Stability Evaluation of Cation Bridging on Muscovite Surface for

Apr 10, 2017 - covered muscovite surface significantly enhances the adsorp- .... mineral muscovite in various cation exchange states, which depends on...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/JPCC

Stability Evaluation of Cation Bridging on Muscovite Surface for Improved Description of Ion-Specific Wettability Alteration Kazuya Kobayashi,†,‡ Yunfeng Liang,*,‡,§ Sumihiko Murata,*,‡ Toshifumi Matsuoka,‡,∥ Satoru Takahashi,⊥ Ken-ichi Amano,† Naoya Nishi,† and Tetsuo Sakka† †

Department of Energy and Hydrocarbon Chemistry, Kyoto University, Kyoto 615-8510, Japan Environment and Resource System Engineering, Kyoto University, Kyoto 615-8540, Japan § Center for Engineering, Research into Artifacts (RACE), University of Tokyo, Chiba 277-8568, Japan ∥ Fukada Geological Institute, Tokyo 113-0021, Japan ⊥ Japan Oil, Gas and Metals National Corporation (JOGMEC), Chiba 261-0025, Japan ‡

S Supporting Information *

ABSTRACT: By using molecular dynamics (MD) simulations and potential of mean force (PMF) calculations, we studied the stability of model acidic oil molecules (C9H19COOH or C9H19COO−) adsorbed on muscovite surfaces in aqueous solution. The muscovite surfaces are covered by different cations (Na+, K+, Mg2+, and Ca2+). It was found that Ca2+covered muscovite surface significantly enhances the adsorption of C9H19COO− with adsorption Gibbs energy 1 order of magnitude higher than that of Na+-covered surface and 3 times higher than that of K+-covered surface. Furthermore, we found clear evidence that Ca2+ and K+ cause cation bridging, whereas Mg2+ and Na+ cause water bridging. The adsorption Gibbs energy is much higher for cation bridging than that of water bridging. The ion specific effect is not observed when the carboxyl group is protonated (i.e., C9H19COOH). These results well explain the results of previous wettability and core flooding experiments and support their key findings that adsorption of Ca2+ cations induces a macroscopic wetting transition either on a flat mineral surface (in wettability experiments) or in a porous media (in core flooding experiments). The insight obtained in this study leads us to optimal design of low-salinity water flooding for enhanced oil recovery.



observed;10,11 (5) permeability decreases, which is likely a consequence of selective plugging of pore network throats by the fine particles.8 To unify the mechanisms of EOR by LSWF, considerable effort is required to investigate crude oil/brine/ mineral interactions.15,16 Crude oil/brine/mineral interactions are sensitive to ionic composition and concentration in the aqueous phase. These factors have been studied by contact angle measurements,17−21 visual observation of crude oil deposits,22,23 and measurement of the interaction forces between minerals and oil droplets by surface force apparatus, 24,25 atomic force microscopy (AFM),26−29 and adhesion testing.30 Drummond and Israelachvili25 used a surface force apparatus to show that the wettability and measured force between a crude oil-coated mica surface and clean mica surface correlate well with each other at various salt concentrations and pH except at high salt concentration and high pH, where the natural surfactants in crude oil influenced this correlation.

INTRODUCTION The structural and dynamic properties of aqueous solution− mineral interfaces are fundamental to geology, biology, chemistry, and engineering.1 These properties depend on a sensitive balance of noncovalent interactions between water, specific ions, surface groups, and other dissolved substances.1−5 Water flooding is currently one of the most common methods to improve oil recovery from reservoirs. Either formation water (produced) or seawater is injected into the reservoir, yielding an average recovery factor of around 35%.6 However, it has been reported that low-salinity water flooding (LSWF) can result in additional oil recovery of 5% to 38% over conventional water flooding.6−14 Although use of LSWF is desirable because of its low cost and abundant water sources, the origin of enhanced oil recovery (EOR) by LSWF is still controversial. Seventeen mechanisms of LSWF have been proposed.6 Many core-flooding experiments revealed the following five conditions for EOR from sandstone reservoirs by LSWF:6,7 (1) sandstones must contain clay;8−11 (2) the oil must contain polar components (i.e., organic acids and/or bases);10,12−14 (3) the formation water must contain divalent cations (Ca2+ and Mg2+);13 (4) the production or migration of fine particles is © XXXX American Chemical Society

Received: December 1, 2016 Revised: April 8, 2017 Published: April 10, 2017 A

DOI: 10.1021/acs.jpcc.6b12116 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Figure 1. Overview of the simulation system. (a) Snapshot of the whole system. Water molecules have been omitted for clarity. H, C, O, Na, Si, Cl, K, and Al are represented by light gray, black, red, blue, yellow, purple, pink, and light blue spheres, respectively. (b) Acidic oil molecule at low pH (C9H19COOH). (c) Acidic oil molecule at high pH (C9H19COO−).

adsorbed on mineral surfaces with different cation exchange states. Molecular simulations are expected to reveal the molecular scale structures and stability of acidic oil molecules on mineral surfaces.40−42 First-principles studies assessed the stability of acidic oil molecules on calcite surfaces with the aim of applying LSWF to carbonate reservoirs.40,41 Underwood et al.42 performed molecular dynamics (MD) simulations that indicated Ca2+ makes montmorillonite surfaces oil-wet. They also obtained a cation bridging structure from the simulations.42 Here, we examine the cation bridging structure on the molecular scale and energetic stability of the adsorption structures of acidic oil/brine/mineral systems by MD simulations. The quantitative information obtained helps us to understand the affinity of acidic oil molecules for the clay mineral muscovite in various cation exchange states, which depends on ionic composition and concentration of the formation water and the components in crude oils.

Wettability alteration in crude oil/brine/mineral system under various conditions has been examined by AFM using functionalized tips (R−CH 3 or R−COOH) and rock models.27−29 Each total interaction measured by AFM is considered to be a combination of van der Waals interactions and the effects of the electrical double layer and cation bridging. For muscovite, which is suitable as a model surface of a clay mineral, the adhesion force for a −CH3 tip changes from 226 pN in artificial seawater with high salinity (36 500 mg/L total dissolved solids) to 166 pN at low salinity (1400 mg/L total dissolved solids).28 The change in the adhesion force is governed by the electrical double layer force.29 The interaction between a −COOH tip and muscovite is a little different; the adhesion force is higher and dependence on salinity is weaker.28 These differences can be explained by cation bridging (COO−− Ca2+−clay).13,28,31 Indeed, contact angle measurements of muscovite, decane with stearic acid, and aqueous solution (NaCl aq or CaCl2 aq) demonstrated that Ca2+ markedly enhances the adsorption of stearic acid,21 making the substrate oil-wet. Yildiz and Morrow31 suggested that Ca2+ makes a core sample oil-wet in the case of Moutray crude oil (acid number = 0.6 mg of KOH/g of oil).16 Clay minerals are layer-type minerals, like muscovite, which are constructed of tetrahedral and octahedral sheet structures.1 Depending on the number of tetrahedral and octahedral sheets that combine to form a layer, clay minerals can be classified into three types: 1:1, 2:1, and 2:1 with a hydroxide interlayer. Most 2:1 clay mineral layers have net negative charges, and the space between crystal layers is occupied by cations or hydrated cations.1 For example, in muscovite, substitution of Si4+ with Al3+ at tetrahedral sites yields negative charge that is compensated by K+. Compensating cations can be exchanged in interlayer spaces and on surfaces of clay minerals.32−39 Such cation exchange alters the properties of clay minerals (e.g., swelling35 and hydration structure at surfaces36,37), providing cation-specific properties. Furthermore, it has been suggested that cation exchange causes ion-specific cation bridging, which affects the wettability of clay mineral surfaces.13 Therefore, it is important to determine the stability of acidic oil molecules



COMPUTATIONAL METHODS The stability of acidic oil molecules (decanoic acid, C9H19COOH, or sodium decanoate, C9H19COO−) adsorbed on muscovite surfaces in aqueous solutions was studied by MD and potential of mean force (PMF) calculations. The interface system was constructed with an aqueous phase and muscovite slab (Figure 1a). The aqueous phase consisted of 1400 water molecules, 25 NaCl ion pairs, and an oil molecule. The corresponding NaCl concentration was 0.99 mol/kg. The reason that we used NaCl solutions instead of pure water was to avoid complications when comparing with experiments. The oil molecule was C9H19COOH/C9H19COO− (Figure 1b,c), which has a pKa of 4.90.43 We used these two molecular models to examine the effect of pH (i.e., −COOH corresponds to low pH and −COO− corresponds to high pH) on surface wettability. The crystal structure of muscovite was taken from reported X-ray diffraction measurements.44 Hydrogen atoms were inserted to saturate the Al−OH groups in octahedral layers. While keeping a constant O−H distance, the O−H orientation is allowed to be random. Then, positions of hydrogen atoms were optimized by minimizing energy with the B

DOI: 10.1021/acs.jpcc.6b12116 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Figure 2. Initial structures of muscovite cleavage surfaces with various cation exchange states. (a) Na+, (b) K+, (c) Mg2+, and (d) Ca2+ are represented by blue, pink, orange, and green spheres, respectively. The color scheme for the other atom species is the same as Figure 1a.

force field described below. A 6 × 3 × 1 supercell with a size of 3.12 × 2.71 × 2.00 nm was used for muscovite. The cleavage basal plane (i.e., (001) surface) was exposed to the aqueous phase with a size of 3.12 × 2.71 × 5.16 nm. The various cations (Na+, K+, Mg2+, and Ca2+) were regularly distributed on the cleavage surface so that the net negative surface charge was neutralized by adsorbed cations (Figure 2) to mimic different cation exchange states. Cations do not adsorb randomly but form preferentially ordered structures such as rows or geometrical domains.38 The preferential lateral positions of cations has been discussed by Ricci et al.38 and in our previous paper.45 No restraint was imposed on the cations, but they remained at the sites during the simulations without being replaced by Na+. This is presumably because simulation time in our study is at tens of nanoseconds, while the actual exchange takes place in tens of minutes.46 Besides, the reaction is limited by the availability of cations (in solutions) to replace the cations on the surface. According to the work by Bowers et al.,47 the exchange rate should be controlled by total amount of K+ on the surface and Na+ in solution. In 0.99 mol/kg of NaCl solution of our simulation system, we have only 25 Na+ in solution phase. On the muscovite surface, the number of cation initially at the surface are 36 and 18 for monovalent cation and for divalent cation, respectively. In this particular system, assuming the exchange happens in simulations, we will obtain a 0.99 mol/kg KCl solution on the Na-muscovite surface. By noting the K+ has lower hydration free energy,48 hence, higher affinity to the muscovite surface, the final equilibrium system may still be K+-rich surface. In the case of divalent covered surface, a mixed adsorption surface might be anticipated.

The GROMACS package49 (version 4.5.6) was used for the MD simulations. The SPC/E model50 and CLAYFF force field51 were used for water and muscovite, respectively. For monovalent ions and divalent ions, the Joung and Cheatham model52 and Åqvist model53 were used, respectively. The ion models well reproduce hydration free energy of the ions52,53 and solubility of salt of the ions.52,54 The CGenFF force field was used for the oil molecules.55−58 The Lorentz−Berthelot combination rule was employed for the Lennard-Jones (LJ) parameters for different atom species. The cutoff distances were 1.0 nm for both LJ and electrostatic potentials. The particle mesh Ewald summation method59 was used for long-range electrostatic interaction. For all of the simulations, the temperature was controlled by the Nosé−Hoover thermostat,60,61 where the time constant was 2.0 ps. A Parrinello− Rahman barostat62 was used to control the pressure with the time constant of 5.0 ps, where anisotropic scaling without shear deformation was implemented. Snapshots of the simulation systems were prepared by Visual Molecular Dynamics software.63 We evaluated the adsorption Gibbs energy (ΔG) by PMF calculations using the umbrella sampling method.64−67 The distance (D) between the muscovite surface and center of mass (COM) of the oil molecule, C9H19COOH or C9H19COO−, was defined as a reaction coordinate for PMF. The simulation involved the following steps. First, the oil molecule was placed at the center of the aqueous phase. Then, the oil molecule was pulled toward the muscovite surface. D was controlled by a harmonic potential kh of 4000 kJ/mol nm2 on the oil molecule. The center of the harmonic potential was moved toward the surface at a constant rate (0.001 nm/ps). Second, we restrained C

DOI: 10.1021/acs.jpcc.6b12116 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Figure 3. Potential of mean force (PMF) and adsorption Gibbs energy (ΔG). PMF as a function of the distance (D) between the muscovite surface and the center of mass (COM) of the oil molecule for (a) C9H19COOH and (b) C9H19COO−. (c) Comparison of ΔG of the different protonated states and cation exchange states. × indicates that ΔG is not negative. The inset figure in (c) is taken from ref 21, where filled symbols represent contact angle results with CaCl2 aq (cyan, blue, red, and black symbols represents concentration of 1 mM, 10 mM, 100 mM, and 1 M, respectively), and open red symbols represent contact angle results with NaCl aq (100 mM). The arrow c in the inset represents the increase in concentration of electrolytes.

the oil molecule at D = 0.44 nm, which was the minimum distance obtained by the pulling simulation, for 10 ns to obtain an equilibrium configuration for the oil molecule at the surface of muscovite. In the next step, the oil molecule was pulled toward the aqueous bulk phase to generate snapshots with different D, which are initial configurations for the following umbrella sampling simulation. In this step, the pull rate was also 0.001 nm/ps. Finally, we implemented the umbrella sampling simulations using the initial configurations. For the umbrella sampling simulations, we used an interval of D = 0.025 nm from 0.44 to 1.79 nm (55 windows) with kh = 4000 kJ/mol nm2. It should be noted that the parameter of the umbrella sampling is not universal.66 For example, the larger kh and smaller window intervals (i.e., increase in the number of windows) are necessary when the distance between the muscovite surface and the headgroup is chosen for a reaction coordinate. It is because more structured PMF, W(r), is expected by the relation W(r) = −kBT ln g(r), where g(r) is number density normalized by bulk concentration of component at a reaction coordinate, r, and kB and T are Boltzmann constant and temperature, respectively.66 For the purpose to compare affinities of an oil molecule on a surface, the choice of COM is beneficial to reduce the number of the windows. Although the best choice of a window interval, number, and simulation time for each window cannot be known a priori, overlapping of umbrella histograms of a reaction coordinate from each window is more critical for the umbrella sampling than the simulation time.66,67 We have examined the overlapping of histograms from each window in our calculation setting, where at least 3 or 4 histograms cover certain value of the reaction coordinate (D) of interest (see Figure S1). Besides, the bootstrapping analysis was used to quantify calculation errors.68,69 The errors in the original data were estimated by standard deviation of bootstrap samples (Nb

= 200). The harmonic potential mentioned above was imposed normal to the surface (i.e., the oil molecule was free to move in a lateral direction). The oil molecule was free to rotate in all simulation procedures. When we started a calculation at each window, we assigned the initial velocities of atoms with random numbers. Therefore, the calculations are independent for each window. The data were taken from a 6.1 ns calculation for a single window. The total computational time for one cationcovered surface for one state of the oil is around 335 ns. Therefore, it becomes 2680 ns for all the systems considered in this study. To generate PMF, the weighted histogram analysis method was implemented.69−71 Temperature and pressure were ambient conditions (298 K and 1 bar). The simulation details on the convergence of PMFs are described in the Supporting Information (Figures S2−S4).



RESULTS AND DISCUSSION Results of PMF calculations for the different cation exchange states of the muscovite surface are presented in Figure 3. The adsorption Gibbs energy (ΔG) was defined as the difference between the minimum PMF value and that at a long separation distance (D = 1.74 nm). Figure 3 clearly reveals that both ΔG and the shape of the PMF curve depend on the cation exchange state and protonation state of the oil molecule. At low pH, where the oil molecule is protonated, the PMF curves are similar regardless of the cation exchange state (Figure 3a). Moreover, ΔG is not negative. This means the oil molecule has no affinity for the muscovite surface independent of its cation exchange state. The overall repulsive force can be explained by the overlap of the muscovite surface and hydration layers of the −COOH groups of the acid.45 Interestingly, at high pH, where C9H19COOH is deprotonated (C9H19COO−), the PMF curves strongly depend on the cation exchange state (Figure 3b). By comparing ΔG of the D

DOI: 10.1021/acs.jpcc.6b12116 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Figure 4. Adsorption structures of the deprotonated oil molecule with different cation exchange states: (a) Na-muscovite, (b) K-muscovite, (c) Mgmuscovite, and (d) Ca-muscovite. The atoms at the origin of the RDFs represent the carbon atom of the carboxylate group. RDFs between the carbon atom of the carboxylate group and the oxygen atom of water (blue) and cations (red) are depicted. The middle and right column represent representative snapshots and schematic images, respectively.

different cation exchange states (Figure 3c), we find that adsorption of Ca2+ at a muscovite surface markedly strengthens the adsorption of the deprotonated oil molecule (ΔG = −22.0 kJ/mol). Using the Boltzmann factor to estimate the effect of this difference of ΔG indicates that oil molecules at the surface with Ca2+ are 500 times more concentrated than that at the surface with K+. Our calculated ΔG values agree with experimental results. Mugele et al.21 measured the contact angles of NaCl and CaCl2 solutions in decane with stearic acid on muscovite. For the

same cation concentration (100 mM), the contact angle of CaCl2 solution was ∼5° at pH = 3 and ∼30° at pH = 10, whereas that of NaCl solution was ∼0° at pH = 3 and ∼5° at pH = 10. These values suggest that the substrate is water-wet at low pH (i.e., no adsorption of stearic acid) and becomes more oil-wet at high pH in the presence of Ca2+ (i.e., enhanced adsorption of stearic acid). It seems that the effect of cation type on contact angle is larger at high pH. When more concentrated CaCl2 solution (500 mM) was used, the contact angle was ∼5° at pH = 3 and ∼60° at pH = 10. The height E

DOI: 10.1021/acs.jpcc.6b12116 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

inner-sphere surface complexes (see Figure S7): K+ is on top of ditrigonal cavities, whereas Ca2+, Mg2+, and Na+ are above substituted Al sites. These observations qualitatively agree with experiments to investigate adsorption of cations on muscovite surfaces.33,34,38,73 The adsorption structure is well correlated with ΔG of the systems (Figures 3 and 4). |ΔG| is higher when the oil molecule and mineral surface are directly bridged by cations (Kmuscovite and Ca-muscovite) rather than water. Comparing the systems with the same bridging structure, |ΔG| is higher when the ionic valence is higher. It was noted that small cations (which have a high degree of hydration) bind more tightly to COO− than large cations (which possess a low degree of hydration) do.2,3 However, we show here that the small cations (Na+ and Mg2+) at the mineral surface interact primarily with water, which bridges the adsorbed cation and COO−. Consequently, the local adsorption structure differs considerably from that found in bulk aqueous solutions.4,5,74,75 It was reported that Mg2+ interacts with water so strongly that contact ion pairs could not be observed experimentally even in bulk aqueous solutions.75 For K+ and Ca2+, the adsorbed cations are bound directly to COO−, which is similar to the case in bulk aqueous solutions. The order of interaction of the cations to the putative acetate binding sites in aqueous solutions, measured by lower critical solution temperature of polypeptides, is reported as Ca2+ > Mg2+ > Na+ > K+, with corresponding estimated dissociation constants Kd of 0.0036, 0.0076, 0.222, and 0.339, respectively.3 According to this order, Mg2+ and Na+ are expected to interact strongly with COO−. The different order of the cation−COO− interactions (Ca2+ > K+ > Mg2+ > Na+) in the present study should result from the different bound structures on the mineral surface (i.e., water bridging) and in aqueous solutions (i.e., contact ion pairs similar to cation bridging). By comparison with the free energy to form ion pairs in bulk aqueous solutions (−13.9 and −2.7 kJ/mol for Ca2+ and K+ at 298 K, respectively), which are estimated from Kd,3 the interaction of Ca2+ and K+ with COO− is enhanced in the presence of the muscovite surface. Namely, ΔG for the acidic oil/brine/mineral systems with Ca2+ and K+ are of the same order and of larger magnitude than those in bulk aqueous solutions. The changes in the magnitude of the cation−COO− interactions might be caused by the presence of the surface, which changes the hydration structure around the cations. The prediction of the interaction between the divalent cations (Ca2+ and Mg2+) and COO− from Kd of the bulk aqueous solutions indicates that both Ca2+ and Mg2+ yield wettability alteration, as suggested in requirement 3 for LSWF. However, this study demonstrated that only Ca2+ is able to induce wettability alteration. The above discussion is also applicable to other clay minerals such as illite and smectite, which possess similar structures to that of muscovite. Different from swelling clays, the interlayer cations cannot be exchanged for muscovite. Underwood et al. demonstrated that Ca2+ in interlayers of montmorillonite (a specific form of smectite) enhances adsorption of decanoate by the cation bridging mechanism.42 Muscovite shares 2:1 structure with montmorillonite, but the origin of negative charge is different: the negative charge for muscovite results from isomorphous substitutions in tetrahedral plane whereas that for montmorillonite results from the substitution in octahedral plane. The cation bridging by Ca2+ is observed for both clay minerals no matter how they are negatively charged. Therefore, cation exchange at the clay surfaces may in turn

profiles measured by AFM suggested the formation of a monolayer of stearic acid near three-phase (oil, water, mineral) contact line in this case. The PMF curves and positive ΔG calculated for C9H19COOH (Figure 3a,c) are consistent with the water-wet substrate and smaller effect of cation type at low pH.21 Furthermore, the high stability of the oil molecules in the presence of Ca2+ at a muscovite surface (Figure 3c) agrees with the more oil-wet substrate in the presence of Ca2+ at high pH.21 The results of experiments21 and our calculation suggest that conditions 1, 2, and 3 listed above are strongly related to ionspecific wettability alteration. In addition, our quantitative analysis emphasizes the importance of Ca2+ rather than Mg2+ among divalent cations with respect to condition 3. The presence of clay, organic acids, and Ca2+ seems to make reservoirs oil-wet by forming Ca2+ cation bridging structures. The injection of low-salinity water would then destabilize the structure, which, in turn, changes the wettability of clay to weakly water-wet. EOR can be realized by wettability alteration.72 Indeed, a spontaneous imbibition experiment demonstrated that Ca2+ in formation water made a core oilwet with crude oil with a high acid number.31 Lager et al.13 demonstrated that oil recovery is not increased by LSWF without Ca2+ in the formation water. We believe that these core-scale observations can be explained by the mechanism mentioned above. We evaluated the structure of the adsorbed deprotonated oil molecule by analysis of the MD trajectory for the window at or near the PMF well. Figure 4 shows the radial distribution function (RDF) between the carbon atom of the carboxylate group (COO−) and the oxygen atom of water or cations. Corresponding snapshots and schematic images are also presented. For Na-muscovite (Figure 4a), the peak from the oxygen atom of water (r = 0.35 nm) is observed first, followed by the peak from Na+ (r = 0.50 nm). This means COO− does not directly interact with Na+ (i.e., there are bridging water molecules, as depicted in the right panel in Figure 4a). In the case of K+, COO− directly interacts with K+ (cation bridging) when the oil molecule is adsorbed on the muscovite surface. The left panel in Figure 4b shows that the peak from K+ comes first in the RDF curve, unlike in the case of Na-muscovite. For the divalent cations, we observed water bridging for Mgmuscovite and cation bridging for Ca-muscovite. In the case of Mg-muscovite, the peak from the oxygen atom in water is seen first, whereas the peak from the oxygen atom in water follows that from Ca2+ in the case of Ca-muscovite. The difference in the bridging structure (i.e., cation bridging for K+ and Ca2+, water bridging for Na+ and Mg2+) has influenced the hydration shell of COO−group as shown in Figures S5 and S6. It is found that the major change from Na-muscovite to K-muscovite is that about 1−2 water molecules are replaced by K+ as anticipated (since the latter has cation bridging) (Figure S5). The major change from Mg-muscovite to Ca-muscovite is the expansion of the hydration shell: the first minimum of RDF is 0.405 nm in the case of Mg-muscovite, whereas it is 0.435 nm in the case of Ca-muscovite (Figure S6). In addition, a shoulder (between 0.405 and 0.435 nm) can be clearly seen in the case of Ca-muscovite. Models (i.e., cation bridging for K+ and Ca2+, water bridging for Na+ and Mg2+) provided in Figure 4 agree with the adsorption mechanisms proposed by Lager et al.13 Underwood et al. have shown that cations have inner-sphere surface and outer-sphere surface complexes on montmorillonite, and formation of the complexes is cation specific.39 In this study, it is noted that cations involved in bridging structure are F

DOI: 10.1021/acs.jpcc.6b12116 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Japan Science and Technology Agency (JST)/Japan International Cooperation Agency (JICA)−Science and Technology Research Partnership for Sustainable Development (SATREPS), and Japan Petroleum Exploration Co., Ltd. (JAPEX).

result in the formation (or deconstruction) of a cation bridging structure. Similar results were obtained for minerals in outcrop sandstone (Berea sandstone) with high cation-exchange capacity.13 This further indicates that mineral surface wettability is modified to less water-wet by the presence of cations in the order of Na+ < Mg2+ < K+ < Ca2+. Moreover, it was found that more oil was produced even in highly saline NaCl brine if the formation water contained divalent cations.13 All of these findings are consistent with our simulation results; that is, the adsorption Gibbs energy of acidic oil molecules to the cationcovered clay mineral surface has the order Na+ < Mg2+ < K+ < Ca2+. For clays with a relatively low cation exchange capacity, such as kaolinite, a different effect of cation type on ΔG may be anticipated.76





CONCLUSION In summary, molecular dynamics simulations revealed the ionspecific affinity of oil molecules for a clay mineral. Muscovite surfaces were covered by different cations (Na+, K+, Mg2+, Ca2+) that mimic different cation exchange states. Structural analysis of the adsorbed oil molecules and evaluation of ΔG revealed that (i) the presence of Ca2+ on the muscovite surface markedly enhances the adsorption of C9H19COO−, (ii) an ionspecific effect is hardly observed when the carboxyl group of the acid is protonated (i.e., C9H19COOH), and (iii) Ca2+ and K+ form cation bridging, whereas Mg2+ and Na+ cause water bridging. These results readily explain the results of previous wettability21 and core flooding experiments.13,31 They provide interpretations that wettability alteration of mineral surfaces by organic acids depends on the type of ions in solutions. Among the divalent cations (Ca2+ and Mg2+) in formation water, which are required for LSWF, Ca2+ plays a larger role in wettability alteration than Mg2+.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcc.6b12116. Umbrella histograms, convergence of PMF curves with a 0.5, 1.0, and 3 ns time window and corresponding convergence of ΔG; the hydration number of COO− group and cations (involved in bridging structure) on four different muscovite surfaces (PDF)



REFERENCES

(1) Sposito, G. The Chemistry of Soils; Oxford University Press: 1989. (2) Salis, A.; Ninham, B. W. Models and Mechanisms of Hofmeister Effects in Electrolyte Solutions, and Colloid and Protein Systems Revisited. Chem. Soc. Rev. 2014, 43, 7358−7377. (3) Kherb, J.; Flores, S. C.; Cremer, P. S. Role of Carboxylate Side Chains in the Cation Hofmeister Series. J. Phys. Chem. B 2012, 116, 7389−7397. (4) Aziz, E. F.; Ottosson, N.; Eisebitt, S.; Eberhardt, W.; JagodaCwiklik, B.; Vacha, R.; Jungwirth, P.; Winter, B. Cation-Specific Interactions with Carboxylate in Amino Acid and Acetate Aqueous Solutions: X-Ray Absorption and ab initio Calculations. J. Phys. Chem. B 2008, 112, 12567−12570. (5) Hess, B.; van der Vegt, N. F. A. Cation Specific Binding with Protein Surface Charges. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 13296−13300. (6) Sheng, J. J. Critical Review of Low-Salinity Waterflooding. J. Pet. Sci. Eng. 2014, 120, 216−224. (7) RezaeiDoust, A.; Puntervold, T.; Strand, S.; Austad, T. Smart Water as Wettability Modifier in Carbonate and Sandstone: A Discussion of Similarities/Differences in the Chemical Mechanisms. Energy Fuels 2009, 23, 4479−4485. (8) Morrow, N.; Buckley, J. Improved Oil Recovery by Low-Salinity Waterflooding. JPT, J. Pet. Technol. 2011, 63, 106−112. (9) Bernard, G. G. Effect of Floodwater Salinity on Recovery of Oil from Cores Containing Clays. The 38th Annual California Regional Meeting of the Society of Petroleum Engineers of AIME, Los Angeles, CA, 1967, SPE1725. (10) Tang, G.-Q.; Morrow, N. R. Influence of Brine Composition and Fines Migration on Crude Oil/Brine/Rock Interactions and Oil Recovery. J. Pet. Sci. Eng. 1999, 24, 99−111. (11) Berg, S.; Cense, A. W.; Jansen, E.; Bakker, K. Direct Experimental Evidence of Wettability Modification by Low Salinity. Petrophysics 2010, 51, 314−322. (12) Sharma, M. M.; Filoco, P. R. Effect of Brine Salinity and CrudeOil Properties on Oil Recovery and Residual Saturations. SPE Journal 2000, 5, 293−300. (13) Lager, A.; Webb, K. J.; Black, C. J. J.; Singleton, M.; Sorbie, K. S. Low Salinity Oil Recovery - An Experimental Investigation. Petrophysics 2008, 49, 28−35. (14) Suijkerbuijk, B. M. J. M.; Hofman, J. P.; Ligthelm, D. J.; Romanuka, J.; Brussee, N.; vander Linde, H. A.; Marcelis, A. H. M. Fundamental Investigations into Wettability and Low Salinity Flooding by Parameter Isolation. The 18th SPE Improved Oil Recovery Symposium, Tulsa, OK, 2012, SPE154204. (15) Wu, L. M.; Zhou, C. H.; Keeling, J.; Tong, D. S.; Yu, W. H. Towards an Understanding of the Role of Clay Minerals in Crude Oil Formation, Migration, and Accumulation. Earth-Sci. Rev. 2012, 115, 373−386. (16) Buckley, J. S.; Liu, Y. Some Mechanisms of Crude Oil/Brine/ Solid Interactions. J. Pet. Sci. Eng. 1998, 20, 155−160. (17) Lashkarbolooki, M.; Ayatollahi, S.; Riazi, M. The Impacts of Aqueous Ions on Interfacial Tension and Wettability of an Asphaltenic−Acidic Crude Oil Reservoir during Smart Water Injection. J. Chem. Eng. Data 2014, 59, 3624−3634. (18) Standal, S.; Haavik, J.; Blokhus, A. M.; Skauge, A. Effect of Polar Organic Components on Wettability as Studied by Adsorption and Contact Angles. J. Pet. Sci. Eng. 1999, 24, 131−144. (19) Bera, A.; Kissmathulla, S.; Ojha, K.; Kumar, T.; Mandal, A. Mechanistic Study of Wettability Alteration of Quartz Surface Induced by Nonionic Surfactants and Interaction between Crude Oil and Quartz in the Presence of Sodium Chloride Salt. Energy Fuels 2012, 26, 3634−3643.

AUTHOR INFORMATION

Corresponding Authors

*(Y.L.) Tel +81-3-7136-4271; e-mail [email protected]. *(S.M.) Tel +81-75-383-3204; e-mail murata.sumihiko.6v@ kyoto-u.ac.jp. ORCID

Yunfeng Liang: 0000-0002-8832-1778 Naoya Nishi: 0000-0002-5654-5603 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This research was supported by the Japan Society for the Promotion of Science (JSPS) through a Grant-in-Aid for JSPS Fellows (no. 16J00156), Grant-in-Aid for Scientific Research A (no. 24246148), and Grant-in-Aid for Scientific Research C (no. 16K06925). We further acknowledge funding from the G

DOI: 10.1021/acs.jpcc.6b12116 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C

Low Salinity Recovery Mechanism. J. Phys. Chem. C 2014, 118, 19180−19187. (42) Underwood, T.; Erastova, V.; Cubillas, P.; Greenwell, H. C. Molecular Dynamic Simulation of Montmorillonite−Organic Interactions under Varying Salinity: An Insight into Enhanced Oil Recovery. J. Phys. Chem. C 2015, 119, 7282−7294. (43) Barratt, M. D. Quantitative Structure−Activity Relationships (QSARs) for Skin Corrosivity of Organic Acids, Bases and Phenols: Principal Components and Neural Network Analysis of Extended Datasets. Toxicol. In Vitro 1996, 10, 85−94. (44) Richardson, S. M. Crystal Structure of a Pink Muscovite from Archer’s Post, Kenya: Implications for Reverse Pleochroism in Dioctahedral Micas. Am. Mineral. 1982, 67, 69−75. (45) Kobayashi, K.; Liang, Y.; Amano, K.; Murata, S.; Matsuoka, T.; Takahashi, S.; Nishi, N.; Sakka, T. Molecular Dynamics Simulation of Atomic Force Microscopy at the Water−Muscovite Interface: Hydration Layer Structure and Force Analysis. Langmuir 2016, 32, 3608−3616. (46) Raviv, U.; Laurat, P.; Klein, J. Time Dependence of Forces between Mica Surfaces in Water and Its Relation to the Release of Surface Ions. J. Chem. Phys. 2002, 116, 5167−5172. (47) Bowers, G. M.; Bish, D. L.; Kirkpatrick, R. J. Cation Exchange at the Mineral-Water Interface: H3O+/K+ Competition at the Surface of Nano-Muscovite. Langmuir 2008, 24, 10240−10244. (48) Marcus, Y. Thermodynamics of Solvation of Ions. J. Chem. Soc., Faraday Trans. 1991, 87, 2995−2999. (49) Hess, B.; Kutzner, C.; van der Spoel, D.; Lindahl, E. GROMACS 4: Algorithms for Highly Efficient, Load-Balanced, and Scalable Molecular Simulation. J. Chem. Theory Comput. 2008, 4, 435−447. (50) Berendsen, H. J. C.; Grigera, J.; Straatsma, T. The Missing Term in Effective Pair Potentials. J. Phys. Chem. 1987, 91, 6269−6271. (51) Cygan, R.; Liang, J.; Kalinichev, A. Molecular Models of Hydroxide, Oxyhydroxide, and Clay Phases and the Development of a General Force Field. J. Phys. Chem. B 2004, 108, 1255−1266. (52) Joung, I. S.; Cheatham, T. E., III Determination of Alkali and Halide Monovalent Ion Parameters for Use in Explicitly Solvated Biomolecular Simulation. J. Phys. Chem. B 2008, 112, 9020−9041. (53) Åqvist, J. Ion−Water Interaction Potentials Derived from Free Energy Perturbation Simulations. J. Phys. Chem. 1990, 94, 8021−8024. (54) Kobayashi, K.; Liang, Y.; Sakka, T.; Matsuoka, T. Molecular Dynamics Study of Salt-Solution Interface: Solubility and Surface Charge of Salt in Water. J. Chem. Phys. 2014, 140, 144705. (55) Vanommeslaeghe, K.; Hatcher, E.; Acharya, C.; Kundu, S.; Zhong, S.; Shim, J.; Darian, E.; Guvench, O.; Lopes, P.; Vorobyov, I.; MacKerell, A. D., Jr. CHARMM General Force Field: A Force Field for Drug-Like Molecules Compatible with the CHARMM All-Atom Additive Biological Force Fields. J. Comput. Chem. 2010, 31, 671−690. (56) Yu, W.; He, X.; Vanommeslaeghe, K.; MacKerell, A. D., Jr. Extension of the CHARMM General Force Field to SulfonylContaining Compounds and Its Utility in Biomolecular Simulation. J. Comput. Chem. 2012, 33, 2451−2468. (57) Vanommeslaeghe, K.; MacKerell, A. D., Jr. Automation of the CHARMM General Force Field (CGenFF) I: Bond Perception and Atom Typing. J. Chem. Inf. Model. 2012, 52, 3144−3154. (58) Vanommeslaeghe, K.; Raman, E. P.; MacKerell, A. D., Jr. Automation of the CHARMM General Force Field (CGenFF) II: Assignment of Bonded Parameters and Partial Atomic Charges. J. Chem. Inf. Model. 2012, 52, 3155−3168. (59) Essmann, U.; Perera, L.; Berkowitz, M. L.; Darden, T.; Lee, H.; Pedersen, L. G. A Smooth Particle Mesh Ewald Method. J. Chem. Phys. 1995, 103, 8577−8593. (60) Nosé, S. A Molecular Dynamics Method for Simulations in the Canonical Ensemble. Mol. Phys. 1984, 52, 255−258. (61) Hoover, W. G. Canonical Dynamics: Equilibrium Phase-Space Distributions. Phys. Rev. A: At., Mol., Opt. Phys. 1985, 31, 1695−1697. (62) Parrinello, M.; Rahman, A. Strain Fluctuations and Elastic Constants. J. Chem. Phys. 1982, 76, 2662−2666. (63) Humphrey, W.; Dalke, A.; Schulten, K. VMD: Visual Molecular Dynamics. J. Mol. Graphics 1996, 14, 33−38.

(20) Nasralla, R. A.; Bataweel, M. A.; Nasr-El-Din, H. A. Investigation of Wettability Alteration and Oil-Recovery Improvement by LowSalinity Water in Sandstone Rock. J. Can. Petrol. Technol. 2013, 52, 144−154. (21) Mugele, F.; Bera, B.; Cavalli, A.; Siretanu, I.; Maestro, A.; Duits, M.; Cohen-Stuart, M.; van den Ende, D.; Stocker, E.; Collins, I. Ion Adsorption-Induced Wetting Transition in Oil-Water-Mineral Systems. Sci. Rep. 2015, 5, 10519. (22) Lebedeva, E. V.; Fogden, A. Nano-scale Structure of Crude Oil Deposits on Water-Wet Substrates: Dependence on Aqueous Phase and Organic Solvents. Colloids Surf., A 2011, 380, 280−291. (23) Fogden, A. Effect of Water Salinity and pH on the Wettability of a Model Substrate. Energy Fuels 2011, 25, 5113−5125. (24) Drummond, C.; Israelachvili, J. Surface Forces and Wettability. J. Pet. Sci. Eng. 2002, 33, 123−133. (25) Drummond, C.; Israelachvili, J. Fundamental Studies of Crude Oil−Surface Water Interactions and Its Relationship to Reservoir Wettability. J. Pet. Sci. Eng. 2004, 45, 61−81. (26) Basu, S.; Sharma, M. M. Measurement of Critical Disjoining Pressure for Dewetting of Solid Surfaces. J. Colloid Interface Sci. 1996, 181, 443−455. (27) Hassenkam, T.; Pedersen, C. S.; Dalby, K.; Austad, T.; Stipp, S. L. S. Pore Scale Observation of Low Salinity Effects on Outcrop and Oil Reservoir Sandstone. Colloids Surf., A 2011, 390, 179−188. (28) Hassenkam, T.; Mitchell, A. C.; Pedersen, C. S.; Skovbjerg, L. L.; Bovet, N.; Stipp, S. L. S. The Low Salinity Effect Observed on Sandstone Model Surfaces. Colloids Surf., A 2012, 403, 79−86. (29) Hilner, E.; Andersson, M. P.; Hassenkam, T.; Matthiesen, J.; Salino, P. A.; Stipp, S. L. S. The Effect of Ionic Strength on Oil Adhesion in Sandstone - The Search for the Low Salinity Mechanism. Sci. Rep. 2015, 5, 9933. (30) Yang, S.-Y.; Hirasaki, G. J.; Basu, S.; Vaidya, R. Statistical Analysis on Parameters that Affect Wetting for the Crude Oil/Brine/ Mica System. J. Pet. Sci. Eng. 2002, 33, 203−215. (31) Yildiz, H. O.; Morrow, N. R. Effect of Brine Composition on Recovery of Moutray Crude Oil by Waterflooding. J. Pet. Sci. Eng. 1996, 14, 159−168. (32) Karaborni, S.; Smit, B.; Heidug, W.; Urai, J.; van Oort, E. The Swelling of Clays: Molecular Simulations of the Hydration of Montmorillonite. Science 1996, 271, 1102−1104. (33) Loh, S.-H.; Jarvis, S. P. Visualization of Ion Distribution at the Mica−Electrolyte Interface. Langmuir 2010, 26, 9176−9178. (34) Lee, S. S.; Fenter, P.; Nagy, K. L.; Sturchio, N. C. Changes in Adsorption Free Energy and Speciation during Competitive Adsorption between Monovalent Cations at the Muscovite (001)Water Interface. Geochim. Cosmochim. Acta 2013, 123, 416−426. (35) Boek, E. S.; Coveney, P. V.; Skipper, N. T. Monte Carlo Molecular Modeling Studies of Hydrated Li-, Na-, and K-Smectites: Understanding the Role of Potassium as a Clay Swelling Inhibitor. J. Am. Chem. Soc. 1995, 117, 12608−12617. (36) Sakuma, H.; Kawamura, K. Structure and Dynamics of Water on Li+-, Na+-, K+-, Cs+-, H3O+-Exchanged Muscovite Surfaces: A Molecular Dynamics Study. Geochim. Cosmochim. Acta 2011, 75, 63−81. (37) Lee, S. S.; Fenter, P.; Nagy, K. L.; Sturchio, N. C. Monovalent Ion Adsorption at the Muscovite (001)−Solution Interface: Relationships among Ion Coverage and Speciation, Interfacial Water Structure, and Substrate Relaxation. Langmuir 2012, 28, 8637−8650. (38) Ricci, M.; Spijker, P.; Voïtchovsky, K. Water-Induced Correlation between Single Ions Imaged at the Solid−Liquid Interface. Nat. Commun. 2014, 5, 4400. (39) Underwood, T.; Erastova, V.; Greenwell, H. C. Ion Adsorption at Clay-Mineral Surfaces: The Hofmeister Series for Hydrated Smectite Minerals. Clays Clay Miner. 2016, 64, 472−487. (40) Sakuma, H.; Andersson, M. P.; Bechgaard, K.; Stipp, S. L. S. Surface Tension Alteration on Calcite, Induced by Ion Substitution. J. Phys. Chem. C 2014, 118, 3078−3087. (41) Sánchez, V.; Miranda, C. R. Modeling Acid Oil Component Interactions with Carbonate Reservoirs: A First-Principles View on H

DOI: 10.1021/acs.jpcc.6b12116 J. Phys. Chem. C XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry C (64) Torrie, G. M.; Valleau, J. P. Monte Carlo Free Energy Estimates Using non-Boltzmann Sampling: Application to the Sub-Critical Lennard-Jones Fluid. Chem. Phys. Lett. 1974, 28, 578−581. (65) Torrie, G. M.; Valleau, J. P. Nonphysical Sampling Distributions in Monte Carlo Free-Energy Estimation: Umbrella Sampling. J. Comput. Phys. 1977, 23, 187−199. (66) Kästner, J. Umbrella Sampling. Wiley Interdiscip. Rev. Comput. Mol. Sci. 2011, 1, 932−942. (67) Beutler, T. C.; van Gunsteren, W. F. The Computation of a Potential of Mean Force: Coice of the Biasing Potential in the Umbrella Sampling Technique. J. Chem. Phys. 1994, 100, 1492−1497. (68) Coveney, P. V.; Wan, S. On the Calculation of Equilibrium Thermodynamic Properties from Molecular Dynamics. Phys. Chem. Chem. Phys. 2016, 18, 30236−30240. (69) Hub, J. S.; de Groot, B. L.; van der Spoel, D. g_wham−A Free Weighted Histogram Analysis Implementation Including Robust Error and Autocorrelation Estimates. J. Chem. Theory Comput. 2010, 6, 3713−3720. (70) Kumar, S.; Rosenberg, J. M.; Bouzida, D.; Swendsen, R. H.; Kollman, P. A. The Weighted Histogram Analysis Method for FreeEnergy Calculations on Biomolecules. I. The Method. J. Comput. Chem. 1992, 13, 1011−1021. (71) Souaille, M.; Roux, B. Extension to the Weighted Histogram Analysis Method: Combining Umbrella Sampling with Free Energy Calculations. Comput. Phys. Commun. 2001, 135, 40−57. (72) Jadhunandan, P. P.; Morrow, N. R. Effect of Wettability on Waterflood Recovery for Crude-Oil/Brine/Rock Systems. SPE Reservoir Eng. 1995, 10, 40−46. (73) Schlegel, M. L.; Nagy, K. L.; Fenter, P.; Cheng, L.; Sturchio, N. C.; Jacobsen, S. D. Cation Sorption on the Muscovite (001) Surface in Chloride Solutions Using High Resolution X-ray Reflectivity. Geochim. Cosmochim. Acta 2006, 70, 3549−3565. (74) Iskrenova-Tchoukova, E.; Kalinichev, A. G.; Kirkpatrick, R. J. Metal Cation Complexation with Natural Organic Matter in Aqueous Solutions: Molecular Dynamics Simulations and Potential of Mean Force. Langmuir 2010, 26, 15909−15919. (75) Huynh, U. T. D.; Lerbret, A.; Neiers, F.; Chambin, O.; Assifaoui, A. Binding of Divalent Cations to Polygalacturonate: A Mechanism Driven by the Hydration Water. J. Phys. Chem. B 2016, 120, 1021− 1032. (76) Lebedeva, E. V.; Fogden, A. Adhesion of Oil to Kaolinite in Water. Environ. Sci. Technol. 2010, 44, 9470−9475.

I

DOI: 10.1021/acs.jpcc.6b12116 J. Phys. Chem. C XXXX, XXX, XXX−XXX