2416 Energy & Fuels 2005, 19, 2416-2422

Zeta Potential and Langmuir Films of Asphaltene Polar

Fractions
Heriberto Grijalva-Monteverde,† Oscar V. Arellano-Tánori,† and
Miguel A. Valdez*,†,‡
Departamento de Investigación en Materiales and Departamento de Fı́sica, Universidad de
Sonora, Rosales y transversal, C. P. 83000, Hermosillo, Sonora, México

Received April 25, 2005. Revised Manuscript Received July 24, 2005

In this work, we study the surface and electrokinetic properties of asphaltenes and some of
their polar fractions. Using different mixtures of pentane/methylene chloride as a precipitant,
we obtain four fractions (F1-F4), which contain different proportions of iron, nickel, copper, cobalt,
aluminum, and vanadium. Modifications of asphaltenes and fractions F1 and F2 are performed
by treatment with the salt form of ethylenediamine tetraacetic acid (EDTA) and, after this, with
aqueous solutions of copper sulfate. The results indicate differences in the metal content between
fractions before and after the treatments. Besides these differences, we observed differences in
the zeta potential measurements and differences in the isotherms at the air/water interface. Our
results show that, after the EDTA treatment, asphaltenes and fractions F1 and F2 can retain
copper (>200% of the original copper content of the samples). This high copper content of the
treated asphaltenes and the F1 and F2 samples is reflected in a change of the electrical charge
from a negative charge to a positive charge. The hydrophobic behavior of the samples was analyzed
with a Langmuir balance. The results show that the asphaltenes and the F1 fractions are more
hydrophilic, with a smaller molecular area, and the F4 fractions are more hydrophobic, with a
larger molecular area, before and after EDTA and copper sulfate treatments. Both the zeta
potential measurements and Langmuir films of the samples indicate the possibility that EDTA
molecules remain attached to the asphaltene molecules.

Introduction different subfractions have the same type of molecules

but different molecular population distributions. An-
The decline of the oil production with conventional other interesting procedure to separate the soluble and
methods, has obligated producers to exploit heavy oil
insoluble fractions of asphaltenes was used by Acevedo
reservoirs, and, at the same time, the corresponding
et al.,5 who used the PNP (p-nitrophenol) method. They
problems that are associated with the deposition of the
found interesting results in the composition of the
asphaltenes have been increased. Asphaltenes, which
subfractions. The resulting toluene-insoluble fraction
are defined as the heaviest fraction of the crude, are
obtained by precipitation with n-heptane or n-pentane, had poor hydrogen aromaticity and rich carbon aroma-
among other compounds.1 Many methods to obtain ticity, which is consistent with a more rigid and planar
asphaltenes have been described in the literature,2,3 and structure; in addition, both soluble and insoluble frac-
it has been shown that the physicochemical properties tions contained a high content of hydrogen bound to
of asphaltenes are sensitive to the separation techniques aliphatic carbons that were joined to aromatic com-
and solvents that are used. pounds. The fraction that was soluble in toluene was
Different attempts have been realized to investigate rich in hydrogen that was bound to aliphatic chains,
the physical and chemical properties of the different which is consistent with a more flexible structure.
fractions of asphaltenes. Groenzin and Mullins4 have Similar results of the behavior of asphaltenes treated
investigated the molecular size of different asphaltenes using PNP were obtained by Gutierrez et al.6
fractions obtained via precipitation by modifying the Recently, Douda et al.7 analyzed the asphaltene
toluene/n-pentane proportions and determined that fractions by the Soxhlet method and found, by different
techniques, a series of maltenes fractions with different
* Author to whom correspondence should be addressed. Tel.: +52- aromatic proportions and showed that the original
6622592108. Fax: +52-6622592109. E-mail address: valdez@
[Link]. asphaltenes contained a greater amount of heterocyclic
† Departamento de Investigación en Materiales.
molecules and ketones than the maltenes subfractions.
‡ Departamento de Fı́sica.

(1) Zhang, L. Y.; Lawrence, S.; Xu, Z.; Masliyah, J. H. J. Colloid

Interface Sci. 2003, 264, 128-140. (5) Acevedo, S.; Escobar, O.; Echavarrı́a, L.; Gutiérrez, L. B.;
(2) Parra-Barraza, H.; Hernández-Montiel, D.; Lizardi, J.; Hernán- Méndez, B. Energy Fuels 2004, 18, 305-311.
dez, J.; Herrera-Urbina, R.; Valdez, M. A. Fuel 2003, 82, 869-874. (6) Gutiérrez, L. B.; Ranaudo, M. A.; Méndez, B.; Acevedo, S. Energy
(3) Alboudwarej, H.; Beck, J.; Svrcek, W. Y.; Yarranton, H. W. Fuels 2001, 15, 624-628.
Energy Fuels 2002, 16, 462-469. (7) Douda, J.; Llanos, M. E.; Alvarez, R.; Navarrete-Bolaños, J.
(4) Groenzin, H.; Mullins, O. C. Energy Fuels 2003, 17, 498-503. Energy Fuels 2004, 18, 736-742.

10.1021/ef050120y CCC: $30.25 © 2005 American Chemical Society

Published on Web 08/24/2005
Surface Properties of Asphaltene Polar Fractions Energy & Fuels, Vol. 19, No. 6, 2005 2417

In contrast, the maltenes contain more alkyl side chains, paper (10-15 µm). The asphaltenes were dried at room
which contribute to the solubility of the asphaltenes. temperature.
Asphaltenes precipitated with both polar (acetone) Fractionation. The asphaltene sample obtained was dis-
and nonpolar (n-heptane) solvents have been analyzed solved in methylene chloride (10% w/w). Pentane then was
added until a 3/7 methylene chloride/pentane ratio was
by Buenrostro-González et al.,8 and interesting differ-
obtained. The precipitate is termed the F1 fraction. This
ences in solubility and molecular weight distributions fraction was collected by filtration with a Millipore 0.45 µm
were obtained, depending on the toluene/n-heptane and filter and was allowed to dry at room temperature. The F2
toluene/acetone proportions used to precipitate the fraction was precipitated using the supernatant and adding
asphaltene fractions of the solvent. Boukir et al.9 enough pentane to obtain a ratio of 1:3. Other fractions were
observed important differences in the constituents of obtained by continuing with the method.10
resins and asphaltenes. Resins have more-aliphatic and Metal Chelation. We used 1.5 g of asphaltenes, F1 fraction,
less-aromatic compounds, and they are composed of and F2 fraction, respectively, and separately mixed with 500
polar end groups with O, S, and N heteroatoms and mL of a 0.15 M of EDTA in deionized water at pH 9. The
nonpolar paraffinic groups. Their proposed model of solutions were agitated for 3 h at room temperature. The
precipitates then were filtered and vacuum-dried for 4 h.
resin/asphaltene interaction suggests that resins are
The effect of a copper sulfate solution on the samples treated
attached to asphaltene micelles through their polar end with EDTA was investigated by mixing the samples with a
groups (hydrogen bonding, dipole-dipole interactions). 0.2 M CuSO4‚5H2O solutions. The solutions were agitated for
Kaminski et al.10 have classified asphaltenes according 3 h at room temperature and twice washed with deionized
to the polarity of the solvent to precipitate asphaltenes. water and filtered. Samples then were vacuum-dried for 3 h.
Using a mixture of pentane/methylene chloride, they Aqueous suspensions of asphaltenes for zeta potential
obtained different fractions with different solubility measurements were prepared as follows. First, 50 mg of
properties, as a consequence of their polarity properties. asphaltenes were dispersed in 15 mL of ethanol under ultra-
On the other hand, surface properties of asphaltenes sound during 20 min. A quantity (1.5 mL) of this suspension
have been shown to be very important in the absorption was added to 100 mL of triply distilled water containing 0.001
M NaNO3 to fix the ionic strength. The pH then was adjusted,
process on different rocks,11 and, recently, Jada and
using nitric acid or sodium hydroxide diluted solutions, to the
Salou12 used the zeta potential to demonstrate that the desired value and the suspension was conditioned for 20 min
proportion of resins to asphaltenes determines the under magnetic agitation. During the conditioning time, if the
charge of the bitumens and this proportion contributes pH changed, it was readjusted to the desired value.
to maintain the stability in the crude oil. In a recent Zeta potentials were measured with a Zeta Meter 3.0+ unit
paper,2 we have used zeta potential measurements on (manufactured by Zeta-Meter, Inc., New York). This apparatus
different asphaltenes samples, obtained with different includes a microprocessor that first measures the electro-
n-heptane/crude proportions, and demonstrated the phoretic mobility of colloidal particles dispersed in aqueous
influence of the precipitation method on the electrical solutions, and then automatically calculates the zeta potential
behavior of asphaltenes. (in millivolts) using the Smoluchowski equation. This is the
most elementary expression for zeta potential (ζ) calculations
In this work, we investigate different asphaltene
from electrophoretic mobility (Ue) measurements. The equation
fractions, using the method of Kaminski et al.10 to obtain can be written as
different polar fractions. We modify the asphaltenes and
the fractions by treatment with the salt form of ethyl-
enediamine tetraacetic acid (EDTA) and with aqueous ζ) (η)U e
solutions of copper sulfate and perform zeta potential
measurements and Langmuir isotherms for the as- where η and  represent the viscosity and the permittivity of
phaltenes and some fractions. Our aim is to understand the aqueous medium. Zeta potentials reported here are the
the influence of polarity and metal content of asphal- average values obtained by tracking at least 20 different
tenes and the fractions with the corresponding surface particles.
electrical charge and the behavior of their monolayers A Langmuir balance (Nima, Langmuir Blodgett through,
on the air/water interface. model type 611), whose surface tension precision is 0.1 mN/
m, was used to determine two-dimensional isotherms of
asphaltenes at the air/water interface. Each asphaltene sample
Experimental Section was dissolved in toluene, and then 100 µL of a 1 mg/mL
Asphaltenes were obtained by precipitation with n-heptene solution was deposited at the clean water surface (using
with a modified procedure contained in ASTM 2007D.13 Milli-Q water, Millipore) with a Hamilton microsyringe. The
Asphaltenes samples were obtained from a crude oil (Maya) temperature was maintained at 25 ( 0.1 °C.
from the Instituto Mexicano del Petróleo, México. Briefly, one Sample preparation for atomic adsorption spectroscopy
volume of crude oil was mixed with 40 volumes of n-heptane (AAS) was conducted as follows. We prepared 1.25 g of an
for 1 h at room temperature and kept sealed for 2 days at room equimolar mixture of Na2CO3 and K2CO3, then mixed them
temperature. The mixture was filtered with Whatman #4 filter with 250 mg of the sample in a mortar with a pestle. The
mixture was fused at 900 C, then 5 mL of nitric acid (HNO3)
(8) Buenrostro-González, E.; Andersen, S. I.; Garcı́a-Martı́nez, J. A.; and 1 mL of hydrochloric acid (HCl) were added at room
Lira-Galeana, C. Energy Fuels 2002, 16, 732-741. temperature. The solution is boiled with deionized water.
(9) Boukir, A.; Aries, E.; Guiliano, M.; Asia, L.; Doumenq, P.; Mille, Finally, deionized water was added to bring the sample volume
G. Chemosphere 2001, 43, 279-286. to 100 mL. Sample analysis was performed using a Perkin-
(10) Kaminski, T. J.; Folger, H. S.; Wolf, N.; Wattana, P.; Mairal,
A. Energy Fuels 2000, 14, 25-30. Elmer instrument (model 2380).
(11) Kokal, S.; Tang, T.; Schramm, L.; Sayegh, S. Colloids Surf. A
1995, 94, 253-265.
(12) Jada, A.; Salou, M. J. Pet. Sci. Eng. 2002, 33, 185.
Results
(13) ASTM Annual Book of ASTM Standards, Vol. 5.0; American
Society for Testing and Materials (ASTM): Philadelphia, PA; 1983; p Metal Content of Samples. The results of the
158. content of some metals (iron, nickel, copper, cobalt,
2418 Energy & Fuels, Vol. 19, No. 6, 2005 Grijalva-Monteverde et al.

Figure 1. Proposed asphaltene structure with incorporated Figure 2. Zeta potential measurements of the asphaltenes
metal ions. The asphaltene structure without metal ions has and fractions F1 and F2, for different pH in triply distilled
been proposed in an earlier publication. [Strausz, P. O.; water and 0.001 M NaNO3 to fix the ionic stregnth.
Mojelesky, W. T.; Faraji, F.; Louren, M. E.; Perg, P. Additional
Structural Details on Athabasca Asphaltene and Their Rami-
The effect of the treatment of asphaltenes and the F1
fications. Energy Fuels 1999, 13, 207-227.]
and F2 fractions with a CuSO4 solution showed that the
Table 1. Metal Content of the Asphaltenes and the First amount of copper was increased by >200%, in compari-
Two Fractions (F1 and F2) before and after Treatment son with the copper present in the original samples (527
with Ethylenediamine Tetraacetic Acid (EDTA) mg/mL for asphaltenes, 340 mg/mL for the F1 fraction,
Metal Content (mg/mL) and 165 mg/mL for the F2 fraction after the treatment).
Before EDTA Treatment After EDTA Treatment Zeta Potential Results. Zeta potential measure-
metal asphaltene F1 F2 asphaltene F1 F2
ments have been proven to be a good reference to detect
the presence of resins in the bitumens composition, as
iron 713 905 865 421 520 514
nickel 267 329 285 264 325 277
demonstrated by Jada and Salou.12 Bitumens with a
copper 220 80 59 131 76 50 high resins/asphaltenes ratio show a higher positive zeta
cobalt 43 50 50 31 44 44 potential at low pH, compared with samples with a low
vanadium 1462 1778 1680 1141 1752 1654 resins/asphaltenes ratio. This behavior was demon-
aluminum 440 560 396 290 427 268 strated earlier2 with different asphaltenes samples by
vanadium, and aluminum) in the aqueous phase of the precipitating asphaltenes with different crude/n-hep-
digestion of the samples of asphaltenes and two frac- tane proportions. The same behavior has been observed
tions (F1 and F2) before and after treatment with EDTA by González et al.15 for asphaltenes from different
at pH 9, as described previously, are shown in Table 1. Brazilian crude oils. The pH at which the zeta potential
The results are in accordance with those observed by is zero defines the isoelectric point (IEP) of the material;
Kaminski et al.10 Taking into account the effect of the at pH values smaller than the IEP, the zeta potential
EDTA treatment, we can identify two groups of met- is positive whereas at pH values higher than the IEP,
als: one group is formed by iron and aluminum, which the zeta potential is negative. The pH effect on the zeta
are removed considerably (>40%). The other group is potential of asphaltenes is explained if one considers
formed with nickel, copper, cobalt, and vanadium, which that the surfaces of these materials contain pH-depend-
practically retain their original concentrations. This ent ionizable functional groups, acidic and basic, that
could be explained by considering that nickel, copper, can undergo dissociation and protonation.16 A negative
cobalt, and vanadium are being sequestered in central surface charge on asphaltenes results upon dissociation
porphyrin-like structures, where these metal ions show of acidic surface functional groups such as carboxylic,
a strong stability. In Figure 1 (extracted from ref 14), which exhibit pKa values of e4,12,14 and phenolic groups,
these metal ions are represented by the symbol M in which exhibit pKa values of 4 < pKa< 9.17 The positive
the porphyrin ring. We consider that one important surface charge results upon protonation of basic nitrogen-
fraction of Fe and Al must be joined to the asphaltene containing functional groups, such as pyridine.12
molecules in a different form (represented by the symbol In Figure 2, we show the behavior of the zeta potential
X in Figure 1). We suggest that the fraction of Fe and for the asphaltene sample and the two fractions F1 and
Al extracted with EDTA could be ions joined with other F2 for different pH. We observe that the behavior of the
heteroatoms such as sulfur and oxygen, which are complete asphaltenes is similar to that which has
present in the asphaltenes structures and, therefore, the previously been observed.2 We notice a different behav-
interactions between metal X and the asphaltenes could ior for the fractions at low pH. The zeta potential value
be weaker than the interactions between metal X and
(15) González, G.; Neves, G. B. M.; Saraiva, S. M.; Lucas, E. F.; dos
the EDTA. Anjos de Sousa, M. Energy Fuels 2003, 17, 879-886.
(16) Abraham, T.; Christendat, D.; Karan, K.; Xu, Z.; Masliyah, J.
(14) Cotton, F. A.; Wilkinson, G.; Murillo, C. A.; Bochmann, M. Ind. Eng. Chem. Res. 2002, 41, 2170.
Advanced Inorganic Chemistry, 6th Edition; Wiley: New York, 1999. (17) Jada, A.; Chau, A. A. J. Pet. Sci. Eng. 2003, 39, 287-296.
Surface Properties of Asphaltene Polar Fractions Energy & Fuels, Vol. 19, No. 6, 2005 2419

Figure 3. Zeta potential measurements of the asphaltenes Figure 4. Zeta potential measurements of the asphaltenes
and fractions F1 and (2) F2 treated with EDTA, for different and fractions F1 and F2 treated with CuSO4 after treatment
pH in triply distilled water and 0.001 M NaNO3 to fix the ionic with EDTA, for different pH in triply distilled water and 0.001
strength. M NaNO3 to fix the ionic strength.
is slightly higher (in magnitude) for the F1 fraction than for the F1 fraction without EDTA treatment and is not
for the F2 fraction and the asphaltenes. This could be a very sensitive to the increasing pH. This could mean
consequence of the higher polarity of the F1 fraction, that this fraction, when reduced of cations and resins,
because of its lower resins proportion, which leaves free is not significantly affected by EDTA molecules and is
more carboxylic or phenolic groups at the interface. The now completely saturated by negative charges at lower
polarity of the F1 fraction is probably increased with pH, and additional OH- radicals in solution do not
the higher metal content (10% more than fraction F2 contribute significantly to the net charge of the particles.
and 15% more than asphaltenes). We notice that, for A similar effect is produced in the F2 sample. However,
the lowest pH used, asphaltenes particles find their IEP. it seems that the charge saturation is produced at
As long as we increase the OH- concentration, we higher pH (∼9), which was indicative that EDTA
observe a decrease in the zeta potential values of the strongly modified the F2 structure and charge sites. At
F1 and F2 fractions. This phenomenon could be under- higher pH, there is a slight increase of the zeta potential
stood if we assume that, for the F1 and F2 frations, values; however, this variation can be considered as a
asphaltene molecules and resins could be joined by fluctuation inside the error bars. The asphaltene par-
hydrogen or dipole-dipole bondings9 and when OH - ticles showed a decrease in the zeta potential values as
radicals are added to the water, they neutralize the the pH was increased, and, for pH ∼9, we observed a
protons or the metal ions, destroying these bonds and relative maximum. This could mean, again, that, first,
leaving free negatively charged resin molecules. This at low pH, the OH- radicals break hydrogen bonds
would leave F1 and F2 particles less negative charged between asphaltenes and negatively charged resins and/
until this process ends and a relative maximum around or negatively charged EDTA-resin complexes, leaving
pH 4 and 4.5 is reached as observed in Figure 2. Up to less negatively charged asphaltene molecules at the
this pH, the behavior is similar for the three samples, interface of asphaltene particles. Thus, with the addition
and for pH >10, it seems that the surface charge of the of more OH- radicals, a partial neutralization was
particles of the three samples is saturated with the OH- produced and less-negative zeta potential values were
radicals from the solution. obtained. For higher pH values, the behavior of asphal-
The effect of the EDTA, as shown in Table 1, is to tenes is similar to that observed in Figure 2.
eliminate a great proportion of some metals from Finally, we show the results of zeta potential mea-
asphaltenes and both fractions. These changes in the surements for the three samples after they were treated
metal composition of asphaltenes and fractions produce with the copper sulfate solution, as described in the
variations on the zeta potential measurements of the Experimental Section. As can be observed in Figure 4,
same three samples, as can be observed in Figure 3. We we notice that the sign of the charge is inverted for all
notice that, for low pH, the behavior of asphaltenes is the pH range for both fractions and asphaltenes. From
very different, in comparison to that shown in Figure the results obtained of the copper content of the samples
2. At low pH, the zeta potential value of the asphaltenes with the copper sulfate treatment (527 mg/mL for
and F2 samples is higher in magnitude (g40 mV) asphaltenes, 340 mg/mL for F1 and 165 mg/mL for F2)
comparing with the values without the EDTA treat- and the corresponding values shown in Table 1, we
ment. This probably happens because EDTA molecules conclude that samples treated with the copper solution
remove some cations outside of the kernel of the contain ∼3-4 times the amount of copper than the
asphaltene molecules and the two fractions and prob- samples treated only with EDTA. This effect suggests
ably some type of EDTA-resin bonds are formed at the that copper is included in the particles of asphaltenes
surface of the particles. Observe that the zeta potential and both fractions, producing a positive zeta potential.
value at low pH for F1 is almost the same as that shown The small changes of the zeta potential values with the
2420 Energy & Fuels, Vol. 19, No. 6, 2005 Grijalva-Monteverde et al.

increasing pH suggests that very few charges are added

to the particles from the solution in the entire range of
pH analyzed, and the added OH- radicals are not
enough to change the sign of the particles. Another
possible effect of the copper treatments on the asphal-
tenes and the two fractions could be that most of resins
were absent in the samples, increasing the positive sign
of the charge. This possibility will be discussed by
analyzing the corresponding monolayers at the air/water
interface.
Langmuir Isotherms. The molecular behavior of
monolayers of asphaltenes and fractions F1-F4 were
investigated at the air/water interface, using a Lang-
muir balance. The two-dimensional isotherms are ob-
tained by measuring the pressure, which is defined as

π ) γ0 - γ
Figure 5. Isotherms in the air/water interface of the asphalt-
where γ0 is the surface tension of pure water and γ is ene sample and the four fractions at 25 °C. Samples of 0.1
the surface tension when some hydrophobic molecules mg/100 µL in toluene were deposited at the interface.
are adhered at the air/water interface and varies
depending of the molecular concentration and the dimensional-like phase. This behavior is not observed
exposed area in the balance. One of the most important for the F1 and F2 samples, probably because the F1 and
parameters to characterize isotherms is the compress- F2 molecules are less extended and cannot be over-
ibility of the films at the interface. This is defined as18 lapped or folded at the minimum area reached by the
balance.
1 dA Another important parameter of isotherms is the
Cs ) - (1)
A dπ limiting area (A0), which is obtained by extrapolation
at the intersection of the abscissa axis with the tangent
where A is the cross-sectional area and π is the surface at the π-A isotherm at π ) 2 mN/m. We observe a
pressure. This parameter is the inverse of the film regular behavior of A0 for asphaltenes and the four
elasticity  that has been used by other authors.19 fractions: 170 cm2 for asphaltenes, 185 cm2 for the F1
The shape of the π-A isotherms is similar to that fraction, 225 cm2 for the F2 fraction, 260 cm2 for the F3
reported by Zhong et al.1 and Mohammed et al.20 The fraction, and 280 cm2 for the F4 fraction. These values
monolayers behave as a two-dimensional gas phase, indicate the monolayer expansion of the films, and, from
where the molecules are far apart. In this phase, the these results, we conclude that the F4 molecules occupy
surface pressure is very low and when the area is the largest surface area at the air/water interface and,
further compressed, molecules are closer, creating the as suggested by Fauconnier et al.,19 this order could
liquid expanded (LE) phase (at a pressure of ∼5 mN/m correspond to the corresponding order of the hydrody-
when the curves begin to rise). Under further compres- namic radius of the molecules. In contrast, the original
sion, asphaltene molecules are closely packed and the asphaltenes sample seems to occupy the smallest area
resins or alkyl chains are assumed to be in the air and at the interface, compared to the area of the fractions,
more-ordered, building the liquid-condensed (LC) phase probably because of the influence of the more polar and
(at a pressure of ∼30 mN/m in Figure 5). After this abundant F1 fraction at the air/water interface. We
phase, we find the critical area (Ac), where molecules should remember that asphaltenes are built as a
start overlapping and folding, which leads to a new mixture of different proportions of molecules, and the
three-dimensional phase. No abrupt collapse is observed small average molecular area at the air/water interface
for any sample, as in the case of simple lipids such as indicates that small polar molecules dominate, produc-
arachidic or stearic acid monolayers.21 Of course, no ing a more condensed and rigid structure. The hydro-
specific phase transition points were observed in any phobic properties of the asphaltenes and the fractions
case that was analyzed, because of the wide molecular can be compared by observing the corresponding pres-
distribution of the samples. sure at small areas. We notice that the pressure of the
In Figure 5, where the results of the isotherms for films increases according to the following order: as-
the asphaltenes and the four fractions are shown, we phaltenes < F1 < F2 < F3 < F4. This sequence
observe that only fractions F3 and F4 show the presence corresponds to the same order in the limiting area as a
of Ac, which is obtained by intersecting tangent lines of consequence of the increasing hydrophobicity from
the isotherms from the LC phase and from the three- asphaltenes to fraction F4. These results are in cor-
respondence with the results of solubility obtained by
(18) Gaines, G. L., Jr. Insoluble Monolayers at Liquid-Gas Inter-
faces; Wiley-Interscience: New York, 1966. Kaminski et al.,10 showing that fraction F1 and unfrac-
(19) Fauconnier, M. L.; Blecker, C.; Gryone, J.; Razamindralmbo, tionated asphaltene are less toluene soluble than frac-
H.; Vanzerven, E.; Marlier, M.; Paquot, M, J. Agric. Food Chem. 2000, tion F4, probably because more resin molecules remain
48, 2709-2712.
(20) Mohammed, R. A.; Bailey, A. I.; Luckham, P. F.; Taylor, S. E. attached to fraction F4.
Colloids Surf. A 1993, 80 (2-3), 237-242. From the isotherms, we evaluated the compressibility
(21) Valdes-Covarrubias, M. A.; Cadena-Nava, R. D.; Vásquez-
Martinez, E.; Valdes-Pérez, D.; Ruiz-Garcı́a, J. J. Phys. Condens. of the samples in the entire range of measured pres-
Matter 2004, 16, 2097-2017. sures, according to the definition and procedure ex-
Surface Properties of Asphaltene Polar Fractions Energy & Fuels, Vol. 19, No. 6, 2005 2421

Figure 7. Isotherms in the air/water interface of the asphal-

Figure 6. Compressibility of the asphaltenes and the four tene sample and the four fractions treated with EDTA at 25
fractions (F1-F4) for different pressures. Curves were ob- C. Samples of 0.1 mg/100 µL in toluene were deposited at the
tained from eq 1 directly from Figure 4. interface.

plained in eq 1. The results are shown in Figure 6 and is the same as that shown in Figure 5. However, the
Table 2. We found that, for low π, the compressibility values of A0 are larger for each sample: 255 cm2 for
values for the asphaltenes and fraction F1 seems to be asphaltenes, 260 cm2 for F1, 280 cm2 for F2, 290 cm2
higher than that for fractions F2, F3, and F4. For higher for F3, and 360 cm2 for F4. The tendency of the curves
pressures (∼30 mN/m), the compressibility of fraction show that, at 100 cm2, the pressure is higher, compared
F4 takes the largest value (∼0.11 m/mN) and that of with the pressure reached for the samples at the same
fraction F1 is ∼0.05 m/mN. The compressibility values area before using EDTA. This means that EDTA modi-
of asphaltenes and fractions F2 and F3 are almost the fies the samples, thereby increasing the hydrophobic
same at this pressure. These values are larger, com- properties of the asphaltenes and fractions monolayers
pared to the values obtained by Zhang et al.22 using at the air/water interface. It seems that the molecular
asphaltenes extracted from Athabasca oil sands bitu- area of the samples is larger or bigger aggregates are
men. The compressibility values that they determined formed as a consequence of eliminating metal atoms
at the pressure of π ) 30 mN/m were <0.02 m/mN, from asphaltenes and its fractions. Another reason could
which means that our asphaltenes and fractions films be that asphaltenes and EDTA form a complex molec-
are more flexible than those used by Zhang et al.22 This ular system with a larger area. Kaminski et al.10 did
can be due to the method that they used for extracting not find any evidence of such interaction via Fourier
and purifying asphaltenes. We assume they obtained transform infrared spectroscopy; however, a weak in-
asphaltenes that were almost free of resins, producing teraction could be responsible for such an increase in
more-rigid films at the air/water interface. molecular area, as we have already demonstrated while
These results are expected because, at larger areas investigating the interaction of cholesterol and chitosans
(low π), small hydrophilic molecules of fraction F1 do at the air/water interface.23
not touch each other enough, contrary to the expanded It is important to mention that the collapse like phase
molecules of fraction F4, which contribute to a more is now present in all samples, indicating the presence
relatively rigid monolayer. At larger π values, small and of folded or overlapped large molecules, which is due to
rigid molecules of fraction F1 are closer and interact the influence of EDTA molecules.
strongly to produce a more rigid and compact monolayer The compressibility values at 30 mN/m are almost the
than molecules of fraction F4, which probably contain same as those shown in Figure 6, except for fraction F1,
more-flexible resin residues. which, at this pressure, now shows a larger value (0.08
The effect of EDTA on the asphaltenes and fractions m/mN). This result is consistent with the fact that the
F1-F4 is observed in Figure 7. The behaviors of all of average molecular area has grown due to the influence
the isotherms are similar to those before treatment with of EDTA, producing more-flexible films. At low pressure
EDTA. The sequence of the corresponding limiting area (π ) 2 mN/m), the compressibility values are different,

Table 2. Compressibility for the Monolayers of the Asphaltenes and the Four Fractions Obtained by Different Methods
(F1-F4) at the Air/Water Interface at 25 °C
Compressibility (m/mN)
sample pressure (mN/m) asphaltenes F1 F2 F3 F4
untreated sample 2 3.10 4.57 2.1 4.03 3.12
untreated sample 30 0.08 0.05 0.08 0.09 0.11
sample with EDTA 2 4.2 2.97 2.78 4.9 7
sample with EDTA 30 0.09 0.08 0.07 0.07 0.09
sample with CuSO4‚5H2O 2 1.64 3.96 4.8
sample with CuSO4‚5H2O 30 0.03 0.04 0.05
2422 Energy & Fuels, Vol. 19, No. 6, 2005 Grijalva-Monteverde et al.

Conclusions

In this work, we have analyzed the effect of ethyl-

enediamine tetraacetic acid (EDTA) and copper sulfate
treatments on the asphaltenes and some of its polar
fractions. Differences in the metal content of the as-
phaltenes and fractions were observed, with a predomi-
nance of iron and vanadium. After the EDTA treatment,
we observed that the proportions of vanadium, cobalt,
copper, and nickel did not change appreciably. In
contrast, iron and aluminum showed a significantly
decrement in the samples. This difference suggests that
the former metals are being located in central porphy-
rin-like structures, where these metal ions show strong
stability, whereas iron and aluminum are metals that
could be joined with other heteroatoms such as sulfur
and oxygen, showing weaker interactions than the
Figure 8. Isotherms in the air/water interface of the asphal- metal-EDTA interactions. These differences in metal
tene sample and the four fractions treated first with EDTA
and then with CuSO4 at 25 C. Samples of 0.1 mg/100 µL in composition and the method of precipitation created
toluene were deposited at the interface. differences in the zeta potential measurements and
Langmuir films at the air/water interface. Our zeta
compared to the values of the samples without EDTA. potential measurement results, and the behavior of the
All samples, except fraction F1, showed larger com- asphaltenes and the four fractions at the air/water
pressibility values, as can be observed in Table 2. This
interface, indicate the possibility that EDTA molecules
is reasonable for larger areas, in the sense that bigger
remain attached to the different samples. One of the
or more hydrophobic molecules at the air/water interface
most interesting results was the fact that, when samples
remain closer than smaller or hydrophilic ones.
Finally, we discuss the isotherms for the asphaltenes are treated with copper sulfate, the zeta potential values
and fractions F1 and F2 that have been treated with changes sign and becomes positive for all pH analyzed,
copper sulfate. In Figure 8, we show the isotherms of indicating the effect of the metal content. The isotherms
the asphaltenes and fractions F1 and F2. As observed, of the asphaltenes and the four fractions showed dif-
the limiting area A0 does not show important variations, ferences in hydrophobicity and compressibility. The
compared to the areas of the monolayers obtained with hydrophobicity (limiting area) increased in the sequence
the original asphaltenes and fractions. On the contrary, of asphaltenes < F1 < F2 < F3 < F4 before the EDTA
they now appear smaller than the limiting areas shown treatment; after the EDTA treatment, the same se-
for the samples treated with EDTA. The largest differ- quence was preserved but with larger limiting area
ence in limiting area is shown by the isotherm that values, indicating the formation of complexes between
corresponds to asphaltenes (∼100 cm2). This is a very the samples and EDTA molecules and an enhancement
important difference in the hydrophobicity of the samples of the hydrophobic character of the samples. The effect
and the molecular structure at the air/water interface. of the copper sulfate on the isotherms, as expected,
The size of the aggregates of asphaltenes and the showed a decreasing effect on the limiting area of
fractions now seem to be reduced or/and the influence asphaltenes and both fractions F1 and F2, because of
of the electrical charge of the asphaltene-copper and the larger amount of copper in the samples. Other
fraction-copper complexes become the molecules in effects of the copper on the monolayers was the en-
relatively more hydrophilic solutions. The compress- hancement of the monolayer rigidity, in comparison
ibility values shown for the isotherms at π ) 30 mN/m with that observed in samples that were treated with
are shown in Table 2. All these values were smaller, EDTA.
compared to those derived from Figures 5 and 7.
Observe that now our asphaltenes monolayer has
almost the same compressibility as that observed by
Zhang et al.22 It is also noticeable that the buckling (the Acknowledgment. We thank the Departamento de
zone of three-dimensional domains) disappears at higher Investigación en Materiales and Departamento de Fı́sica
π, as a consequence of the increasing rigidity of the at the Universidad de Sonora, where this work was
monolayers, because of the effect of the copper addition performed. We appreciate the help of Dr. Francisco
and the partial elimination of resins. Brown and Silvia Burruel. M.A.V. appreciates the
financial support from the CONACyT, México (under
(22) Zhang, L. Y.; Xu, Z.; Masliyah, J. H. Langmuir 2003, 19, 9730- Grant No. ER074).
9741.
(23) Parra-Barraza, H.; Burboa, M. G.; Sánchez-Vázquez, M.;
Juárez, J.; Goycoolea, F. M.; Valdez, M. A. Biomacromolecules 2005,
in press. EF050120Y