Pore Characteristics and Gas Preservation of the Lower Cambrian Shale in a Strongly Deformed Zone, Northern Chongqing, China

: The Lower Paleozoic marine shale in southern China has undergone several strong tectonic transformations in an extensive region outside the Sichuan Basin. Although some shale strata underwent strong deformation, they still contain a signiﬁcant amount of shale gas. The gas preservation mechanism in the strongly deformed shale has become the focus of attention. In this paper, the Lower Cambrian gas-bearing shale samples with a strong deformation taken from an exploration well in northern Chongqing, China, were investigated on their pore types and structure, with the aim to reveal the reason for the gas preservation. The pore types of the Lower Cambrian shale are dominated by microfractures and interparticle (interP) pores occurring mainly between clay minerals and between organic matter (OM) and clay minerals, while pores within OM that can be observed by FE-SEM (ﬁeld emission-scanning electron microscopy) are rare. The shale has a low porosity, with an average of 1.56%, which is mainly controlled by the clay mineral content. The adsorption experiments of low pressure N 2 (LPNA) and CO 2 (LPCA) indicate that the shale is rich in micropores and small mesopores (<2–3 nm) provided mainly by OM, but mesopores with a size range of 3–50 nm are underdeveloped. The shale, as revealed by LPNA data, has dominant slit-like or plate-like pores and an obvious low-pressure hysteresis (LPH), indicating a low gas diffusion. The deformed shale samples with a removal of OM by oxidation and their isolated kerogen further indicate that the LPH is completely related to OM, without any relationship with minerals, while an undeformed shale sample, taken from another well for a comparison, has no obvious LPH for both of its OM-removed sample and kerogen. Based on a comprehensive analysis of the relative data, it is suggested that the nanopores related to OM and clay minerals in the shale were signiﬁcantly altered owing to the deformation, with a result of the pores being squeezed into the slit-like shape and converted into micropores. This extraordinary pore structure of the shale formed during the deformation process should be the main preservation mechanism of shale gas. composition, and porosity of shale samples.


Introduction
As an unconventional natural gas resource, shale gas has been exploited commercially for more than several decades. However, the occurrence and enrichment mechanism of shale gas have still remained unclear since the formation that shale gas plays is very complicated and controlled by a variety of geological and geochemical factors [1][2][3][4][5][6][7][8][9]. In southern China, the Lower Cambrian and Lower Silurian shales are considered to have great potential [10][11][12][13], but currently, only the Lower Silurian shale gas has been exploited commercially [14,15]. The Lower Cambrian shale is distributed more widely and displays greater TOC (total organic carbon) content compared with the Lower Silurian shale [5,[16][17][18], but  [19] and M [36], respectively. Well YC2 is close to the Chengkou fault, where tectonic compression is stron Lower Cambrian strata have an inclined angle of as high as 35°-80° (Figure 1b). A t 10 samples collected from the well were all black shale, with a burial depth of 9 1147.68 m. They experienced a heavy deformation, with abundant microfracture by irregular veins (Table 1; Figure 2). According to the data from Yu et al. [19], th had an average desorbed gas amount of 0.79 m 3 /t, with a maximum of 1.23 m 3 methane is the main gas component (84.20-99.87%), while N2 and other non-hydro gases are approximatively less than 10% [48]. For a comparison, a black shale sam  [19] and Ma et al. [36], respectively.
Well YC2 is close to the Chengkou fault, where tectonic compression is strong. The Lower Cambrian strata have an inclined angle of as high as 35 • -80 • (Figure 1b). A total of 10 samples collected from the well were all black shale, with a burial depth of 941.34-1147.68 m. They experienced a heavy deformation, with abundant microfractures filled by irregular veins (Table 1; Figure 2). According to the data from Yu et al. [19], the shale had an average desorbed gas amount of 0.79 m 3 /t, with a maximum of 1.23 m 3 /t, and methane is the main gas component (84.20-99.87%), while N 2 and other non-hydrocarbon gases are approximatively less than 10% [48]. For a comparison, a black shale sample of the Lower Cambrian Niutitang Formation from well YC9 in southeastern Chongqing, China, was selected in this study. The area where well YC9 was located has had a relatively stable geological background [49], and the shale had no obvious deformation, with a complete sedimentary texture. However, its desorbed gas is much less (0.01-0.13 m 3 /t) and rich in nitrogen (>90%) [49]. The basic geological information and visible features of the 11 samples mentioned above are shown in Table 1. the Lower Cambrian Niutitang Formation from well YC9 in southeastern Chongqing, China, was selected in this study. The area where well YC9 was located has had a relatively stable geological background [49], and the shale had no obvious deformation, with a complete sedimentary texture. However, its desorbed gas is much less (0.01-0.13 m 3 /t) and rich in nitrogen (>90%) [49]. The basic geological information and visible features of the 11 samples mentioned above are shown in Table 1. LECO CS-200 carbon-sulfur analyzer was applied for the TOC content analysis of the prepared samples, then the peak area of CO 2 generated from combustion of the organic matter and calibrated by carbon steel (TOC = 0.812 ± 0.006%) was calculated, with analytical precision better than 10%.

Solid Bitumen Reflectance Measurement
Although the Lower Cambrian shale samples lack vitrinite, but solid bitumen are abundant, and the solid bitumen reflectance (BRo, %) can be used as an indicator of their maturity. A total of 4 samples were highly polished for sample preparation and reflectance measurements. The used instrument was a 3Y-Leica DMR XP microphotometer, with an objective of 50 × 0.85 in oil immersion and under incident light. The calibration standard samples were YAG-08-57 (Ro = 0.904%) and cubic zirconia (Ro = 3.11%). Large solid bitumen particles with a smooth surface were selected for the measurement in order to reduce the interference from surrounding minerals. More than 50 individual particles were measured for each sample, and their average was deemed as its representative reflectance value.

Mineral Composition Analysis
The mineral composition of the shale samples was analyzed by a Bruker D8 Advance X-ray diffractometer from Germany. The samples (about 2 g for each) were crushed to 200 mesh and dried at 378.15 K for 12 h to remove the moisture and volatile substances. The measured conditions were as follows: a Cu target, an acceleration voltage of 40 KV, a laser diffraction slit width of 1 mm, and a scanning speed of 4 • /min. The relative content of minerals was calculated according to their spectral peak areas integration approach with correction of Lorentz Polarization [50,51].

Porosity Measurement
The porosity of the samples (Φ, %) was determined by their skeletal density and apparent density [29], and the calculation formula is as follows: where ρb is apparent density (cm 3 /g), and ρs is skeletal density (cm 3 /g). The apparent density and skeleton density were measured, respectively, by the wax wrapping method and helium injection method, as suggested by Tian et al. [52].

FE-SEM Observation
The samples were prepared using an argon ion milling instrument (Hitachi High-Tech IM4000), with conditions of an acceleration voltage of 3 kV and a grinding duration of 4 h. The field emission scanning electron microscopy (FE-SEM) applied for the observation was Hitachi S4800. The working voltage was 1-2 kV, and the working distance of the electron gun was 1-8 mm. Both the backscattering (BSE) mode and secondary electron (SE) mode were used in order to identify different minerals and facilitate pore observation.

Kerogen Isolation and OM Oxidation
In order to reveal the organic and inorganic pore structure characteristics of the shale samples, some of them were selected for kerogen isolation and OM oxidation. The method to isolate kerogen from shales suggested by Liu et al. [53] was followed to treat the samples. The purpose for the OM oxidation to the samples is to remove their OM, while leaving their mineral framework. The method recommended by Chen et al. [54] and Gu et al. [55] was followed and is described as below. Shale samples (about 5 g for each) were crushed to 20-40 mesh, dried at 383.15 K for 12 h, and then placed into a muffle furnace to be oxidized at 698.15 K for 24 h. The treated samples were deemed to be OM-free [54].

Low-Pressure Gas Adsorption
The experiments of low-pressure N 2 adsorption (LPNA) and low-pressure CO 2 adsorption (LPCA) were performed with a Micromertics ASAP 2020 automatic specific surface analyzer. The experimental procedures are as follows. The samples (including original shale, isolated kerogen, and oxidized samples) were crushed to 20-40 mesh and dried at 383.15 K for 12 h. The samples (1.5 g for each) were transferred into the instrument sample tube and degassed for 10 h at 383.15 K under vacuum conditions (<10 mm Hg). The LPNA experiment was performed at the liquid nitrogen temperature (77.4 K) and a relative pressure (P/P 0 ) range of 0.001-0.995. The LPCA experiment was conducted at the temperature of ice-water mixture (273.15 K) and a relative pressure (P/P 0 ) range of 0.0001-0.03.
Based on LPNA adsorption branching data, the Barret-Joyner-Hallenda (BJH) model and Brunauer-Emmett-Teller (BET) model were used, respectively, to calculate the pore size distribution (PSD) and specific surface area, with an effective pore size range of 1.7-230 nm. Based on the LPCA data, the micropore volume and specific surface area were calculated, respectively, using density functional theory (DFT) model and the Dubinin-Astakhov (DA) equation [56][57][58].

TOC Content, Mineral Composition, and Maturity
The TOC content of the YC2 well samples ranges from 1.37% to 3.79%, with an average of 2.47% (Table 2), greater than the lowest limit (2%) of the gas-bearing shale with commercial potential [5]. The TOC content varied within a small range, without an obvious relationship with depth, different from the Lower Cambrian shales from other blocks where their TOC increases with depth [59]. The YC2 well samples are composed of quartz, clay minerals, plagioclase, calcite, pyrite, and a small amount of dolomite (Table 2), and the quartz and clay minerals are their main components. The quartz content ranges from 21.5% to 37.1%, with an average of 28.9%. The clay minerals include illite and chlorite, and their total content ranges from 17.0% to 45.2%, with an average of 31.7%. The carbonate minerals are mainly calcite and dolomite, and their total content ranges from 5.4% to 31.3%, with an average of 16.6%. The samples also contain a certain amount of plagioclase and pyrite, with a content of 11.1-23.2% (an average of 16.9%) and 1.5-35.8% (an average of 5.8%), respectively. Compared with the Lower Silurian shale in typical blocks (e.g., Jiaoshiba block) in southern China [60,61], the YC2 samples have a high content of clay minerals and can be considered to be a clay-rich shale. Figure 3 shows the correlation between the TOC and four main mineral contents of the YC2 samples. The TOC is positively correlated with quartz, total carbonate minerals, and  (Figure 3a,b,d), while negatively correlated with the clay minerals ( Figure 3c). This is slightly different from the Lower Cambrian shale from well HY1 in northern Guizhou, China, reported by Tian et al. [62]. For their samples, the TOC shows a weak positive correlation with the clay minerals but no obvious correlation with quartz. A positive correlation between TOC and quartz content is generally believed to contain a significant amount of biogenic quartz in shales [60,61,63,64]. The quartz in the YC2 samples may have two origins: biological and detrital, but the latter should be dominant because of the weak positive correlation between the TOC and quartz (Figure 3a). dolomite, and their total content ranges from 5.4% to 31.3%, with an average of 16.6%. The samples also contain a certain amount of plagioclase and pyrite, with a content of 11.1-23.2% (an average of 16.9%) and 1.5-35.8% (an average of 5.8%), respectively. Compared with the Lower Silurian shale in typical blocks (e.g., Jiaoshiba block) in southern China [60,61], the YC2 samples have a high content of clay minerals and can be considered to be a clay-rich shale. Figure 3 shows the correlation between the TOC and four main mineral contents of the YC2 samples. The TOC is positively correlated with quartz, total carbonate minerals, and dolomite (Figure 3a,b,d), while negatively correlated with the clay minerals ( Figure  3c). This is slightly different from the Lower Cambrian shale from well HY1 in northern Guizhou, China, reported by Tian et al. [62]. For their samples, the TOC shows a weak positive correlation with the clay minerals but no obvious correlation with quartz. A positive correlation between TOC and quartz content is generally believed to contain a significant amount of biogenic quartz in shales [60,61,63,64]. The quartz in the YC2 samples may have two origins: biological and detrital, but the latter should be dominant because of the weak positive correlation between the TOC and quartz (Figure 3a). Thermal maturity is of great importance for influencing shale porosity, pore structure, and gas-bearing property [65,66]; thus, an accurate evaluation of shale thermal maturity is critical. The maturity of the Lower Cambrian shale in the studied area has been reported in previous studies [36,44,67,68], but with quite different results. For example, the measured BRo value had a range of 3.28-4.25% from Ma et al. [36] and 4.22-5.34% from Cui et al. [67]; the EqRo value (calculated equivalent vitrinite reflectance from bitumen reflectance or other parameters) was between 2.4% and 3.0% from Han et al. [44] and between 3.34% and 3.78% from Wang et al. [68]. The BRo value can be influenced by a series of factors except for maturity, such as the optical anisotropy, size, and internal structure of the measured bitumen particles, as well as the numbers of measured points [69][70][71], which may be the reason for the great variation in the above results. In order to reduce the influence of the above factors, only large particles, especially band-shaped bitumen without internal structure were selected for the reflectance measurement in this Thermal maturity is of great importance for influencing shale porosity, pore structure, and gas-bearing property [65,66]; thus, an accurate evaluation of shale thermal maturity is critical. The maturity of the Lower Cambrian shale in the studied area has been reported in previous studies [36,44,67,68], but with quite different results. For example, the measured BRo value had a range of 3.28-4.25% from Ma et al. [36] and 4.22-5.34% from Cui et al. [67]; the EqRo value (calculated equivalent vitrinite reflectance from bitumen reflectance or other parameters) was between 2.4% and 3.0% from Han et al. [44] and between 3.34% and 3.78% from Wang et al. [68]. The BRo value can be influenced by a series of factors except for maturity, such as the optical anisotropy, size, and internal structure of the measured bitumen particles, as well as the numbers of measured points [69][70][71], which may be the reason for the great variation in the above results. In order to reduce the influence of the above factors, only large particles, especially band-shaped bitumen without internal structure were selected for the reflectance measurement in this study. The results are shown in Figure 4. Although the BRo ranges of the selected four samples are different, the main peak values are similar, between 3.0% and 3.2%. The averaged BRo values of the four samples are from 3.01% to 3.12%, which gives the EqRo values of 3.10-3.20% by the formula: EqRo = (BRo + 0.2443)/1.0495 [69], similar to the maturity of Lower Silurian shales in the Jiaoshiba block of the Sichuan Basin (a commercial shale gas reservoirs in southern China) [72].  Figure 4. Although the BRo ranges of the selected four samples are different, the main peak values are similar, between 3.0% and 3.2%. The averaged BRo values of the four samples are from 3.01% to 3.12%, which gives the EqRo values of 3.10-3.20% by the formula: EqRo = (BRo + 0.2443)/1.0495 [69], similar to the maturity of Lower Silurian shales in the Jiaoshiba block of the Sichuan Basin (a commercial shale gas reservoirs in southern China) [72]. Xiao et al. [5] proposed that a maturity with an EqRo value > 3.5% would be a high risk to shale gas exploration for Lower Paleozoic shales in southern China, which has been verified by pyrolysis experiments and shale gas exploration and development [73][74][75]. Although the maturity of the YC2 shale samples is quite high, it does not exceed this maturity limit, which is believed to be the basis of the storage of shale gas.

Pore Types, Porosity and Controlling Factors
The FE-SEM observation revealed that microfractures, interparticle (interP) pores, and intraparticle (intraP) pores occurred widely in the YC2 samples. Some minerals were deformed or fractured to form micro-fractures due to intense extrusion (Figure 5a,b), and some pyrite particles were elongated and oriented (Figure 5a). The interP pores occurred mainly between different particles, especially between clay minerals and between clay minerals and OM. This type of pores is mainly sheet, triangle, and irregular shapes, sizing from a few to hundreds of nanometers (Figure 5c,f). They may be formed by extrusion deformation. The OM in the samples was found mainly in the form of mineral-OM aggregates, with irregular or reticular pores within them (Figure 5d,e), while pores within Xiao et al. [5] proposed that a maturity with an EqRo value > 3.5% would be a high risk to shale gas exploration for Lower Paleozoic shales in southern China, which has been verified by pyrolysis experiments and shale gas exploration and development [73][74][75]. Although the maturity of the YC2 shale samples is quite high, it does not exceed this maturity limit, which is believed to be the basis of the storage of shale gas.

Pore Types, Porosity and Controlling Factors
The FE-SEM observation revealed that microfractures, interparticle (interP) pores, and intraparticle (intraP) pores occurred widely in the YC2 samples. Some minerals were deformed or fractured to form micro-fractures due to intense extrusion (Figure 5a,b), and some pyrite particles were elongated and oriented (Figure 5a). The interP pores occurred mainly between different particles, especially between clay minerals and between clay minerals and OM. This type of pores is mainly sheet, triangle, and irregular shapes, sizing from a few to hundreds of nanometers (Figure 5c,f). They may be formed by extrusion deformation. The OM in the samples was found mainly in the form of mineral-OM aggregates, with irregular or reticular pores within them (Figure 5d,e), while pores within pore OM particles were rare (Figure 5d  pore OM particles were rare (Figure 5d,f). IntraP pores developed mainly in carbonate minerals (Figure 5g), with triangular and multilateral shapes, and a size of hundreds of nanometers. Their formation may be related to dissolution. Figure 5h,i shows the pores from sample YC9-1. Compared with the YC2 samples, its microfractures were underdevelopment (Figure 5h), but OM-hosted pores with a diameter of 20-50 nm were widely present (Figure 5i), quite similar to the Lower Cambrian shale from the northern Guizhou, China [62]. The porosity of the YC2 samples is quite small, ranging from 1.08% to 2.15%, with an average of 1.56% ( Table 2). The porosity has no obvious correlation with either the TOC or quartz content, but it has a good positive correlation with the clay mineral content (R 2 = 0.71) and a weak negative correlation with the carbonate content (R 2 = 0.31) ( Figure 6). It can be inferred from the results that the porosity of the shale is mainly controlled by its clay minerals. The porosity of the YC2 samples is quite small, ranging from 1.08% to 2.15%, with an average of 1.56% ( Table 2). The porosity has no obvious correlation with either the TOC or quartz content, but it has a good positive correlation with the clay mineral content (R 2 = 0.71) and a weak negative correlation with the carbonate content (R 2 = 0.31) (Figure 6). It can be inferred from the results that the porosity of the shale is mainly controlled by its clay minerals. Previous studies have shown that although the porosity of Lower Cambrian shale without obvious deformation in southern China is not greater as a whole, it generally reaches 2-4%, and the TOC content is a dominant controlling factor [12,16,62,76]. For example, Tian et al. [62] reported the porosity of Lower Cambrian shale samples from well HY1 in the northern Guizhou, China, ranged from 2.51% to 4.13%, having a clear positive correlation with their TOC content (R 2 = 0.77) but no obvious relationship with their clay minerals. In contrast, the porosity of the YC2 samples is lower, and the clay minerals provide more pore spaces, with a limited contribution from OM pores. The reason may be related to the deformation of clay minerals to form some larger deformed pores and the compaction of larger OM pores into smaller pores. Under the action of intense compressive stress, clay minerals in shales may creep [77], with their original interlayer pores being compacted or deformed to increase spaces due to the existence of unbalanced deformation and distortion [78]; especially, some interP pores protected by brittle minerals (such as pyrite, quartz, etc.) may be enlarged significantly. Although clay minerals may form some larger deformation pores, and some brittle minerals may form tensile microfractures in deformed shales, their total porosity generally shows a decreasing trend [39,79]. Liu et al. [80] also pointed out that the stronger the tectonic stress compression, the lower the porosity of shales. The FE-SEM observation from the YC2 samples and sample YC9-1 are shown in Figure 5, and more data on their pore structure, which is presented in next sections, all support the above interpretation. In addition, the weak negative correlation between porosity and carbonate minerals (Figure 5d) may be related to the cementation of carbonate minerals, which may be another reason for the low porosity of the YC2 samples. Previous studies have shown that although the porosity of Lower Cambrian shale without obvious deformation in southern China is not greater as a whole, it generally reaches 2-4%, and the TOC content is a dominant controlling factor [12,16,62,76]. For example, Tian et al. [62] reported the porosity of Lower Cambrian shale samples from well HY1 in the northern Guizhou, China, ranged from 2.51% to 4.13%, having a clear positive correlation with their TOC content (R 2 = 0.77) but no obvious relationship with their clay minerals. In contrast, the porosity of the YC2 samples is lower, and the clay minerals provide more pore spaces, with a limited contribution from OM pores. The reason may be related to the deformation of clay minerals to form some larger deformed pores and the compaction of larger OM pores into smaller pores. Under the action of intense compressive stress, clay minerals in shales may creep [77], with their original interlayer pores being compacted or deformed to increase spaces due to the existence of unbalanced deformation and distortion [78]; especially, some interP pores protected by brittle minerals (such as pyrite, quartz, etc.) may be enlarged significantly. Although clay minerals may form some larger deformation pores, and some brittle minerals may form tensile microfractures in deformed shales, their total porosity generally shows a decreasing trend [39,79]. Liu et al. [80] also pointed out that the stronger the tectonic stress compression, the lower the porosity of shales. The FE-SEM observation from the YC2 samples and sample YC9-1 are shown in Figure 5, and more data on their pore structure, which is presented in next sections, all support the above interpretation. In addition, the weak negative correlation between porosity and carbonate minerals (Figure 5d) may be related to the cementation of carbonate minerals, which may be another reason for the low porosity of the YC2 samples.

LPNA and LPCA Curves and Pore Size Distribution
According to the gas adsorption mechanism of porous materials, with the increase in relative pressure (P/P 0 ), gas in pores will be micropore filling, and there will be monolayer and multilayer adsorption on the pore wall, and then capillary condensation up to a high relative pressure [33,81]. LPNA curves of shales can reflect their overall pore characteristics. According to IUPAC, LPNA curves can be divided into four categories based on their hysteresis ring morphology [33]. The LPNA curves of the YC2 samples all are close to type H4 in the IUPAC classification (Figure 7a-e), indicating that their nanopores have silt-like or layered shales. Their adsorption capacity increased rapidly (P/P 0 < 0.1) and then slowly (0.1 < P/P 0 < 0.9), reflecting the presence of a large number of micropores and a very limited multilayer adsorption, respectively. As P/P 0 > 0.90, their adsorption capacity increased rapidly, implicating the capillary condensation occurring in the macropores. Compared with sample YC9-1 (Figure 7f), the YC2 samples present a unique hysteresis ring, although the maximum adsorption quantity between the two well shales is also different-e.g., that of sample YC9YC is significantly greater than that of sample YC2-9 (both have a similar TOC content). Sample YC9-1 has a hysteresis loop as P/P 0 > 0.5 and presents a strong "forced closure" phenomenon due to "tensile strength effect" (TSE) [82] in a P/P 0 range of 0.4-0.5, exhibiting a totally reversible adsorption-desorption process as P/P 0 < 0.4. However, as for the YC2 samples, their hysteresis rings were not closed, owing to only a slight TSE at the P/P 0 of 0.4-0.5, which resulted in a low pressure hysteresis phenomenon (LPH) [83].

LPNA and LPCA Curves and Pore Size Distribution
According to the gas adsorption mechanism of porous materials, with the increase in relative pressure (P/P0), gas in pores will be micropore filling, and there will be monolayer and multilayer adsorption on the pore wall, and then capillary condensation up to a high relative pressure [33,81]. LPNA curves of shales can reflect their overall pore characteristics. According to IUPAC, LPNA curves can be divided into four categories based on their hysteresis ring morphology [33]. The LPNA curves of the YC2 samples all are close to type H4 in the IUPAC classification (Figure 7a-e), indicating that their nanopores have silt-like or layered shales. Their adsorption capacity increased rapidly (P/P0 < 0.1) and then slowly (0.1 < P/P0 < 0.9), reflecting the presence of a large number of micropores and a very limited multilayer adsorption, respectively. As P/P0 > 0.90, their adsorption capacity increased rapidly, implicating the capillary condensation occurring in the macropores. Compared with sample YC9-1 (Figure 7f), the YC2 samples present a unique hysteresis ring, although the maximum adsorption quantity between the two well shales is also different-e.g., that of sample YC9YC is significantly greater than that of sample YC2-9 (both have a similar TOC content). Sample YC9-1 has a hysteresis loop as P/P0 > 0.5 and presents a strong "forced closure" phenomenon due to "tensile strength effect" (TSE) [82] in a P/P0 range of 0.4-0.5, exhibiting a totally reversible adsorptiondesorption process as P/P0 < 0.4. However, as for the YC2 samples, their hysteresis rings were not closed, owing to only a slight TSE at the P/P0 of 0.4-0.5, which resulted in a low pressure hysteresis phenomenon (LPH) [83].  The LPCA curves of the samples are shown in Figure 8, and they are similar to the type I isothermal adsorption curve suggested by IUPAC. The slope of the curves has a trend of gradual decline with increasing pressure, indicating that the growth rate of adsorption capacity decreases. Based on the adsorption capacity at the maximum relative pressure (P/P 0 ≈ 0.03), the adsorption capacity of the samples to CO 2 can be preliminarily judged. Compared with the samples from well YC2 with a similar TOC content (e.g., YC2-9), the adsorption capacity of sample YC9-1 is also greater.
Energies 2022, 15, x FOR PEER REVIEW 1 The LPCA curves of the samples are shown in Figure 8, and they are similar type I isothermal adsorption curve suggested by IUPAC. The slope of the curves trend of gradual decline with increasing pressure, indicating that the growth r adsorption capacity decreases. Based on the adsorption capacity at the maximum re pressure (P/P0 ≈ 0.03), the adsorption capacity of the samples to CO2 can be prelimi judged. Compared with the samples from well YC2 with a similar TOC content (e.g. 9), the adsorption capacity of sample YC9-1 is also greater. As can be seen from the pore size distribution (PSD) curves characterized by th model (Figure 9), the peak value of the YC2 samples is mainly located at about 2 nm the peak width is very narrow (1.7-3 nm), indicating that the shale is dominat micropores and small mesopores. The PSD curve of the sample YC9-1 in the microp and mesoporous ranges (1.7-50 nm) all are relatively high, without an obvious (Figure 9), indicating that both its micropores and mesopores developed well. Th (0.35-1.1 nm) curves characterized by DFT model based on LPCA data show that th no significant difference between the YC2 samples and sample YC9-1 (Figur According to these PSD data, mesopores were greatly reduced in the deformed YC2 samples. In a study of the Lower Silurian shale samples with and without deform from the southeast area of Sichuan Basin, China, Liang et al. [84] also found that te deformation caused an alteration of pore structure in the deformed shale-i.e., f relatively equal pore distribution, even an advantage of mesopores of the undefo shale to a pore structure with obvious macropore and micropore advantages deformed shale. As can be seen from the pore size distribution (PSD) curves characterized by the BJH model (Figure 9), the peak value of the YC2 samples is mainly located at about 2 nm, and the peak width is very narrow (1.7-3 nm), indicating that the shale is dominated by micropores and small mesopores. The PSD curve of the sample YC9-1 in the microporous and mesoporous ranges (1.7-50 nm) all are relatively high, without an obvious peak (Figure 9), indicating that both its micropores and mesopores developed well. The PSD (0.35-1.1 nm) curves characterized by DFT model based on LPCA data show that there is no significant difference between the YC2 samples and sample YC9-1 ( Figure 10). According to these PSD data, mesopores were greatly reduced in the deformed YC2 shale samples. In a study of the Lower Silurian shale samples with and without deformation from the southeast area of Sichuan Basin, China, Liang et al. [84] also found that tectonic deformation caused an alteration of pore structure in the deformed shale-i.e., from a relatively equal pore distribution, even an advantage of mesopores of the undeformed shale to a pore structure with obvious macropore and micropore advantages of the deformed shale.

Pore Structure Parameters and Their Controlling Factors
According to the pore structure data of the YC2 samples (Table 3), their V BJH (BJH pore volume), S BET (BET specific surface area), Vmic (micropore volume), and Smic (micropore specific surface area) vary in a range of 0.32-0.64 cm 3 /100 g, 2.68-9.79 m 2 /g, 0.25-0.46 cm 3 /100 g, and 6.80-11.96 m 2 /g, respectively, with an average of 0.45 cm 3 /100 g, 5.55 m 2 /g, 0.36 cm 3 /100 g, and 9.57 m 2 /g, respectively. The Smic and Vmic are on average 1.72 and 0.80 times the S BET and V BJH , respectively. The V BJH , S BET , Vmic, and Smic values of the sample YC9-1 are 1.00 cm 3 /100 g, 10.91 m 2 /g, 0.46 cm 3 /100 g, and 7.35 m 2 /g, respectively. The Smic and Vmic are on average 0.62 and 0.46 times of the S BET and V BJH , respectively (Table 3). Compared with the sample YC9-1, the V BJH and S BET of the YC2 samples are much less, while their Vmic and Smic are only slightly lower (Table 3), which further indicates that the YC2 samples are dominated by micropores, while sample YC9-1 has an obvious advantage of mesopores, as presented by their PSD curves in Figure 9. The V BJH and Vmic of the YC2 samples have no significant correlation with their porosity (Figure 11), indicating that their nanopores (<220 nm, the maximum pore size characterized by LPNA in this study) have a limited contribution to their porosity and that their porosity should be provided mainly by macropores with a size of >220 nm. the sample YC9-1 are 1.00 cm 3 /100 g, 10.91 m 2 /g, 0.46 cm 3 /100 g, and 7.35 m 2 /g, respectively. The Smic and Vmic are on average 0.62 and 0.46 times of the SBET and VBJH, respectively (Table 3). Compared with the sample YC9-1, the VBJH and SBET of the YC2 samples are much less, while their Vmic and Smic are only slightly lower (Table 3), which further indicates that the YC2 samples are dominated by micropores, while sample YC9-1 has an obvious advantage of mesopores, as presented by their PSD curves in Figure 9. The VBJH and Vmic of the YC2 samples have no significant correlation with their porosity (Figure 11), indicating that their nanopores (<220 nm, the maximum pore size characterized by LPNA in this study) have a limited contribution to their porosity and that their porosity should be provided mainly by macropores with a size of >220 nm. Although The TOC content of the YC2 samples has no obvious correlation with porosity, it has clear positive correlations with VBJH, Vmic, SBET, and Smic, and moreover, its correlation with Vmic is better than that with VBJH, and its correlation with Smic is better than that with SBET ( Figure 12). Combined with the PSD curves of the samples (Figure 9), it can be believed that the TOC controls not only the micropores but also some of small mesopores (<5-10 nm, the limit of FE-SEM observation) [29,85]. Although The TOC content of the YC2 samples has no obvious correlation with porosity, it has clear positive correlations with V BJH , Vmic, S BET , and Smic, and moreover, its correlation with Vmic is better than that with V BJH , and its correlation with Smic is better than that with S BET (Figure 12). Combined with the PSD curves of the samples (Figure 9), it can be believed that the TOC controls not only the micropores but also some of small mesopores (<5-10 nm, the limit of FE-SEM observation) [29,85]. The clay mineral content of the YC2 samples has weak negative correlations with their VBJH and SBET, but no obvious correlation with their Vmic and Smic (Figure 13a-d).
Considering that the TOC content has a dominant control on pore structure parameters (Figure 12), the pore structure parameters were normalized by the TOC content. The results show that the clay mineral content has no correlation with the VBJH/TOC (ratio of VBJH to TOC) and SBET/TOC (ratio of SBET to TOC), but positive correlations with both the Vmic/TOC (ratio of Vmic to TOC) and Smic/TOC (ratio of Smic to TOC) (Figure 13e-h), indicating that the clay minerals in the samples contain less mesopores but are rich in micropores. This result is obviously different from previous research on clay minerals. For example, Ji et al. [86,87] documented that micropores in chlorite and illite (both of them are present in theYC2 samples, see Table 2) were underdeveloped, and their pores were dominated by pores with a size of 20-100 nm (mesopores and macropores); Li et al. [88] suggested that 2-50 nm pores in clay minerals provided their main pore volume and specific surface area. Considering that the porosity of the YC2 samples is mainly controlled by the clay minerals (Figure 6c), it can be further deduced that the tectonic deformation resulted in a significant reduction in mesopores in clay minerals, with the formation of some larger deformation macropores and the transformation into micropores. The clay mineral content of the YC2 samples has weak negative correlations with their V BJH and S BET , but no obvious correlation with their Vmic and Smic (Figure 13a-d).
Considering that the TOC content has a dominant control on pore structure parameters (Figure 12), the pore structure parameters were normalized by the TOC content. The results show that the clay mineral content has no correlation with the V BJH /TOC (ratio of V BJH to TOC) and S BET /TOC (ratio of S BET to TOC), but positive correlations with both the Vmic/TOC (ratio of Vmic to TOC) and Smic/TOC (ratio of Smic to TOC) (Figure 13e-h), indicating that the clay minerals in the samples contain less mesopores but are rich in micropores. This result is obviously different from previous research on clay minerals. For example, Ji et al. [86,87] documented that micropores in chlorite and illite (both of them are present in theYC2 samples, see Table 2) were underdeveloped, and their pores were dominated by pores with a size of 20-100 nm (mesopores and macropores); Li et al. [88] suggested that 2-50 nm pores in clay minerals provided their main pore volume and specific surface area. Considering that the porosity of the YC2 samples is mainly controlled by the clay minerals (Figure 6c), it can be further deduced that the tectonic deformation resulted in a significant reduction in mesopores in clay minerals, with the formation of some larger deformation macropores and the transformation into micropores. Similar to the correlation between the quartz content and porosity of the YC2 samples, there is also no obvious relationship of the quartz content with the VBJH and Vmic (Figure 14a,b). Although the carbonate content is positively correlated with the VBJH and Figure 13. Relationships of clay mineral content with pore structure parameters (a-d), and pore structure parameters/TOC (ratios of pore structure parameters to TOC) (e-h) for the YC2 samples. See details in text.
Similar to the correlation between the quartz content and porosity of the YC2 samples, there is also no obvious relationship of the quartz content with the V BJH and Vmic (Figure 14a,b). Although the carbonate content is positively correlated with the V BJH and Vmic (Figure 14c,d), its correlations with V BJH /TOC and Vmic/TOC were also not found (Figure 14e,f). Thus, it can be inferred that the quartz and carbonate minerals have no obvious influence on the pore structure parameters for the YC2 samples. This is consistent with previous studies on the undeformed Lower Cambrian shales in other areas of southern China [62,89]. Vmic (Figure 14c,d), its correlations with VBJH/TOC and Vmic/TOC were also not found (Figure 14e,f). Thus, it can be inferred that the quartz and carbonate minerals have no obvious influence on the pore structure parameters for the YC2 samples. This is consistent with previous studies on the undeformed Lower Cambrian shales in other areas of southern China [62,89].  LPNA experiments were performed on the kerogen and OM-oxidized samples. According to the adsorption-desorption curves, the OM-oxidized samples (YC2-6F, YC2-7F, YC9-1F) became very similar-without a LPH, but with a hysteresis loop in the high pressure section (P/P 0 > 0.5)-while the two kerogens (YC2-6K and YC2-7K) show an obvious LPH, but the kergenYC9-1K does not ( Figure 15). As presented in Figure 7 for their original shale samples, both the YC2-5 and YC2-6 have a LPH (Figure 7c,d), while YC9-1 does not (Figure 7f). Therefore, it can be believed that the LPH of the YC2 samples should be caused by their OM with a pore structure having been altered during strong deformation.
In addition, the adsorption equilibrium duration of those samples seems to depend on whether there is a LPH ( Figure 16). As P/P 0 is less than 0.1, the adsorption equilibrium duration of the samples with a LPH (YC2-6, YC2-6K, YC2-7, YC2-7K) is much greater than those without a LPH (YC2-6F, YC2-7F, YC9-1, YC9-1K, YC9-1F), and the data of the three OM-oxidized samples (YC2-6F, YC2-7F, YC9-1F) became overlapped, indicating the minerals in the shales have a very similar adsorption behavior. For a specific shale sample, the equilibrium duration order is kerogen > original shale > OM-oxidized sample.  In addition, the adsorption equilibrium duration of those samples seems to depend on whether there is a LPH ( Figure 16). As P/P0 is less than 0.1, the adsorption equilibrium duration of the samples with a LPH (YC2-6, YC2-6K, YC2-7, YC2-7K) is much greater than those without a LPH (YC2-6F, YC2-7F, YC9-1, YC9-1K, YC9-1F), and the data of the three OM-oxidized samples (YC2-6F, YC2-7F, YC9-1F) became overlapped, indicating the  Gas diffusion in the shale nanopore system is mainly characterized by Fick diffusion, Knudsen diffusion, and surface diffusion [93][94][95]. As a shale is dominated by micropores, its gas diffusion is dominated by surface diffusion, owing to the influence of surface adsorption potentials, with a diffusion coefficient to be closely related to pore diameter and morphology [95][96][97]. If the diffusion is insufficient in a finite time, the resulting LPNA curve will show a LPH [83]. Therefore, the LPH of a shale indicates that its pore structure is characterized by low diffusion.
The leakage process of shale gas mainly includes desorption and diffusion [97]. As the pore diameter of a shale becomes smaller, its gas adsorption potential energy will increase, meaning that its gas desorption will overcome a greater resistance. As discussed in above sections, the YC2 samples are mainly characterized by micropores and small mesopores, with undeveloped mesopores, and their pores, such as OM-hosted pores and clay-hosted pores may be flattened by extrusion and/or compaction to have silt-like or layered shapes. This unique pore structure is obviously not conducive to gas loss and would play an important role in the preservation of shale gas. On the other hand, the Lower Cambrian shale gas in southern China is usually rich in N2 [25][26][27]. A high content of N2 in shale strata is generally believed to be sourced from the thermal decomposition of their organic matter and NH4 + -rich minerals, and/or from the atmospheric N2 infiltration [24,25,27,74]. The thermal decomposition-derived N2 occurs mainly at a very high maturity stage, e.g., EqRo > 3.5% according to Gai et al. [74], while the YC2 samples have an EqRo of 3.1-3.2% (see Section 2.2), within the maturity range of methane generation and preservation [5,75]. In addition, the Lower Cambrian shale at well YC2 is currently shallowly buried and has a high dip angle, which is suitable for the infiltration of atmospheric N2. However, the shale gas contains only a minor amount of N2. This further suggests that the unique pore structure of the shale has prevented N2 from entering its nanopore system and that the reserved methane has been preserved. These Gas diffusion in the shale nanopore system is mainly characterized by Fick diffusion, Knudsen diffusion, and surface diffusion [93][94][95]. As a shale is dominated by micropores, its gas diffusion is dominated by surface diffusion, owing to the influence of surface adsorption potentials, with a diffusion coefficient to be closely related to pore diameter and morphology [95][96][97]. If the diffusion is insufficient in a finite time, the resulting LPNA curve will show a LPH [83]. Therefore, the LPH of a shale indicates that its pore structure is characterized by low diffusion.
The leakage process of shale gas mainly includes desorption and diffusion [97]. As the pore diameter of a shale becomes smaller, its gas adsorption potential energy will increase, meaning that its gas desorption will overcome a greater resistance. As discussed in above sections, the YC2 samples are mainly characterized by micropores and small mesopores, with undeveloped mesopores, and their pores, such as OM-hosted pores and clay-hosted pores may be flattened by extrusion and/or compaction to have silt-like or layered shapes. This unique pore structure is obviously not conducive to gas loss and would play an important role in the preservation of shale gas. On the other hand, the Lower Cambrian shale gas in southern China is usually rich in N 2 [25][26][27]. A high content of N 2 in shale strata is generally believed to be sourced from the thermal decomposition of their organic matter and NH 4 + -rich minerals, and/or from the atmospheric N 2 infiltration [24,25,27,74]. The thermal decomposition-derived N 2 occurs mainly at a very high maturity stage, e.g., EqRo > 3.5% according to Gai et al. [74], while the YC2 samples have an EqRo of 3.1-3.2% (see Section 2.2), within the maturity range of methane generation and preservation [5,75]. In addition, the Lower Cambrian shale at well YC2 is currently shallowly buried and has a high dip angle, which is suitable for the infiltration of atmospheric N 2 . However, the shale gas contains only a minor amount of N 2 . This further suggests that the unique pore structure of the shale has prevented N 2 from entering its nanopore system and that the reserved methane has been preserved. These two aspects are believed to be the basic reasons for the preservation of the Lower Cambrian shale gas at well YC2. However, it should be pointed out that whether this mechanism has a general significance for shale gas exploration and development in strongly deformed shale strata in southern China needs to be verified by more cases.

Conclusions
In this research, the pore characteristics and pore structure and the controlling factors of the Lower Cambrian gas-bearing shale samples with strong tectonic deformation taken from the YC2 well in southern China were investigated, and the main conclusions have been drawn as follows: (1) The pore types of the shale are microfractures-interP pores occurring mainly between clay minerals and OM-clay mineral composite pores within their aggregates-and all of them were formed or obviously affected by extrusion deformation. The porosity of the shale is low, and it is mainly contributed by macropores with a size of >220 nm and controlled by clay mineral content. (2) The pore structure of the shale is characterized by predominant micropores and 2-3 nm mesopores that are mainly contributed by organic matter and an underdevelopment of 3-50 nm mesopores, with dominant pores being a slit-like or layered shape. (3) The alteration of OM and clay mineral-related pores by strong deformation appears to be the main mechanism for the formation of the unique pore structure of the shale. This results in an obvious low pressure hysteresis of the LPNA curve caused by low gas diffusion, which is the main reason for the preservation of shale gas. This explanation gives a new significance to the LPH phenomenon, that is, it indicates a lower gas diffusion rate. This study provides new enlightenment for the preservation mechanism of shale gas under a background of complex structure.