Correlation Study Between Minerals Composition and Porosity Evolution of Black Shale Under Different Oxidation Mechanisms

The correlation between rock minerals composition and porosity evolution is essential to black shale oxidation,chemical and biological effects may act different roles which is unclear. In present study, the acid solution and Acidithiobacillus ferrooxidans was applied to simulate the chemical and biological effects, the total organic carbon (TOC), pyrite and clay minerals content, porosity characteristics, micro-surfacemorphologyof black shale sampleswere employed to analyze the minerals-porosity correlation. Results showed that the chemical effect led to a greaterdecrease of pyrite and increase of clay minerals content than biological effect, which accompanied by a higherspecic surface areas and pore volume. These difference may attribute to the TOCcontent, which decreasedby acid erosion and led to the increase of macropores in chemical group, then promoted the minerals reactions; whereas the TOC content increased byattached and dead bacteria cell bodies on rock surface, led to the increase of micropores and less minerals reactions in biological group. It suggests that the correlation between minerals composition and porosity evolution is different in chemical and biological effects, is mainly due to the different alter trends of TOC content.

Correlations between the minerals composition and porosity evolution of black shale have been well established, the decrease of pyrite and increase of clay minerals were proven to increase the shale porosity [4,5,14,15], but the role of organic matters remained open to debate. For example, Fisher et al, proposed the organic matters increased the speci c surface area of black shale samples [11], Yang et al, made a similar conclusion based on laboratory studies [16]. But based on eld investigations, researchers reported a negative correlation between organic matter content and rock porosity on weathered pro les [4,5]. Interestingly, in spite the total organic carbon (TOC) content decrease on regolith pro les, Jin et al, also reported an addition of modern organic matters on the pro les surface and attributed this phenomenon to the biota turbulence [4]. Therefore, the contradict roles of organic matters in shale porosity evolution may be associate with the biological oxidation mechanism.
The chemical and biological effects are major mechanisms for black shale oxidation [1,3,4,11,13,[17][18][19][20][21], but the minerals-porosity correlations are mostly focus on the chemical mechanism [4,5,11,18,22,23]. Our lab recently reported that Acidithiobacillus ferrooxidans (A. ferrooxidans) bacteria led to an increase of elemental C and a decrease of porosity on the shale micro-surface, which in contract to the oxidation behavior of acid solution [7]. However, that study mainly investigated the pyrite oxidation rate and micro-surface porosity characteristics at the initial oxidation stage (7 days), the overall alterations of other minerals and porosity evolution during black shale oxidation process was still unclear.
In present study, the acid solution and A. ferrooxidans bacteria were applied to simulate the chemical and biological effect during black shale oxidation process. The XRD analysis, organic carbon measurement, nitrogen absorption experiment and high-pressure mercury intrusion test, SEM observation and EDX line scanning analysis were carried out to measure the minerals composition, total organic carbon (TOC) content, porosity characteristics, micro-surface morphology and elemental concentrations of black shale samples. The targets of this study are, i) to analyse the different minerals composition and shale porosity characteristics under chemical and biological effects; ii) and to establish the correlation between minerals alteration and porosity evolution of black shale under biological oxidation mechanisms.

Sample preparation
The black shale samples were collected from a mountainous area in Chengkou County, Chongqing City, China. The sample site, previously described by Li, et al [7], located at the elevation of 1000 m above sea level and the latitude and longitude of 31°57'N and 108°37'E, respectively (Fig. 1). The fresh rock fragments were obtained at 10 m below the natural ground level, and cut into 0.3 cm (width) × 0.3 cm (length) × 0.3 cm (height) cubes. After cleaned by an ultrasonic cleaner with ethanol (75%, v/v) for 20 min and rinsed by deionized water for 5 times, samples were dried in a vacuum drying incubator at 40 °C and sterilized by UV irradiation for 12 h. Approximately 30 g rock cubic was used for each group.

Bacterial strain and culture media
The A. ferrooxidans strain used in this study was isolated from the acidic groundwater near the sampling site [7]. A. ferrooxidans was cultured in 9K medium, which contained (per litre): 3 g (NH 4 ) 2 SO 4 , 1 g KCl, 0.5 g K 2 HPO 4 , 0.5 g MgSO 4 ·7H 2 O, 0.001 g Ca(NO 3 ) 2 , and 44.2 g FeSO 4 ·7H 2 O, with the pH adjusted to 2.50 with H 2 SO 4 . After A. ferrooxidans was cultivated to the logarithmic growth phase, the solid materials were removed using sterilized qualitative lter paper, and the cells were harvested by centrifugation (2376 × g).
The resuspended cells were counted using a haemocytometer and the bacterial density was adjusted to 4 × 10 8 cells/mL with a H 2 SO 4 solution (pH = 2.50).

Experimental grouping
100 mL 9K medium without ferrous sulphate as (per litre): 3 g (NH 4 ) 2 SO 4 , 1 g KCl, 0.5 g K 2 HPO 4 , 0.5 g MgSO 4 ·7H 2 O and 0.001 g Ca(NO 3 ) 2 was used for each group. The pH value in control group (G1) was approximately at 6.50, and adjusted to pH = 2.50 with H 2 SO 4 for the acid solution group (G2) and A. ferrooxidans group (G3). 10% of A. ferrooxidans cell suspension was inoculated to G3. All samples were immersed in a 250-mL Erlenmeyer ask and incubated in a constant temperature shaker at 28 °C, 150 r/p.

Organic carbon measurement
Total organic carbon (TOC) was measured using a C230 Elemental Analyzer (Leco Corporation, USA) as follows analysis procedure: approximately 100 mg of dried and powdered black shale at 100-mesh was treated with 5% HCl at 80 °C to remove any carbonate, and then washed six times with deionized water to remove residual HCl. The treated powder was mixed with iron powder and a tungsten-tin alloy as additives, then combusted in the analyser at 3000 °C. The oxygen was added as the combustionaccelerating gas and nitrogen as the carrier gas. The TOC was calculated by integrating the area of the peak representing the CO 2 generated from the combustion of organic matter.

XRD analysis
The XRD diffraction analysis (Bruker, Karisruhe, Germany) was performed for the minerals composition analysis. The X-rays used Cu Kα radiation (λ = 1.54060) at a voltage of 40 kV and a current of 40 mA, with a scanning range 2.5 52° (2θ) and a scanning step length of 0.01° (2θ), a Cu lament and Ni lter was performed for the qualitative mineral analysis.

Nitrogen adsorption and high-pressure mercury intrusion tests
Nitrogen adsorption tests by a Quadrasorb SI Surface Area and Porosity Analyzer (Quantachrome, American) were performed to measure the samples speci c surface area and the pore size distribution with range of 0.35 nm ~ 400 nm. The test procedures were as follows: approximately 0.5 g of powdered shale was initially degassed at 150 °C for 3 hours to remove moisture and residual volatiles. Then the pure nitrogen at -197.3 °C with a purity greater than 99.999% was used for relative absorption properties measurement. For speci c surface area (surface area per unit gram) measurement, a modi ed BET technique was applied to invert the adsorption isotherm and calculate the speci c surface area, as suggested by Rouquerol et al. [15,24]. The pore volume was obtained from the Gurevich rule [25,26] and the pore size distribution was obtained by applying the Barrett, Johner and Halenda (BJH) method [27] from the adsorption isotherm, in the pore size range of 1.7 nm ~ 150 nm and relative pressures (P/P 0 ) of 0.06 ~ 0.99.
The high-pressure mercury intrusion (HPMI) test was performed to obtain pore size distributions of macropores larger than 100 nm [28,29] , crushed and dried samples were injected mercury at given pressure, a Micromeritics AutoPore IV 9500 Porosimeter (USA) was used to record the injected volume and calculated the pores diameter using the Laplace-Washburn equation [28].

SEM analyses and EDX line scanning
The scanning electron microscopy (SEM) with energy dispersive X-ray spectroscopy (EDX) was used for rock surface microstructure analysis and speci c elements identi cation (Zeiss, Heidenheim, Germany).
The working condition of SEM was at acceleration voltage of 20 kV, extraction voltage of 4.9 kV and emission current of 10 µA. The EDX line scanning was operated at an excitation voltage of 20 kV, and an energy range of 0 ~ 20 keV, the acquisition time was 100 s (point analysis) and 1800s (line scan), with processing and dead time of 35% ~ 40%. Table 1 shows the content of total organic carbon (TOC), pyrite and clay minerals before and after treatment. The minerals content is signi cantly altered after treatment, the most change presents in the chemical group (G2). As shown in Table 1, the decreased ratio of pyrite content in G2 is near triple and double greater than that in control group (G1) and biological group (G3), respectively. Whereas the increased ratio of total clay minerals (TCM) is about 8-folds and 1.5-folds greater than that in G1 and G3, respectively, the increased TCM is mainly Smectite. The alter trends of TOC content are different in G2 and G3, which decrease by 44.2% in G2 but increase by 48.3% in G3, respectively. Notably, at the initial oxidation stage, the minerals reaction rates are similar in G2 and G3, the pyrite decreased by 21.6% and 20.8%, and the TCM increased by 7.9% and 11.9% in G2 and G3 at the 7th day, respectively, but these reactions seem to subsequently slow down in G3, result in a less change of pyrite and TCM content than that in G2 at the 30th day.  Figure 2 shows the adsorption isotherms of nitrogen at liquid temperature (-197.3 °C) of each sample. The porosity characteristics can be qualitative assessed by the shape of isotherms [15,16], which similar in original samples (Fig. 2a). At the initial stage, the adsorption isotherms raise rapidly, indicating a certain amount of micropores existed. With the increase of relative pressure, the adsorption isotherm shows a signi cant hysteresis pattern closed to the Type IV isotherm [26], and follows by a steep slopes at the relative pressure range of 0.98 ~ 1.00, indicating the samples contained both mesopores and macropores.

Adsorption isotherms
After treated by different solution at predetermined period, the isotherms shape exhibits differential trends at predetermined periods ( Fig. 2b-2d). In G1, the absorbed volume is stable at low pressure (< 0.1), but gradually increases at the intermediate-and high-relative pressure. At the 30th day, the absorbed volume increases by 44.8% and 69.1% at the relative pressure of 0.5 and 1.0, respectively (Fig. 2d), indicating the mesopores and maropores all increased. In G2, the absorbed volume remarked increases throughout the range of relative pressure, and increases by 97.2%, 98.9% and 95.1% at relative pressure of 0.1, 0.5 and 1.0 at the 30th day, respectively (Fig. 2d), indicating a more increase in all scales of pores occurred in G2.
In G3, the absorbed volume increases initially, by 77.3%, 81.4% and 68.5% at relative pressure of 0.1, 0.5 and 1.0 at the 15th day, these increments even greater than that in G2 at the same time (Fig. 2c).
However, the absorbed volume seems "stead" subsequently, and it nally increases by 74.7%, 73.9% and 81.3% at relative pressure of 0.1, 0.5 and 1.0 at the 30th day (Fig. 2d). Figure 3 shows the pore size distribution characteristics results combined the nitrogen absorption isotherms and high pressure mercury injection analysis. The bimodal distribution characteristics of original samples are similar, including a major peak around 30 nm and a minor peak at 60 nm ~ 100 nm (Fig. 3a). Experimental treatments have not changed the bimodal patterns but increased the porosity ( Fig. 3b-d). The most increment of pore volume range is observed in G2, is by 72.7% and 82.9% at major and minor peak (Fig. 3d), close to the total pore volume increment (89.1%) ( Table. 2). In biological group (G3), the pore volume range increases by 36.9% at major peak but decreases by 45.7% at the minor peak (Fig. 3d), which may explain a 21.5% increment of total pore volume in G3 (Table. 2). A moderate change occurs in G1, the pore volume range increases by 12.5% and 20.6% at the major and minor peak, respectively (Fig. 3d), corresponds a 29.5% increase of total pore volume (Table. 2). Table 2 shows the calculated speci c surface areas, pore volume and average pore diameter before and after treatment. The chemical effect has the most increment both in the speci c surface areas and pore volume. Compared with original samples, these parameters increase by 5.4%, 29.5%, and 1% in G1, by 61%, 89.1%, and 2.8% in G2, and by 31.4%, 21.5%, and 1.5% in G3 (Table 2), respectively. The micro-, meso-and macro-pores differently in uence the speci c surface areas and pore volume. As shown in Fig. 4, the mesopores predominately (more than 70%) determine the speci c surface areas and pore volume, and have not signi cantly changed after treatment (less than 1%). The ratio of micropores in speci c surface areas was higher than macropores (21.35% ~ 27.51% vs. 1.5% ~ 4.35%) (Fig. 4a), but is lower (4.38% ~ 7.69% vs. 16.24% ~ 19.93%) in pore volume (Fig. 4b). The micropores decrease and the macropores increase in G2, so the increase of pore volume is more than the increase of speci c surface areas (89.1% vs. 61%). Whereas the micropores increase and the macropores decrease in G3, so the increase of pore volume is less than the increase of speci c surface areas (21.5% vs. 31.4%) ( Table 2).

SEM image
The microstructure of original samples is dominated by clay minerals, a multiple scale of pores, such as triangular pores, slit pores and cracks are observed on shale micro-surface (Fig. 5a) and signi cantly increase after treatment. In G1, the continuous clay matrix is destroyed and formed a lot of intragranularses dissolved holes, with wedge-shaped, oval, or irregular structures (Fig. 5b). The chemical effect lead to a similar but more remarked damage than G1, a large amount of channels and dissolved pores is observed on shale surface, with elliptical, cylindrical or irregularly shape (Fig. 5c), as reported by Ling, et al on weathered black shale pro les [5]. In G3, a lot of loosen, round cake-shaped and short rodlike materials are cover on the rock surface and form many irregular polygonal pores (Fig. 5d).
3.6. EDX line scans analysis Figure 6 shows the relative elemental concentrations of black shale samples from surface to interior subsurface (Fig. 6e). Elemental Si, O, C, Al, S, Fe, Ca, K, and Mg are heterogeneous distributed at each point (Fig. 6a), experimental treatments signi cantly change the relative proportion with respect to a 2 µm depth ( Fig. 6b-d). The elemental Si and O are among the most stable elements which kept a high proportion throughout the experiment. Whereas the relative proportion of elemental Al, S, Fe, Ca, K, and Mg decrease by more than 90% in G2 (Fig. 6c), and by approximately 50% in G1 (Fig. 6b) and in G3 (Fig. 6d), respectively. In consistent to the minerals alterations, the relative proportion of elemental C decreased in G1 and G2, but increased in G3.

Discussion
The correlation between minerals composition and porosity evolution is an essential topic for black oxidation process. A series of minerals reactions change the shale porosity, in turn, the evolution of porosity, especially the "active" pores -enable oxygen and uids ow into -determine the uid-rock reactions areas and control the kinetic rate of minerals reactions [3,4,12,[30][31][32][33][34][35][36]. Previous studies have well established the minerals-porosity correlations based on the chemical oxidation mechanism, but as a basic oxidation mechanism, the role of biological effect is seldom to report.
In present study, two classic effects -acid solution and A. ferrooxidans-were applied to simulate the chemical and biological oxidized black shale process, respectively. The results suggest that the chemical and biological effects led to varied alterations of minerals composition and porosity evolution, mainly due to the different alter trends of TOC content.
The acid solution was proven to accelerate the organic matters oxidation [9,10,13], that can penetrate the decrease of TOC content in G2. But the A. ferrooxidans, to our acknowledgements, majorly promoted the pyrite oxidation [37][38][39][40][41], so it hard to directly explain the increase of TOC content in G3. Previous studies proposed that the A. ferrooxidans cell bodies will attach on the rock surface when oxidize the pyrite [37,39], so the coverage of rod-like material on the shale surface (Fig. 5d) and the accumulation of elemental C (Fig. 6d) on the shale microsurface may be the attached A. ferrooxidans cell bodies. Liu et al.
[42] also reported an accumulation of organic matters on the rock surface during A. ferrooxidans oxidized ore contained pyrite; Jin et al. proposed the addition of modern organic matters at the regolith surface was associated with the biological turbulence such as the plant leaves, animal feces and microbial carcasses [4]. Therefore, the increase of TOC in G3 can be contribute to the attachment and dead A. ferrooxidans cell bodies.
Due to the ability of blocking rock pores and inhibiting uid-rock interactions, the organic matter is generally recognized as a negative factor for porosity evolution [4,10,15,43]. But researchers also reported a positive relation between the speci c surface area and organic matters content [11,16], probably since the different contributions of micropores and macropores to the speci c surface areas and pore volume. Kuila et al and Fisher et al reported a double value of speci c surface areas obtained by nitrogen absorption experiment than by high-pressure mercury intrusion test which can not measure the micropores [11,15].
We found that the chemical effect mainly increases the amount of macropores, whereas the biological effect mainly increases the amount of micropores. As shown in Fig. 4, the ratio of micropores in the speci c surface areas (more than 20%) is much more than in the pore volume (less than 5%), that can explain a more reduction of pore volume than the speci c surface areas occurred in G3. However, these micropores may be "inactive"-where oxygen and uid can not ow into, resulting in the lower uid-rock reaction areas and rates. Therefore, we found that the most reduction of TOC in G2 accompanied by the most decrease of pyrite and increase of clay minerals content among groups. Whereas the increase of TOC in G3 resulted in less change in pyrite and clay minerals content than that in G2.
Pearson correlation coe cients for relationships between pores volume and minerals content indicated that the TOC content is positively correlated with micorpores and mesopores in biological group (R = 0.72, 0.97), whereas it negatively correlates with mesopores and macropores in chemical group (R= -0.93, -0.94) (Fig. 7).
In summary, we propose that the increase of TOC in biological group impacts the correlation between minerals alteration and porosity evolution during black shale oxidation. The probably pathway is as follows: Initially, A. ferrooxidans promotes the pyrite oxidation and porosity formation, so the minerals content and porosity characteristics are similar between chemical and biological groups. With the processing of oxidation, the attached and dead A. ferrooxidans bodies blocked the large pores and decrease the rock-uid reaction areas, then slow down the minerals reaction rate in biological group.

Conclusion
A laboratory study on black shale from the Chengkou County in the Southwest Mountains of China was carried out, revealed the different correlations between minerals alteration and porosity evolution under chemical and biological effects. The summarized results indicate that, (1) The decrease of pyrite content and increase of clay minerals was more in chemical group than that in biological group, may be attributed to the chemical effect decrease but the biological effect increased the TOC.
(2) The increase of TOC increases the amount of macropores in chemical group but the decrease of TOC increase the amount of micropores in biological group, resulting in the more speci c surface areas and pore volume in chemical group.
(3) The attached and dead A. ferrooxidans may contribute to the increase of TOC and surface elemental C in biological group, which is not favour to the pore volume and led to a less uid-rock reaction areas, resulting in a lower minerals reaction rate in biological group than that in chemical group.
It suggested that the difference of chemical and biological effect in the correlation of minerals alteration and porosity evolution during black shale oxidation process may be attributed to the organic matters content alteration. Nitrogen adsorption-desorption isotherms of black shale samples in different groups at 0 day (a), 7thday (b), 15thday (c) and 30th day (d).

Figure 3
Pore size distribution characteristics obtained from the nitrogen absorption isotherm of samples in different groups at 0 day (a), 7thday (b), 15thday (c) and 30th day (d).

Figure 4
The relative retio of micro-, meso-and macro-pores in the speci c surface areas (a) and pore volume (b) of black shale samples.