1
1. What is the problem? Garnet is not uncommon in granitoid rocks, but its origin is contentious. What mechanisms have been proposed? Essentially, 4 principal scenarios... In a review, Clarke (1981) listed the following mechanisms: Derivation as a refractory restite phase from the zone of partial melting. Derivation as refractory xenocrysts from porphyroblasts in high-level pelitic country rocks. Generation in the marginal facies of a granite body as a result of reaction between magma and pelitic xenoliths rich in Al and Mn relative to the melt. Reaction between early-formed phases and silicate melt, for example: Liq + Bt Grt + Ms or Liq + Als + Bt Grt . Nucleation directly from the silicate melt, especially in aplites and pegmatites, as late-stage differentiation increases the Mn / (Mn + Fe2 + + Mg) ratio sufficiently to stabilise spessartine-rich garnet. l l l l l Þ Þ 1. Xenocrysts. 2. Garnet crystals grown by solid-state replacement. 3. Garnet formed by reaction between solid or early–formed phases and granitoid melt. 4. Garnet crystallized directly from the granitoid melt. 6. Discussion Did garnet form as xenocrysts? Did garnet grow by solid-state replacement? Did garnet grow by reaction between melt & Bt? Did garnet grow from the granitoid melt? For Against For Against For Against For Against Xenocrysts of other phases are present Xenocryst grain sizes are much larger than those included in garnet. Faceted outlines - no evidence of reaction with melt. Similar quartz inclusion and groundmass grain sizes and shapes. garnet appears to cut across xenocryst outlines. Poikilitic texture. Qtz inclusion sizes similar to Qtz in groundmass. Absence of feldspar and muscovite inclusions. If the clustered garnets are skeletal. If the garnets are not discrete grains. Qtz inclusion SPO different to groundmass SPO. No evidence of biotite or aluminium silicate having been present. No evidence of muscovite intergrown with garnet. Subhedral and euhedral grain shapes. Poikilitic texture. Similar chemical composition to published “granite garnets”, high MnO contents and normal zoning. If garnet clusters are discrete grains, possibility of glomerocrystic aggregation? Absence of feldspar and muscovite inclusions. Glomerocrysts – poikilitic texture suggests granite largely crystallised making movement of garnets through melt difficult? clustered but are (the reactions: Absence of biotite. Absence of aluminium silicate. Liq + Bt Grt + Ms or Liq + Als + Bt Grt) Þ Þ 7. Conclusions Decision Time - melt or solid state - or both? Putnis (2002, Min Mag): “The precondition for a dissolution- reprecipitation reaction is merely that a solid phase or phase assemblage is out of equilibrium with a fluid phase, and that re- equilibration may be achieved by the crystallisation of a “different” solid phase. This difference may be chemical or textural, and the fluid may be a simple aqueous solution, a melt or anything in between.” Both solid and liquid? Perhaps the Mn-rich garnet crystallises from a final stage granitoid melt/fluid, but rather than produce lots of small spatially separate crystals it has made a few large ones. This will have involved dissolution (Pl+Kfs+Ms) and precipitation (Grt), supported by the “clean” Grt in contact with groundmass and xenocrystic Pl, and favouring a skeletal geometry. If the garnet is skeletal, might it have grown along end-stage melt/fluid l l l Origin of poikilitic garnet in a leucogranite dike Dave J. Prior, Aziz Rahimi-Chakdel*, Alan P. Boyle**. Department of Earth & Ocean Sciences, University of Liverpool, Liverpool, L69 3GP, U.K. * now at ** email: [email protected] Department of Geology, Faculty of Science, University of Agricultural Science, GORGAN – IRAN. 4. Garnet chemistry Garnet shows zoning profiles with high Mn (Sps) between 40-50 mol% in cores and significant Ca- contents (Grs) between 12-25 mol%. Fe-content (Alm) is antipathetic with Sps. Mg-content (Prp) is very low. In general, garnet compositions are similar to those reported from granites. Fe Mn Mg This study Igneous garnets defined by Miller & Stoddard (1981) Grt 1 X Y Grt 1 X Y 3 14 m Grt 1 0 10 20 30 40 50 60 0 300 600 900 1200 Distance in microns XMol%component X Y 0 10 20 30 40 0 100 200 300 Distance in microns Grt2-a X Y Grt2-a X Y Grt2-a 50 60 X Y Grt2-a 50 60 SPS Alm Gru Prp SPS Alm Gru Prp X Y 100 m m Fe Mn 100 m m 2. Geological setting, petrography & chemistry The sample studied comes from a Caledonian 1-10 metre wide peraluminous leucogranite dike at St. Ruuan on the Isle of Man. It is characterised by three main textural components: 1. an equigranular sub-100 m groundmass with m a poorly defined shape fabric; 2. xenocrysts and aggregates of coarse, >1mm, Qtz, Pl, Ksp and Ms; 3. garnet grains typically showing faceted margins and tending to occur in >1mm clumps or clusters. The garnets are strongly poikilitic with included grains being almost exclusively quartz, but inclusion free where in contact with xenocrystic plagioclase. Composition SiO 2 74.63 TiO 2 0.04 Al 2 O 3 14.96 Fe 2 O 3 0.51 MgO 0.03 MnO 0.04 CaO 0.67 Na 2 O 5.41 K 2 O 3.62 P 2 O 5 0.05 LOI* 0.48 Total 100.44 A/CNK = 1.54 Groundmass Mineral Mode Qtz = 37% Pl = 28% Ksp = 25.5 Ms = 6% Bt = 1% Grt = 1.5% Acc. = 1% 0.1 Gpa 0.2 Gpa 0.5 Gpa 1 Gpa 2 Gpa 3 Gpa S-type I-type M-type A-type Avg.Crust Avg.Upper Crust Avg.lower Crust Modern Archean Plagiogranite Nigeria Ascension + + SRMGD + Ternary minimum Ternary eutectic + Ternary minimum + + + Or Qtz Ab Foxdale Granite Calf of Man St Michael’s Island Port St Mary Port Cornaa Dhoon Granite Douglas Ramsey Peel 70 80 90 20 30 40 1 2 3 7 4 5 6 Fault Post-Silurian Tract boundary Major intrusions Niarbyl Fm (Wenlock) Key to other symbols: Manx Group stratigraphy not formalised 5km N Manx Group Lithostratigraphical units in tracts 1-7: Microgranite sample locality Port Enin Creggan Mooar and Glion Cam Lady port Santon and Lonan Mull Hill and Port Erin Maughold, Creg Agneash and Ny Garvain Injebreck and Barrule Injebreck and Glen Rushen 4 5 3 1 2 4 4 6 7 PPL & XPL Images 1mm Grt Pl 1mm Grt Pl Qtz Qtz l l l Xenocrystic phases show undulose extinction, as does quartz in the groundmass. Quartz inclusions in garnet do not show undulose extinction. Garnet appears to partly replace plagioclase. Quartz inclusions look similar size to matrix quartz grains. High SiO content suggests Qtz may have been first phase on the liquidus as also suggested by groundmass mode. However, need to be careful that composition and mode not affected by addition of xenocrysts... 2 3. Quartz size & shape data Quartz crystal size distributions (CSDs) and shapes were measured in the groundmass and in garnet inclusions to investigate relative timing of garnet growth. l l Groundmass quartzes show a distinct SPO, whereas inclusion quartzes do not. SPO function of flow during emplacement or subsequent deformation? l l l CSD analysis reveals that groundmass and inclusion quartz have similar size/population slopes, but inclusion quartzes extend to finer sizes. No evidence for garnet attempting to include >1mm xenocrystic quartz Finer grain sizes in inclusions are not a simple function of sequential entrapment during quartz growth Quartz Grain CSDs 0 2 4 6 8 10 12 0 0.2 0.4 0.6 0.8 Length (mm) Ln(n) Inclusion Groundmass Population density Quartz Grain SPOs 0 0 Inclusion N=221 Groundmass N=324 Grt OC image of Qtz inclusions Note that there is no simple relation between size (or shape) of Qtz crystals and distance from the core. l Garnet growth is after groundmass quartz growth, but before the development of Qtz SPO and undulose extinction. 5. Garnet & quartz crystal orientations Electron backscatter diffraction was used to determine crystallographic orientations in garnet and in both groundmass and inclusion quartzes. Groundmass quartz CPOs Groundmass N= 341 Inclusion N= 400 ZO XO YO Inclusion quartz CPOs Quartz crystallographic orientations were measured for 400 inclusions in the two labelled garnets above, and for 341 quartz grains in the surrounding groundmass. For both datasets the crystallographic preferred orientation (CPO) is uniform. l l The development of an SPO has not resulted in any significant CPO. The inclusion forming process has not resulted in any “selection” of grains on a crystallographic basis The clumped garnets have similar crystal orientation… l l Discrete glomerocrystic grains or connected as skeletal/dendritic growths? Small size of outliers favours discrete grains? Crystal orientations of 15 clumped garnets 1mm Grt outlier circled in {100} pole figure ZO XO YO BSE OC Clumped or clustered garnets have similar crystallographic orientation, suggesting: l l they either “know” something about each other, or they are interconnected in 3D

Origin of poikilitic garnet in a leucogranite dikepcapboyle/home/2003-AGU-magmatic-grt-incs.pdf · 2003-12-04 · The garnets are strongly poikilitic with included grains being almost

  • Upload
    others

  • View
    4

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Origin of poikilitic garnet in a leucogranite dikepcapboyle/home/2003-AGU-magmatic-grt-incs.pdf · 2003-12-04 · The garnets are strongly poikilitic with included grains being almost

1. What is the problem?Garnet is not uncommon in granitoid rocks, but its originis contentious.

What mechanisms have been proposed?

Essentially, 4 principal scenarios...

In a review, Clarke (1981) listed the followingmechanisms:

Derivation as a refractory restite phase from the zone ofpartial melting.Derivation as refractory xenocrysts from porphyroblasts inhigh-level pelitic country rocks.Generation in the marginal facies of a granite body as aresult of reaction between magma and pelitic xenoliths richin Al and Mn relative to the melt.Reaction between early-formed phases and silicate melt,

for example: Liq + Bt Grt + Ms or Liq + Als + Bt Grt .Nucleation directly from the silicate melt, especially inaplites and pegmatites, as late-stage differentiationincreases the Mn / (Mn + Fe2+ + Mg) ratio sufficiently tostabilise spessartine-rich garnet.

� �

1. Xenocrysts.2. Garnet crystals grown by solid-state replacement.3. Garnet formed by reaction between solid or early–formed

phases and granitoid melt.4. Garnet crystallized directly from the granitoid melt.

6. DiscussionDid garnet form as xenocrysts?

Did garnet grow by solid-statereplacement?

Did garnet grow by reaction between melt& Bt?

Did garnet grow from the granitoid melt?

For

Against

For

Against

For

Against

For

Against

Xenocrysts of other phases are present

Xenocryst grain sizes are much larger than those included in garnet.Faceted outlines - no evidence of reaction with melt.Similar quartz inclusion and groundmass grain sizes and shapes.garnet appears to cut across xenocryst outlines.

Poikilitic texture.Qtz inclusion sizes similar to Qtz in groundmass.Absence of feldspar and muscovite inclusions.If the clustered garnets are skeletal.

If the garnets are not discrete grains.Qtz inclusion SPO different to groundmass SPO.

No evidence of biotite or aluminium silicate having been present.No evidence of muscovite intergrown with garnet.

Subhedral and euhedral grain shapes.Poikilitic texture.Similar chemical composition to published “granite garnets”, high

MnO contents and normal zoning.If garnet clusters are discrete grains, possibility of glomerocrysticaggregation?

Absence of feldspar and muscovite inclusions.Glomerocrysts – poikilitic texture suggests granite largely crystallised

making movement of garnets through melt difficult?

clustered but are

(the reactions:Absence of biotite.Absence of aluminium silicate.

Liq + Bt Grt + Ms or Liq + Als + Bt Grt)� �

7. ConclusionsDecision Time - melt or

solid state - or both?

Putnis (2002, Min Mag):“The precondition for a dissolution-reprecipitation reaction is merely that asolid phase or phase assemblage is out ofequilibrium with a fluid phase, and that re-equilibration may be achieved by thecrystallisation of a “different” solid phase.This difference may be chemical ortextural, and the fluid may be a simpleaqueous solution, a melt or anything inbetween.”

Both solid and liquid?Perhaps the Mn-rich garnet crystallisesfrom a final stage granitoid melt/fluid, butrather than produce lots of small spatiallyseparate crystals it has made a few largeones.This will have involved dissolution(Pl+Kfs+Ms) and precipitation (Grt),supported by the “clean” Grt in contact withgroundmass and xenocrystic Pl, andfavouring a skeletal geometry.If the garnet is skeletal, might it havegrown along end-stage melt/fluid

Origin of poikilitic garnet in a leucogranite dikeDave J. Prior, Aziz Rahimi-Chakdel*, Alan P. Boyle**.Department of Earth & Ocean Sciences, University of Liverpool, Liverpool, L69 3GP, U.K.* now at

** email: [email protected] of Geology, Faculty of Science, University of Agricultural Science, GORGAN – IRAN.

4. Garnet chemistryGarnet shows zoning profiles with high Mn (Sps) between 40-50 mol% in cores and significant Ca-contents (Grs) between 12-25 mol%. Fe-content (Alm) is antipathetic with Sps. Mg-content (Prp) isvery low. In general, garnet compositions are similar to those reported from granites.

Fe

Mn

Mg

This study

Igneous garnets defined byMiller & Stoddard (1981)

Grt 1X

Y314 m

Grt 1X

Y314 m

Grt 1

0

10

20

30

40

50

60

0 300 600 900 1200

Distance in microns

XM

ol%

co

mp

on

en

t X Y

0

10

20

30

40

0 100 200 300

Distance in microns

Grt 1

0

10

20

30

40

50

60

0 300 600 900 1200

Distance in microns

XM

ol%

co

mp

on

en

t X Y

0

10

20

30

40

0 100 200 300

Distance in microns

Grt2-a X

Y

Grt2-a X

Y

Grt2-a

50

60 X YGrt2-a

50

60

SPSAlmGruPrp

SPSAlmGruPrp

X Y

100 m�

Fe

Mn

100 m�

2. Geological setting, petrography & chemistryThe sample studied comes from a Caledonian 1-10 metre wide peraluminous leucogranite dike at St.Ruuan on the Isle of Man. It is characterised by three main textural components:

1. an equigranular sub-100 m groundmass with� a poorly defined shape fabric;2. xenocrysts and aggregates of coarse, >1mm, Qtz, Pl, Ksp and Ms;3. garnet grains typically showing faceted margins and tending to occur in >1mm clumps or

clusters. The garnets are strongly poikilitic with included grains being almost exclusively quartz,but inclusion free where in contact with xenocrystic plagioclase.

Composition

SiO2 74.63

TiO2 0.04

Al2O3 14.96

Fe2O3 0.51

MgO 0.03

MnO 0.04

CaO 0.67

Na2O 5.41

K2O 3.62

P2O5 0.05

LOI* 0.48

Total 100.44

A/CNK = 1.54

GroundmassMineral ModeQtz = 37%Pl = 28%Ksp = 25.5Ms = 6%Bt = 1%Grt = 1.5%Acc. = 1%

0.1 Gpa0.2 Gpa

0.5 Gpa1 Gpa

2 Gpa3 Gpa

S-type I-type M-type A-type

Avg.Crust Avg.Upper Crust Avg.lower Crust

Modern ArcheanPlagiograniteNigeria

Ascension

+

+

SRMGD

+ Ternary minimum

Ternary eutectic

+ Ternary minimum

++

+

Or

Qtz

Ab

Foxdale Granite

Calf of Man

St Michael’s IslandPort St Mary

PortCornaa

DhoonGranite

Douglas

Ramsey

Peel

70

80

90

20 30 40

1 2 3

74 5 6

Fault

Post-Silurian

Tract boundary

Major intrusions

Niarbyl Fm (Wenlock)

Key to other symbols:

Manx Group stratigraphy

not formalised

5km

N

Manx Group Lithostratigraphical units in tracts 1-7:

Microgranite samplelocality

Port Enin

Creggan Mooar andGlion Cam

Ladyport

Santon andLonan

Mull Hill andPort Erin

Maughold, Creg Agneashand Ny Garvain

Injebreck andBarrule

Injebreck andGlen Rushen

4

5

3

1

2 4 4

6

7

PPL & XPL Images

1mm

Grt

Pl

1mm

GrtPl

Qtz Qtz

Xenocrystic phases showundulose extinction, asdoes quartz in thegroundmass. Quartzinclusions in garnet donot show unduloseextinction.Garnet appears to partlyreplace plagioclase.Quartz inclusions looksimilar size to matrixquartz grains.

High SiO content suggests Qtzmay have been first phase onthe liquidus as also suggestedby groundmass mode. However,need to be careful thatcomposition and mode notaffected by addition ofxenocrysts...

2

3. Quartz size & shape dataQuartz crystal size distributions (CSDs) and shapes weremeasured in the groundmass and in garnet inclusions toinvestigate relative timing of garnet growth.

Groundmass quartzes show a distinct SPO,whereas inclusion quartzes do not.SPO function of flow during emplacement orsubsequent deformation?

CSD analysis reveals that groundmass andinclusion quartz have similar size/populationslopes, but inclusion quartzes extend to finersizes.No evidence for garnet attempting to include>1mm xenocrystic quartzFiner grain sizes in inclusions are not a simplefunction of sequential entrapment duringquartz growth

Quartz Grain CSDs

0

2

4

6

8

10

12

0 0.2 0.4 0.6 0.8

Length (mm)

Ln

(n)

Inclusion

Groundmass

Po

pu

latio

nd

en

sity

Quartz Grain SPOs0 0

Inclusion

N=221

Groundmass

N=324

Grt

OC image of Qtz inclusions

Note that there is no simple relationbetween size (or shape) of Qtzcrystals and distance from the core.

� Garnet growth is after groundmass quartz growth, but before

the development of Qtz SPO and undulose extinction.

5. Garnet & quartz crystal orientationsElectron backscatter diffraction was used to determine crystallographic orientations in garnetand in both groundmass and inclusion quartzes.

Groundmass quartz CPOs

GroundmassN= 341

InclusionN= 400

ZO

XOYO

Inclusion quartz CPOs

Quartz crystallographic orientations were measured for 400

inclusions in the two labelled garnets above, and for 341

quartz grains in the surrounding groundmass. For both datasets

the crystallographic preferred orientation (CPO) is uniform.�

The development of an SPO has not resulted in any significant CPO.

The inclusion forming process has not resulted in any “selection” of grains

on a crystallographic basis

The clumped garnets have similar crystalorientation…�

Discrete glomerocrystic grains or connected asskeletal/dendritic growths?Small size of outliers favours discrete grains?

Crystal orientations of 15 clumpedgarnets

1mm

Grt outlier circled in{100} pole figure

ZO

XOYO

BSE OC

Clumped or clustered garnets have similar crystallographic orientation, suggesting:�

they either “know” something about each other, or

they are interconnected in 3D