Content uploaded by H.-J. Förster
Author content
All content in this area was uploaded by H.-J. Förster on Nov 08, 2017
Content may be subject to copyright.
Apatite–monazite relations in the Kiirunavaara
magnetite–apatite ore, northern Sweden
Daniel E. Harlov
a,
*, Ulf B. Andersson
a,c
, Hans-Ju
¨rgen Fo¨rster
a,b
, Jan Olov Nystro¨m
c
,
Peter Dulski
a
, Curt Broman
d
a
GeoForschungsZentrum Potsdam, Telegrafenberg, D-14473 Potsdam, Germany
b
Institute of Earth Sciences, University of Potsdam, D-14415 Potsdam, Germany
c
Naturhistoriska Riksmuseet, Box 50007, SE-10405 Stockholm, Sweden
d
Institute of Geology and Geochemistry, Stockholm University, SE-10691 Stockholm, Sweden
Abstract
The magnetite–apatite ores in the Kiruna area, northern Sweden, are generally considered to be of magmatic origin formed
in a subvolcanic–volcanic environment during the early Proterozoic. They are thought to have crystallised from volatile-rich
iron oxide magmas derived by immiscibility in calc-alkaline to slightly alkaline parental magmas. Three major morphological
types of the magnetite –apatite ore (primary, brecciated, and banded) have been investigated for textural relations and mineral
chemistry using transmitted light, back-scattered electron imaging (BSE), electron microprobe analysis (EMPA), and laser
ablation–inductively coupled plasma-mass spectrometry (LA–ICPMS). In all three types, Th- and U-poor monazite is present
as small inclusions in the apatite. Larger (up to 150 Am) recrystallised monazite grains, both along apatite grain boundaries and
intergrown with magnetite and silicate minerals, are present in the brecciated and banded samples. Primary apatite grains,
without monazite inclusions, are generally enriched in light rare earth elements (LREEs) together with Na and Si.
Petrological and mineralogical evidence suggest that the Kiruna magnetite–apatite ore experienced successive stages of
fluid–rock interaction. The first stage occurred under high-temperature conditions (700 – 800 jC) shortly after emplacement and
crystallisation of the ore magmas and involved concentrated, probably Cl-dominated brines expelled from the magma. This
fluid is held to be responsible for the nucleation of the numerous small monazite inclusions within the apatite due to high-
temperature leaching of Na and Si, while the LREEs were concentrated in the monazite. The large monazite grains in the
brecciated and banded samples are proposed to be the product of recrystallisation from the much smaller monazite inclusions.
During greenschist-facies metamorphism (T= 300 –400 jC), fluids from the surrounding country rock caused strong
(LREE +Na + Si) depletion along apatite grain boundaries and cracks in the apatite. LREEs were either redeposited as monazite
grains along apatite grain boundaries or were flushed out of the ore. This fluid interaction also introduced the silicate
components responsible for the interstitial formation of allanite, talc, tremolite, chlorite, serpentine, muscovite, quartz, and
carbonates along apatite grain boundaries.
D2002 Elsevier Science B.V. All rights reserved.
Keywords: Kiruna; Apatite; Magnetite; Monazite; Brines; Orebody
0009-2541/02/$ - see front matter D2002 Elsevier Science B.V. All rights reserved.
PII: S 0009-2541(02)00148-1
*
Corresponding author.
E-mail address: dharlov@gfz-potsdam.de (D.E. Harlov).
www.elsevier.com/locate/chemgeo
Chemical Geology 191 (2002) 47 – 72
1. Introduction
The major proportion of the lanthanides in both
intermediate to felsic igneous and metamorphic rocks
are contained in the accessory minerals, among which
the phosphates often play a crucial role (e.g. Bea,
1996; Bingen et al., 1996; Pan and Fleet, 1996; Bea
and Montero, 1999; Hoskin et al., 2000). The most
common are fluorapatite, Ca
10
(PO
4
)
6
F
2
(generally
referred to as apatite), monazite, (Ce, LREE)PO
4
,
and, more rarely, xenotime (Y,HREE)PO
4
. The min-
eralogical relationship between apatite and these Y+
rare earth elements (Y + REE) phosphate minerals is
variable and not always well understood. In addition
to forming genetically independent minerals, they
sometimes form close associations.Previously, a
number of workers have noted the presence of typi-
cally small ( < 1 10 Am) grains of monazite and/or
xenotime in metamorphic rocks both as inclusions
within the apatite as well as along apatite grain
margins (e.g. Pan et al., 1993; Pan, 1997; Fo¨rster
and Harlov, 1999; Harlov and Fo¨rster, 2002). These
monazite grains are generally characterised by very
low abundances of Th and U and relatively low La/Nd
ratios (e.g. Pan et al., 1993; Harlov and Fo¨ rster, 2002).
Moreover, these metamorphic rocks show evidence
for interaction with fluids under relatively high-tem-
perature conditions (T>500–600jC) (Pan et al.,
1993; Harlov and Fo¨rster, 2002). Studies of natural
systems (e.g. A
˚mli, 1975; Pan et al., 1993; Harlov and
Fo¨rster, 2002) as well as experimental studies (e.g.
Harlov et al., 2002; Harlov and Fo¨rster, unpublished
data) indicate that these inclusions have originated in
the apatite itself, from the (Y + REE) budget available,
as the product of coupled substitution and mass trans-
fer during metasomatic alteration. There is no evi-
dence to suggest that they are the result of exsolution
in the apatite during cooling or that they could
represent independent grains later overgrown by the
apatite (e.g. A
˚mli, 1975; Pan et al., 1993; Harlov and
Fo¨rster, 2002; Harlov et al., 2002).
Inclusions of monazite and xenotime in apatite
have been reported in metamorphosed pegmatites
(A
˚mli, 1975), intermediate and mafic granulites (Har-
lov and Fo¨rster, 2002), hydrothermally altered schists
associated with Au deposits (Pan et al., 1993; Pan,
1997), and a highly metamorphosed magnetite – apa-
tite ore deposit (McKeown and Klemic, 1957). Mon-
azite inclusions in apatite have also been briefly noted
in Kiruna-type magnetite – apatite ores of Proterozoic
age including those at Kiruna itself (e.g. Para
´k, 1973,
1975a,b) as well as Pea Ridge, Missouri (Kerr and
Samson, 1998; Kerr, 1998). Magnetite–apatite ores of
the Kiruna-type occur worldwide with formation ages
ranging from the early Proterozoic to the Pliocene.
They show strong evidence of having formed from
volatile-rich iron oxide magmas which separated as
immiscible iron-rich melts from calc-alkaline to
slightly alkaline parental magmas during cooling
(Henrı
´quez and Martin, 1978; Nystro¨m and Henrı
´-
quez, 1994; Frietsch and Perdahl, 1995; Naslund et
al., 2000).
This study reports on and discusses the textures and
composition of phosphate, oxide, carbonate, and sili-
cate minerals for a series of texturally different samples
from magnetite –apatite iron ore deposits in the Kiruna
area, northern Sweden. Particular attention is paid to
the relationship between apatite and monazite. This is
investigated using transmitted light, back scattered
electron imaging (BSE), electron microprobe analysis
(EMPA), and laser ablation– inductively coupled
plasma-mass spectrometry (LA – ICPMS). Links are
established between the most probable metasomatising
fluids (as well as the origin of such fluids) and the
formation of Th-poor monazite inclusions ( < 1–10
Am) in the apatite as well as later stage fluids and the
depletion of Light rare earth elements (LREE) along
apatite grain rims. Lastly, compositionally identical but
much larger (10–150 Am) monazite grains, found
intergrown with magnetite and silicates as well as along
apatite grain boundaries, are discussed.
2. Geological setting
The Kiirunavaara deposit and other iron ores in the
Kiruna area of northern Sweden (1880 1890 Ma;
Cliff et al., 1990; Romer et al., 1994) represent one of
the greatest concentrations of magnetite – apatite ore in
the world (Fig. 1). The ores occur as tabular bodies
intercalated in the upper part of a thick volcanic
sequence dominated by acid pyroclastic rocks known
as the Porphyry Group (Geijer, 1910, 1931, 1967;
Frietsch, 1978, 1984). Similar (Y +REE) patterns for
the magnetite, apatite, and volcanic host rock sur-
rounding the deposits support a magmatic origin
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–7248
Fig. 1. Iron deposits in the Kiruna area, northern Sweden (simplified from Para
´k, 1975a) with sample locations. The tabular ore bodies dip 50 –
70jto the east and are hosted by acid to intermediate volcanic rocks of the Porphyry Group. The Lower Hauki volcanics consist of pyroclastic
and sedimentary rocks.
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–72 49
(Frietsch and Perdahl, 1995). Additional evidence
presented by Para
´k (1975a,b) supports the hypothesis
that the ores formed in a volcanic system due to the
extrusion of iron oxide lava and ash as well as
emplacement of subvolcanic orebodies (see also
Nystro¨m and Henrı
´quez, 1994). The Kiirunavaara
orebody (along with Luossavaara, Fig. 1) overlies
trachyandesitic lavas traditionally referred to as syen-
ite porphyries (cf. Geijer, 1910). These, in turn, are
overlain by rhyodacitic ignimbrites which are usually
termed quartz porphyries. Iron deposits located near
the top of the Porphyry Group (Rektorn, Haukivaara,
Henry and Nukutusvaara) differ from Kiirunavaara
and Luossavaara in that they have higher apatite and
hematite contents as well as significant amounts of
quartz and carbonate (cf. Frietsch, 1967). The ubiq-
uitous presence of magnetite relicts in the hematite,
however, shows that the latter is a secondary oxidation
product. Hydrothermal overprinting at lower temper-
atures is expressed by the presence of pyrite and
gypsum in the form of veinlets and individual grains
in what were originally porous regions of the ore.
However, there is no evidence that any part of the
orebody itself has been hydrothermally deposited
(Nystro¨m and Henrı
´quez, 1994).
The rocks of the Porphyry Group in the Kiruna
area and associated iron ores have been regionally
metamorphosed without penetrative deformation, i.e.
with preservation of primary volcanic structures and
textures (Nystro¨m and Henrı
´quez, 1994). Character-
istic minerals in the associated mafic rocks, e.g.
chlorite, zoisite/clinozoisite, epidote, actinolite, and
albite (Martinsson, 1997; Bergman et al., 2001), are
consistent with greenschist-facies metamorphism.
According to Frietsch (1984), the orebodies are spa-
tially connected with extensive fault zones that, to a
large extent, guided their emplacement. Reactivation
of the fault systems has resulted in local deformation
and recrystallisation. Moreover, since both quartz and
carbonates are most abundant in foliated regions of
the ore (see Figs. 52, 57 and many other photographs
in Para
´k, 1975a), they were likely introduced during
the deformation. Primary textures and structures sur-
vived as relicts in the largest deposits, especially in
Kiirunavaara, but are less common in the apatite-rich
and less competent smaller deposits at the top of the
Porphyry Group.
3. Petrography
Three different morphological types of the Kiruna
magnetite–apatite ore were investigated. These con-
sisted of two samples, in which the original texture
has been preserved, and four samples of the ore
affected by deformation. Five of the six samples come
from Kiirunavaara with the remaining one (PG-K5)
taken from Rektorn (Fig. 1). Samples lacking defor-
mation, and which could be termed ‘‘primary’’ (PG-
36:2 and KUJ-4), are characterised by dendrites of
magnetite in an apatite matrix (Nystro¨ m and Henrı
´-
quez, 1989). Of the four deformed samples, two are
breccias (PG-K6 and PG-K7) and consist of magnetite
fragments in apatite. The other two deformed samples
(PG-618 and PG-K5) can be described as tectonically
banded ore with alternating magnetite-rich, apatite-
poor and magnetite-poor, apatite-rich bands or streaks.
Hydrous secondary silicate minerals, consistent
with the regional greenschist facies metamorphism
that characterises the surrounding country rock, are
typically observed interstitially along apatite grain
boundaries in the ores. Chlorite occurs in samples
from all three ore types, talc occurs in the dendritic ore
and one of the brecciated ore samples (PG-K6), while
serpentine is typically present in the other brecciated
ore sample (PG-K7). Tremolite and muscovite, in
addition to quartz, are found in the banded ore.
3.1. The primary magnetite – apatite ore
The two nondeformed samples show different
textures. Sample PG-36:2 is from the so-called skel-
Fig. 2. Series of BSE photographs from sample PG-36:2 showing a large-scale view of the skeleton ore, a primary ore type with platy magnetite
dendrites (white) enclosed by apatite (grey) (a). Also shown are high-contrast images of the apatite between the magnetite (Mt) plates (b– d).
Numerous monazite inclusions are seen in the centre of most of the apatite grains. Note the sporadic dark areas (depleted in LREE; cf. Table 1) along
apatite grain boundaries, and apparent cracks in the apatite as well as patchiness in the apatite grain interiors (b). These dark areas are also typically
associated with sheet silicates (talc and chlorite) along the apatite grain boundaries (c) which show up as interstitial black regions along apatite grain
boundaries on the BSE images. In many of the apatite grains, the monazite inclusions are elongated and show a preferred orientation (d).
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–7250
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–72 51
eton ore (Geijer, 1910; Nystro¨m and Henrı
´quez, 1989,
1994). It consists of platy crystals of dendritic magnet-
ite enclosed in a matrix of apatite. The plates, a few
tenths to a few millimeters thick, tend to be oriented
either randomly or form subparallel arrays with the
plates locally displaying branching offshoots (Fig.
2a). Sample KUJ-4 comes from an irregular concen-
tration of apatite within the skeleton ore (cf. Nystro¨m
and Henrı
´quez, 1989, their Fig. 3). The magnetite
dendrites form scattered plates in the apatite but also
occur as fringes of columnar habit surrounding small
( < 1 cm in cross-section) concentrations of magnetite
(cf. Nystro¨m and Henrı
´quez, 1989, their Fig. 8). In
both samples, the magnetite is evenly peppered with
small (10 20 Am) inclusions of apatite (Fig. 2a).
Under high-contrast BSE imaging, the rims of a
majority of the apatite grains in both samples tend to
be darker than the core (Fig. 2b and c). These dark
rims are both discontinuous as well as varying in
thickness. In addition, the interior of the apatite grains
shows a certain patchiness in the form of alternating
bright and dark regions which can take the form of
linear zones related to tiny fractures in the apatite (Fig.
2b). In apatite from sample PG-36:2, inclusions of
monazite are common (Fig. 2b –d). These inclusions
are abundant inside the apatite but are less common in
dark areas and along grain boundaries (Fig. 2c and d).
However, most occur in areas of the apatite that
appear to be largely homogeneous under high-contrast
BSE imaging (Fig. 2d). In many cases, the monazite
inclusions show varying degrees of parallel elongation
along a preferred direction within the apatite which
appear to be controlled by the apatite crystal structure
(Fig. 2d).
In contrast to PG-36:2, only a few of the apatite
grains in sample KUJ-4 have monazite inclusions and
then only sparsely scattered in the interior. However,
the apatite shows the same sort of darkening along
grain rims as well as the patchy interiors seen in PG-
36:2 (Fig. 3a). In addition, a selection of apatite grains
in KUJ-4 is observed which are relatively darker than
the neighbouring apatites (Fig. 3a). Allanite is observed
Fig. 3. High-resolution BSE photographs of primary ore sample KUJ-4 showing typical apatite grains (a) and inhomogeneous allanite growing
interstitially along apatite grain boundaries (b). Note the presence of dark areas along apatite grain boundaries, patchiness in the apatite interiors
as well as apatite grains which are relatively darker than neighbouring apatite grains. Dark areas are depleted in (LREE + Na + Si). Black areas
along apatite grain boundaries are talc and chlorite. The inhomogeneity in the allanite is due primarily to varying REE contents and not
variability in the abundance of Th which is below the microprobe detection limit.
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–7252
Fig. 3 (continued ).
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–72 53
only in sample KUJ-4, where it typically occurs inter-
stitially along apatite –apatite and apatite –magnetite
grain boundaries (Fig. 3b). In some cases, it is inti-
mately intergrown with talc, as well as, to a lesser
extent, chlorite and calcite.
3.2. Brecciated magnetite–apatite ore
Sample PG-K7 is made up of angular or brecciated
fragments of magnetite ore in apatite. It appears to be
the product of brittle deformation at relatively lower
temperatures compared to sample PG-K6. In sample
PG-K6, brecciation of the magnetite has resulted in
less distinct and sharply delimited fragments (Fig. 4a),
suggesting higher temperatures and more plastic con-
ditions than for PG-K7. PG-K6 also contains the same
interstitial silicate mineralogy as the primary samples,
namely talc and chlorite. Similar to the undeformed
primary ore samples, the magnetite in either sample
contains numerous inclusions of apatite ranging from
hundreds to < 10 Aminsize(Fig. 4a). In either
sample, the interior of the apatite grains contains
abundant monazite inclusions which show the same
textural patterns as those in the primary ore (Fig. 4b
and c). Under high-contrast BSE imaging, the rims of
most of the apatite again show discontinuous, dark-
ened rims of varying thickness as well as patchy
interiors (Fig. 4b). The rim areas and apatite grain
boundaries contain relatively fewer but occasionally
larger monazite grains (Fig. 4b).
Though not common, monazite also occurs as
relatively large (10 100 Am), isolated, relatively
homogeneous grains in both samples (Figs. 4b, c
and 5). These grains occur principally as interstitial
grains with respect to apatite (Fig. 4b and c) and, to a
lesser extent, are associated with sheet silicates such
as serpentine and chlorite. They are also present both
as rare skeletal grains intergrown with the brecciated
magnetite (Fig. 5a) as well as included in the magnet-
ite. Under high-contrast BSE imaging, the large mon-
azite grains show some faint zoning (Fig. 5b). This is
typical of large monazite grains in the deformed
samples in general.
In addition to serpentine along interstitial apatite
grain boundaries, a unique feature of sample PG-K7 is
the presence of patches (Fig. 6a) and veins of dolo-
mite, which contain numerous, evenly distributed,
small (1 10 Am) grains of magnetite and monazite,
some of which are intergrown with each other (Fig.
6b).
3.3. Banded magnetite–apatite ore
Samples PG-K5 and PG-618 bear the imprint of
plastic deformation. PG-K5 shows fine alternating
apatite- and magnetite-rich bands of relatively con-
stant thickness (1 2 mm; Fig. 7a), except in folds
where the bands tend to be thicker. The bands and
streaks in PG-618 are discontinuous, giving the sam-
ple an almost gneiss-like appearance. The magnetite
in both samples lacks apatite inclusions and occurs as
almost equant (mostly < 200 Am) grains evenly dis-
tributed within the bands. PG-K5 also contains hem-
atite, which is typical of the Rektorn and the other
deposits emplaced stratigraphically higher than Kiir-
unavaara (Frietsch, 1967).
The apatite in the banded samples generally has a
smaller grain size than in the primary samples and is
often elongated parallel to the banding, particularly in
the more strongly deformed sample PG-K5. Monazite
inclusions are considerably less abundant than in the
nonbanded samples (except KUJ-4), particularly in
PG-K5, in which more than 3/4 of the apatite grain
interiors are free of monazite inclusions (Fig. 7b). The
remaining apatite contains scattered and somewhat
larger monazite inclusions, without any obvious crys-
tallographic orientation, and a few monazite grains
along apatite grain boundaries. As in the primary and
brecciated ores, under high-contrast BSE imaging, the
apatite grain rims show discontinuous dark areas of
varying thickness as well as darkened patchy interiors
(Fig. 7b).
Fig. 4. Series of BSE photographs from sample PG-K6 showing a large scale view of the brecciated ore (a) and a high-contrast close-up of
typical apatite grains with numerous monazite inclusions in their interior as well as the presence of dark areas along apatite grain boundaries (b).
Note the darkened patchiness in the interiors of many of the apatite grains. Note also the presence of a few larger interstitial grains of monazite
(unlabeled) along apatite grain boundaries in both (b) and (c). A field of elongated monazite inclusions from the interior of one of the apatite
grains is shown in (c). Black regions are areas where grains have been plucked out.
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–7254
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–72 55
Fig. 5. (a) BSE and (b) high-contrast BSE images of a large skeletal monazite grain intergrown with magnetite and apatite in sample PG-K6. The
very faint zoning seen here is typical of large monazite grains intergrown both with magnetite and silicates in the other deformed samples.
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–7256
Fig. 6. BSE high-contrast image of dolomite (Dol) patch in apatite in sample PG-K7 (a) along with a close-up (b). Note the high concentration of
monazite intermingled with less common magnetite. In some cases, monazite is intergrown with magnetite (e.g. monazite –magnetite grain top-
centre in (b)). Note also the presence of numerous small monazite inclusions in the apatite in (a).
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–72 57
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–7258
A typical feature of the banded samples is the
presence of large (10 150 Am) relatively homoge-
nous monazite grains intergrown with magnetite
(commonly as inclusions in the magnetite; cf. Fig.
7c), apatite and silicate minerals such as quartz (PG-
K5; Fig. 7c), and tremolite (PG-618; Fig. 7d). Under
high-contrast BSE imaging, these grains show very
faint, if any, zoning similar to what is seen in Fig. 5b.
4. Mineral chemistry
4.1. Analytical techniques
4.1.1. EMPA
Analyses were made with SX-50 and SX-100
CAMECA electron microprobes at the GeoFor-
schungsZentrum Potsdam operating in the wave-
length-dispersive mode. The operating conditions
during analysis of monazite were an acceleration
voltage of 20 kV, a beam current of 40 nA, and a
beam diameter of 1 2Am. Counting times, data
reduction, analysing crystals, standards, analytical
precision, and detection limits have been described
in detail by Fo¨rster (1998) (see also Fo¨rster, 2000).
Detection limits for the rare earth elements are around
300 500 ppm. Concentrations of Tb, Ho, Er, and Lu
in monazite were routinely measured but not detected.
The concentrations of U and Pb also were below
detection limits (200 and 150 ppm, respectively).
Analyses of apatite were performed at 10 kV and 10
nA using a defocused 20-Am diameter beam. The
REE in apatite were measured separately using an
acceleration voltage of 20 kV and a beam current of
20 nA. Under these conditions, the detection limits for
the REE are between about 800 and 1200 ppm.
Therefore, only La, Ce, and Nd were systematically
detected in the bright areas of the apatite. However, in
the dark areas of the apatite, none of the routinely
sought REE (La, Ce, Nd, Sm, Gd, Dy) were present at
concentrations above their detection limits.
Mean apatite and monazite compositions, includ-
ing ranges for each sample, are given in Tables 1 and
2, respectively. Operating conditions for the analysis
of magnetite (Table 3), dolomite (Table 4), and a
series of silicate minerals (Table 4) (i.e. allanite,
tremolite, talc, chlorite, muscovite, and serpentine)
included using a 20-nA, 15-kV, 1-Am beam. Standards
were taken from the Smithsonian and the Cameca
standard sets. Standards for allanite were the same as
those used for apatite. Chondrite-normalised REE
abundances for the mean monazite values (Table 2)
and the mean allanite values (Table 3) are plotted in
Fig. 8.
4.1.2. LA–ICPMS
Analyses were performed using a UV Microprobe I
laser ablation system coupled to a PQX-S quadrupole
inductively coupled plasma-mass spectrometer (both
VG Elemental, Winsford, UK). The laser microprobe
incorporates a pulsed Nd:YAG laser (Continum, Santa
Clara, USA) operating at 266 nm (UV). The energy of
the beam focused onto the sample surface is attenu-
ated in accordance with the energy appropriate to a
suitable crater size (approximately 40 Am) and
requires samples of approximately 300-Am thickness.
Due to the large crater size from the laser ablation,
coupled with the rather high density of monazite
inclusions in the apatite grains from most of the
samples, only sample KUJ-4 was considered appro-
priate for laser ablation analysis as 99% of the apatite
grains are essentially monazite inclusion-free (see Fig.
3a).
To achieve the maximum benefit from LA–
ICPMS, time-resolved data acquisition was performed
for all analyses (Longerich et al., 1996). A quadrupole
settling time of 5 ms, a dwell time of 25 ms for each
analyte, and a total acquisition time of 150 s were
applied. For the first 60 s, a pre-ablation background
was acquired. After inspection of the spectra, signal
intervals for integration were selected in order to
avoid regions, which are obviously contaminated or
Fig. 7. Series of BSE photographs from the banded ore samples PG-K5 and PG-618. (a) A large-scale view of the banding which is defined by
different proportions of magnetite and apatite (PG-K5). Interstitial dark grey areas are quartz and black areas represent plucking out of grains.
(b) A high-contrast close-up of one of the apatite bands poor in magnetite (PG-K5). White grains not labelled are magnetite. Dark areas not
labelled are muscovite. Note the almost total lack of monazite inclusions in the apatite grains, the darker depletion areas along the apatite grain
boundaries as well as the darkened patchiness in many of the apatite grains. Many apatite grains show a preferred orientation. Close-ups of
magnetite-rich bands show monazite intergrown with magnetite, apatite and quartz (c). Monazite is also commonly found intergrown with
tremolite in sample PG-618 (d).
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–72 59
Table 1
Mean composition of bright and dark areas in the apatite
Sample Area P
2
O
5
SiO
2
SO
3
Y
2
O
3
La
2
O
3
Ce
2
O
3
Nd
2
O
3
CaO MnO FeO
a
SrO Na
2
O F Cl H
2
O
b
Sum Oj(F + Cl) Total (La + Ce + Nd)
c
Dendritic ore
PG-36:2 bright (n= 17) 41.6 0.15 0.25 (0.05) 0.13 0.32 0.16 54.7 0.02 0.03 0.04 0.19 3.51 0.05 0.11 101.3 1.49 99.7 0.61
0.31 0.02 0.08 0.03 0.05 0.07 0.05 0.29 0.01 0.03 0.01 0.05 0.14 0.01 0.07
dark (n= 7) 42.1 0.03 0.02 nd (0.03) (0.03) (0.02) 55.4 (0.01) (0.01) nd (0.02) 3.57 0.02 0.10 101.2 1.50 99.7 (0.08)
0.13 0.02 0.03 0.02 0.04 0.02 0.42 0.01 0.01 0.03 0.11 0.01 0.05
KUJ-4 bright (n= 18) 41.6 0.17 0.24 (0.06) 0.18 0.46 0.19 54.2 0.02 0.05 0.04 0.22 3.55 0.06 0.08 101.1 1.52 99.6 0.83
0.19 0.04 0.14 0.03 0.05 0.08 0.07 0.29 0.02 0.07 0.02 0.05 0.11 0.01 0.05
dark (n= 4) 42.0 0.03 (0.01) (0.02) (0.02) (0.03) 0.02 55.3 0.02 0.06 0.05 0.04 3.76 0.04 0.02 101.4 1.59 99.8 (0.07)
0.31 0.03 0.01 0.02 0.03 0.04 0.02 0.23 0.01 0.04 0.02 0.01 0.08 0.01 0.03
Brecciated ore
PG-K6 bright (n= 12) 41.7 0.12 0.14 (0.04) 0.11 0.26 0.19 55.1 0.02 0.05 0.06 0.14 3.71 0.06 0.03 101.7 1.58 100.1 0.56
0.33 0.04 0.05 0.02 0.04 0.05 0.07 0.25 0.02 0.08 0.02 0.04 0.12 0.01 0.04
dark (n= 11) 42.2 0.04 0.03 (0.01) (0.02) (0.02) (0.03) 55.6 0.02 0.02 0.04 0.03 3.76 0.04 0.02 101.9 1.60 100.3 (0.07)
0.28 0.02 0.02 0.01 0.02 0.02 0.02 0.33 0.01 0.02 0.02 0.02 0.10 0.01 0.03
PG-K7 bright (n= 18) 41.8 0.14 0.20 (0.04) 0.13 0.32 0.17 54.8 0.02 0.02 0.04 0.17 3.83 0.04 0.01 101.6 1.63 100.0 0.62
0.34 0.02 0.04 0.03 0.03 0.07 0.05 0.25 0.02 0.02 0.02 0.04 0.10 0.01 0.01
dark (n= 7) 42.2 0.04 0.02 (0.02) (0.02) (0.03) (0.02) 55.3 0.02 0.04 0.05 0.03 3.88 0.02 101.6 1.65 100.0 (0.07)
0.19 0.02 0.02 0.03 0.03 0.04 0.03 0.28 0.01 0.02 0.03 0.02 0.07 0.01
Banded ore
PG-618 bright (n= 6) 42.0 0.16 0.20 (0.05) 0.17 0.33 0.20 54.6 0.02 0.03 nd 0.17 3.23 0.08 0.24 101.4 1.39 100.2 0.70
0.25 0.05 0.02 0.03 0.07 0.12 0.06 0.33 0.02 0.03 0.03 0.07 0.01 0.04
dark (n= 5) 42.1 0.12 0.06 (0.03) (0.03) (0.06) (0.05) 55.0 0.02 0.04 nd 0.07 3.26 0.07 0.23 101.0 1.39 99.6 (0.14)
0.19 0.02 0.04 0.03 0.03 0.03 0.02 0.33 0.01 0.03 0.03 0.11 0.00 0.05
PG-K5 bright (n= 16) 41.6 0.14 0.25 (0.06) 0.14 0.29 0.16 54.8 0.02 0.05 0.05 0.19 4.05 0.03 102.1 1.76 100.3 0.59
0.40 0.03 0.07 0.02 0.03 0.06 0.04 0.20 0.02 0.02 0.03 0.03 0.10 0.01
dark (n= 10) 41.9 0.05 0.08 (0.04) (0.02) (0.06) (0.04) 55.6 (0.01) 0.03 0.08 0.07 4.04 0.02 101.9 1.80 100.5 (0.12)
0.28 0.03 0.05 0.02 0.03 0.04 0.04 0.19 0.01 0.03 0.02 0.04 0.10 0.01
1-rstandard deviation in italics.
Concentrations in parentheses are not statistically significant in terms of detection limits.
a
Total Fe as FeO.
b
Calculated assuming the (F,Cl,OH) site is filled.
c
Sum of oxide weight percents; nd = not detected; blank = not measured.
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–7260
Tab le 2
Mean composition of monazite inclusions in the apatite and large monazite grains outside the apatite
Sample Type P
2
O
5
SiO
2
ThO
2
Y
2
O
3
La
2
O
3
Ce
2
O
3
Pr
2
O
3
Nd
2
O
3
Sm
2
O
3
Gd
2
O
3
Dy
2
O
3
CaO FeO
a
Total
Dendritic ore
PG-36:2 inclusion 29.7 0.37 0.09 0.20 15.8 35.7 3.45 12.0 0.99 0.42 (0.03) 1.36 0.05 100.2
(n=4) 0.14 0.11 0.12 0.11 1.41 0.76 0.23 1.33 0.25 0.26 0.05 0.26 0.06
range 29.5 – 29.7 0.27 – 0.45 0.00 – 0.27 0.06 – 0.33 14.7 – 17.8 35.1 – 36.7 3.12 – 3.62 10.1 – 13.2 0.67 – 1.22 0.06 – 0.63 0.00 – 0.11 0.98 – 1.57 0.01 – 0.14
Brecciated ore
PG-K6 inclusion 29.8 0.31 0.25 0.46 14.9 34.9 3.53 12.9 1.32 0.66 (0.02) 0.86 0.02 100.0
(n=6) 0.18 0.03 0.17 0.09 0.28 0.30 0.11 0.24 0.08 0.05 0.03 0.15 0.03
range 29.5 – 30.0 0.26 – 0.36 0.03 – 0.42 0.30 – 0.56 14.6 – 15.3 34.5 – 35.4 3.62 – 3.34 12.6 – 13.2 1.25 – 1.43 0.61 – 0.75 0.00 – 0.06 0.70 – 0.99 0.00 – 0.07
outside 29.6 0.36 0.41 0.34 15.4 35.5 3.50 12.1 1.18 0.60 0.04 0.22 0.19 99.5
(n= 10) 0.28 0.07 0.34 0.13 1.61 1.21 0.26 1.50 0.28 0.20 0.05 0.17 0.21
range 29.3 – 30.1 0.21 – 0.44 0.00 – 0.87 0.11 – 0.56 13.6 – 19.6 34.2 – 38.3 2.80 – 3.81 9.09 – 13.8 0.57 – 1.51 0.26 – 0.84 0.00 – 0.17 0.07 – 0.74 0.00 – 0.70
PG-K7 inclusion 29.7 0.31 0.12 0.26 15.0 36.3 3.58 12.2 1.08 0.49 0.05 0.89 0.01 100.1
(n=6) 0.14 0.05 0.06 0.07 0.62 0.44 0.19 0.61 0.17 0.11 0.05 0.18 0.01
range 29.5 – 29.9 0.24 – 0.39 0.07 – 0.21 0.20 – 0.40 14.1 – 15.8 35.6 – 36.8 3.25 – 3.82 11.0 – 12.7 0.83 – 1.26 0.40 – 0.69 0.00 – 0.12 0.61 – 1.15 0.00 – 0.02
outside 29.7 0.38 0.29 0.26 14.2 36.6 3.66 12.5 1.06 0.47 0.03 0.25 0.12 99.6
(n=9) 0.26 0.08 0.19 0.09 0.82 0.59 0.12 0.58 0.13 0.08 0.04 0.08 0.10
range 29.3 – 30.0 0.29 – 0.52 0.00 – 0.55 0.15 – 0.46 13.4 – 16.1 35.8 – 37.7 3.40 – 3.84 11.6 – 13.3 0.87 – 1.22 0.34 – 0.59 0.00 – 0.10 0.11 – 0.34 0.00 – 0.33
dolomite vug 29.6 0.36 0.19 0.16 14.3 37.9 3.66 11.8 0.85 0.39 (0.02) 0.25 0.22 99.7
(n= 15) 0.21 0.07 0.21 0.11 0.94 1.77 0.20 1.41 0.31 0.16 0.03 0.10 0.15
range 29.2 – 29.9 0.27 – 0.47 0.00 – 0.72 0.00 – 0.28 12.9 – 16.1 35.9 – 41.6 3.27 – 3.94 9.53 – 13.6 0.30 – 1.27 0.09 – 0.58 0.00 – 0.12 0.07 – 0.42 0.06 – 0.47
Banded ore
PG-618 inclusion 29.8 0.33 0.10 0.35 15.3 34.9 3.52 12.5 1.27 0.59 0.03 0.93 0.11 99.8
(n=5) 0.14 0.08 0.17 0.12 0.56 0.65 0.08 0.72 0.18 0.14 0.02 0.31 0.15
range 29.6 – 30.0 0.28 – 0.42 0.00 – 0.40 0.28 – 0.55 14.7 – 16.0 34.1 – 35.8 3.44 – 3.64 11.4 – 13.0 1.02 – 1.46 0.41 – 0.75 0.00 – 0.07 0.65 – 1.39 0.00 – 0.32
outside 29.8 0.37 0.20 0.37 14.9 34.8 3.59 13.1 1.30 0.65 0.03 0.19 0.04 99.4
(n= 14) 0.13 0.03 0.17 0.11 0.70 0.51 0.11 0.57 0.17 0.13 0.04 0.09 0.04
range 29.6 – 30.0 0.32 – 0.41 0.01 – 0.55 0.17 – 0.56 14.1 – 16.6 33.7 – 35.3 3.45 – 3.72 11.9 – 13.9 0.93 – 1.59 0.37 – 0.84 0.00 – 0.12 0.09 – 0.42 0.00 – 0.15
intergrown 29.9 0.35 0.16 0.32 15.0 35.4 3.55 12.9 1.27 0.58 0.03 0.14 0.03 99.6
with tremolite 0.09 0.04 0.15 0.06 0.37 0.37 0.10 0.44 0.10 0.11 0.03 0.03 0.03
(n= 12) range 29.9 – 30.0 0.31 – 0.43 0.00 – 0.49 0.16 – 0.42 14.5 – 15.8 34.5 – 35.7 3.41 – 3.73 12.0 – 13.5 1.12 – 1.45 0.31 – 0.74 0.00 – 0.11 0.10 – 0.20 0.00 – 0.06
PG-K5 inclusion (n=10) 29.90.250.170.5215.234.43.4912.81.530.900.080.480.0499.8
0.20 0.02 0.12 0.13 0.68 0.30 0.06 0.43 0.13 0.10 0.04 0.31 0.04
range 29.7 – 30.2 0.22 – 0.29 0.05 – 0.40 0.35 – 0.70 14.3 – 15.8 34.0 – 34.9 3.35 – 3.59 11.9 – 13.4 1.32 – 1.76 0.70 – 1.02 0.02 – 0.16 0.21 – 1.15 0.00 – 0.13
outside (n=9) 29.90.230.200.4016.034.13.3212.31.490.910.060.310.4099.7
0.26 0.04 0.14 0.17 1.06 0.67 0.14 0.71 0.25 0.23 0.04 0.21 0.28
range 29.4 – 30.2 0.18 – 0.33 0.10 – 0.43 0.12 – 0.74 14.4 – 17.5 33.1 – 34.9 3.10 – 3.57 11.7 – 13.0 1.21 – 1.99 0.63 – 1.45 0.00 – 0.13 0.12 – 0.78 0.08 – 0.83
1-rstandard deviation in italics. Concentrations in parentheses are not statistically significant.
a
Total Fe as FeO.
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–72 61
represent signals from small cracks or the glass
mount.
The glass standard NIST-610 (National Institute of
Standards and Technology, Gaithersburg, USA) was
used as reference sample for calculating element
concentrations. A series of workers (e.g. Pearce et
al., 1997; Rocholl et al., 1997, 2000) have shown that
this standard is homogeneous for Y and REE on the
microscale. Reference values for Ca were taken from
Pearce et al. (1997) and for (Y + REE) from Dulski
(2001).
Each measurement sequence consisted of 2
acquisitions on the glass standard followed by 10
acquisitions on 10 individual apatite grains and then
again 2 acquisitions on the glass standard. Raw
count rates from both the apatite grains and the glass
standards were gas blank-corrected and normalised
to the Ca net count rates and the Ca concentration.
For apatite, the Ca concentration used was that
obtained by microprobe analysis for apatite in sam-
ple KUJ-4, i.e. CaO = 54.2 wt.% (Table 1). Element
concentrations for each of the apatite grains are
calculated by comparison of the Ca-normalised count
rates from the apatite and the mean of the four glass
standard acquisitions bracketing the 10 apatite anal-
yses.
Mean (Y + REE) concentrations for bright and
dark areas in KUJ-4 are given in Table 5, and mean
chondrite-normalised distribution patterns for the
bright and dark areas from apatite in sample KUJ-
4, taken from LA–ICPMS data, are plotted in Fig.
9. An accompanying plot of REE depletion factors
for individual apatite grains in KUJ-4 is shown in
Fig. 10.
4.2. Phosphate chemistry
The Kiruna fluorapatites have Cl and H
2
O abun-
dances generally lower than 0.08 and 0.24 wt.%
(Table 1), respectively. This confirms previous studies
on bulk mineral separates (Para
´k, 1973; Frietsch,
1974). In all samples, the brighter internal areas of
the apatite have significantly higher contents of
LREE, Si, Na, and S than the dark areas. The bright
areas of the apatite in the primary dendritic ore
samples tend to be somewhat higher in the LREE
than the deformed samples. Values for the minor
elements Na
2
O and SiO
2
range about 0.14–0.22 and
0.12–0.17 wt.%, respectively, in the bright areas, and
about 0.02–0.07 and 0.03–0.12 wt.%, respectively, in
the dark areas (Table 1). Also, SO
3
is reduced from
0.14–0.25 wt.% in the bright areas to V0.08 wt.% in
the dark areas. There is, thus, a clear covariation
between Na, Si, and the LREEs which is balanced
by an increase in Ca and P in the dark areas by 0.5 –
1.0 and 0.5 wt.%, respectively. Y and heavy rare earth
element (HREE) contents are generally low, below
microprobe detection limits.
The LA–ICPMS data of the apatite from sample
KUJ-4 are in good agreement and support the validity
Table 3
Representative analyses of magnetite and hematite from the Kiruna ore deposit
PG-36:2 KUJ-4 PG-K6 PG-K7 PG-618 PG-K5
Magnetite Magnetite MagnetiteMagnetite Magnetite Magnetite
End-member Ti-enriched End-member Ti-enriched Hematite
SiO
2
0.03 0.05 0.03 0.03 0.02 0.04 0.05 0.08 0.04
TiO
2
0.02 0.02 0.02 0.03 3.31 0.01 0.02 1.36 0.95
Cr
2
O
3
0.02 0.02 0.02 0.01 0.02 0.02 0.01 0.02 0.03
Al
2
O
3
0.01 0.01 0.01 0.01 0.03 0.01 0.02 0.04 0.03
Fe
2
O
3a
69.9 69.3 69.8 69.7 61.3 70.0 69.8 64.9 93.3
FeO
a
31.4 31.1 31.4 31.2 33.4 31.4 31.4 31.7 nd
MgO 0.03 0.02 0.02 0.07 0.07 0.03 0.01 nd nd
MnO 0.10 0.08 0.07 0.13 0.03 0.12 0.01 0.01 nd
NiO 0.04 0.05 0.04 0.04 0.03 0.04 0.08 0.01 nd
Total 101.5 100.6 101.4 101.2 98.2 101.7 101.4 98.1 99.4
nd = Not detected.
a
Calculated by charge balance.
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–7262
Table 4
Representative analyses of silicate and carbonate minerals from the Kiruna ore deposit
PG-36:2 KUJ-4 PG-K6 PG-K7 PG-618 PG-K5
Talc Chlorite Talc Allanite Chlorite Talc Chlorite Serpentine Dolomite Tremolite Chlorite Muscovite
Grain Range Interstitial Range
SiO
2
62.1 33.0 62.3 29.8 27.4 – 30.1 29.5 28.8–30.3 29.0 63.6 32.8 45.8 0.02 56.6 33.9 50.7
TiO
2
0.02 0.01 0.25 0.22 – 0.43 0.50 0.22 – 0.50 0.05 nd 0.01 nd 0.01 0.01 0.01 0.46
ThO
2
nd nd
UO
2
nd nd
Al
2
O
3
0.57 16.4 0.20 14.1 12.5 – 14.9 11.2 10.6 –14.5 19.0 0.06 16.5 0.51 nd 1.04 15.4 28.4
Cr
2
O
3
0.03 0.01 nd 0.01 0.03 nd 0.01 0.01 nd 0.01 0.02
Y
2
O
3
0.07 0.04 – 0.16 0.02 0.00 – 0.16
La
2
O
3
5.30 5.10 – 7.14 5.66 5.16 – 7.98
Ce
2
O
3
12.7 12.0 – 13.4 13.0 11.6 – 14.2
Pr
2
O
3
1.41 1.16 – 1.41 1.42 1.14 – 1.48
Nd
2
O
3
4.88 3.75 – 5.08 4.80 3.28 – 5.66
Sm
2
O
3
0.53 0.34 – 0.57 0.61 0.15 – 0.67
Gd
2
O
3
0.22 0.08 – 0.42 0.22 0.11 – 0.32
Dy
2
O
3
nd 0.00 – 0.11 0.01 0.00 – 0.07
FeO 2.06 5.63 4.64 15.5 14.7 – 16.7 17.8 14.6 – 18.4 12.4 3.35 5.74 1.64 0.79 4.21 4.17 5.20
MgO 30.5 32.4 28.1 0.50 0.41 – 0.72 0.25 0.20–0.70 25.5 29.7 32.0 41.0 22.0 22.8 33.9 2.36
MnO nd 0.09 0.02 0.16 0.04 – 0.29 0.53 0.06 – 0.63 0.27 nd 0.13 0.13 0.50 0.05 0.13 nd
CaO 0.04 0.05 0.12 10.5 9.82 – 10.7 9.72 9.49 – 10.9 0.10 0.22 0.21 0.11 34.3 11.6 0.09 0.12
BaO nd nd nd nd 0.03 0.04 nd nd 0.02 nd 0.06
Na
2
O 0.03 0.01 0.08 nd 0.05 0.01 nd nd 1.35 nd 0.11
K
2
O 0.01 0.01 nd nd 0.02 0.01 nd nd 0.39 nd 9.96
F 0.18 nd 0.57 nd nd 0.04 0.34 nd 0.81 nd 0.56 nd nd
Cl 0.01 nd nd 0.03 0.01 0.01 0.03 0.02 0.02 0.01 0.01
Sum 95.5 87.6 96.0 96.0 95.4 86.5 97.4 87.4 90.0 57.6 98.6 87.5 97.4
OjF 0.08 0.24 0.02 0.14 0.34 0.24
OjCl 0.01 0.01 0.01
Total 95.4 87.6 95.8 96.0 95.4 86.4 97.3 87.4 89.7 57.6 98.4 87.5 97.4
nd = Not detected; blank = not measured.
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–72 63
of the microprobe data for the LREE as well as
confirm the low abundances of the HREEs and Y
(compare Tables 1 and 5). The chondrite-normalised
patterns (Fig. 9) are thus LREE-enriched, and signifi-
cantly more strongly so for the bright areas (La
N
/
Yb
N
= 14.5) than for the dark areas (La
N
/Yb
N
= 5.5).
The present data compare well with the data on apatite
separates from the Kiruna area, but show slightly
higher abundances for the LREE and lack the small
negative La anomaly reported by Frietsch and Perdahl
(1995).
The composition of monazite in each of the sam-
ples, whatever their textural position, is virtually
indistinguishable (Table 2). The slightly higher Ca
contents of the inclusions in the apatite may be
attributed to their small size with some influence from
the surrounding apatite in the analysis since there is no
systematic covariation with Th or Si. The Th contents
are always very low (ThO
2
V0.41 wt.%) with a weak
tendency to be lowest in the inclusions within the
apatite. SiO
2
and Y
2
O
3
contents are also low (0.23–
0.38 and 0.16–0.52 wt.%, respectively), both without
any systematic variation with respect to the textural
position of the monazite. This shows that the huttonite
(ThSiO
4
) and brabantite (CaTh(PO
4
)
2
) substitution
mechanisms (e.g. Fo¨rster, 1998; Fo¨rster and Harlov,
1999) are of very restricted importance in these
monazites which are nearly pure (LREE)PO
4
mon-
azites. HREE contents are also low, after Gd mostly
Fig. 8. Chondrite-normalised REE patterns (filled circles) of monazite (using the mean values for both inclusions and the large grains listed in
Table 2) and allanite (see Table 4; filled squares). Chondrite values are taken from Anders and Grevesse (1989).
Table 5
Mean REE contents (ppm) in bright and dark areas in apatite from
sample KUJ-4 measured using LA – ICPMS
Element Dark areas (n= 5) Bright areas (n= 29)
Mean 2-rMean 2-r
Y 844 69 909 147
La 454 47 1307 511
Ce 1137 134 2856 1002
Pr 163 13 377 128
Nd 670 57 1407 439
Sm 138 10 223 56
Eu 19 1 29 6
Gd 164 10 207 43
Tb 23 1 27 5
Dy 135 8 149 27
Ho 28 2 30 5
Er 77 8 83 15
Tm 10 1 11 2
Yb 56 6 61 11
Lu 8 1 9 2
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–7264
Fig. 9. Chondrite-normalised REE patterns of apatite from sample KUJ-4 measured by LA – ICPMS (see Table 5). Filled squares connected by a
thick line and outlined by dashed lines above and below represent mean and 2-rcompositional boundaries for apatite with ‘‘high’’
concentrations of LREE (n= 29), i.e. bright regions. Filled circles connected by a thick line and outlined by dotted line above and below
represent mean and 2-rcompositional boundaries for apatite with the lowest LREE concentrations (n= 5), i.e. dark regions. See text for a more
complete description of the bright and dark regions. Chondrite values are taken from Anders and Grevesse (1989).
Fig. 10. Plot of depletion factors for 34 apatite grains from sample KUJ-4. Depletion factors are calculated as the ratio of the REE concentration
of each individual apatite grain normalised to the mean of the five apatite grains with highest LREE concentrations. The 29 apatite grains with
variably ‘‘high’’ concentrations of LREE are indicated by thin lines, whereas the five apatite grains with ‘‘low’’ LREE concentrations are
indicated by thick lines. Intermediate values of bright areas indicate either partial LREE depletion of the apatite associated with the formation of
the interior monazite or represent the natural compositional variability of the magmatic apatite itself.
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–72 65
below the microprobe detection limits (Table 2). The
chondrite-normalised REE patterns (Fig. 8) of the
monazites are, therefore, strongly LREE-enriched
which is typical for monazite from most geological
environments, although with varying abundances of
the individual LREE (e.g. Chang et al., 1996; Fo¨ rster,
1998, and references therein).
4.3. Oxide chemistry
Magnetite in all samples is dominantly character-
ised by a near end-member composition (Table 3).
However, samples PG-K7 and PG-K5 also contain
magnetite with significant TiO
2
contents (3.3 and 1.4
wt.%, respectively). In PG-K7, the Ti-enriched type
occurs as veins or blebs within or along the margins of
the normal magnetite, while in PG-K5, it forms either
separate individual grains or grains intergrown with
magnetite. The hematite of sample PG-K5 has a TiO
2
content generally lower than 1 wt.%.
4.4. Silicate chemistry
The allanite in sample KUJ-4 is Fe-rich (Table 4)
and has low abundances of MgO and MnO ( V0.72
and V0.63 wt.%, respectively). TiO
2
is also low
(V0.50 wt.%), while U and Th contents are very
low, below the detection limits for the microprobe.
The sum of (Y + REE) oxides is generally 25 26
wt.%, with a strong preference for the LREE. The
(Y + REE) pattern (Fig. 8) is similar to that for
allanite from other occurrences (e.g. Exley, 1980;
Deer et al., 1986; Petrı
´k et al., 1995). The (Y + REE)
pattern mimics that of the monazites at somewhat
lower values. The patchy zoning pattern of the
interstitial allanite displayed in Fig. 3b is related to
small variations in the LREE content.
The composition of the other silicates shows some
variation between the samples and is most iron-rich
in sample KUJ-4 (Table 4). In talc (Mg/Mg + Fe)
varies between 0.92 (KUJ-4) and 0.96 (PG-36:2),
and in chlorite, between 0.79 (KUJ-4) and 0.94 (PG-
618). In the tremolite from PG-618 and the Serpen-
tine from PG-K7, (Mg/Mg + Fe) equals 0.91 and
0.98, respectively. Muscovite from sample PG-K5
has appreciable amounts of Fe and Mg, and minor
Ti, while the dolomite from PG-K7 has only sub-
ordinate Fe. The sheet silicates are devoid of Cl but
the talc, tremolite, and serpentine contain minor F
(0.2–0.6 wt.%).
5. Discussion
5.1. Textures and chemistry
The primary and deformed samples of magnetite –
apatite ore investigated in this study have two textural
features in common: (i) clusters of monazite inclu-
sions in the interior of the apatite (though less com-
mon in samples KUJ-4 and PG-K5) as well as, to a
lesser extent, large monazite grains along the apatite
grain rims and (ii) the presence of dark areas along the
apatite grain rims, cracks and as well as more diffuse
patchy dark regions within the grains. EMPA and
LA–ICPMS data have demonstrated that the dark
regions in the apatite (cf. Figs. 2b – d, 3a, 4b, and
7b) represent areas strongly depleted in the LREE.
Mobilisation of the lanthanides from the apatite is
interpreted as the result of interaction with a fluid
percolating through the rocks. The textural develop-
ment of the dark areas in the apatite appears to be
related to the most probable routes for these fluids,
namely along grain boundaries, cracks, and crystal
defects. Similar dark, LREE-depleted regions have
been observed by Darling and Florence (1995) along
the margins of apatite grains in the Port Leyden
nelsonite, though with no accompanying monazite
inclusions.
The depletion of LREE in the apatite is strongly
correlated with a corresponding depletion in both Na
and Si (Table 1). In the apatite structure, trivalent
cations such as Y
3+
and the REE
3+
can proxy for
Ca
2+
with electrostatic neutrality maintained by the
substitution of Si
4+
for P
5+
or Na
+
for additional
Ca
2+
. This can be expressed in the form of two
coupled substitution reactions (Roeder et al., 1987;
Rønsbo, 1989; Fleet et al., 2000):
NaþþðYþREEÞ3þ¼2Ca2þð1Þ
and
Si4þþðYþREEÞ3þ¼P5þþCa2þð2Þ
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–7266
Nucleation of monazite in apatite in an open
system involves the following general reaction modi-
fied from Pan et al. (1993):
apatiteð1ÞþðCa2þ
;P5þÞin a fluid
¼apatiteð2Þþmonazite
þðSi4þand NaþÞin a fluid ð3Þ
where apatite(1) is enriched in the LREEs, Na and Si
relative to apatite(2) (see also Harlov and Fo¨ rster,
2002; Harlov et al., 2002). Reaction (3) assumes that
in an open system, Ca and P are added to those areas
of the apatite depleted in LREE. This is a reasonable
assumption since a portion of the apatite should
dissolve in the fluid as well, e.g. creating space for
the nucleation and growth of the monazite inclusions.
Excess amounts of dissolved Ca and P in the fluid
may partly reprecipitate along grain boundaries or in
the alteration zones around the ores, where impreg-
nations of apatite are occasionally observed (e.g.
Geijer and O
¨dman, 1974). The orientation of the
monazite inclusions appears to be at least partly
controlled by the crystallography of the apatite. The
apparent elongation of the monazite inclusions along a
preferred crystallographic orientation in the apatite is
probably related to the preferential ease with which
the fluid is able to dissolve the apatite in this direction
during dissolution –precipitation growth of the mon-
azite.
Previous studies of monazite inclusions in fluo-
rapatite, both experimentally (Harlov and Fo¨rster,
unpublished data; see also Harlov et al., 2002) and
in natural systems (A
˚mli, 1975; Pan et al., 1993;
Harlov and Fo¨rster, 2002), have stressed the impor-
tance of metasomatising fluids as agents promoting
coupled substitution reactions (1) and (2), leading to
the removal of Na and Si from the apatite structure.
This, consequently, results in charge imbalance with
respect to the less mobile LREE, which induces the
subsequent nucleation and growth of the monazite in
the apatite according to reaction (3). This requires
that the fluid must also be capable of transporting P
and Ca such that mass balance is maintained. In
addition, there appears to be a temperature threshold
( < 850 jC), below which monazite will not nucleate
within the apatite (Harlov and Fo¨ rster, unpublished
data). Instead, the LREEs are simply leached from
the apatite structure together with Na and Si. In such
an open system, these elements could either be
transported away, or else, depending on the nature
of the fluids, a portion of the now mobile lanthanide
budget could be diverted into growing monazite
grains interstitially on the surface of the apatite
grains.
The similarity in (Y + REE) pattern of the apatite
and the enclosing host rock porphyries, and dissim-
ilarities with sedimentary apatites of, e.g. phosphorites
(which have much lower (Y + REE) contents), support
the magmatic origin of the Kiruna apatite (cf. Frietsch
and Perdahl, 1995). Iron oxide magmas start to
crystallise at temperatures above 800 jC, as indicated
from experimental and fluid/melt inclusion evidence
(Broman et al., 1999; Naslund et al., 2000). However,
volatiles, such as CO
2
,H
2
O, Cl, F, SO
4
, and presum-
ably P, can depress the crystallisation temperature of
iron oxide melts significantly (Frietsch and Perdahl,
1995; Naslund et al., 2000). For example, O’Farrelly
(1990) has estimated an emplacement temperature of
700 jC for Kiirunavaara, based on stable isotope data.
The implication would be that the originally homoge-
neous, inclusion-free, (LREE + Na + Si)-rich apatite
should have finished crystallising at T>700 jC. The
composition of the original apatite is difficult to
assess, as almost all the grains seem to have been
affected by depletion to some degree. However, they
must have contained at least the amount measured in
the bright areas with the highest (Y + REE) contents in
sample KUJ-4 (i.e. 7000–8000 ppm total (Y + REE);
Table 5).
The very low abundances of Th and U, as well
as Si and Ca, are typical of monazite associated
with apatite, and provide strong evidence that their
chemical components are derived from the apatite
(Pan et al., 1993; Harlov and Fo¨rster, 2002; Harlov
et al., 2002). This contrasts with most independently
formed monazite in igneous and metamorphic rocks
which typically show various degrees of enrichment
in these elements (e.g. Watt, 1995; Fo¨rster, 1998;
Bea and Montero, 1999). The numerous small
monazite inclusions within the interior of apparently
homogeneous apatite are interpreted to have formed
at high temperatures shortly after crystallisation
when magmatic fluids, expelled during crystallisa-
tion, were still present along grain boundaries.
Fluid-induced depletion along apatite grain bounda-
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–72 67
ries and cracks, coupled with the associated forma-
tion of additional larger monazite grains, also along
apatite grain boundaries, followed at lower temper-
atures. This is supported by their association with
low-temperature silicate mineral assemblages (e.g.
Fig. 2c).
The large Th- and U-poor monazite grains, which
occur intergrown with magnetite, apatite, and silicate
minerals primarily in the banded ore samples, are
interpreted to have formed by recrystallisation and
coalescence of the previously formed smaller mona-
zite inclusions. Such an origin is supported by the fact
that they have a composition identical to that of the
monazite inclusions (cf. Table 2). It is also supported
by the fact that the abundance of these grains appears
to be related to temperature and deformation. This is
seen in the lower temperature brecciated samples
where the apatite contains numerous inclusions of
monazite similar to that seen in the undeformed,
primary ore and only a very few large monazite
grains. Whereas in the banded samples (plastically
deformed at relatively higher temperatures), monazite
inclusions are dramatically less abundant and large
monazite grains are much more abundant, particularly
in the most strongly deformed sample PG-K5. This
further supports the argument that plastic deformation
and recrystallisation of the apatite facilitated the
growth of larger monazite grains.
The extreme scarcity of monazite inclusions in the
apatite, coupled with the presence of allanite in
primary ore sample KUJ-4, is puzzling. The simplest
explanation is that most of the apatite in sample KUJ-
4 did not come into contact with fluids of sufficiently
high temperature and/or of a composition capable of
activating coupled substitution reactions (1) and (2).
This is supported by the relatively higher LREE
abundances in the apatite grain interiors of this sample
compared with that for the apatite from PG-36:2 and
the deformed ore samples (Table 1). The irregular,
patchy zoning seen in the allanite reflects small-scale
variations in the REE content and is consistent with
precipitation from a fluid. Its occurrence here may be
related to a localised fluid with an exceptionally high
activity in allanite-forming components, particularly
Al. The source of the REE incorporated into the
allanite most likely comes from the depleted regions
along the apatite grain boundaries. This is supported
by the REE pattern for the allanite which is subpar-
allel to that of the monazite and independent of
textural type (Fig. 8).
5.2. Fluids
Studies of magnetite – apatite ores of the Kiruna-
type suggest that they are likely to have been depos-
ited from iron oxide magmas rich in volatiles, partic-
ularly P (Naslund et al., 2000). During the
crystallisation of such magmas, other volatile phases
such as H
2
O are concentrated in the residual melt and
may separate from it in response to a decrease in P – T
conditions or increasing crystallisation (Roedder,
1992). Subsequent separation of H
2
O-rich fluids from
the melt results in the extensive hydrothermal over-
print typically associated with magnetite – apatite ores
(Broman et al., 1999; Naslund et al., 2000; Bergman
et al., 2001, and references therein). At the same time,
Cl will be preferentially partitioned into this H
2
O-rich
fluid phase (cf. Cline and Bodnar, 1991; Webster,
1997). While no rigorous fluid inclusion study has
been made of the Kiruna magnetite – apatite ores to
date, highly saline aqueous fluids are a characteristic
feature associated with the formation of other ore-
bodies in the Kiruna region (Frietsch et al., 1997;
Lindblom et al., 1996; Broman and Martinsson,
2000). Such fluid inclusions are typical of Kiruna-
type magnetite –apatite ores in general. For example,
studies of fluid inclusions in both pyroxene and
apatite from the Kiruna-type Pliocene El Laco mag-
netite–apatite ore deposits in northern Chile indicate
that highly saline fluids were present as late-stage
magmatic fluids shortly before and during crystallisa-
tion (T>800 jC; Broman et al., 1999). In a prelimi-
nary reconnaissance study of fluid inclusions in
sulphide- and hematite-bearing quartz veins cross-
cutting the iron ores at Nukutusvaara and Rektorn
(Fig. 1), apatite in skeleton ore from Rektorn and
apatite in a magnetite – apatite banded ore type from
Kiirunavaara was reported to contain high-salinity
fluid inclusions similar to those observed in the
epigenetic deposits in the surrounding area (Broman
and Lindblom, 1995; see also Broman and Martins-
son, 2000). Primary fluid inclusions in the quartz
veins were found to have a salinity of between 32
and 41 eq. wt.% (CaCl
2
+ NaCl), whereas secondary
fluid inclusions in the apatite contained up to 22 eq.
wt.% (CaCl
2
+ NaCl).
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–7268
We propose that the magnetite– apatite ore deposits
in the Kiruna area experienced two subsequent stages
of fluid–rock interaction. The first episode occurred at
high temperature (700 800 jC), shortly after the
emplacement and crystallisation of the orebodies.
Fluid inclusion studies of Kiruna-type magnetite–
apatite ores (e.g. Broman et al., 1999), as well as
the apatite grain chemistry itself (Tables 1 and 5),
suggest that the fluid present along the apatite grain
boundaries shortly after crystallisation was complete
would have been both Cl-rich as well as contain F,
SO
2
, and P components. Such fluids are very efficient
in the transportation of ore metals as well as other
cations such as Si, Na, K, Mg, Ca, and (Y + REE) in
the form of chloride complexes (Seward and Barnes,
1997). These could have been responsible for the
leaching of Na and Si from the apatites and the
subsequent precipitation of the monazite inclusions
as well as the growth of monazite along apatite grain
boundaries, depending on the temperature. This is
corroborated, in part, by experiments involving fluo-
rapatite and a variety of fluids including H
2
O, CO
2
–
H
2
O mixtures, and KCl brines (Harlov and Fo¨rster,
unpublished data). Use of these fluids have resulted in
the formation of monazite either as inclusions in the
fluorapatite at T>850 jC or as grains growing on the
surface of the fluorapatite at lower temperature. Mon-
azite formation has subsequently been coupled with
an associated depletion in (Y + REE + Na + Si) in the
apatite. The results from these experiments strongly
suggest that the salt component in the fluid could not
have been dominated by NaCl as this would tend to
stabilise Na in the apatite and subsequently the LREE
via coupled substitution reaction (1). The theoretical
calculations of Haas et al. (1995) suggest that Cl
complexes with the LREE are more stable than with
the HREE which could explain the LREE depletion
pattern in the apatite (cf. Figs. 9 and 10). Moreover,
Giere
´(1996) proposes that phosphate ions are impor-
tant complexing agents of the REE, especially at
higher temperatures. This may explain why the REEs
were not removed from the rock but dominantly
reprecipitated in the phosphate-rich ore. Expulsion
of these magmatic fluids from the cooling orebody
is probably responsible for the variable albitisation,
sericitisation and the formation of secondary actino-
lite, titanite, biotite–chlorite, carbonates, sulphides
and sometimes allanite, scapolite, and tourmaline
observed in the surrounding country rock (Frietsch
et al., 1997; Bergman et al., 2001, and references
therein).
Greenschist-facies regional metamorphism, in the
Kiruna area, took place during a separate episode
which followed within a few million years after the
formation of the orebody. This fluid activity intro-
duced components of the altered country rocks (Si,
Mg, Al, K, Na) into the ores, as reflected by their
interstitial greenschist-facies silicate assemblages
(Table 4), including allanite talc, tremolite, chlorite,
muscovite, quartz, carbonates, and serpentine which
are, in many cases, closely associated with the
depleted, dark, apatite grain rims (e.g. Figs. 2c, d
and 3a). This would suggest that leaching of Na, Si,
and LREE from the apatite grain rims and cracks most
likely took place at lower fluid temperatures, i.e.
greenschist grade (300–450 jC). The effects of the
subsequent regional greenschist facies metamorphism,
overprinting the ores and their country rocks, are
difficult to separate from those alterations immedi-
ately associated with and following emplacement of
the ore magma. As a consequence, the deformed
samples most likely represent localised early defor-
mation associated with local faulting which occurred
shortly after the emplacement of the orebody.
6. Conclusions
In general, assuming that all magnetite apatite ores
of the Kiruna-type form under the same volatile-rich
conditions, interstitial fluids should be found along
apatite grain boundaries in these orebodies at temper-
atures of 800 jC or higher shortly after crystallisation.
This would imply that one characteristic of Kiruna-
type magnetite–apatite ores is that the apatite grains
should contain monazite and/or xenotime inclusions.
However, two criteria must be met. Namely, the
concentration of (Y + REE) in the apatite must be
sufficiently high enough (>0.2 –0.3wt.%) and the
fluid composition must allow for coupled substitution
reactions (1) and/or (2) to be activated. For example,
monazite and/or xenotime inclusions have never been
documented in apatites from two extensively studied,
much younger ( < 35 Ma) Kiruna-type magnetite –
apatite orebodies, i.e. El Laco, Chile (Broman et al.,
1999) and Durango, Mexico (Young et al., 1969).
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–72 69
This is despite the fact that the apatites in either
orebody are even more enriched in (Y + REE) than
in Kiruna. One reason for this lack of monazite
inclusions could be that, at least in the case of El
Laco, the interstitial fluid present during crystallisa-
tion was enriched in NaCl (cf. Broman et al., 1999)
which would have retarded monazite nucleation. The
implication then is that the presence or absence of
monazite inclusions in apatite from magnetite – apatite
ore deposits of the Kiruna-type can serve to add
constraints on the nature and chemistry of the fluids
associated with these particular orebodies both during
their emplacement and directly after their crystallisa-
tion. Such insights, combined with fluid inclusion
data, could be a powerful tool in understanding how
these ores and their associated fluids form in general,
as well as understanding their subsequent history from
the standpoint of their interaction with later external
fluids under different pressure –temperature and tec-
tonic conditions.
Acknowledgements
We thank Dieter Rhede and Oona Appelt for
assistance with the microprobe as well as for their
efforts in developing the standards used in analysis of
the mineral species studied in this paper. Helga
Kemnitz and Ursula Glenz are thanked for support
with the scanning electron microscope. U.B. Ander-
sson acknowledges a grant from STINT (Stiftelsen fo¨r
internationalisering au ho¨gre utbildning och for-
skning). Gerhard Franz is thanked for a very careful
and thorough review of the original manuscript. An
anonymous reviewer is thanked for encouraging the
use of LA–ICPMS on the apatites, which greatly
increased the analytical dimensions of the paper. [CA]
References
A
˚mli, R., 1975. Mineralogy and rare earth geochemistry of apatite
and xenotime from the Gloserheia granite pegmatite, Froland,
southern Norway. Am. Mineral. 60, 607 – 620.
Anders, E., Grevesse, N., 1989. Abundances of the elements: mete-
oric and solar. Geochim. Cosmochim. Acta 53, 197– 214.
Bea, F., 1996. Residence of REE, Y, Th and U in granites and
crustal protoliths: implications for the chemistry of crustal melts.
J. Petrol. 37, 521 – 552.
Bea, F., Montero, P., 1999. Behaviour of accessory phases and
redistribution of Zr, REE, Y, Th, and U during metamorphism
and partial melting of metapelites in the lower crust: an example
from the Kinzigite formation of Ivrea – Verbano, NW Italy. Geo-
chim. Cosmochim. Acta 63, 1133 – 1153.
Bergman, S., Ku
¨bler, L., Martinsson, O., 2001. Description of re-
gional geological and geophysical maps of northern Norrbotten
county (east of the Caledonian orogen). Sver. Geol. Unders., Ba
56, 110 pp.
Bingen, B., Demaiffe, D., Hertogen, J., 1996. Redistribution of rare
earth elements, thorium, and uranium over accessory minerals in
the course of amphibolite to granulite facies metamorphism: the
role of apatite and monazite in orthogneisses from southwestern
Norway. Geochim. Cosmochim. Acta 60, 1341 –1354.
Broman, C., Lindblom, S., 1995. Fluid inclusion studies of volcan-
ite hosted gold, sulphide and iron oxide deposits in northern
Norrbotten. Unpublished research report, NUTEK project 94-
2739, 16 pp.
Broman, C., Martinsson, O., 2000. Fluid inclusions in epigenetic
Fe – Cu – Au ores in northern Norrbotten. In: P. Weihed, O. Mar-
tinsson (Eds.), Abstract volume and Field trip guidebook, 2nd
annual GEODE-Fennoscandian shield field workshop on Palae-
oproterozoic and Archean greenstone belts and VMS districts in
the Fennoscandian Shield, Ga¨llivare-Kiruna, Sweden. Lulea
˚
University of Technology, Research Report 2000:6, 7.
Broman, C., Nystro¨m, J.O., Henrı
´quez, F., Elfman, M., 1999. Fluid
inclusions in magnetite – apatite ore from a cooling magmatic
system at El Laco, Chile. Geol. Fo¨ren. Stockh. Fo¨rh. 121,
253 – 267.
Chang, L.L.Y., Howie, R.A., Zussman, J., 1996. Rock-Forming
Minerals, 2nd ed. Non-silicates. Sulphates, Carbonates, Phos-
phates and Halides, vol. 5B. Longman, Harlow, UK.
Cliff, R.A., Rickard, D., Blake, K., 1990. Isotope systematics of the
Kiruna magnetite ores, Sweden: part 1. Age of the ore. Econ.
Geol. 85, 1770 – 1776.
Cline, J.S., Bodnar, R.J., 1991. Can economic porphyry copper
mineralization be generated by a typical calc-alkaline melt? J.
Geophys. Res. 96, 8113 – 8126.
Darling, R.S., Florence, F.P., 1995. Apatite light rare earth element
chemistry of the Port Leyden nelsonite, Adirondack Highlands,
New York: implications for the origin of nelsonite in anorthosite
suite rocks. Econ. Geol. 90, 964 – 968.
Deer, W.A., Howie, R.A., Zussman, J., 1986. Rock-Forming Min-
erals, 2nd ed. Disilicates and Ring Silicates, vol. 1B, Geological
Society, London.
Dulski, P., 2001. Reference materials for geochemical studies: new
analytical data by ICP-MS and critical discussion of reference
values. J. Geostand. Geoanal. 25, 87 – 125.
Exley, R.A., 1980. Microprobe studies of REE-rich accessory
minerals: implications for Skye granite petrogenesis and REE
mobility in hydrothermal systems. Earth Planet. Sci. Lett. 48,
97 – 110.
Fleet, M.E., Liu, X., Pan, Y., 2000. Rare-earth elements in chlor-
apatite [Ca
10
(PO
4
)
6
Cl
2
]: uptake, site preference and degradation
of monoclinic structure. Am. Mineral. 85, 1437 – 1446.
Fo¨rster, H.-J., 1998. The chemical composition of REE – Y – Th –
U-rich accessory minerals in peraluminous granites of the Erz-
gebirge – Fichtel gebirge regio n, Germany Part I. The mona-
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–7270
zite – (Ce) – brabantite solid solution series. Am. Mineral. 83,
259 – 272.
Fo¨rster, H.-J., 2000. Cerite – (Ce) and thorian synchysite – (Ce) from
the Niederbobritsch granite, Erzgebirge, Germany: implications
for the differential mobility of the LREE and Th during alter-
ation. Can. Mineral. 38, 67 – 79.
Fo¨rster, H.-J., Harlov, D.E., 1999. Monazite – (Ce) – huttonite solid
solutions in granulite-facies metabasites from the Ivrea – Verbano
Zone, Italy. Mineral. Mag. 63, 587 – 594.
Frietsch, R., 1967. The relationships between magnetite and hema-
tite in the iron ores of the Kiruna type and some other iron ore
types. Sver. Geol. Unders., Ser. C (625) 28 pp.
Frietsch, R., 1974. The occurrence and composition of apatite with
special reference to iron ores and rocks in northern Sweden.
Sver. Geol. Unders., Ser. C (694) 49 pp.
Frietsch, R., 1978. On the magmatic origin of iron ores of the
Kiruna type. Econ. Geol. 73, 478 – 485.
Frietsch, R., 1984. Petrochemistry of the iron ore-bearing metavol-
canics in Norrbotten county, northern Sweden. Sver. Geol.
Unders., Ser. C (802) 62 pp.
Frietsch, R., Perdahl, J.-A., 1995. Rare earth elements in apatite and
magnetite in Kiruna-type iron ores and some other iron ore
types. Ore Geol. Rev. 9, 489 – 510.
Frietsch, R., Tuisku, P., Martinsson, O., Perdahl, J.-A., 1997. Early
Proterozoic Cu – (Au) and Fe ore deposits associated with re-
gional Na – Cl metasomatism in northern Fennoscandia. Ore
Geol. Rev. 12, 1 – 34.
Geijer, P., 1910. Igneous rocks and iron ores of Kiirunavaara, Luos-
savaara and Tuolluvaara. Scientific and practical researches in
Lapland arranged by Luossavaara– Kiirunavaara Aktiebolag.
PhD Thesis, Univ. Uppsala, Uppsala, Sweden.
Geijer, P., 1931. The iron ores of the Kiruna type: geographical
distribution, geological characters, and origin. Sver. Geol.
Unders., Ser. C (367) 39 pp.
Geijer, P., 1967. Internal features of the apatite-bearing magnetite
ores. Sver. Geol. Unders., Ser. C (624) 32 pp.
Geijer, P., O
¨dman, O.H., 1974. The emplacement of the Kiruna iron
ores and related deposits. Sver. Geol. Unders., Ser. C (700) 48
pp.
Giere
´, R., 1996. Formation of rare earth minerals in hydrothermal
systems. In: Jones, A.P., Wall, F., Williams, C.T. (Eds.), Rare
Earth Minerals. Chemistry, Origin and Ore Deposits. The Min-
eralogical Society Series, vol. 7, pp. 105 – 150.
Haas, J.R., Shock, E.L., Sassani, D.C., 1995. Rare earth elements in
hydrothermal systems: estimates of standard partial molal ther-
modynamic properties of aqueous complexes of the rare earth
elements at high pressures and temperatures. Geochim. Cosmo-
chim. Acta. 59, 4329 – 4350.
Harlov, D.E., Fo¨rster, H.-J., 2002. High-grade fluid metasomatism
on both a local and regional scale: the Seward Peninsula, Alaska
and the Val Strona di Omegna, Ivrea – Verbano zone, northern
Italy. Part II. Phosphate mineral chemistry. J. Petrol. 43,
801 – 824.
Harlov, D.E., Fo¨ rster, H.-J., Nijland, T.G., 2002. Fluid induced nu-
cleation of (Y + REE) – phosphate minerals in apatite: nature and
experiment. Part I. Chlorapatite. Am. Mineral. 87, 245 – 261.
Henrı
´quez, F., Martin, R.F., 1978. Crystal-growth textures in mag-
netite flows and feeder dykes, El Laco, Chile. Can. Mineral. 16,
581 – 589.
Hoskin, P.W.O., Kinny, P.D., Wyborn, D., Chappell, B.W., 2000.
Identifying accessory mineral saturation during differentiation in
granitoid magmas: an integrated approach. J. Petrol. 41, 1365 –
1396.
Kerr, I.D., 1998. Mineralogy, chemistry and hydrothermal evolution
of the Pea Ridge Fe – oxide – REE deposit, Missouri, USA. MS
Thesis, University of Windsor, Windsor, Ontario, Canada.
Kerr, I.D., Samson, I.M., 1998. REE mineralogy of the Pea Ridge
Fe – REE deposit, Missouri. Geol. Soc. Am., Abst. Prog. 30, 370.
Lindblom, S., Broman, C., Martinsson, O., 1996. Magmatic-hydro-
thermal fluids in the Pahtohavare Cu–Au deposit in greenstone
at Kiruna, Sweden. Mineral Deposita 31, 307 – 318.
Longerich, H., Jackson, S.E., Gu
¨nther, D., 1996. Laser ablation
inductively coupled plasma mass spectrometric transient signal
data acquisition and analyte concentration calculation. J. Anal.
At. Spectrom. 1, 899 – 904.
Martinsson, O., 1997. Tectonic setting and metallogeny of the Kir-
una greenstones. PhD Thesis, Lulea
˚Univ. of Technology, Lulea
˚,
Sweden.
McKeown, F.A., Klemic, H., 1957. Rare-earth-bearing apatite at
Mineville, Essex County, New York. Bull. US Geol. Surv.
1046-B, 9 – 23.
Naslund, H.R., Aguirre, R., Dobbs, F.M., Henrı
´quez, F., Nystro¨m,
J.O., 2000. The origin, emplacement and eruption of ore mag-
mas. IX Congreso Geologico Chileno, Actas 2, 135 – 139.
Nystro¨m, J.O., Henrı
´quez, F., 1989. Dendritic magnetite and mini-
ature diapir-like concentrations of apatite: two magmatic fea-
tures of the Kiirunavaara iron ore. Geol. Fo¨ren. Stockh. Fo¨rh.
111, 53 – 64.
Nystro¨m, J.O., Henrı
´quez, F., 1994. Magmatic features of iron ores
of the Kiruna type in Chile and Sweden: ore textures and mag-
netite geochemistry. Econ. Geol. 89, 820 – 839.
O’Farrely, K.S., 1990. A stable isotopic investigation of the origin
and evolution of the Kiirunavaara iron mine, northern Sweden.
PhD Thesis, Univ. of Wales, Cardiff, Wales, United Kingdom.
Pan, Y., 1997. Zircon- and monazite-forming metamorphic reactions
at Manitouwadge, Ontario. Can. Mineral. 35, 105 – 118.
Pan, Y., Fleet, M.E., 1996. Rare earth element mobility during pro-
grade granulite facies metamorphism: significance of fluorine.
Contrib. Mineral. Petrol. 123, 151 – 262.
Pan, Y., Fleet, M.E., Macrae, N.D., 1993. Oriented monazite inclu-
sions in apatite porphyroblasts from the Hemlo gold deposit,
Ontario, Canada. Mineral. Mag. 57, 697–707.
Para
´k, T., 1973. Rare earths in the apatite iron ores of Lappland
together with some data about the Sr, Th and U content of these
ores. Econ. Geol. 68, 210 – 221.
Para
´k, T., 1975a. The origin of the Kiruna iron ores. Sver. Geol.
Unders., Ser. C (709) 209 pp.
Para
´k, T., 1975b. Kiruna iron ores are not ‘‘intrusive-magmatic ore
of the Kiruna type’’. Econ. Geol. 70, 1242 – 1258.
Pearce, N.J.G., Perkins, W.T., Gorton, M.P., Jackson, S.E., Neal,
C.R., Chenery, S.P., 1997. A compilation of new and published
major and trace element data for NIST SRM 610 and NIST
SRM 612 glass reference samples. J. Geostand. Geoanal. 21,
115– 144.
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–72 71
Petrı
´k, I., Broska, I., Lipka, J., Siman, P., 1995. Granitoid allanite –
(Ce): substitution relations, redox conditions and REE distribu-
tions (on an example of I-type granitoids, western Carpathians,
Slovakia). Geol. Carpath. 46, 79 – 94.
Rocholl, A.B.E., Simon, K., Jochum, K.P., Bruhns, F., Gehan, R.,
Kramar, U., Luecke, W., Molzahn, M., Pernicka, E., Seufert, M.,
Spettel, B., Stummeier, J., 1997. Chemical characterisation of
NIST silicate glass certified reference material SRM 610 by
ICP-MS, TIMS, LIMS, SSMS, INAA, AAS and PIXE. J. Geo-
stand. Geoanal. 21, 101 – 114.
Rocholl, A., Dulski, P., Raczek, J., 2000. New ID-TIMS, ICP-MS
and SIMS data on the trace element composition and homoge-
neity of NIST certified reference material SRM 610. J. Geo-
stand. Geoanal. 24, 261 – 274.
Roedder, E., 1992. Fluid inclusion evidence for immiscibility in
magmatic differentiation. Geochim. Cosmochim. Acta 56, 5– 20.
Roeder, P.L., MacArthur, D., Ma, X.-P., Palmer, G.R., Mariano,
A.N., 1987. Cathodoluminescence and microprobe study of
rare-earth elements in apatite. Am. Mineral. 72, 801 – 811.
Romer, R.L., Martinsson, O., Perdahl, J.-A., 1994. Geochronology
of the Kiruna iron ores and hydrothermal alterations. Econ.
Geol. 89, 1249 – 1261.
Rønsbo, J.G., 1989. Coupled substitutions involving REE’s and Na
and Si in apatites in alkaline rocks from Ilimaussaq, South
Greenland, and the petrological implications. Am. Mineral.
74, 896 – 901.
Seward, T.M., Barnes, H.L., 1997. Metal transport by hydrothermal
ore fluids. In: Barnes, H.L. (Ed.), Geochemistry of Hydrother-
mal Ore Deposits, 3rd ed. Wiley, NY, New York, pp. 435 – 486.
Watt, G.R., 1995. High-thorium monazite – (Ce) formed during dis-
equilibrium melting of metapelites under granulite-facies condi-
tions. Mineral. Mag. 59, 735 – 743.
Webster, J.D., 1997. Exsolution of magmatic volatile phases from Cl-
enriched mineralizing granitic magmas and implications for ore
metal transport. Geochim. Cosmochim. Acta 61, 1017 – 1029.
Young, E.J., Myers, A.T., Munson, E.L., Conklin, N.M., 1969.
Mineralogy and geochemistry of fluorapatite from Cerro de
Mercado, Durango, Mexico. U.S. Geol. Surv. Prof. Pap.
650-D, D84 – D93.
D.E. Harlov et al. / Chemical Geology 191 (2002) 47–7272
