Myron G. Best. Igneous and metamorphic 2003 Blackwell Science

Подождите немного. Документ загружается.

Evolution of Imposed Metamorphic Fabrics: Processes and Kinetics

537

17.19 (Continued ).

(f)

32% 36°

44% 55°

58% 71°

Pure

shear

Simple

shear

0°

Undeformed array;

random slip directions

80° 40° 40° 80°

two-dimensional sheet of paper it is common prac-

tice to assume a state of plane strain in which all

changes in shape occur in only two dimensions and

none in the third perpendicular to the plane; if this

assumption is valid, then strain can be represented

by a reference circle and a strain ellipse.

Structural geologists recognize two distinct styles

of plane strain in rocks (Passchier and Trouw, 1996,

pp. 15–19). On the left side of Figure 17.20, if a

circle and circumscribed square are progressively

ﬂattened by compressive forces they become an

ellipse and a parallelogram, respectively, in coaxial

pure shear. On the right, shear stresses progres-

sively deform the reference shapes in noncoaxial

simple shear. In some instances, shear stresses may

act on distributed surfaces throughout the deform-

ing body in an analogous manner to slipping a deck

of playing cards. Both modes of deformation are

homogeneous in that the reference circle is changed

uniformly into a regular ellipse, rather than some

irregular shape. Clearly, whether the applied stress

results in a homogeneous deformation depends on

the scale of observation (Figure 17.21); a large body

may have been inhomogeneously deformed, so that

there is uneven, or nonuniform, strain partitioning,

but parts of the body were subjected to homoge-

neous deformation.

Figure 17.20 reveals an important principle of

deformation. The ﬁnal end-product—that is, the

ﬁnite homogeneous deformation—carries no in-

formation about the deformation path—that is, the

details of the history of progressive displacements

and changes wrought in the body by the applied

stress. This can be seen by examination of the con-

trasting paths of rotation and stretching followed

by reference lines during progressive deformation

in the pure and simple shear modes. Although the

paths differ, the end result is identical.

In rocks, we see only the end-product of what

may have been a complex path of deformation,

perhaps resulting from successive overprinting

processes in polymetamorphism. Because all rocks

are inhomogeneous in terms of how they respond

to applied stress, the resulting deformation is also

inhomogeneous. The tectonite fabric in metamor-

phic rocks can be imprinted by differing patterns

of ductile ﬂow during inhomogeneous deformation.

Fabric resulting from coaxial pure shear tends to

have an orthorhombic symmetry; a sphere or cube

is deformed into a triaxial ellipsoid or rectangular

prism. The effects of simple shear as typically

developed in mylonitic shear zones are discerned

by their monoclinic symmetry in overturned, small-

scale folds and other sense-of-shear indicators (see

Figure 17.35).

Special Interest Box 17.1 Deformation and

strain: patterns, styles, and deﬁnitions

It may be recalled from Section 8.1.2 that deforma-

tion is the result of the stress applied to a body and,

in general, consists of three components—a trans-

lation, a rotation, and a distortion. Distortion is

more commonly called strain and is the change in

shape or volume of the stressed body. A convenient

way of visualizing strain is to compare a reference

sphere in the initial undeformed state with its form

after deformation. If the applied stress is nonhy-

drostatic, a homogeneously deformed state is

represented by a strain ellipsoid (Figure 14.12).

Because it is difﬁcult to show ellipsoids on a

promote slip on at least one slip plane. This concept

accounts for the well known fact that most metals are

more easily strained plastically than rocks. Metals, in

fact, are ductile at atmospheric conditions, whereas

rocks are brittle. Most metal crystals have high sym-

metry, such as the isometric halite in Figure 17.13,

in contrast to major rock-forming minerals, such as

triclinic and monoclinic feldspars. An additional factor

contributing to ductility of metals is their weaker

atomic bonds compared to covalently bonded silicates.

Different rock-forming minerals have different

plasticities. Hexagonal carbonates have enough slip

and twin systems to allow for considerable plasticity in

monomineralic aggregates (marble) over a wide range

of T, P, and strain rate. Although hexagonal quartz

has about as many slip systems as calcite, the Si–O

bonds are much stronger; consequently monomineralic

538 Igneous and Metamorphic Petrology

45°

45°

45°

45°

1

1

1

1

Line

rotation

Line

stretch

PURE SHEAR SIMPLE SHEAR

17.20 Contrast between progressive deformation by coaxial pure shear and noncoaxial simple shear. Plots in lower part of ﬁgure of rotation

and schematic stretching of thick and thin reference lines in deformed objects in upper part disclose different deformation paths for the

two styles of plane strain. Nonetheless, the ﬁnal ﬁnite strain is identical. These models also conform to plane strain in that there are no

changes in their shapes perpendicular to the two-dimensional plane of the diagram.

Undeformed

Homogeneous

deformation

SIMPLE SHEAR

NONCOAXIAL DEFORMATION

Inhomogeneous

deformation

(strain

partitioning)

PURE SHEAR

COAXIAL DEFORMATION

17.21 Schematic contrasts between homogeneous deformation and

inhomogeneous deformation in pure and simple shear. Strain

ellipses could represent pebbles in a metaconglomerate and

band could represent a dike or vein.

aggregates of quartz (quartzite) are much stronger and

less easily strained plastically. Triclinic and monoclinic

feldspars in quartzo-feldspathic rocks are still less

plastic and their ductile strength is greater (Figure 8.8)

because of the paucity of slip systems. This contrast in

the ductility of the principal minerals in mylonitized

granitoids accounts for the typical porphyroclasts of

feldspar in the quartz-rich matrix (Figure 14.16c).

17.2.3 Crystal Defects

All real crystals are defective or imperfect to varying

degrees. Their lattices are never perfect arrays as com-

monly pictured, as for example in mineralogy text-

books. Nonrational boundaries of grains in aggregates

are one type of imperfection (Figure 6.9a). Other de-

fects lie within the crystal volume, distort the regular

periodicity of a lattice, introduce local strain, and

increase the internal stored elastic energy. Point de-

fects (Figure 6.5) lie near or at lattice points and are

essential for intracrystalline diffusion of ions through

the volume of a crystal. Line defects, more commonly

called dislocations, are responsible for the plasticity of

crystals and reduce plastic yield strengths. A disloca-

tion is not simply a linear collection of point defects,

but is instead a special category of irregularity that may

be visualized as a line in the slip plane separating

slipped from unslipped parts of the crystal. The nature

of dislocations and how they relate to slip are shown in

Figures 17.22 and 17.23.

In an ideally perfect crystal lacking dislocations, if

all the atoms along the entire slip plane were to be

moved simultaneously to a new equilibrium position,

considerable stress would be required to break all their

bonds with neighboring atoms. But slip in real crystals

by movement of dislocations requires far less shear

stress, about one thousand times smaller than theore-

tically predicted in perfect crystals, because only a line

of atoms needs to be moved at any instant and because

of the large number of possible dislocations. In real

crystals, slip nucleates in a small region of high stress

concentration and then spreads out in an expanding

loop across the slip plane until it ultimately intersects a

grain boundary where it produces a small step (Figure

17.23b). Propagation of an edge dislocation accompany-

ing slip through a stressed crystal has been likened to

the movement of a caterpillar, which humps up the rear

of its body by moving its tail forward, and then moves

the hump progressively forward to the head. Another

analogy is provided by the old parlor trick of how to

Evolution of Imposed Metamorphic Fabrics: Processes and Kinetics

539

Unstrained

(a)

Edge dislocation(b)

Extra

half plane

Completely

slipped crystal

Edge dislocation

atomic bond ac ﬂips over to bc

Screw dislocation(c)

Completely

slipped crystal

Slip

plane

Screw

dislocation

17.22 Two types of dislocations. These are line defects on a slip plane (dark shaded) between sheared and unsheared crystal. (b) During

leftward (sinistral) shear of the top part of the crystal, the edge dislocation at the bottom edge of the extra half plane of lattice points

(light shaded) moves through the crystal by ﬂipping atomic bonds. (c) During leftward tearing of the upper part of the crystal, a screw

dislocation moves from front to rear through the crystal.

remove the carpet from a room without ﬁrst taking out

all the heavy furniture, including the grand piano. The

easy way is to wrinkle up the carpet along one edge and

move the wrinkle across the room by lifting up corners

of furniture one at a time as necessary. This procedure

would have to be repeated numerous times to remove

the entire carpet from the room, but it does not require

much work (energy) at any instant. A screw dislocation

may be likened to tearing a sheet of paper, starting at

one edge and propagating the tear to the opposite edge.

Dislocations are thermodynamically unstable be-

cause they represent a local high energy region in a

crystal where the atomic bonds are elastically distorted.

But all crystals have them from the time of their

growth; they are built into the crystal from the begin-

ning, like the extra partial rows of kernels in some cobs

of corn. In undeformed crystals of quartz, the “den-

sity” of dislocations, expressed as the total length of

dislocation lines per volume, is about 10

cm/cm

(Fig-

ure 17.23d). Not only does shear stress cause existing

540 Igneous and Metamorphic Petrology

17.23 Dislocations in crystals. (a) Most dislocations in real crystals are not pure edge or screw dislocations but instead are of mixed character.

A dislocation loop nucleated within a nonhydrostatically stressed crystal propagates outward along the slip plane. (b) Along the front and

left side faces of the crystal block the loop is a pure screw and pure edge dislocation. (c) Completely slipped crystal showing offset step

along margin. Never at any time during the slip process is the crystal cleaved in two pieces; that is, it retains cohesion throughout the

plastic deformation. (d) Transmission electron micrograph of moderately strained quartz showing stringlike dislocations (line defects)

that are piled up at subgrain margins, the largest of which in right center is about 4 m across. Photograph courtesy of M. S. Weathers.

Dislocation

loop

Slip

plane

Completely

slipped crystal

(a)

(b)

(c)

(d)

dislocations to move, it also produces new ones at grain

boundaries, cracks, and other sources within the crys-

tal. Applied stress does work on the crystal, part of

which is stored as energy at newly created dislocations.

An intensely deformed quartz crystal, for example,

might contain 10

cm/cm

dislocations. As the dislo-

cation density increases during progressive deforma-

tion, greater shear stress is required to move interfering

piled up dislocations and to slip parts of the crystal that

have been effectively strengthened by the tangled dis-

locations. This phenomenon of continued plastic strain

requiring ever higher applied stress is called strain, or

work, hardening and can be seen in the stress–strain

curve for ductile behavior of real rocks and minerals

compared to the mathematical idealized plastic re-

sponse in Figure 8.3b.

17.2.4 Recovery during Dislocation Creep

At high temperatures above about 0.5 T

, where T

the melting T of the crystal in Kelvins, rates of atomic

diffusion become appreciable so that dislocations can

be mobilized, dislocation density reduced, and a more

stable, lower energy state attained by the crystal. This

recovery phenomenon counteracts work hardening

and allows for a more or less continuous deformation

to occur at more or less uniform applied nonhydrosta-

tic stress in what is called dislocation creep. The rate at

which new dislocations are generated in the stressed

crystal is approximately balanced by their elimination

in dislocation creep.

One recovery mechanism involves dislocation climb,

where an edge dislocation moves perpendicular to the

slip plane by the addition or removal of atoms, one at

a time, by diffusion along the edge of the extra half-

plane (Figure 17.24). Two dislocations of opposite sign

climbing toward each other are annihilated.

In other recovery processes, new grains are created

from pre-existing strained ones during deformation.

This can occur in two ways. In the ﬁrst, climbing dislo-

cations of the same sign accumulate to form dislocation

walls lying at low angles to the less strained lattice on

either side. These walls deﬁne subgrains (Figures

17.18b and 17.25). Several subgrains are more stable

than a whole strained grain because the organized dis-

locations and lattice between them have lower energy.

Somewhat reoriented subgrains are responsible for

most of the undulatory extinction seen in strained sili-

cates under cross-polarized light in thin section. In the

second process of syntectonic, or dynamic, recrystal-

lization, dislocations in a segment of the margin of a

strained grain are mobilized and migrate to form small-

scale bulges that expand into adjacent grains, producing

irregular interlocking knobby outlines. New, small,

unstrained grains may also nucleate, initially along

margins of highly strained deformation bands and

ribbons, adding to the irregularity of the sutured grain

boundaries (Figures 17.26b and 15.17). Eventually,

small, unstrained polygonal grains may consume the

entire higher energy ribbon or subgrain (Figure 17.26c).

Thus, during dynamic recrystallization, deformation

is contemporaneous with recrystallization; as new

unstrained grains are created, they in turn become

deformed and are replaced by strain-free grains, and

so on. Very large amounts of strain become possible.

Dynamic recrystallization is especially evident in mylon-

ites developed in ductile shear zones.

With increasing magnitude of the shear stress, dis-

location density increases and the size of subgrains and

dynamically recrystallized grains decreases (Twiss and

Moores, 1992, p. 409; Snoke et al., 1998, p. 13). If the

magnitude of nonhydrostatic stress driving dislocation

creep increases, recovery processes will adjust to main-

tain a steady state by gathering the faster-forming

dislocations into more closely spaced subgrain walls;

the dislocations have a smaller distance to migrate

and the recovery rate is increased. Calibration against

experimentally deformed rocks suggests that stresses

on the order of 1 kbar are responsible for dynamically

recrystallized rocks in mylonitic ductile shear zones

(Figure 17.26). Dislocation creep produces signiﬁcant

strain weakening in initially coarse-grained aggregates

and tends to partition strain (Figure 17.21) into ﬁner-

grained zones that can evolve into mylonite (Snoke

et al., 1998, p. 13).

Recovery and recrystallization can continue after the

high-T grain aggregate has ceased to be deformed. This

annealing that can occur isothermally is driven by the

residual energy in the still strained grains and creates

new, strain-free polygonal grains with straight bound-

aries. In single-phase aggregates of minerals low in the

crystalloblastic series, such as quartz and feldspars,

grain boundaries meet at 120° triple junctions.

17.2.5 Hydrolytic Weakening of Silicates during

Plastic Slip

Widespread textural evidence of intracrystalline plastic

slip in silicate rock-forming minerals, especially quartz,

in tectonites had been observed for many decades,

but until the mid-1960s quartz had resisted plastic

deformation in the laboratory; in fact, it displayed

exceedingly large brittle strengths in excess of 20 kbar

at elevated P and T. This paradox was resolved when

D. T. Griggs and J. D. Blacic discovered the phenome-

non of hydrolytic weakening (Griggs, 1967) in quartz

crystals deformed in an apparatus (Figure 8.5) using

talc as a surrounding conﬁning-pressure medium.

(At high T, the talc dehydrated and the liberated water

diffused into the deforming quartz, unlike in previous

experiments using “dry” pressure media). Subse-

quently, deformation experiments combined with opti-

cal and transmission electron microscopy and infrared

spectroscopy have been employed to investigate the

Evolution of Imposed Metamorphic Fabrics: Processes and Kinetics

541

hydrolytic weakening phenomenon. Silicate minerals

such as quartz, feldspar, and olivine are nominally

anhydrous, but infrared spectroscopy reveals that

signiﬁcant concentrations of water can be dissolved,

apparently in both molecular form and as (OH)



ions,

even under the high T and P that prevail in the mantle

(Section 11.2.1). Laboratory deformation of hydrous

quartz crystals containing as much as 0.13 wt.% H

(9000 H/10

Si atoms) occurs plastically above 400°C

at stress differences about an order of magnitude less

than required to deform dry quartz crystals. These

contrasts in strength are shown in Figure 8.8.

There has been considerable debate regarding ex-

actly how hydrolytic weakening takes place in “wet”

nonhydrostatically stressed silicates. The residence of

water may differ in deformed single quartz crystals

versus aggregates of grains. In quartzites, water can

exist as adsorbed molecules on grain boundaries and

as intergranular ﬁlms and within individual grains in

aqueous ﬂuid inclusions (Special Interest Box 16.3), as

structural water at defects, and bonded to Si and O

ions as H



and (OH)



ions. Post et al. (1996) found

the fugacity of water to be the most important fac-

tor controlling dislocation creep in quartzite and

542 Igneous and Metamorphic Petrology

Atom above extra half plane of atoms

Atom below

Extra

half

plane

(a)

(b)

(c)

17.24 Recovery by dislocation climb and annihilation. (a) Dislocation climb is accomplished by losses or gains of atoms one at a time by diffu-

sion along a jogged half plane edge. An atom from the edge of the extra half plane of atoms has diffused into a nearby vacancy, causing

the edge dislocation to climb upward. (b) The reverse occurs, extending the area of the half plane and causing the dislocation to climb

downward. (c) Two edge dislocations, now viewed from their ends, climbing toward one another are annihilated and the two extra half

planes become one whole. T symbolizes the end of an edge location. Redrawn from Nicolas and Poirer (1976).

Evolution of Imposed Metamorphic Fabrics: Processes and Kinetics

543

(a)

(b)

0.010mm

17.25 Low-angle dislocation walls and subgrains. (a) Schematic of strained lattice with edge dislocations of the same sign (half planes all

extend in the same direction) concentrated into low-angle dislocation walls between three subgrains (shaded). Compare with bent

lattice in Figure 17.18 and with transmission electron photomicrograph of strained quartz grain that has dislocations accumulated along

subgrain margins (Figure 17.23d). (b) Photomicrograph of quartz subgrains viewed in cross-polarized light under a petrographic

microscope. (See also Figures 17.12b and 17.26a, b.)

fugacities of hydrogen and oxygen and hydrogen activ-

ity unimportant.

Locally higher water fugacities, perhaps related to

a passing front, or wave, of thermally induced dehy-

dration reactions, could weaken quartz-rich rocks and

allow partitioning of deformation into thin shear zones,

creating mylonite. Alternatively, dry, strong, high-grade

rocks might be invaded in some manner by aqueous

ﬂuids that would hydrolytically weaken them and

allow focused deformation. In either case a change in

the state of stress or increasing stress magnitude is not

required to create the shear zone, only inﬂux of water.

17.2.6 Power Law in Ductile Flow

Experiments on ductile ﬂow of rocks by diffusive

transfer and plastic dislocation mechanisms in the

laboratory cannot exactly duplicate conditions in real

rock bodies because of the many variables existing in

geologic systems. One variable that is vastly different

is the time scale of deformation. Deformation in the

laboratory is accomplished over time periods of weeks

or, at most, months at strain rates, d/dt, of 10

7

/s.

But judging from rates of lithospheric plate movement,

strain rates in natural rock bodies are more likely to be

on the order of 10

12

–10

15

/s. In order to evaluate the

effects of these orders-of-magnitude slower strain rates,

the steady-state ﬂow or creep seen in experimentally

deformed rocks and minerals is ﬁtted to a exponential

rate equation, or power law (Paterson, 1987) of the

form

d/dt  A

exp(E

/RT) 17.1

where A is the rate constant,  is the stress difference

producing strain, n is a number between 2 and 8, E

the activation energy for steady-state ﬂow, R is the gas

constant, and T is absolute temperature. A, n, and E

are empirical parameters that can be evaluated from

laboratory experiments (Twiss and Moores, 1992,

Chapter 18). Using this equation, the ductile strength

of rocks and minerals can be extrapolated through the

lithosphere for a geologically reasonable strain rate

(Figure 8.8). It is quite obvious that rocks at elevated P

and T in the lower crust and mantle that are strained

very slowly in a more or less viscous manner do so

under very small stress differences. Large nonhydrostatic

stresses cannot be sustained in these rocks.

17.3 INTERACTIONS BETWEEN

DEFORMATION, CRYSTALLIZATION,

AND FLUIDS IN TECTONITES

The previous two major sections examined recrystalliza-

tion and deformation as distinct and essentially separate

processes, as they may be in some metamorphic systems.

However, in the most widespread class of metamorphic

rocks and fabrics—tectonites and tectonite fabric—

these two processes were broadly contemporaneous.

Nonhydrostatic stresses have no direct bearing on

mineral equilibria (as once believed) but the magnitude

of these stresses in concert with strain rate, P, T, and

volatile fugacities during deformation can strongly

inﬂuence how—and even if—a rock recrystallizes. For

example, it is not unusual to ﬁnd, in an otherwise

penetratively deformed metamorphic terrane, pods of

weakly or nondeformed rock (Figure 15.19) in which

a peak metamorphic mineral assemblage has not been

retrogressively overprinted or destroyed as in the sur-

rounding deformed rocks.

Interactions between deformation and recrystal-

lization can go both ways (Rutter and Brodie, 1995).

Deformation can enhance mineral reactions through:

1. Reducing grain size, increasing surface-area reactivity.

2. Increasing grain energy through work done on

grains, increasing dislocation density, thus making

grains less stable and more amenable to dissolution

in the presence of a ﬂuid or to other modes of

breakdown with other reactants. Also, providing

high-energy sites for nucleation where dislocations

have migrated to grain margins.

3. Enhancing rock permeability in the intergranular

network and in hydraulic fractures (Holness, 1997).

Transient pressure gradients can be established by

opening and closing of cracks, driving ﬂuid ﬂow

(dilatancy pumping, Section 16.7.2). Moving ﬂuids

not only catalyze mineral reactions but are essential

to their progress.

4. Increasing the T of the rock by shear heating (Sec-

tion 11.1.1).

544 Igneous and Metamorphic Petrology

(a)

0 0.1mm

(b)

0 0.1mm

17.26 Progressive mylonitization of the basal quartzite unit at the

Stack of Glencoul, Moine thrust zone, Scotland, viewed

under cross-polarized light. (a) Sample 110 m from the thrust

contact between quartzite and overlying Moine schist has

5% new recrystallized grains along margins of relict grains.

(b) Sample 5–10 m from the thrust contact has elongated rib-

bons (deformation bands) of highly plastically slipped quartz

grains that show undulatory extinction. Dynamic recrystalliza-

tion involving grain boundary migration has produced

knobby, or sutured, margins on quartz ribbons and local

domains of small polygonal grains that are essentially strain

free. (c) At the thrust contact, the quartzite is composed

almost entirely of dynamically recrystallized grains of about

the same size as in (b). For a discussion of the quartz-lattice

fabric in these mylonitic rocks see Law (1990). Photographs

courtesy of M. S. Weathers, from Weathers et al. (1979).

On the other hand, mineralogical factors can

inﬂuence deformation in the following ways:

1. Lower grade metamorphic rocks may be weaker

than higher grade. Phyllosilicates are weaker than

chemically equivalent, but volatile-free, assem-

blages of feldspars, pyroxenes, and so on. Rocks

rich in phyllosilicates, regardless of their grade, as

well as rocks that contain abundant carbonates and

quartz, can be weaker than feldspathic rocks.

2. Mineral composition governs the mode of ductile

deformation. Minerals generally more soluble in

the presence of ﬂuids under metamorphic condi-

tions, such as carbonate minerals and quartz, are

more susceptible to pressure solution. Carbonate

minerals readily deform by twinning and disloca-

tion glide. Quartz is one of the more plastic min-

erals in rocks, whereas feldspars possess fewer slip

systems and do not readily deform plastically in

felsic rocks.

3. Most important, prograde devolatilization reac-

tions provide what is probably the major source of

ﬂuids in the deeper continental crust. Fluids, espe-

cially water, are so important in fabric development

that their role is considered further in the following

section.

17.3.1 Role of Fluids in Tectonite Fabric

Development

Fluids are liberated throughout most of prograde

metamorphism up to granulite-facies conditions in vir-

tually all chemical rock groups. Pelites and hydrated

metabasites release water and calc–silicate rocks mixed

C–O–H ﬂuids. The amount and fate of the liberated

volatiles depends on the depth (and P) at which the

devolatilization reactions occur, their nature, modal

composition of the rock, and its permeability. Small

amounts of ﬂuid may only reside along grain edges and

corners but nonetheless could enhance diffusion and

pressure solution. Larger released amounts, especially

where produced by mixed-volatile ﬂuid reactions in

calc–silicate rocks, can enhance intergranular perme-

ability and reduce the effective stress in the rock (Sec-

tion 8.2.1). This embrittles the rock, possibly creating

hydraulic fractures (Figure 8.2) that can be ﬁlled with

precipitates from the ﬂuid, forming veins of various

mineral compositions. The effects of repeated cracking

and vein ﬁlling—the crack–seal process—is evident in

many veins (Passchier and Trouw, 1996, Chapter 6).

Reduced frictional resistance in ﬂuid overpressured

shear zones may allow grain boundary sliding or even

cataclasis, resulting in dilatance and increase in perme-

ability (de Roo and Williams, 1990). These conduits for

advective ﬂow allow solution transfer, metasomatic

alteration, and other ﬂuid-related processes. If the

ﬂuid eventually escapes from the rock system, ductile

conditions can be restored as T continues to rise in

dehydrated rocks. Veins formed by brittle fracturing

might become folded or otherwise deformed. These

changing metamorphic conditions can be viewed in a

spatial setting in Figure 17.27.

In some cases, undeformed masses of rock sur-

rounded by pervasively deformed rock (as cited above)

escape deformation because of restricted access of

water. Rock and mineral strengths are signiﬁcantly

reduced in the presence of water (Figure 8.8). Higher

water fugacities promote hydrolytic weakening in

silicates and ductile ﬂow that otherwise might not

happen by dislocation creep under a particular state of

stress.

17.3.2 Timing of Deformation and Crystallization:

Larger Scale Implications

Determination of the chronology of deformation and

crystallization during the development of tectonites is

of fundamental signiﬁcance because of its bearing on

larger scale metamorphism and tectonism in the evolu-

tion of orogens. In tectonites, the effects of different

styles and orientations of deformation together with

the effects of mineral equilibration under different

metamorphic conditions are overprinted on one an-

other. Superposition of successive overprints may not

completely erase the effects of prior events in poly-

metamorphism so all are preserved to varying degrees

in the rock fabric and structure. Unraveling these

overprints and elucidating their chronology is a major

challenge, particularly when it comes to translating

fabric features into distinct episodes of tectonism and

heating/burial that shaped an evolving, continent-scale

orogen over tens or perhaps hundreds of millions of

years.

Establishing relative chronologies of deformation

and recrystallization in metamorphic rocks, our focus

in this section, depends on accurate reading and inter-

pretation of critical fabric elements corresponding to

different overprints. The next step, establishing an

absolute chronology using the tools of isotope geo-

chemistry, must be based on datable mineral grains,

but their time of growth or destruction relative to the

tectonism must be determined by the petrologist for

the age determinations to have any relevance in the

chronologic framework of the evolving orogen.

In an evolving single-cycle orogen, the major fabric-

forming deformation typically precedes, or is concur-

rent with, the peak thermally induced recrystallization

of the regionally metamorphosed rocks (Yardley,

1981). During heating as the peak metamorphism is

approached, dehydration mineral reactions promote

ductile deformation in the prevailing nonhydrostatic

stress ﬁeld. Continuing recrystallization at the thermal

peak overprints, to varying degrees, even destroying,

features of the ductile deformation that occurred en

Evolution of Imposed Metamorphic Fabrics: Processes and Kinetics

545

route. However, larger scale structures, such as folds,

boudinage, and coarse compositional layering, are typ-

ically larger than the sphere of diffusional transfer and

dislocation migration and so mostly escape oblitera-

tion. Once the metamorphic terrane has been largely

dewatered through prograde reactions, and may be

also cooling, pervasive ductile deformation of the

stronger rocks is limited, if not completely curtailed.

However, local retrograde metamorphism can occur

where ﬂuids are introduced, perhaps initially along

fractures; accompanying hydrating mineral reactions

and the overall weakening effect of water on the rock

cause further strain to be focused or partitioned so as

to produce local ductile deformation in shear zones.

Before we attempt to deal with speciﬁc fabrics from

which relative timing can be determined, two caveats

should be pointed out. First, deformation in inhomo-

geneous anisotropic metamorphic rocks is itself inho-

mogeneous and, therefore, its expression depends on

the scale of observation. Strain partitioning can result

in contrasting expressions of tectonite fabric in differ-

ent domains of a rock (Figures 17.21 and 14.15). For

example, folia of phyllosilicates can generally accom-

modate more strain in a particular state of stress than

can neighboring quartzo-feldspathic layers, especially

if the latter are dry. Where there is a local inﬂux of

water into initially dry rocks, for whatever reason, they

can be weakened. Because deformation can be speciﬁc

to particular mineral and modal compositions and ﬂuid

conditions a particular tectonic episode may not be

recorded in all rocks.

Second, deformation–crystallization interactions

observable on the scale of a thin section may be

difﬁcult to translate accurately to the larger scale of an

entire metamorphic terrane unless numerous samples

are examined carefully.

17.3.3 Pre-, Syn-, and Postkinematic Fabrics

There are three possibilities as to when recrystallization

takes place in relation to the time of deformation; it

may be prekinematic, synkinematic, or postkinematic.

(Some petrologists use the term tectonic rather than

kinematic in this context; however, differential move-

ment of rock masses, as in local shear zones, is certainly

a kinematic event but it may not correspond to a re-

gional tectonic event of signiﬁcance throughout an

orogen.) Establishing which of these three chronolo-

gies applies in a metamorphic rock hinges on accurate

recognition of critical cross-cutting relations, over-

growths, and contrasts in expression and orientation in

its fabric (Passchier and Trouw, 1996).

Much of this critical chronologic record is preserved

in the pattern and types of mineral inclusions within

poikiloblasts relative to the texture of the matrix

surrounding them. As fruitful as these porphyroblast–

matrix relations can be, many provide no unambigu-

ous, single interpretation (Vernon, 1988). In addition,

a single thin section may not accurately portray the

546 Igneous and Metamorphic Petrology

Crack-seal

vein

Stronger, higher grade

rock

Hydraulic

fractures

Devolatilized

mineral

assemblage

Reactant

volatile-

bearing

mineral

assemblage

Ductile zone

Depth (P)

interval

Reaction

Geotherm

17.27 Strain partitioning into a 2–3 km thick interval of the stressed continental crust where release of volatile ﬂuids from prograde mineral re-

actions has reduced the ductile strength. Fluids escaping upwards promote hydraulic fracturing and crack–seal vein formation,

featured in enlarged inset in upper left, involving successive fracturing and vein ﬁlling. As the crust is continually heated from below

the devolatilizing reactions progress upwards so that earlier planar veins emplaced under brittle conditions are overtaken and deformed

by the upward advancing ductile shear. Redrawn from Rutter and Brodie (1995).