REE Geochemical Characteristic of Apatite: Implications for Ore Genesis of the Zhijin Phosphorite

: Phosphorite-type rare earth deposits, which are one of the important types of rare earth elements (REE) ore deposits, have attracted increasing attention because of the extreme enrichments in heavy rare earth elements (HREE), including Yttrium (Y). In this study, in situ geochemical analyses of apatite grains from Zhijin phosphorites were conducted using electron probe microanalysis (EMPA) and laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS). Based on EPMA mapping analysis, we show that rare earth elements and Yttrium (REY) entering into the crystal lattice by isomorphism rather than by inclusions of REY-bearing accessory minerals. The post-Archean Australian Shales (PAAS)-normalized REY patterns of the apatite grains are characterized by hat-shaped MREE-enriched patterns. We interpret that this pattern may reflect the REE distribution of seawater at that time. We propose that in a local, reducing environment, dramatically increased the concentration of REY in seawater, and resulted in the MREE-enriched patterns in the ancient ocean. The main mechanism for the genesis of the Zhijin phosphorite deposit is the apatite crystallizes during the mixing process of REY- and P-rich fluid and oxidizing seawater. the PAAS-normalized REE La, Gd and are are a drastic increase of Pr  Gd (the slight decrease of Gd  (the and  Lu (the


Introduction
China is a principal producer of the majority of the REY (rare earth elements and Y) in the world, up to 95% of the world's demand, and has 136 Mt of REY resources in total, sourced from two types of ore deposits [1]: (1) Bayan Obo carbonatite-type deposits, which provide 80% of the LREE in the world, but are low in HREE and Y; (2) sedimentary phosphorite-type deposits of South China, which dominate the HREE and Y Market (1.3 million tons (Mt) [2]. In the Guizhou Province, South China, abundant sedimentary phosphorite deposits of the Cryogenian system Doushantuo Stage of and the Terreneuvian Series Meishucunian Stage include plentiful REY and P, a new and important type of REY ore deposit [3][4][5]. In the Zhijin area, REY resources are mainly produced from the Meishucunian Stage Gezhongwu Formation. There are 3.5 Mt of REY oxide (REY2O3) resources in phosphate rocks of the Zhijin area, which is highly enriched in Y2O3 (35.33%) [4]. Compared with other rare earth deposits with high Th and U concentrations, the Zhijin phosphorite-type rare earth deposit has a lesser environmental impact on mines [2]. Thus development and utilization have far-reaching significance.
Dating back to the 1960s, three aspects of this type of deposit have been questioned frequently: (1) Is the occurrence of REY due to isomorphic substitution or adsorption? Some studies show that REY enters the lattice of apatite mainly by isomorphic substitution [2,[5][6][7][8]. Phase analysis showed that 97% of the total REY is present in francolite [9,10]. Through dissolution experiments of three phosphorites using dilute acid, Gong [9] found identical dissolution efficiency of P, Y and Ce for the samples, which inferred that REY is isomorphic in apatite. But there is a lack of direct information about REY's occurrence state. (2) What is the enrichment mechanism for REY in phosphorite deposits? Many researchers have shown that the deviation from Australia's post-Archean average shale normalized REE pattern of modern seawater is a result of several processes, including depositional (e.g., hydrothermal, biological) and post-depositional (e.g., diagenetic, weathering) [2,[11][12][13][14][15][16][17][18][19]]. Yet other researchers show the differences in REY patterns are the direct results of redox conditions in ocean chemistry [2,[20][21][22][23][24][25][26][27]. (3) What are the sources of P and REY? Do they occur together? Some experts proved that there are several sources of phosphorous: terrestrial sedimentary [3,28], igneous [3], hot spring biologic (e.g., submarine volcanic, [29]) and oceanic (e.g., upwelling current, [11,12]). As one of the two oldest phosphorus-forming events in the earth's history, the early Cambrian phosphorus-forming event is interpreted to be closely related to the dramatic changes of ocean, atmosphere and biological systems during the period [30,31]. Evidence showed that these changes were closely related to the breakup of the Rodinia supercontinent [11][12][13][32][33][34][35][36], which can provide useful information to explain the source of P and REY.
The focus of this study is to apply novel in situ geochemical analyses of apatite from the phosphorite deposits in the Zhijin region to answer these three outstanding questions. Electron probe microanalysis (EPMA) mapping will provide information on the occurrence and distribution of REY in apatite grains. Laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) analyses allow us to characterize the enrichment mechanisms and the source of the P and REY. In this study, 27 spots of 20 apatite grains were used for EPMA analysis, and these grains were also used for LA-ICP-MS analysis. The PAAS-normalized REY pattern has long been used as a tool to compare sedimentary rocks [37]. Therefore, all the samples were normalized to PAAS, and the results were compared with those from different sources to determine whether the REY redistribute after deposition. Moreover, the samples were also normalized to seawater, which could provide important information for the interpretation of the distinctive REY patterns. Redox conditions may control the distribution of REY in apatite. Ce and Eu are well accepted as a useful tracer for paleo-redox conditions [14,38]. Therefore, we use Ce/Ce*(Ce/Ce* = CeN/(LaN*PrN) 1/2 ), Eu/Eu*(Eu/Eu* = EuN/(SmN*GdN) 1/2 ) [37], Y/Ho parameters to restrict the redox environment during apatite formation. These parameters give vital information for us to understand the enrichment mechanism of REY. The average apatite-seawater (2500 m in-depth) REY partition coefficients reveal the correlation of REY between apatite and seawater.

Geologic Background
The South China Craton lies in the center of Rodinia, between Australia and Laurentian ( Figure  1A, B). Rodinia assembled through global orogenic events between 1300 Ma and 900 Ma, while the breakup of the supercontinent began at ca. 750 Ma and was not completed until ca. 530 Ma [34,35]. Previous studies have shown that many rift basins within the Yangtze Block are directly related to the breakup of Rodinia [11,34,35]. The Yangtze Block is an important part of the South China Craton (Figure 2A). The two most obvious periods of the Yangtze block drifting out from the breakup of Rodinia are ca. 600-550 Ma and 540 Ma [11]. Many experts [11,39,40] have shown that the breakup of the Rodinia supercontinent affected the carbon dioxide cycle in the earth's atmosphere and hydrosphere and changed the global climate at that time. Two glacial processes (750-700 Ma and 610-575 Ma) controlled the prospering and decaying of the biosphere and also affected the deposition of carbonate, iron, manganese and phosphorus on the surface of the lithosphere. When the glacial age ended, and the temperature rose, two important biological explosion events occurred (Doushantuo age and Meishucunian age) [11,35].  The Zhijin phosphorite-type REY deposit in Guizhou Province (26°40′24′′ N and 105°51′30′′ E) is located 14 km east of Zhijin County. The width of the ore body is 0.4-4.0 km, and it has an area of 32 km 2 , extending 20 km along the NE-SW direction [12]. The occurrence of the REY deposits in the phosphorous-bearing rock series was between the erosion surface of paleo-karst formed by black shale of the Terreneuvian Serious Stage2 Niutitang Formation and dolomite of the Ediacaran system Dengying Formation took place at the bottom of Meishucunian Stage [41]. Structural development mainly consists of the NE-SW trending Xinhua anticline and a series of normal faults, with locally developed small structures [11,14,42]. These faults are at the southwestern end of the 'central Guizhou uplift' and belong to the western margin of the Yangtze Block ( Figure 2A).
From SW-NE, the Zhijin phosphorite-type REY deposit can be divided into the Gaoshan, Xinhua, Gezhongwu, Jiakua, Guohua and Daga ore blocks ( Figure 2B). The distribution of phosphorus and REY are mainly along with the core of Guohua-Gezhongwu anticline and Damachang anticline and are strictly controlled by the strata. From bottom to top, the strata are as follows ( Figure 2C): (1) Upper Dengying Formation, which exposes super-thick fine-grained dolomite; (2) Gezhongwu Formation mainly exposes interbedded phosphorus dolomite and dolomitic phosphorite with thick and thin beds ranging from millimeters to tens of centimeters, which they are the main ore-bearing beds; (3) Lower Niutitang Formation mainly develops quartz sandstone.
The intraclastic apatite is mainly elliptical, with irregular quadrilateral shapes. Particle sizes range from 30~400 μm ( Figure 3A-C). They are interpreted as aggregates of microcrystalline carbonate fluorapatite (CFA, shown in the XRD analysis Supplementary Data S1). Hand specimens show that the structural feature is banded with different thickness ( Figure 3D). The light-colored band is phosphorous dolomite, and the dark-colored band is of dolomitic phosphorite ( Figure 3D). Dolomite and quartz are the main gangue minerals, which exist as cement around the ores. Dolomitic phosphorite and phosphorus dolomite are mainly clastic structured, described in detail by previous researchers [5,42]. The rocks have been interpreted to have formed in the sedimentary water, consisting of internal clastic rather than terrigenous clastic particles [42].

Samples and Methodology
In this study, 10 samples of the Motianchong ore block (MTC-1-MTC-10), mainly consisting of phosphorus-bearing dolomite and dolomitic phosphorite, were collected from the Gezhongwu Formation. There are no obvious morphology differences, so we randomly selected two samples (MTC-7, MTC-9) for single mineral separation and subsequent microanalysis pretreatment. Apatite separation was completed by the Langfang Tuoxuan Rock and Mineral Testing Service Co., Ltd., which adopts conventional gravity separation and manual selection under a microscope. From the separated apatite grains, 20 grains were selected for EPMA analysis (M1-1-M20-1) and LA-ICP-MS (M1-M20), and 3 of these grains were selected for EPMA mapping analysis (M9, M11, M20).
The general chemical formula of apatite is Ca5 (PO4) 3 (F, Cl, OH) [43], where Z = F, Cl, OH. Ca 2+ can exist in 9-, 7-or 6-fold coordination, and can be substituted by Sr 2+ , Pb 2+ , Ba 2+ , Mg 2+ , Mn 2+ , Fe 2+ , REE 3+ , Na + , K + and other cations [44]. (PO4) 3+ can be replaced by (SiO4) 4− , (SO4) 2-, (CO3) 2− and other complex anions [45]. We calculated concentrations of OH and CO2 assuming the ideal formula Ca10P6O24Z2, where the cation coefficient at Ca, P and F position are denoted by X, Y, Z. The content of CO2 and OH groups need to be calculated as they are undetectable by the EPMA analytical method. We use the stoichiometric ratio calculation method to calculate the mineral formula with X = 10. The mineral composition contains F and Cl, and these anions replace oxygen, so the correction should be done on the oxygen by deducting the content of oxygen replaced by F and Cl from the total amount. The deduction formula is: O = 0.23 × Cl and 0.42 × F.
Analysis of major elements in mineral micro-regions was completed in the EPMA Laboratory of Geoscience College of Zhejiang University using a SHIMADZU 1720 H microprobe. The data acquisition time for each element is 30 s with a beam spot diameter of 5 μm, accelerating voltage of 15 kV, and beam current of 20 nA (For EPMA mapping analysis, 20 kV, 100 nA, and10-30 ms/point). The ZAF method was used for correction. The analytical accuracy is 1-2%, and the detection limit is hundreds of ppm. The spectroscopic crystals used are as follows: LSA55(F), RAP (Na, Mg, Al), PET (P, Si, Cl), ADP (S), LIF (Ti, Fe, Ca, Mn). The mineral standard samples used are plagioclase (Na), spodumene (Al), willemite (Si), pyrope (Mg), apatite (Ca, P, F, S), benitoite (Ti), orthoclase (K), almandine (Fe), tugtupite (Cl), rhodonite (Mn).
Some volatile elements such as F and Cl in apatite migrate easily with 20 nA beam exposure, so the F and Cl contents not accurate. Researchers analyzed the random crystallographic orientations of the apatite, which imply there could have been an increase in F X-ray counts of up to ~30-40% near parallel to the c-axis, while near perpendicular to the c-axis could be quite accurate [46][47][48], the same as Cl data. We list the F and Cl data in Table 1 (marked F a and Cl a ) to remind the readers that these measured values, while clearly demonstrate the basic compositional variability and can provide basic calculations for minerals structural formulas, have large uncertainties. Table 1. Electron probe microanalysis (EPMA) analytical results for apatite grains (wt.%). LA-ICP-MS analyses (spot M1-M20) were carried out at the State Key Laboratory of Ore Deposit Geochemistry at the Institute of Geochemistry Chinese Academy of Sciences (Guiyang). The instrument is a GeoLasPro laser ablation system (193 nm, premixed gas, IIIb safety level, 95% confidence level). The laser spot switch is a manual version. The laser ablation system is an Agilent 7900 inductively coupled plasma mass spectrometer. Helium was used as carrier gas, and helium and argon are mixed into the ICP-MS. Spot ablation was carried out using a 44 μm spot size at 5 J/cm 2 and 10 Hz with an 18 s baseline and 50 s of ablation. Using NIST610, NIST612, NIST614 and Durago apatite as standard samples, apatite trace element compositions were quantified by using the multiexternal standard-internal standard method. The detection limit of REY varies from 0.03-0.4 ppm; the specific values were listed in Liu et al. (2010) [49]. We were careful to avoid analyzing mineral inclusions in apatite. The recommended standard values are from the GeoReM database (http://georem.mpch-main z. gwdg. de/). The off-line processing of analytical data was accomplished using ICPMSDataCal [50,51]

Results
Three grains of EPMA mapping analysis ( Figure 4)   The elemental composition based on the EPMA analysis (

Occurrence of REY in Phosphorite
Previous energy-dispersive X-ray spectroscopy (EDS), phase analysis, and experimental dissolution studies have inferred that REY enters the lattice of apatite mainly by isomorphic substitution [9,10,53]. Our LA-ICP-MS signals are constant, which precludes the existence of REYbearing mineral inclusions that should be characterized by bumps in the spectra. Here, a direct distribution map of REY is useful to characterize the occurrence of REY in apatite. From the EPMA mapping analysis, a uniform distribution of the elements P and Ca can be clearly seen in Figure 4B-C. The EPMA mapping analysis of La, Ce, Pr, Nd, Gd, Yb and Y also are displayed in Figure 4. In these graphs, each picture's element has its color scale, so element contents cannot be horizontally compared by color in different pictures. In contrast, in an independent picture, different colors can represent the content of an element. Compared with REY, P and Ca have high counts because they are the main components of the apatite ( Figure 4B,C). Epoxy resins (the surrounding of CFA grains) can be ignored at the high counts of P and Ca data, while at the low counts of REY data, epoxy resins have obvious background values. It is a pity that we did not remove the background colors of epoxy resins in REY data, impacting the results. However, we can still recognize the distribution of elements in apatite from colors except for background colors. For example, in Figure 4H, we can see that the distribution of Yb, the green and yellow can represent the relative content of Yb except for the blue and black background. If Yb occurs as mineral inclusions, there will be an agglomerate distribution with colors of yellow or green. The same as the other REY, the results showed that Yb is evenly distributed in the mineral, which indicates REY are associated with isomorphism rather than inclusions of the independent minerals in CFA.

Impacts of Depositional or Post-Depositional Processes on REY Patterns
Suppose REY patterns are owed to depositional or post-depositional processes. There are various variables affecting REY patterns during these processes: (1) Diagenetic inputs-Previous studies [2,5,18,54] shown the original REY characteristics of seawater can be well preserved by finegrained apatite, and the REY compositions are not easily altered during later burial and diagenesis. Moreover, the influence of burial and diagenesis can be detected by the negative correlation between Ce/Ce* and DyN/SmN [14], while the samples in this study do not show this correlation ( Figure 7A).
(2) Hannigan and Sholkovitz's [17] experiment inferred preferential removal of MREE by the surface water, and in the long-term chemical differentiation process, it will inevitably cause the loss of MREE rather than enrichment. Obviously, aeolian inputs also do not interpret the unique REY patterns in this study. (3) Biological inputs-Previous studies on REY in biological fossils had found that most of the REY patterns in the fossils did not show HREE depletion ( Figure 6) [52]. (4) Hydrothermal inputs: In general, 100% of midocean hydrothermal REE fluxes [2,23] are quickly adsorbed to settling oxyhydroxide particles, which show strong positive Eu anomalies. Our Eu/Eu* data obviously deviate from the above values (Figures 6 and 7D), which indicates REY in apatite has no direct relation with hydrothermal inputs.

The source of REY
Excluding all the above processes, we prefer to favor the deviation of REY patterns due to longterm marine chemistry variations. In general, the PAAS-normalized REY pattern has long been used as a reliable tool for sedimentary rocks [51,55]. However, Takahashi et al. [56] suggested that for deepsea sediments, the PAAS normalization was not suitable because different geochemical processes of submarine materials limited REY. Hence, they adopted a normalized model of deep-sea water [56]. In this study, we use both PAAS (Figure 8) and the deep-sea water model (Figure 9) [52] as standard models to normalize the REY patterns of apatite. We compared PAAS-normalized patterns with seawater-normalized patterns (2500 m) adopted for the REY in apatite. When normalized to PAAS through comparative analyses, the samples show significant negative Ce anomalies, slight enrichments of LREE, obvious enrichment of MREE and depletions of HREE. While normalized to deep-sea water, the CFA grains display significant positive Ce anomalies, obvious enrichments of LREE, slight enrichments of MREE and clear depletion of HREE. Our REY patterns are similar to Emsbo's data [2] of phosphorites (or apatite) from different ages and regions ( Figure 8). Furthermore, Emsbo's [2] data are also normalized to seawater (Figure 9) to compare with our patterns. The results showed that the overall curve is smooth from left to right, except Ce, which indicates the single source of REY.  Moreover, previous studies imply the mineral/seawater REY partition coefficients and the logarithm of ionic radius display a simple parabolic relation if REY is a single source [57][58][59]. According to our result, most data fall on the parabola (Figure 10) (as the Ce value deviates from other elements in Figure 9, it was removed from the scatter diagram). Thus, it also suggests that the REY in apatite is from seawater rather than other contributors.

Marine Redox Conditions
Evidence suggests that REY was derived from seawater. Modern seawater is considered to be an oxidizing environment. Hence, it is very important to determine the redox environment when REY enters into apatite. In general, Ce, Eu anomalies and Y/Ho ratio are considered to restrict the redox environment of paleo-seawater [51,60,61].
1. Ce anomaly: The Ce anomaly of phosphate minerals can represent the paleo-seawater conditions at the time of its formation, only when LaN/SmN > 0.35, and the LaN/SmN is not related to the Ce/Ce* [62,63]. In this study, the LaN/SmN ratio ranges from 1.18 to 1.47 (significantly greater than 0.35), and there is no correlation between the LaN/SmN and Ce/Ce* ( Figure 7B). Hence, the Ce anomaly can be used in determining the paleo-redox environment. When the Ce/Ce* < 0.95, it indicates the oxidizing environment in seawater. On the contrary, the Ce/Ce* > 1.05 interprets the reducing environment. The Ce/Ce* value (0.30~0.32) of Zhijin apatite, much less than 1, indicates the deposition of apatite in an oxidizing marine environment.
2. Y/Ho: In an oxidizing environment, Fe-Mn hydroxides from the seafloor are more likely to adsorb Ho than Y, which makes the adsorption rate for Ho being close to twice that of Y, leading to higher Y/Ho ratios (44~83) in oxidizing seawater [64,65]. However, Ho will be released from Fe-Mn hydroxides in the reducing environment, resulting in the decrease of Y/Ho ratio (reduced to 30) [15,64,65]. The inputs of terrigenous clastic materials (Y/Ho: [23][24][25][26][27] can also lead to low Y/Ho values [64]. However, the Ce/Ce* of terrigenous clastic materials can come up to 0.9. Meanwhile, there are no terrigenous clastic particles under the petrographic observation. Our data show a high Y/Ho value (ranges from 50.67~54.37, average 52.81), which suggests an oxidizing environment during the formation process of apatite.
3. Eu Anomaly: Under a strongly reducing medium, Eu 3+ will be reduced to Eu 2+ . The redox potential of Eu 3+ /Eu 2+ decreased significantly at a higher temperature. When the temperature was higher than 200 °C, Eu 2+ became relatively stable. Therefore, Eu 2+ is found to be enriched in highly reducing hydrothermal fluids [66]. As the same to Ce/Ce*, when the Eu/Eu* < 0.95, it indicates the oxidizing environment in seawater. In seawater, however, Eu is not significantly fractionated during the precipitation-dissolution processes (it exists solely in the trivalent state) [14], leading to no obvious Eu anomaly. Therefore, Eu/Eu* anomaly cannot be used to identify the redox environment in seawater but can determine whether REY comes from seawater.
From the cross-plot of Ce/Ce* vs. Y/Ho and Ce/Ce* vs. Eu/Eu* ( Figure 7C,D), all the data in this study fall into the relative oxidizing area, indicate that apatite in the Zhijin phosphorite deposit was formed in an oxidizing environment, which is similar to modern seawater.

Mechanisms of REY Patterns
During the Early Cambrian, the surface seawater's oxidation degree and even the whole shallow water area was high, while the deep bottom seawater was anoxic environment [11,12]. Researchers revealed phosphorites in the Yangtze block mainly deposited at a shallow shelf environment, while no phosphorus deposits were found in the deep-sea environment [11,12]. These distributions of deposits indicate that the redox environment restricted the formation of phosphorite. In general, HREE exists by clathrate, and Fe-Mn hydroxides preferentially adsorb LREE and MREE, resulting in the unique REY pattern at oxidizing seawater [17,20,21]. The dissolution of adsorbing hydroxides in deep water can locally increase LREE and MREE concentrations by as much as ten times at a reducing environment [22,23]. When these large numbers of LREE and MREE arrived at the upper oxidizing water, due to the change of redox environment, partial Ce and Eu precipitate, leading to the negative Ce and Eu anomalies, resulting in the distinctive REY patterns.
The crystallization of apatite needs an oxidizing environment, while the release of LREE and MREE requires a reducing environment. Thus, these LREE and MREE need a medium to reach upper oxidizing water. Previous studies have revealed the relationship between the formation of phosphorite and the breakup of the Rodinia supercontinent [11,12]. Sm-Nd and Rb-Sr isochron ages of the Zhijin phosphorites are 542 ± 23 Ma and 541 ± 12 Ma, respectively [11]. The ore-forming ages of apatite coincide with the Yangtze block splitting and drifting from the Rodinia supercontinent, implying the formation of phosphorite deposits was closely related to the breakup events [11,12]. Previous studies [24][25][26][27]67] showed that the volcanic activities in the transitional period of Ediacaran-Cambrian might play significant roles in global geological events, and these activities may be responsible for triggering or strengthening the deep-sea reduction degree. Furthermore, these activities accelerated the release of LREE and MREE from Fe-Mn hydroxides, heated the deep seawater and formed ocean currents and columns. Carrying the large numbers of LREE and MREE, the currents and columns upwelled to upper oxidizing water, leading to distinctive REY patterns of CFA.

The Source of P and the Diagenesis Mechanism
Nd isotopic model ages (tDM) can be used to trace the sediment source, and the variation of Nd model age (εNd(t)) is a reliable tracer of mantle-derived materials [68]. The tDM of the Zhijin phosphorite deposit (1453-1477 Ma) is obviously lower than that of around seawater in Cambrian, and the εNd(t) (−2.37~−2.65) is higher than that of seawater during the Cambrian period, which suggests that the initial material for large-scale phosphorous mineralization may be mainly derived from the mantle, and also indicates the interaction between crust and mantle is strong during the breakup of Rodinia. Liu and Shi [11,12] inferred that the breakup of the Rodinia supercontinent led to the upwelling of the hydrothermal fluids, which carried a large number of P-rich materials from the mantle. When the P-rich fluid overflowed into deep-sea water, due to the crystallization of apatite requires an oxidizing environment, these P-rich bearing and REY would not precipitate but floated and upwelled with currents to shallow oxidizing seawater. Other researchers suggested Fe-Mn hydroxides could efficiently adsorb not only REY but also P and then sank into the seafloor sediments, which is the main mechanism of dissolved phosphorus and REY entering the seafloor sediments [69,70]. No matter P-rich materials were from the mantle or the release of Fe-Mn hydroxides, the upwelling P-rich bearings provided abundant materials for the crystallization of apatite. Therefore, the P and released LREE and MREE upwelled to upper shallow water with the ocean currents and columns. Due to the change of redox environment, these P-rich and REY materials crystalized to apatite, resulting in the distinctive REY patterns in the Zhijin phosphorite deposit.

Conclusions
1. Based on the EPMA mapping analysis, it suggested that the REY occurring in the Zhijin phosphorite originate from isomorphism rather than mineral inclusions in the CFA; 2. There is a simple parabolic relationship between the mineral/seawater REE partition coefficients and the ionic radii, so to a certain extent, it implies that the REY in apatite is originated from seawater rather than other sources; 3. In a local reducing environment, the released REY from hydroxides dramatically increases the concentration of REY in the ocean seawater; 4. The breakup of the supercontinent leads to the upwelling of the hydrothermal fluids, which was carrying abundant P-rich materials. These upwelling processes will enhance or lead to the reducing environment and dissolution of considerable LREEs and MREEs. When the P-rich fluid and REY reach the upper oxidizing water with the upwelling currents to upper oxidizing shallow water, REY-rich apatite formed during this mixing process.

Conflicts of Interest:
The authors declare no conflicts of interest.