Molecular Dynamic Simulations of the Effect on the ... - ACS Publications

May 17, 2012 - overcoming the water management problems due to the water back-diffused .... identity of dissociated species on the surface is the subj...
2 downloads 0 Views 7MB Size
Article pubs.acs.org/JPCC

Molecular Dynamic Simulations of the Effect on the Hydration of Nafion in the Presence of a Platinum Nanoparticle Myvizhi Esai Selvan, Qianping He, Elisa M. Calvo-Muñoz, and David J. Keffer* Department of Chemical and Biomolecular Engineering, University of Tennessee, Knoxville, Tennessee 37996-2200, United States ABSTRACT: Platinum catalysts play a critical role in fuel cell technology. Current optimization efforts focus on reducing the amount of Pt in the system and optimizing the utilization of that which remains. The effect of the presence of Pt nanoparticles on the local structure and morphology of the polymer electrolyte membrane, water, and hydronium ions has been studied at molecular level in this work. Classical molecular dynamics simulation has been used to examine a system containing a 4 nm fcc cubic ({100} face) platinum nanoparticle at the center surrounded by Nafion polymer, water molecules, and hydronium ions at λ = 3, 6, 9, 15, and 22. The changes in density and orientation distribution of sulfonic acid groups in the side-chains, water, and hydronium as a function of distance from platinum surface are analyzed in this study. Sulfonic acid groups and hydronium ions show a very high increase in density near the platinum surface, and they approach the bulk value as they move away from the platinum surface. At lower water contents (λ = 3 and 6), water is strongly attracted to platinum surface, increasing the density near the platinum surface. However, at the highest humidity level studied, the density of water farthest from the platinum surface (>20 Å) is higher than the bulk value and the density of water nearest to the platinum surface. The above observed phenomenon is explained using the probability distribution of orientation computed at various distances from platinum surface. As the water content increases, sulfonic acid groups show a preferential orientation for the sidechains to align themselves horizontally on the platinum than vertically, thereby covering a larger area of platinum and pushing the water away from the platinum surface. This causes the water density to drop near the platinum surface. Water molecules and hydronium ions show preference to certain orientations near the platinum surface, but usually to none as they move away from the platinum surface.

I. INTRODUCTION Fuel cells are one of the most promising alternative power sources, with efficiencies of up to 60%, higher energy densities relative to batteries, and the ability to operate on clean fuels while producing no pollutants. Among fuel cells, polymer electrolyte membrane fuel cells (PEMFC) have been especially studied because they are cheaper to operate and give the highest efficiency as compared to other types of fuel cells.1−4 In these fuel cells, Pt acts as one of the most effective electrocatalysts for the oxygen reduction reaction and fuel (including hydrogen, methanol, ethanol, and formic acid) oxidation reaction.5 Pt also plays an important role in overcoming the water management problems due to the water back-diffused and reducing crossover of H2 and O2 through the membrane, caused by the finite solubilities of those gases in the membrane, which are two key aspects of the PEMFCs to enhance their performance and to suppress their degradation.6 Watanabe et al.7−9proposed the self-humidifying membranes with highly dispersed nanoparticles of Pt and/or other metal oxides (which today is known as polymer nanocomposite membranes).10 The Pt nanocrystals catalyze the recombination of the crossover H2 with O2 generating water that can directly humidify the PEM.6 Several types of PtPEMs have been proposed, and improvements in the cell performance have been reported.6,11,12 © 2012 American Chemical Society

However, in a world where optimization and proper and controlled use of natural resources have become not just a luxury but a priority, there are still many technical and marketrelated issues that must be overcome before fuel cells can be a commercially viable technology on a large scale. These challenges include decreasing the fuel cell cost, choosing the appropriate fuel source and infrastructure, and increasing its performance at higher temperatures of about 100 °C (PEMs require water to maintain proton conductivity, which also makes it impossible for this type of fuel cells to function at high temperatures).10 One of the main reasons of the high cost of PEMFCs is the utilization of precious metals as catalyst, especially platinum, as it was mentioned before. To decrease the fuel cell cost, as well as to avoid Pt supply limitations, ways to reduce the amount of Pt in the fuel cell, both by increasing its catalytic activity and by enhancing the fuel cell performance, must be found.5,13 Regarding the catalytic activity improvement of Pt, it has been shown that it is strongly dependent on the particle size, shape, and morphology because different crystal surfaces of platinum can have very different electronic structures and Received: March 1, 2012 Revised: May 17, 2012 Published: May 17, 2012 12890

dx.doi.org/10.1021/jp3020436 | J. Phys. Chem. C 2012, 116, 12890−12899

The Journal of Physical Chemistry C

Article

space group of Fm3m̅ , with four atoms per unit cell. This unit cell structure was replicated in the three dimensions until the desired size of the nanoparticle was achieved. The basic nanoparticle thus obtained is cubical in shape but is not symmetrical. Therefore, three cuts were made with planes parallel to the 100, 010, and 001 faces, which removed the first layer of atoms in each direction. Thus, a cubic particle enclosed by six 100 faces was built. A snapshot of the Pt nanoparticle used in the simulation is shown in Figure 1. It should be mentioned that in this work Pt nanoparticles remain isolated. The agglomeration of platinum nanoparticles has not been considered.

atomic arrangements, leading to drastically different reactivity toward the same reaction.14 The potential improvement of the performance of the fuel cell by using platinum nanoparticles with well-controlled shape and size has not been found, although these Pt nanoparticles inside the membrane as well as on the electrodes are expected to enhance the performance of the fuel cells. Among the information needed, key data to be found were the crystal structure, size, shape, favored faces, and surface chemistry of platinum nanoparticles as well as the role of platinum nanoparticles in the performance of the PEMFCs. Molecular dynamics simulation of the catalyst layer consisting of polymer system, carbon/graphite, and platinum has been performed earlier to understand the structural and dynamical changes in the system due to the presence of platinum nanoparticle. Cheng et al. studied the effect of platinum nanoparticle size on the microstructure of the catalyst layer and transport properties of water and ion.15 They examined systems that had platinum particles with truncated cub-octahedral geometry at the size of 1, 2, and 3 nm. The Watanabe and Uchida group studies if Pt nanoparticles inside the membrane focused on the performance advantages rather than the molecular-level relationship between Pt nanoparticles with different shapes and the general enhancement of the fuel cell.7−9 In this work, we have mainly concentrated on capturing the structural changes of the polymer electrolyte due to the presence of the nanoparticle. This Article is organized as follows: The details of the simulations are described in section II. Results and discussion are presented in section III, and the findings are summarized in section IV.

II. SIMULATION METHODOLOGY NVT simulations were performed with a single nanoparticle of platinum surrounded by Nafion at 300 K at five hydration levels (λ = 3, 6, 9, 15, 22). The platinum nanoparticle in the system was first built separately. Pt particle used in the simulation is a cubic nanoparticle bounded by low index facets {100} with a length of 3.92420 nm and contains 4631 atoms. The unit cell in the nanoparticle has a face-centered cubic structure, which is the structure dominantly retained in spherical, cubic, and cuboctahedral Pt nanoparticles.16 Platinum nanoparticles can be synthesized with various shapes (cubes, tetrahedrons, octahedrons, decahedrons, icosahedrons) bounded by a different number of facets and with different defects.5,17 The size and shape of the platinum nanoparticle determine the properties of the nanostructure. In this work, we have mainly concentrated on a cubic nanoparticle, which is one of the nanostructures whose shape and size can be controlled with high yield.18,19 It is widely known that, in general, the catalytic and electrocatalytic activity of metal particles depends on the ratio of surface area to volume, and, as a result, great effort has been made to increase the ratio of surface area to volume by reducing the size of the Pt nanoparticles.5 Up to now and for the different shapes studied, the investigations have focused on controlling their sizes in the range 2−10 nm.20−22 Currently, we have chosen approximately 4 nm as the standard length of our nanoparticle. Further, the platinum model has a clean surface without any defects. The nanoparticle was built by the repetition of the unit cell on the three dimensions of the space, x, y, and z. A unit cell was represented with a fcc structure, corresponding lattice parameters23 such as the length of the edges and angles between the edges (a = b = c = 3.92420 Å, α = β = γ = 90°) and

Figure 1. Snapshot of the face-centered cubic ({100} faces) platinum nanoparticle modeled for the simulation. The cube has 4631 atoms and a length of 3.92420 nm.

The Nafion polymer studied in this study consists of 15 monomers with an equivalent weight of 1144. The system is analyzed at five water contents, λ = 3, 6, 9, 15, and 22, each containing 40 Nafion ionomers, 600 H3O+ ions, and 3000, 4800, 8400, and 12 600 H2O molecules, respectively. The model of Nafion, water, and the hydronium ions used in this work is identical to that of our previous work.24 In Nafion, we used united atoms for CF 3, CF 2, and CF to reduce computational costs. The potential parameters of the Nafion model25−28 have been reported previously.24,29 The water is modeled using the TIP3P model30 with a flexible OH bond,31 and the hydronium ion is similar to that of Urata et al.32 In particular, the partial charges for the oxygen and hydrogen atoms are taken from Urata et al.32 The bond distance, bond angles, and force constants are the same as in the TIP3P model.30 This potential does not allow for structural diffusion of the proton. We included bond stretching, bending, torsion, and intramolecular and intermolecular nonbonded interactions via the Lennard-Jones (LJ) potential and Coulombic interactions. The Lorentz−Berthelot mixing rules are invoked for all interspecies interactions, to maintain uniformity in the interaction potential and to remain consistent with the Nafion potential, which used them in its parametrization.25−28We note that the Lorentz−Berthelot mixing rules are incorporated in most transferrable potentials (UFF, AMBER, MM3, TraPPE, and OPLS),33−36 but that structural and thermodynamic 12891

dx.doi.org/10.1021/jp3020436 | J. Phys. Chem. C 2012, 116, 12890−12899

The Journal of Physical Chemistry C

Article

Figure 2. Snapshots of Nafion/Pt system. (a)−(c) Cross-section showing Pt surrounded by water and polymer at λ = 3, 9, and 22. (d)−(f) Pt cube and water only with the Nafion and hydronium ions rendered invisible at λ = 3, 9, and 22. Pt are pink, CF2 and CF3 pseudoatoms are gray, H are white, S are orange, and O are red, except the O of H3O+, which are green.

properties can be affected by other choices of mixing rules.37−39 To validate our interaction potentials, including mixing rules, we have compared our simulation results with recent quantum and experimental studies of the orientation of water/hydronium ion species on the surface of Pt or other noble metals, and we are able to match our results with their findings quite well. Details are given in the Results and Discussion. The site−site reaction field40 is applied for the calculation of the electrostatic interactions. A constant temperature of 300 K is maintained by employing a Nosé−Hoover thermostat.41,42 The two time scale r-RESPA43 method is incorporated to integrate the equations of

motion with 2 fs for the large time step size and 0.2 fs for the intramolecular motions. The atoms in the platinum nanoparticle are held rigid, and the Pt interactions with other atoms in the system are represented by LJ 12-6 potentials (εPt/k = 2336.0 K, σPt = 2.41 Å).44 The choice of Pt potential is made to maintain continuity with previous work, in which we examined three-phase systems with Nafion/vapor/Pt and Nafion/vapor/ graphite,45 as well as Nafion/vapor/Pt/graphite systems.46 Lamas et al.47 have also used this potential for a similar system. Of course, Pt is a catalyst and water or hydronium ions at the Pt surface can react. In this work, we have not included the 12892

dx.doi.org/10.1021/jp3020436 | J. Phys. Chem. C 2012, 116, 12890−12899

The Journal of Physical Chemistry C

Article

Figure 3. Schematic defining the five regions in the system parallel to the platinum surface. Region 1 is closest to the Pt surface, and region 5 is farthest from Pt.

III. RESULTS AND DISCUSSION In this Article, we are mainly interested in the effect of the presence of Pt nanoparticle on the structural changes in the Nafion polymer, water molecule, and hydronium ions. A few snapshots of the systems studied are shown in Figure 2a−f. Figure 2a−c shows the cross-section of Pt surrounded by water and polymer at λ = 3, 9, and 22. The distribution of water molecules surrounding the platinum and in the PEM at the same water contents (λ = 3, 9, and 22) is shown in Figure 2d−f. The PEM atoms and hydronium ions are omitted to show the clustering of water molecules clearly. A. Density Distribution. The density distribution of sulfonic acid group, water, and hydronium is studied as a function of distance from the platinum surface. This analysis helps in understanding the effect of the presence platinum particle on the morphology of the membrane. A density distribution of the above molecules and ions is obtained by dividing the system into five regions as shown in Figure 3. The portions of the system parallel to each face of the cubic particle are divided into five regions, and the results are averaged over time and the six faces to get the distribution in each region. The portions of the system that may overlap being parallel to two faces are ignored, and the molecules present in those portions are not involved in the calculation of the density. Dividing the system into regions as suggested above enables us to capture any change in the density as we move toward the platinum surface. The first region is defined 5 Å from the surface of platinum, the regions bounded by 5−10, 10−15, 15−20 Å from the platinum surface are defined as the second, third, and fourth regions, respectively, and the remaining region is denoted as the fifth region. The density of a particular type of molecule in each region is calculated by dividing the number of molecules in that region by the volume of the corresponding region. The

chemical reactivity of the Pt surface. That reactivity and the identity of dissociated species on the surface is the subject of ab initio studies.48 These classical simulations are capable of describing how the nanophase segregation that occurs in “bulk” Nafion is perturbed by the presence of the solid nanoparticle. Many aspects of liquid ordering near the liquid/solid interface are captured by theory and simulation across various levels of description of the solid surface. For example, models of a simple fluid near an atomistic wall using either hard-sphere or Lennard-Jones potentials yield a similar layering structure.49 The initial configuration of the studied system was obtained by placing the cubic nanoparticle in the center of the system, and the polymers, water molecules, and hydronium ions were randomly placed around the system. Initially, all of the atoms except platinum were given zero size by making the corresponding LJ parameters zero. The atoms were then grown slowly over 10 000 steps with their LJ parameters gradually increased and electrostatic interactions switched off. Equilibration using this configuration was then performed with the appropriate LJ parameters and electrostatic interaction switched on. To achieve faster polymer relaxation, the systems were subjected to temperature annealing. The annealing procedure consisted of heating the system from 300 to 600 K over 0.1 ns with the temperature reassigned every 200 fs, running NVT simulation at 600 K for 1 ns, and cooling the system from 600 to 300 K over 0.5 ns with the temperature reassigned every 200 fs. Equilibration using these configurations was performed for another 0.6 ns before any production runs were began. The data production mode followed for an additional 2 ns. 12893

dx.doi.org/10.1021/jp3020436 | J. Phys. Chem. C 2012, 116, 12890−12899

The Journal of Physical Chemistry C

Article

Figure 4. Density of (a) SO−3 group, (b) H2O molecule, and (c) H3O+ ion as a function of distance from Pt surface at all water contents. Region 1 is nearest to the Pt surface, and region 5 is farthest from Pt.

is big enough for the components to reach the equilibrium density and that beyond 15 Å from the platinum surface the density distribution of the sulfonic acid groups is not affected by the platinum. The hydronium distribution presented in Figure 4c follows the same trend as the sulfonic acid groups and leads to conclusions similar to those discussed above. However, the water distribution represented in Figure 4b follows a very different trend. At lower humidity levels (λ = 3 and 6), water is attracted to the platinum surface, resulting in a high density in region 1, and slowly approaches the bulk value at region 5. At λ = 9, there is an increase in water density above the bulk value both at the nearest and at the farthest from the platinum surface with a low density value in region 3. At higher water contents (λ = 15 and 22), the density of water at the region 5 is much higher than the bulk value. Moreover, density at region 5 is higher than that at region 1. Molecular dynamic study of the PEM−catalyst−carbon interface containing Pt256 clusters approximately 1.6 nm supported over fixed two layered graphite at λ = 5, 9, 13, 24, and 64 indicates that the strong interactions of sulfonic sites with Pt increase the concentration of hydrophilic sites near the metallic particles, favoring the location of stable water clustering near catalytic active sites.47 However, a cubic Pt nanoparticle at the center of membrane in this study seems to cause water depletion near the Pt surface at high water contents. This behavior can be explained using the orientation distribution discussed below.

allocation of the molecule/ion to a region is based on the position of sulfur atom for the sulfonic acid group and oxygen atom for water molecule and hydronium ion. The volume of each region is the product, length of platinum cube × length of platinum cube × thickness of the region. The thickness of the regions 2, 3, 4, and 5 can be obtained directly from the definition. However, the volume of region 1 is not calculated using the thickness of region as 5 Å. A modified thickness is used, to accommodate the inaccessible region due to the platinum−membrane interaction. Approximately, the thickness of the inaccessible region is taken as the average of σPt and σS(SO−3 ) (2.98 Å) or σPt and σO(H2O) (2.78 Å) or σPt and σO(H3O+) (2.78 Å) depending on the density distribution required. Figure 4a−c represents the density distribution of S(SO−3 ), O(H2O), and O(H3O+), respectively, as a function of regions from the platinum surface for all of the water contents. Region 1 is closest to the platinum surface, and region 5 is the furthest from the platinum. The bulk density obtained from pure membrane simulation without any platinum nanoparticle is also included as a reference.24 The sulfonic acid group distribution (Figure 4a) shows that there is a very high increase in the density near platinum surface (region 1), which causes a decrease in density in regions 2 and 3 lower than the bulk value. Except for λ = 3, all other water contents seem to approach the bulk value as they move further away from the platinum surface (regions 4 and 5). This distribution also shows that system size 12894

dx.doi.org/10.1021/jp3020436 | J. Phys. Chem. C 2012, 116, 12890−12899

The Journal of Physical Chemistry C

Article

Figure 5. Probability distribution of the orientation of the SO−3 group with respect to the Pt nanoparticle (perpendicular to the surface) at λ = 3, 6, 9, 15, and 22. (a) Region 1, (b) region 2, and (c) region 3. Region 1 is closest to the Pt surface, and region 3 is farthest from Pt.

figure for a better understanding. In region 1, the highest probable orientation for side-chains is to be aligned perpendicular to the platinum surface in which all three oxygen atoms face the platinum surface. At cos θ = −0.25 and 0.25, the side-chains are aligned horizontally with respect to the platinum surface and have two oxygen atoms and a single oxygen atom closer to the platinum surface, respectively. The “standing” and “lying” orientations of the side-chains near the platinum surface have also been observed in the simulation by Cheng et al.15 The absence of significant peaks in regions 2 and 3 (Figure 5b and c) except at λ = 3 shows that there is no preferential orientation once the sulfonic acid group moves away from the platinum surface. At λ = 3, there might be strong morphological changes of the polymer even in region 2 due to very few water molecules in the system and an increase in density near the platinum surface. We believe that the orientation of the side-chain containing the sulfonic acid group in region 1 might play a significant role in the water density distribution near the platinum surface. Figure 5a indicates that the height of the peak at cos θ = −1 decreases with an increase in the humidity level, while the peak height at cos θ = −0.25 increases with an increase in humidity level. This might suggest that the side-chains prefer to align

B. Orientation Profile. The orientation of water molecules, hydronium ions, and sulfonic acid group with respect to the axis perpendicular surface of the platinum nanoparticle is studied in this work. The analysis is based on three regions rather than five regions as in the density calculation. The number of regions is decreased to obtain better statistics because the probability distribution of orientation in each region is examined. The three regions denote 0−5 Å from the surface of platinum, 5−15 Å, and 15 Å, the remaining length of the system. 1. Sulfonic Acid Group. The sulfonic acid group axis is defined to originate at the midpoint of the three oxygen atoms and terminate at sulfur atom. An angle of 180° corresponds to the sulfur atom facing the platinum surface with its three oxygen atoms in close contact with the platinum surface, while an angle of 0° corresponds to the sulfur atom close to platinum surface with its three oxygen atoms buried into the membrane. Figure 5a−c shows the probability distribution of the orientation of sulfonic acid group in the three regions at the hydration level of λ = 3, 6, 9, 15, and 22. We can see that there are three peaks in the probability distribution in region 1 at cos θ = −1, −0.25, and 0.25 from Figure 5a. A snapshot of the alignment of the acidic group with respect to platinum surface corresponding to each peak in region 1 is also provided in the 12895

dx.doi.org/10.1021/jp3020436 | J. Phys. Chem. C 2012, 116, 12890−12899

The Journal of Physical Chemistry C

Article

Figure 6. Probability distribution of the orientation of the water molecule with respect to the Pt nanoparticle (perpendicular to the surface) at λ = 3, 6, 9, 15, and 22. (a) Region 1, (b) region 2, and (c) region 3. Region 1 is nearest to the Pt surface, and region 3 is farthest from Pt.

with respect to the platinum surface in the three regions at the water content of λ = 3, 6, 9, 15, and 22 is plotted in Figure 6a− c. In region 1, all water contents show an increase in the probability distributions at angles of 90° and 0°. A peak at 90° implies water molecules favor orienting themselves to form a 2D hydrogen-bonding network parallel to the platinum surface. This kind of orientation has also been previously observed in the molecular dynamics simulation of the membrane/Pt catalyst interface.45 Tatarkhanov et al. have also confirmed the preferred orientation of water molecules on Pd and Ru surface using scanning tunneling microscopy (STM) as well as DFT total energy calculations, and they concluded that individual water on top of the noble metal surface bonding through the O atom like to remain in the molecular plane parallel to the metal surface.50 Glebov et al. also had the same observation when they were studying the adsorption behavior of water on Pt(111) surface using high-resolution He atom scattering at substrate temperatures between 20 and 140 K; they detected a loss feature at about 27 cm−1, which they assigned to an FT mode of the molecule parallel to the

themselves horizontally to the platinum surface rather than vertically at high water contents. The horizontal preferential orientation of the side-chains causes the platinum surface to be mostly covered by the side-chains, thereby causing a decrease in water concentration near the surface and pushing it farther away from the surface. The shape, concentration, and distribution of the Pt nanoparticle might also play an important role in the water distribution around the platinum surface. However, our initial study in understanding the effect of the presence of the Pt nanoparticle on the morphology of the membrane is limited to a single model system containing only one cubic nanoparticle. 2. Water. The water molecule axis is defined to originate at the midpoint of the two hydrogen atoms and terminate at the oxygen atom. An angle of 180° corresponds to the oxygen atom being buried into the membrane with its two hydrogen atoms in close contact with the platinum surface, while an angle of 0° corresponds to the oxygen atom close to the platinum surface with its two hydrogen atoms buried into the membrane. The probability distribution of the orientation of water molecule 12896

dx.doi.org/10.1021/jp3020436 | J. Phys. Chem. C 2012, 116, 12890−12899

The Journal of Physical Chemistry C

Article

Figure 7. Probability distribution of the orientation of the hydronium ion with respect to the Pt nanoparticle (perpendicular to the surface) at λ = 3, 6, 9, 15, and 22. (a) Region 1, (b) region 2, and (c) region 3. Region 1 is closest to the Pt surface, and region 3 is farthest from Pt.

surface.51 The second significant peak at 0° shows that the water molecules also prefer an orientation where the hydrogen atoms are embedded in the membrane. This is confirmed by research groups of Ogasawara and Norskov that another preferential orientation of the water molecules on the solid metal surface is with the oxygen-end down and one or both of the O−H bonds oriented away from the surface.52,53 In their studies, core level spectroscopies (XAS, XES, and XPS) with density functional theory (DFT) are combined to obtain element-specific information on molecular orientation. Henderson also reached the same conclusion in his review paper.54 Good agreement of our simulation results with available publications involving DFT calculations and atomic level experiments makes us feel assured of the choice of employing mixing rules to treat the intermolecular interactions. From Figure 6b, we can see that there are no preferential orientations in region 2 except at λ = 3. In region 3, where the influence of platinum is minimum, the orientation distribution is homogeneous.

3. Hydronium Ion. The hydronium ion axis is defined to originate at the midpoint of the three hydrogen atoms and terminate at the position of oxygen atom. An angle of 180° corresponds to the oxygen atom being buried into the membrane with its three hydrogen atoms in close contact with the platinum surface, while an angle of 0° corresponds to the oxygen atom protruding toward platinum surface with its three hydrogen atoms embedded in the membrane. The probability distribution of the orientation of hydronium ions in the three regions at the water content of λ = 3, 6, 9, 15, and 22 is represented in Figure 7a−c. Figure 7a depicts that the hydronium ion prefers orientations either with all three hydrogen atoms in close contact with the platinum or with two hydrogen atoms aligned parallel to platinum surface and the third hydrogen atom perpendicular. The flat-lying conformation of hydronium ion on the Pt surface has also been found in the work of Schiros et al.55 They used density functional theories to study the surface and hydrogen bonding of water and hydroxyl at metal surfaces. They concluded that the expected physisorption of H3O+ on metal surface is H12897

dx.doi.org/10.1021/jp3020436 | J. Phys. Chem. C 2012, 116, 12890−12899

The Journal of Physical Chemistry C

Article

(grant number DE-FG02-05ER15723). This work benefited from the resources of the Center for Computational Sciences at Oak Ridge National Laboratory, which is supported by the Office of Science of the DOE under contract DE-AC0500OR22725.

down and with a weak bond between Pt and O atoms. Their conclusion is confirmed by DFT calculations by Skulason et al., which give a DFT optimized hydronium conformation above the Pt surface with a H-down flat-lying structure, and a Pt−O bond distance of 4.5 + 0.2 Å.53 The highest peak at 0.65 suggests that the hydronium ions show a strong preference to contribute to the 2D hydrogen-bonding network parallel to the platinum surface. In regions 2 and 3, the orientation distribution is uniform with no preferential orientation. There are also previous experimental56 and simulation57 studies showing the preferential orientation of polar species at the solid/aqueous interface. Miranda et al.56 concluded that the interaction between liquid molecules and solid wall will affect the liquid surface structure, resulting in a more ordered aqueous/solid interface structure as compared to the bulk. They observed increased concentration of polar groups at the interface and orientations that maximized hydrogen bonding. A similar affinity for the interface and preferential orientation of the hydronium ion has also been observed at the interface between a liquid and hydrophobic media or vapor from both simulation58−60 and experiment.61,62 The orientations of water and hydronium ions shown in Figures 6 and 7 also serve to maintain as much as possible the hydrogen-bonding network with the surrounding fluid.



IV. CONCLUSIONS The purpose of this Article was to give a molecular-level insight into the structure of sulfonate ions, water molecules, and hydronium ions around the surface of platinum nanoparticles embedded in a PEM. NVT molecular dynamics simulations were performed on a system containing a 4 nm fcc cubic ({100} face) platinum nanoparticle surrounded by Nafion polymer, water molecules, and hydronium ions at λ = 3, 6, 9, 15, and 22 to analyze morphological features such as density and orientation distribution as a function of distance from the platinum surface. The density of sulfonic acid groups and hydronium ions increases near the platinum surface and reaches the bulk value as they move away from the platinum surface at all water contents. The water density increases near the Pt at low water contents. However, at higher water contents, the water density is lower near the Pt surface due to it being excluded by the polymer attached to the sulfonate ions, which are concentrated at the surface. Preferential orientation has also been observed in water, hydronium, and sulfonic acid groups near the Pt surface, which disappears as the distance from the Pt increases. The orientation distribution of the sulfonic acid groups near the Pt surface shows that at high water contents the sulfonic acid groups prefer to orient themselves in such a way that the side-chains are aligned horizontally to the Pt surface. This causes a decrease in the water density near the Pt surface because most of the area is covered by the polymer sidechains.



REFERENCES

(1) Cui, H. F.; Ye, J. S.; Zhang, W. D.; Wang, J.; Sheu, F. S. J. Electroanal. Chem. 2005, 577, 295. (2) Anantaraman, A. V.; Gardner, C. L. J. Electroanal. Chem. 1996, 414, 115. (3) Sugishima, N.; Hinatsu, J. T.; Foulkes, F. R. J. Electrochem. Soc. 1994, 141, 3325. (4) Plzak, V.; Rohland, B.; Wendt, H. Advanced Electrochemical Hydrogen Technologies: Water Electrolyzers and Fuel Cells; Plenum Press: New York, 1994. (5) Chen, J. Y.; Lim, B.; Lee, E. P.; Xia, Y. N. Nano Today 2009, 4, 81. (6) Watanabe, M.; Uchida, H.; Emori, M. J. Electrochem. Soc. 1998, 145, 1137. (7) Watanabe, M.; Uchida, H.; Seki, Y.; Emori, M.; Stonehart, P. J. Electrochem. Soc. 1996, 143, 3847. (8) Watanabe, M.; Uchida, H.; Emori, M. J. Phys. Chem. B 1998, 102, 3129. (9) Uchida, H.; Ueno, Y.; Hagihara, H.; Watanabe, M. J. Electrochem. Soc. 2003, 150, A57. (10) Cele, N.; Sinha Ray, S. Macromol. Mater. Eng. 2009, 294, 719. (11) Hagihara, H.; Uchida, H.; Watanabe, M. Electrochim. Acta 2006, 51, 3979. (12) Yang, T. Int. J. Hydrogen Energy 2008, 33, 2530. (13) Gasteiger, H. A.; Kocha, S. S.; Sompalli, B.; Wagner, F. T. Appl. Catal., B: Environ. 2005, 56, 9. (14) Stamenkovic, V. R.; Fowler, B.; Mun, B. S.; Wang, G. F.; Ross, P. N.; Lucas, C. A.; Markovic, N. M. Science 2007, 315, 493. (15) Cheng, C. H.; Malek, K.; Sui, P. C.; Djilali, N. Electrochim. Acta 2010, 55, 1588. (16) Qi, W. H.; Huang, B. Y.; Wang, M. P.; Yin, Z. M.; Li, J. J. Nanopart. Res. 2009, 11, 575. (17) Peng, Z. M.; Yang, H. Nano Today 2009, 4, 143. (18) Fu, X. Y.; Wang, Y. A.; Wu, N. Z.; Gui, L. L.; Tang, Y. Q. Langmuir 2002, 18, 4619. (19) Subhramannia, M.; Pillai, V. K. J. Mater. Chem. 2008, 18, 5858. (20) Wang, Q.; Ostafin, A. E. Encyclopedia of Nanoscience and Nanotechnology; American Scientific Publishers: Valencia, CA, 2004. (21) Sugimoto, T. Monodispersed Particles; Elsevier: Amsterdam, 2001. (22) Miyazaki, A.; Balint, L. Metal Nanoclusters and Materials Science: The Issue of Size Control; Elsevier: Amsterdam, 2008. (23) Calvo-Munoz, E.; Esai Selvan, M.; Xiong, R.; Ojha, M.; Keffer, D. J.; Nicoloson, D. M.; Egami, T. Phys. Rev. E 2010, 83, 011120. (24) Liu, J. W.; Suraweera, N.; Keffer, D. J.; Cui, S. T.; Paddison, S. J. J. Phys. Chem. C 2010, 114, 11279. (25) Li, H.-C.; McCabe, C.; Cui, S. T.; Cummings, P. T.; Cochran, H. D. Mol. Phys. 2003, 101, 2157. (26) Cui, S. T.; Siepmann, J. I.; Cochran, H. D.; Cummings, P. T. Fluid Phase Equilib. 1998, 146, 51. (27) Vishnyakov, A.; Neimark, A. V. J. Phys. Chem. B 2001, 105, 9586. (28) Vishnyakov, A.; Neimark, A. V. J. Phys. Chem. B 2001, 105, 7830. (29) Cui, S. T.; Liu, J. W.; Esai Selvan, M.; Keffer, D. J.; Edwards, B. J.; Steele, W. V. J. Phys. Chem. B 2007, 111, 2208. (30) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L. J. Chem. Phys. 1983, 79, 926. (31) Neria, E.; Fischer, S.; Karplus, M. J. Chem. Phys. 1996, 105, 1902. (32) Urata, S.; Irisawa, J.; Takada, A.; Shinoda, W.; Tsuzuki, S.; Mikami, M. J. Phys. Chem. B 2005, 109, 4269.

AUTHOR INFORMATION

Corresponding Author

*E-mail: dkeff[email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This research project is supported by the U.S. Department of Energy’s (DOE) Office of Basic Energy Sciences program 12898

dx.doi.org/10.1021/jp3020436 | J. Phys. Chem. C 2012, 116, 12890−12899

The Journal of Physical Chemistry C

Article

(33) Martin, M. G. Fluid Phase Equilib. 2006, 248, 50. (34) Basdevant, N.; Borgis, D.; Ha-Duong, T. J. Phys. Chem. B 2007, 111, 9390. (35) Xue, C. Y.; Zhong, C. L. Chin. J. Chem. 2009, 27, 472. (36) Njo, S. L.; Koegler, J. H.; vanKoningsveld, H.; vandeGraaf, B. Microporous Mater. 1997, 8, 223. (37) Banaszak, M.; Chiew, Y. C.; Radosz, M. Fluid Phase Equilib. 1995, 111, 161. (38) Reis, R. A.; Paredes, M. L. L.; Castier, M.; Tavares, F. W. Fluid Phase Equilib. 2007, 259, 123. (39) von Solms, N.; Koo, K. Y.; Chiew, Y. C. Fluid Phase Equilib. 2001, 180, 71. (40) Hummer, G.; Soumpasis, D. M.; Neumann, M. Mol. Phys. 1992, 77, 769. (41) Nosé, S. Mol. Phys. 1984, 52, 255. (42) Nosé, S. J. Chem. Phys. 1984, 81, 511. (43) Tuckerman, M.; Berne, B. J.; Martyna, G. J. J. Chem. Phys. 1992, 97, 1990. (44) Liem, S. Y.; Chan, K. Y. Surf. Sci. 1995, 328, 119. (45) Liu, J.; Esai Selvan, M.; Cui, S.; Edwards, B. J.; Keffer, D. J.; Steele, W. V. J. Phys. Chem. C 2008, 112, 1985. (46) Liu, J.; Cui, S.; Keffer, D. J. Fuel Cells 2008, 8, 422. (47) Lamas, E. J.; Balbuena, P. B. Electrochim. Acta 2006, 51, 5904. (48) Taylor, C. D.; Janik, M. J.; Neurock, M.; Kelly, R. G. Mol. Simul. 2007, 33, 429. (49) Statistical Mechanics of Phases, Interfaces, and Thin Films; Davis, H. T., Ed.; VCH Publishers, Inc.: Minneapolis, MN, 1995; p 420. (50) Tatarkhanov, M.; Ogletree, D. F.; Rose, F.; Mitsui, T.; Fomin, E.; Maier, S.; Rose, M.; Cerda, J. I.; Salmeron, M. J. Am. Chem. Soc. 2009, 131, 18425. (51) Glebov, A. L.; Graham, A. P.; Menzel, A. Surf. Sci. 1999, 427− 28, 22. (52) Schiros, T.; Andersson, K. J.; Pettersson, L. G. M.; Nilsson, A.; Ogasawara, H. J. Electron Spectrosc. Relat. Phenom. 2010, 177, 85. (53) Skulason, E.; Karlberg, G. S.; Rossmeisl, J.; Bligaard, T.; Greeley, J.; Jonsson, H.; Norskov, J. K. Phys. Chem. Chem. Phys. 2007, 9, 3241. (54) Henderson, M. A. Surf. Sci. Rep. 2002, 46, 5. (55) Schiros, T.; Ogasawara, H.; Naslund, L. A.; Andersson, K. J.; Ren, J.; Meng, S.; Karlberg, G. S.; Odelius, M.; Nilsson, A.; Pettersson, L. G. M. J. Phys. Chem. C 2010, 114, 10240. (56) Miranda, P. B.; Shen, Y. R. J. Phys. Chem. B 1999, 103, 3292. (57) Vacha, R.; Horinek, D.; Berkowitz, M. L.; Jungwirth, P. Phys. Chem. Chem. Phys. 2008, 10, 4975. (58) Petersen, M. K.; Iyengar, S. S.; Day, T. J. F.; Voth, G. A. J. Phys. Chem. B 2004, 108, 14804. (59) Esai Selvan, M.; Liu, J.; Keffer, D. J.; Cui, S.; Edwards, B. J.; Steele, W. V. J. Phys. Chem. C 2008, 112, 1975. (60) Lee, H. S.; Tuckerman, M. E. J. Phys. Chem. A 2009, 113, 2144. (61) Gopalakrishnan, S.; Liu, D. F.; Allen, H. C.; Kuo, M.; Shultz, M. J. Chem. Rev. 2006, 106, 1155. (62) Tian, C. S.; Ji, N.; Waychunas, G. A.; Shen, Y. R. J. Am. Chem. Soc. 2008, 130, 13033.

12899

dx.doi.org/10.1021/jp3020436 | J. Phys. Chem. C 2012, 116, 12890−12899