Geochemistry and mineralogy coupling reveals the Fe-S cycle in a receding methane seep

1.   Introduction

Receding Methane seeps is an ideal environment to record the biogeochemical processes and products during and after the methane seep activity, especially for the iron and sulfur cycle (Wang M et al., 2015; Zhao J et al., 2021). The main reaction in methane seep is sulfate reduction coupled with the anaerobic oxidation of methane (SR-AOM) and thus defines the position of sulfate–methane transition zone (SMTZ) [Eq. (1)]. Besides, in normal deep-sea sediment, organoclastic sulfate reduction (OSR) is the predominant mode for organic matter consumption [Eq. (2); Canfield DE, 1991)]. Both of these biogeochemical processes generate hydrogen sulfide (H₂S) and subsequently pyrite by combining with free Fe³⁺ and Fe²⁺ in sediments (Jorgensen BB et al., 2004; Peckmann J and Thiel V, 2004).

Pyrite provides the net sink of sulfur from the oceans to marine sediments and thus plays a key role in the global sulfur cycle (Bottrell SH and Newton RJ, 2006). Due to intense SR-AOM in the SMTZ, the content of pyrite rises steeply to reach a high level compared to the non-SMTZ section (Dewangan P et al., 2013; Miao XM et al., 2021). During pyrite formation, ³²S favors combination with free Fe²⁺ instead of ³⁴S; therefore, the formed FeS and later pyrite are ³⁴S-depleted and expressed as negative δ³⁴S values (Rickard DT, 1975). In typical marine sediments, sulfur isotopic fractionation is controlled by OSR up to 70‰ (Canfield DE et al., 2010; Sim MS et al., 2011). However, situation varies significantly in methane seep sediment: generally, while the SMTZ remains stable at or near the surface of the seafloor, the rapid feed from downward diffusing seawater sulfate inhibits the exhaustion of porewater sulfate (Gong SG et al., 2018). As a result, the produced pyrite has extremely negative δ³⁴S value, which means it is difficult to distinguish the OSR and SR-AOM by ³⁴S-³²S fractionation alone (Formolo MJ and Lyons TW, 2013; Lin ZY et al., 2017); in contrast, seawater sulfate has difficulty diffusing downward into the deeper SMTZ (Gong SG et al., 2018). The produced pyrite in such an environment generally exhibits a positive δ³⁴S value relative to that of seawater of 21‰ (Tostevin R et al., 2014) or even higher (Borowski WS et al., 2013; Lin Q et al., 2015; Lin ZY et al., 2016; Gong S et al., 2022).

A part of previously formed pyrite in methane seep sediments is reoxidized to sulfate and Fe (hydrogen) oxide under different conditions (Gartman A and Luther GW, 2014; Mahoney C et al., 2019). Certainly, this process generates various intermediates, such as ferric hydroxide (Kirk Nordstrom D, 1982). Studies in the field (Zhang YC et al., 2012; Zhao J et al., 2021) and in laboratories (Nyavor K et al., 1996; Mitsunobu S et al., 2021) further showed both biotic and abiotic pathways stimulating pyrite oxidation. In many cases, the oxidized pyrite formed Fe (hydrogen) oxide coatings (rim) that surrounded tiny pyrite cores (Schieber J, 2007; Kamata A and Katoh M, 2019; Dantas RC et al., 2022) and thus possibly prevented or retarded the inner core from being further oxidized (Huminicki DMC and Rimstidt JD, 2009). Accordingly, the oxidation rate of pyrite varies greatly between early alteration and later erosion (Gartman A and Luther GW, 2014). Consequently, such intermediate species formed during pyrite oxidation can be preserved for long periods of time (Morse JW, 1991) and be gradually transformed into goethite as a terminal product (pyrite pseudomorphs; Merinero R et al., 2008; Huminicki DMC and Rimstidt JD, 2009).

Although studies aiming to explore the intimate relationship between methane seeps and diagnostic minerals in South China Sea (SCS) have yielded numerous outcomes (Lin ZY et al., 2017; Chen TT et al., 2021), correlation studies among geochemistry and multiple minerals, for instance, pyrite, framboid gothite, elemental sulfur (ES) and intermediate species of pyrite oxidation both in natural and laboratory environments, are still lack and therefore prevent our further understanding of Fe-S cycle in wanning methane seep in this region. To make clear these processes, we carried out diverse experiments into a 14-m core sampled from SCS. The comprehensive geochemical and mineralogical results exhibited two methane seep activities during the past 46 kyr, basic on this result, this paper demonstrated an intrinsic relationship among these minerals and constructed the Fe-S cycle, which is dominated by the position change of the SMTZ in this core. Therefore, our results provide a viable reference for studying the Fe-S cycle in methane seep sediments.

2.   Regional setting

The South China Sea is one of the largest marginal seas in the west Pacific under the complex impact of the Eurasian, Pacific and India-Australian plates drift (Fig. 1). The Shenhu area is located at northern slope of SCS, between the Xisha trough and the Dongsha islands, adjacent to several large oil and gas fields (He JX et al., 2009; Li YJ et al., 2016), with water depth ranges between 800 m and 2,000 m and complex submarine topography (sea valleys, domes, and erosion grooves; Liu CL et al., 2012). The structural setting of the Shenhu area is controlled by complex fractures, folds, and diapirs that favor the flow of methane-rich fluids and the formation of gas hydrates (Wang JH et al., 2006; Liang JQ et al., 2014; Su P et al., 2018). The sedimentary stratum is 1000−7000 m thick with organic matter contents of 0.46%–1.9% (Wang HB et al., 2003; Su Z et al., 2012; Wu LS et al., 2013). This organic matter is considered to be the source of early microbial-generated methane, found throughout the sediment sequences (Wu NY et al., 2008). Moreover, the geological and geophysical data and drilling and gas hydrate exploration results all show that the Shenhu area has great gas hydrate exploration and development prospects (Wang XJ et al., 2014; Su M et al., 2016; Li Y et al., 2019; Zhang W et al., 2020b; Song YR et al., 2022).

Figure  1.  Bathymetric map of core Z22-3. The map was made with the GeoMapApp (www.geomapapp.org) / CC BY / CC BY (Ryan WBF et al., 2009).

DownLoad: Full-Size Img PowerPoint

3.   Materials and methods

3.1   Sample

The sample was collected from Shenhu area, South China Sea conducted by Qingdao Institute of Marine Geology in 2022. The sampling depth is 1650 m, and the column length is 14.3 m. Clay is the main composition in this column and gradually consolidate with depth. After collected, the sediment column is immediately sealed and stored at 4°C, and then divided at an interval of 2 cm in the laboratory, then freeze-dried and refrigerated until experiments.

3.2   Total organic carbon (TOC) determination

To determine δ¹³C_TOC and TOC, 5 mg weighed samples is putted into a 5×9 silver cups, add appropriate concentration of 1N hydrochloric acid, remove inorganic carbon, after completely removing inorganic carbon, put in a sealed box with magnesium perchlorate and caustic soda, remove hydrochloric acid and water. The silver cup containing the sample was wrapped in a 5×9 tin cup. The tin cup was sealed and compacted and the sample was determined by the Element Analyser-Stable Isotope Mass Spectrometer online (EA Isolink-253 Plus, Thermo Scientific). The isotope reference material is IVA33802151 (δ¹³C=−28.85‰), and the δ¹³C value is the international standard material VPDB (Vienna Peedee Belemnite). The analysis accuracy was < 0.1‰ (n=8). The standard substance contents were IVA33802180 (No.133506, TC=0.83%), IVA33802186 (No.133505, TC=2.75%) and IVA33802188 (No. 305012, TC=4.57%). According to the weight, content and test integral area of the standard material, as well as the test integral area of the sample, the weight of the total organic carbon of the sample sediment was obtained. The weight was divided by the weight of the sample to get the total organic carbon content of the sediment. The analysis accuracy was less than 0.1%.

3.3   Chromium reducible sulfur (CRS) extraction and δ³⁴S determination

Reduced inorganic sulfur was extracted from the sample using CrCl₂ and HCl in Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences. This reduction method releases sulfur from all sulfide lattices until the sulfur content equals the total amount of CRS (Canfield DE et al., 1986). Silver sulfide was analyzed for δ³⁴S using an elemental analysis-isotope ratio mass spectrometer at the State Key Laboratory of Biogeology and Environmental Geology, China University of Geosciences (Wuhan, China). All the results are reported in standard delta notation per mil deviation using the Vienna-defined Canyon Diablo Troilite (VCDT) scale [Eq. (3)]. The standard deviation of the measurements is <0.3‰ (VCDT). A measurement error of about0.2‰ (1σ) was calculated based on repeated analyses of IAEA International Standards IAEA S2 (22.7‰), IAEA S3 (−32.3‰), and National Reference Materials (−0.07‰).

3.4   SEM and EDS test of hand-picked minerals

4 g of each freeze-dried sample was weighed and soaked in deionized water for 48 h. The fully dispersed samples were placed in a standard sieve with a sieve aperture of 0.15 mm, washed with running water and shaken until no sediment remained on the minerals, then transferred into a drying vessel and dried. The multiple minerals [pyrite, Fe (hydrogen) oxides etc.] were picked out from sieved and dried samples in the vessel under a stereoscope. This study used the Cold Field Emission Scanning Electron Microscope (S-4800, Hitachi, Japan) and corresponding Energy Dispersive Spectrometer in Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences. Minerals firstly were fixed at the sample stage by conductive glue, then gilded. Further, the sample stage was putted into the sample holder and then observed carefully. Subsequently, taking photos of specific areas and finally performed the EDS test. Furthermore, the size distributions of framboidal pyrite were measured based on SEM images. SEM images provide a view of whole framboids so the apparent diameter is equal to the true diameter, avoiding skews in size distributions associated with two dimensional slices as in thin-section petrography (Wilkin RT et al., 1996).

3.5   In situ sulfur isotope determination of pyrite

In situ sulfur isotope analyses of bulk pyrite were performed on a Neptune Plus MC-ICP-MS (Thermo Fisher Scientific, Bremen, Germany) equipped with a Geolas HD excimer ArF laser ablation system (Coherent, Göttingen, Germany) at the Wuhan SampleSolution Analytical Technology Co., Ltd, Hubei, China. In the laser ablation system, helium was used as the carrier gas for the ablation cell and was mixed with argon (makeup gas) after the ablation cell. The single spot ablation mode was used. Then the large spot size (44 μm) and slow pulse frequency (2 Hz) were used to avoid the down hole fractionation effect which have been reported by (Fu J 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 the short-term variation of the signal, especially for the slow pulse frequency condition (Hu ZC et al., 2015). The laser fluence was kept constant at about5 J/cm². The Neptune Plus was equipped with nine Faraday cups fitted with 10¹¹Ω resistors. Isotopes ³²S, ³³S and ³⁴S were collected in Faraday cups using static mode. The newly designed X skimmer cone and Jet sample cone in Neptune Plus were used to improve the signal intensity. The nitrogen (4 ml/min) was added to the central gas flow to reduce the polyatomic interferences. All measurements were performed using the medium-resolution with a revolving power (as defined by a peak edge width from 5%–95% of the full peak height) that was always greater than 5000.

A standard-sample bracketing method (SSB) was employed to correct for instrumental mass fractionation. To avoid the matrix effect, a pyrite standard PPP-1, a chalcopyrite standard GBW07268 (a pressed pellet) and a synthetic Ag₂S standard IAEA-S-1 (a pressed pellet) were chosen as reference materials for correcting the natural pyrite, pyrrhotite and pentlandite samples, the natural chalcopyrite samples and the natural Ag₂S samples, respectively. The reference values of δ³⁴S_V-CDT in these standards have been reported by (Fu J et al., 2016). In addition, the in-house references of a pyrrhotite SP-Po-01 (δ³⁴S_V-CDT=1.4‰±0.4), a chalcopyrite SP-CP-01 (δ³⁴S_V-CDT=5.45‰±0.3) and two synthetic Ag₂S standard IAEA-S-2 (δ³⁴S_V-CDT=22.58‰±0.39) and IAEA-S-3 (δ³⁴S_V-CDT=−32.18‰±0.45) were analyzing repeatedly as unknown samples to verify the accuracy of the calibration method. The more detail of the in situ Pb isotopic ratios analysis was described in (Fu J et al., 2016). All data reduction for the MC-ICP-MS analysis of S isotope ratios was conducted using “Iso-Compass” software (Zhang W et al., 2020a). The analysis accuracy is less than 0.4‰.

3.6   Dating of planktonic foraminifera Accelerator mass spectrometry ¹⁴C (AMS ¹⁴C)

This study considered the Globigerinoides ruber (G. ruber) and Globigerinoides sacculifer (G. sacculifer) for AMS ¹⁴C dating. Each sample containing at least 8 mg of G. ruber and G. sacculifer foraminifera shells was sent to the Pilot National Laboratory for Marine Science and Technology (Qingdao) for radiocarbon dating. The ¹⁴C content of graphitized samples after being targeted was determined by accelerator mass spectrometry, and the instrument used was the 0.5MV tandem accelerator mass spectrometer of NEC (National Electrostatics Corporation, USA) (produced in 2018). Data processing utilizes NEC’s “abc” data processing software and “Fudger” data processing software developed by Lawrence Livermore National Laboratory in the United States. The original data results for all samples were the internationally accepted Fraction Modern (F¹⁴C; Stenström K et al., 2011). The standard for Modern compounds used is the internationally accepted oxalic acid, OXII (SRM 4990C). According to the measured F¹⁴C, the ¹⁴C content, Δ¹⁴C (‰; Stenström K et al., 2011), and ¹⁴C age (year before present, yr BP; CALIB rev.8; Stuiver M and Reimer PJ, 1993) were calculated, and the error value (σ value) was calculated.

3.7   Determination of carbon and oxygen isotopes of planktonic foraminifera

A total of 10 (about1 mg) G. ruber and G. sacculifer foraminifera shells were selected for each sample to determine the carbon and oxygen isotopes present in planktonic foraminifera. The δ¹³C and δ¹⁸O associated with the bulk carbonate fraction of planktonic foraminifera were determined using Gasbench II-IRMS (a Gasbench II interfaced with a 253 plus mass spectrometer) at the Key Laboratory of Climate, Resources, and Environment in Continental Shelf Sea and Deep Sea of Department of Education of Guangdong Province, Guangdong Ocean University, China. Isotopic ratios are expressed in δ-notation as per mil (‰) deviation relative to the Vienna Pee Dee Belemnite standard. The analytical precisions of δ¹³C and δ¹⁸O were less than 0.2‰.

4.   Results

4.1   AMS ¹⁴C dating and carbon and oxygen isotopes data of planktonic foraminifera

Based on the AMS ¹⁴C dating results (Fig. 2a and Table 1), the age of the sediment gradually increased with depth, and the age of the bottom sediment is 46.43 kyr. Moreover, by comparing these results with previous studies in the SCS (Wang PX, 1999; Zheng HB et al., 2008; Li BH and Wang XY, 2009), this paper further divides this core into Marine Isotope Stages (MIS) 1‒3 (Fig. 2b and c). It is obvious that the deposition rate during MIS1 was extremely low (10.805 cm/kyr), and the sediment rate first increased with depth during MIS2 (29.350 cm/kyr) and then slightly declined during late MIS3 (22.120 cm/kyr); however, the sediment rates then rapidly increased to the bottom of the core (47.82 and 104.972 cm/kyr, respectively).

Figure  2.  AMS ¹⁴C dating and corresponding sedimentation rate and carbon (δ¹³C) and oxygen (δ¹⁸O) isotope profile of planktonic foraminifera.

DownLoad: Full-Size Img PowerPoint

Table  1.  The results of AMS ¹⁴C dating

Sample number	F14C	σ-F14C	14C age (kyr BP)	σ-14C age (yr)	Δ14C (‰)	σ-Δ14C (‰)
Z22-3-71	0.1970	0.0011	13.05	45	−804.7	1.1
Z22-3-281	0.0332	0.0008	27.36	190	−967.1	0.8
Z22-3-401	0.0086	0.0007	38.21	650	−991.5	0.7
Z22-3-511	0.0048	0.0007	42.81	1140	−995.2	0.7
Z22-3-701	0.0031	0.0007	46.43	1780	−996.9	0.7

 | Show Table

DownLoad: CSV

4.2   Redox indicators

The main redox indicators of core Z22-3 are displayed in Fig. 3 and Table 2 and S1. Most of the CRS contents (represents pyrite hereafter) reveal a significant high value (up to 0.52%) in the SMTZs (Fig. 3a) compared to those of non-SMTZ sections. While the δ³⁴S value of CRS show an extremely negative trend from the top to the bottom in this core (Fig. 3b; range from −47.63‰ to −18.78‰). However, a gradually increasing trend of δ³⁴S value of CRS in the upper SMTZ is also nonnegligible. The TOC apparently increases in the top sediment and then gradually declines in the lower intervals (Fig. 3c). However, both of the TOC content and δ¹³C value varies completely within a narrow range (0.63%–1.35% and −21.69‰–−19.90‰, respectively). The CRS/TOC (S/C) ratio changes significantly from the top of the sediment to the bottom (Fig. 3d; from 0 to 0.52). Most of the S/C values in the SMTZ are higher than those in the non-SMTZ. This study only utilize the results that involve at least 30 measurements of framboid size, and in fact, most of the measurements exceed 100 (Rickard D, 2019). The mean diameter (MD) of the framboid vary significantly (Fig. 3e; 7.82−20.78μm). Furthermore, the selected framboids from hand-picked pyrite generally reveal negative δ³⁴S values (Fig. 3f and Table S2; low to −49.73‰).

Figure  3.  redox profiles of core Z22-3. (a) and (b)‒the content and corresponding δ³⁴S values profiles of CRS. (c)‒the content and δ¹³C value of TOC. (d)‒S_CRS/C ratios with depth. (e)‒mean diameter of framboids. (f)‒δ³⁴S values of hand-picked pyrite. The grey zone indicates the position of paleo-SMTZ.

DownLoad: Full-Size Img PowerPoint

Table  2.  Distribution of pyrite framboid

Sample number	Depth (cmbsf)	Distribution of pyrite framboid sizes
		Minimum size (μm)	Maximum size (μm)	Mean diameter (μm)	standard deviation (SD/μm)	Number
Z22-3-81	161	6	16	11.60	2.09	67
Z22-3-101	201	4	30	11.47	6.66	60
Z22-3-111	221	3	23	9.87	3.33	325
Z22-3-131	261	3	33	11.12	7.29	109
Z22-3-151	301	5	18	11.29	2.89	145
Z22-3-171	341	6	25	11.22	2.90	174
Z22-3-181	361	7	58	20.62	13.05	185
Z22-3-191	381	7	17	10.15	1.91	113
Z22-3-211	421	5	44	14.87	7.77	159
Z22-3-221	441	5	41	13.46	6.73	138
Z22-3-231	461	8	59	20.78	12.91	172
Z22-3-241	481	5	46	16.44	5.80	221
Z22-3-251	501	6	30	12.35	3.55	100
Z22-3-271	541	8	21	12.49	2.21	229
Z22-3-281	561	8	48	13.87	5.24	84
Z22-3-291	581	5	18	10.19	2.80	201
Z22-3-321	641	6	17	10.86	1.66	137
Z22-3-331	661	5	20	7.94	3.42	140
Z22-3-351	701	5	37	8.72	3.38	125
Z22-3-371	741	6	15	9.61	2.27	92
Z22-3-381	761	5	25	9.55	3.56	62
Z22-3-391	781	5	48	9.15	5.16	145
Z22-3-401	801	6	18	8.39	1.42	114
Z22-3-411	821	5	15	9.33	2.79	86
Z22-3-481	961	4	11	7.82	1.82	136
Z22-3-561	1121	6	30	11.11	2.51	203
Z22-3-601	1201	7	41	13.32	6.82	130
Z22-3-611	1221	8	39	11.66	3.88	101
Z22-3-641	1281	6	78	10.84	6.30	168
Z22-3-671	1341	9	20	12.94	2.38	114
Z22-3-701	1401	6	45	13.47	8.08	116

 | Show Table

DownLoad: CSV

4.3   Various morphologies of pyrite

The various morphologies of pyrite that emerged in this core are described in Fig. 4. Most of the bulk pyrite is irregularly shaped, and the particle size varies greatly (Fig. 4a; 200−500 μm). The dominant pyrite morphology throughout the core is octahedral (Fig. 4b), which corresponds with previous studies on methane hydrate-bearing sediments in the northern SCS (Xie L et al., 2012; Wu L et al., 2014). Moreover, modified pyritohedron microcrystals (Fig. 4c; Barnard AS and Russo SP, 2007) and cubic pyrite are found in several intervals (Fig. 4d‒g) in or closed to the SMTZ. In addition, euhedral pyrite with large sizes is distributed in both SMTZs, and the single crystal size reaches 4−5 μm (Fig. 4h and i).

Figure  4.  Diverse morphologies of pyrite. (a)‒the stereoscope image of pyrite (1281 cmbsf). (b)‒framboid formed by the octahedral microcrystals (261 cmbsf). (c)‒framboid formed by the intermediate between optimized truncated cubic shape and octahedron (301 cmbsf; Barnard AS and Russo SP, 2007; Chang JY et al., 2022). (d) and (f)‒framboids formed by the cubic microcrystals (421 cmbsf and 461 cmbsf respectively). (e) and (g)‒the enlarged areas of white boxes in (d) and (f) respectively. (h) and (i)‒euhedral pyrite with large size (221 cmbsf and 1261 cmbsf respectively), single particle size reaches 4‒5 μm.

DownLoad: Full-Size Img PowerPoint

4.4   Elemental sulfur (ES)

Elemental sulfur emerged in multiple intervals (Fig. 5 and Table S3), and its morphological change with depth is significant (Fig. 5a‒c). However, the morphological variations in ES in and below the upper SMTZ reveal an apparent order from anhedral to semi-euhedral and then euhedral (rhombus). In addition, ES can exist on the surface of minerals other than pyrite (Fig. 5d). Moreover, octahedral pyrite with a small size (<1 μm) emerging from the ES surface prevails throughout the core (Fig. 5e and f). In particular, the octahedral microcrystals are generally clustered and are only found in the oxidized area of ES. The typical EDS result of ES is shown in Fig. 5g.

Figure  5.  Elemental sulfur discovered in the multiple intervals. (a)‒anhedral elemental sulfur coexists with pyrite (281 cmbsf). The red arrows indicate the elemental sulfur for example (similarly hereinafter). (b)‒subhedral-euhedral elemental sulfur (681 cmbsf). (c)‒euhedral elemental sulfur (781 cmbsf). (d)‒subhedral elemental sulfur distribute on the surface of other minerals (281 cmbsf). (e)‒octahedral pyrite germinated from the elemental sulfur surface (701 cmbsf). (f)‒the enlarged area of white box in (e). (g)‒the typical EDS result of elemental sulfur (701 cmbsf), the yellow box represents where the EDS is done (similarly hereinafter).

DownLoad: Full-Size Img PowerPoint

4.5   Intermediate species of pyrite oxidation

Intermediate species of pyrite oxidation appeared in several intervals (Fig. 6). The morphologies of these species include (hemi)spherical (Fig. 6a and b) and irregular (Fig. S1). Furthermore, all the particles of this mineral reveal a unique color dominated by blue-green to dark-blue. The SEM image of this intermediate species covers many framboids that remain their original form (Fig. 6c). The EDS results of this species show a distinctively high content of oxygen (Fig. 6d and e; Table S4). Additionally, the oxidation of former pyrite is apparent, and thus, some pyrite microcrystals did not retain their original morphologies (Fig. 6e).

Figure  6.  Intermediate species of pyrite oxidation appeared in the several intervals. (a)‒the stereoscope image of intermediate species enclosed in the foraminiferal tests and exhibit complete spherical shape (481 cmbsf). (b)‒intermediate species shows hemispherical shape (481 cmbsf). (c)‒SEM image of the hemispherical intermediate species (481 cmbsf). Numerous framboids keep as the initial shape. (d)‒the EDS result of enlarged area of white box in (c). (e)‒the EDS result of loosely cemented intermediate species (221 cmbsf). The alteration of former pyrite is apparent.

DownLoad: Full-Size Img PowerPoint

4.6   Framboid goethite (pyrite pseudomorphs)

Fe (hydrogen) oxides (mostly goethite) are common in this core (Fig. 7). The bulk Fe (hydrogen) oxides show a typical dark-reddish brown color under stereoscopy (Fig. 7a). More specifically, the framboid goethite (pyrite pseudomorphs) found in diverse intervals indicate the complete oxidation of framboid pyrite (Fig. 7b and c; Merinero R et al., 2008, 2009; Cavalazzi B et al., 2012). The EDS results of framboid goethite are exhibited in Fig. 7d and e.

Figure  7.  Framboid goethite (pyrite pseudomorphs) distribute in the multiple intervals. (a)‒the dark-reddish brown Fe (hydrogen) oxides (including goethite) under stereoscope (661 cmbsf). (b) and (c)‒the SEM results of framboid goethite (401 cmbsf and 761 cmbsf respectively). (d) and (e)‒the EDS results of amplified areas of white boxes in (b) and (c) respectively.

DownLoad: Full-Size Img PowerPoint

5.   Discussion

5.1   Geochemical indicators reveal potential SR-AOM and limited OSR

Berner RA (1982) showed that the contents of TS and TOC in typical marine sediments controlled by OSR generally show a positive correlation with a relatively constant S/C ratio of 0.36. However, a deviation from this correlation to a higher S/C ratio is usually seen in methane seep sediments due to SR-AOM (Sato H et al., 2012; Hu Y et al., 2017). Although no detailed TS data were obtained in this study, the S_CRS/C trend (Fig. 3d) corresponded to the previous cases where the S_CRS/C ratio in the SMTZ was higher than that in the non-SMTZ (Miao XM et al., 2021). In addition, the weak correlations between CRS and TOC (Fig. 8a; R²=0.004) and δ¹³C_TOC (Fig. 8b; R²=0.095) indicate that OSR is not the predominant biogeochemical process for pyrite formation (Miao XM et al., 2021; Chang X et al., 2023).

Figure  8.  Scatter plot of solid-phase data (a)‒TOC vs CRS, (b)‒δ¹³C_TOC vs CRS.

DownLoad: Full-Size Img PowerPoint

5.2   Mineralogical features reveal the position of the SMTZ

In this study, high CRS contents in 81−421 cmbsf and 1021−1401 cmbsf (Fig. 3a) reveal the approximate position of paleo-SMTZs. Notably, the high deposition rate (104.972 cm/kyr) that emerged in 42.81−46.43 kyr prevented the organic matter from being consumed quickly and therefore contributed to the later formation of diagenetic framboids via OSR (Canfield DE, 1991). Although the content of pyrite in the lower SMTZ formed by OSR rose, in dysoxic or anoxic sediment, such as the Shenhu area (Lin ZY et al., 2016), framboids formed by this pathway typically have small sizes (MD<10 μm) and narrow distribution ranges (SD<3 μm; Chang JY et al., 2022). However, framboidal pyrite formed in the SMTZ generally has a large mean diameter (15−80 µm) and a wide size range (3−35 µm) (Chang JY et al., 2022). The mean diameter (9.87−20.62 μm and 10.84−13.47 μm, respectively) and standard deviation of framboid (1.91−13.05 μm and 2.51−8.08 μm, respectively) in SMTZ in our study are more corresponding to this conclusion. Furthermore, it has been widely found in the Shenhu area of the SCS that the mean diameters of framboids in the SMTZ are larger than those in the non-SMTZ of methane seep sediments (Lin Q et al., 2016a; Lin ZY et al., 2016; Wang B et al., 2022). By combining these factors, it was concluded that the lower SMTZ was impacted by SR-AOM to some extent, that’s meaning the methane seep was once active in these intervals.

Regarding the morphology of pyrite, octahedra are most prevalent in marine sediments, even in methane seep environments, including in the South Atlantic Ocean (Dantas RC et al., 2022), South China Sea (Xie L et al., 2013; Lin Q et al., 2016b; Lin ZY et al., 2016; Lin ZY et al., 2017), East China Sea (Wang M et al., 2015), and Baltic Sea (Bottcher ME and Lepland A, 2000; Neumann T et al., 2005; Liu JR et al., 2022). However, cubic pyrite usually exists in methane seep sediments (Liu JR et al., 2022) and references therein; Chen DF et al., 2006) even in permafrost areas (Wang PK et al., 2014). Typically, pyrite synthesis via the “polysulfide pathway” generates an octahedral morphology (Luther GW, 1991; Gartman A and Luther GW, 2013), while synthesis via the “H₂S pathway” predominantly results in various morphologies, including cubes and framboids (Rickard D and Luther GW, 1997; Gartman A and Luther GW, 2013). Different morphologies of pyrite, including octahedrons and cubes discovered in the diverse intervals in or closed to the SMTZ (Fig. 4), indicate that the methane seep remained stable for a rather long period of time in this section. The vast H₂S generated by SR-AOM in the SMTZ extent to out diffuse and dissolved the iron oxide in the sediment (Jorgensen BB et al., 2004; Peckmann J and Thiel V, 2004; Riedinger N et al., 2017) and thus provided the Fe²⁺. Due to the sufficient and continuous supply of iron and sulfur, pyrite gradually grew with multiple morphologies (Merinero R et al., 2008; Liu JR et al., 2022). In addition, the euhedral microcrystals in clusters had larger sizes (Fig. 4h and i; single crystal size reached about 4−5 μm) relative to non-clustered crystals (Merinero R et al., 2008; Lin ZY et al., 2016); thus, the authors think these microcrystals directly precipitated from porewater (Taylor KG and Macquaker JHS, 2011; Lin ZY et al., 2016).

In this core, the extremely negative δ³⁴S values of both CRS (Fig. 3b; low to −47.69‰) and hand-picked pyrite (Fig. 3f; low to −49.73‰) observed in various intervals in both SMTZs with respect to sea water sulfate (21‰; Borowski WS et al., 2013; Tostevin R et al., 2014) show sulfur isotopic fractionation greater than 70‰. This is also found in Site 4B, a core in Shenhu area very colsed to our study, where high pyrite weight reveal significant negative δ³⁴S values (−49.16‰–−41.69‰ CDT) in the methane affected sections and it’s attributed to episodic eruption of a mud volcano (Xie L et al., 2013). An obvious positive trend of δ³⁴S_CRS values with depth (−47.69‰–−36.54‰) in the upper SMTZ also indicates the potential SR-AOM (Fernandes S et al., 2020; Li N et al., 2021). Therefore, the authors suggest that both SMTZs once stabilized at or near the surface of the seafloor. The impact of disproportionation of intermediate sulfur species and sulfide reoxidation promote the sulfur isotope fractionation, and the vast elemental sulfur found in the SMTZs of this core (Fig. 5) also supports this viewpoint (Peckmann J and Thiel V, 2004; Liu J et al., 2020).

5.3   Multiple minerals reveal waning methane seepage

5.3.1   Decreasing methane seepage revealed by elemental sulfur analysis

Recently, diverse studies, especially in methane seeps, have revealed that elemental sulfur is preserved as a rather stable solid mineral in sediments (Lin Q et al., 2015; Riedinger N et al., 2017; Lin ZY et al., 2018). However, studies aiming to explain the morphology of elemental sulfur in methane seep sediment are still limited (Kamyshny A and Ferdelman TG, 2010; Lichtschlag A et al., 2013). As previous studies concluded, the sulfide-polysulfide-dissolved elemental sulfur system reaches equilibrium within seconds (Kamyshny A et al., 2003); however, both the sulfide-polysulfide-colloidal elemental sulfur system (Kamyshny A and Ferdelman TG, 2010) and sulfide-polysulfide-rhombic elemental sulfur system (Boulegue J and Michard G, 1977) require hours to reach equilibrium. In addition, elemental sulfur in deeper sediments tends to be further from equilibrium than that near the sediment surface (Lichtschlag A et al., 2013).

In this core, elemental sulfur was observed in multiple intervals (Fig. 5) and had anhedral, subhedral and euhedral (rhombic) morphologies. Moreover, the large amount of elemental sulfur coexisted with octahedral pyrite in all sulfur-bearing samples (Fig. 5a‒c), implying that methane seepage had ceased. Otherwise, the continuously produced H₂S from SR-AOM would have stimulated the forward reaction in polysulfide equilibrium (Lichtschlag A et al., 2013) and thus abundant elemental sulfur is difficult to preserve, particularly throughout SMTZs. The lack of H₂S thus benefited the preservation of elemental sulfur in the methane seep sediments (Riedinger N et al., 2017; Lin ZY et al., 2018). In this case, when the intensity of methane seepage weakened, a part of the remaining H₂S was oxidized to elemental sulfur by several of the pathways (Biogenic and Abiogenic; Pyzik AJ and Sommer SE, 1981; Otte S et al., 1999; Schippers A and Jørgensen BB, 2001) and then remained. The elemental sulfur was only distributed on the surface of other minerals, which also indicated that it formed later (Lin Q et al., 2015). In addition, only octahedral pyrite crystals formed on the oxidized surface of elemental sulfur found in various intervals (Fig. 5e and f), supporting the belief that elemental sulfur reacted with hydrogen sulfide and produced polysulfides and subsequently pyrite (Luther GW, 1991; Liu J et al., 2020). Furthermore, the morphology shifts among anhedral, subhedral and euhedral (rhombic) elemental sulfur forms in the sediments possibly correlated to redox condition variation, that is, changes from oxic to anaerobic with depth. However, this hypothesis requires more evidence based on experimental cultures and natural observations.

5.3.2   Iron-sulfur cycle revealed by intermediate species and Fe (hydrogen) oxide

The oxidation of pyrite involves multiple intermediate species (szomolnokite, jarosite, and possibly green rust; Huggins FE et al., 1980; Gartman A and Luther GW, 2014; Koeksoy E et al., 2019) but does not involve elemental sulfur (Anderson TF and Pratt LM, 1995). During the oxidative processes, Fe (hydrogen) oxide coatings (or rims) surrounding pyrite cores have been found in multiple studies (Morse JW, 1991; Mahoney C et al., 2019; Dantas RC et al., 2022) and may provide a feasible interpretation for this core. Huminicki DMC and Rimstidt JD (2009) suggested that when pyrite oxidizes at near neutral pH (normal marine sediment) in the presence of sufficient alkalinity, Fe oxyhydroxide coatings develop on the surface. As the coatings grow thicker, the oxidation rate of pyrite is obviously restricted, preventing or retarding further oxidation of the inner pyrite (Huminicki DMC and Rimstidt JD, 2009; Mahoney C et al., 2019; Wang WT et al., 2020). Furthermore, Gartman A and Luther GW (2014) proposed precise electron transfer mechanisms in the later step between the pyrite core and O₂ via either O₂ diffusion through the pore space or electron shuttling within Fe(II)/(III) hydroxides, which contributes to the complete oxidation of pyrite to goethite. Overall, the terminal product of pyrite oxidation is Fe(Ⅲ) (hydrogen) oxide (mostly goethite; Huggins FE et al., 1980; Huminicki DMC and Rimstidt JD, 2009; Zhang P and Yuan SH, 2017; Dantas RC et al., 2022) and sulfate (Schippers A and Jørgensen BB, 2001). The characteristic morphology of Fe(Ⅲ) (hydrogen) oxide (especially many framboid goethite) serves as a proxy for former pyrite in methane seeps (Merinero R et al., 2008, 2009; Zhong Y et al., 2017). Generally, Fe (hydrogen) oxide is hard to preserve in active methane seeps due to the large amount of H₂S generated by SR-AOM (Riedinger N et al., 2005; Dewangan P et al., 2013; Roberts AP, 2015). The iron ions are supplied by dissolving Fe (hydrogen) oxide, which then participate in the later processes of pyrite formation (Roberts AP, 2015). Furthermore, Fe (hydrogen) oxide is completely consumed within a distinct sediment interval in the long term under consistent H₂S flux (several millennia; Riedinger N et al., 2005).

In this study, all of the intermediate species of pyrite clearly contained S, Fe and O (Fig. 5 b, d and e). The results of EDS showed that the Fe/S and Fe/O rates for % and At% of the intermediate species were greater than those of pyrite and goethite, respectively (Fig. 9a and b). This means that the composition of intermediate species fell between that of pyrite and goethite. In contrast, in ‘sulfur-depleted’ pyrite crystals, the Fe/S of At% is much greater than 1, but that of ‘pristine’ pyrite crystals is less than 1, precisely equal to 0.5 (Mahoney C et al., 2019). However, the EDS results showed both situations (Fig. 9b and Table S2). Therefore, by combining these results with those of previous studies, the authors suggest that this intermediate species is a mixture bearing both FeS₂ and Fe (hydrogen) oxide. The outer coatings formed by Fe (hydrogen) oxide exhibited a unique blue-green to dark-blue color, which was different from that of the inner pyrite (Fig. 6a and b). Although the intermediate species is represented as a mixture, the correlation coefficient of % (R²=0.5625) and At% (R²=0.8217) of S-O implied a decrease in sulfur corresponding to an increase in oxygen (Fig. 9c and d) and thus reflected the oxidation degree to a certain extent. Moreover, the evident alteration of pyrite observed in Fig. 6e revealed that the coatings possibly sloughed off and that the original morphology of the pyrite crystals changed (Mitsunobu S et al., 2021; Zhao ZY et al., 2022). However, regardless of whether the elemental content was modified or the morphology changed, the position of oxidized crystals in bulk pyrite, especially in the framboid, remained fixed and thus were ‘pyrite pseudomorphs’ (Merinero R et al., 2008).

Figure  9.  Element ratio and content scatter plot of the intermediate species. (a)‒the data of Fe (%) /S (%) and Fe (%)/O (%) of intermediate species. Fe (%)/S (%)_pyrite=0.88; Fe (%)/O (%)_goethite=1.75. (b)‒the data of Fe (At%)/S (At%) and Fe (At%)/O (At%) of intermediate species. Fe (At%)/S (At%)_pyrite=0.5; Fe (At%)/O (At%)_goethite=0.5. At: Atomic. (c) and (d)‒the % (R²=0.5625) and At% (R²=0.8217) data of S and O of intermediate species respectively. Elemental % and At% rates of intermediate species are listed in Table S2.

DownLoad: Full-Size Img PowerPoint

The abundant framboid Fe(Ⅲ) (hydrogen) oxide that was observed in multiple sections displayed a rufous color, and Fe/O (At%) was approximately 1∶2, which implied that the mineral was goethite (Fig. 7d and e). Moreover, the unaltered surface and euhedral crystals of framboid goethite (Fig. 7b and c) indicated that they formed in situ rather than originating from terrigenous detritus (Sun ZL et al., 2019), unless they transformed into anhedral hematite after undergoing long-distance transport due to their low thermal stability (Gualtieri AF and Venturelli P, 1999). In addition, rather than goethite, poorly crystallized Fe (hydrogen) oxide (mainly ferrihydrite) may preferably form during lavas alteration, and the goethite is then transformed from these Fe (hydrogen) oxide (Yokoyama T and Nakashima S, 2005). However, the framboid goethite in this core was not the product of lavas alteration based on the vast amounts of framboid pyrite and intermediate species found in the various intervals.

Interestingly, all of the intermediate species were found in or closed the upper SMTZ (Fig. 6), while both SMTZs contained Fe (hydrogen) oxide, particularly framboid goethite (Fig. 7). The reason is related to the fluctuations in the SMTZ with respect to the sedimentation rate and deep methane flux intensity. Once the methane seepage intensity of the lower SMTZ decreased, pyrite was exposed to the oxic condition and was oxidized, and intermediate species possibly formed. As discussed above, the Fe (hydrogen) oxide coating inhibited or slowed the reaction rate; therefore, such intermediate species can be preserved for many years (Morse JW, 1991; Gartman A and Luther GW, 2014). Based on AMS ¹⁴C dating, the time interval between the two active methane seeps is at least 15 kyr, and such a long duration means that any pyrite present exposed to O₂ was completely oxidized before methane seepage resumed. Moreover, the coexistence of pyrite, intermediate species and framboid goethite in the upper SMTZ revealed the pyrite oxidation process. As Fig. 9 c and d show, different degrees of pyrite oxidation occurred. The sulfur content of pyrite decreased as the oxygen content rose. When sulfur was completely eliminated from the intermediate species, Fe (hydrogen) oxide (particularly goethite) was formed.

5.4   Indication of methane seep activities in geological scale

The distribution of δ³⁴S values of CRS in this core is significantly different than that of the sediment only controlled by OSR where high sedimentation rate induces more positive δ³⁴S values and show more negative δ³⁴S values when sedimentation rate is low (Liu XT et al., 2019). Though this study excludes the possibility of only terrigenous input defining the diagenetic environment, the sedimentation rates vary following sea-level changes during MIS3 and between MIS1 and MIS2 (Rohling EJ et al., 2009; Dan XP et al., 2023) indeed impacts the position of SMTZ which determined by both sedimentation rates and methane flux. Interesting, both SMTZs determined by high contents of CRS and S/C ratio, cubic pyrite is situated in the low sea-level periods when sedimentation rate is rather high (Figs. 2a and 3a). High sedimentation rate hampers the SMTZ keeps at a certain interval and thus decreases the formation of pyrite. However, studies from South China Sea also proposed that vibrant methane seep activities also occurred in glacial periods perhaps caused by gas hydrate decomposition below (Tong HP et al., 2013; Li N et al., 2021). Therefore, the authors suggest the methane seep activity observed in this core balanced by both sedimentation rates and methane flux.

6.   Conclusion

So far, Fe-S cycle in subsided methane seep is still unclear. By utilizing various methods, this paper studied a core sampled from the Shenhu area, South China Sea. The AMS ¹⁴C dating and carbon and oxygen isotope test results of planktonic foraminifera indicated continuous sedimentation from MIS3 to MIS1. Low correlations between CRS and TOC and δ¹³C_TOC limited the dominant influence of the OSR. The rising contents of CRS as well as the mean diameter and standard deviation of framboids coupled with cubic crystals found in several intervals demonstrated the rough location of the SMTZ. The extremely negative δ³⁴S value of both CRS and pyrite that was found in both SMTZs implied that the SMTZ was once located at or near the surface of the seafloor, but the extensive disproportionation of intermediate sulfur species and sulfide reoxidation stimulated the fractionation of sulfur isotopes. Moreover, the widespread distribution of elemental sulfur in the core, particularly in the SMTZ, suggested that the SR-AOM had subsided. The presence of framboid goethite and intermediate species form during pyrite oxidation further confirmed this assumption. Additionally, the difference in the distribution of intermediate species and Fe (hydrogen) oxide (framboid gothite) indicated that a part of the pyrite was oxidized, but the rest remained. Thus, subsided methane seep is a progressing environment to interpret the Fe-S cycle in marine sediment by multiple geochemical and mineralogical results (Fig. 10).

Figure  10.  Schematic diagram of the formation, distribution and interrelationship of various minerals during the evolution of methane seep. Oxidation of pyrite in low oxygen concentration is referred to Nyavor K et al. (1996).

DownLoad: Full-Size Img PowerPoint

Supplementary data set

Supplementary data of this article can be found online at doi: 10.31035/cg20240110.