extrusion of some basalts, unsupported lead with extremely high207Pb/206Pb ratios was added to the lunar surface.

Sm-Nd and Rb-Sr Chronology of Continental Crust Formation

Samarium-neodymium and rubidium-strontium isotopic systematics together with plausible assumptions regarding the geochemical evlution of continental crust material, have been used to ascertain the times at which segments of continental crust were formed. Analyses of composites from the Canadian Shield representing portions of the Superior, Slave, and Churchill structural provinces indicate that these provinces were all formed within the period 2.5 to 2.7 aeons. It has been possible to determine the mean age of sediment provenances, as studies of sedimentary rocks suggest that the samarium-neodymium isotopic system is not substantially disturbed during sedimentation or diagenesis.

Negative thermal ion mass spectrometry of osmium, rhenium and iridium

We report on a technique for obtaining intense ion beams of negatively charged oxides of Os, Re and Ir by thermal ionization, in a conventional surface ionization mass spectrometer. It was found that the principal ion species of Os, Re and Ir produced are OsO_3^−, ReO_4^− and IrO_2^−. The sharp distinction in the masses of the dominant molecular species produced by this technique permits the measurement of isotopic compositions of each element from mixtures of platinum-group elements without significant isobaric interferences. For ^(187)Re-^(187)Os isotope studies, this technique offers the advantage of isotopic analyses without prior chemical separation of Re from Os, as no isobaric interference between the oxides of ^(187)Os and ^(187)Re exists under these conditions. For 4 ng Os, stable ion currents of 3 × 10^(−12) A can be maintained for over one hour, which allows determination of isotopic ratios with a Faraday collector to a precision of better than ±2‰ (2 σ_m. For 70 pg Os, isotopic ratios can be measured with a precision of better than ±5‰ using a secondary electron multiplier. The detection limit for Os is estimated to be below 10^(−14) g. Osmium isotopic ratios have also been determined by direct loading of natural iridosmine with a precision of ±0.5‰ or better. We have obtained ionization efficiencies of 2–6% for Os and >20% for Re; these are superior to those reported for other techniques available to date and demonstrate that negative thermal ion mass spectrometry will have widespread application to ^(187)Re-^(187)Os chronometry and to studies of the geochemistry and environmental chemistry of the platinum-group elements.

The Eastern Mediterranean paleoclimate as a reflection of regional events: Soreq cave, Israel

The climate of the Eastern Mediterranean region of the last 60 ky was determined by a high resolution study of the oxygen and carbon isotopic composition (1500 measurement pairs) of speleothems from the Soreq cave, Israel, with chronology provided by 53 precise ^(230)Th–^(234)U (TIMS) ages. The high precision of the speleothem TIMS ages permits us to determine the timing of regional climatic events in the Eastern Mediterranean region and to see if they correlate with global events. During the period from 60 to 17 ky, the δ^(18)O and δ^(13)C values were generally 2–2.5‰ higher than during the period from 17 ky to present. This is consistent with the climatic transition from glacial to interglacial. Within the 60 to 17 ky period, the Soreq cave stable isotope profile includes four cold peaks (at 46, 35, 25 and 19 ky) and 2 warm peaks (at 54 and 36 ky). In addition, the period <17 ky has two more cold peaks at 16.5 and from 13.2 to 11.4 ky. The ages of four of the six cold peaks correlate well with the ages of three Heinrich events (H1, H2, H5) and with the age of the Younger Dryas. However, the other two Heinrich events are not reflected in the Soreq cave record. Several other isotope peaks which appear during the last 7 ky are contemporaneous with regional climatic events in the Middle East and North Africa. In addition to the drop in δ^(18)O and δ^(13)C observed between the last glacial and the Holocene, sharp simultaneous drops in (^(234)U/^(238)U)_0 ratios, Sr concentrations and in ^(87)Sr/^(86)Sr are also observed, suggesting that the latter are climate related. These variations are interpreted in terms of major changes in the temperature, the mean annual rainfall and its isotopic composition, the isotopic composition of the Mediterranean vapor source, the soil moisture conditions, and in the mixing proportions of sources with different ^(87)Sr/^(86)Sr ratios (sea spray, dust particles and dolomitic host rock).

Isotopic evidence for a terminal lunar cataclysm

Most highland total rock samples define a single UPb isochron which corresponds to a metamorphism age of ∼ 3.9AE. This age is also obtained for internal UPb isochrons for some of these samples. The data on 18 rock samples range from concordant samples with238U/206Pb∼ 1.2 to discordant ones with238U/206Pb∼ 0.02. This feature coupled with a correlated pattern of 238U/204Pb ratios, indicates that Pb was extensively mobilized at ∼ 3.9AE. The observed PbU fractionation is essentially due to Pb volatilization during the metamorphic events. Volatile Pb transport is not accompanied by similar effects in Rb and must therefore be attributed to a specific process. RbSr internal isochrons for the same rocks determine distinct metamorphic events in the interval 3.85–4.00 AE. We conclude that highland samples from widely separated areas bear the imprint of an event or series of events in a narrow time interval which can be identified with a cataclysmic impacting rate of the moon at ∼ 3.9AE, although differentiation by internal magma generation cannot be excluded. This cataclysm is associated with the Imbrium impact and very possibly the formation of Crisium and Orientale and possibly several other major basins in a narrow time interval (∼2 × 108yr or less). The UPb data indicate formation of the lunar crust at ∼ 4.4AE, which is distinctly younger than 4.6 AE, the time generally associated with planetary formation. If the lunar crust started being formed at ∼ 4.6AE, then this process must have continued until times significantly younger than 4.4 AE. RbSr data also indicate formation of the lunar crust around 4.5 AE with only minor additions of high Rb/Sr materials to the crust at times younger than 4.3 AE. Using the UPb systematics, K/U and the average U concentration of the moon as obtained from heat-flow measurements, we estimate the lunar concentrations: primordial Pb= 35ppb;Rb= 0.5ppm with Rb/Sr= 0.006.

Sm-Nd isotopic evolution of chondrites and achondrites. II

The ^(147)Sm-^(143)Nd and ^(146)Sm-^(142)Nd isotope systematics have been investigated in five chondrites and the achondrites Moama and Angra dos Reis (ADOR). The new chondrite data and those we have reported before are all consistent with our previously reported reference values for CHUR (“chondritic uniform reservoir”) of (^(143)Nd/^(144)Nd)_(CHUR)^0 = 0.511847 and (^(147)Sm/^(144)Nd)_(CHUR)^0 = 0.1967. Most of the bulk chondrites analyzed have ^(143)Nd/^(144)Nd and ^(147)Sm/^(144)Nd within 0.5 e-units and 0.15% of the CHUR values, respectively. This strongly suggests that the CHUR evolution is now known to within these error limits throughout the history of the solar system. The St. Severin chondrite yields an Sm-Nd internal isochron age of T = 4.55 ± 0.33AE and an initial e_(Nd) = 0.11 ∓ 0.26. Much larger variations in Sm/Nd ratios were measured in mineral separates of the Moama and ADOR achondrites. Thus, very precise ages of 4.46 ± 0.03 AE and 4.564 ± 0.037 AE were obtained for these meteorites, respectively. The initial e_(Nd) values obtained for Moama and ADOR are 0.03 ∓ 0.25 and 0.14 ∓ 0.20, respectively. The values obtained on these meteorites are fully consistent with the CHUR evolution curve. Initial e_(Nd) data on terrestrial igneous and meta-igneous rocks demonstrates that positive initial e_(Nd) values occur throughout the past 4 AE. This confirms our earlier report that a light rare earth element-depleted layer has existed throughout most of the Earth history and is the source of present-day mid-ocean ridge basalts. The inferred shape of the e_(Nd) vs. age curve for the depleted mantle suggests profound changes in tectonic regimes with time; in particular, it suggests a much higher rate of recycling of continental materials into the mantle during the Archean as compared to later time periods. ^(146)Sm-^(142)Nd systematics of ADOR and Moama are supportive of the hypothesis that ^(146)Sm was present in the early solar system and suggests a ^(146)Sm/^(144)Sm ratio of about 0.01 for the solar system ∼ 4.56 AE ago. This inferred high ^(146)Sm abundance cannot be explained as a late injection from a supernova and must be due to galactic nucleo-synthesis.