E.B. Watson

Rutile saturation in magmas: implications for TiNbTa depletion in island-arc basalts

Abstract The TiO 2 contents of rutile-saturated melts ranging from basalt to rhyodacite have been investigated at P = 8–30 kbar and T = 1000–1300°C under hydrous, CO 2 -saturated, and volatile-absent conditions. Dissolved TiO 2 is positively correlated with T and not strongly dependent on P total . For fixed P and T , TiO 2 content decreases markedly as the melts become more felsic. The distribution of TiO 2 between rutile and liquid, expressed as a wt.% concentration ratio, D (rut/liq), is given by: In D = −3.16 + (9373T) + 0.026P − 0.152FM where T is in Kelvins, P in kbar and FM is a melt composition parameter, FM = [Na+K+ 2(Ca+Fe+Mg)]/Al· 1/Si in which the chemical symbols represent cation fractions. The first term expresses the competition of aluminate and titanate anions for charge-compensating cations, and the second term expresses the inverse dependence of dissolved TiO 2 on SiO 2 content. There is no apparent dependence of rutile solubility on water content. For ranges of probable solidus conditions, rutile saturation in basaltic, andesitic, and dacitic liquids requires 7–9, 5–7, and 1–3 wt.% TiO 2 , respectively. These concentrations are well in excess of those found in the respective rock types, so depletion in Nb, Ta, and Ti and reduced Nb/U and Nb/Th ratios in volcanic rocks erupted at convergent plate margins cannot be attributed to residual rutile in their source regions. Thus, Nb, Ta and Ti depletion must be an inherent property of the source region. We suggest that the island-arc source region has been depleted in Nb and Ta by a previous episode of melt extraction (MORB), zoning refining, or equilibration with a percolating melt or fluid. Such a process markedly depletes the LILE and HFSE element concentrations in the residuum, but ratios such as Nb/U, Nb/Th and U/Th remain relatively constant due to similar solid-melt partition coefficients. The depletion of Nb relative to Th in the source regions of island-arc magmas occurs during hybridization of the source by rutile-saturated (Nb/Ta-depleted) melts or aqueous fluids. If the hybridizing agent is a melt, a relatively felsic composition, produced under low T (900°) hydrous conditions, is required.

Rare-earth diffusion in zircon

Diffusion rates for three rare-earth elements (REEs: Sm, Dy, Yb) have been measured in synthetic and natural zircon. REE-phosphate powders were used as the source of diffusant, with Rutherford backscattering spectrometry (RBS) used to measure REE depth profiles. Over the temperature range 1150–1400°C, the following Arrhenius relations were obtained (diffusion coefficients in m2 s−1 ): log DYb = (7.40±1.15)+(−769±34 kJ mol−12.303 RT) log DDy = (5.36±0.21)+(−734±35 kJ mol−12.303 RT) log DSm = (8.46±1.61)+(−841±57 kJ mol−12.303 RT) Results for synthetic: and natural zircons were quite similar, and no evidence of significant anisotropy was observed when comparing transport normal and parallel to the c-axis. The data show a systematic increase in diffusivity with decreasing ionic radius (i.e. faster diffusion rates for the heavier REEs). Given these trends the diffusion rates of Lu and La should differ by over two orders of magnitude. Diffusive fractionation is unlikely in the Sm-Nd system because differences in diffusivities are relatively small, but may be a factor in the Lu-Hf system given the much slower diffusion rates of tetravalent cations. The very slow diffusion rates measured for the REEs suggest that they are essentially immobile under most geologic conditions, thus permitting the preservation of fine-scale chemical zoning and isotopic signatures of inherited cores.

OXYGEN DIFFUSION IN ZIRCON

Oxygen diffusion in natural, non-metamict zircon was characterized under both dry and water-present conditions at temperatures ranging from 765°C to 1500°C. Dry experiments were performed at atmospheric pressure by encapsulating polished zircon samples with a fine powder of18O-enriched quartz and annealing the sealed capsules in air. Hydrothermal runs were conducted in cold-seal pressure vessels (7–70 MPa) or a piston cylinder apparatus (400–1000 MPa) on zircon samples encapsulated with both18O-enriched quartz and18O water. Diffusive-uptake profiles of18O were measured in all samples with a particle accelerator, using the18O(p, α)15N reaction. For dry experimental conditions at 1100–1500°C, the resulting oxygen diffusivities (24 in all) are well described by:Ddry (m2/s) = 1.33 × 10 −4exp(−53920/T) There is no suggestion of diffusive anisotropy. Under wet conditions at 925°C, oxygen diffusion shows little or no dependence uponPH2O in the range 7–1000 MPa, and is insensitive to total pressure as well. The results of 27 wet experiments at 767–1160°C and 7–1000 MPa can be described a single Arrhenius relationship:Dwet (m2/s) = 5.5 × 10−12exp(−25280/T) The insensitivity of oxygen diffusion toPH2O means that applications to geologic problems can be pursued knowing only whether the system of interest was ‘wet’ (i.e.,PH2O > 7MPa) or ‘dry’. Under dry conditions (presumably rare in the crust), zircons are extremely retentive of their oxygen isotopic signatures, to the extent that δ18O would be perturbed at the center of a 200 μm zircon only during an extraordinarily hot and protracted event (e.g., 65 Ma at 900°C). Under wet conditions, δ18O may or may not be retained in the central regions of individual crystals, cores or overgrowth rims, depending upon the specific thermal history of the system.

Pre-eruption recharge of the Bishop magma system

The 650 km 3 rhyolitic Bishop Tuff (eastern California, USA), which is stratigraphically zoned with respect to temperatures of mineral equilibration, refl ects a corresponding thermal gradient in the source magma chamber. Consistent with previous work, application of the new TitaniQ (Ti-in-quartz) thermometer to quartz phenocryst rims documents an ~100 °C temperature increase with chamber depth at the time of eruption. Application of TitaniQ to quartz phenocryst cores, however, reveals lower temperatures and an earlier gradient that was less steep, with temperature increasing with depth by only ~30 °C. In many late-erupted crystals, sharp boundaries that separate low-temperature cores from high-temperature rims cut internal cathodoluminescent growth zoning, indicating partial phenocryst dissolution prior to crystallization of the high-temperature rims. Rimward jumps in Ti concentration across these boundaries are too abrupt (e.g., 40 ppm across a distance of <10 µm) to have survived magmatic temperatures for more than ~100 yr. We interpret these observations to indicate heating-induced partial dissolution of quartz, followed by growth of high-temperature rims (made possible by lowering of water activity due to addition of CO 2 ) within 100 yr of the climactic 760 ka eruption. Hot mafi c melts injected into deeper parts of the magma system were the likely source of heat and CO 2 , raising the possibility that eruption and caldera collapse owe their origin to a recharge event.

A study of strontium diffusion in K-feldspar, Na-K feldspar and anorthite using Rutherford Backscattering Spectroscopy

Abstract Sr chemical diffusion has been measured in orthoclase, anorthoclase and anorthite under dry, 1-atm conditions. A strontium oxide-aluminosilicate powder mixture was used as the source material, with Rutherford Backscattering Spectroscopy (RBS) used to measure diffusion profiles. Over the temperature range 725–1075°C the following Arrhenius relations were obtained. Orthoclase (normal to (001)):D=5.97 × 10+0.0060−0.0030exp−67900 ± 1600RTcm 2 s −1 Measurements of diffusion perpendicular to (010) and (100) indicate little diffusional anisotropy in orthoclase. Anorthoclase: normal to (010): D=4.51 × 10 +1 +358−5.6exp−89100 ± 4800RT cm 2 s −1 normal to (001): D=2.25 × 10 +2 +1615−31exp − 89300 ± 4600RT cm 2 s −1 Anorthite (normal to (010)): D=3.85 × 10 −2 +0.380−0.0039 exp − 78800 ± 5400RT cm 2 s −1 The correlation of Sr uptake with a reduction of K and Ca in orthoclase and anorthite, respectively, indicates that Sr is exchanging with these species. Nuclear reaction analysis (NRA) measurements of Al in orthoclase suggests that the coupled exchange Sr 2+ + Al 3+ → K 1+ + Si 4+ is taking place, as expected. Sr closure temperatures for feldspars, calculated with the above parameters, are relatively high compared to those of some other minerals (e.g., biotite, muscovite, apatite). However, closure temperatures of feldspars may be significantly depressed if they are exsolved or twinned, leading to small effective diffusion lengths. Feldspar crystals of millimeter size should retain 87 /Sr 86 Sr ratios except when subjected to thermal events of extreme temperature or duration.

The incorporation of Pb into zircon

Abstract The incorporation of Pb into zircons grown from Pb-rich solutions was evaluated using three different approaches: (1) high-temperature growth of large crystals from Pb-silicate melts; (2) hydrothermal overplating of thin epitaxial layers on substrates of natural zircon; and (3) growth of small, homogeneously nucleated crystals from aqueous fluids. The melt-grown zircons (50–400 μm) were crystallized from PbOSiO2ZrO2 (±P2O5) liquid at atmospheric pressure by cooling from 1430° to 1350°C. In the P2O5 free system, despite 66 wt% PbO in the melt, these zircons contain 3 atom% Pb, with apparent zircon/fluid partition coefficients of 4.2 and 2.6, respectively, for Pb4+ and Pb2+. In contrast to the case of melt-grown zircons, available P is excluded from the aqueous epitaxial zircon, suggesting that charge balance is accomplished by H+ instead. Small (2–5 μm) zircons grown by cooling aqueous solutions (PbO + SiO2 + ZrO2 ± P2O5) from 800°C or 900°C contain ∼ 0.25–0.5 atom% Pb (∼ 2–4 wt% PbO), yielding apparent DPb values of ∼ 0.2–0.3. Available P5+ is incorporated in a 2:1 ratio with Pb2+, suggesting a specific charge-balance mechanism: [2P5+ + Pb2+] = [2Si4+ + Zr4+]. However, Pb enters the zircon even when P is unavailable, so H+ may again play a charge-balancing role. Because of the rapid, polythermal modes of zircon growth and the high Pb content of the experimental systems, the apparent partition coefficients should not be viewed as equilibrium values, but as qualitative indicators of Pb compatibility under various growth circumstances. The overall results are consistent with the low but variable levels of non-radiogenic (common) Pb in natural zircons. The increased compatibility of Pb in fluid-grown, low-temperature zircons suggests a possible fingerprint for zircons from hydrothermal and wet-metamorphic rocks, i.e., high concentrations of common Pb.

Adcumulus dunite growth in a laboratory thermal gradient

Laboratory experiments near 1450° C at 1 bar (QFM) on komatiite bulk composition show olivine and liquid in cumulus textures which evolve with experiment duration. Orthocumulus texture with settled olivine crystals separated by liquid matrix is developed within a day. Experiments quenched after a few days to a week show a progression of textures which include development of columns of olivine crystals separated by channels of liquid. Olivine grain sizes increase with the cube root of time suggesting that dissolution and reprecipitation of olivine may be involved in the organization into columns and channels. Experiments quenched after two weeks have well developed adcumulus texture. The basal polycrystalline granular olivine aggregate forms from the decay of the olivine columns. Melt expulsion from the aggregate can be virtually complete, leaving 1% or less of the melt originally present.Buoyancy-driven compaction of olivine is not the mechanism responsible for this textural evolution because the final basal aggregate sometimes contains vesicles. An addition proof of the inadequacy of buoyancy is provided by raising the crucible slightly above the thermal symmetry point of the furnace. The aggregate then compacts on top of a crystal-free liquid. The thermal gradients above and below the furnace hot spot are thought to be primarily responsible for the olivine redistributions observed. Diffusion of olivine components in the liquid is driven along a saturation gradient resulting from the temperature gradient. The process, called thermal migration in geological literature, is essentially the same as traveling solvent zone refining in metallurgy. Differential solubility and Soret fractionation both contribute to olivine redistribution to the cold region of the crystal-liquid aggregate. There may be some applications of these results to natural cumulate rock petrogenesis.

Lattice diffusion of Ar in quartz, with constraints on Ar solubility and evidence of nanopores

Abstract Diffusion and solubility of Ar in optically clear natural and synthetic quartz crystals were examined at ∼500 to 1200 °C by treating polished specimens in pressurized Ar (1–185 MPa) and characterizing the resulting diffusive-uptake (or subsequent diffusive-loss) profiles using Rutherford backscattering spectroscopy (RBS). Analytical uncertainty leads to significant scatter in the data, but the Ar diffusivity, D, is reasonably well constrained by the following Arrhenius relationship:D = 8.2−4.2+8.8 × 10−19 exp[(−6150 ± 750)/T(K)] m2/sNo effects of crystallographic orientation or quartz structural form (α or β) are discernible. Apparent solubilities typically fall between 1000 and 3000 ppm (by mass), with large uncertainties (±50–60% 2σ), but some lower values (∼700 ppm) are observed near the low end of the Ar pressure range investigated. Occasional high-concentration outlier values fall between 5000 ppm and 3.8 wt.% Ar. These do not correlate with Ar pressure, suggesting extrinsic (non-lattice) siting of Ar in some cases. Field-emission SEM images and numerical simulations of the diffusion process document isolated nanopores as the hosts for the occasional very high concentrations of Ar (observable pores range down to ∼10 nm in diameter; indirect evidence points to smaller ones as the more common sinks for Ar). The systematics of the data suggest an actual (lattice) solubility of ∼2000 ppm at 100- to 200-MPa Ar pressure, which is equivalent to a partition coefficient of ∼0.001cm3STP/g · atm. Using either organic clathrate or fullerene as the Ar sources, 1-GPa experiments in a piston-cylinder apparatus result in similar uptake of Ar into quartz, in this case through partitioning equilibrium with C-O-H fluid (clathrate source) or amorphous carbon (fullerene source). The ability of quartz, relative to other minerals, to incorporate significant amounts of Ar may allow this ubiquitous and abundant mineral to serve as a local sink for Ar in crustal rocks lacking a free fluid phase. The diffusion data permit open-system behavior of Ar in quartz below the closure temperature of biotite and other 40Ar source minerals.

Sr, Y, and REE diffusion in fluorite

The diffusion of Sr, Y and three rare earth elements (Nd, Dy and Yb) has been measured in natural fluorite under dry 1-atm conditions. Sources of diffusant consisted of mixtures of Sr, Y, or REE fluorides and CaF2 powders, pre-reacted in vacuo in silica glass tubes. Experiments were performed by sealing the prepared source material with pieces of cleaved fluorite in silica glass tubes under vacuum, and annealing in vertical tube furnaces for times ranging from 20 min to a few months. Following diffusion anneals, concentration profiles were measured with Rutherford Backscattering Spectroscopy (RBS). Over the temperature range 700–1050°C, the following Arrhenius relations are obtained: DSr=2.3×101 exp(−450 kJ mol−1/RT) m2/s DY=1.2×102 exp(−454 kJ mol−1/RT) m2/s DNd=4.8×10−2 exp(−385 kJ mol−1/RT) m2/s DDy=3.2×100 exp(−419 kJ mol−1/RT) m2/s DYb=3.1×10−1 exp(−395 kJ mol−1/RT) m2/s Diffusion rates for the REE (and Y) are all quite similar, and faster than diffusion of Sr. This is in contrast to observations in other mineral systems (e.g., feldspars, zircon) where diffusion of more highly charged cations proceeds more slowly than that of cations with lower charge. However, the ionic radius of Sr2+ (1.26 A) is larger than that of Y3+ (1.019 A) or the investigated REE (0.985–1.109 A), suggesting that differences in cationic size may exert greater influence on diffusion rates in fluorite than do differences in cation charge. This size dependence is weak, however, perhaps due to similarities in size between these cations and Ca, for which they likely substitute in the fluorite lattice, or as a consequence of the relatively flexible fluorite lattice. The diffusion findings also have important implications for Rb–Sr and Sm–Nd isotopic dating and can be useful in interpreting zoning patterns observed in fluorite. These data indicate that REE zoning on the 100 μm scale will be retained for times on order of the age of the earth at temperatures below typical crystallization temperatures for fluorite in igneous systems (i.e., ∼500°C), provided that the zoning is modified solely by volume diffusion.

Olivine/silicate melt partitioning of germanium: an example of a nearly constant partition coefficient

Abstract The partitioning of germanium between forsterite (Fo) and liquids in the diopside-anorthiteforsterite join was investigated by electron microprobe analysis of Ge-doped samples equilibrated at 1300°–1450°C. Germanium is somewhat incompatible in Fo relative to the haplobasaltic melts, with a grand mean for all simple partition coefficients (DFo-lGe) of 0.68 ± 0.06. For the melt composition range studied, DFo-lGe is virtually constant in isothermal series of experiments, and shows only minor overall temperature dependence. The exchange reaction partition coefficient K D = ( Mg 2 GeO 4 ) Fo (SiO 2 ) l (Mg 2 SiO 4 ) Fo (GeO 2 ) l ] is near unity in all cases, with a grand mean of 0.93 ± 0.11. One exploratory run at 20 kbar yielded a distinctly lower partition coefficient (DFo-lGe = 0.54 ± 0.04), which confirms the negative pressure dependence predicted by the thermodynamics of Ge ai Si exchange. These new data indicate that absolute Ge enrichment must occur in terrestrial magmas undergoing olivine fractionation, while Ge Si remains nearly constant.