Introduction

Monazite [(LREE, Th, U, Ca) (P, Si)O4] is a common accessory mineral in various types of crustal rocks, and is extremely useful for understanding and timing fluid infiltration events and subsequent mineral-fluid interaction processes (e.g., Harlov et al. 2005, 2007, 2011; Williams et al. 2007; Budzyń et al. 2010, 2011, 2017; Upadhyay and Pruseth 2012). The complex compositional zone patterns and decomposition textures preserved in monazite are generally considered to be the products of fluid-mediated element mass transfer that record the local chemistry of metasomatic/metamorphic reactions as well as date them (e.g., Broska et al. 2005; Rasmussen and Muhling 2009; Hetherington et al. 2010; Harlov et al., 2011; Williams et al., 2011; Ondrejka et al. 2012; Upadhyay and Pruseth 2012). Moreover, based on net-transfer equilibrium with garnet or xenotime and/or differences in element partitioning coefficients, monazite has been utilized to estimate metamorphic temperatures (Gratz and Heinrich 1997, 1998; Pyle et al. 2001), indicate porphyroblast (e.g., garnet and xenotime) growth or breakdown, and to investigate melt crystallization (Zhu et al. 1999; Stepanov et al. 2012; Xing et al. 2013).

Ti–Fe oxide phases (e.g., magnetite, ilmenite, and rutile) and pyrite, are pervasive in magmatic and fluid-related ore deposits. Recent studies have revealed that the crystal growth morphology of and trace element concentrations in Ti–Fe phases are sensitive to changing fluid chemistry, temperature, and pressure (e.g., Nadoll et al. 2012, 2014a,b; Hu et al. 2014, 2015b, 2017; Wen et al. 2017; Xie et al. 2017; Chen et al. 2020). However, reports of metasomatically induced partial alteration of Ti–Fe oxide minerals closely associated with the formation of apatite–allanite–epidote coronas around monazite have so far not been reported in the literature.

The aim of this study is to investigate fluid-aided processes during the metasomatic alteration and subsequent metamorphism of a granite, associated with a nearby sulphate-bearing evaporate deposit, to a metagranite in the Jiao-Liao-Ji orogenic belt, North China Craton. This is accomplished via the systematic integration of metasomatic processes involving monazite and Ti-bearing magnetite with stable S isotopic data from pyrite and geochronological data from zircon and monazite in a series of detailed petrographic observations, microtextural investigations, X-ray element mapping, electron microprobe and Raman spectral analysis.

Geological setting

The North China Craton consists of three Palaeoproterozoic tectonic belts: the khondalite belt, the trans North China orogeny, and the Jiao-Liao-Ji (JLJ) orogenic belt (Fig. 1a, Zhao et al. 2005). In the eastern North China Craton, the JLJ orogenic belt is more than 1000 km long, with a NNE-SSW strike. It divides the Eastern Block of the North China Block into the Longgang Block to the north in China and the Nangrim Block to the south in Korea (Fig. 1b; Li et al. 2005; Zhao et al. 2005, 2012).

Fig. 1
figure 1

modified from Tian et al. 2017), eastern Liaoning Province, with the metagranite location (16KD67) marked by a star in a circle. Some granite and pegmatite ages are also marked

a Tectonic subdivision of the North China Craton and adjacent regions (modified after Zhao et al. 2005, 2012). b The occurrence of the evaporitic sequence and graphite-rich sediments associated with the Liaohe Group (modified after Peng and Palmer 2002). The location of the salt dome and related sulfide deposits are based on Wang et al. 1998. c Sketched geological map of the Sanjiazi region (

Full size image

In the JLJ orogenic belt, vigorous metamorphic fluid activity has played a key role in the formation of many large or super-large deposits, such as the Dashiqiao magnesite-talc deposit (Zhang et al. 1988; Misch et al. 2018) and the Lianshanguan uranium deposit (Cuney et al. 2012). Uranium mineralization in this deposit is closely associated with the albitization of an associated early Palaeoproterozoic granite (Zhong and Guo 1988; Cuney et al. 2012).

A series of metasedimentary and metavolcanic successions in the JLJ orogenic belt are referred to as the Liaohe Group. They are commonly considered to have been deposited between 2200 and 2000 Ma (Hu et al. 2015a; Liu et al. 2017b, 2019b; Wang et al. 2017; Xu et al. 2019). The lower portion of the Liaohe Group (Zhang 1988) contains an evaporitic sequence made up of borate, sulphides, halides, Mg-rich carbonates, and Ca–Ba sulfates (Fig. 1b; Jiang et al. 1997; Wang et al. 1998; Peng and Palmer 2002; Liu et al. 2012; Yan et al. 2014; Dong et al. 2017). The metamorphic evolution of the JLJ orogenic belt has been established from the metabasite and metapelitic rocks (Tam et al., 2011, 2012a,b,c; Cai et al., 2017; Liu et al. 2017b, 2019b; Zou et al. 2017, 2018, 2019). It displays a clockwise P–T-t path with prograde (2100–1960 Ma), peak (1960–1900 Ma), isothermal decompression (1900–1850 Ma), and cooling stages (1850–1800 Ma).

The metagranite described in this study is a member of the Lieryu formation, which is a part of the Liaohe Group, and is located north of the town of Taziling (N 40°51′19.75″; E 123°18′49.95″) (Fig. 1c). This location is closely associated with evaporate deposits of the Liaohe Group and is less than 30 km from the largest known uranium deposit in northeast China (Fig. 1b). The metagranite varies in width from 0.5 to ~ 1.0 m and occurs parallel to a kyanite-bearing, garnet–sillimanite micaschist and a monzonite (Fig. 2a). Due to the dissemination of goethite within mineral grain interiors, along grain boundaries, and along microfractures, the metagranite exhibits a reddish color. Thin dark veins, mainly composed of muscovite and chlorite with a thickness of ca. 1–3 mm, can be clearly observed in hand specimens (Fig. 2b).

Fig. 2
figure 2

Field photographs showing the occurrences of the metagranite and surrounding rocks. a Spatial relations between the metagranite, kyanite-bearing, garnet–sillimanite micaschist, and monzonite. b Muscovite and chlorite vein in the albitized metagranite. The sample position is shown in the red rectangles in Fig. 2a. Mineral abbreviations: Chl chlorite, Ms  muscovite, Grt garnet, Sil sillimanite

Full size image

Analytical methods

Back scattered electron (BSE) imaging

Backscattered electron (BSE) imaging of monazite breakdown textures was conducted using a ZEISS ultra plus scanning electron microscope (SEM) with an electron beam voltage of 15 kV. A 50 mm2 OXFORD energy-dispersive spectrometer (EDS) was utilized to determine the elements present. The images were processed using INCA 4.4 (OXFORD) software. Two different brightness and contrast-level BSE images were taken in order to discern mineral zonation, mineral inclusions, and the surrounding mineral reaction textures. Discrete-color BSE images were also obtained for directly distinguishing different kinds of minerals by color.

Electron microprobe (EMP) analysis

Representative mineral compositions from the metagranite were obtained using a JEOL JXA-8230 electron microprobe (EMP) at the Wuhan Sample Solution Analytical Technology Co., Ltd., Wuhan, China. The EMP is equipped with 6 wavelength-dispersive spectrometers. The operating conditions for REE-bearing minerals (monazite, xenotime, allanite, epidote, apatite, and huttonite), silicate minerals, ilmenite, rutile, magnetite, and pyrite were as follows: accelerating voltage of 15 kV, beam current of 50 nA, and spot size of 3 μm. The counting times for Th, U, F, Cl, Y, and REE were 30 s on the peak and 10 s at each background position. For elements in titanite, feldspar, muscovite, and chlorite, the counting times for the peak and the background positions were 10 and 5 s, respectively. Lanthanum and Ce were measured on a PETJ crystal and Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu were measured on an LIFH crystal. Each element was calibrated using well-characterized natural and synthetic standards and reduced using the ZAF method.

EMP analyses of monazite, xenotime, apatite, britholite, thorite, allanite, REE-epidote, epidote, titanite, rutile, ilmenite, magnetite, goethite, pyrite, K-feldspar, plagioclase, muscovite, and chlorite along with their reaction products or inclusions are listed in Table 1 (monazite) and Supplementary Appendix S1.

Table 1 Representative monazite and xenotime EMP Analyses (wt%)
Full size table

X-ray compositional mapping

The internal structure of the select monazite grains and elements mapping was done using a Nova Nano SEM 450 with 50 mm2 Max OXFORD energy-dispersive spectrometer. The operating conditions included a low-vacuum state, 15 kV acceleration voltage, and 10 nA beam current. Mapping scans used the Ca-Kα, P-Kα, Th-Mα, La-Lα, Ce-Lα, Pr-Lβ, Nd-Lα, Y-Lα, and Si-Kα X-ray intensities. Element-mapping images were processed using the software AZtecLive version 3.0.

Raman spectral analysis

Fine-grained mineral inclusions in pyrite and magnetite were identified by Laser Raman at the Mirco-Raman Lab in the Institute of Geology, Chinese Academy of Geological Sciences, Beijing, China using an Horiba spectrometer LabRAM HR evolution equipped with an Olympus BX41 light microscope. Raman spectra were excited using a 532.02 nm Nd: YAG laser with a beam diameter of 1 μm, 100 mW laser power, and acquired via a 600 gr/mm optical grating through an 80 m confocal hole. The data were processed using LabSpec 6 software.

In situ S isotope analysis

In situ S isotopic analyses of pyrites were performed using a Neptune Plus MC–ICP–MS (Thermo Fisher Scientific, Bremen, Germany) equipped with a Geolas HD excimer ArF laser ablation system (Coherent, Göttingen, Germany) and nine Faraday cups at the Wuhan Sample Solution Analytical Technology Co. Ltd, Wuhan, China. Helium was used as the carrier gas for the ablation cell and was mixed with Ar after the ablation cell. Single spot ablation mode was used. Large spot size (44 μm) and slow-pulse frequency (2 Hz) were used to avoid the down-hole fractionation effect (Fu et al. 2016). 100 laser pulses were completed in one analysis. A new signal-smoothing device was used downstream from the sample cell to efficiently eliminate short-term variations in the signal, especially for slow-pulse frequency conditions (Hu et al. 2015c). The laser fluence was kept constant at  ~ 5 J/cm2. Pyrite standard PPP-1 (Fu et al. 2016) was used as the reference material for correcting the acquired data. In addition, the in-house reference material SP-Py-01 (δ34SV-CDT = 2.0‰ ± 0.4‰), was analyzed repeatedly as an unknown sample to verify the accuracy of the calibration. Sulfur isotope ratios are reported as δ34S relative to the Vienna Canyon Diablo Troilite (V-CDT) and are listed in Supplementary Appendix S2.

U–Pb laser ablation inductively coupled plasma-mass spectrometer (LA-ICPMS) dating of monazite and zircon

U–Pb dating of monazite from the metagranite was using a GeolasPro laser ablation system that consists of a COMPexPro 102 ArF excimer laser (wavelength of 193 nm) and a MicroLas optical system. An Agilent 7700e ICP-MS instrument was used to acquire ion-signal intensities. Helium was applied as a carrier gas. Argon was used as the make-up gas and mixed with the carrier gas via a T-connector before entering the ICP. A “wire” signal smoothing device is included in this laser ablation system. The spot size and frequency of the laser were set to 16 µm and 2 Hz, respectively. Monazite standard 44,096 and glass NIST610 were used as external standards for U–Pb dating and trace element calibration, respectively. Each analysis incorporated a background acquisition of approximately 20–30 s followed by 50 s of data acquisition from the sample. An Excel-based software ICPMSDataCal was used to perform off-line selection and integration of background and analyzed signals, time-drift correction and quantitative calibration for trace element analysis and U–Pb dating (Liu et al. 2010). Monazite U–Pb ages of the monzonite and kyanite-bearing, garnet–sillimanite micaschist were measured using an ELEMENT-XR inductively coupled plasma-mass spectrometer (ICP-MS) attached to a 200 nm femtosecond laser ablation (LA) system at the Department of Solid Earth Geochemistry, Japan Agency for Marine-Earth Science and Technology, Yokosuka, Japan. A secondary multiplier with counting mode was utilized for measuring 202Hg, 204(Pb + Hg), 206Pb, 207Pb, 208Pb, 232Th, and 238U. A 10 μm laser beam with 6 J/cm2 was rotated along the circumference of a circle with a 5 μm radius at a velocity of 2.1 μm at 1 and 3 Hz, resulting in a 20 μm ablation pit. To reduce elemental fractionation due to defocusing during the laser ablation process, a rotation raster ablation protocol was applied (Kimura et al. 2011). In this protocol, each analysis consisted of 30 s gas blank measurements followed by ca. 90 s laser ablation sampling. In total, three analyses of monazite 44,069 (Aleinikoff et al. 2006), one analysis of the Manangotry monazite (Horstwood et al. 2003), and one analysis of the 16-F-6 monazite (Simonetti et al. 2006) were inserted at intervals of every 10 analyses for the monazite. The specific correction technique for common Pb and analytical uncertainties has been described by Itano et al. (2016).

LA-ICPMS U–Pb dating of zircons was conducted using an AnlytikJena PQMS Elite ICPMS instrument with an ESI NWR 193 nm LA system at Beijing Createch Testing Technology Co. Ltd., Beijing, China. The laser beam diameter was 25 μm and was operated at a frequency of 10 Hz. Helium was applied as a carrier gas. Argon was used as the make-up gas and mixed with the carrier gas via a T-connector before entering the ICP. Each analysis consisted of about 15 s of background acquisition and 45 s of data acquisition. Each set of 5–10 analyses was followed by analyses of the glass standard NIST610 and the zircon standards GJ-1, 91,500, and Plešovice (Hu et al. 2008). The U–Pb concordia plots were processed with Isoplot 3.0, and the data are presented in Supplementary Appendix S3 with 1σ errors and 95% confidence limits (Ludwig 2003).

Silicate petrology and petrography of the metagranite

On the thin-section scale, the most striking feature in the metagranite is the extensive albitization of the plagioclase (Fig. 3a, b) and the heterogeneous distribution of minerals (see Fig. S1 in Supplementary Appendix S4). Albitization dominantly occurs in the cores of the plagioclase. The boundaries between the Na-rich core and Ca-richer rims are distinct and sharp (Fig. 3a, b). In the Ab–Or–An diagram (see Fig. S2 in Supplementary Appendix S4), the albitized plagioclase cores plot between An1 and An10, with an average of An4. The BSE brighter rims have distinctly higher An values (An = 18–19) corresponding to oligoclase (Supplementary Appendix S1). Occasionally, anti-perthite is seen in the core of the plagioclase grains. Feldspar in the immediate vicinity of muscovite–chlorite veins is near end member K-feldspar. Relative to plagioclase, K-feldspar is minor, and only a few fine grains are found near the muscovite–chlorite vein or occur peripherally to the plagioclase grains (see Fig. S1b in Supplementary Appendix S4).

Fig. 3
figure 3

X-ray element maps for a Ca, b Na, c K, and corresponding d BSE image of albitization and chloritization in the metagranite. Mineral abbreviations: Ab albite, Ep epidote, Mnz monazite, Pl plagioclase, Qz quartz

Full size image

Thin sections from the metagranite can be subdivided into a coarse muscovite-rich region, a quartz- and chlorite-rich domain, and a quartz-rich chlorite-absent domain. Muscovite in the metagranite shows a broad grain size range from 0.1 mm to 5 mm. In the quartz-rich region, varying degrees of chloritization accompany the magnetite and epidote along the muscovite cleavages (Fig. 3c, d). In the thinner veins, muscovite fills the centre, and chlorite occurs along the flanks. The muscovite–chlorite veins also cut through pre-existing coarse-grained muscovite. Allanite occurs along cleavages in the muscovite. Locally, fine-grained monazite and allanite can be found sporadically distributed in the chlorite zone (see Fig. S1b in Supplementary Appendix S4).

Coarse muscovite is characterized by the lowest SiO2, FeO, and MgO and the highest Al2O3 and TiO2 contents, which range between 45.17 and 47.46 wt%, 4.09–4.86 wt%, 0.59–0.90 wt%, 30.34–33.40 wt%, and 0.82–1.33 wt%, respectively (Supplementary Appendix S1). Compared with the coarse muscovite, fine-grained muscovite, accompanied by albite, shows a slight increase in FeO but still has a lower FeO content than muscovite near magnetite and pyrite. The muscovite vein exhibits a narrow SiO2 and Al2O3 compositional range and has the highest FeO content (6.95–9.51 wt%). Retrograde, fine-grained muscovite shows a slightly lower Ti content than that of fresh coarse-grained muscovite. However, the skeletal muscovite associated with magnetite has a higher Ti and Fe content compared to the coarse muscovite. Chlorite in muscovite cleavages has Al contents between 17.66 and 19.52 wt%, such that Al on the tetrahedral position ranges from 1.00 to 1.16 (apfu).

Quartz grains commonly exhibit irregular, lobate grain boundaries or occur as inclusions in muscovite or albitized plagioclase. Pervasive fractures and vugs or interstices quartz may remain after the dissolution of some minerals.

Phosphate, oxide, and sulphide mineralogy of the metagranite

Monazite, xenotime, and huttonite

In the metagranite, monazite occurs as euhedral to subhedral inclusions within the feldspar and muscovite. The monazite crystals are 30–200 μm long with length to width ratios of  ~ 1:1–4:1. Inclusions in the monazite dominantly consist of fluorapatite and huttonite (ThSiO4) (Fig. 4) where the huttonite is a metastable phase. Inclusions of xenotime can also occur in the monazite (Fig. S3 in Supplementary Appendix S4). More than 300 monazite grains were documented under BSE imaging in two thin sections. Most of the grains display a distinct BSE dark core with a lighter area, which in turn is surrounded by a BSE gray concentric oscillatory zone, and then surrounded by a BSE bright mantle with a BSE darker gray rim (Fig. 4a, b, c). Altered domains in the monazite can be easily distinguished from other regions by extensive, BSE bright, fine-grained inclusions, and a high porosity.

Fig. 4
figure 4

BSE images showing various examples of altered monazite from the metagranite. a Monazite showing regions of alteration rich in porosity along with huttonite inclusions. b Monazite with fine-grained inclusions and extensive porosity in the altered area, which is rimmed by an apatite, allanite, and epidote corona. c Monazite with BSE bright, BSE gray, BSE dark, and BSE light regions. A high porosity and fine-grained huttonite inclusions are characteristic of the altered areas. d Monazite with a BSE dark core and BSE bright mantle showing alteration along fractures and numerous huttonite inclusions. A BSE gray rim truncates earlier domains. e Primary monazite partially replaced by secondary apatite with a corona of REE epidote and muscovite. f Monazite partially replaced by secondary apatite with small remnants of secondary monazite. Mineral abbreviations: F-Ap  fluorapatite, Aln allanite; Hut huttonite

Full size image

A three-layered corona structure consisting of successive concentric zones of first apatite, followed by allanite, REE-epidote, and then epidote surrounds a subset of the monazite grains. The apatite layer consists of a narrow band (< 20 μm) that follows the original shape of the monazite grain. Apatite can also partially/totally replace the monazite leaving behind fine-grained (< 0.5 μm) remnants of monazite in an apatite matrix (Fig. 4e, f). The apatite zone is surrounded by a few- to tens-of-microns-wide allanite ring, which grades into REE epidote and then epidote (Fig. 4a, d). Locally, allanite rimmed by epidote can also be observed along muscovite cleavage planes (see Fig. S1b in Supplementary Appendix S4). The width of the epidote zone can vary from a few microns to several hundreds of microns. Formation of these apatite–allanite–epidote coronas surrounding the monazite took place via the following generalized reaction:

$$\begin{aligned} &{\text{Monazite}} + {\text{Muscovite}} + {\text{Anorthite}}\left( {{\text{in}}\;{\text{Plagioclase}}} \right)\\& \quad + F\left( {\text{in fluid}} \right) \to {\text{Fluorapatite}} + {\text{Allanite}}\\& \quad + {\text{Epidote}} + {\text{Th - silicate}} + {\text{Quartz}}{.} \end{aligned}$$
(1)

Five distinct domains (the BSE dark core, along with BSE light areas, followed by a BSE gray concentric oscillatory zone, BSE bright mantle, and BSE gray rim) of seven zoned monazite grains from the metagranite were systematically analyzed by EMP. The variation trends from core to rim of four representative grains are illustrated in Fig. S4 (Table 1, see also Supplementary Appendix S1 and S4). Among the five regions, the BSE dark core was characterized by the highest LREE and P2O5 (28.29–29.54 wt%) and lowest ThO2 (2.78–4.94 wt%), SiO2 (0–0.03 wt %), and PbO (0.53–0.74 wt%) concentrations. It was also the only domain in which the S content exceeded the detection limit (Table 1). In contrast, BSE light areas in the BSE dark core show significant differences in higher UO2 (2.54–2.67 wt%) and CaO (2.18–2.44 wt%), and lower Nd2O3 (9.12–9.29 wt%) and Sm2O3 (1.37–1.40 wt%) compared to the other zones (Fig. 5). The concentric BSE gray oscillatory zone surrounding the BSE dark core has the lowest UO2 content (0.43–0.60 wt%) in all domains. The altered BSE bright mantle surrounding the BSE gray oscillatory zone has high ThO2 (13.06–16.67 wt%) and SiO2 (2.17–2.93 wt%) contents with correspondingly lower P2O5 (23.61–25.70 wt%), LREE, and CaO (0.92–1.16 wt%) values. In contrast, the BSE gray rim shows element values intermediate between those of the core and the mantle (Table 1). The Gd fraction (XGd = Gd/∑REE) exhibits an extremely narrow range (0.037–0.041) no matter which zone it is measured in (Supplementary Appendix S1).

Fig. 5
figure 5

BSE images and Th, Si, Ca, P, La, Ce, Pr, Nd, and Y mapping of two monazites from the metagranite. Warmer colors indicate higher element concentrations

Full size image

On the REE + P + Y vs. Th + U + Si diagram (Fig. 6a), the concentric zone, BSE gray rim, and BSE bright mantle indicate that the huttonite substitution dominates. The ThO2 content exhibits the broadest variation (2.78–16.67 wt %) between the BSE dark core, BSE bright mantle, and BSE gray rim. It has a positive correlation with SiO2 (Fig. 6b).

Fig. 6
figure 6

Compositional plot of monazite and xenotime from the albitized granite. a Th + U + Si vs. REE + Y + P plot (atomic proportions) showing the ideal cheralite and thorite/huttonite substitution vectors (straight lines). b Th + U vs. Si substitution diagram for monazite and xenotime. The thorite substitution is given by the dashed line

Full size image

The xenotime inclusions in the monazite have Y2O3 contents ranging from 34.47 to 44.70 wt% and molar proportions of Y3+ between 34.0 and 40.6 (Supplementary Appendix S1). The positive correlation between (U + Th) and Si (Fig. 6b) indicates the presence of the thorite (ThSiO4) and coffinite (USiO4) components in xenotime. Thorite inclusions have ThO2 contents ranging from 64.15 to 73.32 wt% and SiO2 contents ranging from 10.45 to 16.86 wt%. The REE contents in thorite ranges 1.75–11.21 wt% and exhibit inverse correlation with ThO2 contents (Supplementary Appendix S1). Utilizing the Y in monazite geothermomenter of Gratz and Heinrich (1997) for monazite co-existing with the xenotime inclusions gives a range of estimated temperatures ranging from 633 to 703 °C and 519 to 584 °C for the BSE dark core and BSE bright mantle, respectively (Table 1).

Apatite, allanite, and epidote

Apatite in the coronas surrounding the monazite have LREE (La-Sm) + Y2O3 contents ranging from 0.32 to 6.44 wt% that exhibit a negative correlation with the P2O5 and CaO. The F content in apatite ranges from 0.98 to 3.95 wt%, corresponding to 0.41 to 1.66 atoms per formula unit (apfu) (Supplementary Appendix S1). The Cl content is mostly below the EMP detection limit. Based on charge balance calculated on the halogen site, the OH content are ranges from 0.27 to 0.59. Apatite is also found as inclusions in the monazite and well as intergrown with monazite (Fig. 4b, d). However it was too small for accurate EMP analysis, without being affected by the surrounding monazite.

Allanite and epidote cations and type were calculated using the WinEpclas program (Yavuz and Yildirim 2018). The mean composition of each epidote type, and the corresponding cation ratios, are listed in Supplementary Appendix S1 and illustrated in Fig. S5 (Supplementary Appendix S4). Based on the REE + Y2O3 content, the epidote can be subdivided into allanite, REE-bearing epidote, epidote, and clinozoisite. According to Ce–Y–Nd classification, the allanite-(Ce) exhibits Ce contents between 6.49 and 11.93 wt%, SiO2 contents between 31.31 and 35.69 wt%, Al2O3 contents between 14.39 and 21.01 wt%, and REE + Y2O3 contents between 14.65 and 25.29 wt% (Supplementary Appendix S1). Allanite in the corona grades into REE-epidote (REE + Y2O3 = 5.91 – 12.55 wt%).

Magnetite, ilmenite, rutile, and titanite

In the metagranite, the monazite breakdown textures are often intimately associated with magnetite alteration microstructures. These magnetite alteration microstructures are concentrated in the chlorite- and quartz-rich domains (Fig. S1a). Six types of magnetite can be identified in the metagranite (Supplementary Appendix S1). These include (1) primary euhedral, magmatic magnetite (Mag1–1), which contains ilmenite exsolution lamellae as well as ilmenite along the grain rim (Fig. 7a); (2) Mag1–2, which has a partly preserved exsolution texture with abundant, elongated, parallel-aligned ilmenite and rutile lamellae (Fig. 7b, c) (3) Mag2, which ranges in size from 200 to 800 μm and commonly exhibits intergrowth with lamellae consisting of rutile–ilmenite symplectite along the {111} plane (Fig. 7d), (4) Subhedral inclusion-free magnetite (Mag3), which is found in titanite associated with retrograde Mag1–2, (5) Ragged magnetite (Mag4), which features etch pits and fine-grained inclusions of barite, galena, and sphalerite (Fig. 7e); and lastly (6) euhedral magnetite (Mag5) grains, which range in size from 50 to 200 μm, and are associated with chloritized muscovite (Fig. 7f). A three-layered corona texture surrounds Mag1–2 consisting first of rutile, followed by ilmenite + rutile, then titanite, and finally, epidote. Surrounding this reaction texture, flaky muscovite and chlorite are also observed (Fig. 7c).

Fig. 7
figure 7

BSE images showing six types of magnetite in the albitized metagranite. a Ilmenite lamellae in the magnetite and ilmenite along the rim of magnetite (Mag1–1). b Magnetite (Mag1–2) rimmed by a rutile–magnetite symplectite and locally retrograded to titanite and Mag3. c Magnetite (Mag1–2) with ilmenite and rutile lamellae rimmed by a rutile and ilmenite symplectite and titanite. d Rutile and ilmenite intergrowths along the {111} planes in magnetite (Mag2), which replace the original ilmenite lamellae. e Ragged magnetite (Mag4) with inclusions of galena and sphalerite. f Euhedral magnetite (Mag5) in chloritized muscovite. Mineral abbreviations: Mag magnetite, Ilm ilmenite, Ttn titanite, Rt rutile, Gth  goethite, Zrn zircon, Py  pyrite, Ttn titanite, Gn galena, Sp sphalerite

Full size image

Chemically, the different magnetite types can be distinguished from each other by their TiO2, V2O3, and Cr2O3 content (Fig. S6a; Supplementary Appendix S1). Magmatic Mag1–1 is characterized by high TiO2 contents ranging from 6.90 to 7.31 wt %, while the partly metasomatically altered Mag1-2 has TiO2 contents ranging from 1.03 to 5.43 wt % (average 1.99 wt %). Both Mag1-1 and Mag1-2 exhibit high V2O3 values, which range from 0.15 to 0.17 and 0.10 to 0.16 wt%, respectively. Mag1–2 and Mag2 also show the highest Cr2O3 contents. Skeletal-shaped Mag2 has TiO2 contents, which range from 0.60 to 1.37 wt% (average 1.14 wt%). Mag3 exhibits lower TiO2 contents from 0.25 and 0.97 wt% (average 0.60 wt%). Mag4 and Mag5 show similar TiO2 contents of below 0.2 wt%. Mag4 also contains NiO. In general, Mag1 always has the highest Ti V, and Cr contents. Mag2 has intermediate Ti and Cr contents, while Mag3 and Mag4 approximate that of pure Fe3O4.

Ilmenite lamellae in magnetite have variable TiO2 contents ranging from 49.84 to 55.35 wt% (XIlm = 0.87–0.98) and MnO contents ranging from 1.15 to 4.14 wt% (XPyr = 0.02–0.08) (Supplementary Appendix S1, Fig. 7).

Titanite rimming ilmenite–rutile symplectites and the magnetite has FeO, F, and Al2O3 contents ranging from 1.11 to 1.71 wt %, 1.04 to 1.21 wt%, and 3.66 to 4.70 wt%, respectively (Supplementary Appendix S1). The Al and Fe content show an inverse correlation with Ti, which is in accordance with the coupled substitution (Al, Fe)3+ + F− ↔ Ti4+ + O2− in the titanite octahedral site (Enami et al. 1993; Fig. 7 and S6b). REE are below the EMP detection limit. The rutile associated with ilmenite and in the symplectite approximate endmember rutile (Supplementary Appendix 1).

Pyrite and goethite

In order to investigate a possible fluid origin for pyrite formation, in situ S isotopic analysis of pyrite in the metagranite were conducted. The pyrites yield δ34SV-CDT values ranging from 13.03 ± 0.08‰ to 13.41 ± 0.07‰ (Supplementary Appendix S2). They show a range similar to those from the sillimanite-bearing schist and feldspathic rock in the Liaohe Group (Hao et al. 2017; Sun et al. 2020; Zhao et al. 2009), which are interpreted as representing a typical metamorphic signature.

Pyrite in the metagranite is rimmed by goethite with widths of 20–300 μm. In some cases the pyrite has been totally replaced by goethite (Fig. 8a; Supplementary Appendix S1). Some of the goethite shows apparent concentric zoning, and monazite grains can be found as inclusions in the goethite (Fig. 8a). Since goethite has a larger molar volume than that of pyrite, its formation results in radial and concentric micro-fractures (Figs. 7e, 8a, b). Occasionally, in intensively chloritized muscovite, thin barite films may surround the goethite and pyrite (Fig. 8b). Barite crystals are only found near goethite rims around pyrite or as inclusions in Mag4 (Fig. 8a, b, c). Goethite rims can be surrounded by a 5–20 μm thick layer of muscovite and epidote (Fig. 8c). In addition, some Mag4 and Mag1–2 appears to be intergrown with pyrite, which is being partially replaced by goethite (Fig. 8c, d). Small inclusions of synchysite-(Y) [Ca(Ce, Y)(CO3)2F] are also found in pyrite (Fig. S7).

Fig. 8
figure 8

Representative BSE images showing micro-textures in altered pyrite. a Concentric zones of goethite with monazite inclusions, along with the adjoining magnetite. b Pyrite surrounded by a goethite corona, which is enclosed by barite. c Pyrite partially replaced by goethite along the fractures and rims. The pyrite is associated with magnetite. d Pyrite, which is associated with magnetite, has been mostly replaced by goethite. Mineral abbreviations: Brt  barite

Full size image

Zircon and monazite U–Pb geochronology

In order to date the crystallization and subsequent metasomatic alteration of the metagranite, U–Pb dating of monazite and zircon in the metagranite, and the neighboring monzonite and kyanite-bearing, garnet–sillimanite micaschist were conducted (Fig. 2b; Supplementary Appendix S3). Corresponding U–Pb concordia diagrams and cathodeluminesence (CL) images are shown in Fig. 9.

Fig. 9
figure 9

206Pb/238U vs. 207Pb/235U diagram and histogram showing monazite and zircon ages from the JLJ orogenic belt. a–b metagranite sample (16KD67-1), Eu/Eu* in b referred as Eu/\(\sqrt{\mathrm{Sm}\times \mathrm{Gd}}\). c–d monzonite sample (16KD67-4). e kyanite-bearing, garnet–sillimanite micaschist (16KD67-2). f Age histogram of monazite from JLJ orogenic belt. Representative CL and BSE images and analysis positions are also marked. Red circles represent the initial crystallized age in the melt and the blue circles represent a metamorphic/metasomatic age

Full size image

Under CL imaging, most zircons in the metagranite show distinct zones with bright cores and gray/dark rims (Fig. 9a). From these, 32 CL bright cores with high Th (71–2339 ppm, average 541 ppm) and U contents (32–2336 ppm, average 650 ppm), and Th/U ratios ranging from 0.48 to 1.47, yield 207Pb/206Pb ages of 2090 ± 33 to 2239 ± 32 Ma with an upper intercept at 2163 ± 17 Ma (Fig. 9a). This age is interpreted to be the crystallization age of the original granite. In contrast, 19 analyses from the CL dark rims exhibited extremely low Th contents ranging from 8 to 158 ppm, and higher U contents (355–2593 ppm, average 1316 ppm) than those of the CL bright cores, which results in low Th/U ratios of 0.01–0.09. The 207Pb/206Pb ages of these CL gray/dark rims are between 1772 and 1880 Ma with an upper intercept at 1849 ± 10 Ma (Fig. 9a).

Monazite grain separates from the metagranite also show distinct compositional zones (Fig. 9b). Thirty-six spot analyses on the BSE dark cores yielded 207Pb/206Pb ages of 1902 ± 22–1829 ± 23 Ma with an upper intercept at 1876 ± 36 Ma (MSWD = 0.62). The BSE bright mantles gave 207Pb/206Pb ages of 1874 ± 24–1798 ± 26 Ma with an intercept age at 1836 ± 14 Ma (MSWD = 1.10) (Fig. 9b). In addition, the BSE bright mantles exhibit distinct negative Eu anomaly than the BSE dark cores (Fig. 9b).

Zircons from the monzonite show fewer effects from metasomatic alteration and recrystallization. Here the metasomatic rims are narrower than those seen for zircons from the metagranite. The melt crystallization age of the monzonite was obtained from the zircon oscillatory zoned cores using U–Pb dating. A total of 28 dates from the 30 analyzed grains yielded 207Pb/206Pb ages from 2251 to 2026 Ma (Fig. 9c), with a weighted mean age of 2166 ± 12 Ma and a Th/U ratio of 0.34–1.15. This age is very close to the presumed crystallization age of the adjoining granite from which the metagranite is derived (2163 ± 17 Ma), which suggests that both the granite and monzonite crystallized at the same time.

The monazites from the monzonite show a bright core and a dark rim in the BSE images (Fig. 9d). Except for six discordant ages, 20 U–Pb dating spots yielded an upper intercept age of 1873 ± 23 Ma (MSWD = 0.88) (Fig. 9d).

Twenty four monazites from the surrounding kyanite-bearing, garnet–sillimanite micaschist give an intercept age of 1869 ± 17 Ma (Fig. 9e). These two sets of dates are nearly identical within the error bars and lie about halfway between the two metamorphic monazite ages for the metagranite suggesting that they may represent a mix of these two ages.

Discussion

Monazite growth and alteration in the metagranite

In the metagranite, the monazite, the Ti–Fe phases, and the plagioclase exhibit features typical of a coupled dissolution–precipitation process. These include a pervasive porosity and extensively developed, fine-grained mineral inclusions in the altered areas of the monazite. These altered areas are delineated from unaltered areas by sharp curved or irregular compositional boundaries (Putnis 2002, 2009; Hetherington and Harlov 2008; Putnis and Austrheim 2010; Harlov et al. 2011; Guillaume et al. 2012; Ruiz-Agudo et al. 2014; Altree-Williams et al. 2015; Grand'Homme et al. 2018).

The complex compositional relationships between different domains in the monazite suggest that monazite growth and fluid alteration occurred simultaneously (Fig. 4). Taking into account the micro-textures, X-ray mapping, geothermometry, and geochronology, we propose that the BSE dark, S-rich monazite cores represent the original monazite the crystallized out with the granite and was later metasomatically altered by a high temperature S-bearing fluid/melt during metamorphism of the granite to a metagranite (Chakhmouradian and Mitchell 1999; Laurent et al. 2016). The BSE bright mantle surrounding the BSE dark magmatic core always exhibits an inconsistent orientation with the BSE dark core, such that it sometimes cuts across BSE dark core and concentric zones (Fig. 4a, c), suggesting that the BSE bright mantle must have formed during a subsequent metasomatic/thermal stage, which could have been induced by anatectic veins and S-type granite crystallization in the JLJ orogenic belt from 1870 to 1840 Ma (Liu et al. 2019a). Distinct Eu depletion in the bright mantles compared to BSE dark cores (Fig. 9b) may indicate these areas formed along with Ca-rich minerals and/or fluid infiltration (Kirkland et al. 2016). Following formation of the BSE bright mantle, fluids along fractures in the monazite reacted to form the BSE gray domains along cracks and the BSE gray monazite grain rims (Fig. 4d). Such textures have been reproduced synthetically in experiments involving monazite in which fluid migration occurred along preferential pathways via cracks in the unaltered monazite interior resulting in the alteration of the monazite along these cracks (Harlov et al. 2007, 2011; Budzyn et al. 2011; Williams et al., 2011; Grand’Homme et al. 2018). Sodium-rich fluid alteration of the monazite can result in a decrease in the Ca, Y, and Dy concentrations and an increase in the Th/U ratio in the altered monazite (Grand’Homme et al. 2018), which is consistent with the chemical variation trend seen between the BSE dark core and the BSE bright mantle (Figs. 5 and 10a, b). The irregular BSE light areas in the BSE dark monazite cores may be related to fluid propagation along cracks in the monazite during another metasomatic stage. This is supported by the distinct high UO2, CaO, and low Nd and Sm contents in these BSE light areas, which plot linearly with the same elements from the BSE dark cores (Fig. 6a, b). When normalized to the average composition of the BSE dark cores (Supplementary Appendix S1; Fig. 10a), a converse LREE variation trend from the concentric zones in the monazite BSE dark core to the BSE bright mantle, BSE gray rim, and, finally, BSE light areas in the core, takes the form of a gradual increase in the Eu2O3, SrO, and CaO contents (Supplementary Appendix S1; Fig. 10a). This variation trend most likely occurred during the albitization of the plagioclase by a Na-rich fluid and the release Ca and elements with an affinity for Ca.

Fig. 10
figure 10

taken from Heinrich et al. (1997). c Al + Mn vs. Ti + V diagram (Nadoll et al. 2014a) for the different magnetite types. d Sulfur isotopic compositions from the metagranite, the evaporite, and the metamorphic rocks in the Liaohe Group

Analysis diagrams of monazite, magnetite, and pyrite, showing the probable evolutionary process of the magnetite. a The average compositional ratios between the BSE gray concentric zone, BSE bright mantle, BSE gray rim, BSE dark cores, and BSE light area in the BSE dark core in the monazite. b Y2O3 vs. Dy2O3 diagram of monazite from the metagranite along with previously reported monazite from apatite–allanite–epidote coronas (Broska et al. 2005; Lo Pò et al. 2016; Ondrejka et al. 2012; Upadhyay and Pruseth 2012). The area for granulite facies monazite is

Full size image

Both experimental results and thermodynamic modeling by phase equilibria suggest that the CaO/Na2O ratio is one of the most crucial parameters controlling whether or not monazite is altered to allanite (Finger et al. 1998; Spear 2010; Budzyń et al. 2011, 2017; Richard et al. 2015). Excess Na in the fluid will prevent the growth of allanite and promote the growth of secondary monazite (Budzyń et al. 2011, 2017). Fluids responsible for the extensive albitization in the plagioclase cores were also responsible for the formation of the high Th and Si and low Ca BSE bright monazite mantles via the huttonitic (Th, U)SiREE-1P-1 substitution (Fig. 6a, b) P5+  + (Y3+ + REE3+) ↔ Si4+  + Th4+/U4+ (Zhu et al. 1999; Förster 2006) (Fig. 6a). Calcium released into the fluid during the albitization of the plagioclase permeated the monazite core along micro-fractures to form the BSE light patchy, high Ca and U domains via the cheralitic substitution 2(Y3+ + REE3+) ↔ 2Ca2+  + Th4+/U4+. This Ca-rich fluid may also have helped to initiate the formation of the apatite inclusions occasionally seen in the BSE dark core (Fig. 4b, d). From the BSE dark core to the BSE gray concentric zones to the BSE bright mantle to the BSE gray outer rim the Y2O3 and Dy2O3 content gradually increases (Figs. 5, 10a), which could also be due in part to an increase in temperature (Heinrich et al. 1997).

In the most likely scenario monazite and zircon crystallized out together at 2160 Ma in the original granitic magma. During the JLJ orogeny at 1960–1800 Ma, the BSE dark core (1902–1870 Ma) was first metasomatized by a S-bearing fluid from the local sulphate-bearing evaporates, such that the monazite geochronometer was reset. During the subsequent isothermal decompression stage (1900–1840 Ma) of the JLJ orogeny, a similar Na-rich fluid originating in the local evaporites induced albitization of the plagioclase, the formation of the monazite BSE bright mantles (1836 ± 14 Ma), and the formation of U and Ca-rich BSE light areas along fractures in the BSE dark cores. Formation of metamorphic rims on the zircon (1849 ± 10 Ma) also occurred at this time. The composition of the altered domains (BSE bright mantles, BSE light areas in the BSE dark cores and pore developed areas) of the monazite shows a distinct compositional decrease of around 1.4–1.5 wt% Y2O3 compared to the original monazite (BSE dark core) (Fig. 10b).

Apatite–allanite–epidote coronas around monazite in the metagranite

Apatite–allanite–epidote coronas around monazite are a characteristic alteration texture produced in response to specific P–T conditions, local mineral and fluid chemistry, and grain boundary permeabilities. Since Finger et al. (1998) first reported this texture in a granitic gneiss from the eastern Alps, similar textures have been successively found in metagranites (Broska et al. 2005; Budzyń et al. 2010, 2011; Ondrejka et al. 2012; Upadhyay and Pruseth 2012), metasedimentary rocks (Majka and Budzyń 2006; Gasser et al. 2012; Lo Pò et al. 2016), and metamorphosed BIF deposits (Xu et al. 2015).

Finger et al. (1998) proposed that this alteration texture formed under amphibolite-facies regional metamorphism and had a relatively slow reaction rate, which was controlled by diffusion, whereas Upadhyay and Pruseth (2012) proposed that the apatite–allanite–epidote corona textures surrounding the monazite were related to high-pressure amphibolite-facies (10–11 kbar and 587–695 °C) fluid-induced retrogression and could be used as a genetic indicator of high-pressure metamorphism. In metasedimentary rocks, these multi-layered coronas associated with monazite can also form under lower P–T conditions (e.g., greenschist-facies) during either the prograde or retrograde stage (Majka and Budzyń 2006; Rasmussen and Muhling 2007, 2009; Gasser et al. 2012; Lo Pò et al. 2016). Broska et al. (2005) concluded that the fluid-induced alteration and breakdown of monazite partly occurred in response to the alteration of anorthite and biotite. They observed that the breakdown of monazite is more dependent on the fluid composition and the ratio of silicate minerals than on the P–T conditions, which seems to be the one basic conclusion based on above observations and which is probably the most relevant here in this study.

In the metagranite, the apatite–allanite–epidote corona textures associated with monazite most likely formed as a result of amphibolite-facies metamorphism during the isothermal decompression stage of the JLJ orogeny (1870–1840 Ma) (Liu et al. 2017a, 2019a, b) in the presence of Ca-rich fluids released during the albitization of the plagioclase (Finger et al. 1998; Broska et al. 2005; Ondrejka et al. 2012).

Alteration of oxides and sulphides in the metagranite

Fe–Ti oxides and sulphides are sensitive to changes in the infiltrating fluid chemistry (Hu et al. 2014; Wen et al. 2017) and are crucial indicators of the redox state (Harlov 1992, 2000; Harlov et al. 1997; Harlov and Hansen 2005; Nadoll et al. 2014a; Guo et al. 2017).

The magnetite Ti + V vs. Al + Mn plot of Nadoll et al. (2014a) reveals that Mag1-2, Mag2, and Mag3 was altered between 300 and 500 °C (Fig. 10c). This temperature range is consistent with previous temperature estimations for symplectitic rutile formation in a retrograde metabasite (Guo et al. 2017). Mag4 and Mag5 formed at temperatures below 300 °C (Fig. 10c). When the Al tetrahedral site (AlIV) chlorite geothermometer (Bourdelle et al. 2013) was applied, it gave late-stage alteration temperatures of 260–310 °C (Supplementary Appendix S1), which coincides with the formation of Mag4 and Mag5.

The sulfide minerals from the surrounding country rocks of the Lieryu formation, associated with the metagranite, consist of pervasive gypsum, barite, ludwigite, and anhydrite. Sulfur isotopes from these minerals exhibit a δ34SV-CDT value of 20.7 to 24.9 ‰ (Supplementary Appendix S3; Hu et al. 2015a; Peng and Palmer, 2002). In the metagranite, the δ34S values for pyrite (11.6–17.33 ‰) (Supplementary Appendix S2) are distinctly higher than for a typical magmatic fluid (0–5 ‰) (Chen et al. 2019; Ding et al. 1992; Duan et al. 2017; Li et al. 2019; Liu et al. 2019c; Yu et al. 2018; Zhang et al. 2020). In the marbles and schists from the Liaohe Group, which were metasomatised by a Triassic magmatic fluid, 98% of the pyrites have δ34S values ranging from 2.8 to 9.1 ‰ with a weighted mean value of 6.8 ‰ (Chi 2002; Duan et al. 2017; Li et al. 2019; Ma et al. 2016; Song et al. 2017). These values are lower than the δ34S value for pyrite from the metagranite in this study (Fig. 10d). This would suggest that the S in the pyrites from the metagranite was mainly derived from the regional metamorphism of the Lieryu Formation with some contribution from the Liaohe Group.

Integrating together all of these observations, along with the U–Pb ages of the zircon CL gray/dark grain rims and monazite BSE bright mantles (1870–1860 Ma), we propose that the extensive evaporites located in the Lieryu Formation of the Liaohe Group were a crucial external source of Na-, Ba-, Cl-, and SO42−-bearing fluids/melts during the high-grade regional metamorphism related to the JLJ orogeny from 1960 to 1900 Ma and subsequent isothermal decompression (1900–1840 Ma) (Peng and Palmer 1995, 2002; Dong et al. 2016, 2017; Hu et al. 2017). These fluids were responsible for the albitization of the original granite to metagranite. The SO42− component in these fluids from the evaporites provided a major oxidizing agent, which allowed for the formation of pyrite from pre-existing magnetite (Mag1 or Mag1-2) in the original granite via the reaction (Li et al. 2014, 2015; Wen et al. 2016):

$${\text{Fe}}_{3} {\text{O}}_{4} + 6{\text{SO}}_{4}^{2 - } \to 3{\text{FeS}}_{2} + 14{\text{O}}_{2} .$$
(2)

Infiltration of fluids from the evaporite would have coincided with the appearance of fine-grained barite, galena, and/or sphalerite in the vicinity of the pyrite or Mag4 (Fig. 8b, c). It is also possible from the textures presented in Fig. 8 that some of the pyrite could later have been partially oxidized back to magnetite due to the high oxygen fugacity (Whitney 1984; Harlov et al., 1997; Harlov and Hansen 2005; Drūppel et al. 2006). Finally, the pyrite and magnetite were both partially converted to goethite under high fO2 and fH2O conditions during greenschist facies P–T conditions at some later stage.

Geochronology and regional implications

Coupled dissolution–reprecipitation reactions in monazite can induce redistribution of radiogenic Pb and gave rise to meaningless individual dates (Harlov et al. 2011; Williams et al. 2011; Weinberg et al. 2020). Therefore, it can be difficult to distinguish the exact formation age of the monazite core and mantle. Especially, during the post-collision exhumation stage of an orogeny when multiple anatexis and metasomatic events are common (Imayama et al. 2012; Liu et al. 2012; Symington et al. 2014; Poujol et al. 2016; Melo et al. 2017). In JLJ orogenic belt, the ca. 1900 Ma regional tectonic–metamorphic event is well-recognized by many types geochronological dating techniques (Li et al. 2016 and references therein). Especially in the LiaoJi granite (magnetite monzogranite), SHRIMP dating of zircon overgrowth rims yielded a 1914 ± 13 Ma metamorphic age (Li and Zhao 2007). In addition, a compilation of all published monazite ages (N = 1121) of post-tectonic magma and decompression from the JLJ orogenic belt shows more than 78% are located in the range between 1870 and 1800 Ma (Fig. 9f) with a major peak at ca. 1860 Ma. Specific to our study area (Fig. 1c), monazite from a granitoid, a garnet amphibolite, and a metapeltic granulite yield an age range between 1920 and 1820 Ma (Liu et al. 2017b, 2019b), which is interpreted to represent post-peak retrogression. In addition, two zircon age peaks at 1870 Ma and 1840 Ma are also recognized and interpreted to represent two episodes of anatexis (Liu et al. 2017a, 2019a).

Based on the monazite micro-textures in Fig. 4c and the Dy2O3 vs Y2O3 diagram in Fig. 10b, we speculate that the BSE dark core and gray concentric zone was metasomatically reset during the initial exhumation JLJ orogenic belt at 1900–1870 Ma (Yin and Nie 1996). The BSE bright mantle and gray rim formed during the second episode of anatexis at  ~ 1840 Ma both rimming and partly replacing the BSE dark core (Fig. 4c).

Summary

The present results, along with previous investigations, allow us conclude that the evolution of textures associated with monazite, magnetite, and sulphides in the albitized metagranite were a combination of metasomatic processes associated with regional magmatic and metamorphic events (Fig. 11).

Fig. 11
figure 11

Schematic illustrations showing the formation, alteration, and breakdown of monazite, magnetite, pyrite, and the related ion exchange reactions in monazite. a Oxy-exsolution of ilmenite from the magnetite and monazite crystallization during cooling of the pre-tectonic granite at ~ 2160 Ma. b Pyrite and synchysite-(Y) crystallized and pre-existing monazite was metasomatically altered under S- and CO2-rich conditions during the early stage of the JLJ orogeny (1960–1900 Ma). This occurred when evaporites from the Lieryu Formation of the Liaohe Group underwent extensive partial alteration, which released Na-, Cl-, F-, and SO4-bearing fluids/melts. c During the decompression of the JLJ orogenic belt (1900–1830 Ma), BSE bright mantles formed on the monazite and BSE light areas were metasomatically induced to form in the BSE dark core along cracks. At the same time or shortly after apatite–allanite–epidote coronas formed around the monazite. Ilmenite–rutile–titanite–epidote coronas formed around magnetite from the ilmenite exsolution lamellae in the magnetite. d During late stage cooling of the JLJ orogenic belt under approximate greenschist-facies conditions goethite partially replaces pyrite and Ca, K, Fe, Al, and Si are precipitated to form allanite and Mag5 along chlorite cleavages and fine-grained muscovite around epidote

Full size image

Ti-bearing magnetite and monazite crystallized out with the original pre-tectonic granite at ~ 2160 Ma (Fig. 11a). The Ti-bearing magnetite then underwent an ilmenite oxy-exsolution process as the granite cooled allowing for ilmenite lamellae to form in the magnetite. During the high-grade (> 700 °C) regional metamorphic event associated with the JLJ orogeny from 1900 to 1870 Ma, evaporites from the Lieryu Formation of the Liaohe Group underwent extensive partial alteration, which released Na-, Cl-, F-, and SO4-bearing fluids/melts. This event albitized the pre-tectonic granite to a metagranite while at the same time inducing pyrite to form from pre-existing magnetite (Fig. 11b). A CO2 component in this fluid induced the formation of synchysite-(Y) inclusions in the pyrite. The original monazite (BSE dark cores and concentric BSE gray oscillatory zone) was metasomatically enriched in S during this time and the geochronometer reset. During exhumation of the JLJ orogenic belt (1870–1830 Ma) (Fig. 11c) BSE bright mantles formed around the monazite and the BSE light areas enriched in Ca and U formed in the BSE dark cores. These, along with the metamorphic rims on zircon, record a fluid-mineral interaction process. During or shortly after this stage, apatite–allanite–epidote coronas formed around monazite and ilmenite–rutile–titanite–epidote coronas formed around magnetite (Mag1) from the exsolved ilmenite lamellae (Fig. 11c). Concurrent with corona formation, some of pyrite was partly oxidized to Mag4. Finally, under later greenschist-facies conditions, a decrease in temperature led to LREE, Ca, K, Fe, Al, and Si being precipitated to form allanite and Mag5 along chlorite cleavages and fine-grained muscovite around epidote, while goethite partly replaced pyrite (Fig. 11d).