Organic matter (OM) pores are widely considered to be important for gas storage and transportation. In this work, we quantitatively analyze the pore structure of OM and its controlling factors through geochemical and petrologic analyses, optical microscope, OM isolation, and adsorption isotherms. These analyses were carried out on lacustrine shale samples from the Lower Cretaceous Shahezi Formation, which is located in the Changling Fault Depression in Songliao Basin. The results show that the content of soluble OM (SOM) is low, accounting for 0.26%-3.75% of total OM. The contribution of pore development from SOM itself is limited. After extraction of SOM by chloroform, pore volume (PV), specific surface area (SSA), and average pore diameter (APD) exposed to gas molecules greatly increase. The existence of SOM has an obvious effect on pores of >10 nm, especially the clay mineral-related pores that contribute the most to the total PV. The content of kerogen is higher than SOM and accounts for 9.9%-65.5% (averaging 24.0%) of total PV in bulk shale, only second to minerals. More importantly, kerogen is the dominant contributor to organic PV, accounting for 95.8%-99.7% (averaging 98.3%) of the total organic PV. The pores developed in the kerogen peak at 0.4-0.7 nm and 10-30 nm. The solid bitumen (SB) followed by vitrinite and inertinite in kerogen contributes the most to the total kerogen PV.
According to the classification of Loucks et al. , pores in shales can be divided into interparticle mineral pores, intraparticle mineral pores, and organic matter (OM) pores in terms of pore occurrence in different matrices . Among these pores, OM-hosted pores are widely recognized as critical for gas storage and transport [2–10]. As people conduct in-depth research on shale reservoirs, growing importance and attention is given to organic pores.
In terms of solubility in organic solvents, OM can be divided into soluble OM (SOM) and kerogen [11–13], which also called macerals by organic petrologists. There are overlaps between the concepts referring to organic petrology and organic geochemistry. Specifically, SOM includes liquid hydrocarbons and some soluble solid bitumen (SB) from the secondary OM; kerogen includes the original macerals and insoluble SB from the secondary OM, i.e., pyrobitumen. There is general agreement on the importance of SOM in the shale pore system [14–19]. SOM is generally thought to occupy a certain pore volume (PV), blocking the path of fluid migration. But there is no consensus on how SOM affects pore size distribution (PSD) [14–19]. Wei et al.  suggested that the meso-PV increases more with higher maturity after dichloromethane extraction and the micro-PV increases more with TOC. Liu et al.  and Li et al.  both indicated that small pores tend to be occupied by extractable OM, i.e., SOM. Xiong et al.  demonstrated that the increments of PV after extraction with dichloroform are mainly related to the >30 nm and <10 nm pores, in Yanchang shales of the Ordos Basin. Kerogen is difficult to distinguish from OM by direct visual inspection. The effects of kerogen on pore development are mainly derived from measurements of isolated kerogen thorough extraction methods and gas adsorption isotherms. Rexer et al.  measured the pore structures of kerogens of the Jurassic Posidonia Shale from northern Germany using CO2 and N2 adsorption and found that the pore parameters (e.g., PV) of kerogen increase with thermal maturity. Ji et al.  suggested that the PV and specific surface area (SSA) of isolated OM after extraction are approximately three times greater than those of corresponding bulk shale before extraction. However, few studies have studied the effects of separate SOM and kerogen and quantitatively evaluated their impacts on pore spaces.
The terrestrial shale is enriched in various types of OM. The Shahezi shale in Lower Cretaceous lacustrine shales in Songliao Basin provides an excellent example to study the effects of OM components on pore structures. The Shahezi shale is in the mature and overmature stages, of which the OM contains both SOM and kerogen. In order to clarify the influences of SOM and kerogen on pore structures, this work employs the Shahezi shale in the Changling Fault Depression, attempting to mainly emphasize the roles of various kinds of OM present in the shale pore system. The specific objectives of this work are (1) quantitative characterization of the effects of SOM on pore structure through Soxhlet extraction and gas adsorption isotherms and (2) quantitative elucidation of the influence of kerogen on the pore structure using the kerogen isolation method and gas adsorption isotherms.
2. Geological Setting
The Changling Fault Depression, which is located towards the south of the Central Depression Zone, is the largest secondary structural unit (with an area of ) in Songliao Basin (Figure 1). Consistent with the basin tectonic movements, the Changling Fault Depression experienced three distinct stages of evolution: rifting, subsidence, and tectonic reversal [22, 23]. The rifting stage can be further divided into three distinct segments: initial rifting, maximum rifting, and late rifting, which correspond to the sedimentary deposition of the Huoshiling Formation (J3h), Shahezi Formation (K1sh), and Yingcheng Formation (K1yc), respectively  (Figure 2). Considering the difference in intensity of faulting, the Changling Fault Depression can be divided into seven secondary half-graben depressions: Qian’an sag, Qianshenzi sag, Heidimiao sag, Changling sag, Longfengshan sag, Zhaganhua sag, and Fulongquan sag. Taking the Shuangtuozi uplift zone as a boundary, the western area of the zone is dominated by a half-graben of faults to the west and overlaps to the east. The eastern area is dominated by a half-graben of faults to the east and overlaps to the west. During the Shahezi stage, the depression continued to expand, and the sedimentary environment became a braided delta-lake-subaqueous fan system [22, 25]. The lake deposit is dominated by a thick set of dark shales with thin interbedded sandstone layers.
3. Sample and Methodology
3.1. Shale Samples
A total of nine core samples from three wells of the Lower Cretaceous Shahezi Formation were subjected to geochemical experiments, among which six samples were subjected to organic petrology analysis, OM isolation, and CO2 and N2 adsorption isotherms. The detailed petrological characteristics were reported in Table 1 in Gao et al. . Clay minerals and quartz are the two most abundant minerals in samples. The content of clay and quartz ranges from 46.1% to 67.6% (averaging 57.3%) and 30.1% to 49.7% (averaging 34.8%), respectively. Feldspar is dominated by plagioclase (2.3%-10.1%). The clay minerals primarily include mixed illite-smectite (68%–82%), illite (15%–24%), and chlorite (5%–13%). In addition, small amounts of calcite, siderite, and pyrite are present.
3.2. Geochemical Analysis
TOC content was obtained from a LECO CS-230 carbon-sulfur analyzer. Before measurement, about 100 mg of powder samples was immersed in HCl with a concentration of 5% for 2 h to remove the inorganic carbon. Powders cleaned with distilled water were then dried at 70°C. Rock-Eval pyrolysis was conducted using an OGE-II rock pyrolyzer to investigate the hydrocarbon generation potential. With the continuous heating of each sample in the crucible, we recorded the peaks of residual hydrocarbon, S1, and the peak of residual TOC, S2, at 300 and 600°C, respectively [26, 27].
3.3. Optical Microscope Analysis
The optical microscopy observation under reflected light and oil immersion is aimed at determining the type of OM and the statistics of organic petrographic composition. The samples that were used to quantify maceral composition were crushed and then prepared as whole-rock pellets. The detailed procedure of maceral identification was summarized in Luo et al. . The classification of organic maceral is based on Cheng et al. .
3.4. Chloroform Extraction and Kerogen Isolation
In order to investigate the influence of SOM on pore development, a chloroform extraction method was used based on the theory of similarity and intermiscibility [19, 21]. Approximately 60 g of core samples was ground to powder with a size of 60-80 mesh (i.e., 0.18-0.25 mm). Each set of powder samples was then extracted over 24 h using chloroform. The remnant shale samples were dried under vacuum at atmospheric temperatures for 12 h. In order to investigate the contribution of kerogen to pore development, kerogen needs to be isolated. After extraction with chloroform, the remnant shale samples are further treated to remove inorganic carbon by HCl and inorganic mineral using HF and CrCl2. The preparational process is detailed in Ji et al. . The residue was used after washing, centrifuging, and drying.
3.5. CO2 and N2 Adsorption Isotherms
Before CO2 and N2 adsorption isotherm measurements, each set of powder samples with sufficient quantity was divided into three parts. Two of the three parts were prepared as described in Section 3.4. The other untreated part was used to compare with the other two parts. The experimental conditions of CO2 and N2 adsorption for the three parts were kept the same, which followed the settings in Jiang et al. . The particle sizes are used as 60-80 mesh. The measurable range of pore diameters is 0.30-1.47 nm for the CO2 adsorption isotherm and 3-275 nm for N2 adsorption isotherms. The SSA obtained from the N2 adsorption isotherm was calculated via the Brunauer-Emmett-Teller (BET) method . The PV and PSD were obtained from N2 adsorption isotherms through the Brunauer-Joyner-Halenda (BJH) theory [31, 32]. The CO2 adsorption isotherms were interpreted using the density functional theory (DFT), because DFT is more accurate when describing the thermodynamic properties of the fluid in the pores .
4. Results and Discussion
4.1. Geochemical Characteristics
The geochemical parameters of the samples are listed in Table 1. The TOC content is in the range of 1.08−3.57 wt%, indicative of an overall organic-medium shale. Rock-Eval values range from 462 to 577°C and Ro ranges from 1.68% to 2.10%, both of which indicate stages of overmature or gas window. The type of kerogen is identified as type III according to the identification of maceral and the formula proposed by Hu and Cao . S1 ranges between 0.09 and 0.61 mg/g. The values of S1 and S2 are low, which also indicate an overmature stage . Although the value of S1 is different from the extractable OM by chloroform, chloroform bitumen “A,” S1 and chloroform bitumen “A” have a good positive correlation (Figure 3(a)). TOC is different from the content of total OM (TOM), which can be empirically calibrated from TOC by multiplying by 1.18 based on Tissot and Welte . The value of TOM ranges from 1.28% to 4.21%. The content of kerogen can be calculated from TOM minus chloroform bitumen “A” () and TOM minus S1 (), both of which show excellent correlation (Figure 3(b)), indicating that the calculation of kerogen is reasonable. Here, we use as the content of kerogen. The percentage of SOM to TOM roughly lies in the range of 0.26%-3.75%, while kerogen accounts for 96.25%-99.74% of the TOM (Table 1).
4.2. Maceral Compositions and Characteristics
The maceral composition of the samples is listed in Table 2. The OM particles of samples are mainly vitrinite, inertinite, and SB. The content of SB using the point-counting method is generally low with values between 0.80 vol. % and 5.56 vol. %. The vitrinite is the most common OM particle in the samples, with the content in the range of 83.33%-89.80%. Normal vitrinite and perhydrous vitrinite coexist. There is also a certain amount of inertinite, with content varying between 8.40% and 14.00%. The characteristics of specified macerals are shown in Figure 4.
4.3. Organic Pore Characteristics from Gas Adsorption Isotherms
4.3.1. CO2 and N2 Adsorption Isotherms
All CO2 adsorption isotherms for each set of samples exhibit characteristics of type I (Figure 5), indicating micropore filling . The adsorbed volumes of CO2 for extracted shales are generally higher than those of untreated bulk shales, indicating that the SOM itself occupies a certain micropore volume. The total CO2 adsorption amount of untreated shale does not increase with TOC, indicating that the micropore development may be more affected by the inorganic minerals, especially clay minerals. Among them, the isotherms of B2-N04 and SL2-N09 for untreated and extracted shales converge at high relative pressure but still diverge at low relative pressure. The CO2 adsorption curve of isolated kerogen is much higher than that of untreated and extracted shale except S103-N05, indicating that kerogen itself is very microporous. The isotherm of kerogen in sample S103-N05 is lower than that of the corresponding bulk shale, suggesting the micropores in kerogen may contribute less to the bulk shale in the case.
All curves of low-pressure N2 adsorption isotherms demonstrate an anti-S shape, a significant hysteresis pattern (), and an absence of a plateau at high pressures () (Figure 6). The hysteresis loop and the absence of an adsorption plateau indicate the development of mesopores and macropores, respectively . The isotherms generally are recognized as type IV (isotherm with hysteresis loop) and similar for H3 and H4 , indicating the existence of slit-like and ink-bottle pores . Additionally, the three types of isotherms show obvious adsorption at very low relative pressures (), indicating the abundance of micropores. The N2 isotherms of extracted shale are also higher than those of bulk shale, indicative of occupation of SOM in mesomacropores. The isotherms of isolated kerogen are generally still higher than the other two kinds of curves (Figure 5), especially at high relative pressures, indicative of more mesomacropore pores in isolated kerogen. In sample SL2-N09, the adsorption volume of kerogen is lower than that of extracted shale, indicating a lower contribution of pores associated with kerogen.
4.3.2. Quantitative Characterization of Pores in Different Matrices
The pore structure parameters of bulk shale, the shale after extraction with chloroform, and the kerogen are listed in Table 3. The total PV and SSA of bulk shale range between 0.0094-0.0143 ml/g and 5.52-28.44 m2/g, respectively, which are lower than those of the corresponding extracted shale. The total PV and SSA of extracted shale are 0.0385-0.0717 ml/g and 28.55-41.92 m2/g, respectively. For isolated kerogen, the parameters are 0.0437-0.1522 ml/g and 18.59-143.51 m2/g, respectively.
In equation (1), is the total PV of bulk shale, is the PV associated with inorganic minerals of bulk shale, is the PV associated with SOM in bulk shale, and is the PV associated with kerogen in bulk shale. In equation (2), is the total PV of corresponding extracted shale; is the volume occupied by SOM itself, which can be obtained by subtracting equation (1) from equation (2); and is the total PV developed associated with kerogen, which is obtained from the CO2 and N2 measurements on isolated kerogen. is a correction factor, which is equal to .
We assume that the degree of organic pore development per unit of mass sample in SOM is comparable to that in kerogen. Then, the total organic PV () can be approximately calculated by multiplying by TOM in shale. The results may be slightly higher than the real values, but the trend will not change. The calculated results are shown in Table 4 and Figure 7. The total organic PV ranges from 0.0012 to 0.0064 ml/g, with an average of 0.0024 ml/g, accounting for 9.9%-65.5% (averaging 24%) of the total PV. The PV of pores related to inorganic minerals is 0.0034-0.0129 ml/g with an average value of 0.0082 ml/g, accounting for 34.5%-90.1% (averaging 76%) of the total PV. The pores associated with SOM are far less than those in kerogen. The PV of SOM and kerogen in shale is, respectively, 0.000007-0.000058 ml/g and 0.0011-0.0064 ml/g, accounting for 0.27%-3.70% (averaging 1.65%) and 95.80%-99.69% (averaging 98.32%) of the total organic PV, respectively.
4.3.3. PSD from CO2 and N2 Adsorption Isotherms
Figure 8 shows the micropore PSD of bulk shale, extracted shale, and isolated kerogen. Overall, all three curves of each sample exhibit very similar trends. The curves show an obvious bimodal pattern, in which the peaks are in the wide range of 0.4-0.7 nm and the narrow range of 0.8-0.9 nm, indicating that micropores mainly come from the range of 0.4-0.7 nm and 0.8-0.9 nm. The difference of PSD curves between bulk shale and extracted shale is not significant for pores of >0.5 nm, while for pores of <0.5 nm, the PSD curves of extracted samples are slightly higher than those of bulk samples, which indicate that SOM preferably occurs in . But for sample S103-N05, the PSD curve of extracted shale is higher than that of corresponding bulk shale for the whole micropore range (0-2 nm), indicating that SOM may exist in the whole micropore range. The PSD curves of isolated kerogen are highest among three curves except for SL2-N09, and the peaks are located in 0.4-0.7 and 0.8-0.9 nm as well.
Unlike the PSD curves from CO2 adsorption isotherms, all curves from N2 adsorption isotherms after extraction are dramatically higher than those before extraction (Figure 9). Especially for , the increment due to extraction in PV increases linearly and then keeps stable, suggesting that SOM mainly occupies >10 nm pores in addition to the micropore pores discussed before. The PV curves of kerogen are also significantly higher than those of bulk shale, especially in the >10 nm range. The reason could be that kerogen itself develops . Three kinds of SSA PSD curves for each sample show a similar trend and present a bimodal pattern. The PSD curves of bulk shale show that the main contributor to SSA mainly comes from , especially 2-4 and 10-30 nm. Many small pores can provide a larger SSA than a few large pores . The SSA of extracted shale in 2-10 nm is obviously higher than that of kerogen, indicating that the occupied pores by SOM are mainly related to the inorganic minerals. The subpeak at 10-30 nm is lower than that of kerogen, which may be due to the removal of SOM from kerogen-related pores.
4.4. Effects of SOM on Pore Development
After extraction of SOM, both PV and SSA of the remaining shale samples increase. The increment of PV and SSA from both CO2 and N2 adsorption isotherms is positively correlated with chloroform bitumen “A” except in the B2-N04 and SL2-N09 samples (Figures 10(a) and 10(b)). This phenomenon has been observed in many earlier studies [12, 15, 16, 19]. The SOM itself occupies a certain pore space. After extraction, the initially occupied pores will be reexposed, resulting in the overall increase of PV and SSA. In addition, the correlation between PV and SSA from CO2 adsorption and chloroform bitumen “A” is better than that from N2 adsorption and chloroform bitumen “A,” indicating that the micropores may be fully filled with SOM and the mesomacropores may be partially filled with SOM. The increment of micropores in B2-N04 and SL2-N09 is negative, and the absolute value is small . The increment of micropores before and after extraction is generally lower than that of mesomacropores from N2 adsorption isotherms. The APD of shale before and after extraction increased from 4.7-13.89 nm (averaging 6.06 nm) to 9.07-11.07 nm (averaging 10.76 nm), which indicates that pores with large sizes will be preferentially occupied by SOM (Figure 9). Generally, the pore parameters of bulk shale show no obvious correlation with TOC (Figures 10(c) and 10(d)), for either micropores or mesomacropores, indicating that pore development may be more influenced by inorganic minerals, especially clay minerals. The good correlation with clay minerals (correlation coefficients are 0.80 and 0.49 for PV and SSA, respectively) and negative correlation between pore parameters of extracted shale and kerogen confirm this explanation (Figures 10(e) and 10(f)). But it cannot be denied that, to some extent, some SOM is adsorbed onto the kerogen, because the PSD trend of extracted shale and isolated kerogen is highly consistent (Figures 8 and 9). For each sample, the increment in different pore size ranges is different. The increment of PV in the range of >10 nm is the highest. The increment of SSA mainly comes from the pores in the range of 0.3-0.5 and 2-30 nm.
4.5. Effects of Kerogen on Pore Development
Quantitative characterization shows that the contribution of kerogen to bulk shale is second to clay minerals (Table 4), which is consistent with the previous conclusions obtained through other correlation analyses . However, for the OM, the pores associated with kerogen are prevailing, accounting for 96.25%-99.73% of the total OM PV, with an average of 98.34%. The kerogen of the samples in the study area was identified as vitrinite, inertinite, and SB. Although observation via FE-SEM shows that porous OM is dominated by SB, and partially by organoclay complexes and mixtures of organic and inorganic minerals , correlation analyses show that the SB weakly correlates with the PV and SSA of isolated kerogen (Figures 11(a) and 11(b)). The reason may be the counting statistics of the content of SB (Table 2), because the SOM is difficult to distinguish from SB under reflected light. Vitrinite also possesses a weak correlation with the PV and SSA of kerogen (Figures 11(c) and 11(d)), indicating that vitrinite contributes slightly to the pore development of kerogen. This is somewhat similar to the study of Teng et al. , in which PV and SSA are shown to be positively correlated. The reason could be that there is a certain amount of hydrogen-rich vitrinite in the shale, which has been proven to have the potential of hydrocarbon generation. Inertinite is negatively correlated with the PV and SSA of kerogen (Figures 11(e) and 11(f)). In general, the hydrocarbon generation potential of inertinite is low. The inertinite with pores in the selected samples is rare. Therefore, if the pores of kerogen develop, it will be beneficial to the bulk shale. Otherwise, the nonporous kerogen will occupy the pores and increase the tortuosity of gas transport in the bulk shale.
This work combined integrated geochemical analyses, optical microscopy, OM isolation, and gas adsorption isotherms to quantitatively characterize the pores in different OM of Shahezi shale. Specifically, we draw the following main conclusions from this comprehensive suite of experiments:
The content of SOM is low, accounting for 0.26%-3.75% of the total OM. The contribution of SOM to organic pore development is limited. After extraction of SOM by chloroform, the PV, SSA, and APD greatly increase. Pores with diameters above 10 nm will be preferentially occupied and blocked by SOM. SOM tends to affect the pores related to clay minerals
TOM is dominated by kerogen, accounting for more than 97% of total OM. Kerogen provides a second contribution to the total PV and is the dominated contributor to organic PV. The pores developed in kerogen are mainly distributed in the range of 0.4-0.7 and 10-30 nm
The maceral composition of Shahezi shale is mainly vitrinite, followed by inertinite and SB. The solid bitumen (SB) followed by vitrinite and inertinite in kerogen contributes most to total kerogen PV based on correlation analysis
The [DATA TYPE] data used to support the findings of this study are included within the article.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
The work is supported by the National Science and Technology Major Project (2016ZX05034001-005) for providing financial support and the Northeastern Petroleum Bureau of SINOPEC for providing necessary basic geological data and samples. Special acknowledgment is also given to associate professors Qingyong Luo and Lin Wei for their guidance on identification of the type of OM.