In this study, a methane (CH4) cracking experiment in the temperature range of 425–800°C is presented. The experimental result shows that there are some alkane and alkene generation during CH4 cracking, in addition to hydrogen (H2). Moreover, the hydrocarbon gas displays carbon isotopic reversal () below 700°C, while solid carbon appears on the inner wall of the gold tube above 700°C. The variation in experimental products (including gas and solid carbon) with increasing temperature suggests that CH4 does not crack into carbon and H2 directly during its cracking, but first cracks into methyl (CH3) and proton (H+) groups. CH3 shares depleted 13C for preferential bond cleavage in 12C–H rather than 13C–H. CH3 combination leads to depletion of 13C in heavy gas and further causes the carbon isotopic reversal () of hydrocarbon gas. Geological analysis of the experimental data indicates that the amount of heavy gas formed by the combination of CH3 from CH4 early cracking and with depleted 13C is so little that can be masked by the bulk heavy gas from organic matter (OM) and with enriched 13C at . Thus, natural gas shows normal isotope distribution () in this maturity stage. CH3 combination (or CH4 polymerization) intensifies on exhaustion gas generation from OM in the maturity range of . Therefore, the carbon isotopic reversal of natural gas appears at the overmature stage. CH4 polymerization is a possible mechanism for carbon isotopic reversal of overmature natural gas. The experimental results indicate that although CH4 might have start cracking at , but it cracks substantially above 6.0% in actual geological settings.

1. Introduction

The notable development in natural gas exploration in the 21st century is the significant discovery of shale gas. Furthermore, some new geochemical and geological characteristics of shale gas have been encountered during exploration, such as the isotopic rollover and/or isotopic reversal of shale gas at very high thermal maturity levels. These features have also been observed in coal-derived tight gas [13]. In recent years, substantial research has been conducted to unravel the mechanism of isotopic rollover and reversal of overmature natural gas. However, the question is still not fully understood. Some researchers have suggested that the isotopic reversal of shale gas at overmature stages is caused by mixing of primary gas from kerogen cracking and secondary gas from the cracking of remained hydrocarbons [46], while others proposed that the redox reactions between water and CH4 at 250-300°C generate isotopically light carbon dioxide and hydrogen in the overmature phase, which further interacts to form isotopically light ethane (C2H6) and finally causes the isotopic reversal of shale gas [7, 8].

Based on oil-prone kerogen pyrolysis experiments under hydrous and anhydrous conditions, Gao et al. [9] and Sun et al. [10] illustrated that mixing between primary and cracking gases could not yield isotopic reversal, only isotopic rollover. Therefore, the mixing mechanism for isotopic reversal is not experimentally supported. Xia et al. [6] proposed that the mixing between two gas samples with organic origin can lead to carbon isotopic reversal. However, one end member involved in the mixing model is very dry gas and another is very wet gas. Generally, this phenomenon of the mixing of very dry gas and wet gas does not cooccur in shale system, especially in one with a higher maturity than 2.5% [10]. Thus, mixing cannot be the real mechanism for isotopic reversal of overmature gas. The alternate view on the abnormal isotopic characteristic is impossible in theory. As geological temperature reaches 250–300°C at which CH4 could react with water, C2H6 would react with water prior to CH4 owing to the comparative lower activation energy. The reaction between C2H6 and water would cause residual C2H6 to bear heavy carbon isotope. Thus, the carbon isotopic reversal could not occur.

Natural gas with the isotopic reversal feature was originally considered to be of abiogenic origin before shale gas with such feature was found. The abnormal isotopic feature of abiogenic gas was thought to be caused by polymerization (or combination) of CH4 formed by Fischer–Tropsch synthesis between H2 and CO2 (or CO) ([1113]. Zeng et al. [3] proposed that the carbon isotopic reversal for the deep layer natural gas developed in the Xujiaweizi fault depression, in the Songliao Basin, China, was attributed to the polymerization of CH4 originating from overmature coal formation source rock. However, the theoretical view of CH4 polymerization causing the isotopic reversal for abiotic gas or organic gas has not been proven experimentally. Nevertheless, as the direct synthesis of higher hydrocarbons by using CH4 and metallic catalysts is possible in modern petrochemical industry [9, 14, 15]), the natural formation of heavy hydrocarbon gas by CH4 polymerization may also be possible. Cheng et al. [16] proved that heavy hydrocarbon gases could generate in isothermal cracking experiments of CH4, CH4+Ni, and CH4+shale conducted in gold tube at 450_580°C. Moreover, three gaseous products in the fifteen groups of CH4 cracking experiments at different temperature showed isotopic reversal phenomenon.

Therefore, it is important to prove in lab whether CH4 polymerization can form heavy hydrocarbon gas depleted in δ13C, and it is necessary to determine the geological conditions under which CH4 polymerization occurs. In this study, a cracking experiment of mixed gas of CH4 and N2 is presented. The result shows that several hydrocarbon gaseous components and H2 are generated. Moreover, the hydrocarbon gases show carbon isotopic reversal before substantial CH4 cracking. Therefore, we suggest that the combination of methyl from CH4 early cracking is a possible mechanism for carbon isotopic reversal in overmature natural gas.

2. Sample and Experiment

2.1. Sample

To quantify the content of CH4 in products at different temperatures, N2 reference gas was mixed with CH4. N2 was selected as reference gas in this experiment as it is the most stable diatomic molecule with a decomposition temperature more than 3000°C. Therefore, it could not crack at the highest temperature used in this experiment (800°C). The gas mixture was purchased from Zhaoge Gas Company, Beijing, China, and the CH4 and N2 () mixing ratio was 89.67% and 10.33%. The carbon isotopic value of CH4 was -26.07‰, and no other hydrocarbon components were detected in the gas mixture by gas chromatography (GC).

2.2. Experiment

The experiments were performed in a gold tube with a length, outer diameter, and thickness of 100, 5.5, and 0.25 mm, respectively. Prior to loading the gas mixture, one end of the tube was sealed by argon-arc welding. To load gas into the gold tube, a special device was developed to connect a vacuum pump and the open tube mouth. In a typical gas loading procedure, the air in the tube was first removed by the vacuum pump to a pressure less than 2 kPa. Then, the open mouth of the tube was closed by a pincer after the mixing gas was injected into the tube from a gas cylinder. Finally, the open end of tube was sealed by argon-arc welding while the tube was immersed in ice water. The pressure of gas injected in the tube was approximately 4 atm in this experiment, which was monitored via a pressure gauge on the gas cylinder. The exact mass of gas sealed in the tube was accurately calculated by subtracting the mass of the tube before gas loading (Table 1).

The gold tube loaded with gas was put into reactor autoclave, and the temperature was controlled by a computer. Each reactor had an independent temperature controller, and all reactors were kept at a constant pressure of 50 MPa during the course of the experiment. Experimental temperature was programmed as follows: the tube was first heated from 20°C to 300°C in 1 h and held for 30 min and then heated at a heating rate of 20°C/h to the target temperature. After the target temperature was reached, the autoclave heating was stopped, and tubes were moved for analysis at room temperature.

Duplicate gold tubes were used at every temperature point in case of breakage. Moreover, additional tubes were also used as replicates for gas analysis and quantification.

Identification and quantification of individual hydrocarbon and nonhydrocarbon gas components were carried out using a two-channel Wasson-Agilent 7890 series gas chromatograph. The heating program for the GC oven heated from room temperature to 68°C (held isothermal for 7 min), then to 90°C (at a rate of 10°C/min and then held isothermal for 1.5 min.), and finally to 175°C (at a rate of 15°C/min and then held isothermal for 1.5 min). An external standard was used for the chromatographic calibration. Certified gas standards were prepared at a precision of better than ±0.1 mol% for each component made by BAPB Inc., United States.

The stable carbon isotopes of the hydrocarbon gases were determined by using an Isochrom II GC-IRMS coupled with a Poraplot Q column with helium as the carrier gas. The heating started from an initial temperature at 30°C (isothermal for 3 min) with heating at 15°C/min to 150°C and held isothermal for 8 min. The measurement of δ13C1 was relatively easy, and 8–10 μl gaseous products were enough for the δ13C1 measurement as a high concentration of CH4 was present in all gaseous products. To guarantee the accuracy of δ13C2, more than 400 μl of gas was injected to the inlet of GC-IRMS to maintain the height of the δ13C2 peak more than half of the reference peak (flat peak in the left of Figure 1) for the low C2H6 content in gaseous products. If the height of the δ13C2 peak was less than half of the reference height, the value of δ13C2 might have been inaccurate and was thus discarded. The measurement of δ13C for each component was repeated at least twice to ensure that the error of each component was less than 0.3‰. The δ13C values for each component presented in this work were the average value of two measurements. Trace carbon isotopes of other hydrocarbon gases were not measured in this study.

3. Experimental Results

3.1. Gas Components

The gaseous products of CH4 cracking experiments at different temperatures are summarized in Table 2. Besides the original components (CH4 and N2) in gaseous products, heavy hydrocarbon gases (including alkane and alkene) and H2 were generated during the experiment at different temperatures. The CH4 content clearly decreased with increasing temperature (Figure 2(a)). The CH4 content decreased by only 1.26% at 700°C, reducing further by 4.91% from 725 to 800°C. Responding to the reduction in CH4 content, N2 content increased slightly (0.43%) at 700°C; however, it increased by 4.77% from 725 to 800°C. The percentages of C2H6 and H2 were high in newly formed components and in the same order of magnitude (×10-1). The content of C3H8 was second highest (×10-2). Other hydrocarbon gases were only presented in trace amounts (×10-3). The content of all heavy gases increased initially and then decreased with increasing temperature (Figures 2(b) and 2(c)). The corresponding temperatures at which the contents of C2H6 and other heavy gas components reached their maximums were 750 and 675°C, respectively. The highest content of total heavy gas was 1.05% at 750°C. The content of total heavy gas decreased to 0.73% at the highest experimental temperature of 800°C. H2 was detectable only above 500°C, and its content increased with increasing temperature (Figure 2(b)).

3.2. Carbon Isotopes of Gases

As presented in Table 2, δ13C1 and δ13C2 generally increased with increasing temperature (Figure 3). The variation of δ13C1 was not observable within the analytic error (±0.3‰) below 700°C. The value of δ13C1 clearly increased above 700°C, with an increase of 1.23‰ from 700 to 800°C. The variation in δ13C2 with increasing temperature could be divided into three segments, gradual increase in δ13C2 below 625°C, rapid increase from 650 to 725°C, and slow increase above 725°C. The distribution between δ13C1 and δ13C2 showed a distinct reversal trend () for below 725°C and a normal one () above 725°C.

3.3. Solid Products

The inner walls of the tubes heated to different temperatures are showed in Figure 4. Some black solid can be observed on the inner walls of the tubes above 725°C and was identified as carbon via energy spectrum analysis (Figure 5). The mass of the solid carbon was calculated by subtracting the mass of the tube before gas loading (one end being already welded) from the mass after gas analysis. Although solid carbon could be observed at the inner wall of tubes heated to more than 700°C, the mass could only be measured after heating at 775 and 800°C, using an analytical balance with a measurement error of 0.1 mg, with a mass of 0.2 mg and 0.4 mg, respectively.

4. Discussion

4.1. Heavy Hydrocarbon Gas Generation

The synthesis of heavier hydrocarbons by using CH4 and metallic catalysts is a popular method in the modern petrochemical industry [9, 14, 15]. Chen et al. [16] reported that heavy hydrocarbon gas could be generated in CH4 cracking experiments. The overall reaction through thermal coupling or polymerization of CH4 is generally believed to involve the following stepwise dehydrogenation at high temperatures [17].

In recent years, many researchers have proposed that the CH4 cracking and formation of higher carbon number hydrocarbons follows a free radical reaction mechanism [16, 18, 19]. The total radical reaction paths are summarized as follows:

All aforementioned analysis indicates that CH4 does not crack into carbon and H2 directly during early cracking, but first forms CH3 radicals and protons, and the CH3 radicals then combine or polymerize into heavy gas. The heavy gas further dehydrogenates and forms unsaturated heavy gas (alkene gas). The unsaturated heavy gas dehydrogenates at increasing temperatures and forms carbon. The experimental result presented in Table 2 shows that alkane and alkene gases are detected in the gaseous products during experimental CH4 cracking. Solid carbon is only observed above 700°C. This indicates that the reaction mechanism of CH4 cracking experiment is consistent with the aforementioned reaction paths. Although the above reaction might follow a free radical reaction mechanism, the formation of heavy alkane gas during early CH4 cracking could also be considered as direct CH4 polymerization. Therefore, isotopic reversal of natural gas owing to CH4 polymerization is chemically plausible.

The variation of different product components with increasing temperature (Figure 2) could be explained by aforementioned reaction mechanism. Although CH4 polymerization to form heavy hydrocarbon gas can occur at a relatively low temperature of 425°C, this reaction clearly increases at 625°C. The obvious decrease in CH4 content above 700°C and the appearance of solid carbon on the tube inner wall at 725°C indicate that the substantial cracking of CH4 begin in temperature range of 700–725°C. However, the content of heavy hydrocarbon gas continuously increases until 750°C. This result appears to conflict with the general knowledge in chemistry that the thermal stability of heavy gases is lower than that of CH4. Shuai et al. [20] proved by coal pyrolysis experiment in a confined system that the generation and cracking of C2H6 could coexist at high temperatures. Our experimental result does not mean that the thermal stability of heavy gases is higher than that of CH4, only that the ratio of heavy gas formation by CH4 polymerization is greater than that of heavy gases cracking into solid carbon in temperature range of 700–750°C. No matter how high the pyrolysis temperature or maturity level of a source rock is, there is no gas sample completely without any heavy gas component from in both OM pyrolysis experiments and in natural gas found in sedimentary basins. Some of these heavy gas components could originate from CH4 polymerization. This is also consistent with the hypothesis proposed by Xia and Gao [19] that heavy gas cracking and generation is a partly reversible reaction at high temperatures. The phenomenon could affect the isotopic composition of C2H6 generated in OM pyrolysis experiment in closed system, which becomes more negative at relatively high temperatures (>600°C). Low δ13C2 values observed may be considered to be erroneous and deleted, as the content of C2H6 is very low in gaseous products and the height of δ13C2 peak is often less than that of the half-height of reference (as in Figure 1). The real reason for the decrease in δ13C2 in OM pyrolysis experiments in a closed system at high temperatures is the CH4 polymerization that forms heavy hydrocarbon with a light isotopic composition. The amount of solid carbon generated by heavy gas cracking is so low that the carbon film on the inner wall of tube is very thin and presents a brown to dark brown color at 725 and 750°C (Figure 3), respectively. Above 750°C, the ratio of heavy gas cracking into solid carbon increases, compared to heavy gas generation, and thus, the concentration of heavy gas decreases. The amount of solid carbon increases obviously above 750°C, and the color of tube inner wall darkens.

4.2. Carbon Isotope Fractionation in Heavy Hydrocarbon Gas Generation

Carbon isotope evolution of CH4 cracking is attributed to kinetic isotope fractionation [16]. The 12C–H and 12C–12C bonds are more chemically active than 13C–H and 12C–13C, respectively [21, 22]. CH4 undergoes preferential bond cleavage at the 12C–H position rather than 13C–H, leading to the produced CH3 being depleted in 13C while residual CH4 is enriched in 13C. When two CH3 groups combine into C2H6, they preferentially form 12C–12C bond, further enhancing depletion in 13C and thus potentially leading to gas isotopic reversal.

Polymerization of CH4 is the primary reaction in experiment below 700°C. C2H6 presents a light carbon isotopic composition, and gas presents a reversal feature (Figure 3). The gradual enrichment of δ13C2 below 625°C is attributed to weak CH4 polymerization. The increase in polymerization of CH4 above 625°C causes rapid enrichment of δ13C2.

The appearance solid carbon on the tube inner wall at 725°C indicates that the cracking rate is greater than the heavy gas formation rate by CH4 polymerization in temperature range at 700–725°C. Little cracking of C2H6 would cause an obvious enrichment in δ13C2 in its low concentration (<1.1%). The change in the distribution between δ13C1 and δ13C2 from reversal trend to a normal one above 700°C is attributed to C2H6 cracking, but not to direct cracking of CH4 into solid carbon.

4.3. Geological Significance of the Experiment

The ultimate objective of the simulation experiment was to unravel the geological background at which isotopic reversal occurs. To achieve that aim, experimental temperatures must be correlated with actual geological temperatures or maturity (). Mi et al. [23] proposed a method to reconstruct of coal residues generated in a gold tube pyrolysis experiment at the same experimental conditions as our experiment. The relationship between and the experiment temperature in experiment at a heating rate of 20°C/h in the gold tube system is described by

Figure 6 shows the variations in CH4 content, C2H6 content, δ13C1, and δ13C2 with increasing calculated by Equation (11). Natural gas exploration has proven that both shale gas and coal formation tight gas with carbon isotopic reversal features are often found in areas where the source rock maturity is more than 2.5–3.0% [1, 2, 4, 8]. This is in good agreement with our experimental result (Figure 6(c)). Although CH4 polymerization might have started at the relatively low maturity of (425°C), CH4 polymerization is slight and only trace heavy gas forms (Figure 5(b)). However, OM has an intense hydrocarbon gas generation potential at this maturity stage. The trace heavy gas produced by CH4 polymerization with depleted δ13C would be masked completely by bulk heavy gas from OM enriched in δ13C. Thus, the natural gas presents normal carbon isotopic distribution. The obvious increase in C2H6 content indicates that CH4 polymerization intensified at (550°C, Figure 6(b)). The gas generation potential from oil-prone OM (type I and II) weakens, and the gas from OM is very dry above 2.5% [24, 25]. Therefore, the heavy gas from CH4 polymerization with depleted in δ13C makes the δ13C value of heavy gas low and displays isotopic reversal of natural gas. Although some gas generation potential exists for type III OM or coal, the gas from this type of OM is very dry with a higher maturity of 2.5% . The trace ethane depleted in δ13C by CH4 polymerization leads to carbon isotopic reversal of coal formation tight gas. This was proved by an mixing experiment of two end member natural gases, one is the gas with carbon isotopic reversal and another is coal-derived gas with high dryness and normal carbon isotopic distribution [26]. Although CH4 polymerization might be an important mechanism of carbon isotopic reversal for natural gas in the maturity range of 2.5–6.0% , it cannot cause the apparent decrease in natural gas resources as only little CH4 experiences early cracking (or polymerization) in this maturity range. The experiment result shows that the maturity limit for CH4 bulk cracking in a geological setting is approximately 6.0% (700–725°C). Therefore, the theoretical maturity limit for natural gas exploration should be 6.0% . However, the reported maximum maturity level for shale gas exploration is only 4.0% (in Fayetteville gas field, United States) [4] at present, which is much lower than the maturity of 6.0% . The upper maturity limit for gas exploration is also controlled by other geological and chemical factors (such as potential and rate of gas generation, gas diffusion rate, preserved condition, and shale catalysis for methane cracking). As the upper maturity limit of gas generation for oil-prone OM is 3.5% [24], therefore, it is acceptable in theory to explore natural gas exploration in marine shale with maturity levels in 3.5–6.0% . However, we suggest to not lightly practice natural gas exploration in oil-prone shale with the maturity space of 3.5–6.0% as the tiny CH4 molecule stored in shale is easily lost to exhaustion without the generated gas supply from shale in very long geological history.

5. Conclusion

Our experimental result shows that heavy hydrocarbon gases (including alkanes and alkenes) are generated during CH4 cracking, in addition to H2. Moreover, the heavy gas displays the feature of depleted carbon isotope composition in the early stage of CH4 cracking. CH4 does not crack into carbon and H2 directly, but first into CH3 and H+ during early cracking. CH4 undergoes preferential bond cleavage of 12C–H rather than 13C–H owing to the relatively higher chemical activity of 12C–H compared with that of 13C–H. The produced CH3 is depleted in 13C, and thus, C2H6 formed by CH3 combination (or CH4 polymerization) is also depleted in 13C. CH4 polymerization at the relatively low maturity stage of is so weak that the trace heavy gas by CH4 polymerization with depleted 13C can be masked by bulk heavy gas with enriched 13C from OM. Natural gas generated in this maturity range would exhibit a normal trend of carbon isotopic distribution. Relatively intense CH4 polymerization accompanied with the weak gas generation from OM causes carbon isotopic reversal appear in natural gas beyond of 2.5%. CH4 polymerization is a possible mechanism for carbon isotopic reversals of overmature natural gas. The maturity threshold for substantial CH4 cracking under actual geological settings is approximately 6.0% .

Data Availability

The data in paper titled “Combination of methyl from methane early cracking: a possible mechanism for carbon isotopic reversal of overmature natural gas” is the result of relative experiment involved in this study. All the data is in the paper. There is no restriction on data access and reader could get easily.

Additional Points

Highlights. (1) Heavy hydrocarbon gas is generated and bears depleted δ13C during CH4 early cracking. (2) CH4 does not crack directly into carbon and H2, but first forms CH3 and H+ groups during early cracking. (3) The CH3 shares depleted 13C owing to preferential bond cleavage of 12C–H rather than 13C–H. (4) The CH3 combination leads to depleted 13C2 and causes isotopic reversal for overmature natural gas. (5) The threshold of substantial CH4 cracking is 6.0% under actual geological settings.

Conflicts of Interest

The authors declare that they have no conflicts of interest.


This research is supported by the State Key Program of National Natural Science of China (Grant No. 42030804) and the National Natural Science Foundation of China (Grant No. 41673047). I would like to thank Doctor Keyu Liu for polishing our manuscript.