The Enrichment Mechanism of REE at Sulfate and Methane Transition Zone (SMTZ) of the Northern Part of the South China Sea

Southern Marine Science and Engineering Guangdong Laboratory (Guangzhou), Guangzhou 511458, China MLR Key Laboratory of Marine Mineral Resources, Guangzhou Marine Geological Survey, Guangzhou 510075, China Key Laboratory of Gas Hydrate, Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences, Guangzhou 510640, China Academy of South China Sea Geological Science, China Geological Survey, Sanya 572025, China


Introduction
Anaerobic oxidation of methane (AOM) is a very important biogeochemical process in marine sediments. It is considered to be of significant impact on climate change, atmospheric evolution, and related element cycle [1][2][3][4][5]. During AOM, the residual porewater sulfate was consumed at the sulfate-methane transition zone (SMTZ). Here, an upward flux of methane reacts with a descending flux of sulfate [6][7][8][9][10], which is mediated by the metabolisms of methaneoxidizing archaea (MOA) and sulfate-reducing bacteria (SR) [3,[11][12][13][14][15][16]. During the AOM, the methane-driven authigenic carbonates (MDAC), pyrite, and other minerals are enriched at the sulfate and methane transition zone (SMTZ) [11,[17][18][19][20]. Of course, SMTZ was not static and moved with the accumulation of sediments and the continuous supply of methane from below (Xie et al., 2021). Therefore, the sediment profile appeared multistage carbonate, and pyrite increases with the depth. The anomalies of the content of pyrite can help identify the paleo-SMTZ. Previous researches have shown that metal oxides and hydroxides could be served as the electron acceptor during methane oxidation [15,[21][22][23][24]. In addition, seafloor observation at seeps revealed that chemical and physical parameters, such as fluid flux, temperature, thermal diffusivity, pH, and alkalinity, are also changed at the SMTZ [25][26][27]. Numerous studies have also demonstrated the AOM plays an important role in the cycle of P, S, Fe, Ca, Mg, B, and Cl [28][29][30][31]. However, up to now, there is little research about the influence of AOM on REEs.
Kato et al. [32] suggested that the deep-sea sediments would be an important source of REEs in the future [32]. Therefore, it is particularly important to understand the REE enrichment mechanism and the influencing factors in marine sediments. At present, the REE patterns are often used in indicating the source, reconstructing the redox environment, and investigating the impact of climate change [33][34][35][36]. While carbonates formed in the oxygenated bottom water usually show a smooth seawater REE profile characterized by the enrichment of heavy REEs (HREEs) and positive La and negative Ce anomalies [37]. And Pol et al. [38] suggested that four lighter REEs (LREE S ) supported the growth of acidophilic methanotroph in volcanic mud pots, however, the higher atomic weight REEs (HREEs) were less conducive to growth [38]. These conclusions confirmed that the methane-rich marine sediments influence the REE distribution. Keltjens et al. also showed that light REE (LREEs) are an important constitute of methanotrophs [39,40]. Shiller et al. [40] found the depletion of LREEs at methane-rich sediments of Gulf of Mexico [40]. Bayon et al. [41] observed Nd in the methane enriched plums above the cold seeps on the Niger Delta margin [41]. Otosaka and Noriki [42] discussed the compositional variations among REEs in settling particles to elucidate the sources of the REEs ( [42]; Xiong et al., 2012; [10]). Although the REE enrichment mechanism has been studied in some sea areas by predecessors, the REE enrichment mechanism in the methane seeps has been rarely studied. In general, the Fe-Mn oxyhydroxides are sensitive to the early diagenetic remobilization in marine sediments [15,16,36,43], which can be used in understanding the REEs influenced by the early diagenetic processes. In this paper, we want to answer the question of the mechanism of REE enrichments at the methane seep sediments and reveal the relationship between the AOM and REE cycle.

Geological Settings
The South China Sea is the largest marginal sea in the Western Pacific Ocean, formed under the convergent plate boundary of the Eurasian plate, the India Australia plate, and the Pacific plate [44][45][46]. Large and medium-sized Cenozoic sedimentary basins were developed on the northern slope of the South China Sea. Since the late Pliocene, a large amount of terrigenous organic matter has been transported and deposited on the northern slope of the South China Sea, which results in rapid sediment deposition and provides favorable conditions for the formation of gas sources [10,47]. The studied sites SH1 and SH3 are distributed in the Shenhu area ( Figure 1). The Shenhu area is located in the zhu II depression area of the Pearl River Mouth Basin on the northern slope of the South China Sea. Influenced by the formation and evolution of the Cenozoic sedimentary basin in the Pearl River Estuary, a wide range of submarine canyons, central diapir belts, and various types of diapir structures were developed in the Shenhu area [48], which provide good channels for the migration of deep source fluids [49]. The deposition rate in the Shenhu area increased and the content of organic matter was high during the Neocene. Guangzhou Marine Geological Survey carried out five gas hydrate drilling expeditions in the Shenhu area [50]. Geochemical analysis of the sediments and pore water at several stations in the Shenhu area also indicated that methane enriched fluid is active in this area [51,52]. Among them, SH3 is of particular interest because of its high gas hydrate saturation and thick hydrate-bearing sediments. The sediments at sites SH1 and SH3 are mainly silty clay. In this study, 31 samples were obtained from site SH3 at 0-25 mbsf, and 11 samples were collected from site SH1.

Major and Rare Earth Element Analysis of Bulk
Sediments. In all samples, the major elements were determined at the Analytical and Testing Center of the Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences. A Thermo ARL ADVANTta IntelliPower TM 2000 X-ray diffraction spectrometry (XRF) instrument was used to determine the contents of the major elements (Fe, Mn, and Ti) in sediments. Measured XRF spectral data were converted to elemental and oxide contents by UniQuant semiquantitative analysis software.
The contents of rare earth elements in the whole rock were analyzed by inductively coupled plasma mass spectrometry (ICP-MS) on an Agilent 7700e instrument at the Wuhan Shangpu Analysis Technology Co., Ltd. And rare earth element analysis was carried out by the following procedure: (1) The 74 μm sample was placed in an oven at 105°C for 12 h; (2) 50 mg of the powder sample was weighed in a Teflon bomb; (3) 1 mL of high-purity nitric acid followed by 1 mL high-purity hydrofluoric acid was added to the sample; (4) the Teflon sample bomb was put into the steel sleeve, tightened, and placed in an oven at 190°C for more than 24 h; (5) the sample bomb was cooled and place on a 140°C hotplate after opening the lid, the mixture was evaporated, and 1 mL of HNO 3 was added prior to conducting another evaporative step (to a liquid free state); (6) 1 mL of high-purity nitric acid, 1 mL of Milli-Q water, and 1 mL of the internal standard (concentration, 1 ppm) were then added, and the Teflon bomb material was dissolved again, after which the bomb was placed into a steel jacket, tightened, and placed in an oven at 190°C for more than 12 h; (7) Table 2. At site SH3, the concentration of Fe (oxl) and Mn (oxl) decreased at SMTZ (Figure 2). The concentration of Mn (carb) increased and the concentration of Fe (carb) slight increase at the SMTZ ( Table 2). What is more, the content of REE has a positive correlation with the change of Fe-Mn oxides (hydroxides) and Fe-Mn carbonate phase.  Figure 1: The location of the study site SH1 and site SH3. In this paper, the upper continental crust-(UCC-) normalized REE patterns at sites SH1 and SH3 are characterized by slight light REE (LREE (La, Ce, Pr, and Nd)) and heavy REE (HREE (Ho, Er, Tm, Yb, and Lu)), and the enrichment of middle REE (MREE (Sm, Eu, Gd, Tb, and Dy)) ( Figure 3). This MREE-bulge type pattern is a typical feature of anoxic pore water in the marine sediments [61], and the anoxic   [64,65]. In the studied area, all the samples at sites SH1 and SH3 showed weak negative Ce anomalies, indicating a reduced environment (Figure 3). And this conclusion is consistent with the previous result. Furthermore, sites SH1 and SH3 show lower content of HREE and LREE compared with those of the SW Taiwan, Luzou Rivers, and Parel Rivers. The result suggests that the REE receives less influence of the input of Pearl River, SW Taiwan, and Luzhon Rivers (Figure 4(b)). Ce 3+ is oxidized to Ce 4+ under oxidizing conditions, resulting in a negative Ce anomaly, which therefore is an effective indicator of oxic conditions, and negative Ce anomalies in marine carbonates have been shown to reflect seawater oxygenation. At site SH1 and SH3, the δCe and δEu were obviously higher than the δCe and δEu of the Parel rivers, SW Taiwan, and Luzhon rivers (Figure 4(a)). Indicating an anoxic and methane seep sediment environment at site SH1 and SH3, such a sediment environment is very favorable for AOM. The conclusion is consistent with the previous study. Therefore, the contents of REE at sites SH1 and SH3 are mainly influenced by the authigenic minerals and later diagenesis induced by the AOM. Combined with the reduced environment, we believe that the REEs are mainly influenced by the authigenic minerals and diagenetic alteration rather than detrital minerals that usually contain much higher REE S [61]. However, the northern part of The South China Sea is a high incidence area of detrital input; to eliminate the impact of detrital     Figure 5). This characteristic indicates that the abrupt increase at 25 mbsf of site SH1 and 17 mbsf of site SH3 is not influenced by the input of terrigenous material. It is more likely to be influenced by the later diagenesis or effects of authigenic minerals.

The
Influence of AOM on the Enrichment of REEs. REE content is usually increased with the sediment depth [66,67]. However, the content of REEs does not increase with the depth at sites SH1 and SH3. As mentioned above, the∑REE/Ti increases abruptly at sites SH1 and SH3 at 25 mbsf and 17 mbsf, respectively ( Figure 5). However, the sulfate and methane transition zone (SMTZ) is at 27-36 msbf of site SH1 and the SMTZ at 18-25 mbsf of site SH3 [15,16,24]. The results show that the ∑REE/Ti increased above the SMTZ. As the previous discussion, the REE is mainly influenced by the later diagenetic and authigenic minerals. Theoretically, REE should be enriched in SMTZ instead of above of SMTZ. Hence, the main question is to explain the enrichment of REE at the top of SMTZ. Akagi et al. [53] found that the content of REE contributed from the carbonate with a higher REE concentration at the North Pacific Ocean and Bering Sea [53]. Liao et al. [68] considered that biogenetic apatite is one of the most important REY carriers in REY-rich deep-sea sediment [68]. Kato et al. [32] also suggested that the main host of the REE is an Fe-hydroxide precipitate from hydrothermal plumes that have taken up rare earth elements from ambient seawater. No matter how the REE forms, as a major process of early diagenesis, AOM often results in the enrichment of authigenic minerals in marine sediments. Meanwhile, AOM also influences the metal oxides and the apatite [15], indicating that it is likely to affect the distribution of REE. To further investigate the relationship between AOM and the content of REE, the Fe-Mn oxides (hydroxides) and Fe-Mn carbonate phase at site SH3 were analyzed. At site SH3, there is no correlation between the content of REE and the content of Fe-Mn oxides (hydroxides) at 0-6 mbsf. However, at the paleo-SMTZ (8-15 mbsf) and current-SMTZ, it showed an obviously positive correlation (Figure 6). At the same time, the Fe-Mn carbonate phase also showed a positive correlation with the content of REE ( Figure 6). This characteristic shows that the REEcarbonate phase and REE-metal oxides (hydroxides) have  (Figure 7). The result showed the HREE has a positive correlation with the authigenic mineral (Figure 7). Therefore, it is proved that the REE-rich Fe-Mn carbonate phase and REE-rich Fe-Mn oxides (hydroxides) affect the HREE. During the AOM, HSand HCO 3 are generated, and the HCO 3 will react with the Ca 2+ to form the carbonate. Therefore, it is often found the vast carbonate at methane seep areas. Meanwhile, HSis also produced by the AOM, which reductively dissolves iron oxides (FeOx), leading to the precipitation of Fe sulfides (e.g., greigite, mackinawite, and pyrite) (Riedinger et al., 2014). During the reductive of Fe ox dissolution, the Fe ox -associated REE is released to the ambient pore water and accumulates in different forms, such as the REE-rich carbonate phase.

Quantitative Calculation of Enrichment of REE.
During the AOM, the Fe-Mn oxides (hydroxides) were reduced by the HS-and the content of authigenic minerals increased at SMTZ. The specific reaction is as follow: According to formulas (4) and (5), it can be seen that the content of rare earth elements in the sediment will decrease by 50 ppm when 1 mol carbonate is produced and 4-mole Fe-Mn oxide reduction. The above discussion has confirmed that REE is primarily affected by the carbonate phase and Fe-Mn oxides (hydroxides). Therefore, the content of REE change should follow this quantitative relationship at SMTZ. At site SH1, the content of REE at 20 mbsf is 162.35 ppm, and the content of REE at 37.98 mbsf is about 113 ppm, the difference is about 49 ppm. The same quantitative relationship can also be found in site SH3. Therefore, the conclusion is consistent with the above.

The Enrichment Mechanism of REE Influenced by AOM.
Previous studies have demonstrated that sulfate-driven anaerobic oxidation of methane (SD-AOM) can cause the enrichment of carbonate. At the same time, most of the metal oxides will convert to metal sulfide in the presence of excess hydrogen sulfide [2,3,70]. With the deepening of research, metal-driven anaerobic oxidation of methane (MD-AOM) is found at the bottom of the SMTZ worldwide [15,24,38,71]. During the metal-driven anaerobic oxidation of methane  Figure 7: The correlation of HREE and different forms metal oxides or carbonate presented at paleo-SMTZ sediments or modern-SMTZ sediments.

Geofluids
(MD-AOM), more metal oxides are converted to metal sulfides compared to the SD-AOM. The previous discussion suggested that the REE-rich carbonate phase and REE-rich Fe-Mn oxides (hydroxides) have an important contribution to the REE. The question is that why the REEs are not enriched at the SMTZ while enriched on top of the SMTZ. Based on our results, it is inferred that the metal oxides (hydroxides) are reduced at the SMTZ; meanwhile, the REE-rich Fe-Mn oxides (hydroxides) will release REE in the pore water during the AOM. The released REE will reconvert to the REE-rich carbonate phase. At the SMTZ, the REE-rich Fe-Mn carbonate phase increased and the REE-rich Fe-Mn oxides (hydroxides) decreased. Therefore, the REE is not enriched in the SMTZ while enriched at the top of the SMTZ (Figure 8). And this study revealed the influence of AOM on REE at methane seeps areas. It can help us to better understand the mechanism of REE in methane seep area.

Conclusion
In this paper, there is an obvious increase of ∑REE/Ti at the top of SMTZ of sites SH1 and SH3. Based on the analysis of δEu vs. δCe, HREE vs. LREE, and the REE-UCC normalized patterns, the REE characteristics of SH1 and SH3 are different from the Pearl River, SW Taiwan, and the Luzhon river, indicating that the terrigenous sediment input has little contribution to the REE enrichments. Hence, the REE was suggested to be more influenced by the authigenic minerals and later diagenesis. Combined the analysis of the different forms of Fe-Mn oxides (hydroxides) minerals, carbonate phase, and the REE patterns, the Fe-Mn oxides (hydroxides) and Fe-Mn carbonate phase have a positive correlation with the REE content, indicating that the REE-rich carbonate phase and REE-rich Fe-Mn oxides (hydroxides) phase have the important contribution of REE enrichment at the SMTZ. However, there is little increase of REE at the SMTZ, it increased at the top of the SMTZ. It is inferred that the AOM results in the decrease of the REE-rich Fe-Mn oxides (hydroxides) phase and a decrease of the REE-rich carbonate phase. This is the reason why the ∑REE/Ti is enriched at the top of the SMTZ. Our results are of great scientific significance for understanding the REE enrichment mechanism in the methane seeps.

Data Availability
All the data have been supplied in the article in the form of a table.

Conflicts of Interest
The authors declare that they have no conflicts of interest.  Figure 8: At first, it is inferred that the metal oxides (hydroxides) are reduced at the SMTZ, meanwhile, the REE-rich Fe-Mn oxides (hydroxides) will release REE in the pore water during the AOM. The released REE will reconvert to the REE-rich carbonate phase. At the SMTZ, the REE-rich Fe-Mn carbonate phase increased and the REE-rich Fe-Mn oxides (hydroxides) decreased. Therefore, the REE is not enriched in the SMTZ while enriched at the top of the SMTZ. 10 Geofluids