Genesis and transformations of monazite, florencite and rhabdophane during medium grade metamorphism: examples from the Sopron Hills, Eastern Alps Ge ´za Nagy a, * , Erich Draganits b , Attila Deme ´ny a , Gyo ¨rgy Panto ´ a , Pe ´ter A ´ rkai a a Laboratory for Geochemical Research, Hungarian Academy of Sciences, Budao ¨rsi u ´t 45, H-1112 Budapest, Hungary b Institut fu ¨r Geologie, Universita ¨t Wien, Althanstrasse 14, A-1090 Vienna, Austria Abstract Electron microprobe studies on the age, mineral chemistry and alteration on accessory LREE-phosphate minerals have been carried out in medium-grade metamorphic rocks of the Sopron Hills belonging to the Lower Austroalpine tectonic unit. Monazite (and xenotime) is relatively common, whereas rhabdophane and florencite are restricted to certain rock types. A first generation of monazite was formed in mica schists during the pre-Alpine, Hercynian metamorphism at 575 – 700 jC and 1.8– 3.8 kbar as evidenced by P – T data from the literature, their mineral paragenetic and textural characteristics and supported by Th – U – total Pb ages of ca. 300 Ma. In orthogneisses, monazite is rare and of igneous origin. Kyanite quartzites and leucophyllites that were formed by Mg metasomatism contain inherited monazite from the precursor rocks. A new generation of monazite was also formed during the Alpine metamorphism at V 550 jC, 13 kbar according to the literature data, giving ages around 75 Ma. Pronounced negative Eu anomalies were found in the igneous monazites (Eu/Eu * < 0.35), while most of the metamorphic monazites have moderately negative Eu anomalies (Eu/Eu*>0.4). Small differences have been observed in Y and HREE contents, whereas the LREE sections of the rare-earth element (REE) patterns nearly coincide. Th and Ca enter the monazite structure at the expense of REE, nearly according to the brabantitic replacement 2REE 3+ X Th 4+ + Ca 2+ . In some mica schists, monazite is altered to rhabdophane. Rhabdophane, distinguished from monazite by quantitative electron microprobe analysis by low-oxide total, is found in many mica schists and orthogneisses. It forms fine-grained aggregates, often attached to apatite or monazite. It usually has higher Y and Ca contents and a less pronounced negative Eu anomaly than that of coexisting monazite. It may have been formed either by crystallization from REE-containing hydrous solutions or from monazite reacting with Y – Ca-containing solutions. Florencite appears only in some leuchtenbergite-bearing leucophyllites, kyanite quartzites and REE-rich clasts. It is often idioblastic and may be grown on apatite or monazite. It is chemically close to its ideal composition, but Ca, Sr and Th may replace REE in minor amounts. In some grains, ThO 2 may reach 10 wt.%. The data indicate that the charge balance is maintained by different mechanisms in low- and high-thorian florencite. No Y or HREE (above Gd) could be measured in florencite. No fractionation was observed between coexisting monazite and florencite; however, monazite inclusions in florencite are depleted in La –Ce and enriched in HREE. D 2002 Elsevier Science B.V. All rights reserved. Keywords: REE minerals; Polymetamorphism; Eastern Alps; REE geochemistry; Th – U – total Pb geochronology 0009-2541/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. PII:S0009-2541(02)00147-X * Corresponding author. E-mail address: [email protected] (G. Nagy). www.elsevier.com/locate/chemgeo Chemical Geology 191 (2002) 25 – 46
22
Embed
Genesis and transformations of monazite, florencite and ... · Genesis and transformations of monazite, florencite and rhabdophane during medium grade metamorphism: examples from
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Genesis and transformations of monazite, florencite and
rhabdophane during medium grade metamorphism:
examples from the Sopron Hills, Eastern Alps
Geza Nagy a,*, Erich Draganits b, Attila Demeny a, Gyorgy Panto a, Peter Arkai a
aLaboratory for Geochemical Research, Hungarian Academy of Sciences, Budaorsi ut 45, H-1112 Budapest, HungarybInstitut fur Geologie, Universitat Wien, Althanstrasse 14, A-1090 Vienna, Austria
Abstract
Electron microprobe studies on the age, mineral chemistry and alteration on accessory LREE-phosphate minerals have been
carried out in medium-grade metamorphic rocks of the Sopron Hills belonging to the Lower Austroalpine tectonic unit.
Monazite (and xenotime) is relatively common, whereas rhabdophane and florencite are restricted to certain rock types. A first
generation of monazite was formed in mica schists during the pre-Alpine, Hercynian metamorphism at 575–700 jC and 1.8–
3.8 kbar as evidenced by P–T data from the literature, their mineral paragenetic and textural characteristics and supported by
Th–U–total Pb ages of ca. 300 Ma. In orthogneisses, monazite is rare and of igneous origin. Kyanite quartzites and
leucophyllites that were formed by Mg metasomatism contain inherited monazite from the precursor rocks. A new generation of
monazite was also formed during the Alpine metamorphism at V 550 jC, 13 kbar according to the literature data, giving ages
around 75 Ma. Pronounced negative Eu anomalies were found in the igneous monazites (Eu/Eu * < 0.35), while most of the
metamorphic monazites have moderately negative Eu anomalies (Eu/Eu*>0.4). Small differences have been observed in Y and
HREE contents, whereas the LREE sections of the rare-earth element (REE) patterns nearly coincide. Th and Ca enter the
monazite structure at the expense of REE, nearly according to the brabantitic replacement 2REE3 + XTh4 + +Ca2 + . In some
mica schists, monazite is altered to rhabdophane. Rhabdophane, distinguished from monazite by quantitative electron
microprobe analysis by low-oxide total, is found in many mica schists and orthogneisses. It forms fine-grained aggregates, often
attached to apatite or monazite. It usually has higher Y and Ca contents and a less pronounced negative Eu anomaly than that of
coexisting monazite. It may have been formed either by crystallization from REE-containing hydrous solutions or from
monazite reacting with Y–Ca-containing solutions. Florencite appears only in some leuchtenbergite-bearing leucophyllites,
kyanite quartzites and REE-rich clasts. It is often idioblastic and may be grown on apatite or monazite. It is chemically close to
its ideal composition, but Ca, Sr and Th may replace REE in minor amounts. In some grains, ThO2 may reach 10 wt.%. The
data indicate that the charge balance is maintained by different mechanisms in low- and high-thorian florencite. No Y or HREE
(above Gd) could be measured in florencite. No fractionation was observed between coexisting monazite and florencite;
however, monazite inclusions in florencite are depleted in La–Ce and enriched in HREE.
Mineral abbreviations after Kretz (1983): Bt—biotite; And—andalusite; Sil—sillimanite; Grt—garnet; Chl—chlorite; Ser—sericite.a Typical error, calculated for anal 2.b In apatite.c Monazite-like grain with low oxide total.d Including Y.
G. Nagy et al. / Chemical Geology 191 (2002) 25–4630
and orthogneisses. Florencite appears only in some of
the leuchtenbergite-containing leucophyllite, kyanite
quartzite and REE-rich clast samples. Selected ana-
lytical results are given in Tables 1 and 3; the whole
set of analyses is available on request.
4.1. Monazite
As monazite does not contain considerable amount
of H2O or any other light element which could not be
detected by electron microprobe, the analyses with
oxide totals above 97.5 wt.%were taken to be monazite
only. Mineral grains with appearance and composition
similar to monazite, however, with lower oxide totals
will be treated in the last paragraph of this section.
In mica schists and kyanite quartzites, monazite is
abundant at certain localities (Brennberg, Voroshıd,
etc.); however, it can also be absent in other places. It
usually appears as 10–100 Am, mainly xenoblastic
grains. Some idioblastic grains were found in or among
well-preserved biotite or andalusite, which formed du-
ring Hercynian metamorphism (Kishazi and Ivancsics,
1985) and biotite may also be included in monazite
(Fig. 3a) indicating their contemporaneous formation.
In mica schists with Alpine mineral assemblages and in
kyanite quartzites, the majority of the monazite grains
have similar sizes and compositions (Fig. 3b). We
assume that these are inherited pre-Alpine Hercynian
grains. A newmonazite generation can also be found in
some of the samples with slightly (rarely highly)
different compositions and small sizes that have been
formed during Alpine events (Fig. 3c, Table 1.)
In most of the gneisses and leucophyllites, mon-
azite is rare and appears in the form of small (f 10
Am) rounded grains often included in apatite together
with zircon (Fig. 3d).
In some of the REE-rich clast samples, the Alpine
monazite has grown to over 50 Am in the form of
idioblastic grains (Fig. 4a). In other REE-rich clasts,
10–50 Am monazite grains are found together with
florencite of similar size and may contain leuchten-
bergite inclusions. Monazite can also form small
inclusions in large florencite grains.
4.1.1. Chemical compositions
The majority of the monazites analysed are chemi-
cally similar (Table 1). In most cases, only small
differences were found in the relative abundances of
REE and Y, which can be seen in chondrite-normal-
ised REE patterns (Fig. 5a) or, slightly better, on REE
patterns normalised to one of the monazites (Fig. 5b).
The following characteristics of monazite chemistry
have been observed.
a) Differences in Eu anomalies. The majority of the
monazite grains in orthogneisses and leucophyllites
have high negative Eu anomalies with Eu/Eu * < 0.4.
The monazite inclusions in apatite are characterised
by even lower Eu/Eu* ratios ( < 0.3). On the other
hand, for monazites in mica schists and kyanite
quartzites, Eu/Eu* ratios are between 0.4 and 1 with
a few exceptions (Figs 5a and 6).
A positive Eu anomaly (Eu/Eu*>1) has been
observed in some of the Alpine monazites: in a few
grains in kyanite quartzite and in each point analysis
of a REE-rich clast (sample S-2, which was used for
Th–U–total Pb age determination).
b) Differences in Y and HREE contents. The LREE
parts of the chondrite-normalised patterns nearly coin-
cide, the HREE parts are, more or less, steeper. Y and
HREE change parallel to each other. In the case of
monazites in gneisses and leucophyllites with the
strongest negative Eu anomalies, the range of Y con-
tents may exceed 4 oxide wt.%. The monazites with
moderately negative Eu anomalies, i.e. in kyanite
quartzites, in mica schists and partly in gneisses and
leucophyllites Y2O3 < 2 wt.% with a few exceptions
(Fig. 6). Zoning in Y content was observed in two
monazite grains in mica schists, as the Y decreased, Eu/
Eu* increased slightly towards the grain boundaries.
Monazite inclusions in florencites or monazite
found among florencite grains is depleted in La and
Ce (and enriched in the heavier REE and Y) (Fig. 7).
c) Differences were also found in Th and Ca
contents. In most igneous and metamorphic monazite,
ThO2>1 wt.%. However, in the REE-rich clast sam-
ples, the ThO2 content of the medium size (>10 Am)
monazite changes between 0 and 20 wt.%. The Th and
Ca contents show the same trend (Fig. 8a), at the
expense of the REE.
Th and Ca can be incorporated by monazite simul-
taneously with Ca and/or Si in order to maintain
charge balance according to the following reactions
(Van Emden et al., 1997):
2 � REE3þ X ðTh;UÞ4þ þ Ca2þ ð1ÞREE3þ þ P5þ X ðTh;UÞ4þ þ Si þ ð2Þ
G. Nagy et al. / Chemical Geology 191 (2002) 25–46 31
Fig. 3. Backscattered electron pictures of monazite and rhabdophane grains. The numbers correspond to analyses in Table 1. (a) Idioblastic,
inhomogeneous monazite with biotite inclusions among biotite (Bt) and andalusite (And) grains. (b) Xenoblastic pre-Alpine monazite in kyanite
quartzite. (c) Alpine monazite grains in the same kyanite quartzite. (d) Monazite (Mnz) and zircon (Zrn) inclusions in apatite (grey rounded
grains), and rhabdophane (Rha) attached to apatite in gneiss. (e) Monazite (brighter) and monazite-like grains with low oxide total (darker). (f)
Monazite (brighter core) altered to rhabdophane (darker margin) in biotite–andalusite– sillimanite schist. Note that the biotite (Bt) has been
chloritised (Chl) next to the altered grain.
G. Nagy et al. / Chemical Geology 191 (2002) 25–4632
Fig. 4. Backscattered electron pictures of LREE minerals. Numbers in (a) and (b) refer to analyses in Table 1 and in (c)– (f) to Table 3. (a)
Alpine monazite (white), Cl–apatite (grey) and kyanite (black) in a REE-rich clast sample (S-2). (b) Rhabdophane (white) among muscovite
grains in a schist with Alpine minerals, not containing monazite (sample NM-6). (c) Florencite (grey) grown on pre-Alpine monazite (white) in
kyanite (Ky). The space among the two marked grains is filled with leuchtenbergite (dark grey). The rock is kyanite quartzite. (d) Idioblastic
florencite grains (white) among muscovite (brighter) and leucophyllite (darker) lamellae, in leucophyllite. The circle on the upper florencite was
caused by the electron beam during analysis. (e) Zoned florencite in a REE-rich clast. The white zones contain numerous small monazite
inclusions. (f) High-thorian parts (encircled) in low-thorian florencite in the same REE-rich clast.
G. Nagy et al. / Chemical Geology 191 (2002) 25–46 33
In monazites of the Sopron Hills, the ‘‘brabantitic’’
replacement, i.e. the joint entry of Th and Ca:
2 � REE3þ X Th4þ þ Ca2þ ð1aÞ
is the most important. Deviations from this relation
are caused in most cases by entry of U and Si (Fig.
8b). In addition, most inclusions in apatite have
excess Ca content.
4.1.2. Th–U–total Pb ages
An idioblastic monazite inside biotite has been
measured in one of the biotite–andalusite–sillimanite
schists (sample DE-10). Repeated measurements gave
two similar ages: 300F 41 Ma (Table 2) and 310F 34
Ma (Fig. 9a), which support the opinion that the
monazite was formed in the main pre-Alpine Hercy-
nian metamorphism. In a kyanite quartzite (grey
quartzite, sample DE-6) which has two different
monazite generations, the medium size (15–100
Am) primary grains gave nearly the same age:
296F 41 Ma (Fig. 9b), overlapping well within the
limits of error and supporting the theory that they
were inherited from the precursor mica schists.
In the kyanite quartzite mentioned above, the
small (V 7 Am) grains of the second-generation
monazite have lower Pb contents (Fig. 9b), often
below the detection limit. This indicates a younger
age of formation; however, the data points cluster
around a straight line only with weak correlation
(r2 = 0.54). Slightly larger, 10-Am size monazite
grains in a white quartzite (sample DE-12) that was
presumably formed from different source rock
(gneiss) by Mg metasomatism gave an age of
83F 19 Ma (Table 2, Fig. 9b). The biggest size
grains (f 50 Am) of the new generation monazite
(with wide range of Th contents) appear in one of the
REE-rich clast samples (S-2). Some grains are char-
acterised by very low Th* and Pb contents
(Pb < 0.012 wt.%, i.e. the detection limit), proving
that this monazite type contained no excess Pb and
has not suffered Pb loss; thus, the Pb vs. Th*
regression line used for age calculation can be forced
through the origin. Excluding the data below detec-
tion limit, however, forcing the fitted line through the
origin, yields 71F 8 Ma. Using the whole data set
(including those below detection limit) without forc-
ing the fitted line through the origin (as in the other
cases) yields 70F 11 Ma (Fig. 9a). This range of
error seems to be more realistic, taken the high
uncertainties of the individual analyses into consid-
eration (alternative treatment: replacing of the data
below detection limit by Pb = 0 yielded a similar age:
69F 11 Ma). We presume that the younger monazite
formed simultaneously in the kyanite quartzites and
REE-rich clasts and that the differences are caused by
analytical uncertainties. The age data obtained on the
second-generation monazite indicate their formation
during the Eo–Alpine metamorphic event.
The Th–U–total Pb ages obtained are in good
agreement with our knowledge of Hercynian and
Alpine metamorphism. Extension of the method to
other rock samples will presumably supply new or
more detailed knowledge on the development of the
Eastern Alps.
Fig. 5. REE patterns of selected monazite (A–D) and rhabdophane
(E) analyses. A= in biotite–andalusite– sillimanite schist; B = in
biotite–muscovite–garnet schist; C = independent grain; D = inclu-
sion in apatite; E = rhabdophane; C–D–E are in the same gneiss
(metagranite). (a) Normalised to chondrite. (b) Normalised to
monazite ‘‘A’’.
G. Nagy et al. / Chemical Geology 191 (2002) 25–4634
4.1.3. Monazite-like grains with low oxide totals
Mineral grains with oxide totals between 93.7 and
96.5 wt.% were found in abundance in three mica
schist samples (Fig. 2a), one of them also containing
monazite. They are similar to monazite both in appear-
ance, grain size (10–60 Am) and composition (Table
1). In BSE images, they seem slightly darker than
coexisting monazite (Fig. 3f), indicating the presence
of some light component. This is most probably H2O,
though other light elements can also be present in
approximately 5 wt.% amount, based on the difference
from 100%. This group of grains may be rhabdophane
with 0.5 H2O, or an unknown variety of monazite.
4.2. Rhabdophane
LREE-phosphate grains with oxide totals below
96.5 wt.% and with appearance different from the
Fig. 6. Eu anomaly vs. Y2O3 contents in monazite and rhabdophane (a) in mica schists and ‘‘grey’’ quartzites, (b) in gneisses, leucophyllites and
‘‘white’’ quartzite (originating from gneiss). Legend: monazites: A= in mica schist; B = in kyanite quartzite; C = inclusions in apatite (in gneiss
and in leucophyllite); D = in gneiss; E = in leucophyllite. F =monazite-like grains with low oxide totals (in three mica schists).
G = rhabdophanes. Note that each monazite inclusions in apatite fall below Eu/Eu * = 0.3. Five of the rhabdophane analyses, marked with
arrows, fall between Y2O3 = 4.5 and 6.0 wt.%.
G. Nagy et al. / Chemical Geology 191 (2002) 25–46 35
monazite are widespread both in mica schists and in
orthogneisses. They form fine-grained aggregates con-
taining usually < 10-Amgrains with similar appearance
to rhabdophane described by Sawka et al. (1986) or
Banfield and Eggleton (1989). The aggregates are often
attached to apatite in gneisses (Fig. 3d), andmay appear
on altered monazite grains in mica schists (Fig. 3f) or
individually as pseudomorphs (Fig. 4b).
Quantitative analyses revealed that most of these
grains have a rather constant, nearly stoichiometric P
contents (between 3.8 and 4.1 pfu in 47 cases from 58
analyses) and high Y+REE contents (between 3 and 4
pfu in 45 cases, Fig. 10a). No elements other than REE,
Y, P, Th, U, Ca and Si could be determined by EMPA.
Fe and Mn may be present in small concentrations
( < 0.5 wt.%), as the REE-dominated part of the EMPA
spectra did not allowmore precise analysis. The spectra
would also allow small amounts of Cr and V, but their
low concentrations in the host rock (V 10 ppm) would
argue against it. The oxide totals fall between 85.7 and
96.4 wt.% (Fig. 2a). The only knownmineral with such
or similar composition is rhabdophane.
In the case of EMPA of rhabdophane, the difference
of the oxide totals from 100% is nearly equal to the H2O
Fig. 7. Chondrite-normalised REE-patterns of coexistent monazite
and florencite (a) in a REE-rich clast sample and (b) in a grey
quartzite. Legend: M1=medium size monazite; M2= small mon-
azite grain included in florencite [on (a)] or in the middle of
florencite aggregate [on (b)]; F = florencite.
Fig. 8. (a) Ca vs. Th, (b) Ca + Si vs. Th +U in monazite and
rhabdophane, expressed in ionic numbers recalculated on basis of
16 oxygens. Legend: A=monazite excluding inclusions in apatite,
B =monazite inclusions in apatite, C = rhabdophane. The dashed
lines correspond to entry in equal amounts. Four rhabdophane
analyses fall between Ca = 1.5 and 2.5, and three of them between
Ca + Si = 2.1 and 2.6, marked with arrows on the respective figures.
G. Nagy et al. / Chemical Geology 191 (2002) 25–4636
content (in about 1 wt.%) as demonstrated by Bowles
and Morgan (1984) determining the water contents by
differential thermal analysis. The sample that was
analysed by them has low Ca content, while our
LREE-phosphate grains have notable Ca contents (up
to 9 oxide wt.%) that exceed the equivalent Th (plus U)
content according to the brabantitic replacement reac-
tion (1) (Fig. 8a). Mineral grains with similar compo-
sition as our analysis 7 in Table 1 have been identified
by electron diffraction by Dorfman et al. (1993) and
described as calcium rhabdophane. Among the mon-
azite alteration products, Poitrasson et al. (2000) also
found a LREE-phosphate phase with similar composi-
tion (MO133) and with hexagonal structure, deter-
mined by laser Raman spectroscopy, which is the struc-
ture of rhabdophane. They observed a REE, Th and Ca
substitution, coupled with monoclinic to hexagonal
Table 2
Selected analytical results (in wt.%) used for Th–U– total Pb age
determination
Biotite–andalusite–
sillimanite schistaKyanite quartziteb
Sample: DE-10 Sample: DE-12
One f 100-Am grain Several f 10-Am grains
Th
(wt.%)
U
(wt.%)
Pb(1)
(wt.%)
Pb(2)
(wt.%)
Th
(wt.%)
U
(wt.%)
Pb(1)
(wt.%)
Pb(2)
(wt.%)
5.60 1.36 0.122 0.134 1st group
3.12 0.46 0.024 0.052 3.18 0.86 0.029 0.029
5.66 0.87 0.115 0.136 2.71 0.89 0.029 0.015
7.43 0.76 0.114 0.135 8.73 1.01 0.032 0.043
6.19 0.67 0.113 0.118 11.75 3.64 0.090 0.092
2.93 0.38 0.060 0.053 8.02 1.88 0.045 0.065
5.12 0.54 0.096 0.098 4.40 0.98 0.022 0.050
5.89 0.85 0.107 0.112 10.21 3.16 0.080 0.078
3.22 0.52 0.057 0.052
3.49 0.39 0.034 0.058 2nd group
3.61 0.48 0.051 0.058 1.45 3.19 0.109 0.098
3.30 0.40 0.061 0.056 5.07 1.16 0.071 0.076
5.16 0.45 0.081 0.100 5.75 0.48 0.072 0.079
5.46 0.56 0.097 0.099 5.01 0.36 0.083 0.090
2.79 0.44 0.058 0.060
3.15 0.33 0.059 0.050
5.32 0.47 0.071 0.071
4.66 0.47 0.055 0.078
5.34 0.80 0.088 0.097
5.36 1.38 0.128 0.132
6.75 0.89 0.104 0.095
6.82 0.78 0.085 0.118
2.92 0.38 0.041 0.048
4.18 0.53 0.080 0.080
a The first of the two data sets measured on the same grain,
which is not plotted on Fig. 9a. The equation of the fitted line: