You are on page 1of 10

Article

pubs.acs.org/Langmuir

Interfacial Rheology of Asphaltenes at Oil−Water Interfaces and


Interpretation of the Equation of State
Jayant P. Rane,†,‡,∥ Vincent Pauchard,§,∥ Alexander Couzis,‡ and Sanjoy Banerjee*,†,‡,∥

Energy Institute, City College of New York, New York, New York 10031, United States

Department of Chemical Engineering, City College of New York, New York, New York 10031, United States
§
Flow Technology Group, Department of Process Technology, SINTEF Materials and Chemistry, Trondheim, Norway

FACE, the Multiphase Flow Assurance Innovation Center, Trondheim, Norway
*
S Supporting Information

ABSTRACT: In an earlier study,1 oil−water interfacial tension was measured


by the pendant drop technique for a range of oil-phase asphaltene
concentrations and viscosities. The interfacial tension was found to be related
to the relative surface coverage during droplet expansion. The relationship was
independent of aging time and bulk asphaltenes concentration, suggesting that
cross-linking did not occur at the interface and that only asphaltene monomers
were adsorbed. The present study extends this work to measurements of
interfacial rheology with the same fluids. Dilatation moduli have been
measured using the pulsating droplet technique at different frequencies,
different concentrations (below and above CNAC), and different aging times.
Care was taken to apply the technique in conditions where viscous and inertial
effects are small. The elastic modulus increases with frequency and then
plateaus to an asymptotic value. The asymptotic or instantaneous elasticity has
been plotted against the interfacial tension, indicating the existence of a unique relationship, between them, independent of
adsorption conditions. The relationship between interfacial tension and surface coverage is analyzed with a Langmuir equation of
state. The equation of state also enabled the prediction of the observed relationship between the instantaneous elasticity and
interfacial tension. The fit by a simple Langmuir equation of state (EOS) suggests minimal effects of aging and of nanoaggregates
or gel formation at the interface. Only one parameter is involved in the fit, which is the surface excess coverage Γ∞ = 3.2
molecules/nm2 (31.25 Å2/molecule). This value appears to agree with flat-on adsorption2 of monomeric asphaltene structures
consisting of aromatic cores composed of an average of six fused rings and supports the hypothesis that nanoaggregates do not
adsorb on the interface. The observed interfacial effects of the adsorbed asphaltenes, correlated by the Langmuir EOS, are
consistent with the asphaltene aggregation behavior in the bulk fluid expected from the Yen−Mullins model.3,4

1. INTRODUCTION membrane;10,14−16 (ii) upon compression/decompression in a


Water-in-oil emulsions are formed in flows through restrictions Langmuir/Blodgett trough, asphaltenes layers exhibit hetero-
like choke valves and pumps during oil production. The water geneous domains which behave as cohesive solids.17 The
in such emulsions must be separated out in order to attain pendant drop technique used in the current study is different
adequate export oil quality. Indigenous components of crude from a conventional Langmuir trough. In pendant drop
oils such as resins,5,6 solidlike waxes, and inorganic particles7−9 measurements, asphaltenes are adsorbed on the interfaces
stabilize such emulsions and hinder gravity separation of the oil slowly by what appears to be a diffusion/barrier controlled
from the water. Asphaltenes, a solubility class of polyaromatic mechanism,1 whereas in the case of the Langmuir trough
species bearing heteroatoms, that naturally occur in crude oil application for the water−air interface, toluene-containing
are a major emulsion stabilizer. Asphaltenes will be of asphaltenes solutions are evaporated on the water surface.
increasing importance as a significant portion of the new oil Such procedures may force nanoaggregates to deposit at the
reserves coming into production are heavy crudes and are of an water surface and perhaps to rearrange as discussed in the
asphaltenic nature (oil sands, heavy oils). literature.18,19 Once toluene is readded to the Langmuir trough,
The mechanism of emulsion stabilization by asphaltenes is the state of the asphaltenes at the interface could be different
thought to be related to the formation of a kind of cross-linked from that in the pendant drop experiments, where considerable
gel phase at the water/oil interface.10−13 This interpretation is
supported by several phenomenological observations: (i) upon Received: December 10, 2012
contraction of water droplets covered by asphaltenes, wrinkles Revised: March 14, 2013
appear that are similar to those seen during buckling of a Published: March 18, 2013

© 2013 American Chemical Society 4750 dx.doi.org/10.1021/la304873n | Langmuir 2013, 29, 4750−4759
Langmuir Article

amounts of asphaltene remain in the bulk solution. The frequency dependence of the dilatational modulus of asphaltene
pendant drop is therefore closer to depicting real systems. covered interfaces26−28 that the complex modulus follows a
Further, Langmuir trough experiments for an oil−water power law dependency over frequency:
interface are difficult to interpret, and asphaltene nano-
|E| ∝ ωn (4)
aggregates can fall onto the interface due to gravity settling
and/or solvophobic interactions between asphaltene molecules The exponent, n, appears to be proportional to the phase shift:
on the interface and nanoaggregates.20 ⎛ E″ ⎞ π
Shear rheology experiments also show significant elastic φ = A tan⎜ ⎟ ≈ n
response after long aging times,6,16,21,22 which is a sign of ⎝ E′ ⎠ 2 (5)
typical solidlike behavior. This behavior influenced previous These observations are in line with the predictions of a
analyses of the dilatational rheology data obtained for rheological model for a 2D gel that is close to the gel point.
asphaltenes.23−31 For example, observation of predominantly On the other hand, a somewhat different argument was made
elastic behavior (the elastic modulus E′ being much larger in by analyzing asphaltenes dilatational rheology data by
most instances to the “dissipative” modulus E″) was generally Szurukowski and Yarranton.29,30 They observed behavior
ascribed to the formation of a membrane with solidlike indicative of the LVDT model. At low asphaltenes concen-
behavior. tration (<50 ppm), the total modulus increased with time and
This interpretation of the results of dilatational rheology concentration and is rather insensitive to frequency. The elastic
experiments is not necessarily the only possible explanation. modulus was high, and the dissipative modulus was low. At
Other adsorbents can generate significant elastic moduli higher asphaltenes concentration (>50 ppm), the total modulus
without cross-linking. By analogy with the Gibbs modulus of increased with frequency and decreased with concentration.
a thin liquid film, Lucassen and Van den Tempel (LVDT) The ratio of the elastic modulus to dissipative modulus tended
established a theory for dilatational modulus based upon to decrease. They interpreted their experimental data with the
diffusional exchange between the interface and bulk solution.32 LVDT model. However, this approach only worked for short
Upon instantaneous expansion/contraction of a droplet, surface aging times (up to about 10 min). At longer times the results
coverage changes, as it is inversely proportional to the area of deviated and was ascribed to gelling. Freer and Radke31 took a
the droplet. This causes an instantaneous change in interfacial similar approach and reached a similar conclusion.
tension following the equation of state, which can be translated In addition to the difficulty in obtaining the parameters in the
into an instantaneous elastic modulus: LVDT model (acknowledged by Szurukowski and Yarranton),
−∂Π it should be recalled that the validity of LVDT model relies on
E0 = Γ the exchange between the adsorbed layer at the interface and
∂Γ (1)
the bulk solution being fully controlled by diffusion. In our
where Π is the surface pressure and Γ is surface coverage. recent study,1 we concluded that the adsorption of asphaltenes
For very slow oscillations, surfactants desorb/adsorb to at the water/oil interface is initially controlled by diffusion, so
maintain equilibrium between the adsorbed layer and bulk the LVDT theory should apply at this early stage. However, we
solution. In this case, the interfacial tension is constant and also found that the adsorption process slowed down as the
dilatational modulus vanishes. For finite frequencies, a dynamic surface coverage increased. Here the dilatational rheology
equilibrium establishes between variations of surface coverage results might deviate from the LVDT theory as concluded by
due to area changes and surfactant exchange. This generates a Szurukowski and Yarranton. We also concluded from the
dissipative (imaginary component) modulus E″ and elastic expansion experiments that interfacial tension was a unique
(real) modulus E′. By integrating the diffusion equations, one function of surface coverage irrespective of adsorption
obtains the LVDT model: conditions (time, viscosity of solvent, and concentration
below and above critical nanoaggregation concentration,
(ωτdiff ) + (ωτdiff /2)1/2
E′ = E 0 CNAC). This conclusion led to the hypothesis that there
1 + (ωτdiff ) + (2ωτdiff )1/2 (2) could be a unique equation of state for asphaltene adsorption in
terms of which the dilatational rheology results could perhaps
(ωτdiff /2)1/2 be explained.
E″ = E 0 To test this hypothesis, a regime of experiments was
1 + (ωτdiff ) + (2ωτdiff )1/2 (3)
undertaken using the same model systems used in our previous
where τdiff ≈ (1/D)(∂Γ/∂c) is the characteristic time of
2
study. The pulsating drop technique was used to measure the
diffusion and ω is the pulsation. interfacial rheological properties taking care to verify that
For low surface coverage (i.e., low surface pressure), τdiff is viscous and inertial effects do not alter measurements.
always very large, and E′ ∼ E0 for all practical frequencies (and Oscillation amplitude and frequency were varied for a range
E″ ∼ 0). When surface coverage and surface pressure increase, of oil viscosities, bulk asphaltene concentrations, and aging
τdiff decreases, E0 increases, and E′ tends to decrease (while E″ times. The elastic and dissipation moduli were derived from
increases). The LVDT model has been verified for a wide these data as discussed later. Of course, increasing amplitude,
variety of surfactants and proteins.32−35 Also, according to frequency, and viscosity also generates hydrodynamic effects,
LVDT, ω increases, E′ ≈ E0, which is called as instantaneous which render the use of the usual pendant drop equation
elasticity of the interface. invalid for extraction of interfacial tension and modulus data
With the LVDT model in mind, we discuss the interpretation from droplet shape analyses. In all cases we must verify that
of dilatational rheology results for asphaltenes containing oil− such hydrodynamic effects remain negligible compared to
water systems. Proof of cross-linking at the oil−water interface surface forces.
is not necessarily provided by the behavior of the elastic Several studies show that asphaltenes are a mixture of many
moduli. Perhaps a supporting argument can be found from the compounds.18,36−38 However, approaches to capture some
4751 dx.doi.org/10.1021/la304873n | Langmuir 2013, 29, 4750−4759
Langmuir Article

aspects of asphaltene behavior, particularly at the nanoscale, are are not accounted for in the calculations of interfacial tension and
often based on the average structure of asphaltene monomers, elastic modulus.
some of which associate to form nanoaggregates, and these in There are several ways to assess the importance of those forces in
turn can form clusters, the so-called Yen−Mullins model.3,4 As the chosen test conditions (up to 3 Hz and viscosities up to 28 cP)
with respect to the use of the Young−Laplace equation for
the current study focuses on macroscopic measurements of interpretation of the pendant drop results.
interfacial behavior, the results obtained primarily elucidate the First, in all test conditions, the profile analysis algorithm was able to
average state and composition of asphaltenes at the interface. fit the droplet shape well. (Droplet images and the Young−Laplace fit
As discussed later, theories for surfactant mixtures39−41 are to them are presented in the Supporting Information.) Second,
difficult to evaluate using these data. experiments have been carried out on the behavior of interfaces
between water and pure fluids (i.e., without asphaltenes) varying both
2. MATERIALS AND METHODS frequency and viscosity. These results are summarized in Table 2. It is
2.1. Chemicals. Chemicals used in this study are the same as
reported in our previous study. The water was deionized Milli-Q Table 2. Interfacial Moduli for Clean Water−Oil Interfaces
grade, with a conductivity of ∼0.05 μS/cm. It was premixed with 43 g/ over the Frequency Range Used in the Study
L of NaCl and 7 g/L of CaCl2, and the final pH was adjusted to 7 with
0.1 M NaOH. The asphaltenes were extracted by precipitation with 20 Nexbase 2002 (6 cP) Nexbase 2008 (28 cP)
vol of n-heptane to 1 vol of crude oil from the Norwegian continental freq (Hz) E′ (mN/m) E″ (mN/m) E′ (mN/m) E″ (mN/m)
shelf, stirring overnight at room temperature, followed by filtration and
0.02 0.68 0.02 0.252 0.102
rinsing with n-heptane. A stock solution of 1000 ppm asphaltenes in
toluene was prepared by sonication for 5 min (Sonics Vibra-Cell, 0.05 0.76 0.11 0.31 0.02
Newtown, CT) and then stored in a sealed glass vial wrapped in 0.1 1.22 0.67 0.529 0.221
aluminum foil to prevent exposure to light. Before use, the stock 0.25 0.45 0.233 0.854 0.233
solution is sonicated again for 5 min. Then the appropriate volume of 0.5 0.74 0.112 0.535 0.306
stock solution is withdrawn and further diluted with toluene before 1.25 1.05 0.532 0.786 0.3
mixing with an aliphatic synthetic oil of the Nexbase oil 2000 series 2.5 1.11 0.23 1.56 −1.34
(Neste Oil, Finland). The obtained asphaltene solution is then 3 0.98 0.002 1.453 −2.22
sonicated for a further 1 min. Asphaltene solutions are discarded after
24 h in order to avoid the effect of long-term phase segregation on
measurements. evident from the table that the moduli for oil−water interfaces without
The choice of the Nexbase 2000 series oil relies on the availability of asphaltenes are very small and when compared to data, presented later,
grades with varying viscosity without varying the chemical nature of on the moduli for asphaltene-bearing interfaces are essentially
the base oil (all are poly-α-olefins but with different chain lengths). negligible. The same range of frequencies was used for the asphaltene
The composition and viscosity of the different mixtures studied here bearing systems so that the pure fluid system could be considered as a
are reported in Table 1. Viscosity was measured using a 2° cone and reference.
In general, the method loses accuracy at high frequencies and
Table 1. Viscosities of Different Nexbase 2000 Series Oils highest viscosity tested. This is likely to be due to hydrodynamic
with 15% Toluene effects associated with the oscillation; i.e., the flow does not remain
purely radial. Such effects can be detected as they lead to negative loss
oil viscosity (Pa·s) moduli values (down to −2 mN/m). Data taken under these high-
85% Nexbase 2002 +15% toluene 0.0065
frequency and high-viscosity conditions are not considered further in
the current study. Only conditions at which we get positive values of
85% Nexbase 2008 +15% toluene 0.0280
moduli of clean interfaces are used. As a note, the elastic modulus
values do not, however, seem to be much impacted by hydrodynamic
plate configuration on a TA Instruments AR 2000 ex rotational effects. At low surface pressure (π = γ0 − γ) it is expected that E′ ∼ E0
rheometer at room temperature (298 K). Asphaltene concentration ∼ π, where E0 is Gibb’s elasticity and Π is the surface pressure. This is
ranged from 10 to 500 ppm, i.e., way below and way above CNAC (80 indeed the case: γ0 ∼ 41 mN/m, γ ∼ 40 mN/m, E′ ∼ 1 mN/m. It is
ppm1). also expected that the effect of hydrodynamic forces will be low on the
2.2. Pulsating Drop Module. Interfacial rheology experiments larger values of moduli measured for asphaltenes containing systems
were performed using pulsating drop module (PD200) in a KSV (both E′ and E″).
tensiometer. Pulsation of the droplet is carried out using a piezo crystal The amplitude dependence of the elastic modulus in systems
system, which is connected to the end of the needle. Amplitude and containing asphaltenes was also checked. For both membranes and
frequency of the sinusoidal oscillation can be monitored using the surfactants (with LVDT type elasticity), the complex dilatational
Attention Theta software by KSV. Droplet shape is analyzed by means modulus, E, should not depend on amplitude at least within the limit
of the usual pendant drop method which is based on the Young− of the linear elastic domain, which is the case with asphaltenes in the
Laplace equation42 to extract interfacial tension. Area and interfacial low-viscosity solvent.45,46 On the other hand, hydrodynamic forces do
tension variations are converted into a complex dilatational modulus depend upon the amplitude of oscillations (i.e., the flow rate of liquid
defined as in and out of the droplet). Figure 1 shows that even up to 3 Hz, elastic
dγ dγ moduli remain unchanged for strains between 2.0 and 4.5%, suggesting
E= =A that hydrodynamic effects are negligible if we stay in this frequency
d ln A dA (6)
range. Further, this also suggests that the stress−strain relationship is
From the phase shift between area and interfacial tension oscillations linear in this range of percent strain. (The amplitude range throughout
Φ, one can calculate the real (elastic) and imaginary (viscous) moduli: this study is kept at 2−3%.)
E′ = |E| cos ϕ , E″ = |E| sin ϕ (7) 2.3. Pendant Droplet as a Langmuir Trough (Expansion
Experiments). To investigate the relationship between interfacial
The use of high frequencies and viscosities could minimize exchange tension and surface coverage in various adsorption conditions, the
between the interface and bulk fluid (so that E′ ∼ E0, i.e., the pendant droplet apparatus is used as a Langmuir trough following a
instantaneous elasticity arising from the equation of state). However method developed for surfactants47,48 and applied to asphaltenes.1
this could lead to a bias in the measurement of interfacial tension due After a certain adsorption time, the droplet is rapidly expanded to the
to increasing hydrodynamic forces (both viscous and inertial43,44) that point of droplet detachment. Interfacial tension is measured as a

4752 dx.doi.org/10.1021/la304873n | Langmuir 2013, 29, 4750−4759


Langmuir Article

Figure 1. Effect of amplitude on the elastic modulus measurements for


15 and 100 ppm after 1 and 3 h, respectively, at extreme frequencies of Figure 2. Frequency dependence of elastic modulus of the interface
0.02−3 Hz. for 15−100 ppm asphaltene-containing model oil solutions for
different aging times.

function of interfacial area. During rapid area expansion (typically 5 s


for a 200% area expansion), it is reasonable to assume that no
asphaltenes are exchanged between the bulk and the interface.
Therefore, the amount of asphaltenes on the droplet surface Γ(t)·A(t)
remains constant, where Γ(t) is the interfacial coverage of asphaltenes
at time t and A(t) is the surface area of the droplet at the same time. If
a reference area is chosen, one can plot interfacial tension vs relative
coverage [Γ(t)/Γ(Aref) = Aref/A(t)] for each test condition. To
compare different test conditions (different adsorption times, different
asphaltenes mass fraction), a common reference interfacial tension γref
is chosen that dictates the choice of the reference area for each test Aref
= A(γref). If interfacial tension is a unique function of surface coverage,
then Γ(Aref) is constant. In turn, all curves of interfacial tension vs
relative coverage [Γ(t)/Γ(Aref) = Aref/A(t)] will overlap. On the
contrary, if adsorbed species undergo relaxation/reorganization over
time, curves will not overlap; instead they will intersect at Γ(t)/Γ(Aref)
= 1.
As reported in the earlier study,1 it has been found that the curves of
interfacial tension vs relative coverage obtained with variable
asphaltenes concentration and adsorption time systematically overlap, Figure 3. Frequency dependence of elastic modulus interface for
revealing a unique dependence of interfacial tension on relative surface higher viscosity bulk oil phase Nexbase 2008 at different
coverage. For the present article, it was further verified that the droplet concentrations of asphaltenes and at different aging times.
expansion with pure oil phase (i.e., no asphaltenes) does not cause any
significant change in interfacial tension. This confirmed that the (Figure 4). Instantaneous elasticity increases with asphaltenes
presented observations arise from the behavior of asphaltenes at the concentration at any particular time. This is another indication
water−oil interface. Some further tests with asphaltenes in higher that the elastic modulus measured at or above 1.25 Hz is not
viscosity oil (Nexbase 2008) were also performed. much impacted by relaxation (and hence corresponds to
instantaneous elasticity). On comparison with Yarranton et
3. RESULTS AND DISCUSSION al.,23,24 the elastic modulus measured by them starts decreasing
3.1. Interfacial Rheology. 3.1.1. Definition of Instanta- from 20 ppm (with some slight variations with aging time,
neous Elasticity. Figure 2 shows the frequency dependence of solvent, and frequency from 0.033 to 0.5 Hz). Besides, the time
the elastic modulus with Nexbase 2002 (6 cP) at 15, 50, and evolution of instantaneous elasticity is very similar to the
100 ppm asphaltene (below and above CNAC) for various dynamic interfacial tension curves presented in ref 1 (and
adsorption time. In all cases, elastic modulus increases with replotted in Figure 5). The rate of initial increase of
frequency and seems to reach a plateau. Such asymptotic instantaneous elasticity increases significantly for concentra-
behavior had been previously reported.29,30,46 In all cases the tions from 10 to 50 ppm but much less from 100 to 200 ppm.
plateau is reached or almost reached at about 1.25 Hz. Figure 3 At 200 ppm the instantaneous elasticity is almost constant after
shows similar trend with the oil of higher viscosity (Nexbase 80 min just like interfacial tension. After 120 min, instantaneous
2008). In all cases plateau is reached or almost reached at 0.5 elasticity is nearly the same (40 mN/m) at 100 and 200 ppm,
Hz. Elastic moduli from now on will be called “instantaneous” just like interfacial tension.
elasticity when measured at 1.25 Hz or above. Similar measurements are performed with higher viscosity oil
3.1.2. Concentration and Time Dependence of Instanta- (Nexbase 2008) (see Figure 6). The main observable difference
neous Elasticity. The elastic modulus was measured over time with the higher viscosity oil is the slower increase in
at 1.25 Hz for concentrations of 10, 20, 50 ppm (below instantaneous elasticity than that of lower viscosity for the
CNAC) and 100, 200 ppm (above CNAC) in Nexbase 2002 same asphaltenes concentration. This is particularly true for 10
4753 dx.doi.org/10.1021/la304873n | Langmuir 2013, 29, 4750−4759
Langmuir Article

Figure 7 shows that renormalized curves superpose at early


times and then diverge: the curve for the lowest viscosity slows

Figure 4. Instantaneous elasticity (f = 1.25 Hz) vs adsorption time for Figure 7. Instantaneous elastic modulus against square root of time/
different concentrations in Nexbase 2002. viscosity for 100 ppm in Nexbase 2002 and Nexbase 2008.

down compared to the other one. Initial superposition can be


seen as an indication of initial diffusion control of adsorption.
Divergence can be seen as an indication of a secondary
mechanism slowing down adsorption, e.g., steric hindrance to
asphaltene adsorption.1
All the elastic modulus data presented in the current study
shows repeatability within experimental uncertainties with a
maximum error of ±5% which occurs at lower coverage.
Typical repeats for frequency and time dependence for 100
ppm are shown in the Supporting Information.
3.2. Instantaneous Elasticity versus Interfacial Ten-
sion. Instantaneous elasticity measured at 1.25 Hz after
variable aging time for variable concentrations above and
below the CNAC vs interfacial tension is plotted in Figure 8

Figure 5. Dynamic interfacial tension for different asphaltene mass


fractions (10−200 ppm) in Nexbase 2002.

Figure 8. Instantaneous elasticity (1.25 Hz) for Nexbase 2002 at


concentrations ranging from 10 to 200 ppm.
Figure 6. Elastic modulus against time for different concentrations of
asphaltenes in Nexbase 2008 at 2 Hz.
(average interfacial tension during oscillations for each
interfacial rheology experiment; it can be verified that
ppm. At 100 ppm the initial increase in elasticity is much faster oscillations do not change the decrease in interfacial tension
at lower viscosity but curves tend to plateau out at longer times. with timesee Supporting Information). All curves superpose
This was already observed for dynamic interfacial tension1 and over the whole experimental range. Furthermore, Figure 9
could be rationalized by renormalizing time by viscosity. The shows that the relaxation modulus vs interfacial tension curves
same renormalization can be done for instantaneous elasticity. also superpose. Quantitatively, it means interfacial tension,
4754 dx.doi.org/10.1021/la304873n | Langmuir 2013, 29, 4750−4759
Langmuir Article

long adsorption time only or continuous oscillations during


adsorption, comprising around 100 experimental points
about 250 points if repeats are included. Again, all curves
collapse within experimental uncertainties.
3.3. Equation of State. 3.3.1. Expansion Experiments.
Figure 11 shows the data of relative coverage Γ(t)/Γ(γref) =

Figure 9. Relaxation modulus (1.25 Hz) for Nexbase 2002 at


concentrations ranging from 10 to 200 ppm.

elastic modulus, and relaxation modulus are the same for


asphaltene-stabilized interface aged for 10 min at 500 ppm and
24 h at 10 ppm. As the interfacial adsorption of asphaltenes
proceeds, interfacial coverage increases, interfacial tension Figure 11. Interfacial tension against relative coverage for all runs
decreases, and elastic modulus increases monotonically. showing collapse of all curves on single master curve.
Therefore, as interfacial tension is a unique function of
surface coverage,1 instantaneous elasticity is also a unique A(γref)/A(t) against the interfacial tension obtained for two oils,
function of surface coverage. Both relationships may be thought different concentrations (below and above CNAC), and
of as fingerprints of an appropriate equation of state (EOS) and different aging times. Various equations of state have been
should yield the same EOS with the same parameters. This will tested, and a Langmuir equation of state is preferred because it
be analyzed in the next section. Furthermore, this means that has fewer adjustable parameters and the parameters that are
nanoaggregates seem to have no influence on elasticity. present are possible to interpret physically:
Figure 10 displays the relationship between instantaneous
elasticity and interfacial tension for a wide variety of γ(Γ) = γ0 + kT Γ∞ ln(1 − Γ/Γ∞) (8)
experimental conditions: different viscosities, different frequen-
which can be written in terms of measurable quantities as
cies above 1.25 Hz, different concentrations, oscillations after a
⎛ Γ/Γ ref ⎞
γ(Γ/Γ ref ) = γ0 + kT Γ∞ ln⎜⎜1 − ⎟⎟
⎝ Γ∞/Γ ref ⎠ (9)
where Γ∞ is maximum interfacial coverage and γ0 is the clean
surface interfacial tension. γ0 = 41 mN/m being known from
the dynamic interfacial tension measurement, we are left with
two unknowns: Γ(Aref)/ Γ∞ and Γ∞. The parameters were
determined by fitting the data by the least-squares method.
Figure 12 shows an example of the fit obtained. The
corresponding standard deviation is 0.164 mN/m (for an
interfacial tension in the 20−40 mN/m range) and seems only
due to uncertainty in the experimental data. Γ(Aref) depends on
the choice of γref, but Γ∞ is remarkbly constant at 3.2
molecules/nm2 (Table 3).
To test the fit, one can fix Γ∞ at various other values and try
to determine Γ(Aref) only. Figure 13 shows the results after
decreasing Γ∞ down to 1.54 molecules/nm2 to illustrate the
quality of the fit. This is one of the lowest value for which
fitting works reasonably over the whole interval. The
corresponding standard deviation is 2.76 mN/m, and the
Figure 10. Instantaneous elasticity modulus versus interfacial tension shape of the curve is changed significantly as compared to the
for all the measurements including different concentrations at fixed
frequency, different frequencies and different viscosities at different
experimental data. There is a systematic deviation in the
aging times and concentrations In legend, for a particular predictions from the experimental results as evident from the
concentration, when aging time is reported; different frequencies are figure.
studied at that aging time and elastic moduli at frequencies higher than 3.3.2. Interfacial Rheology. Using the Langmuir equation of
1.25 Hz are plotted, and when a frequency is reported, elastic moduli is state, one can calculate the Gibbs elasticity (instantaneous
measured over time for that frequency. elasticity) at any surface coverage by the following equation:
4755 dx.doi.org/10.1021/la304873n | Langmuir 2013, 29, 4750−4759
Langmuir Article

nm2 (fit from the expansion experiments discussed earlier) and


Γ∞ = 1.54 molecules/nm2 are both compared to experimental
results in Figure 14. At low surface pressures (higher interfacial

Figure 12. Langmuir equation of state fit to the experimental data of


interfacial tension versus relative coverage (fitting both Γ(Aref) and Γ∞,
Γ∞ = 3.2 molecules/nm2).

Table 3. Γ∞ as a Function of the Reference Interfacial


Tension Chosen for the Fit
Figure 14. Parametric curve instantaneous elasticity vs interfacial
γref (mN/m) Γ∞ (molecules/nm2)
tension from for Langmuir equation of state (optimized Γ∞ to 3.2
21.2 3.22 molecules/nm2 and arbitrary Γ∞ = 1.54 molecules/nm2) and
26.58 3.2 comparison with experimental results.
32.2 3.19
36 3.2 tensions 40−35 mN/m), both curves agree with the
experimental results. This is expected because regardless of
the equation of state and its parameters, instantaneous elasticity
is equal to the surface pressure at low surface coverage. This
result can be obtained by expanding the argument of the
logarithm in eq 11 for Γ → 0 and then using eq 12 for Γ → 0.
From Figure 14, it is clear that E0 → γ0 − γ at low surface
pressures until 5 mN/m. At higher surface pressure, the curve
obtained with Γ∞ = 3.2 molecules/nm2 remains fairly close to
experiments while the curve obtained with Γ∞ = 1.54
molecules/nm2 diverges from the data.
3.3.3. Significance of EOS Parameters. In the Langmuir
equation of state, the excess surface coverage Γ∞ is the inverse
of the cross-section area occupied by the adsorbing species.
Considering Γ∞ = 3.2 molecules/nm2, a cross-section area of
0.3125 nm2 is obtained. This value remains constant over the
range of interfacial tension (20−40 mN/m) studied which
covers bulk asphaltene concentrations well above and below
CNAC, suggesting that asphaltene nanoaggregates are absent
Figure 13. Langmuir equation of state fit to the experimental data of on the interface. An interpretation of this finding can be made
interfacial tension versus relative coverage (fitting Γ(Aref) only and by comparing the estimated cross-section area to the structural
setting Γ∞ = 1.54 molecules/nm2). description of asphaltenes and analysis of their conformation at
water−oil interfaces. Sum frequency generation spectroscopy
−∂γ studies suggest that the polyaromatic core of asphaltenes lies
E0 = Γ flat on the water surface whereas the aliphatic chains stand
∂Γ (10)
perpendicular to the water surface.2 In other words, the cross-
Combined with sectional area of an asphaltene molecule at the oil/water
γ(Γ) = γ0 + kT Γ∞ ln(1 − Γ/Γ∞) interface could be the area of its aromatic core. There is an
(11)
emerging consensus for describing asphaltenes as a core of
yields seven aromatic rings on the average.4 It is reasonable to assume
the area of an aromatic ring is 0.05 nm2, so these estimates
⎡ Γ∞ ⎤ ultimately yield 0.35 nm2 for the asphaltenes core, which is very
E0 = kT Γ⎢ ⎥
⎣ Γ∞ − Γ ⎦ (12) close to the estimate from the equation of state fit.
It can be observed in Figure 14 that the parameters in
It is then possible to calculate a parametric curve of E0 = f(Γ) vs Langmuir equation of state would fit be better at low surface
γ = f(Γ). The resulting curves using either Γ∞ = 3.2 molecules/ pressure (<10 mN/m) with a slightly higher cross-sectional
4756 dx.doi.org/10.1021/la304873n | Langmuir 2013, 29, 4750−4759
Langmuir Article

area. Spectroscopy shows that there is a transition (between 5


and 10 mN/m) from disoriented aliphatic chains to parallel
chains.49 It could mean that at low surface pressure lateral
chains occupy some space on the interface, which could explain
this relatively small effect.
It appears from these data that the simple Langmuir equation
of state can capture the measured interfacial behavior of
asphaltene-bearing systems over the range we have studied.
Asphaltenes are of course mixtures of many compounds. The
existence of a simple equation of state for instantaneous elastic
modulus and interfacial tension with coverage suggests that a
parametrization based on an average monomer structure can
suffice. More complex theories39,41exist to predict these data for
mixtures at interfaces, but these models are difficult to assess
with our macroscopic data.
In more detail, the early model for surfactant mixtures from Figure 15. Schematic of asphaltene stabilized oil−water interface
Joos50 has recently been demonstrated to work for binary showing that asphaltenes monomers with polyaromatic core size
mixtures. A simplified version for more complex mixtures has corresponding to the Yen−Mullins model adsorbed “flat on”. In this
picture nanoaggregates and clusters do not adsorb as readily because of
even been developed by Fainerman and Miller,40 but it has only protruding aliphatic chains, which must be accommodated, as
been validated for conditions at which elastic modulus is close illustrated in the figure.
to instantaneous elasticity. In any case, these models require a
priori knowledge of all parameters (surface excess coverage and
adsorption parameter) for each of the individual components in
the mixture, which is difficult to obtain for asphaltenes and that the primary adsorbing species and interface stabilizer
seems to be monomeric asphaltenes only.


which we do not have for our systems.

4. CONCLUSION ASSOCIATED CONTENT


The dilatational rheology of asphaltenes adsorbed at model *
S Supporting Information

oil−water interfaces was studied using the oscillating pendant Droplet images and the Young−Laplace fit to them. Graphs
drop technique, care being taken to ensure that the range of which show repeatability of moduli. This material is available
free of charge via the Internet at http://pubs.acs.org.


conditions was such that the technique was applicable. The
elastic modulus is found to be a function of frequency and
AUTHOR INFORMATION
adsorption conditions (concentration and aging time). For each
test condition, the elastic modulus was found to reach a plateau Corresponding Author
with increasing frequency (the instantaneous elasticity). The *E-mail banerjee@che.ccny.cuny.edu.
plateau value was found to be a unique function of interfacial Notes
The authors declare no competing financial interest.


tension, which in turn is a unique function of surface coverage.
These findings agree with aspects of the LVDT model at short
adsorption times and also reveal a unique equation of state ACKNOWLEDGMENTS
irrespective of adsorption conditions. The fit of the relationship This work is performed at the City College of New York under
between interfacial tension and surface coverage by a Langmuir the auspices of the FACE centrea research cooperation
equation of state allowed prediction of the relationship between between IFE, NTNU, and SINTEF. The centre is funded by
instantaneous elasticity and interfacial tension. This suggests The Research Council of Norway and by the following
that cross-linking is absent. The corresponding parameter (Γ∞) industrial partners: Statoil ASA, GE Oil & Gas, Scandpower
corresponds well to the expected size and interfacial Petroleum Technology AS, FMC, CD-adapco, and Shell
conformation of asphaltenes molecules adsorbed with their Technology Norway AS.
polyaromatic cores lying flat on the interface and the aliphatic The authors thank Professor Charles Malderelli for useful
chains perpendicular. This interpretation of the results is discussions and PhD student Sharli Zarkar for help with the
consistent with the picture proposed in the Yen−Mullins model experiments.
for asphaltene aggregation in the bulk fluid. The EOS fits
suggest that asphaltene monomers adsorb at the oil−water
interface but the nanoaggregates do not. Perhaps this is because
■ REFERENCES
(1) Rane, J. P.; Harbottle, D.; Pauchard, V.; Couzis, A.; Banerjee, S.
these nanoaggregate structures, as visualized in the Yen− Adsorption Kinetics of Asphaltenes at the Oil-Water Interface and
Mullins model, have the polyaromatic cores in the interior with Nanoaggregation in the Bulk. Langmuir 2012, 28 (26), 9986−9995.
the aliphatic legs protruding outward into the oil phase (see (2) Andrews, A. B.; McClelland, A.; Korkeila, O.; Demidov, A.;
Figure 15). Such nanoaggregate structures would be expected Krummel, A.; Mullins, O. C.; Chen, Z. Molecular Orientation of
to adsorb with more difficulty than the monomers, which could Asphaltenes and PAH Model Compounds in Langmuir-Blodgett Films
Using Sum Frequency Generation Spectroscopy. Langmuir 2011, 27
explain our findings in the framework for the bulk fluid (10), 6049−6058.
behavior in the Yen−Mullins model.3,4 Our results suggest that (3) Mullins, O. C. The Modified Yen Model. Energy Fuels 2010, 24
nanoaggregates, if at all seen on the interface, might be because (4), 2179−2207.
of solvophobic interactions between aliphatic chains of (4) Mullins, O. C.; Sabbah, H.; Eyssautier, J.; Pomerantz, A. E.; Barre,
monomer monolayer of asphaltenes at the interface with L.; Andrews, A. B.; Ruiz-Morales, Y.; Mostowfi, F.; McFarlane, R.;
those of nanoaggregates and clusters (Figure 15). It appears Goual, L.; Lepkowicz, R.; Cooper, T.; Orbulescu, J.; Leblanc, R. M.;

4757 dx.doi.org/10.1021/la304873n | Langmuir 2013, 29, 4750−4759


Langmuir Article

Edwards, J.; Zare, R. N. Advances in Asphaltene Science and the Yen- (24) Bauget, F.; Langevin, D.; Lenormand, R. Dynamic Surface
Mullins Model. Energy Fuels 2012, 26 (7), 3986−4003. Properties of Asphaltenes and Resins at the Oil-Air Interface. J. Colloid
(5) Sjoblom, J.; Aske, N.; Auflem, I.; Brandal, O.; Havre, T.; Saether, Interface Sci. 2001, 239 (2), 501−508.
O.; Westvik, A.; Johnsen, E.; Kallevik, H. Our current understanding of (25) Aske, N.; Orr, R.; Sjoblom, J. Dilatational Elasticity Moduli of
water-in-crude oil emulsions. Recent Characterization Techniques and Water/Crude Oil Interfaces Using the Oscillating Pendant Drop. J.
High Pressure Performance. Adv. Colloid Interface Sci. 2003, 100−102, Dispersion Sci. Technol. 2002, 23 (6), 809−825.
399−473. (26) Bouriat, P.; El Kerri, N.; Graciaa, A.; Lachaise, J. Properties of a
(6) Spiecker, P.; Kilpatrick, P. Interfacial Rheology of Petroleum Two-Dimensional Asphaltene Network at the Water-Cyclohexane
Asphaltenes at the Oil-Water Interface. Langmuir 2004, 20 (10), Interface Deduced from Dynamic Tensiometry. Langmuir 2004, 20
4022−4032. (18), 7459−7464.
(7) Bridie, A.; Wanders, T.; Zegveld, W.; Heijde, H. v. d. Formation, (27) Dicharry, C.; Arla, D.; Sinquin, A.; Graciaa, A.; Bouriat, P.
Prevention and Breaking of Sea Water in Crude Oil Emulsions Stability of Water/Crude Oil Emulsions Based on Interfacial
‘Chocolate Mousses’. Mar. Pollut. Bull. 1980, 11 (12), 343−348. Dilatational Rheology. J. Colloid Interface Sci. 2006, 297 (2), 785−791.
(8) Papirer, E.; Bourgeois, C.; Siffert, B.; Balard, H. Chemical Nature (28) Pauchard, V.; Sjöblom, J.; Kokal, S.; Bouriat, P.; Dicharry, C.;
and Water Oil Emulsifying Properties of Asphaltenes. Fuel 1982, 61 Müller, H.; al-Hajji, A. Role of Naphthenic Acids in Emulsion
(8), 732−734. Tightness for a Low-Total-Acid-Number (TAN)/High-Asphaltenes
(9) Rondon, M.; Bouriat, P.; Lachaise, J.; Salager, J. Breaking of Oil. Energy Fuels 2008, 23 (3), 1269−1279.
Water-in-Crude Oil Emulsions. 1. Physicochemical Phenomenology of (29) Sztukowski, D.; Yarranton, H. Rheology of Asphaltene -
Demulsifier Action. Energy Fuels 2006, 20 (4), 1600−1604. Toluene/Water Interfaces. Langmuir 2005, 21 (25), 11651−11658.
(10) Yarranton, H. W.; Hussein, H.; Masliyah, J. H. Water-in- (30) Yarranton, H. W.; Sztukowski, D. M.; Urrutia, P. Effect of
Hydrocarbon Emulsions Stabilized by Asphaltenes at Low Concen- Interfacial Rheology on Model Emulsion Coalescence: I. Interfacial
trations. J. Colloid Interface Sci. 2000, 228 (1), 52−63. Rheology. J. Colloid Interface Sci. 2007, 310 (1), 246−252.
(11) Siffert, B.; Rageul, P.; Papirer, E. Acid-Base Characteristics of (31) Freer, E. M.; Radke, C. J. Relaxation of Asphaltenes at the
Asphaltenes Isolated from Mechanical Sheared Oil Distillation Toluene/Water Interface: Diffusion Exchange and Surface Rearrange-
Residues. Fuel 1996, 75 (14), 1625−1628. ment. J. Adhes. 2004, 80 (6), 481−496.
(12) Nordli, K. G.; Sjoblom, J.; Kizling, J.; Stenius, P. Water-in-Crude (32) Lucassen, J.; Van Den Tempel, M. Dynamic Measurements of
Oil Emulsions from the Norwegian Continental Shelf 4. Monolayer Dilational Properties of a Liquid Interface. Chem. Eng. Sci. 1972, 27
Properties of the Interfacially Active Crude Oil Fraction. Colloids Surf. (6), 1283−1291.
1991, 57 (1), 83−98. (33) Lucassen-Reynders, E. H.; Cagna, A.; Lucassen, J. Gibbs
(13) Sheu, E. Y.; De Tar, M. M.; Storm, D. A.; DeCanio, S. J. Elasticity, Surface Dilational Modulus and Diffusional Relaxation in
Aggregation and Kinetics of Asphaltenes in Organic Solvents. Fuel Nonionic Surfactant Monolayers. Colloids Surf., A 2001, 186 (1,Ä i2), ̀
1992, 71 (3), 299−302. 63−72.
(14) Khristov, K.; Taylor, S. D.; Czarnecki, J.; Masliyah, J. Thin (34) Lucassen-Reynders, E. H.; Benjamins, J.; Fainerman, V. B.
Liquid Film Technique-Application to Water-Oil-Water Bitumen Dilational Rheology of Protein Films Adsorbed at Fluid Interfaces.
Emulsion Films. Colloids Surf., A 2000, 174 (1−2), 183−196. Curr. Opin. Colloid Interface Sci. 2010, 15 (4), 264−270.
(15) Yeung, A.; Zhang, L. Shear Effects in Interfacial Rheology and (35) Lucassen-Reynders, E. H.; Fainerman, V. B.; Miller, R. Surface
Their Implications on Oscillating Pendant Drop Experiments. Dilational Modulus or Gibbs’ Elasticity of Protein Adsorption Layers.
Langmuir 2006, 22 (2), 693−701. J. Phys. Chem. B 2004, 108 (26), 9173−9176.
(16) Mohammed, R. A.; Bailey, A. I.; Luckham, P. F.; Taylor, S. E. (36) Andersen, S. I.; Birdi, K. S. Aggregation of Asphaltenes As
Dewatering of Crude Oil Emulsions 2. Interfacial Properties of the Determined by Calorimetry. J. Colloid Interface Sci. 1991, 142 (2),
Asphaltic Constituents of Crude Oil. Colloids Surf., A 1993, 80 (3), 497−502.
237−242. (37) Rogel, E.; Carbognani, L. Density Estimation of Asphaltenes
(17) Fan, Y.; Simon, S.; Sjoeblom, J. Influence of Nonionic Using Molecular Dynamics Simulations. Energy Fuels 2003, 17 (2),
Surfactants on the Surface and Interfacial Film Properties of 378−386.
Asphaltenes Investigated by Langmuir Balance and Brewster Angle (38) Bardon, C.; Barre, L.; Espinat, D.; Guille, V.; Li, M. H.;
Microscopy. Langmuir 2010, 26 (13), 10497−10505. Lambard, J.; Ravey, J. C.; Rosenberg, E.; Zemb, T. The Colloidal
(18) Zhang, L. Y.; Lawrence, S.; Xu, Z.; Masliyah, J. H. Studies of Structure of Crude Oils and Suspensions of Asphaltenes and Resins.
Athabasca Asphaltene Langmuir Films at Air-Water Interface. J. Colloid Fuel Sci. Technol. Int. 1996, 14 (1−2), 203−242.
Interface Sci. 2003, 264 (1), 128−140. (39) Ravera, F.; Ferrari, M.; Santini, E.; Liggieri, L. Influence of
(19) Orbulescu, J.; Mullins, O. C.; Leblanc, R. M. Surface Chemistry Surface Processes on the Dilational Visco-elasticity of Surfactant
and Spectroscopy of UG8 Asphaltene Langmuir Film, Part 1. Langmuir Solutions. Adv. Colloid Interface Sci. 2005, 117 (1−3), 75−100.
2010, 26 (19), 15257−15264. (40) Fainerman, V. B.; Aksenenko, E. V.; Petkov, J. T.; Miller, R.
(20) Zhang, L.; Xu, Z.; Masliyah, J. Characterization of Adsorbed Adsorption Layer Characteristics of Multi-component Surfactants
Athabasca Asphaltene Films at Solvent-Water Interfaces Using a Solutions. Soft Matter 2010, 6 (19), 4694−4700.
Langmuir Interfacial Trough. Ind. Eng. Chem. Res. 2005, 44 (5), 1160− (41) Miller, R.; Ferri, J. K.; Javadi, A.; Kraegel, J.; Mucic, N.;
1174. Wuestneck, R. Rheology of Interfacial Layers. Colloid Polym. Sci. 2010,
(21) Eley, D. D.; Hey, M. J.; Lee, M. A. Rheological Studies of 288 (9), 937−950.
Asphaltene Films Adsorbed at the Oil/Water Interface. Colloids Surf. (42) Barnes, G.; Gentle, I. Interfacial Science: An Introduction; Oxford
1987, 24 (23), 173−182. University Press: New York, 2005; p 247.
(22) Acevedo, S. c.; Escobar, G. n.; Gutierrez, L. B.; Rivas, H.; (43) Freer, E. M.; Wong, H.; Radke, C. J. Oscillating Drop/Bubble
Gutierrez, X. Interfacial Rheological Studies of Extra-Heavy Crude Oils Tensiometry: Effect of Viscous Forces on the Measurement of
and Asphaltenes: Role of the Dispersion Effect of Resins in the Interfacial Tension. J. Colloid Interface Sci. 2005, 282 (1), 128−132.
Adsorption of Asphaltenes at the Interface of Water-in-Crude Oil (44) Alexandrov, N.; Marinova, K. G.; Danov, K. D.; Ivanov, I. B.
Emulsions. Colloids Surf., A 1993, 71 (1), 65−71. Surface Dilatational Rheology Measurements for Oil/Water Systems
(23) Freer, E. M.; Svitova, T.; Radke, C. J. The Role of Interfacial with Viscous Oils. J. Colloid Interface Sci. 2009, 339 (2), 545−550.
Rheology in Reservoir Mixed Wettability. J. Pet. Sci. Eng. 2003, 39 (1), (45) Jafari, M. Interfacial Rheology of Asphaltenes in Water-in-
137−158. Hydrocarbon Emulsions. MSc, University of Calgary, Calgary, 2005.

4758 dx.doi.org/10.1021/la304873n | Langmuir 2013, 29, 4750−4759


Langmuir Article

(46) Quintero, C. G.; Noik, C.; Dalmazzone, C.; Grossiord, J. L.


Formation Kinetics and Viscoelastic Properties of Water/Crude Oil
Interfacial Films. Oil Gas Sci. Technol. 2009, 64 (5), 607−616.
(47) Kumar, N.; Couzis, A.; Maldarelli, C. Measurement of the
Kinetic Rate Constants for the Adsorption of Superspreading
Trisiloxanes to an Air/Aqueous Interface and the Relevance of
These Measurements to the Mechanism of Superspreading. J. Colloid
Interface Sci. 2003, 267 (2), 272−285.
(48) Pan, R.; Green, J.; Maldarelli, C. Theory and Experiment on the
Measurement of Kinetic Rate Constants for Surfactant Exchange at an
Air/Water Interface. J. Colloid Interface Sci. 1998, 205 (2), 213−230.
(49) Orbulescu, J.; Mullins, O. C.; Leblanc, R. M. Surface Chemistry
and Spectroscopy of UG8 Asphaltene Langmuir Film. Part 2. Langmuir
2010, 26 (19), 15265−15271.
(50) Joos, P.; Fainerman, V. B. Dynamic Surface Phenomena; VSP:
Utrecht, 1999.

4759 dx.doi.org/10.1021/la304873n | Langmuir 2013, 29, 4750−4759

You might also like