Microbial Mineralization of Montmorillonite in Low-Permeability Oil Reservoirs for Microbial Enhanced Oil Recovery

Microbial mineralization is ubiquitous in the natural environment. Microbes in low-permeability reservoirs are able to facilitate alteration of the structure and phase of the Fe-poor minerals by reducing Fe(III) and inhibiting clay swelling, which is still poorly studied. This study aimed to reveal the interaction mechanism between Fe(III)-reducing bacterial strains and Ca-montmorillonite under anaerobic conditions and to investigate the extent and rates of Fe(III) reduction and phase changes with their activities. Application of Fe(III)-reducing bacteria will provide a new way to inhibit clay swelling, to elevate reservoir permeability, and to reduce pore throat resistance after water flooding for enhanced oil recovery in low-permeability reservoirs.

ABSTRACT

Microbial mineralization (corrosion, decomposition, and weathering) has been investigated for its role in the extraction and recovery of metals from ores. Here we report our application of biomineralization for the microbial enhanced oil recovery in low-permeability oil reservoirs. It aimed to reveal the etching mechanism of the four Fe(III)-reducing microbial strains under anaerobic growth conditions on Ca-montmorillonite. The mineralogical characterization of Ca-montmorillonite was performed by Fourier transform infrared spectroscopy, X-ray powder diffraction, scanning electron microscopy, and energy-dispersive spectrometry. Results showed that the microbial strains could efficiently reduce Fe(III) at an optimal rate of 71%, alter the crystal lattice structure of the lamella to promote interlayer cation exchange, and efficiently inhibit Ca-montmorillonite swelling at a rate of 48.9%.

IMPORTANCE Microbial mineralization is ubiquitous in the natural environment. Microbes in low-permeability reservoirs are able to facilitate alteration of the structure and phase of the Fe-poor minerals by reducing Fe(III) and inhibiting clay swelling, which is still poorly studied. This study aimed to reveal the interaction mechanism between Fe(III)-reducing bacterial strains and Ca-montmorillonite under anaerobic conditions and to investigate the extent and rates of Fe(III) reduction and phase changes with their activities. Application of Fe(III)-reducing bacteria will provide a new way to inhibit clay swelling, to elevate reservoir permeability, and to reduce pore throat resistance after water flooding for enhanced oil recovery in low-permeability reservoirs.

INTRODUCTION

Many oil reservoirs worldwide are increasingly depleted every year. These reservoirs feature low stratum pressure, low permeability, low single well production, and increased water absorption, which have made further crude oil production more difficult. Low-permeability reservoirs mainly contain montmorillonite, illite, mixed montmorillonite/illite, chlorite, and other irregular clay minerals. Among these, montmorillonite and its mixed layer of minerals may swell with water contents, while illite and chlorite barely swell. The clay swelling would significantly decrease the reservoir porosity, permeability, and oil recovery in low-permeability reservoirs. There is thus a need to develop efficient strategies to inhibit montmorillonite swelling for enhanced reservoir permeability and oil recovery when water flooding exists.

A combination of well arrangement, optimization of water flooding, fracturing, and acidizing is used in low-permeability oilfields (1). Cationic polymers and quaternary ammonium salt are major chemical agents to inhibit clay swelling in low-permeability reservoirs. Despite their inhibitory or shrinking effect on montmorillonite in reservoirs, these chemical agents can cause permanent damage to reservoirs and induce adverse effects on the ecological environment. These technologies are also laborious and entail high energetic, economic, and environmental costs (2). In addition, the injection of chemical agents into low-permeability formations is somewhat challenging and may produce secondary environmental pollution. In recent years, the interaction between microbes and montmorillonite minerals has become a hot spot in the research community of geological biology (3). Microbial enhanced oil recovery (MEOR) is an environmentally friendly tertiary recovery method that involves the use of microbial communities and their metabolic products, including biogas, biosurfactants, biomass, and acids, to extend the production life of oil wells (4–6). Indigenous Fe(III)-reducing microbes, such as Shewanella (7), Bacillus (8), Thermoanaerobacter ethanolicus (9), and Desulfovibrio (10), which are highly adaptable to the oil reservoir environment, are widely used in MEOR (6). These bacteria alter the mineral structure in clay and convert montmorillonite to illite or other secondary minerals for inhibition of clay swelling in low-permeability reservoirs.

At present, the mechanism of Fe(III) reduction in low-permeability oil reservoirs is still poorly understood. By definition, microbial dissimilatory Fe(III) reduction refers to the process of oxidizing organic substance under anaerobic conditions with Fe(III) as the terminal electron acceptor. The relevant literature (11) reviewed three kinds of possible mechanisms of microbial enzyme to catalyze dissimilatory Fe(III) reduction and their molecular regulation by the interaction among the microbes, Fe(III) oxides, siderophores, and electron-shuttling compounds. Many membrane-bound cytochrome proteins are involved in this process, resulting in a complicated regulation network.

The mineralization (corrosion, decomposition, and weathering) of montmorillonite mediated by the microbes represents an emerging opportunity for researchers to design innovative solutions for the problems associated with MEOR. The mineralization activities on montmorillonite by Fe(III) reduction are realized by the catalyzed conversion of montmorillonite to illite and quartz (12–14). These activities will improve the seepage environment of oil water fluid, strengthen pore flow resistance, repress clay swelling, and increase oil recovery in low-permeability reservoirs. The mechanism of microbial etching on the structure and phase of Fe-poor minerals in low-permeability reservoirs has been poorly studied.

Our novel study involves four Fe(III)-reducing bacterial strains for the structure alteration of Ca-montmorillonite (Ca-10) in low-permeability reservoirs (Fig. 1). The mineralogical characterization of the microbial cell metabolism on the minerals was performed using Fourier transform infrared spectroscopy (FTIR), X-ray powder diffraction (XRD), scanning electron microscopy (SEM), and energy-dispersive spectrometry (EDS). It established the interaction between microbes and reservoir minerals from the perspective of biochemistry to improve oil recovery in low-permeability reservoirs.

FIG 1.

Schematic diagram of Ca-montmorillonite phase transformation.

RESULTS AND DISCUSSION

Fe(III) reduction by bacteria. (i) Changes in pH and total protein content.

The physiological activities of the four Fe(III)-reducing bacterial strains—CA, KB, TC, and TD—were measured on the basis of their acid production and the total protein content (15). As shown in Fig. 2a, the pH in the control remained at about 6.9 during the whole experimental period, while the pHs of CA, KB, TC, and TD dropped from the initial pH of 6.9 to about 5.8, 4.9, 5.9, and 4.3, respectively. Our measurements have shown that the four Fe(III)-reducing bacterial strains could produce organic acids, such as acetic acid, propionic acid, butyric acid, isobutyric acid, and valeric acid (data not shown), which caused the decrease in pH (16). The growth of CA, KB, TC, and TD in acidic reservoir environments enables their Ca-10 dissolution to enhance the contact between microbial cells and Ca-10 surface area and to promote Fe(III) reduction (17). It was found that due to elevated oxidation reduction potential in the acidic environment of the reservoir, Fe(III)-reducing bacteria regenerate ferric iron to provide more energy to sustain their growth for increased pore diameter and improved clay permeability.

FIG 2.

Changes of pH (a) and total protein concentration (b) in the system.

As shown in Fig. 2b, the total protein in the control remained at about 5.5 mg/liter throughout the whole experimental period. The total protein changes in the experimental group can be divided into two stages. At stage 1, the total protein concentration increased rapidly due to the activities of CA, KB, TC, and TD under anaerobic growth conditions after inoculation (Fig. 2b). At stage 2, growth of CA, KB, TC, and TD upon the consumption of nutrients and Fe(III) from Ca-10 reached the stable state and then gradually began to decline, eventually reaching steady state. This phenomenon may be attributed to bacterial overpopulation, resulting in the lack of space and nutrition (18). From the fifth day to the end of the experiment, the total protein concentrations for the four bacterial strains were maintained at 31.5, 33.2, 28.3, and 21.2 mg/liter (Fig. 2b).

(ii) Fe(III) reduction and Fe(II) production during Ca-10 alteration.

CA, KB, TC, and TD used Fe(III) in the Ca-10 as their electronic receptor to reduce it to Fe(II) in the etching experiment. Fe(II) accumulated in the solution as ions. It was then taken up by the microorganisms to provide the energy required for their cell metabolism (19). Therefore, the concentrations of Fe(II) and Fe(III) in the suspension were used to evaluate the degree and efficiency of Fe(III) reduction in Ca-10 by CA, KB, TC, and TD.

The concentrations of Fe(II) and Fe(III) in the growth media were measured to determine the change of Fe(III) content in Ca-10 by microbial erosion versus time (Fig. 3a), a similar comparison was made for the reduction of the generated Fe(II) concentration (Fig. 3b), ratio of Fe(II) to Fe(III) (Fig. 3c), and ratio of Fe(II) to total iron (Fig. 3d). The Fe(III) and Fe(II) concentrations in the control were measured as 5.8 and 9.1 mg/liter, respectively, which may be due to the contents of Fe(II) and Fe(III) ions in the Ca-10 during the experimental period (Fig. 3a and b).

FIG 3.

Changes of the extent of Fe(II) and the reduction rate of Fe(II) in the system. (a) Fe(III) concentration; (b) Fe(II) concentration; (c) ratio of Fe(II) to Fe(III); (d) ratio of Fe(II) to total Fe.

After addition of the four bacterial strains into the reactors, the concentration profile of Fe(III) and Fe(II) was divided into two stages. At the first stage, Fe(III) and Fe(II) were rapidly increased from the lower initial concentration. The highest Fe(III) concentration in the dissolved and reduced Ca-10 and the maximum Fe(II) concentration were obtained at the 4th day (Fig. 3b). It was seen that the maximum Fe(II) concentration appeared at the same time as that of the total protein, indicating that the reduction rate of Fe(III) in Ca-10 was closely related to the enzymatic activities of the bacterial strains. At the second stage, the Fe(III) accumulation rate and the Fe(II) reduction rate decreased gradually, with the concentrations of Fe(III) stabilizing at 225.7, 322.5, 301.2, and 233.8 mg/liter, respectively, while Fe(II) reductions were at 103.8, 215.8, 263.6, and 203.5 mg/liter, respectively. In addition, it was found that the production of Fe(II) in Ca-10 via the Fe(III)-reducing bacteria was positively correlated with the reduction of Fe(III), indicating that the cumulative Fe(II) is mainly produced from the free state of ferric ion in the reaction system (20).

From Fig. 3c, it is evident that the rates of reduction of Fe(III) in Ca-10 for TC and TD (87.1% and 86.8%, respectively) were higher than those of CA and KB. However, the microbial reduction time of Fe(III) in Ca-10 by CA and KB was short, and the Fe(III) reduction rates (68.1% and 80.3% by the two microbes, respectively) were reached by 48 h; so were the maximum amounts of protein produced by CA and KB (Fig. 2b). In addition, as shown in Fig. 3d, the reduction rates of total Fe(III) in Ca-10 gradually decreased and finally stabilized, reaching the highest levels—47.1%, 62.9%, 67.1%, and 71.1%—at the 4th day. These values were higher than the iron reduction rates of Shewanella putrefaciens CN32 (19% to 40%) and Thermus scotoductus (32% to 44%) in the literature (21). In addition, the peak Fe(III) reduction rates in Ca-10 by CA, KB, TC, and TD were in accordance with the maximum enzymatic activities of the four microbial strains. These results suggest that the production and concentration of Fe(II) are closely related to the activities of CA, KB, TC, and TD.

Characteristics of Ca-10 before and after treatment. (i) FTIR analysis of chemical bond vibration.

Comparing with the infrared spectra of Ca-10 before and after pitting by the four Fe(III)-reducing bacterial strains, we observed similar characteristic peak distributions, both for the Ca-10 samples and the control, corresponding to the chemical bond vibration in the crystal structure. After the reduction of Fe(III), there appeared to be no change in the absorption bands, while there was a slight reduction in intensity (Fig. 4). A previous study (22) has shown that the FTIR spectrum of Ca-10 can be broken down into four main regions spanning from 4,000 to 400 wavenumbers (per centimeter), that is, 3,800 to 3,200, 1,800 to 1,400, 1,400 to 1,000, and 950 to 700 cm⁻¹ (23). This absorption peak range is characterized to be the νM-OH (M = Mg, Al, or Fe) stretching bands (24), the O-H deformed vibrating belt, the Si-O stretching vibration zone, and the νM-OH deformation vibration band in the octahedron, respectively, in which Si-O peak is a typical characteristic peak of silicate minerals (25).

FIG 4.

FTIR changes of Ca-10 before and after reduction by strains CA (a), KB (b), TC (c), and TD (d).

As shown in Fig. 4, the intensity of the νM-OH stretching vibration at 3,750 cm⁻¹ was sharply weakened after the erosion of Ca-10 by CA, KB, TC, and TD. The O-H stretching vibration peaks at 3,460 and 3,320 cm⁻¹ were slightly shifted (Fig. 4c and d) or disappeared (Fig. 4a and b) after the reduction of Ca-10. The water peak at 3,560 cm⁻¹ was weakened, due to the loss of crystal water in Ca-10 causing hydration swelling. The figure shows also that CA and KB became weakened in the νM-OH stretching zone (Fig. 4a and b), which was correlated with the high Fe(III) reduction rate in CA and KB in the experimental period.

In Fig. 4, the O-H deformed vibrational peaks were significantly attenuated between 1,570 and 1,360 cm⁻¹ after Ca-10 reduction; this peak disappeared at 1,360 cm⁻¹ due to the action of KB (Fig. 4b), indicating that the microbes impacted the distribution of O-H and interlayer H₂O in Ca-10. At the same time, the Si-O peaks at 1,120 and 1,043 cm⁻¹ were reduced substantially after the microbial mineralization. The results described above indicate that the bacterial reduction increased the susceptibility of mineral dissolution, changed in short-range order within the clay crystal, and altered the vibrational energies of structural O-H and Si-O (26). This shift to a lower frequency was consistent with previous studies of bacterially reduced smectites (24). Thus, we observed changes in the relative positions and the intensities of the parallel layer stretching vibration and vertical layer stretching vibration of Si-O in the crystal structure of Ca-10, as well as changes in the vertical and horizontal vibrations of the vertical layer.

Furthermore, it is evident from Fig. 4 that the symmetry of the silicon tetrahedron and the strength of the Si-O in the Ca-10 crystal structure were altered, with the intensification of Si-O telescopic vibration zone observed for TC and TD cultures. In general, the shift of the structural OH and Si-O vibration, as well as the loss of intensity of OH groups, indicates that the bonding and/or symmetry properties in the octahedral and tetrahedral sheets change as Fe(III) is reduced to Fe(II). This observation may be interpreted to mean that as Fe reduction has modified the crystal structures of Fe-bearing smectites, the overall effects are modest and of about the same extent as for dithionite at similar levels of reduction.

For Fig. 4, it should be noted that some νM-OH deformation bands (M = Mg, Al, or Fe) at 700 to 950 cm⁻¹ in FTIR spectra were present, which is characteristic of the vibrations of octahedral cations. In addition, there were strong bands observed at 746, 812, 845, and 924 cm⁻¹ in the M-OH deformation region, which could be assigned to FeMg-OH, FeFe-OH, AlMg-OH, and AlFe-OH, respectively (27). Moreover, the peaks corresponding to the period after reduction were obviously shortened, indicating that the lattice structure of Ca-10 had shrunk. In summary, the oxidation state of Fe in smectite clay minerals results in significantly altered clay structure, including decrease in structural OH content, migration of Fe from cis- to trans-octahedral sites, formation of trioctahedral domains, and defect sites within the octahedral sheet.

In addition, the absorption peaks at 2,892 and 2,364 cm⁻¹ after microbial mineralization are seen to be in accordance with the stretching vibration of lipid C-H (28). This may be due to the certain organic products adsorbed on the mineral surface. These results showed that the chemical bond vibration of Ca-10 was altered after the mineral etching by CA, KB, TC, and TD at the cost of Si and/or O atom disassociation around the Fe element, which also altered the structural Si-O vibrational and nuclear energies. This phenomenon resulted in the collapse of the silicon tetrahedron and the aluminum oxide octahedral framework in Ca-10 and led to the deformation of the overall crystal structure. Additionally, the sensitivity of Al sites to the change in the Fe sites is consistent with the change in the band at 3,560 cm⁻¹ observed in the OH stretching region. It indicates that Fe reduction altered the structure of non-Fe octahedral sites.

(ii) XRD analysis of mineralogical composition.

The mineralogical composition and morphological characteristics of the solid phases, as well as the effects of the bacteria and their products in Ca-10, were investigated by XRD (29–32) for changes of Fe(III) in the form of octahedral and tetrahedral structures caused by the erosion by the four bacterial strains in Ca-10 (Fig. 5). The resultant XRD data were converted with Origin software to obtain a diffraction peak spectrum.

FIG 5.

XRD changes of Ca-10 samples before and after reduction by strains CA (a), KB (b), TC (c), and TD (d).

In Fig. 5, it can be seen that the d₍₀₀₁₎ (the interlayer spacing of clay minerals) peaks in Ca-10 were shifted from 14.486 to 11.861, 11.827, 11.7825, and 11.7369 Å during structural Fe(III) reduction via the pitting of CA, KB, TC, and TD, respectively. This shift led to both the decrease in peak intensity and the increase in half-peak width. This phenomenon indicated that the crystal structure of Ca-10 was deformed, the crystal properties changed, and the crystallinity was reduced. The decrease in interlayer spacing also increases the thickness of smectite particles, raising the number of layers per packet from 2 to 4 to more than 50 (33) in bioreduced smectite. This transformation clearly impacts particle size and settling properties of the clay, as well as permeability and hydraulic conductivity (34). Additionally, it suggested that the mechanism of the newly formed illite was closely related to the bioreduction, also implying the reductive dissolution of Ca-10 and the simultaneous precipitation of illite.

In Fig. 5, it can be seen that the peaks at the same time points appeared at d₍₀₀₃₎ of 11.01 Å (2θ = 8.07°), d₍₀₀₃₎ of 10.776 Å (2θ = 8.23°), d₍₀₀₄₎ of 10.52 Å (2θ = 8.42°), d₍₀₀₄₎ of 10.28 Å (2θ = 8.62°), and d₍₀₀₄₎ of 10.09 Å (2θ = 8.776°). The layer spacing of these peaks was similar to that of illite (10 Å), indicating that a small amount of Ca-10 was converted into illite minerals by microbial mineralization. According to these phenomena, we found that the crystal size distribution (CSD) of neoformed illite phase is different from that of Ca-10 by microbial reduction. In addition, two additional peaks at d₍₀₀₅₎ of 7.561 Å (2θ = 11.638°) (Fig. 5a) and d₍₀₀₅₎ of 6.413 Å (2θ = 11.771°) (Fig. 5d) were observed, which were possibly the Fe(II) formed during Fe(III) reduction in Ca-10. It has been observed that ferrous ions are able to combine with Na⁺, CO₃²⁻, and PO₄³⁻, as well as potassium phosphate to form the siderite [Na₃Fe(PO₄)(CO₃)] (35) and phosphorous ore (FeHPO₄) (Fe²⁺ + HPO₄²⁻→FeHPO₄) (36). These results agree with previous studies on minerals such as illite (31), siderite, and carbonate minerals in the literature (32).

In Fig. 5, there appear to be some peaks in the spectra at d₍₀₀₆₎ of 4.511 Å (2θ = 19.96°), d₍₀₀₇₎ of 3.419 Å (2θ = 26.77°), d₍₀₀₈₎ of 2.923 Å (2θ = 31.789°), and d₍₀₀₈₎ of 2.692 Å (2θ = 35.12°), with an enhanced peak intensity. This indicates that the crystal structure of Fe(III) in Ca-10 octahedral and tetrahedral lamellae was etched by CA, KB, TC, and TD, resulting in its structural deformation. Some of silicon tetrahedron and aluminum oxide octahedron layers in Ca-10 have even been eroded to form small d-value minerals, such as quartz, calcite, calcite, and dolomite (37). Conversely, the formation of calcite was unexpected considering the potentially abundant Fe(II) during iron dissolution in the solution, which should yield Fe-rich carbonate before calcium carbonate precipitates (31).

(iii) SEM analysis of morphology changes.

The surface morphology of the Ca-10 samples was observed by SEM. As shown in Fig. 6a, SEM revealed that the abiotic control did not undergo any mineralogical changes, displaying a typical lamellar structure and similar flaky texture and uniform distribution.

FIG 6.

SEM images of Ca-10 samples before and after reduction by strains CA (b), KB (c), TC (d), and TD (e), with untreated Ca-10 (a) as a control. The red squares correspond to the EDS scan area in the corresponding panels in Fig. 7.

In comparison, the surface of Ca-10 after the mineralization by CA and TC became dissipated and the large area of the lamellar structure collapsed or disappeared (Fig. 6b and d). It is interpreted that Fe(III) reduction in Ca-10 by CA enabled direct contact between the cell membrane and the clay surface areas, leading to the production of a chemical chelating agent (38). We further hypothesized that the erosion of CA may be caused by a close association between bacterial cells and biofilm which has been reported elsewhere (39, 40), which enables the enzymatic mediation of electron transfer to Fe(III) in Ca-10 (Fig. 6b) (41). Similarly, TC may erode and reduce Fe(III) in Ca-10 by its cell attachment through the extracellular polymer, using nanowire as the electron-chelating agent and electron acceptors as the electron transfer for reduction (42). The above assumption was supported by the fact that when using structural Fe(III) as the sole electron acceptor, cells may boost Fe reduction by direct contact with the clay mineral or by releasing organic acids or chelators that facilitate solubilization of the structural Fe(III) (32, 43). In addition, a decrease in specific surface area and the associated collapse of superimposed smectite layers clearly will also affect other important clay mineral properties. As the layers collapse, a fraction of the interlayer cations also likely will become less exchangeable.

As shown in Fig. 6c and e, the surface of Ca-10 treated with KB and TD shows pitting that is distorted into loose porous flocs. The surface structure and electrical properties of the microorganisms allowed their adherence to the mineral surface (44), generating a large number of corrosion pits and causing significant corrosion. These corroded pits would not only increase the contact surface area between the bacteria and Ca-10 (45) but also reduce Ca-10 crystallization. This would promote the Fe(III) reduction, cause serious deformation of the mineral structure, and increase the binding sites for the generation of secondary minerals. Figure 6b to e show that erosion broke the charge balance between the trivalent iron ion and other cations, changed the bond price structure and balance among Si, Al, and Fe, and caused the collapse of the silicon tetrahedron and aluminum oxide octahedral frame and destruction of the overall crystal structure, thus enabling the formation of illite and other nonswelling secondary minerals (46).

(iv) Analysis of the element contents by EDS.

To explain in depth the changes in the element contents in Ca-10 before and after the reduction of bacteria, EDS was performed for the Ca-10 samples originating from SEM (Fig. 6).

As shown in Fig. 7, the Si and Al contents in Ca-10 minerals were drastically reduced due to the microbial mineralization of CA, KB, TC, and TD. It is shown that the Si/Al mass ratios decreased from 3.04 to 2.67, 1.96, 1.88, and 1.75 and the Fe(III) contents (by mass percent) were reduced from the initial 2.29% to 1.17%, 0.94%, 0.85%, and 0.66%, respectively. These results were consistent with the reduction efficiency of Fe(III) in Ca-10, and excess Fe from the reductive dissolution of Ca-10 resulted in the precipitation of minerals as by-products. It seems that the decrease in Si, Al, and Fe contents in Ca-10 with the aid of CA, KB, TC, and TD caused the distortion and collapse of the silicon tetrahedron and aluminum oxide octahedral crystal in the structure in Ca-10 (47). It is further deduced that a large amount of Si, Al, and Fe elements was released (such elements will support bacterial growth in the depletion of nutrients and vitamins) by microbial activities and the new minerals were formed. The distortion and collapse of crystal structure of Ca-10 thus triggered its conversion into secondary minerals, such as illite and calcite, which reduced hydration expansion.

FIG 7.

EDS changes of Ca-10 samples (corresponding to Fig. 6) before and after reduction by strains CA (b), KB (c), TC (d), and TD (e), with untreated Ca-10 (a) as a control.

In addition, it was observed that the P and K contents of Ca-10 were increased. On one hand, the increase in P content was mainly due to the formation of blue iron ore via the reaction of the phosphate and Fe(II) in the medium (48), which was consistent with the appearance of blue iron ore in XRD. On the other hand, the increases in K content (by mass percent) from the initial 0.12% to 1.99%, 3.02%, 1.3%, and 3.46%, respectively, might be caused by the exchange of K⁺ and Na⁺ in Ca-10. Additionally, when the positive charge in the octahedral sheet was decreased as a result of reduction of Fe(III), K was taken into the interlayers to balance the charge and for the formation of illite, which was similar to the results of previous studies (31). The bioreduction increased the cation exchange capacity in Ca-10 (49) and triggered the replacement of intermolecular Na⁺ by K⁺ in the culture medium.

Finally, the conversion of montmorillonite to illite and other secondary minerals was confirmed by the data shown in Fig. 6 and 7. At first, the microbes reduced the montmorillonite octahedron and tetrahedral structure of Fe(III) by cell metabolism, which promoted the substitution of other ions in the lattice to increase the negative charge between the layers and allow the entrance of K⁺ into the interlayer structure. Then the smectite-illite transformation occurred and subsequently formed illite and other secondary minerals. Finally, the microbial erosion of Fe(III) in montmorillonite, the microbial erosion of Fe(III) in montmorillonite caused the distortion and collapse of the silica tetrahedron and aluminum oxide octahedron structure and released a large amount of Si and Al and other elements in montmorillonite crystals. These cations were recrystallized from K⁺ in the culture medium to form the small particles of illite and other secondary minerals.

Changes of swelling properties of Ca-10.

The swell inhibition efficiency of Ca-10 was determined by the static test method in accordance with the Chinese National Standard (SY/T5613-2000) (50). The Ca-10 samples were measured in both water and kerosene before and after microbial reduction by CA, KB, TC, and TD. The differences in the swelling volume in the blank versus test groups for 24 h were compared and are shown in Table 1. It can be seen that CA, KB, TC, and TD displayed a high efficiency of inhibition for Ca-10 swelling, while the Fe(III) in Ca-10 was successfully reduced by bacterial mineralization. Among the four bacterial strains, TD showed the best effect, with optimal inhibitory rates of 48.9% in water and 92.6% in kerosene, respectively. The results showed that the swelling volume of Ca-10 was significantly reduced; in particular, it was of a low degree in kerosene due to Fe(III) reduction in low-permeability reservoirs to change the crystalline structure. The process resulted in the mixed layer of montmorillonite-illite in minerals or illite and other secondary minerals. The inhibitory rate was similar to that of the chemical antiswelling agent used in oilfields as described in the literature (51).

TABLE 1.

Swelling properties of Ca-10 samples in water and kerosene before and after reduction by CA, KB, TC, and TD^a

Strain	Solution	Swelling vol (ml) at:		Difference in swelling vol (ml)	Inhibition of Ca-10 (%)
		0 h	24 h
None (blank)	Water	0.50	0.96	0.46	0
CA	Water	0.64	0.935	0.295	35.87
KB	Water	0.65	0.925	0.275	40.22
TC	Water	0.65	0.90	0.25	45.65
TD	Water	0.68	0.915	0.235	48.91
None (blank)	Kerosene	0.494	0.63	0.136	0
CA	Kerosene	0.46	0.48	0.02	85.29
KB	Kerosene	0.45	0.465	0.018	86.76
TC	Kerosene	0.43	0.448	0.015	88.97
TD	Kerosene	0.45	0.46	0.01	92.65

^aCa-10 was used at 0.1 g.

Conclusions.

Microbial erosion by four Fe(III)-reducing bacterial strains on the Fe-poor Ca-10 was carried out under anaerobic conditions in this study. It was observed that the microbes could secret organic acids to etch Ca-10 clay for the deformation of the surface area to promote Fe(III) reduction, with an optimal reduction rate of 71%. The microbial mineralization changed the surface morphology of Ca-10 and induced the distortion of the lattice and the structure for enhanced interlayer cation exchanges to decrease the interlayer space for reduced clay swelling.

In addition, mechanism underlying the conversion of montmorillonite to illite and other secondary minerals was found. First, the microbial reduction of montmorillonite octahedron and tetrahedral structure of the Fe(III) would promote the substitution of other ions in the lattice to increase the negative charge between the layers and the process of K⁺ entering the interlayer structure. This would enhance the smectite-illite mixture to form the illite and other secondary minerals. Second, the microbial reduction of montmorillonite Fe(III) would cause the structure of the silica tetrahedron and aluminum oxide octahedron to be distorted and collapse to release a large amount of Si and Al, as well as other elements in montmorillonite crystals. As a result, these cations were recrystallized from K⁺ in the culture medium to form the small particles of illite and other secondary minerals. Finally, the resultant inhibition efficiency for the Ca-montmorillonite swelling showed an optimal inhibitory rate of 48.9%.

MATERIALS AND METHODS

Bacteria and clay minerals.

The microbes were isolated from samples of (i) production fluids in the Da Qing oilfield in China, (ii) a circulating drilling fluid collected at a 1,000-m depth in the Chinese continental deep drilling, (iii) production fluid from the Liao He oilfield in China, and (iv) production fluid from the Da Gang oilfield (52). They were identified by 16S rRNA gene analysis to exhibit a similarity of 98% to 99% to Enterobacter aerogenes (NCBI accession number EFJ63628.1), 99% to Pseudomonas (NCBI accession number KPW20975.1), 98% to 99% to Bacillus cereus (NCBI accession number COB27440.1), and 99% to Proteus (NCBI accession number OSB73291.1), respectively, and are referred to here as CA, KB, TC, and TD.

The experimental protocol was as follows. Ten milliliters of production fluids and 90 ml of iron-reducing bacteria enrichment medium were mixed and kept for 5 to 7 days under anaerobic conditions at 35°C. When the color of the enrichment medium changed from yellow to light green or the medium became colorless, 2.0 ml of the mixture was transferred to a fresh iron-reducing enrichment medium for 3 times. It was then stretched on the petri dish to pick a single colony. Iron-reducing bacterial enrichment medium was made based on the modification of the medium in the literature (53) as follows: C₆H₁₂O₆·H₂O, 10 g/liter; NH₄Cl, 1.0 g/liter; KH₂PO₄, 0.25 g/liter; K₂HPO₄, 0.72 g/liter; MgSO₄·7H₂O, 0.5 g/liter; CaCl₂, 0.1 g/liter; FeC₆H₅O₇·5H₂O, 3.3 g/liter; and trace element solution, 1.0 ml.

Ca-10 clay was purchased from the Jian Qing District of Liao Ning Province in China. The total Fe content in Ca-10 is 3.7%. The chemical composition of Ca-10 as determined by electron microprobe analysis was (Ca_0.19Na_0.17K_0.08)_0.44(Al_1.36Mg_0.32Fe_0.26)_1.94(Si_3.87Al_0.13)_4.0 O₁₀ (OH) _2.0·nH₂O. The Ca-10 clay was thoroughly soaked and slightly ground or sonicated in an ultrasonic water bath to disperse clay aggregates but not to break individual particles. Bulk clay was size fractionated at 0.5 to 2.0 μm. After drying at 80°C for 2 h, the size fractions of Ca-10 were ready for experiments.

Interaction of Fe(III)-reducing bacteria and clay minerals.

The microbial Fe(III) reduction on Ca-10 was carried out under anaerobic conditions. Sucrose was used as the main carbon source, sodium acetate as the electron donor, and the Fe(III) in Ca-10 as the final electron acceptor in the culture medium. Growth medium consisted of (per liter of deionized water) sucrose (10 g), NaCl (5 g), NH₄Cl (2 g), KH₂PO₄ (1.4 g), K₂HPO₄ (3.7 g), MgSO₄·7H₂O (0.5 g), CaCl₂ (0.07 g), NaCH₃COOH (10 mM), yeast extract (0.5 g), vitamin solution (1.0 ml), trace element solution (1.0 ml), and 0.1% resazurin (redox indicator; 1.0 ml) (53). The final pH of the medium was adjusted to 7.0 with 0.1 N NaOH. The medium was made anoxic in 125-ml serum bottles with O₂-free N₂-CO₂ gas mix (80:20) via passage through a sterile needle. The bottles were sealed with thick butyl rubber stoppers and then sterilized via autoclaving at 121°C for 15 min. The serum bottles contained 0.1 g of Ca-10 and 80 ml of culture medium [C_Fe(III) = 404 mg/liter] [C_Fe(III) refers to the concentration of total Fe(III) from Ca-10 in the reaction system]. Before inoculation, the bottles were stored at 65°C overnight to allow the sulfur functional group of the yeast extract to remove any residual O₂. The experimental group was inoculated in LB medium (10 g/liter of peptone, 10 g/liter of yeast powder, and 10 g/liter of NaCl) containing enrichment cultures of CA, KB, TC, and TD for 36 h under anaerobic conditions. The final concentration of 10⁷ to 10⁸ cells/ml (54) was based on acridine orange direct counts, whereby acridine orange was injected into the tubes in the experimental group and the blank group (inoculated with the same amount without bacterial medium). The tubes were incubated at 35°C with shaking at 150 rpm for 10 days. To obtain reliable data, all treatments were performed in duplicate. All solutions and cultures were transferred by using sterile needles and syringes.

Analysis of fluid composition and characterization of Ca-10.

Analysis of fluid chemistry included pH, total protein content, and Fe(II) and Fe(III) concentrations. First, 3.0 ml of the cell-mineral mixed suspensions was taken from the anaerobic console each day, 0.5 ml of which was collected using a sterile syringe. It was added to plastic tubes containing 0.1 ml of 5.5 M NaOH, and then 0.5 ml of 10% physiological saline was added. The mixture was lysed in boiling water for 10 min, cooled to room temperature, and then centrifuged. The total protein was determined in accordance with the modified Coomassie brilliant blue method. Absorbance at 595 nm was measured on a UV-2802 PC UV-visible (UV-VIS) spectrophotometer, and the total protein concentration was calibrated by absorbance. The viability of the microbes grown in the mixture was evaluated on the basis of change in total protein content. Second, the extent of bacterial reduction of Ca-10 was monitored by measuring Fe(II) and Fe(III) concentrations with the use of improved 1,10-phenanthroline spectrophotometry (55). At selected time points, 0.5 ml of the cell-mineral mixture suspension was collected using a sterile syringe and added to plastic tubes containing 0.5 ml of 1 N HCl. The mixture suspension was allowed to stand in HCl for 24 h before the measurement of Fe(II) and Fe(III) concentrations. Then 1.0 ml of the mixture suspension of the solution was filtered through a 0.22-μm filter and then centrifuged at 6,000 rpm for 10 min. Fe(II) and total Fe contents of a 0.2-ml suspension were measured. These data were used to calculate the amount of Fe(II) and Fe(III) reduction in the system. Before the remaining 1.0 ml of cell-mineral mixture suspension was allowed to settle, the pH of the supernatant was measured in an oxygen-free glove bag to avoid artifacts resulting from the reoxidization of sulfides to sulfate.

Finally, after the mineralization was carried out for 10 days, the solid product was collected, dried at 60°C, and ground in an agate mortar. FTIR and XRD were conducted to determine the changes in chemical bonds and crystal structures in Ca-10 before and after bacterial erosion. The surface morphology of the Ca-10 clay was observed by SEM and EDS for mineral chemical composition analysis.

Analytical methods. (i) FTIR.

FTIR spectroscopy was used to monitor changes in chemical bond vibrations and changes in the strength of functional groups in Ca-10 mineral structures before and after bacterial treatment. For this analysis, 200 mg of KBr and 2.0 mg of sample powder were weighed, mixed, and ground to fully disperse Ca-10 clay in KBr. The samples were analyzed in the diffuse reflectance mode using a Perkin-Elmer Frontier infrared spectrometer. Fifty scans over the range 400 to 4,000 cm⁻¹ with a spectral resolution of 4 cm⁻¹ were accumulated for each spectrum.

(ii) XRD.

Both control and Ca-10 samples were examined by XRD to identify the morphological and structural changes resulting from microbial etching. Samples for XRD analysis were prepared from specimens that were dispersed and washed in 95% ethyl alcohol six to eight times to remove growth medium and salts. The treated sample was dispersed in ethyl alcohol, pipetted onto circular glass slides, and dried in air at room temperature. Powder XRD patterns were collected by a Scintag X-ray powder diffractometer using Cu-Kα (λ = 0.15418) radiation, a fixed slit scintillation detector, and 1,200 W of power (40 kV; 100 mA). XRD slides were scanned in 0.02 2θ steps with a count time of 2 s per step and a scanning range of 5 to 50° 2θ.

(iii) SEM and EDS.

Mineralogical changes were further studied via SEM. Ca-10 samples were prepared following a previously published procedure (29). In brief, the cell-mineral mixture suspensions were fixed in 2.5% glutaraldehyde in a bicarbonate solution and one droplet of fixed suspension was placed onto the surface of a glass coverslip that was precleaned with 1.0 mg/ml of polylysine solution. Ca-10 particles were allowed to settle onto the coverslip for 15 min. The sample-coated coverslips were sequentially dehydrated using various proportions of ethanol and distilled water, followed by critical point drying. The coverslips were mounted onto SEM stubs and Au coated for observation using a Zeiss low-vacuum SEM instrument, which was operated at an accelerating voltage of 10 to 15 kV. The EDS (equipment model X-Max 50) provided a primary means for mineralogical identification. A certain working distance (8.5 mm) and higher beam current (50 to 70 μA) were used for qualitative EDS.

(iv) Swell inhibitory efficiency of Ca-10.

The swell inhibition efficiency of Ca-10 was determined by the static test method in accordance with the Chinese National Standard (SY/T5613-2000) by the static test method (50). Inhibitory efficiency was calculated using the following equation: η = [(V₀ − V)/V₀] × 100, where V₀ is the swell volume of Ca-10 without bioreduction in the reaction solution, V is the swell volume of Ca-10 with bioreduction in the reaction solution, and η is the swell inhibition efficiency.