Time scales of geological processes from modelling the zoning patterns of crystals

1. Introduction to zoning in crystals 2. Diffusion equation 3. Diffusion coefficient 4. Modeling natural crystals: isothermal case * Initial conditions * Boundary conditions 5. Problems, pitfalls and uncertainities

1. Zoning in crystals

X-ray distribution map of olivine from lava lake in Hawaii Moore & Evans (1967)

Development of electron microprobe 1960, first traverses and X-ray Maps

Among the first applications to obtain time scales those related to cooling histories of meteorites (e.g. Goldstein and Short, 1967)

1. Zoning in crystals
- Major elements, trace elements, and isotopes

- Increasingly easier to measure gradients with good precision and spatial resolution (LA-ICP-MS, SIMS, NanoSIMS, FIB-ATEM, e-probe, micro-FTIR)

Major and trace element zoning in Plag

Stable isotope zoning 18O in zircon from Yellowstone magmas

Bindeman et al. (2008)

Sr isotope zoning in plagioclase

Tepley et al. (2000)

Compositional Zoning in Grt

Lower part of Grt-zone Sample, WAH Peak P-T: 8.10.9 kbar, 57550 oC

Compositional Zoning in Grt

Upper Part of Grt-zone Sample, WAH Peak P-T: 8.10.5 kbar, 58065 oC

Compositional Zoning in Grt

Upper Part of Grt-zone Sample, WAH Peak P-T: 8.10.5 kbar, 58065 oC

Compositional Zoning in Grt

Ky-zone Sample, WAH Peak P-T: 8.30.6 kbar, 59565 oC

Compositional Zoning in Grt

Ky-zone Sample, WAH Peak P-T: 8.30.6 kbar, 59565 oC

Compositional Zoning in Grt

Ky-migmatite zone Sample, WAH Peak P-T: 8.60.8 kbar, 67535 oC

Compositional Zoning in Grt

Kfs-Ky-Kfs Migmatite zone Sample, WAH Peak P-T: >7 kbar, 79555 oC

Compositional Zoning in Grt

Kfs-Ky-Kfs Migmatite zone Sample, WAH Peak P-T: >7 kbar, 79555 oC

Garnet Zoning
Garnet: use of elements of contrasting D for initial conditions
Garnet: effect of crystal size

Sc

Mn
Spear & Kohn (1996), Geology 24 Spear & Daniel (1998), GMR 1

Compositional Zoning in Garnet Due to Garnet Growth

Stability of Grt-bearing assemblages in Different Metamorphic Grades

1. Zoning in crystals
Diffusion driven by a change in P, T, or composition t0, T0, P0, X0 tt, Tt, Pt, Xt

distance

distance

1. Zoning in crystals
- The compositional zoning will reequilibrate at a rate governed by the chemical diffusion (Ficks laws)

- Because D is Exp dependent on T and Ds in geological materials are slow, minerals record high T events (as opposed to room T)
distance

1. Zoning in crystals

Crystals record the changes in variables and environments: gradients are a combined record of crystal growth and diffusion

MEASURING TIMESCALES: THE CHARACTERISTICS OF CLOCKS

Stage 1: A short-time branch, R~0%, No measurable diffusion Stage 2: An intermediate branch, R=f(t), Measurable diffusion, System frozen, Compositionally zoned mineral

Stage 3: A long-time branch, R~100%, Diffusion complete, System at equilibrium, Compositionally Homogeneous mineral
Time recorder: Diffusion, Deformation, Chemical Reaction, Atomic Ordering, Radioactive Decay

2. The diffusion equation

2. Diffusion, flux, and Ficks law

Diffusion: (1) motion of one or more particles of a system relative to other particles (Onsager, 1945)

(2) It occurs in all materials at all times at temperatures above the absolute zero

(3) The existence of a driving force or concentration gradient is not necessary for diffusion

2. Diffusion, flux, and Ficks law

GAS

LIQUID

SOLID

2. Diffusion, flux, and Ficks law

Random motion leads to a net mass flux when the concentration is not uniform: equalizing concentration is a consequence, NOT the cause of diffusion

Ficks first law Flux has units of mass or moles or volume * distance/time Diffusion coefficient has units of distance2/time

Ficks Law of Diffusion: General Features

Diffusion Equation: Application of Continuity Relation

where

moles/cc

Diffusion Coefficients: Different Types

Tracer diffusion coefficient Self diffusion coefficient Chemical diffusion coefficient Interdiffusion coefficient

Tracer diffusion coefficient

A tracer diffusion coefficient is defined as the diffusion coefficient of a tracer isotope of an element that describes its flux in response to the specific isotopic concentration gradient in a medium that is homogeneous with respect to the concentration of the chemical elements. Such a medium is referred to as a chemically homogeneous medium. In practice, this can be ensured only when an isotope of an element diffuses by exchanging position with another isotope of the same element. However, this exchange mechanism may not always hold as an isotope of an element may exchange position with an isotope of a different element, in which case one deals with the problem of interdiffusion. As long as the diffusing isotope is very dilute so that its thermodynamic behavior is within the Henrys law limit, it can be shown that the interdiffusion is effectively the same as the isotopic or tracer diffusion. A tracer diffusion coefficient of an element will be designated by the symbol D(*i) here the superscript * indicates the isotope and i stands for the element.

Self diffusion coefficient

A self diffusion coefficient is the one that describes the flux of an element solely in response to its own concentration gradient and under condition that its thermodynamic interaction with the matrix is independent of the change of its own concentration. A tracer diffusion coefficient of an element is the same as its self diffusion coefficient when all the isotopes of the element have the same diffusivity. In this work, the self and tracer diffusion coefficients Will be treated as effectively the same quantity. A self diffusion coefficient of an element i is denoted by the symbol D(i).

Chemical diffusion coefficient

A chemical diffusion coefficient of an element (i) represents a product of its self diffusion coefficient and a thermodynamic factor (TF) that is given by the quantity (1 + lni/lnCi) where i and Ci are, respectively, the activity coefficient and concentration (moles per unit volume) of the component i. If the molar volume of the medium remains constant, then the lnCi = lnXi, where Xi is the mole fraction of the component i.

Interdiffusion coefficient
An interdiffusion coefficient of two elements, denoted by D(i-j), represents a product of the weighted average of the two self diffusion coefficients, D(i) and D(j), and the thermodynamic factor. For the special case of interdiffusion of two equally charged species in a medium of fixed molar volume, it is given by

Factors Controlling Diffusion Coefficient

1. Temperature: D = Do exp(-Q/RT), where Q: activation
energy of diffusion, generally expressed at 1 bar Do= Pre-exponential factor;

2. Pressure: 3. fO2:

Qp= Q1bar + (p-1) * V+, where p= pressure in bar; V+ = activation coefficient of diffusion in cc

Diffusion Coefficient and Arrehenius Relation

Summary of self-diffusion data of Fe, Mg, Mn and Ca determined by modeling multicomponent diffusion profiles in diffusion- couples annealed at high P-T conditions within graphite capsules. All data are normalized to a pressure of 10 kb. Solid and dashed lines: Arrehenius relations to the data obtained from AlmPrp (Ganguly et al. 1998b) and Alm-Sps (Chakraborty and Ganguly 1992) diffusion couples, respectively.

Arreheneus Relation: Calculate Q1bar

P=10 kbar

Multicomponent diffusion

Multicomponent Garnet Diffusion

Grt D Matrix
Mg Mg Fe Mn DMgMg DFeMg DMnMg Fe DMgFe DFeFe DMnFe Mn DMgMn DFeMn DMnMn

nth component is the dependent component, here DCa

Multicomponent Garnet Diffusion

DMgMg= D*Mg-[(D*MgXMg)/(D*MgXMg+ D*FeXFe+ D*MnXMn + D*Ca*XCa)] (D*MgD*Ca)

DMgMn= -[(D*MnXMn)/(D*MgXMg+ D*FeXFe+ D*MnXMn + D*Ca*XCa)] (D*Mg-D*Ca)

Analytical Solution of Diffusion Equation:

1. The constant source solution for constant D:

where Cs = Surface concentration, C=initial concentration in the crystal; C(x,t) = concentration at any point (x) within the crystal at time t; D=Diffusion coefficient; T=time

2. Two semi-infinite media with a common interface and initial homogeneous but contrasting compositions

where Ci(o)= lower of the two initial concentrations, Ci,o =Diff in initial concentrations of the species i at the interface; D(i-j)= binary inter-diffusion coefficient between the species i and j .

Numerical Solution of Diffusion Equation: Finite Difference Method

Courant condition, r=Dt /(x)2 0.5.

Advantages and Limitations of the Numerical Methods

1. If an explicit method is to be used, we cannot choose D, x and t arbitrarily. As D and x are usually specified by the problem at hand, the choice of t is limited by the Courant condition. 2. Inspection of the formula above shows the necessity to have the two boundary conditions and one initial condition. Boundary condition: Knowledge of compositions at the 1st (i-1 th) and last point (i+1 th point). Concentrations held constant or there is no flux etc 3. As these calculations proceed by explicitly determining the concentrations at each position at each time step, it is easy to add variations in the model. For example, to start off a calculation we need to define the concentrations at each grid point (the initial condition). Because each concentration is specified individually, it is irrelevant whether the concentration profile is homogeneous, has a regular geometric form, or is of an arbitrary shape. Similarly, if we wish to change the boundary conditions subject to which diffusion occurs after, say, the 20th time step, we can directly do so. This is where the flexibility of the numerical method noted in the text comes from.

Outline-I

1. Application to isothermal problems

(a) boundary and initial conditions (b) effect of geometry and multiple dimensions

2. Application to problems with cooling history (a) approach of Lasaga and co-workers

Outline- II
3. Uncertainties

(a) in model parameters (e.g., T)

(b) in the modeling approach (e.g., boundary conditions) 4. Examples of applications (a) olivine in magmatic rocks (c) garnet in metamorphic rocks

1. Isothermal diffusion problem

t0, X0 tt, Xt

Ci Ci (Di ) t x x
Boundary ?

Initial ?
distance distance

D does not depend on t but can depend on C or other variables

Boundary conditions

I. (a) open or (b) closed to flux with the environment II. may also include the effects of: (a) crystal growth or (b) dissolution

Boundary conditions
Example of Fe/Mg diffusion in garnet

O O
What are the effects of boundary conditions for calculated times ?
Fe/Mg

Garnet

O
C C

Florence & Spear (1995) EPSL 134

Boundary conditions
Closed boundary

Open boundary

distance

Conc.

Conc.

Boundary conditions: effects on time scales

Equilibration in the closed system occurred much faster

Incorrectly applying a no flux condition to an open system can lead to underestimation of time by factors as large as an order of magnitude (this has happened in the literature!)

Boundary conditions: effects of crystal growth or dissolution

(a) Neglecting crystal growth tends to overestimate time scales (b) Neglecting crystal dissolution tends to underestimate time scales (e.g., shortening of diffusion profiles)

t1

t2

distance

distance

Initial conditions
tt, Xt
Boundary OK Initial ? distance

Need to be evaluated using geological criteria; vary depending on the problem at hand. At least 4 end-member possibilities
1. Compare profiles of elements with extremely different diffusivities; and also the zoning profile of crystals of different sizes

Initial conditions
Plagioclase - Diffusivity of anorthite is ~ 10^7 slower than Mg
DMg~ 100 x DSr ~ 100,000x DNaSi-CaAl Different elements can record different aspects of the thermal history experienced by a plagioclase crystal. Modeling the differing degrees of modification of Compositional zoning of various elements in one crystal provides a unique tool to constrain timescales of processes in a Magma chamber.

- Small (Pl4) crystal more equilibrated than larger ones (Pl7)

Costa et al (2003), GCA 67

Initial conditions
Garnet: use of elements of contrasting D for initial conditions
Garnet: effect of crystal size

Sc

Mn
Spear & Kohn (1996), Geology 24 Spear & Daniel (1998), GMR 1

Initial conditions
2. (a) Using arbitrary maximum initial concentration range in natural samples (this provides maximum time estimates). (b) Some times possible to use an oscillatory zoned model. Elimination of smaller wavelengths by diffusion Preservation of sharp compositional gradients in igneous garnet, unmodified by diffusion, can be used as a time constraint (i.e. maximum permissible duration of a thermal event without diffusive modification), if the cation diffusion coefficients are known and if sharpness can be quantified.

Trepmann et al (2004), EJM 16

Initial conditions
3. Model of Spear et al (1990). Initial concentrations are constrained by thermodynamic equilibria + mass balance to establish the initial zoning of garnet during growth 4. Homogenous initial profiles has proved a good approach when dealing with minerals that have experienced high-T episodes prior to the event that finally drove diffusion, e.g., Fe-Mg in garnet or olivine (e.g., Chakraborty & Ganguly, 1991; Costa & Dungan, 2005)

What are the effects of initial profiles in calculated time scales?

Initial conditions

100

Equil %

0 time distance

Conc.

Conc.

Initial conditions: effects on time scales

Despite the difference in shapes of the initial profiles the maximum difference on calculated time scales is a factor of ~1.5

Thus, although the initial profile that we assume controls the time that we obtain, the error can be evaluated and is typically not very large

Effects of geometry and multiple dimensions

These are important depending on the type, shape, and size of the crystal that we are studying and on the diffusion time Careful with the orientation of the diffusion front with respect to the traverse- one can create unnecessary and artificial 2D effects
Not OK OK

Neglecting multidimensional effects tends to overestimate time scales

Effects of geometry and multiple dimensions

Example of plagioclase 1D, t = 225 y

800 700 600 500

ini

Mg ppm

400 300 200 100

calc
equi
0 50 100 microns 150 200

Costa et al (2003), GCA 67

Effects of geometry and multiple dimensions

ini 800
700 600 500 400 300 200 100 0

calc
Mg ppm

800 700 600 500 400 300 200 100 0

Mg ppm

1D, t = 225 y ------> 2D, t = 60 y almost a factor of 4!

Fe-Mg zoning in olivine: diffusion anisotropy and 2 dimensional effects

Mg zoning

Model Initial

b
Fo zoning

150 m

Isothermal diffusion problem summary

A clear understanding of the initial and boundary conditions plus the effects of multiple dimensions in the natural system is very important because it can have large effects on the calculated time scales Careful petrographic observations, analytical work and detailed modeling is necessary for obtaining reliable time estimates from diffusion modeling

Next step:cooling history

The effect of a slow cooling history

Important for metamorphic and some igneous rocks (plutons and dikes) t0, T0 tt, Tt G B t0, T0

G
B

tt, Tt

G
K0
distance

G
B
distance

Kt B

The effect of a slow cooling history

The driving force for diffusion is related to the T dependence on K. T(t) ---> K(t)

tt, Tt G K(t) B
distance

G B

Fe# Fe#

K(t) Ko e

Q ] RT(t)

The effect of a slow cooling history

But because of T(t) ----> D(t)

tt, Tt

G
K(t)

D(t) Do
B
distance

E ] RT(t)

C D(t) t

C 2 x
2

The effect of a slow cooling history

1.
G Fe# B Fe#

K(t) Ko
[ E ] RT(t)

Q ] RT(t)

2. D(t) Do e 3. C D(t) t
T(t) 4.

2C x 2

To ~ T(t) To s t Near- linear cooling, s= cooling rate 1 (s t/To)

Next step, find analytical or numerical solution

The effect of a slow cooling history

Approach of Lasaga et al., (1977) and Lasaga, (1983) for analytical solution

Create a new variable t (compressed time scale ),

t ' D( ) d
0

Diffusion equation becomes

C 2 C 2 t' x

The effect of a slow cooling history

C 2 C 2 t' x
With this new time variable t, the initial conditions are the same as with t, but the boundary conditions are not. This requires the mathematical work shown in Lasaga et al., (1977) and Lasaga (1983)

Jaoul & Sautter (1999) and Jaoul & Bejina (2005) basically the same approach, analytical

The effect of a slow cooling history

Chakraborty & Ganguly (1992) reduce the nonisothermal problem to an isothermal one by defining a characteristic diffusion coefficient Dch

Dch t D( )
0

Total amount of diffusion can be obtained by calculating at a given Tch at which D=Dch. Tch~ 0.95 Tpeak

The effect of a slow cooling history

Finally, there is the possibility of constructing finitedifference algorithms. These are extremely flexible and rather simple to implement

What are the effect of cooling history on time scales and shapes of diffusion profiles ?

The effect of a slow cooling history

Blindly modeling a profile with a cooling history at a single T underestimates the time by orders of magnitude

The effect of a slow cooling history

Shapes of profiles are quite different when one incorporates K(t)

The effect of a slow cooling history

Neglecting K(t) largely underestimated the time

The effect of a slow cooling history

A cooling history oscillating around a trend is indistinguishable from that of the simpler trend (Jiang & Lasaga, 1995)

Errors and uncertainties associated with time determinations

One can distinguish two types: (a) those associated with the parameters used in the model (e.g., T, D), and

(b) those associated with how well we understand and reproduce the natural physical conditions (e.g., multiple dimension etc)

Errors and uncertainties associated with time determinations

1. T uncertainties from geothermometers are ~ 50 oC
E (kJ/mol) Uncer.Factor 500 4 200-300 2 100 1.3

2. Experimental D determination at a given T: within a factor of 2 3. Uncertainties in other variables that D depends on, e.g., oxygen fugacity, pressure. Typically much smaller

4. One can expect overall uncertainties on calculated times s of a factor of 2 to 4, e.g. 10 years might mean between 5 - 20 years

Errors and uncertainties associated with time determinations

Uncertainties related to initial, boundary conditions or multidimensional effects are very variably and can be very large, more than 1 order of magnitude

However, by careful petrographical, analytical, and modeling work one can diminish these uncertainties to almost null; it just takes work!

Examples of applications

I. Olivine zoning in lavas from Chilean volcano (Costa & Dungan, 2005). Time scales between assimilation of mafic crust by ascending basalts and eruption

II. Garnet zoning in metamorphic rocks. Exhumation rates of metamorphic core complexes (Spear, 2004)

Example of cooling and exhumation rates of metamorphic core complex using Fe-Mg zoning in garnet

Basic Geology

Valhalla metamorphic core complex part of the Shuswap terrane of British Colombia (Cretaceous)

High grade migmatitic paragneisses, with 820 oC and 8 kbar as peak metamorphic conditions

Mineral zoning
Garnets show Fe-Mg , Mn and also Ca and Cr

Ca

Fe-Mg

Mn

Cr

Ca and Cr diffuse very slowly through garnet: they record the growth history (high T)

Mineral zoning
In contrast, the Fe-Mg and Mn zoning at the garnet rims and record the cooling history of the complex

Modeling of zoning
Diffusion model constructed using only Fe-Mg (it could have been multicomponent in garnet) First attempt using a single T did not reproduce the zoning, so a model with a cooling history was constructed (using the explicit methods)

Modeling of zoning
Best fits when an initial fast cooling (> 200 oC/ Ma) period followed by a longer and slower one (~ 25 oC/ Ma)

Age dating
F. Spear not convinced, so in-situ Th/Pb dating of different generations of Monazite supposed to have crystallized at different Ts

Agreement between radiometric and diffusion times. We understand everything and Ds are correct!

Thermal modeling
How to produce such fast cooling rates? By simple erosion is too slow, and in part by tectonic denudation

Thrust ramp onto a cooler footwall does the work

Thermal modeling

Thermal model fits well with diffusion and radiogeneic dating.

Summary geospeedometry
- Its a method based on thermodynamic and kinetic laws that can access the time scales and rates of geological processes spanning more than 15 orders of magnitude, from seconds to millions of years - Such a wide applicability is unrivaled by any other geochronometric method - The precision and accuracy of the results varies depending on how well we know T, but also Ds .

Summary geospeedometry
- In the case of metamorphic rocks there seems to be a quite good agreement between diffusion data, radiogenic data and thermal modeling

- In the case of magmatic rocks this is much less so. - There is a lot of room for doing new and interesting work related to geospeedometry and diffusion: experiments, modeling, interpretations. Its a good field to invest ones future time and energy

Grt Diffusion: Application in Geothermobarometry

Greenschist to Lower Amphibolite facies Condition: Growth zoning preserved and not significantly modified by diffusion. Strategy: Grt rim composition reflects its peak metamorphic T. Granulite facies Condition: Growth zoning obliterated due to volume diffusion at high T; Core compositionally homogeneous; retrograde zoning near rim. Strategy: Grt core composition reflects possible preservation of its peak metamorphic T. Use of rim compositions of garnet for thermobarometric purpose would yield PT conditions at which diffusion effectively ceased in the garnet grains. These diffusion closure temperatures tend to be 500 C or even lower.

Problem in Granulite facies Grt: For very HT metamorphism, Grt Core may be affected by diffusion during cooling; Grt core no longer suitable for retrieval of TMax. What to be done?

Peak T from HT Granulite Facies Grt

Choice of Grt Grain Size: Large enough for Grt cope to escape diffusive modification. This can be predicted by using the equation below,

where, M= dimensionless parameter, R =gas constant, D(Tp) =diffusion coefficient at the peak temperature (Tp), E =activation energy of binary diffusion, dT/dt =cooling rate at T and =radius of the garnet crystal. Ganguly and Tirone (1999) showed that there is no significant resetting of garnet core composition for M0.1 if Tp1000 C.

Peak T from HT Granulite Facies Grt

Minimum grain size of spherical of garnet crystal surrounded by biotite (or homogeneous infinite matrix) that is required to preserve the peak metamorphic composition at the grain center. Tp stands for the peak temperature. The numbers beside the curves indicate cooling rates (C per Myr).

Equilibrium Composition for Estimation of Pressure

A thin section map of a metapelite granulite showing internal and external Grt decomposition.

Pressure-Controlled Reaction Textures

Grt Compositional zoning