Chemical and Laser Ablation Synthesis of Monometallic and Bimetallic Ni-Based Nanoparticles

: The catalytic properties of nanoparticles depend on their size, shape and surface / defect structure, with the entire catalyst performance being governed by the corresponding distributions. Herein, we present two routes of mono- and bimetallic nanoparticle synthesis that enable control of the structural parameters, i.e., wet-chemical synthesis and laser ablation in liquid-phase. The latter is particularly suited to create defect-rich nanoparticles. Impregnation routes were applied to prepare Ni and NiCu nanoparticles, whereas nano- and femtosecond laser ablation in liquid-phase were employed to prepare Ni and NiAu nanoparticles. The e ﬀ ects of the Ni:Cu ratio in impregnation and of laser ﬂuence and liquid-medium on laser ablation are discussed. The atomic structure and (surface) composition of the nanoparticles were characterized by electron microscopic (BF-TEM, DF-TEM, HRTEM) and spectroscopic / di ﬀ raction techniques (EDX, SAED, XPS, IR), complemented by theory (DFT). The chemically synthesized bimetallic NiCu nanoparticles initially had Cu-rich surfaces, which changed to Ni-rich upon reaction. For laser ablation, depending on conditions (ﬂuence, type of liquid), highly defective, ordered, or core / shell-like nanoparticles were produced. The case studies highlight the speciﬁc beneﬁts of each preparation method for catalyst synthesis and discuss the potential of nanoparticles produced by pulsed laser ablation for catalytic applications.


Introduction
Methane dry reforming is a promising direction of utilizing CO 2 , and is thus a topic of intense green energy research [1][2][3]. Although Ni/ZrO 2 catalysts work well, they suffer from coking, and designing a highly active and coke-resistant catalyst is of importance. Previous studies have emphasized using bimetallic nanoparticles (NPs), due to their modified physicochemical properties, which can be quite different from those of the single elements [4][5][6][7][8]. A central idea is to decorate the sites that lead to coking with less active metals such as Cu or Au. Wet-chemical synthesis is frequently used and the bimetallic nanoparticles often undergo structural/compositional changes during activation and reaction, resulting from the different surface energies of the applied metals.
(a,b,d) Adapted with permission from [75]; published by Springer, 2016. (c) Adapted with permission from [3]; published by Elsevier, 2016. Figure 2a,b show electron micrographs of 1:1 NiCu/ZrO 2 after ex situ reduction. The diameter of the ZrO 2 particles was 50-100 nm, whereas the alloy particles were~20 nm in size. In order to confirm that NiCu alloy (rather than separated Ni and Cu) nanoparticles formed upon synthesis, various ex situ and in situ methods were applied. Elemental energy-dispersive X-ray (EDX) fluorescence maps proved that the particles contained both Ni and Cu. Because the reduced catalyst was exposed to air prior to TEM/EDX, the NiCu particles are expected to be partially oxidized, which was confirmed by X-ray diffraction (XRD) (JCPDS card no.   [29]. (a) HRTEM image of a bimetallic NiCu nanoparticle on ZrO 2 . The NiCu particle in the circle was identified EDX. (b) Scanning TEM overview and EDX maps of Ni, Cu, Zr and O of the region marked by the red square. Imaging after ex situ oxidation at 773 K followed by reduction at 673 K (sample exposed to air before imaging). Adapted with permission from [8]; published by The Royal Society of Chemistry, 2015.
In addition to electron microscopy, surface-specific methods, such as Fourier transform infrared (FTIR) and near ambient pressure X-ray photoelectron (NAP-XPS) spectroscopy, were employed to characterize the surface composition. Figure 3a shows room temperature FTIR spectra in up to 5 mbar CO and after evacuation to <2 × 10 −6 mbar, with the sample pre-reduced in 5 mbar H 2 at 400 • C. . IR and NAP-XPS spectra and DFT-derived models of Ni or NiCu nanoparticles supported by ZrO 2 : room temperature IR spectra of (a) reduced Ni/ZrO 2 and (b) various NiCu/ZrO 2 in up to 5 mbar CO (solid lines) and after evacuation (dashed lines). (c) Ni 2p3/2 and (d) Cu 2p3/2 in 0.25 mbar H 2 at 400 • C (E kin of~150 eV; incident photon energies were 1010 and 1100 eV for Ni and Cu, respectively). The lower trace in (c) was acquired at 450 • C in methane. Structure models derived by density functional theory (DFT) are shown in (c,d,e), with (f,g) illustrating Ni segregation to the surface, induced by adsorbed CH x (2435 Cu atoms and 840 Ni atoms, 3275 atoms in total; color code: Ni-blue, Cu-redwood). (a) Adapted with permission from [75]; published by Springer, 2016. (b,e,f,g) Adapted with permission from [29]; published by Elsevier, 2017. (c,d) Adapted with permission from [8]; published by The Royal Society of Chemistry, 2015.
These spectra, together with further CO-FTIR of Cu/ZrO 2 , provide the basis for characterizing bimetallic NiCu nanoparticles. Figure 3b displays room temperature FTIR spectra in 5 mbar CO (highest pressure selected due to weak CO bonding on Cu), with the samples again pre-reduced at 400 • C. Apart from peaks characterizing CO on reduced Ni and Cu, bands of CO on oxidized Ni and oxidized Cu species are observed, pointing to incomplete reduction. On the monometallic (oxidized) Ni/ZrO 2 sample the typical peaks of Ni 0 and Ni + appeared, as discussed before. CO on reduced Cu/ZrO 2 exhibits (apart from the Zr 4+ -CO interaction at 2190 cm −1 [29,[80][81][82][83]) only a single peak around 2111 cm −1 , consisting both of Cu 0 -CO and stronger Cu + -CO contributions. Surprisingly, on all copper containing bimetallics, the typical bands characterizing Ni 0 and Ni + are largely absent, whereas the weaker CO interactions with reduced Cu 0 (below 2110 cm −1 ) [84][85][86][87] and Cu + (2120-2143 cm −1 ) are observed [84][85][86].
Upon alloying Ni and Cu, the Ni 0 -CO peak, located at~2070 cm −1 in 5 mbar CO, red-shifts by 50-70 cm −1 to 2012-2023 cm −1 on the bimetallics (Figure 3b), in line with results of Dalmon et al. [88]. On the bimetallic samples, the CO adsorption on Cu exceeds CO interacting with Ni, suggesting that on the bimetallic surface more Cu than Ni atoms are present. Accordingly, FTIR spectroscopy proved that surface alloy formation occurred, but with Cu surface-enrichmented in Cu, independent of the initial Ni:Cu ratio in the synthesis.
To explain the Cu enrichment, theoretical calculations by density functional theory (DFT) were performed for nanometer sized NiCu NPs, shown as insets in Figure 3c,d, allowing one to determine the optimal chemical ordering (for details see ref [29]). Figure 3c shows the results of DF calculations by Neyman and coworkers for a ca. 1.4 nm Ni 105 Cu 35 particle with truncated octahedral fcc structure. The nanoparticle has 96 atoms at the surface and 44 atoms in the interior. Cu segregation on surface positions was found energetically favourable. The largest energy gain was obtained for Cu segregation to 6-coodinated corners (−413 meV). Lower energy gain was associated with Cu at 7-coodinated edge positions (−300 meV) and 9-coordinated terrace (−193 meV) sites. This clearly shows that Cu is more stable on the surface relative than in the bulk, preferring corner and edge sites to terrace positions ( Figure 3c). When the Cu amount is increased, e.g., in NiCu 1:1, Ni is thus present on the surface solely when the number of Cu atoms in the NP is insufficient to occupy all surface sites, as shown for the minimum-energy structure of Ni 70 Cu 70 (inset of Figure 3d).
The descriptors obtained through the precise electronic structure calculations of NPs with 140 atoms also allowed to predict the chemical ordering in larger NPs with truncated octahedral fcc structure (~6 nm), which are more similar to the chemically synthesized NPs. Such particles have 3275 atoms in total and 1020 surface positions and exhibit preferential Cu segregation. This is shown in Figure 3e for a Ni 2416 Cu 819 (3:1) particle, with a pure Ni interior and Cu occupying 80% of the surface (at 0 K). More precisely, Cu sits at all 24 corner and 156 edge sites, whereas Ni atoms form islands on some of the surface terraces. From 225 to 725 • C, the Ni concentration on the surface nearly doubles (from 20 to 35%), showing that Ni segregation on the surface is facilitated by higher temperature [29]. For Ni 1637 Cu 1638 (1:1) and Ni 819 Cu 2416 (1:3) (Figure 3f) NPs, at 0 K the outer shell is fully covered with Cu, but even at 725 • C the Ni surface concentration was calculated to be only 15% and 5%, respectively. Thus, the calculations corroborate the experimentally determined Cu-rich surfaces of the NPs after synthesis, even when the NP is overall rich in Ni.
When the NiCu nanoparticles were applied for methane reactions up to 500 • C [8,29], NAP-XPS indiciated that Ni segregated to the surface, replacing the Cu surface atoms (lower trace of Figure 3c,f,g). NAP-XPS also detected various carbonaceous species at the particle surface and DFT calculations revealed that the adsorption of CH x groups (x = 0-3) provided the driving force for Ni segregation to the surface. CH 3 provides the lowest and C the highest stabilization, in addition to the temperature increase favouring a Ni-rich surface [29].
Taking all effects into account, an "ideal" coking-resistant NiCu alloy particle would consist of edges and corners from Cu, and terraces of Ni, but the interior must be Cu (Figure 3g), so that segregating Ni cannot replace all Cu surface atoms and thus reverse the intended coking resistance. To meet these criteria a~6 nm large truncated octahedral NiCu nanoparticle should have 840 Ni atoms and 2435 Cu atoms, 3275 atoms in total. After synthesis and pretreatment, the Ni atoms would be mostly located inside the particle ( Figure 3f) and only after reaction-induced segregation Ni atoms are on the surface (Figure 3g). Nevertheless, the Ni atoms would just be sufficient to cover the terrace positions on the nanoparticle surface, while the Cu atoms should occupy 180 edge/corner positions and the 2215 interior sites (Figure 3g).
Such an ideal Ni 840 Cu 2435 particle corresponds to a 1:3 Ni:Cu composition, which indeed has experimentally shown much better coking resistance [8,29]. According to temperature-programmed oxidation (re-oxidizing carbon), the amount of coke was about half and it was less-stable (reoxidized at~100 • C lower temperature) than that formed on Ni/ZrO 2 . However, the better coking resistance of 1:3 Ni:Cu was accompanied by a five-fold reduction in activity. Apparently, if most of the active Ni sites (edges, steps, corners) are replaced by Cu, coking is reduced, but at the expense of activity.

Nanosecond Laser Synthesis of Ni/NiO x Nanoparticles
Laser ablation of Ni targets in liquid environment is an alternative route of nanoparticle synthesis [55,[100][101][102]. Ni nanoparticles are prepared and collected in the liquid, e.g., ethanol, before they are dropcast on carbon-coated Cu grids for TEM analysis (this facilitates imaging, but oxide supports could also be used). As mentioned, the ablation conditions (pulse length, energy and number) are crucial, but also the type of liquid environment may affect the nanoparticle generation. Figure 4 shows the effect of increasing nanosecond laser fluence on the size distribution of Ni NPs synthesized in ethanol by 1000 pulses. Table 1 collects the median and mean sizes and the polydispersity indicators. For the lowest fluence of 309 J cm −2 , there is a maximum at~10 nm with some tailing to larger size (diameter), best represented by the median size of 10.6 nm, rather than the simple average of 12.8 nm. The polydispersity in lognormal distributions can be described by the interquartile range (IQR), that is representative of deviations from median sizes [12] (Table 1), whereas standard errors are presented to show deviations from mean values. Such Ni NPs would be well-suited for catalytic studies. With increasing fluence, the median increases up to 20.6 nm, enabling to tune particle size in studies of size effects. This trend agrees with previous experimental and theoretical studies in the nanosecond and femtosecond regime [24,50,55,72,103,104]. The local vaporization of the target forms a high temperature-pressure plasma forming NPs (see Section 3 below). The plasma creates a supercritical temperature liquid that triggers the formation of a cavitation bubble and subsequently of a low-density metal-fluid mixing region [45]. Higher laser fluences lead to a higher plasma temperature, higher temperature in the bubble and subsequently in increased Brownian motions, which increases the NP collisions and thus particle sizes [105]. Accordingly, increasing laser fluence (F) resulted in larger median sizes and broader size distribution, due to increased probability of particle-particle collisions. In the studied range, there is even a linear relationship between F and the median and IQR values for ns laser-synthesised Ni NPs in ethanol.  To analyze the structure and composition of the NPs, combined bright-field (BF)/dark-field (DF)-TEM, selected area electron diffraction (SAED) and EDX studies were performed, indicating crystalline fcc nickel cores (JCPDS card no. 04-0850) with stacking fault defects, likely surrounded by thin Ni oxide shells ( Figure 5) (JCPDS card no. 78-0643). Figure 5b depicts the corresponding SAED pattern, Figure 5c the related dark-field image (for the objective aperture located on the bright Ni spot on the 5th virtual ring in Figure 5b). The dark-field image of one nickel NP, characteristic of a (222) fcc reflection, clearly reveals that SF defects extend to the particle surface (Figure 5c), where they create highly active sites. It was previously reported that the laser fluence can affect the abundance of crystalline imperfections. For example, ns laser ablation of Ag in water led to the formation of highly-strained Ag nanocatalyst with a high density of stacking faults, which were active for hydrogen generation [71]. The EDX spectrum in Figure 5d confirms the presence of NiO (shells), based on the atomic Ni:O ratio of~2:1 (note the different sensitivity of EDX to Ni and O). Similar as the impregnated NPs discussed above, the NP surface seems dominated by Ni/NiO x ensembles. In LAL, nucleation occurs within tens of microseconds [44]. Studies on cavitation bubble dynamics [5,44,103,[106][107][108] revealed a formation of primary NPs (~10-20 nm) and secondary NPs (~50-60 nm) resulting from rapid solidification of initial NPs in the low-density mixing region and rupture of the superheated molten layer due to Rayleigh-Taylor instabilities [45]. Secondary nanoparticles (larger NPs) can also be formed by agglomeration of small NPs confined inside a bubble [107,109], due to bubble damping forces towards the bottom of the bubble close to the target. Several processes may lead to larger sizes and asymmetric size distributions, as particle formation may be followed by coalescence or Ostwald ripening. However, based on theoretical kinetic studies, coalescence is favoured for the majority of NPs [110]. Ripening processes can occur after the final collapse of the cavitation bubbles [108].
Different fluids may thus affect particle growth differently. The effect of the type of fluid on the size distribution and polydispersity was studied for Ni/NiO x NPs, at constant fluence of 663 J cm −2 and 1000 pulses ( Figure 6 and Table 1) [55]. Compared to the same ablation parameters in ethanol, lower median sizes and IQR values were obtained in isopropanol and butanol, whereas water yielded the broadest distribution. This suggests an effect of laser-induced pyrolysis and photolysis of liquids. During LAL, organic molecules may dissociate on the metallic surface and carbonaceous species can be deposited on the laser-irradiated area. The generated carbonaceous layer increases interfacial absorption, decreasing the ablation threshold fluence and enhancing the incubation effect [56]. Furthermore, agglomeration may be hampered by the formation of a thin graphite shell, acting as capping agent stabilizing smaller colloidal NPs, or of amorphous carbon flakes, inducing steric hindrance [12,72,111].

Femtosecond Laser-Synthesis of Bimetallic (Oxidized) NiAu Nanoparticles
Corresponding ablation experiments were also carried out for a target of Au-coated Ni foil, yielding bimetallic NiAu nanoparticles. Higher ablation efficiency was obtained by fs-pulses at repetition rates ≤ 200 kHz [112], due to a negligible heat affected zone (HAZ) and lower energy losses. Furthermore, the capability of ultrashort pulses for bimetallic nanoparticle synthesis was explored. Figure 7 shows the influence of the fluid nature on the number-weighted size distribution and polydispersity of fs laser-synthesized (oxidized) NiAu. The distribution analysis revealed a trend similar to that of ns ablation in various fluids, with lowest median sizes and IQR values obtained in isopropanol and butanol (Table 2). Ethanol seemed not to provide the growth-limiting carbonaceous species. Furthermore, EDX analyses of the various nanoparticle samples (Figure 7d; Table 2) revealed a Ni/Au ratio of~25 and a Ni/O ratio of~1 in isopropanol and butanol, indicating that Ni was more oxidized in the bimetallic than in the monometallic NPs.  Table 2. The influence of fluid nature on the size distribution and polydispersity of femtosecond laser-synthesized oxidized NiAu nanoparticles [12] (F = 1 J cm −2 , N = 1000,) along with summary of EDX results (atomic%). HRTEM images of the NiAu NPs, fs laser-synthesized in 2-butanol, are shown in Figure 8a,b. Two specific types of nanoparticles were observed (see also Figures 7c and 8d): with and without a graphite shell. The intensity profile displayed as inset in Figure 8b reveals a lattice spacing of 0.333 nm, characteristic of graphite (JCPDS card no. . The FFT inset in Figure 8b shows that the NP is single crystalline. However, regions of Ni and Au in a particle cannot be easily differentiated in the current HRTEM images (displaying mostly lattice resolution in one direction). The lattice spacings of cubic Ni (111)   Based on previous studies [38], it is reasonable to assume that Au-containing particles do not develop graphite layers. Accordingly, Ni-rich and Au-rich NPs can be differentiated by the existence/absence of graphite layers (Figure 8a,b), that can also be discerned from BF-and DF-images (Figure 8d-g). In fact, Ni NPs can serve as an effective catalyst for carbonization, which is why this is a critical issue in methane reforming processes, as coking triggers poisoning, filament growth, reactor clogging and even catalyst disintegration. Laser-induced pyrolysis of organic molecules on the hot metallic surface [113] may add to the graphitization of Ni NPs. Ultrashort laser pulses induce a series of nonlinear effects that contribute to solvent photoionization [114,115]. Previous studies of laser ablation also reported a graphitization of NPs [12,72] and formation of amorphous carbon networks [42,111,[116][117][118][119]. Moreover, direct photolysis of solvents due to interaction with an intense beam [116] and thermal pyrolysis of liquid molecules due to formation of high temperature plasma (~5000-7000 K) [120] may occur. Apparently, alloying with Au atoms significantly reduces these unwanted effects, but this should be better controlled. Whether the absence of graphite layers is due to an Au shell or to low-coordinated Ni surface atoms partially replaced by or coordinated to Au atoms will be subject of further studies. The Ni nanoparticles enclosed in a graphite shell also hold promise for studies of catalysis in confined spaces [121,122]).

Wet-Chemical Synthesis of ZrO 2 Supported (Oxidized) Ni and NiCu Nanoparticles
ZrO 2 was obtained by calcination of Zr(OH) 4 (MEL chemicals XZO 880/01) at 973 K. After calcination, the support has a specific surface area of 36.6 m 2 /g. To prepare supported Ni nanoparticles by impregnation [8,29,75], an appropriate amount of Ni-nitrate to yield 5 wt% loading was dissolved in water, before suspending ZrO 2 powder in the solution and drying at 373 K overnight. The powder was then heated to 723 K in air (heating rate of 5 K/min) and calcined 2 h at 723 K. Before each characterization, the catalyst was oxidized at 773 K and reduced at 673 K (at~1 bar in 20% O 2 in Ar and 20% H 2 in Ar, respectively, with a flow rate 50 mL/min).

Nanosecond Pulsed Laser Ablation of Ni Target in Liquids
A nickel target (Alfa Aesar, Kandel, Germany, 50 × 50 mm 2 ; thickness of~2 mm; purity 99.5%) was placed at the bottom of a glassy container, which was closed with an optical quartz window ( Figure 9). A Q-switched Nd:YAG laser (SPECTRA-PHYSICS, Vienna, Austria, model no. GCR-130, @532 nm, power of~1.2 W, pulse duration of 5 ns, repetition rate of 20 Hz, beam diameter ca. 5 mm) was employed to produce Ni NPs in various liquid media (ethanol, isopropanol, butanol, water; Sigma-Aldrich, Vienna, Austria). A power meter (OPHIR Photonics, Darmstadt, Germany) was located after the polarizer (THORLABS, Newton, NJ, USA) to measure the output power. The focal position in various media were experimentally evaluated by microscopically measuring the ablation area on a silicon target (Zeiss AxioVision software, Vienna, Austria) as a function of the distance of the focusing plano-convex lens (focal length of 92 mm) from the target.

Femtosecond Pulsed Laser Ablation of NiAu Target in Liquids
Nickel foil (50 × 50 mm; 0.5 mm thickness; purity ≥ 99.5%) was obtained from Alfa Aesar, Kandel, Germany. Gold wire (purity ≥ 99.95%) was used for physical vapour deposition (PVD) onto the nickel foil. The Au thickness on Ni was evaluated by time-of-flight secondary ion mass spectrometry IONTOF (ToF-SIMS 5), Münster, Germany and a DektakXT stylus profilometer (BRUKER, Karlsruhe, Germany) indicating~50 nm. Ultrasonication in ethanol was used to clean the surface of the target, which was placed at the back side of the glassy reactor at a height of~15 mm (Figure 9b). Pure alcohols (ethanol, butanol and isopropanol) were provided by Sigma-Aldrich, Vienna, Austria.
Pulsed laser ablation of NiAu target in fluids was carried out by means of a femtosecond titanium-sapphire (Ti:sapphire) laser (@800 nm, 30 fs, 1 kHz). Laser ablation in liquid was done with horizontal beam transfer to reduce optical fluctuations and variations due to formation of optic cavitation. For amplification, a second (commercial) femtosecond laser (SPECTRA-PHYSICS, Vienna, Austria, ≤400 mW, 800 nm, 10 fs, 75 MHz) was used. The entire system is equipped with a motorized XYZ-scanning stage that was controlled by a bCNC program, San Francisco, CA, USA. In order to modify the output power, a variable atten uator including a beam splitter and a Brewster (thin film) polarizer with quartz half waveplate was used. To measure the final output power, a powermeter (OPHIR Photonics, Darmstadt, Germany) was positioned after a variable attenuator. The focus point in various contact phases was experimentally assessed by microscopically measuring the ablated areas on a silicon target as a function of the distance of a focusing parabolic mirror (focal length of 101.6 mm) to the target. Upon ablation, craters were formed on the target (dimensions depending on the conditions) and thus the laser spot was automatically moved to a pristine area after finishing each sequence. The various components of the femtosecond laser device were described in detail before [12].

Impregnated Ni and NiCu Nanoparticles
The size, shape and distribution of the metal particles on the zirconia support were determined by high resolution transmission electron microscopy (HRTEM) [130,131]. TEM measurements of catalysts deposited on carbon-coated Au grids were carried out using an analytical field emission TECNAI F20 (200 kV), Hillsboro, OR, USA, equipped with an energy dispersive X-ray fluorescence detector (EDX).
Synchrotron-based near atmospheric pressure NAP-XPS experiments were done at the ISSIS-PGM beamline at the Helmholtz-Zentrum Berlin [8,29,132]. The in situ setup is based on a differentially pumped electrostatic lens system and a SPECS hemispherical electron analyser. To prevent sample charging, the catalyst 31NiCu/ZrO 2 50 with 50 wt% metal loading was used for the XPS experiments. Measurements at photon energies of 1010 eV for Ni 2p and 1100 eV for Cu 2p yielded a photoelectron kinetic energy of 150 eV, leading to the same probing depth in both cases.
Transmission IR spectra were acquired on a Vertex 70 spectrometer (Bruker, Vienna, Austria), equipped with a mercury cadmium telluride (MTC) detector [29,75,129]. Catalysts were pressed to small discs and placed in the IR cell. Oxidation was carried out in 100 mbar O 2 by heating to 500 • C (10 K min −1 ) for 1 h. For reduction, the IR cell was evacuated, the catalyst cooled to 300 • C and heated in 5 mbar H 2 and 900 mbar N 2 to 400 • C (30 min), followed by evacuation. All infrared spectra had a resolution of 4 cm −1 in the 4000-900 cm −1 range, averaging 128 scans to obtain good signal to noise ratio. CO (5 mbar at 30 • C) was used as probe molecule to detect nanoparticle surface composition and oxidation state.
Electronic structure calculations of NiCu alloy nanoparticles of various composition were carried out using the periodic plane-wave code VASP [133] and the PBE [134] exchange-correlation functional. For details we refer to [75,135].

Laser-Synthesized (Oxidized) Ni and NiAu Nanoparticles
Ni/NiO x and bimetallic (oxidized) NiAu nanoparticles were characterized by transmission electron microscopy (TEM). Samples for TEM imaging were prepared by drop casting of the colloidal dispersion on a lacey carbon film (copper grid) after 5 min ultrasonication, before TEM grids were placed in a vacuum desiccator and dried at room temperature. Bright-field (BF-TEM) and dark-field (DF-TEM) images (particularly detecting defects [136][137][138]), along with selected area electron diffraction (SAED) patterns, were taken by a Philips CM200 TEM, Hillsboro, OR, USA (LaB 6 cathode, acceleration voltage of 200 kV) equipped with a Gatan Orius CCD camera, Pleasanton, CA, USA. PASAD-tools along with Gatan digital micrograph software were applied to evaluate reciprocal lattice distances and acquire crystallographic data [68]. The measured lattice distances were compared to those in a database [139]. Number-weighted size distributions of nanoparticles were obtained by evaluation of at least seven TEM frames (700 × 700 nm; ca. 500 particles) using Gatan microscopy software. Elemental composition studies were performed by energy-dispersive X-ray spectroscopy (EDX) in the TEM. Furthermore, high resolution transmission electron microscopy (HRTEM) was performed with a FEI Titan 80-300, Hillsboro, OR, USA equipped with an imaging Cs-corrector.

Conclusions
Ni, NiCu and NiAu (oxidized or graphite-coated) nanoparticles were prepared by wet-chemical impregnation or laser ablation in liquid media (LAL). Electron microscopy and diffraction, X-ray fluorescence and photoemission, as well as infrared and density functional theory were used to characterize the (surface) structure and composition of the synthesized nanoparticles. The chemically synthesized bimetallic NiCu nanoparticles initially had Cu-rich surfaces, independent of the original Ni:Cu ratio, but Ni segregated to the surface upon heating and especially reaction. This effect must thus be accounted for in nanoparticle synthesis to reach the desired surface composition under working conditions. Laser ablation allows to prepare nanoparticles with controlled size (median) and a high density of bulk defects, that propagate to the surface, which should be relevant for catalytic applications. The parameters of LAL (pulse length, fluence and type of liquid) can be used to control the resulting particle size distributions and surface structures/compositions. In case of NiAu nanoparticles, either clean (presumably Au-rich) or graphite-covered (presumably Ni-rich) surfaces were obtained. The latter samples may be particularly interesting for further studies of intercalated molecules or catalytic reactions in confined spaces, i.e., in between the metallic surface and the graphite overlayer. Clearly, the properties of LAL-synthesized nanoparticles must be further explored by additional adsorption and catalytic reaction studies, similar to those of the impregnated catalysts, but their defect structure holds promise for beneficial properties.
Author Contributions: All authors listed have made a substantial, direct and intellectual contribution to the work, and approved it for publication. All authors have read and agreed to the published version of the manuscript.