Thermochemical oxidation of methane induced by high-valence metal oxides in a sedimentary basin

Thermochemical oxidation of methane (TOM) by high-valence metal oxides in geological systems and its potential role as a methane sink remain poorly understood. Here we present evidence of TOM induced by high-valence metal oxides in the Junggar Basin, located in northwestern China. During diagenesis, methane from deeper source strata is abiotically oxidized by high-valence Mn(Fe) oxides at 90 to 135 °C, releasing 13C-depleted CO2, soluble Mn2+ and Fe2+. Mn generally plays the dominant role compared to Fe, due to its lower Gibbs free energy increment during oxidation. Both CO2 and metal ions are then incorporated into authigenic calcites, which are characterized by extremely negative δ13C values (−70 to −22.5‰) and high Mn content (average MnO = 5 wt.%). We estimate that as much as 1224 Tg of methane could be oxidized in the study area. TOM is unfavorable for gas accumulation but may act as a major methane sink in the deep crustal carbon cycle.

mudstone (interchannel flood deposits) 6 . Calcareous mudstones and sandstones, and playa sediments such as dolomite and gypsum do not occur in T1b and adjacent formations. These features distinguish the T1b Formation from the non-marine Triassic redbeds of NW Europe and the Triassic-Jurassic redbed sequences of North America, which comprise calcretes and playa lake sediments [12][13][14][15] .
Successive global atmospheric-CO2 greenhouse episodes with abnormally high temperatures occurred during the Early Triassic 16,17 , and the T1b Formation contains sedimentary assemblages that are typically produced by debris flows, sheet floods, and high-energy stream flows, indicating that it was deposited in an arid-semiarid paleoclimate 10 . The unusually warm paleoclimate might have resulted in the depletion of organic matter during deposition, as recorded by abundant subsurface brown mudstones and conglomerates of the formation ( Supplementary Fig. 2).
Shan et al. 18 investigated paleo-current directions based on analyses of dip-meter logging data, indicating that the provenance of the T1b Formation includes paleo-uplifts along the northwestern margin of the basin. Further petrological analysis indicates that conglomerates in the study area are composed mainly of clasts of tuff, sedimentary rock, and granite, while coarse-grained sandy components are dominated by quartz, feldspar, and granite rock fragments 10,18 . There was no carbonate supply to the sediments of the T1b Formation. These observations indicate that sediments of the T1b Formation were derived from two provenance areas with different rock types, namely the basin basement comprising granite and mafic-ultramafic igneous rocks, and underlying Carboniferous to Permian sedimentary sequences 2 .
The weathering of mafic-ultramafic igneous rocks in the basin basement, and tuffaceous components in the Carboniferous and Permian sedimentary rocks, formed abundant high-valence Fe-Mn oxides before deposition of the T1b Formation.

Supplementary Note 3: Occurrence of high-valence Mn(Fe) oxides
As mentioned Supplementary Note 2, the T1b Formation comprises conglomerate, sandstone, and mudstone. The gravel-sized clasts of the conglomerate are mainly mafic-ultramafic tuff, sedimentary rock, and granite, while the sand-sized grains are dominated by quartz, feldspar, and granite rock fragments 6,10,18 . The matrix among the conglomeratic and sandy grains is mainly smectite and mixed-layer illite/smectite (I/S), with minor kaolinite and hematite ( Supplementary Fig.   3a). In the mudstones, silty quartz, tuff, and granite rock fragments occur with various clay minerals such as smectite and I/S. Minor hematite is disseminated throughout the clay. Common heavy minerals (e.g., ilmenite, leucosphenite, rutile, and epidote) also occur in various rocks in the T1b Manganese valence was investigated by in situ X-ray photoelectron spectroscopy (XPS). Useful Mn 2p and 3s signals were not obtained due to the low Mn content 19,20 , but the reliable Fe 2p spectra indicate that Fe is trivalent in hematite ( Supplementary Fig. 3d). The reduction potentials of Mn 3+ and Mn 4+ are higher than that of Fe 3+ 21, 22 , so both Mn 3+ and Mn 4+ can oxidize Fe 2+ to Fe 3+ . Therefore, in the presence of Fe 3+ , Mn generally occurs as an isomorphous Mn 3+/4+ substitution in hematite [23][24][25][26] . Where Fe 2+ is present in the chlorite, Mn is expected to be bivalent as Fe 2+ substitution 27 .
In weakly acidic fluid, Mn 3+/4+ is released into formation water through the dissolution of hematite 28,29 . The reduction of Mn 3+/4+ also promotes its dissolution and release. However, aqueous Mn 3+ is unstable and tends to undergo disproportionation into Mn 2+ and MnO2 29,30 . In contrast to insoluble MnO2 (oxidation potential, Eh = 1.224 V), aqueous Mn 3+ (Eh = 1.542V) is characterized by a stronger oxidization capacity 21 . This may be why methane can be oxidized at such low temperatures.

Supplementary Note 4: Burial history and Diagenesis
Tectonic activity has not resulted in significant uplift of the Mahu Sag since the Early Triassic 8 . Fig. 4).

The depth of the Baikouquan Formation generally increased after deposition (Supplementary
However, the geothermal gradient decreased gradually during geological history, from 36.3 °C km -1 in the late Permian to 33.8°C km -1 in the Late Triassic, 28.4 °C km -1 in the Late Jurassic, 24.8 °C km -1 in the Late Cretaceous, and eventually dropping to 22.8 °C km -1 in the Neogene, which is similar to the present geothermal gradient 22 . The thermal evolution of the T1b Formation is indicated by its geothermal gradient and burial history. During the first period of methane charging (J1), the formation temperature had reached above 83 °C, and by the second charging of reservoir K1 it had reached almost 100 °C ( Supplementary Fig. 4).
After deposition, the early diagenesis had major influences on the supply of high-valence Mn in the T1b Formation after deposition. The increasing pressure of overlying sediments resulted in compaction, promoting the mechanical infiltration of clay-rich pore water and the formation of clay grain coatings. The clays are presently illite or mixed-layers I/S 6 , although they are likely to have been more smectite-rich upon deposition. Dissolution of labile detrital grains, such as anorthite components in plagioclase, lithic fragments, and ferromagnesian grains, would have been possible at that time 31,32 . Ferric iron derived from dissolution of iron-bearing grains or clay, precipitated as hematites through the dehydration of hydroxides, and reddening the rocks 31 After charging by oil-and gas-bearing fluids, the high-valence Mn reacted to oxidize CH4 to CO2 at temperatures above 80 °C. Organic acids and the generated CO2 in petroleum-bearing fluids may also have promoted the dissolution of orthoclase 6 , and the dissolution of detrital feldspars would have released substantial amounts of Ca 2+ into the formation water. In some reservoir beds of the T1b Formation, Mn-rich calcites then precipitated in the weakly alkaline fluid environment, along with kaolinite. During the two stages of reservoir oil and gas charging in the Jurassic and Cretaceous, respectively, the formation temperature of the latter was obviously higher than that of the former. As a higher temperature is favorable for the thermochemical oxidation of methane, more high-valence Mn could have been reduced to Mn 2+ by methane during the later stage, with the Mn content of late-stage calcite cements thus being higher than that of early-stage calcites. Supplementary Fig. 1. Typical oil and gas reservoir profile showing the combination of source rock, reservoir, and cap rocks in the study area. Oil and gas migrated into T1b along steeply dipping faults.

Supplementary Tables
Supplementary Table 1 In situ δ 13 C and δ 18   "bdl"denotes below detection limit.    a) Fe2O3 T, total iron as Fe2O3., b) Mn2O3T, total manganese as Mn2O3. LOI is the weight loss on ignition.

Supplementary Table 2 Composition and C1-C4 δ 13 C of natural gas from
Supplementary