Kinetics and Properties of Asphaltene Adsorption on Surfaces

Article pubs.acs.org/EF

Kinetics and Properties of Asphaltene Adsorption on Surfaces Atoosa Zahabi and Murray R. Gray* Department of Chemical and Materials Engineering, University of Alberta, Edmonton, Alberta, Canada T6G 2G6

Tadeusz Dabros CanmetENERGY, Devon, Alberta, Canada T9G 1A8 ABSTRACT: A quartz crystal microbalance (QCM) was used to probe asphaltene adsorption on gold surfaces and hydrophilic and hydrophobic silica particles. The adsorption studies were conducted with solutions of various ratios of pentane solvent to model oil (S/O) and asphaltene concentrations. The adsorption of asphaltenes on solid surfaces at different S/O showed multilayer deposition without reaching equilibrium after 16 h. Adsorption of asphaltenes and other components on the surfaces were detectable well below the onset of precipitation. The amount of material adsorbed increased significantly after the onset of precipitation. Adsorption was more pronounced on the gold surface than on the silica particles. The swelling of the asphaltene aggregates on the surface of the quartz crystal led to restructuring of the adsorbed material on the surfaces, depending on the S/ O. For S/O > 0, the swelling effect suggested open nanoaggregate or polymer brush structure. The adsorbed asphaltenes demonstrated more viscoelastic behavior as S/O was increased. The kinetics of the adsorption of asphaltenes on gold at room temperature suggested the formation of very large, slowly diffusing aggregates, even in toluene. These results were not consistent with optical microscopy or other methods.

■

INTRODUCTION The asphaltene solubility class of petroleum contains large and polar molecules made up of different aromatic groups and aliphatic chains and bridges. This fraction contains heteroatoms, such as nitrogen, sulfur, and oxygen, and the metal components vanadium and nickel. The asphaltene fraction also includes surface-active components with a hydrophobic structure that contains some hydrophilic functional groups.1,2 The stable dispersion of asphaltenes in crude oil can be destabilized by adding a paraffinic solvent, such as pentane or hexane, which results in precipitation. Precipitated asphaltenes can flocculate water droplets and solid particles, which is instrumental in producing dry and solids-free bitumen from a water-rich froth3 and has potential for removal of fine solids from refinery streams.4 Asphaltenes adsorb on surfaces as individual molecules or colloidal aggregates of various sizes.5−11 Accumulation of this material forms solid deposits that can foul pipelines and process equipment. Asphaltenes can also adsorb on mineral surfaces and reservoir rocks, forming deposits that limit extraction of heavy oils from reservoirs and changing the wettability of the mineral surface.12,13 These undesirable behaviors motivate research activities on the deposition of asphaltenes onto various surfaces. Information on the equilibrium behavior of asphaltene adsorption from solution and rate of adsorption is valuable for pipeline flow design. Fundamental studies of the deposition of asphaltenes on surfaces, under controlled conditions, help to identify the underlying mechanisms. The amount of asphaltene deposited on solid surfaces has been measured by UV−vis spectroscopy,14−17 quartz crystal microbalance (QCM),7,18−23 and X-ray photoelectron spectroscopy (XPS).7,24 These studies are mainly on the type of asphaltene adsorption on solid surfaces, the kinetics of deposition, and the amount of adsorption on © 2011 American Chemical Society

different surfaces. The liquid phases in these studies were mainly asphaltene in toluene or a mixture of toluene and alkane. In some cases, the relationship between deposition and the onset of precipitation was not defined. Surfaces employed in previous adsorption studies included metallic and mineral surfaces. Asphaltene adsorption has been studied on a variety of mineral surfaces, including clay,25,26 mica,27 silica,28 montmorillonite,26 quartz,29 dolomite, and calcite.12 Both monolayer and multilayer adsorption have been observed in asphaltene deposition studies. Langmuir adsorption of asphaltenes from toluene solutions was observed on clay minerals,26 kaolin, CaCO3, BaSO4, FeS, Fe3O4, TiO2, SiO2,30 gold,7 stainless steel, alumina, and iron.16 Multilayer adsorption has been observed on oxide surfaces, including silica, quartz, dolomite, calcite, kaolin, Fe2O3, and TiO2.12,13,28 Multilayer adsorption requires significant attraction between asphaltene nanoaggregates or particles; therefore, this behavior was observed more commonly for higher concentrations of asphaltenes and in toluene−alkane solutions.21 No equilibrium was observed for the deposition of asphaltenes from alkane solutions, even after 300 and 700 min in Ekholm et al. and Xie et al., respectively.21−23 Rudrake et al. reported equilibrium deposition for a C7-asphaltene-in-toluene solution after 200 min, with a thickness around 6−8 nm.7 Goual et al. reported that, for low concentrations (8 wt %), however, their QCM results showed that the steady-state condition can be achieved after 800 s. From the literature, it is obvious that asphaltene adsorption on solid Received: September 27, 2011 Revised: December 19, 2011 Published: December 26, 2011 1009

dx.doi.org/10.1021/ef2014698 | Energy Fuels 2012, 26, 1009−1018

Energy & Fuels

Article

where f 0 is the fundamental resonance frequency (f 0 = 5 × 106 Hz), ρq is the specific density of the quartz (2650 kg/m3), and νq is the shear wave velocity of the quartz (3340 m/s).18,23 The density is ρ and the viscosity is η, where the subscript S refers to solvent solutions of toluene + pentane and L refers to oil− toluene−pentane. In some QCM instruments, the dissipation (ΔD) or resistance (ΔR) is measured. The shift in the dissipation factor and the resistance in a liquid medium are given as follows:35,40

surfaces is a complex and highly inconsistent phenomenon and there is still no simple explanation for these experimental results. Clearly, the nature of the solvents, the asphaltene concentration,7 and the proximity to the onset of asphaltene precipitation31 can be major factors determining whether an isotherm is observed. Dudasova et al. estimated the diffusion coefficients of an asphaltene-in-toluene solution by NMR to be on the order of 10−10 m2/s.30 Abudu et al.23 estimated the diffusion coefficient by QCM to be in the same range as that reported by Dudasova et al.,30 but an error in their calculation of the time of diffusion led to incorrect diffusion coefficients. Xie et al.22 estimated much smaller diffusion coefficients by QCM, in the range 10−11 to 10−12 m2/s for asphaltene-in-toluene solutions of 50−200 ppm. The reason for such low values for the diffusion coefficient from QCM has not been explained. In this study, the QCM technique is used to measure the amount and rate of asphaltene deposition on gold and silica surfaces from a model oil solution of pitch in toluene. Rates of deposition were used to estimate the diffusion coefficient and the particle size. A range of ratios of pentane solvent to model oil solution (S/O) were studied, up to and including the onset of precipitation. Our previous work showed that the onset of visible aggregation by optical microscopy was at S/O = 0.43.4 However, destabilization of a dispersion of silica particles occurred at S/O below the onset of asphaltene precipitation. Our hypothesis is that asphaltene aggregates adsorb on surfaces even at S/O below the onset of asphaltene precipitation.

f 0.5 ⎡ ρ η ⎤0.5 ΔD = 2 0 ⎢ L L ⎥ ρ νq ⎣ π ⎦

⎡ 1 ρ η ⎤0.5 ⎢ · L L⎥ ΔR = 2 8e 26AM ρq νq ⎢⎣ f0 π ⎥⎦ πμqhq

(4) −4

where AM= is the area of the crystal (0.3149 × 10 m ), hq is thickness for coupling (3.31701662 × 10−4 m), e26 is the piezoelectric constant (0.0966 C/m2), and νq = (μq/ρq)0.5, where μq is the shear modulus of the quartz (2.947 × 1010 Pa). When mass is adsorbed on the quartz, a frictional energy is created that increases the dissipation (ΔD). If the film is viscous, energy is dissipated by the oscillatory motion induced in the film. Hence, a rigid adsorbed layer gives no change in dissipation, and eq 1 applies, while a loose layer gives a dissipation or resistance increase. The QCM instrument measures either the resistance change (ΔR) with frequency or the dissipation (ΔD). In eqs 3 and 4, the term (ρLηL)0.5 would change according to the density and viscosity of the medium; however, the rest of the parameters in the equations are constants and depend only on the characteristics of the quartz crystal. Therefore, measurements of either the dissipation or the resistance can be used to characterize the rigidity of the material deposited on the crystal surface. Kinetics of Asphaltene Deposition. The total amount of surface deposit can be calculated using the Ward and Tordai equation, as follows:42

■

THEORY Quartz Crystal Microbalance (QCM). Detailed studies of QCM have been described by many authors.32−40 Here, we give a brief discussion of its behavior and governing equations in liquid media. Sauerbrey was the first to use QCM technology to measure mass changes in gas phase. The quartz crystal oscillates at a certain frequency when an AC voltage is applied. If any mass is adsorbed on the surface of the crystal, the frequency decreases. From the change in frequency, the mass is calculated from the Sauerbrey equation:32−40

t D M (t ) = 2 [c0 t − π 0

∫

Δf = − C f Δm (1) where Cf is the sensitivity factor for the quartz crystal and depends on the properties of the crystal (56.6 Hz μg−1 cm2 for a 5 MHz AT-cut quartz crystal at room temperature), Δm is the adsorbed mass on the surface of the quartz crystal, and Δf is the change in frequency upon adsorption of mass on the crystal surface. For many years, QCMs were used as gas-phase mass detectors; however, Konash and Bastiaans41 showed that the QCM can be used in liquid media as well. In liquid media, the characterization of the adsorbed layer is different from that in the gas phase; therefore, both resonance frequency and dissipation of the quartz oscillator must be considered. The loading of the liquid on the crystal causes the frequency to shift from its frequency in air. The change in frequency is not due to mass adsorption in this case, but rather to the liquid phase damping, as follows:23,34 f 3/2 (ρ η )0.5 − (ρSηS)0.5 Δfloading = − 01/2 L L ρq νq π

(3)

q

d[(t − τ)1/2 ]]

2

1/2

φ(τ) (5)

where M(t) is the amount of material adsorbed at time t, c0 is the bulk concentration, D is the diffusion coefficient, c0 is the solute concentration in the bulk solution, τ is a variable for time, ϕ(τ) is the subsurface concentration. If ϕ(τ) = 0, eq 5 is reduced to

⎛ Dt ⎞1/2 M(t ) = 2c0⎜ ⎟ ⎝π⎠

(6)

The condition ϕ(τ) = 0 is true only if there is no backdiffusion of the solutes to the bulk, that is, when the adsorption is irreversible. Thermal motion of the particles may lead to the detachment of the particles from the surface. The irreversible adsorption in the case of asphaltenes corresponds to adsorption in a deep primary energy minimum with possible chemical bonds being formed between the asphaltene and the substrate. In our work, asphaltenes are assumed to have strong attachment to the surfaces. Therefore, the probability of removal of deposited asphaltene from the gold surface is very low, and ϕ(τ) can be assumed to be zero.

(2) 1010


Energy & Fuels

Article

QCM to measure the amount of asphaltene deposition on solid surfaces. The probe used had 5 MHz AT-cut quartz crystals coated with gold and a QCM25 crystal oscillator (QCM200 from Stanford Research Systems (SRS), Santa Barbara, CA, U.S.A.). The constants of the crystals were C = 33 fF, L = 30 mH, and R = 10 Ω (for a dry crystal). The capacitance, C0, was zeroed each time the crystal was changed. Immersion Experiments. In all the experiments, both resonance frequency (f) and resistance (R) data were collected. After the crystal was mounted, the response was measured in air for at least 10 min to obtain a stable baseline (stage 1). The probe was then immersed vertically in a pentane−toluene solution for 10−15 min until a stable reading was obtained (stage 2); then, it was transferred to an oil− toluene−pentane solution with the same pentane−toluene ratio as in stage 2 (stage 3). The frequency was measured in the toluene− pentane−oil solution for 2 h, and then, the probe was rinsed with pure toluene and immersed in toluene until a stable reading was obtained (stage 4). In all cases, the sample container was covered carefully to avoid evaporation of the solvents. The probe was then removed from the toluene, washed with excess toluene, and allowed to stand in air until the frequency became constant (stage 5). The mass adsorbed on the crystal was calculated after washing with toluene and drying in air at the end of the experiment. Flow-Through Cell. To measure the amount of adsorption on silica particles at various S/O, the flow-through cell method was used to avoid losing the silica particles during immersion in the solution. To compare the results from adsorption on silica particles with results on gold alone, the same experiments were repeated for the gold surface as well. The flow-through cell (SRS QCM, Santa Barbara, CA, U.S.A.) is used in place of the crystal retainer ring of the crystal holder. Once installed, the volume of the cell is around 0.15 mL. The flow-through cell includes two inlet and outlet ports with a 0.040 in. i.d. throughhole and is fitted with barbed hose adapters for 0.062 in. i.d. tubing. A 5-mL syringe was used for injecting the sample into the cell. The sample was injected at a rate of around 5 mL/min. First, a stable baseline in the air was established, then the oil solution was mixed with pentane at various ratios and injected into the cell for a period of nearly 2 h while the data were collected. Spin coating (Laurell Technologies Corporation, U.S.A.) was used to coat the gold quartz crystal with silica particles (AngstromSphere from Fiber Optics Center, Inc.). Hydrophilic particles had been acid washed and had a contact angle of 20 ± 10°, while the hydrophobic particles were treated in 5% (v/v) dichlorodimethylsilane in toluene for 10 min to give a contact angle over 90°.4 A solution of 0.1 wt % silica particles in isopropyl alcohol (IPA) was prepared, and the sample was shaken with a vortex shaker for 20 min. Then, droplets of the sample were placed on the surface of the gold crystal. The quartz was spun at 500, 1000, 2000, 3000, 4000, and 5000 rpm for 30 s each and then for 2 min at 6000 rpm. The coating was checked with a microscope to make sure that there was monolayer adsorption of silica particles on the surface of the probe. For concentrations of silica particles in IPA higher than 0.1 wt %, we observed multilayer accumulation of particles in some spots, while reducing the concentration to less than 0.1 wt % resulted in empty areas on the surface of the crystal. Although the surface was not covered completely with silica particles, we avoided increasing the concentration of silica particles in IPA to avoid multilayer accumulation of silica particles on the surface of the gold-plated quartz crystal.

This condition (ϕ(τ) = 0) should be satisfied not only at the beginning of the adsorption but also at all times during the adsorption process. If this condition is true, then the apparent diffusion coefficient may be calculated from the slope of M versus √t.

⎛ D ⎞1/2 Slope = 2c0⎜ ⎟ ⎝π⎠

(7)

For a spherical particle, the diffusion coefficient is given by the Stokes−Einstein equation as follows:

D=

kBT 6πηr

(8) −23

where kB is the Boltzmann constant (1.3806503 × 10 m kg s−2 K−1), T is the temperature (296 K), η is the viscosity of the surrounding fluid, and r is the radius of the hard sphere. From this equation and the diffusion coefficient, one can estimate the radii of the particles, asphaltene molecules, or aggregates as equivalent spheres.

■

2

MATERIALS AND METHODS

Model Oil Solution. The oil solution was a 5 wt % solution of pitch material in toluene (model oil). The pitch was the vacuum residue fraction of thermally cracked Cold Lake bitumen, as a representative asphaltene-containing refinery stream. The samples of model oils were shaken on a horizontal table shaker for 2 h and left overnight to make sure that all of the pitch was dissolved in toluene. These samples were kept as a stock solution for the entire series of experiments. Solutions of asphaltenes and maltenes in toluene were prepared the same way. Asphaltene Extraction. Asphaltenes were separated by first mixing pitch material with chloroform at a 1:0.4 mass ratio of pitch to chloroform. The mixture was shaken with a table shaker for 2 h until all the pitch material was dissolved in chloroform. Then, pentane was added to the mixture at a 40:1 volume ratio of pentane to pitch mixture. The mixture was stirred for 4 h and left overnight to allow the precipitated asphaltenes to settle. Then, the supernatant was removed, and the remaining precipitate was further diluted with pentane at a 4:1 volume ratio of pentane to asphaltenes. After 4 h shaking, the final mixture was filtered using 0.22-μm Millipore filter paper to remove the asphaltenes, which were washed thoroughly with pentane until the filtrate became colorless. The precipitated asphaltenes were dried in an oven at 80 °C.43 The solvents were separated from the filtrate by rotary evaporation, and the remaining maltenes were then dried in an oven at 80 °C. The asphaltenes and maltenes were kept in a desiccator for further experiments. Solutions of asphaltenes and maltenes were prepared to determine the contributions of these fractions to adsorption on surfaces. The original asphaltene content of the pitch material was 60 wt %. To examine the contributions of each fraction in a 5 wt % solution of pitch, a sample with 3 wt % asphaltene in toluene and a sample with 2 wt % maltenes in toluene were prepared. Property Measurements. Solution densities were measured using a DMA 4500 density meter (Anton Paar, U.S.A.). The viscosity was measured by using a Viscolab 3000 rheometer (Cambridge Applied Systems, MA, U.S.A.). All the sample viscosities were measured in a thermostatic bath using a glass capillary viscometer as well. The viscosities were measured at 22 °C. For the samples in which pentane was added to the oil solution, the mixture was left for 1 h to allow any flocculated material to settle before taking the density and viscosity measurements of the supernatant. QCM Experiments. QCM is a sensitive technique used to measure the adsorbed and adhered mass on different surfaces. Two types of measurements are feasible: QCM-D with measurement of dissipation and R-QCM with measurement of resistance. In this study we used R-

■

RESULTS AND DISCUSSION Stages in QCM Immersion Experiment. Each QCM immersion experiment in this work had five stages, which are shown in Figure 1; recording the signal in air, immersion in solvent, deposition of asphaltene, followed by rinsing of the surface material in toluene and air drying. The data in Table 1 summarize the significant shifts in frequency due to changes in the medium surrounding the probe and due to adsorption and removal of asphaltenes. In the 1011


Energy & Fuels

Article

The solvent in the oil mixture in stage 3 is similar to the solvent in stage 2, that is, toluene−pentane in the same ratio. The change in frequency due to liquid loading for transferring from stage 2 to 3 was calculated by using eq 2 and is shown in Table 2. Comparing the experimental change in frequency for stage 2 to 3 (Table 1) with calculated Δf (2 → 3) due to liquid loading (Table 2) shows that most of the frequency change in the transition from stage 2 to the end of stage 3 for S/O > 0 is due to the deposition of asphaltene on the surface of the probe. At S/O = 0, the total frequency shift was −114.4 Hz, compared to an expected shift of −68.5 Hz due to the change in the solvent medium. Therefore, the effect of solvent loading at S/O = 0 was larger than the contribution of asphaltene deposition on the surface. In contrast, at higher values of S/O, the shifts due to the solvent alone were smaller; therefore, the contribution of the asphaltene deposits was dominant. At S/ O = 0.5, for instance, the calculated change in frequency due to liquid loading is about −38.6 Hz, whereas the experimental change in frequency is about −1768 Hz. Therefore, the effect of asphaltene deposition on the surface is much larger than the effect of liquid loading. In stage 4, the probe was washed with toluene and remained in toluene for 2 h. If any materials adsorbed on the surface reversibly, they would be dissolved in toluene during this period. After stage 4, the probe was washed with excess toluene and brought back to air (stage 5). The probe remained in the air until a stable frequency reading was obtained. The frequency change in this step is due to irreversibly adsorbed material; no other factor affects this signal. The adsorption in stage 3 shows continuous deposition of asphaltenes on the surface of the gold without achieving a plateau or stable mass after 2 h. To determine whether an equilibrium in the adsorption could be achieved, one experiment for 0.2 g/L pitch in toluene solution (model oil) was carried out for 16.3 h. The results are shown in Figure 2.

Figure 1. Sequence stages for the QCM immersion experiment (S/O = 0.5) (solid line, frequency; dashed line, resistance). Stage 1: mounting the probe in air. Stage 2: exposing the probe to pentane and toluene. Stage 3: transferring the probe to the oil−toluene−pentane solution. Stage 4: washing the probe with toluene. Stage 5: drying the probe in air.

Table 1. Experimental Changes in Frequency Due to Immersion in Different Solutions for Various S/O S/O

Δf (Hz) (1 → 2)

Δf (Hz) (2 → 3)

Δf (Hz) (4)

Δf (Hz) (5)

0.5 0.43 0.36 0.0

−443 −450 −460 −585

−1768 −771 −213 −114

−1609 −1065 −686 −638

−195 −72 −29.5 −1.8

transition from stage 1 to stage 2, exposing the probe to solvent (pentane and toluene in this study) gave a change in both frequency and resistance due to the changes of the density and viscosity of the medium. As there is no adsorption, the frequency reaches a new equilibrium in few seconds and becomes stable, with a value of Δf in Table 1 that was comparable to the predicted value from eq 2, within 10%. The effect of S/O on Δf in Table 1 was consistent with the predictions from eq 2 due to changes in density and viscosity of the solutions (data provided in Table 2). The change in resistance follows the same pattern as the frequency. Table 2. Density and Viscosity of Toluene−Pentane (Stage 2) and Oil−Toluene−Pentane (Stage 3) for Various S/Oa

a

S/O

density of stage 2 (kg/m3)

viscosity of stage 2 ( × 103) (Pa s)

density of stage 3 (kg/m3)

viscosity of stage 3 ( × 103) (Pa s)

calcd Δf for stage (2→ 3), (Hz)

0 0.36 0.43 0.5

867 789 779 770

0.62 0.43 0.41 0.39

876 796 785 776

0.78 0.5 0.48 0.47

−68.5 −36.4 −34.9 −38.6

Figure 2. Adsorbed mass for 0.2 g/L pitch in toluene (model oil) for 16.3 h (S/O = 0).

Δf (2 → 3) as calculated by using eq 2.

These data illustrate that no equilibrium condition was achieved, so that asphaltene-on-asphaltene deposition continued well beyond the formation of a monolayer, which is estimated to occur at a deposited mass on the order of 3.2 mg/ m2. In the context of multilayer adsorption, the gradual decrease in slope in Figure 2 could indicate changing structure of the deposit or depletion of components from the liquid solution. Dudasova et al. showed that adsorption of asphaltene on gold is a multilayer deposition phenomenon without reaching equilibrium.44 On inorganic particles, however, they observed

When the probe was immersed in the oil solution (oil− toluene−pentane) in stage 3, the frequency steadily decreased and the resistance gradually increased. The initial small change in the frequency was due to the small increase in the viscosity and density of the medium, from toluene−pentane to oil− toluene−pentane. The slower progressive change over a period of 2 h was due to the deposition of asphaltenes on the surface of the probe (gold-coated quartz). 1012


Energy & Fuels

Article

Langmuir type adsorption.30 Goual et al.18 reported that the steady-state condition can be achieved; however, it depends on the concentration of asphaltenes in the solution. Abudu et al.23 reported that no plateau was observed in alkane solution of asphaltene even after 3.5 h. Ekholm et al.21 and Xie et al.22 reported that no equilibrium can be achieved in asphaltene in alkane or toluene solutions. Our observation of continued deposition on the surface, attributed to multilayer adsorption of nanoaggregates as well as restructuring of the surface material with time, is consistent with most of the previous studies that used a similar range of concentrations in static, nonflowing conditions. Amount of Adsorbed Asphaltene at Various S/O by Immersion Method. For each S/O, the adsorbed mass was calculated in the air after washing the crystal with toluene (stage 5). When the mass is calculated in the air, we can assume that there is no viscosity and density effect, and that the mass of the deposit can be calculated without considering the liquid loading, using the Sauerbrey equation (eq 1). In this case, the change in frequency is due to irreversibly adsorbed mass. The data in Figure 3 show that, even below the onset of asphaltene

Table 3. Comparison of Masses Adsorbed from Asphaltenes, Maltenes, and Pitch in Toluenea

a

material

concn in toluene (wt %)

concn in pitch (wt %)

density (kg/m3)

viscosity ×103 (Pa s)

adsorbed mass (mg/m2)

pitch asphaltene maltenes

5 3 2

59.5 15.8

876.19 870.63 867.04

0.78 0.74 0.73

7.95 13.09 4.86

Deposition time = 2 h.

(S/O = 0), the amount of adsorption in asphaltene solution (13.09 mg/m2) was higher than that in both the maltene solution (4.86 mg/m2) and the pitch solution (7.95 mg/m2). Although the wt % of asphaltenes in both pitch and asphaltene solution was the same, the amount of adsorption was consistently lower in the pitch solution. This result is consistent with competitive adsorption between the two materials. Adsorption of maltene components, mainly as molecular species from solution, would fill some of the available sites on the surface of the gold, reducing the mass of adsorbed asphaltene nanoaggregates in the pitch solution and reducing the total adsorbed mass. We expect that if pentane is added to destabilize the asphaltenes, the maltenes would still contribute a portion of adsorbed mass, but that this mass would not increase with pentane concentration, indicating that the fraction due to the asphaltenes is dominant. Adsorption on Silica Particles with Different Hydrophobicities. In our previous work,4 for the FTIR spectroscopic analysis, the hydrophilic silica particles were immersed in the oil solution for 2 h and washed with toluene and the amount of asphaltene adsorption was then determined. The same time was chosen here to measure the amount of asphaltene adsorption on silica particles with different hydrophobicities using QCM. The gold-coated quartz crystal probe was covered with a monolayer of silica particles by spin coating. The probe was mounted in the flow-through cell and the oil solutions at various S/O were injected into the cell carefully with a very low flow rate (5 mL/min) to minimize removal of the silica particles from the surface of the probe. The probe was in contact with the oil solution for 2 h. As with the immersion method, no equilibrium was observed for the experiments using the flow-through cell. Comparisons of the adsorbed masses of asphaltenes on silica particles with different hydrophobicities and on the gold surface, with no surface correction, did not show significant differences between the adsorbed masses. The surface of the silica particles can be calculated based on a hexagonal monolayer of 1 μm silica particles on the probe. The results considering the correction to the coated area are shown in Figure 4. From these results, it is clear that the gold surface adsorbed more material than the silica surfaces. The amounts of adsorption on the hydrophilic and hydrophobic silica particles, however, were almost identical. Figure 5 shows the amount of asphaltene adsorption on hydrophilic silica particles measured with FTIR4 and QCM at various S/O. The amount of adsorbed material in the QCM experiments is higher than the FTIR measurements only at the highest S/O values, simply because of the washing of the particles with excess toluene in the FTIR experiments. In the FTIR experiments, after suspending the silica particles in the model oil for 2 h, the particles were separated by centrifugation and were washed with toluene until the supernatants were

Figure 3. Amount of adsorbed mass on gold at various S/O after 2 h of deposition. Quantity measured after washing with toluene and drying in air.

precipitation, which is at S/O = 0.43, there is evidence of adsorbed materials on the surface of the gold but the amount of the adsorbed mass was much larger above S/O = 0.43. The masses reported here are the result of adsorption after 2 h, because no equilibrium was observed and the frequency did not level off for the duration of stage 3 of each experiment. The mass of adsorbed asphaltenes was consistent with previous studies at a given solubility parameter of the liquid solution. With a solution of solubility parameter of 18 MPa1/2, Abudu et al.23 observed deposition of 4.6 mg/m2, compared to 2.9 mg/m2 after 2 h in our study. A larger discrepancy was observed for low values of solubility parameter, below 17.4 MPa1/2, and could be due differences in the onset of flocculation for asphaltenes from different sources. Ekholm et al.21 observed continuous adsorption over a period of 50 min to reach 3.5 mg/m2 of deposit, comparable to our result of 5.1 mg/m2 in toluene. The difference can be attributed to the different sources of the asphaltenes. Adsorption of Asphaltenes and Maltenes in Toluene Solution. In all of our deposition experiments, the solution of pitch in toluene was used as the model oil, and we assumed that most of the adsorbed material on the solid surfaces was asphaltenes. To check this assumption, the amount of adsorption on gold was measured in toluene solutions of C5asphaltenes and maltenes (Table 3). When no pentane is added 1013


Energy & Fuels

Article

of material from the model oil and part of it is due to solvent trapped in the deposited layer. This latter contribution is particularly clear when we analyze the results from the end of stage 4, where a stable layer of adsorbed material remains in toluene (the frequency became stable) in stage 5, which is in air. The experimental and predicted frequency shifts for Δf (4) are given in Table 4. Table 4. Experimental and Expected Changes in Frequency in Stage 4 for Various S/Oa

Figure 4. Comparison of masses deposited from model oil solutions on different surfaces using flow-through cell method with QCM after 2 h without achievement of equilibrium.

S/O

expt. Δf (Hz)

expected Δf (Hz)

apparent solvent−asphaltene ratio (g/g)

0 0.36 0.43 0.5

−638 −686 −1065 −1609

−511 −539 −582 −705

106 5 6.7 4.6

a Expected Δf = Δf (liquid loading due to toluene by using eq 2) + Δf (Stage 5).

When the probe is loaded with toluene only, the calculated frequency due to toluene (by using eq 2) is around 509.8 Hz. In addition, we have the experimental amount of deposited mass on the quartz crystal at each S/O in the air (stage 5). The expected Δf at the end of stage 4 is the sum of these contributions. For instance, at S/O = 0.5, the frequency due to deposited mass on the probe in stage 5 is −195.4 Hz; therefore, we expect that the Δf (stage 4) for S/O = 0.5 to be −705.2 Hz. The experimental Δf (stage 4) is much larger, at −1609 Hz, which is likely due to trapping of toluene in the asphaltene layer on the probe to give a larger effective mass. From the data of Table 4, the apparent ratio of solvent to asphaltene in the deposits was in the range 4.5−6.7 g/g. The estimate at S/O = 0 was unreliable due to the errors in estimating the masses. These ratios of solvent to asphaltene imply porosities of 0.85−0.9, which are consistent with an open nanoaggregate45 or a polymer brush structure.46 Changes in Dissipation (ΔD) and Resistance (ΔR). The change in dissipation factor, ΔD, indicates how much energy is dissipated in the layer adjacent to the surface of the probe. In our work, R-QCM was used with the ability to record the resistance. Using eqs 3 and 4, the dissipation and resistance parameters were calculated, and the results for some of the solvent ratios are given in Table 5.

Figure 5. Amount of asphaltene adsorption on hydrophilic silica particles with FTIR4 and QCM methods at various S/O (deposition time = 2 h in both cases).

colorless (washing for 3 days). In the QCM method, we did not wash the probe coated with silica particles with toluene to avoid losing the silica particles from the surface. However, the trends of increasing amounts of adsorbed material on the particle surfaces before and after the onset of asphaltene precipitation are similar in both methods. Solvent Swelling Effect. The frequency behavior at stage 4 of the immersion sequence reveals some interesting features. The data of Figure 6 show the change in frequency for different

Table 5. Calculated Dissipation and Calculated and Measured Resistance Due to Oil−Toluene−Pentane Solution Loading S/O

ΔD (calcd) × 104

meas. ΔR (Ω)

calcd ΔR (Ω)

0 0.36 0.43 0.5

2.35 1.8 1.75 1.73

287 226 232 234

308 235 229 225

When liquid is coupled to the surface of the gold crystals, this liquid loading behavior is manifested by a linear variation of ΔD versus Δf. Aung et al.47 found that ΔR and ΔD are correlated well with each other and both parameters give similar characterizations of the viscoelastic properties of the adsorbed material on the surface of the probe. If we assume that this analysis applies to the case for asphaltenes as well, we can use the resistance data to get information about the viscoelastic

Figure 6. Change in frequency with stages of immersion for S/O = 0.43.

stages of the QCM for S/O = 0.43 at the onset of asphaltene precipitation. The linear change in Δf with time indicates steady linear increase in the mass of the deposit during stage 2. In stage 4, part of the frequency change is due to the adsorption 1014


Energy & Fuels

Article

properties of the layers attached to the gold crystals. Using eqs 2 and 4, we can calculate the ratio of change in frequency to change in resistance:35 2 2 8e 26 f0 AM |Δf | = = 1.92 ΔR πμqhq

(9)

where e26 is the piezoelectric constant, AM is the electrode area, μq is the shear modulus of the quartz crystal, and hq is the thickness for 5 MHz resonance. The ratio of the changes in ΔR and Δf leads to a constant that is independent of the liquid characteristics and can be used as a verification for proper operation of the instrument. The data of Table 6 show |Δf |/ΔR for some values of S/O for both the liquid loading and the adsorption stages of each experiment.

Figure 7. Comparison of the amount of adsorbed mass of asphaltenes on gold versus (t − t0)1/2 for model oils with 0.05 and 1.0 g/L pitch in toluene. Time (t − t0)1/2 = 0 corresponded to the instant that the probe was immersed in the solution. The horizontal line shows the estimated monolayer mass.

Table 6. Ratio of Δf to ΔR for Different Stages of the QCM Experiments S/O

|Δf |/ΔR liquid loading (1 → 2)

|Δf |/ΔR during adsorption process (stage 3)

0 0.36 0.43 0.5

2.33 2.25 2.36 2.3

3.93 5.53 5.67 8.29

1 h is on the order of 7.4 nm for the model oil with 0.05 g/L pitch-in-toluene solution. For the asphaltene solution at 0.05 g/ L, the amount of monolayer deposition of asphaltene after 204 s is around 3.07 mg/m2. However, this amount kept increasing, and there was no leveling off in the frequency. Therefore, it is obvious that we have multilayer adsorption of asphaltenes on gold. The continuous adsorption on a poorly organized adsorbed layer could account for deviation of the adsorbed mass from linearity. Restructuring of the deposited material of the surface due to solvent swelling effects could be another reason for the incorrect diffusion prediction based on QCM data. Figure 8 shows comparisons of the amounts of adsorbed mass versus (t − t0)1/2 for various concentrations of the oil

The calculated ratios for liquid loading are close to the expected value of 1.92, with a mean value of 2.31 ± 0.05 that was not dependent on S/O. This constant relative response showed that the instrument was working properly. When asphaltenes are adsorbed on the crystal, however, the values of |Δf |/ΔR in stage 3 are significantly greater than 1.92, and increase systematically with S/O, which shows that the viscoelastic properties of the adsorbed material has increased. Previous work with QCM has shown that when the adsorbed mass on the surface of the probe is elastic, ΔR = 0.48 However, when the adsorbed materials form a viscoelastic layer on the surface, the change in ΔR becomes significant. For all the S/O, the change in resistance during the adsorption part (stage 3) was significant, which shows that the adsorbed materials on the surface of the gold for all S/O are not rigid and have viscoelastic behavior. This behavior is completely consistent with the changes in frequency in Table 4, which indicated a large fraction of solvent in the asphaltene layers deposited on the surface of the probe. Kinetics of Asphaltene Adsorption. Model oil solutions of 0.05, 0.1, 0.25, 0.5, 1, and 1.5 g/L pitch in toluene were prepared to measure the rate of deposition. The adsorption experiments were run for 1 h at each concentration. The adsorbed mass was plotted versus (t − t0)1/2 (where t0 is time of immersion of the probe in the solution), following eq 6. Figure 7 shows the adsorbed mass versus (t − t0)1/2 plots for 0.05 and 1 g/L pitch in toluene, which are nonlinear. From QCM data, we can determine the mass adsorbed on the gold surface versus time. If we consider the density of the adsorbed material to be around 1200 kg/m3, the height of the adsorbed material can be estimated versus time. If we assume that (a) there are limited adsorption sites on the surface of the gold, (b) the adsorption of asphaltenes on the gold surface is uniform, and (c) the size of asphaltene nanoaggregates is 3 nm, we can calculate the length of time required to obtain a monolayer adsorption of asphaltenes on the surface of the gold. The calculated thickness of the asphaltene aggregate layer after

Figure 8. Mass of asphaltenes deposited on gold during the first 20 s with different concentrations of pitch material in toluene.

solution for the first 20 s of each experiment. Each of these plots is quite linear, which is consistent with a diffusioncontrolled deposition rate with infinite adsorption kinetics at the surface. Nanoaggregates about 3 in nm diameter with density of 1200 kg/m3 would give a monolayer concentration of 3.2 mg/m2; therefore, at a concentration of 1.5 g/L, the deposits during this linear phase exceed monolayer coverage. These data show that the slopes of the adsorbed mass with (t − t0)1/2 increased with increasing concentration, consistent with the results of Abudu et al.23 Using eq 7 to calculate the apparent diffusion coefficient and eq 8 to determine the apparent diameter of the aggregates, in the data of Table 7, suggest rapid increase in mean diameter of asphaltenes with concentration in toluene. 1015


Energy & Fuels

Article

Potential Errors in Estimated Diffusion Coefficients for Asphaltene Aggregates. Asphaltene deposition on surfaces is a complex phenomenon and many factors should be considered when any equation is being used for calculating the diffusion coefficient of asphaltene aggregates. The derivation of the equation includes two assumptions: one regarding the kinetics of the adsorption step that governs the term ϕ(t) and the other that the mass of deposited material does not change. We assumed irreversible adsorption, so that material once attached remains on the surface and ϕ(t) = 0. The behavior during stage 4 of the experiment provides some support for this assumption; when the probe is immersed in toluene there is little loss of mass from the surface. The data of Figure 8 indicate masses at a significant fraction of the estimated monolayer coverage of ca. 3 mg/cm2, which could lead to errors in the assumption of irreversible adsorption. In addition, asphaltene films at interfaces are not static with time. Imaging of surface layers by atomic force microscopy shows evidence for reorganization of asphaltene surface layers over a period of hours,51 and similar results have been reported for interfacial films.52 If the flux of asphaltene aggregates toward the surface of the probe is accompanied by restructuring of the surface layer, solvent may be released with time and reduce the net rate of accumulation of mass. The result would be a significant underestimation of the diffusivity of asphaltene aggregates by eq 7 due to the reverse flux of solvent. Restructuring of the surface layer is unlikely to occur over short time periods, that is, 20 s, as in Figure 8, but it could be very significant in explaining the longer-time behavior (Figures 2 and 3). Modeling of this mechanism would require a change in the governing equations and would require data on the solvent associated with aggregates in the solution versus on the surface. Because such data are not available by current methods, we conclude that QCM data by the probe immersion method on kinetics of asphaltene deposition are subject to large and illdefined biases, rendering estimates of diffusivity and hydrodynamic diameter unreliable.

Table 7. Effect of the Concentration of Pitch in Toluene on the Apparent Diffusivity during the First 20 s of Deposition and Apparent Hydrodynamic Diameter of Adsorbed Particlesa concn pitch (g/L)

slope × 107 (kg m−2 s−1)

D × 1012 (m2/s)

viscosity (mPa s)

d (nm)

0.05 0.1 0.25 0.5 1 1.5

3.05 3.83 6.39 9.24 11.6 14.1

81.3 32.0 14.2 7.4 2.9 1.9

0.586 0.590 0.596 0.607 0.625 0.631

9.1 23.0 51.1 96.0 236 359

d (nm) from ref 22 34 78 154b

a

Diffusion coefficients (D) are estimated from eq 7 with asphaltene concentration of 60% of the pitch concentration, and the diameter (d) is from eq 8, using data from Figure 8. bExtrapolated from data at 0.15 and 0.2 g/L.

Abudu et al.23 and Xie et al.22 used QCM data and the same equations to calculate the hydrodynamic diameter of the asphaltene aggregates. The data of Xie et al.22 are included in Table 7 and are three times larger than the values from the present study over the common range of concentration from 0.05 to 0.2 g/L. Abudu et al.23 gave sizes 1−2 orders of magnitude smaller; however, they did not use a necessary time correction in their calculations. For determination of the slope by eq 7, we used the difference between the cumulative recorded time and the time when the samples were immersed in the solution to calculate (t − t0)1/2, giving values on the ordinate axis from 0 to 10 s1/2, as illustrated in Figure 8. In contrast, Abudu et al.23 did not subtract the time before immersion, so their x-axis values included times before the samples were immersed, giving apparent slopes that were 1 order of magnitude smaller. For example, using their data at a concentration of 0.278 g/L and a corrected x-axis of (t − t0)1/2, the actual apparent hydrodynamic diameter was 1220 nm, not the reported 0.8 nm. Dechaine and Gray45 reported the size of the asphaltene aggregates to be 5−9 nm in 1 g/L asphaltenes in toluene. Small-angle neutron scattering (SANS) and X-ray scattering (SAXS) gave estimates of the asphaltene aggregate size in the range 3−11 nm49 and 3.3−25.2 nm,50 respectively. Therefore, the size of asphaltene aggregates can be assumed to be around 5−25 nm in toluene solution at concentrations ca. 1 g/L, and almost certainly, a distribution of values in this range represents the size of asphaltenes rather than a single value. The calculated size of the asphaltene species in our work at higher concentrations, data from Xie et al.,22 and results from the corrected data of Abudu et al.23 greatly exceeds the estimates from scattering and diffusion. While our estimate from QCM at low concentration is comparable to other methods, the rapid increase in d with concentration in a strong solvent (toluene) is inconsistent. We would expect more molecular species and fewer nanoaggregates at very low concentrations, but the formation of very large aggregates over 300 nm in diameter at only 1 g/L asphaltenes in toluene is completely inconsistent with the appearance under light microscopy4 and measurements by other methods.2,45 Consequently, we conclude that the apparent diameters in Table 7 are invalid estimates, and we need to analyze the factors that could account for unreasonably small apparent diffusion coefficients from eq 7. Some of the possible reasons for significant deviation of the apparent size of the asphaltene aggregates are presented below.

■

CONCLUSIONS A quartz crystal microbalance (QCM) was used to measure asphaltene adsorption on gold surfaces and silica particles having different hydrophobicities. The adsorption of asphaltene at different S/O showed multilayer adsorption, without reaching a steady state after 16 h. The amount of adsorption on gold and silica increased monotonically with concentration below the onset of precipitation, then showed a rapid increase when the asphaltene solution became unstable. Adsorption was more pronounced on the gold surface than on the silica particles. For S/O = 0.43, for instance, adsorption on the gold surface was 160, 18.79, and 11.37 mg/m2 for hydrophilic and hydrophobic silica particles, respectively. We studied the swelling effect of the adsorbed layer on the surface, which showed aggregate with porosity in the range 0.85−0.9, indicating open nanoaggregate or polymer brush structure. The adsorbed materials at S/O > 0 showed viscoelastic characteristics. The predicted size of the aggregates obtained by calculating the diffusion coefficient using QCM was in the range 18−472 nm for 0.05−1 g/L asphaltene in toluene solutions. The calculated diffusion coefficient was in agreement with previous QCM estimates. However, estimates of the diffusion coefficient of asphaltenes in toluene by other methods, such as NMR, and microscopic observations of the size of the aggregates are 1016


Energy & Fuels

Article

(20) Hannisdal, A.; Ese, M.; Hemmingsen, P. V. Particle-Stabilized Emulsions: Effect of Heavy Crude Oil Components Preadsorbed onto Stabilizing Solids. Colloids Surf., A 2006, 276 (1−3), 45−58. (21) Ekholm, P.; Blomberg, E.; Claesson, P. A Quartz Crystal Microbalance Study of the Adsorption of Asphaltenes and Resins onto a Hydrophilic Surface. J. Colloid Interface Sci. 2002, 247 (2), 342−350. (22) Xie, K.; Karan, K. Kinetics and Thermodynamics of Asphaltene Adsorption on Metal Surfaces: A Preliminary Study. Energy Fuels 2005, 19 (4), 1252−1260. (23) Abudu, A.; Goual, L. Adsorption of Crude Oil on Surfaces Using Quartz Crystal Microbalance with Dissipation (QCM-D) under Flow Conditions. Energy Fuels 2009, 23 (3), 1237−1248. (24) Abdallah, W. A.; Taylor, S. D. Study of Asphaltenes Adsorption on Metallic Surface using XPS and TOF-SIMS. J. Phys. Chem. C 2008, 112, 48. (25) Gaboriau, H.; Saada, A. Influence of Heavy Organic Pollutants of Anthropic Origin on PAH Retention by Kaolinite. Chemosphere 2001, 44 (7), 1633−1639. (26) Pernyeszi, T.; Dekany, I. Sorption and Elution of Asphaltenes from Porous Silica Surfaces. Colloids Surf., A 2001, 194 (1−3), 25−39. (27) Drummond, C.; Israelachvili, J. Fundamental Studies of Crude Oil-Surface Water Interactions and its Relationship to Reservoir Wettability. J. Pet. Sci. Eng. 2004, 45 (1−2), 61−81. (28) Acevedo, S.; Ranaudo, M. A.; Garcia, C. Importance of Asphaltene Aggregation in Solution in Determining the Adsorption of this Sample on Mineral Surfaces. Colloids Surf., A 2000, 166 (1−3), 145−152. (29) Gonzalez, G.; Middea, A. Peptization of Asphaltene by Various Oil Soluble Amphiphiles. Colloids Surf. 1991, 52 (3−4), 207−217. (30) Dudasova, D.; Simon, S.; Hemmingsen, P. V.; Sjöblom, J. Study of Asphaltenes Adsorption Onto Different Minerals and Clays. Part 1. Experimental Adsorption with UV Depletion Detection. Colloids Surf. 2008, 317 (1), 1−9. (31) Pernyeszi, T.; Patzko, A.; Berkesi, O. Asphaltene Adsorption on Clays and Crude Oil Reservoir Rocks. Colloids Surf. 1998, 137 (1−3), 373−384. (32) Martin, S. J.; Granstaff, V. E.; Frye, G. C. Characterization of a Quartz Crystal Microbalance with Simultaneous Mass and Liquid Loading. Anal. Chem. 1991, 63 (20), 2272−2281. (33) Reed, C. E.; Kanazawa, K. K.; Kaufman, J. H. Physical Description of a Viscoelastically Loaded AT-Cut Quartz Resonator. J. Appl. Phys. 1990, 68 (5), 1993−2001. (34) Kanazawa, K. K.; Gordon, J. G. The Oscillation Frequency of a Quartz Resonator in Contact with a Liquid. Anal. Chim. Acta 1985, 175, 99−105. (35) Cho, N.; D’Amour, J. N.; Stalgren, J. Quartz Resonator Signatures Under Newtonian Liquid Loading for Initial Instrument Check. J. Colloid Interface Sci. 2007, 315 (1), 248−254. (36) Arnau, A., Piezoelectric Transducers and Applications. SpringerVerlag: New York, 2008. (37) Martin, B. A.; Hager, H. E. Velocity Profile on Quartz Crystals Oscillating in Liquids. J. Appl. Phys. 1989, 65 (7), 2630−2635. (38) Martin, B. A.; Hager, H. E. Flow Profile Above a Quartz Crystal Vibrating in Liquid. J. Appl. Phys. 1989, 65 (7), 2627−2629. (39) Muramatsu, H.; Tamiya, E.; Karube, I. Computation of Equivalent-Circuit Parameters of Quartz Crystals in Contact with Liquids and Study of Liquid Properties. Anal. Chem. 1988, 60 (19), 2142−2146. (40) Viitala, T.; Hautala, J. T.; Vuorinen, J. Structure of Anionic Phospholipid Coatings on Silica by Dissipative Quartz Crystal Microbalance. Langmuir 2007, 23 (2), 609−618. (41) Konash, P. L.; Bastiaans, G. J. Piezoelectric-Crystals as Detectors in Liquid Chromatography. Anal. Chem. 1980, 52 (12), 1929−1931. (42) Ward, A. F. H.; Tordai, L. Time-Dependent of Boundary Tensions of Solutions. I. the Role of Diffusion in Time-Effects. J. Chem. Phys. 1946, 14 (7), 453−461. (43) Rahmani, N. H. G.; Masliyah, J. H.; Dabros, T. Characterization of Asphaltenes Aggregation and Fragmentation in a Shear Field. AIChE J. 2003, 49 (7), 1645−1655.

inconsistent with QCM results. Therefore, the diffusion coefficient for asphaltene solutions measured by QCM by the probe immersion method is not considered reliable.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected].

■

REFERENCES

(1) Rahmani, N. H. G.; Dabros, T.; Masliyah, J. H. Settling Properties of Asphaltene Aggregates. Energy Fuels 2005, 19 (3), 1099−1108. (2) Yarranton, H. W.; Masliyah, J. H. Molar Mass Distribution and Solubility Modeling of Asphaltenes. AIChE J. 1996, 42 (12), 3533− 3543. (3) Long, Y.; Dabros, T.; Hamza, H. Stability and Settling Characteristics of Solvent-Diluted Bitumen Emulsions. Fuel 2002, 81 (15), 1945−1952. (4) Zahabi, A.; Gray, M. R.; Czarnecki, J.; Dabros, T. Flocculation of Silica Particles from a Model Oil Solution: Effect of Adsorbed Asphaltenes. Energy Fuels 2010, 24 (6), 3616−3623. (5) Jamialahmadi, M.; Soltani, B.; Muller-Steinhagen, H. Measurement and Prediction of the Rate of Deposition of Flocculated Asphaltene Particles from Oil. Int. J. Heat Mass Transfer 2009, 52 (19− 20), 4624−34. (6) Hamadou, R.; Khodja, M.; Kartout, M. Permeability Reduction by Asphaltenes and Resins Deposition in Porous Media. Fuel 2008, 87 (10−11), 2178−2185. (7) Rudrake, A.; Karan, K.; Horton, J. H. A Combined QCM and XPS Investigation of Asphaltene Adsorption on Metal Surfaces. J. Colloid Interface Sci. 2009, 332 (1), 22−31. (8) Amin, J. S.; Ayatollahi, S.; Alamdari, A. Fractal Characteristics of an Asphaltene Deposited Heterogeneous Surface. Appl. Surf. Sci. 2009, 256 (1), 67−75. (9) Nabzar, L.; Aguilera, M. E. The Colloidal Approach. A Promising Route for Asphaltene Deposition Modelling. Oil Gas Sci. Technol. 2008, 63 (1), 21−35. (10) Ramirez-Jaramillo, E.; Lira-Galeana, C.; Manero, O. Modeling Asphaltene Deposition in Production Pipelines. Energy Fuels 2006, 20 (3), 1184−1196. (11) Wang, S.; Civan, F. Modeling Formation Damage by Asphaltene Deposition during Primary Oil Recovery. J. Energy Resour. Technol. 2005, 127 (4), 310−317. (12) Marczewski, A. W.; Szymula, M. Adsorption of Asphaltenes from Toluene on Mineral Surface. Appl. Surf. Sci. 2002, 196 (1), 301− 311. (13) Saraji, S.; Goual, L.; Piri, M. Adsorption of Asphaltenes in Porous Media Under Flow Conditions. Energy Fuels 2010, 24 (11), 6009−6017. (14) Nassar, N. N. Asphaltene Adsorption onto Alumina Nanoparticles: Kinetics and Thermodynamic Studies. Energy Fuels 2010, 24 (8), 4116−4122. (15) Marchal, C.; Abdessalem, E.; Tayakout-Fayolle, M. Asphaltene Diffusion and Adsorption in Modified NiMo Alumina Catalysts Followed by Ultraviolet (UV) Spectroscopy. Energy Fuels 2010, 24 (8), 4290−4300. (16) de Castellanos-Ramirez, I. C.; Ortiz-Tapia, A. Study of Monolayer to Multilayer Adsorption of Asphaltenes on Reservoir Rock Minerals. Colloids Surf., A 2009, 340 (1−3), 149−54. (17) Alboudwarej, H.; Pole, D.; Svrcek, W. Y., Adsorption of Asphaltenes on Metals. Ind. Eng. Chem. Res. 2005, 44(15). (18) Goual, L.; Horvath-Szabo, G.; Masliyah, J. H. Adsorption of Bituminous Components at Oil/Water Interfaces Investigated by Quartz Crystal Microbalance: Implications to the Stability of Water-inOil Emulsions. Langmuir 2005, 21 (18), 8278−8289. (19) Dudaova, D.; Silset, A.; Sjoblom, J. Quartz Crystal Microbalance Monitoring of Asphaltene Adsorption/Deposition. J. Dispersion Sci. Technol. 2008, 29 (1), 139−146. 1017


Energy & Fuels

Article

(44) Dudasova, D.; Silset, A.; Sjoblom, J. Quartz Crystal Microbalance Monitoring of Asphaltene Adsorption/Deposition. J. Dispersion Sci. Technol. 2008, 29 (1), 139−146. (45) Dechaine, G. P.; Gray, M. R. Membrane Diffusion Measurements Do Not Detect Exchange between Asphaltene Aggregates and Solution Phase. Energy Fuels 2011, 25 (2), 509−523. (46) Shengqun, W.; Jianjun, L.; Liyan, Z. Colloidal Interactions between Asphaltene Surfaces in Toluene. Energy Fuels 2009, 23 (2), 862−869. (47) Aung, K. M. M.; Ho, X.; Su, X. DNA Assembly on Streptavidin Modified Surface: A Study Using Quartz Crystal Microbalance with Dissipation Or Resistance Measurements. Sens. Actuators, B 2008, 131 (2), 371−378. (48) Zhou, L.; Chen, H.; Jiang, X. Modification of Montmorillonite Surfaces Using a Novel Class of Cationic Gemini Surfactants. J. Colloid Interface Sci. 2009, 332 (1), 16−21. (49) Gawrys, K. L.; Blankenship, G. A.; Kilpatrick, P. K. On the Distribution of Chemical Properties and Aggregation of Solubility Fractions in Asphaltenes. Energy Fuels 2006, 20 (2), 705−714. (50) Fenistein, D.; Barre, L. Experimental Measurement of the Mass Distribution of Petroleum Asphaltene Aggregates using Ultracentrifugation and Small-Angle X-Ray Scattering. Fuel 2001, 80 (2), 283−287. (51) Long, J.; Xu, Z.; Masliyah, J. H. Single Molecule Force Spectroscopy of Asphaltene Aggregates. Langmuir 2007, 23 (11), 6182−6190. (52) Li, X.; Elliott, J. A. W.; McCaffrey, W. C.; Yan, D.; Li, D.; Famulak, D. Dynamic Surface Tensions of Athabasca Bitumen Vacuum Residue. J. Colloid Interface Sci. 2005, 287, 640−646.

1018


Kinetics and Properties of Asphaltene Adsorption on Surfaces

Recommend Documents