Symposium NM01 : Nanomaterials and Devices by Cluster Beam Deposition

If we are to translate the exquisite physics of atomic clusters into the synthesis of a new class of functional nanomaterials, massive scale-up of the rate of cluster depositon is required. The prize is a set of applications ranging from theranostics and water treatment to catalysis and memristors. Compared with colloidal synthesis of nanoparticles, the cluster beam approach is green - it involves no solvents and no effluents; particles can be size-selected as desired; and challenging combinations of metals (nanoalloys) can readily be produced. Yet to date the cluster approach has been held back by extremely low rates of particle production, largely confined to the sub-microgram per hour range. As one example, industrial catalysis R&D typically requires a gram of catalyst, or 10 mg of clusters at 1% loading on a suitable catalyst support. The cluster beam community is now rising to this scale-up challenge. I will discuss the development of a new kind of super-intense nanoparticle beam source, the “Matrix Assembly Cluster Source” (MACS). The MACS is based on ion beam sputtering of a solid rare gas matrix into which metal atoms are pre-loaded by evaporation. A scale-up of five orders of magnitude in cluster beam intensity has been achieved. I will address the mechanism of cluster formation in the matrix, highlighting the roles of multiple ion impact events in stimulating the ripening of the clusters, as well as the size-dependence of the cluster escape from the matrix to form the beam. However, the generation of intense cluster beams is not sufficient for the production of functional cluster-based materials. There is also a processing challenge to address. I will discuss a number of routes by which size-controlled clusters may be presented in a form matching the desired functional application, with specific reference to catalysis and theranostics. These examples of "formulation engineering on the nanoscale" include direct deposition of metal cluster beams onto micron-scale oxide powder support particles; bottom-up synthesis of novel cluster-decorated carbon powders; and functionalisation of deposited clusters to achieve biological compatibility. In all cases the synthetic work is supported by aberration-corrected Scanning Transmission Electron Microscopy (STEM) imaging. FInally I will discuss the validation challenge that we need to meet if we want to show that cluster-based functional materials are competitive with, or ideally superior to, advanced materials created by more traditional routes. I will focus on the hydrogenation (in both gas and liquid phases) of organic molecules over cluster catalysts as a basis for applications in the fine chemicals and pharmaceuticals sectors. Comparison with reference materials synthesised by standard methods begins to highlight the types of reaction in which cluster materials show a notable competitive edge. It may be that these three challenges will define the future of the field of cluster science.

Recent advances in understanding of aerosol formation and growth through discrete element modeling and molecular dynamics allow now optimal reactor design, away from the Edisonian approaches of the past. This leads to scalable synthesis of sophisticated nanoparticles with controlled composition, size and morphology resulting in a few high value products (nanosilver and carbon-coated Co nanoparticles) in the market already, while several promising ones are emerging such as single atom catalysts and breath sensors.
This lecture will highlight the interplay of surface growth and coagulation revealing the evolution of nascent to mature soot formation by mesoscale simulations, in excellent agreement with experimental data in premixed and diffusion flames. Such an understanding facilitates to elucidate the effects of primary particle polydispersity and chemical bonding (aggregation) on soot morphology and optical properties using the Discrete Dipole Approximation. This is of critical importance in monitoring soot emissions, carbon black manufacture, climate modelling and operation of fire detectors. Agglomerates of polydisperse aggregates and spheres produced by agglomeration and surface growth have more compact structure and larger fractal dimension, D_f, than agglomerates of monodisperse spheres. Primary particle polydispersity and aggregation enhance soot scattering (of critical importance to fire detectors) up to 50 and 30 %, respectively!
Quantitative understanding of the formation of bimetallic nanoparticles is important as they exhibit catalytic, optical, electronic and magnetic synergy between their constituent metals. Typically, that synergy is traced to the domain structure and surface characteristics of such particles. Here these characteristics of coalescing Ag-Au nanoparticles of various initial sizes, morphologies (segregated or alloys) are investigated by atomistic molecular dynamics (MD). Silver atoms exhibit increased mobility over Au and occupy gradually the surface of the coalesced (or sintered) bimetallic particle, consistent with scanning electron microscopy and selective O₂ chemisorption experiments for heterogeneous catalysis of ethylene oxidation. The characteristic sintering time of equally-sized Ag-Au nanoparticles is similar to that of pure Au but shorter than that of Ag nanoparticles. When the latter coalesce with substantially bigger Au ones, a patchy Ag layer is formed at the Au particle surface. However, when Ag are bigger, then Au is rather embedded into Ag consistent with microscopy data. Most notably, X-ray diffraction patterns of Ag-Au nanoparticles are obtained, for the first time to our knowledge by MD distinguishing segregated from alloyed ones. The latter form smaller crystal size (highly polycrystalline) than coalescing pure and segregated Ag & Au nanoparticles, quantitatively explaining the structure of flame-made Ag-Au nanoparticles for biomaterial applications.

Since the seminal work of Haberland’s group [1], magnetron based gas aggregation sources (GAS) have been used in an increasing number of nanoparticles studies. As compared to other GAS, the magnetron based GAS have gained popularity probably thanks to their relative ease to use, and to the fact that the majority of the generated nanoparticles are charged which allows their manipulation (filtering and deflection) [2].

In this talk we will address issues not fully investigated nor understood of the magnetron based GAS. This includes the control of the race-track that is formed in circular magnetron sputtering through the use of a Full Face Erosion (FFE) magnetron that has been developed for the first time in GAS. We will show how such design allows a more stable generation of nanoparticles over extended periods of time. We will also address the use of High-Power Impulse Magnetron Sputtering (HiPIMS) in GAS, focusing on its use for the generation of alloyed nanoparticles. In particular we will show for the first time that HiPIMS allows the generation of a variety of Co₅₀Au₅₀ nanoparticles not accessible in Direct Current GAS. Finally, we will address the assisted generation of nanoparticles by controlled injection of gases in GAS. Although already know, the injection of gases in the GAS that can have important effects on the generation of nanoparticles is not fully understood. Here we will present new results to illustrate the drastic impact of gas injection on the generation of Au nanoparticles in GAS.
References

[1] H. Haberland, M. Karrais, M. Mall, Zeitschrift für Physik D Atoms, Molecules and Clusters 1991, 413-415; H. Haberland, M. Karrais, M. Mall, Y. Thurner, Journal of Vacuum Science and Technology A: Vacuum, Surfaces, and Films 1992, 10, 3266-3271.
[2] Gas-Phase Synthesis of Nanoparticles, Wiley, 2017, Ed. Y. Huttel.

Combustion engines are still required in power-generation and automobile industry for more decades even in switching to electrification. For the sustainable development, lean combustion at larger air-to-fuel ratios suppresses fuel consumption due to higher heat efficiency. Its exhaust gas is cooler accordingly, so that catalytic conversion must be accomplished at lower temperatures. We are tackling this crucial issue with new materials, i.e. Pt cluster disks chemically bound to a Si semiconductor substrate, Pt_N/Si (N=5-60) [1], where electrons are accumulated at their sub-nano interface [2,3]. In this talk, their prominent low-temperature catalytic activity is unveiled on a basis of surface-chemistry measurements.

We have found that CO oxidation by atomic O species on the Pt₃₀/Si disks starts at 130 K [4], which is lower by 150 K than that on the Pt(111) single-crystal surface [5]. Furthermore, NO reduction proceeds at lower by 100 K than on supported Pt nano-particles. This particularity was observed also in turnover rates (TOR) under a steady-state condition of continuous supply of CO and O₂ [6]. Furthermore, hysteresis in the TOR was discernible in the heating and subsequent cooling periods due to the bistability switching between O- and CO-rich regimes on Pt_N/Si, while no hysteresis for Pd nano-particles on an MgO substrate [7]. Considering a report that a smaller CO-oxidation rate than the dissociative adsorption rates of O₂ damps the hysteresis of nano-reactors [8], the atomic O species produced by Pt_N/Si are highly reactive to reduce the reaction temperature as well as to maintain the hysteresis even in sub-nano ranges. This is also true in the NO reduction, in which reactive atomic N species recombine into N₂ at lower temperatures. It is probable that the low-temperature and highly-efficient catalytic activities of Pt_N/Si derive from their electron accumulation.


References
[1] H. Yasumatsu, Encyclopedia of Interfacial Chemistry: Surface Science and Electrochemistry, ed. Klaus Wandelt, accepted (2017).
[2] H. Yasumatsu, T. Hayakawa and T. Kondow, Chem. Phys. Lett. 487, 279 (2010).
[3] H. Yasumatsu, P. Murugan and Y. Kawazoe, Phys. Stat. Solidi B, 6, 1193 (2012).
[4] H. Yasumatsu and N. Fukui, J. Phys. Chem. C 119, 11217 (2015).
[5] J. Yoshinobu and M. Kawai, J. Chem. Phys. 103, 3220 (1995).
[6] H. Yasumatsu and N. Fukui, Catal. Sci. Technol. 6, 6910 (2016).
[7] V. Johánek, M. Laurin, A.W. Grant, B. Kasemo, C.R. Henry and J. Libuda, Science 304, 1639 (2004).
[8] V.P Zhdanov and B. Kasemo, Surf. Sci. 496, 251 (2002).

Deposition of size-selected clusters in ultrahigh vacuum is used to create catalysts and electrocatalysts with well defined initial catalytic site size. Spectroscopy in UHV is used to characterize the catalysts, and then reactions are studied using either gas-surface interaction methods in UHV, or by electrochemical methods. For gas-surface catalysis, the focus will be on methods for post-deposition modification of clusters, allowing preparation of size-selected bimetallic clusters. Alloying of Pt clusters by boron and tin, and the effects on alkene dehydrogenation will be discussed. Electrochemistry can be studied either in situ in the vacuum system, thus avoiding issues of contamination and electrochemical cleaning, or ex situ, allowing studies using more elaborate electrochemical techniques. For example, Scanning ElectroChemical Cell Microscopy (SECCM) is used to study reactions on electrode areas down to a few hundred nanometers in diameter, and at cluster coverages where processes such as Ostwald ripening or cluster diffusion/agglomeration are negligible. We are, thus, able to study fundamental processes such as proton reduction on well defined catalytic sites down to single atoms.

Nanowire based field effect transistors (FETs) have been extensively studied for their applications in biosensing. Due to the large surface to volume ratio of these structures, even a single biomolecule attached to the nanowire surface is able to change the electronic properties of the FET. Metal oxide nanowires such as CuO and ZnO are biocompatible, nontoxic, and stable when grown via thermal oxidation methods. However, the surface of the nanowire must be properly functionalized in order to fabricate a highly specific nanowire biosensor. We propose instead to use gold nanoparticles (GNPs) as the interface material between the nanowire and biomolecules of interest. In this study, we compare the performance between FETs with plain metal oxide nanowires and GNP-decorated nanowires fabricated using a cluster beam deposition (CBD) process via a magnetron sputtering system. These devices are used as DNA sensors. The CBD process allows us to exert size control over the GNPs to isolate and sense single DNA strands attached to the nanowire. We also show that GNP decoration increases the surface area to volume ratio of the nanowire and therefore the sensitivity of the device. This CBD process is integrated into a CMOS compatible procedure to fabricate arrays of nanowire DNA sensors.

Light interaction with metal nanoparticles (NPs) gives rise to a number of fascinating optical phenomena. One of the directions of research is the search for configurations that enable efficiently interconvert propagating and localised plasmon resonances and thereby promote the enhanced local fields. Therefore, there is a strong interest to the formation of nanoparticle arrays for practical applications in plasmonics, surface enhanced Raman spectroscopy and optical sensing.
To form arrays, cluster beam deposition is combined with electron beam lithography. The substrates are covered by thin (100 nm) layer of photoresist followed by thin poly(methyl methacrylate) PMMA layer using standard spin-coating. In the first series of experiments simple patterns representing 100 nm wide stripes with varying distance between the stripes (period) are written in PMMA by electron beam, thus, forming periodic structures of linear trenches down to the photoresist layer. Size-selected silver NPs (of about 18 nm in diameter in the current experiments) produced by the magnetron sputtering cluster apparatus are deposited in low-energy regime on the patterns reaching the coverage of a monolayer of NPs across the surface. After the deposition the samples are annealed at 180 C for 5 minutes to facilitate partial immersion of the deposited NPs into the polymer films. This treatment improves adhesion making the NPs resistant against wet chemical procedures. After that a standard lift-off is performed to remove the PMMA yielding sharp linear stripes of metal NPs slightly embedded into the photoresist layer. The samples demonstrate strong localised surface plasmon resonance bands at wavelength of ca. 400 nm which is good agreement with the simulations using Mie theory. Properties of propagating plasmon resonances of the stripes formed by individual NPs as well as possible optical coupling between the stripes are currently under the study.
In summary, by combining the electron beam lithography and cluster beam deposition techniques linear arrays (nanoscale stripes) of size-selected metal NPs are successfully obtained. The approach paves a way for formation of a variety of different geometrical arrays with the limitations in dimensions dictated only by lithography, thus, providing an excellent method for production of plasmonic systems with required configurations.

Metal-oxide (MOx) nanowire (NW) based gas sensors have proven to be highly sensitive to different gases due to their high surface-to-volume ratio. This sensitivity restricts the MOx capacity to distinguish between different gases. To overcome this shortcoming, noble metal nanoparticles (NPs) deposited on MOx NWs increases the sensitivity and selectivity of the gas sensor [1]. In previous work [2], single CuO NWs were functionalized by 5nm Pd NPs to increase carbon monoxide (CO) sensing. While the size effect of metallic NPs on gas sensing performance has already been intensively investigated in literature [1], there are only few reports on the metallic NP coverage effect, especially on CuO NW based gas sensors. Thus, in this work, CuO NW devices were decorated with different amounts of size-selected Pd NPs using a cluster beam deposition (CBD) method based on magnetron sputtering processes. This allows us to study the coverage effect on the sensing performance towards different gases. Also, this study provides an insight for the optimum coverage needed to functionalize CuO NWs and enhance their sensitivity and selectivity.
[1] Franke M. E. et al. Small. 2, 36-50 (2006).
[2] Steinhauer S. et al. Nanotechnology 26, 175502 (2015).

Metal nanomaterials have a variety of useful chemical, optical, catalytic, and electrical properties. Among them, silver nanomaterials are widely used today for printed electronics due to their high conductivity. However, silver is a relatively expensive material. To reduce costs, alloying silver with other conductive metals is beneficial. In this work, we have synthesized different multicomponent metal nanopowders and nano-inks including copper-silver and copper-nickel using the High Temperature Reducing Jet (HTRJ) process. The (HTRJ) reactor has been developed in our group to enable continuous one-step gas-phase (aerosol) synthesis of alloy metal nanoparticles from metal salt precursors. In this process, a fuel-rich hydrogen flame passes through a converging-diverging nozzle. An aqueous metal precursor solution injected at the throat section of the nozzle is atomized by the high velocity gas stream. The resulting droplets evaporate and the precursor decomposes, initiating nucleation of particles in a reducing environment containing excess H₂. After the reaction zone, particles are cooled immediately to prevent further particle growth and coalescence. Recently, we have expanded the capabilities of this system to enable in situ functionalization of the synthesized bare nanopowders with organic ligands. In the first demonstration of this approach, we atomize short-chain amine molecules that bind to metal nanoparticles and deliver them at the reactor exit, where they evaporate. This initiates the gas-surface reaction between the ligands and nanoparticles. The resulting nanopowders are readily dispersed in organic non-polar solvents to make stable, uniform nano-inks. These stable nano-inks can be used as conductive inks for printed electronics applications or in many other applications where a stable metal nanoparticle dispersion is needed for low-cost processing.

Gas aggregation cluster source (GAS) based on planar magnetron is very popular system for synthesis of various nanoparticles (NPs). Such source was used for production of NPs from different materials ranging from metals or metal oxides to organic compounds. In recent years, the focus shifts to the preparation of heterogeneous NPs. In this study, we report on the two strategies for the production of heterogeneous multicomponent NPs based on the GAS systems. The first is based on the utilization of the composite magnetron target of the GAS magnetron. The second is based on in-flight deposition of the shell on the already formed NPs after they left the GAS.
The composite target was made of copper sheet 3 mm thick and 3 inch in diameter with the tungsten pellets placed in the erosion zone. The sputtering took place in the water cooled aggregation chamber in argon atmosphere at pressures from 30 to 210 Pa and magnetron current up to 500 mA. Aggregation chamber was ended with conical orifice 2 mm in diameter facing on substrate placed in the distance 20 cm in the deposition chamber with the background pressure below 1 Pa. The amount of tungsten in the NPs was predetermined by the amount of pellets and tuned by plasma parameters. Surface chemistry was determined by XPS and NPs morphology by SEM and HRTEM. Very good segregation of the W core and Cu shell was observed by STEM that proved core@shell structure of produced NPs.
In the second case Ni NPs prepared by GAS with Ni magnetron target 1.5 mm thick were coated by copper by means of tubular magnetron (TMG). The magnetron was connected directly to the GAS and was ended by a nozzle similarly to the GAS. The diameter and length of the nozzle was adjusted in such a way that the pressure inside the TMG was about one half of the pressure inside the GAS. Due to that Ni NPs leaving the GAS were decelerated to the drift velocity of the gas and their residence time was increased. It was found that below 150 mA DC current the NPs can easily flight through the TMG but the Cu coverage of Ni NP is negligible. Self-pulsing of the deposition rate and voltage was observed between 150 and 250 mA. Above 250 mA no NPs passed through the TMG. However, if the magnetron current was pulsed it was possible to deposit NP also above this limit. It was shown that change of the pulse duty cycle leads to the change of Cu/Ni ratio and the size of fabricated NPs. Due to high miscibility of Ni and Cu core-shell structure was not observed in this case and produced NPs were alloy type with higher Cu content on the surface.

Acknowledgement:
This work was supported by grant GACR 17-22016S from the Grant Agency of the Czech Republic.

Theoretical prediction of the structure of nanoparticle (NP) based thin films via atomistic simulations is challenging, due to computational limitations. To overcome these limitations, we are developing a meso-scale technique that integrates atomistic and microscopic scale methods. Herein, as a first step, we propose an approach that encompasses the random nature of NP deposition, which results in a statistical cancellation of individual sintering mechanisms between coalescing NPs upon deposition. As a result, we present a simple and intuitive analytical method that describes the average coalescence behavior of NPs, regardless of constituent element or crystallinity, emphasizing only the predominant coalescence dependencies on temperature and size-dependent NP melting points.^[1] When two NPs touch each other (i.e. impact in flight, soft-land on one another or surface-diffuse on a substrate) a part of them melts due to free surface annihilation, forming a region that, when re-solidified, binds the two NPs together. The exact configuration may depend on many parameters (such as relative orientation, defect content, constituents of the nanoparticles, etc.) and can be quite complex, but the thesis of our proposed model is that it mainly depends on the melting temperature of the NP: the lower it is, the bigger the portion of the NP that melts. In other words, if the NP melting temperature as a function of size is known, no other physical properties are needed in order to describe the sintering, and thus one can circumvent the complex physics involved and simulate the NP deposition on a much coarser scale. Typically, NP melting temperatures follow a T_m (1-constant/diameter) dependence with size, approaching asymptotically the bulk melting temperature above a certain size threshold.
We assess our model using molecular dynamics (MD) computer simulations of dissimilar systems, and find good agreement between its predictions and the MD results. Its simplicity makes our model a suitable starting point for the development of a meso-scale simulation technique that can describe the growth of nanoparticulated films and the prediction of their properties such as structure, porosity, and percolation threshold.

[1] “Simple analytical model of nanocluster coalescence for porous thin film design”
P. Grammatikopoulos, E. Toulkeridou, K. Nordlund, M. Sowwan
Modelling Simul Mater Sci Eng 23 (2015) 015008.

With the most powerful tool of a software containing a powerful command-line interface, any nanoparticle is possible. The software is capable of creating atomistic models of nanostructures by receiving common input files of pristine crystal structures. By using the command line the user can manipulate and create any possible nanoparticle which afterwards can be exported in any format. In order to implement a fast and easy to control computational tool, C++, Win32API, Assembly and OpenGL are used, providing the freedom to develop and design a user oriented software package dedicated to nanoparticles.
The software has the most common nanoparticle forms ready to be used with adjustable planes with the press of a button. An example of that is a cubic structure with all {100}, {110}, {111} families. Every family plane can easily be translated parallel to the <100>, <110>, <111> directions, each family individually, by pressing the proper buttons. This way any possible combination of those planes can be formed. Other shapes, like the rod-like, a typical sphere and any type of nanowires with adjustable parameters are possible.
The command line is capable of creating all of the above options with countless more possibilities. Obviously, the aim of creating those nanoparticle forms, described above, is to make them easier to use considering they are the most common ones. The command line is fully programmable, meaning we can create and add any user-requested command in a form of update. The list of commands includes among others:
●Plane(int h, int k, int l, int A)
Cuts the crystal along the (hkl) plane with “A” (Angstrom) distance from the center.
●Vector(int h, int k, int l, int A)
Cuts the crystal along the plane which is perpendicular to the [hkl] vector with “A” distance from the center.
●Super Cell(int a, int b, int c)
Multiplies the content of the input file a,b,c times along the tree primitive vectors respectively.
Some examples that we have created are: cuboctahedron¹, rhombic dodecahedron², rod-like³, cube⁴, nanowires⁵ as well as a variety of very specific shapes⁶.
1. H. Hofmeister, G.L. Tan, M. Dubiel, J. Mater. Res., 20(6), 1551 (2005)
2. E. Auyeung, T. I. N. G. Li, A. J. Senesi, A. L. Schmucker, B. C. Pals, M. Olvera de la Cruz and C. A. Mirkin Nature 505,73 (2014)
3. S. J. Kim, P. Lei, K. Zhang, C. Zhou, G. W. Graham, and X. Pan, Chemistry of Materials 29 (5), 2016, (2017)
4. J. Zhao, E. Baibuz, J. Vernieres, P. Grammatikopoulos, V. Jansson, M. Nagel, S. Steinhauer, M. Sowwan, A. Kuronen, K. Nordlund, and F. Djurabekova, ACS Nano, 10(4), 4684(2016)
5. J. Kioseoglou, T. Pavloudis, T. Kehagias, P. Komninou, T. Karakostas, C.D. Latham, M.J. Rayson, P.R. Briddon, M. Eickhoff, Journal of Applied Physics, 118(3), 034301 (2015)
6. P. Grammatikopoulos, C. Cassidy, V. Singh and M. Sowwan Scientific Reports 4, 5779 (2014)

Metal hydrides are a promising material for hydrogen storage, given their relatively low cost, abundance and high weight percent hydrogen absorption (e.g. 7.6 wt. % for magnesium hydride). However, to date, commercial application of metal hydrides has been limited by the relatively slow absorption and desorption kinetics, requiring high pressure and high temperature, respectively. A first approach for enhancing the properties of metal hydrides is nanostructuring, with the associated increase in surface area and reactivity that this inherently creates. Pd nanoparticles on Mg nanofilms may increase the efficiency of bulk Mg and offer an attractive alternative to bulk Pd for catalysis and hydrogen storage.
In this work, the diffusion of H in Mg nanofilms decorated with Pd nanoparticles is investigated by DFT using the VASP ab initio simulation package. Models of Pd nanoparticle/Mg film interfaces and their variations, i.e. Pd(H)/Mg(O)Pd and PdH/MgH2, are examined and the specific reconstructions of the nanoparticle/nanofilm interface arising from the strain are deduced. The diffusion path of a H atoms starting from the nanoparticle region and ending deep in the nanofilm region is investigated and the respective diffusion barriers are extracted. The Pd nanoparticle is found to facilitate the diffusion of H in the first layers of Mg. This barrier is even smaller for PdH, suggesting that the diffusion is dependent on the H content of PdHx. Finally, H diffusion is prohibited in MgO for both Pd and PdH, suggesting that the nanofilm surfaces should not be oxidized before the nanoparticle deposition, and significantly suppressed through previously formed MgH2 domains.

Photoluminescence spectroscopy (PL), X-Ray Diffraction (XRD) and Scanning Electron Microscopy (SEM) were used for the comparative characterization of ZnO and ZnO Er nanocrystals (NCs) obtained by spray pyrolysis method. High temperature annealing at 400 C for 2 hours in nitrogen flow follows the spray pyrolysis process. SEM study has confirmed the nanocrystal (NC) structure of obtained ZnO and ZnO Er films. X-Ray diffraction has revealed the wurtzite crystal structure in ZnO and ZnO Er NCs. The XRD intensity of all XRD peaks monotonically decreases with Er concentrations. PL spectra demonstrate the near band edge (NBE) and defect related PL bands in visible (2.0-3.0 eV) and infrared (0.80-0.85eV) spectral ranges: The PL spectra were studied in dependence on Er atom concentrations and at different temperatures within the range 10-300K for the identification the optical transition responsible for the emission bands. The applications of this semiconductor nanomaterial could be very useful as high temperature light emitting diodes

Single glazed windows are a significant source of energy loss, resulting in $20 billion/year (in the US) of unnecessary energy expenditure¹_, and for a variety of reasons it is often not possible to simple replace these windows with multi-glazing alternatives. This work is a demonstration of the development of a flame pyrolysis tool, integrated with a low vacuum deposition chamber, for the purpose of synthesizing and depositing silica nanoparticles (NPs) in a one-step-process in order to fabricate thermally insulating and transparent thin-films. A primary focus for this technology would be as insulating layers on windows. Sufficient thermal insulation can be achieved by depositing a high porosity silica coating (analogous to an aerogel) on a standard window. To use these coatings for thermal insulation on windows it is a requirement that the coating be sufficiently thick so as to make a meaningful contribution to the thermal insulation of the window, relative to a standard single glazed window. Therefore, in order to maintain the mechanical strength of the coating it is necessary to include binding material for cross-linking. At this early stage of the project coatings of greater than 160 μm have been demonstrated, and initial thermal conductivity measurements are taking place.
In addition to thermal insulation, the optical properties of these coatings must be of high enough quality to maintain sufficient transmission in the visible region of the electromagnetic spectrum, so as to be considered practical for a window coating. This means that properties such as optical haze, reflection and transmission need to be carefully maintained. It is possible to control these optical properties by carefully controlling NP size and size distribution (allowing for even pore size distribution), varying the porosity of the NP coatings as the distance to material interfaces closes (thus preventing dramatic shifts in refractive index) and lastly by controlling the surface roughness to prevent significant scattering losses from the coatings surface. Our early samples currently show high haze and have highly diffuse transmittance in the visible spectrum, suggesting more needs to be done to control pore size uniformity. Pore size distribution and surface roughness are controllable by limiting the NP size, which in turn is controllable by varying the residence time of the particles within the synthesizing flame after the initial particle nucleation. Furthermore, by controlling the pressure differential between the upper and lower chambers of the deposition system, it is possible to vary the speed at which the silica NPs impact a substrate (mounted inside the deposition chamber), and thus, vary the porosity and refractive index of the NP coating.

References
1) Consumer prices for natural gas are compiled by the Bureau of Labor Statistics. For the current summary for major cities, see http://www.bls.gov/regions/midwest/data/AverageEnergyPrices_SelectedAreas_Table.htm .

Gas sensors based on metal oxide (MOx) nanowires are a promising area of research, as their high aspect ratio allow a high degree of sensitivity, while synthesis via thermal oxidation is possible at temperatures which are compatible with silicon technology[1]. However, these sensors have issues with cross sensitivity, reacting to different gases in similar ways. One solution to differentiate the response is to develop an array of nanowires each functionalised with different nanoparticles[2].

Nanoparticles on MOx supports have been used as inorganic heterogeneous catalysts across a variety of gas-based applications for decades including chemical synthesis, industrial production processes and removing pollutants from car exhaust. Nanoclusters synthesised through cluster beam deposition techniques have been emerging over the past two decades as a new tool in undertanding of nanoparticle catalysts. This technique is advantageous for silicon technology as the nanoparticles produced are free from surfactants, solvents and ligands.

This presentation will detail the decoration of CuO nanowire sensing elements using physically deposited gas condensed clusters of catalytic nanoparticles, and the systems subsequent performance as a sensor to CO and acetone. Further work using aberration corrected HDAAF S/TEM and XPS to elcuidate the changes in the nanowire/nanocluster system with increasing temperature will also be shown.

Results from our work demonstrate that CuO nanowires functionalised with nanoclusters including Pt and Ru do exhibit changed sensor behaviour in response to test gases. In addition, as this synthesis method is free from additional ligands and other stabilises, we give some consideration and discussion to changes in the nanoparticles (i.e. possible oxidation state[3]) and the interface between the nanoparticle and nanowire as a result of running these sensing elements at different temperatures.

[1] S. Steinhauer, A. Chapelle, P. Menini, and M. Sowwan,
ACS Sensors, 1(5), (2016), 503-507

[2] S. Steinhauer, V. Singh, C. Cassidy, C. Gspan, W. Grogger, M. Sowwan, and A. Köck,
Nanotechnology, 26, (2015), 175502

[3] S. Steinhauer, J. Zhao, V. Singh, T. Pavloudis, J. Kioseoglou, K. Nordlund, F. Djurabekova, P. Grammatikopoulos, and M. Sowwan,
Chemistry of Materials, 29(14), (2017), 6153-6160

Nanoparticle thin film coatings provide an opportunity for a wider range of designer-specified optical, electrical, and catalytic properties compared to conventional thin films synthesized by methods such as sputtering, evaporation, and thermal oxidation. In cluster beam deposition techniques, the physical and chemical properties of resulting thin films can be tuned by controlling the size and arrangement of nanoparticle clusters during deposition. To attain the prescribed characteristics, extensive studies of their tunable parameters are often required.

In this study, silicon (Si) nanoparticle thin films are deposited onto a glass substrate via supersonic impaction following a plasma synthesis of Si nanocrystals. The custom-designed deposition tool has evolved from previous work presented in refs. [1-2]. The film porosity is tuned by adjusting the distance between the exit nozzle and the substrate during deposition, which alters the speed (and therefore the cluster size) upon impact with the substrate. Using this technique, we have previously demonstrated the tunability of the effective refractive index (n_eff) for nanoparticle thin films between that of air and their constituent material [3]. However, tuning the effective refractive index through porosity control in NP thin films introduces additional optical effects that have not yet been studied in detail.

This work explores how tuning the porosity of Si nanoparticle thin films alters their scattering behavior. The porosity of Si nanoparticle thin films depends on the size and packing density of Si nanoparticle clusters that compose it, resulting in varying pore size distributions. Through experimental and simulation methods, we study the link between porosity, pore size distribution, and scattering properties of the Si nanoparticle thin films. We start with three Si nanoparticle samples of varying porosities (70%, 80% and 90%) and similar effective thicknesses (t_eff). The pore size distribution of the three NP thin film samples is determined using the nitrogen adsorption technique based on the Brunauer–Emmett–Teller (BET) theory. Their scattering properties are determined by measuring the angular intensity distribution of transmitted light at normal incidence. These experimental results serve as a basis for developing a finite-difference time-domain (FDTD) model with pore size distribution as input, allowing us to expand the analysis beyond what we can measure experimentally.

[1] L. Mangolini et. al., High-yield plasma synthesis of luminescent silicon nanocrystals, Nano Lett. 5, 655–659 (2005).

[2] Z. C. Holman and U. R. Kortshagen. A flexible method for depositing dense nanocrystal thin films: Impaction of germanium nanocrystals, Nanotechnology 21, 335302 (2010).

[3] M. Boccard et. al. Low-refractive-index nanoparticle interlayers to reduce parasitic absorption in metallic rear reflectors of solar cells, Phys. Status Solidi A, 1700179 (2017).

The ability to precisely engineer well-defined nanoparticles and nanostructures in ever more complex configurations will allow increased functionality for nanoscale devices. We have taken advantage of plasma characteristics and alloy chemistry to control the nucleation and growth behavior during inert gas condensation (IGC) to create complex and unique nanoparticles with a variety of shapes and atomic structures as well as controlled chemical distributions. For example, core/shell nanoparticles can be designed by proper selection of alloy constituents (that is, immiscible components) and processing parameters, which have been demonstrated in a number of systems such as Fe/Ag, Fe/W, Co/Mo, and Co/Zn. We have also discovered, using plasma diagnostics, that two distinct nucleation zones can exist during the formation of nanoparticles. The first nucleation event occurs inside the plasma and fosters a more thermodynamically-controlled growth process, while the second occurs outside the plasma region and the particles grow in a more kinetically-controlled environment. With this phenomena, we can then create a two-step nucleation process and engineer even more complex structures. During the first stage, we can condense one phase or atomic species, and even control its shape, without mixing or co-condensing the second, immiscible species. These nanoparticles then provide heterogeneous nucleation sites for the secondary nucleation stage. If the first nanoparticle is not a sphere, preferential site condensation occurs, leading to, for example, core/frame structures. As an example of this, we have grown Fe(Co)/Ag core/frame structures, where Fe(Co) cubes are first condensed from the Fe-Co-Ag vapor (plasma), and Ag condenses during the second stage, first at corners to form core/corner structures, then along edges to form core/frame structures, and then on faces to form core/shell structures. In this talk, we will more fully describe the processing of these unique nanoparticles, and detail some functional properties of these complex nanoparticles/nanostructures.

The major importance of surface atoms in small nanoparticles (NPs) offers the opportunity to tailor the magnetic properties by playing with the interface between nanomagnet and its surrounding. The system FeRh has attracted a lot of attention because, when it is in the chemically ordered B2 phase (CsCl-like), it presents an antiferromagnetic to ferromagnetic order (AFM-FM) transition close to room temperature. Recently for epitaxially grown FeRh films both strain and field effects from an underneath BaTiO3 ferroelectric crystal has been exploited to electrically drive, with only few volts, the FeRh metamagnetic transition temperature (just above room temperature) ¹. This paper deal with structural and intrinsic magnetic properties of FeRh nanoclusters, prepared using Mass-Selected Low Energy Cluster Beam Deposition (MS-LECBD) available at the PLYRA platform of Institut Lumière Matière at Lyon. In sharp contrast to film and bulk studies, we have put into evidence the persistence of FM order down to 3 K (ferromagnetic alignment of the Fe and Rh magnetic moments of respectively 3 and 1 μB per atom) in size-selected 3.3 nm diameter FeRh clusters crystallized in the B2 phase ². This anomalous magnetic order has been ascribed to finite size induced structural relaxation. Very recently, FeRh nanoclusters have been deposited on a crystalline BaTiO3 layer epitaxially grown on a Nb-doped SrTiO3 layer, using a MBE technique (at INL-ECL). Grazing incidence x-ray diffraction measurements (GIXRD, on the BM32 beamline at ESRF) have been performed on different particle sizes thanks to the size-selection capabilities of the MS-LECBD setup. The epitaxial relationships FeRh[100]//BaTiO3[110] and FeRh[001]//BaTiO3[001] were determined by X-Ray Diffraction (XRD), with a possible lattice parameter variation for FeRh NPs. One can underline that it is the first evidence of epitaxial relationship between the faces of nano-crystallites pre-formed in gas phase and a mono-crystalline oxide surface. Moreover, these specific interface coupling between FeRh clusters and the BaTiO3 surface would pave the way towards the manipulation of the atomic structure, and hence the magnetic properties, through a voltage-driven control of the ferroelectric (and piezoelectric) substrate.
¹ R. O. Cherifi et al., Nature Materials 13, 345–351 (2014)
² A. Hillion et al., Phys. Rev. Letters 110, 087207 (2013)

Key physical and chemical phenomena that govern the performance of complex functional systems occur at meso and nanoscales and a crucial requisite for understanding such phenomena is exploring materials structure and composition at their natural length scales. In this respect complementary capabilities of synchrotron-based X-ray microscopes in terms of imaging, spectroscopy, spatial, depth and time resolution have opened unique opportunities to explore the properties of micro- and nano-structured objects as a function of their dimensions, morphology and operation conditions [1]. Some recent achievements will be illustrated by selected studies addressing intrinsic heterogeneity, structural and chemical evolutions of free-standing nanostructures [2], multicomponent functional materials relevant to catalysis and electrode/electrolyte interfaces [3]. The results will show how experiments combing chemical imaging with photoelectron, fluorescence or near edge absorption micro-spectroscopy at relevant length scales have led to comprehension of the system properties and the events occurring during operation conditions. Ongoing efforts for pushing the lateral resolution by implementing ptychography and development of set-ups for characterization under realistic working conditions will be outlined [4].

[1] A. Barinov et al, Nucl. Instr. Meth. Phys. Res. A, 601 195 (2009); B. Kaulich et al, J. Phys.: Cond. Mat. 23 (2011) 083002; M. Amati et al, J. Electron. Specrosc. Rel. Phenom (2017) and references therein.
[2] A. Barinov et al, Adv. Mater. 21 (2009) 1916, F. Jabeen et al, Nano Research 3 (2010) 706.
[3] B. Bozzini et al, Scientific Reports 3 (2013) 2848; ACS Appl. Mater. Interf. 6 (2014) 19621; J. Mat. Chem. A 3 (2015) 19155.
[3] G. Kourousias et al, Nano Research 9 (2016) 2046; A. Kolmakov et al, Topics in Catalysis 59 (2016) 448 and references therein.

An emerging approach for having highly efficient catalytic systems relies in the use of metal oxide substrates with metallic clusters deposited on the surface [1]. In this work we have investigated cluster-assembled zirconia (ns-ZrO₂) and atomic layer deposited zirconia (ALD- ZrO₂) substrates on which size selected Cu_n clusters (n=4, 12) have been deposited (coverage about 5%).
The focus of the study is towards conversion of CO₂ to methane and methanol by monitoring the catalytic activity under thermal ramps up to 375 ^oC [2]. All the samples show methane production whereas no signal for methanol is detected. The Cu clusters are more active on ns-ZrO₂ during the first ramp, but the activity decreases over consecutive cycling. In in situ XANES measurements the oxidation state of copper deposited on ns-ZrO₂ changes gradually and on reduction drops to ~ 1.0, whereas the change on ALD- ZrO₂ is sharp and on reduction drops to ~ 0.7-0.8.
XPS and UPS have been adopted for the characterization of the electronic structure of the samples (Cu2p, Zr3d and O1s edges) before and after the thermal treatments. XPS outcomes show a stronger interaction of Cu clusters with ns-ZrO₂ than with ALD-ZrO₂. In UPS, which is a more surface sensitive technique, Cu oxides features are evident only in ALD-ZrO₂ samples. This can be ascribable to a diffusion of Cu clusters in the nanostructured matrix, which is inhibited for ALD-ZrO₂.
This work gives considerable insights into the catalytic activity of Cu clusters on different supporting substrates offering new perspective in the use of cluster-assembled oxides for the development of highly efficient catalysts.

[1] A. Ruiz Puigdollers, P. Schlexer, S. Tosoni, G. Pacchioni, Increasing Oxide Reducibility: The Role of Metal/Oxide Interfaces in the Formation of Oxygen Vacancies, ACS Catal. 7 (2017) 6493–6513. doi:10.1021/acscatal.7b01913.
[2] B. Yang, C. Liu, A. Halder, E.C. Tyo, A.B.F. Martinson, S. Seifert, P. Zapol, L.A. Curtiss, S. Vajda, Copper Cluster Size Effect in Methanol Synthesis from CO ₂, J. Phys. Chem. C. 121 (2017) 10406–10412. doi:10.1021/acs.jpcc.7b01835.

Nanoparticle films can have a range of applications due to their tunable properties by changing material, structure, or functionalization. Nanoparticle films are especially interesting optically when their refractive index approaches 1 from the high porosity of the film. A major optical challenge is keeping the haze low when nanoparticles aggregate into films upwards of 100 µm. To inform the synthesis and deposition of nanoparticles, electron microscopy can be used not only to determine the size of nanoparticles and the thickness of thin films, but also to analyze the shape of nanoparticles and necking between particles with transmission electron microscopy. Gas phase synthesis was performed with both PECVD and gas pyrolysis to produce 5 nm diameter, spherical, silicon nanoclusters, 10-15 nm silica nanoparticles, or nanoclusters and larger fluorine-doped tin oxide nanoparticles. The freshly synthesized nanoparticles in the gas stream were then fed into a custom spray coating tool for high throughput up to 5 µm/min of 10-15 nm nanoparticles on up to 5 in² surfaces. The shapes and necking of nanoparticles before and after deposition were obtained and compared to determine the relationship with haze as determined by UV-Vis-NIR spectroscopy. We found that the shape and necking were also related to porosity of the film as determined by ellipsometry, affecting the refractive index. We outline ways to vary the synthesis and deposition parameters to vary porosity and minimize haze to 5% at 550 nm for a 100 µm film.

In this work, we study the mechanisms of growth of Ag nanoclusters in a solid Ar matrix and the emission of these nanoclusters from the matrix by a combination of experimental and theoretical methods. The molecular dynamics simulations show that the cluster growth mechanism can be described as "thermal spike-enhanced clustering" in multiple sequential ion impact events. We further show that experimentally observed large sputtered metal clusters cannot be formed by direct sputtering of Ag mixed in the Ar. Instead, we describe the mechanism of emission of the metal nanocluster that, at first, is formed in the cryogenic matrix due to multiple ion impacts, and then is emitted as a result of simultaneous effect of interface boiling and a spring force effect. We also develop an analytical model describing this size-dependent cluster emission. The model bridges the atomistic simulations and experimental time and length scales, and allows increasing the controllability of fast generation of nanoclusters in experiments with high production rate.

Silicon underpins nearly all microelectronics today and will continue to do so for some decades to come. However, for silicon photonics, the indirect band gap of silicon and lack of adjustability severely limit its use in applications such as broadband photodiodes. Here, a high-performance p-Si/n-ZnO broadband photodiode working in a wide wavelength range from visible to near-infrared light with high sensitivity, fast response, and good stability is reported. The absorption of near-infrared wavelength light is significantly enhanced due to the nanostructured/textured top surface. The general performance of the broadband photodiodes can be further improved by the piezo-phototronic effect. The enhancement of responsivity can reach a maximum of 78% to 442 nm illumination, the linearity and saturation limit to 1060 nm light are also significantly increased by applying external strains. The photodiode is illuminated with different wavelength lights to selectively choose the photogenerated charge carriers (either electrons or holes) passing through the depletion region, to investigate the piezo-phototronic effect on electron or hole transport separately for the first time. This is essential for studying the basic principles in order to develop a full understanding about piezotronics and it also enables the development of the better performance of optoelectronics.

The possibility to sensitize wide band gap oxides to visible light has stimulated the research community in view of efficiently converting solar to chemical energy, with the aim of obtaining optimized materials for photo-catalysis and sensor applications. Cerium oxide has been started to use considering the presence of localized Ce 4f states between the filled O 2p valence band and the empty Ce 5d conduction band that can make the material a very sensitive probe to identify possible charge transferred to/from neighboring metal atoms. The occupation of the 4f levels is in turn expected to modify the material properties, decreasing the oxygen vacancy formation energy and modifying its optical response. The material can be coupled with plasmonic nanoparticles (NPs). Irradiation with photon energies which excite the localized surface plasmon resonance (LSPR) and hot carries generation can induce energy and/or charge transfer from the metal to the oxide, though undoubtedly responsible for the enhancement of the activity of the material [1]-[3].
Cluster assembling in inert gas condensation chamber and mass selection filter allows to obtain silver NPs with different size down to 10 nm. NPs were characterize using TEM in order to control the morphology and the crystallinity. These nanostructures were fully embedded in CeO2 matrix grown on MgO and Quartz substrates (material with higher band gap) using molecular beam epitaxy in order to maximize the interfaces between the oxide and the metal and to prevent the oxidation of NPs. More layer of CeO2- Ag NPs- CeO2 with different thickness of ceria were realized to have also information on possible collective excitations. The oxidation state of the cerium were obtained using XPS in situ experiment during the growth.
Transient absorption spectroscopy was used to explore mechanisms of energy/charge transfer at femtosecond/picosecond timescale. Towards observing charge/energy transfer, we performed these pump&probe experiments pumping first at 410 nm to excite the LSPR and then at 275 nm in order to excite cerium oxide intraband; for all the excitation were acted using different fluencies of the pump in order to excite more hot electron and so to observe higher signal of transfer.

[1] B. Li, et al. ACS NANO 8, 8152 (2014).
[2] S. M. Kim et al. JPCC 119, 16020 (2015).
[3] D. C. Ratchford et al. Nano Lett. 17, 6047-6055(2017)

There is currently a great deal of interest in building ‘neuromorphic’ computers (i.e. computers based on similar principles to the mammalian brain) because of their potential to outperform traditional computers in, for example, pattern recognition tasks. [1] In most cases, CMOS circuitry is used to emulate the neurons and synapses of the brain, but a radical alternative approach is to build chips which are themselves organised into brain-like structures. [2, 3]

When clusters (nanoparticles) are randomly deposited on a surface they produce percolating films [4] comprising complex networks of atomic switches. [5, 6] These memristor-like switches exhibit synapse-like qualities [7] that provide potential for neuromorphic behaviour. [2, 3] Here we focus on devices based on percolating films of Sn clusters, and especially those that are deliberately constructed so as to guarantee that the film is close to the percolation threshold (onset of conduction). We have developed methods for stabilizing these devices so that they consistently operate for periods of months. [8] Application of controlled pulsed voltage sequences allows us to explore the neuromorphic (i.e. brain-like) properties of the devices.

We show that groups of connected clusters can be viewed as analogous to neurons, and that tunnel gaps between groups provide sites at which atomic scale switching elements can act as synapses. [8, 9] We explore the complex switching dynamics in detail and show that the inherent complexity of the percolating networks ensures that - like the brain - they are poised near criticality. [9] Therefore percolating networks of clusters provide a unique and straightforward route to fabrication of neuromorphic chips.

[1] R. Service, Science 345, 614 (2014); P. A. Merolla et al, Science 345, 668 (2014).
[2] A. V. Avizienis, H. O. Sillin, C. Martin-Olmos, H. H. Shieh, M. Aono, A. Z. Stieg, and J. K. Gimzewski, PLoS One 7, e42772, (2012).
[3] A. Z. Stieg, A. V. Avizienis, H. O. Sillin, C. Martin-Olmos, M. Aono, and J. K. Gimzewski, Adv. Mater. 24, 286, 2012.
[4] J. Schmelzer jr, S. A. Brown, A. Wurl, M. Hyslop, and R. J. Blaikie, Phys. Rev. Lett. 88, 226802 (2002).
[5] A. Sattar, S. Fostner, and S. A. Brown, Phys. Rev. Lett. 111, 136808 (2013).
[6] S. Fostner and S. A. Brown, Phys. Rev. E 92, 052134 (2015).
[7] T. Ohno, T. Hasegawa, T. Tsuruoka, K. Terabe, J. K. Gimzewski, and M. Aono, Nat. Mater. 10, 591 (2011).
[8] S. K. Bose, J. Mallinson, R. M. Gazoni and S. A. Brown, in press, IEEE Trans Nano Devices (2017).
[9] S. K. Bose, S. Shirai, J. Mallinson, S. Acharya, S. A. Brown, to be submitted.

In this presentation, I will give an overview of some our recent progress in making nanoparticles alloys and intermetallic compounds for catalysis, particularly in relation to conversion of sustainable energy [1]. First, we shall discuss new catalysts for electrochemical Oxygen Reduction Reaction [2, 3], which is really the limiting reaction in Proton Exchange Membrane Fuel Cells. Here we have proved that it is possible to make mass-selected nanoparticles of the Pt-early transition alloys like PtY alloys [4] and PtGd alloys [5]. We shall demonstrate that as similar approach can be used for studying size dependence and efficiency for catalysts related to water splitting [6-7] and evaluate the scalability of scarce and expensive elements like Platinum and Ruthenium. Similarly, size dependence and isotope labelled experiments will be presented for NiFe nanoparticles for oxygen evolution under alkaline conditions [8]. Here we shall demonstrate a new principle for dynamic detection of gas evolution [9]. This has also been used to investigate electrochemical CO hydrogenation on mass-selected Copper nanoparticles to study the dynamical influence of surface oxygen on the selectivity for methane/ethene production [10].

References
1) Z. W She, J. Kibsgaard, C. F. Dickens, I. Chorkendorff, J. K. Nørskov, T. F. Jaramillo, Science (2017) 355.
2) J. Greeley, I.E.L. Stephens, A.S. Bondarenko, T. P. Johansson, H. A. Hansen, T. F. Jaramillo, J. Rossmeisl, I. Chorkendorff, J. K. Nørskov,
Nature Chemistry. 1 (2009) 522.
3) M. Escudero-Escribano, P. Malacrida, M- Hansen, U. G. Vej-Hansen, A. Velázquez-Palenzuela V. Tripkovic, J. Schøitz, J. Rossmeisl, I. E.L.
Stephens, I. Chorkendorff, Science 352 (2016) 73.
4) P. Hernandez-Fernandez, F. Masini, D. N. McCarthy, C. E. Strebel, D. Friebel, D. Deiana, P. Malacrida, A. Nierhoff, A. Bodin, A. M. Wise, J.
H. Nielsen, T. W. Hansen, A. Nilsson, I. E.L. Stephens, I. Chorkendorff, Nature Chemistry 6 (2014) 732.
5) A. Velázquez-Palenzuela, …. D. Friebel, A. Nilsson, I. E.L. Stephens, I. Chorkendorff, J. Catal. 328 (2015) 297.
6) E. Kemppainen, A. Bodin, … C. K. Vesborg, J. Halme, O. Hansen, P.D. Lund and I. Chorkendorff, Energy & Environmental Science, 8 (2015)
2991.
7) E. A. Paoli, F. Masini, R. Frydendal, D. Deiana, C. Schlaup, M. Malizia, S. Horch, I. E.L Stephens, I. Chorkendorff, Chemical Science, 6
(2015) 190.
8) C. Roy, …,J. Kibsgaard, I. E. L. Stephens, and I. Chorkendorff; In Preparation (2017).
9) D. T. Bøndergaard, T. Pedersen, O. Hansen, I Chorkendorff and P. C. K. Vesborg, Rev. Sci. Inst. 86 (2015) 075006.
10) S. B. Scott, D. B. Trimarco, …. P. C. K. Vesborg, Jan Rossmeisl, and I. Chorkendorff, In Preparation (2017).

With the intensifying global need for alternative energy, there is strong interest in new approaches for generating and using energy efficiently, and for sustainably storing energy as chemical fuels. Efficient energy transformations involving chemical bonds rely on effective catalysts to lower reaction kinetic barriers. The increased variety of available structural and chemical configurations in vapor phase condensed nanoparticles has the potential for new and improved heterogeneous catalysts. Here we report on elemental and alloy nanoparticles formed via vapour phase condensation. The vapor phase source is a three-target magnetron sputter source operating into a relatively high-pressure inert gas environment. The sputtered atoms are cooled by collisions with the inert gas molecules, resulting in condensation into nanoparticles. By controlling the inert gas pressure and composition, the length of the condensation zone and other experimental variables the particle size distribution can be tuned. A quadrupole mass spectrometer is used to characterize the size distribution and select a size to impinge on the sample surface. We report on copper-silver alloy nanoparticles for the oxygen reduction reaction, and on rhenium nanoparticles incorporated into an electrode structure for electrochemical ammonia production. We discuss use of appropriate underlayers to facilitate particle adhesion in chemical environments. The effect of structure, size and composition on electrochemistry is explored.

Smart nanomaterials with designed properties based on nanoparticles have the potential to revolutionize applications in magnetics, catalysis, and optoelectronics. But implementing nanoparticles’ potential for such applications requires realizing and understanding nanoparticles with controllable size, morphology, crystal structure and chemical composition on a large scale, at low costs and in a safe and environmentally friendly way. So far, this has not been done to a great extent because few methods exist that can enable large-scale production of materials with high enough control of the designed particle properties mentioned, thus hampering the exploration of a wide range of nanoparticle-based materials.

Very recently aerosol generation by the spark discharge method has been identified, within the large EU-program Buonapart-e, as the method that can fulfill all of the above-declared requirement, but so far mainly the large-scale production capabilities have been demonstrated but work remains when it comes to the designed nanoparticles.

Nanoparticle generation by spark discharge where the formation of a plasma channel between two conducting electrodes leads to a spark discharge that evaporates material is a fairly simple method. To create a spark discharge between the electrodes a self-pulsed circuit is used, consisting of a capacitor bank driven by a high voltage DC power supply connected in parallel to the electrode gap. Several parameters can be adjusted to affect and control the particle production including the capacitance, the discharge frequency (by the output current), the electrode distance, the electrode material, the carrier gas flow rate, the type of carrier gas and the geometry of the spark discharge generator.

Here, we report on how the generation parameters, as well as the post-treatment of the produced particles, can be used to control their crystal structure, size, morphology and chemical composition. In addition, some examples of applications where the as-produced particles with designed properties have been used will be presented.

Lithium-oxygen batteries are a promising new class of rechargeable Li batteries with a potentially very high achievable energy density. A major challenge in their development remains the high charge overpotential, which results in a low energy efficiency.¹ It has recently been shown that lithium superoxide (LiO₂), which is a better electronic conductor than the usual lithium peroxide (Li₂O₂) discharge product, can be grown in a Li-O₂ battery using iridium as a catalyst. However, the LiO₂ formation mechanism is not well understood as the Ir particles in the catalyst varied within the size range from a dimer up to hundreds of nm.² To understand if the working catalytic particles are subnanometer sized clusters with dominantly under-coordinated atoms³ or larger particles, we prepared cathodes by depositing size-selected Ir clusters (Ir₂, Ir₄, and Ir₈) on reduced graphene oxide (rGO) supports with varying coverages.⁴ Select samples were annealed with the goal to produce larger nanoparticles through the agglomeration of the deposited clusters. Testing of the Ir/rGO cathodes show that annealed cathodes possess the lowest charge potential of 3.5 eV, comparable with the one reported in [2]. Detailed characterization of the annealed samples using high resolution transmission electron microscopy (HRTEM) and scanning transmission electron microscopy (STEM), complemented by Raman, titration and other techniques confirms the morphology of the working catalysts and the reaction mechanism predicted by DFT calculations. I will also discuss Li₂O₂ based Li-O₂ cells, where the exact size of Ag clusters (Ag₃, Ag₉, and Ag₁₅) deposited on passivated carbon surface, is found to play a dramatic role influencing the morphologies of Li₂O₂, and hence the charge process.³

Reference:
1.D. Aurbach, B. D. McCloskey, L. F. Nazar, and P. G. Bruce. Advances in understanding mechanisms underpinning lithium–air batteries, Nature Energy 1, 1-11 (2016).
2.J. Lu et. al. A lithium–oxygen battery based on lithium superoxide, Nature 529, 377–382 (2016).
3.J. Lu. et. al. Effect of the size-selective silver clusters on lithium peroxide morphology in lithium–oxygen batteries. Nature Communications 5, 4895 (2014).
4.A. Halder et. al. Evolution of cathode nanostructures in lithium superoxide batteries. (under revision).
<!--![endif]---->

Nanotechnology is a rapidly progressing field with tremendous impact on fields such as materials, electronics and medicine. In the case of Li-Ion Batteries (LIBs), much effort is focused on the design of nanostructured electrodes that accelerate the charge-discharge processes, as well as prevent the fracture of the electrode caused by the volumetric expansion-contraction during the lithiation-delithiation mechanisms. This last effect is particularly important for high capacity electrodes, such as Si anodes, which can increase their volume up to four times during lithiation. Nevertheless, the synthesis of nanostructured electrodes is associated with new challenges such as reproducibility, use of harsh chemicals, aggregation and manufacturing difficulties.

Herein, we propose a new approach to synthesize nanostructured electrodes based on the use of Nanoparticle (NP) scaffolds interlayered within amorphous Si (a-Si) layers in order to exploit the advantages of both nanoparticulated and continuous Si films without suffering from their respective shortcomings. The current prototype is fully synthesized by Physical Vapor Deposition (PVD), comprising of an a-Si layer deposited by RF-sputtering and Ta NP scaffolds fabricated by Cluster Beam Deposition (CBD). This prototype anode benefits from being synthesized directly on the substrate (making it easy to incorporate to any microelectromechanical device) without the addition of any binder or chemical solvent, allowing versatile, one-pot synthesis with excellent control of thickness and size of the deposited material. The high chemical stability of the Ta combined with size selection during synthesis and post-growth behavior allow an in-depth evaluation of the NP scaffold’s effect on the structure, morphology, and nanomechanical properties of the a-Si layers, and of the electrochemical performance of the synthesized anode.

Sequential deposition of alternating Ta NP and a-Si layers provides an anode with a multilayer configuration. The a-Si layers show particular structural features such as increased surface roughness, nano-granularity and porosity that are dictated by the nanoparticle scaffolds, boosting the lithiation process due to fast Li diffusion and low electrode polarization. Consequently, the proposed anode shows a remarkable charge/discharge speed, 1200 mAh g^-1 at 10C. Also, nanomechanical heterogeneity self-limits the capacity at intermediate charge/discharge rates; providing exceptional cycleability at 0.5 C, with 100% retention over 200 cycles with 700 mAh g^-1. Higher capacity can be obtained when the first cycles are performed at 0.2C, due to the formation of micro-islands.

This study shows that through engineering underlayers of porous NP scaffolds it is possible to manipulate the morphological, mechanical, and electrochemical properties of the a-Si layers, providing a versatile synthesis approach with promising perspectives to customize nanostructured Si anodes.

Mass-selected cluster deposition is used to prepare model “inverse” catalysts comprised of small metal oxide (M_xO_y: M = Ti, Nb, Mo, Ce, W) and sulfide (M_xS_y: M = Mo, W) clusters deposited on Cu and Au surfaces, respectively, for studies related to the water-gas-shift reaction and CO₂ activation. A key advantage of cluster deposition is that it allows control over cluster stoichiometry which provides a means of introducing oxygen/sulfur “vacancies” and varying the average cation oxidation state. Moreover, the use of well-ordered supports and size-selected clusters is ideally suited for computational modeling of structure and reactions using DFT electronic structure theory. Results will be presented for recent studies of water dissociation on oxide clusters deposited on Cu and Cu₂O surfaces, including mechanistic studies of reactions at elevated pressures using ambient pressure XPS. Investigations of the binding and reaction of CO₂ on alkali-modified surfaces of metal sulfide clusters deposited on Au(111) will also be presented.

This work was carried out at Brookhaven National Laboratory under Contract No. DE-SC0012704 with the U.S Department of Energy, Office of Science, and supported by its Division of Chemical Sciences, Geosciences, and Biosciences within the Office of Basic Energy Sciences.

Highly filled particulate nanocomposite films consisting of metal nanoparticles in a dielectric organic or ceramic matrix have unique functional properties with hosts of applications. In most applications, a high filling factor close to the percolation threshold with control of the particle separation on the nm scale is essential because the functional properties often require short-range interaction between nanoparticles. The present talk demonstrates how vapor phase deposition techniques can be employed for tailoring the nanostructure and the resulting properties. Vapor phase deposition, inter alia, allows excellent control of the metallic filling factor and its depth profile as well as the incorporation of alloy nanoparticles with well-defined composition. We applied various methods such as sputtering, evaporation, and plasma polymerization for the deposition of the matrix, and the metallic component was mostly sputter-deposited or evaporated. Here we put emphasis on generation of the nanoparticles by means of high-rate gas aggregation cluster sources to obtain independent control of filling factor and size of the embedded nanoparticles. Results include findings from UV vis in situ measurements on the early stages of formation of plasmonic nanoparticles, formation of multiple core-shell particles and in situ control of the composition of alloy nanoparticles. Examples of fabricated nanocomposites range from optical composites with tuned particle surface plasmon resonances for plasmonic applications and magnetic high frequency materials with cut-off frequencies well above 1 GHz through sensors for volatile organic compounds and photoswitchable devices that are based on the huge change in the electronic properties near the percolation threshold to biocompatible antibacterial coatings with tailored release rate.

Graphene’s two-dimensional nature makes it very susceptible to adparticles: adsorbed atoms or molecules, either individual or clustered. For instance, graphene’s electronic properties have been shown to be susceptible to gas molecule adsorption with a sensitivity down to single molecule detection.¹ Conversely, the properties of deposited zero-dimensional adparticles are strongly affected by the interaction with the support. In a graphene–adparticle system, both low-dimensional components define the characteristics of this hybrid structure.

Small metallic clusters exhibit distinct electronic and structural properties, that vary in a non-scalable way with their size. Theoretical investigations of well-defined few-atom metallic clusters as adparticles on graphene suggest that the cluster’s size-dependent properties get carried over in, for instance, graphene’s electronic properties.²

We investigated the interaction between size-selected Au₂, Au₃, and Au₆ clusters and graphene. Hereto
preformed clusters are deposited on graphene field-effect transistors, an approach which offers a high control over the number of atoms per cluster, the deposition energy and the deposited density.³ In situ electronic transport measurements on cluster-graphene devices show that charge transfer between the clusters and the graphene leads to p-doping and enhanced charge carrier scattering. The results also indicate that a major part of the deposited clusters remains on the graphene flake as either individual entities. Cluster size dependencies are correlated with the electronic structure of the isolated clusters.

This approach provides perspectives for electronic and chemical sensing of metallic clusters down to their atom-by-atom size-specific properties, and exploiting the tunability of clusters for tailoring desired properties in graphene.

¹ F. Schedin, A. K. Geim, S. V. Morozov, E. W. Hill, P. Blake, M. I. Katsnelson and K. S. Novoselov, Nat. Mater. 6 (2007) 652.
² M. K. Srivastava, Y. Wang, A. F. Kemper and H.-P. Cheng, Phys. Rev. B 85 (2012) 165444.
3 J.E. Scheerder, T. Picot, N. Reckinger, T. Sneyder, V.S. Zharinov, J.F. Colomer, E. Janssens, and J. Van de Vondel, Nanoscale 9 (2017) 10494.

Recently we reported on the formation of metastable configurations of Ag-Cu NPs grown by concurrent magnetron sputtering inert-gas condensation from neighbouring sputtering targets, and showed that energetically unfavourable structures can be generated as a result of nanoparticle coalescence at a later stage of the nucleation and growth process, away from the target and the hot plasma region surrounding it [1]. Ag and Cu form a eutectic system with a pronounced miscibility gap; as such, the two materials are not prone to mix upon coalescence, forming core-satellite or Janus structures, or, in the case of Cu-rich structures of ~5%Ag, a patterned core-partial shell structure which we named ukidama, after traditional Japanese glass floats.
A question that was naturally raised, based on this study, was what would happen if the system was a miscible one. Would it behave in the same way, or would the atoms of the two elements actually mix, as expected from thermodynamics considerations?
To answer this question, we performed a similar study, using the same experimental setup, for the Ni-Pt system; in parallel, we developed nanophase diagrams using thermodynamics considerations. Ni-Pt is a well-established miscible system, as it satisfies all Hume-Rothery criteria: the atomic radii difference is only ~10%, both elements form FCC crystalline structures and have comparable valency and electronegativity values. We found that both stable solid solution and metastable segregated configurations were possible, which we attributed to the frequency of coalescence events. We also observed the existence of ukidama structures, similar to those for the eutectic Ag-Cu system, which can be explained using our previously reported extended cluster heating model of nanoparticle coalescence [2].
[1] “Kinetic trapping through coalescence and the formation of patterned Ag-Cu nanoparticles”
P. Grammatikopoulos, J. Kioseoglou, A. Galea, J. Vernieres, M. Benelmekki, R. E. Diaz, M. Sowwan
Nanoscale 8 (2016) 9780-9790.
[2] “Simple analytical model of nanocluster coalescence for porous thin film design”
P. Grammatikopoulos, E. Toulkeridou, K. Nordlund, M. Sowwan
Modelling Simul Mater Sci Eng 23 (2015) 015008.

Nanocomposite magnets consisting of a fine mixture between a hard magnetic phase and a high saturation magnetization phase would be a promising way to go beyond performances of nowadays single magnetic phase magnets, on top of which NdFeB-based magnets. Theoretical descriptions of nanocomposite magnets [1] revealed the necessity of confining the soft phase in grains of typically less than 10 nm. Yet standard fabrication processes do not permit to have such a control on the microstructure. In this context nanomaterial-dedicated synthesis could permit to realize model films to experimentally explore fine mechanisms that govern magnet performances in such nanocomposite magnets [2].
Here we report on results we obtained from synthesis of Co:FePt nanocomposite by combining low energy cluster beam deposition (LECBD) technique and e-beam evaporation. Indeed to separately adjust size, composition and concentration, two independent beams: one for the mass-selected Co clusters preformed in gas phase and one for the hard L1₀-FePt matrix produced by alternative electron gun evaporation, are in-situ deposited on a same substrate. Doing so, the soft phase is restricted to the nanocluster size, selected from 2 to 8 nm while the volume fraction of nanoclusters in L1₀-FePt matrix was varied up to 30%.
In this paper, we present results of standard structural characterizations (XRD, SEM, TEM) and magnetic characterizations (SQUID magnetometry, MFM) performed on our nanocomposite samples. In addition, synchrotron characterizations (EXAFS, XMCD and XLD on the K-edges of Fe and Co) show the fine structure of the nanocomposite. Especially we will discuss specific magnetization reversal of soft and hard phases and the evolution at the interface of Co nanoparticles and the FePt matrix introduced by annealing at high temperature.

[1] Skomski R. and Coey J. M. D. Giant energy product in nanostructured two-phase magnets. Phys. Rev. B 48, 15812–15816 (1993).
[2] Balasubramanian B., Mukherjee P., Skomski R., Manchanda P., Das B. and Sellmyer D. J. Magnetic nanostructuring and overcoming Brown's paradox to realize extraordinary high-temperature energy products. Scientific reports 4, 6265 (2014)

A novel biosensor for the ultra-sensitive electrochemical detection of para(p)-aminophenol based on an array of size selected Au clusters deposited on Fluorine doped Tin Oxide (FTO) was produced using a novel magnetron cluster beam deposition (CBD) setup equipped with a particle size-selector. Detailed characterization of the structural and electrochemical properties of the biosensor was performed by a combination of aberration-corrected Scanning Transmission Electron Microscopy (STEM), Scanning Electron Microscope (SEM), Electrochemical Impedance Spectroscopy (EIS) and Cyclic Voltammetry.
The fabrication process of the Au-cluster based biosensor is carried out under ultra-high vacuum conditions and involves the production of metal particles in the gas-phase through gas borne aggregation of sputtered gold atoms from a pristine metal target. Next, the formed charged clusters are size-selected (2.5-3.5 nm) based on their kinetic energy using a quadrupole size selector and deposited at a typical coverage of 1 atomic monolayer (~10^Λ15 atoms/cm²) on FTO supports. Unique surface modification of FTO by an array of isolated and monodisperse Au clusters with 3.1±0.3 nm size was confirmed by electron microscopy. EIS study reveals that controlled immobilization of these surfactant free Au clusters efficiently decreases the electron exchange resistance between the analyte and bare FTO from 43.12 Ohms to 29.65 Ohms. Cyclic Voltammetry measurements demonstrate the excellent redox activity of Au-cluster modified FTO that is capable of oxidizing para-aminophenol at 0.14 V compared to 0.67 V for bare FTO. This novel biosensor enables the ultra-sensitive electrochemical detection of para-aminophenol in a concentration range of 10 to 80 µg ml^-1 with a detection limit of 404 ng ml^-1(S/N = 3).



Acknowledgement
The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) under grant agreement n° 607417 (Catsense).

Nanoclusters are an interesting research area due to their different catalytic, electronic, magnetic, and optical properties from their corresponding bulk counterparts. Among the various fabrication methods, inert gas condensation (IGC) effectively produces monodispersed nanoclusters with sizes typically in the 5-20 nm range. Further, IGC is a proven route to produce core/shell structures directly from the gas phase. Here, we produce Co core/ZnO shell nanoclusters to study electric field-induced resistive switching as these structures are candidates for resistive random-access- memory, which has shown great potential to lead the next generation nonvolatile memory technology.
IGC was utilized to produce initially Co core/Zn shell nanoclusters. The Zn was subsequently oxidized to form ZnO both in situ by bleeding oxygen into the gas condensation chamber and ex situ by exposing the nanoclusters to air. The core-shell Co/ZnO nanoclusters were characterized by transmission electron microscopy (TEM) to study the crystal structure of nanoparticles and scanning transmission electron microscopy (STEM) for imaging and elemental mapping. High resolution TEM images displayed Moiré fringes in core and shell regions indicated both core and shell are crystalline, with fcc cobalt in the core and epitaxially grown zinc oxide in the shell. STEM using high angle annular dark field (HAADF) imaging and elemental energy dispersive X-ray spectroscopy (EDS) mapping provided more evidence of the core-shell structure of the nanoparticles. The resistive switching behavior in Co/ZnO nanoclusters were studied by using two methods to obtain I-V characteristics of these nanoparticles. In the first method, which is done in electrical modes of an atomic force microscope (AFM), the conductive tip of the AFM acts as a mobile top electrode, and the voltage can be varied on a single nanocluster deposited on a silicon substrate. The second method deposited the nanoclusters between two metal electrodes and the I-V characteristics were obtained using a two probe configuration. I-V graphs obtained in both methods indicated the bipolar resistive switching behavior which can be explained by the formation and disruption of conducting nano-filaments induced by oxygen vacancy migration and the contribution of Co nanoclusters. With applying a positive voltage, a “SET” process from high resistance OFF state to low resistance ON state appears. However, upon applying negative voltage a transition from a low resistance state to a high resistance state as a “RESET” process appears.

Gas phase synthesis based on DC Magnetron sputtering is a promising method to produce nanoparticle and bimetallic nanoparticles (NPs) due to its single step generation of high purity nanoparticles compared to other physical and chemical methods. Although fine tuning of bimetallic NPs with different structural motifs and sizes is very well established and reported for noble and some transition metals, it still remains a challenge to extend this approach towards alkali and alkaline earth metals. This is because these metals with their low reduction potential readily react with oxygen and water. Moreover, limitation associated with nanoparticle production rate and insufficient information on the mono/ and bimetallic NP nucleation mechanisms is a major drawback in understanding the formation process of bimetallic NPs and improve its production rate for large scale synthesis.
We show the influence of CH₄/H₂ as a trace gas on the nucleation and formation of bimetallic nanoparticles (NPs) prepared by gas phase synthesis. We show this as a strategy to nucleate bimetallic nanoparticles (NPs) made by gas phase synthesis of elements showing difficulty in homogenous nucleation. We illustrate the above mentioned strategy for the case of Mg based bimetallic NPs, which are interesting as hydrogen storage materials and exhibit both nucleation and oxidation issues even at ultra-high vacuum conditions. The above mentioned strategy is associated with tuning dimer bond energy of the formed species to produce stable nuclei for NP formation. We show that not only issue associated with nucleation of NP can be solved but diverse variety of structural motifs can be obtained from alloy to core\shell structures with good control of the NP morphology, size and chemical distribution. Moreover, by tailoring the composition of Ti, Mg and the type of employed trace gas, the as prepared MgTi NPs can be tuned from hexagonal pyramid to platelet shapes. We elucidate the reason based on (i) defects, and (ii) hydrogen and carbon adsorption on (111) planes that alters the growth rates and surface facet stabilization. The shape of MgTi NPs is identified from selected area electron diffraction (SAED) pattern and tomography that is a 3D reconstruction based on a tilt series of Bright-Field transmission electron microscopy (TEM) micrographs. Finally based on our experimental observations and generic geometrical model analysis, we prove that the formation of the various structural motifs is based on two spatially and temporally separated nucleation and growth processes instead of phase separation. This is shown to be associated with the dimer bond energies of the various formed species and the vapor pressures of the metals, which are key factors for NP nucleation.

Pure silver and gold nanoclusters have been studied widely in recent years for their size- and shape-dependent catalytic and optical properties. With straightforward synthesis, light scattering ability and potential for high stability, gold nanoparticles also lend themselves to molecular biology. Alternatively, the implementation of silver nanoparticles in consumer products due to their anti-bacterial activity has stirred investigation into their effects on the environment and on human health. Bimetallic gold-silver nanoparticles are currently being investigated for their feasibility as next-generation catalysts and sensors, whose properties are subject to fine tuning of the gold to silver ratio. Due to the rise in applications for these nanoparticles and potential increase in human exposure, it is critical to have an atomic-level understanding of the reactivity of nanoparticles in biological milieu, especially their reaction mechanism with bio-molecules. While the interaction of noble metal clusters with DNA has been studied extensively, to date, the reactivity studies of these metal nanoparticles with neurotransmitters have not been reported. In this presentation, we discuss results of our recent DFT-based computational study on the reactivity of size-selected Au_n, Ag_n, Au_xAg_y (n = 8, 10; x + y = n) nanoclusters with dopamine. We address the effects of size, shape, and alloying on the reactivity, stability, and optical properties of these clusters. We also address the solvent effects on the interactions between these nanoclusters and dopamine.

Soft-landing of mass-selected ions is a powerful tool for precisely-controlled preparation of cluster-based interfaces. We have developed a high-intensity electrospray ionization source that delivers nanoamperes of mass-selected cluster ions onto surfaces. This capability allows us to efficiently populate two- and three-dimensional (2D and 3D) interfaces with cluster ions. We are particularly interested in studying the deposition of stable anionic species, which efficiently retain their ionic charge on surfaces. Although it is generally assumed that charging of the surface should prevent ions from reaching the deposition substrate, we observe a steady increase in the deposition rate with ion exposure. In situ surface characterization using infrared spectroscopy indicates a gradual accumulation of hydrocarbons during anion deposition. We demonstrate that co-adsorption of neutral molecules stabilizes the charged layer and results in formation of either solid- or liquid-like layers with bulk properties strongly dependent on the soft-landed anion, co-adsorbed neutral, and surface. Furthermore, we demonstrate that anion deposition is an effective tool for populating 2D and 3D interfaces with redox-active clusters. This capability is particularly important for preparation of high-performance electrode-electrolyte interfaces for applications in supercapacitors and lithium ion batteries.

One of the most technologically relevant application of oxide supported transition metals lies in heterogeneous catalysis. This is particularly true in emission control systems - such as the catalytic converter of automobiles - where they are usually highly dispersed and known to be oxidized under ambient conditions. One component in the automobile three-way catalytic converter is Pt/CeO₂; where Pt serves to oxidize hydrocarbons and carbon monoxide, while ceria (CeO₂) acts as an oxygen storage component. Although control at the nm-scale is desirable to open new technological possibilities, there is limited knowledge, both experimentally and theoretically, regarding the geometrical structure and stability of sub-nanometer platinum Pt_N/CeO₂(111) supported clusters.

In this talk, I will describe the implementation of an unbiased Density Functional Theory based global optimization Basin Hopping Monte Carlo algorithm (BH-DFT) to study growth trends of CeO₂(111) supported Pt_N clusters. From our results, we observe a clear preference for planar 2D structures up to size Pt₈, followed by a structural transition to 3D structures at size Pt₉. This remarkable trend is explained by a subtle competition between the formation of strong Pt-O bonds and the cluster internal Pt-Pt bonds. Our calculations show the reducibility of CeO₂(111) provides a mechanism to anchor Pt_N clusters where they become oxidized in a two-way charge transfer mechanism: (a) an oxidation process, where O_surface atoms withdraw charge from Pt atoms forming Pt-O bonds, (b) surface Ce⁴⁺ atoms are reduced to Ce³⁺. In this way, the active role of CeO₂(111) support in modifying the structural properties and eventual chemical reactivity of sub-nanometer Pt_N clusters is computationally demonstrated. Finally, global optimization strategies in order to deal with bimetallic platinum-copper (Pt-Cu) supported clusters within our BH-DFT implementation will be also discussed.

There is a growing need to produce nanoparticles at a scale and cost sufficient to allow their use in industrial catalysts. A very important process is the electrolysis of water, a well-established technology making possible the production of hydrogen from water by utilizing suitably nanostructured electrocatalysts [1]. To make the process viable, these catalysts would need to be produced at low cost and with high throughput. In the present research, we have used a flexible, clean and potentially scalable method for the continuous fabrication of metal nanoparticles, based on a spark discharge process [2,3]. This method enables the mass-production of nanoparticles by striking an electrical discharge between two metal electrodes as the only precursor materials. Therefore, no wet chemical processes are needed, and there is no production of hazardous chemical waste. The nanoparticles produced are easily collected on suitable substrates and can be utilized and characterised for specific catalytic applications. Here, we demonstrate that Pt nanoparticles, produced with the spark discharge method, show excellent electrocatalytic activity in the hydrogen evolution reaction (HER) in an acidic environment [4]. The nanoparticles produced immediately after the spark ablation have a diameter of approximately 3-5 nm, and subsequently merge into larger aggregates before the eventual deposition. It was found that a loading of Pt nanoparticles on the substrates below ~100 ng/cm² gave a high activity in the HER, with ~70 mV overpotential at 10 mA/cm². The method of producing nanoparticles by spark discharge was found to be extremely flexible and environmentally-friendly, and was utilized as well to produce alloyed nanoparticles with tuneable size and stoichiometry. Eventually, nanoparticles produced by spark discharge could be coupled to more complex electrochemical systems such as photoelectrochemical water splitting devices or fuel cells.


References
[1] Roger I., Shipman M. A., Symes M. D., Nature Reviews Chemistry 1, Article number: 0003 (2017)
[2] Tabrizi N. S., Ullmann M., Vons V. A., Schmidt-Ott A., J Nanopart Res (2009) 11: 315
[3] Pfeiffer T. V., Feng J., Schmidt-Ott A., Advanced Powder Technology 25 (2014) 56–70
[4] Li X., Hao X., Abudula A., Guan G., J. Mater. Chem. A, 2016, 4, 11973

The degradation of size-selected Pt nanoclusters is studied under electrochemical conditions. Insight into the early stage of degradation mechanism is given by scanning transmission electron microscopy (STEM). Size selected clusters catalyst mimic carbon supported Pt nanoclusters and nanoparticles typically employed in proton exchange membrane fuel cells (PEMFCs).

In contrast to common assumptions, it is demonstrated that even extremely small Pt clusters made of 22 to 68 atoms exhibit a remarkable stability under electrochemical conditions.
In mild accelerated degradation tests conditions, a main effect is particle migration without agglomeration.
At higher potential limits (in the typical operation range of PEMFCs) particle detachment and/or minor Pt dissolution is observed.
By investigating mixed cluster samples it is clearly demonstrated that no preferential dissolution of Pt₂₂ by Ostwald ripening - usually held responsible to be the main mechanism for activity loss in Pt fuel cell catalysts - is observed.

The results indicate that subnanometer Pt clusters catalysts indeed might be a feasible option as extremely high-dispersed PEMFCs as long as the operation conditions are carefully adjusted and the active phase is effectively immobilized onto the carbon support.

Polymer nanoparticles (NPs) can offer numerous advantages in the field of photonics, water purification, nanomedicine and many others. Performance of polymer NPs strongly depends on their size, shape, architecture, physical and chemical properties; these should be tailor-made to meet the requirements of specific applications. In certain cases, such as drug delivery and imaging technology, polymer NPs need to be functionalized with chemical groups to achieve coupling of drugs or contrast agents to the NP surface. Carboxyl groups are known to be of particular importance in biomedicine since they may participate in covalent coupling with the terminal primary amines of peptides or proteins via a dehydration reaction.
In this work, we utilized a gas aggregation cluster source (GAS) for the synthesis of COOH-functionalized NPs by plasma polymerization of acrylic acid. Vapors of acrylic acid mixed with argon were supplied into the GAS and an rf discharge was ignited both in a continuous wave and in a pulsed mode to trigger the formation of the plasma polymer NPs. Depending on the effective power delivered to the discharge, spherical NPs were obtained with the size ranging from 15 nm to 100 nm, higher power resulting in higher fluxes of smaller NPs. In the CW mode, 10% retention of the COOH groups was detected by XPS which was constant regardless the discharge power applied. By contrast, the pulsed mode allowed tuning of the carboxyl concentration from 1 to 16% by decreasing the power at constant duty cycle. Duty cycle itself was found to be a powerful parameter to control the size, the flux and the chemical composition of the NPs at constant effective power. The plasma polymerization mechanism was found to be contributed by the radical recombination as well as by reactions of intact acrylic acid molecules with unquenched radicals.

Acknowledgement:
This work was supported by the grant GACR 17-12994S from the Grant Agency of the Czech Republic.

Human activities in recent days are mainly conducted indoors. From various objects inside the buildings, many kinds of harmful gases are emitted and damage human body. Especially, organic compounds with a high vapor pressure, therefore, existing in vapor phase at room temperature, called volatile organic compounds (VOCs) are the most significant harmful substances regarding indoor air quality. Although VOCs are usually emitted in a low concentration, their emission is continuous and prolonged exposure to them leads to severe damages in human body. Thus, highly sensitive and selective detection of VOCs is demanded. In order to detect various VOCs with high sensitivity and selectivity, various efforts on metal oxide semiconductors have been conducted on a platform of chemiresistive type gas sensors. Among various metal oxide semiconductors, due to catalytic effect to promote selective oxidation of VOCs, p-type metal oxide semiconductors are strong candidate for VOCs detection. Especially, Co₃O₄ is well-known for its catalytic effect in oxidization of CO, alkanes, and C₇H₈ to effectively detect VOCs as well as its large amount of oxygen adsorption for promising gas sensing properties. However, p-type metal oxides have comparatively lower gas response than n-type metal oxide semiconductors. Since low gas response can yield to poor selectivity of various VOCs, there have been intensive studies on enhancing response of gas sensors based on Co₃O₄. Especially, heterojunctions between different materials are very effective in enhancing gas response by promoting the adsorption of target gases or resistance modulation upon gas adsorption. Over past decades, various gas sensors using p-n heterojunctions have been reported but very few studies have been reported for those based on p-p isotype heterojunctions. Compared to p-n heterojunctions, where recombination between majority charge carriers takes place and total number of charge carriers are decreased, isotype heterojunctions can inject majority charge carriers from wide band gap materials to narrow band gap materials, leading to spatially separated charge carriers without any decrease in total number of them. Therefore, more electronic interactions between semiconductors and gas molecules, and enhanced catalytic effects to VOCs by combination of isotype p-type semiconductors through p-p heterojunctions can be expected. Herein, we report highly sensitive and selective VOCs gas sensors based on vertically aligned Co₃O₄ nanostructures decorated with NiO, fabricated by multiple-step glancing angle deposition method using e-beam evaporator. The oxidization and agglomeration of Ni into NiO during annealing result in numerous p-p heterojunctions with increased active site density and conductivity of Co₃O₄ for better performances. Our results here suggest that the p-p isotype heterojunctions can be effective structures for fabricating highly sensitive and selective VOCs gas sensors.

Nanostructured films obtained by the assembling of preformed atomic clusters are of strategic importance for a wide variety of applications. The deposition of clusters produced in the gas phase onto a substrate offers the possibility to control and engineer the structural and functional properties of the cluster-assembled films. To date the microscopic mechanisms underlying the growth and structuring of cluster-assembled films are poorly understood, and in particular the transition from the sub-monolayer to the thin film regime is experimentally unexplored. Here we report the systematic characterization by Atomic Force Microscopy of the evolution of the structural properties of cluster-assembled films deposited by Supersonic Cluster Beam Deposition¹. As a paradigm of nanostructured systems, we have focused our attention on cluster-assembled zirconia films, investigating the influence of the building blocks dimensions on the growth mechanisms and on the roughening of the thin films, following the growth process from the early stages of the sub-monolayer to the thin film regime². Our results demonstrate that the growth dynamics in the sub-monolayer regime determines different morphological properties of the cluster-assembled thin film. The evolution of roughness with the number of deposited clusters reproduces exactly the growth exponent of the ballistic deposition in the 2+1 model, from the sub-monolayer to the thin film regime.

¹ K. Wegner, P. Piseri, H.V. Tafreshi, and P. Milani, J. Phys. Appl. Phys. 39, R439 (2006).
² F. Borghi, A. Podestà, C. Piazzoni, and P. Milani, ArXiv170606771 Cond-Mat (2017).

Dewetting is a nontrivial challenge in thermal and chemical processing of thin films. Conventional methods in dewetting prevention concentrate on interfacial modification and surface capping. The first approach suppresses dewetting by placing a new chemical species at the interface between films and substrates to decrease their interfacial energy; the later one achieves the goal through exerting additional tensile strain on the film with a capping layer.

Here we prevent dewetting by doping via ion implantation. Silicon atoms of different doses are used at dopant-level concentrations (1, 2, 4×10¹⁵ atoms/cm²) implanted into 100 nm Ag thin films grown at room temperature directly on single crystal sapphire substrates by high-vacuum sputtering (~10^-6 torr). SEM, TEM, AFM and Raman are used to characterize film morphology before or after vacuum annealing treatment (803K, 30 - 1300 min, 10^-4 torr). The interfacial energy of the Ag film during heat processing is calculated based on Miller’s close-packed hexagonal cylinder model. The influence of dopant ions on grain growth is measured as well as simulated according to Cahn’s solute drag theory.

It is found that 10¹⁴cm^-2 Si⁴⁺ ions are sufficient to stop 100nm-thick Ag film from rupture even after 21 hours annealing, which can be attributed to those additional ions successfully retard Ag grain growth during heating resulting from their solute drag effects. Downsized Ag grains significantly decrease grain diameter-to-height ratio of the film, thereby enabling the continuous Ag film to possess a lower interfacial energy than the discrete Ag islands and to maintain its initial shape. Furthermore, a simulation of the solute drag effect on grain growth successfully matches our experiments, proving that small amounts of dopants are capable of decelerating grain growth and shrinking average grain size and thus preventing the thin film dewetting. These results show the potential for novel contact processing for advanced semiconductor device applications.

The ability to synthesize nanomaterials of particular shape, size, and composition is an important scientific goal and necessary for developing advanced materials in applications ranging from the biomedical to catalytic. Solution-based methods have commonly been employed to tune shape of nanoparticles via the selective adsorption of ligands; however, newer cluster beam gas condensation techniques have also shown promise in this area. Here we have developed a synthesis for copper nanocubes using a commercial sputtering-based gas condensation system. We demonstrate that copper nanoparticle shape in this system can be manipulated by tuning a few key growth parameters. Copper nanoparticle shape is shown to be highly sensitive to the geometry of the sputtering target, carrier gases, and length of the aggregation zone. This work adds to the toolbox of nanomaterials that can be grown in a cluster beam-type system and highlights the need for careful consideration of the many parameters involved in such an experiment.

We present results from density functional theory calculations referring to the magnetic properties of 13, 55, 147 and 309 atoms Fe-X (with magnetic Co and Mn or non magnetic Cu substitutions) icosahedral nanoclusters and coating on Cu111 surface in order to stabilize the film in fcc-like structure. Aiming in finding the nanocluster with the optimum magnetic moment (m_B) we explored the various sizes considering several compositions and atomic conformations for Fe with Co, Mn and Cu substitutions. Starting with the FeCu nanoclusters, it came out that configurations with agglomerated Fe atoms inside the Cu-Fe and the pure Cu surface shell are energetically favoured while the highest magnetic moment, 3.2 m_B on the Fe atom, was found in the Cu₁₂Fe case with the Fe atom located at the surface cell. In addition, increasing the size of the cluster m_B converges to the value of the surface Fe atom. The magnetic moment is mainly due to Fe's spin up - down electronic density of states difference close to the Fermi level(E_F). In particular, the spin-up Fe d electronic density of states are fully occupied yielding wavefunctions with homogeneous change distribution and Fe 4p occupation while the spin-down is almost unoccupied exhibiting dangling bonding states close to E_F. The 13 and 55 clusters are almost half-metallic, while the 147 and 309 cluster exhbit mainly metallic features, in line with the Fe clusters. In all cases the nano-clusters exhibit larger m_B than the Fe thin films and bulk systems while the alloy Side cluster's m_B was found to decrease towards the Fe monolayer on Cu(111) value, reaching a plateau above 120 atoms.
The structural and magnetic properties of the Fe nanoclusters with magnetic Co or Mn substitutions were also evaluated. Contrary to Cu₁₂Fe, Co₁₂Fe and Mn₁₂Fe are more stable when Fe is in the shell. The Co₁₂Fe total magnetic moment is higher than Co₁₃ and in both cases the spin are coupled ferromagnetic. On the other hand, Mn₁₂Fe total magnetic moment is smaller compared to Mn₁₃ and both cases show a ferrimagnetic behavior. The structural and magnetic properties of the bigger FeCo and FeMn nanoclusters as well as the FeCo/Cu111 and FeMn/Cu111 thin coatings were also evaluated. These results could be used for the design of environmental sustainable smart nanoclusters and coatings with superior magnetic properties suitable for target applications.
Acknowledgement
This work was supported by the SELECTA (No. 642642) H2020-MSCA-ITN-2014 project.
References:
C.S. Cutrano and Ch.E. Lekka J. Alloys and Compounds 707, 2017, 114-119
H. Choi, S.G. Lee, Y.C. Chung, Comp. Mater. Sci. 49 , 2010, 291-296.
T. Kish, H. Nakamoshi, H. Kasai, A. Okij, J. Phys. Soc. Jpn. 71 (2002) 2983e2985.
Th. Detzel, M. Vonbank, M. Donath, V. Dose, J. Magn. Magn. Mater. 147 (1995) LleL6.
X. Yin, K. Hermann, Phys. Rev. B 63 (2001) 115417.

It is rather clear from basic surface-energy minimization considerations why nanoparticles can grow into close-to-spherical shapes during gas phase inert gas condensation. However, sometimes the particles grow into morphologies far from a spherical one, and it is not obvious why this would happen. We have study the formation mechanisms of iron and iron-gold nanoparticles grown by magnetron sputtering inert gas condensation. Our results emphasize the decisive kinetics effects that give rise specifically to cubic morphologies. Our experimental results, as well as computer simulations carried out by two different methods, indicate that the cubic shape of Fe and FeAu NPs is explained by basic differences in the kinetic growth modes of {100} and {110} surfaces, rather than surface formation energetics. Both our experimental and theoretical investigations show that the final shape is defined by the combination of the condensation temperature and the rate of atomic deposition onto the growing nanocluster. We also construct a comprehensive deposition rate−temperature diagram of Fe NP shapes and develop an analytical model that predicts the temporal evolution of these properties. Adding Au into the system, we find that the cluster corners are decorated with Au atoms, with different possible outcomes depending on temperature. At high temperatures, due to entropy effects the Au atoms distribute evenly on the surface, while at low ones they decorate mainly the corners. The final structure of the particles can be explained by the joint forces of energy minimization and kinetic trapping. The Fe nanocubes have been demonstrated as potential candidates for gas sensing applications.

Determining whether a given nanoparticle structure truly corresponds to thermodynamic equilibrium is often quite difficult from the experimental point of view. This problem is even more evident for nanoalloys, in which both geometric shape and chemical ordering must be equilibrated. For example, in AuCo nanoalloys, experiments have produced quite different types of chemical ordering, and which one corresponds to equilibrium is still under debate [1]. For analyzing this kind of problems, computer simulations can be very helpful. To this end, combining several different computational techniques, from global optimization searches, to molecular dynamics simulations, to DFT calculations, and to statistical-mechanical modeling, is often necessary [2]. Here we consider a series of systems (AuCo, AgCo, AgCu, AgNi, AgPt, PdPt) and analyze their equilibrium structures [2,3] comparing them to the structures obtained in non-equilibrium conditions.

[1] A. Mayoral et al., Chem. Commun. 51, 8442 (2015); L.E. Marbella et al., Adv. Funct. Mater. 24, 6532 (2014), V. Dupuis et al., Phys. Chem. Chem. Phys. 17, 27 996 (2015).
[2] J.P. Palomares-Baez, E. Panizon, R. Ferrando, Nano Letters 17, 5394 (2017).
[3] E. Panizon, R. Ferrando, Nanoscale 8, 15911 (2016); Phys. Rev. B 92, 205417 (2015).