Enhanced diffusion of uranium and thorium linked to crystal plasticity in zircon

Abstract

The effects of crystal-plasticity on the U-Th-Pb system in zircon is studied by quantitative microstructural and microchemical analysis of a large zircon grain collected from pyroxenite of the Lewisian Complex, Scotland. Electron backscatter diffraction (EBSD) mapping reveals a c.18 degree variation in crystallographic orientation that comprises both a gradual change in orientation and a series of discrete low-angle (<4 degrees) boundaries. These microstructural data are consistent with crystal-plastic deformation of zircon associated with the formation and migration of dislocations. A heterogeneous pattern of dark cathodoluminescence, with the darkest domains coinciding with low-angle boundaries, mimics the deformation microstructure identified by EBSD. Geochemical data collected using the Sensitive High Resolution Ion MicroProbe (SHRIMP) shows a positive correlation between concentrations of the elements U, Th and Pb (ranging from 20-60 ppm, 30-110 ppm, and 14-36 ppm, respectively) and Th/U ratio (1.13-1.8) with the deformation microstructure. The highest measured concentrations and Th/U coincide with low-angle boundaries. This enrichment is interpreted to reflect enhanced bulk diffusion of U and Th due to the formation and migration of high-diffusivity dislocations. 207Pb/206Pb ages for individual analyses show no significant variation across the grain, and define a concordant, combined mean age of 2451 +/- 14 Ma. This indicates that the grain was deformed shortly after initial crystallization, most probably during retrograde Inverian metamorphism at amphibolite facies conditions. The elevated Th over U and consistent 207Pb/206Pb ages indicates that deformation most likely occurred in the presence of a late-stage magmatic fluid that drove an increase in the Th/U during deformation. The relative enrichment of Th over U implies that Th/U ratio may not always be a robust indicator of crystallization environment. This study provides the first evidence of deformation-related modification of the U-Th system in zircon and has fundamental implications for the application and interpretation of zircon trace element data.

Figure 1

Simplified geological map of NW Scotland showing the location of sample GST15. The Assynt terrane is bounded to the north by the Laxford Front (LF) and to the east by the Moine thrust (MT). Adapted from Kinny and Friend [34].

Figure 2

Maps of the whole zircon grain derived from EBSD data. Maps contain artifacts resulting from stitching several smaller maps. a) Band contrast map where pixel values correspond to the contrast in hough space, and is a measure of pattern quality. This map shows the external morphology of the grain position of intragrain fractures (e.g., F). The x,y indicates the arbitrary sample reference frame used throughout the study. b) Lower hemisphere equal area stereographic projection of the poles to low-index planes for the reference orientation shown by a cross in (a). Data plotted in the sample reference frame. c) Cumulative misorientation map in which each pixel is coloured for minimum misorientation relative to the reference orientation shown in (a) and (b). This shows lattice distortion is predominantly localised at the grain tips. The right hand box defines a domain (area a) of cumulative misorientation up to 18°, and is the focus of this study. (i) Shows a cross-cutting linear feature that defines an abrupt change in orientation. d) Cumulative misorientation profile for line a-a' shown in (b). The position of a cross-cutting linear feature (i) is indicated by an arrow. e) Lower Hemisphere equal area stereographic projection of showing the dispersion of poles to low-index planes across the whole grain.

Figure 3

Maps of area A shown in Fig. 2c. a) Cumulative misorientation map derived from EBSD data showing disorientation relative to reference orientation shown by a cross. Boundaries between adjacent pixels with disorientation angles of >0.5°, >1°, and >2° are shown as solid lines. Misorientation axes for boundaries (i) to (iv) are shown in Fig. 5. (v) Stereographic projection of poles to low index crystallographic planes for all pixels indexed as zircon in shown in (a). b) Local misorientation map derived from EBSD data. Each pixel is coloured according to the mean disorientation value from a surrounding 5 × 5 pixel grid. Misorientations that define coarse vertical grid lines in (a) and (b) are an artefact of data collection (see text), and were disregarded for misorientation analysis. The position of SHRIMP analyses are indicated on each map. Profile along line x-x' is shown in Fig. 6. c) Panchromatic cathodoluminescence (CL) image. The locations of SHRIMP analyses are shown. Pale linear features are scratches in the C-coat.

Figure 4

Lower hemisphere equal area stereographic projections of disorientation axes along low-angle boundaries (i) to (iv) indicated on Fig. 3a. Grey symbols are low index poles of the reference orientation 'x' in Fig. 2b. Refer to Fig. 2b for legend. Thick black lines extending outside the primitive circle indicate the range of boundary trace orientations on the sample surface.

Figure 5

(a) Wavelength CL map of detailed area shown in Fig. 3c, shaded for integrated CL intensity. (b) CL spectra for points 1 to 8 shown in (a). Each spectrum represents mean values from a 3 × 3 pixel local grid. A broad double peak at ~500nm and ~540nm is common to all spectra. See text for discussion.

Figure 6

(a) to (e) Maps to show spatial variations in trace element geochemistry and isotope data. a) U concentration.(b) Th concentration. c) Th/U ratio. d) Pb concentration. e) ²⁰⁷Pb/²⁰⁶Pb age. Base image is panchromatic cathodoluminescence image shown in Fig. 3a. (f) Concordia plot of all U-Pb SHRIMP analyses for whole grain (n = 50). Error boxes are 1σ. All data are ≥ 94% concordant. The combined mean ²⁰⁷Pb/²⁰⁶Pb age is 2,451 ± 14 Ma, with an MSWD of 1.3.

Figure 7

Profile of misorientation, panchromatic CL intensity, and geochemical data along line x-x' shown in Fig 3. Vertical grey bars represent areas covered by SHRIMP analyses. Errors in ²⁰⁷Pb/²⁰⁶Pb ages are 1σ. Mean ²⁰⁷Pb/²⁰⁶Pb age is for the whole grain (n = 50). Trace element concentrations for SHRIMP analyses are given nominal ±20% errors. Grey geochemistry symbols show values for analyses immediately adjacent line x-x'. Sharp reductions in CL intensity associated with brittle fractures are shown. Artefact misorientations at map edges are shown. Dashed lines are to guide the eyes only. See text for further interpretation.

Figure 8

Plots of local misorientation versus (a) U and (b) Th concentration. Open boxes show analyses that do not lie over low-angle boundaries, closed boxes analyses over low-angle boundaries. Nominal ± 20% error bars are assigned to trace element concentration measurements. Data has been fitted with linear trend lines. See text for discussion.