Assessment and implications of (dis)equilibrium in metamorphic rocks

Open Access
Cruz-uribe, Alicia Marie
Graduate Program:
Doctor of Philosophy
Document Type:
Date of Defense:
July 21, 2014
Committee Members:
  • Maureen Feineman, Dissertation Advisor
  • James David Kubicki, Committee Member
  • David H Eggler, Committee Member
  • John Richard Hellmann Jr., Committee Member
  • metamorphism
  • kinetics
The ability to quantify the rates at which metamorphic reactions occur is critical to assessing the extent to which equilibrium is achieved and maintained in a variety of dynamic settings. Here I investigate the kinetics of rutile replacement by titanite during amphibolite-facies overprinting of eclogite, garnet amphibolite and anorthosite from Catalina Island, CA, the Tromsø Nappe, Norway, the North Qaidam terrane, China, and the Guichicovi Complex, Mexico. Trace element concentration profiles across rutile rimmed by titanite, as determined by laser ablation (LA) ICP-MS, reveal Nb zoning in rutile that are interpreted to be the result of Nb back-diffusion from the rutile-titanite boundary. I present new field-based reaction rates calculated from grain boundary velocities, which in turn were calculated using a 1-D diffusion model for Nb back-diffusion into rutile during titanite replacement over the temperature range 670–770 °C. These data are consistent with or slightly faster than previous estimates of regional metamorphic reaction rates, and extend the temperature and composition range over which regional metamorphic reaction rates are known. Further investigation of the kinetics of all metamorphic reactions reveals a strong correlation between heating rate (dT/dt) and net reaction rate for regional, contact, and subduction zone metamorphic settings. This relationship is used to formulate a new expression relating net reaction rate (Rnet) temperature (T) and heating rate (dT/dt): Rnet=(4.24*10^-11)*((dT/dt)^1.25)*exp(0.00668T) This expression provides a universal prediction of metamorphic reaction kinetics at the tectonic scale and accounts for the 4-7 orders of magnitude range in metamorphic reaction rates observed in natural systems. Based on this analysis, the dynamic nature of subduction zones (high dT/dt) may result in conditions that are far from equilibrium (large ΔGrxn), which drives faster reaction rates in these systems. Examination of the relative abundances of trace elements in rutile and titanite reveal striking differences between high temperature (amphibolite-facies) and low temperature (blueschist-facies) overprinting. I find that trace element distributions approach equilibrium partition coefficients in rocks from amphibolite-facies overprinted terranes, whereas trace element distributions did not approach equilibrium in rocks that experienced blueschist-facies overprinting. Calculated Zr-in-titanite temperatures for amphibolite-facies overprinted rocks are consistent with those reported in the literature, while Zr-in-titanite temperatures for blueschist-facies overprinting consistently overestimate temperature by 50-250 °C, suggesting that Zr does not approach equilibrium distributions during blueschist-facies overprinting. I conclude that single phase thermometers that rely upon slow-diffusing high field strength elements should not be applied to rocks equilibrated at ≤550 °C unless attainment of trace element equilibrium can be demonstrated. In situations where trace element equilibrium has been attained, the use of single element thermometers can be quite useful. Here I present a new method for determining 48Ti concentrations in quartz by LA-ICP-MS at the 1 ppm level, relevant to quartz in HP-LT terranes. Titanium contents in low-CL rims in the Bishop Tuff quartz grains were determined to be homogenous by EPMA (41 ± 3 ppm Ti, 2σ), and are a potential natural reference material. I suggest that natural quartz such as the homogeneous low-CL rims of the Bishop Tuff quartz are more suitable than NIST reference glasses as an in-house reference material for low Ti concentrations because matrix effects are limited and Ca isobaric interferences are avoided, thus allowing for the use of 48Ti as a normalizing mass. Titanium concentration from 33 analyses of low-temperature quartz from the Czech Erzgebirge is 0.9 ± 0.2 ppm (2σ) using 48Ti as a normalizing mass and the Bishop Tuff quartz rims as a reference material. The 2σ average analytical error for analyses of 48Ti is 8% for 50 µm spots and 7% for 100 µm spots, which offers much greater accuracy than the 35% error (2σ) incurred from using 49Ti as a normalizing mass.