Journal of Petroleum Science and Engineering 188 (2020) 106953

Contents lists available at ScienceDirect

Journal of Petroleum Science and Engineering

journal homepage: http://www.elsevier.com/locate/petrol

Crystalizing the interface – The first X-Ray structure of an oil/surfactant/

brine transition layer
Theis I. Sølling a, *, Kristian B. Olesen a, Gustav Palm-Henriksen a, Safwat Abdel-Azeim a,
Anders B. Skov b
a
Center for Integrative Petroleum Research, College of Petroleum & Geosciences, King Fahad University of Petroleum & Minerals, Dhahran 31261, Saudi Arabia
b
Department of Chemistry, University of Copenhagen, Universitetsparken 5, DK-2100, Copenhagen, Denmark

A R T I C L E I N F O A B S T R A C T

Keywords: Liquid-liquid interfacial phenomena are central in a wide variety of scientific disciplines and encompass down-
Interface to-earth aspects such as mayonnaise production, more complex processes in cellular biology and determine what
Interfacial tension drives the flow below the surface of the earth in an oil reservoir. It is the actual molecular structure at the
Surfactants
boundary between two immiscible liquids that determines the interfacial properties. The experimental deter­
Crystal structure
mination of such a structure has previously been unsuccessful. We have designed an experiment to enable
Zink
crystallization of the interface between brine and a surfactant/organic phase. A single-crystal structure reveals,
for the first time, how the structure and orientation of surfactants impact macroscopic measures such as inter­
facial tension.

1. Introduction can be thought of as molecular unhappiness in the sense that it expresses

Seeing is believing; in almost all scientific disciplines that involve
chemistry there is a desire to go beyond the physical limits that prevent
the observation of real-time and real-space motion of molecules (Zewail,
2000). In, for example, ultrafast sciences this has almost been accom­
plished, at least by inference, with the advent of pulsed and very bright
X-ray sources operating in the femtosecond time domain (Neutze and
Moffat, 2012), which can, in principle, produce snapshots of chemical
reactions while they occur. However, the experimental conditions are
very specific and in reality it is only gaseous unimolecular decomposi­
tion reactions that can be interrogated. Observing dynamic
molecular-level structure or even only molecular structure for that
matter, in liquids is a challenge. The challenge becomes even larger
when it comes to the structural properties of interfaces, that is, right at
the boundary between two immiscible liquids. The chemical bonding
environment and associated chemical structure at interfaces are central
factors in determining macroscopically observable parameters such as
the interfacial tension. Such aspects are also central in a range of cellular
processes and determines for example the pressure profiles in mem­
branes - these are central to anesthesia and for some of the
self-assembling process that are key in membrane signaling and traf­
ficking (Cantor, 1997; Lingwood and Simons, 2010). Interfacial tension
the energetic expense of expanding the interfacial layer - an expansion
gives rise to an enhancement of the repulsive forces the two layers in
between (Andersson et al., 2014). So far, no determination of the
structure of the interfacial layer with Angstrom scale resolution has been
reported. However, several X-ray or neutron scattering studies that
determine the nanometer-scale aspects of the interfacial structure have
been reported in what is mostly small angle or grazing incidence scat­
tering experiments. The results have indeed shed light on the nature of
the interfacial repeat units in brine/oil/surfactant systems and, since
such interfaces do scatter, a high degree of orientation is in play
(Schlossman and Tikhonov, 2014; Scoppol et al., 2015). This structure at
the interface can be perturbed by even small modifications of the sur­
factants that are involved to result in perturbation of the macroscopic
observables such as the interfacial tension between the involved two
phases. Another technique that has been employed in addressing inter­
facial properties is the vibrational sum frequency spectroscopy method
(VSFS); this has primarily been employed to shed light on the structure
and bonding of molecules at aqueous surfaces via their IR fingerprint
(Richmond, 2001; Moore and Richmond, 2008). It is then possible to
obtain an IR spectrum of the interfacial layer and thereby provide input
on the bonding environment exactly at the boundary, but since an IR
spectrum does have a one to one correspondence with structure the

* Corresponding author.
E-mail address: theis.solling@kfupm.edu.sa (T.I. Sølling).

https://doi.org/10.1016/j.petrol.2020.106953
Received 3 November 2019; Received in revised form 16 December 2019; Accepted 11 January 2020
Available online 16 January 2020
0920-4105/© 2020 Elsevier B.V. All rights reserved.
T.I. Sølling et al. Journal of Petroleum Science and Engineering 188 (2020) 106953

method does fully disclose the molecular identity of the interface and an aqueous phase (a model of the oil-surfactant/brine system) was
(Richmond, 2001). Techniques such as atomic force microscopy (AFM) measured in a pendant droplet experiment that involves the formation of
(Manne et al., 1994; Sakai et al., 2016) and transmission electron mi­ a decane droplet containing surfactants into an aqueous solution (Berry
croscopy (TEM) (Schacht et al., 1996) also have been employed to et al., 2015). The analysis of the drop shape provides a value for the
obtain structural information pertaining to interfaces, and TEM has interfacial tension between the two phases and gives a measure of the
gotten very close in a study where the oil-water interface is used as a interactions that takes place at the interface. The experiment is
template for an inorganic growth medium to result in a negative mate­ described in greater detail in the supporting information which is
rial that was shown to be highly structured in a TEM experiment available as an appendix along with the electronic version of the paper.
(Schacht et al., 1996). This, on the other hand, was taken to indicate that The surfactant was chosen as an aliphatic carboxylic acid to mimic the
the oil-water interface has a specific structure too (Schacht et al., 1996). natural surfactants in most crude oils: It is of utmost importance for
The experimentation involving AFM is leading to less conclusive results efficient oil production to understand the underlying mechanisms that
when it comes to the structural properties of the interfaces under study influence interfacial tension. In order to span a range of interfacial
(Manne et al., 1994; Sakai et al., 2016). All of the very comprehensive properties the nature of the acid is systematically varied by sequentially
experimental attempts to obtain the molecular structure of the interfa­ adding CH2 groups as shown in Fig. 1. All compounds are employed in
cial layer between liquids have been backed up by a very large portfolio 1.66 M% relative to decane.
of computational suite of methods to arrive at a deeper atomic level
understanding of the interactions that take place at various interfaces. 2. Results and discussion
Substantial focus has been put on aqueous interfaces for application in
for example catalytic problems and corrosion where the chemistry that 2.1. IFT and electrolytes
takes place is determined by interfacial processes (Bjo €rneholm et al.,
2016; Smit et al., 1991; Shelley and Shelley, 2000). The main theoretical The interfacial tension between the decane/acid phase and pure
approach is molecular dynamics simulations and the methodology will water varies in a remarkable way. It decreases from 30.3 mN/m in the
provide information at the atomic level, however, the experimental case of n ¼ 0 to reach 27.6 mN/m for n ¼ 2, then it goes back up again
validation of the computational results still seems to be lacking. for n ¼ 3, 4. This is taken to indicate a structurally induced turnover at n
¼ 2. The behavior is the exact same when it comes to the involvement of
1.1. The motivation for the experimental procedure dilute LiCl, NaCl and KCl solutions (up to 0.1 M). For higher concen­
trations (here up to 1.0 M) there is a steady IFT decrease; the lowest
Our approach here is to use the potential ordering at a brine/oil- value is found for 1.0 M KCl in the case of the acid where n ¼ 4. The
surfactant interface that was employed by (Schacht et al., 1996) in change in trend as a function of salinity is taken to indicate the presence
forming templates to obtain atomic level information of the of at least two effects impacting the interfaces each in different
oil-surfactant/brine interfacial structure. The specific structure of a directions.
range of aqueous divalent metal ions (M2þ) has been investigated in a
recent computational study where the variation of the acidities of [M
2.2. Crystalization at the interface
(H2O)n]2þ were addressed and related to structure (Jackson et al.,
2015). The idea is that when there is a good match between the elec­
To shed light on how the surfactant structure influences the structure
tronic structure of the incoming cation and the surface active compo­
at the interface we brought forward a novel approach – the Zn2þ ion to
nents that are residing at the oil surface there will be a strong
try to get a hold of the interfacial structure. Already at intermediate
aggregation and potential for crystal formation at the interface. This
concentrations the distinctive nature of Zn2þ reveals itself by giving rise
strategy has been employed by others to amplify X-ray scattering from a
to a lowering of IFT, which is indicating a significant interaction be­
brine/oil-surfactant interface where the surfactant had a phosphorus
tween the electrolyte and the liquid-liquid boundary. Extending the
head group. Here, the adhesion of the electrolyte on to the oil surface is
measurement times in the pendent drop experiments beyond 90 s result
pH dependent because of the solvation structure around the acidic [Sr
in a remarkable finding: We observe crystallization of the interface in
(H2O)n]2þ ion and because of the basic phosphonate group (Bu et al.,
real time, that is, we are able to record an actual movie of the crystal
2014). In the case of Sr2þ the experiment provides direct information
growth with the camera that is being used to observe the drop shape of
about the lamellar structure at the interface but not a real atomic-scale
the decane droplet that is submerged in the brine. An example of such a
structure neither of the surfactants nor of their orientation. If the
movie can be seen in the supporting information and Fig. 2 depicts a
interface had solidified into a crystal form such a crystal could be iso­
series of still images.
lated and analyzed in for example a single crystal diffraction experi­
It is only when the linker between the ring and that acid group at­
ment. Presently, in an attempt to achieve such an interfacial
tains a certain length that the crystallization takes place. It is found that
crystallization to arrive at atomic-scale structure, we employ Zn2þ ions
the crystals form for n ¼ 2–4 whereas n ¼ 0 and 1 clearly do not result in
in the form of ZnCl2. This have been shown to provide the most acidic
crystallization even after extended exposure times. We do admit that we
aqua complex – that is, if one disregards the highly toxic Hg2þ and Be2þ
disregard the fact that Zn2þ could alter the initial brine-surfactant/oil
ions (Jackson et al., 2015). The acidity is to a large extent determined by
the interaction between the lone-pairs of the oxygen in water and the
center metal ion: The more acidic the metal ion, the larger the interac­
tion. Thus the aggregation of the metal ion onto the surfactant con­
taining oil surface will be particularly pronounced for Zn and the
crystallization is most likely to be favorable - the lattice energy of the
interfacial Zn-surfactant crystal counter-balances the solvation energies.
This approach has turned out to be a success as detailed in the following
description of the experiment and its outcome.

1.2. Experiments

We have studied interfacial behavior in a series of measurements

where the interfacial tension (IFT) between a decane/surfactant phase Fig. 1. The surfactants have structures with n ranging from 0 to 4.

2
T.I. Sølling et al.
Fig. 2. Crystalization of the decane/acid-brine interface. The acid in the depicted case has n ¼ 2 and the concentration of Zn2þ is 0.5 M.
interface. However, Zn2þ cannot be the main driver in organizing the
interfacial molecules for crystal formation because in that case the
observed large effect of the acid structure would be surprising. The
change in behavior for the acid with two methylene groups (n ¼ 2) when
it comes to the crystalizing behavior resonates well with the observation
that the measured IFT for the decane/acid-pure water systems also
shows a turnover point at n ¼ 2. The results are pointing towards an
impact interfacial structure with an onset at n ¼ 2 that either reveals
itself in the measurement of the surface stress (reflected in IFT) or by
providing a structure that is favorable for Zn2þ adhesion and subsequent
crystallization. In order determine if surfactant alignment enables
crystallization one would have to deduce the structure at the interface
and in the present study we have done exactly that. We notice that it is
only under the present set of experimental conditions that the crystal­
lization is observed. Neither, adding an aqueous solution of Zn2þ to a
decane solution of the acid nor to an aqueous solution that has been in
touch with the acid results in precipitation.
2.3. Structural determination
By extending the exposure time (Zn2þ(aq) towards decane/acid
phase) to 30 min for the acid with n ¼ 4 and aZn2þ concentration of0.5
Fig. 3. Scattering pattern and (top) Resulting crystal structure (bottom, CCDC, 1917011).
3
Journal of Petroleum Science and Engineering 188 (2020) 106953
M the crystals that form grow to a thickness that is sufficient to enable
isolation. The crystal was isolated from the tip of the needle that hosts
the droplet in the pendent droplet experiment, stored in a beaker and
mounted in a diffractometer. It was shown to scatter the incoming X-rays
quite efficiently indicating a distinct order of the formed crystal. The
solved structure has be deposited with the Cambridge crystrallographic
data center (CCDC, 1917011, see supporting information for details).
The maximum resolution was found to be 0.95 Å, yielding highly
disordered cyclohexane rings, but being well resolved around the most
efficiently scattering Zn2þ atoms. The scattering pattern is shown in
Fig. 3(top) and the experimental and refinement details are described in
the supporting information. Zn2þ ions are found to be hexa-coordinated
with the acid oxygens contributing differently to the binding environ­
ment in the sense that the C– – O and C–O units are still structurally
unique in the complex which is showing that the binding entity is COOH
rather than COO .Hydrogen atoms could not be resolved due to the
relatively low resolution of the crystal data. The Zn2þ complex seems to
further consist of two additional oxygen based ligands which arise from
the aqueous medium. Zn2þ complexes are found in both tetra- and
hexacoordination with neutral ligands favoring the hexacoordination in
perfect agreement with the crystal structure.
T.I. Sølling et al.
Fig. 4. Schematic drawing to illustrate the interpretation of the interplay be­
tween structure and electrolyte concentration. When the larger acid leaves
more room for interaction with the electrolytes at the surface the IFT goes down
as the concentration goes up.
Fig. 5. Molecular dynamics simulation of a Zn crystal with the n ¼ 0 acid (left)
and the n ¼ 4 acid (right) at the decane water interface after 300 ns. The initial
state was a uniformly distributed lattice in both cases. Water is at the bottom
and it can be seen that the water breaks through in the case of the small acid
and results in dissolution.
2.4. Structure property relationship
The structure shows that there is a requirement for a rather flat and
flexible orientation of the organic acids at the interface to match the
most favorable coordination angles for Zn. This is exactly what can be
achieved in the case of the larger acids because the larger fatty chains
allows for more favorable stacking and hydrophobic interaction with the
decane phase while at the same time enabling a hydrophilic interaction
via the COOH group to provide the best compromise and thereby
enabling a capture of the Zn2þ ions because of this favorable structure.
The flat structure for the larger acids is in line with the IFT findings. Acid
1 will have to orient itself quite specifically to ensure that hydrophobic
ring is fully submerged in the decane phase. This leaves little room for
the electrolytes to stabilize the interface as the hydrophilic ends will be
quite close and more electrolytes will not stabilize the system but rather
destabilize the system through coulombic repulsion. In the case of the
longer chain acids, the position of the acid at the interface is as found in
the crystal structure more forgiving, leaving room for more ions to
interact at the interface. This idea is illustrated with the schematic in
Fig. 4 showing that structural parameters, such as packing at the surface,
are important. The surfactant has to be directed very specifically for the
small acid compared to what is the case for the longer system: that leaves
some flexibility to the orientation of the surfactant at the interface.
2.5. A theoretical verification – molecular dynamics (MD)
A molecular dynamics study has been found to be in agreement with
all the experimental findings and has tied the final knot by providing an
4
Journal of Petroleum Science and Engineering 188 (2020) 106953
answer to why the crystal does not immediately dissolve in the sur­
rounding water. The computational details are provided in the sup­
porting information. Fig. 5 shows a model of the crystal interface for n ¼
0 and n ¼ 4 and provides a reason for the lack of solubility of the Zn
crystal once it is formed.
It can be seen that water molecules simply cannot penetrate the
crystal due to hydrophobic interactions in the case of n ¼ 4 whereas this
is not a problem the case of n ¼ 0 where the penetration subsequently
leads to dissolution.
3. Conclusions
The combined body of data including the first atomically resolved
interface structure clearly shows that a wealth of parameters need to
come together to enable the crystallization of an interface. However, by
balancing the hydrophobic and hydrophilic interactions it is indeed
possible to solidify interfaces with all the possible applications and side-
effects it will have. In the present case we were able to design the system
such that the IFT was impacted significantly only by perturbing struc­
tural parameters of a range of very simple and naturally occurring sur­
factants. In the context of increasing the performance of a subsurface
reservoir lowering the IFT of oil against brine will act very favorable
towards increased production, admittedly the lowering here is not
substantial enough to for the latter to be a factor but in the right di­
rection. Controlled crystallization involving naturally occurring surfac­
tants (acids) will provide the means of making oil flow in a
predetermined direction to avoid sections with large water content. The
potential side-effects are for example plugging of otherwise permeable
sections – a challenge that hopefully will be manageable under the
further development of concept.
Author contributions
Theis I. Sølling has been respondsible for scoping the project, ana­
lysing the data and for writing the manuscript. Kristian B. Olesen has
done all the initial labwork and observed the crystalization for the first
time. Gustav Palm-Henriksen has isolated the crystals. Safwat Abdel-
Azeim has done all modelling. Anders B. Skov has solved the X-ray
structure.
Funding
The Center for Integrative Petroleum research is gratefully
acknowledged for financial support along with the Research and
Development (R&D) program (Research Pooling Initiative), Ministry of
Education, Riyadh, Saudia Arabia. KBO and GPH acknowledge the
Danish Chemical Society Travel Fund for travel grants #2019-03 (GPH)
and #2018-26 (KBO)
Acknowledgements
Prof. Sine Larsen and Niels Vissing Holst are gratefully acknowl­
edged for kindly assisting in the acquisition of the x-ray data set and its
analysis:
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.petrol.2020.106953.
References
Andersson, M.P., Bennetzen, M.V., Klamt, A., Stipp, S.L.S., 2014. First-principles
prediction of liquid/liquid interfacial tension. J. Chem. Theory Comput. 10,
3401–3408.
T.I. Sølling et al.
Berry, J.D., Neeson, M.J., Dagastine, R.R., Chan, D.Y.C., Tabor, R.F., 2015. Measurement
of surface and interfacial tension using pendant drop tensiometry. J. Colloid
Interface Sci. 454, 226–237.
Bj€
orneholm, O., Hansen, M.H., Hodgson, A., Liu, L.-M., Limmer, D.T., Michaelides, A.,
Pedevilla, P., Rossmeisl, J., Shen, H., Tocci, G., Tyrode, E., Walz, M.-M., Werner, J.,
Bluhm, H., 2016. Water at interfaces. Chem. Rev. 116, 7698–7726.
Bu, W., Mihaylov, M., Amoanu, D., Lin, B., Meron, M., Kuzmenko, I., Soderholm, L.,
Schlossman, M.L., 2014. X-ray studies of interfacial strontium–extractant complexes
in a model solvent extraction system. J. Phys. Chem. B 118, 12486–12500.
Cantor, R., 1997. The lateral pressure profile in membranes: A physical mechanism of
generalanesthesia. Biochemistry 36, 2339–2344.
Jackson, V.E., Felmy, A.R., Dixon, D.A., 2015. Prediction of the pKa’s of aqueous metal
ion þ2 complexes. J. Phys. Chem. A 119, 2926–2936.
Lingwood, D., Simons, K., 2010. Lipid rafts as a membrane-organizing principle. Science
327, 46–50.
Manne, S., Cleveland, J.P., Gaub, H.E., Stucky, G.D., Hansma, P.K., 1994. Direct
visualization of surfactant hemimicelles by force microscopy of the electrical double
layer. Langmuir 10, 4409–4413.
Moore, F.G., Richmond, G.L., 2008. Integration or segregation: how do molecules behave
at oil/water interfaces? Acc. Chem. Res. 41, 739–748.
Neutze, R., Moffat, K., 2012. Time-resolved structural studies at synchrotrons and X-ray
free electron lasers: opportunities and challenges. Curr. Opin. Struct. Biol. 22,
651–659.
Journal of Petroleum Science and Engineering 188 (2020) 106953
Richmond, G.L., 2001. Structure and bonding of molecules at aqueous interfaces. Annu.
Rev. Phys. Chem. 52, 357–389.
Sakai, K., Okada, K., Misono, T., Endo, T., Abe, M., 2016. Characterizing solid/ionic
liquid interfaces in the presence of water and nonionic surfactants Colloids and
Surfaces A. Physicochem. Eng. Aspects 509, 433–439.
Schlossman, M.L., Tikhonov, A.M., 2014. Molecular ordering and phase behavior of
surfactants at water-oil interfaces as probed by X-ray surface scattering. Annu. Rev.
Phys. Chem. 59, 153–177.
Schacht, S., Huo, Q., Voigt-Martin, I.G., Schuth, F., 1996. Oil-water interface templating
of mesoporous macroscale structures. Science 273, 768–771.
Scoppol, E., Watkins, E., Li Destri, G., Porcar, R., Campbell, A., Konovalov, G.,
Fragneto, O., Diat, O., 2015. Structure of a liquid/liquid interface during solvent
extraction combining X-ray and neutron reflectivity measurements. Phys. Chem.
Chem. Phys. 17, 15093–15097.
Shelley, J.C., Shelley, M.Y., 2000. Computer simulation of surfactant solutions. Curr.
Opin. Colloid Interface Sci. 5, 101–110.
Smit, B., Hilbers, P.A.J., Esselink, K., Rupert, L.A.M., van Os, N.M., Schlijper, A.G., 1991.
Structure of a water/oil interface in the presence of micelles: a computer simulation
study. J. Phys. Chem. 95, 6361–6368.
Zewail, A.H., 2000. Femtochemistry: atomic-scale dynamics of the chemical bond using
ultrafast lasers (Nobel Lecture). Angew. Chem. Int. Ed. 39, 2587–2631.