Studies on Bitumen−Silica Interaction in Aqueous ... - ACS Publications

Langmuir 2003, 19, 3911-3920

3911

Studies on Bitumen-Silica Interaction in Aqueous Solutions by Atomic Force Microscopy Jianjun Liu, Zhenghe Xu,* and Jacob Masliyah Department of Chemical & Materials Engineering, University of Alberta, Edmonton, Alberta, Canada, T6G 2G6 Received November 6, 2002. In Final Form: February 9, 2003 The forces between spin-coated bitumen on a silica wafer and a silica particle in aqueous solutions were measured with an atomic force microscope. The effect of solution pH, salinity, divalent ion addition, and temperature on the interaction/adhesion forces was studied. The results showed that higher solution pH and temperature and lower salinity and calcium concentration resulted in a system of a stronger longrange repulsive force and a weaker adhesive force, which is favorable for bitumen detachment from the silica surface and the subsequent stabilization. The long-range interaction forces between bitumen and silica can be well described with the classical Derjaguin-Landau-Verwey-Overbeek theory, suggesting that the electrostatic forces play a dominant role in a bitumen-silica colloidal system. The best-fitted Stern potentials of bitumen and silica were in excellent agreement with the corresponding zeta potential values measured independently using an electrophoresis technique. An additional repulsive force was observed at a relatively short separation. This additional repulsion can be attributed to a polymer-like steric force. The implication of the interaction forces measured by atomic force microscopy was confirmed by zeta potential distribution measurements. The quantitative description of bitumen/silica interaction provided fundamental insights into the bitumen extraction mechanism in a water-based system for bitumen extraction from oil sands and justified the industrial use of caustics.

Introduction Colloidal forces play a crucial role in a variety of disciplines, ranging from biological systems to industrial processes. Taking bitumen extraction from Athabasca oil sands ore as an example, liberation of bitumen from sand grains and the subsequent stabilization against heterocoagulation of the liberated bitumen with sand grains or clay particles are prerequisites for bitumen recovery using a hot water extraction process (HWEP).1,2 Bitumen liberation is largely determined by the adhesion strength of bitumen with sand grains, while the heterocoagulation is controlled by both long-range interactions and adhesion forces between them. As a result, controlling these interactions between bitumen and sand grains is of great importance to the bitumen extraction process and of vast interest to researchers. In early 1944, Clark3 emphasized the effect of surfactants on the displacement of bitumen from sand grains in the HWEP and listed a number of surface active species released into the water during the extraction process. Later, it was proved that the presence of surfactants is not always beneficial as it may induce hydrophobicity by undesired adsorption and subsequent flotation of unwanted solids.4 A number of researchers5-9 identified a * Corresponding author. E-mail: [email protected]. (1) Hepler, L. G.; Hsi, C. AOSTRA Technical handbook on oil sands, bitumen and heavy oils; AOSTRA Technical Publication Series No. 6; Alberta Oil Sands Technology and Research Authority: Edmonton, AB, Canada, 1989. (2) Hepler, L. G.; Smith, R. G. The Alberta oil sands: industrial procedures for extraction and some recent fundamental research; AOSTRA Technical Publication Series No. 14; Alberta Oil Sands Technology and Research Authority: Edmonton, AB, Canada, 1994. (3) Clark, K. A. Trans. Can. Inst. Min. Metall. 1944, 47, 257-274. (4) Baptista, M. V.; Bowman, C. W. Proceedings of the 19th Canadian Chemical Engineering Conference, Canadian Society for Chemical Engineering, Edmonton, Alberta, 1969. (5) Schramm, L. L.; Smith, R. G.; Stone, J. A. Colloids Surf. 1984, 11, 247-263. (6) Schramm, L. L.; Smith, R. G. Can. J. Chem. Eng. 1987, 65, 799811.

critical surfactant concentration for bitumen recovery, which corresponded well to favorable surface charge conditions.10 Sanford and Seyer7 showed that the role of NaOH addition is to generate natural surfactants. These surfactants are believed to facilitate bitumen flotation. Smith and Schramm11 on the other hand indicated that only a small fraction of NaOH is needed to produce sufficient amounts of natural surfactants essential for bitumen flotation, and that the bulk of added NaOH reacts with the clays and divalent ions such as carbonate and sulfate, thereby creating a favorable bitumen flotation environment. Dai and Chung (1996)12 studied the effect of NaOH addition on the liberation of bitumen from sand grains and the emulsification of bitumen. In their work, a critical amount of NaOH needed to achieve an optimal effect was identified. Early studies with Whiterocks tar sands13 also showed that high aqueous pH facilitates the detachment of bitumen from sand grains as a result of lowering bitumen-water interfacial tension to 2-3 mN m-1. A spontaneous detachment of bitumen from sand grains was observed with the addition of sodium triphosphate at optimal pH. It was also found14 that the attachment of air bubbles to bitumen enhances the detachment of the bitumen from sand grains, in particular in the presence of kerosene. In 1967, Bowman15 assumed that the surface charge of bitumen was determined by the natural surfactants with carboxylic groups while the surface charge of sands was affected by divalent cations (7) Sanford, E. C.; Seyer, F. A. CIM Bull. 1979, 72, 164-169. (8) Sanford, E. C. Can. J. Chem. Eng. 1983, 61, 554-567. (9) Schramm, L. L.; Smith, R. G.; Stone, J. A. AOSTRA J. Res. 1984, 1, 5-13. (10) Schramm, L. L.; Smith, R. G. Colloids Surf. 1985, 14, 67-85. (11) Smith, R. G.; Schramm, L. L. Fuel Process. Technol. 1992, 30, 1-14. (12) Dai, Q.; Chung, K. H. Fuel 1996, 75, 220-226. (13) Drelich, J.; Miller, J. Fuel 1994, 73, 1150-1155. (14) Drelich, J.; Lelinski, D.; Hupka, J.; Miller, J. D. Fuel 1995, 74, 1150-1155. (15) Bowman, C. W. Progress 7th World Petroleum Congress 1967, 3, 583-604.

10.1021/la0268092 CCC: $25.00 © 2003 American Chemical Society Published on Web 03/21/2003

3912

Langmuir, Vol. 19, No. 9, 2003

present in the pulp. The dissociation of carboxylic acid groups and the speciation of divalent ions on sand grains at different pHs determined the heterocoagulation and hence the bitumen-sand separation. By application of the classical Derjaguin-Landau-Verwey-Overbeek (DLVO) theory to the zeta potential data measured separately with bitumen emulsions and sand suspensions, it was proposed16 that the presence of calcium ions in the solution would cause the attachment of bitumen to sand grains, thereby deteriorating the bitumen-sand separation and bitumen attachment to air bubbles. Brown and Neustadter17 suggested that the presence of surfactants in crude oils in the form of protonated nitrogenous species might be responsible for the coagulation of fine silica with the oil over the acidic pH range. Buckley et al.18 carried out some adhesion tests of crude oil on the glass surface at different pHs and ion strengths. Their results were explained by electrical double layer calculations involving the surface ionization model. In the visual pickup test, Dai and Chung (1995)19 found that the interactions between bitumen and silica sand were affected by solution pH, temperature, solvent addition, and the size of particles. Basu et al.20,21 investigated the effect of pH on the resultant contact angle of bitumen on sand grains and its relation to the process of bitumen film retracting and detaching from the sand grains. They suggested that a pH cycle might be desirable for bitumen liberation from oil sand grains. Recently, Zhou et al. (1999)22 investigated using a model system the coagulation of bitumen with fine silica as a function of pH and calcium ion addition. In this study, a synergetic effect of various surfactants and added calcium ions was observed. They offered a qualitative explanation to their observations with the postulation of various surface forces. Although these early studies have advanced our knowledge on bitumen-sand interactions, they are mostly indirect and/or qualitative in nature. The recent development of the surface forces apparatus (SFA) and atomic force microscope (AFM) made quantitative investigation of colloidal particle interactions possible. A number of publications on quantitative investigations related to surface forces between a solid particle and a model oil droplet23-26 appeared in the open literature. Basu and Sharma27 used an AFM along with an oil dip-coating technique to investigate the interactions between oil (octadecane and crude oil) and mineral (glass and mica) surfaces, in an attempt to understand the alteration of wettability in an oil reservoir, while Rabinovich et al.28 established a simple theoretical expression to predict the oil-mediated particle adhesion to a silica substrate. For (16) Takamura, K.; Chow, R. S. J. Can. Pet. Technol. 1983, 22, 2230. (17) Brown, C. E.; Neustadter, E. L. J. Can. Pet. Technol. 1980, JulySept, 100-110. (18) Buckley, J. S.; Takamura, K.; Morrow, N. R. SPE Reservoir Eng. 1989, Aug, 332-340. (19) Dai, Q.; Chung, K. H. Fuel 1995, 74, 1858-1864. (20) Basu, S.; Nandakumar, K.; Masliyah, J. H. J. Colloid Interface Sci. 1996, 182, 82-94. (21) Basu, S.; Nandakumar, K.; Masliyah, J. H. Can. J. Chem. Eng. 1997, 75, 476-479. (22) Zhou, Z. A.; Xu, Z.; Masliyah, J. H.; Czarnecki, J. Colloids Surf., A 1999, 148, 199-211. (23) Mulvaney, P.; Perera, J. M.; Biggs, S.; Grieser, F.; Stevens, G. W. J. Colloid Interface Sci. 1996, 183, 614-616. (24) Hartley, P. G.; Grieser, F.; Mulvaney, P.; Stevens, G. W. Langmuir 1999, 15, 7282-7289. (25) Synder, B. A.; Aston, D. E.; Berg, J. C. Langmuir 1997, 13, 590593. (26) Aston, D. E.; Berg, J. C. J. Pulp Pap. Sci. 1998, 24, 121-125. (27) Basu, S.; Sharma, M. M. J. Colloid Interface Sci. 1996, 181, 443-455. (28) Rabinovich, Y. I.; Esayanur, M. S.; Johanson, K. D.; Adler, J. J.; Moudgil, B. M. J. Adhes. Sci. Technol. 2002, 16, 887-904.

Liu et al.

bitumen extraction, Yoon et al. performed pioneering measurements using both the SFA (1995)29 and the AFM (1999).30 The force between two bitumen surfaces in aqueous solutions was measured directly. In their study, the Langmuir-Blodgett (LB) deposition technique was used to prepare bitumen and/or asphaltene surfaces. A strong, long-range repulsive force was detected. The repulsive force was considered to be of steric origin, resulting from protruding tails of asphaltene molecules. These protruding tails on bitumen surfaces were suggested to stabilize bitumen emulsions. Wu et al.31 determined the colloidal forces between bitumen droplets by analyzing droplet-droplet collision trajectories32 and the hydrodynamic forces required to break up bitumen doublets.33 They found that in addition to an electrostatic double layer repulsive force, heterogeneous protrusion of asphaltene molecules on a bitumen surface further increased the repulsive force. These investigations provided valuable information to advance our understanding of bitumen/ bitumen interactions in the sequential process (flotation stage) of bitumen extraction using the HWEP. To gain a better understanding of bitumen recovery from oil sands in a water-based bitumen extraction system, quantitative studies of bitumen/sand interactions are highly relevant and beneficial. However, to our best knowledge, such a study has not been reported in the open literature. In this paper, we focus on the measurement of interaction and adhesion forces between bitumen and silica in aqueous solutions using an atomic force microscope. The measured force profiles are interpreted in terms of the classical DLVO theory and related zeta potential distribution measurements. Experimental Section Materials. Coker feed bitumen provided by Syncrude Canada Ltd. was used in this study. Silica microspheres (∼8 µm) purchased from Duke Scientific Co. (USA) were used as the model of sand grains for colloidal force measurement. Silica sands (-40 µm), provided by Pacific, MO (USA), were ground to -5 µm size and used in the zeta potential distribution measurements. Silicon wafers of 100 crystal planes were purchased from MEMC Electronic Materials (Italy) and used as the substrate for supporting bitumen by the spin-coating technique. Reagent grade HCl and NaOH (Fisher) were used as pH modifiers. Ultrahigh purity KCl (>99.999%, Aldrich) was used as the supporting electrolyte, while reagent grade CaCl2 (99.9965% Fisher) was used as the source of calcium ions. Reagent grade toluene (Fisher) and in-house-distilled absolute ethanol were used as the dilution and cleaning solvents, respectively. Deionized water with a resistivity of 18.2 MΩ cm, prepared with an Elix 5 followed by a Millipore-UV Plus Ultra water purification system (Millipore Inc., Canada), was used throughout this study. Probe Particle Preparation. The silica sphere was glued with a two-component epoxy (EP2LV, Master Bound, Hackensack, NJ) onto the tip of a short, wide beam AFM cantilever. The spring constant of the cantilever was calculated from its geometry determined from scanning electron micrographs.34 The glued probe particle was allowed to set in a vacuum desiccator for more (29) Yoon, R. H.; Guzonas, D.; Aksoy, B. S.; Czarnecki, J.; Leung, A. In Proceedings of the 1st UBC-McGill Bi-Annual International Symposium; Canadian Institute of Mining, Metallurgy and Petroleum: Montreal, Quebec, 1995; pp 277-289. (30) Yoon, R. H.; Rabinovich Y. I. In Proceedings of the 3rd UBCMcGill Bi-Annual International Symposium; Canadian Institute of Mining, Metallurgy and Petroleum: Montreal, Quebec, 1999. (31) Wu, X.; Laroche, I.; Masliyah, J.; Czarnecki, J.; Dabros, T. Colloids Surf. A 2000, 174, 133-146. (32) Wu, X.; Czarnecki, J.; Hamza, N.; Masliyah, J. Langmuir 1999, 15, 5244-5250. (33) Wu, X.; Dabros, T.; Czarnecki, J. Langmuir 1999, 15, 78068713. (34) Veeramasuneni, S.; Yalamanchili, M. R.; Miller, J. D. J. Colloid Interface Sci. 1996, 184, 594-600.

Bitumen-Silica Interaction in Aqueous Solutions than 24 h. Prior to each set of experiments, the probe particles were thoroughly rinsed with deionized water and ethanol, followed by blow-drying with ultrapure grade nitrogen. The particles were then exposed to an ultraviolet light for more than 5 h to remove any possible organic contaminants and to facilitate surface hydrolysis upon the exposure of the particle to the aqueous environment. The exact size of the silica particles used in each set of experiments was determined with a scanning electron microscope after conducting the force measurement. Substrate Preparation. The bitumen surface was prepared with a P6700 spin-coater (Specialty Coating Systems Inc.). A silicon wafer was first oxidized in a well-ventilated furnace at 1100 °C for 12 h to a deep blue color. The oxidized wafer, referred to as the silica wafer, was cut into 15 × 15 mm square pieces. The wafer was washed with chloroform to remove adsorbed organic contaminants, rinsed with deionized water followed by ethanol, and finally blow-dried with ultrapure grade nitrogen. At this stage, the contact angle of an air bubble on the wafer in deionized water was ca. 10°. Bitumen was dissolved in toluene to a concentration of 2.5 mg of bitumen per mL of toluene. The resultant bitumen solution was centrifuged at 20 000g force for 30 min to remove all remaining contained fine solids. About 0.1 mL of the prepared bitumen/toluene solution was dropped slowly on the silica wafer spinning on the spin-coater which ran at 2000 rpm for 20 s and 5000 rpm for 1 min. Extra care was taken in placing the bitumen on the substrate in such a manner that bitumen loss during the coating was minimized. The spin-coated bitumen had a mirrorlike surface and was further dried for about 1 h in a particle-free horizontal laminar hood under an ambient environment. From the volume of the bitumen used in spincoating and the area of the wafer coated, the thickness of the bitumen layer was estimated to be ca. 100 nm. The prepared bitumen surface was found to be acceptably smooth for colloidal force measurement, as shown by its AFM image in Figure 1. After the prepared bitumen was immersed in deionized water for more than 30 min, the contact of an air bubble with the bitumen was made. After a 5-min contact of the air bubble with the bitumen, the static contact angle values were determined to be 70-75° using the captive bubble method. The measured contact values were within the range of the reported contact angle values for bulk bitumen,35 suggesting a full coverage of the silica wafer by a bitumen film. After each set of the surface force measurements, the contact angle values of the air bubble on bitumen remained the same, confirming that the bitumen layer remained on the substrate during the force measurement. The contact angle values obtained in this study are slightly smaller than the values reported for the bitumen from Utah oil sands, determined by the sessile drop method.36 In this study by Drelich et al., a reducing contact angle value with increasing contact time of the water drop with bitumen was observed. Our results were obtained at a much longer contact time to ensure interfacial equilibrium. Therefore, the observed difference can be attributed to the different methods used and the contact time between water and bitumen before the measurement was made. Surface Force Measurement (AFM Technique). Surface force measurement was conducted using a Nanoscope E AFM (Digital Instruments, Santa Barbara, CA) with a vendor-supplied fluid cell. A detailed description of using an AFM to measure colloidal forces can be found elsewhere.37,38 In our experiment, the solution was injected into the fluid cell slowly with great care to avoid trapping air bubbles. Both surfaces immersed in the solution were allowed to equilibrate for 1 h. A piezo translation stage was used to cause the bitumen surface to approach or retract from the spherical silica particle attached to the tip of an AFM cantilever. The force between the bitumen and particle deflects the cantilever upward or downward, depending on the characteristics of the force between them. The deflection of the cantilever was detected to a subnanometer resolution by an optical system. From the displacement of the piezo translation stage and the (35) Kasongo, T.; Zhou, Z.; Xu; Z.; Masliyah, J. H. Can. J. Chem. Eng. 2000, 78, 674-681. (36) Drelich, J.; Bukka, K.; Miller, J. D.; Hanson, F. V. Energy Fuels 1994, 8, 700-704. (37) Ducker, W. A.; Senden, T. J.; Pashley, R. M. Langmuir 1992, 8, 1831-1836. (38) Rabinovich, Y. I.; Yoon, R. H. Langmuir 1994, 10, 1903-1909.

Langmuir, Vol. 19, No. 9, 2003 3913

Figure 1. A typical AFM image of the bitumen substrate in water: (a) three-dimensional, (b) section analysis, and (c) a protrusion at high spatial resolution. actual deflection of the cantilever, the separation between the bitumen and silica sphere and the corresponding force between the two can be obtained. To calculate the forces from the measured cantilever deflection, the spring constant of the AFM cantilever was determined from the measured geometry of the cantilever from its scanning electron micrograph and the Young modulus of 1.5 × 1011 N/m2 given in the literature for the cantilever material.34,39 Using the spring constant determined as such, a

3914

Langmuir, Vol. 19, No. 9, 2003

maximum relative error in the calculate forces is expected to be less than 15%. The measured force profiles were reproducible with a maximum relative error of less than 10%. To have a better comparison from test to test, both the measured interaction force and adhesive force (pull-off force) were normalized with the radius R of the silica sphere. For each test condition, the measurement was performed at a number of different locations on bitumen to ensure the measured force profiles were representative. All the experiments were conducted at room temperature (22 ( 0.1 °C), unless otherwise specified. It is important to note that the contamination of silica spheres by bitumen during the force measurement was found to be negligible as reproducible force profiles were obtained when solution conditions were brought back to the original after a series of measurements and using different spheres in different sets of measurements. The reason for the minimal contamination may be attributed to the use of a ultrathin layer of bitumen on the substrate and a short contact time of 40-60 ms during the AFM force measurement, although the contact was repetitive. The contact time was estimated from the scan rate and load force. The measured pull-off force was found to be independent of the contact time over the range of variation. Zeta Potential Measurement. Zeta potentials of bitumen emulsions and silica suspensions were measured with a Zetaphoremeter III (SEPHY/CAD). The detailed experimental procedures were given elsewhere.40 The average of four independent measurements was reported. The zeta potential distributions of bitumen droplets and silica particles, individually or as a mixture, were also measured to study the interactions between the two species by following the procedure and concept reported elsewhere.41

Theoretical Analysis To further understand the origin of bitumen/silica interactions, a theoretical analysis of the experimentally measured force profile is required. In this study, the analysis was performed in terms of the classical DLVO theory.42 In this theory, only the van der Waals forces and electrostatic double layer forces are considered. Deviations from the classical DLVO theory have been reported for a number of systems in which other forces were present. The reported non-DLVO forces include repulsive hydration forces operating over a 2-3 nm separation for hydrophilic surfaces,43,44 repulsive steric forces and attractive bridging forces operating over the corresponding polymer conformational length scale for polymer-bearing systems,45 and attractive hydrophobic forces over a much longer range than van der Waals forces for hydrophobic surfaces,46-49 just to name a few. In this study, the measured force profiles were fitted to the classical DLVO theory for an asymmetric system. The van der Waals forces were calculated by Hamaker’s microscopic approach.42 Using 6 × 10-20,50 3.7 × 10-20, and 6.5 × 10-20 J42 as the Hamaker constants for bitumen, water, and silica, respectively, the combined Hamaker constant for the bitumen/water/silica system was calcu(39) Sader, J. E.; Larsoni, I.; Mulvaney, P.; White, L. R. Rev. Sci. Instrum. 1995, 66, 3789-3798. (40) Liu, J.; Zhou, Z.; Xu, Z. Ind. Eng. Chem. Res. 2002, 41, 52-57. (41) Liu, J.; Zhou, Z.; Xu, Z.; Masliyah, J. H. J. Colloid Interface Sci. 2002, 252, 409-418. (42) Isrealachvili, J. N. Intermolecular and surface forces, 2nd ed.; Academic Press: San Diego, 1992. (43) Pashley, R. M.; Quirk, J. P. Colloids Surf. 1984, 8, 1-17. (44) Meagher, L. J. Colloid Interface Sci. 1992, 152, 293-295. (45) Ingersent, K.; Klien, J.; Pincus, P. Macromolecules 1990, 23, 548-560. (46) Israelachvili, J. N.; Pashley, R. M. J. Colloid Interface Sci. 1984, 98, 500-514. (47) Parker, J. L.; Claesson, P. M. Langmuir 1994, 10, 635-639. (48) Yoon, R. H.; Flinn, D. H.; Rabinovich, Y. I. J. Colloid Interface Sci. 1997, 185, 363-370. (49) Arich, B. N. Hydrophobic Interactions; Plenum Press: New York, 1980. (50) Vincent, B. J. Colloid Interface Sci. 1973, 42, 270-285.

Liu et al.

lated to be ca. 3.3 × 10-21 J. The electrostatic double layer force, on the other hand, was calculated numerically by solving the nonlinear Poisson-Boltzmann equation for both surfaces with constant surface potential as boundary conditions. During the fitting exercise, surface potentials for both surfaces as well as the decay length (κ-1) were set as adjustable parameters. A Visual Basic program running on an Excel spreadsheet was developed in our laboratory for a general DLVO theory of asymmetric surfaces in symmetrical/asymmetrical electrolyte solutions. The program was checked with known systems. The sensitivity study with the current system showed that the fitted surface potentials and decay length are not a strong function of the Hamaker constant used in the calculations. Results System Characteristics. It is known that bitumen is a deformable, extremely viscous honeylike material. To avoid complication of force analysis incurred by bitumen deformation, using a thin layer of bitumen on a smooth, hard substrate is desirable. For this reason, a molecularly smooth silica wafer was used as the substrate for supporting bitumen film. To avoid forming a thick layer of bitumen, it was necessary to fine-tune the toluene-tobitumen ratio and spin speed for the bitumen coating. Under the optimal spin-coating conditions, the bitumencoated wafer had a black, mirrorlike appearance. A typical image of spin-coated bitumen on a silica wafer is shown in Figure 1. On average, the roughness of the bitumen surface over a 10 × 10 µm area was found to be less than 2 nm (Figure 1a,b). A few protrusions (Figure 1a) up to 20 nm high and 100 nm in diameter (Figure 1c) were observed over the scanned area. A similar observation was reported by Wu et al.31-33 It is important to note that the sparse protrusions at this scale did not affect the results of our AFM force measurement as the contact area of the two surfaces in our measurement was sufficiently small as to avoid interference by these sparsely found protrusions. It is also known that when bitumen is brought in contact with an aqueous phase, its surface would become relaxed or reorganized to have polar molecules concentrate at the bitumen/water interface or even migrate to the aqueous phase, depending on the polarity of their headgroups and the length of the carbon chain in the molecules. In this manner, the bitumen surface becomes brushlike. During the initial force measurement, it was observed that it took about half an hour for the measured force profile to be reproducible. This time-dependent behavior is attributed to the presence of polar molecules in bitumen, confirming the reorganization of the surface active molecules at the bitumen/water interface. It is conceivable that the surfactant molecules in very viscous bitumen take time to migrate to the interface and to be dissociated or dissolved into the aqueous phase when the bitumen is contacted with aqueous solutions. To ensure reaching equilibrium, the system was allowed to incubate at each condition for at least 1 h before collecting force profile data. Deformation of the Bitumen Surface. When dealing with soft, deformable surfaces such as oil droplets23-29 or air bubbles51 in a colloidal force measurement, the surface deformation is always a concern. The degree of surface deformation depends to a large extent on the geometry and the elasticity of the deformable solid or fluid interface. For a given system, a higher loading force would also cause a greater degree of deformation. Although a number of (51) Ducker, W. A.; Xu, Z.; Israelachvili, J. N. Langmuir 1994, 10, 3279-3289.

Bitumen-Silica Interaction in Aqueous Solutions

Figure 2. Raw data from a typical force measurement between the bitumen and silica and between silica and silica in 1 mM KCl solution at pH 8.2.

approaches including the elastic model,51 contact mechanic theory,52,53 and the augmented Young-Laplace equation54 have been proposed to deal with the deformation of a deformable surface, the application of these approaches remains rather limited to a few special cases. To ensure that the deformation does not contribute to errors in the force analysis, the prepared bitumen film thickness was kept sufficiently thin so that the deformation under the applied force could be considered negligible, as shown by AFM raw force profiles in Figure 2. Presented in this figure are the force profiles obtained using a probe silica particle against both silica and bitumen-coated silica surfaces in an aqueous KCl solution at pH 8.2. The open symbols represent the approaching branch, and the solid symbols, the retracting branch. Even though a small adhesion force was measured between bitumen and silica in contrast to the absence of adhesion between a silica-silica pair, the constant compliance region for the two pairs of surfaces showed the same slope (solid lines). This observation confirms the absence of any noticeable deformation of bitumen film under the applied force from the silica sphere. Also, in the range covered the adhesion force did not show a significant scatter among the measurements under varying applied loads. Effect of pH. The solution pH is a critical operating parameter in bitumen recovery and, in most cases, the controlling parameter for surface charge. In a 1 mM KCl solution, the effect of pH on the interaction forces between the bitumen and silica particle is shown in Figure 3. Over the pH range tested, the measured long-range force profiles are monotonically repulsive. The repulsion increases with increasing pH. Even at pH 3.5, a weak long-range repulsive force was observed. As shown by the solid lines of Figure 3, at separations greater than ca. 2-3 nm, all the measured force profiles can be reasonably well fitted with the classical DLVO theory. The excellent fit shown in this figure by solid lines suggests that the long-range repulsive forces are predominantly from the electrostatic double layer interactions. The fitted Debye decay length (κ-1) in the range of 8.5-10 nm agrees well with that calculated for a 1 mM KCl solution used in the experiment, further confirming the electrostatic nature of the long-range repulsive force. The observed repulsive force at separation distances less than 2-3 nm is inconsistent with the (52) Derjaguin, B. V.; Muller, V. M.; Toporov, Y. P. J. Colloid Interface Sci. 1975, 53, 314-326. (53) Johnson, K. L.; Kendall, K.; Roberts, A. D. Proc. R. Soc. London, Ser. A 1971, 324, 301-313. (54) Aston, D. E.; Berg, J. C. J. Colloid Interface Sci. 2001, 235, 162-169.

Langmuir, Vol. 19, No. 9, 2003 3915

Figure 3. Interaction forces (F/R) between bitumen and silica as a function of separation distance in 1 mM KCl solution at different solution pHs. Solid lines represent the DLVO fitting using A132 ) 3.3 × 10-21 J with the best-fitted decay length and Stern potential being as follows: for pH 3.5 (down triangle), κ-1 ) 9.4 nm, ψB ) -20 mV, and ψS ) -25 mV; for pH 5.7 (circle), κ-1 ) 9.4 nm, ψB ) -58 mV, and ψS ) -48 mV; for pH 8.2 (up triangle), κ-1 ) 9.1 nm, ψB ) -76 mV, and ψS ) -59 mV; and for pH 10.5 (square), κ-1 ) 8.6 nm, ψB ) -83 mV, and ψS ) -64 mV. Insert: the adhesive force (Fad/R) as a function of pH, at the loading force of 8-10 mN/m.

attractive force regime as predicted by the DLVO theory, suggesting the presence of an additional repulsive force. Although the exact reason for this contradiction is not clear, considering a 2-3 nm range, this additional repulsive force would appear to originate from brushlike surfaces or small protrusions at the bitumen/water interface, resulting in a steric type of repulsion. To fully understand the interaction of colloidal particles in a dynamic system, the adhesive forces between them have to be considered. The adhesive (pull-off) forces between bitumen and silica measured by the retracting branch of AFM force profiles are shown in the insert of Figure 3. The adhesion force was measured under a loading force of 8-10 mN/m and was found to be independent of loading force in the range examined. A strong adhesion of 8 mN/m was measured at pH 3.5. A drastic decrease of adhesion force occurred from pH 5.7 to 8.2, and adhesion disappeared eventually at pH 10.5. The close correspondence between the long-range interactions and adhesive forces suggests the important role of interfacial chemistry in controlling colloidal interactions between bitumen and silica particles. It is interesting to note that the measured adhesion forces in the current study were in general much smaller than any anticipated capillary forces. Considering the clear trend of adhesion forces observed with changing solution pH for a given system, the capillary force is unlikely to play a role in our study. The presence of asperities on the bitumen film could be a reason for the absence of the capillary forces. The pH-dependent dissociation of cationic/anionic surfactants at the bitumen/ water interface could be the reason for the observed variations of adhesive forces with pH. Effect of Electrolyte (KCl) Concentration. The concentration of an electrolyte is also an important factor to consider in a surface force analysis. The effect of electrolyte (KCl) concentration on the interaction forces between bitumen and silica at pH 8.2 is shown in Figure 4. All the force profiles were repulsive but the range of the repulsive forces decreased with increasing electrolyte concentration, as anticipated. The force profiles at a separation greater than 2-3 nm can be well fitted by the classic DLVO theory with the fitted decay lengths (κ-1) of 9.1, 3.1, and 1 nm. These values corresponded well with

3916

Langmuir, Vol. 19, No. 9, 2003

Figure 4. Interaction forces (F/R) between bitumen and silica as a function of separation distance in different KCl solutions at pH 8.2. Solid lines represent DLVO fitting using A132 ) 3.3 × 10-21 J with the best-fitted decay length and Stern potential being as follows: for 1 mM KCl (square), κ-1 ) 9.1 nm, ψB ) -76 mV, and ψS ) -59 mV; for 10 mM KCl (circle), κ-1 ) 3.1 nm, ψB ) -56 mV, and ψS ) -44 mV; and for 100 mM KCl (up triangle), κ-1 ) 1.0 nm, ψB ) -35 mV, and ψS ) -30 mV. Insert: the adhesive force (Fad/R) as a function of KCl concentration, at the loading force of 8-10 mN/m.

those evaluated for electrolyte concentrations of 1, 10, and 100 mM, respectively, as used in the measurement, further confirming that long-range repulsive forces did arise from the overlap of the electric double layers. Although the long-range repulsive forces were suppressed by increasing electrolyte concentration, the adhesive forces between bitumen and silica were affected only marginally, as shown in the insert of Figure 4. This observation is anticipated as the electrolyte added should have only a minimal effect on surface ionization/dissociation and hence cause little change of interfacial chemistry. Effect of Calcium Ions. Calcium is one of the divalent ions normally present in a bitumen extraction system of oil sands ores. A substantial amount of calcium ions are introduced into the bitumen extraction system by recycling the process water from tailings treatment in which gypsum was added as a process aid to form consolidated tailings.55 The effect of calcium ions on the long-range colloidal and adhesion forces between bitumen and silica spheres was studied by measuring these forces at different solution pHs. In general, the measured long-range repulsive forces reduced much more significantly with increasing calcium concentration than with increasing monovalent electrolyte (KCl) concentration. At pH 8.2, as shown by the results of Figure 5, for example, adding calcium not only depressed the long-range repulsive forces but also increased adhesive forces substantially. The long-range repulsive forces remained to be well fitted by the DLVO theory with a decrease of decay length from 9.1 nm without calcium addition to 8.5 and 4.6 nm with 0.1 and 1 mM calcium ion addition, respectively. These decay length values are in excellent agreement with the theoretical predications. The fitted Stern potential, on the other hand, decreased from -76 to -30 mV for bitumen and from -59 to -25 mV for silica with 1 mM calcium ion addition. The adhesion force increased accordingly from 0.2 to 5 mN/m. A much more dramatic effect of calcium addition on both the long-range colloidal and the contact adhesive forces was observed at pH 10.5, as shown in Figure 6. The long-range colloidal forces changed progressively from repulsive to attractive with increasing calcium ion addition to 1 mM, while the adhesion force changed from 0 to 8 (55) Sheeran, D. E. In Advances in Oil Sands Tailings Research; Alberta Department of Energy: Edmonton, 1995; Vol. III, p 1-56.

Liu et al.

Figure 5. Interaction forces (F/R) between bitumen and silica as a function of separation distance in 1 mM KCl and different CaCl2 solutions at pH 8.2. Solid lines represent the DLVO fitting using A132 ) 3.3 × 10-21 J with the best-fitted decay length and Stern potential being as follows: for 0 mM CaCl2 (square), κ-1 ) 9.1 nm, ψB ) -76 mV, and ψS ) -59 mV; for 0.1 mM CaCl2 (circle), κ-1 ) 8.5 nm, ψB ) -45 mV, and ψS ) -48 mV; and for 1 mM CaCl2 (up triangle), κ-1 ) 4.6 nm, ψB ) -30 mV, and ψS ) -25 mV. Insert: the adhesive force (Fad/R) as a function of calcium ion concentration, at the loading force of 8-10 mN/ m.

Figure 6. Interaction forces (F/R) between bitumen and silica as a function of separation distance in 1 mM KCl and different CaCl2 solutions at pH 10.5. Solid lines represent the DLVO fitting using A132 ) 3.3 × 10-21 J with the best-fitted decay length and Stern potential being as follows: for 0 mM CaCl2 (square), κ-1 ) 9.1 nm, ψB ) -83 mV, and ψS ) -64 mV; for 0.1 mM CaCl2 (circle), κ-1 ) 7.8 nm, ψB ) -45 mV, and ψS ) -35 mV; and for 1 mM CaCl2 (up triangle), κ-1 ) 4.5 nm, ψB ) -38 mV, and ψS ) -8 mV. Insert: the adhesive force (Fad/R) as a function of calcium ion concentration, at the loading force of 8-10 mN/m.

mN/m. Again, the long-range forces can be well fitted by the DLVO theory, indicating that the change of interaction force profiles can be simply attributed to the diminished electrical double layer forces by the specific adsorption of calcium ions. To account for an attractive force profile with 1 mM calcium addition, the charge on the silica and bitumen surface at the Stern plane must have been significantly different, which will be further illustrated by the zeta potential measurements of corresponding colloidal suspensions. Effect of Temperature. Temperature is another critical operating parameter in bitumen extraction by the hot water process. Elevated temperature is usually considered to reduce the viscosity of bitumen in order to facilitate bitumen-air bubble attachment and engulfment.12-13,19,36 To appreciate the role of increased temperature in bitumen digestion/liberation, the colloidal forces were measured at different temperatures in a

Bitumen-Silica Interaction in Aqueous Solutions

Figure 7. Interaction forces (F/R) between bitumen and silica as a function of separation distance in 1 mM KCl solution at pH 8.2 and different temperatures. Solid lines represent the DLVO fitting using A132 ) 3.3 × 10-21 J with the best-fitted Stern potential being as follows: for 22 °C (square), κ-1 ) 9.1 nm, ψB ) -76 mV, and ψS ) -59 mV; for 35 °C (circle), κ-1 ) 9.0 nm, ψB ) -81 mV, and ψS ) -62 mV; and for 50 °C (up triangle), κ-1 ) 9.2 nm, ψB ) -83 mV, and ψS ) -64 mV. Insert: the adhesive force (Fad/R) as a function of temperature, at the loading force of 8-10 mN/m.

Langmuir, Vol. 19, No. 9, 2003 3917

Figure 8. Comparison of best-fitted decay length from the measured force profiles with calculated decay length using Debye-Huckel theory from the actual electrolyte concentration used in the corresponding force measurement.

solution of pH 8.2. As shown in Figure 7, increasing the temperature made the long-range forces between bitumen and silica slightly more repulsive. The increased repulsion can be accounted for by considering an increase in surface charge density at the bitumen/water interface. The longrange force can be reasonably well fitted with the classical DLVO theory, and a decrease in the fitted Stern potential of bitumen droplets from -76 to -83 mV corresponding to a temperature change from 22 to 50 °C was obtained. It appears that higher temperature enhanced the migration of surfactant molecules through bitumen and their accumulation at the bitumen/water interface. The deviation of the measured force profiles from those calculated using the classical DLVO theory was observed at larger separation distances at a higher temperature as shown by the dashed arrows in Figure 7. Qualitatively, this derivation agrees with the predicted temperature dependence of steric forces resulting from the brushlike surfaces. The adhesive forces as shown in the insert of Figure 7, on the other hand, disappeared with the solution temperature being increased to 35 °C. Discussion Interaction Forces and Data Fitting. The above results and analysis of colloidal forces clearly show that the long-range interactions between bitumen and silica sand grains are largely controlled by electrostatic double layer forces. This general finding suggests that to avoid detrimental heterocoagulation of bitumen with silica sand grains in a water-based bitumen recovery process, it is necessary to create a suitable surface charge condition by controlling the pulp slurry chemistry. To further confirm the electrostatic nature of the long-range repulsive forces, the fitted decay length from the measured force profiles was compared with those calculated using the well-known Debye-Huckel approximation formula42 based on the actual electrolyte concentration used in the force measurement. As shown in Figure 8, an excellent agreement between the two sets for all the force profiles presented in this paper confirms our interpretation of the electrostatic nature of the long-range repulsion. Further tests were performed to compare the fitted Stern potentials with the measured zeta potentials of bitumen droplets and silica particles. As shown in Figure 9a for

Figure 9. Effect of calcium ion addition on the zeta potential of (a) bitumen droplets and (b) silica particles in a 1 mM KCl solution.

bitumen droplets, an isoelectric point (iep) of pH 3.0 was obtained. Above the iep, the zeta potential becomes progressively more negative as pH increases till around 8, above which it levels off at ca. -82 mV. The addition of calcium ions caused a slight shift of iep toward a higher pH with the zeta potentials of the bitumen droplets above the iep becoming much less negative. From the shift of the iep, it is clear that calcium ions offer specific affinity to a bitumen surface, mainly through the binding with carboxylic groups of natural surfactants contained in bitumen. In the case of a silica suspension as shown by Figure 9b, no iep was observed and silica particles exhibited an increasingly negative zeta potential with

3918

Langmuir, Vol. 19, No. 9, 2003

Liu et al.

Figure 10. Comparison of the best-fitted Stern potential of the measured force profiles with the classical DLVO theory and the measured corresponding zeta potentials of silica (open symbols) and bitumen droplets (filled symbols).

Figure 11. Zeta potential distribution of a bitumen emulsion and a silica suspension in a 1 mM KCl solution at pH 10.5 without calcium ion addition, measured (a) separately and (b) as a mixture.

increasing pH over the range studied. The addition of calcium ions also suppressed zeta potentials of silica particles, and zeta potentials become less negative with increasing pH from 7 to 11. These results indicate a strong adsorption of calcium on the silica surface at these pHs, possibly in the form of CaOH+.56 The zeta potential results presented above are fairly informative to interpret the measured force profiles. Without calcium addition, repulsive force profiles are anticipated between negatively charged bitumen and negatively charged silica as shown in Figure 3. With the addition of 1 mM calcium ions, on the other hand, the electrically less charged silica and more negatively charged bitumen at pH 10.5 would result in an attractive force profile, which was indeed measured as shown in Figure 6 (triangles). Quantitatively, the plot of the measured zeta potentials against the corresponding fitted Stern potentials with the classical DLVO theory of the measured force profiles exhibited an excellent agreement as shown in Figure 10. This finding confirms that the measured long-range repulsive forces do originate from electrostatic double layer interactions. The deviations of measured forces from the classical DLVO theory at separations less than 2-5 nm, on the other hand, can be partially accounted for by the assumption of the constant surface potential boundary used for solving the Poisson-Boltzmann equation and/or surface roughness. In addition, the presence of steric forces, predominating at a range of less than 2-5 nm, cannot be overlooked. The presence of surface active macromolecules at the bitumen/water interface may result in a brushlike surface conformation as observed in polymer systems, inducing a brush type of electrosteric repulsive force. A shorter range interaction in the present case as compared with typical polymer systems is indicative of much smaller surface brushes of asphaltene molecules. The quantification of this additional short-range force is beyond the scope of this study and will be discussed elsewhere. The attribution of the deviation of short-range repulsive forces from DLVO theory to brush-type steric forces is supported by its temperature dependence of the range at which deviation starts, as shown by dashed arrows in Figure 7. Adhesion Forces. The adhesive force (pull-off force) was found to be highly dependent on solution pH and calcium addition. The adhesive force between two particles generally originates from the molecular/atomic interac-

tions in the contact area, such as electrostatic interaction, chemical bonds, and hydrogen bonding. These contact forces are highly sensitive to surface composition. In the present system, silica surfaces are dominated by the OHgroup and can specifically adsorb calcium ions at higher pH, while the bitumen surface bears various types of natural surfactants,40 which could become protonated or dissociated, depending on the solution pH. At low pH, the cationic surfactants on the bitumen surface are protonated to generate cationic sites (RNH3+), which can interact with OH- groups on the silica surface to induce a strong adhesive force.57 At high pH, the dissociated anionic surfactants (RCOO- and ROSO3-) dominate the bitumen surface. These anionic surface groups have little affinity to the OH- groups on the silica surface, making the system nonadhesive. These interpretations are consistent with our experimental observations shown in the insert of Figure 3. With the addition of calcium ions, the specific adsorption of calcium ions on the silica surface in the form of CaOH+ at high pH (at pH 10.5 in Figure 9b) is responsible for the observed strong adhesion (at pH 10.5 in the insert of Figure 6) between bitumen and silica through chemical binding of carboxylate groups on the bitumen surface with calcium ions on the silica surface. Particle Heterocoagulation. The current study clearly shows that tuning colloidal forces by solution pH, salt concentration, divalent cation addition, and elevated temperature can be confidently attributed to their ability to alter electrostatic double layer and adhesive forces. To demonstrate the interaction in a colloidal bitumen and silica system, the zeta potential distributions of a bitumen emulsion and a silica suspension, individually and in a mixture, were measured and the results were interpreted in terms of colloidal interactions.41 For illustrative and confirmative purposes, only the results with and without calcium addition at a solution pH of 10.5 are presented here. The zeta potential distributions measured with bitumen droplets or silica particles alone in the electrolyte solution without calcium addition were centered at -82 and -67 mV, respectively, as shown in the overlaid histogram of Figure 11a. A similar zeta potential distribution histogram was obtained for the mixture of the two species as shown in Figure 11b. The presence of two distinct distribution peaks at -82 and -69 mV indicates that bitumen droplets and silica particles in the mixture are noncoagulative, that is, they are present separately as individual particles. This observation agrees with the

(56) King, R. P. Principles of Flotation; South Africa Institute of Mining and Metallurgy: Johannesburg, 1982.

(57) Noy, A.; Vezenov, D. V.; Lieber, C. M. Annu. Rev. Mater. Sci. 1997, 27, 381-421.

Bitumen-Silica Interaction in Aqueous Solutions

Figure 12. Zeta potential distribution of a bitumen emulsion and a silica suspension in a 1 mM KCl solution at pH 10.5 with 1 mM calcium ion addition, measured separately (a) and as a mixture (b).

presence of a strong repulsive force between the two as shown in AFM force measurements (Figure 3). When 1 mM calcium ions were added, zeta potential distribution histograms for bitumen droplets and silica particles measured separately were found to be much less negative, as shown in Figure 12a. In contrast, when the zeta potential distribution was measured with the mixture of the two, only a single broad distribution peak located between the two original peaks was observed. The disappearance of the original distribution peaks of silica and bitumen and the appearance of a new broad distribution peak between them suggest that the silica particles and bitumen droplets were heterocoagulated to form composite aggregates. This behavior indicates a strong attraction between the two components, as anticipated for a system of an attractive long-range force profile with a finite adhesion measured with the AFM (see Figure 6). Relevance to the Oil Sands Industry. The adhesive force determines bitumen attachment onto silica surfaces, while long-range forces are the key for dispersion or coagulation of silica-bitumen colloidal systems. In general, long-range forces described by the DLVO theory or extended DLVO theory can be used satisfactorily to interpret the stability of colloidal systems. However, for a dynamic colloidal system such as the bitumen extraction process, both the long-range forces and adhesive forces have to be considered. For instance, the interaction forces between bitumen and silica are repulsive in a simple electrolyte solution at a pH less than 8 (Figure 3), where no coagulation between them was anticipated. This anticipation is in contrast to the visual observations of poor bitumen detachment from a sand grain in a neutral pH aqueous medium made by other researchers.19,22 To reconcile the observed discrepancy and to further understand bitumen digestion from silica sand grains, the adhesive forces between bitumen and silica have to be considered. At a pH below 8, strong adhesion compounded with a weak long-range repulsive force between bitumen and silica is responsible for the observed poor bitumen digestion and coagulation of bitumen with silica.19,22 This finding suggests that bitumen is unlikely to be liberated from sand grains at a solution pH below 8. At a pH greater than 8, the adhesive forces become extremely weak and approach 0 with a strong long-range repulsive force profile. Such a system facilitates bitumen liberation from sand grains and is consistent with the required caustic addition to adjust pulp pH above 8 in industrial bitumen extraction processes.

Langmuir, Vol. 19, No. 9, 2003 3919

However, excess caustic addition to increase pulp pH may not always be beneficial for bitumen digestion when calcium ions are present in the system. The stronger adhesion and long-range attraction incurred by calcium addition at pH 10.5 as shown in Figure 6 will not facilitate bitumen liberation from sand grains. Clearly the presence of calcium ions even at 1 mM (40 ppm) concentration in the bitumen extraction process is detrimental to bitumen separation, as it would hinder the detachment of bitumen from sand grains due to increased adhesion forces between the two. Our study provides a legitimate justification as to why the industrial scale bitumen extraction system operates at a pulp pH of around 8.5 as a compromise. It is also evident from the direct force measurement that increasing the electrolyte concentration is unfavorable for bitumen droplet and silica particle separation as the liberated bitumen may become more easily heterocoagulated with sand grains by reducing long-range repulsive forces without affecting the adhesion forces as shown in Figure 4. The elevated temperature could increase the repulsive interaction force and decrease the adhesive force as shown in Figure 7. The results obtained in this study suggest that in addition to the commonly accepted role of high temperature in reducing bitumen viscosity, the elevated temperature can to some extent facilitate bitumen liberation by controlling interaction and adhesive forces. It is evident that for a bitumen extraction system, a repulsive colloidal force profile and zero adhesive force represent a “stable” colloidal state, resulting in an easy liberation and absence of coagulation between bitumen and silica sand. From the direct force measurement, this condition can be realized by operating the process at high temperature and/or high pH without calcium addition. This favorable condition for bitumen extraction is practiced in these commercial bitumen extraction operations.1,2 For a system of repulsive colloidal force with a finite adhesive force, a careful balance between the hydrodynamic force and the colloidal force is required to create a favorable condition for bitumen separation. This would be a difficult situation to create for a dynamic system, as the energy barrier is normally smaller than the adhesive force, should it exist. The particles usually can obtain sufficient kinetic energy to overcome the repulsive force barrier and become coagulated with each other, but not enough shear to overcome the adhesive force and detach from each other. This case was seen in the low solution pH in Figure 3. For a system with an attractive interaction force and a finite adhesive force as in the case of 1 mM calcium ion addition and pH 10.5 (Figure 6), it is extremely difficult if not impossible to separate the bitumen from the sand grains. Summary The research presented here demonstrated that AFM force measurement in conjunction with the spin-coating technique is a powerful tool to quantitatively investigate the bitumen/silica interaction, which provides further insight into the bitumen digestion mechanism in HWEP technology. The following are general conclusions derived from the current study. 1. Solution pH is a critical factor that controls the bitumen/silica interaction. With increasing solution pH, the long-range repulsive force increases and adhesive force decreases. 2. With the addition of calcium ions, the repulsive force between bitumen and silica becomes smaller and even reversed from repulsive to attractive, depending on the

3920

Langmuir, Vol. 19, No. 9, 2003

pH and calcium ion concentration. The adhesive force increases profoundly with calcium ion addition. 3. At temperatures higher than room temperature, increasing the temperature leads to a slight increase in the repulsive interaction force and a decrease in the adhesive force. 4. Increasing the salinity decreases the repulsive forces between bitumen and silica significantly but has only a marginal effect on the adhesive forces. 5. The measured repulsive forces can be well fitted with the classical DLVO theory, with the fitted Debye lengths and Stern potentials being in excellent agreement with those calculated and measured, respectively, indicating that the electrostatic double layer forces play a dominant role in bitumen/silica interaction. 6. The surface force measurement identified bitumen-

Liu et al.

silica heterocoagulation conditions above pH 10 with 1 mM calcium ion addition, which is substantiated by the zeta potential distribution measurement. 7. Deviation from the classical DLVO theory is observed at separations of less than 2-5 nm and is attributed to short-range steric repulsion. Acknowledgment. The authors acknowledge the financial support from COURSE with Syncrude Canada and Albian Sands as industrial partners, the Natural Sciences and Engineering Research Council of Canada (NSERC), and the Industrial Research and NSERC Chair in Oil Sands Engineering. Provision of oil sands samples by Syncrude Canada Ltd. is also acknowledged. LA0268092