Colloids and Surfaces A: Physicochem. Eng.

Aspects 220 (2003) 9
/27

www.elsevier.com/locate/colsurfa

Effects of petroleum resins on asphaltene aggregation and

water-in-oil emulsion formation
P. Matthew Spiecker a, Keith L. Gawrys b, Chad B. Trail c,
Peter K. Kilpatrick b,*
a
ExxonMobil Upstream Research Company, Houston, TX 77027, USA
b
Department of Chemical Engineering, North Carolina State University, Raleigh, NC 27695, USA
c
University of Kentucky, Lexington, KY 40506, USA

Received 22 July 2002; accepted 30 December 2002

Abstract

Asphaltenes from four crude oils were fractionated by precipitation in mixtures of heptane and toluene. Solubility
profiles generated in the presence of resins (1:1 mass ratio) indicated the onset of asphaltene precipitation occurred at
lower toluene volume fractions (0.1
/0.2) than without resins. Small-angle neutron scattering (SANS) was performed on
solutions of asphaltene fractions in mixtures of heptane and toluene with added resins to determine aggregate sizes.
Water-in-oil emulsions of asphaltene
/resin solutions were prepared and separated by a centrifuge method to determine
the vol.% water resolved. In general, the addition of resins to asphaltenes reduced the aggregate size by disrupting the
p
/p and polar bonding interactions between asphaltene monomers. Interaction of resins with asphaltenic aggregates
rendered the aggregates less interfacially active and thus reduced emulsion stability. The smallest aggregate sizes
observed and the weakest emulsion stability at high resin to asphaltene (R/A) ratios presumably corresponded to
asphaltenic monomers or small oligomers strongly interacting with resin molecules. It was often observed that, in the
absence of resins, the more polar or higher molecular weight asphaltenes were insoluble in solutions of heptane and
toluene. The addition of resins dissolved these insolubles and aggregate size by SANS increased until the solubility limit
was reached. This corresponded approximately to the point of maximum emulsion stability. Asphaltene chemistry plays
a vital role in dictating emulsion stability. The most polar species typically required significantly higher resin
concentrations to disrupt asphaltene interactions and completely destabilize emulsions. Aggregation and film formation
are likely driven by polar heteroatom interactions, such as hydrogen bonding, which allow asphaltenes to absorb,
consolidate, and form cohesive films at the oil
/water interface.
# 2003 Elsevier Science B.V. All rights reserved.

Keywords: Asphaltenes; Resins; Emulsion stability; Small-angle neutron scattering

1. Introduction
* Corresponding author. Tel.: /1-919-515-7121; fax: /1-
919-515-3465. Emulsion challenges during petroleum recovery
E-mail address: [email protected] (P.K. Kilpatrick). have been attributed to colloidal aggregation of
0927-7757/03/$ - see front matter # 2003 Elsevier Science B.V. All rights reserved.
doi:10.1016/S0927-7757(03)00079-7
10 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27

asphaltenes and waxes [1
/6]. Many early studies cise shape difficult to discriminate. With the
were performed on the film forming and emulsify- archipelago-like structure of Fig. 1, it seems
ing behavior of crude oil
/water systems [1,7
/15]. probable that asphaltenic aggregates possess a
These and later studies often pointed to the porous reticulated microstructure.
asphaltenic constituents in crude oil as being While the effects of temperature, solvent aro-
responsible for film formation and stabilization maticity, and polarity have received much atten-
[10,11,16,17]. tion, the solvation of asphaltene aggregates by
Asphaltenes, or n -heptane insolubles and tolu- resins has not been fully explored. What we mean
ene solubles, are the most refractory compounds here by the expression ‘solvation’ is the strong
present in crude oil and are generally distinguished local interaction of asphaltenic aggregates by resin
by a fused aromatic core with polar heteroatom molecules, a phenomenon referred to in previous
functionality [18,19]. Studies indicate the presence studies by the curious term ‘peptized’; we do not
of carboxylic acids, carbonyls, phenols, pyrroles mean by solvation the ‘swelling’ of asphaltenic
and pyridinic functional groups capable of accept- aggregates, although as we show in a subsequent
ing or donating protons [20
/24]. Vapor pressure publication, resins definitely fill in solvent voids in
osmometry measurements suggest number average asphaltenic aggregates created by the reticulated
molecular weights range from approximately 800 structure of asphaltenes. Espinat et al. measured
to 3000 Da [25
/32]. Even lower asphaltene neutron and X-ray scattering intensities from
molecular weights are observed in hot, polar asphaltenic aggregates in several solvents, at low
solvents suggesting that aggregates formed and high temperatures and with added resins [37].
through stacking interactions between asphaltene The scattering curves were fit using a thin disc
monomers. The most plausible mechanisms of form factor model. In toluene, 2 wt.% Boscan
asphaltene aggregation involve p
/p overlap be- asphaltene solutions formed larger aggregates at
tween aromatic sheets, hydrogen bonding between room temperature (234 Å diameter) than at 76 8C
functional groups and other charge transfer inter- (175 Å). At room temperature, aggregate sizes
actions. were two to four times smaller in high polarity
Small-angle neutron scattering (SANS) has been solvents such as pyridine and tetrahydrofuran than
applied to probe solvent and temperature effects in benzene. They also found that resin
/asphaltene
on asphaltene aggregation [33
/44]. Proper analy- ratios of 2:1 by mass reduced the scattering
sis of the scattering intensity curves can provide intensity at low Q suggesting the formation of
aggregate size, shape, molecular weight, and smaller aggregates. However, the effect of resin
fractal dimension. Asphaltenic aggregates are solvation on asphaltene aggregate size was not
comprised of cofacial stacks of planar, fused explicitly reported. Bardon et al. observed resin
aromatic ring moieties connected by aliphatic solvation of Safaniya asphaltenes using SANS and
chains and rings. Recent structural and molecular SAXS [47]. Weight average molecular weights of 2
modeling seem to confirm the so-called ‘archipe- wt.% asphaltene solutions in toluene with R/A
lago’ model of asphaltenes, as opposed to the more ratios of 2, 4, and 8:1 were reduced by factors of
widely implied and invoked ‘island’ model [45]. In 2.6, 4.6, and 7.5, respectively.
the ‘archipelago’ model, individual asphaltene Previously [44,49] we examined the aggregation
monomers are comprised of aromatic and fused and emulsion stabilizing behavior of asphaltenes
aromatic ring moieties, some with polar functional and their more and less soluble fractions in
groups, connected to each other by aliphatic mixtures of heptane and toluene (so-called ‘hep-
polymethylene chains and rings that likely contain tol’). Gravimetric solubility measurements indi-
some sulfide and carbonyl functional groups (see cated that asphaltenes start to precipitate at
Fig. 1). Asphaltenic aggregates have been modeled concentrations between 45 and 52% (v/v) toluene
as mono- and polydisperse spheres [35,46], flat in heptol. Precipitation concentrated the most
disks [37,47,48], and prolate cylinders [38]. As- aromatic and polar constituents as measured by
phaltene polydispersity, however, makes the pre- H/C ratio and nitrogen content. In addition,
 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27 11

Fig. 1. Schematic illustration of archipelago model of asphaltene monomers, asphaltenic aggregate in absence of resins, and
asphaltenic aggregate in presence of resins.

SANS indicated that the less soluble fraction tion [51
/55]. Resins are the most polar and
formed the largest aggregates in heptol while the aromatic species present in deasphalted oil and,
more soluble fraction formed considerably smaller it has been suggested, contribute to the enhanced
aggregates. Asphaltenes and their various sub- solubility of asphaltenes in crude oil by solvating
fractions are well known to stabilize water-in-oil the polar and aromatic portions of the asphaltenic
emulsions near the point of incipient flocculation molecules and aggregates [25,56,57]. The solubility
[50]. The degree of aggregation and proximity to of asphaltenes in crude oil is mediated largely by
the solubility limit governs the stability of emul- resin solvation and thus resins play a critical role
sions prepared in heptol. in precipitation, and emulsion stabilization phe-
In this study, we have taken another step nomena [37,40,58
/60]. Resins, although quite
towards understanding the mechanisms of asphal- surface-active, have not been found to stabilize
tene aggregation and emulsion formation in pet- significantly water-in-oil emulsions by themselves
roleum and petroleum-derived systems through in model systems, a fact that mitigates somewhat
the addition of solvating molecules. Crude oils are against the notion that resins and asphaltenes form
usually characterized by SARA fractionation a simple continuum of molecular structures and
where asphaltenes are removed by precipitation functions [61,62]. However, the presence of resins
with a paraffinic solvent and the deasphalted oil in solution can destabilize emulsions via asphal-
(DAO or maltenes) is separated into saturates, tene solvation and/or replacement at the oil
/water
aromatics and resins by chromatographic separa- interface [62
/66].
12 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27

Here we probe the effects of resins on asphaltene dissolving in toluene and inducing partial precipi-
aggregation and emulsion formation. Aggregation tation through heptane addition. Enough heptane
is likely controlled by the ability of asphaltene was added during fractionation to generate ap-
monomers to interact through aromatic and polar proximately 33% (w/w) insoluble asphaltenes from
forces. By fractionating asphaltenes into solubility a 0.75% (w/v) asphaltene solution in toluene. All
classes, the most polar and aromatic species can be solvents were HPLC grade and obtained from
concentrated. The effectiveness of solvating resins Fisher Scientific. Details of the precipitation and
on these more and less soluble asphaltenes aids in fractionation procedures can be found in another
elucidating the mechanisms of colloid formation publication [44]. A summary of the elemental
and emulsion stabilization in petroleum derived composition of the asphaltene fractions is pro-
fluids. vided in Table 2.

2.2. SARA fractionation

2. Experimental
Petroleum resins were isolated via the SARA
2.1. Asphaltene precipitation and fractionation technique where DAO is charged to silica gel and
extracted with solvents of increasing polarity [51
/
Asphaltenes were precipitated from four crude 53,67]. After a two stage filtration to ensure
oils in a 40:1 excess of n-heptane. The crude oils complete removal of the asphaltenes, the hep-
were obtained from several locations around the tane-diluted crude oil was rotary evaporated until
world: B6 and Hondo (off-shore California), Arab dry. The DAO was dissolved in methylene chloride
Heavy (Safaniya), and Canadon Seco (Argentina). (Fisher-HPLC grade) and adsorbed to activated
For brevity the following abbreviations will be silica gel (Chromatographic silica gel, 35
/60 mesh,
used to describe the asphaltenes generated from Fisher). Silica gel activation proceeded under
Hondo, Arab Heavy, and Canadon Seco crude vacuum at 120 8C for 48 h. The silica gel
/DAO
oils, respectively: HO, AH, and CS. These crude slurry was shaken for 24 h then rotary evaporated
oils are asphaltene rich and vary in viscosity, resin until dry and placed in a nitrogen flushed vacuum
content and asphaltene H/C ratio. Basic crude oil oven at 50 8C for 24 h.
and asphaltene properties can be found in Table 1. Chromatography columns (2 /100 cm with 250
Resin content was determined by sequential elu- ml solvent reservoir) were initially filled with a
tion chromatography (discussed in the next sec- mixture of 68:32 heptane
/toluene (v/v). Clean
tion), H/C ratios were calculated from combustion activated silica gel was added until the depth
elemental analysis (Perkin Elmer Series II reached
/20 cm. Finally, silica gel with adsorbed
CHNSO), and viscosity measurements were per- DAO was transferred to the column until full. A
formed on a Rheometrics Dynamic Stress Rhe- solvent mixture containing 68% (v/v) heptane and
ometer with concentric cylinder geometry. 32% (v/v) toluene eluted saturates, mono-, di-, and
Asphaltenes precipitated from the crude oils were triaromatics from the silica gel. Once the saturates
separated into more and less soluble fractions by and aromatics were extracted, a more polar

Table 1
Crude oil properties

Crude wt.% Asph R/A ratio H/C Asph Viscosity (cP) 1008F

AH 6.7 1.12 1.14 33.8

B6 13.1 0.92 1.24 2030
CS 7.5 1.19 1.11 70
HO 14.8 1.39 1.29 363
 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27 13

Table 2
Asphaltene fraction composition in wt.%, except H/C

Asphaltene H/C Nitrogen Sulfur

Sol Whole Ppt Sol Whole Ppt Sol Whole Ppt

AH 1.17 1.14 1.13 0.92 1.02 1.08 8.06 8.32 7.66

B6 1.30 1.24 1.22 1.81 1.87 1.93 7.25 6.68 6.33
CS 1.12 1.11 1.09 1.32 1.32 1.39 0.52 0.52 0.48
HO 1.30 1.29 1.24 1.95 1.99 2.11 8.42 8.53 8.48

solvent (40:30:30 acetone:toluene:methylene chlo- at the same toluene volume fraction. To ensure all
ride) was applied to elute the resins. The resin- of the resins were removed, the precipitate was
solvent mixture was filtered to remove any silica rinsed with neat heptane prior to dissolution in
gel fines and rotary evaporated until dry. The methylene chloride. The % precipitated was deter-
resins were transferred to jars and placed in a mined from the mass ratio of precipitated asphal-
nitrogen flushed vacuum oven at 60 8C for 48 h or tenes to the original asphaltene mass.
until completely dry. Combustion elemental ana-
lyses of the resins appear in Table 3.
2.4. Small-angle neutron scattering

2.3. Asphaltene and resin solubility Neutron scattering of asphaltenic aggregates

solvated by resins was performed on the NG7
The solubility of asphaltenes and their sub- and NG1 small angle spectrometers at the NIST
fractions were determined in heptol with added Center for Neutron Research (Gaithersburg, MD)
resins. Resins from the crude oil were only added or on the Small Angle Neutron Diffractometer
to their complementary asphaltenes. Solubility (SAND) at Argonne National Laboratory (IPNS,
profiles of the asphaltenes without resins were Argonne, IL). Samples were measured in cylind-
obtained in another study [44] and will be used for rical quartz cells (NSG Precision Cells) with a path
comparison. Resin
/asphaltene solutions were pre- length of 5 mm (all NG7 samples) or 2 mm (all
pared in various mixtures of heptane and toluene. SAND samples).
The asphaltene concentration was 0.75% w/v (
/1 Mixtures of asphaltenes and resins were pre-
wt.%) in 15 mL solvent and the resin
/asphaltene pared at resin:asphaltene (R/A) ratios between
ratio was 1:1 by mass. Resins and asphaltenes were 0.25:1 and 10:1. Asphaltene solutions (1 wt.%)
dissolved together in toluene and allowed to shake were prepared in perdeuterated heptane and
for 12 h prior to heptane addition. After an toluene solutions (CDN Isotopes, Canada) and
additional 12 h, the solutions were vacuum filtered studied at 25 and 80 8C. Scattering intensity versus
through 1.5 mm Whatman 934 AH filter paper to scattering angle (I(Q) vs. Q) data were fit to
collect precipitates and rinsed with 7.5 ml of heptol Lorentzian line shapes using a non-linear least
squares regression to determine the aggregate
Table 3 correlation lengths. Following Ornstein
/Zernike
Resin composition in wt.%, except H/C (O by difference) formalism the scattering intensity, I, can be related
Resin source H/C N S O to the scattering vector, Q, by:

AH 1.31 0.81 6.49 1.53 I0

I(Q)  (1)
B6 1.51 1.48 6.91 1.98 1  (Qj)2
CS 1.39 1.52 0.88 2.77
HO 1.51 where j is the correlation length, I0 is the scatter-
ing intensity at Q /0 [42,68]. In some asphaltene
14 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27

systems, a Porod upturn was observed at low Q 2.5. Resin
/asphaltene emulsions
where the scattering intensity increased monoto-
nically with decreasing Q. Furthermore, incoher- Water-in-oil emulsions were prepared by homo-
ent scattering of all nuclei in the solvent and solute genizing water and model oil solutions containing
with non-zero spin was manifested in the scatter- asphaltenes and resins. Asphaltenes and resins
ing curves at large Q values (typically Q /0.1) as were dissolved together in toluene for approxi-
an isotropic background signal. The Q values that mately 12 h followed by heptane addition. A 4 mL
marked the transition from the Guinier regime to aliquot of this solution with an asphaltene con-
the Porod or incoherent scattering regimes were centration of 0.37% w/v (
/0.5% w/w) was homo-
determined from inflection points in the scattering genized with a 6 ml aliquot of deionized water. A
curves. The Lorentzian line shape described in Eq. Virtis Virtishear Cyclone I.Q. homogenizer with a
(1) was applied over the intermediate range of Q 6 mm rotor
/stator emulsion generator assembly
values between the inflection points. was lowered into the oil
/water system and run at
In the absence of a significant low Q upturn, 15 000 rpm for 3 min.
Guinier analysis was also performed to calculate After aging 24 h, the emulsions were centrifuged
the radius of gyration, Rg, which is defined as the for 1 h at 15 000 rpm. The stability of the
mean squared distance from the center of gravity emulsions was calculated from the volume of
of the scatterer. The Guinier approximation ap- water resolved:
plies to aggregates in dilute solution and is strictly
Volume resolved
valid in the low Q range where QRg is less than 1. % Water resolved 100 (4)
The Guinier approximation has the form: Initial volume
2 2 Complete details on solution preparation and
I(Q) I0 e(Q Rg =3) (2)
homogenization can be found elsewhere [49].
where

I0 Np Vp (Dr)2 (3) 3. Results and discussion

and Np is the number of scatterers, Vp is the
scatterer volume, and Dr2 is the coherent scatter- 3.1. Asphaltene
/resin solubility
ing contrast between the solvent and solute. Given
that elemental compositions of the asphaltenes The effect of adding resins (1:1 mass ratio) on
and resins are available, it is possible to calculate asphaltene solubility is shown in Fig. 2(a
/d). The
the scattering contrast terms and subsequently, the half filled markers represent solubility data deter-
weight-average molecular weight of the aggregates mined for asphaltenes in neat heptol while the
in the Guinier regime. These analyses will be filled markers represent asphaltene solubility with
addressed in a future paper. resins. As mentioned before, the solubility limits of
Comparing the analytical form of the Guinier the whole asphaltene fractions were approximately
approximation and Lorentzian lineshapes predicts 50% toluene. The more soluble fraction or ‘Solu-
that Rg should be proportional to j by a scale ble’ asphaltenes precipitated at toluene volume
factor of â3 (or
/1.73). A comparison of the Rg fractions between 0.3 and 0.4 while the less soluble
and j values obtained from experiments on or ‘Precipitate’ asphaltenes precipitated at consid-
asphaltene and resin mixtures in heptol provides erably higher aromaticity (0.6
/0.8). This indicates
the relation Rg /1.71j. This suggests that Guinier that the Soluble fraction cooperatively solvates the
analysis and Lorentzian fits are equally suitable Precipitate fraction in solution.
for extracting relevant aggregation behavior from The process of fractionation generated unique
scattering curves. Further details concerning the asphaltene classes distinguished by their solubility
SANS instruments, experimental conditions, and behavior. The Precipitate fractions were character-
data analysis methods are provided elsewhere [44]. ized by higher aromaticity, polarity, molecular
 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27 15

Fig. 2. Solubility of 0.75% (w/v) Whole asphaltenes and their more and less soluble subfractions in heptol with and without resins.
Half filled markers denote systems without resins. R/A ratio was fixed at 1:1 by mass. a) AH, b) B6, c) CS, d) HO.

weight, and aggregate size than the Whole or Precipitate fractions. These interactions suggest
Soluble asphaltene fractions. Their behavior in the that CS resins play a significant role in asphaltene
presence of resins should help elucidate the solvation in the crude oil.
molecular mechanisms of solvation and aggrega- Asphaltenes form aggregates in solution
tion. through intermolecular p-p and hydrogen bonds
The asphaltene solubility limit, after resin addi- between asphaltene monomers. Resins reduce the
tion, was reduced as much as 10% (v/v) toluene. In tendency for asphaltenes to aggregate by disrupt-
the precipitated regime, resins were capable of ing these intermolecular interactions. From Table
enhancing asphaltene solubility (reducing the per- 3 we see that resins contain polar heteroatoms
centage of precipitates) between 10 and 50%. within a mixed aromatic
/aliphatic carbon matrix.
Resins appear to enhance the solubility of the Similar to asphaltenes, resins are polydisperse and
more polar and aromatic Precipitate asphaltenes only average chemical properties can be measured.
more so than the Soluble asphaltenes. Soluble Polar functional groups give resins the capacity to
asphaltenes are less polar and aromatic and do not disrupt the electron donor
/acceptor interactions
respond as favorably to resin addition in highly partly responsible for asphaltene aggregation.
aliphatic solvents. Of note is the considerable Resins, however, are less aromatic than asphal-
solvating effect of CS resins on CS Whole and tenes as gauged by H/C ratios between 1.31 and
16 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27

1.51. Due to decreased aromaticity, their solubility linear least squares fit of the Lorentzian lineshape
in more strongly aliphatic solvents is considerably to the data. Based on the shape of the neutron
higher than asphaltenes. Aromatic moieties in scattering curve alone, one can readily distinguish
resins likely solvate the fused ring portion of the two length scales of aggregate sizes in typical
asphaltenes, producing a solvated, stabilized, re- asphaltene solutions. For example, the scattering
sin
/asphaltene aggregate. The ability of resins to curve for CS Whole asphaltenes with no added
dissociate intermolecular asphaltene bonds results resins in 60% toluene (Fig. 3(a)) is a superposition
in reduced aggregate sizes. of a Guinier plateau at intermediate Q and an
intense power law feature at low Q. The Guinier
plateau region indicated scattering from soluble,
3.2. SANS: asphaltenes and resins in heptol
non-flocculating aggregates on the order of ap-
proximately 20
/100 Å. The presence of a low Q
SANS allows us to probe the effects of resins on
feature indicated flocculation of a portion of the
asphaltene aggregation. In another SANS study,
we found that asphaltene Precipitate fractions soluble aggregates. The absence of a second
formed larger aggregates than the unfractionated plateau region in the lowest Q range suggests
or more soluble fractions [44]. Scattering curves that the largest flocs had a size greater than the
were fit with Lorentzian lineshapes to determine order of 1/Qmin (or ]/200 Å). The decrease in
the solute correlation lengths. Fig. 3 shows typical intensity of the low Q feature with increasing resin
I(Q) versus Q neutron scattering curves for CS content indicated that resins were effective at
Whole asphaltenes with added B6 resins in 60% dissolving the larger flocs into non-interacting
toluene at 25 8C. The solid lines represent the non- aggregates. A reduction in the correlation length

Fig. 3. SANS fits (2 mm path length) using Lorentzian lineshapes. Canadon Seco Whole asphaltenes at 258C in 60% toluene with: (a)
No Resins: I0: 11.79/0.4, j: 1019/3, R2: 0.9958. (b) 0.5 wt.% B6 Resins: I0: 6.99/0.1, j: 709/1, R2: 0.9962. (c) 2 wt.% B6 Resins: I0:
4.569/0.06, j: 48.59/0.8, R2: 0.9967. (d) 4 wt.% B6 Resins: I0: 3.809/0.03, j: 36.99/0.5, R2: 0.9970. (e) 6 wt.% B6 Resins: I0: 4.029/0.05,
j: 31.19/0.6, R2: 0.9940. (f) 10 wt.% B6 Resins: I0: 3.309/0.04, j: 22.59/0.4, R2: 0.9922. Note: Every 2nd data point plotted above Q/
0.006. Units of I. and j are cm 1 Å respectively.
 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27 17

from 101 Å (no resins) to 23 Å (10% resin) than B6 resins. Above a 4:1 R/A ratio, the
indicated that resins were also effective at solubi- correlation lengths obtained for the asphaltene
/
lizing the individual asphaltene aggregates. resin systems were within 10%, regardless of the
One trivial explanation for the decrease in resin type. This suggests that resins from different
SANS scattering intensity with increasing resin sources may be approximately equal in effective-
content is the notion that resins themselves form ness at solvating asphaltenic aggregates. Modest
small aggregates similar to asphaltenes. Upon differences in resin aromaticity and polarity (see
mixing the asphaltene aggregates with an increas- Table 3) are secondary to differences in asphaltene
ing number of smaller resin aggregates the average chemistry and solvent conditions for dictating
particle size is expected to decrease. While resin aggregate size in crude oil systems. The aggrega-
aggregation is certainly a plausible explanation for tion behavior of B6 Whole asphaltene
/resin solu-
reduced correlation lengths, asphaltene
/resin in- tions shown in Fig. 4 is typical of the other Whole
teraction in solution cannot be fully discounted. asphaltenes as well. In the absence of resins,
The solubility studies discussed previously have Whole asphaltene correlation lengths followed
shown that resins play a role in enhancing the trend: CS /B6 /HO /AH. Resin addition
asphaltene solubility in solution. We will also was effective at disrupting the intermolecular
show that resins are capable of modifying the bonding and aggregation of each asphaltene to a
surface-activity of asphaltenes, thus affecting their similar extent. The greatest decrease in aggregate
ability to stabilize emulsions. The resins used in size occurred between 0.5 and 2:1 R/A suggesting
this study were incapable of forming stable emul-
resins strongly solvate asphaltenes at ratios close
sions at any concentration or solvent condition.
to those found in crude oil (Table 1). As the R/A
Results from both the solubility and emulsion
ratio approached 10:1 the Whole asphaltene
studies suggest that some interactions with resins
correlation lengths neared a common value of
modify the asphaltene aggregates to some extent.
approximately 11
/14 Å, suggesting this may be
Bardon et al. compared the scattering curve of an
close to a solvated monomer or irreducible oligo-
asphaltene
/resin mixture to the sum of the scat-
mer.
tering intensities from pure asphaltenes and pure
In less aromatic solvents, asphaltenes form
resins [47]. Since the sum of the individual
larger aggregates due to solvent
/solute incompat-
scatterers was larger than the scattering by the
mixture, they concluded that asphaltenes were ibility. Heptane
/toluene mixtures of increasing
solvated by resins. Furthermore, they assumed aliphaticity solvent possess a lower degree of p-
the pure asphaltene scattering intensity was the bond solvating capability and polarity than a pure
difference between the mixed asphaltene
/resin and toluene solvent. As a result, asphaltenes without
pure resin (same concentration) scattering pat- resins in 60% (v/v) toluene have larger correlation
terns. The correlation lengths reported in this lengths than in pure toluene. For example, B6
study were calculated using scattering curves Whole asphaltenic aggregates have correlation
from the mixed asphaltene
/resin solutions, assum- lengths of 79 Å at a toluene volume fraction of
ing that the scattering from resin-only aggregates 0.6 as compared with 43 Å in pure toluene.
in negligible. However, at a R/A ratio of 10:1 the B6 Whole
The effect of resins from B6 and AH crude oils correlation length appears to plateau at similar
on asphaltene aggregation is shown in Fig. 4. B6 values (
/14
/18 Å) regardless of the solvent
Whole asphaltenes in pure toluene and 60% (v/v) aromaticity. This indicates that high resin concen-
toluene in heptol were combined with AH and B6 trations are apparently more effective at reducing
resins at 808C. As shown in the figure, correlation aggregate size than solvent alone. This comes as no
lengths of B6 Whole asphaltenes solvated by resins surprise since resins are more chemically similar to
in pure toluene were identical with either AH or asphaltenes and have been linked to asphaltene
B6 resins. In the mixed solvent, AH resins may solubility in a variety of systems including crude
solubilize B6 asphaltenes slightly more effectively oil.
18 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27

Fig. 4. Aggregate j of 1 wt.% B6 Whole asphaltenic aggregates with AH and B6 Resins in pure toluene and 60% toluene at 80 8C
determined from SANS (5 mm path length). Legend displayed as (resin type, % toluene).

3.3. Asphaltene
/resin emulsion stability: effect of Conversely, B6 Whole aggregates in 60%
varied R/A ratio toluene were sufficiently surface-active (due to
high polarity), even at R/A ratios approaching 5:1,
Asphaltene emulsions were prepared in the to adsorb at oil
/water interfaces and form
presence of resins at R/A ratios from 0.5 to 10 emulsion-stabilizing films with 72% water
and their stabilities were gauged by measuring vol. resolved. Correlation lengths observed for B6
% water resolved after centrifugation. These Whole asphaltenes in 60% toluene at 80 8C
emulsion stabilities were compared with aggregate decreased with resin addition from 47 Å (0.5:1 R/A
correlation lengths determined by SANS (Figs. 5
/ ratio) to 18 Å (10:1 R/A ratio). Correlation lengths
7). As shown in Fig. 5, resin addition decreased observed for HO Whole asphaltenes in 60%
both the aggregate correlation length and stability toluene at 80 8C decreased slightly with resin
of emulsions formed by CS Whole asphaltenes in addition from 38 Å (0.5:1 R/A ratio) to 14 Å (10:1
60% toluene. This trend was also observed for B6 R/A ratio). Both B6 and HO Whole asphaltenes
and HO Whole asphaltenes, although CS Whole were high in polarity but HO Whole lacked the
asphaltenes formed much weaker emulsions than surface activity to maintain a cohesive oil
/water
both B6 and HO Whole at similar solvent condi- interfacial film at R/A ratios greater than 2:1 (92%
tions. CS asphaltenic aggregates were the most water resolved). However, emulsions formed by
aromatic and least polar of the Whole asphaltenes HO Whole asphaltenes were still more stable than
studied. The lack of polarity and possible inability those formed by CS Whole asphaltenes under
to form a network of hydrogen bonds likely similar conditions. Based on the emulsion stability
reduced interfacial film strength. Resin addition results of the Whole asphaltenes, it is apparent
up to R/A ratios of 2:1 effectively solvated the that higher resin concentrations are needed to
aggregates and further reduced their emulsion destabilize emulsions when asphaltenes are more
stabilizing ability. polar.
 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27 19

Fig. 5. Emulsion stability (% water resolved) and aggregate j of CS Whole asphaltenes in 60% toluene with added resins. R/A
represents mass ratio of resins to asphaltenes. Emulsions tested using CS resins, 0.5 wt.% asphaltenes. SANS j measured at 25 8C using
B6 resins, 1 wt.% asphaltenes, 2 mm path length.

B6 Precipitate emulsions prepared in pure trated the most aromatic and polar asphaltenes
toluene remained very stable in systems containing in the Precipitate fraction and, as a result, they
up to 2:1 resins (Fig. 6). This was due primarily to tended to aggregate, adsorb and consolidate into
the high proportion of film forming species in B6 elastic films at oil
/water interfaces. As the R/A
Precipitate. The fractionation process concen- ratio approached 2:1, the film forming portion of

Fig. 6. Emulsion stability (% water resolved) and aggregate j of B6 Precipitate asphaltenes in toluene with B6 resins. Emulsions tested
at 0.5 wt.% and j determined by SANS at 25 8C, 1 wt.% asphaltenes, and 2 mm path length.
20 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27

Fig. 7. Emulsion stability (% water resolved) and aggregate j of B6 Precipitate asphaltenes in 30% toluene with B6 resins. Emulsions
tested at 0.5 wt.% and j determined by SANS at 25 8C, 1 wt.% asphaltenes, and 2 mm path length.

Precipitate asphaltenes were in sufficient supply to lity increased. The maximum in correlation length
maintain nearly complete emulsion stability. The did not coincide with the maximum in emulsion
slight increase in stability from 0 to 0.25:1 R/A was stability, perhaps because different initial asphal-
likely due to enhanced asphaltene solubility and tene concentrations were used for the SANS and
lability of the aggregates. In the absence of resins, emulsions experiements. The emulsions were
the aggregate size may have been too large for tested at an asphaltene concentration of 0.5 wt.%
asphaltenes to effectively cover water droplet and SANS experiments were performed at 1 wt.%.
interfaces and for w/o emulsions to achieve high Fewer resins were needed to dissolve asphaltenes
stability. As the large aggregates were solvated by at the lower concentration, thus the maximum in
resins, they became more interfacially active and emulsion stability was shifted to a R/A ratio of 5:1.
formed a cohesive film. Beyond 2:1 R/A the At R/A of 20:1, the additional resins solvated and
asphaltenic aggregates became increasingly solu- dissociated the asphaltenic aggregates rendering
ble, less surface-active, and consequently formed them less surface-active, and reducing the emul-
weaker emulsions. The most dramatic decrease in sion stability.
asphaltene correlation length occurred between R/ Similar trends in aggregate size and emulsion
A ratios of 0.5 and 2. A minimum aggregate size stability were observed for CS Precipitate asphal-
was approached above 2:1 R/A and the asphal- tenes in pure toluene and in 50% toluene. In pure
tenes were sufficiently solvated that they lost their toluene, correlation lengths increased from 101 Å
interfacial activity. (no resins) to 128 Å (0.5:1 R/A ratio) as resin
B6 Precipitate asphaltenes were beyond the limit addition dissolved the insoluble species followed
of solubility in 30% toluene and the emulsions by a monotonic decrease with further resin addi-
prepared without resins were not particularly tion to 26 Å (10:1 R/A ratio). CS Precipitate
stable (Fig. 7). The increase in correlation length asphaltenes did not form particularly stable emul-
with addition of resins to a R/A ratio of 10:1 sions in pure toluene and the addition of resins had
indicated that resins facilitated the dissolution of a modest effect on emulsion stability. As expected,
the insoluble asphaltenes. As more interfacially CS Precipitate asphaltenes were also partially
active material was dissolved, the emulsion stabi- insoluble in 50% toluene and the addition of resins
 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27 21

up to a R/A ratio of 2:1 aided the dissolution of

the insoluble species. Correlation lengths observed
for CS Precipitate in 50% toluene increased from
98 Å (0.5:1 R/A ratio) to 108 Å (2:1 R/A ratio)
followed by a decrease to 31 Å (10:1 R/A ratio). As
more interfacially active material was dissolved,
the emulsion stability increased markedly from
52% water resolved (no resins) to 25% water
resolved (5:1 R/A ratio).

3.4. Asphaltene (Soluble, Whole, Precipitate)
/

resin emulsion stability

The stability of emulsions prepared with AH

Whole and its fractions at several heptol ratios are
shown in Fig. 8(a
/c). As with the previous
experiments, the asphaltene concentration was
0.37% (w/v), or equivalently 0.5% (w/w), and the
resin
/asphaltene ratio was varied. The volume
fractions of toluene were chosen such that the
asphaltenes were either above or below their
solubility limits. AH Soluble asphaltenes in 60%
toluene formed weak emulsions that became
increasingly unstable as resin concentration in-
creased (Fig. 8(a)). The asphaltenes were very
soluble at these conditions and could not form
interfacial films capable of withstanding rupture.
AH Soluble asphaltenes in 30% toluene were
partially precipitated. Resin addition initially sol-
vated the flocculated aggregates at R/A ratios of
0.25 and 0.5 resulting in enhanced emulsion
stability. Further solvation by resins completely
destabilized the emulsions at a R/A ratio of 2. AH
Whole asphaltenes formed emulsions in 55%
toluene with nearly 50% water resolved (Fig.
8(b)). In the insoluble regime, (30% toluene) the
emulsions formed were initially weaker due to
non-surface-active, flocculated asphaltenic aggre-
gates. Resin addition up to 0.5:1 R/A increased the
emulsion stability as insolubles were dissolved into
solution. At higher R/A ratios, the 30% toluene
system became highly solvated and emulsions were
destabilized. Much the same effect was seen with
AH Precipitate asphaltenes (Fig. 8(c)) where the Fig. 8. Emulsion stability (% water resolved) of AH asphal-
emulsions containing the most soluble asphaltenes tenes in heptol with AH resins. Emulsions tested at 0.5 wt.%,
became unstable while partially insoluble systems 25 8C. (a) Soluble: 30, 60% toluene. (b) Whole: 30, 55% toluene.
actually formed stronger emulsions. (c) Precipitate: 30, 60% toluene.
22 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27

Fig. 9. Emulsion stability (% water resolved) of CS asphaltenes in heptol with CS resins. Emulsions tested at 0.5 wt.%, 25 8C. (a)
Soluble: 40, 50, 60% toluene. (b) Whole: 30, 45, 60% toluene. (c) Precipitate: 50, 100% toluene.
 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27 23

Fig. 10. Emulsion stability (vol.% water resolved) of B6 asphaltenes in heptol with B6 resins. Emulsions tested at 0.5 wt.%, 25 8C. (a)
B6 Soluble: 10, 30, 60% toluene. (b) B6 Whole: 40, 60, 100% toluene. (c) B6 Precipitate: 30, 60, 100% toluene. (d) HO Precipitate: 30,
60, 100% toluene.

The emulsions formed by CS Whole asphaltenes that CS asphaltenes are inherently weak emulsi-
and its more and less soluble asphaltene fractions fiers possibly due to large aggregate formation,
were typically unstable. In the soluble regime, the low polarity and an inability to form hydrogen
ability of each asphaltene to form stable emulsions bonds.
decreased with resin addition (Fig. 9(a
/c)). The Unlike the weak emulsions formed by AH and
emulsion stability of CS Soluble at 40% toluene CS asphaltenes over a wide range of solvent
appeared to increase slightly up to an R/A of 0.5:1 conditions, B6 and HO asphaltene-stabilized emul-
followed by a decrease. CS Whole asphaltenes sions were considerably stronger. The effects of
formed unstable emulsions beyond an R/A of 1:1 resins on emulsion stability in the soluble and
due to resin solvation. The behavior of partially insoluble regimes are quite apparent. In Fig. 10(a),
insoluble CS Precipitate at 50% toluene suggests resins are shown to destabilize emulsions in the
that resins were capable of a modest enhancement soluble regime of B6 Soluble. In the insoluble
of emulsion stability. Even when the systems were regime (30% toluene), emulsion stability is reduced
rendered more soluble with resins, emulsion stabi- only after reaching a 5:1 R/A ratio. Resins initially
lity did not increase appreciably. High aromaticity, dissolved precipitated asphaltenes and rendered
as suggested by low H/C ratios, likely caused the them interfacially active until the system became
formation of large aggregates through p
/p inter- too soluble and the driving force for film forma-
actions between asphaltene monomers and re- tion disappeared. In a highly aliphatic solvent
duced aggregate lability. These properties suggest (10% toluene:90% heptane) B6 Soluble asphaltenes
24 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27

did not form stable emulsions even with an R/A of stabilize emulsions in the presence of large con-
5:1. centrations of strongly solvating resins. In parti-
B6 Whole asphaltenes were studied at three cular, a balance between aromaticity and
solvent conditions in the presence of resins (Fig. hydrogen bonding capacity must exist to allow
10(b)). B6 Whole asphaltenes were completely stability in both pure toluene with no resins (high
soluble at both 60 and 100% toluene. B6 Whole aromaticity, low polarity) and in highly aliphatic
asphaltenes were very soluble in pure toluene but solvents with substantial resin content (low aro-
still formed stable emulsions up to a R/A of 0.25:1. maticity, high polarity).
Beyond this point the emulsion stability rapidly The effectiveness of resins at solubilizing as-
decreased as a result of enhanced solubility. At phaltenes surely depends on asphaltene and resin
60% toluene, B6 Whole asphaltenes approached chemical composition and on the supporting
their limit of solubility in solution and, conse- aliphatic and aromatic solvent. B6 Whole asphal-
quently, formed larger aggregates and stronger tenes formed stable emulsions at 40% toluene
emulsions than observed in pure toluene. As without resins but were only 60% soluble (Fig.
discussed in another paper [44], aggregate size 2). The amount of precipitated material decreased
was observed to increase with decreasing solvent below 10% at an R/A of 1:1. Above this resin
aromaticity up to the asphaltene solubility limit as content, the asphaltenes were soluble and formed
the soluble asphaltenes attempted to minimize stable emulsions up to R/A of 5:1. Further
interactions with the increasingly aliphatic solvent. solvation by addition of resins eventually destabi-
As shown in Fig. 10(b), the proximity to the lized the emulsions. At 30% toluene, B6 Precipitate
solubility limit also corresponds to a maximum asphaltenes did not form stable emulsions. Appar-
in emulsion stability. Emulsions of B6 Whole ently, either the amount of soluble material was
asphaltenes in 60% toluene eventually became insufficient and/or the soluble material was in
unstable at a 10:1 R/A ratio, while a R/A ratio aggregates of too large a size to effectively cover
of 5:1 was needed to completely destabilize emul- water droplet surfaces. Emulsion stability in-
sions in pure toluene. Solutions containing a creased slightly by adding resins of equal mass
fraction of insoluble asphaltenes (40% toluene) ratio to asphaltenes. Solubility measurements in-
were also able to form very stable emulsions up to dicated that the asphaltenes were nearly insoluble
a R/A ratio of 5:1. The dissolution of these at these conditions. However, at an R/A of 5:1, B6
insoluble asphaltenes facilitated strong interfacial Precipitate solubility increased from essentially
film formation up to R/A /10, at which point the insoluble to 55% soluble and emulsion stability
asphaltenes were so strongly solvated by resins reached a maximum. Further solvation with resins
that they ceased to be effectively surface-active. did not enhance gravimetric solubility, but emul-
B6 Precipitate asphaltenes were appreciably sion stability decreased substantially.
insoluble at both 30 and 60% toluene but soluble HO Whole and its more and less soluble
in pure toluene. B6 Precipitate formed stable fractions behaved in a very similar fashion to B6
emulsions in 60 and 100% toluene; however, at asphaltenes; however, HO Precipitate asphaltenes
30% toluene, B6 Precipitate asphaltenes were not were observed to form even stronger emulsions
sufficiently soluble to stabilize emulsions (Fig. than B6 Precipitate fractions (Fig. 10(d)). HO
10(c)). The effect of adding resins at 30% toluene Precipitate asphaltenes were solubilized with resins
was to effectively dissolve the insoluble asphal- at 30% toluene and formed stable emulsions up to
tenes and form strongly stable emulsions above a R/A ratios of 20:1. Again, in the regime in which
R/A of 5:1. Even at 20:1 R/A, there still existed a HO Precipitate asphaltenes were soluble, the
sufficient supply of film forming material to form addition of resins led to emulsion destabilization.
relatively stable emulsions, a remarkable observa- The power of resins to solvate asphaltenes of
tion. This level of solubility is quite high and varied chemical composition is evident from
suggests that B6 Precipitate asphaltenes have SANS, solubility, and emulsion stability measure-
unique molecular structures that allow them to ments. In addition, the ability to dissolve insoluble
 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27 25

asphaltenes and render them capable of emulsion increased with resin addition up to the solubility
stabilization is remarkable. However, this resin
/ limit as the more polar, insoluble asphaltenes were
asphaltene interaction appears to act primarily dissolved into solution. Once the insolubles were
with HO and B6 Precipitate, and CS to a lesser dissolved, additional resins solvated the largest of
degree, and must be linked significantly to the the soluble aggregates as observed by a decrease in
molecular structure of these asphaltenes. HO and the correlation length.
B6 Precipitate fractions have the highest polarity Asphaltene stabilized emulsions were susceptible
of any of the fractions previously prepared (see to the effects of resin solvation. Soluble asphal-
Table 2). The H/C ratio of these fractions is also tenes typically formed the weakest emulsions
lower than B6 or HO Whole but not as low as AH because the aggregates were small, well solvated,
or CS asphaltenes, indicating a moderate level of less aromatic, and less polar than the Whole and
aromaticity. These chemical properties suggest Precipitate fractions, and consequently, less inter-
that aggregation and film formation may be driven facially active. Soluble and Whole asphaltenes at
more so by polar heteroatom interactions such as toluene concentrations in the insoluble regime
hydrogen bonding than by p
/p bonding between could be solvated by resin addition to produce
asphaltenic aromatic moieties. AH and CS asphal- slightly more stable emulsions. Resins completely
tenes contain low concentrations of polar nitrogen destabilized emulsions of Whole and Soluble
and likely aggregate via aromatic stacking. With- asphaltenes in the portion of phase space at which
out sufficient proton donor
/acceptor sites, the these asphaltenes were soluble at a R/A of 2:1,
asphaltenes cannot adsorb and form a cohesive except in the case of B6 Whole, which formed
film at the oil
/water interface. stable emulsions up to a R/A of 10:1.
AH and CS Precipitate asphaltenes were poor
emulsion formers without resins and their emul-
4. Conclusions sions were only modestly more stable with resins.
Their lack of polarity, even when solvated to
In this study, resins isolated by SARA fractio- become more labile, prevented them from inter-
nation had a strong solvating effect on asphaltenes acting strongly at oil
/water interfaces. B6 and HO
and their more and less soluble subfractions */so- Precipitate, however, formed very stable emulsions
called ‘‘Soluble’’ and ‘‘Precipitate’’ asphaltenes. At in their soluble regime that were resistant to the
an R/A ratio of 1:1, the toluene concentration at solvating effects of resins. Both approached com-
which asphaltenes began to precipitate in heptol plete instability near a R/A of 10:1. At 60% toluene
was reduced by as much as 10%. The sizes of B6 and HO Precipitate asphaltenes were partially
asphaltenic aggregates, as gauged by correlation insoluble but formed very stable emulsions due to
lengths from neutron scattering, were observed to their highly polar and H-bonding nature. Resins
decrease significantly upon addition of resins. At added to these systems dissolved the insoluble
room temperature, the correlation lengths of asphaltenic flocs, rendering them interfacially
Whole asphaltenes approached a value of approxi- active. Once B6 Precipitate asphaltenes were
mately 14
/18 Å at high resin concentrations (R/A solubilized, emulsions remained stable even at a
of 10:1) regardless of asphaltene, resin type, or % R/A of 20:1. In highly aliphatic solvents (30%
toluene (in the soluble regime). Furthermore, the toluene), B6 and HO Precipitate asphaltenes were
greatest decreases in aggregate size occurred at largely insoluble and did not form stable emul-
conditions found in typical crude oils (i.e. R/A sions. Resins were able to solvate these systems to
ratios between 0.5 and 2:1). It is not known to a considerable extent and led to stable emulsions
what extent the aggregates can be dissociated by at R/A ratios of 20:1.
resin addition. The presumed minimum would be a Comparisons between aggregate sizes and the
single asphaltene monomer solvated completely by stability of oil-in-water emulsions formed at var-
resins. In the portion of the phase diagram in ious R/A ratios suggested that the most stable
which asphaltenes were insoluble, aggregate sizes emulsions formed from soluble aggregates of
26 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27

maximum size and greatest interfacial activity. Department of Energy, Argonne National La-
These aggregates invariably formed at the limit boratory, and particularly to Pappannan Thiya-
of asphaltene solubility. Modest differences in garajan and Denis Wozniak for their assistance on
resin aromaticity and polarity are secondary to the SAND instrument. We also want to thank
asphaltene chemistry and solvent conditions for Marit-Helen Ese and Jihong Tong for helping with
dictating aggregate size and emulsion stability in the SANS data collection.
asphaltene solutions.
The remarkable ability of resins to solvate
asphaltenes can be attributed to the polar and
References
dispersive nature of the resin molecules. In highly
aromatic solvents, where p
/p interactions between [1] S.A. Berridge, M.T. Thew, A.G. Loriston-Clarke, J. I.
asphaltenic aggregates are largely mitigated, resins Petrol. 54 (1968) 333.
serve to further disrupt polar and hydrogen- [2] C.M. Blair, Chem. Ind. 20 (1960) 538.
bonding interactions within asphaltene aggregates. [3] F.J. Nellensteyn, J. I. Petrol. Technol. 10 (1924) 311.
At low solvent aromaticity, resin solvation aids to [4] J.P. Pfeiffer, R.N.J. Saal, J. Phys. Chem. 44 (1940) 139.
[5] A.S.C. Lawrence, W. Killner, J. I. Petrol. 34 (1948) 281.
break flocs by disruption of dispersion and hydro- [6] M.O. Denekas, F.T. Carlson, J.W. Moore, C.G. Dodd,
gen bonding interactions between aggregates. For Ind. Eng. Chem. 43 (1951) 1165.
systems containing asphaltenes with high polarity [7] C.G. Dodd, J. Phys. Chem. 64 (1960) 544.
and surface activity (e.g. B6 and HO Whole and [8] H. Neumann, Petrochemie 18 (1965) 776.
Precipitate), resin solvation is sometimes insuffi- [9] O.K. Kimbler, R.L. Reed, I.H. Silberberg, Soc. Petrol.
Eng. J. 6 (1966) 153.
cient to induce complete emulsion destabilization. [10] J.E. Strassner, J. Petrol. Technol. 20 (1968) 303.
Weak emulsion formers and Soluble asphaltenes [11] G.D.M. Mackay, A.Y. McLean, O.J. Betancourt, B.D.
possessing a lower polarity and aromaticity are Johnson, J. I. Petrol. 59 (1973) 164.
easily destabilized in the presence of resins. Ag- [12] R.J.R. Cairns, D.M. Grist, E.L. Neustadter, in: A.L.
Smith (Ed.), Theory and Practice of Emulsion Technology,
gregation and film formation in petroleum fluids
Academic Press, Brunel University, 1974, p. 135.
are likely driven by polar heteroatom interactions, [13] J.J. Oren, G.D.M. MacKay, Fuel 56 (1977) 382.
such as hydrogen bonding. The most polar [14] T.J. Jones, E.L. Neustadter, K.P. Whittingham, J. Can.
asphaltenes, typically concentrated in the least Petrol. Technol. 17 (1978) 100.
soluble fraction, require the largest concentration [15] A.L. Bridie, T.T. Wanders, W. Zegveld, H.B. van der
of resins to completely destabilize asphaltene Heijde, Mar. Pollut. Bull. 11 (1980) 343.
[16] S.E. Taylor, Chem. Ind. 20 (1992) 770.
emulsions and likely cause many petroleum pro- [17] C.S. Shetty, A.D. Nikolov, D.T. Wasan, J. Disper. Sci.
duction problems such as pipeline deposition and Technol. 13 (1992) 121.
water-in-crude oil emulsion stabilization. [18] D.L. Mitchell, J.G. Speight, Fuel 52 (1973) 149.
[19] T.F. Yen, J.G. Erdman, S.S. Pollack, Anal. Chem. 33
(1961) 1587.
[20] R.V. Barbour, J.C. Petersen, Anal. Chem. 46 (1974) 273.
Acknowledgements [21] M.M. Boduszynski, J.F. McKay, D.R. Latham, Proc.
Assoc. Asphalt Paving Technol. 49 (1980) 123.
This work was supported by grants from the [22] T. Ignasiak, O.P. Strausz, D.S. Montgomery, Fuel 56
National Science Foundation (CTS-981727), (1977) 359.
[23] S.E. Moschopedis, J.G. Speight, Fuel 55 (1976) 187.
PERF (97-07), and shared consortium funding
[24] J.C. Petersen, Fuel 46 (1967) 295.
from ExxonMobil, Ondeo-Nalco Energy Systems, [25] M.M.H. Al-Jarrah, A.H. Al-Dujaili, Fuel Sci. Technol.
Shell Oil Company and Texaco. Special thanks go Int. 7 (1989) 69.
to George Blankenship and Semaj McIver who [26] S. Acevedo, B. Mendez, A. Rojas, I. Layrisse, H. Rivas,
assisted with the experimental work. We would Fuel 64 (1985) 1741.
like to thank Eric Sirota of ExxonMobil Corpo- [27] M.M. Boduszynski, Energy Fuels 2 (1988) 597.
[28] I.A. Wiehe, K.S. Liang, Fluid Phase Equilib. 117 (1996)
rate Research for acquiring the SANS beam time 201.
at NIST and Min Lin for his assistance on the [29] J.F. McKay, P.J. Amend, T.E. Cogswell, P.M. Harnsber-
NG-7 and NG-1 beamlines. We are grateful to the ger, R.B. Erikson, D.R. Latham, in: P.C. Uden, S. Siggia
 P.M. Spiecker et al. / Colloids and Surfaces A: Physicochem. Eng. Aspects 220 (2003) 9
/27 27

(Eds.), Analytical Chemistry of Liquid Fuel Sources: Tar [49] P.M. Spiecker, Ph?.D. Thesis. North Carolina State
Sands, Oil Shale, Coal, and Petroleum, vol. 170, American University, 2001 .
Chemical Society, Washington, 1979, p. 128. [50] M. van der Waarden, Kolloid Z. Z. Polym. 156 (1958) 116.
[30] S.I. Andersen, A. Keul, E. Stenby, Petrol. Sci. Technol. 15 [51] R. Miller, Anal. Chem. 54 (1982) 1742.
(1997) 611. [52] J.G. Reynolds, Fuel Sci. Technol. Int. 5 (1987) 593.
[31] H.W. Yarranton, J.H. Masliyah, AICHE J. 42 (1996) [53] M.F. Ali, A. Bukhari, M.U. Hasan, Fuel Sci. Technol. Int.
3533. 7 (1989) 1179.
[32] H.W. Yarranton, H. Alboudwarej, R. Jakher, Ind. Eng. [54] M. Alula, M. Diack, R. Gruber, G. Kirsch, J.C. Wilhelm,
Chem. Res. 39 (2000) 2916. D. Cagniant, Fuel 68 (1989) 1330.
[33] P. Herzog, D. Tchoubar, D. Espinat, Fuel 67 (1988) 245. [55] W.R. Middleton, Anal. Chem. 39 (1967) 1839.
[34] J.C. Ravey, G. Ducouret, D. Espinat, Fuel 67 (1988) 1560. [56] J.G. Reynolds, W.R. Biggs, Fuel Sci. Technol. Int. 4 (1986)
[35] R.E. Overfield, E.Y. Sheu, S.K. Sinha, K.S. Liang, Fuel 749.
Sci. Technol. Int. 7 (1989) 611. [57] L.X. Nghiem, M.S. Hassam, R. Nutakki, A.E.D. George,
[36] E.Y. Sheu, D.A. Storm, M.M.D. Tar, J. Non-Cryst. Solids S. Petrol. Eng. 5 (1993) 375.
131
/133 (1991) 341. [58] S.I. Andersen, K.S. Birdi, J. Colloid Interf. Sci. 142 (1991)
[37] D. Espinat, J.C. Ravey, V. Guille, J. Lambard, T. Zemb, 497.
J.P. Cotton, J. Phys. IV 3 (1993) 181. [59] J. Murgich, J.A. Abanero, O.P. Strausz, Energy Fuels 13
[38] P. Thiyagarajan, J.E. Hunt, R.E. Winans, K.B. Anderson, (1999) 278.
J.T. Miller, Energy Fuels 9 (1995) 829. [60] J. Murgich, O.P. Strausz, Petrol. Sci. Technol. 19 (2001)
[39] H. Rassamdana, M. Sahimi, AICHE J. 42 (1996) 3318.
231.
[40] L. Barre, D. Espinat, E. Rosenberg, M. Scarcella, Rev. I. [61] J.D. McLean, P.K. Kilpatrick, J. Colloid Interf. Sci. 196
Fr. Petrol. 52 (1997) 161.
(1997) 23.
[41] M.Y. Lin, E.B. Sirota, H. Gang, Abstr. Pap. Am. Chem. S.
[62] H. Fordedal, Y. Schildberg, J. Sjoblom, J.-L. Volle,
213 (1997) 66.
Colloids Surf. A 106 (1996) 33.
[42] E.B. Sirota, Petrol. Sci. Technol. 16 (1998) 415.
[63] R.A. Mohammed, A.I. Bailey, P.F. Luckham, S.E. Taylor,
[43] D. Fenistein, L. Barre, Fuel 80 (2001) 283.
Colloids Surf. A 80 (1993) 237.
[44] P.M. Spiecker, K.L. Gawrys, P.K. Kilpatrick, J. Colloids
[64] H. Fordedal, Y. Schildberg, J. Sjoblom, J.-L. Volle, J.
Interf. Sci., 2003, in press.
[45] O.P. Strausz, P. Peng, J. Murgich, Energy Fuels 16 (2002) Colloid Interf. Sci. 182 (1996) 117.
[65] J.D. McLean, P.K. Kilpatrick, J. Colloid Interf. Sci. 189
809.
[46] E.Y. Sheu, K.S. Liang, S.K. Sinha, R.E. Overfield, J. (1997) 242.
Colloids Interf. Sci. 153 (1992) 399. [66] O.V. Gafonova, H.W. Yarranton, J. Colloid Interf. Sci.
[47] C. Bardon, L. Barre, D. Espinat, V. Guille, M.H. Li, J. 241 (2001) 469.
Lambard, J.C. Ravey, E. Rosenberg, T. Zemb, Fuel Sci. [67] J.D. McLean, P.K. Kilpatrick, Energy Fuels 11 (1997) 570.
Technol. Int. 14 (1996) 203. [68] P.-G. de Gennes, Scaling Concepts in Polymer Physics,
[48] J.C. Ravey, D. Espinat, Prog. Coll. Pol. Sci. 81 (1990) 127. Cornell University Press, Ithaca, 1979.