Effect of Surface Chemistry on Water Interaction with Cu(111

Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

Effect of surface chemistry on water interaction with Cu(111) Andrew C. Antony, Tao Liang, Sneha A. Akhade, Michael J. Janik, Simon R. Phillpot, and Susan B. Sinnott Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.6b01974 • Publication Date (Web): 21 Jul 2016 Downloaded from http://pubs.acs.org on July 30, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Effect of Surface Chemistry on Water Interaction with Cu(111) Andrew C. Antony1†, Tao Liang1†, Sneha A. Akhade2, Michael J. Janik2, Simon R. Phillpot1, and Susan B. Sinnott1†* 1

Department of Materials Science and Engineering, University of Florida, Gainesville, Florida, 32611, USA. 2

Department of Chemical Engineering, Pennsylvania State University, University Park, Pennsylvania, 16802, USA †

Indicates the author has moved and is now affiliated with Department of Materials Science and Engineering, Pennsylvania State University, University Park, Pennsylvania, 16802, USA *Corresponding author email: [email protected] Keywords: molecular dynamics, spreading rate, copper, water, surface chemistry

1 ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract The interfacial dynamics of water in contact with bare, oxidized, and hydroxylated copper surfaces are examined using classical molecular dynamics (MD) simulations. A third generation charge optimized many-body (COMB3) potential is used in the MD simulations to investigate the adsorption of water molecules on Cu(111) and the results are compared to the findings of density functional theory (DFT) calculations. The adsorption energies and structures predicted by COMB3 are generally consistent with those determined with DFT. The COMB3 potential is then used to investigate the wetting behavior of water nanodroplets on Cu(111) at 20, 130 and 300 K. At room temperature, the simulations predict that the spreading rate of the base radius, R0, of a water droplet with a diameter of about 1.5 nm exhibits a spreading rate of R0 ~ t0.16 and a final base radius of 3.5 nm. At 20 and 130 K, water droplets are predicted to retain their structure after adsorption on Cu(111) and to undergo minimal spreading in agreement with scanning tunneling microscopy data. When the same water droplet encounters a reconstructed, oxidized Cu(111) surface, the classical MD simulations predict wetting with a spreading rate of R ~ t0.14 and a final base radius of 3.0 nm. Similarly, our MD simulations predict a spreading rate of R ~ t0.14 and a final base radius of 2.5 nm when water encounters OH-covered Cu(111). These results indicate that oxidation and hydroxylation causes a reduction in the degree of spreading and final base radius that is directly associated with a decreased spreading rate for water nanodroplets on copper.


Page 2 of 33

Page 3 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

1. Introduction Solids are in constant contact with water in a vast number of applications with important consequences. Corrosion of alloys in underwater environments1, electrocatalysis2-3, and electrodeposition all may occur at water/solid interfaces. Solution phase heterogeneous catalysis4, as well as frictional and wear characteristics of metals under water lubricating conditions5-6 are of interest for these ubiquitous applications. A molecular level understanding of this significant interface is desirable to guide control of processes in these surface science applications. Many surface science studies7-10 of water/metal interfaces have focused on the first water layer formed on wettable metal surfaces. The structure of the first water layer on metals, characterized using low-energy electron diffraction (LEED), was thought to be comprised of a two-dimensional (2D) epitaxial arrangement of molecules that resembles the (001) basal plane of ice I, often called the bilayer model11. With the use of high resolution spectroscopy techniques, including scanning tunneling microscopy (STM), improved structural insight of the local water arrangement on the metal surface was developed. Low-temperature STM studies report different water arrangements on Pt(111), Cu(110), and Cu(111)12-14, indicating that the structure of the first water layer varies with both the type of metal and the surface facet. Analysis of the STM images further indicates that water forms three-dimensional (3D) clusters when adsorbed at or below 20 K and annealed to 130 K on Cu(111) which has been defined7 as non-wetting behavior. This observation is in contrast to data from contact angle measurements of water droplets on oxygen-free, polycrystalline Cu films by Schrader15 under ultra-high vacuum conditions, where the droplets exhibited a 0° contact angle, which is an indication of complete wetting.


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 33

The behavior of liquid droplets on solid surfaces can provide detailed insight into the underlying structure and dynamics at the metal-liquid interface. The dynamic motion of precursor films (PFs) originating from liquid droplets on solid substrates has been extensively reviewed by Popescu et al.16 and observed experimentally using ellipsometry17, atomic force microscopy18, and more recently fluorescence microscopy19.

Classical molecular dynamics

(MD) studies report the propagation mechanism of the water PF on Au substrates as molecules diffuse from the droplet surface to the outer edge of the PF, finding that once molecules become part of the PF region, they are no longer mobile along the metal surface20. These MD simulations have typically used rigid fixed charge water potentials such as extended single point charge (SPC/E)21 and TIP4P22 which do an excellent job of describing molecular adsorption to the underlying surface but allow for only limited (if any) variations in atomic charges in response to changes in the surrounding environment. Additionally, the standard Lennard-Jones (LJ) pairwise potential used to describe liquid/solid interactions23-24 lacks the complexity needed to capture both the underlying physics of the liquid water and chemical bonding in solid metals and metal oxides. Here, we use a single, classical potential to model water molecules and liquid water in contact with surfaces of bare, oxidized, and hydroxylated copper using a single set of parameters. In particular, we use the formalism of the third generation charge optimized many body (COMB3) potential to calculate the energy of our systems. First, the adsorption of water molecules and molecular clusters on Cu(111) surfaces are determined and the COMB3 predictions are compared to adsorption energies and geometries calculated using DFT. Second, the dynamic wetting of Cu(111) by a water droplet is considered at 20, 130 and 300 K. The findings are compared to predictions of previous MD studies and to STM data. Third, the same


Page 5 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

water droplet is deposited at 300 K on reconstructed, oxidized Cu(111) and hydroxylated Cu(111) and the spreading behavior is characterized as a function of initial oxygen and hydroxide coverage. The paper concludes with a summary of the new insights gained from the classical simulations performed here. 2. Experimental Section/Simulation Methods 2.1. DFT Calculations Electronic structure calculations were performed using the Vienna Ab-initio Simulation Package (VASP v. 5.3.5)25-27, a DFT plane-wave pseudopotential code implementing the Projector Augmented Wave (PAW)28-29 method for core-valence treatment. The Perdew-BurkeErnzerhof (PBE)30 exchange-correlation functional described within the Generalized Gradient Approximation (GGA)

31

was employed for all calculations. The plane-wave cutoff energy was

450 eV. A Fermi smearing (σ) of 0.2 eV for Cu containing systems and 0.003 eV for isolated water molecules was applied using the Methfessel-Paxton32 scheme. Geometric optimizations were conducted using the Quasi-Newton33 optimization algorithm until the nuclei force convergence limit of 0.02 eV Å-1 was achieved. The self-consistent electronic convergence limit was set to 1×10-5 eV. The potential energy, interatomic forces and the stress tensor were corrected to include van der Waals contributions using the DFT-D3 method34 with BeckeJonson35 damping. The Brillouin zones of surfaces were sampled using a 5×5×1 MonkhorstPack36 mesh. A periodic copper slab model was constructed using an experimental lattice constant of 3.62 Å. The Cu(111) surface consisted of a 3×3 hexagonal supercell with four atomic layers that contained 9 atoms per layer. The bottom two layers of the slab were fixed while the top two layers were allowed to relax. A 15 Å vacuum thickness was added in the surface normal


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 33

direction to exclude the interaction between periodic slab models. Water adsorption on the Cu(111) slab was considered at coverages of 1/9 (monomer), 2/9 (dimer), and 2/3 (hexamer). 2.2. Formalism and Fitting of the COMB3 Potential The total potential energy (Utot) within the COMB3 formalism includes the electrostatic energy (Ues), charge-dependent short-range energy (Ushort), the van der Waals energy (UvdW) and correction terms (Ucorr), as described by the following equation: , = , + , + +

(1)

where q and r are the charge and coordinate arrays of the system. The dynamic charge on each atom is determined by the charge equilibration scheme first introduced in the work of Rick et al.37. A detailed description of the potential formalism along with the charge equilibration scheme can be found in Ref.

38

. The parameterization for COMB3 used the Parameterization

Optimization Software for MATerials (POSMat)39 to determine the most acceptable parameter set. POSMat uses an iterative optimization technique (either simplex or simulated annealing) to minimize a weighted cost function imposed on each structure and property in the fitting database. To ensure transferability of the COMB3 O/H parameters, we included heats of formation for the following structures in the fitting database: hydroxyl radical (OH), hydrogen peroxide (H2O2), H2O, H3O, and H3O2 molecules and crystalline ice-Ih. Table 1 reports energies and structural details predicted by COMB3 for a single H2O molecule and crystalline ice-Ih. This newly developed COMB3 potential for H2O systems was seamlessly coupled with previously developed COMB3 potentials for the copper-hydrocarbon systems40 and Cu2O41. The supplemental material contains a table of all the calculated heats of formation of the molecular structures listed above and a list of COMB3 parameters used in the present work.


Page 7 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

The COMB3 potentials reported here were implemented in the open source Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) software package42 distributed by Sandia National Laboratory.

Table 1. List of properties of H2O systems calculated with COMB3 and compared with the predictions of common empirical water potentials TIP3P22, TIP4P22, SPC21, and SPC/E21, and experimental data. The dipole moment within TIP3P, TIP4P, SPC, SPC/E, and COMB3 was estimated using qO × 0.6, where 0.6 Å is the distance between positive and negative charge centers. All experimental properties for molecular H2O are from Ref. 43. Properties H2O O-H length H-O-H angle (º) ∆Hf (eV/mole.) Charge on O, qO (e) Dipole moment (eÅ) Ice Ih a (Å) c (Å) O-O distance (Å) ∆Hf (eV/mole.) Charge on O, qO (e) Dipole moment (eÅ) (basal plane) (prism plane) a Ref. 44 b Ref. 45 c Ref. 46 d Ref. 47 e Ref. 48 f Ref. 49 g Ref. 50, calculated value h Ref. 51, calculated value i Ref. 52

TIP3P

TIP4P

SPC

SPC/E

Exp./DFT

COMB3

0.96 104.5 -0.83 0.50

0.96 104.5 -1.04 0.62

1.0 109.5 -0.82 0.49

1.0 109.5 -0.82 0.49

0.95 104.5 -2.50 0.39

0.95 107.6 -2.41 -0.71 0.42

2.74d -0.83 0.50 -

4.49a 7.32a 2.68d -1.04 0.62 -

2.70d -0.82 0.49 -

4.63b 7.18b -0.82 0.49 -

4.52c 7.36c 2.76e -3.02f g 0.52 , 0.64h 0.010-0.018i 0.012-0.016i

4.55 7.39 2.78 -3.16 -0.87 0.53 0.010 0.017


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The data in Table 1 indicates that COMB3 accurately reproduces the energetics and geometry of a single H2O molecule and Ih-type crystalline ice. More importantly, the charges on the atoms dynamically adjust in response to the environment, a feature absent in non-reactive fixed charge water models such as TIP4P and SPC/E. For example, the charge on an oxygen atom in ice-Ih is -0.71 e, which is lower than its value of -0.87 e in an isolated H2O molecule. The O-O dipole moment of H2O can be estimated using a point charge model as qO × 0.6 eÅ, where qO is the charge on an O atom and 0.6 Å is the distance between positive and negative charge centers. Using this estimation, the calculated dipole moment using COMB3 is 0.42 eÅ (2.02 D) for an isolated H2O molecule and 0.53 eÅ (2.55 D) in Ih ice, which is in close agreement with the predictions of first principles calculations50 and about 0.11 eÅ less than the theoretical value calculated by Batista et al. 51 using a self-consistent induction model. It should be emphasized that COMB3 is a reactive potential, which means that the water molecules are not rigid and can change geometry or dissociate in response to changes in their physical environment. These unique features enable COMB3 to describe environmental dependent properties of water such as the standard heat of formation, induced dipole moment, and hydrogen bonds in water, as discussed in the next section. 2.3. COMB3 Water The efficacy of COMB3 is tested by computing molecular and bulk scale properties of liquid water. Table 2 presents some properties of liquid water obtained from canonical ensemble MD simulations with COMB3. The system consisted of 297 water molecules in a 18.17 Å × 18.17 Å × 18.17 Å supercell (ρH2O = 0.99 g/cm3). A Langevin thermostat was applied to all of the molecules to regulate a simulation temperature of 300 K. The simulation ran for 5 ns and included 1 ns of equilibration time. The table also provides a comparison of liquid water


Page 8 of 33

Page 9 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

properties calculated using COMB3 and those predicted with the rigid, fixed charge water potentials TIP3P22, TIP4P22, SPC21, and SPC/E21 as well as with neutron diffraction experimental data. The predicted heat of formation for liquid water using COMB3 is -2.81 eV/H2O or -271 kJ/mol, which is about 5% higher than the experimental value of the heat of formation for liquid water at 300 K and 1 atm. The optimal density of water at 300 K predicted by COMB3 is 1.00 g/cm3, in agreement with experimental values. The COMB3 potential yields a dipole moment that is smaller than the value measured experimentally with a deviation of comparable magnitude to the deviation of predictions from both the TIP4P and SPC/E potentials from experiment. While TIP4P and SPC/E both overestimate the diffusion coefficient somewhat, TIP3P overestimates the diffusion coefficient by more than a factor of two. The value predicted by COMB3 also deviates from the experimental diffusion by underestimating the experimental value by a factor of two.

Table 2. Comparison of bulk water properties from MD simulations using COMB3 or nonreactive water potentials (TIP3P, TIP4P, SPC, and SPC/E) with data from neutron diffraction experiments. Properties ∆Hf (eV/H2O) Density (g/cm3) Average water dipole moment (eÅ) Diffusion coefficient (10-5 cm2/s) a Ref. 54 b Ref. 55

TIP3P53

TIP4P

SPC53

SPC/E

Experiment

0.99 0.49

0.99 0.45

0.98 0.47

1.02 0.49

-2.96 1.00 0.60a

COMB 3 -2.81 1.00 0.45

5.30

3.9

4.02

2.52

2.3b

0.96


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 33

The radial distribution function g_ij(r) (RDF) was computed to provide a structural description of the bonding interactions between water molecules. Figure 1 illustrates the radial distribution functions g_OO(r), g_HH(r) and g_OH(r) calculated using either COMB3 or SPC/E, along with data from neutron diffraction experiments56. The RDFs for liquid water predicted by COMB3 have qualitative merit but Fig. 1 shows that peak heights are inconsistent with, and shifted relative to, both the SPC/E model and neutron diffraction data.

Figure 1. Radial distribution functions of liquid water: (a) oxygen-oxygen, (b) hydrogenhydrogen, and (c) oxygen-hydrogen for experimental neutron diffraction data, SPC/E, and COMB3. Experimental and SPC/E results overlap in (a).

These comparisons illustrate that the COMB3 potential provides a reasonable qualitative description of the properties of liquid bulk water, but shortcomings with regard to self-diffusion coefficient and RDFs are apparent. Both of these inaccuracies may be attributed to the transferability of this complex, reactive, dynamic charge potential. 10 ACS Paragon Plus Environment

In this instance,

Page 11 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

transferability is used to mean that the O-H parameters used for liquid water simulations are the same elemental parameters used for the simulations in section 3.2 of this work where OH is adsorbed at a metal surface. These parameters can also be employed for any system containing oxygen and hydrogen atoms (including ice-Ih, C-H-O systems, ZrH, Cu-SiO2, Cu adsorption on ZnO, among several others38). We did not fit the COMB3 potential to reproduce specific properties of liquid water, but rather included O/H molecules and crystalline ice-Ih for transferability. For reactive, dynamic charge potentials, generating a single parameter set for atoms in different physical environments is an arduous task due to the inclusion of many varied structures (and thus, bonding environments) in the fitting database. Thus, although COMB3 sacrifices a degree of quantitative accuracy for some liquid water properties, we highlight the importance of its ability to retain transferability. In particular, as this work is focused on obtaining a mechanistic understanding of the behavior of wetting on bare and chemically variant copper surfaces, the influence of the variable charge scheme on the interfacial structure and dynamics of water is important. Future efforts to improve upon the reproducibility of the water self-diffusion coefficient are planned but are outside the scope of the present work. 2.4. Water Adsorption on Cu(111) In this section we report adsorption energies and structural information for molecular water on Cu(111). The results are used to determine whether the previously developed COMB3 Cu/O/H parameters40 reproduce the correct geometries and energetics of molecular adsorption on Cu. Thus, the results from the DFT calculations reported here were not included in the COMB3 fitting database (i.e. adsorption energy (Eads) of H2O hexamer on Cu(111)).


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 33

Water adsorption energies and geometries on Cu(111) using the COMB3 potential were determined using a conjugate gradient energy minimization method. The substrate for these calculations was a 31 Å × 31 Å Cu(111) surface consisting of 168 surface atoms and nine atomic layers for a total of 1512 atoms in the Cu slab; 10 Å of vacuum was included above the surface to prevent periodic slab-to-slab interactions in the surface normal direction. Adsorption energy minimization was performed for a water monomer, dimer, and a 2/3 (112 water molecules) surface coverage arranged in a hexagonal ice (ice Ih) configuration. In particular, adsorption energies (Eads) were calculated as: =

!"#$

(

* +,-.

where

%#$(&&&) % ! ∗) ! ) ) !

(2)

is the total energy of the Cu(111) surface with adsorbed H2O molecules,

-.(///) is the total energy of the bare Cu(111) slab, * water molecules, and 0*

+

+

is the per-molecule energy of isolated

is the number of water molecules adsorbed on the surface.

For single water molecule adsorption, two Cu(111) surface sites were identified as possible adsorption locations: one directly above a Cu surface atom (‘atop’), and one directly above the center of three adjacent surface atoms (‘hollow’)57-59. DFT and COMB3 predict negative adsorption energies for both of these configurations indicating that both atop and hollow sites are favorable for molecular water adsorption. Figure 2 illustrates the geometric configuration of a single water molecule absorbed on atop and hollow sites as predicted by COMB3. For the atop adsorption site, the angle (θdipole) that the water molecule forms with the surface normal direction is 83.7° as calculated by DFT and 78.9° using COMB3 (see Fig. 2). DFT calculations indicate a Cu-O distance of 2.35 Å while COMB3 predicts a distance of 2.52 Å, a value within 0.2 Å of the DFT results.


Page 13 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 2. Configuration of water monomers adsorbed on the (a) atop site and (b) hollow site of Cu(111). Side view (above) and top-down view (below) are shown for both adsorption sites. The molecule’s dipole vector forms an angle, θdipole, with the Cu(111) surface normal vector. Cu atoms are yellow, O atoms are red, and H atoms are grey.

Adsorption energies of a H2O monomer, dimer, and 2/3 surface coverage (hexamer) are reported in Table 3 and compared to DFT calculations that include van der Waals corrections, as discussed in section 2.1. COMB3 predicts the opposite order of adsorption site stability (atop versus hollow) for monomers compared to DFT. We attribute this discrepancy to using universal, fully transferrable Cu-O and Cu-H parameters that were previously developed for organic-copper interactions41. This observation may result in isolated H2O molecules at the Cu(111) surface occupying a larger fraction of hollow sites.


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 33

Table 3. Adsorption energies (eV/molecule) for H2O monomer, dimer, and hexamer. Atop and hollow site adsorption energies are reported for a single water molecule on Cu(111). The charges of atoms in the H2O monomer, qO and qH, are reported for both adsorption sites.

Cu(111)+H2Omonomer Cu(111)+H2Odimer Cu(111)+H2Ohexamer

DFT atop -0.40 -0.55 -0.73

COMB3 atop -0.49 -0.59 -0.91

DFT hollow -0.28 -

COMB3 hollow -0.54 -

For adsorbed dimer and hexamer water clusters, the average Cu-O distances predicted by COMB3 are 2.77 Å and 2.80 Å, respectively. DFT calculations predict distances of 2.60 Å (dimer) and 3.14 Å (hexamer) between O atoms and Cu(111) surface atoms. We define the 2/3 ML coverage as two absorbed water molecules per three surface Cu atoms, as illustrated in Fig. 3. The hexagonal geometry of the 2/3 ML coverage is retained after energy minimization for both DFT and COMB3 calculations.

Figure 3. Equilibrated 2/3 ML water surface structure from DFT calculations (top-down view). The figure includes periodic images from the 3×3 hexagonal unit cell (blue region). The color scheme is: Cu atoms are yellow, O atoms are red, H atoms are white.


Page 15 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

The data in Table 3 indicates that COMB3 adequately predicts trend of stronger molecular adsorption with increasing water surface coverage, a result of hydrogen bonding among water molecules. 3. Results and Discussion 3.1. Water Droplet Spreading On Cu(111) as a Function of Temperature Classical MD simulations with COMB3 are used to examine the dynamics associated with water droplets with an initial diameter of 2.82 nm interacting with Cu(111) at three different temperatures; the droplet contains 576 H2O molecules. The structure of the initial H2O configuration is obtained by running an NPT MD simulation of ice-Ih at 300 K for 200 ps, causing the hexagonal structure to become amorphous. The ice-Ih simulation cell is orthorhombic, resulting in a non-spherical shape for the initial water droplet. The copper surface consists of three layers with a surface area of 142 Å × 143 Å and 10,752 Cu atoms. Throughout this work, the Z-axis is defined as the surface normal direction and the radial direction is defined as √2 3 + 4 3 (labeled as the XY-axis). These axes are clearly marked in a side-view simulation snapshot depicted in Fig. 4. The system is maintained at a constant-volume, constant-temperature ensemble at 300 K. The bottom layer of the Cu slab is held rigid while the water droplet and the top two Cu layers are coupled to a Langevin thermostat with a damping time constant of 100.0 ps. Using a Langevin thermostat adds thermal energy forces to the preexisting forces from the employed COMB3 potential. Thermal energy transfer occurs between hydrogen bonds in water, which in COMB3 are represented within the Coulombic energy term. The forces added from the Langevin thermostat are a small fractional number of the bond energy found in Cu (short range metallic bonding) but large compared to hydrogen bonding in water. To account for the larger relative 15 ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 33

forces, and to reduce the amount of charge fluctuation, the input simulation temperature for liquid water is 370 K in order to produce an output of 300 K. On the other hand, an input temperature of 300 K for Cu atoms results in an output temperature of 300 K. Simulations were carried out at three different temperatures: 300, 130, and 20 K. The simulations considering the interaction of the water droplet with the Cu surface at 300 K are comparable to conditions used in the contact angle experiments of Schrader15. A simulation temperature of 130 K is used to compare simulation results with annealing conditions of experimental STM conditions. The results of these experiments12 carried out by Morgenstern’s group report the onset of crystallization from amorphous solid water (ASW) at 130 K. At this annealing temperature, STM images show formation of islands on the Cu(111) surface. More importantly, these images show a higher degree of clustering at increasing temperatures, indicating a tendency of H2O molecules to spread across Cu(111) and form large clusters at or above 130 K. The simulation results at 20 K are compared to low-temperature STM data where molecular water is dosed onto Cu(111). Results of low-temperature STM experiments from the Morgenstern group60-61 show that for temperatures at or below 20 K, water forms large amorphous clusters with no apparent order. During dissociation experiments, the group reports that the Cu(111) samples are cooled to below 20 K in order to avoid clustering of water molecules during deposition.


Page 17 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 4. Side-view snapshot from the MD simulation of a water droplet forming on a Cu(111) surface at 300 K. Dsurf is the thickness of droplet surface water and Dint is the thickness of interfacial water. RXY(Z) is the effective radius of droplet at height Z, the area of such a circle is equal to the area of the water droplet at that height. RZmax is the maximum height of the droplet. Water droplet regions are identified as: (1) droplet surface (rZ > Dint and rXY > RXY(Z)-Dsurf or rZ > Dint and rZ >RZmax-Dsurf); (2) PF (rZ RXY(Dint)); (3) interfacial region (rZ ?_24B)C D =>?_24.C

(3)

The predicted MSD_XY and MSD_W is illustrated in Fig. 8(a) for the four regions of the droplet at 0.4 ns. In addition, MSD_W is reported as a function of droplet height in Fig. 8(b). Since the number of molecules in each region is dynamic and unequal, we present the data in Fig. 8 relative to the average MSD_XY and MSD_W of all molecules in the droplet. Analyzing the data in this manner ensures that we are not overestimating the displacement of molecules in one region solely because it contains more molecules than the others. 21 ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 33

Figure 8. (a) MSD_XY and MSD_W of each region of the droplet illustrated in Fig. 4. The values are reported relative to MSD_XY and MSD_W of all the droplet molecules. (b) MSD of wetting (MSD_W) is calculated for droplet surface molecules as a function of droplet height. The dotted horizontal line distinguishes molecules in the droplet surface from PF molecules. Dashed vertical lines represent average MSD_W of the two regions as seen in (a). All calculations are taken at 0.4 ns.

We address the issue of whether the PF itself is not diffusive, as described by Yuan and Zhao20.

By including molecules directly below bulk water as part of the PF, their work

concludes that the PF is not diffusive and exhibits the least mobility relative to the other molecules in the droplet. Our results are in agreement with these findings for the interfacial and PF regions, both of which exhibit the smallest relative MSD_XY of the water droplet. Similarly, both studies find significant diffusion of droplet surface molecules to the nominal contact region and metal surface. As shown in Fig. 8(a), the radial diffusivity of the droplet’s surface is five times higher than the diffusivity of any other region in the droplet. Besides the droplet surface region, only the PF region exhibits higher MSD_W relative to MSD_XY. The difference


Page 23 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

between MSD_W and MSD_XY is 0.1 in the PF region and 2.1 in the surface droplet region. These results indicate that the PF is relatively stationary while the droplet surface region constitutes a larger fraction of molecules at the surface near the end of the simulation. Additionally, although bulk water is more diffusive than the PF in the entire XY-plane, molecular movement parallel to the surface is random and is predicted to exhibit negative MSD_W, meaning that the radial displacement of bulk molecules has almost no contribution to the water droplet wetting. The dashed vertical lines in Fig. 8(b) indicate that the average MSD_W of the droplet surface and PF coincide with the MSD_W values from Fig. 8(a). Furthermore, Fig. 8(b) indicates that molecules near the top of the droplet (~12 Å above the Cu surface) display greater radial displacement in order to move to the outer edge of the droplet’s surface instead of moving directly toward the Cu surface. Conversely, molecules closer to the nominal contact region (intersection of PF region and droplet surface region) do not demonstrate as much radial displacement as they are already near the outermost edge of the droplet surface. From the above results, we note that the diffusion of water molecules from the surface of the droplet to the nominal contact region supplies additional molecules that can expand the PF region. Contrary to the highly diffusive molecules near the droplet’s surface, molecules closer to the nominal contact region remain stationary. From these observations and the 2D vantage point of Fig. 4, we conclude that the droplet “unwraps” itself so that molecules originally near the top of the droplet surface region (region 1 in Fig. 4) become part of the outer edge of the advancing water PF (see inset of Fig. 6). This conclusion is consistent with Yuan’s report concluding that the molecules near the nominal contact region do not propagate across the surface, but rather, molecules from the droplet’s surface move downward to form the final PF. 3.2. Water interacting with oxidized and hydroxylated Cu(111)


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 33

To demonstrate transferability of the COMB3 potential, we investigate the effect of Cu(111) surface oxidation and hydroxyl coverage on the spreading rate of water. We reduce the dimensions of the Cu(111) slab to have surface dimensions 11.0 nm x 11.0 nm in the X- and Ydirections and increase the thickness to 1.0 nm (6 atomic layers). Oxygen atoms are placed 1.9 Å from the Cu surface atoms at a coverage of 0.50 ML. In this work, we define O* atoms as adsorbed oxygen atoms in order to differentiate them from O atoms found in H2O. The O* atoms are placed in random adsorption sites to replicate an oxidized Cu surface63. Energy minimization is performed on the system using a conjugate gradient algorithm implemented in LAMMPS to allow the surface layer atoms to optimize their positions. The surface Cu atoms reconstruct to form a corrugated surface in good agreement with experimental data64 and the predictions of other computational studies65 for high coverage O adsorption on Cu(111). For the final simulation, we consider a Cu(111) surface with OH* coverage of 0.50 ML (OH* representing adsorbed hydroxyl groups). Again, a conjugate gradient energy minimization used in the previous sections is performed on the Cu(111)-OH* surface prior to introducing water molecules into the system and running molecular dynamics. After structural relaxation and using simulation details similar to those outlined in section 3.1, we perform NVT simulation of liquid water interacting with O* and OH* covered Cu(111). Snapshots of the initial MD simulation structure for both O* and OH* surfaces and their respective structures after 1.0 ns of simulation time are shown in Fig. 9.


Page 25 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 9. Perspective views of the initial setup of H2O interacting with (a) an oxidized Cu(111) surface (0.50 ML O*) and (b) 0.50 ML OH* Cu(111) surface. Snapshots after 1.0 ns of simulation time are shown in (c) and (d). Oxygen atoms from O* and OH* species are blue and O atoms from H2O are red. Hydrogen atoms from OH* species are white and H atoms from H2O are grey. Cu atoms are yellow. 3.3. Analysis of O* and OH* effects on spreading rate For both O* and OH* covered Cu(111), the droplet exhibits a lower spreading rate compared to the bare surface. These spreading rates are compared in Fig. 10, which indicates that water on O* and OH* covered Cu(111) has a spreading rate in closer agreement to the molecular-kinetic theory prediction66 of R ~ t1/7 compared to bare Cu(111). More interestingly, these droplets exhibit a smaller final base radius (~3.0 nm for O* surface and ~2.5 nm for OH* surface), indicating that the degree to which the droplet spreads is reduced on these surfaces. Classical MD simulations analyzing spontaneous spreading of a nanodroplet on surfaces with varying wettability indicate that decreasing the wettability of a surface reduces the rate at which 25 ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 33

the droplet spreads23. Our results indicate that both the degree of spreading and spreading rate are reduced on O* and OH* covered Cu(111). This finding indicates that O* and OH* decreases the wettability of water on Cu(111).

Figure 10: Spreading rate of single crystal Cu(111) surface (bare) and 0.50 ML O* and OH* covered Cu(111) after 1.5 ns of simulation time. The surfaces with pre-adsorbed species exhibit a lower degree of spreading and a slower spreading rate. A log-log plot of the data to verify the quality of exponent extraction is included in the supporting information.

Here, we propose a mechanism by which the spreading rate changes based on the chargecharge interactions at the interface. On bare Cu, hydrogen bonds within H2O compete with Cu-O Coulombic interactions at the surface. The Cu surface atoms have an approximate charge of +0.15 e and the O atoms from the H2O molecules in contact with the surface exhibit a charge of 0.88 e. These Cu-O interactions may therefore be thought of as relatively weak which allows the H2O molecules to readily move across the surface.


Page 27 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

When the Cu(111) surface is oxidized the simulation results indicate that O* atoms carry a charge of -0.75 e and H atoms have an approximate charge of +0.33 e. The resulting stronger Coulombic interaction present means that more energy is required to break the O*-H bonds than was the case for the bare Cu surface, which slows the movement of H2O molecules and reduces the rate at which they spread across the surface. Similarly, at the hydroxylated Cu(111)/H2O interface the H atoms from OH* carry a charge of +0.40 e and O atoms from the surface contact layer of H2O have an approximate atomic charge of -0.91 e, which results in a greater charge difference than was present in either the bare or oxidized Cu surface/water interface. The resulting strong Coulombic interactions at this interface cause the greatest reduction in water spreading rate. A detailed figure of the chemical species and their charges at the interface is given in the supporting information. 4. Summary and Conclusions We have presented a COMB3 potential capable of simulating solid copper and liquid water within the same dynamic charge framework. Results from COMB3 calculations indicate the ability of the potential to capture changes in oxygen charge and dipole moment with varying environments (Table 1). COMB3 adsorption properties are compared to DFT calculations for adsorbed water molecules on a Cu(111) surface. The results show that COMB3 overestimates binding energies for a single molecule, dimer, and 2/3 ML coverage. We apply our potential to simulate the spreading of a water droplet over a Cu(111) substrate at surface areas greater than 50 nm2. The dynamic description of a liquid water nanodroplet wetting across the Cu surface is summarized by the pronounced radial diffusivity of droplet surface molecules so that the final PF (image at 1.6 ns in Fig. 5) primarily consists of molecules originally part of the droplet’s surface. We also report that pre-adsorbed O* and OH* coverage decreases the degree of spreading over


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 33

Cu(111) and reduces the spreading rate to a value in closer agreement to the molecular kinetic theory value of R ~ t1/7. The work presented here, especially the development of a dynamic charge potential capable of capturing liquid water in contact with a chemically variant metal surface, provides an advanced computational technique for probing the properties of interfacial phenomena where charge transfer plays a crucial role. ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge via the Internet at http://pubs.acs.org. Additional O/H compounds and their heats of formation included in COMB3 fitting database, log-log plot of the R ~ tx relation from Fig. 10, a schematic of the charge-based spreading rate mechanism, and a complete set of COMB3 parameters used during MD simulations in LAMMPS, AUTHORINFORMATION Corresponding Author Email: [email protected] Tel: (814) 863-3117 ACKNOWLEDGEMENTS The authors gratefully acknowledge the support of the National Science Foundation CBET1264173. We also acknowledge helpful discussion and feedback regarding LAMMPS thermostat settings from T.R. Shan. REFERENCES (1) Syrett, B. C., Erosion-Corrosion of Coper-Nickel-Alloys in Sea-Water and Other Aqueous Environments - Literature Review. Corrosion 1976, 32, 242-252. (2) Gattrell, M.; Gupta, N.; Co, A., A review of the aqueous electrochemical reduction of CO2 to hydrocarbons at copper. J. Electroanal. Chem. 2006, 594, 1-19. (3) Welch, C. M.; Compton, R. G., The use of nanoparticles in electroanalysis: a review. Anal. Bioanal. Chem. 2006, 384, 601-619. (4) Chinchen, G. C.; Waugh, K. C.; Whan, D. A., The Activity and State of the Copper Surface in Methanol Synthesis Catalysts. Appl. Catal. 1986, 25, 101-107.


Page 29 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

(5) Shangguan, B.; Zhang, Y. Z.; Xing, J. D.; Sun, L. M.; Chen, Y., Study of the Friction and Wear of Electrified Copper against Copper Alloy under Dry or Moist Conditions. Tribol. T. 2010, 53, 927-932. (6) Mazurkiewicz, M.; Kubala, Z.; Chow, J., Metal Machining with High-pressure Water-jet Cooling Assistance - A New Possibility. J. Eng. Ind.-T. ASME 1989, 111, 7-12. (7) Carrasco, J.; Hodgson, A.; Michaelides, A., A molecular perspective of water at metal interfaces. Nat Mater 2012, 11, 667-674. (8) Henderson, M. A., The interaction of water with solid surfaces: fundamental aspects revisited. Surf. Sci. Rep. 2002, 46, 1-308. (9) Hodgson, A.; Haq, S., Water adsorption and the wetting of metal surfaces. Surf. Sci. Rep. 2009, 64, 381-451. (10) Verdaguer, A.; Sacha, G. M.; Bluhm, H.; Salmeron, M., Molecular structure of water at interfaces: Wetting at the nanometer scale. Chem. Rev. 2006, 106, 1478-1510. (11) Doering, D. L.; Madey, T. E., The Adsorption of Water on Clean and Oxygen-dosed Ru(001). Surf Sci. 1982, 123, 305-337. (12) Mehlhorn, M.; Morgenstern, K., Faceting during the Transformation of Amorphous to Crystalline Ice. Phys. Rev. Lett. 2007, 99, 246101. (13) Nie, S.; Feibelman, P. J.; Bartelt, N. C.; Thurmer, K., Pentagons and Heptagons in the First Water Layer on Pt(111). Phys. Rev. Lett. 2010, 105, 4. (14) Yamada, T.; Tamamori, S.; Okuyama, H.; Aruga, T., Anisotropic water chain growth on Cu(110) observed with scanning tunneling microscopy. Phys. Rev. Lett. 2006, 96, 036105. (15) Schrader, M. E., Ultrahigh vacuum techniques in the measurement of contact angles. III. Water on copper and silver. The Journal of Physical Chemistry 1974, 78, 87-89. (16) Popescu, M. N.; Oshanin, G.; Dietrich, S.; Cazabat, A. M., Precursor films in wetting phenomena. J. Phys. Condens. Matter 2012, 24, 243102. (17) Cazabat, A.; Fraysse, N.; Heslot, F.; Carles, P., Spreading at the microscopic scale. J. Phys. Chem. 1990, 94, 7581-7585. (18) Xu, H.; Shirvanyants, D.; Beers, K.; Matyjaszewski, K.; Rubinstein, M.; Sheiko, S. S., Molecular motion in a spreading precursor film. Phys. Rev. Lett. 2004, 93, 206103. (19) Hoang, A.; Kavehpour, H. P., Dynamics of Nanoscale Precursor Film near a Moving Contact Line of Spreading Drops. Phys. Rev. Lett. 2011, 106, 254501. (20) Yuan, Q. Z.; Zhao, Y. P., Precursor Film in Dynamic Wetting, Electrowetting, and Electro-ElastoCapillarity. Phys. Rev. Lett. 2010, 104, 246101. (21) Berendsen, H. J. C.; Grigera, J. R.; Straatsma, T. P., The Missing Term in Effective Pair Potentials. J. Phys. Chem. 1987, 91, 6269-6271. (22) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L., Comparison of Simple Potential Functions for Simulating Liquid Water. J. Chem. Phys. 1983, 79, 926-935. (23) Sedighi, N.; Murad, S.; Aggarwal, S. K., Molecular dynamics simulations of spontaneous spreading of a nanodroplet on solid surfaces. Fluid Dyn. Res. 2011, 43, 015507. (24) Wu, C.; Qian, T.; Sheng, P., Droplet spreading driven by van der Waals force: a molecular dynamics study. J. Phys.: Condens. Matter 2010, 22, 325101. (25) Kresse, G.; Furthmüller, J., Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 1996, 54, 11169. (26) Kresse, G.; Hafner, J., Ab initio molecular dynamics for liquid metals. Phys. Rev. B 1993, 47, 558. (27) Kresse, G.; Hafner, J., Ab initio molecular-dynamics simulation of the liquid-metal–amorphoussemiconductor transition in germanium. Phys. Rev. B 1994, 49, 14251. (28) Blöchl, P. E., Projector augmented-wave method. Phys. Rev. B 1994, 50, 17953. (29) Kresse, G.; Joubert, D., From ultrasoft pseudopotentials to the projector augmented-wave method. Phys. Rev. B 1999, 59, 1758. 29 ACS Paragon Plus Environment

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(30) Perdew, J. P.; Burke, K.; Ernzerhof, M., Generalized gradient approximation made simple. Phys. Rev. Lett. 1996, 77, 3865. (31) Perdew, J. P.; Chevary, J.; Vosko, S., K. Jackson A., Pederson MR, Singh DJ, et Fioljais. C.,“Atoms, molecules, solids ans surfaces: Applications of generalized gradient approximation for exchange and correlation. Phys. Rev. B 1992, 46, 6671. (32) Methfessel, M.; Paxton, A., High-precision sampling for Brillouin-zone integration in metals. Phys. Rev. B 1989, 40, 3616. (33) Pulay, P., Convergence acceleration of iterative sequences. The case of SCF iteration. Chem. Phys. Lett. 1980, 73, 393-398. (34) Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H., A consistent and accurate ab initio parametrization of density functional dispersion correction (DFT-D) for the 94 elements H-Pu. J. Chem. Phys. 2010, 132, 154104. (35) Grimme, S.; Ehrlich, S.; Goerigk, L., Effect of the damping function in dispersion corrected density functional theory. J. Comput. Chem. 2011, 32, 1456-1465. (36) Monkhorst, H., Physical Review B (Solid State) 13 (12), 5188 (1976); JD Pack and HJ Monkhorst. Physical Review B (Solid State) 1977, 16, 1748. (37) Rick, S. W.; Stuart, S. J.; Berne, B. J., Dynamical Fluctuating Charge Force-Fields - Application to Liquid Water. J. Chem. Phys. 1994, 101, 6141-6156. (38) Liang, T.; Shan, T. R.; Cheng, Y. T.; Devine, B. D.; Noordhoek, M.; Li, Y. Z.; Lu, Z. Z.; Phillpot, S. R.; Sinnott, S. B., Classical atomistic simulations of surfaces and heterogeneous interfaces with the charge-optimized many body (COMB) potentials. Mat. Sci. Eng. R 2013, 74, 255-279. (39) Martinez, J. A.; Yilmaz, D. E.; Liang, T.; Sinnott, S. B.; Phillpot, S. R., Fitting empirical potentials: Challenges and methodologies. Curr. Opin. Solid State Mater. Sci. 2013, 17, 263-270. (40) Liang, T.; Devine, B.; Phillpot, S. R.; Sinnott, S. B., Variable charge reactive potential for hydrocarbons to simulate organic-copper interactions. J. Phys. Chem. A 2012, 116, 7976-7991. (41) Liang, T.; Cheng, Y. T.; Nie, X. W.; Luo, W. J.; Asthagiri, A.; Janik, M. J.; Andrews, E.; Flake, J.; Sinnott, S. B., Molecular dynamics simulations of CO2 reduction on Cu(111) and Cu/ZnO(10 (1)over-bar 0) using charge optimized many body potentials. Catal. Commun. 2014, 52, 84-87. (42) Plimpton, S., Fast Parallel Algorithms for Short-range Molecular Dynamics. J. Comput. Phys. 1995, 117, 1-19. (43) NIST Computational Chemistry Comparison and Benchmark Database. 2013. (44) Vega, C.; McBride, C.; Sanz, E.; Abascal, J. L., Radial distribution functions and densities for the SPC/E, TIP4P and TIP5P models for liquid water and ices I h, I c, II, III, IV, V, VI, VII, VIII, IX, XI and XII. Phys. Chem. Chem. Phys. 2005, 7, 1450-1456. (45) Rick, S. W., Simulations of proton order and disorder in ice Ih. J. Chem. Phys. 2005, 122, 094504. (46) Röttger, K.; Endriss, A.; Ihringer, J.; Doyle, S.; Kuhs, W., Lattice constants and thermal expansion of H2O and D2O ice Ih between 10 and 265 K. Acta Crystallogr. Sect. B: Struct. Sci. 1994, 50, 644-648. (47) Dong, S.; Wang, Y.; Li, J., Potential lattice dynamical simulations of ice. Chem. Phys. 2001, 270, 309317. (48) Bergmann, U.; Di Cicco, A.; Wernet, P.; Principi, E.; Glatzel, P.; Nilsson, A., Nearest-neighbor oxygen distances in liquid water and ice observed by x-ray Raman based extended x-ray absorption fine structure. J. Chem. Phys. 2007, 127, 174504. (49) Vega, C.; Abascal, J.; Sanz, E.; MacDowell, L.; McBride, C., Can simple models describe the phase diagram of water? J. Phys.: Condens. Matter 2005, 17, S3283. (50) Rusnak, A. J.; Pinnick, E. R.; Calderon, C. E.; Wang, F., Static dielectric constants and molecular dipole distributions of liquid water and ice-Ih investigated by the PAW-PBE exchange-correlation functional. J. Chem. Phys. 2012, 137, 034510.


Page 30 of 33

Page 31 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

(51) Batista, E. R.; Xantheas, S. S.; Jónsson, H., Molecular multipole moments of water molecules in ice Ih. J. Chem. Phys. 1998, 109, 4546-4551. (52) Ding, P.; Li-Min, L.; Gareth, A. T.; Ben, S.; Angelos, M.; Enge, W., Surface energy and surface proton order of the ice Ih basal and prism surfaces. J. Phys.: Condens. Matter 2010, 22, 074209. (53) Wu, Y.; Tepper, H. L.; Voth, G. A., Flexible simple point-charge water model with improved liquidstate properties. J. Chem. Phys. 2006, 124, 024503. (54) Badyal, Y. S.; Saboungi, M. L.; Price, D. L.; Shastri, S. D.; Haeffner, D. R.; Soper, A. K., Electron distribution in water. J. Chem. Phys. 2000, 112, 9206-9208. (55) Price, W. S.; Ide, H.; Arata, Y., Self-diffusion of supercooled water to 238 K using PGSE NMR diffusion measurements. J. Phys. Chem. A 1999, 103, 448-450. (56) Ichikawa, K.; Kameda, Y.; Yamaguchi, T.; Wakita, H.; Misawa, M., Neutron-Diffraction Investigation of the Intramolecular Structure of a Water Molecule in the Liquid-phase at High-temperatures. Mol. Phys. 1991, 73, 79-86. (57) Michaelides, A.; Alavi, A.; King, D. A., Insight into H2O-ice adsorption and dissociation on metal surfaces from first-principles simulations. Phys. Rev. B 2004, 69, 113404. (58) Michaelides, A.; Ranea, V. A.; de Andres, P. L.; King, D. A., General model for water monomer adsorption on close-packed transition and noble metal surfaces. Phys. Rev. Lett. 2003, 90, 216102. (59) Tang, Q. L.; Chen, Z. X., Density functional slab model studies of water adsorption on flat and stepped Cu surfaces. Surf Sci. 2007, 601, 954-964. (60) Morgenstern, K.; Rieder, K.-H., Dissociation of water molecules with the scanning tunnelling microscope. Chem. Phys. Lett. 2002, 358, 250-256. (61) Morgenstern, K.; Rieder, K.-H., Formation of the cyclic ice hexamer via excitation of vibrational molecular modes by the scanning tunneling microscope. J. Chem. Phys. 2002, 116, 5746-5752. (62) Yeh, K. Y.; Janik, M. J.; Maranas, J. K., Molecular dynamics simulations of an electrified water/Pt(111) interface using point charge dissociative water. Electrochimica Acta 2013, 101, 308325. (63) Soon, A.; Todorova, M.; Delley, B.; Stampfl, C., Oxygen adsorption and stability of surface oxides on Cu(111): A first-principles investigation. Phys. Rev. B 2006, 73, 165424. (64) Niehus, H., Surface reconstruction of Cu (111) upon oxygen adsorption. Surf Sci. 1983, 130, 41-49. (65) Jeon, B.; Sankaranarayanan, S. K. R. S.; van Duin, A. C. T.; Ramanathan, S., Influence of surface orientation and defects on early-stage oxidation and ultrathin oxide growth on pure copper. Philos. Mag. 2011, 91, 4073-4088. (66) de Ruijter, M. J.; Blake, T. D.; De Coninck, J., Dynamic wetting studied by molecular modeling simulations of droplet spreading. Langmuir 1999, 15, 7836-7847.


Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For Table of Contents Only


Page 32 of 33

Page 33 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

320x227mm (150 x 150 DPI)

ACS Paragon Plus Environment

Effect of Surface Chemistry on Water Interaction with Cu(111

Recommend Documents