Brief thermal pulses during mountain building recorded by Sr diffusion in apatite and multicomponent diffusion in garnet

https://doi.org/10.1016/j.epsl.2007.07.017Get rights and content

Abstract

We use the records of Sr diffusion preserved in apatite and corroborating multicomponent Fe-Mg–Ca–Mn diffusion preserved in garnet from the classic Barrovian metamorphic zones (Scotland) to quantify the timescale for the thermal peak of crustal heating c. 465 Ma. We show that Sr diffusion in apatite is a powerful means to determine thermal timescales, and provide the first set of diffusion-based timescale estimates across a wide range of metamorphic grades for Barrovian metamorphism. The results demonstrate that the thermal peak was extremely brief and lasted a few hundred thousand years. This timescale is one to two orders of magnitude shorter than peak timescales predicted by conventional models based on conductive relaxation of overthickened crust. The short peak thermal pulse or pulses probably involved advective heat transfer driven by magmas and associated fluid flow, followed by rapid exhumation. Peak thermal pulses may represent a very short part (< 1 Ma) of overall mountain building cycles lasting ∼ 107 yr or more, but play a dominant role in determining mineralogy, geochemical fluxes, and fluid production in mountain belts.

Introduction

Profound questions surround the timescales, rates, and mechanisms of heat transfer through Earth's crust required to drive chemical reactions that produce metamorphic rocks and liberate fluids during plate convergence and mountain building (Chamberlain and Rumble, 1988, Baxter and DePaolo, 2000, Camacho et al., 2005, Kelly, 2005, Wilbur and Ague, 2006, Bjornerud and Austrheim, 2006). Frequently cited models of conductive thermal relaxation in tectonically overthickened crust predict that attainment of peak (maximum temperature) conditions should span ∼10 million years or more across collisional orogenic belts (Thompson and England, 1984). However, recent work in the Barrovian type locality (Scotland) and in related rocks in Ireland suggests much shorter timescales (Friedrch et al., 1999, Oliver et al., 2000, Baxter et al., 2002) which may have resulted from advective magmatic heating and associated fluid flow ∼ 465 million years ago (Baxter et al., 2002). This result has broad implications because Barrovian metamorphic sequences are the most commonly recognized type in mountain belts worldwide.

Very precise estimates of the duration of heating are required to determine mechanisms of heat transfer and tectonism, but conventional radiometric dating methods have typical uncertainties of several million years or more for rocks of this age. However, the record of diffusion preserved within minerals has the potential to resolve extremely short-lived events. Diffusion is a thermally-activated process that operates to smooth compositional heterogeneities within crystals. Thus, preservation of pre-existing heterogeneities constrains the timescales of thermal events, if diffusion rates, temperature, and initial concentration profiles are well constrained (Watson et al., 1985, Cherniak and Ryerson, 1993, Lasaga and Jiang, 1995, Ganguly et al., 1996, van Haren et al., 1996, Carlson and Schwarze, 1997, Graham et al., 1998, Dachs and Proyer, 2002, Faryad and Chakraborty, 2005). For example, diffusive profiles in garnet have been used to estimate retrograde cooling and exhumation rates (Lasaga and Jiang, 1995).

Recent diffusion-based results suggest that some metamorphic processes may operate over much shorter timescales than previously thought. Examples include fluid infiltration and vein formation during Barrovian metamorphism in New England (van Haren et al., 1996), fluid-triggered granulite to eclogite reaction, Holsnøy, Norway (Camacho et al., 2005), thermal pulses on Naxos, Greece (Wijbrans and McDougall, 1986), and rapid burial and/or exhumation of eclogite and associated conduction-dominated heat transfer during alpine orogenesis (Dachs and Proyer, 2002, Faryad and Chakraborty, 2005). Nonetheless, hypotheses of short-lived metamorphism have been controversial (Bjornerud and Austrheim, 2006), largely because quantitative interpretation of intracrystalline diffusion may depend on a variety of factors including: (1) compositional variations within mineral solid solutions, (2) initial and boundary conditions, (3) uncertainties in placing the observed diffusion profiles into the context of prograde, peak, and/or retrograde stages of the orogenic cycle, and (4) in some cases, the presence or absence of water. Furthermore, experimental calibrations of diffusion rates can have significant uncertainties which are often not propagated when field-based estimates of timescales are made. It begs the question whether these inferences of short-lived metamorphic processes are rare, isolated exceptions or whether they may characterize continental metamorphism in general.

We test the hypothesis of short-lived thermal pulses during typical regional metamorphism by focusing on a widespread mineral not previously considered in this regard: fluor–apatite (apatite). Sr diffusion in apatite is a very promising means to quantify timescales of thermal events, as discussed here. (1) The Sr diffusion is described by a tightly-constrained, linear Arrehnius relationship calibrated over a large T range of 550 °C (Watson et al., 1985, Cherniak and Ryerson, 1993). (2) Three different experimental techniques all produce a single, internally-consistent diffusion calibration (Watson et al., 1985, Cherniak and Ryerson, 1993). The diffusion exchange vector appears to be very predictable and may simply be: Sr2+←→Ca2+. (3) The diffusion coefficient (D) is calibrated down to a relatively low experimental T of 700 °C (Cherniak and Ryerson, 1993). Diffusion modeling for our samples involves only a small projection (100 to 200 °C) to lower T. It is extremely unlikely that the diffusion mechanism would change significantly over this relatively small projection range. (4) The effects of radiation damage on diffusivity are negligible at high temperatures where such damage is quickly annealed (Weber et al., 1997). (5) Natural fluor-apatites have compositions that closely approach the pure end-member (see below), so complications due to extensive solid solution are not an issue. (6) Diffusion of Sr may be weakly anisotropic (faster perpendicular to c-axis), but anisotropic effects are within experimental error (Watson et al., 1985, Farver and Giletti, 1989, Cherniak and Ryerson, 1993) and are accounted for in our data reduction. (7) Elemental (not isotopic tracer) Sr diffusion was measured experimentally (Watson et al., 1985, Cherniak and Ryerson, 1993), directly analogous to the natural Sr diffusion that we study. (8) Sr diffusion slows sufficiently below about 500 °C, so the system is effectively non-responsive to diffusive modification during most lower greenschist and sub-greenschist facies retrogression. (9) Sr concentrations can be easily determined using the electron microprobe. (10) Pressure effects should be negligible at the relatively low pressures of interest here (Farver and Giletti, 1989). (11) Apatite is a common phase that can grow during prograde metamorphism across a range of grades and persist through the peak metamorphic conditions we seek to constrain. Finally, it is worth emphasizing that diffusion timescales can be preserved in the compositions of many other minerals as well, thus providing the potential for multiple corroborating chronometers.

Section snippets

Geologic relations

Samples are from the Dalradian metasediments in Barrow's biotite (Bt), garnet (Grt), and staurolite (St) zones north of Stonehaven, and from the sillimanite (Sil) zone in the type locality near Glen Clova (Barrow, 1893, Atherton, 1977) (Fig. 1). Grt zone specimens JAB1 and JAB101L lie above the chloritoid isograd; 101L contains chloritoid crystals. Detailed sample locations and descriptions are published elsewhere (McLellan, 1985, Ague, 1997, Masters et al., 2000, Ague et al., 2001, Baxter et

Imaging and electron microprobe analyses

Chemical analyses and backscattered electron images of minerals were done using the JEOL JXA-8600 electron microprobe at Yale University. Analyses for Sr in apatite used 15 kV accelerating voltage, 25 nA cup beam current, wavelength-dispersive spectrometers, natural and synthetic standards, φ(ρz) matrix corrections, off-peak background corrections, 5 μm diameter beam spot (to minimize beam damage), 150 s on-peak counting times for Sr, and 75 s on each off-peak position. Multiple Sr

General approach

Here we outline our general diffusion modelling approach for apatite and garnet. Details are then provided in the following sections. To be conservative, our calculations assume initial step functions in composition across core-rim boundaries. Taking initially smoothed profiles would result in shorter timescales. Forward modelling was used to compute the time required to relax the initial step functions to match the measured chemical profile in each grain given known D during peak metamorphic T

Short peak metamorphic timescales

The best-estimates for peak timescales for eight apatite (Grt and St zones) and two garnet (Sil zone) crystals are remarkably short, ranging from < 1.0 × 105 to 6.5 × 105 yr with a geometric (logarithmic) mean of 2.3- 0.8+ 1.4 × 105 yr (95% confidence) or an arithmetic mean of 2.7 × 105 ± 1.3 × 105 yr (Table 1, Fig. 8). Notably, the timescales agree regardless of metamorphic grade, at least within our uncertainties. The results for apatite and garnet are statistically indistinguishable, although there is a

Petrological and tectonic implications of extremely brief thermal pulses

The Grampian orogeny lasted for only c. 10–15 Ma in Scotland (Friedrch et al., 1999, Oliver et al., 2000, Baxter et al., 2002) and probably involved subduction along the Highland Boundary Fault (Dempster et al., 2000, Masters and Ague, 2005). While a background (i.e.,<  500 °C) regional conductive heating and cooling due to overthickening probably spanned nearly the entire 10–15 Ma orogeny, the thermal peak conditions lasted for only 200–300 kyr and were attained c. 465 Ma close to the end of

Acknowledgements

We thank J.O. Eckert, Jr., for valuable assistance with the electron microprobe, A. Camacho for thoughtful comments on an earlier version of this paper, two anonymous referees for their critical and constructive reviews, and the National Science Foundation Directorate for Geosciences for support (NSF EAR-0509934 to JJA and EAR-0547999 to EFB).

References (52)

  • M.P. Atherton

    The variation in garnet, biotite, and chlorite composition in medium grade pelitic rocks from the Dalradian, Scotland, with particular reference to zoning in garnet

    Contrib. Mineral. Petrol.

    (1968)
  • M.P. Atherton

    The metamorphism of the Dalradian rocks of Scotland

    Scott. J. Geol.

    (1977)
  • M. Ayres et al.

    A comparative study of diffusion profiles in Himalayan and Dalradian garnets: constraints on diffusion data and the relative duration of the metamorphic events

    Contrib. Mineral. Petrol.

    (1997)
  • G. Barrow

    On an intrusion of muscovite–biotite gneiss in the south-eastern highlands of Scotland, and its accompanying metamorphism

    Q. J. Geol. Soc. Lond.

    (1893)
  • E.F. Baxter

    Natural constraints on metamorphic reaction rates

  • E.F. Baxter et al.

    Field measurement of slow metamorphic reaction rates at temperatures of 500° to 600 °C

    Science

    (2000)
  • E.F. Baxter et al.

    Prograde temperature–time evolution in the Barrovian type-locality constrained by Sm/Nd garnet ages from Glen Clova, Scotland

    J. Geol. Soc. Lond.

    (2002)
  • M.G. Bjornerud et al.

    Hot fluids or rock in eclogite facies metamorphism?

    Nature

    (2006)
  • C.M. Breeding et al.

    Isotopic and chemical alteration of zircon by metamorphic fluids; U–Pb age depth-profiling of zircons from Barrow's garnet zone, northeast Scotland

    Am. Mineral.

    (2004)
  • A. Camacho et al.

    Short-lived orogenic cycles and the eclogitization of cold crust by spasmodic hot fluids

    Nature

    (2005)
  • W.D. Carlson

    Rates of Fe, Mg, Mn, and Ca diffusion in garnet

    Am. Mineral.

    (2006)
  • W.D. Carlson et al.

    Petrological significance of homogenization of prograde growth zoning in garnet; an example from the Llano uplift

    J. Metamorph. Geol.

    (1997)
  • C.P. Chamberlain et al.

    Thermal anomalies in a regional metamorphic terrane: An isotopic study of the role of fluids

    J. Petrol.

    (1988)
  • C.P. Chamberlain et al.

    The influence of fluids on the thermal history of a metamorphic terrain: New Hampshire, USA

  • C.B. Chernoff et al.

    Disequilibrium for Ca during growth of pelitic garnet

    J. Metamorph. Geol.

    (1997)
  • G.A. Chinner

    Pelitic gneisses with varying ferrous/ferric ratios from Glen Clova, Angus, Scotland

    J. Petrol.

    (1960)
  • Cited by (146)

    View all citing articles on Scopus
    View full text