Article

Cite This: J. Phys. Chem. C 2019, 123, 12753−12761 pubs.acs.org/JPCC

Quantum Mechanical Prediction of Wettability of Multiphase

Fluids−Solid Systems at Elevated Temperature
Jin You Lu,†,§ Qiaoyu Ge,†,‡,§ Aikifa Raza,† and TieJun Zhang*,†,‡
†
Department of Mechanical Engineering, Masdar Institute, Khalifa University of Science and Technology, P.O. Box 54224, Abu
Dhabi, UAE
‡
Alibaba Cloud-Khalifa University Joint Innovation Laboratory of Artificial Intelligence for Clean Energy, Abu Dhabi, UAE
*
S Supporting Information
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

ABSTRACT: Physiochemical insights into solid−liquid in-

terfaces are essential for characterizing surface wettability and
multiphase fluid behaviors in diverse applications. We propose
a first-principles approach to predict the polar and thermal
Downloaded via KHALIFA UNIV on May 11, 2020 at 09:55:05 (UTC).

effects on wetting properties of crystalline surfaces for a variety

of polar or nonpolar liquids. By directly applying the approach
to multiphase systems, we simultaneously predict the
macroscopic contact angles, the work of adhesion at the
solid−liquid interface, and the interfacial tension at the
liquid−liquid interfaces. A unique feature of our approach lies
in its capability of quantifying the electrostatic interaction at
the interfaces of polar liquids and solid surface. Our results
reveal a linear relation between the adsorption energy and the
electrostatic interaction at the solid−polar liquid interface, which provides a more effective prediction than classical surface free
energy calculations. By using quantum molecular dynamics simulation, we predict the variation of surface wettability in
multiphase systems at elevated temperature and validate them with experiments. This approach opens a new avenue to probe
the mechanism of sophisticated wetting phenomena in multiphase systems with direct quantum mechanical simulation.

■ INTRODUCTION
Wettability of a solid surface for different liquids plays a
determinant role in many real-world applications, such as
crystal growth,1 material fabrication,2 condensation heat
transfer,3−5 water harvesting,6,7 oxygen electrocatalysis,8 and
enhanced oil recovery.9,10 Although surface wettability can be
experimentally characterized by measuring the contact angle
on different surfaces, the underlying mechanism of intrinsic
wettability and its alteration needs to be revealed at an atomic
level by looking into the polar/nonpolar interactions at the
solid−liquid interfaces. Plenty of work has been done to
investigate the effect of surface polarity, electronic structure,
and roughness on the wetting performance of graphene,11−14
ceramics,15−17 crystals,10,18 and liquid metal−solid interfa-
ces19−21 owing to their significance in various applications.
Surface hydrophobicity is determined not only by surface
energy but also by surface polarity and solid−liquid interfacial
interaction.16,22 Most ceramic and metal oxides are hydrophilic
due to the hydrogen bonding with interfacial water molecules.
However, rare-earth oxides exhibit intrinsic hydrophobicity
owing to their unique electronic structures, which minimize
the interfacial polar interaction.16,23 Similarly, polar interaction
plays an important role in the wettability of crystals, e.g.,
calcite. Calcite surface exhibits intrinsic hydrophilicity resulting
from the electrostatic interaction between calcium ions and
interfacial water molecules.10
Understanding the surface wettability and interfacial
interaction of liquid−liquid−solid multiphase systems is
essential for many applications such as condensation on an
oil-infused surface24 and enhanced oil recovery.10,25 Recently,
density functional theory (DFT)-based quantum molecular
dynamics (QMD) simulation26−28 has been carried out to
predict the wetting properties of different surfaces, such as
oxide surfaces with defects,29 ion-substituted mineral surfa-
ces,30 and ionic water on mineral surfaces.31 For liquid−liquid
interactions, DFT was used to predict interfacial tensions for
various binary liquid−liquid interfaces with assistance of a
COSMO-RS implicit solvation model.32 For interactions at the
solid−liquid interfaces, a general methodology33 was devel-
oped to predict the surface wettability for different liquids
directly through the standard DFT simulation. DFT simulation
was also used to address the effect of metallic bonding on the
wetting performance of the interface between gallium-based
liquid metal and its intermetallic layer.20,21 It was shown that
liquid gallium spread on CuGa2 and FeGa3 surfaces due to the
metallic bond interaction, although it partially wet pure Cu and
Fe surfaces. However, these studies considered either a solid−
liquid or a liquid−liquid interface only under a thermodynami-
Received: January 30, 2019
Revised: May 1, 2019
Published: May 2, 2019

© 2019 American Chemical Society 12753 DOI: 10.1021/acs.jpcc.9b00937

J. Phys. Chem. C 2019, 123, 12753−12761
The Journal of Physical Chemistry C Article

Figure 1. DFT-based simulation of water−calcite, ethanol−calcite, and decane−calcite systems. (a) Schematic diagram to predict the surface
wettability of a solid for a liquid. The adhesion energy USL at the solid−liquid interface and the cohesion energy ULV/2 within the liquid layers can
be predicted using the DFT solid−liquid slab model. (b) Adsorption energy at different liquid layers for two polar and one nonpolar liquids: water,
ethanol, and decane. The adsorption energy at the first liquid layer is equal to the adhesion energy at the solid−liquid interface: USL = Um=1
ad . As the
number of liquid layers is greater than 1, the adsorption energy is essentially the cohesion energy at the liquid−liquid interface: ULV/2 = Um>1
ad . (c−
e) Charge density difference distribution of polar water, ethanol, and nonpolar decane, respectively.

cally stable condition.34,35 Thermal effect on surface wettability
is another major factor in practical multiphase systems.
Previous experimental25 and computational10 studies on an
oil−water−mineral system have shown that carbonate
becomes more water-wet with increasing temperature, which
is applaudable for enhanced oil recovery. Specifically, for
ionized water, this positive effect is more pronounced owing to
the relatively larger polar interaction at elevated temperature.10
Therefore, a comprehensive method is needed to predict the
wetting properties of multiphase systems involving polar
methods.42,43 Moreover, we investigate the thermal effect in
the decane−water−calcite multiphase system by using the
QMD simulation. The temperature-dependent surface wett-
ability is predicted by inputting the QMD-predicted time-
averaged adsorption energy into the Young−Dupré equation.
Our QMD predictions agree well with the experimental
measurements at elevated temperature. The proposed
quantum mechanical approach provides a new way to
comprehensively predict the wetting properties of liquid−
liquid−solid multiphase systems at elevated temperature.

interactions between solid and two different liquids at elevated
temperature.
In this work, we extract the surface free energy components
of the calcite (10.4) surface from the DFT-predicted calcite
surface wettability. Our results show that classical surface free
energy calculations36,37 are applicable to predict the wetting
properties of the calcite surface for nonpolar liquids; however,
the prediction results are not effective for polar liquids due to
complicated polarized charge differences at the solid−liquid
interface. Therefore, we take advantage of the first-principles
DFT simulation to quantify the effect of polar interaction on
the mineral surface wettability with Bader charge analysis38,39
for various polar/nonpolar liquids, including water, ethylene
glycol, ethanol, glycerol, and decane. Our results reveal that
surface adhesion correlates linearly with electrostatic inter-
action between polarized charges of calcite surface and
different polar liquids. To obtain more physical insights into
oil−water−mineral multiphase systems, we perform the DFT-
based quantum simulation for decane−water−calcite and
decane−water−dolomite systems and demonstrate that the
predicted adsorption energy at all of the interfaces can be
linked to the macroscopic work of adhesion at the solid−liquid
interface and interfacial tension between the two liquids.
Meanwhile, another multiphase system of mercury−water−
calcite is simulated to consider the effect of metallic bonding.
The interaction at the mercury−water interface plays an
important role in the description of a number of electro-
chemical phenomena involving voltammetry and amperom-
etry.40,41 In fact, both oil−water−mineral and mercury−
water−mineral systems are valuable for the applications in
oil recovery and mercury injection capillary pressure
■
RESULTS AND DISCUSSION
Polar Effect on Adsorption Energy and Wettability.
We start with performing DFT simulation for two polar and
one nonpolar liquids: water, ethanol, and decane on calcite
(10.4) surface through the methodology we previously
proposed.33 Calcite crystal is a common constituent of
sedimentary rocks, particularly limestone. The reason for
choosing the (10.4) plane of calcite crystal as the substrate
surface is that it is the most thermodynamically stable and
most commonly studied compared to other planes. All DFT
simulations are performed with the Quantum Espresso
package.44,45 To keep the completeness, the method is briefly
introduced here. The computational details and data for the
cleaved calcite (10.4) surface are described in Sections S.I−
S.III (Supporting Information). In the following context,
calcite (10.4) is referred to calcite in short. The simulation is
based on a slab model of the solid surface and multiple layers
of liquids, as shown in Figure 1a. The liquid is added to the
solid surface layer by layer. Configurations for the multiple
liquid layers on the calcite surface for different liquids are
shown in Section S.IV. For a system with m layers of liquids,
the adsorption energy per molecule for the mth layer of liquid
(Umad) is calculated by subtracting the total energy of a single
layer of liquid (UmL) and that of the solid associated with (m −
1) layers of liquid (US+(m−1)L) from the total energy of the
whole system (US+mL)
US + mL − US + (m − 1)L − UmL
m
Uad =
nL (1)

12754 DOI: 10.1021/acs.jpcc.9b00937

J. Phys. Chem. C 2019, 123, 12753−12761
The Journal of Physical Chemistry C Article

where nL is the number of molecules in one liquid monolayer. Table 1. DFT-Predicted Contact Angle and Analytical
When m = 1 (first layer of liquid), the adsorption energy is Solutions Compared to Experimental Data for Various
equal to the adhesion energy at the solid−liquid interface: USL Polar/Nonpolar Liquids on Calcite Surface
= Um=1
ad . As the number of liquid layers is greater than one, the
γ analytical DFT experiment
adsorption energy is essentially the cohesion energy at the liquids (mN/m) (deg) (deg) (deg)
liquid−liquid interface: ULV/2 = Um>1 ad . Figure 1b shows the
water 72.80 40.40 37.70 40.80 ± 4.00
adsorption energy of different liquid layers for water, ethanol,
ethanol 22.40 spread spread spread
and decane. The adhesion energies per liquid molecule for
decane 23.83 spread spread spread
water, ethanol, and decane are −96.63, −87.60, and −88.27
glycerol 64.00 40.70 62.90 61.10 ± 3.00
kJ/mol, respectively. For both water and ethanol, the cohesion
ethylene glycol 48.00 spread 72.30 61.92 ± 3.00
energy is obtained by taking the average magnitude of the
benzene 28.90 spread spread
adsorption energy from the third to seventh layer. However,
methanol 22.50 spread spread
for nonpolar decane, the cohesion energy is taken from the cyclohexane 25.24 spread spread
average magnitude of the adsorption energy from the third to chlorobenzene 33.60 spread spread
fourth layer. In general, the accuracy of the cohesion energy
can be improved by increasing the number of liquid layers.
However, for the case of decane (C10H22), due to considerably respectively. The charge density difference distributions in
higher computational efforts, we simulated up to four layers. Figure 1c,d show that the interaction between water/ethanol
The original simulation cell is increased by doubling the cell and calcite is dominated by the intermolecular interaction of
length in the y-direction to facilitate the relatively long-chain the dipole in water/ethanol and calcium ions in the calcite
hydrocarbon liquid. In fact, owing to the nonpolar nature of surface. The polar interaction between decane and calcite is
decane, the fluctuation of the cohesion energy is much less extremely weak due to the dominant van der Waals
than that of polar water and ethanol. The relatively high interactions. The stronger adsorption for polar liquid
fluctuation for ethanol can potentially be minimized by molecules on the calcite surface is also reflected by smaller
increasing the size of the simulation cell. interfacial atomic distances: 2.07 and 2.20 Å between calcium
According to the Young−Dupré equation: WSL = γLV(1 − and oxygen for water and ethanol, respectively, and 3.07 Å
cos θ), the contact angle θ can be calculated as in ref 33 once between calcium and carbon for decane, as marked in Figure
the work of adhesion WSL and the surface tension of the liquid 1c−e. More details of the charge difference distribution can be
γLV are known. Here, the work of adhesion WSL = USL/A, liquid found in Figure S2 (Section S.VI of the Supporting
surface tension γLV = ULV/2A, where A is the interaction area Information).
at the interfaces. The contact angle can be obtained from the Partial-density-of-state (PDOS) calculations on one layer of
DFT-predicted adsorption energy Umad despite difficulty in the water with and without sitting on the calcite surface are
estimation of A, which is canceled out for prediction of θ performed and shown in Figure S5a. The results indicate that
O 2p PDOS of water interacting with calcite has more
USL − ULV /2 U m=1 − U m>1
cos θ = = ad m > 1 ad electronic states than those of pure water. The water−calcite
ULV /2 Uad (2) interfacial polar interaction leads to the hybridization of the 4s
PDOSs of the calcium ion and the O 2p PDOS of water. The
On the other hand, once the contact angle is known, the electrostatic potential along the c-axis of the calcite crystal is
macroscopic work of adhesion at the solid−liquid interface can also calculated to evaluate the polar effect on the electrostatic
be predicted by inputting the measured surface tensions of potential. Compared with the interface at the vacuum calcite,
liquids that are available in the literature.37 the polar interaction lowers the electrostatic potential at the
We extract the surface energy components of the calcite interface between water and calcite, as shown in Figure S5b.
surface (Table S1) from the DFT-predicted work of adhesion More details are included in Section VII of the Supporting
and tabulate the values of water, ethanol, and decane (Table Information.
S2) based on the vOCG method.36,37 Analytical solutions for The surface tension components in Table S2 indicate that
the contact angles of various liquids on the calcite surface are water, ethanol, glycerol, and ethylene glycol have relatively
obtained using these surface energy components. Details of the high polarity compared to other selected liquids. Specifically,
calculation of the surface free energy components are included for the three partially wetting polar liquids, i.e., water, ethylene
in Section S.V of the Supporting Information. The DFT- glycol, and glycerol, a counter-intuition wetting phenomenon
predicted contact angles based on eq 2 and the analytical is identified: water has the lowest contact angle on calcite
solutions for various liquids are summarized in Table 1 and despite having the highest surface tension, and ethylene glycol
compared to the experimental data. We can find that the has the highest contact angle but the lowest surface tension.
analytical solutions are valid for various nonpolar liquids on the The calculated adsorption energy at different layers of ethylene
calcite surface; however, they do not work well for polar glycol and glycerol on the calcite surface is shown in Figure 2a.
liquids, such as ethylene glycol, and glycerol. It is essential to Similar to water and ethanol on calcite, the adhesion at the
look into the interactions between calcite and such liquids, as solid−liquid interface for ethylene glycol and glycerol is
both of them play major roles in calcite crystal growth and dominated by the binding between the dipole of the polar
dissolution.46,47 The charge density difference distributions Δρ liquid and calcium ion, as shown in Figure 2c,d. The difference
at the solid−liquid interfaces are calculated is that ethanol has low surface tension due to low cohesion
Δρ = ρSL − ρSV − ρLV (3) energy, which makes it fully wet the calcite surface. For water,
ethylene glycol, and glycerol, which partially wet the calcite
where ρSL, ρSV, and ρLV are charge densities of the solid−liquid surface, the contact angle is dominated by adhesion energy,
system, solid substrate, and the adsorbed liquid molecules, shown in Figures 1b and 2a. Note that the adhesion between
12755 DOI: 10.1021/acs.jpcc.9b00937
J. Phys. Chem. C 2019, 123, 12753−12761
The Journal of Physical Chemistry C Article

Figure 2. Effect of polar interaction on adsorption energy at liquid−solid interfaces based on Bader charge analysis. (a) Adsorption energy of the
mth liquid layer on the calcite surface covered with (m − 1) layers of liquids for ethylene glycol and glycerol. (b) Adhesion energy of different polar
liquid molecules at the solid−liquid interface versus the electrostatic potential calculated by considering the integrated Bader charge difference and
the distance among atoms. Black symbols: adhesion energy; blue symbols: contact angle; and dotted line: a linear relation between adsorption
energy and electrostatic potential difference. (c, d) Charge density difference distribution of ethylene glycol and glycerol on calcite, respectively.

ethylene glycol and calcite is much weaker than others as in
Figure 2b, and the fluctuation in Figure 2a can potentially be
minimized by increasing the size of the simulation cell.
To quantitatively analyze the polar effect on the interaction
at the solid−liquid interface, we use the Bader charge
analysis38,39 to calculate the amount of charge difference per
atom at the interface owing to the binding between liquid and
solid, as given by
Δq = qSL − qSV − qLV (4)
Similar to the calculation of the charge density difference
distribution, the calculation of the integrated charge difference
per atom also requires three calculations of integrated charge
per atom in the three systems: the full solid−liquid system qSL,
the solid substrate qSV, and the adsorbed liquid molecules qLV.
Among the four polar liquids including water, ethanol, ethylene
glycol, and glycerol, the magnitude of adhesion energy at the
solid−liquid interface is proportional to the magnitude of the
interaction between polarized charges, as indicated by the
linear relation between the electrostatic potential difference
and adhesion energy at the solid−liquid interface, as shown in
Figure 2b. The electrostatic potential difference is obtained by
calculating ΔUe = ∑i,jΔqiΔqj/rij, where qi is the polarized
charge in calcium cation of calcite, qj is the charge in hydrogen
and oxygen atoms of liquid molecules, and rij is the distance
between atom i in the calcite molecules and atom j in the
liquid molecules. The predicted contact angles for water,
glycerol, and ethylene glycol are also shown in Figure 2b to
illustrate the correlation with adhesion energy. The similar
trend for contact angle and adhesion energy indicates that for
polar liquids with a relatively high surface tension, electrostatic
interactions affect the contact angle more significantly than the
surface tension.
Multiphase Fluids−Solids Systems. Multiphase systems
consisting of water on mineral crystal surfaces (i.e., calcite and
dolomite) in decane are taken as benchmark studies to reveal
the role of interfacial interaction in multiphase systems. In our
previous work, the DFT-predicted adhesion and cohesion
energies (USL and ULV/2) of the calcite (10.4) surface for
12756
liquids, including water and hexane, were used as input
parameters to predict the surface wettability of a water−
hexane−calcite system, as given by33
(γOV − γWV ) − (WOS − WWS) = γOW cos θ (5)
where γOV and γWV are the surface tensions of hexane and
water liquids, respectively. γOW is the interfacial tension at the
hexane−water interface. The work of adhesion at the hexane−
calcite and water−calcite interfaces, WOS and WWS, are
predicted with
ULS
WLS = γLV
ULV /2 (6)
where the surface tension of the liquids γLV is experimentally
measured. In fact, WOS and WWS are predicted separately in
two solid−liquid systems with one single type of liquid in each
system. In addition, the interfacial tensions at the water−oil
interface was taken from the literature. In this section, the
method is extended to an integrated system with two different
liquids on the solid surface simultaneously, which helps us to
gain more insights into the interactions at the interface
between two different liquids on the solid surface, as shown in
Figure 3a. We take three benchmark studies: multiple liquid
layers of decane on the water−calcite surface and the water−
dolomite surface and multiple liquid layers of mercury on the
water−calcite surface.
A slab model with multiple layers of decane on the water−
calcite surface, as shown in Figure 3b, is first simulated with
standard DFT simulation. The adsorption energies for different
liquid layers on calcite, consisting of three water layers and four
decane layers, are calculated using eq 1 and shown in Figure
3c. Several interfaces can be identified: water−calcite, water−
water, water−decane, and decane−decane. First-principles
quantum simulation of this system predicts the adhesion
energy at the water−calcite interface, cohesion energy at the
water−water interface, adhesion energy at the water−decane
interface, and cohesion energy at the decane−decane interface,
respectively, as highlighted with different-colored zones in
DOI: 10.1021/acs.jpcc.9b00937
J. Phys. Chem. C 2019, 123, 12753−12761
The Journal of Physical Chemistry C
Figure 3. DFT prediction for adsorption energies in decane−water−
calcite and mercury−water−calcite systems. (a) Schematic for a
liquid−water−solid system that is directly predicted with DFT
simulations. (b) Configuration of three-layer decane on three-layer-
water-covered calcite surface. (c) Adsorption energy predicted at
different interfaces for decane−water−calcite and decane−water−
dolomite systems, including calcite/dolomite−water, water−water,
water−decane, and decane−decane interfaces. (d) Configuration of
five-layer mercury on three-layer-water-covered calcite surface. (e)
Adsorption energy predicted at different interfaces, including calcite−
water, water−water, water−mercury, and mercury−mercury inter-
faces. The blue dotted lines in (c) and (e) show the predicted
cohesion energies of decane and mercury by individual decane−
calcite and mercury−calcite systems, respectively.
Figure 3c. The average adsorption energy of the topmost two
layers of decane gives the cohesion energy at the decane−
decane interface, which has the same value as predicted by the
system of calcite−decane shown in Figure 1b and the blue
dotted line in Figure 3c. This agreement verifies the feasibility
of direct simulation of the decane−water−calcite system. The
adsorption energy for the first layer of decane corresponds to
the adhesion energy at the water−decane interface. The
adsorption energy increases gradually with the number of
decane layers until reaching the cohesion energy at the third
layer of decane. Similar simulation is performed for the
decane−water−dolomite (10.4) system. The dolomite (10.4)
surface is cleaved from a bulk dolomite (CaMg(CO3)2) crystal
with details included in Section S.II of the Supporting
Information. As expected, the solid surface has no effect on
the cohesion energy and adhesion energy between the liquid
layers, including water−water, water−decane, and decane−
decane interfaces. The major difference is the adhesion energy
at the water−calcite and water−dolomite interfaces. The lower
12757
Article
adhesion at the water−dolomite interface indicates that
dolomite is less hydrophilic than calcite.
By following similar procedures, the case of mercury liquid
on the water−calcite surface is also investigated, as shown in
Figure 3d. The adsorption energy at different liquid layers on
the calcite surface in Figure 3e represents the adhesion energy
at the mercury−water interface and the cohesion energy at the
mercury−mercury interface. The predicted cohesion energy for
mercury agrees well with that predicted in a mercury−calcite
system as previously reported.33 Again, it is worth highlighting
that the adsorption energies in the green zone of Figure 3e are
very different from those in Figure 3c. The large variation from
layer 4 to layer 6 indicates a relative complex interface between
water and mercury due to the strong polar interactions in this
system. Effects of polar interactions for these two systems are
discussed next.
The mechanism of polar interactions in these triphase
systems is significantly different. Figure 4 shows the charge
difference distribution at different interfaces. For the decane−
water−calcite system, the polar interaction of dipole-induced
dipole at the decane−water interface (Figure 4a) is much
stronger than the nonpolar interaction at the decane−decane
interface (Figure 4b). This makes decane−water adhesion
energy higher than decane−decane cohesion energy, as shown
in Figure 3c. For the mercury−water−calcite system, the
strong metallic bonding between mercury atoms (Figure 4d)
overcomes the dipole−dipole interaction at the mercury−
water interface (Figure 4c) as the number of mercury layers
increases. This leads to a much higher mercury cohesion
energy than the mercury−water adhesion energy and a large
variation of the adsorption energy at the water−mercury
interfaces as shown in Figure 3e. Other views of the charge
difference distribution can be found in Figures S3 and S4.
It is important to link the energy obtained from DFT
simulation, as shown in Figure 3c,e, to the physical quantities
of macroscopic surface wettability of a liquid−liquid−solid
triphase system. For a system with one type of multiple liquid
layers on the solid surface, we assume the surface area of the
liquid molecule to be the same at the solid−liquid and liquid−
liquid interfaces in the atomic simulation, as shown in Figure
1a, which works well for the prediction of the surface
wettability of calcite crystal surfaces for different polar and
nonpolar liquids. Here, to predict the interfacial tension
between the two types of overlaying liquids on the calcite
surface, we assume that the interaction area between the two
liquids A′ is equal to that of the calcite−water interface A, that
is, A = A′ in Figure 3a. This assumption is valid if the size of
liquid molecules on the calcite surface is much smaller than
that of the other liquid molecules on the water surface, which
leaves the interaction area determined by the surface area of
topping larger molecules. This condition is applicable to both
decane−water−calcite and mercury−water−calcite systems.
The Dupré equation is used to predict the interfacial tension
between two different liquids
γWL = γWV + γLV − WWL (7)
where γWV and γLV are the surface tensions of the two liquids;
γWL is the interfacial tension between them; and WWL is the
work of adhesion at the water−liquid interface, where the
subscript “L” represents the topping liquid, decane or mercury,
in this work. An equation similar to eq 6 is used to calculate
WWL, as given by
DOI: 10.1021/acs.jpcc.9b00937
J. Phys. Chem. C 2019, 123, 12753−12761
The Journal of Physical Chemistry C Article

Figure 4. Charge difference distribution in multiphase systems: (a) decane−water interface; (b) decane−decane interface; (c) mercury−water
interface; and (d) mercury−mercury interface.

UWL and the isolated mth liquid monolayer at temperature T,

WWL = γ respectively. An accurate calculation of the average adsorption
ULV /2 LV
where UWL is the adhesion energy at the water−decane
interface, ULV/2 is the cohesion energy at the decane−decane
interface, and γLV is the surface tension of decane liquid.
Surface tensions for water, decane, and mercury are taken from
the experimental data, which are 72.80, 23.83, and 425.41 mN/
m, respectively. Using eqs 7 and 8, the predicted water−decane
interfacial tension is 62.22 mN/m, slightly higher than our
experimental measurement of 50.02 mN/m (Table S3). For
the mercury−water interface, the predicted interfacial tension
is 385.10 mN/m, while the experimental value is 375.00 mN/
m.48 The predicted interfacial tensions for both systems are
slightly higher than the experimental data, probably due to the
assumption for the interfacial interaction areas. This work
provides an alternative method for the prediction of interfacial
tension and an integrated slab model involving three different
phases. To the best of our knowledge, this is the first time that
multiphase fluids on solid are directly simulated with the first-
principles DFT simulation.
Thermal Effect on Surface Wettability. With advanced
computational resources and QMD simulation, we quantita-
tively study the temperature dependence of wetting properties
for a decane−water−calcite triphase system. The temperature-
(8)
energy Umad(t, T) requires two QMD calculations for US+mL(t,
T) and US+(m−1)L(t, T) and thousands of self-consistent
calculations of UmL(t, T) based on the configuration in
different time steps.
To predict the contact angle of water on the calcite surface
in the presence of decane at different temperature, all of the
parameters in eq 5 are dependent on temperature
(γOV(T ) − γWV(T )) − (WOS(T ) − WWS(T ))
= γOW(T )cos(θ(T ))
(10)
where each term on the left-hand side of the equation is
estimated with a first-order linear approximation over the
temperature deviation from ambient (T0). For example,
′
UOV
γOV(T ) ≅ γOV(T0) + γOV(T0) U ΔT , where the temper-
OV(T0)
ature difference ΔT = T − T0. The gradient of cohesion
energy between oil layers U′OV = dUOV/dT = ΔUOV/(T2 −T1),
which is predicted with reference temperature T1 and T2.
Substituting the expressions of linear approximation for
γOV(T), γWV(T), WOS(T), and WWS(T) into eq 10 and using
the relation of WOS = γOV(UOS/UOO) and WWS = γWV(UWS/
UWW), eq 11 can be eventually obtained. The detailed

ÄÅ É
ÅÅÅ ij yzÑÑÑÑ
dependent work of adhesion and surface tensions predicted for derivation can be found in Section S.VIII.
yz ij
ÅÅγ jj1 + UOV zz − γ jj1 + U ′ zzÑÑ
the water−calcite and decane−calcite systems are linked by a
ÅÅ OV jj UOV zz{ WV j j UWV zz{ÑÑÑÑÖ
ÅÅÇ k
ΔT ΔT
k
temperature-corrected governing equation to calculate the ′
ÄÅ ÉÑ
WV
contact angle of water on calcite in a decane environment. A
ÅÅ ij ΔT yzz ij ΔT yzzÑÑÑÑ
Å
Å j j
time-dependent adsorption energy Umad(t, T) of the mth liquid
− ÅÅWOSjj1 + UOS z − WWSjj1 + UWS zÑ
ÅÅÅÇ j UOS zz{ j UWS zz{ÑÑÑÑÖ
layer on the surface covered with (m − 1) layers of liquids can
k k
′ ′
be calculated at different temperature

m
US + mL(t , T ) − US + (m − 1)L(t , T ) − UmL(t , T ) = γOW(T )cos(θ(T ))
Uad(t , T ) = (11)
nL
where γOV = γOV(T0), γWV = γWV(T0) and so for WOS, WWS,
(9)
UOS, and UWS, where (T0) is omitted for the ease of expression.
where US+mL(t, T), US+(m−1)L(t, T), and UmL(t, T) are the time- Here, γOW(T) is measured for a temperature range of 25−85
dependent total energies of the solid surface with m overlaying °C with a Goniometer equipped with a heater (experimental
liquid layers, the surface with (m − 1) overlaying liquid layers, data are listed in Table S3 in Section S.IX). Surface tensions
12758 DOI: 10.1021/acs.jpcc.9b00937
J. Phys. Chem. C 2019, 123, 12753−12761
The Journal of Physical Chemistry C
γOV and γWV are experimentally measured as well. Work of
adhesion WOS and WWS are predicted from the DFT energy
calculation as mentioned before. The calcite−water and
calcite−decane systems are simulated separately at temper-
atures of 200 and 300 K for 3000 steps to calculate the
adsorption energy gradients: U′OV, U′WV, U′OS, and U′WS, which
are further used in eq 11 to predict the contact angle of water
on the calcite surface in the presence of decane at different
temperature. The predicted temperature range is the same as
that used in the experimental measurement with variation in
temperature, ΔT = 0−70 K.
The predicted adsorption energy gradients are listed in
Table 2. The adhesion and cohesion energies of decane and
Table 2. QMD-Predicted Reduction Rates of Adsorption
Energies for Decane and Water
decane water
adhesion/cohesion gradient ′
UOS ′
UOV ′
UWS
QMD predicted reduction rates 0.0579 0.0580 0.0407
(kJ/(mol K))
water increase with temperature at different increasig rates,
which in fact is a magnitude reduction rate as the adhesion and
cohesion energies are negative. For decane, the reduction rates
for both adhesion and cohesion energies are very similar, which
indicates that the thermal effect on the adsorption of decane
on the calcite surface is not significant due to the nonpolar
nature of decane. For water, the reduction rate of cohesion
energy is significantly higher than that of the adhesion energy
due to the presence of polar interaction. Figure 5 shows the
Figure 5. QMD-predicted and experimentally measured contact
angles of water on the calcite surface in the presence of decane at
different temperatures. Experimental measurements are performed on
three different positions of an Iceland spar crystal with a flat cleaved
surface (10.4).
predicted and experimentally measured contact angles of water
on the calcite surface in the presence of decane. Note that the
experimental measurements are taken at three different
positions on the same Iceland spar crystal with a flat cleaved
surface (10.4), and these measurement data are different from
each other due to nonhomogeneous roughness on the crystal
surface. However, the prediction from our quantum simulation
is intrinsic without considering the surface roughness. To make
a fair comparison, the average roughness factor of the crystal
surface is measured with an atomic force microscope (AFM)
and used in a Wenzel model49 to estimate the intrinsic contact
angles from the measurements, as described in Section S.X.
Article
The symbols in Figure 5 show the estimated intrinsic contact
angles, which are different for different positions due to the
averaged roughness factor. Nevertheless, the comparison
between DFT predictions and the estimated values based on
the experimental measurement shows good agreement, with a
maximum deviation of less than 10%. Both results indicate that
increase of temperature makes the calcite surface more water-
wet. According to the previous study,30 Mg2+ ion substitutions
make the surface more water-wet as well. Although it is beyond
the scope of this work to combine the two effects into the
QMD simulation, a more water-wet calcite surface under Mg2+
ion substitutions at an elevated temperature is foreseeable.
■
CONCLUSIONS
In summary, we quantitatively investigated the effect of
polarity on the solid surface wettability for various liquids by
using a DFT slab model-based first-principles methodology.
′
UWV The quantitative charge analysis for atoms at the solid−liquid
0.0759 interfaces reveals a linear correlation between surface
adsorption and electrostatic interaction. A counter-intuition
phenomenon between contact angle and liquid surface tension
is identified and studied, that is, water has a lower contact
angle on calcite than ethylene glycol or glycerol despite its
higher surface tension. We find that this is attributed to the
larger electrostatic interaction at the calcite−water interface
according to the distribution of charge density difference. The
quantum simulation approach is developed, for the first time,
to directly predict the interfacial interactions in liquid−liquid−
solid triphase systems involving two different liquids
simultaneously. By using the predicted adsorption energies at
different liquid−liquid interfaces, interfacial tensions for
water−decane and water−mercury are calculated, which are
in reasonable agreement with the experimental data. The
thermal effect on solid surface wettability in a multiphase
system is predicted using the first-principles quantum
molecular simulation and validated against the contact angle
measurements of water on the calcite surface in the presence of
decane. Overall, the proposed methodology reveals the
mechanism of complicated surface wetting and interfacial
behaviors, particularly polar interactions in multiphase systems.
More importantly, this first-principles approach is general and
applicable to any crystal surface and various surrounding fluids.
■ METHODS
Contact Angle Measurements. For contact angle
measurements, we carefully choose calcite samples from
Iceland spar crystals with a flat cleaved surface (10.4) to test
the generality of the measurements. We also confirmed the
presence of a robust crystalline phase of (10.4) using X-ray
diffraction patterns.33 This observation confirmed the domi-
nance of a thermodynamically stable plane in the sample. The
samples are prepared by following the same cleaning
procedure. First, the calcite crystals are cleaned with acetone,
ethanol, and deionized water in a sonicator for 10 min,
sequentially. Then, the wet surfaces are dried immediately by
blowing dry nitrogen gas. After the cleaning procedures, the
calcite crystals are immediately used to measure the liquid
contact angles in air and decane, respectively. It should be
noted that for water contact angle measurement on the calcite
surface in decane, the water droplet was injected in the decane
liquid near the calcite surface to avoid the existence of air at
12759 DOI: 10.1021/acs.jpcc.9b00937
J. Phys. Chem. C 2019, 123, 12753−12761
The Journal of Physical Chemistry C
the solid−liquid interfaces. Measurements are performed at
three different positions of a crystal surface.
■
ASSOCIATED CONTENT
*
S Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jpcc.9b00937.
Calcite (10.4) and dolomite (10.4) surfaces cleaved
from bulk calcite and dolomite crystals; pseudopotentials
in the Quantum Espresso package; configurations of
multiple liquid layers on the calcite surface for different
liquids; calculations of surface free components of the
calcite solid surface; charge density difference distribu-
tions of multiple liquid layers on the calcite surface for
different liquids; representative configurations for solid−
liquid systems for different polar/nonpolar liquids;
charge density difference distribution at solid−liquid
interfaces; surface free energy components of the calcite
(10.4) surface based on DFT-vOCG prediction results;
and surface tension components for various polar/
nonpolar liquids (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: tiejun.zhang@ku.ac.ae.
ORCID
Qiaoyu Ge: 0000-0002-7467-7327
TieJun Zhang: 0000-0001-9102-0178
Author Contributions
§ (18) Liu, J.; Zhu, C.; Liu, K.; Jiang, Y.; Song, Y.; Francisco, J. S.;
J.Y.L and Q.G are first authors.
Notes
The authors declare no competing financial interest.
■ ACKNOWLEDGMENTS
This work was supported by the Abu Dhabi National Oil
Company R&D Department (RDProj.081-RCM) and also by
High Performance Cloud Computing Platform of Alibaba
Cloud.
■
REFERENCES
(1) Lee, S. G.; Lee, H.; Gupta, A.; Chang, S.; Doyle, P. S. Site-
Selective In Situ Grown Calcium Carbonate Micromodels with
Tunable Geometry, Porosity, and Wettability. Adv. Funct. Mater.
2016, 26, 4896−4905.
(2) Lu, J. Y.; Raza, A.; Noorulla, S.; Alketbi, A. S.; Fang, N. X.; Chen,
G.; Zhang, T. Near-Perfect Ultrathin Nanocomposite Absorber with
Self-Formed Topping Plasmonic Nanoparticles. Adv. Opt. Mater.
2017, 5, No. 1700222.
(3) Li, H.; Yang, W.; Aili, A.; Zhang, T. Insights into the Impact of
Surface Hydrophobicity on Droplet Coalescence and Jumping
Dynamics. Langmuir 2017, 33, 8574−8581.
(4) Anand, S.; Paxson, A. T.; Dhiman, R.; Smith, J. D.; Varanasi, K.
K. Enhanced Condensation on Lubricant-Impregnated Nanotextured
Surfaces. ACS Nano 2012, 6, 10122−10129.
(5) Li, H.; Aili, A.; Alhosani, M. H.; Ge, Q.; Zhang, T. Directional
Passive Transport of Microdroplets in Oil-Infused Diverging
Channels for Effective Condensate Removal. ACS Appl. Mater.
Interfaces 2018, 10, 20910−20919.
(6) Parker, A. R.; Lawrence, C. R. Water Capture by a Desert Beetle.
Nature 2001, 414, 33−34.
(7) Dai, X.; Sun, N.; Nielsen, S. O.; Stogin, B. B.; Wang, J.; Yang, S.;
Wong, T.-S. Hydrophilic Directional Slippery Rough Surfaces for
Water Harvesting. Sci. Adv. 2018, 4, No. eaaq0919.
12760
Article
(8) Stoerzinger, K. A.; Hong, W. T.; Azimi, G.; Giordano, L.; Lee, Y.
L.; Crumlin, E. J.; Biegalski, M. D.; Bluhm, H.; Varanasi, K. K.; Shao-
Horn, Y. Reactivity of Perovskites with Water: Role of Hydroxylation
in Wetting and Implications for Oxygen Electrocatalysis. J. Phys.
Chem. C 2015, 119, 18504−18512.
(9) Alvarado, V.; Manrique, E. Enhanced Oil Recovery: An Update
Review. Energies 2010, 3, 1529−1575.
(10) Liu, J.; Wani, O. B.; Alhassan, S. M.; Pantelides, S. T.
Wettability Alteration and Enhanced Oil Recovery Induced by
Proximal Adsorption of Na+, Cl−, Ca2+, Ma2+, and SO42− Ions on
Calcite. Phys. Rev. Appl. 2018, 10, No. 034064.
(11) Wang, J.; Sun, L.; Zou, M.; Gao, W.; Liu, C.; Shang, L.; Gu, Z.;
Zhao, Y. Bioinspired Shape-Memory Graphene Film with Tunable
Wettability. Sci. Adv. 2017, 3, No. e1700004.
(12) Wang, J.; Gao, W.; Zhang, H.; Zou, M.; Chen, Y.; Zhao, Y.
Programmable Wettability on Photocontrolled Graphene Film. Sci.
Adv. 2018, 4, No. eaat7392.
(13) Lu, J.-Y.; Lai, C.-Y.; Almansoori, I.; Chiesa, M. The Evolution
in Graphitic Surface Wettability with First-Principles Quantum
Simulations: The Counterintuitive Role of Water. Phys. Chem.
Chem. Phys. 2018, 20, 22636−22644.
(14) Lu, J.-Y.; Olukan, T.; Tamalampudi, S. R.; Al-Hagri, A.; Lai, C.-
Y.; Mahri, M. A. Al; Apostoleris, H.; Almansouri, I.; Chiesa, M.
Insights into Graphene Wettability Transparency by Locally Probing
Its Surface Free Energy. Nanoscale 2019, 7944−7951.
(15) Giovambattista, N.; Debenedetti, P. G.; Rossky, P. J. Enhanced
Surface Hydrophobicity by Coupling of Surface Polarity and
Topography. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 15181−15185.
(16) Azimi, G.; Dhiman, R.; Kwon, H.; Paxson, A. T.; Varanasi, K.
K. Hydrophobicity of Rare-Earth Oxide Ceramics. Nat. Mater. 2013,
12, 315−320.
(17) Hass, K. C. The Chemistry of Water on Alumina Surfaces:
Reaction Dynamics from First Principles. Science 1998, 282, 265−268.
Zeng, X. C.; Wang, J. Distinct Ice Patterns on Solid Surfaces with
Various Wettabilities. Proc. Natl. Acad. Sci. U.S.A. 2017, 114, 11285−
11290.
(19) Tang, J.; Zhao, X.; Li, J.; Zhou, Y.; Liu, J. Liquid Metal
Phagocytosis: Intermetallic Wetting Induced Particle Internalization.
Adv. Sci. 2017, 4, No. 1700024.
(20) Cui, Y.; Liang, F.; Yang, Z.; Xu, S.; Zhao, X.; Ding, Y.; Lin, Z.;
et al. Metallic Bond-Enabled Wetting Behavior at the Liquid Ga/
CuGa2 Interfaces. ACS Appl. Mater. Interfaces 2018, 10, 9203−9210.
(21) Cui, Y.; Liang, F.; Xu, S.; Ding, Y.; Lin, Z.; Liu, J. Interfacial
Wetting Behaviors of Liquid Ga Alloys/FeGa3 Based on Metallic
Bond Interaction. Colloids Surf., A 2019, 569, 102−109.
(22) Giovambattista, N.; Debenedetti, P. G.; Rossky, P. J. Effect of
Surface Polarity on Water Contact Angle and Interfacial Hydration
Structure. J. Phys. Chem. B 2007, 111, 9581−9587.
(23) Carchini, G.; García-Melchor, M.; Łodziana, Z.; López, N.
Understanding and Tuning the Intrinsic Hydrophobicity of Rare-
Earth Oxides: A DFT + U Study. ACS Appl. Mater. Interfaces 2016, 8,
152−160.
(24) Kreder, M. J.; Alvarenga, J.; Kim, P.; Aizenberg, J. Design of
Anti-Icing Surfaces: Smooth, Textured or Slippery? Nat. Rev. Mater.
2016, 1, No. 15003.
(25) Strand, S.; Høgnesen, E. J.; Austad, T. Wettability Alteration of
CarbonatesEffects of Potential Determining Ions (Ca2+ and SO42−)
and Temperature. Colloids Surf., A 2006, 275, 1−10.
(26) Saville, G. Computer Simulation of the Liquid−solid−vapour
Contact Angle. J. Chem. Soc., Faraday Trans. 2 1977, 73, 1122−1132.
(27) Maruyama, S.; Kurashige, T.; Matsumo, S.; Yamaguchi, Y.;
Kimura, T. Liquid Droplet in Contact with a Solid Surface. Microscale
Thermophys. Eng. 1998, 2, 49−62.
(28) Pandey, P. R.; Roy, S. Is It Possible to Change Wettability of
Hydrophilic Surface by Changing Its Roughness? J. Phys. Chem. Lett.
2013, 4, 3692−3697.
(29) Sarkar, T.; Ghosh, S.; Annamalai, M.; Patra, A.; Stoerzinger, K.;
Lee, Y.-L.; Prakash, S.; Motapothula, M. R.; Shao-Horn, Y.; Giordano,
DOI: 10.1021/acs.jpcc.9b00937
J. Phys. Chem. C 2019, 123, 12753−12761
The Journal of Physical Chemistry C Article

L.; Venkatesan, T. The Effect of Oxygen Vacancies on Water

Wettability of Transition Metal Based SrTiO3 and Rare-Earth Based
Lu2O3. RSC Adv. 2016, 6, 109234−109240.
(30) Sakuma, H.; Andersson, M. P.; Bechgaard, K.; Stipp, S. L. S.
Surface Tension Alteration on Calcite, Induced by Ion Substitution. J.
Phys. Chem. C 2014, 118, 3078−3087.
(31) Prabhakar, S.; Melnik, R. Wettability Alteration of Calcite Oil
Wells: Influence of Smart Water Ions. Sci. Rep. 2017, 7, No. 17365.
(32) Andersson, M. P.; Bennetzen, M. V.; Klamt, A.; Stipp, S. L. S.
First-Principles Prediction of Liquid/Liquid Interfacial Tension. J.
Chem. Theory Comput. 2014, 10, 3401−3408.
(33) Lu, J. Y.; Ge, Q.; Li, H.; Raza, A.; Zhang, T. Direct Prediction
of Calcite Surface Wettability with First-Principles Quantum
Simulation. J. Phys. Chem. Lett. 2017, 8, 5309−5316.
(34) Llovell, F.; Vilaseca, O.; Jung, N.; Vega, L. F. Water + 1-Alkanol
Systems: Modeling the Phase, Interface and Viscosity Properties.
Fluid Phase Equilib. 2013, 360, 367−378.
(35) Vilaseca, O.; Vega, L. F. Direct Calculation of Interfacial
Properties of Fluids Close to the Critical Region by a Molecular-
Based Equation of State. Fluid Phase Equilib. 2011, 306, 4−14.
(36) van Oss, C. J.; Good, R. J.; Chaudhury, M. K. Additive and
Nonadditive Surface Tension Components and the Interpretation of
Contact Angles. Langmuir 1988, 4, 884−891.
(37) Preston, D. J.; Song, Y.; Lu, Z.; Antao, D. S.; Wang, E. N.
Design of Lubricant Infused Surfaces. ACS Appl. Mater. Interfaces
2017, 9, 42383−42392.
(38) Bader, R. F. W. Atoms in Molecules: A Quantum Theory; Oxford
University Press: New Yourk, 1990.
(39) Yu, M.; Trinkle, D. R. Accurate and Efficient Algorithm for
Bader Charge Integration. J. Chem. Phys. 2011, 134, No. 064111.
(40) Nazmutdmov, R. R.; Probst, M.; Hemzinger, K. Quantum
Chemical Study of the Adsorption of an H2O Molecule on an
Uncharged Mercury Surface. J. Electroanal. Chem. 1994, 369, 227−
231.
(41) Zhang, B.; Liu, J.; Zheng, C.; Chang, M. Theoretical Study of
Mercury Species Adsorption Mechanism on MnO2(1 1 0) Surface.
Chem. Eng. J. 2014, 256, 93−100.
(42) Pittman, E. D. Relationship of Porosity and Permeability to
Various Parameters Derived from Mercury Injection-Capillary
Pressure Curves for Sandstone (1). AAPG Bull. 1992, 76, 191−198.
(43) Schmitt, M.; Fernandes, C. P.; da Cunha Neto, J. A. B.; Wolf, F.
G.; dos Santos, V. S. S. Characterization of Pore Systems in Seal
Rocks Using Nitrogen Gas Adsorption Combined with Mercury
Injection Capillary Pressure Techniques. Mar. Pet. Geol. 2013, 39,
138−149.
(44) Giannozzi, P.; Baroni, S.; Bonini, N.; Calandra, M.; Car, R.;
Cavazzoni, C.; Ceresoli, D.; Chiarotti, G. L.; Cococcioni, M.; Dabo,
I.; et al. Q UANTUM ESPRESSO: A Modular and Open-Source
Software Project for Quantum Simulations of Materials. J. Phys.:
Condens. Matter 2009, 21, No. 395502.
(45) Giannozzi, P.; Andreussi, O.; Brumme, T.; Bunau, O.; Nardelli,
M. B.; Calandra, M.; Car, R.; Cavazzoni, C.; Ceresoli, D.; Cococcioni,
M. Advanced Capabilities for Materials Modelling with Quantum
ESPRESSO. J. Phys.: Condens. Matter 2017, 29, No. 465901.
(46) Flaten, E. M.; Seiersten, M.; Andreassen, J.-P. Polymorphism
and Morphology of Calcium Carbonate Precipitated in Mixed
Solvents of Ethylene Glycol and Water. J. Cryst. Growth 2009, 311,
3533−3538.
(47) Konopacka-Łyskawa, D.; Kościelska, B.; Karczewski, J.
Controlling the Size and Morphology of Precipitated Calcite Particles
by the Selection of Solvent Composition. J. Cryst. Growth 2017, 478,
102−110.
(48) Henry, D. C.; Jackson, J. Interfacial Tension between Mercury
and Water. Nature 1938, 142, 616−617.
(49) de Gennes, P.-G.; Brochard-Wyart, F.; Quere, D. Capillarity
and Wetting Phenomena Drops, Bubbles, Pearls, Waves; Springer-Verlag:
New York, 2004.

12761 DOI: 10.1021/acs.jpcc.9b00937

J. Phys. Chem. C 2019, 123, 12753−12761