Lattice Strain and Surface Activity of Dislocation‐Distorted AgPd Nanoalloys Under Preoxidation and Catalysis Condition

Although line defects endow excellent catalytic performance by undercoordinated sites with compressive tensile strain, few studies have systematically unraveled the relationship between dislocation, strain, and electrochemical activity for formate oxidation reactions (FOR). Herein, a novel approach for synthesizing defect‐rich nanomaterials at room temperature is proposed for the first time. The heated and dealloyed AgPd nanoparticles (hd‐AgPd NPs) substantially improve the intrinsic electrocatalytic activity by introducing compressive strain to tune its electronic structure. Electrochemical experiments show that the mass activity of hd‐AgPd NPs for FOR is 5.3 times higher than that of pure Pd nanoparticle catalysts. Following a 3600 s chronoamperometric process, a portion of the dislocation vanishes, but the strain persists on the AgPd (111) facet. The mechanisms for activity enhancement are further explored through density functional theory and molecular dynamics calculations, which show that compressive strain effectively alters its electronic structure and decreases the energy of the rate‐determining step during the reaction, significantly enhancing the FOR performance and stability. The results of electrochemical performance and physical characterization show that lattice strain has a more significant impact on FOR performance than alloying and preoxidation. This study presents a new approach to produce high‐performance catalysts by inducing strain into nanoparticles.


Introduction
The overuse of fossil fuels has resulted in a range of environmental and societal issues.To address this, many countries have implemented strategies to reduce emissions of greenhouse gases. [1]Highperformance electrocatalysts for utilizing renewable energy sources are becoming increasingly important. [2]Bimetallic nanostructures are considered promising candidates for fuel cells due to their unique physicochemical properties and synergistic effects compared to monometallic catalysts. [3]Recently, alloying Pt or Pd with other non-noble metals such as Cu, Fe, and Ni has been shown to significantly improve catalytic activity through geometric effects, ligand, and electronic structure. [4]However, non-noble metals are prone to dissolve in strong acid or alkaline solutions.To overcome this challenge, many studies have found that Ag is a suitable candidate due to its resistance to corrosion and relatively affordable price compared to Pt and Au. [5]A variety of strategies for synthesizing bimetallic catalysts have been reported. [6]However, traditional wet chemical synthetic methods often require harsh conditions, such as corrosive chemicals and harmful gas. [7]To avoid aggregation of catalysts on the electrode, polymers, ligands, and surfactants are used to improve dispersion, but these can also reduce the activity of catalysts due to the presence of organic matter on the surface.In contrast, pulsed laser deposition (PLD) offers a convenient and environmentally friendly method for preparing kinetically stable alloy catalysts at room temperature. [8]train engineering is a recently developed method to optimize the electronic structure and improve the electrochemical activity of catalysts, thereby increasing reaction efficiency. [9]Many efforts have been made to manipulate the strain to enhance catalytic activity. [10]For example, Chen et al. reported that defects or lattice strain generated at the PdAg nanowires with grain boundaries would change the binding energy toward O/OH species, resulting in improved performance for oxygen reduction reaction. [11]ore-shell structures are an effective strategy for creating strain due to the large lattice mismatch between different metals.
Zhang et al. successfully produced defect-rich Pd aerogels through laser-assisted synthesis technology, which significantly enhanced catalytic activity for ethanol oxidation reactions by introducing defects and lattice strain. [12]9a] However, strain can also be gained through external mechanics or varied temperature, which is not ideal for catalysts due to strict requirements. [13]urface oxide, native oxide, and 2D oxides on metal materials were frequently overlooked in previous studies on noble metal catalysts, which is a process of atomic reconstitution. [14]These preparation processes were often carried out under ultrahighvacuum conditions to prevent oxidation.To take advantage of the synergistic effect of Pt group metals and metallic oxide, researchers have tried to use doped metal oxides as alternative catalyst supports.For example, Hirai et al. reported spin-orbit torque in Ta/Co/Pt/oxide films by modifying the adjacent oxide material to alter the electronic structure at the interface between Pt and oxide. [15]Li deposited low-content Pt on tungsten oxide and reduced graphene oxide aerogel to prepare low-Pt content/tungsten oxide/reduced graphene oxide aerogel catalysts through a facile solvothermal process, which exhibited high HER activity and excellent stability. [16]However, these are still oxide-supported nanoparticle (NP) mixtures that have high demands for dispersity of oxide-supported and catalysts.Few studies focus on the interaction between the catalyst and surface oxide, which is a kind of ultrathin oxide structure on the surface of the catalyst.
We have successfully synthesized high-performance and stable AgPd catalysts by introducing strain through local dislocations, without using any capping or stabilizing ligands.The hd-AgPd NPs, which were synthesized in this study, demonstrate outstanding electrocatalytic activity for the formate oxidation reaction (FOR).Dislocation was introduced into the NPs, which is a typical line defect formed when an extra half-plane is inserted into the crystal as an edge dislocation plays a primary role in the local strain field. [17]density functional theory (DFT) was carried to investigate the effect of strain on the electrochemical activity.To further understand the internal mechanism of strain generation, we used molecular dynamics (MD) simulations to calculate the strain variation of the AgPd crystal with dislocations at different temperature. [18]The experiments and theoretical calculations demonstrate that the dislocation defects in AgPd NPs can cause compressive strain, downshift the d-band center, and optimize the adsorption of intermediates, which can enhance the electrocatalytic activity and stability for FOR.This work provides a new guideline for tuning the microstructure of catalysts and effectively improving their activity.

Preparation Process and Microstructure
The AgPd NPs were synthesized using a PLD system, a method of physical vapor deposition. [19]In this process, a high-energy density laser is directed at an AgPd target, causing a small amount of the material to be ablated and form a plasma plume.
The ejected AgPd material provides a high forward-directed flux, enabling the growth of high-quality AgPd NPs on the substrate.The as-prepared AgPd NPs were first dealloyed (d-AgPd NPs) in a 0.5 M sulfuric acid solution using a potential range from 0.3 to 0.85 V at a scan rate of 50 mV s À1 for 15 cycles and then heated at 673.15 K (hd-AgPd NPs) according to an improved process, as shown in Figure 1a.The heated AgPd NPs without dealloying (h-AgPd NPs) were conducted as control sample.The hd-AgPd NPs were also investigated after 3600 s CA test (labeled as w-AgPd NPs).As shown in the schematic diagram of Figure 1b,c, the X-ray diffraction (XRD) peaks of the obtained AgPd NPs catalyst gradually shift to the right, indicating reduction of lattice spacing and presence of compressive strain in the treated AgPd catalyst.Atomic force microscopy (AFM) was used to study the morphology and thickness of AgPd NPs, as shown in Figure 2a and S1, Supporting Information.The AFM thickness profile of the pristine AgPd NPs film (deposited for 40 min) was around 6 nm.Low-transmission electron microscopy (TEM) images (Figure 2b) show that the synthesized AgPd NPs are evenly distributed with a size range of 2-15 nm.Additionally, a high-angle-annular dark-field-scanning transmission electron microscopy (HAADF-STEM) image (Figure S2c, Supporting Information) of the AgPd NPs was obtained, and the size of the NPs was evaluated using nanomeasurer software based on the TEM image.The size histogram summary is shown in Figure S2d, Supporting Information, with an average size of around 5 nm.It is worth noting that fine grain can not only enhance the mechanical properties, but also increase the specific surface area of the catalyst.The inductively coupled plasma optical emission spectrometry (ICP-OES) test results, presented in Figure S3, Supporting Information, reveal that the atomic percentages of Pd in the pristine AgPd, d-AgPd, h-AgPd, and hd-AgPd NPs are 50.63%,52.65%, 51.58%, and 54.12%, respectively.
The formation of bimetallic AgPd NPs in the codeposition of Pd and Ag metals on a glassy carbon electrode (GCE) was investigated using XRD. Figure 2c displays the XRD patterns of AgPd, d-AgPd, h-AgPd, and hd-AgPd NPs.The standard XRD peaks of Ag and Pd are also drawn for comparison.It can be seen that all the peaks align with the peaks of the monometallic Ag and Pd, indicating that the NPs are in an alloy state.The main peaks correspond to the face-centered-cubic (fcc) A1 phase with (111), (200), (220), and (311) planes.Figure 2d shows a zoomed-in section of the (111) peak for AgPd, d-AgPd, h-AgPd, and hd-AgPd NPs.After dealloying and heat treatment, the (111) peaks of the d-AgPd, h-AgPd, and hd-AgPd NPs shift slightly to the peaks of Pd, likely due to the insertion of Ag and O into Pd lattice structures.

Effect of Processing on Dislocation Density Variations
Figure 3 in the given text displays the morphology, strain, and composition analysis of as-prepared AgPd NPs.The lowresolution image (Figure 3a) shows that the NPs consist of self-assembled sub-10 nm AgPd NPs.The larger AgPd NPs are formed through the coalescence of smaller particles during the crystalline grain growth process.Figure 3b depicts the highresolution transmission electron microscopy (HRTEM) microstructure of the 5 nm AgPd NPs with well-defined lattice fringes,  indicating their good crystallinity.The fast-Fourier transform (FFT) images in Figure 3c,e correspond to the marked square in Figure 3b.These images demonstrate clear lattice spacing of 0.234 and 0.232 nm, which correspond to the (111) planes of AgPd alloy.This spacing suggests the formation of a bimetallic AgPd alloy, with a lattice distance between the Pd (111) and Ag (111) planes.The inverse fast Fourier transform (IFFT) images (Figure 3f,h) were also obtained from the FFT patterns shown in Figure 3c,e, respectively.The strain distribution maps were created from the IFFT images.These maps provide information about the dislocations in the atomic lattice (Figure 3f,h) and the corresponding strain distributions toward lattice planes (Figure 3g,i).The strain analysis was performed using strainþþ software.It was found that the PLD process, which involves rapid cooling of the substrate, generates some dislocations and introduces small strain into the AgPd NPs. Figure 3g,i shows typical strain maps of the AgPd NPs, showing compressive tensile strain.Compared to monometallic NPs, the AgPd NPs exhibit some dislocations due to the lattice distortion caused by the different atomic radii, which blocks the motion of dislocations.This is similar to the concept of the "solution strengthening theory." [20]The AgPd NP catalyst was further analyzed using HAADF-STEM with a transmission electron microscope which had an aberration correction system.Figure 3j shows the HAADF-STEM image and energy-dispersive X-ray spectroscopy (EDS) elemental mappings of AgPd NPs.The composition analysis indicates that the AgPd NPs mainly consist of Pd and Ag elements, with a minor presence of oxygen on the surface.The EDS line profile in Figure 3d reveals an atomic ratio of Pd to Ag at %1:1.Additionally, both Ag and Pd are evenly distributed throughout the NPs.
The TEM characterization of AgPd NPs after electrochemical dealloying in a 0.5 M H 2 SO 4 aqueous solution is shown in Figure 4.As seen in Figure 4a, the NPs' morphology experienced substantial alterations.The larger NPs are composed of sub-10 nm NPs, which are connected in a subtle manner.In comparison to the as-prepared AgPd NPs, there are distinct gaps between the sub-10 nm NPs.The dealloyed AgPd NPs (d-AgPd NPs) were examined using high-magnification TEM (Figure 4b), which revealed good crystallinity.Figure 4c displays the FFT pattern of the yellow square marked in Figure 4b, corresponding to the (111) and ( 200) planes of the AgPd alloy.The IFFT image (Figure 4e,f,h) was obtained from the FFT pattern in Figure 4c, revealing clear interplanar spacing of 0.230 nm and 0.200 nm corresponding to (111) and (200) planes of AgPd alloy.It is evident that the interplanar spacing of the (111) planes in d-AgPd NPs is less than that of AgPd NPs, indicating that some Ag on the surface was dissolved away during the electrochemical dealloying process.The IFFT images of Figure 4f,h also reveal dislocations ("T"-shaped symbols).Figure 4g,i shows the typical strain maps of the d-AgPd NPs, displaying compressive tensile strain distribution.The presence of strain is frequently observed in areas with defects.Dealloying is not primarily employed to attain lattice strain.Electrochemical dealloying does, however, expose additional Pd sites.Compared to Figure 3g and 4g.It should be noted that a portion of lattice strain may diminish during the dealloying process.The HAADF-STEM image and EDS elemental mapping images (Figure 4j) show that Ag and Pd are evenly distributed in the d-AgPd NPs.The EDS linescan (Figure 4d) was also conducted along the selected region in Figure 4j, revealing that both Pd and Ag elements are distributed in the d-AgPd NPs, with Pd being slightly more prevalent than Ag.The atomic ratio of Ag and Pd was also determined by EDS, which is 45.8: 54.2 for Ag: Pd in d-AgPd NPs.
The microstructure of d-AgPd NPs after heat treatment (labeled as hd-AgPd NPs) was analyzed in The IFFT patterns (Figure 5f,h) are also obtained from the red square iand ii, respectively.The strain distribution maps (Figure 5g,i) were created from Figure 5f,h.An interesting phenomenon can be observed that, where the red area on the left of Figure 5g is assigned to the region of Ag 2 O, and the blue area corresponds to the region of AgPd.This indicates that compressive tensile dislocations are generated in the oxidation process of Ag.The insertion of oxygen atoms into the lattice of AgPd alloy on the surface of NPs results in strain in the AgPd NPs.The strain distributions of pure Pd catalysts were also examined as a control, as depicted in Figure S7, Supporting Information.Notably, no significant strains were observed in the pure Pd catalyst in contrast to the hd-AgPd NPs.The addition of Ag may play a critical role in inducing strain.Figure 5d presents an EDS linescan of the selected area in Figure 5j, which illustrates that the percentage of oxygen in this silver oxide-rich region for hd-AgPd NPs is notably higher than that of AgPd and h-AgPd NPs.It is worth noting that the variation in oxygen content along this linescan profile in this area is directly proportional to the silver content.EDS testing revealed that the large region of hd-AgPd NPs contains %49.09% Ag.Furthermore, as shown in Figure 5j, the oxygen content is significantly higher at the top of the arrow (Ag 2 O area), which corresponds to the observed lattice spacing in square i of Figure 5b.

Influence of Dislocation Density on for Catalytic Performance
The electrocatalytic activity of catalysts is significantly influenced by their electrochemical specific surface area (ECSA).The ECSA of the catalysts were evaluated based on the reduction charge of PdO in cyclic voltammetry (CV) curves, as shown in Figure 6a.The ECSAs of hd-AgPd, d-AgPd, AgPd, and pure Pd NPs were determined to be 119.8,93.7, 78.5, and 41.3 m 2 g Pd À1 , respectively (Figure S8, Supporting Information).Figure 6b illustrates CV activity curves for the FOR in a nitrogen-saturated 1.0 M HCOOK and 1.0 M KOH aqueous solution at room temperature.As shown in Figure 6c, the mass activity of the hd-AgPd NPs reached as high as 4.8 A mg Pd À1 , which is higher than those of d-AgPd NPs (2.7 A mg Pd À1 ), AgPd NPs (2.1 A mg Pd À1 ), and pure Pd NPs (0.93 A mg Pd

À1
). Table S1, Supporting Information, provides a literature survey of previously reported catalytic properties of palladium-based and silver-based FOR catalysts.The hd-AgPd NPs exhibits the highest mass activity of 4.8 A mg Pd À1 among the previously reported palladium-based and silver-based FOR catalysts, indicating its potential commercial value as an anode catalyst in direct formate fuel cells (DFFCs).During the process of alloying and oxidation, the surface of hd-AgPd NPs undergoes significant roughening, potentially leading to the formation of point defects.Additionally, the introduction of oxygen into hd-AgPd NPs alters their electronic structure, consequently influencing their catalytic performance.
Stability is another important factor to consider for a catalyst to be commercially viable.The cycling stability of catalysts was also evaluated, as shown in Figure S9, Supporting Information.After 500 CV cycles, the hd-AgPd NPs exhibited the highest retained activity among all the catalysts.The results from the 500 CV cycles align with the CA curves, as shown in Figure 6d, where the hd-AgPd NPs exhibits the highest activity retention.CA curves were used to evaluate the stability of the electrocatalytic properties in 1.0 M KOH and 1.0 M HCOOK water solution at 0.53 V, as a function of working time (Figure 5d).After 3600 s, the residual activity was 440, 214, 109, and 11 mA mg Pd À1 for hd-AgPd, d-AgPd, AgPd, and pure Pd NPs, respectively.Among all the prepared catalysts, the hd-AgPd NPs exhibits remarkable stability.Figure 6e illustrates the regeneration ability of the hd-AgPd NPs in comparison to the pure Pd NPs.After an hour of potentiostatic polarization, the catalysts were reactivated by 10 times CV in a nitrogen-saturated 1.0 M KOH water solution.The hd-AgPd NPs recovered 93% of its initial activity after four potentiostatic polarizations, while the pure Pd catalyst recovered only 26% of its initial activity under the same conditions.This roughening increases the specific surface area and active sites.The higher ratio of activity recovery suggests that the hd-AgPd NPs have better reproducibility than the pure Pd catalyst. [21]igure S10, Supporting Information, presents the CV and CA curves of hd-AgPd, pure Pd, and pure Ag NPs in an electrolyte of 1.0 M KOH þ 1.0 M HCOOK.The electrocatalytic activity toward FOR is closely related to the active sites of Pd, and there is no formate oxidation peak for the pure Ag samples, indicating that alloying with Ag can enhance FOR performance, even though monometallic Ag does not act as a catalyst for the electrochemical reaction.Additionally, CV is a useful technique to confirm the chemical composition of the catalyst surface.The AgO reduction peak of hd-AgPd NPs is much less than that of pure Ag, suggesting that Pd occupies most of the active sites on the hd-AgPd NPs surface, effectively blocking the oxidation-reduction of Ag.As shown in Figure S11, Supporting Information, an EDS line scan was also performed along the edge of hd-AgPd NPs.The resulting Pd/Ag ratio of %58.53/41.47indicates that Pd is enriched on the surface of the particles.The enrichment of Pd on the hd-AgPd NPs surface can be attributed to two factors: 1) the dealloying process of AgPd NPs; and 2) during the heat treatment and upper limiting potential scanning process, Ag atoms on the surface were oxidized into Ag 2 O, with only a few Ag atoms remaining on the surface.In addition, the electrocatalytic activity of hd-AgPd NPs toward FOR was examined at different upper limiting potentials, ranging from 1.02 to 1.62 V versus RHE (Figure S12, Supporting Information).The peak mass activity for the hd-AgPd NPs increased from 1.9 to 4.8 A mg Pd

À1
, indicating that the activity of hd-AgPd NPs increased significantly with an increase of the upper limiting scanning potential.This upper limiting potential scanning method also confirms that preoxidation treatment on the surface of the catalyst can enhance the electrocatalytic performance toward FOR.

Influence of Strain on the Performance of FOR
To better understand why the hd-AgPd NPs with strain exhibit the highest FOR activity among the samples prepared, three heterojunction models were constructed to simulate the hd-AgPd NPs with metallic surface oxide on the surface.The strain rate was varied from À5% to þ5% based on the lattice parameter in the HRTEM image analysis and strain distribution maps.Figure 7a,b shows the adsorption configuration of FOR path and the free energy graph for the FOR process on hd-AgPd NPs with À5%, 0, and þ5% strain, respectively. [22]he FOR process was studied through the following steps: and "mo" represent double and single oxygen adsorption, respectively.The step from HCOO bi !HCOO mo is the ratedetermining step for the entire reaction. [23]The modulation of the electronic structure by compressive strain is worth noting, as hd-AgPd NPs with À5% strain has the lowest energy barrier of 0.36 eV for the rate-determining step.This low barrier can speed up the dissociation of HCOO À at a lower potential, thus accelerating the FOR rates thermodynamically.Figure 7c shows the partial density of state distributions for hd-AgPd NP with À5%, 0, and þ5% strain, with d-band centers of À3.04, À3.28, and À3.31 eV, respectively.The downshift of the d-band center is generally thought to result from electron transfer. [24]The hd-AgPd Ns with -5% strain is closest to the Fermi level among the three models, indicating that the electronic structure is related to the modification of strain. [25]igure S13, Supporting Information, shows the calculated state densities and d-band centers of Pd and Ag NPs under the same conditions as hd-AgPd NPs.The positive shift in the d-band center of hd-AgPd NPs suggests that partial surface electrons of Ag transfer to Pd.The shift of the d-band center for catalysts may adjust the adsorption capacity of reactants, intermediates, and reaction products and improve the performance of the electrocatalytic process. [26]The adsorption energies for OH À on the catalysts are shown in Figure 7d.The OH À adsorption energy was determined to be À1.62,À1.57, and À1.26 eV for hd-AgPd NPs with À5%, 0, and þ5% strain, respectively.The higher OH À adsorption energy induced by the surface metal oxide contributes to providing oxygen functional groups to efficiently desorb the H* intermediates during the electrochemical reaction.The DFT results indicate that strain can subtly alter the electronic properties of the surface atoms in hd-AgPd NPs, thereby improving the performance of the FOR.
Early research suggested that oxidation may decrease the performance and stability of nano alloy catalysts.However, recent studies have shown that the surface of AgPd nanoalloys can be reconstituted after in-situ electrochemical or gas-phase oxidation.The new catalytic active phase is derived from high-oxidation-state metal atoms and surface adlayers under strain.Specifically, high-oxidation-state Ag atoms can increase the adsorption of OH À , while negative strain can increase the adsorption energy of OH À and decrease the adsorption energy of H þ .These effects have been confirmed through extensive research on the electronic and strain influences on FOR activity.

Relationship between Dislocation Density and Strain
To understand the relationship between dislocation and strain, MD simulations were conducted to calculate the strain of AgPd with dislocation using the open-source software LAMMPS, while the perfect AgPd crystal was also simulated for comparison, as seen in Figure S14, Supporting Information.Based on the MD calculations, as shown in Figure S14b,d, Supporting Information at 673.15 K, some patchs of Pd atoms dispersed on the top of the AgPd NPs.Only a small quantity of Pd atom, Pd dimer, or Pd trimer can be observed.Figure 8a,b and S15, Supporting Information, shows the strain of the AgPd crystal in the range of 300-1500 K, with a step of 10 K.When the temperature is below 1320 K, it can be observed that AgPd crystal with one and three dislocations have compressive strain, while the comparison models exhibit tensile strain under the same conditions.Figure 8c,d and S16, Supporting Information, illustrate the volumetric strain of perfect AgPd and AgPd with a dislocation heated from 300 to 673.15 K.In comparison to AgPd, the surface of AgPd with a dislocation is dominated by compressive strain at 673.15 K. AgPd with a dislocation shrinks after MD simulations to become a lower-energy state than that of the perfect AgPd, which eventually results in the compressive atomic strains on the surface. [27]Burgers reported that when the movement of an array of dislocations is stopped, it leads to a large-scale local stress concentration. [28]dditionally, temperature is a critical factor in determining both stability and creep resistance.It is observed that the AgPd catalyst with dislocations experienced less deformation compared to the perfect AgPd crystal.Kanninen reported that dislocations on a glide plane can be blocked by a single locked dislocation, stopping the movement of the slip plane by the dislocation pile-up. [29]The reduced deformation is beneficial for the structural stability of the catalyst, and this is the first report of the strain for the AgPd catalyst using the method of thermal oxidation and MD simulations.Therefore, the presence of dislocations can induce strain, which not only modifies the electronic structure properties of the AgPd catalyst, but also enhances its stability.

Durability of hd-AgPd NPs Catalyst during FOR Process
The above experiments have shown that surface modification of the AgPd NPs can effectively improve FOR performance.In order to further explore the mechanism behind the observed enhancement, X-ray photoelectron spectroscopy (XPS) analysis was conducted to investigate the surface electronic structure and valence states of AgPd NPs with different treatments processes, as shown in Figure 9 and S17, Supporting Information.Figure 9a shows the high-resolution Pd 3d XPS spectrum of AgPd NPs catalysts referred to as various AgPd alloys, which can be deconvoluted into native oxide of Pd, Pd (0) and alloyed Pd, represented by red, white, and yellow, respectively.Alloyed Pd is only a part of Pd atoms from the solid solution alloy.In addition to being a homogeneous mixture (atomically mixed alloys), the combination of silver and palladium can also be referred to as an alloy.A pure Pd sample was also analyzed under the same conditions as a control, as shown in Figure 9b.The spectrum can be deconvoluted into native oxide of PdO, native oxide of Pd and Pd (0), which was represented by rose red, red, and white, respectively.In contrast to the pure Pd catalyst, it can be observed that the alloyed Pd peaks (represented by yellow) are present in the AgPd catalysts.In combination with XPS and HRTEM, there was some surface oxide presented in the surface of catalyst in the hd-AgPd.The O atoms on the surface of AgPd and d-AdPd NPs were mainly adsorbate species.
The area of alloyed Pd peaks increased after CV, indicating that zero-valent Ag dissolved in zero-valent Pd.After CA, the area of the alloyed Pd peak in w-AgPd NPs was also larger than that of AgPd NPs after heat treatment, showing that a portion of alloyed Pd generated by kinetic factors can be maintained on the surface of the catalyst.This is an intriguing phenomenon in which AgPd alloy is formed not only during PLD, but also during heat treatment process.When the h-AgPd and hd-AgPd NPs undergo the pretreatment scan of CV, the surface of AgPd NPs acquires a negative charge during the second half of the scan process, potentially impacting the XPS measurement results.It can be observed that the peaks of Pd and Ag shift negatively after CV and CA, as shown in Figure 9a-c.XPS is an important technology to discover changes in metallic state.This study presents the first report on the variation of alloyed Pd content with different treatment processes.Additionally, there is a PdO peak that appears at about 338 eV in Figure 9b, which is not present in Figure 9a.This comparison experiment confirms that the oxidation resistance of Pd is dramatically enhanced when Ag dissolves in Pd.
The ratio of high-valence state of Pd was calculated by integration of the oxidation area, as shown in Figure 9d.In the stages of AgPd NPs and d-AgPd NPs, the high-valence Pd in both AgPd NPs and pure Pd NPs catalysts are similar, being less than 10%.After heat oxidation, the ratio of the high-valence Pd in the pure Pd catalyst increases significantly, almost twice that of the AgPd catalyst.This suggests that Pd becomes more stable when Ag is dissolved in it.This study reveals the reason why Pd in AgPd NPs catalysts has more stable performance than pure Pd catalysts.The blue curve in Figure 9d shows the percentage of alloyed Pd in AgPd NPs.Alloyed Pd is a particular state of Pd that has been rarely reported and confirms that atomic-level miscible AgPd alloy can be successfully obtained through this preparation process.The mass activities of AgPd, d-AgPd, h-AgPd, hd-AgPd, and w-AgPd NPs are depicted in the histogram.As the number of dislocations increased, the catalytic performance for FOR was observed to improve.Despite a decrease in dislocation density after 3600 s of operation, w-AgPd NPs exhibited excellent catalytic performance.
Figure S18, Supporting Information, shows the HAADF-STEM and strain map characterization of hd-AgPd NPs after CA. Figure S18a,b, Supporting Information, shows the TEM and HRTEM of the w-AgPd NPs, with two FFT images in Figure S18c,e, Supporting Information, corresponding to the red dashed line boxes i and ii in Figure S18b, Supporting Information, respectively.The morphology experienced significant changes compared to that of hd-AgPd.Distinct gaps between NPs have disappeared and the particles have become larger than hd-AgPd NPs.The distribution of oxygen signal does not match the NP perfectly in Figure S18j, Supporting Information, which is different from previous catalysts of AgPd, d-AgPd and hd-AgPd NPs.The elemental EDS mapping and linear scanning profile are displayed in Figure S18j,d, Supporting Information.It can be seen that the oxygen content decreased significantly after CA.This is because the CA was conducted at a reduction potential, which reduced most of the PdO to metallic Pd.Thus, the surface roughened substantially by the consistent electrochemistry behavior.
The IFFT patterns in Figure S18f,h, Supporting Information, of area i and ii show d-spacings of 0.225 and 0.234 nm, respectively, which are indexed to Pd (111) and AgPd (111) facets.The strain distributions of w-AgPd NPs were also analyzed using strainþþ.Despite the reduction of oxide on the catalyst surface, some compressive strain remained on the Pd and AgPd facets, as shown in Figure S18g,i, Supporting Information.In comparison with Figure 4i, the strain on the AgPd (200) facets (Figure S19, Supporting Information) was found to have significantly decreased for w-AgPd NPs.It can be observed that the strain on the AgPd (111) facet was more stable than that of the AgPd (200) facet.

Conclusion
In summary, we have prepared AgPd NPs with rich compressive strain through thermal oxidation.The resulting hd-AgPd NPs exhibit excellent electrochemical formate reduction with a mass activity of 4.8 A mg Pd À1 , which is significantly higher than that of AgPd NPs (2.1 A mg Pd

À1
) without strain.Additionally, the hd-AgPd NPs display superior stability with less activity deterioration compared to Pd NP catalysts.The introduction of oxygen atoms and the formation of partial oxides on the surface effectively altered the strain distributions through heat oxidation.Electrochemical experiments demonstrate that the two processes mentioned above substantially improve the performance of FOR.The DFT results indicate that the incorporation of strain into AgPd NPs can expedite the rate of FOR reaction.Additionally, the alloyed Pd was confirmed in the AgPd alloy, which is a particular state distinguished from metallic Pd, and rarely reported.It is proposed that the enhanced electrocatalytic activity can be attributed to the compressive strain resulted by dislocation defects, thus substantially reducing the d-band center of AgPd NPs, thereby reducing the energy of the rate-determining step in the entire reaction.This discovery offers a novel approach for creating high-performance catalysts through the introduction of strain into NPs.

Experimental Section
Materials: Pd-Ag alloy (with a 1:1 ratio of Pd to Ag), pure Pd (99.99 wt%), and pure Ag (99.99 wt%) were used as the laser-ablated target materials.The substrate used was a GCE.Potassium hydroxide (KOH) and sulfuric acid (H 2 SO 4 ) were obtained from Guangdong Guanghua Polytron Technologies Inc. Potassium formate (HCOOK > 99%), ethanol (CH 3 CH 2 OH), and acetone (CH 3 COCH 3 ) were sourced from Sinopharm Chemical Reagent Co Ltd and Tianjin Fuchen Co Ltd respectively.All aqueous solutions were prepared using ultrapure deionized water (18.25 MΩ cm).
Synthesis of AgPd NPs: The Pd-Ag alloy was fabricated using PLD on GCE in a stainless steel vacuum chamber (pressure 2.0 Â 10 À4 Pa).A nanosecond Q-switched Nd: YAG laser (EKSPLA, Lithuania) with a wavelength of 266 nm and 200 mJ pulse À1 was used in this study.The diameter of the laser beam was 1 mm and the operation frequency was set to 10 Hz, with a pulse delay of 2150 units (unit = 0.25 μs) between Q-switched and flashlight.Before the PLD process, the GCE was cleaned in acetone solution for 30 min, soaked in ethanol for 10 min, washed in deionized water for 10 min, and dried in a vacuum drying oven for 2 h.
The GCE was placed on a rotating sample platform of the PLD, which was positioned 5 cm from the target.The Pd-Ag target was created by directly combining the elements in the desired stoichiometric ratio.The laser beam was focused onto the rotating target at a 45°angle.Both the target and the sample platform kept rotating at 10 and 5 revolutions min À1 , respectively.The ambient temperature inside the chamber was monitored by a K-type thermocouple throughout the deposition process, which could be observed through a viewport window.Once the deposition process was completed, high-purity nitrogen was introduced into the chamber, and the sample was slowly cooled to room temperature.
Preparation of d-AgPd NPs: The dealloyed AgPd catalyst was synthesized in a three-electrode electrochemical setup at room temperature, using a platinum foil and a silver-silver chloride electrode as the counter and reference electrodes, respectively.The potentials were recorded against the reversible hydrogen electrode (RHE) using the equation: E (RHE) = E (applied) þ 0.235 þ 0.0592 pH.The dealloying process was performed in a nitrogen-saturated 0.5 M sulfuric acid solution using a potential range of 0.3-0.85V at a scan rate of 50 mV s À1 for 15 cycles.After the process, the dealloyed AgPd was washed with ultrapure water and dried naturally in air.
Preparation of h-AgPd NPs and hd-AgPd NPs: The as-prepared AgPd and d-AgPd catalysts were placed in a quartz tube furnace and heated at 673.15 K for 15 min.The heating process was performed in a low-oxygen environment, which was maintained at 100 parts per million, as measured by an oxygen analyzer.
Physical Characterization: The composition and structure of the as-prepared catalyst samples were characterized using various techniques, including HRTEM, EDS, XRD, XPS, and AFM.HRTEM images were obtained using a FEI Talos FRG 200F transmission electron microscope and EDS was used to examine the elemental composition.XRD was performed using a PANalytical X'Pert Pro MPD spectroscope, with Cu Kα radiation (λ = 1.5406Å) at a grazing incidence of 1°.XPS was conducted in an ultrahigh-vacuum environment (10 À9 Torr) using an ESCALAB 250 spectroscope and an Al Kα source (1486.6 eV).The binding energies were calibrated using the carbon peak (C 1s) at about 284.6 eV.The geometric phase analysis (GPA) was investigated using the strainþþ software, based on HRTEM images. [30]AFM was conducted using a Bruker Dimension Icon in tapping mode, and the size of the NPs was evaluated using the nanomeasurer software based on TEM images of the catalysts.
Electrochemical characterization: The electrochemical experiments were performed using a CHI660c electrochemical work station that was equipped with a three-electrode electrochemical cell.The GCE coated with the samples, a platinum foil electrode, and a Hg/HgO electrode in 1 M NaOH solution were used as the working electrode, counter electrode, and reference electrode, respectively.Prior to each test, the electrolyte was purged with high-purity nitrogen for at least 20 min to remove any oxygen.CV measurements were conducted at a sweep rate of 50 mV s À1 .The stable CV curve was obtained in 1 M KOH solution after repeatedly cycling between À0.9 and 0.6 V for ten cycles.The stability of the FOR was recorded through CA tests.The initial potential of the CA was set at 0.53 V versus RHE.The hd-AgPd catalyst after CA for 3600 s was labeled as w-AgPd.The electrochemically active surface area (ECSA) of the catalysts was determined from the integrated area of the PdO reduction peak, which was recorded in a 1 M KOH solution at a sweep rate of 50 mV s À1 .The electrochemical performance of the catalyst toward FOR was evaluated by CV and CA in a 1 M KOH þ 1 M HCOOK aqueous solution.
To ensure accuracy, all the tests were repeated at least three times.
DFT Calculations: In the DFT calculations, we utilized the DMol3 package, which is based on atomic orbitals. [31]We employed the generalized gradient approximation (GGA) of Perdew-Burke-Ernzerhof (PBE) functions for the exchange and correlation energies. [32]We also selected the spin unrestricted option and used formal spin as the initial input in the DMol3 calculations.For core treatment, we used DFT semicore pseudopotentials (DSPPs) and double-numerical plus polarization (DND) as the basis set.The self-consistent-field (SCF) tolerance was set to a value of 10 À5 Hartrees (1 Hartree = 27.2114eV).For multipolar expansion, we selected the hexadecapole option.To model AgPd, we used a four-layer slab model and inserted a vacuum layer of 20 Å to prevent interaction between the slabs. [33]Additionally, for hd-AgPd modeling, we built a heterojunction model using the four-layer slab AgPd model and a monolayer Ag 2 O.The bottom two-layer atoms were fixed during the geometry optimization and adsorption process, while the remaining atoms were fully relaxed.
Molecular Dynamics Simulations: In the MD simulations, we constructed AgPd crystal samples, as shown in Figure S12, Supporting Information.The samples had a size of 6 Â 6 Â 6 nm 3 and consisted of 13 500 atoms.26a] We used the embedded atom method potential (EAM) to explore the deformation behavior of the samples and to describe the atomic interaction. [34]We applied fully periodic boundary conditions in our simulation models and used the isothermal-isobaric (NPT) ensemble. [35]We processed dump files generated from LAMMPS using OVITO basic software and employed the common neighbor analysis (CNA) to analyze the evolution of the samples' local atomic structure. [36]

Figure 1 .
Figure 1.a) Preparation schematic illustration of AgPd NPs with compressive strain through dealloying and heat oxidation technology.b) Schematic of how pristine AgPd NPs transform into hd-AgPd NPs after heat oxidation, which can introduce strain effects into the particles.c) The XRD peak shifts of (220) peak after dealloying.The h-AgPd, d-AgPd, and hd-AgPd NPs represent heated, dealloyed, and heated and dealloyed AgPd NPs, respectively.w-AgPd NPs represent the sample that hd-AgPd NPs worked after 3600 s CA test.

Figure 2 .
Figure 2. a) Tapping-mode AFM images of AgPd NPs and thickness of AgPd NPs synthesized by PLD for 40 min with 24 000 pulses.b) TEM image of AgPd NPs.c) XRD patterns of AgPd, d-AgPd, h-AgPd, and hd-AgPd NPs catalysts.The standard XRD patterns for Ag and Pd are also given for reference.d) Zoomed-in view of the (111) peak region corresponding to AgPd, d-AgPd, h-AgPd, and hd-AgPd NPs catalysts.

Figure 3 .
Figure 3. TEM and strain characterization of as-prepared AgPd NPs.a) TEM image of AgPd NPs.b) HRTEM image recorded from a part of region in image (a).c,e) The FFT patterns of the framed part (red dashed line) in panel (b).d) EDS line scan profile along the red arrow in the HAADF-STEM image.f,h) IFFT patterns of the (111) planes corresponding to the FFT patterns in (c,e) respectively.g,i) Strain distributions maps of (f,h) (the scale bar on the right of strain distribution maps represents compressive and tensile strain values).j) HAADF-STEM image and energy-dispersive X-ray (EDX) elemental mapping images of AgPd NPs.

Figure 4 .
Figure 4. TEM and strain characterization of d-AgPd NPs.a) TEM image of d-AgPd NPs.b) HRTEM image recorded from a part of region in image (a).c,e) FFT and IFFT patterns of the framed part (yellow dashed line) in panel (b).d) EDS linescan profile along the red arrow in the HAADF-STEM image.f,h) IFFT patterns of the (111) and (200) planes corresponding to the FFT patterns in (c), respectively.g,i) Strain distributions maps of (f,h).j) HAADF-STEM image and EDX elemental mapping images of d-AgPd NPs.
Figure 5 using 100 ppm oxygen at 673.15 K for 15 min, with technical parameters adjusted and optimized, as shown in Figure S4-S6, Supporting Information.The size of the crystallites is in the range of 4-5 nm, as shown in Figure 5a,b.The IFF and IFFT patterns in Figure 5c,e correspond to the red marked square i in Figure 5b and the lattice spacing distances presented in the IFFT pattern of Figure 5e are 0.273 and 0.234 nm, which are approximately assigned to the (111) planes of Ag 2 O and AgPd.

Figure 5 .
Figure 5. TEM and strain characterization of hd-AgPd NPs.a) TEM image of hd-AgPd NPs.b) HRTEM image recorded from a part of region in (a).c,e) The FFT and IFFT patterns of the above framed part "i" (red dashed line) in panel (b).d) EDS linescan profile along the red arrow in the HAADF-STEM image (j).f,h) IFFT patterns of the (111) planes corresponding to the FFT patterns in (c) respectively.g,i) Strain distributions maps of (f,h).j) HAADF-STEM image and EDX elemental mapping images of hd-AgPd NPs.

Figure 6 .
Figure 6.The electrocatalytic performance of hd-AgPd NPs, compared to d-AgPd NPs, AgPd NPs, as well as Pd NPs control sample.a) CV curves recorded in nitrogen-saturated 1 M KOH aqueous solution.b) CV curves performed in nitrogen-saturated 1 M HCOOK and 1 M KOH aqueous solution.c) Mass activities of various AgPd and Pd NPs.d) CA curves of various AgPd NPs and Pd NPs for 3600 s in nitrogen-saturated 1 M HCOOK and1 M KOH aqueous solution at 0.53 V versus RHE as a function of work time.e) Regeneration capacity of hd-AgPd NPs compared to Pd NPs catalyst.

Figure 7 .
Figure 7.The DFT calculations were performed on hd-AgPd NPs under À5%, 0%, and þ5% strain.a) Adsorption configuration of FOR path on hd-AgPd NPs, with red, gray, white, blue, and baby blue denoting O, C, H, Pd, and Ag, respectively.b) Free energy diagram for the FOR steps with the ratedetermining step of the entire reaction transitioning from HCOO bi * to HCOO mo *. c) State densities of hd-AgPd NPs under À5%, 0%, and þ 5% strain.The d-band centers are marked with vertical dashed line.d) Adsorption energy of OH À species.

Figure 8 .
Figure 8. a) Strain of AgPd with one dislocation at the heating rate of 5 Â 10 À12 K s À1 with a 10 K step.b) Strain of AgPd with three dislocations under the same conditions.AgPd with no dislocation was examined as a control.c) Surface atomic strain distributions on AgPd at 673 K. d) Surface atomic strain distributions on AgPd with one dislocation at 673 K.The strain intensity is indicated by the color pattern, with negative values representing compressive strain and positive values representing tensile strain.

Figure 9 .
Figure 9. XPS spectra of Pd and Ag in AgPd NPs alloy and pure Pd NPs.a) High-resolution XPS spectrum of Pd 3d for AgPd NPs alloy.b) High-resolution XPS spectrum of Pd 3d for pure Pd NPs.c) High-resolution XPS spectrum of Ag 3d for AgPd NPs alloy.d) Effect of dislocation density, alloying, and oxidation during preoxidation and catalysis on the catalytic activity of AgPd nanoalloys.The lattice strain from dislocation, oxidation from PdO x in AgPd, and alloying from alloyed Pd in AgPd.The mass activity of different AgPd NPs is the column in the chart.It is clear that lattice strain has a greater impact on FOR performance than that of alloying and oxidation.