Complex Metallic Alloys are metallic solids of high structural complexity, consisting of large numbers of atoms in the unit cell. The consequences of this structural complexity are manifold and give rise to a variety of exciting physical properties. The influence that such structural complexity may exert on the lattice dynamics of CMAs will be exemplified for two different CMA systems, which evidence peculiarities in their dynamical behavior, clearly having their origin in the respective structural building blocks. In the Ba-Ge-Ni clathrate system the dynamics of encaged Ba guest atoms in the surrounding Ge-Ni host framework will be discussed with respect to the strongly reduced lattice thermal conductivity which is experimentally evidenced. Furthermore the surprising dynamical flexibility of Tsai-type clusters and especially of their innermost, tetrahedral shell will be addressed for the (Mg-)Sc-Zn system. For both systems experimental results from inelastic neutron scattering are analyzed and interpreted on atomistic scale by means of ab-initio and molecular dynamics simulations.

We present a detailed study of the diffuse scattering in the icosahedral ZnSc quasicrystal. We show that a large amount of diffuse scattering is present in the diffraction pattern. It can be interpreted using the hydrodynamic theory of quasicrystal in term of phason modes. The phason elastic constant and their ratio is found to be close to a 3-fold instability. The results will be compared to the one obtained for other quasicrystals. Implication on the stability ofthe Zn-Sc icosahedral quasicrystal will also be discussed.

The atomic structure of F-type icosahedral quasicrystals (QC&’s) in Al-based alloys, e.g., i-Al-Pd-Mn, still poses a great challenge in the research field of QC&’s. A high-quality stable approximant to these QC&’s was previously unavailable, hindering the structural information regarding the constituent clusters and their packing nature from being obtained via ordinary X-ray crystallography. In this work, a synthesis of a stable approximant has been attempted by substituting a blend of Cr and Fe for Mn, thus allowing an adjustment of the valence electron concentration. The attempt has succeeded in synthesizing a large cubic approximant at a certain mixing ratio between Cr and Fe, where fine single crystals with facets were obtained by slow cooling. The specimen has been subjected to single-crystal X-ray diffraction and ab initio structure refinement, whereby a complex crystal structure containing 4,320 atoms per unit cell has been revealed. The space group is Pa-3 and the lattice constant ~40.5 Å. The structure can be pictured as one formed through dense packing of clusters, which are allowed to overlap with each other on their peripheries. Two kinds of cluster known as the pseudo-Mackay (M) type and the Bergman (B) type [1] are involved; there is no need for glue atoms. The full icosahedral symmetry does not exist in each cluster; for instance, only 6~8 sites of the inner dodecahedral shell of each M type cluster is occupied, while the outermost dodecahedral shell of each B type cluster has uneven distribution of Pd(/Fe) and Al. The asymmetry originates from the local coordination of each cluster. It has been shown that the cluster centers are located at the vertices of a canonical cell tiling, which corresponds to the 2x2x2 superstructure of the 3/2-packing [2]. In particular, 128 M (resp. 136 B) type clusters are centered at the vertices of the even (resp. odd) parity, giving the full explanation of the F-type order. This is the first decoration model of a canonical cell tiling for a real approximant to Al-based F-type icosahedral QCs. A straightforward generalization of the model leads to a number of hypothetical approximant structures, some of which may be obtained in reality. Possible formation of these hypothetical approximants as well as quasicrystals in the present alloy system will be further investigated and the results will be reported at the meeting. [1] V. Elser, Philosophical Magazine B 73, 641 (1996). [2] C. L. Henley, Physical Review B 43, 993 (1991).

We have studied and resolved occupancy correlations in the existing average structure model of the complex metallic alloy xi;'-Al-Pd-Mn. It has approximately 320 atoms in the unit cell and many fractionally occupied sites. Model variants were constructed systematically in a tiling-decoration approach and subjected to simulated annealing by use of both density functional theory and molecular dynamics with empirical potentials. We reproduce the Al-Pd-Mn phase diagram at T = 0 K and derive an enthalpy of formation for each experimentally known stable structure. This yields a direct measure for thermodynamic stability for our structure models with respect to competing phases. The main structural building blocks of xi;' are pseudo-Mackay icosahedral (PMI) clusters which form columns along the [0 1 0] direction and, when projected, are vertices of a hexagonal tiling. The tiling representation allows us to extend structure models for the xi;'-phase to structure models of other closely related phases where the projections of the PMI columns in addition form pentagon and nonagon tiles. Our refined tiling-decoration model applies to any structure within the ε-phases family. Extending our tiling-decoration model further we present a detailed atomistic description of all tiles observed in the decagonal quasicrystalline phase of Al-Pd-Mn.

Bergman-type structures have great interest not only for their structural complexity as shown in the body-centered-cubic 1/1 approximant Mg32(Al,Zn)49 [1], but also because they represent one of the three crystalline approximant types (Bergman-, Mackay-, and Tsai-types) from which icosahedral quasicrystals can be tuned electronically. In particularly, higher order cubic 2/1 crystalline approximants, which contain similar building clusters as 1/1 approximants but show complex structural motifs closer to those of quasicrystals, play more important roles in understanding the structures and properties of corresponding icosahedral quasicrystals. However, Bergman-type cubic 2/1 approximant structures have been available only from the Mg-Al-Zn system. Here we present the structure of Na27(Au,Ga)58, Pa-3, a = 23.446(5) Å, Z = 8, a Bergman-type cubic 2/1 approximant of recently discovered Na26Au24Ga30 icosahedral quasicrystal. In the structure, the major building clusters include (1) multiply endohedral triacontahedral clusters that are centered at Wyckoff 8c (~ 0.346 x x) sites. Each triacontahedral cluster contains, from the center out, an empty icosahedron, a dodecahedron, a larger icosahedron, a buckyminister-like 60-atom polyhedron, and the outmost Na triacontahedron; and (2) prolate rhombohedra that are centered by Ga-Ga dimers. The prolate rhombohedra are the “fillers” around the condensed triacontahedral clusters. Such structural motifs are consistent with the 2/1 approximant structure of Mg27Al11Zn47 [2], with only difference in that the additional decorating atom on the prolate rhombohedron splits into two equal parts along 3-fold axis in Mg27Al11Zn47, whereas the present structure is free of configurational disorder. Noteworthy also is that the arrangements of triacontahedral clusters and prolate rhombohedral clusters in Bergman-type 2/1 approximants are exactly same as those in Tsai-type cubic 2/1 approximants in the Sc-Mg-Zn and Ca-Au-T (T = In, Ga, Sn) systems [3], regardless the different short-range-orders (geometries and decorations) within the so-called Bergman- and Tsai-type clusters. [1] Bergman, G.; Waugh, J. L. T.; Pauling, L. Acta Crystallogr. 1957, 10, 254. [2] Lin, Q.; Corbett, J. D. Proc. Natl. Acad. Sci. USA 2006, 103, 13589. [3] Lin, Q.; Corbett, J. D. Struct. & Bond. 2009, 133, 1.

Surface structures often result from surface energy minimization. When dealing with a simple metal, this leads to the emergence of preferred surface orientations, and for a given (h,k,l) surface, interlayers relaxations and possible surface reconstructions. In metallic alloys, the introduction of different chemical elements offers additional mechanisms to minimize the surface energy, through a chemical surface segregation or by the selection of specific planes. In the latter case, the possible covalency of bondings might play a role too. In this work, the surface structures of some intermetallics, with different complexities among complex metallic alloys, are investigated by both (i) experimental methods under ultra-high vacuum - low-energy electron diffraction, photoemission spectroscopy, scanning tunneling microscopy - and (ii) first-principles calculations. The relation between the bonding covalency and the surface structure will be discussed on various examples. Among them, we will show (i) that the unusual reconstruction on the Al₂Cu(001) surface results from the covalent character of bondings highlighted in the description of bulk Al₂Cu as a system of interpenetrating graphite-like aluminium 6³ net with copper atoms in the channels between the nets [1] and (ii) that the plane selection in Al₉Co₂ can be related to the covalent character of the Al-Al bonding between the antiprism of the Al₉Co₂ bulk structure [2]. In conclusion, we will discuss these results in relation to previous observations at quasicrystalline surfaces. [1] L. N. Serkovic Loli et al., Phys. Rev. Lett. 108 , 146101 (2012). [2] S. Alarcon Villaseca et al., J. Phys. Chem. C 115 (30) , 14922 (2011).

Over the last 20 years, intensive efforts have been dedicated to the study of quasicrystal surfaces [1]. With only few exceptions, a consensus has emerged that quasicrystal surfaces should be considered as bulk-terminated. The selection of bulk planes as surface terminations obeys experimentally established rules. Compared to surfaces of “classical” crystalline materials, aperiodic surfaces offer a wealth of structurally and chemically inequivalent sites. Their particular electronic structure and topography have led to unusual growth phenomena [2]. Now, the successful synthesis of single grains of approximants allows the direct comparison of structural and growth phenomena on related periodic and aperiodic surfaces. Here, we will report on the recent adsorption works performed on the orthorhombic Al13Co4 (100) and the monoclinic Al13Fe4 (010) surfaces for various substrate temperatures. Both complex metallic alloys are approximant to the decagonal quasicrystals. For the Al-Co system, Pb adatoms remain highly mobile at 300K and adsorb preferentially within the hollow site situated in between adjacent Al pentagonal clusters present at the surface. These experimental findings are supported by ab initio calculations based on density-functional theory (DFT). For two temperature regimes (300K and 573K), Pb adsorption leads to the formation of pseudomorphic monolayers above which the high adsorbate mobility prohibits the growth of additional layers. For the high temperature deposition, we propose a structural model for the Pb film and discuss its relationship with the underneath substrate. For the Al-Fe system, we have characterised Pb adsorption at 300K from submonolayer to multilayer regime. While the initial stages of adsorption differs to the Al-Co system, the deposition leads to the formation of a pseudomorphic monolayer. The resulting atomic structure of the Pb film will be discussed in relation to the arrangement of the substrate bipentagonal motifs. Finally, we will report on our latest work on molecular adsorption on the Al9Co2 (001). [1] R. McGrath, J.A Smerdon, H.R Sharma, W. Theis and J. Ledieu, J. Phys.: Condens. Matter 22 (2010) 084022. [2] V. Fournée and P.A. Thiel, J. Phys. D: Appl. Phys. 38 (2005) R83-R106.

Film growth by deposition of metals on quasicrystalline surfaces has been motivated in part by the possibility of transferring the quasiperiodic structure of the substrate template into the growing film. Modeling is complicated by the complex non-periodic structure of the template, although DFT analysis provides partial insight into the PES for adsorbate binding. However, a key challenge is to describe the kinetics (not just the adsorption thermodynamics) of these non-equilibrium growth processes in simulations on the relevant time- and length-scales. For submonolayer growth of 2D pseudomorphic islands, it is viable to extend standard atomistic lattice-gas modeling to incorporate the quasi-periodic array of adsorption sites describing the diffusion paths between adjacent sites by a “disordered-bond-network”. Model input involves distinct adsorption energies for each site, hopping barriers between all neighbors, as well as interactions between adspecies. Atomistic-level modeling is difficult for multilayer growth, but instead coarse-grained step dynamics modeling is viable where one tracks just step edges in each layer. This approach can incorporate surface diffusion-mediated growth, step edge attachment barriers, layer-dependent energetics (including QSE), and strain effects.

Gadolinium, and alloys of Gd with various post-transition elements, are considered promising materials for magnetic refrigeration near or below room temperature. Gd₅Ge₄, an orthorhombic system with 36 atoms per unit cell, is in this group. We have characterized the (010) surfaces of two different samples of Gd5Ge4 using scanning tunneling microscopy and x-ray photoelectron spectroscopy. Although there are differences between data from the two samples (especially in the progression of surface structure as a function of temperature), they have the following features in common: (1) Surface composition is the same as bulk composition to within + 5 atomic %, both after sputtering and after annealing at various temperatures; (2) The surface exhibits two main types of terraces, which have much different fine structures. The two types of terraces alternate across the surface. Under some conditions they have comparable areas and the alternation is visually striking. Under other conditions a single type dominates, but the second type is still discernible as small protrusions at step edges. Step heights between similar terraces correspond well to the separation between equivalent layers along the <010> direction in the bulk structure, and step heights between dissimilar terraces correspond to the distance between certain dissimilar but dense layers. However, the fine structure on the terraces does not correlate with the bulk structure, and we suggest that there is significant surface reconstruction.

Complex compounds can be found in a large variety of intermetallic systems. They have unit cells containing hundreds or even thousands of atoms [1]. Several quasicrystal approximants and some modulated structures can be found among these compounds. Highly-symmetric lattices, such as cubic [2] and hexagonal, seem to be preferred by complex intermetallic compounds. In order to better understand complex intermetallics, we describe their structures in different ways, for example in terms of the cluster approach, as superstructures, or as layered compounds. When these compounds are interpreted as cluster packings, two recurring structural motifs are most prominent: multiple-shell fullerene-like clusters packed closely in different ways and blocks of smaller polyhedra (mainly Friauf polyhedra and similar clusters) filling the remaining gaps. Another general property of complex intermetallics is a complex diffraction pattern, which contains a small subset of strong reflections, indicating that the structure can be regarded as a superstructure. The clearly visible layered structure that many intermetallics exhibit also reflects the periodicity of the underlying basic structure of such a superstructure description. By comparing structures which are geometrically similar but differ chemically, we hope to reveal the building principles responsible for the formation of highly complex intermetallic structures. On the other hand, we observe the occurrence of multiple very complex phases in some intermetallic systems. We compare these compounds of similar stoichiometry using the mentioned complementary methods of structure description, in order to relate their different geometries to one another. [1] Weber, T., Dshemuchadse, J., Kobas, M., Conrad, M., Harbrecht, B., Steurer, W. (2009). Acta Cryst. B65, 308-317. [2] Dshemuchadse, J., Jung, D. Y., Steurer, W. (2011). Acta Cryst. B67, 269-292.

Recent exploratory syntheses of polar intermetallic compounds containing gold have established gold&’s tremendous ability to stabilize new phases that exhibit diverse and fascinating structural motifs. In particular, Au-rich polar intermetallics contain Au atoms condensed into tetrahedra, rods of hexagonal stars, and diamond-like three-dimensional frameworks. In Au-poor intermetallics, on the other hand, Au atoms tend to segregate which leads to structural transformations or superstructures in Au-containing gamma brasses. Lastly, for polar intermetallics with intermediate Au content, complex networks of icosahedra have emerged, including discovery of the first sodium-containing, Bergman-type, icosahedral quasicrystal. Gold&’s behavior in this metal-rich chemistry arises from its high electronegativity among metals, the significant relativistic effects on its valence orbitals, its low valence s, p electron count, as well as its filled 5d orbitals. Thus, the structural versatility of gold and the accessibility of various Au fragments within intermetallics are opening new insights toward elucidating the relationships among atoms, clusters, and bulk solids. This presentation will summarize our group&’s recent efforts on the synthesis and characterization of several Au-containing intermetallics.

An emerging theme in the study of intermetallic compounds is a link between structural complexity and the coexistence of mutually exclusive bonding or packing modes. In this presentation, we will describe new theoretical tools developed by our research group for analyzing such frustration in crystal structures and providing explanations for their geometrical features. Methods to be discussed will include DFT-calibrated Hückel models, the DFT Chemical Pressure analysis, and the mu;3-acidity analysis.

The electron localizability indicator (ELI) is a key quantity used in determining the features of chemical bonding in molecules and solids. When combined with the electron density (ED), it enables a real space multifacetted description of chemical bonding providing information on local interactions. Therefore, this ELI/ED approach is an ideal method for studying the nature of atomic interactions in intermetallic compounds with large and complex crystal structures. Chemical bonding in two groups of compounds are investigated using highly accurate first-principles electronic structure calculations with an emphasis on ELI/ED approach. One of the groups consists of aluminum-rich binary compounds with Co and Fe, which are considered as approximants of the quasicrystalline decagonal phases. The analysis shows that two-center bonds implying covalent interactions are found only for the shortest contacts, and that the nature of bonding evolves from mainly covalent towards mainly metallic (multi-center) as the number of valence electrons increases. The other group is based on type I clathrate compounds of composition Ba₈T_xE_46-y (usually with framework vacancies) where T is Li, Mg, Ga or a transition metal (groups 9-12), E is Si or Ge, and 0 le; x,y le; 6. Calculations for some members of this group require supercells containing of the order of 200 atoms in the primitive cell. The analysis reveals two-center T-E interactions in all cases, and the degree of the polarity of these T-E covalent bonds can be quantified by computing the individual T and E contributions to the electron populations of the bond basins. We find that in addition to the electronegativity difference, the local geometry (the coordination) also contributes to the polarity of a covalent bond. For the cases T = Cd, Ag, Pt, Au, T-Ba interactions are found, as well, which may have implications for some transport properties.

Quasicrystalline precipitates in ZrAlNiCuNb alloy [1] were systematically studied by transmission electron microscopy. Electron diffraction patterns show that precipitates always contain various linear phason strains [2, 3], caused by which the diffraction spots&’ shifting can be divided into two types, parallel and perpendicular to the incident beam direction. The latter one can be subdivided into transverse and longitudinal shifting. Experimentally observed spots&’ shifting induced by linear phason strains can be parallel, transverse, longitudinal or mixed. After meticulous measurement, quantitative fitting and calculation of the diffraction spots&’ shifting, the phason strain matrices were determined. Within some single grains, nanoscale domain structures formed as a result of linear phason strain variants along different directions with equal probability. These phason strain variants can counterbalance each other to produce low macroscopic strain for the whole grain. In a unique case, phason strain induces subtle opposite displacements along the incident beam direction, leading to the breaking of central symmetry in the electron diffraction pattern. This study will enrich the knowledge of linear phason strain and may also deepen the understanding about the intrinsic characteristic of quasicrystals and its related bulk metallic glass matrix composites. [1] Y.F. Sun, C.H. Shek, B.C. Wei, W.H. Li and Y.R. Wang, J. Alloy Compd. 403, 239 (2005). [2] P.A. Bancel and P.A. Heiney, J. Phys. Colloques 47, C3-341 (1986). [3] J.E.S. Socolar and D.C. Wright, Phys. Rev. Lett. 59, 221 (1987).

The discovery of stable binary i Cd-Yb and i Cd-Ca QCs by Tsai et al.[1] has enabled us to obtain i-QCs and approximants(APs) structurally as well as chemically ordered. It has now become possible to investigate the influence of chemical disorder and local distortion induced by chemial disorder, on the structural and physical poroerties of the QCs and APs. Subsequent works have shown an occurrence of a unique low-temperature structural transition associated with a group motion of atoms, i.e., a tetrahedron, inside an icosahedral cluster in a series of Cd6R(R=Rare-earth) APs[2]. In addition to the structural transitions, the Cd6R APs also exhibit long-range magnetic orders at low temperatures: Successive antiferromagnetic/ferrimagnetic transitions were observend in Cd6Tb[3] and Cd6Sm[4], respectively, showing different behaviors depending on the type of the R element. In this talk, we first present the recelntly determined low-temperature structures of the Cd6R APs and then describe the magnetic transitions of various Cd6R with respect to the low-temperature structures and the de Genne factor of the R element. The influence of the phase transition on the magnetic ordering will be also addressed. [1] Tsai et al., Nature 408, 537 (2000).[2] Tamura et al., Jpn. J. Appl. Phys. 41, L524 (2002).[3] Tamura et al., Phys. Rev. B 82, 220201(R) (2010).[4] Tamura et al., Phys. Rev. B 85, 014203 (2012).

Little is known about the influence of hydrogen on structure and magnetic properties of TiZrNi quasicrystals, while much attention has been focused on hydrogen storage in Ti-based quasicrystals. Pieces of Ti, Zr and Ni with purities of higher than 99.9% were arc-melted in an argon atmosphere. The nominal compositions of TiZrNi quasicrystals are 53, 27 and 20 for Ti, Zr and Ni, respectively. The magnetic moment and magnetic hysteresis of hydrogen absorbed Ti53Zr27Ni20 quasicrystal were evaluated by using a vibrating sample magnetometer. Hydrogen decreases the magnetic moments of both Pd added and pure quasicrystals, while its contribution is larger than that of Pd. As increasing the amount of hydrogen absorption, the maximum magnetization values of Ti53Zr27Ni20 quasicrystals significantly decreased after hydrogenation. It is believed that the decreased magnetization values for the hydrogenated samples are attributed to the reduced interactions of magnetic dipole moments between Ni atoms arose from the expansion of the quasilattice constants after absorbing hydrogen in interstitial sites.

State of the art magnets for interior traction motors applications are based on Nd2Fe14B. Although this compound is adequate for vehicle applications, the natural abundance of Nd is insufficient to meet the market requirement. It is therefore desired to look for a similar material with a better natural abundance and enhanced performance/properties. We have performed a first principles calculation on a pseudo-ternary (R,Ce)_{2}Fe_{14}B [R=Nd,La] compound, looking for improved properties. Due to the highly localized f-electrons of the rare-earth elements in this compound, we have used a density functional theory based GGA+U approach to properly capture the concerned effects. We have first studied the phase stability of this compound, showing a partial disordered phase to be energetically the most favorable one. Interestingly, the Ce-substitution indicates a site-preference towards one of the two rare-earth (R) sublattices. This may help in improving the magnetic properties via the exchange coupling of hard and soft phases, if carefully controlled during experiments. We shall also show the sensitivity of some of the magnetic properties both on the Fe-sites and the R-sites as the additional Ce-atom is doped.

Considerable efforts have been devoted to bimetallic nanoparticles owing to their enhanced catalytic properties, surface plasmon band energy, and magnetic properties. However, materials that are derived from core-shell particles are of extensive scientific and technological interest, due to their unique and tailored properties for various applications in materials science. Creating heterostructured nanoparticles is an attractive way to design systems possessing diverse physical and chemical properties. It has recently been demonstrated that a unique electronic interaction takes place between gold and silver in the core@shell (Au@Ag) structure, which results in enhanced resistance to oxidation for the Ag. Such a phenomenon allows a greater degree of control in creating highly active NPs with enhanced properties and stability, for example as bio-molecular probes. Building on the phenomenon, Au@Fe core@shell NPs were synthesized to study the resulting electronic properties and determine whether the Fe would also gain enhanced resistance to oxidation. In addition to the unique electronic properties induced by Au, the Fe is expected to possess unique magnetic properties as well, which have applications in magnetic separation, MRI contrast agents, targeted drug delivery as well as many others. Au NPs were first synthesized in aqueous medium via the well-known citrate reduction method. These Au NPs were used as monodispersed seeds for the further deposition of an Fe shell. The Fe was essentially grown on the surface of the Au NPs via seed mediated growth to form Au@Fe NPs. The resulting NP morphology and structural properties were studied using TEM as well as STEM-HAADF and elemental mapping techniques, revealing a discrete core@shell structure. Finally, XPS was used to gain a fundamental understanding of how the core@shell structure impacts the resulting electronic properties for Au and Fe in the NPs. This presentation will discuss the synthetic approach as well as the unique electronic properties of the Au@Fe NP system. The results are discussed in terms of synthesis technique, and characterization of the optical/magnetic properties using techniques such as SQUID, TEM, elemental mapping, XRD, XPS, UV-Vis as well as many others.

Metadislocations are considered the most complex defects in materials science. They are pivotal defects mediating plastic deformation in complex metallic alloys (CMAs). Due to their size, a single step of movement implies the coordinated jumps of hundreds of atoms per elementary step. To date, the underlying atomic mechanisms are far from being understood. In this contribution, recent results on the structure of metadislocations involving novel mechanisms of plastic deformation are presented. They were obtained using aberration-corrected high-angle annular dark field scanning transmission electron microscopy (HAADF-STEM). A new deformation mechanism was found in the CMA phase T-Al-Mn-Pd with 156 atoms per unit cell. It is based on the movement of a metadislocation core mediating strain and separate escort defects. Upon deformation, the escort defects move along with the dislocation core and locally transform the material structure. Although the mechanism is very complex, it can be described by a simple jigsaw-puzzle-like rearrangement of basic structural subunits. Furthermore, the atomic structure of a metadislocation core in the CMA phase ε6-Al-Pd-Mn was investigated. The core is highly ordered, consists of about 1500 atoms and is based on atomic clusters. It possesses a multiscale structure with an inner core comprising the strain of the metadislocation and an outer core which is not strained but has a modified cluster structure with respect to the bulk. Based on this structural data, a first atomic model for metadislocation motion is presented which involves the coordinated atomic movement of hundreds of atoms by means of characteristic repetitive jump patterns.

Quasicrystals have some exotic properties such as high strength and hardness, implying to have a potential to be used for structural applications. However, due to their brittleness, quasicrystals themselves are difficult to be used as structural materials. Instead, quasicrystal can be one of the good candidates for reinforcing phase in structural composite materials. To be used for this purpose, there are some conditions: 1) quasicrystals should be thermally stable; 2) there should be wide two phase region of quasicrystal and matrix crystalline phase; and 3) the interface between quasicrystal and matrix crystalline phase should be strong enough. In the present study, the role of icosahedral (I-) phase in enhancing mechanical properties will be presented in Mg-Zn and Mg-Sn based systems. I-phase in Mg-Zn-RE composites has some unique intrinsic properties such as orientation relationship with α-Mg matrix and low interfacial energy with matrix. As a result, I-phase particle reinforced in situ composites provide beneficial microstructural characteristics such as refinement of grain size, suppression of basal texture development and high thermal stability, which lead to combination of high strength and ductility and enhancement of formability and high temperature properties. High strength of Mg-Sn based wrought alloys is generally acquired by the presence of precipitates formed during extrusion process or artificial aging treatment. However, little attention has been given in investigation on the proper selection of secondary solidification phase in achieving high strength level in Mg-Sn based alloys. Recently, we found that the secondary solidification phase changes from Mg2Sn and MgZn2 to Mg2Sn and Mg44Zn15Al41 I-phase when the content of Zn and Al is properly adjusted. As a result, we developed high strength Mg-Sn-Zn-Al wrought alloy exhibiting high level of yield strength in as-rolled state. The high level of strength is attributed to the presence of finely distributed Mg2Sn and Mg44Zn15Al41 I-phase particles embedded in the matrix.

Deformation in Complex Metallic Alloys, such as Al13Co4, occurs readily above a ductile to brittle transition temperature by dislocation motion on the (001)[010] slip system [1]. At lower temperatures, bulk material of this phase failed by fracture before plasticity could be observed. However it is generally thought that such dislocation motion requires a mixture of both glide and climb and that the onset of flow is associated with the increasing ease of climb at higher temperatures. The aim of this paper is to investigate whether deformation can occur at lower temperatures, and to establish the mechanism by which this occurs. Micropillar compression is used as a way of suppressing fracture, allowing the plastic deformation of Al13Co4 at temperatures below its ductile to brittle transition temperature to be studied. Micropillars of single crystal Al13Co4, with their orientation favouring slip on the active slip system observed in the ductile regime at higher temperatures were made by focused ion beam milling and compressed in-situ in the SEM at a range of temperatures below the brittle to ductile transition temperature. The micropillar compression results are used in combination with nanoindentation and transmission electron microscopy to understand the underlying deformation mechanisms and relate the results to the high temperature data measured on the bulk material of this phase. 1 M. Heggen, D. Deng, M. Feuerbacher (2007) Intermetallics 15 (11) p. 1425-1431

Dilute Mg alloys containing a few atomic per cent of TM (Transition Metal) and RE (Rare Earth) have attracted great attention because of their excellent mechanical properties. The key microstructural feature common for Mg-TM-RE alloys is the formation of novel type of long period polytype structures [1]. They are unique long period stacking structures synchronized with chemical order; a long period stacking/order (LPSO) structure. Z-contrast STEM observations of the LPSO structures clearly showed up chemical order along the stacking direction, as monitored by significant TM/RE enrichment at the particular close-packed layers relevant to faulting of the original 2H type stacking of the hcp structure [1]. For the well-ordered LPSO phases, the chemical order is further developed within the TM/RE-enriched close-packed layers. This is well represented by Zn₆Y₈ clusters with L1₂ type order [2] that are successfully embedded in the local fcc stacking layers in the LPSO. We attempted structure optimizations based on first-principles calculations (VASP), showing significant relaxation of the Zn₆Y₈ clusters. We further investigate local structural details of the Mg-Zn-Y LPSO phases, based on both the theoretical calculations and STEM experiments. Looking carefully the relaxed configuration of the Zn₆Y₈ cluster, we found that significant displacements of Zn/Y atoms cause formation of internal “voids” at the center of clusters, which are capable for placing one extra atom. On this basis, we put an extra atom X at the center of the Zn₆Y₈ cluster to construct modified Mg₁₄₀Zn₁₂Y₁₆X₂ models, and evaluate the phase stability for all the possible extra atoms (i.e, X = Mg, Zn, Y). From the VASP calculation results, for all the elements, Mg, Zn, Y, the Mg₁₄₀Zn₁₂Y₁₆X₂ structures are found to be significantly stabilized from the original structure. The above theoretical implication can be tested by STEM experiments, including a unique ABF imaging [3] as well as HAADF imaging. Since the present LPSO structures are rather disordered due to the incommensurability [2], we used template-matching procedure (cross-correlations basis) to reconstruct the detailed intensity map relevant to the Zn₆Y₈ clusters. In the averaged ABF image, there appears a fairly dark dot at the center of Zn₆Y₈ clusters, which verifies the existence of Mg atoms at the relevant position. This means that, in order to compensate the void caused by significant cluster relaxation, the extra atomic sites have been generated within the original close-packed structure, leading to a modified Mg₁₄₂Zn₁₂Y₁₆ structure model (14H-type). All the present procedures clearly demonstrate that a combination of advanced STEM and energy-based structural tuning is quite powerful for determination of complex intermetallic structures. References [1] E. Abe et al., Philo. Mag. Lett., 91 (2011) 690-696. [2] D. Egusa and E. Abe, Acta Mater., 60 (2012) 166-178. [3] R. Ishikawa et al., Nature materials, 10 (2011) 278-281.

Within recent years, complex metallic alloys (CMAs), intermetallic compounds with giant unit cells, comprising up to more than a thousand atoms per unit cell, have attracted much attention ranging from scientific curiosity about their complex structure, physical and mechanical properties to technological aspects of preparation and potential applications. CMAs exhibit several attractive properties for engineering applications, such as high strength-to-weight ratio, good oxidation resistance and high-temperature strength. However, as with most of intermetallic phases, one major drawback for their use in engineering applications is their limited plastic deformability at room temperature. A way to improve the room temperature plastic deformability is the development of a heterogeneous microstructure combining a soft metallic matrix with second-phase CMA particles. The CMA particles act as strength-bearing component, while the metallic matrix supplies ductility. In this overview, selected examples of the mechanical deformation behavior for multi-phase materials containing intermetallic compounds having different structural complexity ranging from Laves phases, to complex metallic alloys and quasicrystals (QCs) are presented, revealing that the development of a heterogeneous microstructure combining intermetallic particles with a ductile matrix phase, can be employed to produce materials with promising properties in terms of strength as well as of room temperature plastic deformation.

The zinc-rich domain of the Fe-Zn system between 67 and 95 at.% Zn has been intensively studied in the past [1-4], owing to its technological importance in the field of anticorrosion techniques. It comprises several different phases, among which the δ phase exists in the homogeneity range between 86.5 and 92 at.% Zn (at 300 oC). Recently, a structural model has been elaborated for the composition FeZn10 [5], showing that δ-FeZn10 crystallizes in the hexagonal space group P63/mmc (No. 194) with the lattice constants a = 1.2787 nm and c = 5.7222 nm and 556 atoms in the giant unit cell (52 Fe and 504 Zn). The unit cell contains 47 Zn and 5 Fe crystallographic sites, seven of which are fractionally occupied. The structure can be viewed as very dense packing of four types of polyhedra, (1) a statistically disordered distorted icosahedron spinning around a 3-fold axis inside a large Euler polyhedron, (2) a Frank-Kasper 16-vertex polyhedron, also referred to as the icosioctahedron, (3) a more or less regular icosahedron and (4) a bicapped pentagonal prism. The seven fractionally occupied sites are all located on the statistically disordered distorted icosahedron and introduce randomness and intrinsic disorder into the structure. Physical properties of the δ-FeZn10 phase remain unknown and we have determined its electrical, magnetic and thermal properties (the electrical resistivity, the magnetoresistance, the thermoelectric power, the Hall coefficient, the specific heat and the thermal conductivity). The temperature-dependent electrical resistivity exhibits a maximum at 220 K, which can be explained by the theory of slow charge carriers. Magnetoresistance increases approximately linearly with the magnetic field up to 9 T and amounts about 1.9 % at the lowest investigated temperature of 2 K and 9 T field. magnetic study has shown that the δ-FeZn10 phase remain paramagnetic down to 2 K with an antiparallel coupling between the spins. The thermopower is small and negative, whereas the Hall coefficient is positive, indicating complex structure of the Fermi surface. The specific heat indicates a slight increase of the electron thermal effective mass at low temperatures. Thermal conductivity originates predominantly from the conduction electrons. [1] J.K. Brandon, R.Y. Brizard, P.C. Chieh, R.K. McMillan, W.B. Pearson, Acta Crystallogr. B 30 (1974) 1412. [2] O. Kubaschewski, Phase Diagram of Binary Iron Alloys (ASM International, Materials Park, OH, 1993) p. 459. [3] G.F. Bastin, F.J.J. van Loo, G.D. Rieck, Z. Metall. 65 (1974) 656. [4] G.F. Bastin, F.J.J. van Loo, G.D. Rieck, Z. Metall. 68 (1977) 359. [5] C.H.E. Belin, R.C.H. Belin, J. Solid State Chem. 151 (2000) 85.

Because of their complex crystal structure and peculiar physical properties, quasicrystals are an unusual type of metallic alloys and they keep attracting great attention since their discovery in 1982. Their transport properties are defined by the local structure complexity rather than long-range quasicrystalline order. Decagonal quasicrystals are a special class of quasicrystals that have quasicrystalline planes (QCP) of decagonal symmetry which are periodically stacked along the direction perpendicular to QCP. Among the most investigated decagonal quasicrystals is d-AlCoNi phase. It has two quasicrystalline layers of atoms along the stacking direction and shows highly anisotropic transport properties.[1] Theoretical work on quasicrystals is limited due to the lack of periodicity, and approximant phases are of major importance here, because they have local atomic order very similar to that of quasicrystals, still being periodic in all directions. Crystal structure of decagonal approximants can be viewed as periodic stacking of quasidecagonal planes. Here we present transport properties (thermal conductivity, electrical resistivity and thermopower) of decagonal quasicrystal d-AlCoNi,[2] and approximant phases Y-AlCoNi,[3] o-Al13Co4,[4] m-Al13Fe4,[5] m-Al13(Fe,Ni)4[5] and T-AlMnFe[6]. For all alloys investigated the stacking direction (periodic for d-QCs) is the most conductive one for charge and heat transport, and the in/out-of-plane anisotropy is much larger than the in-plane anisotropy. There is a strong relationship between periodicity length along stacking direction and anisotropy of transport properties in both quasicrystals and their approximants, suggesting the decrease of anisotropy with increasing number of stacking layers.[7] Structural quality appears to be highly important, because small amount of disorder can substantially influence transport properties, as shown. Thermal transport properties of quasicrystals and their approximants are especially interesting because of various mechanisms that take place in different temperature regimes. References: [1] M. Bobnar, et al, Phys.Rev. B, 85 (2012) 024205. [2] P. Pop#269;evicacute;, et al, Isr. J. Chem., 51 (2011) 1340. [3] A. Smontara, et al, Phys. Rev. B, 78 (2008) 104204. [4] J. Dolinscaron;ek, et al, Phys. Rev. B, 79 (2009) 184201. [5] P. Pop#269;evicacute;, et al, Phys. Rev. B, 81 (2010) 184203. [6] M. Heggen, et al, Phys. Rev. B, 81 (2010) 184204. [7] P. Pop#269;evicacute;, et al, in preparation.

Crystalline high-entropy alloys (HEA) are the thermodynamically stable counterparts to bulk metallic glasses. HEA are approximately equiatomic single-phase solid solutions of four or more different metallic elements, which crystallize in rather simple average structures such as bcc cP2-W. The random lattice distortions resulting from the significantly differing atomic diameters lead to high hardness and strength that can persist even to high temperatures. Contrary to amorphous HEA (i.e., bulk metallic glasses), which have been intensively studied for decades, the investigation of their crystalline counterparts started less than ten years ago (Huang et al., 2004, Adv. Eng. Mater. 6, 74; Yeh et al., 2004, Adv. Eng. Mater. 6, 299). The rationale behind this new research field was that the high mixing entropy would favor random solid solutions in simple structures rather than the formation of multinary complex intermetallic compounds. The more than 250 papers published so far (about 200 of them within the last three years) mainly deal with the microstructure and properties of HEA, in particular high-temperature oxidation resistance and mechanical properties (for a review see Yeh at al. 2007, Mater. Sci. Forum 560, 1). We will present first results of our studies of high-melting HEA, mainly constituted of the refractory metals W, Mo, Nb, Ta, V, Hf, Zr and Ti. Refractory-based HEA have potential applications in the aerospace industry, particularly, where there is a need for structural materials with sufficient strength at high temperature (see also Senkov et al. 2011, Intermetallics 19, 698).

We have constructed a unified picture for B- and Al-based icosahedral cluster solids [1]. The first of common features are as below; they should be structurally complex materials such as giant unit cell crystals or quasicrystals because of inconsistency between the icosahedral symmetry and the periodicity. The above first feature induces the self-compensation phenomena. The crystalline structure of beta-rhombohedral boron (0/1-2/0 approximant) is constructed from the electron deficient B12 icosahedra and the electron excess B28 clusters, and the former is compensated by the interstitial boron atoms (max 25%) and the latter is compensated by the vacancies at the certain site (also max 25%). When electrons are doped from Li or Mg atoms in the doping sites, first the interstitial boron atoms are removed, and second the vacancies are introduced at the other site, to compensate the electrons [2]. This is also considered to be covalent-ionic bonding conversion. The similar self-compensation phenomena have been reported in the approximant crystals of Al-based icosahedral quasicrystal. In Al-Mn-Fe-Si 1/1-cubic approximants, an increase in the valence electrons due to the replacement of Mn atoms by Fe atoms is compensated by an increase in the number of vacancies in the glue sites [3]. Al-based icosahedral quasicrystals and approximant crystals are structurally complex materials and have very wide distribution of bonding strength form strong covalent to weak metallic in the same solid [4]. The intra-cluster bonds are rather stronger than the inter-cluster ones. They are considered to be intermediates between typical metallic, covalent and molecular bonding solids. The guiding principle to increase thermoelectric performance, weakly bonded rigid heavy clusters (WBRHCs), has been proposed [5]. According to the guiding principle, the thermoelectric figures of merit for Al-Pd-Re and Al-Pd-Mn quasicrystals increase by substituting Re with Ru or Fe, or Al with Ga [6-8]. WBRHCs also should widen the pseudogap to a real gap and should increase the Seebeck coefficient like a narrow gap semiconductor Al2Ru or Ga2Ru. A quasicrystalline Semiconductor should be a candidate for good thermoelectric material. We have searched a quasicrystalline semiconductor and found a B-Ti-Ru decagonal quasicrystal [9]. [1] K. Kimura et al., J. Solid State Chem. 133, 302 (1997). [2] H. Hyodo et al., Solid State Sci., (2012) in press. [3] T. Takeuchi et al., Mater. Trans. 42, 933 (2001). [4] K. Kirihara et al., Phys. Rev. B 68, 014205 (2003). [5] K. Kimura et al., Mater. Res. Soc. Sym. Proc. 886, F06-10 (2006). [6] T. Nagata et al., J. Appl. Phys. 94, 6560 (2003).. [7] J. T. Okada et al., J. Appl. Phys. 101, 103702 (2007). [8] Y. Takagiwa, et al., J. Electron. Mater. 39, 1885 (2010). [9] Y. Miyazaki, et al., J. Phys. Soc. Jpn. 79, 073601 (2010).

Complex intermetallic compounds are a topic of intense fundamental and application-oriented research. Cage compounds such as intermetallic clathrates or skutterudites are, with tens to hundreds of atoms per unit cell, structurally relatively simple representatives. However, vacancies and site disorder add to the complexity. Some of these materials have shown promising thermoelectric properties [1, 2]. Here I present recent results on transition metal substituted Ge- and Si-based clathrates. Financial support from the Austrian Science Fund (FWF project TRP 176-N22) and from the European Integrated Center for the Development of new Metallic Alloys and Compounds (C-MAC) is gratefully acknowledged. [1] S. Paschen, “Thermoelectric materials” in Mechanical Properties of Complex Intermetallics (World Scientific Publishing Co. Pte. Ltd., Singapore, ISBN 978-981-4322-16-4), p. 425-457 (2011). [2] S. Paschen, C. Godart, and Yu. Grin, “Recent progress in the development of thermoelectric materials with complex structures” in Complex Metallic Alloys (Wiley, Eds. Dubois, Belin-Ferre, Weinheim, ISBN 978-3-527-32523-8), p. 365-384 (2011).

Intermetallic clathrate systems are currently actively investigated due to their special thermoelectric properties. They are composed of periodically arranged cages, the host framework, which encloses single guest atoms. The low thermal conductivity of these structures has been attributed to the scattering of phonons on the local vibration modes ("rattling") of the guest atoms. For computational studies of dynamic properties long simulation times and large samples are required. This makes first principle calculations of these structures, even with high performance computers, very unfeasible. Classical molecular dynamics, however, can meet the requirements. The potentials needed for such MD simulations can be obtained from ab-initio calculations with the force-matching method. It uses large numbers of reference data to fit an effective potential that reproduces the forces, energies and stresses of ab-initio calculations. To model the strongly directional atomic interactions in clathrates, angular dependent potentials are required. We have fitted analytic effective potentials which yield results in good agreement with those of ab initio simulations; especially the phonon dynamics is well reproduced. Molecular dynamics simulations and the Green-Kubo method are used to study the thermal conductivity of germanium- and silicon-based clathrate systems. The results with these new potentials are compared to previous calculations with other three-body potentials and the influence of the size of the simulation box is presented. Structural complexity as a source of the low thermal conductivity and the contribution of the guest atoms will be discussed.

The ground-state energies of electron-doped BaFe_2-2xM_2xAs₂ (M = Co, Ni, Cu) for 0 < x < 0.18 in non-magnetic (NM), antiferromagnetic (AFM), and disordered local moment (DLM) configurations for both tetragonal (I4/mmm) and orthorhombic (Fmmm) structures are compared using density functional theory (DFT). Due to common-band behavior, we find Co and Ni doped compounds obey similar trends versus e/Fe, while Cu is manifestly different due to split-band behavior. At low e/Fe, the DLM is the preferred high-T phase, with competing magnetic states beyond ~0.06 e/Fe for Co and Ni. We find that the calculated concentration for a transition from AFM to NM in the low-T phase agrees with experiment. We study the evolution of DLM and NM Fermi surfaces (both electron and hole) with doping; for NM case, Cu states do not exhibit rigid-band behavior and are dispersed.

Development in heterogeneous catalysis is often hindered by the complexity of the applied catalytic systems. While this is necessary for industrial use to maximise the surface area and stability, a valuable basic science approach is to make the catalysts “as simple as possible, but not simpler” (A. Einstein). A class of compounds not often investigated in catalysis are intermetallic compounds, which differ strongly from substitutional alloys, since they possess crystal structures different from the constituent elements. The more complicated structures result from the partly covalent bonding within the compounds, which also enhances their stability leading to high stability under reaction conditions and reduced segregation. The structural changes and the bonding situation result in much stronger changes of the electronic structure compared to substitutional alloys. Combining the “as simple as possible, but not simpler” idea with the huge potential of the intermetallic compounds leads to their unsupported use in heterogeneous catalysis. Due to their large number (more than 10,000 binary intermetallic compounds are known and many more with more than two elements) this opens a vast matrix to explore relationships between electronic and crystal structure on the one hand and the catalytic properties on the other. Uncovering these relationships leads to a deep understanding and the gained knowledge can be used to select intermetallic compounds with optimised crystal and electronic structure in the next step. This is named the knowledge-based approach [1]. This approach has been successfully applied in the development of a new class of highly selective semi-hydrogenation catalysts based on Ga-Pd intermetallic compounds. These compounds represent the first example, where the knowledge-based approach was applied starting from the bulk materials, studying their stability under reaction conditions, developing an approach to unsupported nanoparticulate materials and inventing an industrially feasible route to highly active materials while preserving the selectivity and stability of the bulk compounds [2]. Due to the knowledge gained during the investigations, the Pd-based systems could finally be replaced by more ubiquitous non-noble metal based materials like Al13Fe4 [3]. [1] K. Kovnir, M. Armbrüster, D. Teschner, T.V. Venkov, F.C. Jentoft, A. Knop-Gericke, Yu. Grin, R. Schlögl, Sci. Technol. Adv. Mater. 8, 2007, 420. [2] M. Armbrüster, K. Kovnir, M. Behrens, D. Teschner, Yu. Grin, R. Schlögl, J. Am. Chem. Soc. 132, 2010, 14745-14747. [3] M. Armbrüster, K. Kovnir, M. Friedrich, D. Teschner, G. Wowsnick, M. Hahne, P. Gille, L. Szentmikloacute;si, M. Feuerbacher, M. Heggen, F. Girgsdies, D. Rosenthal, R. Schlögl, Y. Grin, Nature Mater. 2012, DOI: 10.1038/NMAT3347.

Most experimental and computed phase diagrams for the Ti-O system indicate a very large oxygen solubility of up to 30% in hexagonal closed packed (HCP) Ti at low temperatures. Yet already much smaller amounts of oxygen on the order of just fractions of a percent are known to cause rather dramatic changes of mechanical properties, most notably ductility. Using a combination of first-principles calculations and Monte Carlo simulations based on lattice Hamiltonians, we have systematically investigated the Ti-rich end of the Ti-O phase diagram. The simulations predict three distinct Ti-O phases that are based on the HCP lattice: Ti6O, Ti3O, and Ti2O. The structures of these phases are in exact agreement with the results of neutron diffraction experiments that hitherto have been integrated incompletely and inconsistently into available phase diagrams. Using our approach we obtain a revised equilibrium phase diagram in the concentration range up to 33% oxygen. Our results show that the effective equilibrium solubility of O in Ti is fact less than 1% at room temperature, beyond which Ti6O precipitation should occur.