ARTICLE pubs.acs.

org/EF

LiquidLiquid Phase Equilibria in Asphaltene + Polystyrene + Toluene Mixtures at 293 K

M. Khammar and John M. Shaw*
Chemical and Materials Engineering, University of Alberta, Edmonton, T6G2 V4, Canada ABSTRACT: The phase behavior of hydrocarbon mixtures where one of the constituents self-aggregates is a subject of signi cant industrial and academic interest. Here, a nonintrusive acoustic phased-array technique operated in pulse echo mode is used to investigate the phase behavior of asphaltenes, a well-known self-aggregating species, in mixtures with polystyrene and toluene at 293 K and atmospheric pressure. This mixture exhibits liquidliquid phase behavior where both liquids are opaque to visible light, are of uniform composition, and are stable over broad ranges of composition. One phase is asphaltene rich and the other phase is polystyrene rich. Varying the polystyrene mean molar mass had little impact on the liquid to liquidliquid phase boundaries. Liquidliquid critical points were identi ed and phase compositions were con rmed for a xed global composition using the UVvisible spectrophotometry and mass balance equations. This is the rst report of liquidliquid phase behavior for such mixtures. Depletion occulation is hypothesized to be the mechanism causing phase separation in this ternary mixture.

1. INTRODUCTION Asphaltenes are de ned as the fraction of crude oil insoluble in alkanes and soluble in benzene and toluene on the basis of ltration experiments (ASTM D4055). However, in toluene, asphaltenes aggregate to form sterically stabilized colloidal particles.15 The size of these particles has been the subject of several investigations. For example, Barre et al.6 measured the radius of gyration of asphaltenes colloidal particles in a 3 vol % toluene solution using small-angle X-ray scattering (SAXS). They found that the asphaltene aggregates fall in the size range 6.3 to 16 nm. In heavy oils (15 to 20 wt % asphaltenes7) nano ltration experiments showed that asphaltene nanoaggregates form large aggregates up to 50100 nm in both Athabasca bitumen and Maya crude oil.8 Espinat et al.9 measured the size of asphaltenes in toluene using small-angle X-ray scattering (SAXS), small-angle neutron scattering (SANS), and dynamic light scattering to investigate the e ect of temperature and pressure on the size of asphaltene aggregates in toluene. They found that the size of asphaltene aggregates decreased with temperature, while pressure did not appear to have a signi cant e ect on size. The interaction forces between asphaltene colloidal particles in toluene are dominated by repulsion. For example, Wang et al.10 measured steric repulsive forces between asphaltene-coated surfaces in toluene. If asphaltene-containing mixtures are treated as colloidal solutions, conventional separation methods include ultracentrifugation, variation of pressure, temperature, solvent evaporation, and addition of antisolvent.11,12 Alternatives include polymer addition.12 In a recent work, Lima et al.13 focused on the impact of adsorbing polymers, polycardanol and sulfonated polystyrene, on asphaltene solutions. Polycardanol polymers were added to asphaltene in toluene solutions (60 mg/L) and sulfonated polystyrene was added to asphaltene in toluene and acetone solution. Asphaltene + polymer solutions were left for 24 h and then centrifuged at 3000 rpm for 30 min. The e ect of polymer addition was estimated by measuring the concentration of
r 2011 American Chemical Society

asphaltenes remaining dispersed using UVvisible spectrophotometry. At low polymer concentrations (<0.7 w/vol% for polycardanol and <0.2 w/vol% for sulfonated polystyrene) both polymers acted as occulants, i.e., individual polymer molecules adsorb on more than one nanoaggregate simultaneously. The nanoaggregates connected in this way form ocs. At higher polymer concentrations, these polymers behave as dispersants. It was reported that polystyrene without polar groups did not modify asphaltene behavior in dilute solution. This experimental result is contrary to the anticipated impact of a nonadsorbing polymer (polystyrene) in a good solvent (toluene)14,15 on concentrated sterically stabilized colloidal particles (asphaltenes)10 and may re ect the dilute composition range investigated. Siegla et al.16 found that when linear polystyrene is added to a cross-linked polystyrene swollen particlea microgel particle sterically stabilized in toluene, the mixture separates into two stable liquid phases. Clarke et al.17 obtained a similar result for a mixture of polystyrene microgel particle + linear polystyrene + ethylbenzene. Phase separation was attributed to entropic e ects. The mechanism driving phase separation for these mixtures, called depletion occulation, was rst described by Asakura and Oosawa.18,19 The microgel polystyrene particles are assumed to be su ciently cross-linked to prevent penetration by linear polymer chains. Con gurations for which the polymer penetrates the microgel particle are excluded and a polymer-free depletion layer forms around the dispersed particles. When two depletion layers overlap, the osmotic pressure around the particles becomes unbalanced giving rise to a net attractive force. For a su ciently high polymer concentration, mixtures can separate into colloid-poor and colloid-rich stable phases of uniform composition. The colloid-rich phase can be either liquid-like or solid-like. The addition of nonadsorbing polymers can therefore
Received: October 10, 2011 Revised: November 16, 2011 Published: November 16, 2011
1075
dx.doi.org/10.1021/ef201545q | Energy Fuels 2012, 26, 10751088

Energy & Fuels

ARTICLE

Figure 1. Experimental variation of mixture composition in the phase diagram.

Figure 3. Speed of sound pro les in toluene + polystyrene mixtures at 293 K and 1.01 bar. Polymer volume fraction is a parameter. (a) Mw = 393 400 g/mol, (b) Mw = 700 000 g/mol.

Figure 2. Speed of sound pro les in asphaltene + toluene mixtures at 293 K and 1.01 bar. Asphaltene volume fraction is a parameter.

be used to control the attractive forces between the colloidal particles and to manipulate their phase behavior. The phase behavior of mixtures of colloids and nonadsorbing polymers has been the subject of a substantial number of experimental and theoretical investigations.20 For example, Ramkrishnan et al.21 measured the phase behavior for silica particles, with a radius (a) of 50 nm, coated with 1-octadecanol + polystyrene + toluene ternary mixtures. They observed uidgel, uidcrystal, and uid uid transitions as the ratio of polymer radius of gyration (Rg) to colloidal particle radius (Rg/a) increased from 0.026 to 1.395. Hennequin et al.22 measured the phase behavior of silica particles sterically stabilized with 1-octadecanol in toluene + polystyrene for Rg/a = 4.1 and 5.2. They found that, after nearly 1 h of homogenization and over a range of compositions, the

mixtures separated into two stable liquid phases: one colloid-rich and one colloid-poor. The method developed by Bodnar et al.23 was then used to estimate the composition of the phases in equilibrium. In this work, the behavior of Maya pentane asphaltene + toluene + polystyrene mixtures is investigated at asphaltene concentrations greater than 5.0 vol %, at 293 K. At these concentrations, asphaltenes are expected to form sterically stabilized colloidal particles, and liquidliquid phase behavior is anticipated, as long as the asphaltene particles fall in an appropriate size range relative to the polystyrene molecules for the depletion occulation phase separation mechanism to occur.

2. EXPERIMENTAL METHODOLOGY
2.1. Chemicals and Solution Preparation. Toluene 99% was purchased from Fisher Scientific. Pentane asphaltenes were prepared from Maya crude oil according to ASTM standard D4055.24,25 Asphaltene in toluene mixtures were prepared by adding toluene to vials containing asphaltenes. They were mixed with a vortex mixer and hand shaken for at least 45 min until they
1076
dx.doi.org/10.1021/ef201545q |Energy Fuels 2012, 26, 10751088

Energy & Fuels

ARTICLE

Figure 5. Phase behavior observations for asphaltene + toluene + polystyrene mixtures on trajectories p, q and r, v, ae, and additional mixtures for polystyrene molar masses (a) Mw = 393 400 g/mol, (b) Mw = 700 000 g/mol. Liquidliquid phase behavior is denoted by crosses and liquid-phase behavior is denoted by triangles.

Figure 4. Experimental speed of sound data at 293 K and 1.01 bar for the binary mixtures: (a) asphaltene () + toluene, (b) polystyrene () (Mw = 393 400 g/mol) + toluene, (c) polystyrene () (Mw = 700 000 g/mol) + toluene.

appeared homogeneous and liquid like. Polystyrene with two different average molecular weights (Mw): Mw = 393 400 g/mol (Rg 25.6 nm) and Mw = 700 000 g/mol (Rg 36.1 nm)26 were purchased from Aldrich. The volume fractions of asphaltenes and polystyrene were calculated assuming an asphaltene density of 1.17 g/cm34 and using the density of polystyrene specified by the supplier, 1.047 g/cm3. 2.2. Acoustic Apparatus and Measurement Methodology. A detailed description of the acoustic apparatus and the measurement methodology is available elsewhere.27,28 Key points are summarized here. Phase boundaries are detected on the basis of two independent measurements, namely speed of sound differences between phases, and spikes in acoustic wave attenuation,

that arise at liquidliquid interfaces. Uniformity of compositions within phases is evaluated from speed of sound profiles. Composition gradients are readily detected as are composition differences between phases. Both sets of measurements are obtained from reflected waveforms measured with a phased array acoustic probe attached to the walls of a PolyBenzImidazole cell. Measurements were performed at 113 elevations (300 m apart) simultaneously. Acoustic wave attenuation is less accurate than speed of sound for interface detection but plays an important supporting role in cases where the volume of a phase is too small to obtain a speed of sound measurement within the phase. The temperature of the cell interior was controlled to within (0.1 K by circulating thermostatted fluid through tubes in the cell. Asphaltene + polystyrene + toluene mixtures were prepared in the cell by introducing a xed mass of an asphaltene + toluene + polystyrene mixture to which prepared mixtures of polystyrene in toluene and toluene were added using syringes. This permitted roughly orthogonal movements in the phase diagrams as shown in Figure 1. The contents were stirred for 5 min before the stirrer was removed. The acoustic measurements were started after removing the stirrer to capture the acoustic properties just after mixing at time (t) equals zero min, and were then obtained every 5 min for at least the rst hour after homogenization. The rate of the measurements was then decreased. 2.3. Speed of Sound in Toluene + Asphaltenes and Toluene + Polystyrene Binary Mixtures. Figure 2 shows the
1077
dx.doi.org/10.1021/ef201545q |Energy Fuels 2012, 26, 10751088

Energy & Fuels Table 1. Experimental Data for Mixtures (ae) Asphaltene + Polystyrene (Mw = 393 400 g/mol) + Toluene
global composition speed of sound (m/s) interface elevation vol. fraction time for phase (mm) Hinterface of upper phase R separation (min)

ARTICLE

mix a mix b mix c mix d mix e 0.217 0.142 0.119 0.107 0.102

0.0058 0.0305 0.0383 0.0419 0.0438

asphaltene rich phase polymer-rich phase di erence 1352.7 1348.5 1347.1 1346.2 1344.7 1352.7 1345.2 1343.8 1343.0 1341.8 3.3 3.3 3.2 2.9

nal time phase (min) behavior 1129 LL LL LL LL 0000LL

11.0 9.2 8.6 8.0

0.55 0.68 0.73 0.76

38 24 33 40

1413 156 427 1614

Table 2. Experimental Data for a Mixture of Asphaltene + Polystyrene (Mw = 393 400 g/mol) + Toluene (Composition, Speed of Sound Per Phase, LiquidLiquid Interface Elevation, and Volume Fraction of the Upper Phase)
global composition asphaltene-rich 0.069 0.098 0.093 0.089 0.085 0.078 0.077 0.073 0.070 0.196 0.185 0.167 0.153 0.146 0.128 0.118 0.109 0.148 0.142 0.134 0.127 0.118 0.113 0.108 0.098 0.168 0.139 0.118 0.172 0.154 0.142 0.0544 0.0450 0.0429 0.0410 0.0392 0.0361 0.0354 0.0335 0.0323 0.0123 0.0115 0.0105 0.0095 0.0091 0.0080 0.0074 0.0068 0.0284 0.0272 0.0258 0.0245 0.0227 0.0217 0.0208 0.0188 0.0204 0.0169 0.0142 0.0254 0.0340 0.0398 1349.8 1344.7 1341.4 1352.7 1353.1 1353.8 1353.6 1351.3 1348.6 1346.8 1344.9 1342.3 1341.2 1339.7 1349.7 1347.9 1346.1 1345.0 1343.9 1342.5 1346.1 1345.0 1343.5 1342.7 1342.0 1341.2 1341.4 1339.9 1346.9 1342.9 1340.9 1348.6 1349.0 1349.6 2.9 1.8 0.5 4.1 4.1 4.2 14.6 17.0 18.8 17.9 17.3 16.4 0.33 0.35 0.39 0.36 0.45 0.52 41 36 37 62 75 89 3.6 2.9 2.6 2.3 1.9 1.3 11.6 12.2 12.2 12.2 12.0 10.5 0.43 0.43 0.46 0.49 0.53 0.61 39 41 39 37 32 49 1346.0 1344.7 1343.4 1342.6 1342.0 1342.6 phase speed of sound (m/s) polymer-rich phase 1342.4 1342.7 1341.6 1340.6 1339.7 1339.0 1338.7 1338.0 1338.5 1349.4 1348.4 1345.9 1344.6 1342.8 di erencemix 1342.3 3.31345.0 3.11343.5 2.81342.3 2.91341.3 3.01340.7 3.91339.7 1339.0 1338.6 4.21354.2 2.91351.5 2.71348.8 2.2 2.1 elevation interface (mm) Hint <1.2 4.1 3.8 3.2 2.6 2.0 1.7 <1.2 <1.2 14.3 15.2 16.1 17.0 17.6 vol. frac. of upper phase R >0.94 0.83 0.85 0.88 0.91 0.93 0.95 >0.96 >0.96 0.25 0.25 0.28 0.30 0.31 time for phase separation (min) 46 37 32 35 36 41 37 47 48 50 43 41 41 46 nal time (min) 904 374 136 129 126 222 144 1082 1182 306 144 120 187 127 688 1334 1705 785 133 170 203 126 125 1211 1643 677 108 1400 2030 1104 8918 phase behavior LL LL LL LL LL LL LL LL LL LL LL LL LL LL L L L LL LL LL LL LL LL L L LL LL LL LL LL LL

variation of the speed of sound with elevation for five asphaltene + toluene mixtures. Each array of measurements was obtained at least 1 h after introducing the mixtures into the cell. For 23.4 vol % asphaltene in toluene, the measurement was performed 17 h

later. Speed of sound is independent of elevation. Thus at each composition, the asphaltene + toluene mixtures are stable and macroscopically homogeneous. Results for polystyrene + toluene mixtures are comparable, as shown in Figure 3a and b
1078
dx.doi.org/10.1021/ef201545q |Energy Fuels 2012, 26, 10751088

Energy & Fuels

ARTICLE

Table 3. Experimental Data for a Mixture of Asphaltene + Polystyrene (Mw = 700 000 g/mol) + Toluene (Composition, Speed of Sound Per Phase, LiquidLiquid Interface Elevation, and Volume Fraction of the Upper Phase)
speed of global composition asphaltene 0.075 0.095 0.108 0.104 0.098 0.094 0.090 0.087 0.089 0.081 0.199 0.188 0.181 0.171 0.163 0.154 0.148 0.141 0.135 0.127 0.119 0.156 0.156 0.152 0.135 0.130 0.121 0.112 0.107 0.104 0.099 0.094 0.0528 0.0462 0.0419 0.0405 0.0384 0.0368 0.0351 0.0341 0.0323 0.0315 0.0115 0.0109 0.0105 0.0099 0.0094 0.0089 0.0086 0.0081 0.0078 0.0074 0.0069 0.0258 0.0258 0.0250 0.0222 0.0214 0.0200 0.0185 0.0176 0.0171 0.0163 0.0155 rich sound (m/s) polymer rich 1339.0 1339.4 1340.7 1340.6 1338.8 1337.9 1337.1 1336.7 1336.7 1338.0 1350.3 1348.2 1347.0 1345.9 1344.7 1343.8 1343.1 1342.5 1341.7 di erence elevation (mm) Hint vol. frac. of upper phase R 0 0.90 0.83 0.85 0.88 0.89 0.92 0.94 0.96 >0.96 0.25 0.26 0.26 0.28 0.29 0.32 0.32 0.34 0.34 time for phase separation (min) nal time (min) 1440 394 169 688 182 404 133 137 604 1571 124 309 127 128 128 129 127 123 845 3053 2695 1343.6 1344.0 1342.1 1340.0 1339.3 1339.1 1338.2 1337.6 1337.7 1337.4 1336.6 3.2 3.2 2.9 2.4 2.2 1.9 1.6 1.0 0.7 11.3 11.3 11.6 12.2 12.0 11.7 10.4 8.3 5.3 0.46 0.45 0.46 0.49 0.52 0.56 0.64 0.73 0.83 50 50 50 40 27 24 19 33 95 425 942 204 334 167 740 158 127 1069 1331 727 phase behavior L LL LL LL LL LL LL LL LL LL LL LL LL LL LL LL LL LL LL L L LL LL LL LL LL LL LL LL LL L L

1342.0 1343.2 1343.4 1341.1 1339.9 1338.6 1338.6 1339.8 1353.7 1351.3 1349.7 1348.3 1346.8 1345.7 1344.8 1344.0 1342.7 1340.0 1338.3 1346.8 1347.2 1345.0 1342.4 1341.5 1341.0 1339.8 1338.6 1338.4

2.6 2.5 2.8 2.3 2.0 1.5 1.9 3.1 3.4 3.1 2.7 2.4 2.1 1.9 1.7 1.5 1.0

2.3 4.4 4.1 3.5 3.2 2.6 2.0 1.4 <1.2 15.2 15.8 16.4 17.0 17.6 17.9 18.3 18.8 19.5

59 50 45 49 45 44 43 41 53 61 60 50 50 52 53 55 59 72

Table 4. Phase Volume and Speed of Sound Measurement Repeatability for Mixture d
global composition speed of sound (m/s) asphaltene-rich trial 1 2 3 4 0.107 0.042 phase 1346.2 1346.3 1346.9 1345.6 polymer-rich phase 1343.0 1343.1 1343.5 1342.5 di erence 3.2 3.2 3.4 3.1 elevation interface (mm) Hint 8.6 8.3 8.3 8.3 vol. fract. of upper phase R 0.73 0.74 0.74 0.74 time for phase separation (min) 33 35 34 30 nal time (min) 427 142 685 345

for Mw = 393 400 g/mol and Mw = 700 000 g/mol polystyrene, respectively. Measurements were performed following 2224 h of equilibration. The small gradient in the speed of sound values with elevation, evident for mixtures of 7.7 vol% polystyrene in toluene, may reflect residual thermal or composi-

tion gradients. Averaged speed of sound values as functions of composition for the binaries are presented in Figure 4. Both polymers and the asphaltenes have comparable impacts on the speed of sound of mixtures vis--vis toluene at the a same volume fraction. Thus, detailed phase compositions
1079
dx.doi.org/10.1021/ef201545q |Energy Fuels 2012, 26, 10751088

Energy & Fuels

ARTICLE

Figure 6. Speed of sound pro les (i), acoustic wave attenuation spectra di erence (ii), and attenuation di erence at 7.9 MHz (iii), at the equilibration times given in Table 1, and for global asphaltene + toluene + polystyrene (Mw = 393 400 g/mol) compositions along trajectory ae shown in Figure 5a.

for asphaltene + polystyrene + toluene ternary mixtures are not readily discriminated on the basis of speed of sound, measurements themselves, or models fit to binary data, a technique limitation also noted elsewhere.29 2.4. UV-visible Spectrometry Measurements. A Variant Cary 50 Scan UVvisible spectrophotometer was used to measure the concentration of asphaltenes during only one liquid

liquid equilibrium experiment, as these measurements are destructive. The specified instrument tolerance for absorbance measurement (Abs) is (0.005 Abs at 1 Abs. Samples were introduced in a 10-mm-path quartz cell. Absorbance spectra for toluene, polystyrene (7.8 vol%) + toluene, and apparent absorbance for nine asphaltene + toluene mixtures (ranging from 50 to 1100 mg/L) were used to calibrate composition
1080
dx.doi.org/10.1021/ef201545q |Energy Fuels 2012, 26, 10751088

Energy & Fuels

ARTICLE

Figure 7. (i) Evolution of the liquidliquid, liquidair interfaces, (ii) speed of sound: (triangle) upper phase, (circle) lower phase, and (iii) speed of sound di erence between the phases with time for global compositions (be) shown in Figure 5a: (I) composition (b), (II) composition (c), (III) composition (d), (IV) composition (e).

measurements. Neither toluene nor polystyrene + toluene mixtures absorb significantly in the wavelength range 500800 nm. Their absorbance is 0.03 Abs. Maya asphaltenes also do not contain chromophores that sorb significantly in the range 500900 nm.30 Apparent absorbance in this range is attributed solely to scattering by asphaltene particles. The apparent aborbance of asphaltenes is linearly proportional to composition at low concentrations and an apparent absorbance of 1.7 Abs was attained at 1090 mg/L for scans at 700 nm.

3. RESULTS AND DISCUSSION The phase behavior of asphaltene + toluene + polystyrene mixtures was explored by adding mixtures of toluene + polystyrene to mixtures of asphaltenes + toluene + polystyrene (trajectories ae, and w) and by adding toluene to mixtures of asphaltenes + toluene + polystyrene (trajectories p, q, r, v). These observations are summarized in Figure 5. The bulk of the observations fell in the liquidliquid region of the phase diagram. Global compositions based on the volume fractions
1081
dx.doi.org/10.1021/ef201545q |Energy Fuels 2012, 26, 10751088

Energy & Fuels

ARTICLE

Figure 8. Liquidliquid interface elevation (a), volume fraction of the upper phase (b), and speed of sound di erence between the coexisting phases (c) for global compositions be shown in Figure 5a.

of asphaltenes () and polystyrene (), speeds of sound in each phase, liquidliquid interface elevation (Hint), upper-phase volume fraction (R), and equilibration times for Maya asphaltenes + toluene + polystyrene (Mw = 393 400 and 700 000 g/mol) are presented in Tables 1, 2, and 3. Measurement repeatability is illustrated in Table 4. These data are explored in detail below. 3.1. Phase Behavior of Polystyrene (MW = 393 400 g/mol) + Toluene + Asphaltenes. Aliquots of polystyrene, = 0.077, in toluene were added to an asphaltene, = 0.234, in toluene mixture to construct the trajectory that follows the line segment ae in Figure 5a. Experimental data for these mixtures at equilibrium are summarized in Table 1. Speeds of sound and attenuation profile measurements are shown in Figure 6. Mixture a remained homogeneous and there was no evidence of phase separation even after 19 h. As more of the polystyrene + toluene

mixture was added, an upper liquid phase and a lower liquid phase each with uniform compositions, separated by an interface, appeared. From the sequence of experiments, the upper liquid phase is richer in polymer and the lower liquid phase is richer in asphaltenes. The liquidliquid interface elevation and the air liquid interface elevation were detected most accurately as breaks in the speed of sound measurements at equilibrium; Figure 6i, ae. Sound wave attenuation difference maps Figure 6ii, ae and sound wave attenuation difference values at a specific frequency (7.9 MHz) Figure 6iii, ae provide secondary and less precise measures of interface location. Attenuation differences are computed by subtracting the attenuation just after mixing from attenuations arising at another time, in this case, the equilibration time. These mixtures reach equilibrium within 40 min and no further change in interface elevation, speed of sound in each phase, or speed of sound di erence arises as shown in Figure 7 for points be along trajectory ae. The elevation of the liquid liquid interface, and the speed of sound di erence between the separated phases are stable with time within 0.3 mm and 0.4 m/s, respectively. These variations are comparable to the repeatability of individual phase separation measurements. For example, four sequential equilibration trials were performed with mixture d. The results are summarized in Table 4. For each trial, the mixtures were stirred for 5 min. The phases separated or reseparated within 35 min. Liquidliquid interface elevation was repeatable to within 0.3 mm. Consequently, the volume fraction of the upper phase, R, was repeatable to within 0.01. The speed of sound di erence values between phases at equilibrium were repeatable to within 0.3 m/s even though the speed of sound drifted by as much as 1.5 m/s over the time frame of the experiments. Based on the speed of sound of toluene, such a drift can be accounted for with a temperature drift in the cell of 0.3 K. The liquidliquid separation is time invariant and repeatable. The variation of the elevation of the liquidliquid interface, Hinterface, the volume fraction of the upper phase, R, and the speed of sound di erence between the phases in mixtures be at equilibrium are shown in Figure 8ac. As the polystyrene volume fraction increases along the trajectory ae the liquidliquid interface elevation decreases (Figure 8a), and the volume fraction, R, of the upper-phase increases (Figure 8 b), while the speed of sound di erence between the phases is invariant within experimental error (Figure 8c). 3.1.1. Identification of the LiquidLiquid Critical Behavior for Asphaltene + Polystyrene (Mw = 393 400 g/mol) + Toluene Mixtures. Phase behavior measurements were made along trajectories p, q, and r as shown in Figure 5a. These measurements were performed by adding small aliquots of toluene to three mixtures of asphaltene + polystyrene + toluene with different volume fraction ratios of polystyrene to asphaltenes. If sufficient toluene is added, the trajectories intersect the liquidliquid to single-liquid phase boundary. In addition to providing points defining the boundary, the location of the liquidliquid critical point can be inferred from the phase volume and speed of sound difference data and their trends with composition shown in Figure 9. As a liquidliquid critical point is approached, the compositions of the asphaltene-rich and the polymer-rich phases approach one another. Consequently, the upper-phase volume fraction approaches 0.5 and the speed of sound difference approaches 0. If the global composition is marginally richer in one component, the phase volumes diverge from 0.5 as the
1082
dx.doi.org/10.1021/ef201545q |Energy Fuels 2012, 26, 10751088

Energy & Fuels

ARTICLE

Figure 9. Elevation of the liquidliquid interface (i), volume fraction of the upper phase (ii), and speed of sound di erence between the coexisting phases (iii) for mixtures of asphaltenes + polystyrene (Mw =393 400 g/mol) + toluene. (I) trajectory p, (II) trajectory q, (III) trajectory r.

Figure 10. Phase separation results for a mixture with global composition ( = 14.72 vol%, = 2.88 vol %): (a) speed of sound pro le, (b) acoustic wave attenuation spectrum di erence, and (c) ultrasound attenuation di erence at 7.9 MHz.

boundary is approached. The gradient and direction of the divergence provide directional and proximity information related to the location of the critical point. For trajectory p, the liquid liquid interface elevation decreases from 4.1 to 1.7 mm, as the upperphase volume fraction increases from 0.83 to 0.95. As the phase boundary approached the base of the cell (Hinterface < 1.2 mm),

acoustic wave attenuation was used to detect the presence of the interface because the speed of sound values in the asphaltene-rich phase could not be measured accurately. Trajectory p approaches the two-phase to single-phase boundary with the polymerrich phase near a point ( 0.07, 0.03) remote from the liquidliquid critical point. For trajectory q, the volume fraction
1083
dx.doi.org/10.1021/ef201545q |Energy Fuels 2012, 26, 10751088

Energy & Fuels

ARTICLE

Figure 11. Elevation of the liquidliquid interface (i), volume fraction of the upper phase (ii), speed of sound di erence between the coexisting phases (iii) for mixtures of asphaltenes + polystyrene (Mw = 700 000 g/mol) + toluene. (I) Trajectory p, (II) trajectory q, (III) trajectory r.

of the upper phase increases from 0.43 to 0.61 and the speed of sound difference between the separated phases decreases from 3.6 to 1.3 m/s. Trajectory q crosses the phase boundary into the polymer-rich phase near a point ( 0.110, 0.021). For trajectory r, the upper-phase volume fraction increases from 0.25 to 0.31 and the speed of sound difference decreases from 4.2 to 2.1 m/s. Trajectory r crosses the phase boundary into the asphaltene-rich phase near a point ( 0.137, 0.0086). Thus the liquidliquid critical point falls between the intersection of trajectories q and r with the phase boundary. 3.1.2. Phase Composition (Tie Line) Measurements in the Liquidliquid Region. A mixture of asphaltene + polystyrene (Mw = 393 400 g/mol) + toluene ( = 0.147, = 0.029) with a composition close to that of mixture 1 on trajectory q ( = 0.148, = 0.028) was prepared. Phase separation was completed within 52 min, and the mixture was equilibrated for 167 min. The equilibrium profiles for speed of sound and sound wave attenuation are shown in Figure 10. The volume fraction of the polymerrich phase (R) is 0.42. This value compares well with that measured for mixture 1 on trajectory q, where R = 0.43. Samples from both liquid phases were extracted from the cell and aliquots of 0.4 mL of each phase were diluted in 100 mL of toluene prior to UV analysis. The concentration of asphaltene in the diluted solutions extracted from the lower and upper phases were 938.7 and 346.4 mg/L, respectively. This corresponds to asphaltene volume fractions of I = 0.201 ( 0.006 and II = 0.074 ( 0.003 in the asphaltene-rich and polymer-rich phases, respectively. From

the asphaltene material balance equation, the global volume fraction of asphaltene can be calculated as R II 1 R I 0:147 ( 0:007 1

This value is in close agreement with the global volume fraction of asphaltenes in the mixture = 0.147 ( 0.006. The volume fraction of polystyrene was determined indirectly and by di erence following evaporation of toluene from samples placed in oven for 10 h at 90 C then in a vacuum oven at 80 C for 6 h. The mass of polystyrene is determined by subtracting the mass of asphaltenes, determined directly as noted above, from the residual mass. Consequently, the error is large at small volume fractions. The volume fraction of polystyrene in the asphaltene-rich and polymer-rich phases are I = 0.000 ( 0.007 and II = 0.060 ( 0.003, respectively. The global polystyrene volume fraction in the mixture can be calculated from the polystyrene material balance j RjII 1 RjI 0:025 ( 0:005 2

This value falls within the error of the global volume fraction of polystyrene (0.029 ( 0.003) in the mixture. 3.2. Phase Behavior of Polystyrene (MW = 700 000 g/mol) + Toluene + Asphaltenes Mixtures. The phase behavior results obtained with this mixture are both qualitatively and quantitatively similar to those obtained with the lower molar mass
1084
dx.doi.org/10.1021/ef201545q |Energy Fuels 2012, 26, 10751088

Energy & Fuels

ARTICLE

Figure 12. Phase diagrams for mixtures of asphaltenes + toluene + polystyrene: (a) polystyrene Mw = 393 400 g/mol, (b) polystyrene Mw = 700 000 g/mol. Two-phase region: (crosses), single-phase region: (triangles), phase compositions from UV measurements (small dots), estimated liquidliquid critical points (large dots).

polymer and are presented in summary form here. Phase behavior measurements were performed along trajectories p, q, and r as shown in Figure 5b and liquidliquid interface elevation, the volume fraction of the upper phase and the speed of sound difference values are summarized in Figure 11. For trajectory p, the position of the liquidliquid interface decreases from 4.4 to 1.4 mm, as the volume fraction of the polymer-rich upper phase increases from 0.83 to 0.96. As the liquidliquid interface approaches the base of the cell, acoustic wave attenuation was used to detect the interface. Trajectory p clearly approaches the two-phase to single-phase boundary with the polymer-rich phase near the point ( 0.08, 0.03) that is remote from the liquidliquid critical point. For trajectory q, the volume fraction of the upper phase diverges from 0.46 and rises to 0.83 as the liquidliquid to liquid boundary is approached while the speed of sound difference between the phases decreases from 3.2 to 0.7 m/s. Trajectory q crosses the phase boundary into the polymerrich phase near point ( 0.102, 0.0167). For trajectory r, the volume fraction of the upper phase increases from 0.25 to 0.34 and the speed of sound difference decreases from 3.4 to 1.0 m/s. Trajectory r crosses the phase boundary into the asphaltene-rich phase near point ( 0.131, 0.0076). Again, the liquidliquid critical point falls between the intersection of trajectories q and r. 3.3. Phase Diagrams for Polystyrene + Toluene + Asphaltene Mixtures. Phase diagrams for the two polystyrene + toluene + asphaltene ternaries are presented in Figure 12a

and b along with estimated critical points. The two-phase region occupies much of both phase diagrams. The polymerrich and asphaltene-rich phases are found close to their respective composition axes. As polystyrene does not possess polar groups that can associate with asphaltenes, asphaltene + polystyrene + toluene ternaries appear to be analogous to mixtures of sterically stabilized colloidal particles + polysty rene + toluene mixtures, where phase separation is driven by depletion flocculation. The phase diagrams in Figure 12a and b possess characteristics similar to phase diagrams for welldefined sterically stabilized colloidal particles + non-adsorbing polymer + solvent mixtures.22,23 Unlike these mixtures, though, asphaltene colloidal particles are not monodispersed and their mean size may also depend on global and phase compositions. 3.4. Phase Separation Kinetics. Phase separation times for asphaltene + toluene + polystyrene mixtures range from 30 to 50 min following mixing. These times are slow compared to typical hydrocarbon liquidliquid phase separation times. The evolution of the separation process for one mixture, asphaltene ( = 0.148) + polystyrene (Mw = 39 3 400 g/mol, = 0.028) + toluene, shown in Figure 13 is illustrative. This mixture corresponds to the first composition on trajectory q in Figure 5a and it is remote from the liquidliquid critical point. The horizontal dashed lines in Figure 13 denote the liquidliquid and liquidair interfaces arising at equilibrium. At t = 0 min, the mixture is macroscopically homogeneous. The speed of sound profiles and the attenuation profile are featureless. With increasing time, the speed of sound decreases in the upper part of the cell and increases in the lower part of the cell. Composition gradients in the liquid phase become evident within 14 min (Figure 13c). High attenuation regions, related to the formation of a liquidliquid interface, arise at 14 min and move downward from the upper part of the cell and upward from the lower part of the cell (Figure 13c, ii). They coalsce and form the liquidliquid interface at 35 min (Figures 13cg). No further change is observed in the profiles after 39 min (Figure 13h). This phase separation time is consistent with the values reported in Tables 14 in this work and with prior work on the time reported for completion of phase separation for mixtures of colloidal particles + non-adsorbing polymers + solvents. For example, Hennequin et al.22 reported that after homogenizing mixtures of silica colloidal particles + polystyrene + toluene, a fuzzy interface first formed at the bottom of the cuvette, moved upward, and became sharper as phase separation proceeded. All samples investigated reached equilibrium 1 h after homogenization. They reported that the separated phases were clearly distinguishable and the interface was sharp. Ramakrishnan et al.21 reported that for mixtures of colloidal silica particles + polystyrene + toluene, a meniscus separating two fluid phases appeared within minutes of mixing. The lower phase was richer in colloidal particles than the upper phase and both phases flowed easily. Aarts et al.31 reported that for mixtures of fluorescent PMMA nanoparticles + polystyrene in cis/trans decalin, phase separation was complete within 15 min at intermediate polymer concentrations and took up to few hours at high polymer concentrations. In the future we plan to evaluate options for normalizing the dynamic acoustic profile data to extract interfacial zone thickness and interfacial composition.32
1085
dx.doi.org/10.1021/ef201545q |Energy Fuels 2012, 26, 10751088

Energy & Fuels

ARTICLE

Figure 13. Continued

1086

dx.doi.org/10.1021/ef201545q |Energy Fuels 2012, 26, 10751088

Energy & Fuels

ARTICLE

Figure 13. Phase separation kinetics measurements for a mixture of asphaltene ( = 0.148) + polystyrene (Mw = 393 400 g/mol, = 0.028) + toluene. (i) Speed of sound pro le, (ii) acoustic wave attenuation spectra di erence, and (iii) attenuation di erence at 7.9 MHz. Time elapsed after mixing, in minutes, is a parameter: (a) 0, (b) 4, (c) 14, (d) 19, (e) 24, (f) 29, (g) 35, (h) 39, (k) 784 min.

4. CONCLUSIONS Repreatable and reversible liquidliquid phase behavior was observed for asphaltene + polystyrene + toluene mixtures over a broad range of compositions at 293 K and atmospheric pressure. From speed of sound and UV spectroscopy measurements, one of the phases was found to be asphaltene-rich and the other was polystyrenerich. Liquidliquid critical points were also identi ed along the liquid liquid to liquid-phase boundary. Remote from the critical point, the asphaltene-saturated polymer-rich phase could comprise a volume fraction of asphaltenes as large as 0.06. By contrast, the polymer volume fraction in the asphaltene-rich phase remote from the critical point was less than 0.01. The observed phase behavior conforms with that of sterically stabilized colloidal particles (asphaltenes) + a nonadsorbing polymer (polystyrene) + a good solvent (toluene). Depletion occulation is the phase separation mechanism associated with such mixtures. This is the rst report of liquidliquid phase behavior arising in asphaltene + polystyrene + toluene mixtures and the results are expected to open new lines of inquiry related to property prediction, property measurement, and process development linked to asphaltene behavior in solvents and in live oils.
AUTHOR INFORMATION
Corresponding Author

and setup of the Focus LT acoustic equipment, Professors Murray Gray and Gregory Dechaine from the University of Alberta for suggesting and helping with the UV spectophotometry measurements, and Dr. Carlos Lira-Galena at the Mexican Petroleum Institute for the Maya crude oil sample. This research was supported by the Alberta Innovates Energy and Environment Solutions, ConocoPhillips Inc., Imperial Oil Resources, Halliburton, Kellogg Brown and Root, NEXEN Inc., Shell Canada, Total, the Virtual Materials Group, and the Natural Science and Engineering Research Council of Canada (NSERC).

REFERENCES
(1) Andreatta, G.; Bostrom, N.; Mullins, O. High-Q ultrasonic determination of the critical nanoaggregate concentration of asphaltenes and the critical micelle concentration of standard surfactants. Langmuir 2005, 21, 27282736. (2) Andreatta, G.; Goncalves, C.; Bu n, G.; Bostrom, N.; Quintella, C.; Arteaga-Larios, F.; Perez, E.; Mullins, O. Nanoaggregates and structure-function relations in asphaltenes. Energy Fuels 2005, 19, 12821289. (3) Zeng, H.; Song, Y.-Q.; Johnson, D. L.; Mullins, O. C. Critical Nanoaggregate Concentration of Asphaltenes by Direct-Current (DC) Electrical Conductivity. Energy Fuels 2009, 23, 12011208. (4) Mostow , F.; Indo, K.; Mullins, O. C.; McFarlane, R. Asphaltene Nanoaggregates Studied by Centrifugation. Energy Fuels 2009, 23, 11941200. (5) Lisitza, N. V.; Freed, D. E.; Sen, P. N.; Song, Y.-Q. Study of Asphaltene Nanoaggregation by Nuclear Magnetic Resonance (NMR). Energy Fuels 2009, 23, 11891193. (6) Barre, L.; Simon, S.; Palermo, T. Solution properties of asphaltenes. Langmuir 2008, 24, 37093717.
1087
dx.doi.org/10.1021/ef201545q |Energy Fuels 2012, 26, 10751088

*Tel.: +1-780-492-8236. E-mail: jmshaw@ualberta.ca.

ACKNOWLEDGMENT We aknowledge and thank Dr. Brent Zeller from Eclipse Scienti c for assistance and support related to the con guration

Energy & Fuels

(7) Zhao, B.; Becerra, M.; Shaw, J. M. On Asphaltene and Resin Association in Athabasca Bitumen and Maya Crude Oil. Energy Fuels 2009, 23, 44314437. (8) Zhao, B.; Shaw, J. M. Composition and size distribution of coherent nanostructures in Athabasca bitumen and Maya crude oil. Energy Fuels 2007, 21, 27952804. (9) Espinat, D.; Fenistein, D.; Barre, L.; Frot, D.; Briolant, Y. E ects of temperature and pressure on asphaltenes agglomeration in toluene. A light, X-ray, and neutron scattering investigation. Energy Fuels 2004, 18, 12431249. (10) Wang, S.; Liu, J.; Zhang, L.; Xu, Z.; Masliyah, J. Colloidal Interactions between Asphaltene Surfaces in Toluene. Energy Fuels 2009, 23, 862869. (11) Branco, V.; Mansoori, G.; Xavier, L.; Park, S.; Mana , H. Asphaltene occulation and collapse from petroleum uids. J. Pet. Sci. Eng. 2001, 32, 217230. (12) Myakonkaya, O.; Eastoe, J. Low energy methods of phase separation in colloidal dispersions and microemulsions. Adv. Colloid Interface Sci. 2009, 149, 3946. (13) Lima, A. F.; Mansur, C. R. E.; Lucas, E. F.; Gonzalez, G. Polycardanol or Sulfonated Polystyrene as Flocculants for Asphaltene Dispersions. Energy Fuels 2010, 24, 23692375. (14) Noda, I.; Higo, Y.; Ueno, N.; Fujimoto, T. Semidilute Region for Linear-Polymers in Good Solvents. Macromolecules 1984, 17, 10551059. (15) Luckham, P. Measurement of the Interaction Between Adsorbed Polymer Layers - The Steric E ect. Adv. Colloid Interface Sci. 1991, 34, 191215. (16) Siegla , C. Phase Separation in Mixed Polymer Solutions. J. Polym. Sci. 1959, 41, 319326. (17) Clarke, J.; Vincent, B. Stability of non-aqueous microgel dispersions in the presence of free polymer. J. Chem. Soc., Faraday Trans. I 1981, 77, 18311843. (18) Asakura, S.; Oosawa, F. On Interaction between 2 Bodies Immersed in a Solution of Macromolecules. J. Chem. Phys. 1954, 22, 12551256. (19) Asakura, S.; Oosawa, F. Interaction Between Particles Suspended in Solutions of Macromolecules. J. Polym. Sci. 1958, 33, 183192. (20) Tuinier, R.; Rieger, J.; de Kruif, C. G. Depletion-induced phase separation in colloid-polymer mixtures. Adv. Colloid Interface Sci. 2003, 103, 131. (21) Ramakrishnan, S.; Fuchs, M.; Schweizer, K. S.; Zukoski, C. F. Entropy driven phase transitions in colloid-polymer suspensions: Tests of depletion theories. J. Chem. Phys. 2002, 116, 22012212. (22) Hennequin, Y.; Evens, M.; Angulo, C.; van Duijneveldt, J. Miscibility of small colloidal spheres with large polymers in good solvent. J. Chem. Phys. 2005, 123, 054906 (doi: 10.1063/1.1953548). (23) Bodnar, I.; Oosterbaan, W. Indirect determination of the composition of the coexisting phases in a demixed colloid polymer mixture. J. Chem. Phys. 1997, 106, 77777780. (24) Buenrostro-Gonzalez, E.; Lira-Galeana, C.; Gil-Villegas, A.; Wu, J. Asphaltene precipitation in crude oils: Theory and experiments. AIChE J. 2004, 50, 25522570. (25) Fulem, M.; Becerr, M.; Hasan, M. D. A.; Zhao, B.; Shaw, J. M. Phase behaviour of Maya crude oil based on calorimetry and rheometry. Fluid Phase Equilib. 2008, 272, 3241. (26) Fetters, L.; Hadjichristidis, N.; Lindner, J.; Mays, J. Molecular Weight Dependence of Hydrodynamic and Thermodynamic Properties for Well-De ned Linear Polymer in Solution. J. Phys. Chem. Ref. Data 1994, 23, 619640. (27) Khammar, M. The Phase Behavior of Asphaltene + Polystyrene + Toluene Mixtures at 293 K. PhD Thesis, University of Alberta, 2011 (28) Khammar, M.; Shaw, J. M. Phase Behaviour and Phase Separation Kinetics Measurement Using Acoustic Arrays. Rev. Sci. Instrum. 2011, 82, 104902 (doi: 10.1063/1.3650767 (9 pages)). (29) Beliajeva, O.; Grebenkov, A.; Zajatz, T.; Timofejev, B. Speed of sound in the liquid phase of the R134a/152a refrigerant blend. Int. J. Thermophys. 1999, 20, 16891697.

ARTICLE

(30) Dechaine, G. P.; Gray, M. R. Membrane Di usion Measurements Do Not Detect Exchange between Asphaltene Aggregates and Solution Phase. Energy Fuels 2011, 25, 509523. (31) Aarts, D.; Lekkerkerker, H. Confocal scanning laser microscopy on uid- uid demixing colloid-polymer mixtures. J. Phys.: Condens. Matter 2004, 16, S4231S4242. (32) Zhang, X.; Chodakowski, M.; Shaw, J. Dynamic interfacial zone and local phase concentration measurements in emulsions, dispersions, and slurries. J. Dispersion Sci. Technol. 2004, 25, 277285.

1088

dx.doi.org/10.1021/ef201545q |Energy Fuels 2012, 26, 10751088