Experimental Investigation of Mechanical Properties of Black Shales after CO2-Water-Rock Interaction

The effects of CO2-water-rock interactions on the mechanical properties of shale are essential for estimating the possibility of sequestrating CO2 in shale reservoirs. In this study, uniaxial compressive strength (UCS) tests together with an acoustic emission (AE) system and SEM and EDS analysis were performed to investigate the mechanical properties and microstructural changes of black shales with different saturation times (10 days, 20 days and 30 days) in water dissoluted with gaseous/super-critical CO2. According to the experimental results, the values of UCS, Young’s modulus and brittleness index decrease gradually with increasing saturation time in water with gaseous/super-critical CO2. Compared to samples without saturation, 30-day saturation causes reductions of 56.43% in UCS and 54.21% in Young’s modulus for gaseous saturated samples, and 66.05% in UCS and 56.32% in Young’s modulus for super-critical saturated samples, respectively. The brittleness index also decreases drastically from 84.3% for samples without saturation to 50.9% for samples saturated in water with gaseous CO2, to 47.9% for samples saturated in water with super-critical carbon dioxide (SC-CO2). SC-CO2 causes a greater reduction of shale’s mechanical properties. The crack propagation results obtained from the AE system show that longer saturation time produces higher peak cumulative AE energy. SEM images show that many pores occur when shale samples are saturated in water with gaseous/super-critical CO2. The EDS results show that CO2-water-rock interactions increase the percentages of C and Fe and decrease the percentages of Al and K on the surface of saturated samples when compared to samples without saturation.


Introduction
Increasing attention has been given to the reduction of the emission of carbon dioxide (CO 2 ), which contributes most of the greenhouse effect. Geological storage of CO 2 is one of the most promising ways to mitigate global warming and climate change [1]. Shale gas reservoirs, which are characterized as having ultra low permeability and high storage potential, are suitable for CO 2 sequestration [2][3][4].
As CO 2 is injected into shale reservoirs, it dissolves into waters or brines and changes the acid-base equilibrium which triggers the dissolution and precipitation of minerals [5,6]. Ketzer et al. [7] observed the dissolution of feldspars and calcite cement and the precipitation of dickite, opal and calcite, and reported that the dissolution of Ca and Fe cations limited the precipitation of carbonate. Lu et al. [8] found that concentrations of cations in groundwater presented two trends, and all the concentration variations were dominated by the precipitation of carbonate minerals. Liu et al. [9] experimentally investigated the chemical reactions of shale in CO 2 and brine saturation. Micro-scale observation revealed the dissolution of carbonates and feldspars, and the precipitation of carbonates

Geochemical Interactions
Due to the dissolution of CO2 in water, the weak acid H2CO3 forms and the pH of the fluid decreases, as described by the following equation: When CO2 is injected into a shale reservoir, the underground high pressure and temperature cause the pH of the water to decrease dramatically. According to Toews et al. [32], the pH of water in equilibrium with CO2 is 2.84 when the temperature and pressure are 40 °C and 7 MPa, respectively.

Geochemical Interactions
Due to the dissolution of CO 2 in water, the weak acid H 2 CO 3 forms and the pH of the fluid decreases, as described by the following equation: CO 2paqq`H2 O Ø H 2 CO 3 Ø H``HCO3 Ø 2H``CO 23 (1) When CO 2 is injected into a shale reservoir, the underground high pressure and temperature cause the pH of the water to decrease dramatically. According to Toews et al. [32], the pH of water in equilibrium with CO 2 is 2.84 when the temperature and pressure are 40˝C and 7 MPa, respectively. Ultra low pH will result in the dissolution of some minerals, including feldspar, calcite and pyrite [16,24,33].
However, the CO 2 -water-rock reaction is dependent on the environmental conditions [33]. According to Oomole and Osoba [34], higher CO 2 pressure promotes the dissolution of carbonates, while lower pressure decelerates the dissolution, or accelerates carbonate precipitation. Wigand et al. [35] found that higher pressure and temperature enhance the dissolution of quartz. Iron concentration also influences the reaction [7,36].
Based on the mineral compositions of shale samples (see Table 1), likely reversible reactions are listed in Equations (2)- (5).
Equations (2) and (3) present the reversible reactions that influence the dissolution of k-feldspar (Equation (3)) and calcite (Equation (4)). Equations (4) and (5) indicate the likely CO 2 trapping in mineral through bicarbonate and cations. Kaolinite and pyrite will not dissolve in the CO 2 equilibrium solution.

Sample Preparation
Shale blocks were cored parallel to the beddings. The length of each sample was 60 mm, and the diameter was chosen as 30 mm to ensure a height-diameter ratio of 2:1. All the coring and grinding work was finished in the Institute of Rock and Soil Mechanics, Chinese Academy of Science, China, and the experiments were conducted in the Deep Earth Energy Laboratory in the Department of Civil Engineering at Monash University, Melbourne, Australia.
Twelve samples were divided into two groups and saturated in water in equilibrium with CO 2 at pressures of 7 MPa (gaseous) and 9 MPa (super-critical), respectively. For each group, samples were saturated for different times (10 days, 20 days and 30 days). All the adsorbing conditions had the same temperature of 40˝C. The arrangement of saturation conditions is shown in Table 2. Two samples without saturation were set as the control group. Three slices (with a thickness of 0.4 mm), two of which were saturated together with samples in water with gaseous and super-critical CO 2 for 30 days, were used for SEM tests.
The saturation system consists of five parts: CO 2 cylinder, pump, valves, monitor system and container/heating system, as shown in Figure 2. A Model 500D syringe pump (Monash University, Melbourne, Australia) was used to refill CO 2 into containers. The pressure was controlled by a Teledyne D-series pump controller (Monash University, Melbourne, Australia) with a precision of 1 kPa. The surface of the container was covered by electric resistance wires and the temperature could be adjusted from room temperature to 100˝C.  Without saturation  ---2   Gaseous   10  7  40  2  20  7  40  2  30  7  40  2   Super-critical   10  9  40  2  20  9  40  2  30  9  40  2 Materials 2016, 9, 663 5 of 15

Testing Arrangement
Fourteen UCS tests were performed on shale samples after CO2-water-rock interactions. SHIMADZU AG 9 300 kN compression equipment (Monash University, Melbourne, Australia) was used to conduct the experiments. The loading rate was set at 0.1 mm/min for all the tests. The load and strain were recorded by an advanced data acquisition system. The crack propagations were recorded by acoustic emission (AE) sensors and ARAMIS 3-D technology. The microstructure and X-ray spectra were obtained by a JEOL FE 7001 SEM machine (Monash University, Melbourne, Australia) with a Brucker EDS detector at the Monash MCEM (Monash Centre for Electron Microscopy) center.

Microstructure Alteration after CO2-Water-Interaction
Mineralogical variations associated with microstructure alteration induced by water with gaseous/super-critical CO2 was investigated by SEM together with EDS analysis. Figure 3 shows the microstructure of shale samples without any fluid saturation (a,b), samples with water + gaseous CO2 (c,d) and water + SC-CO2 saturation (e,f). It is clear that the saturation of water with both gaseous and super-critical CO2 creates many pores on the surface of the shale slice, possibly because the dissolution of CO2 in water decreases the pH of the fluid (Equation (1)), which may accelerate the chemical reactions for mineral dissolution (Equations (2) and (3)) and carbonate precipitation (Equations (4) and (5)). Although SEM can only observe the surface of the sample, we can deduce that, with long-term saturation, water, CO2 and ions in the fluids will penetrate into the matrix of the shale and create more pores inside. These pores will create a secondary porosity system, which decreases the strength of the natural pore structure, and the strength of the sample therefore decreases after saturation [37,38]. As SEM analysis can only concentrate on an ultra-small area of the slice's surface, the difference of the effect of gaseous CO2 and SC-CO2 on the microstructure of shale samples is minor in the SEM images.

Testing Arrangement
Fourteen UCS tests were performed on shale samples after CO 2 -water-rock interactions. SHIMADZU AG 9 300 kN compression equipment (Monash University, Melbourne, Australia) was used to conduct the experiments. The loading rate was set at 0.1 mm/min for all the tests. The load and strain were recorded by an advanced data acquisition system. The crack propagations were recorded by acoustic emission (AE) sensors and ARAMIS 3-D technology. The microstructure and X-ray spectra were obtained by a JEOL FE 7001 SEM machine (Monash University, Melbourne, Australia) with a Brucker EDS detector at the Monash MCEM (Monash Centre for Electron Microscopy) center.

Microstructure Alteration after CO 2 -Water-Interaction
Mineralogical variations associated with microstructure alteration induced by water with gaseous/super-critical CO 2 was investigated by SEM together with EDS analysis. Figure 3 shows the microstructure of shale samples without any fluid saturation (a,b), samples with water + gaseous CO 2 (c,d) and water + SC-CO 2 saturation (e,f). It is clear that the saturation of water with both gaseous and super-critical CO 2 creates many pores on the surface of the shale slice, possibly because the dissolution of CO 2 in water decreases the pH of the fluid (Equation (1)), which may accelerate the chemical reactions for mineral dissolution (Equations (2) and (3)) and carbonate precipitation (Equations (4) and (5)). Although SEM can only observe the surface of the sample, we can deduce that, with long-term saturation, water, CO 2 and ions in the fluids will penetrate into the matrix of the shale and create more pores inside. These pores will create a secondary porosity system, which decreases the strength of the natural pore structure, and the strength of the sample therefore decreases after saturation [37,38]. As SEM analysis can only concentrate on an ultra-small area of the slice's surface, the difference of the effect of gaseous CO 2 and SC-CO 2 on the microstructure of shale samples is minor in the SEM images.  Figure 4 shows the EDS results of shale samples with or without fluid saturation. The X-ray spectra results are shown in Figure 5. For each slice, we observed three different areas and here present the one with a moderate percentage of carbon. From Figures 4 and 5 we can see that oxygen and silica are the first and second highest proportion of all the elements of the three kinds of shale slices. For slices after saturation with water with gaseous CO2 and SC-CO2, carbon accounts for the third highest percentage of the elements, which are 9.6% and 9.8%, respectively. However, for samples without saturation, the percentage of carbon is negligible. This is mainly because the precipitation of carbonates attaches on the surface of the shale slice. As the gaseous and super-critical CO2 applied in the study have similar solubility and pH in water (shown in Table 4), they present similar percentages of carbon. Compared to the spectra of samples without saturation, ions like Al  Figure 4 shows the EDS results of shale samples with or without fluid saturation. The X-ray spectra results are shown in Figure 5. For each slice, we observed three different areas and here present the one with a moderate percentage of carbon. From Figures 4 and 5 we can see that oxygen and silica are the first and second highest proportion of all the elements of the three kinds of shale slices. For slices after saturation with water with gaseous CO 2 and SC-CO 2 , carbon accounts for the third highest percentage of the elements, which are 9.6% and 9.8%, respectively. However, for samples without saturation, the percentage of carbon is negligible. This is mainly because the precipitation of carbonates attaches on the surface of the shale slice. As the gaseous and super-critical CO 2 applied in the study have similar solubility and pH in water (shown in Table 4), they present similar percentages of carbon. Compared to the spectra of samples without saturation, ions like Al and K decrease and Fe increase in samples with water and CO 2 saturation. This is caused by the dissolution of minerals, such as K-feldspar. and K decrease and Fe increase in samples with water and CO2 saturation. This is caused by the dissolution of minerals, such as K-feldspar.   and K decrease and Fe increase in samples with water and CO2 saturation. This is caused by the dissolution of minerals, such as K-feldspar.

Effects of CO 2 -Water-Rock Interaction on Mechanical Behaviors
The variation of UCS and Young's modulus of samples without saturation and samples saturated in water absorbed with gaseous and super-critical CO 2 are shown in Table 3 and Figure 6. The standard deviations of UCS values and Young's modulus for each saturation condition are minor, except for the two samples saturated in water with SC-CO 2 for 30 days, which show large UCS variation. Considering the change of strength with saturation time, we chose the higher one for the purposes of discussion. For other groups, the average values are used in all discussions.
According to Table 3 and Figure 6, of all the tested samples, samples without saturation have the highest UCS and Young's modulus values, of 58.82 MPa and 5.22 GPa, respectively. After 10-day saturation in water with gaseous CO 2 , the UCS value decreases to 40  The considerable reductions of strength and Young's modulus for saturated samples are due to the CO 2 -water-rock interactions coupled with chemical and mechanical effects. When shale samples are saturated in gaseous/super-critical CO 2 with the medium of fresh water, clays in the rock absorb water resulting in the shale swelling, which causes the decrease of strength and Young's modulus [39]. According to Heller and Zoback [40] and Luo et al. [41], shale gas, which exists in natural fractures, porous matrices and kerogen, is easier to be replaced by CO 2 as CO 2 has better adsorption ability in shale [42]. Therefore, the adsorption of CO 2 in shale samples will also cause shale swelling and strength decrease [43]. More importantly, the dissolution of CO 2 in water leads to the chemical reactions for mineral dissolution (Equations (2) and (3)) and carbonate precipitation (Equations (4) and (5)). The dissolution and precipitation process creates pores in the rock, as shown in Figure 3, which changes the microstructure of shale samples. This phenomenon will contribute to the reduction of strength and Young's modulus. Meanwhile, longer saturation time will cause greater damage on the shale sample, and the strength and Young's modulus will therefore be lower. This is in accordance with the experimental results.
From Figure 6, we can see that the UCS and Young's modulus values of samples saturated in water with gaseous and super-critical CO 2 have the same variation trend with saturation time. However, with the same saturation time, both of the values of samples soaked in water + SC-CO 2 fluids are smaller than those of samples soaked in water + gaseous CO 2 fluids. The small discrepancies of strength and Young's modulus are mainly caused by the difference of properties between gaseous CO 2 and SC-CO 2 , as shown in Table 4. SC-CO 2 has higher density, viscosity, thermal conductivity and dissolution ability in water than gaseous CO 2 . The pH of water dissoluted with these two fluids is similar, that of SC-CO 2 based water being 2.83 and that of gaseous CO 2 based water being 2.84. These differences overall contribute to the difference in results. Moreover, water, CO 2 and ions under super-critical saturation conditions will more easily penetrate into shale samples than under gaseous saturation conditions because of the 2 MPa higher confining pressure.
Materials 2016, 9, 663 9 of 15 reduction of strength and Young's modulus. Meanwhile, longer saturation time will cause greater damage on the shale sample, and the strength and Young's modulus will therefore be lower. This is in accordance with the experimental results. From Figure 6, we can see that the UCS and Young's modulus values of samples saturated in water with gaseous and super-critical CO2 have the same variation trend with saturation time. However, with the same saturation time, both of the values of samples soaked in water + SC-CO2 fluids are smaller than those of samples soaked in water + gaseous CO2 fluids. The small discrepancies of strength and Young's modulus are mainly caused by the difference of properties between gaseous CO2 and SC-CO2, as shown in Table 4. SC-CO2 has higher density, viscosity, thermal conductivity and dissolution ability in water than gaseous CO2. The pH of water dissoluted with these two fluids is similar, that of SC-CO2 based water being 2.83 and that of gaseous CO2 based water being 2.84. These differences overall contribute to the difference in results. Moreover, water, CO2 and ions under super-critical saturation conditions will more easily penetrate into shale samples than under gaseous saturation conditions because of the 2 MPa higher confining pressure.  Another important mechanical characteristic of reservoir rock is the brittleness index (BI). The brittleness index can be obtained by many methods, including mechanical analysis, energy analysis and mineral composition analysis [46]. In the present study, mechanical analysis was used for the calculation of the brittleness index (BI). It is defined by the following equation [47]. Table 5 shows the brittleness index of all tested samples and Figure 7 presents the variations of brittleness index with saturation time. The standard deviations of samples without saturation and samples saturated in gaseous/super-critical CO2 for 10 and 20 days are minor, varying from 1.2% to 3.8%. However, when the saturation time is 30 days, the samples saturated under gaseous conditions have a high standard deviation of 7.5%, and samples saturated under super-critical conditions have only one value (the other one is excluded because of the ultra-low UCS value). This means that samples with a longer saturation time will have larger variations in the brittleness index, which is caused by chemical-mechanical effects. However, the average value is still more reasonable for analysis.  Another important mechanical characteristic of reservoir rock is the brittleness index (BI). The brittleness index can be obtained by many methods, including mechanical analysis, energy analysis and mineral composition analysis [46]. In the present study, mechanical analysis was used for the calculation of the brittleness index (BI). It is defined by the following equation [47].

BI "
reversible strain total strain (6) Table 5 shows the brittleness index of all tested samples and Figure 7 presents the variations of brittleness index with saturation time. The standard deviations of samples without saturation and samples saturated in gaseous/super-critical CO 2 for 10 and 20 days are minor, varying from 1.2% to 3.8%. However, when the saturation time is 30 days, the samples saturated under gaseous conditions have a high standard deviation of 7.5%, and samples saturated under super-critical conditions have only one value (the other one is excluded because of the ultra-low UCS value). This means that samples with a longer saturation time will have larger variations in the brittleness index, which is caused by chemical-mechanical effects. However, the average value is still more reasonable for analysis.
According to Table 5 and Figure 7, samples without saturation have the highest brittleness index of 84.3%, which is consistent with the mineralogical analysis of samples that contain a high percentage of rigid components. After 10-day saturation, the values of samples in water with gaseous/super-critical CO 2 decrease to 76.4% and 65.2%, respectively. This is mainly because the adsorption of water and CO 2 and the chemical reactions of shale and fluids increase the plasticity and toughness. When the saturation time is 20 days, the brittleness index of both the two saturation conditions continues to increase. For samples saturated under gaseous conditions, the brittleness is 74.0%, slightly smaller than that of samples saturated under the same conditions for 10 days. Samples with SC-CO 2 and water saturation present a much lower brittleness index of 59.9%. Compared to the results of 10-day saturation, 20-day saturation may create more cracks and pores in shale samples, which cause shale samples to have higher plasticity. When the saturation time is extended to 30 days, both kinds of saturated samples present higher axial strains than those with a shorter saturation time before failure, and their brittleness index decreases to 50.9% and 47.9%, respectively. The discrepancy of physical properties (Table 4) between gaseous and super-critical CO 2 has less effect on shale strength and Young's modulus. However, their influence on shale's brittleness index is considerable. Samples with SC-CO 2 and water saturation present a lower brittleness index than that of samples saturated in gaseous CO 2 and water fluids for all three saturation times. This means that the beneficial properties of SC-CO 2 decrease shale's brittleness and increase its plasticity. According to Table 5 and Figure 7, samples without saturation have the highest brittleness index of 84.3%, which is consistent with the mineralogical analysis of samples that contain a high percentage of rigid components. After 10-day saturation, the values of samples in water with gaseous/super-critical CO2 decrease to 76.4% and 65.2%, respectively. This is mainly because the adsorption of water and CO2 and the chemical reactions of shale and fluids increase the plasticity and toughness. When the saturation time is 20 days, the brittleness index of both the two saturation conditions continues to increase. For samples saturated under gaseous conditions, the brittleness is 74.0%, slightly smaller than that of samples saturated under the same conditions for 10 days. Samples with SC-CO2 and water saturation present a much lower brittleness index of 59.9%. Compared to the results of 10-day saturation, 20-day saturation may create more cracks and pores in shale samples, which cause shale samples to have higher plasticity. When the saturation time is extended to 30 days, both kinds of saturated samples present higher axial strains than those with a shorter saturation time before failure, and their brittleness index decreases to 50.9% and 47.9%, respectively. The discrepancy of physical properties (Table 4) between gaseous and super-critical CO2 has less effect on shale strength and Young's modulus. However, their influence on shale's brittleness index is considerable. Samples with SC-CO2 and water saturation present a lower brittleness index than that of samples saturated in gaseous CO2 and water fluids for all three saturation times. This means that the beneficial properties of SC-CO2 decrease shale's brittleness and increase its plasticity.

Effects of CO2-Water-Rock Interaction on Crack Propagation
With the benefits of high sensitivity and non-destructive monitoring, AE analysis is a widely used method to investigate the stages of crack closure, crack initiation and crack damage in rock mechanics studies and engineering applications [37,48]. The crack closure stage is characterized by very small AE energy, which is released by seating and loading adjustment. With the increase of axial load, stable crack growth or dilation occurs and the AE energy increases gradually, leading to the

Effects of CO 2 -Water-Rock Interaction on Crack Propagation
With the benefits of high sensitivity and non-destructive monitoring, AE analysis is a widely used method to investigate the stages of crack closure, crack initiation and crack damage in rock mechanics studies and engineering applications [37,48]. The crack closure stage is characterized by very small AE energy, which is released by seating and loading adjustment. With the increase of axial load, stable crack growth or dilation occurs and the AE energy increases gradually, leading to the beginning of crack initiation. When rock samples reach the crack damage point, the AE energy increases drastically and the unstable crack growth creates considerable damage and samples finally fail. Hence, AE analysis provides an additional way to manifest rock mechanics during UCS tests. Figure 8 shows the variation of cumulative AE energy with axial strain for all kinds of saturated samples. As each group has two samples, here we present only one of the AE results for each saturation condition. The cumulative AE energy and axial stress for samples on both crack initiation and crack damage and the peak cumulative AE energy are listed in Table 6.
Materials 2016, 9, 663 11 of 15 beginning of crack initiation. When rock samples reach the crack damage point, the AE energy increases drastically and the unstable crack growth creates considerable damage and samples finally fail. Hence, AE analysis provides an additional way to manifest rock mechanics during UCS tests. Figure 8 shows the variation of cumulative AE energy with axial strain for all kinds of saturated samples. As each group has two samples, here we present only one of the AE results for each saturation condition. The cumulative AE energy and axial stress for samples on both crack initiation and crack damage and the peak cumulative AE energy are listed in Table 6.  According to Figure 8 and Table 6, samples with longer saturation time show higher axial strain when failure occurs. This is mainly because of the swelling of samples caused by water and CO2  According to Figure 8 and Table 6, samples with longer saturation time show higher axial strain when failure occurs. This is mainly because of the swelling of samples caused by water and CO 2 adsorption, and the occurrence of pores and cracks causes samples to have higher strain variations. This also indicates that CO 2 saturation decreases sample brittleness and increases its plasticity. When samples reach the crack initiation point, their values of cumulative AE energy are low, accounting for a small amount of the peak value ranging from 1.4% to 6.4%. A sample without saturation has the highest axial stress at this point, which is 23.36 MPa. The axial stress for samples saturated in gaseous CO 2 and water increase with the increase of saturation time, and the corresponding proportion of peak axial stress increases from 31.6% to 77.2%. For samples saturated under super-critical conditions, the axial stress decreases with the extension of saturation time. However, the proportion of axial stress on peak stress increases with increased saturation time, meaning that samples after gaseous/super-critical CO 2 and water saturation require a higher percentage of peak UCS value to reach the crack initiation when the saturation time is longer. From Figure 8, we can also observe that samples saturated under gaseous conditions reach crack initiation faster than under super-critical conditions when the saturation durations are 10 days and 30 days. In contrast, for 20-day saturation, the trend is reversed. Samples saturated in water with SC-CO 2 reach the crack damage point earlier than samples saturated in water with gaseous CO 2 in 20 days. However, when saturation durations are 10 days and 30 days, the crack damage points are almost the same for both kinds of saturated samples. The reason for this difference is mainly because shale samples are anisotropic. When samples reach to the crack damage point, the cumulative AE energy increases to a higher level, ranging from 10.6%-38.2% on peak cumulative AE energy.
According to Figure 8, Tables 3 and 6, peak cumulative AE energy shows a negative correlation with the UCS values. A sample without saturation has the lowest peak cumulative AE energy of 44,802 µJ. For samples saturated in water with gaseous/super-critical CO 2 , the peak cumulative AE energy increases with increasing saturation time. Specifically, with 10 days' saturation, the values for samples under gaseous and super-critical conditions doubled and reached 91,227 µJ and 94,840 µJ, respectively. After 20-and 30-day saturation, the values of peak cumulative AE energy were 100,564 µJ and 130,037 µJ for samples under gaseous conditions, and 116,068 µJ and 147,862 µJ for samples under super-critical conditions, respectively. This phenomenon is mainly caused by several reasons. Firstly, the adsorption of water and CO 2 increases the conductivity of AE emissions. Meanwhile, the swelling caused by water and CO 2 adsorption creates more artificial fractures. The propagation of these fractures creates more AE energy. In addition, carbonates created by the precipitation process are crushed during the UCS tests and generate acoustic emissions. More importantly, according to the SEM results in Figure 3, the numerous pores produced by chemical reactions decrease the brittleness and increase the plasticity of samples. Therefore, after the peak strength point, samples with CO 2 saturation can still bear load and create AE energy.

Conclusions
A series of UCS tests was conducted on samples without saturation, and samples saturated in water with gaseous/super-critical CO 2 for different periods of time. AE and SEM analyses were performed to evaluate the influence of gaseous/super-critical CO 2 -water-rock interactions and different saturation times on the mechanical properties of Chinese black shale. Several detailed conclusions can be drawn as follows: Gaseous/super-critical CO 2 -water rock interactions weaken the mechanical properties of shale, and the reductions of UCS, Young's modulus and brittleness index are closely related to the saturation time. With a saturation time of 10 days, water with gaseous CO 2 can cause reductions of 31.28% of UCS and 27.39% of Young's modulus, while water with SC-CO 2 causes UCS and Young's modulus decreases of 33.66% and 30.27%, respectively. By extending the saturation time to 30 days, the UCS and Young's modulus show reductions of 56.43% and 54.21% for water with gaseous CO 2 saturation, and 66.05% and 56.32% for water with SC-CO 2 saturation, respectively. The brittleness also decreases with increased saturation time. After 30-day saturation, the value decreases from 84.3% for samples without saturation to 50.9 for samples saturated under gaseous conditions, to 47.9% for samples saturated under super-critical conditions. The decrease of mechanical properties is partially due to the CO 2 -water-rock interactions, which dissolute mineral components and precipitate carbonates. Because of the small gap of physical properties, the effect of the different phase of CO 2 on shale's UCS and Young's modulus is minor.
The saturation of water with gaseous/super-critical CO 2 increases the total cumulative AE energy of shale samples, which shows a positive correlation with saturation time. Longer saturation time creates higher axial strain when failure occurs. For samples without saturation, the peak cumulative AE energy is 44,802 µJ. After 10 days' saturation in water with gaseous CO 2 , the value increases to 91,227 µJ, which is smaller than for 30-day saturation. Samples saturated in water with SC-CO 2 show a similar trend. However, the peak cumulative AE energy for super-critical saturation is slightly higher than that under gaseous saturation when the saturation time is the same.
Based on the SEM results, many pores occur on the surface of shale samples after 30-day saturation in water with gaseous/super-critical CO 2 . EDS analysis shows that CO 2 -water-rock interactions increase the percentages of C and Fe and decrease the percentages of K and Al on the surface of saturated samples. The changes of microstructure and chemical elements indicate the decrease in mechanical properties of saturated samples. CO 2 sequestration is a long-term program. However, in this study, we can only propose some short-term rules for CO 2 -water-rock interactions. Therefore, for a better understanding of the effects of CO 2 -water-rock interactions on shale's mechanical properties, experiments with a much longer saturation time are necessary.