Visualizing Facet-Dependent Hydrogenation Dynamics in Individual

27 Aug 2018 - Surface faceting in nanoparticles can profoundly impact the rate and selectivity of chemical transformations. However, the precise role ...
0 downloads 0 Views 1MB Size
Subscriber access provided by University of Winnipeg Library

Communication

Visualizing facet-dependent hydrogenation dynamics in individual palladium nanoparticles Katherine Sytwu, Fariah Hayee, Tarun C. Narayan, Ai Leen Koh, Robert Sinclair, and Jennifer A. Dionne Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.8b00736 • Publication Date (Web): 27 Aug 2018 Downloaded from http://pubs.acs.org on August 28, 2018

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 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 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.

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 21 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

Nano Letters

Visualizing facet-dependent hydrogenation dynamics in individual palladium nanoparticles Katherine Sytwu,∗,†,k Fariah Hayee,‡,k Tarun C. Narayan,¶ Ai Leen Koh,§ Robert Sinclair,¶ and Jennifer A. Dionne∗,¶ †Department of Applied Physics, Stanford University, 348 Via Pueblo, Stanford, CA 94305 ‡Department of Electrical Engineering, Stanford University, 350 Serra Mall, Stanford, CA 94305 ¶Department of Materials Science and Engineering, Stanford University, 496 Lomita Mall, Stanford, CA 94305 §Stanford Nano Shared Facilities, Stanford University, 476 Lomita Mall, Stanford, CA 94305 kThese authors contributed equally E-mail: [email protected]; [email protected] Abstract Surface faceting in nanoparticles can profoundly impact the rate and selectivity of chemical transformations. However, the precise role of surface termination can be challenging to elucidate since many measurements are performed on ensembles of particles and do not have sufficient spatial resolution to observe reactions at the single and sub-particle level. Here, we investigate solute intercalation in individual palladium hydride nanoparticles with distinct surface terminations. Using a combination of diffraction, electron energy loss spectroscopy, and dark field contrast in an environmental transmission electron microscope (TEM), we compare the thermodynamics and directly visualize the kinetics of 40-70nm {100}-terminated cubes and {111}-terminated

1

ACS Paragon Plus Environment

Nano Letters 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

octahedra with approximately 2nm spatial resolution. Despite their distinct surface terminations, both particle morphologies nucleate the new phase at the tips of the particle. However, while the hydrogenated phase front must rotate from [111] to [100] to propagate in cubes, the phase-front can propagate along the [100], [1¯10], and [111] directions in octahedra. Once the phase front is established, the interface propagates linearly with time and is rate-limited by surface-to-subsurface diffusion and/or the atomic rearrangements needed to accommodate lattice strain. Following nucleation, both particle morphologies take approximately the same time to reach equilibrium, hydrogenating at similar pressures and without equilibrium phase coexistence. Our results highlight the importance of low-coordination number sites and strain, more so than surface faceting, in governing solute-driven reactions.

Keywords in-situ transmission electron microscopy, surface faceting, single particle, kinetics, palladium hydride, phase transition Nanoparticle surface faceting has profound implications for chemical transformations, particularly in catalysis 1,2 and intercalation-based phase transitions. 3–6 For example, in formic acid oxidation, a potential reaction for fuel cells, palladium {100}-cubes have shown higher catalytic activity and stability over both {111}-octahedral and {110}-dodecahedral shapes. 7,8 On the other hand, in alkyne hydrogenation, a selective reaction used in the food industry, palladium {111}-octahedra have higher catalytic activity than {100}-cubes. 9 Similarly, surface-termination has been reported to control Li-ion intercalation rates and hence electrochemical performance in battery electrodes. 3,5,6 For example, in TiO2 nanocrystals, Li+ insertion/extraction is more efficient in {001}-dominated particles than {101} ones, leading to higher charging/discharging rates. 3 In Co3 O4 nanoparticles, however, better charging/discharging rates and cycling performance have been observed for {111}-octahedral particles over {100}-cubes. 5,6

2

ACS Paragon Plus Environment

Page 2 of 21

Page 3 of 21 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

Nano Letters

Most studies comparing particle morphologies are performed over an ensemble of nanoparticles with varied size and shape. While advances in nanoparticle synthesis have enabled narrower size and shape distributions, studies at the ensemble level are still susceptible to heterogeneity. More importantly, ensemble studies lack the resolution to directly visualize reactions as they occur within a nanoparticle, making it difficult to correlate reaction dynamics with local structural features. Kinetic studies are particularly susceptible, as ensemble studies are insensitive to differences in the start times of a phase transition for various particles. 10 Single particle studies, on the other hand, can identify particle-specific activity and give better insight in the role of structure on chemical reactions. 11 Here, we use the hydrogenation of single-crystalline palladium nanoparticles as a model system to study the effects of surface faceting on the phase transition dynamics at the single and sub-particle level. Palladium nanocrystals have been extensively studied due to their facet-dependent catalytic behavior 9,12–14 as well as their distinct hydrogen-storage thermodynamics compared to bulk. 15–20 The palladium hydrogenation reaction is characterized by three steps: 21,22 first, hydrogen molecules catalytically split into two hydrogen atoms which are chemisorbed to the surface of the palladium nanoparticle; next, the hydrogen atoms diffuse into the hydrogen saturated subsurface (i.e. the outermost 1nm of the particle 19 ); finally, the hydrogen atoms diffuse into the bulk palladium matrix, inducing a phase change at sufficiently high H2 pressures. The palladium-hydrogen system (PdHx ) can be described by its two phases which have distinct properties: the hydrogen-poor, solid-state solution α-phase and the hydrogen-rich β-phase, which at room temperature have solubility limits of x = 0.03 and 0.6, 23 respectively. The mechanism and rate-limiting step of palladium hydrogenation has been shown to vary with size and possibly surface termination. Kinetic studies at the ensemble level have suggested diffusion into the bulk of the nanoparticle (the third step) to be the rate limiting step for phase transitions in particles less than 10nm in diameter. 24 On the other hand, ensemble experiments on palladium particles of ∼10nm have suggested the rate limiting step to

3

ACS Paragon Plus Environment

Nano Letters 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

be the facet-dependent diffusion of hydrogen atoms to the subsurface layer (the second step) since their octahedral particles hydrogenate faster than cubes. 25 Both of these rate-limiting steps support the “shrinking core” hydrogenation mechanism hypothesized and reported in ensembles of small nanoparticles. 26,27 Interestingly, recent single particle studies of 15-80nm palladium cubes and 160-340nm palladium nanoparticles have reported a different loading mechanism; in these larger nanoparticles, the β-phase nucleates at a corner before establishing a phase-front boundary. 28,29 This new understanding begs the question of how particle shape and surface termination affects the phase-transition dynamics when interrogated at the single particle level. To understand the effect of surface faceting on phase transition dynamics, we directly visualize the reaction in-situ using an environmental transmission electron microscope (TEM) which flows hydrogen gas over the sample. In recent years, in-situ and real time environmental TEM techniques have revealed the dynamics behind lithiation of spinel metal oxides 30,31 and silicon nanorods, 32 passivation layer growth on lithium, 33 as well as hydrogenation of palladium nanocubes. 28 Using a suite of techniques, we can identify phases at the single and sub-particle level. To monitor the hydrogenation state of a single particle, we use selected area electron diffraction (SAED) to measure the lattice parameter. Upon hydrogenation from the α- to the β-phase, the bulk palladium lattice isotropically expands by ∼ 3.5%, 34 which corresponds to a ∼ 3.5% contraction in its diffraction pattern. To track and identify phases at the subparticle level with 2nm spatial resolution, we use diffraction contrast in scanning TEM (STEM) and electron energy loss spectroscopy (EELS). Upon hydrogenation, palladium’s electronic properties change, with a bulk plasmon energy of 8eV in the α-phase and 6eV in the β-phase, a 2eV shift that is readily detectable using monochromated-EELS. 16,35 We focus on palladium cubes and octahedra with edge lengths of 40 − 70nm, a size range more relevant for storage materials, 36 which we colloidally synthesize by modifying a previously reported synthesis 37 (see Supplementary Information). Figure 1 shows scanning electron micrographs (SEMs) of the synthesized palladium cubes (Figure 1a) and octahedra

4

ACS Paragon Plus Environment

Page 4 of 21

Page 5 of 21 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

Nano Letters

(Figure 1d), with schematic insets detailing their morphologies. SAED patterns verify the single-crystalline nature and crystallography; cubes have a fcc [100] zone axis diffraction pattern (Figure 1c), while octahedra have a fcc [111] zone axis pattern (Figure 1f). Before we study the dynamics, we first need to establish the near-equilibrium behavior of cubes and octahedra at our given temperature and pressure range. Under constant and minimal electron exposures (see Supplementary Information for Experimental Procedures), we collect SAED patterns of single Pd cubes and octahedra at 238K at different hydrogen pressures, waiting 30 minutes after pressure stabilization at each point. The phase of each particle can be determined by the change in its diffraction pattern relative to its diffraction pattern taken near 0Pa (see Supplementary Information). We then plot out the percentage change in lattice parameter at each pressure step, constructing pressure-composition isotherms for individual particles. In Figure 2, we show six representative isotherms of 3 individual cubes and 3 individual octahedra. As evidenced by the sharp transition between the α- and β-phase, neither cubes nor octahedra support phase coexistence in thermodynamic equilibrium, consistent with previous measurements of cubes, octahedra, and other single crystalline nanoparticles. 17,18 We also observe a ∼ 3% change in lattice parameter in both cubes and octahedra, suggesting that hydrogen storage capacity is not shape dependent within this size range (Figure S2). However, this change is slightly less than the reported ∼ 3.5% reported in bulk Pd at room temperature. 38 Both geometries show variation in the loading pressure, but no discernible trend over the size range studied (see Figures S3 and S4a). The lack of difference between the loading pressure values of cubes and octahedra indicate that surface termination does not affect the equilibrium hydrogenation pressure, also consistent with previous findings. 17 To visualize the reaction in real time, we choose our hydrogen pressure to be near the observed loading pressure and use scanning dark field imaging and EELS to record the phase progression with around 2 second resolution (see Supplementary Information for STEM experimental details). Since the cubes and octahedra lie close to their zone axis, any lattice

5

ACS Paragon Plus Environment

Nano Letters 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

defect or strain (like at the phase interface) can affect image contrast (see Supplementary Information Sec. 3). We verify the phase of each region by pausing the scanning beam, and placing it over specific points to collect the EEL spectra. In both particle morphologies, the β-phase nucleates at one of the particle corners before establishing a phase boundary, as seen in Figure 3a and b. We observed this corner nucleation behavior in 21 octahedra and 24 cubes. Concurrent EELS measurements confirm that the observed contrast is due to phase nucleation, rather than solely lattice strain. To verify the α and β regions, we then freeze the reaction intermediates to 100K during hydrogenation, and use displaced aperture dark field imaging to selectively image the α- and β-phase distribution. 28 As shown in Figure 3c, the β-phase is only found at the corners of the octahedra, confirming that the new phase nucleates at the corners. In contrast to real time measurements, these frozen intermediate states show multiple corner nucleation sites, including nucleation at opposite corners. This is likely due to the slow cool-down (5-10 minutes) while freezing the system, and nucleation being more favorable than phase propagation and growth. While it is possible to have multiple nucleation sites, nucleation at a second corner does not seem to be independent. In our real-time measurements of 21 octahedra, we have not observed two nucleation sites at opposite corners (i.e. the only two corners that are not adjacent). These findings demonstrate that octahedra, similar to cubes, start their phase transition from their corners as opposed to their faces. While cubes load from their [111] directional corners, octahedra load from their [100] directional corners. The existence of corner loading irrespective of surface faceting suggests that initial nucleation is primarily due to a structural or geometric factor, namely coordination number and strain. Indeed Ludwig et al. have shown that hydrogen is more likely to diffuse to the subsurface at sites like corners and edges, which have lower coordination numbers. 39 Our results show that this preferential diffusion at corners also holds true for creating a stable nucleation site in the bulk at a pressure regime when the subsurface is likely already saturated. 40,41 This result also supports similar tip nucleation behavior seen in nanorods. 42 Nucleation can also be due to tensile strain at the

6

ACS Paragon Plus Environment

Page 6 of 21

Page 7 of 21 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

Nano Letters

corners. Experimental studies have revealed that a hydrogen-rich surface layer can result in tensile strain at the corners of α-phase PdHx nanocubes despite compressive strain in pure palladium nanocubes. 41 While molecular dynamic studies have reported compressive strain at the corners of platinum octahedra, 43 tensile strain can also be present at the corners due to a hydrogen saturated surface. After the β-phase nucleates at the corner, it creates an α-β phase interface and propagates through the particle. In Figures 4a-c, we show time-series snapshots of the phase transition in a cube (a) and two octahedra (b,c), with accompanying schematics of our interpretation for octahedra. In cubes, the interface is initially in the < 111 > direction and always rotates to the < 100 > direction before propagating through the rest of the particle. 28 On the other hand, octahedra do not always show the same interface orientation. The direction of propagation and shape suggest a (100) interface in Figure 4b and an initial (111) interface in Figure 4c. In Figure 4b, the (100) interface is maintained throughout, with the β-phase entering like a pyramid with a growing base. At 135s, we can clearly see the projection of the square interface between the two phases; the subsequent progression then pushes the α-phase out through the opposite corner. Notably, (100) is the lowest elastic energy interface in bulk-Pd for a coherent boundary. 28 Since the interface orientation does not align with the zone-axis of octahedra, it is difficult to image the phase boundary; we were able to uniquely identify the phase boundaries for 6 octahedra, but only verify phase propagation for the rest of the 15 octahedra. Though we see (100) interfaces in 3 out of the 6 studied octahedra (Figure S16), we occasionally also observe other interfaces, like (111) and (1¯10). In Figure 4c, the initial stage does not correspond to any of the lowest order interfaces, and is likely a corner nucleation without a defined interface. The β-phase then establishes a (111) interface by 50 seconds, before being joined by what appears to be a (1¯10) interface at 64 seconds. This second phase front would have come from nucleation at an adjacent corner. We can verify this nucleation and growth behavior by again freezing the system during hydrogenation and directly interrogating the phase-distributions with 2nm resolution with

7

ACS Paragon Plus Environment

Nano Letters 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

EELS mapping. We decompose the resulting EELS map using multivariate curve resolution 44 to obtain the relative strengths of the α and β spectral components and plot these weights across the entire particle. In Figure 4d, we show frozen EELS maps of 4 different octahedral particles at various stages of hydrogenation, with superimposed dark-field images to delineate the particle boundaries. As seen in the STEM data, the hydrogenated β-phase nucleates at the corner (Figure 4d(i)), establishes an interface (Figures 4d(ii) and 4d(iii)), and then pushes out the initial α-phase at another corner (Figure 4d(iv)). While the (1¯10) phase-front seen in Figure 4d(ii) is not often seen in our real-time STEM measurements, we again attribute the different behavior to the slow cool-down which allows the hydrogen to redistribute. Note that the β-phase regions at the particle boundaries are an artifact from the spectral overlap of the α-phase surface plasmon with the β-phase bulk plasmon (Figure S5), and dark field images verify that the β-phase is not present at the edges (Figure S6,S7). Using the sequential STEM images, we then quantify and compare the kinetics of the phase transition in individual cubes and octahedra. We calculate the phase transformation time, defined as the time between nucleation and full particle loading, for various cubes and octahedra of similar volumes, taken at 246K under a hydrogen pressure near their loading pressure (471Pa). This reported time is different from ensemble measurements 25 which measure the total phase transition time starting from when pressure is increased; our measurements, on the other hand, pinpoint the exact time the phase transformation starts. In Figure 5a, we see that there is some inherent scatter for both types of particles, but no distinguishable difference between the total phase transformation time in cubes versus octahedra. To understand why octahedra and cubes take the same amount of time to load hydrogen, we next quantify how the phase front propagates within the particle in a separate set of experiments. In Figure 5b and c, we show the position of a single β-phase-front as a function of time for three cubes and octahedra, taken at 380Pa and 423Pa, respectively. We start measurement at the establishment of a phase-front and end when either the phase-front disappears or is joined by another phase-front. For cubes, this position vector is parallel to

8

ACS Paragon Plus Environment

Page 8 of 21

Page 9 of 21 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

Nano Letters

the direction of propagation, while for octahedra, the position is defined as the distance from the interface to its nucleating corner, illustrated by the STEM snapshot diagrams on the right. For both particle shapes, the phase-front position progresses linearly with time, with interface speeds ranging from 0.11 to 0.27nm/s in cubes and from 0.15 to 0.31nm/s in octahedra. We observed this linear trend in three octahedra (limited by the number of particles with discernible contrast throughout the entire phase transition), and in twelve cubes. This linear trend is also observable over multiple cycles of hydrogenation (Figure S11). Even at slightly different pressures, the resultant rates between cubes and octahedra are comparable, which support the comparable phase transformation times. From these single and sub-particle measurements, we can conclude that the phase transformation kinetics are facet independent and limited by a linear process. Therefore, the reaction cannot be limited by the diffusion of hydrogen atoms into the bulk of the parti√ cle, which would follow a t curve. Our phase transformation time is also too long for a diffusion-limited process; at 246K, the diffusion constant of H in β-phase bulk PdHx is 1.9 × 10−8 cm2 /s, 45 which would correspond to the phase front traveling 1.4μm in 1 second, about four orders of magnitude greater than the timescale we observe. The linear dependence could therefore originate from two possible rate-limiting steps. One possibility is that the reaction is limited by the splitting and adsorption of hydrogen at the surface. This result, however, is in contrast to bulk palladium hydride studies which suggest that splitting and adsorption is not the rate-limiting step above 200K on Pd (111)- and (100)-terminated films. 46,47 Another possibility is that the reaction is limited by the diffusion of hydrogen atoms from the surface to the subsurface at the lower coordinated corners and edges, which have higher subsurface diffusion rates. 39 This process would still be linear with time due to the small length scales involved. The ratio of edge and corner atoms to surface atoms for our two particle morphologies is comparable in the 40 − 70nm regimes and slightly higher in smaller particles (see Figure S13), 48 which can explain the lack of difference in transformation time in our particles despite an observed and calculated difference in ensemble

9

ACS Paragon Plus Environment

Nano Letters 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

measurements of ∼ 10nm particles. 25,49 Given that our reaction is much slower than estimated by diffusion, another possibility is that strain or crystallite formation in the nanoparticles slows down the reaction. In our nanocubes, we note a sharp angular splitting of the α and β diffraction points, indicating that there is increased strain or crystallite formation within the α and β regions (see Figure S14). Such results are consistent with previous studies that report degradation of crystal quality and slight reorientation during the hydrogenation process, though these crystallographic imperfections disappear upon full hydrogenation. 28,29 We hypothesize that this strain formation and subsequent disappearance within the α and β regions slows down the propagation of the phase-front since the lattice has to rearrange itself. Accordingly, within these nanoparticles, the effective diffusion rate would be equivalent to this linear phase-front migration rate. In summary, we have investigated the thermodynamics and kinetics of hydrogenation in differently faceted, 40 − 70nm palladium nanoparticles using in-situ TEM. Our approach enables direct visualization of the kinetics and additional characterization at the sub-particle level. In equilibrium, palladium cubes and octahedra show similar loading pressures and storage capacity due to their comparable sizes and single crystalline nature. From our non-equilibrium measurements, we find that irrespective of surface faceting, the β-phase nucleates at the corners of the particles, which we postulate is due to the higher rates at low-coordination number sites and/or tensile strain. The α-β interface propagates linearly with time, indicating that the reaction is not diffusion limited within this temperature and pressure regime, and we hypothesize that the reaction could be limited by surface activity at low coordination number sites or crystallite rotation. On par with this rate-limiting step, cubes and octahedra show comparable phase transformation times. Our results point towards coordination number, more so than surface faceting, as the dominant factor governing phase transformation mechanisms in particles larger than 40nm. Accordingly, to maximize the speeds of reactions, efforts should focus on particles with many low coordination number

10

ACS Paragon Plus Environment

Page 10 of 21

Page 11 of 21 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

Nano Letters

sites as well as geometries and sizes that can minimize crystallite formation.

Figure 1: SEM and TEM images of Pd cubes and octahedra. SEM images of Pd cubes (a) and octahedra (d), with insets describing their morphology and surface faceting. Scale bar is 200nm. (b,c) TEM image of a representative Pd cube and its electron diffraction pattern. (e,f) TEM image of a representative Pd octahedra and its electron diffraction pattern. Since the cubes and octahedra usually rest on one of their faces, the subsequent 2D TEM projection is that of a square (cube) or a hexagon (octahedron). Scale bar is 50nm for the TEM images.

11

ACS Paragon Plus Environment

Nano Letters 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

Figure 2: Near-equilibrium behavior of cubes and octahedra. Pressure-composition isotherms of hydrogen loading measured from selected area diffraction patterns overlaid on TEM images for three cubes (green, left) and three octahedra (purple, right). The ∼ 0% change in lattice parameter corresponds to α-phase while the ∼ 3% change corresponds to β-phase. See Figure S2 for more statistics on final hydrogen storage capacity.

12

ACS Paragon Plus Environment

Page 12 of 21

Page 13 of 21 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

Nano Letters

Figure 3: Corner nucleation in cubes and octahedra. (a,b) Still frames from live STEM recordings of the phase transition in (a) two cubes and (b) two octahedra. α and β markers are placed where we have confirmed the phase using EELS. (c) Overlaid displaced aperture dark field images of two octahedra that show corner loading. See Figure S15 for more examples.

13

ACS Paragon Plus Environment

Nano Letters 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

Figure 4: The intermediate stages of the α-β phase transition in Pd cubes and octahedra. (a-c) Still frames from STEM recordings of the phase transition in a cube (a) and two octahedra (b,c) with accompanying schematics of the reaction in octahedra displaying our interpretation below. In the STEM images, the α and β labels are placed where we have confirmed the phase using EELS. In the schematics, the blue represents the α-phase, the red represents the β-phase, and the purple represents the interface. Solid lines are an octahedron orientation, and dotted lines outline a possible interface orientation. (d) EELS maps of the frozen, intermediate states of four octahedra at different stages of the phase transformation, arranged by increasing β amount. All scale bars are 50nm.

14

ACS Paragon Plus Environment

Page 14 of 21

Page 15 of 21 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

Nano Letters

Figure 5: Single particle kinetics in cubes and octahedra. (a) Comparison of the total phase transformation time in cubes and octahedra for particles of similar volume, taken at 471Pa. (b,c) Position of the interface in three separate cubes (b) with effective edge lengths of 55nm, 61nm, and 58nm (light, medium, and dark green), taken at 380Pa; and three octahedra (c) with effective edge lengths of 60nm, 63nm, and 50nm (light, medium, and dark purple), taken at 423Pa. Time is the number of seconds from the establishment of a defined interface, and the lines are weighted linear fits to the data. On the right of each graph are representative STEM snapshots of each of the cubes and octahedra, with arrows displaying the distance measured. All scale bars are 50nm.

Acknowledgement The authors would like to acknowledge Michal Vadai and Yiyang Li for helpful discussions; Andrew Riscoe, Andrea Baldi, Herman Schreuders, and Bernard Dam for helping with ensemble measurements; and all members of the Dionne group for helpful scientific feedback and support. TEM imaging and spectroscopy were performed at the Stanford Nano Shared Facilities (SNSF), supported by the National Science Foundation (NSF) under award ECCS-1542152. Support from a PECASE Award administered by the Air Force Office of Scientific Research (FA9550-15-1-0006), a National Science Foundation CAREER Award (DMR-1151231), and a Camille and Henry Dreyfus grant are gratefully acknowledged. K.S. was supported by the Gabilan Stanford Graduate Fellowship and the National Science Foundation Graduate Research Fellowship (DGE-1656518). T.C.N was supported by an award from the Department of Energy (DOE) Office of Science Graduate Fellowship

15

ACS Paragon Plus Environment

Nano Letters 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

Program administered by the Oak Ridge Institute for Science and Education for the DOE. ORISE is managed by Oak Ridge Associated Universities (ORAU) under DOE contract number DE-AC05-06OR23100. All opinions expressed in this paper are the authors and do not necessarily reflect the policies and views of NSF, DOE, ORAU, or ORISE.

Author Contributions K.S, F.H, and J.A.D. designed the experiments. K.S. and F.H. performed the experiments and A.L.K. and R.S. helped with the TEM setup and analysis. K.S. and F.H. analyzed the data, and K.S. wrote the initial draft of the manuscript. J.A.D. supervised the project. All authors discussed the results and contributed to the final manuscript preparation.

Supporting Information Available Supporting information includes procedures for nanoparticle synthesis, sample preparation, TEM and STEM imaging conditions, analysis on hydrogen storage capacity, renormalized data in terms of surface area, more details and simulations of EELS data, frozen displaced aperture dark field images, additional STEM snapshots, TEM images, and analysis on STEM contrast, diffraction point widths, and strain/crystallite formation in a nanocube. This material is available free of charge via the Internet at http://pubs.acs.org/.

Competing Interests The authors declare no competing financial interests.

References (1) Zhou, K.; Li, Y. Angewandte Chemie International Edition 2012, 51, 602–613.

16

ACS Paragon Plus Environment

Page 16 of 21

Page 17 of 21 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

Nano Letters

(2) Jin, M.; Liu, H.; Zhang, H.; Xie, Z.; Liu, J.; Xia, Y. Nano Research 2011, 4, 83–91. (3) Sun, C. H.; Yang, X. H.; Chen, J. S.; Li, Z.; Lou, X. W.; Li, C.; Smith, S. C.; Lu, G. Q. M.; Yang, H. G. Chemical Communications 2010, 46, 6129–6131. (4) Kuppan, S.; Xu, Y.; Liu, Y.; Chen, G. Nature Communications 2017, 8, 14309. (5) Xiao, X.; Liu, X.; Zhao, H.; Chen, D.; Liu, F.; Xiang, J.; Hu, Z.; Li, Y. Advanced Materials 2012, 24, 5762–5766. (6) Gao, R.; Zhu, J.; Xiao, X.; Hu, Z.; Liu, J.; Liu, X. The Journal of Physical Chemistry C 2015, 119, 4516–4523. (7) Zhang, X.; Yin, H.; Wang, J.; Chang, L.; Gao, Y.; Liu, W.; Tang, Z. Nanoscale 2013, 5, 8392–8397. (8) Zhang, J.; Feng, C.; Deng, Y.; Liu, L.; Wu, Y.; Shen, B.; Zhong, C.; Hu, W. Chemistry of Materials 2014, 26, 1213–1218. (9) Laskar, M.; Skrabalak, S. E. ACS Catalysis 2014, 4, 1120–1128. (10) Smith, J. G.; Zhang, X.; Jain, P. K. Journal of Materials Chemistry A 2017, 5, 11940– 11948. (11) Sambur, J. B.; Chen, P. Annual Review of Physical Chemistry 2014, 65, 395–422. (12) Klinkova, A.; Cherepanov, P. V.; Ryabinkin, I. G.; Ho, M.; Ashokkumar, M.; Izmaylov, A. F.; Andreeva, D. V.; Kumacheva, E. Small 2016, 12, 2450–2458. (13) Chen, T.; Chen, S.; Song, P.; Zhang, Y.; Su, H.; Xu, W.; Zeng, J. ACS Catalysis 2017, 7, 2967–2972. (14) Xu, Z.-N.; Sun, J.; Lin, C.-S.; Jiang, X.-M.; Chen, Q.-S.; Peng, S.-Y.; Wang, M.-S.; Guo, G.-C. ACS Catalysis 2012, 3, 118–122.

17

ACS Paragon Plus Environment

Nano Letters 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

(15) Bardhan, R.; Hedges, L. O.; Pint, C. L.; Javey, A.; Whitelam, S.; Urban, J. J. Nature Materials 2013, 12, 905–912. (16) Baldi, A.; Narayan, T. C.; Koh, A. L.; Dionne, J. A. Nature Materials 2014, 13, 1143–1148. (17) Syrenova, S.; Wadell, C.; Nugroho, F. A.; Gschneidtner, T. A.; Fernandez, Y. A. D.; ´ Nalin, G.; Switlik, D.; Westerlund, F.; Antosiewicz, T. J.; Zhdanov, V. P. et al. Nature Materials 2015, 14, 1236–1244. (18) Narayan, T. C.; Baldi, A.; Koh, A. L.; Sinclair, R.; Dionne, J. A. Nature Materials 2016, 15, 768–774. (19) Griessen, R.; Strohfeldt, N.; Giessen, H. Nature Materials 2016, 15, 311. (20) Zalineeva, A.; Baranton, S.; Coutanceau, C.; Jerkiewicz, G. Science Advances 2017, 3, e1600542. (21) Jerkiewicz, G.; Borodzinski, J.; Chrzanowskia, W.; Conway, B. Journal of The Electrochemical Society 1995, 142, 3755–3763. (22) Jerkiewicz, G. Progress in Surface Science 1998, 57, 137–186. (23) Switendick, A. Hydrogen in Metals I ; Springer, 1978; pp 101–129. (24) Langhammer, C.; Zhdanov, V. P.; Zori´c, I.; Kasemo, B. Physical Review Letters 2010, 104, 135502. (25) Li, G.; Kobayashi, H.; Dekura, S.; Ikeda, R.; Kubota, Y.; Kato, K.; Takata, M.; Yamamoto, T.; Matsumura, S.; Kitagawa, H. Journal of the American Chemical Society 2014, 136, 10222–10225. (26) Zhdanov, V. P.; Kasemo, B. Chemical Physics Letters 2008, 460, 158–161.

18

ACS Paragon Plus Environment

Page 18 of 21

Page 19 of 21 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

Nano Letters

(27) Bugaev, A. L.; Guda, A. A.; Lomachenko, K. A.; Shapovalov, V. V.; Lazzarini, A.; Vitillo, J. G.; Bugaev, L. A.; Groppo, E.; Pellegrini, R.; Soldatov, A. V. et al. The Journal of Physical Chemistry C 2017, 121, 18202–18213. (28) Narayan, T. C.; Hayee, F.; Baldi, A.; Koh, A. L.; Sinclair, R.; Dionne, J. A. Nature Communications 2017, 8, 14020. (29) Ulvestad, A.; Welland, M.; Cha, W.; Liu, Y.; Kim, J.; Harder, R.; Maxey, E.; Clark, J.; Highland, M.; You, H. et al. Nature Materials 2017, 16, 565. (30) Li, J.; He, K.; Meng, Q.; Li, X.; Zhu, Y.; Hwang, S.; Sun, K.; Gan, H.; Zhu, Y.; Mo, Y. et al. ACS Nano 2016, 10, 9577–9585. (31) He, K.; Zhang, S.; Li, J.; Yu, X.; Meng, Q.; Zhu, Y.; Hu, E.; Sun, K.; Yun, H.; Yang, X.-Q. et al. Nature Communications 2016, 7, 11441. (32) Liu, X. H.; Fan, F.; Yang, H.; Zhang, S.; Huang, J. Y.; Zhu, T. ACS Nano 2013, 7, 1495–1503. (33) Li, Y.; Li, Y.; Sun, Y.; Butz, B.; Yan, K.; Koh, A. L.; Zhao, J.; Pei, A.; Cui, Y. Nano Letters 2017, 17, 5171–5178. (34) Wicke, E.; Brodowsky, H.; Z¨ uchner, H. Hydrogen in Metals II ; Springer, 1978; pp 73–155. (35) Bennett, P.; Fuggle, J. Physical Review B 1982, 26, 6030. (36) Aric`o, A. S.; Bruce, P.; Scrosati, B.; Tarascon, J.-M.; Van Schalkwijk, W. Nature Materials 2005, 4, 366–377. (37) Niu, W.; Zhang, L.; Xu, G. ACS Nano 2010, 4, 1987–1996. (38) Switendick, A. Hydrogen in Metals I ; Springer, 1978; pp 101–127.

19

ACS Paragon Plus Environment

Nano Letters 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

(39) Ludwig, W.; Savara, A.; Madix, R. J.; Schauermann, S.; Freund, H.-J. The Journal of Physical Chemistry C 2012, 116, 3539–3544. (40) Sachs, C.; Pundt, A.; Kirchheim, R.; Winter, M.; Reetz, M.; Fritsch, D. Physical Review B 2001, 64, 075408. (41) Ulvestad, A.; Welland, M.; Collins, S.; Harder, R.; Maxey, E.; Wingert, J.; Singer, A.; Hy, S.; Mulvaney, P.; Zapol, P. et al. Nature Communications 2015, 6 . (42) Hayee, F.; Narayan, T. C.; Nadkarni, N.; Baldi, A.; Koh, A. L.; Bazant, M. Z.; Sinclair, R.; Dionne, J. A. Nature Communications 2018, 9 . (43) Wu, J.; Qi, L.; You, H.; Gross, A.; Li, J.; Yang, H. Journal of the American Chemical Society 2012, 134, 11880–11883. (44) Jaumot, J.; Gargallo, R.; de Juan, A.; Tauler, R. Chemometrics and intelligent laboratory systems 2005, 76, 101–110. (45) Seymour, E.; Cotts, R.; Williams, W. D. Physical Review Letters 1975, 35, 165. (46) Gdowski, G.; Felter, T.; Stulen, R. Surface Science 1987, 181, L147–L155. (47) Okuyama, H.; Siga, W.; Takagi, N.; Nishijima, M.; Aruga, T. Surface Science 1998, 401, 344–354. (48) Van Hardeveld, R.; Hartog, F. Surface Science 1969, 15, 189–230. (49) Matsuda, A.; Mori, H. Chemical Physics Letters 2016, 644, 255–260.

20

ACS Paragon Plus Environment

Page 20 of 21

Page 21 of 21 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

Nano Letters

Graphical TOC Entry

21

ACS Paragon Plus Environment