This study used thermally immature to early mature shales and coals with R_o values from 0.39 to 0.62%. Samples from various basins have different TOC contents, ages, and kerogen types of I, II, and III (Table 1). Samples were crushed into chips (size range from 1 to 5 mm) or powders (0.25–0.42 mm) for use in our experiments (Table 2). Samples sealed in rigid glass tubes were heated at ambient pressures. All rock chips were placed under vacuum at room temperature for 2 days to remove volatile hydrocarbons prior to heating experiments. Heating from 60°C to 140°C tested gas generation at temperatures similar to those typically encountered in sedimentary basins (Table 2).

All generated gas products from glass ampoules were collected using the same glass vacuum line (Figure 1) by following the principles outlined by Wei et al. (2018). After opening the glass ampoule and releasing product gases into the vacuum line, the condensable gas fraction (predominantly CO₂) and the incondensable gas fraction (including hydrocarbon gas and nitrogen gas) were separately collected. The Baratron readings from injected known volumes of pure CH₄ or CO₂ were used to calibrate pressure responses in distinct partial volumes of the vacuum line (not only to calculate gas yields in μmol but also to determine volume split ratios; Figure 1A). All gases were finally transferred to a methane receptacle with charcoal, followed by sealing off with a torch. The resulting sealed gas sample was transferred into a 10 ml/60 ml Pyrex^® culture glass bottle and subsequently crimp-sealed.

Wei et al. (2018, 2019) quantified generated gas products with an SRI Multiple Gas Analyzer #2 (Jin et al., 2010) based on gas chromatography, using a flame ionization detector (FID) to detect C₁−C₄ hydrocarbon gases and a thermal conductivity detector (TCD) to quantify H₂, O₂, N₂, CH₄, CO, and CO₂ (Figure 1A). The GC column oven was initially held at 40°C for 4 min, followed by an increase to 250°C at 30°C min⁻¹, where it was held isothermally for 19 min. The GC precision for standard gases was better than ±400 ppmv for CO₂, ±10 ppmv for CH_4, and ±4 ppmv for C₂H₆ and C₃H₈.

Ma et al. (2021) quantified generated gas product with a SARAD RTM2200 gas analyzer (SARAD GmbH, Dresden, Germany), equipping 1) a madur madIR-D01 CO₂ detector (madur electronics, Zgierz, Poland, www.madur.com) and 2) an Axetris laser OEM module LGC F200 (Axetris AG, Kägiswil, Switzerland) to quantify CH₄. A detailed analyzing procedure is described by Ma et al. (2021). The system was calibrated by injecting known volumes of pure CH₄ or CO₂ with a gas-tight syringe through polymer tubing. The overall CH₄ and CO₂ yields from headspaces in glass tubes were calculated by taking into account the known gas split ratios between small and large gas fractions that had been separated at the vacuum line (Figure 1B).

The liberation of pre-existing trapped methane from source rock chips during heating and fracturing of rock through heating and subsequently mimicking catalytically generated gas. The most commonly used method to evaluate residual gas in source rocks is by crushing in a closed environment, such as a ball mill chamber. Wei et al. (2018) assessed methane volumes released from shales and coals by crushing chips in a modified 65−mL SPEX 8000 ball mill (operated at 1080 cycles per minute) using a SPEX 8001 Hardened Steel Grinding Vial Set (http://www.spexsampleprep.com/equipment-and-accessories/accessory_product/8001) that had been fitted with a septum port to obtain gas samples from the closed mill chamber using a gas-tight syringe (Figure 1A). Pairs of 12.7- and 6.4-mm diameter alloy steel 52–100 crushing balls contained Fe, Cr, Mn, and 0.98–1.1 wt. % of C in the form of metal carbides in steel (http://www.suppliersonline.com/propertypages/52100.asp). Wei et al. (2018) demonstrated that adding deionized water to the ball mill and crush with samples together is effective for reducing the localized heating when the kinetic energy of a steel ball is converted into heat upon impact. Aliquots of 1 g of chips were loaded into the ball mill with an addition of 0.4 ml deionized water, the headspace was flushed with ultra-high purity nitrogen, and milling proceeded for 4 min, followed by methane quantification in headspace gas via GC. Wei et al. (2018) used gas-tight syringes to transfer gas aliquots from a ball mill chamber with a single septum port to a gas chromatograph to quantify CH₄ yields from source rocks after ball milling. They reported varied amounts of pre-existing gases in closed pores of original, un-heated source rocks. For example, Mowry Shale has 0.013 and 2.059 μmol g⁻¹ pre-existing CH₄ and CO₂ in rock chips before heating. They also further quantified the amount of gas released from the gas ampoules and the amount of gas in closed pores from heated source rocks under different heating conditions in the study.

Wei et al. (2019) pre-evacuated Yabulai Shale and NAS_472-RD for 3 days before long-term heating simulation experiments. Those samples were at 60 mesh (0.25–0.42 mm) and were proposed to have removed pre-existing adsorbed gas. Gas yields from those experiments did not require further correction.

Ma et al. (2021) made further efforts in ball milling and gas detection with a highly sensitive SARAD RTM2200 system that did not require any gas transfer using a syringe. They provided new evidence for a largely systematic, non-negligible contribution of pre-existing CH₄ from imperfectly closed pores. They used the source rock–specific average deficits of gases between original and heated source rocks to correct the overall gas yields. Their study concluded that the gas content in closed pores was fixed for each series of samples and used the results from the ball milling experiments to determine the average amounts of CH₄ [and CO₂] that had leaked from imperfectly closed pores. Their reported data were presented as corrected yield and uncorrected yield. The final corrected methane yields were uncorrected yields subtracting certain amounts. For each one of the four source rocks used (i.e., Springfield Coal #2, NAS_IN3, NAS_IN6, Second White Specks Fm.), the respective subtracting amounts in μmol g⁻¹ TOC are 0.05 [4.19], 0.06 [3.04], 0.02 [8.11], and 0.40 [8.96] (CH₄ yield [CO₂ yield]).

To make a better gas generation yield comparison between different studies (Wei et al., 2018; Ma et al., 2021), the amount of newly generated gas in each study is re-calculated by subtracting yields of residual gas in closed pores of original, un-heated samples from total yields of released gas from glass ampoules and residual gas in closed pores from source rock chips after heating experiments. Powdered samples of 60-mesh heated for 1 month (Wei et al., 2019) had removed pre-existing gas and could avoid the calculation above.

Metal concentrations were obtained using a method modified from the one used by Kendall et al. (2009). An aliquot of 40–50 mg of powdered, whole-rock shale was baked for 12 h in a 550°C oven. The ashed powder was then treated with 5 ml of HNO₃ and 1 ml of concentrated hydrofluoric acid (HF) and placed on a 200°C hot plate for 10 h. Due to moderated organic content in samples, repeated HNO3-HF treatment was performed to fully dissolve the organic matter. Samples were dried and treated with HCl after dissolution and were diluted with HNO₃. Metal abundances were measured by inductively coupled plasma-mass spectrometry (ICP-MS) against multiple-element calibration standards. Analyte concentration reproducibility was better than 5%.

Researchers have proposed and proved the existence of catalytic gas generation eluding biogenic gas contribution using in-lab low-temperature long time heating simulation experiments in recent years (Carothers and Kharaka, 1978; Mango, 1996; Rowe and Muehlenbachs, 1999; Tilley and Muehlenbachs, 2006; Xu et al., 2008; Mango and Jarvie, 2009; Wei et al., 2018; Wei et al., 2019; Ma et al., 2021). Wei et al. (2018, 2019) used different coal and shale samples and generated CH₄ and CO₂ gas from 60°C to 140°C based on a series of 1–24 months of heating experiments. Ma et al. (2021) adjusted and improved the gas yield measurement method with additional source rock samples using similar heating experiments. However, pre-existing trapped methane may become liberated during heating and fracturing of rock at high pressure and subsequently mimic catalytically generated gas. Extended evacuation of rocks chips before heating experiment at room temperature cannot remove all pre-existing gas trapped in closed pores. Previous studies reported three types of methane amount, i.e., the amounts of pre-existing gases in closed pores of original (before the heating experiment), un-heated source rocks, the amounts of gases in closed pores of heated source rocks (after the heating experiment), and the amounts of gases released from gas ampoules (after the heating experiment).

The yield of methane from Mowry Shale in Wei et al. (2018)’s experiment is much lower compared to that in the research conducted by Mango and Jarvie (2009) using a short-term heating experiment. Mango and Jarvie (2009) heated Mowry Shale (60-mesh) to 100°C for 24 min in a flow-through reactor. The helium effluent carried a cumulative yield of 25 μg g⁻¹ C₁–C₅ hydrocarbons claimed to be of catalytic origin. Using the same heating apparatus at 50°C with Floyd Shale, Mango and Jarvie (2009) observed a chaotic pattern of distinct methane pulses eluting over time that was interpreted as a characteristic hallmark of catalysis. We concluded that no single microscopic catalytic domain in shale can spontaneously generate enough methane to emit a gas pulse that is detectable with a GC. A more likely explanation for the observed methane pulses from Mowry Shale is the developing mechanical stress within Mowry Shale during exposure to ultrapure helium. The diffusive influx of helium through narrow throats into pores containing pre-existing methane causes pore overpressure.

The method used to evaluate residual gas in source rocks is by crushing in a closed environment, such as a ball mill chamber. Water was present in the ball milling experiments as a coolant to limit the thermal stress during steel ball impacts. Wei et al. (2018) first claimed that dry ball milling experiments greatly introduced inorganic methane from metal carbides from the steel surface of the ball mill chamber, and moist ball milling experiments collected methane yields of 0.021 μmol g⁻¹ can be accounted for by a reaction of metal carbides. Ma et al. (2021) proposed that the absence of any additional solid substrate did not prevent abrasion as steel balls impacted the steel walls and generated 0.05 μmol g⁻¹ CH₄ and 0.24 μmol g⁻¹ CO₂. The differences could be accounted to the cleanness of rock powder attached to the ball mill chamber between runs, the gas collection and transfer process between different runs, and the sensitivity between the SARAD RTM2200 system and traditional gas chromatography. Although varied quantification of pre-existing gas amount exists between the two studies, the real catalytic gas generation amount could be compared by simply mass balance equation. In this study, we chose to effectively compare newly catalytic gas generation yields between those two studies (Wei et al., 2018; Ma et al., 2021) with the following: the amount of catalytic generated gas equals the total amount of liberated gas yields from glass ampoules and residual gas yield in closed pores from source rock chips after heating experiments deducted yields of residual gas in closed pores of original, un-heated samples.

Wei et al. (2018) conducted low-temperature (60 and 100°C) heating experiments lasting 6 months to 1 year and provided evidence for the geo-catalytic generation of hydrocarbons from various source rocks at different hydrostatic pressures. In addition, they performed another suite of low-temperature (100 and 140°C) heating experiments from two types of source rock lasting 1 month and quantified generated CH₄ and CO₂ gas yield. Ma et al. (2021) determined the overall yield of CH₄ generated from heating experiments at 80°C, 100°C, and 120°C from source rocks of different kerogen types lasting more than 14 months. Our study re-quantified final catalytic gas generation yields using the total yield of released gas from glass ampoules and residual gas in closed pores from source rock chips after heating experiments deducting yields of residual gas in closed pores of the original, un-heated samples. Gas generation yields from those different heating experiments were re-organized and presented in Table 3 and Figure 2.

The comparison shows that methane generation yields are highest from Second White Specks Formation Shale, Mowry Shale, and Springfield Coal #2 (Figure 2). Four NAS samples were used, covering TOC content from 1.2% to 15.1%, and their CH₄ yields were comparable. Methane yields did not express direct correlations with kerogen type, TOC content, or initial thermal maturity. Increasing heating time limitedly increased gas generation yields. For one specific source rock type, CH₄ and CO₂ yields show a certain amount of increase with increased temperature (60–120°C). For example, the CH₄ amount in Mowry Shale, Springfield Coal #1, Wilcox Lignite, Mahogany Shale, and NAS_MM3 shows an increasing trend with elevated temperature from 60 to 100°C for 12 months. Methane yields released from Second White Specks Formation Shale, NAS_IN3 and NAS_IN6 also increased with heating temperature from 80°C to 100°C and 120°C.

Gas yields are expressed in μmol g⁻¹ of TOC to emphasize the relation to the organic matter in source rocks (Table 3; Figure 2). In the studied shale samples, the Mowry Shale and Second White Specks Formation liberated the highest methane yields in comparison to other studied source rocks at comparable conditions, even at 60 and 80°C, but especially at 100°C over 12 and 38 months. Samples of Mahogany Shale and NAS produced small amounts of CH₄ even at 120°C over 36 months. Generally, NAS and Mahogany Shale produced comparable methane amounts at similar temperature and time conditions. Samples of NAS_MM3 generated higher methane yields compared to Springfield #1 Coal and Wilcox Lignite at both 60°C and 100°C over 12 months. In addition, NAS_472-RD generated methane yields of ∼0.20 and ∼0.90 μmol g⁻¹ TOC, which is ∼20 and ∼45 times higher compared with methane yields from Yabulai Shale of ∼0.01 and ∼0.02 μmol g⁻¹ TOC at 60 and 100°C over 1 month, respectively. We also noted that Wei et al. (2018) reported CH₄ yields of only ∼0.13–0.26 μmol per g TOC from Springfield Coal #1 at 100°C heating for 12 months, which is much lower than those reported for Springfield Coal #2 in Ma et al.’s (2021) study–CH₄ yields of ∼2.2–2.8 μmol per g TOC after heating for 36 months at 100°C (Figure 2; Table 3). The increased heating time could not fully explain such an obvious yield difference.

Carbon dioxide is another important product liberated from the series of low-temperature heating experiments. Most coals and all shales produced much more CO₂ than CH₄. Among generated gas products, carbon dioxide (CO₂) is at a relative abundance of ∼96.3–99.9 mol%, or about 30 to 10,000 times higher concentrations than all gaseous hydrocarbons combined at 100°C over 12 months of experiments. Compared to other shales, shale from the Second White Specks Formation generated large amounts of both CH₄ and CO₂ at 80 and 100°C over 38 months. Mowry and NAS generated higher amounts of CO₂ than Mahogany Shale by ∼6–15 times. Results from samples of Springfield Coal #2, Springfield Coal #1, and Mahogany Shale generated a relatively low amount of CO₂, especially in Springfield Coal #2 at 80°C, 100°C, and 120°C over 36 months. Another result worth noticing is that the CO₂ yields for NAS_IN3 are higher at 100°C heating conditions than those at 80°C and 120°C over 36 months of experiments (Figure 2; Table 3).

Maceral compositions were analyzed and compared within studied source rocks to investigate whether there is a relationship between maceral content and CH₄ and CO₂ yields (Table 4). Our studied coal samples include Springfield Coal #1, Springfield #2, and Wilcox Lignite. The dominant macerals in coal samples are sporinite, vitrinite, and inertinite, but their relative proportions differ between the samples (Table 4). Springfield Coal #1 contains high contents of vitrinite (82 vol. %) and inertinite (12 vol. %) and a small amount of liptinite (mostly sporinite; 6 vol.%) (Figures 3A–D). Springfield Coal #2 has a lower content of vitrinite (72 vol. %) and inertinite (12 vol. %) and a higher amount of liptinite (mostly sporinite; 16 vol.%) (Figures 3E,F). Maceral in Wilcox Liginite is composed of vitrinite (82 vol. %), inertinite (10 vol. %), and a small portion of liptinite (sporinite and liptodetrinite; 8 vol. %) (Figures 3G,H). Similar maceral composition in Springfield Coal #1 and Wilcox Lignite indicates the limited contribution of vitrinite and inertinite to catalytic methane generation (Table 4). The higher liptinite content in Springfield Coal #2 obviously enhanced catalytic methane yield (Figure 8A).

For the shale samples, the dominant maceral group is liptinite regarded as an oil-prone maceral during thermal maturation. The macerals in studied shales are mainly composed of the liptinite group represented by alginite, AOM, and liptodetrinite, while the relative content of those varied between shale samples. The representative maceral compositions of samples from NAS_472-RD and Yabulai Shale are shown in Table 4. Maceral in Yabulai Shale is composed of liptinite (mostly alginite; 69 vol. %), solid bitumen (10 vol. %), vitrinite (30 vol. %), and inertinite (10 vol. %) (Figure 4B). NAS_472-RD dominantly contains liptinite (Figure 4A). Higher liptinite content facilitated catalytic methanogenesis in NAS_472-RD compared to Yabulai Shale at the same heating time and temperature conditions (Table 4).

Liptinite in samples from the Second White Specks Formation is composed dominantly of AOM and a very low amount of well-preserved alginite (Furmann et al., 2015), in contrast to the NAS_472-RD and Mahogany Shale where alginite (Tasmanites) is the main component. More specifically, Mowry Shale and Second White Specks Formation Shale both contain relatively high content of AOM (30% and 64%) derived from microbially degraded phytoplankton, zooplankton, and bacterial biomass, which appears structureless and occurs dispersed in the matrix or as organic streaks parallel to bedding (Table 4; Figures 5A–D). Besides AOM, Second White Specks Formation Shale and Mowry Shale contain 15% and 35% (vol. %) alginite, respectively, occurring as elongated rods perpendicular to the bedding and as flattened disks parallel to the bedding (Figures 5B,E). Some form discrete lenses or oval-shaped bodies with distinctive internal structures (telalginite, Figure 5F). The alginite shows strong greenish-yellow to yellow fluorescence under blue light irradiation. In addition, Mowry Shale contains relatively more terrestrially-derived material (vitrinite and inertinite; Table 4 and Figure 5A). The CH₄ yields from NAS and Mahogany Shale are relatively lower. NAS_472-RD and Mahogany are full of alginite, especially tasmanites occurring as discrete lenses (Figures 6A–D, 7; Table 4); vitrinite and inertinite are rare. Alginite occurs as lamellae typically <10 μm in thickness, and cell structures are present in some occurrences (Figures 6A,B). NAS_472-RD generated two times higher methane yields than NAS_MM3 at 100°C over 12 months of heating. Maceral composition in NAS_MM3 featured relatively averaged amounts of alginite (28 vol. %), AOM (19 vol. %), other liptinite (19 vol. %) and solid bitumen (33 vol. %), while NAS_472-RD contained 25 vol. % of AOM and 65 vol. % of alginite. Comparing the AOM content within studied shale samples, higher AOM content in the liptinite group positively correlated with higher methane yields (Figure 8B).

Organic matter is the parent material for CH₄ and CO₂ products during thermogenic gas generation. Macerals are divided into oil-prone and gas-prone, showing the quality of a source rock whether makes it more likely to generate oil or gas. Previous studies indicate the liptinite group as oil-prone maceral presents heterogeneity in the oil generation window and potential during maturation. Resinite has the highest oil-generation potential and is usually attributed to early oil generation (R_o ∼0.3–0.4%). Alginite comprises unicellular solitary or colonial algae of planktonic and benthic origin. AOM occurs as a product of variable organic matter which has undergone alteration and degradation. Liu et al. (2019a) conclude that AOM transformed to hydrocarbons earlier than alginite. Alginite derived from tasmanites cysts has higher hydrocarbon generation potential (Revill et al., 1994; Vigran et al., 2008). Cardott et al. (2015) also suggested that a post-oil solid bitumen network could have developed along with the AOM network. Vitrinite, as a gas-prone maceral, generates gas instead of oil, although most thermogenic gases are concluded from oil-cracking instead of kerogen. Hydrous pyrolysis simulation experiments (>350°C) usually indicate that the liptinite group have a higher hydrocarbon generation potential than vitrinite, including both oil and gas. Thermogenic gas generation potential from source rock is strongly controlled by kerogen type. This study further suggests that source rocks of the same kerogen type have strong heterogeneity in catalytic gas generation ability because of their specific maceral composition. Liptinite facilitates methane generation at low-temperature conditions. Moreover, alginite and AOM are the dominant liptinite maceral in marine shales, and their content can vary significantly in shales. Liu et al. (2019) proposed alginite (tasmanites in particular) is more abundant than AOM for NAS in high-energy environments. Our results proposed that AOM undergoes a transformation into gas earlier than alginite for catalytic methanogenesis, i.e., AOM is much more prone to gas than alginite in immature to marginally mature shales. Higher sporinite input also makes a positive contribution to catalytic methane generation. Vitrinite, inertinite, and solid bitumen did not show obvious differences in liberating methane gas at low temperatures.

The most dominant gas product is CO₂, which could be 30 to 10,000 times higher concentrations than hydrocarbon products in studied samples. Liptinite has a more aliphatic structure, hydrogen-rich and lower oxygen content than vitrinite (Walker and Mastalerz, 2004). Springfield Coal #2 contained more maceral liptinite than Springfield Coal #1 and generated much lower CO₂ at comparable heating temperature and longer time, which indicate that vitrinite or inertinite is easly able to liberate CO₂ upon heating than the liptinite group. In addition, higher CO₂ yields from Mowry and NAS with more terrestrial input (vitrinite, inertinite) than Mahogany Shale also proved the same point. The maceral composition can at least partly explain methane yield differences between samples. However, the related factors are still complex and need a better illustration explaining the catalytic reaction mechanism.

Various literature reported economically viable natural gas accumulations in low-rank source rocks. What is worth noticing is that most published literature about catalytic gas were generated from humic organic matter. Galimov (1988) reported that “early thermogenic” natural gas in the Western Siberian Basin comes from humic-type organic matter. Xu et al. (2008) proposed that Jurassic coal was generated and accumulated in the gas reservoir in Turpan-Harmi Basin. Wang et al. (2005) reported immature to low mature natural gas in shallow reservoirs (<2500 m) in Liaohe and Subei Basins. Ramaswamy (2002) concluded natural gas accumulation in the North Cambay Basin in India comes from Mehsana Coal. In Qong Southeast Basin in the South China Sea, the fluid inclusion data shows (80°C–120°C) fluid inclusions providing evidence of low mature gases migrating into the reservoirs of BD 13 gas-bearing structure. The humic-type organic matters from Oligocene-Miocene source rocks (burial depth of 1570–2000 m) have a certain potential to generate biogenetic gas and low mature gas (Huang et al., 2003; Ding et al., 2018). Those reported R_o values mostly range from 0.3 to 0.5%, and reported δ ¹³C_CH4 values are from −55‰ to −40%. Gases are mainly composed of CH₄ (82.48%–86.14%) with relatively high C_2–5 components (about 10%), relatively low CO₂ (0.6%–1.2%) and N₂ (2.4%–6.7%) content, and high dry coefficients (C₁/C_1–5 = 0.88–0.91). Fewer examples of low maturity shale strata in sedimentary basins featuring non-microbially generated hydrocarbon gas accumulations were found. For example, Muscio et al. (1994) proposed that immature shales released high amounts of natural gas in Williston Basin in North America with δ ¹³ C_CH4 values of ∼ –33.7‰. Cardott. (2012) investigated natural gas in western Arkoma Basin in Oklahoma in America from immature Woodford Shale with δ ¹³ C_CH4 values of –52.8‰. In addition, the New Albany Shale from the Illinois basin, Mowry Shale Formation, and Second White Specks Formation in our study are mainly explored as a shale oil play in sedimentary basins (Clarkson and Pedersen, 2011). Wei et al. (2018) pointed out that considering that the New Albany Shale has been at its current depth and thermal regime for approximately 90 million years (Wei et al., 2018). In this way, today’s lack of abundant CH₄ in those basins testifies against significant geocatalytic methanogenesis. The reasonable explanation is that the liptinite group significantly contributes to catalytic gas generation. However, the alginite and AOM enriched shales turned to generate and expelled oil when maturity reached the oil window, while the gas-prone enriched source rock turned to continue generating thermogenic gas. Whether catalytic methanogenesis can be a significant source for natural gas accumulation is not simply related to the kerogen type of source rock but more related to the relative content of specific maceral content in shales and coals. In addition, previous studies (Wei et al., 2018) observed elevated pressures of 100 and 300 MPa caused gas yield reduction, and we proposed that lower source rock deposit with proper depth and good seal positively contributes to low-mature natural gas accumulation.

Studies on catalytic gas have indicated a high yield of 96.3–99.9 mol% carbon dioxide in gas from low-temperature heating experiments (Wei et al., 2018; Ma et al., 2021). The abundance of carbon dioxide in natural gas is highly variable yet often amounts to ∼10 mol% (Lundegard and Land, 1986; Jenden et al., 1988; Smith and Ehrenberg, 1989) and much lower than in our low-temperature heating experiments. In contrast to the confined conditions in gold cells and glass tubes, the natural conditions in sedimentary basins apparently allow carbon dioxide to be removed via secondary processes, such as preferential dissolution in aqueous pore fluids and subsequent migration and precipitation as carbonate minerals (Ferris et al., 1994; House et al., 2006; Gilfillan et al., 2009). In addition, the long exposure to air during storage of source rocks in core lockers may have caused surficial oxygenation of organic and inorganic moieties that gave rise to CO₂ instead of methane.

Results of metal concentration analysis are available in Table 5. Typical variations of sensitive trace metals in NAS_472-RD, NAS_MM3, and Mowry Shale, including Cr, Fe, Ni, Cu, Zn, Cd, Ba, and Pb, are presented. Methane yields differences between studied shale samples did not exhibit obvious relationships with trace metal concentration. Considering the proposed hypotheses for catalytic gas generation, the reduced metallic catalyst content lacking a relationship with methane yields in studied samples thus denied their contribution to aiding catalytic methanogenesis. Both coals used have comparable sulfur contents (∼4 wt.%), thus the difference in gas yields cannot be explained by sulfur-induced lower activation energy for hydrocarbon generation in one coal compared to the other (Ma et al., 2021). In addition, certain types of macerals undergo easier and earlier bond breaking to generate light hydrocarbons. Rahman et al. (2018) conducted a study on source fabric’s influence on clay mineral catalysis as it controls the extent to which organic matter and clay minerals are physically associated. They used 1) a particulate fabric where organic matter (alginite) is present as discrete (>5 μm; sample from the Permian Stuart Range Formation) and 2) a nanocomposite fabric in which AOM is associated with clay mineral surfaces at the sub-micron scale (sample from the Miocene Monterey Formation). Kinetic experiments are performed on paired whole rock and kerogen isolate samples from these two formations. More specifically, no significant difference in the modelled hydrocarbon generation window of paired whole rock and kerogen isolates from the Stuart Range Formation. Extrapolation to a modelled geological heating rate shows a 20°C reduction in the onset temperature of hydrocarbon generation in the Monterey Formation whole-rock samples relative to paired kerogen isolates. The extent to which organic matter and clay minerals are physically associated could have a significant effect on the timing of hydrocarbon generation (Rahman et al., 2018). Following their proposed idea, we also considered the organic matter and clay minerals contact extent in our samples, especially shale samples composed mainly of the liptinite group. Kus et al. (2017) concluded several identification rules distinguishing the AOM and alginite. Alginite presents the optically identifiable and recognizable lamellar structure, and isolated forms clearly distinguishable from one another and from the mineral groundmass. In comparison, the AOM, generally considered to be a degradation product, usually presents thin, small lenses and irregular stripers, porous, was also observed to occur in groundmass having close contact with clay minerals (Taylor et al., 1998). Although liptinite expresses a higher potential for methanogenesis, the differences between AOM and alginite generating catalytic gas could also be related to their contact area with surrounding minerals. New Albany Shale and Mahogany Shale generating lower hydrocarbon gas are characterized as relatively large, discrete organic particles of alginite (tasmanites in particular) and solid bitumen having limited contact with the mineral matrix and the clay mineral. In comparison, Mowry and Second White Specks Formation Shale with higher AOM debris content associated with clay mineral surface intimately which lower its catalytic reaction. However, further quantitative evidence of OM and clay mineral contact area is needed before reaching a solid mechanism explanation.