Self-catalytic growth of one-dimensional materials within dislocations in gold

Significance

Dislocation edges on metal surfaces provide loosely bonded atoms, which make the surface more active for chemical corrosion and enhance the chemical reactivity. In this work, we introduce and demonstrate an autocatalytic method in which the dislocation lines in alloys can serve as a template for the growth of a different inorganic material as nanowires. We explain the formation mechanism of an incorporated system of single crystal gold(I)-cyanide nanowires on nanoporous gold during selective dealloying, using the dislocation ends in the prestrained Au-Ag alloy as catalytic centers for crystallization. This method may provide a possible route for controlling the growth direction, shape, and morphology of a crystal according to the original alloy’s slip system.

Abstract

Dislocations in metals affect their properties on the macro- and the microscales. For example, they increase a metal’s hardness and strength. Dislocation outcrops exist on the surfaces of such metals, and atoms in the proximity of these outcrops are more loosely bonded, facilitating local chemical corrosion and reactivity. In this study, we present a unique autocatalytic mechanism by which a system of inorganic semiconducting gold(I) cyanide nanowires forms within preexisting dislocation lines in a plastically deformed Au-Ag alloy. The formation occurs during the classical selective dealloying process that forms nanoporous Au. Nucleation of the nanowire originates at the surfaces of the catalytic dislocation outcrops. The nanowires are single crystals that spontaneously undergo layer-by-layer one-dimensional growth. The continuous growth of nanowires is achieved when the dislocation density exceeds a critical value evaluated on the basis of a kinetic model that we developed.

Dislocations in metals have been known for almost a century, ever since edge dislocations were postulated in 1934 by Orowan (1), Polanyi (2), and Taylor (3). Dislocations can form during crystal growth as well as due to plastic deformation. Dislocations formed in metals lead to an increase in their strength, hardness, and brittleness (4). Dislocation outcrops on the metal surfaces can be easily etched by corrosive chemicals to form dissolution etch pits (5, 6). Atoms at the dislocation outcrops are more reactive and more loosely bonded than other atoms on the surface and therefore demonstrate enhanced chemical reactivity (7–9). It was shown several decades ago that plastically deformed copper (10) and silver (11) are more catalytically active than their annealed counterparts, which have lower dislocation densities. More recently, it was also shown that the chemical properties of strained metal surfaces differ considerably from those of unstrained surfaces (12) and in particular that catalysis can be enhanced near the points of intersection of edge dislocations with the surface (13). Here, we demonstrate that dislocations formed upon plastic deformation in a Au-Ag alloy can serve as catalytic centers for the nucleation of one-dimensional (1D) semiconducting gold(I) cyanide nanowires (AuCN NWs). We further show that the dislocation lines serve as templates for the formation of nanowires (NWs).

We subjected an Au₅₀Ag₅₀ alloy to varying degrees of plastic deformation, followed by the classical free-corrosion selective dealloying of Ag, a procedure known to form nanoporous (np) Au (14–19). The latter exhibits multiple properties, such as high surface area (20–22), chemical and thermal stability (23, 24), and excellent catalytic activity (23, 25–28). Owing to its wide range of properties, np-Au has been used extensively in sensors (29, 30), catalysis (15, 27, 31–33), rechargeable batteries (34, 35), actuators (36, 37), electrodes (38–40), and more.

Here, we present a method for the formation of AuCN NWs along dislocation lines, which provide a template for their growth and are integrated in a classical np-Au structure. We did this by instigating an autocatalytic reaction on the dislocation outcrops of plastically deformed Au-Ag alloys during selective dissolution of the Ag. This work reports the layer-by-layer 1D growth of AuCN NWs by means of an autocatalytic reaction at dislocation outcrops. We developed a kinetic model that demonstrates the correlation between AuCN NWs formation and the dislocation density in the predeformed alloy.

Results and Discussion

To study the effect of dislocation outcrops on potential changes in the rate and anisotropic leaching of Ag from Au-Ag alloy, we repeatedly loaded 0.25- to 0.5-g pellets of Au_0.5Ag_0.5 alloys (99.99% pure, ACI Alloys) using 1 to 25 loading cycles with uniaxial compressive stresses of 10-tons force each, providing compression strains of 85 to 91%, well above their yield strength (Methods). The loaded samples, as well as a control nonloaded sample, were subjected to selective leaching via classical free corrosion of Ag (16, 17) using 70% aqueous nitric acid (HNO₃) solution for 168 h while stirring at room temperature. Dealloying of the control sample resulted in the formation of classical np-Au morphology (Fig. 1A). In the preloaded samples, we observed, unexpectedly, a completely different and nontypical morphology (Fig. 1 B–E). Their surfaces appeared black (Fig. 1 B, Inset and SI Appendix, Fig. S1) rather than the characteristic reddish-gold color of the control np-Au samples (Fig. 1 A, Inset and SI Appendix, Fig. S1). Upon characterizing the preloaded samples by high-resolution scanning electron microscopy (HR-SEM) we observed, to our surprise, a dense array of NWs on the surface (Fig. 1B). Higher numbers of loading cycles resulted in thinner and longer NWs for identical etching durations. Average NWs thicknesses measured for samples loaded with 1, 10, and >>25 cycles were found to be 45.9 ± 7.3, 37.5 ± 5.5, and 7.7 ± 1.2 nm, respectively (Fig. 1 B–D). The appearance of NWs only when plastic deformation was applied prior to etching indicates that their formation is probably directly related to the plastic deformation of the Au-Ag alloy.

Fig. 1.

HR-SEM micrographs of AuCN NWs samples loaded at 10 tons for different cycles. (A) Reference sample, zero cycles (Inset, reddish-gold np-Au sample). (B) One cycle (Inset, black AuCN NW sample). (C) 10 cycles. (D) >>25 cycles, images showing interdependence of the predeformation and the morphology of the final NWs formed. (E and F) Cross-sectional micrographs of sample loaded for 10 cycles at (E) ∼5 μm below the surface, showing NWs incorporation within classical np structure and (F) ∼20 μm below the surface, showing classic np structure. (Scale bars, 600 nm.)

When the sample seen in Fig. 1C was mechanically fractured and its cross section was imaged, we observed that the NWs are present on the top ∼8 μm of the sample, while a classical np-Au structure is obtained beneath them (Fig. 1F). Nevertheless, at the interface between these two phases, the NWs could be seen growing from within the np-Au morphology (Fig. 1E). The Au volume fraction of the np structure was also measured via focused ion beam (FIB)-cut sections to be 68 ± 2% (Methods).

To decipher the composition and structure of the NWs formed, we further characterized them by high-resolution X-ray diffraction (HR-XRD) (Fig. 2A). Surprisingly, the NWs were identified as gold(I) cyanide (41) (AuCN) despite the fact that no cyanide ions were used in the etching solutions. Surface analysis via high-resolution X-ray photoelectron spectroscopy (HR-XPS) of a loaded sample revealed an intense gold(I) cyanide peak (Au⁺¹) for the Au4f line, whereas in unloaded np-Au control samples that were synthesized without the use of HNO₃ solution (Methods), we observed only the metallic state of gold (Au⁰) (Fig. 2B). Furthermore, no nitrogen was detected in the control sample (N1s line), whereas a clear cyanide peak was observed in the sample of AuCN NWs (Fig. 2C).

Fig. 2.

Structural and chemical characterization of NWs. (A) HR-XRD pattern of np-Au and loaded AuCN NWs samples (*, diffraction peaks of the AuCN NWs). (B and C) HR-XPS of the np-Au and AuCN NWs of (B) Au4f line. For np-Au sample (Au⁰), only the metallic state is observed, while for AuCN NW (Au⁺¹), an Au⁺¹ oxidation state is observed. (C) N1s line. For np-Au sample, no nitrogen was detected, while for AuCN NW, a cyanide function was observed (BE = 398.5 eV). (D) HR-TEM images of AuCN NW emerging from an np-Au ligament taken in cold-stage configuration (Inset, converged beam electron diffraction of a single AuCN NW). (Scale bar, 20 nm.)

A single nanowire was imaged by high-resolution transmission electron microscopy (HR-TEM) in cold-stage configuration (Fig. 2D) (Methods). The NW was found to be a single crystal of AuCN, as seen in the converged-beam electron diffraction in the Inset of Fig. 2D. Imaging of the AuCN NWs in HR-TEM without utilizing a cold-stage configuration led to distinct changes in morphology within several seconds (SI Appendix, Fig. S2 A–D). Previous reports have described the decomposition of cyanide and the formation of Au nanocrystals as a result of the beam damage (42–44). We observed a similar decomposition effect when AuCN NWs samples were thermally annealed at 220 °C under low vacuum (∼10 Torr) for several hours (SI Appendix, Fig. S3 A–D).

Energy dispersive X-ray (EDX) spectroscopy in the HR-SEM further confirmed that the NWs are indeed composed of AuCN (SI Appendix, Fig. S4A). Time-of-flight secondary ion-mass spectroscopy (TOF-SIMS) (SI Appendix, Fig. S4B) also indicated the presence of NWs within the sample up to a depth of ∼10 μm from the surface.

Gold cyanide complexes are relatively stable and have a wide variety of applications, for example in Au electroplating (45), NH₃ sensors (46), and catalysis of water–gas shift reactions (47). The AuCN NWs synthesized in this work are single crystals, which can be easily transformed to Au nanocrystals by decomposition of the AuCN complex via thermal annealing (SI Appendix, Fig. S3) or ion-beam exposure (SI Appendix, Fig. S2). This also offers a unique possible method of synthesizing a classical np-Au core and a size-controlled Au nanocrystals shell, thereby incorporating a high–surface area Au system for possible catalytic applications.

To gain an understanding of the observed phenomenon and the AuCN NWs formation mechanism, we aimed to correlate the degree of the sample’s plastic deformation and the density of the NWs formed. To this end, we examined samples loaded to different strain levels: 22.5, 55, and 90%. Following selective Ag etching of these samples, we detected NWs on the surfaces of all strained samples, while higher strains resulted in higher NWs densities (SI Appendix, Fig. S5 A–D). In the case of the 22.5% deformation level, only a few NWs were formed, and the np-Au structure predominated (SI Appendix, Fig. S5B). The 55%-strained sample exhibited a much higher relative density of NWs, but the np-Au structure beneath it still prevailed (SI Appendix, Fig. S5C). For the 90%-strain sample, the NWs density was even higher (SI Appendix, Fig. S5D), indicating a direct correlation between the applied deformation and the NWs density.

To further verify that the formation of NWs originates in their dislocations, two Au-Ag alloy samples were cleaned and heated at 1,000 °C (close to the solidus line) (48) for 9 h in a forming gas environment (5% hydrogen and 95% nitrogen). This procedure anneals out preexisting dislocations and allows stress relaxation in the alloy. The samples were then cleaned, one sample was loaded while the other remained unstrained, and both were dealloyed as described in Methods. The prior heating of the sample before loading did not affect the formation of AuCN NWs on the surface of the loaded sample (SI Appendix, Fig. S6B), supporting the assumption that the AuCN NWs are formed as a result of the high dislocation density imposed by the high strains provided immediately before dealloying.

To rule out any possible external contaminations in the equipment or the synthesis methods that might have caused the NWs formation, the following samples were also examined: 1) the sample compressed using a stainless-steel diffusion barrier between the presser and the alloy; 2) the sample compressed using high-grade aluminum foil barrier between the presser and the alloy; 3) the sample compressed in the presser while covering the alloy with nonwoven polyester/cellulose wipe (cleanroom wipes); 4) etching samples with a different batch of 70% nitric acid (trace metals basis; Sigma); 5) Repeating the synthesis method with three different Au-Ag ingot batches that were ordered and produced months apart and by different producers (ACI alloys and Surface Preparation Laboratory [SPL.EU]). All sample treatments resulted in the same formation of AuCN NWs on the surfaces of the samples, which led us to conclude that the NWs did not result from a contamination in the system, further confirming our assumption that the formation of NWs is related to the presence of severe plastic deformation in the system.

To strengthen that assumption, we cleaned two samples and loaded them repeatedly with 10 tons, one sample for 10 cycles and the other for 25 cycles. Both samples underwent thermal treatment for 3 h at 900 °C after loading and were then cleaned and etched as described in Methods. No evidence of AuCN NWs formation was detected on the surface of either sample when imaged by HR-SEM or by EDX spectrometry (SI Appendix, Fig. S7 A–C). The absence of AuCN NWs suggests that a recovery process, which occurs during heating and reduces the dislocation density near the surface, prevents AuCN NW formation. These findings point to a clear relationship between the AuCN NWs formation during Au-Ag dealloying and the existence of dislocations formed due to the preceding severe plastic deformation of the alloy.

As expected, in samples after loading and before etching, the grains perpendicular to the compression direction were found to be elongated, and higher loading of a sample led to higher dislocation densities throughout the alloy (Fig. 3 A–D).

Fig. 3.

Dislocations and NWs after different loadings. (A–D) HR-TEM micrographs of samples after uniaxial compressive loading of (A and B) one cycle of 10 tons, showing slightly elongated grains and high density of dislocations. (C and D) >>25 cycles of 10 tons, showing highly elongated grains and very high density of dislocations. (Scale bars, 50 nm.) (E) HR-SEM micrograph of <111>-oriented single crystal showing AuCN NWs formed in {111} planes (Inset, fast Fourier transform showing threefold symmetry of AuCN NWs growth). (Scale bar, 200 nm.)

Having now confirmed the correlation between dislocations and NWs formation, we then aimed to verify that the NWs indeed grow along dislocation lines. We postulated that if a prestrained single crystal of Au-Ag is cut along the {111} crystallographic plane, the etching will lead to directional NWs growth. Since Au-Ag demonstrates a face-centered cubic unit cell, dislocations slide along four octahedral {111} planes. Dislocation lines of full edge dislocations within a {111} plane demonstrate (a/2)<110> Burgers vectors, which lie along the $<11\overline{2}>$ directions. Hence, they form a twofold structure of lines inclined at 60° to each other. Since several active slip systems operate simultaneously during plastic deformation, we envisaged that in the case of a <111>-oriented single crystal of Au-Ag, the formed AuCN NWs would be oriented in the $<11\overline{2}>$ family of directions, and a threefold symmetry would be expected between the AuCN NWs emerging from the same nucleation junction. Moreover, the NWs would be expected to grow parallel to the {111} surface.

Accordingly, we loaded a <111>-oriented single crystal of Au-Ag (Au₈₄Ag₁₆; refer to Methods for preparation procedure) with one cycle of 70% deformation in the [111] direction, followed by cleaning and then etching as described in Methods. As can be seen in Fig. 3E, both the HR-SEM micrographs and fast Fourier transform showed the expected threefold symmetry within the surface of the crystal {111} plane, further confirming the growth of AuCN NWs along the dislocation lines.

As shown by Landau et al. (49), the microstructure of metals after high plastic deformation reveals patterning of dislocations that are divided into cells. The cells are composed of volumes that are almost free of dislocations and are delineated by dislocation walls. The dislocation cells form cell blocks, and within each block a definite group of slip systems is active. With increasing applied plastic strain, the cell size decreases via a mechanism of cell subdivision. In the case of Au, the dislocation cells are already formed at relatively low plastic strain levels of ∼10% (49). The internal structure of dislocation boundaries consists of dislocation tangles, which become sharper with increasing strain. For higher strain levels of ∼25%, the dislocations within the dislocation boundaries become rearranged to form arrays of parallel dislocations. Evolution of the Au-Ag alloy microstructure during large plastic deformations at room temperature was expected to be similar to that of pure Au. This was verified by our investigation of Au-Ag alloy microstructures after very high plastic deformations (Fig. 3 A–D). It is evident that with increasing strain, the grain size decreases, the dislocation boundaries become sharper, the dislocation walls become denser, and therefore the overall dislocation density increases.

One of the enigmas we encountered relates to the source of the cyanide in the solution. None of our etching solutions contained any cyanide. To establish the carbon source in the AuCN NWs, we cleaned and manually polished three Au-Ag samples to remove any surface contamination. We then loaded the samples to 90% strain, cleaned them again, etched them in a standard HNO₃ solution through which each of the three gases (N₂, CO₂ and air) were separately bubbled, and removed the samples from the etching solution after ∼24 h. There was no evidence of any AuCN formation for the sample prepared in an environment of N₂. However, the samples prepared under air or CO₂ environment indeed displayed AuCN NWs on their surfaces (SI Appendix, Fig. S8 A and B). Moreover, in the case of the CO₂ environment, the formation of NWs was already evident at a much earlier time in the etching process from the emergence of a black hue on the surfaces of the samples. These findings indicate that the absence of CO₂, as in the case of N₂ bubbling, prevented the formation of AuCN NWs, leading us to conclude that the carbon source needed for AuCN formation is probably the CO₂ dissolved in the aqueous HNO₃ solution.

Nucleation and growth of AuCN NWs are assumed to be an outcome of the reaction between activated Au atoms formed during etching and selective dissolution of Ag at the alloy/solution interface, and ${\text{CN}}^{-}$ions:

$${\text{Au}}^{+}+{\text{CN}}^{-}=\text{AuCN},$$ [1]

where the CN⁻ ions probably result from interaction between the nitric and carbonic ions formed from HNO₃ and the dissolved CO₂ in the acidic aqueous solution (47, 50–52), with Au acting as a catalyst:

$$10\cdot {\text{Au}}^0+{\text{CO}}_3^{-2}+{\text{NO}}_3^{-}+12\cdot {\text{H}}^{+}\to 10\cdot {\text{Au}}^{+}+{\text{CN}}^{-}+6\cdot {\text{H}}_2\text{O}.$$ [2]

Formation of AuCN NWs.

Highly activated Au atoms are formed on the entire alloy/solution interface. However, suitable conditions for the nucleation and continuous growth of AuCN NWs can appear only in the region of dislocation outcrops, where the concentration of reactive Au atoms is the highest, and a specific atomic arrangement on the surface can provide an accumulation of AuCN complexes that form a rather large nucleus of AuCN. The dealloying process includes dissolution of the less-noble Ag atoms from low-coordination sites such as step edges. Passivation of these sites with Au leads to surface roughening. Also described in the literature (53–55), a bicontinuous nanoporous structure is formed as the dealloying process proceeds. The dealloying at the dislocation outcrop takes place in a similar manner as that of a perfect surface, however, at a higher rate due to the local increase in the Ag chemical potential. The higher chemical potential mainly stems from the elastic field of the dislocations. This faster dissolution leads to the rapid development of etch pits. Au atoms in the etch pits form steps with very narrow terraces, since the ratio of Vs (the lateral displacement velocity of surface steps) to Vn (the dissolution rate) (i.e., Vs/Vn) is sufficiently small (56). The formed steps also include a high density of geometrical kinks. As a result, the Au atoms at the etch pits are poorly coordinated and thermodynamically unfavorable, which most probably makes them catalytically active in the formation of cyanide.

The next step after cyanide complexes formation is the nucleation of AuCN hexagonal phase, which can further grow as the additional AuCN complexes are delivered to nuclei. The process of selective dissolution continues, and thus AuCN complexes also continue to be formed at the solid/liquid interface near the bottom of the original nucleus (see schematics in Fig. 4). In the vicinity of a dislocation outcrop as well as beneath an AuCN nucleus, the density of low-coordinated Au atoms and vacancies is high. It can be assumed that a new cyanide group formed near the lower border of the nucleus is transferred from one active Au atom to another (or from one step/kink to another) and in this way reaches the most favorable place by attaching to existing edge steps at the growing AuCN bottom interface. This mechanism allows layer-by-layer 1D-growth of AuCN NWs along dislocation lines.

Fig. 4.

Schematic illustration of nucleation and 1D-growth of AuCN NWs along dislocations. Au atoms that are thermodynamically unfavorable and have very low coordination are formed at steps and kinks near the dislocation outcrop and act as catalysts in the chemical reaction of cyanide formation. AuCN complexes form and create AuCN nuclei; new AuCN complexes diffuse to the bottom of the nucleus and attach to edge steps at the growing AuCN bottom interface.

A surface-diffusional flux of Au atoms, driven by the difference in Au chemical potentials between the np-Au layer and the dislocation outcrops, takes place during the dealloying process. This diffusion facilitates the continuous growth of the AuCN NWs (Fig. 5). The diffusion flux enables most of the Au atoms from the np layer to be transferred to the dislocation region, where they are consumed by the layer-by-layer growth of the AuCN NWs. This process occurs rather rapidly even at room temperature, since it is controlled by rapid surface diffusion of Au in aqueous solution containing metal anions (57). Over time, coarsening of the np-Au structure results in a decrease in chemical potential of the Au atoms in the np-Au layer. This leads to a diminishing of the diffusion flux to the AuCN NWs. We found that there is a minimum initial dislocation density above which the rate of AuCN NWs growth is higher than that of the coarsening of the np-Au structure. This critical value of dislocation density can be expressed as follows (also refer to SI Appendix, Supporting Text for detailed calculations):

$$\rho \ge {\rho }_c=\frac{{\gamma }_{\text{Au}}{\text{Ω}}_{\text{Au}}}{B_{\rho }{\overline{R_0}}^3\left({\mu }_{\text{Au}}^{\text{np}}-{\mu }_{\text{Au}}^{\text{CN}}\right)},$$ [3]

where ${\mu }_{\text{Au}}^{\text{np}}$ and ${\mu }_{\text{Au}}^{\text{CN}}$ are the respective chemical potentials of Au atoms in the np-Au layer, and beneath the AuCN NWs, ${\gamma }_{\text{Au}}$ is the Au surface energy, ${\text{Ω}}_{\text{Au}}$ is the atomic volume of Au, $\overline{R_0}$ is the average nanopore radius in the np-Au layer, and B_ρ is a numerical constant ∼10 ($B_{\rho }=14\sqrt{6}/3$ from Eqs. S7 and S9 in SI Appendix). Reasonable values of these parameters in the np-Au CN system are the following: $\overline{R_0}$ = 10 nm, γ_Au = 1.0 J/m², Ω_Au= 0.0168 nm³, ${\mu }_{Au}^{np}-{\mu }_{Au}^{CN}=\left(0.1÷1.0\right)k_{B}T$, and T = 300 K. This yields ρ_c = (3.5 ÷ 35) ⋅ 10¹⁴ ⋅ m⁻², which indeed corresponds to common dislocation densities in Au-Ag alloys after severe plastic deformations.

Fig. 5.

Schematic illustration of AuCN NWs growth along dislocations during dealloying of Au-Ag alloy. (A) Three-dimensional illustration of Au-Ag alloy with dislocations perpendicular to the surface. (B) Illustration of a loaded Au-Ag alloy cross section showing the dislocation cores, dislocation outcrops, and possible ions in the acidic solution. (C) Formation of np-Au layer after selective dissolution of Ag. (D) Nucleation of AuCN phase at the dislocation outcrops, accompanied by formation of grooving cone around the dislocation. (E) Growth of AuCN NWs along the dislocation, governed by surface diffusion of Au atoms from the np-Au near-surface layer. Thickening of the AuCN NWs with elongation is due to the different rates of Ag selective dissolution close to and far from the dislocation cores (decrease of the cone apex angle results in increase of the np-Au layer volume transforming to the AuCN volume).

Conclusions

In summary, in this work we have demonstrated a phenomenon, namely the self-catalytic formation of NWs via reaction of activated Au atoms at dislocation outcrops in the presence of nitric and carbonic ions. The unique reactivity of dislocation outcrops in predeformed alloys was harvested to propagate nucleation and crystal growth of inorganic materials. Using this procedure, we were able to synthesize AuCN NWs via uniaxial loading of the Au-Ag sample and common selective chemical etching. Previous works have presented only epitaxial growth of such NWs on graphene templates (42, 43, 58) or orientation-controlled growth of horizontal NWs of several compositions on different planes of sapphire (59–61). We conclude that the initial nucleation and further crystallization of the inorganic phase is strongly dependent on the dislocation density in the original alloy prior to selective etching. It is governed by Au surface diffusion driven by the difference between the chemical potential of Au atoms in the np-Au structure and in the AuCN phase formed near the dislocation outcrops.

The formation of inorganic crystals via self-propagating reactions along dislocation lines, as shown in this work, may also be achievable in other plastically deformed compounds. The possibility of controlling the growth direction, shape, and morphology of the crystal according to the original alloy slip system, and its crystallographic properties may provide a route to create one- or two-dimensional inorganic materials with tunable properties.

Methods

Sample Preparation and Materials.

Au_0.5Ag_0.5 alloys (50:50% wt at 99.99% pure, ACI Alloys) were cleaned in acetic acid (5 mL, AR Glacial, Gadot Chemical Terminals) for 5 min while stirring, then rinsed with deionized water (5 mL) and then with ethanol (5 mL, absolute dehydrate, BioLab). Ingots were then loaded repeatedly with uniaxial 10-ton force loads for different number (0, 10, 25, and >>25) of loading cycles using a hydraulic press (CrushIR, PIKE Technologies). The initial loading cycle of each sample was ∼950 to 1,000 MPa. From the second loading cycle, each additional loading cycle was equivalent to ∼150 MPa. The initial loading cycle led to 85 to 87% of compression strain to the alloy. Samples that were loaded for 10 cycles exhibited 89 to 91% of compression strain. The Ag was then dealloyed from the samples in an aqueous solution of nitric acid (10 mL, 70%, AR grade, BioLab) for 168 h while stirring in room temperature. The final samples were washed for 5 min in deionized water, then 5 min in ethanol while stirring, and then air dried. Cross-sectional HR-TEM samples were prepared by either FEI Helios NanoLab DualBeam G3 UC or Tescan S9000X dual beam plasma FIB.

Preloading and postloading thermal treatments were conducted in a horizontal tube furnace at 900 – 1,000 °C for 3–9 h under forming gas environment (5% H₂—95% N₂) for stress release. Thermal stability experiments for final AuCN NWs samples were conducted in a Jeio Tech OV-11 vacuum oven at 220 °C under low vacuum (∼10 Torr) for up to 48 h.

An Au₈₄Ag₁₆ single crystal sample was synthesized using a <111>-oriented pure Au single crystal (99.99% pure, SPL.EU). Ag (99.999% pure, METALOR) was evaporated on the single crystal using a physical vapor deposition (PVD-4) thermal evaporator (Vinci Technologies) under high vacuum (10⁻⁶ mbar). Diffusion of Ag into the single crystal was then carried out in a horizontal tube furnace under forming gas (5% H₂—95% N₂) until required composition was achieved.

Sample Characterization.

High-resolution scanning electron microscopy (HR-SEM) and energy-dispersive X-ray spectroscopy (EDX).

Surfaces and cross sections were imaged with a Zeiss Ultra-Plus field emission gun scanning electron microscope. Chemical analysis using EXD was done with X-Max 80 (Oxford Instruments) (accelerating voltage 4 kV and working distance 8.5 mm).

High resolution X-ray diffraction (HR-XRD).

High-resolution X-ray diffraction was carried out in a Rigaku SmartLab 9 kW, using monochromatic radiation of 1.5406 Å.

High resolution transmission electron microscopy (HR-TEM).

Cross-sectional images were obtained using a probe-corrected FEI/ThermoFisher Titan Cubed Themis G2 60-300 operated at an acceleration voltage of 200 KeV. AuCN NWs were imaged using FEI Tecnai G² T20 S-Twin high-resolution transmission electron microscopy (HR-TEM) at 200 KeV. Converged-beam electron diffraction was obtained by cold-stage configuration using liquid nitrogen at −170 °C.

High resolution X-ray photoelectron spectroscopy (HR-XPS).

HR-XPS analysis was performed with a VersaProbe III (PHI) at ultra-high vacuum 2 × 10⁻¹⁰ Torr, where samples were irradiated with a focused X-ray Al-Kα monochromatic X-ray source (1486.6 eV) using an X-ray beam size of 200 µm, 25 W, and 15 kV. HR-XPS survey spectra were recorded for the elements Au4f and N1s at PE = 55 eV, step size = 0.125 eV, and dwell time = 20 ms.

Time-of-flight secondary ion mass spectroscopy (TOF-SIMS).

TOF-SIMS results were obtained using TOF.SIMS 5 of IONTOF Gmb. Bi+ (25 KeV) was used as a primary beam with a 30-μm measurement area, while for sputtering we used a Cs+ (2 KeV) ion beam with a crater size of 150 × 150 μm.

Volume fraction of nanoporous structure.

To measure the volume fraction of the Au in the resulting nanoporous Au structures, cross sections of the samples were produced and polished using FEI Helios NanoLab DualBeam G3 UC FIB. Image processing and analysis to measure the volume fraction was done using ImageJ program.

Data Availability

All study data are included in the article and/or SI Appendix.