Preserving surface strain in nanocatalysts via morphology control

Abstract

Engineering strain critically affects the properties of materials and has extensive applications in semiconductors and quantum systems. However, the deployment of strain-engineered nanocatalysts faces challenges, in particular in maintaining highly strained nanocrystals under reaction conditions. Here, we introduce a morphology-dependent effect that stabilizes surface strain even under harsh reaction conditions. Using four-dimensional scanning transmission electron microscopy (4D-STEM), we found that cube-shaped core-shell Au@Pd nanoparticles with sharp-edged morphologies sustain coherent heteroepitaxial interfaces with larger critical thicknesses than morphologies with rounded edges. This configuration inhibits dislocation nucleation due to reduced shear stress at corners, as indicated by molecular dynamics simulations. A Suzuki-type cross-coupling reaction shows that our approach achieves a fourfold increase in activity over conventional nanocatalysts, owing to the enhanced stability of surface strain. These findings contribute to advancing the development of advanced nanocatalysts and indicate broader applications for strain engineering in various fields.

Fig. 1.. Enhancing surface strain stability via nanoscale morphology control.

(A) Schematic illustration of strain preservation by a nanoscale morphology with sharp edges. (B) Atomic-resolution ADF-STEM images of the sharp-core (top left) and round-core (bottom left) Au_cube@Pd_cube NPs. Zoomed-in images from boxed area show a coherent interface (top right) and lattice mismatch (bottom right) between Au and Pd in sharp-core and round-core particles respectively, with the zone axes labeled. (C and D) Schematic showing 4D-STEM with a diffraction pattern (D, top) and an EWPC (D, bottom) from the core-shell NP for precise strain analysis. (E, F, H, and I) Strain maps (ε_xx, ε_yy, ε_xy, and ε_rot) from individual sharp-core (E and F) and round-core (H and I) Au_cube@Pd_cube NPs. The red and blue arrows indicate clockwise and counterclockwise rotation, respectively. (G and J) Lattice parameters that are parallel to the interface between Au and Pd for sharp-core (G) and round-core (J) NPs.

Fig. 2.. Morphology-dependent critical thicknesses.

(A to C) 4D-STEM strain maps (ε_xx with Au lattice as reference) and lattice-constant-ratio maps of the core-shell Au_cube@Pd_cube particles with different shell thicknesses (top to bottom: 1.7 ± 0.3, 5.5 ± 0.5, and 10.5 ± 0.5 nm, labeled aside) and different core sharpness with SI = 0.83 (A), 0.68 (B), and 0.32 (C). (D) Histogram of strain measurements in Pd shells from the samples in (A) to (C) with thickness increasing top to bottom. Bulk Pd lattice is used as the reference. (E) Average shell strain with increasing shell thickness for three core geometries. (F) Critical thickness of core-shell Au_cube@Pd_cube NPs depending on the core sharpness and particle size.

Fig. 3.. MD simulation of dislocation nucleation.

(A and B) 3D views of core-shell Au_cube@Pd_cube MD simulations with different shell thicknesses (1.0 and 5.0 nm) and different core sharpnesses [SI = 0.86 (A) and 0.60 (B)]. Local regions with body-centered cubic (bcc) coordination (blue)indicate large structural deformation of the face-centered cubic (fcc) lattice according to the Bain transformation path model. Regions with hexagonal-close packed (hcp) coordination (green) indicate stacking faults that form as a result of dislocations splitting into partial dislocations. (C) Dislocation density (total length over volume) as a function of the shell thickness from MD simulations. (D) Average longitudinal strain (ε_xx) at the center of the particles’ facets in a radius of 5 nm. (E to H) Strain maps of a slice through a simulation snapshot for a Au_cube@Pd_cube NP (SI = 0.86) with ~4 nm shell thickness. In (H), the dotted line traces the atomic column, which deviates from a straight line (solid), indicating the lattice bulging. The red and blue arrows in (H) indicate clockwise and counterclockwise rotation, respectively. (I and J) Shear maps from simulation snapshots from sharp (SI = 0.86) (I) and truncated (SI = 0.60) (J) NPs at a thin (~1 nm) and thick (~4 nm) shell thickness. Shear and rotation maps are averaged over five subslices to reduce noise.

Fig. 4.. Strain stability under catalytic reactions.

(A) Schematic illustration of the catalytic process of homocoupling in a Suzuki-type reaction on sharp-core core-shell Au_cube@Pd_cube NPs with tensile strain. This type of reaction relies on the dissolution, reduction, and redeposition of Pd, which presents a harsh environment for strain-engineered catalysts. (B) The average strain in sharp- and round-core Au_cube@Pd_cube NPs with a 4 nm shell thickness before and after the catalytic reaction. (C) The kinetic profiles of homocoupling reactions from Au_cube@Pd_cube catalysts with sharp (purple and green) and round cores (red). Top: schematic of homocoupling reaction catalyzed by Pd surfaces. (D) Comparison of the catalytic activities of pure Pd_cube particles (gray) (27) with round-core (red) and sharp-core (purple and green) Au_cube@Pd_cube nanocatalysts. The rate is normalized by the surface area of Pd. (E) Comparison of the yield between round-core (red) and sharp-core (purple and green) Au_cube@Pd_cube nanocatalysts.