Article pubs.acs.org/JPCA
Electric-Field-Induced Phase Transition of Confined Water Nanofilms between Two Graphene Sheets Zhenyu Qian and Guanghong Wei* State Key Laboratory of Surface Physics, Key Laboratory for Computational Physical Sciences (MOE), and Department of Physics, Fudan University, Shanghai 200433, P. R. China ABSTRACT: A recent study reported that confined water nanofilms may freeze continuously or discontinuously depending on their densities. In this study, we report results from molecular dynamics simulations of the structures and the phase transition of water confined between two graphene sheets with a separation of 1.0 nm under the influence of an electric (E) field applied along the direction parallel to the sheets. We find that confined water can form three kinds of ice phases at atmospheric pressure: amorphous, hexagonal, or rhombic bilayer ice, depending on the E-field strength (0−1.5 V/nm). As the E-field strength changes, these ice configurations can transform into each other through a first-order phase transition. These E-field-induced water phases are different from those induced by high pressure (under high density). In addition, we find that all of the three ice nanofilms melt through a first-order transition. The heating and cooling processes are accompanied by a hysteresis loop between the solid and liquid phases. A phase diagram of confined water between two graphene sheets is given in the temperature-E-field plane. by means of Fourier transform infrared spectroscopy (FTIR),31 low-energy electron diffraction (LEED),32 ultrafast electron crystallography (UEC),33 and X-ray diffraction.34 The electric field is widely applied in experiments to modify the interfacial properties of water at the nanoscale and to tune surface wettability. Computational studies showed E-fieldinduced electrostriction can increase water density.35−38 It was also found that water confined to a thickness of 3−8 layers between two parallel plates crystallized under an E-field of 5 V/ nm along the direction parallel to the plates at 300 K.39 The freezing transition of interfacial water was observed by scanning tunneling microscope (STM) at much weaker field strength (∼10 6 V/m), which implied a new type of freezing mechanism.40 It was also reported that the electric field can induce evaporation or electromelting of water confined between two finite hydrophobic sheets when its direction was vertical to the plates.41,42 In the present study, we investigate the impact of lateral Efields on water confined between two graphene sheets with a separation of 1.0 nm. The E-field strength is in the range of 0− 1.5 V/nm. This choice is made according to a recent density function theory study showing that E-fields beyond a threshold of ∼3.5 V/nm are able to dissociate water molecules.43 We find that water can freeze into hexagonal bilayer ice at lower E and rhombic bilayer ice at higher E both through a first-order phase
I. INTRODUCTION Confined water in nanoscale exhibits novel structural properties that are different from those of bulk water. It has attracted recent attention due to its physiochemical and technological importance in many systems, such as water permeation through ion channels,1−8 the design and coating of superhydrophobic surface,9−11 surface lubricating,12−15 and protein folding.16−20 Slit flat graphene pores provide an ideal model system for the study of such quasi-2D water. The behavior of confined water has been studied by means of molecular dynamics (MD) simulations.21−30 Koga et al. found that water confined in a hydrophobic pore could freeze into amorphous (0.1 MPa) or hexagonal (50 MPa to 1 GPa) bilayer ice via a first-order transition.21,22 Zangi et al. reported that water in a narrower pore could form rhombic monolayer ice, whose arrangement was sensitive to the size and separation of the two plates.24 Systematical investigations under different pressure (P), temperature (T), and plate separation (H), revealed that confined water was similar to the structure and behavior of bulk water at a lower density (∼0.9 g/cm3), with a shifted ∼40 K lower in thermodynamics.25 A temperature− density (T−ρ) phase diagram of water confined between two plates was proposed and showed that the freezing and melting occurred abruptly when the density was 1.33 g/cm3.30 It was also reported that the hydrophobic or hydrophilic properties of plates would strongly influence the behavior of confined water.26,28 It was proposed that confined water can form an ordered structure at room temperature (300 K) if only the plate separation and water density are proper.29 Ordered water structures at hydrophobic interfaces and low-density liquid phase of supercooled water were later observed experimentally © XXXX American Chemical Society
Special Issue: International Conference on Theoretical and High Performance Computational Chemistry Symposium Received: January 28, 2014 Revised: May 15, 2014
A
dx.doi.org/10.1021/jp500989t | J. Phys. Chem. A XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry A
Article
calculated by particle-mesh Ewald method (PME)49 with a real space cutoff of 1.4 nm, and the short-range Lennard-Jones potential with a cutoff of 1.4 nm. The carbon−carbon interaction parameters are adopted from a previous work of Hummer.50 The temperature is maintained constant using Nosé−Hoover’s method,51,52 and the pressure is kept at 1 bar using the isotropic Parrinello−Rahman’s method.53,54 All bonds in water molecules and graphene sheets are constrained using the SETTLE55 and LINCS56 algorithms, respectively, and the simulations are performed using the leapfrog algorithm with a time step of 2 fs. Periodic boundary conditions are applied in all directions. The simulation time for each MD run is 30−50 ns. The two graphene sheets are fixed in space throughout. In data analysis, to avoid the interfacial perturbations, we exclude the water molecules within 0.5 nm away from the plate edges and consider the confined water molecules located in the central part of the x−y plane. The area S is constant (S = (4.9 − 2 × 0.5) × (4.9 − 2 × 0.5) = 15.21 (nm2)), while the number of water molecules located in this region changes with simulation conditions.
transition. The alignments of water molecules in the hexagonal and rhombic bilayer ice found here are different from those observed at high pressure in previous studies.22,30 With the increase of the electric field strength, amorphous ice can transform into hexagonal bilayer ice and then to rhombic bilayer ice discontinuously. These results reveal that electric fields can induce a rich phase and phase transition of quasi-2D water.
II. METHODS Extensive all-atom MD simulations are performed in the isothermal−isobaric (NPT) ensemble using the GROMACS 3.3.3 software.44 As in previous studies, the five-site TIP5P model is used to describe the water molecules.24,30,42,45 The TIP5P model46 is chosen for its explicit representation of the electron lone pair sites and its “Nosé-shape” curve of temperature versus crystallization time.47,48 Two parallel square graphene sheets (Lx × Ly, Lx = Ly = 4.9 nm) are immersed in a box of TIP5P water (Figure 1a). The two sheets are separated by a distance of 1.0 nm (H = 1.0 nm). This separation is chosen such that there are effectively two layers of water molecules between the confining sheets. An E-field is applied parallel to the two plates. The molecular interactions are described by the sum of the long-range electrostatic interactions, which are
III. RESULTS AND DISCUSSION The formation of bilayer ice from disordered liquid water is monitored in the absence and presence of an external E-field along the lateral direction (x-axis in our simulation box). We find water can freeze into three kinds of bilayer ice at 230 K: amorphous ice under E = 0 V/nm, hexagonal ice under E = 0.5 V/nm, and rhombic ice under E = 1.5 V/nm (Figure 1b−d). In addition, the formation of a mixed hexagonal-rhombic ice bilayer (Figure 1e) is found under E = 1.0 V/nm, indicating the coexistence of different ice phases at the transition E-field. Such coexistence phenomena of quasi-2D water have been reported in other simulated conditions.24,29 Electric fields can change the orientation of water molecules and make the intermolecular hydrogen bond (H-bond) network be reorganized, and these Efield-induced confined ice phases are morphologically close to but substantially different from those induced by high pressure.22,30 We have also observed a nucleation process, and it takes only a few nanoseconds for the initial disordered water molecules to form bilayer ice when an electric field is applied along the direction parallel to the graphene planes, but the formation time extends to tens of nanoseconds without an E-field. A similar phenomenon of ice formation was reported for the E-field-induced heterogeneous ice nucleation.57,58 Interestingly, a mixture of hexagonal and cubic ice was observed in the E-field-nucleated ice growth process.59 In a recent study on the pressure-induced phase transition of confined 2D water, Bai et al. found that the ice rule was not satisfied for the very high density amorphous bilayer ice (denoted as BL-VHDA),60 where a hydrogen bond is considered to be formed if the O···O distance is not longer than 2.2 Å. This O···O distance is used because under high pressure (∼GPa), water molecules are much closer than those at atmosphere pressure. In our study, two water molecules are considered to form a hydrogen bond if (1) their O···O distance is shorter than 3.5 Å and (2) the corresponding angle O−H···O is larger than 150°.42 Detailed structural examination shows that each water molecule is hydrogen-bonded to exactly four nearest neighbor water molecules, satisfying the ice rule. The probability of water-molecule orientation under different E-field strengths at 230 K is shown in Figure 2a. The watermolecule orientation is described by the angle (θ) between the dipole vector of a water molecule projected in the x−y plane
Figure 1. Top view (left) and side view (right) of the initial setup of the system (a). Snapshots of different bilayer ice (top view) obtained from MD trajectories at 230 K under different E-field strengths are shown in (b)−(e). The bilayer ice conformation is (b) amorphous at E = 0 V/nm, (c) hexagonal at E = 0.5 V/nm, and (d) rhombic at E = 1.5 V/nm. (e) Mixed hexagonal-rhombic under E = 1.0 V/nm, which is located at the transition field strength. Oxygen atoms are in red and hydrogen atoms in white. In (a), the graphene sheets are in cyan. In (b)−(e), the graphene sheets are omitted for clarity. B
dx.doi.org/10.1021/jp500989t | J. Phys. Chem. A XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry A
Article
strength. The water density distribution along the z-axis is mainly determined by the repulsive interaction between water molecules and the graphene sheets. We present in Figure 3a the density (ρ) of water confined between the two sheets as a function of E-field strengths at 230
Figure 2. Structural analysis of bilayer ice formed at 230 K under three different E-field strengths (0, 0.5, 1.5 V/nm). (a) Population of watermolecule orientation θ under an E-field with a strength of 0, 0.5, and 1.5 V/nm. The θ is the angle between the projection of watermolecule dipole vector in the x−y plane and the x-axis. If the projection of a water dipole vector is located in the first and second quadrants of the x−y plane, then 0° ≤ θ < 180°; otherwise, 180° ≤ θ < 360°. (b) Probability density profile (along z-axis) of water molecules confined between the two graphene sheets. The two graphene sheets are respectively fixed at z = −0.5 and +0.5 nm.
Figure 3. Influence of E-field strengths on the properties of confined water at T = 230 K: (a) the density ρ of water; (b) the potential energy U of water (per water molecule). The error bar is also given.
K. We define ρ as the effective density ρ ≡ m⟨N⟩/(LxLyH′), where m is the mass of a water molecule and ⟨N⟩ is the timeaveraged number of water molecules in the cell slab parallel to the sheets. H′ is the effective distance between the two sheets and it is estimated as 0.674 nm because the van der Waals repulsive interactions between the water molecules and the carbon atoms of the sheets prevent water molecules from coming too close to the sheets.25 With the increase of the Efield strength, the water density curve displays three states for E in the range 0−1.5 V/nm. The water density remains at 0.91 g/ cm3 at a small E-field of 0−0.2 V/nm and then suddenly reaches to ∼0.96 g/cm3 at E = 0.3 V/nm, and the density remains at this value until E = 0.8 V/nm. At E = 0.9 V/nm, the water density climbs steeply to ∼1.02 g/cm3 and stays at this value for E = 1.2−1.5 V/nm. The three states correspond to the formation of three different ice configurations: amorphous, hexagonal, and rhombic bilayer ice, respectively. The stepwise feather of water density with the increase of E-field strength reveals that the solid−solid phase transition is discontinuous. Figure 3b gives the potential energy U of water (per water molecule) as a function of E-field strength at 230 K. Similar to the ρ−E curve in Figure 3a, the U−E curve also exhibits three states at E = 0−0.2, 0.3−0.8, and 1.2−1.5 V/nm. The abrupt changes in energy between states indicate that both the transitions from amorphous to hexagonal bilayer ice and from hexagonal to rhombic bilayer ice are first-order transitions. In addition, the coexistence of the hexagonal ice and rhombic ice (Figure 1d) indicates two distinct phases, and such phase separation reflects the transition is first-order.24,29
and the x-axis, and it reflects the in-plane arrangement of dipoles as well as the internal H-bond network. In the absence of an E-field, the distribution of θ is almost homogeneous (black curve in Figure 2a), implying that water molecules in the amorphous ice configuration do not have a preferred orientation and the confinement does not affect the isotropy of water molecules in the xcy plane. However, under E = 0.5 and 1.5 V/nm, there are two peaks centered at θ = 30°/150° (red curve in Figure 2a) and 15°/165° (green curve in Figure 2a), respectively. It reveals that the electric field modulates the orientation of water-molecule dipoles. Although the direction of E-field is fixed along the x-axis, the most populated orientation of water molecules varies with the change of E-field strength. It is noted that the most populated dipole orientations are not the same as the E-field direction, indicating that water molecules arrange themselves to achieve a global energy optimization and the in-plane H-bond network is the result of a combination of water-E-field and water−water interactions. Figure 2b shows the probability density profile (along the zaxis) of water molecules between the two graphene sheets at 230 K under different strengths of E-fields. The two graphene sheets are respectively fixed at z = ±0.5 nm. It can be seen from Figure 2b that the probability density profile has two peaks around z = ±0.15 nm, indicating the presence of a water bilayer. The positions of the two peaks do not change with the increase of E-field, reflecting that the electric field has a minor influence on the water distribution in the direction perpendicular to the E-field within our simulated E-field C
dx.doi.org/10.1021/jp500989t | J. Phys. Chem. A XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry A
Article
To further demonstrate the critical role of the E-field on the phase transition behavior of confined water, we take the polarization energy −D·E per water molecule into consideration, and plot in Figure 4 the total potential energy U − D·E
Figure 5. Potential energy U of confined water (per water molecule) as a function of temperature under different E-field strengths: 0, 0.5, and 1.5 V/nm. The black and red curves indicate the heating and cooling processes, respectively.
Figure 4. Potential energy U − D·E of three kinds of ice configurations (amorphous, hexagonal, and rhombic bilayer ice) at T = 230 K under three given electric fields: E = 0, 0.5, and 1.5 V/nm. D represents the dipole moment per water molecule.
loop becomes wider as the E-field strength increases. It should be noted that the hexagonal ice under E = 0.5 V/nm and the rhombic ice under E = 1.5 V/nm melts discontinuously, whereas previous study reported that the high-density (ρ > 1.33 g/cm3) rhombic bilayer ice induced by high pressure undergoes a continuous phase transition to liquid water.30 The difference may be attributed to the different mechanisms in pressure- and E-field-induced phase transition of confined water. The influence of high pressure on the configuration of the quasi2D confined water is isotropic in the x−y plane. The highpressure-induced high-density water would facilitate hydrogen bond formation, which leads to liquid−solid phase transition. Differently, due to polarization, electric fields with appropriate strength can reorientate the water molecules along the direction of electric field to form hydrogen bond and decrease thermal fluctuations. This E-field-induced freezing mechanism may have significant implications for the ice formation in diverse systems. Moreover, the solid−liquid transition temperatures under E = 0.5 and 1.5 V/nm are about 280 and 300 K, respectively, much higher than the freezing temperature of the bulk water at atmospheric pressure, in agreement with the transition temperature observed by STM experiment under low electric fields.40 Our results suggest that the adjustment of watermolecule orientation by electric field weakens the thermal vibration of water molecules, and thus raises the freezing point of confined water. The dynamic property of confined water is probed by calculating the mean-square displacement (MSD) MSD = ⟨[r(t) − r(t0)]2⟩,61 which may serve as a good measure to distinguish solid and liquid states.62 MSD is calculated from t0 = 49.0 ns and over the last 1 ns of 50 ns MD simulation. The mobility (the slope of the MSD) differences at different temperatures are shown in Figure 6, and the left/right three panels correspond to the heating/cooling process. At lower temperature, the MSD remains almost flat zero, suggesting that water has frozen into ice. The notable mobility differences at higher temperature distinguish the confined water from solid
of three different ice bilayers under the given electric field strength of 0, 0.5, and 1.5 V/nm at 230 K, where D is the dipole moment per water molecule. The upper, middle, and lower panels clearly show that the amorphous, hexagonal, and rhombic bilayer ice has the lowest potential energy under E = 0, 0.5, and 1.5 V/nm, respectively, revealing that each ice configuration is energetically favorable under the corresponding E-field, and the solid−solid phase transition is attributed to the change of electric field. The ice formation process is accompanied by the decrease of entropy and thus it is an enthalpy-driven process. In the amorphous ice, due to the disordering of water orientations, the total dipole moment of water molecules is almost zero so that the contribution of the polarization energy −D·E per water molecule to the total potential energy is approximately zero. In contrast, water molecules in the ordered hexagonal and rhombic ice have a large net dipole moment; therefore, the total potential energy (U + (−D·E)) is reduced due to the negative contribution of polarization energy −D·E. This explains the observation that at E = 0 the amorphous ice has the lowest energy whereas at E = 0.5 and 1.5 V/m the ordered ice has the lowest energy. After examining the effect of E-field on the structures and phase transition of water at low temperatures, we investigate the freezing and melting of confined water under the influence of an E-field. Figure 5 shows the temperature dependence of the potential energy U of water (per water molecule) under three different E-field strengths (0, 0.5, 1.5 V/nm). With T going up, we find an abrupt rise in U (4−5 kJ/mol), and the transition temperature varies as the E-field strength changes. The transition temperature under E = 0 V/nm is in good agreement with earlier MD studies on water confined between hydrophobic plates at atmospheric pressure without an Efield.22,30 The sudden rise in U and the hysteresis in the heating (black curve) and cooling (red curve) processes suggest the presence of a first-order solid−liquid transition. The hysteresis D
dx.doi.org/10.1021/jp500989t | J. Phys. Chem. A XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry A
Article
0.75 and 1.0 V/nm reveals the existence of an intermediate liquid state between the two solid phases. This finding is very similar to the pressure-induced phase transition behavior of a monolayer ice reported by Bai et al.63 Finally, we present the calculated phase diagram of water confined between two graphene sheets with an effective separation of 0.674 nm as a function of temperature T and E-field (Figure 8). It shows that the E-field-induced solid−solid
Figure 6. Mean-square displacement (MSD) of confined water at different temperatures under an E-field of 0, 0.5, and 1.5 V/nm. The left and right three panels correspond to the heating and cooling processes, respectively. The calculation of MSD is started from t0 = 49.0 ns and over the last 1 ns of 50 ns MD simulation.
Figure 8. Calculated phase diagram of water confined between two graphene sheets with an inner surface distance of 0.674 nm in the temperature−E-field plane. The black curve represents the solid− liquid transition line. The red and blue curves represent the solid− solid transition lines between the amorphous and hexagonal bilayer ice, and between the hexagonal and rhombic bilayer ice, respectively. Here, both solid−liquid transition and solid−solid transition are firstorder transitions. The phase diagram is obtained by performing ∼90 independent 10 ns MD runs with an interval of 10 K in temperature and 0.25 V/nm in E-field.
phase to liquid phase and indicate a clear first-order phase transition under respective field strength. In addition, the MSD at the transition temperature (red curves: 255 K at E = 0 V/nm, 280 K at E = 0.5 V/nm, 300 K at E = 1.5 V/nm) shows a marked hysteretic phenomenon when undergoing the heating/ cooling process and provides further evidence that all the three quasi-2D ice configurations found under E = 0, 0.5, and 1.5 V/ nm melt discontinuously. Interestingly, a previous study by Bai et al. reported that, for pressure-induced solid−solid phase transition of monolayer ice, one solid phase could transform into another solid phase through intermediate liquid phase.63 To examine whether the E-field-induced hexagonal−rhombic ice transition can occur via an intermediate liquid phase, we present in Figure 7 the lateral diffusion coefficients of confined water under different electric fields at 270 K. It can be seen that the confined water exists as solid ice under 0.5 V/nm and 1.25/1.5 V/nm, corresponding to respectively the hexagonal bilayer ice and rhombic bilayer ice. The increased lateral diffusion coefficient under an E-field of
phase transition from the amorphous to hexagonal bialyer ice and that from the hexagonal to rhombic bilayer ice are both first-order transitions. This is similar to the quasi-1D system of confined water inside carbon nanotubes under an electric field, where transitions between pentagonal and helical ice nanotubes are discontinuous.64 The solid−liquid transition between the amorphous/hexagonal bilayer ice formed under E = 0−0.75 V/ nm and the disordered liquid water is discontinuous, consistent with the first-order solid−liquid transition of nanoplateconfined water at different pressures (0.1 MPa−4 GPa).22,60,63 It is noted that under certain conditions, confined water could undergo continuous phase transition.60 The rhombic ice bilayer formed under E = 0.75−1.5 V/nm also melts discontinuously, different from the phase transition induced by high pressure (under high water density),30 which indicates that the E-field- and pressure-induced water freezing occurs through different mechanisms. In addition, as the E-field increases from 0 to 1.5 V/nm at 270 K, the confined water undergoes the liquid state twice, similar to the emergence of two quasi-2D stable liquids respectively at low and high densities reported in a previous MD study.25 Meanwhile, the solid−solid transitions from amorphous to hexagonal bilayer or from hexagonal to rhombic bilayer is terminated by a critical end point (T = ∼260 K, E = ∼0.12 V/nm; or T = ∼265 K, E = ∼0.75 V/nm), because at high temperature the ice bialyers lose their identities and become disordered liquid water.
IV. CONCLUSIONS We have investigated the E-field-induced phase transition of water confined between two graphene sheets (H = 1 nm) by carrying out extensive all-atom MD simulations. The results
Figure 7. Lateral diffusion coefficients of confined water under different electric fields at 270 K. Snapshots for disordered liquid, hexagonal ice, and rhombic ice are given as insets, where oxygen atoms are in red, hydrogen atoms in white, and hydrogen bonds in cyan. E
dx.doi.org/10.1021/jp500989t | J. Phys. Chem. A XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry A
Article
(12) Bhushan, B.; Israelachvili, J. N.; Landman, U. Nanotribology: Friction, Wear and Lubrication at the Atomic Scale. Nature 1995, 374, 607−616. (13) Zhu, Y.; Granick, S. Viscosity of Interfacial Water. Phys. Rev. Lett. 2001, 87, 096104. (14) Uri, R.; Suzanne, G.; Joseph, F.; Jacob, K. Viscosity of UltraThin Water Films Confined between Hydrophobic or Hydrophilic Surfaces. J. Phys. Condens. Mater. 2002, 14, 9275. (15) Opitz, A.; Ahmed, S. I. U.; Schaefer, J. A.; Scherge, M. Friction of Thin Water Films: A Nanotribological Study. Surf. Sci. 2002, 504, 199−207. (16) Brooks, C. L.; Onuchic, J. N.; Wales, D. J. Taking a Walk on a Landscape. Science 2001, 293, 612−613. (17) Zhou, R.; Huang, X.; Margulis, C. J.; Berne, B. J. Hydrophobic Collapse in Multidomain Protein Folding. Science 2004, 305, 1605− 1609. (18) Liu, P.; Huang, X.; Zhou, R.; Berne, B. J. Observation of a Dewetting Transition in the Collapse of the Melittin Tetramer. Nature 2005, 437, 159−162. (19) Giovambattista, N.; Lopez, C. F.; Rossky, P. J.; Debenedetti, P. G. Hydrophobicity of Protein Surfaces: Separating Geometry from Chemistry. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 2274−2279. (20) Berne, B. J.; Weeks, J. D.; Zhou, R. Dewetting and Hydrophobic Interaction in Physical and Biological Systems. Annu. Rev. Phys. Chem. 2009, 60, 85−103. (21) Koga, K.; Zeng, X. C.; Tanaka, H. Freezing of Confined Water: A Bilayer Ice Phase in Hydrophobic Nanopores. Phys. Rev. Lett. 1997, 79, 5262−5265. (22) Koga, K.; Tanaka, H.; Zeng, X. C. First-Order Transition in Confined Water between High-Density Liquid and Low-Density Amorphous Phases. Nature 2000, 408, 564−567. (23) Zangi, R.; Mark, A. E. Bilayer Ice and Alternate Liquid Phases of Confined Water. J. Chem. Phys. 2003, 119, 1694−1700. (24) Zangi, R.; Mark, A. E. Monolayer Ice. Phys. Rev. Lett. 2003, 91, 025502. (25) Kumar, P.; Buldyrev, S. V.; Starr, F. W.; Giovambattista, N.; Stanley, H. E. Thermodynamics, Structure, and Dynamics of Water Confined between Hydrophobic Plates. Phys. Rev. E 2005, 72, 051503. (26) Giovambattista, N.; Rossky, P. J.; Debenedetti, P. G. Effect of Pressure on the Phase Behavior and Structure of Water Confined between Nanoscale Hydrophobic and Hydrophilic Plates. Phys. Rev. E 2006, 73, 041604. (27) Rodgers, J. M.; Weeks, J. D. Interplay of Local HydrogenBonding and Long-Ranged Dipolar Forces in Simulations of Confined Water. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 19136−19141. (28) Giovambattista, N.; Rossky, P. J.; Debenedetti, P. G. Effect of Temperature on the Structure and Phase Behavior of Water Confined by Hydrophobic, Hydrophilic, and Heterogeneous Surfaces. J. Phys. Chem. B 2009, 113, 13723−13734. (29) Giovambattista, N.; Rossky, P. J.; Debenedetti, P. G. Phase Transitions Induced by Nanoconfinement in Liquid Water. Phys. Rev. Lett. 2009, 102, 050603. (30) Han, S.; Choi, M. Y.; Kumar, P.; Stanley, H. E. Phase Transitions in Confined Water Nanofilms. Nat. Phys. 2010, 6, 685− 689. (31) Mallamace, F.; Broccio, M.; Corsaro, C.; Faraone, A.; Majolino, D.; Venuti, V.; Liu, L.; Mou, C.-Y.; Chen, S.-H. Evidence of the Existence of the Low-Density Liquid Phase in Supercooled, Confined Water. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 424−428. (32) Kimmel, G. A.; Matthiesen, J.; Baer, M.; Mundy, C. J.; Petrik, N. G.; Smith, R. S.; Dohnálek, Z.; Kay, B. D. No Confinement Needed: Observation of a Metastable Hydrophobic Wetting Two-Layer Ice on Graphene. J. Am. Chem. Soc. 2009, 131, 12838−12844. (33) Yang, D.-S.; Zewail, A. H. Ordered Water Structure at Hydrophobic Graphite Interfaces Observed by 4D, Ultrafast Electron Crystallography. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 4122−4126. (34) Malkin, T. L.; Murray, B. J.; Brukhno, A. V.; Anwar, J.; Salzmann, C. G. Structure of Ice Crystallized from Supercooled Water. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 1041−1045.
show that the confined water can form amorphous, hexagonal, or rhombic bilayer ice at low temperature, depending on the strengths of electric fields. The bialyer ice can transform into each other through a first-order phase transition, which mainly results from the interplay between strong intermolecular hydrogen bonds and collective water dipole orientation along the E-field direction. Interestingly, we find that the solid−solid phase transition between the hexagonal and rhombic bilayer ice can take place via an intermediate liquid phase. The temperature-induced solid−liquid transition from rhombic ice to liquid water or vice versa occurs discontinuously under the influence of E-fields, accompanied by a hysteresis loop, which is different from the continuous solid−liquid phase transition under high pressure.30 The existence of an external electric field significantly raises the freezing point of the quasi 2D-confined water.
■
AUTHOR INFORMATION
Corresponding Author
*G. Wei: e-mail,
[email protected]; tel, 86-21-55665231. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We acknowledge financial supports from the National Natural Science Foundation of China (11074075). Simulations were performed at the Shanghai Supercomputing Center and the National High Performance Computing Center of Fudan University.
■
REFERENCES
(1) Fu, D.; Libson, A.; Miercke, L. J. W.; Weitzman, C.; Nollert, P.; Krucinski, J.; Stroud, R. M. Structure of a Glycerol-Conducting Channel and the Basis for Its Selectivity. Science 2000, 290, 481−486. (2) Murata, K.; Mitsuoka, K.; Hirai, T.; Walz, T.; Agre, P.; Heymann, J. B.; Engel, A.; Fujiyoshi, Y. Structural Determinants of Water Permeation through Aquaporin-1. Nature 2000, 407, 599−605. (3) de Groot, B. L.; Grubmüller, H. Water Permeation across Biological Membranes: Mechanism and Dynamics of Aquaporin-1 and GlpF. Science 2001, 294, 2353−2357. (4) Sui, H.; Han, B.-G.; Lee, J. K.; Walian, P.; Jap, B. K. Structural Basis of Water-Specific Transport through the AQP1 Water Channel. Nature 2001, 414, 872−878. (5) Beckstein, O.; Sansom, M. S. P. Liquid−Vapor Oscillations of Water in Hydrophobic Nanopores. Proc. Natl. Acad. Sci. U. S. A. 2003, 100, 7063−7068. (6) Jensen, M. Ø.; Tajkhorshid, E.; Schulten, K. Electrostatic Tuning of Permeation and Selectivity in Aquaporin Water Channels. Biophys. J. 2003, 85, 2884−2899. (7) Majumder, M.; Chopra, N.; Andrews, R.; Hinds, B. J. Nanoscale Hydrodynamics: Enhanced Flow in Carbon Nanotubes. Nature 2005, 438, 44−44. (8) Li, J.; Gong, X.; Lu, H.; Li, D.; Fang, H.; Zhou, R. Electrostatic Gating of a Nanometer Water Channel. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 3687−3692. (9) Genzer, J.; Efimenko, K. Creating Long-Lived Superhydrophobic Polymer Surfaces through Mechanically Assembled Monolayers. Science 2000, 290, 2130−2133. (10) Erbil, H. Y.; Demirel, A. L.; Avcı, Y.; Mert, O. Transformation of a Simple Plastic into a Superhydrophobic Surface. Science 2003, 299, 1377−1380. (11) Lafuma, A.; Quere, D. Superhydrophobic States. Nat. Mater. 2003, 2, 457−460. F
dx.doi.org/10.1021/jp500989t | J. Phys. Chem. A XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry A
Article
(35) Vaitheeswaran, S.; Rasaiah, J. C.; Hummer, G. Electric Field and Temperature Effects on Water in the Narrow Nonpolar Pores of Carbon Nanotubes. J. Chem. Phys. 2004, 121, 7955−7965. (36) Dzubiella, J.; Allen, R. J.; Hansen, J. P. Electric Field-Controlled Water Permeation Coupled to Ion Transport through a Nanopore. J. Chem. Phys. 2004, 120, 5001−5004. (37) Bratko, D.; Daub, C. D.; Leung, K.; Luzar, A. Effect of Field Direction on Electrowetting in a Nanopore. J. Am. Chem. Soc. 2007, 129, 2504−2510. (38) Bratko, D.; Daub, C. D.; Luzar, A. Field-Exposed Water in a Nanopore: Liquid or Vapour? Phys. Chem. Chem. Phys. 2008, 10, 6807−6813. (39) Zangi, R.; Mark, A. E. Electrofreezing of Confined Water. J. Chem. Phys. 2004, 120, 7123−7130. (40) Choi, E.-M.; Yoon, Y.-H.; Lee, S.; Kang, H. Freezing Transition of Interfacial Water at Room Temperature under Electric Fields. Phys. Rev. Lett. 2005, 95, 085701. (41) Vaitheeswaran, S.; Yin, H.; Rasaiah, J. C. Water between Plates in the Presence of an Electric Field in an Open System. J. Phys. Chem. B 2005, 109, 6629−6635. (42) Qiu, H.; Guo, W. Electromelting of Confined Monolayer Ice. Phys. Rev. Lett. 2013, 110, 195701. (43) Saitta, A. M.; Saija, F.; Giaquinta, P. V. Ab Initio Molecular Dynamics Study of Dissociation of Water under an Electric Field. Phys. Rev. Lett. 2012, 108, 207801. (44) Lindahl, E.; Hess, B.; van der Spoel, D. GROMACS 3.0: A Package for Molecular Simulation and Trajectory Analysis. J. Mol. Model. 2001, 7, 306−317. (45) Kumar, P.; Han, S. Dynamics of Two-Dimensional Monolayer Water Confined in Hydrophobic and Charged Environments. J. Chem. Phys. 2012, 137, 114510. (46) Mahoney, M. W.; Jorgensen, W. L. A Five-Site Model for Liquid Water and the Reproduction of the Density Anomaly by Rigid, Nonpolarizable Potential Functions. J. Chem. Phys. 2000, 112, 8910− 8922. (47) Baez, L. A.; Clancy, P. Phase Equilibria in Extended Simple Point Charge Ice-Water Systems. J. Chem. Phys. 1995, 103, 9744− 9755. (48) Yamada, M.; Mossa, S.; Stanley, H. E.; Sciortino, F. Interplay between Time-Temperature Transformation and the Liquid-Liquid Phase Transition in Water. Phys. Rev. Lett. 2002, 88, 195701. (49) Essmann, U.; Perera, L.; Berkowitz, M. L.; Darden, T.; Lee, H.; Pedersen, L. G. A Smooth Particle Mesh Ewald Method. J. Chem. Phys. 1995, 103, 8577−8593. (50) Hummer, G.; Rasaiah, J. C.; Noworyta, J. P. Water Conduction through the Hydrophobic Channel of a Carbon Nanotube. Nature 2001, 414, 188−190. (51) Nosé, S. A Molecular Dynamics Method for Simulations in the Canonical Ensemble. Mol. Phys. 1984, 52, 255−268. (52) Hoover, W. G. Canonical Dynamics: Equilibrium Phase-Space Distributions. Phys. Rev. A 1985, 31, 1695−1697. (53) Parrinello, M.; Rahman, A. Polymorphic Transitions in Single Crystals: A New Molecular Dynamics Method. J. Appl. Phys. 1981, 52, 7182−7190. (54) Nosé, S.; Klein, M. L. Constant Pressure Molecular Dynamics for Molecular Systems. Mol. Phys. 1983, 50, 1055−1076. (55) Miyamoto, S.; Kollman, P. A. Settle: An Analytical Version of the SHAKE and RATTLE Algorithm for Rigid Water Models. J. Comput. Chem. 1992, 13, 952−962. (56) Hess, B.; Bekker, H.; Berendsen, H. J. C.; Fraaije, J. G. E. M. LINCS: A Linear Constraint Solver for Molecular Simulations. J. Comput. Chem. 1997, 18, 1463−1472. (57) Yan, J. Y.; Patey, G. N. Heterogeneous Ice Nucleation Induced by Electric Fields. J. Phys. Chem. Lett. 2011, 2, 2555−2559. (58) Yan, J. Y.; Patey, G. N. Molecular Dynamics Simulations of Ice Nucleation by Electric Fields. J. Phys. Chem. A 2012, 116, 7057−7064. (59) Yan, J. Y.; Patey, G. N. Ice Nucleation by Electric Surface Fields of Varying Range and Geometry. J. Chem. Phys. 2013, 139, 144501.
(60) Bai, J.; Zeng, X. C. Polymorphism and Polyamorphism in Bilayer Water Confined to Slit Nanopore under High Pressure. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 21240−21245. (61) Hansen, J. P.; McDonald, I. Theory of Simple Liquids, 3rd ed.; Academic Press: London, U.K., 2006. (62) Hedges, L. O.; Jack, R. L.; Garrahan, J. P.; Chandler, D. Dynamic Order-Disorder in Atomistic Models of Structural Glass Formers. Science 2009, 323, 1309−1313. (63) Bai, J.; Angell, C. A.; Zeng, X. C. Guest-Free Monolayer Clathrate and Its Coexistence with Two-Dimensional High-Density Ice. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 5718−5722. (64) Fu, Z.; Luo, Y.; Ma, J.; Wei, G. Phase Transition of NanotubeConfined Water Driven by Electric Field. J. Chem. Phys. 2011, 134, 154507−154506.
G
dx.doi.org/10.1021/jp500989t | J. Phys. Chem. A XXXX, XXX, XXX−XXX