Geoscience Frontiers 5 (2014) 683e695Contents lists available at ScienceDirect
China University of Geosciences (Beijing)
Geoscience Frontiers
journal homepage: www.elsevier .com/locate/gsfResearch paperDeciphering fluid inclusions in high-grade rocks
Alfons van den Kerkhof*, Andreas Kronz, Klaus Simon
Geowissenschaftliches Zentrum der Universität Göttingen, Goldschmidtstr. 1-3, D 37073 Göttingen, Germanya r t i c l e i n f o
Article history:
Received 31 October 2013
Received in revised form
27 February 2014
Accepted 14 March 2014
Available online 29 March 2014
Keywords:
Fluid inclusions
Granulite
Charnockite
Cathodoluminescence
LA-ICPMS
Quartz* Corresponding author.
E-mail address: akerkho@gwdg.de (A. van den Kerkho
Peer-review under responsibility of China University
Production and hosting by Els
1674-9871/$ e see front matter 
 2014, China Univer
http://dx.doi.org/10.1016/j.gsf.2014.03.005a b s t r a c t
The study of fluid inclusions in high-grade rocks is especially challenging as the host minerals have been
normally subjected to deformation, recrystallization and fluid-rock interaction so that primary in-
clusions, formed at the peak of metamorphism are rare. The larger part of the fluid inclusions found in
metamorphic minerals is typically modified during uplift. These late processes may strongly disguise the
characteristics of the “original” peak metamorphic fluid. A detailed microstructural analysis of the host
minerals, notably quartz, is therefore indispensable for a proper interpretation of fluid inclusions.
Cathodoluminescence (CL) techniques combined with trace element analysis of quartz (EPMA, LA-
ICPMS) have shown to be very helpful in deciphering the rock-fluid evolution. Whereas high-grade
metamorphic quartz may have relatively high contents of trace elements like Ti and Al, low-
temperature re-equilibrated quartz typically shows reduced trace element concentrations. The result-
ing microstructures in CL can be basically distinguished in diffusion patterns (along microfractures and
grain boundaries), and secondary quartz formed by dissolution-reprecipitation. Most of these textures
are formed during retrograde fluid-controlled processes between ca. 220 and 500 
C, i.e. the range of
semi-brittle deformation (greenschist-facies) and can be correlated with the fluid inclusions. In this way
modified and re-trapped fluids can be identified, evenwhen there are no optical features observed under
the microscope.

 2014, China University of Geosciences (Beijing) and Peking University. Production and hosting by
Elsevier B.V. All rights reserved.1. Introduction
Fluid inclusions in high-grade metamorphic rocks have been
studied much later than for other lithologies such as hydrothermal
and magmatic rocks, which have been systematically described
since the 19th century. Pioneering studies on geological fluids in
high-grade rocks dated back to the early 1970’s (Touret, 1981, 2001)
and awakened a lively interest, which continues until today. The
study of fluid inclusions in metamorphic rocks was particularly
stimulated by the recognition of their role in stabilizing meta-
morphic mineral assemblages from autoclave experiments coupled
with thermodynamic calculations (e.g. Winkler, 1965). Relic fluids,
preserved as inclusions in natural rocks, are the only directf).
of Geosciences (Beijing)
evier
sity of Geosciences (Beijing) and Pevidence of the role of fluids during metamorphism. However, fluid
inclusions in metamorphic rocks are normally strongly modified
during retrograde conditions such that the corresponding micro-
textures are complex. They therefore pose a challenge for unrav-
elling the role of the metamorphic fluid during rock evolution. The
fluid inclusion inventory within one rock sample may show a large
variation in compositional and density, whichmust reflect different
stages of rock evolution to investigate. In some granulites the fluid
inclusions are extremely difficult, consisting only of small, highly
transposed inclusions. In other granulites, they are much easier to
investigate consisting of relatively large and abundant fluid in-
clusions in many of the rock-forming minerals. It should be
emphasized that since fluid inclusions are a part of the rocks, they
must be studied as much as any other rock constituent (Jacques
Touret pers. comm.).
Carbonic fluid inclusions are typical for many granulite rocks
and were recorded for the first time in the Bamble Sector, southern
Norway (Touret, 1971a, b), i.e. within the zone delimited by the
orthopyroxene-in isograd. The CO2 is assumed responsible for the
lowH2O activities required for the stabilization of orthopyroxene in
these rocks. Following this first discovery, carbonic fluid inclusions
were found in many other granulite-facies rocks worldwide. Theeking University. Production and hosting by Elsevier B.V. All rights reserved.
A. van den Kerkhof et al. / Geoscience Frontiers 5 (2014) 683e695684number of fluid inclusion studies in granulite rocks rose steadily
from <5 to ca. 15 to 35 publications per year in the late 1980’s and
early 1990’s (Fig. 1). In the early 1980’s the introduction of Laser
Raman micro-spectrometry for geological studies allowed for the
fast detection of CO2 and therewith stimulated the study of micron-
scale “gas inclusions” in rocks. The somewhat lower number of
studies nowadays can be explained by the general acceptance of
CO2 being the dominant fluid phase in granulite-facies rocks. Still
some scepticism persists for a number of geologists who consider a
proper interpretation of fluid inclusions in high-grade meta-
morphic rocks problematic, or at the best “difficult”, due to the
strongly modified character of the majority of fluid inclusions with
only a few (if any at all) preserved unchanged from peak-
metamorphic conditions.
In addition to CO2, high-salinity brines are known as an
important constituent of many high-grade metamorphic rocks.
These fluids also help to reduce the rock H2O activity and stabilize
“dry” mineral assemblages (Touret, 1985). Brines are found in many
granulite-facies rocks along side carbonic inclusions.
For more than two decades, fluid phase petrology, i.e. the study
of fluid inclusions and related microtextures, has made good use of
cathodoluminescence (CL) techniques in order to visualize fluid-
induced textures, which cannot be observed otherwise (e.g. van
den Kerkhof and Hein, 2001). In the present paper we give an
overview and a guideline for fluid inclusion studies in quartz from
high-grade metamorphic rocks, from microscopic observations
including detailed fluid inclusion mapping to microthermometry,
CL and trace element analysis.
2. Carbonic fluid inclusions in granulite rocks
Fluid inclusions in granulite-facies rocks are almost always
preserved in quartz as the hostmineral. Some granulite-facies rocksFigure 1. Graphic presentation of the number of publications on fluid inclusions in granulit
Total number of items ¼ 492, from the first publication (Touret, 1971a, b) until 2012.also show carbonic inclusions in other minerals like garnet,
plagioclase, pyroxene, apatite, and zircon, besides fluid inclusions
in quartz. Examples are the Doddabetta charnockite complex,
Southwest India (Touret and Hansteen, 1988), the East African
charnockites (Coolen, 1980; Herms and Schenk, 1998), Victoria
Land, Antarctica (van den Kerkhof et al., 1998) and Varberg char-
nockite, SW Sweden (Harlov et al., 2013). The reason that carbonic
inclusions are sometimes preserved in all the rock-forming min-
erals and in other cases only in quartz only, has not been fully
explained so far, but suggests a high CO2 activity during the growth
of these minerals. Comparison between examples from the litera-
ture suggests that igneous charnockites tend to contain fluid in-
clusions in more than one rock-forming mineral (e.g. garnet,
plagioclase, and quartz), whereas metamorphic charnockites nor-
mally have fluid inclusions in quartz only. However, there are ex-
ceptions and the distinguishing features between igneous vs.
metamorphic charnockites have not been completely resolved
(Touret and Huizenga, 2012).
The first studies of carbonic inclusions in metamorphic rocks
already demonstrated large variations in fluid inclusion densities
on the micron scale (e.g. Touret, 1971a, b). Normally few high-
density CO2 inclusions (>1 g/cm3), originating from the peak-
metamorphic conditions, are preserved and the majority of in-
clusions show densities down to the critical density of CO2 (0.47 g/
cm3), or lower. Inclusions with highest density often show primary
characteristics, i.e. they are isolated, or part of a cluster, as a group
of inclusions formed by partial fluid re-trapping during uplift.
However, fluid inclusions with the highest density are earliest only
in the case of decompression paths, not always in the case for HT-
granulites. Inclusion geochronology should be exclusively
deduced from the inclusion morphology, i.e. the relation with the
hostmineral, and then the density evolution can be discussed in the
light of this timing. We will see in the present study that thee-facies rocks as obtained from the GeoRef data base (American Geosciences Institute).
A. van den Kerkhof et al. / Geoscience Frontiers 5 (2014) 683e695 685highest-density inclusions are not always easy to recognize from
petrological observations and may be “hidden” in between other
re-equilibrated inclusions of varying density.
In granulites with carbonic fluid inclusions preserved in more
than one rock-forming mineral, high-grade metamorphic in-
clusions are typically characterized by negative crystal shapes, e.g.
hexagonal shapes for quartz (Fig. 2). These inclusions are arranged
in diffuse tracks and swarms, and are rarely isolated (e.g. Coolen,
1980; Herms and Schenk, 1998). On the other hand, granulites,
which developed carbonic inclusions only in quartz, almost never
contain inclusions with a negative crystal shape. In the latter case,
early-formed high-density carbonic inclusions are much more
difficult to identify.
Homogenization temperatures (or fluid densities) are often
graphically presented as bar plots, which may show one, two or
more frequency maxima corresponding to episodes of massive CO2
re-trapping during uplift and never under peak-metamorphic
conditions. Touret (1987) earlier preferred curves, which have the
advantage of illustrating perturbations (dissymmetric histograms).
However, any statistical treatment of the Th distribution has shown
to be of limited use so far. The “mean” CO2 density appeared to be
normally not representative for the peak of metamorphism. Peak-Figure 2. Primary high-density carbonic fluid inclusions N2 in quartz from a
granulite-facies rock showing a negative crystal shape. Example from a biotite-garnet
gneiss from the Furua Granulite Complex, Tanzania (sample: C. Coolen). The lowest
ThCO2 of these inclusions is ca. e46

C. (a) View parallel to the quartz c-axis; (b) view
perpendicular to the quartz c-axis.metamorphic fluid inclusions are rare and one of the main tasks
of studying metamorphic fluid inclusions is the identification of
these “exceptions”. In this respect the decompression re-
equilibration experiments of Vityk and Bodnar (1995, 1998)
showed that fluid inclusions never equally adjusted to the change
in pressure-temperature. His conclusion, that a part of the inclusion
always survives the changing P-T conditions, provided a solid base
and optimism for the interpretation of natural fluid inclusions. It
confirms observations made for natural samples that high-density
fluid inclusions are almost always present, even when sometimes
difficult and time-consuming to find.
3. Fluid phase petrography
3.1. Primary and secondary inclusions
Primary fluid inclusions in high-grade metamorphic rock have
been identified as high-density, isolated carbonic inclusions, e.g. in
the Bamble area, southern Norway (Touret, 1971a, b; van den
Kerkhof et al., 2004), Furua Granulite Complex, southern Tanzania
(Coolen, 1980), southern India (Santosh and Tsunogae, 2003), and
East Antarctica (Tsunogae et al., 2002). In these examples, nitrogen
has sometimes been found as an additional fluid component,
resulting in high-density CO2-N2 inclusions (e.g. Kooi et al., 1998).
In other cases, like the charnockitic rocks of Söndrum, SW Sweden,
low-salinity aqueous-carbonic inclusions could be identified as the
earliest, peak-metamorphic fluid (Harlov et al., 2006). These also
show regular shapes, whereas the carbonic inclusions, which make
up the majority of fluid inclusions, must have formed by the uptake
of the H2O into the melt during partial melting of the granitic host
rock. During this process of H2O-loss the fluid phase becomes
enriched in CO2. This process is in accordance with observations
made for rocks which show typical restite reactions, including melt
formation (Touret and Dietvorst, 1983).
Under high-grademetamorphic conditions two immiscible fluid
phases coexist for a wide range of bulk fluid compositions. There-
fore, it is thought that many salt-bearing aqueous fluid inclusions
formed simultaneously with the carbonic ones (CO2 with up to
50mol% H2O, and very low salt concentration). The latter inclusions
show only small H2O volume fractions under the microscope at
room temperature and are recognized as “carbonic”. Experiments
by Shmulovish et al. (1994) and Zhang and Frantz (1989) showed
that bulk fluids with a CO2/(CO2þH2O) ratio of >0.3 and salt con-
centrations of even less than 10 wt.% must plot in the two-phase
field under high-grade conditions. During uplift the aqueous in-
clusions tend to strongly re-equilibrate due to their steep isochores.
These fluids may be largely retrapped as secondary inclusions. The
expansion of the immiscibility field on the lowering of pressure
may result in the further “unmixing” of the aqueous-carbonic fluids
during retrogression.
Secondary fluid inclusions make up the large majority of fluid
inclusions in granulites. Carbonic inclusions are typically better
preserved and may, in part, originate from an early metamorphic
stage. On the other hand, the aqueous inclusions are normally
secondary in character with only a few exceptions. Swanenberg
(1980) distinguished between clusters and trail-bound secondary
fluid inclusions, which was very useful also for later studies. These
textures can be observed for carbonic as well as for aqueous in-
clusions. Clusters are defined as spatially dispersed groups of fluid
inclusions, which occupy a specific volume of the host mineral.
They are also sometimes referred to as decrepitation clusters as
they are thought to form as a result of fluid re-trapping after
leakage due to internal fluid overpressure (“explosion”) during
uplift. The process of fluid re-trapping is still not well understood.
So far, attempts at generating “explosion” clusters in experiments
Figure 3. Histograms presenting CO2 homogenization temperatures for a granulite-facies sample from the Bamble Sector, southern Norway. (a) Results for 10 samples from quartz
segregations; (b) results for one sample, one chip, and one cluster with relatively high-density inclusions, including the highest-density primary inclusions; (c) result for the most
frequent secondary inclusions in “transposed” trails.
A. van den Kerkhof et al. / Geoscience Frontiers 5 (2014) 683e695686have not been successful (Vityk and Bodnar, 1995). On the other
hand, fluid inclusion clusters can be easily reproduced during re-
equilibration experiments which create fluid under-pressure, i.e.
“implosion” decrepitation (Sterner and Bodnar, 1989). In naturalquartz, fluid inclusion clusters can be observed in high-grade rocks,
which were subject to a thermal event during the emplacement
of relatively shallow magmatic bodies, followed by an episode
of approximate isobaric cooling (e.g. Rogaland high-grade
A. van den Kerkhof et al. / Geoscience Frontiers 5 (2014) 683e695 687metamorphic complex, Norway; Swanenberg, 1980; Bakker and
Jansen, 1993). In the latter case, the fluids are also strongly chem-
ically re-equilibrated towards CH4-rich compositions.
Trail-bound inclusions are very common for granulites and
indicate fluid trapping along microfractures. These inclusions are
always secondary and clearly postdate peak-metamorphic condi-
tions. Quartz deformation grades from plastic towards brittle
behaviour during uplift (Passchier and Trouw, 2005) so that fluid
inclusions tend to form along healed fractures at lower tempera-
tures. Continuing re-crystallisation, however, may result in the
forming of new quartz mineralization. Intra-granular trails (see
Simmons and Richter and Simmons, 1977 for trail nomenclature)
are very common in granulite-facies quartz and must represent
fluid trapping during an early stage of deformation, but before the
formation of subgrains. These processes are expected to take place
in the semi-brittle deformation P-T range (essentially greenschist
facies) as the result of retrograde fluid-controlled processes be-
tween ca. 220 to 500 
C.
3.2. Fluid inclusion mapping
The preparation of detailed sketches of fluid inclusions and
related microtextures has been a tradition since the earliest fluid
inclusion descriptions (e.g. Sorby, 1858). This means of documen-
tation is particularly helpful for unravelling the chronology of fluid
inclusion trails as can be deduced from intersecting relationships
(e.g. Touret, 1981). Photographs have the disadvantage that only
one focus level can be recorded, so that more photos must be
combined. In hand-mapped fluid inclusion plates the distribution
of homogenisation temperatures on a sub-millimetre scale can give
insight into the process of fluid inclusion modification and re-
trapping.
The highest-density carbonic inclusions can be found by the
examination of a set of successive, separate doubled polished thick
section samples. As an example we show results obtained from
quartz segregations from granulite-facies rocks in the Bamble
Sector (Southern Norway) (Fig. 3a). The microthermometry of 10
samples results in homogenization temperatures (Th, i.e. the phase
transition liquid þ vapour/ liquid), which cluster in 3 frequency
ranges: the strongest mode (A) between ca. 5 and 9 
C (mean ca.
0 
C), the second-largest mode (B) between ca. þ15 and þ26 
C
(mean about þ21 
C) and a weak mode (C) between ca. 25
and 15 
C (mean ca. 20 
C). These modes can be found also for
each of the individual samples, although their relative frequencies
strongly differ. It is assumed that these frequency modes corre-
spond to regional events during uplift, which resulted in the partial
re-trapping of inclusion fluids. In searching for the highest-density
inclusion, we selected one sample showing only ThCO2 in the low-
temperature A and C modes and only a very few high Th, for a
more detailed study (Fig. 3b). We selected a ca. 5 mm-sized doubly
polished chip containing fluid inclusions with the lowest Th after
measuring three for each sample (Fig. 3b). On this chip we mapped
the fluid inclusion in order to identify those inclusions with the
highest density and to locate their position among re-equilibrated
inclusions of lower density (Fig. 4a). The highest-density inclu-
sion, which consists of nearly pure CO2 (<1% N2) shows a Th
of48.7 
C, which corresponds to a density of 1.15 g/cm3. So far, the
lowest Th that has been recorded for this area is 40 
C (Touret,
1985). An interesting observation from fluid inclusion mapping is
the large spread in Th within one cluster (Fig. 3b). The cluster
containing the highest-density inclusion contains not only in-
clusions with “mode C” homogenisation temperatures, but also a
“mode A” inclusion within a 20 mm distance. A similar observation
was made in granulite-facies quartz from Söndrum, Sweden
(Harlov et al., 2014). Here, the highest-density inclusions(Th 39.3 
C) occur in the same cluster together with low-density
carbonic inclusions. This type of high-density inclusion cannot be
easily recognized as they do not show primary characteristics, but
occur as unremarkable relic inclusions among decrepitated ones.
The majority “mode A” inclusions, however, occur on healed
microfractures and are clearly secondary (Fig. 4b). Oftenwe observe
textures, which indicate multi-stage deformation resulting in the
“transposition” of fluid inclusion trails, forming subparallel or en
echelon short trails. Contrary to the decrepitation clusters, the
density of fluid inclusions within one trail is almost the same. In the
Bamble Sector example, a typical “transposed trail” shows ThCO2
between 5 and þ4 
C with two fine modes, one around 0 
C (A1)
and another around 3 
C (A2) (Fig. 3c). The inclusions with higher
Th are found in the central part of the trail, which indicates that a
part of the inclusionsmust have been retrapped after the reopening
and subsequent healing of the same structure. The short trails ar-
ranged en echelon have a higher density than the inclusions in the
central part. This type of inclusion is common in high-grade rocks
and shows that the same fractures can be repeatedly opened and
sealed during deformation.
3.3. Isochore calculations for carbonic inclusions
Isochores for the highest-density CO2 fluid inclusions found in
granulite-facies rocks are the key for estimating the fluid trapping
depth. Calculations can be made e.g. with equations of state
collected in the “FLUIDS” software (Bakker, 2003). For high-density
CO2 we used the equation of state of Span and Wagner (1996)
whose highest accuracy is in the range of 489e1373 
C and
0.5e800 MPa. The lowest homogenization temperatures collected
for a number of granulite-facies terrains show a remarkably small
range between 53 and 39 
C (Fig. 5) corresponding to densities
between 1.11 and 1.17 g/cm3. When considering peak-metamorphic
temperatures of 750 to 850 
C these densities correspond to trap-
ping pressures of 800 to 1000 MPa, i.e. at 30 to 35 km. These
conditions are in agreement with a geothermal gradient of ca.
25 
C/km, which is typical for the stable continental crust. Excep-
tions are areas with a high thermal flow at lower crustal levels, e.g.
theWest-Uusimaa Complex, southern Finland, which is interpreted
as a thermal dome (Schreurs, 1985) with a corresponding
geothermal gradient of ca. 50 
C/km (number 3 in Fig. 5). Peak
conditions for charnockites in relation to the “wet” granite melting
curve plot at lower temperatures (650e720 
C), but are also in
agreement with the continental geothermal gradient (e.g. the
Swedish charnockites, number 5 and 6 in Fig. 5) and trapped at ca.
30 km depth. Thus most “dry” granulites have preserved high-
density carbonic inclusions with homogenization temperatures
of 40 
C and lower, even when these inclusions are rare.
3.4. Aqueous fluid inclusions in high-grade metamorphic rocks
In addition to carbonic inclusions, remnants of NaCl-rich brines
are found in a great number of granulite-facies rocks (Touret,
1995). There are strong arguments to believe that brine-rich
fluids are present at peak metamorphic conditions. However,
due to their steep isochore slopes, aqueous inclusions almost
never survive uplift unchanged, though some aqueous inclusions
with a primary character can sometimes be found, e.g. in quartz
blebs in garnet (Fu et al., 2003). Other arguments for the existence
of high-grade metamorphic brines are the relationship between
the fluid and certain types of host rocks (e.g. detrital and meta-
evaporite rocks in Bamble granulite), or the traces left by inter-
granular fluids (K-feldspar micro-veins and/or myrmekite, see
Touret and Nijland, 2012). The densities of associated brine in-
clusions are often high (small or no bubble) and indicate (re)
Figure 4. Result from fluid inclusion “mapping” with an indication of the homogenisation temperature distribution. Example from the quartz segregations in granulite-facies rocks
from the Bamble Sector, southern Norway. The insets marked in the photographs are shown in the drawings below. (a) Decrepitation clusters including the highest-density carbonic
inclusions (48.7 
C), possibly dating from peak-metamorphic conditions (dark purple signature); (b) transposed trail of carbonic inclusions with homogenization temperatures
between 5 and þ4 
C corresponding to the most frequent CO2 densities found in this sample material (c.f. Fig. 3c). Note the somewhat higher homogenization temperatures in the
central part of the structure indicating later retrapping.
Figure 4. (continued).
A. van den Kerkhof et al. / Geoscience Frontiers 5 (2014) 683e695 689trapping of aqueous fluids at low temperature and pressure. Fluid
re-trapping does not necessarily imply changes in the salt
composition and concentration, so that the study of aqueous in-
clusions may indeed provide information on the solutes acting
during high-grade metamorphic conditions. The brines play an
important role in controlling the chemical potential of alkalis and
LILE elements. The inventory of aqueous inclusions largely de-
pends on the protolith, while low-salinity aqueous-carbonic fluidsare common in igneous rocks. Metapelites contain almost only
aqueous fluids. Aqueous fluids play an essential role in the
metasomatic changes observed for granulite-facies rocks. In the
localized orthopyroxene-bearing dehydration zone at Söndrum,
SW Sweden, the dehydration event was accompanied by the
depletion of the charnockite in Fe, YþHREE, Na, K, F, and Cl, and
the enrichment in Mg, Mn, Ca, and Ti compared to the sur-
rounding granitic gneiss (Harlov et al., 2006).
Figure 5. Isochores with indicated Th corresponding to the highest-density CO2 found in selected granulite-facies rocks from 1. Furua Granulite Complex, southern Tanzania
(Coolen, 1980); 2. Bamble Sector, southern Norway (Touret, 1985); 3. West-Uusimaa Complex, southern Finland (Schreurs, 1985); 4. Doddabetta, southern India (Touret and
Hansteen, 1988); 5. Söndrum, Halmstad, SW Sweden (Harlov et al., 2006); 6. Varberg, SW Sweden (Harlov et al., 2013). CO2 isochores calculated utilizing the equation of state
from Span and Wagner (1996). The upper limit of immiscibility is given for H2O-CO2, and 2.6 wt.% NaCl brine þ CO2 fluid inclusions (Crawford, 1981). Granite melting curves
(Newton, 1986) are given for dry granite melting, and for H2O activities of 0.7 and 1.
A. van den Kerkhof et al. / Geoscience Frontiers 5 (2014) 683e6956903.5. Retrograde fluids in high-grade rocks
Fluid interactionwith the rock during upliftmay result in changes
in the fluid chemistry. The most obvious are the methane-rich com-
positions found inmany fluid inclusions, e.g. Rogaland (Swanenberg,
1980). These granulites aremetapelitic in origin and contain traces of
organic carbon. The fluid composition represents fluid-rock equilib-
rium, which is essentially a function of temperature and the oxygen
fugacity as determined by the mineralogy in the C-O-H system
(Huizenga, 2001; Bakker, 2003). The pressure is a less sensitive
parameter. In this way the fluid composition, e.g. defined as the CO2/
CH4 ratio, can be used as a thermometer for a given oxygen fugacity
(e.g. van den Kerkhof et al.,1991). At high-grade conditions, however,
the gas composition in these fluid inclusions is essentially pure CO2.
There are also instances of late fluid inclusions having highly
saline CaCl2 compositions. The presence of CaCl2 brines can be
considered as a result of extreme retrograde alteration during low
temperature albitisation. These brines are thought to be the result
of feldspar alteration from the reaction:
CaAl2Si2O8 þ 2Naþ þ 4SiO2 ¼ 2NaAlSi3O8 þ Ca2þ (1)which results in higher Ca-concentrations with increasing fluid
salinity (Yardley and Shmulovich,1995). These fluid inclusions form
typically during the last stages of rock evolution and always show
homogenization temperatures below ca. 200 
C. Although these
high-salinity inclusions are found in many granulites they do not
represent high-grade metamorphic conditions.
4. Cathodoluminescence (CL) and trace element analysis
4.1. Technical details
CL studies with the optical CL-microscope is highly suitable for
distinguishing quartz textures related with different quartz gen-
erations, which differ by their trace element and structural defect
inventory. We used the high-power HC3-LM-Simon-Neuser CL
microscope (Neuser et al., 1995). Additional studies with an elec-
tron microprobe (JEOL JXA 8900RL) equipped with a CL-detector
(200e900 nm) have the advantage of combining CL-imaging and
resolving trace element compositions (EPMA). Quantitative anal-
ysis of Al, Ti, Fe, K and Na can be carried out onmetamorphic quartz
(Müller, 2000; Müller et al., 2003; van den Kerkhof et al., 2004).
A. van den Kerkhof et al. / Geoscience Frontiers 5 (2014) 683e695 691Laser ablation inductively coupled plasma mass spectrometry
(LA-ICPMS) analysis can be used for measuring Al, Ti, Fe, Li, Th, U,
and Ge in quartz for spot sizes of about 100 mm.We used a VG PQ2þ
mass spectrometer equipped with an S-option to raise sensitivity.
The laser ablation system is a 266 nm Nd: YAG VG UV microprobe.
The detection limits in order of the detectability (ppm) are Al (9), Fe
(4), Ti (3), Li (2), Ge (2), U (0.1) and Th (0.05). These values are in
agreement with the findings of Flem et al. (2002) and substantially
better compared to the EPMA results. The laser ablation craters are
30e50 mm in diameter and 100e150 mm deep. Subsequently, this
method has a lower spatial resolution compared to EPMA. On the
other hand, LA-ICPMS has the advantage of eliminating some of the
distortion effects implicit for EPMA and the higher sensitivity of
measuring some other trace elements. In this respect bothmethods
are complementary.
4.2. Microtextures in cathodoluminescence
Granulites show a wide variety of CL-contrasting textures,
which in part may correlate with textures visible under the mi-
croscope, such as micro-fractures and fluid inclusions (Frentzel-
Beyme, 1989; van den Kerkhof and Hein, 2001; van den Kerkhof
et al., 2004). These microtextures are thought to form during
different stages of uplift and show a number of distinguishing
features (see also Fig. 6):
(1) Diffusive textures around grain boundaries or around micro-
fractures characterized by reduced CL-intensity. Concentrations
of Ti and Al were found to be lower in these zones (van den
Kerkhof et al., 2004). Migrating fluids may have taken up
theseelements fromthequartz in a zoneofup to ca.100mmfrom
the fluid-quartz phase boundary.Figure 6. Examples of (a) optical CL recorded for granulite-facies rocks from the Bamble S
terrane (northern Victoria Land, Antarctica) (Sample: F. Talarico; van den Kerkhof et al., 1993
3 ¼ euhedral quartz nuclei; 4 ¼ open, non-healed fractures (see text for explanation).(2) Healed fractures and patches of secondary quartz (non-CL) are
part of a typical dark contrasting pattern consisting of sec-
ondary quartz interconnected by fine healed fractures (e.g.
Behr and Frentzel-Beyme, 1987). These textures have been
referred to as “crackling” textures (Frentzel-Beyme, 1989) or
“decrepitation” textures (Behr, 1989), and are assumed to form
by the dissolution-precipitation of quartz during brittle defor-
mation. The pattern is characterized by dark, practically non-
luminescent quartz patches connected by dark thin lines
(healed fractures). Sometimes several generations of crackling
patterns can be distinguished with later structures showing a
cross-cutting relationship with the older structures and the
younger textures normally being darker than the older ones.
EPMA analysis indicates that the dark patches have reduced
trace element concentrations, particularly in Ti and Al. The host
quartz around the dark patches often shows slightly elevated
concentrations of Fe and K.
(3) Recrystallized quartz shows idiomorphic growth nuclei of
secondary quartz (dark CL). These textures are assumed to
have formed by recrystallization on the cooling of quartz
with overall high trace element contents (van den Kerkhof
et al., 2004). Almost pure quartz nuclei, with notably low
Ti and Al, are assumed to grow by the replacement of
impure quartz. The higher stability of pure quartz is sup-
posed to be the driving force for this quartz growth at low
temperatures.
(4) Open, non-healed fractures are late structures, which formed at
low temperatures and have no CL contrasts.
Structural H2O in quartz is manifested by the strong reddish CL
emission, which typically increases the intensity by modification of
the structural defects in the electron beam (e.g. Stenina et al., 1984;ector, southern Norway and (b) SEM-CL from granulite-facies rocks from the Wilson
). 1 ¼ diffuse alteration textures; 2 ¼ secondary quartz in patches and healed fractures;
A. van den Kerkhof et al. / Geoscience Frontiers 5 (2014) 683e695692Kalceff and Phillips, 1995; Müller, 2000). This type of luminescence
has commonly been observed in hydrothermal and granitic quartz,
but is exceptional in high-grade metamorphic quartz. Here, tex-
tures around aqueous fluid inclusions are typically dark in CL and
indicate reduced trace elements in the host quartz (e.g. van den
Kerkhof and Hein, 2001; van den Kerkhof et al., 2004). It is
thought that these trace elements are enriched along the grain
boundaries and in the fluid inclusions. Carbonic inclusions typically
do not show halo structures under CL (Behr, 1989), which suggests
that small amounts of H2O originally present in the carbonic fluid
must be still present in the inclusions. These small amounts of H2O
normally cannot be observed under the microscope. Experiments
of Doppler et al. (2013) show possible fast diffusion of H2O at
temperatures above 450 
C, if the fugacity gradient is sufficiently
high. In natural rocks this would imply that pore spaces are H2O
saturatedwith an “infinite” fluid source. However, as the porosity in
high-grade metamorphic rocks is very low, the re-equilibration of
fluid inclusions with external fluids can be neglected. On the other
hand, our observations from CL show that trace elements from the
host quartz are able to diffuse into the fluid inclusions during
cooling.Figure 7. Trace element plot (Ti, Al) with application of the Ti-in-quartz geothermometer (W
southern Norway (see also van den Kerkhof et al., 2004) are shown. The granulite-facies tem
this sample are shown in Figs. 3b and 4a.4.3. Trace element-in-quartz geothermometry
Trace elements in quartz can be used as an independent ther-
mometer and compared with isochore calculations obtained from
fluid inclusion studies. Several tests involving Al in quartz have been
made (e.g. Scotford, 1975) but its application as a geothermometer
does not give reliable results. The Ti-in-quartz geothermometer
(Wark andWatson, 2006) is more reliable provided that rutile is also
present in the rock (Ti saturation). The detection limits formeasuring
Ti are in the range 18e43 ppm for EPMA and ca. 3 ppm for LA-ICPMS
(van den Kerkhof et al., 2004), which corresponds to 575e650 
C and
450 
C, respectively. This implies that the Ti-in-quartz geo-
thermometer can be best applied under high-grade conditions.
If the Ti-in-quartz thermometer is applied to that sample from the
Bamble granulite whose fluid inclusions contain the highest CO2
densities (sample FK5T, see above), an overall quartz formation tem-
peratureof810e850 
C(164e218ppmTi) isestimated for theprimary
quartz (Fig. 7). Secondaryquartz showsmuch lowerTi concentrations,
which point to temperatures below ca. 650 
C, i.e. in the amphibolite-
facies range down to greenschist facies. On the other hand, Fe is
somewhat higher in the secondary quartz and around fractures.ark and Watson, 2006). Data from one granulite-facies sample from the Bamble Sector,
perature range as revealed from trace elements is 810e850 
C. Fluid inclusion data for
A. van den Kerkhof et al. / Geoscience Frontiers 5 (2014) 683e695 6935. Discussion
5.1. The preservation of peak-metamorphic fluids
Carbonic fluids trapped in the continental crust at 30e35 km
depth are, even when rare, almost always preserved in granulite-
facies rocks on the Earth’s surface. The density of theseFigure 8. Re-equilibration of high-density (peak-metamorphic) fluid inclusions during up
trapping of fluids in clusters (1, 2 and 3), depending on the inclusion size, i.e. larger inclus
chore which defines the pressure in the inclusion on further cooling. Inset: Estimated maxinclusions, being typically 1.11e1.17 cm3/mol, have pressures of
80e130 MPa at room temperature (Fig. 8, inset). The size of the
inclusions ranges between 5 and 10 mm and the pressures recorded
for these inclusions is in accordance with the maximum pressure
that is required to initiate non-elastic deformation in the quartz
around fluid inclusions of this size (Bodnar et al., 1989) and
therewith may trigger decrepitation. As the maximum pressure oflift projected on to a PeT field (not to scale). Decompression results in stepwise re-
ions may leak earlier than smaller ones. Each newly trapped fluid defines a new iso-
imum pressures in fluid inclusions in quartz (after Bodnar et al., 1989).
A. van den Kerkhof et al. / Geoscience Frontiers 5 (2014) 683e6956945e10 mm inclusions in quartz is ca. 170e220 MPa, none of these
inclusions can survive heating up to 150 
C. During uplift, however,
the pressure difference between the inclusion and lithostatic
pressures (possibly 400e500 MPa) normally must far exceed these
values. In spite of the fluid overpressure, a few inclusions always
survive even these extreme conditions.
The uplift path can be complicated and characterized by meta-
morphic overprints. Retrograde PeT paths, characterised by
decompression-cooling (e.g. Touret and Huizenga, 2012), are
mostly marked in high-pressure granulites rather than isothermal
decompression. However, in high-temperature and especially
ultra-high-temperature granulites, rapid decompression is very
late (below ca. 500 
C) and preceded by a long period of isobaric
cooling. In the latter case, modified inclusions show a density in-
crease (Touret, 1981). In the case of (isothermal) decompression,
conditions corresponding to the normal continental geothermal
gradient of ca. 25 
C/kmmay change to steeper gradients of around
50 
C/km. During this uplift most of the carbonic inclusions will be
retrapped in inclusions of lower density, arranged in clusters as
observed in many granulites. The maximum pressure under which
the fluid inclusion remains largely intact depends on the inclusion
diameter. Larger inclusions will leak earlier than smaller ones. The
pressure difference between fluid inclusions and the ambient
pressure can be estimated from the isochore calculations. The
maximum inclusion pressure is on the order of 150 to 250 MPa
above this pressure and depends on the inclusion size and shape
(Bodnar et al., 1989). Fluid inclusions with an initial equal density
but different size may therefore leak at different stages and form
secondary fluid inclusions with a high variability in density, which
define different isochores. Fig. 8 shows an example of fluid inclu-
sion leakage in 3 steps, which results in a high variability in density
for the newly formed inclusions. This is in accordance with the
observation of high variability in density in fluid inclusion clusters,
which have also preserved high-density inclusions. The majority of
the carbonic fluid inclusions are assumed to re-equilibrate during
the first episode of uplift.
5.2. The retrograde development of aqueous inclusions and
cathodoluminescence structures
Aqueous fluid inclusions in granulites will be equilibrated under
low PT conditions, corresponding roughly to 5e10 km depth. Only
primary inclusions that have been “protected” in the mineral host
are sometimes preserved, such as fluid inclusions in quartz in-
clusions in garnet. Re-equilibration normally occurs “in situ”, with no
change in the salinity. Sometimes even the difference between
seawater or fresh H2O can be identified, e.g. the Dabie Shan eclogites,
which has been confirmed by stable isotope oxygen geochemistry
(Fu, 2002). This indicates that during final equilibration under low PT
conditions does not indicate that fluids have been necessarily
introduced into the rock system at this moment. The only argument
is the timing of inclusion formation in the mineral host. Micro-
textures observed by CL techniques show, in part, direct correlation
with aqueous fluid inclusions, notably the healed fractures and
patches. Aqueous inclusions are typically trapped in secondary
quartz characterised by very low trace element concentrations.
Correlation of carbonic inclusions with CL-textures has not been
observed (see also Behr, 1989). Many aqueous inclusions in high-
grade rocks typically show very high densities, i.e. showing small
or no bubbles and cannot be trapped under high-grademetamorphic
conditions. Homogenisation temperatures for these inclusions are
mostly far below 200 
C. On the other hand, isochores corresponding
with homogenisation temperatures of around 200 
C run across
granulite-facies metamorphic conditions. However, microtextures
demonstrate that they cannot be primary. Therefore it can beconcluded that aqueous inclusions (with Th< 400 
C) are always re-
equilibrated at relatively low pressures during uplift when adapting
to the mean continental geothermal gradient. These conditions
correspond to the greenschist facies (300e450 
C) and lower. Sec-
ondary quartz with trapped aqueous inclusions may show different
generations with clear timing relations. Even when the trace
element contents are too low for doing temperature estimates, it has
been observed that under CL the darker textures (lower tempera-
tures) are always younger than the less dark ones (higher temper-
ature). This observation demonstrates that aqueous inclusions
normally form as a result of re-equilibration during cooling. The
massive recrystallization of granulite quartz is evident from the
occurrence of idiomorphic growth nuclei, visible as darkly con-
trasting quartz crystals in the bright luminescent host quartz. These
textures are assumed to form by the growth of almost pure quartz
crystals replacing impure high-temperature quartz. This recrystalli-
zation is assumed to be triggered by the formation of growth nuclei
around aqueous fluid inclusions. Elevated Na-concentrations found
in recrystallized quartz (van den Kerkhof et al., 2004) support this
interpretation, which implies that the fluid inclusion contents are up
taken in the newly formed quartz after complete disintegration
(healing) of the fluid inclusion.
5.3. Concluding remarks
Extending fluid inclusion petrography and microthermometry
with CL techniques and trace element analysis has shown to be
helpful for deciphering the complex fluid inclusion inventory found
in high-grade metamorphic rocks. It has been demonstrated that a
number of microtextures directly relate to aqueous fluid inclusions,
but that the interpretation of carbonic fluid inclusions almost totally
relies on “traditional” fluid inclusion petrography. A part of the peak-
metamorphic carbonic or aqueous-carbonic fluid inclusions almost
always survives uplift. However, recognizing “primary” fluids is
much more difficult than for other rock types like hydrothermal or
diagenetic rocks. Textures typical for “primary” fluid inclusions, like
negative crystal shape and the isolated or cluster arrangements are
found in a number of igneous charnockites, but in most meta-
morphic granulites the earliest fluid inclusions are “hidden” in fluid
inclusion clusters amongst decrepitated inclusions with highly
varying densities. In the latter case detailed “fluid inclusion map-
ping” of selected samples appears to be quite helpful for finding
relics of the peak-metamorphic fluids.
Acknowledgements
The authors would like to thank the critical reviews and
constructive remarks of Prof. (em.) Jacques L. R. Touret and Prof.
Jan-Marten Huizenga. The editorial work and English corrections
made by Dr. D. E. Harlov are highly appreciated.
References
Bakker, R.J., 2003. Package FLUIDS; 1. Computer programs for analysis of fluid in-
clusion data and for modelling bulk fluid properties. Chemical Geology 194
(1e3), 3e23.
Bakker, R.J., Jansen, J.B.H., 1993. Calculated fluid evolution path versus fluid inclu-
sion data in the COHN system as exemplified by metamorphic rocks from
Rogaland, South-west Norway. Journal of Metamorphic Geology 11-3, 357e370.
Behr, H.J., 1989. Die geologische Aktivität von Krustenfluiden (Gesteinsfluide). Nie-
dersächsische Akademie der Geowissenschaften Veröffentlichungen 1, 7e41.
Behr, H.J., Frentzel-Beyme, K., 1987. Permeability and paleoporosity in crystalline
bedrocks of the Central European Basement e studies of cathodoluminescence.
In: Behr, H.J., Raleigh, C.B. (Eds.), Exploration of the Deep Continental Crust, vol.
2. Springer, Berlin-Heidelberg New York, pp. 477e497.
Bodnar, R.J., Binns, P.R., Hall, D.L., 1989. Synthetic fluid inclusions VI: quantitative
evaluation of the decrepitation behaviour of fluid inclusions in quartz at one
atmosphere confining pressure. Journal of Metamorphic Geology 7, 229e242.
A. van den Kerkhof et al. / Geoscience Frontiers 5 (2014) 683e695 695Coolen, J.J.M.M.M., 1980. Chemical Petrology of the Furua Granulite Complex,
Southern Tanzania. GUA papers of Geology. Universiteit van Amsterdam. Ser. 1,
n
 13, 258pp.
Crawford, M.L., 1981. Fluid inclusions in metamorphic rocks e low and medium
grade. In: Hollister, L.S., Crawford, M.L. (Eds.), Short Course in Fluid Inclusions:
Application to Petrology (Chapt. 7), vol. 6. Mineralogical Association of Canada,
Calgary, pp. 157e181.
Doppler, G., Bakker, R.J., Baumgartner, M., 2013. Fluid inclusion modification by H2O
and D2O diffusion: the influence of inclusion depth, size, and shape in re-
equilibration experiments. Contributions to Mineralogy and Petrology 165,
1259e1274.
Flem, B., Larsen, R.B., Grimstvedt, A., Mansfeld, J., 2002. In situ analysis of trace
elements in quartz by using laser ablation inductively coupled plasma mass
spectrometry. Chemical Geology 182, 237e247.
Frentzel-Beyme, K., 1989. REM-Kathodolumineszenz-Strukturen im Quarzteilgefüge
von Metamorphiten: Bestandaufnahme und geologische Interpretation. Ph.D.
thesis. University of Göttingen, 78 pp.
Fu, B., 2002. Fluid Regime during High- and Ultrahigh-Pressure Metamorphism in
the Dabie-Sulu Terranes, Eastern China. Ph.D. dissertation. Free University,
Amsterdam, 144 pp.
Fu, B., Touret, J.L.R., Zheng, Y.-F., Jahn, B.-M., 2003. Fluid inclusions in granulites,
granulitized eclogites and garnet clinopyroxenites from the Dabie-Sulu ter-
ranes, eastern China. Lithos 70 (3e4), 293e319.
Harlov, D.E., Johansson, L., van den Kerkhof, A.M., Förster, H.J., 2006. Transformation
of a granitic gneiss to charnockite, Söndrum Stenhuggeriet, Halmstad, SW
Sweden: the role of advective fluid flow and diffusion during localised dehy-
dration. Journal of Petrology 47 (1), 3e33.
Harlov, D.E., van den Kerkhof, A.M., Johansson, L., 2013. The Varberg-Torpa char-
nockite-granite association, SW Sweden: mineralogy, petrology, and fluid in-
clusion chemistry. Journal of Petrology 54 (1), 3e40.
Harlov, D.E., van den Kerkhof, A.M., Johansson, L., 2014. High-grade localized
metasomatic alteration of the granitic gneiss surrounding a clinopyroxene-rich
pegmatoid dyke: Söndrum Stenhuggeriet, Halmstad, SW Sweden. Journal of
Metamorphic Geology 32 (4), 389e416.
Herms, P., Schenk, V., 1998. Fluid inclusions in high-pressure granulites of the Pan-
African Belt in Tanzania (Uluguru Mts); a record of prograde to retrograde fluid
evolution. Contributions to Mineralogy and Petrology 130 (3e4), 199e212.
Huizenga, J.-M., 2001. Thermodynamic modelling of C-O-H fluids. Lithos 55 (1e4),
101e114.
Kalceff, S.M.A., Phillips, M.R., 1995. Cathodoluminescence microcharacterization of
the defect structure of quartz. Physical Review B 52/5, 3122e3134.
Kooi, M.E., Schouten, J.A., van den Kerkhof, A.M., Istrate, G., Althaus, E., 1998. The
system CO2-N2 at high pressure and applications to fluid inclusions. Geochimica
et Cosmochimica Acta 62 (16), 2837e2843.
Müller, A., 2000. Cathodoluminescence and Characterization of Defect Structures in
Quartz with Applications to the Study of Granitic Rocks. Ph.D. dissertation.
University of Göttingen, 222 pp.
Müller, A., Wiedenbeck, M., van den Kerkhof, A.M., Kronz, A., Simon, K., 2003. Trace
elements in quartz; a combined electron microprobe, secondary ion mass
spectrometry, laser-ablation ICP-MS, and cathodoluminescence study. Euro-
pean Journal of Mineralogy 15 (4), 747e763.
Neuser, R.D., Bruhn, F., Götze, J., Habermann, D., Richter, D.K., 1995. Kathodolumi-
neszenz: Methodik und Anwendung. Zentralblatt für Geologie und Paläonto-
logie Teil I, H.1/2, 287e306.
Newton, R.C., 1986. Fluids of granulite facies metamorphism. In: Walther, J.V.,
Wood, B.J. (Eds.), Fluid-rock Interactions during Metamorphism (Chapter 2).
Springer-Verlag Berlin Heidelberg, Berlin, pp. 36e59.
Passchier, C., Trouw, R., 2005. Microtectonics. Springer-Verlag Berlin Heidelberg,
Berlin, p. 382.
Richter, D., Simmons, G., 1977. Microcracks in crustal igneous rocks; microscopy. In:
Heacock, J.G., Keller, G.V., Oliver, J.E., Simmons, G. (Eds.), The Earth’s Crust; its
Nature and Physical Properties, Geophysical Monograph, vol. 20, pp. 149e180.
Santosh, M., Tsunogae, T., 2003. Extremely high density pure CO2 fluid inclusions in
a garnet granulite from Southern India. Journal of Geology 111, 1e16.
Schreurs, J., 1985. The West-Uusimaa Low-pressure Thermal Dome. Ph D. disser-
tation. Vrije Universiteit Amsterdam, 179 pp.
Scotford, D.M., 1975. A test of aluminium in quartz as a geothermometer. American
Mineralogist 60, 139e142.
Shmulovish, K.I., Tkachenko, S.I., Plyasunova, N.V., 1994. Phase equilibria in fluid
systems at high pressures and temperatures. In: Shmulovish, K.I.,
Yardley, B.W.D., Gonchar, G.G. (Eds.), Fluids in the Crust: Equilibrium and
Transport Properties (Chapter 8). Chapman & Hall, London, pp. 193e214.
Sorby, H.C., 1858. On the microscopic structure of crystals, indicating the origin of
minerals and rocks, 1st part Geological Society of London Quarterly Journal 14,
453e500.
Span, R., Wagner, W., 1996. A new equation of state for carbon dioxide covering the
fluid region from the triple point temperature to 1100 K at pressures up to 800
MPa. Journal of Physical and Chemical Reference Data 25, 1509e1596.Stenina, N.G., Bazarov, L.S., Shcherbakova, M.Y., Mashkovtsev, R.I., 1984. Structural
state and diffusion of impurities in natural quartz of different genesis. Physics
and Chemistry of Minerals 10, 180e186.
Sterner, S.M., Bodnar, R.J., 1989. Synthetic fluid inclusion; VII, re-equilibration of
fluid inclusions in quartz during laboratory simulated metamorphic burial and
uplift. Journal of Metamorphic Geology 7 (2), 243e260.
Swanenberg, H.E.C., 1980. Fluid inclusions in high-grade metamorphic rocks from SW
Norway. PhD thesis, 25. University of Utrecht. Geologica Ultraiectina, 147 pp.
Touret, J.L.R., 1971a. Contrôle du facies granulite dans le Bamble par le CO2 de la
phase fluide. Compte rendu sommaire des séances de la société géologique de
France 6, 143e144.
Touret, J.L.R., 1971b. Le facies granulite en Norvege méridionale II: Les inclusions
fluides. Lithos 4, 423e436.
Touret, J.L.R., 1981. Fluid inclusions in high grade metamorphic rocks. In:
Hollister, L.S., Crawford, M.L. (Eds.), Short Course in Fluid Inclusions: Applica-
tion to Petrology (Chapt. 8), vol. 6. Mineralogical Association of Canada, Calgary,
pp. 182e208.
Touret, J.L.R., 1985. Fluid regime in southern Norway: the record of fluid inclusions.
In: Tobi, A.C., Touret, J.L.R. (Eds.), The Deep Proterozoic Crust in the North
Atlantic Provinces. NATO ASI Series. Series C: Mathematical and Physical Sci-
ences, vol. 158, pp. 517e549.
Touret, J.L.R., 1987. Fluid inclusions and pressure-temperature estimates in deep-
seated rocks. In: Helgeson, H.C. (Ed.), Chemical Transport in Metasomatic Pro-
cesses. NATO ASI Series, C218, D. Reidel Publishers, Dordrecht, pp. 91e122.
Touret, J.L.R., 1995. Brines in granulites: the other fluid (ECROFI XIII). Boletín de la
Sociedad Española de Mineralogía 18-1, 250e251.
Touret, J.L.R., 2001. Fluids in metamorphic rocks. Lithos 55, 1e25.
Touret, J.L.R., Dietvorst, P., 1983. Fluid inclusions in high-grade anatectic meta-
morphites. Journal of the Geological Society 140, 635e649.
Touret, J.L.R., Hansteen, T.H., 1988. Geothermobarometry and fluid inclusions in a
rock from the Doddabetta charnockite complex, Southwest India. Rendiconti
Società Italiania Mineralogia e Petrologia 43, 65e82.
Touret, J.L.R., Huizenga, J.M., 2012. Charnockite microstructures: from magmatic to
metamorphic. Geoscience Frontiers 3 (6), 745e753.
Touret, J.L.R., Nijland, T.G., 2012. Prograde, peak and retrograde metamorphic fluids
and associated metasomatism in upper amphibolite to granulite facies transi-
tion zones. In: Harlov, D.E., Austrheim, H. (Eds.), Metasomatism and the
Chemical Transformation of Rock e The Role of Fluids in Terrestrial and
Extraterrestrial Processes (Chapter 11). Springer-Verlag Berlin Heidelberg.
Tsunogae, T., Santosh, M., Osanai, Y., Owada, M., Toyoshima, T., Hokada, T.,
2002. Very high-density carbonic fluid inclusions in sapphirine-bearing
granulites from Tonagh Island in the Archean Napier Complex, East
Antarctica: implications for CO2 infiltration during ultrahigh-temperature
(T>1100
C) metamorphism. Contributions to Mineralogy and Petrology
143 (3), 279e299.
van den Kerkhof, A.M., Hein, U.F., 2001. Fluid inclusion petrography. Lithos 55, 27e47.
van den Kerkhof, A.M., Touret, J.L.R., Maijer, C., Jansen, J.B.H., 1991. Retrograde
methane-dominated fluid inclusions from high-temperature granulites of
Rogaland, SW Norway. Geochimica et Cosmochimica Acta 55, 2533e2544.
van den Kerkhof, A.M., Frezzotti, M.-L., Talarico, F., Berdnikov, N.V., 1993. Metastable
phase transitions in the system CO2-CH4-N2 including immiscibility of the
undercooled liquid (L1L2G): an example from granulites of the Wilson Terrane,
North Victoria Land (Antarctica). XLIX, z Archiwum Mineralogiczne T 1,
227e229.
van den Kerkhof, A.M., Kooi, M.E., Schouten, J.A., Istrate, G., Althaus, E., 1998. The
system CO2-N2 at high pressure and applications to fluid inclusions. Geochimica
et Cosmochimica Acta 62e16, 2837e2843.
van den Kerkhof, A.M., Kronz, A., Simon, K., Scherer, T., 2004. Fluid-controlled
quartz recovery in granulite as revealed by cathodoluminescence and trace
element analysis (Bamble Sector, Norway). Contributions to Mineralogy and
Petrology 146 (5), 637e652.
Vityk, M.O., Bodnar, R.J., 1995. Textural evolution of synthetic fluid inclusions in
quartz during re-equilibration, with applications to tectonic reconstruction.
Contributions to Mineralogy and Petrology 121-3, 309e323.
Vityk, M.O., Bodnar, R.J., 1998. Statistical microthermometry of synthetic fluid in-
clusions in quartz during decompression reequilibration. Contributions to
Mineralogy and Petrology 132-2, 149e162.
Wark, D.A., Watson, E.B., 2006. TitaniQ: a titanium-in-quartz geothermometer.
Contributions to Mineralogy and Petrology 152, 743e754.
Winkler, H.G.F., 1965. Petrogenesis of Metamorphic Rocks. Springer-Verlag, Berlin.
Yardley, B.W.D., Shmulovich, K.I., 1995. An introduction to crustal fluids. In:
Shmulovish, K.I., Yardley, B.W.D., Gonchar, G.G. (Eds.), Fluids in the Crust:
Equilibrium and Transport Properties, pp. 1e12.
Zhang, Yi-G., Frantz, J.D., 1989. Experimental determination of the compositional
limits of immiscibility in the system CaCl2-H2O-CO2 at high temperatures and
pressures using synthetic fluid inclusions. Chemical Geology 74, 289e308.