some of the issues with our understanding and

interpretation of metamorphism and metamorphic rocks

disclaimer: these are some of the questions that are being

addressed by metamorphic petrologists today, so it is work in
progress!
I low-grade rocks
I equilibrium model of metamorphism breaks down
I high-grade and UHT rocks
I retrograde re-equilibration
I melt presence or absence
I definitions of UHT
I HP rocks
I fluid-saturation (lawsonite blueschists and eclogites)
I (ignoring) the influence of Fe3+
low-grade metamorphic rocks

I most (all) of our understanding of metamorphism is based on

an equilibrium model of metamorphism
I we assume that rocks and reactions in rocks strive towards
equilibrium
I thermal, mechanical and chemical / diffusive equilibrium
I the equilibrium model holds for most metamorphic grades
I but breaks down under extremely low-grade (sub-greenschist
facies) conditions
I too cold for simple diffusion to maintain equilibrium
I it is only when deformation forces minerals to recrystallise,
that they also equilibrate
I rocks under these conditions effectively record the P–T
conditions under which they were last deformed, rather than
peak P–T conditions
 (a)
I Muscovite zoning
4.5
II (b)

Mg + Fe + Mn + Ti (wt %)
II I
III
4

III
3.5

Qz Ms

3
Ms Chl 0 50 100 150
Distance (μm)
(c)
50 μm
Mg + Fe + Ti + Mn Octahedral Al
Mg + Fe + Ti + Mn Octahedral Al (d)
(c) 0.55
3.66
Pl
0.46 3.55
3.64
0.5
3.60
Ms
0.5 3.6
0.46 0.5 0 0
0.45 3.6
0.46 0.46 3.64 5
3.66 3.50
0.42 3.55
0.42

3.64 3.68 3.50 3.55

0.46 0.55
0.5 3.64 50 μm 50 μm
0.5

Chlorite zoning Biotite zoning

43
Mg + Fe + Ti + Mn (wt %)
(e) (f)
35
Fe + Mg (wt %)

42
34

41
33
100 μm 40 μm
0 50 100 150 0 50 100 150
Distance (μm) Distance (μm)
if mineral compositions do not work, what does?
I illite crystallinity
I examine the degree of transformation
of smectite → illite (muscovite)
I fluid inclusion thermometry
I fluid inclusions consist of liquid, gas
bubbles & crystals
I heat up under the microscope until
homogenisation to determine T of
entrapment
I cool until freezing to determine
salinity of fluid
high-grade and UHT rocks

I long cooling history with lots of time for retrograde

re-equilibration
I high temperatures means that diffusion continues during
retrogression → e.g. reaction textures
I presence or absence of melt in polymetamorphic rocks
I melt loss elevates solidus T & preserves high-grade minerals
I subsequent metamorphism will involve these high-grade
minerals → reworking may appear to occur at granulite grades
when it really occurs at medium grade subsolidus conditions
I recognising UHT metamorphic conditions
re-equilibration during retrogression

I diffusion can continue during the retrograde evolution of

high-T rocks
I they often cool slowly (limited exhumation)
I melt is often not immediately lost
I reaction textures and re-equilibration zoning develops
I these features can obscure the peak P–T history
I but they can be used to investigate the retrograde history
re-equilibration during retrogression

a b Mn
bt
+
sil
pl+qtz
+kfs
grt XSps=0.012

XSps=0.038

kfs
1 mm 1 mm

c Fe/(Fe+Mg) d Ca
pl+bt+sil

0.74 XGrs≈0.020

0.87

1 mm 1 mm
re-equilibration during retrogression
microshear
presence or absence of melt
I high-T rocks will retain the composition and fluid content
from their last melt-loss event unless fluid (or melt?!) is
added from an external source
I they will produce and lose melt during initial metamorphism
I but might not melt again during subsequent metamorphism
I polymetamorphic granulites might not be melt-bearing!

I M1 : cgr opx–sill–
cd–bi–pl–q–liq
I M2 : fgr opx–sill–
bi–pl–q ± liq?
 presence or absence of melt
post M1 melt loss and M2
NCKFMASHTO (+ q + ru) S009 (b) NCKFMASHTO (+ q + ru) S012
11

P (kbar)
g opx ged 5
10 g opx ky bi pl
ky bi pl
g opx sill bi pl liq 11 g opx sill bi pl liq
10 4 6
3 g opx
sill bi pl

ill liq
7

ill bi liq
opx ged ky bi pl
2

opx s
opx sill bi pl liq opx ky bi pl opx sill bi pl liq

ll ksp
l 3

opx s
9

0.16
ip opx

0.16
lb 1 cd

opx si
l
si

0.12
opx sill

pl liq
d cd s 5 bi p
ge pl ill bi
pl liq 9
l liq 1

0.13
x

0.14
bi

0.11
op 4

0.15
0.14

ll 6
0.15

si 8 2
cd 7 8
d pl opx cd sill bi pl
ge opx cd sill bi pl sp
x ik opx sill
op lb
sil bi pl p l opx cd
0.13

cd sp
0.12

x ik
ky ll op ill b
ksp pl
liq p pl liq

ky
si ds
0.11

opx cd 7 liq

liq
xc

ll

0.09
si
ksp liq op

sp pl
9

opx cd bi ksp pl
ks
0.0

0.1
0.1

opx cd px cd bi

0.08

d bi k
ksp pl

opx cd bi ksp pl
6

opx c
12
o

opx cd ksp
opx cd bi pl pl liq (-q)
cd opx
13 8
liq (-q)
5
675 700 725 750 775 800 825 850 875 900 925 950 650 675 700 725 750 775 800 825 850 875 900 925 950
T (oC) T (oC)

M2 assuming melt presence

NCKFMASHTO (+ q + ru + ilm) S030 gives 9 kbar @ 825–900
(a) (b) (c)
g ged 13
9
M2 assuming
ky bi 3 melt absence
g sill bi pl liq 14 gives much lower
1. opx ged cd sill bi pl liq
2. opx ged sill bi pl liq
P–T
1. opx cd sill bi ksp pl liq
2. opx cd sill ksp pl liq
1. g ged opx cd sill bi pl liq
2. g ged opx sill bi pl liq
pl
3. g opx ged sill bi pl liq 3. opx sill bi ksp pl liq 3. g ged ky bi pl liq
10 4. opx cd sill bi ksp pl liq
recognising UHT metamorphic conditions

I UHP is easy to detect → coesite

I no direct indicator of UHT (> 900 ◦ C) conditions
I the assemblages
I sapphirine–quartz
I orthopyroxene–sillimanite
I spinel–quartz
I osumillite
have all been suggested to indicate UHT conditions, but . . .
I sa–q coexists to significantly lower T with increased Fe3+
I opx–sill can occur at 800 ◦ C in K-poor rock compositions
(most UHT rocks are low-K, Mg–Al-rich pelites)
I sp–q can occur at 750 ◦ C in many rock compositions
sapphirine–quartz

8
FMAS (+opx+q)
FMASO (+opx+q)

sp t g
sill mt

m
sa g

sa
sa m ll
si
t
hem
7.5

cd sill
[cd,hem]

sa

sa g
sill
sp sill
7

sp sill
sa g
g
he

mt
sa sill
m

sa sp l
a

P (kbar)

il
gsp

cd

mt s
sa s
[sp,g] cd mt sa
mt sa [sp]
[sill] cd mt hem
sill cd sill cd
m sill [g-hem] sa
he t cd sa
g
6.5 sa mt sp cd

sp ill
m sa g

gs
s ill sp cd sa
t [mt,hem] sp cd

mt
m cd sp cd
sp
sill cd
g
sp

g cd
sp
[sa,hem]
sill cd
g
sillcd mt
mt g cd

g
sp

kerrie taylor-jones
5.5
800 850 900 950 1000 1050
T ( C)
 orthopyroxene–sillimanite
FMASHTO (+ q + ru) S009 (b) NCKFMASHTO (+ ru) S012
11
12 1 3

bi
11

liq q
P (kbar)
4

g opx sill
13 g opx sill 14 10 18
bi liq
g opx sill 2
bi pl liq 10 g opx sill bi pl liq q
ged ky bi pl liq 17
ged ky bi pl liq q

q
i liq
opx

ged ky bi pl liq
opx sill sill liq opx

sill b
11 opx sill bi bi liq 8 9 opx sill bi pl liq q sill liq
pl liq

ged ky bi pl q H2
9

opx
liq 10 opx c 12 16
i pl d sill b
ill b 15 7 i pl liq 13 15
ds opx c 16 q
dc d sill

qq
ge bi pl

q H2O
liq 14

bi pl li
17
8
d pl liq

O
liq

ed cd
5
cd
liq

6 opx
x
opx c

op
bi pl

cd liq

opx g
opx cd bi pl liq q
opx cd
7
d cd

bi pl liq ged cd bi pl liq q

ky

opx cd bi liq
l
sil
ged cd bi pl
e

ged cd bi pl liq
opx g

liq
ged cd bi pl q H2O

opx cd bi pl
18 opx cd 6
liq (-q)
liq q H2O
19
20
5
700 725 750 775 800 825 850 875 900 925 950 650 675 700 725 750 775 800 825 850 875 900 925 950
T (oC) T (oC)

FMASHTO (+ q) opx introduced by breakdown

S030 (a) of gedrite –(b)rather than biotite (c) – in
27 1. ged pa ky bi H O 1. ged pa ky bi q H O 1. g ged pa ky bi pl ru H O 2 2 2
low-K, Mg–Al pelites 28 2. ged pa ky bi liq H O
3. ged pa ky bi liq
2. ged pa ky bi pl q H O 2. ged pa ky bi pl ru H O
3. ged pa ky bi pl q liq H O 3. ged pa ky bi pl ru liq H O
2 2 2

26 2 2
4. ged pa ky bi pl liq 4. ged pa ky bi pl q liq 4. ged pa ky bi pl ru liq
5. ged pa ky bi pl H2O 5. ged cd ky bi pl q H2O 5. ged cd sill bi pl ru H2O
g ged ky
fluid saturation in HP rocks

high-P metamorphic conditions related to subduction are generally

identified from examining mafic rocks
I metapelites are not very sensitive to P
I do not develop characteristic mineral assemblages
I garnet, kyanite, quartz / coesite are stable over wide P range
I mafic rocks can develop glaucophane, lawsonite, omphacite,
jadeite, kyanite, talc etc but . . .
I unlike metasediments, basalts are often not fully hydrated prior
to metamorphism
I we should not assume fluid saturation
I “high-P” minerals might reflect fluid-undersaturated
conditions, rather than very high P
lawsonite blueschists
I lawsonite (CaAl2 Si2 O7 (OH)2 ·H2 O) contains lots of H2 O in its
crystal structure
I the growth of lawsonite requires more H2 O than can be held in
metasomatised, maximally-hydrated basalt before subduction
I the presence of lawsonite is restricted to rocks that had fluid
addition along the prograde path during subduction
I many rocks are fluid-absent at P–T where lawsonite is stable

maximally hydrated basalt

eclogites

I in the field, many eclogites occur as isolated pods and boudins

within lower-grade blueschist or high-P amphibolite
I the traditional interpretation is that the blueschist or high-P
amphibolite was eclogite, but got retrogressed during
exhumation, with the eclogite boudins escaping retrogression
but what if metamorphism involved a patchily hydrated basalt?
I H2 O-saturated portions might form blueschist or high-P
amphibolite assemblages
I and H2 O-undersaturated portions might form eclogite
I this would mean that the difference is because of H2 O content
rather than P–T
I the blueschist or high-P amphibolite is the better reflection of
actual peak P–T conditions than the eclogite
eclogites
eclogites
(b) @M(H2O) = 3.0
‘real’ eclogites have to be (very) 1.0
ilm q H 2O
H2O-undersaturated opx

Modal abundance
0.8 sph ru
pl g

g-amphibolite
hb ‘granulite’
0.6
dio
(a) NCKFMASHTO (+ hb + q) @ 650°C hb-eclogite, sample GB1 o
20 0.4
o g ep ru H2O ru
19 ep o g ep 0.2 hb
0.35 og ky ksp
18 d ep
ru
2 5 10 15 20
17 0.30
ru o g ep Pressure [kbar]
16 dio g ep ru H2O ep ky ru ep ru
g (c) @M(H2O) = 0.5
g o 1.0
15 0.25 dio di sp q
k opx ru
ky

Modal abundance
14 0.8

gabbro / granulite
ilm + mt ky
g

true eclogite
0.20
13 0.6
Pressure [kbar]

pl
g h
12 o p dio g pl ep ru 0.4
di p s
11 e
dio g pl

0.2 dio
sph ru

dio g pl ep sph hb ep o
10 dio g pl ru
9 2 5 10 15 20
dio g
dio pl pl sph Pressure [kbar]
8 (d) @18 kbar
sph H2O 1.0
7 dio g pl ilm H 2O q ky

Modal abundance
dio g opx pl ilm 0.8 ru ksp
6 g

hb-cz ‘eclogite’

true eclogite
5 0.6

hb eclogite
dio pl dio
4 dio pl sph H2O
dio opx pl

sph ilm 0.4 o

ilm mt

b c hb
3

dio
dio opx pl ilm 0.2
2 ep
5 4 3 2 1 0
5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0
M(H2O) [mol%]
M(H2O) [mol %]
ferric iron

I many elements can exist with multiple oxidation states

I Fe2+ or Fe3+
I Mn2+ or Mn3+ or Mn4+
I Cu+ or Cu2+
I C or CO2−
3
I S2− or SO2−
4
I Fe2+ & Fe3+ is the only one of these that is relatively
abundant in rocks & minerals
I routine analytical techniques do not distinguish between Fe2+
and Fe3+
I XRF reports all Fe as Fe2 O3
I microprobe reports all Fe as FeO
I also difficult to include Fe3+ in experiments
I Fe3+ is often treated as a nuisance and is mostly ignored
I “all-Fe-as-FeO” assumption in many thermobarometers
ferric iron

however . . .
I many silicate minerals can take up large amounts of Fe3+
I epidote has all Fe as Fe3+
I omphacite, jadeite, glaucophane & sapphirine can have more
than 50 % of their Fe as Fe3+
I biotite, hornblende & orthopyroxene can take significant
amounts
I Fe3+ can drastically affect the mineral assemblages that are
stable in some rocks
I Fe3+ stabilises glaucophane to lower P (blueschist facies)
I Fe3+ stabilises sapphirine to lower T (UHT)
I including Fe3+ also makes rocks more Mg-rich
I some thermobarometers are extremely sensitive to Fe3+
I garnet–clinopyroxene
stability of blueschists
NCKFMASHTO (+ q + H2O) @ 450 oC
15 o law mu law
o act
chl sph o gl p sph
law mu hl e o gl mu chl ep ru
mu c

o
gl
14 chl sph 2 1. o act law mu chl ep sph

m
sph

u
2. o law mu chl ep sph

ch
1 ep
chl

le
mu 3. o mu chl ep sph

gl mu chl ep ru
p
gl

sp
13 o 4. o act mu bi chl ep sph
3

gl mu chl ep ru hem
ru
5. o gl act mu bi chl ep sph
8 6. o gl act mu chl ep sph
o act mu di gl mu chl ep sph 10
12 7. o di gl act mu chl ep sph
chl ep sph 12 13
7 9 8. o di gl mu chl ep sph
6 11 14 9. di gl act mu chl ep sph
di gl act bi chl ep sph
11 10. di gl mu bi chl ep sph
5 ph gl mu
4 ps 11. gl act mu bi chl ep sph
le chl ep
ch gl mu 12. o gl act mu chl ep sph
bi sph ru
a ct bi chl 13. gl mu chl ep sph
10 gl
P (kbar)

o ep sph 14. gl act mu chl ep sph

gl act bi chl ep sph
15. gl bi chl ab ep sph
gl mu bi 16. gl bi chl ab ep sph ru

gl bi chl ep sph
9 chl ep 17. gl bi chl ab ep sph hem
sph ru 18. gl bi chl ep ru
gl a gl mu
ct b 19. gl mu bi chl ep ru hem
8 i ch
l ab bi chl
ep s gl bi ep ru 20. gl bi chl ab ep ru hem
ph chl ep 21. gl mu bi chl ab ep ru hem
sph ru 18 22. gl bi chl ab ep sph ru hem
7 HM 15 19
23 23. gl mu bi chl ab ep ru
QFM

16 24. gl bi chl ab ep ru
21
25 24 25. gl act bi chl ab ep sph hem
6 20
17
act bi chl ab ep sph 22
5

act bi chl ab ep sph hem

4

3
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Fe3+/FeT
another example from Madagascar
corundum–quartz

I natural corundum–quartz assemblages documented in a

number of high-grade terranes
I cor–q ± sill/ky ± g ± mt–sp–ilm
however
I cor–q assemblages cannot be reproduced experimentally or by
thermodynamic calculations
I ky → cor + q metastable to sillimanite
I sill → cor + q metastable to kyanite
natural occurrences explained by
I sluggish reaction rates, particularly in dry granulites
I corundum exsolved from spinel in the presence of quartz
I minor components that might increase corundum stability
(e.g. Fe3+ )
corundum–quartz
FMASTO (+ g + q)
40

Fe3+/FeT = 0
35 cor-ru
P (kbar)

30

25

20

15

ky
sill
10

5
800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800
T (oC)
corundum–quartz
FMASTO (+ g + q)
40

Fe3+/FeT = 0.01
35

P (kbar)

30

25

20

15

ky
sill
10

5
800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800
T (oC)
corundum–quartz
FMASTO (+ g + q)
40

Fe3+/FeT = 0.02
35 cor-ru
P (kbar)

30

25

20

15

ky
sill
10

5
800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800
T (oC)
corundum–quartz
FMASTO (+ g + q)
40

Fe3+/FeT = 0.05
35

P (kbar)

-ru
30
r
co

25

20

15

ky
sill
10

5
800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800
T (oC)
corundum–quartz
FMASTO (+ g + q)
40

Fe3+/FeT = 0.1
35

P (kbar)

30

ilm
u-
25
r-r
co

20

15

ky
sill
10

5
800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800
T (oC)
corundum–quartz
FMASTO (+ g + q)
40

Fe3+/FeT = 0.2
35

P (kbar)

30

25

-ilm
-ru
cor
20

15

ky
sill
10

5
800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800
T (oC)
corundum–quartz
FMASTO (+ g + q)
40

Fe3+/FeT = 0.4
35

P (kbar)

30

25

20

cor-ru-Ti hem
15

ky
sill
10

5
800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800
T (oC)
corundum–quartz
FMASTO (+ g + q)
40

Fe3+/FeT = 0.75
35

P (kbar)

30

25

20

15

ky
sill
10 cor-ru-hem

5
800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800
T (oC)
corundum–quartz
FMASTO (+ g + q)
40

Fe3+/FeT = 1
35

P (kbar)

30

25

20

15

ky
sill
10 cor-ru-hem

5
800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800
T (oC)