This talk will describe the results of computational studies of the energetics of twin boundaries in hcp transition metals and alloys. The calculations employ a recently developed method for modeling interfacial energies in substitutional alloys from first-principles calculations based on the special quasirandom structure approach. The calculations reveal anomalously low values for hcp metals and alloys with d-band filling near that of the group VII metal rhenium. The origins of this behavior are investigated and linked to the theory of the trends in bulk structural stability across the transition-metal series, which predicts the stability of topologically close-packed phases containing atomic coordination polyhedra similar to those found in the stable twins in this region of the periodic table. The results are discussed in light of experimental observations of deformation microstructures, and the implications for alloy design are described. This research was supported by the Office of Naval Research under Grant N00014-11-1-0886.

For magnesium and some other hexagonal-close-packed metals, twinning on the {10-12} plane is a common mode of plastic deformation. Recently, we have used in situ transmission electron microscope (TEM) to monitor the deformation of submicron-sized single-crystal magnesium, in quantitative compression and tension tests (B-Y. Liu et al., Nature Commun. 2014). We have observed the reorientation of the parent lattice to a “twin” lattice, producing an orientational relationship akin to that of the conventional {10-12} twinning. However, aberration-corrected TEM observations reveal that the boundary between the parent lattice and the “twin” lattice is composed of many segments of semi-coherent basal-prismatic (B-P) interfaces, instead of the {10-12} twinning plane. Both TEM and molecular dynamics simulations suggest that the migration of this boundary is accomplished by B-P interfaces undergoing basal-prismatic transformation, in addition to the migration of the boundary of the {10-12} extension twin. This deformation mode mimics conventional deformation twinning, but is distinct from the latter. It is a form of boundary motion coupled to stresses, but produces plastic strain that is not simple shear. The basal-prismatic transformation appears to be important under deformation conditions when the availability and/or mobility of twinning dislocations/disconnections are limited. As such, this interface-mediated lattice reorientation accompanying deformation twinning enriches the known repertoire of plasticity mechanisms.

Understanding the competition between twinning and dislocation slip is of significant importance for the fundamental understanding of plasticity in hexagonal-close-packed (HCP) materials. Towards developing such an understanding, a unified discrete dislocation dynamics (DDD) framework for modeling dislocations, twinning, and their interactions in magnesium (Mg) polycrystals has been developed. A systematic interaction model between dislocations and {10-12} tension twin boundaries (TBs) was introduced into the DDD framework, and a nominal grain boundary (GB) model based on experimental results was also introduced to mimic the GB&’s barrier effect. This new framework is then utilized to study grain size effects on the competition between dislocation slip and {10-12} twinning in the deformation of micro-twinned polycrystalline Mg. It is shown that twinning deformation exhibits a strong grain size effect; while dislocation mediated slip in untwinned polycrystals displays a weak one. This leads to a critical grain size at 2.7 mm, above which twinning dominates, and below which dislocation slip dominates. Furthermore, the predicted evolution of the twin boundary, the twin boundary velocity, and the morphology of the twin boundary are all compared to experimental results and predictions from molecular dynamics simulations.

Plasticity in zirconium, as well as in many other hexagonal close-packed (hcp) metals, is mainly due to the glide of dislocations with a Burgers vector along the axis. Those dislocations can however not explain any deformation along the axis. When submitted to a strain along the axis, the two possible and often competitive ways for the metal to accommodate deformation are glide of dislocations on the one hand and twinning on the other hand. At low temperatures
twinning can explain the main part of the observed plastic deformation whereas at higher temperatures dislocations appear and become predominant even if twinning is still active.
The present study focuses on the mechanisms controlling twin nucleation and growth. Atomistic simulations are performed using either an empirical interatomic potential of the EAM type or ab-initio calculations. The EAM potential is used to understand size effects due to elastic interactions between the different defects before performing ab-initio calculations.
The approach above is first applied to study perfect twin boundaries for the four twinning systems experimentally observed in zirconium. This study allow us to compare the structures and energies of the different twin boundaries.
The understanding and modeling of the nucleation and growth of twins in zirconium involves the study of disconnections. To this end we model disconnection glide along twinned planes. Studying this process with a similar combined EAM-ab inito approach allows us to calculate the energy variation of a disconnection dipole as a function of its length in order to extract the dipole core energy. We also use the NEB method to determine the energy barrier opposing disconnection glide. This study considers several possible disconnections to gather a broad knowledge of the conditions of twin nucleation and growth for different twinning systems.

HCP metals are widely used as structural materials in many industries, ranging from transport and energy to biomedical applications due to their low density, high specific strength. However, HCP metals show pronounced anisotropy in their mechanical properties and less number of slip systems compared to FCC and BCC metals. In the absence of sufficient number of slip systems in HCP metals, twining is found to be one of the important deformation modes during plastic deformation. At present, we do not have very clear knowledge of properties of twin boundaries in HCP metals at atomic length scale. Understanding atomic structure and chemistry of twin-associated boundaries is crucial to improve mechanical properties of HCP metals. In this work, using first-principles density function theory, we study twinning-associated boundaries (TBs), coherent twin boundaries (CTBs) and coherent basal-prismatic boundary (CBP) in six hexagonal metals (Cd, Zn, Mg, Zr, Ti and Be), with a focus on structure and solute&’s solubility at twin boundaries. We find that the formation of TBs is associated with creation of an excess volume. All the six metals show positive excess volume associated with and CTBs, but the excess volume associated with CTBs and CBP can be positive or negative depending on metal. To understand solubility at TBs, we calculated solubility of solute atoms in Mg, Ti, and Zr for solute positions in bulk, CTB and CBP boundaries and show that, in general, solute atoms have better solubility at CTB and CBP than in bulk. We also found solubility of solute atoms linearly changes with normal strain at CBP, increasing with the normal strain for solute atom with a greater metallic radius than the matrix, and decreasing with the normal strain for solute atom with a smaller metallic radius than the matrix. This suggests that the distribution of solute atoms in bulk, CTB, and CPB varies with stress state, and in turn affects mobility of CTB and CPB boundaries.

Recent studies have shown that doped interfaces can have intriguing structures and that, in some cases, thermodynamically-stable interfacial complexions can form. In this talk, we will explore the usage of nanoscale amorphous intergranular films as structural features that can alter mechanical behavior. Molecular dynamics and Monte Carlo simulations are used to show that amorphous grain boundaries can act as high-capacity sinks for dislocations, and to identify the processing conditions which promote the formation of such boundary structures. Experimental validation will be provided by high resolution transmission electron microscopy in conjunction with energy dispersive x-ray spectroscopy, showing segregation of Zr to the boundaries of Cu-Zr alloys created with mechanically alloying and providing evidence for the formation of amorphous films. Microcompression and in-situ bending experiments are then used to quantify the effect of doping on mechanical behavior, showing that strength, strain-to-failure, and failure mode can be controlled with segregation engineering.

Despite the good ductility exhibited by pure metals, their application as structural materials has been limited due to their poor specific strength. Strengthening methods such as solid solution, precipitation, dispersion, and grain-size reduction hardening have all been employed to improve the strength of metals. In particular, the reduction of grain-size to the nanometer scale has been widely used to produce both pure nanocrystalline metals and nanostructured alloys which exhibit ultra-high strength. In this study, we introduce novel Cu-based ternary nanostructured alloys with a unique microstructure. Cu-Zr-Ti and Cu-Zr-Al composites were synthesized by rapid solidification techniques. The composites show an ultra-fine distribution of three different phases: a ductile Cu crystalline phase; an intermetallic compound and nanocrystalline phase. Quasi-static uniaxial compression tests reveal yield strengths near 2 GPa, comparable or higher than those of Cu-base bulk metallic glasses. Additionally, extensive ductility over 10% was observed. The alloy's exceptional mechanical properties are the result of interface-dominated deformation, such as phase boundary sliding and crack deflection. These unique deformation mechanisms were investigated by use of scanning and transmission electron microscopy.

One strategy to retard grain growth in nanocrystaline alloys is the intentional addition of grain boundary segregating alloying elements. However, the segregation of impurities to grain boundaries is known to strongly affect mechanical behavior in bulk alloys, often leading to embrittlement. As grain sizes decrease, accompanied by increases in yield strengths, the cohesion of grain boundaries is expected to play an increasingly important role in determining overall mechanical properties. To design nanocrystalline alloys with enhanced mechanical performance, there is thus a need to systematically understand the effects of impurities on boundary cohesion. We systematically compiled a large database of existing ab initio simulations on the effects of impurities on grain boundary cohesion to determine global trends in the solutes&’ effects on boundary cohesion. Our analysis of the resulting database shows that the majority of the variation in impurities&’ effects on boundary cohesion is quantitatively predicted by simple bond breaking arguments, and that these trends are robust to the choice of solvent, and computational method. Implications for the design of nanocrystalline alloys with enhanced mechanical properties will be discussed.

Impact-based mechanical surface treatments are commonly used in industry to improve the service lifetime of materials through increasing the local mechanical properties in the near-surface. In this process, the material is exposed to repeated mechanical loadings, producing a severe plastic deformation and local refinement of the microstructure. Such Tribologically Transformed Surfaces (TTS) confer better local mechanical properties against wear or fatigue. In this project, two different procedures are implemented to obtain TTS surfaces on pure iron samples; shot-peening treatments and micro-percussion tests. For the first technique, the whole surface of the sample is impacted repetitively by metallic balls. For the second procedure, every impact is done on the same position by a rigid conical indenter, with control over the number, angle and velocity of the impacts. The main purpose of this work is to establish a complete description of the transformed microstructures for both techniques, and to understand the mechanisms involved on the formation and evolution of TTS and how it influences the material mechanical properties. EBSD crystal orientation mapping and HR-EBSD have been used to determine the size of the plastically deformed zone, the grain size distribution and to estimate the density of geometrically necessary dislocations in the cross section of the impacted zone. Nano-indentation and in-situ micro-pillar compression tests have also been performed to quantify the evolution of mechanical properties starting from the near-surface. A correlation between the grain size, dislocation density, and the mechanical properties in the deformed zone has been done and will be presented.

Metals and alloys used in industry are generally strongly heterogeneous since they include a wide variety of multi-scale physical mechanisms. For example, during the damage of thin-film deposited on substrates, micro-cracking and plasticity occurring at the nanoscale modify the evolution of buckles standing themselves at the microscale [1].
Numerically, in strategies consisting in describing the damage processes (cracks, pore formation, dislocationshellip;) and their effects on the evolution of structures, atomistic simulations usually fail to account for realistic systems with high enough space and time scales. Continuum methods, such as the phase-field thus appear as potentially well adapted alternatives.
Indeed, since the beginning of the last decade, these methods - already used in the study of phase transformations - have been extended to describe crystalline defects, such as cracks [2] and dislocations [3]. They are now used to solve problems involving cracks, voids and dislocations taken independently, often coupled with other kinds of fields related to other physical mechanisms. Recently, we have proposed a new model to explicitly couple cracks and dislocations, within a finite-strain continuum formalism [4].
In this presentation, the latest improvements of the actual model will be exposed (face-cubic centered symmetry, root-finding algorithmshellip;) and illustrated in the buckling context. In particular, we will show how interface plasticity modifies the shape of circular buckles and their destabilization. Since the model is general enough to address other damage problems, additional examples illustrating its potentialities should also be exposed.
[1] L.B. Freund and S. Suresh, Thin film materials: stress, defect formation, and surface evolution, Cambridge University Press, (2003)
[2] H. Henry and H. Lévine, Phys Rev Lett 93 (10), 105504, (2004).
[3] D. Rodney and A. Finel, Acta Mater 51, 17-30, (2003).
[4] A. Ruffini and A. Finel, Acta Mater 92, 197-208, (2015)

For plastic deformation at grain sizes around or below 10 nm, one of the central issues that remains to be resolved is to identify the relative importance of the variety of different possible deformation modes and assign and quantify in which manner the relevant/dominant modes contribute with rising stress to overall strain. In a case study on NC PdAu alloys with grain sizes below 10 nm, we determined activation energy, shear activation volume and shear dilatation to identify and discriminate deformation mechanisms. In-situ deformation and diffraction has been exploited to unravel the interplay of relevant mechanisms and their evolution up to large plastic strains. The unexpected observation of solid solution softening can be shown to has its origin in deformation modes related to grain boundaries and in particular to their inherent shear softening. Grain boundary mediated deformation seems to dominate the overall deformation behavior of metallic nanoscale microstructures at the low end of the nanoscale. Mechanistically, it bears great similarities with shear transformations, the generic flow defect of metallic glasses.

During last decades large experimental and modeling efforts have been made to investigate the rate-limiting deformation mechanisms in nanocrystalline metals. For grain sizes below 50nm these mechanisms usually involve grain boundary mediated processes, such a dislocation nucleation and interaction with grain boundaries, grain boundary sliding and grain boundary migration mechanisms. An important aspect that is often overlooked is the role of recovery. This, however, may play an important role in the constant deformation resistance that is often observed for nanocrystalline metals.
Transient testing has proven to be a suitable tool to gather information on the rate limiting deformation mechanisms that are activated during the deformation path. In particular in stress reduction experiments, after intermediate/large stress reduction thermally activated dislocation slip is suppressed so that recovery mechanisms are brought into foreground during subsequent transient creep. In this work, we combine this testing with in situ x-ray diffraction at the Swiss Light Source. Therefore, the transient responses are captured in terms of evolution of strain rate as well as diffraction peak broadening. The constant flow stress reached during uniaxial deformation of electrodeposited nanocrystalline metals reflects a quasi-stationary balance between dislocation slip and recovery mechanisms. The latter plays an essential role in producing plastic strain [1].
To gain further knowledge on the interplay between dislocation mechanism and grain boundary recovery mechanisms, molecular dynamics simulations (MD) have been performed. Stress reduction experiments have been simulated followed by creep simulations up to 1.7 ns. The simulations confirm the interpretation of the behavior of the peak broadening during in-situ diffraction experiments: after a stress drop strain is produced by grain boundary accommodation mechanisms. Depending on the magnitude of the drop, dislocation mechanism can start again after a certain amount of strain produced by accommodation.
REFERENCE
[1] Z. Sun, S. Van Petegem, A. Cervellino, K. Durt, W. Blum, H. Van Swygenhoven, Acta Materialia 2015; 91: 91-100

Stress-assisted grain growth in nanocrystalline metals transpires collectively with a number of competing deformation mechanisms including lattice and grain boundary mediated dislocation plasticity, grain boundary sliding, and grain rotation. While a number of pioneering studies have experimentally demonstrated mechanical grain growth with complementary simulations confirming simultaneous operation of the aforementioned mechanisms, questions remain concerning their role in facilitating grain boundary migration and its implications for plastic strain accommodation. In this study, molecular dynamics simulations of surface nanoindentation were performed to quantify the distribution of plastic strain among competing deformation mechanisms during stress-assisted grain growth in nanocrystalline nickel as a function of grain size and temperature. The role of segregated solute was also investigated using a Ni-1 at. % P alloy where the vast majority of the P atoms segregated to the grain boundaries following a Monte Carlo energy minimization procedure. Under identical conditions of rate and temperature in nominally the same grain size structure, stress-assisted grain growth found to be prevalent in pure nanocrystalline Ni was virtually absent in the Ni-P alloy. A reduction in the deformation temperature also quelled mechanical grain growth in both nanocrystalline Ni and Ni-P, suggesting thermal activation was inherent to the governing physics. The distribution of plastic strain among the disparate mechanisms was quantified using continuum deformation metrics. From this analysis, plastic strain was found to be highly localized in the grain boundaries during nanoindentation and dislocation activity, while present, did not represent the dominant carrier of plasticity. The addition of P reduced the magnitude of slip in the grain boundaries, thus confirming the ability of segregated solute to stabilize boundaries against mechanical grain growth via grain boundary migration.

The localization of stress at dislocation intersections with high angle grain boundaries in FCC polycrystals was studied through a combination of experimental methods and molecular dynamics modeling. Areas of highly localized stress are generally observed in the vicinity of the interaction area. Both, atomistic models and experiments show significant stress localization in the adjacent grain, decaying with distance from the grain boundary. The implications of these results for the transfer of slip to the adjacent grain are discussed.

Modern high-strength steels (Dual-Phase, low-alloyed TRIP) are generally multi-phase steels, in which the martensitic phase plays a key role.
In a multi-phase steel, the generally brittle martensitic phase clearly improves the strength but it may induce a detrimental effect on the ductility. However, recent literature has shown evidence of ductile fracture of martensite. The mechanism behind this ductile fracture is of great interest, since it may help to considerably improve the ductility of high-strength steels in general. The observed ductility may not be an intrinsic property of a martensitic lath, but rather a composite effect at the subgrain scale, where another phase may be retained. For this reason, the term 'apparent ductility' is used, since the small amounts of retained phases are difficult to observe and identify experimentally.
Using martensite's crystallographic substructure, this contribution aims to identify a physically based mechanism contributing to the apparent ductility of martensite. The proper description of the crystallographic characteristics at the micro-scale and the identification of small amounts of entrapped phases constitutes the basis of the computational model. A martensite grain reveals a well defined internal heterogeneous (crystalline) substructure, i.e. packets, blocks, laths. Thin layers of retained austenite may also be present between the laths, whereby small volume fractions (5\%) of interlath retained austenite may already explain the apparent ductility of martensitic subgrains. A conventional crystal plasticity is used to model austenite as FCC and martensitic laths as BCC phases. The entrapped thin FCC phase is shown to act like a greasy plane on which stiff BCC laths can slide. The role of the orientation relationship between FCC and BCC phases is shown to be of fundamental importance for the observed effect. The typical dimensions of martensite crystals and interlath austenite films range from hundred nanometers down to few nanometers. Therefore, Molecular Dynamics simulations have been carried out to analyse the small scale behaviour of a martensite-austenite bicrystal.
A two-scale framework results, in which the mesoscale behaviour is obtained by homogenization of the microscale bicrystal by means of a lamella rule of mixture. The model has been validated with experimental results on a martensitic steel. By accounting for the presence of interlath austenite, the main features of the experimentally observed deformation behaviour (stress-strain curve, slip activity, roughness pattern) are qualitatively well reproduced by the model. It is also shown that, by neglecting the presence of interlath austenite, the observed experimental stress-strain response is not captured. The presence of interlath retained austenite can remarkably enhance the local deformation of martensite. Finally, the potential influence of the thin layers retained austenite on the damage and fracture behaviour of martensite is explored.

Dual phase (DP) steels composed of ferrite and martensite are widely used in industries because of their high strength, adequate ductility and formability. It is expected that grain refinement can be an effective way to enhance mechanical properties of DP steels. In the present study, tensile behavior of a low-carbon DP steel with different grain sizes ranging from about 60 to 4 micrometers was investigated. The fine-grained DP steel was fabricated through a thermomechanical process including repetitive phase transformation and plastic deformation. It was found that the fine-grained DP steel had higher strength and larger total elongation than those of coarse-grained DP steel. In order to clarify details of the mechanical features in tensile tests, local strain distributions between ferrite and martensite phases were investigated by Digital Image Correlation (DIC) method. The strain value of the ferrite phase was always higher than that of the martensite phase due to the difference of the strength. On the other hand, it was found that the local strain distributions of two phases became more homogeneous in the DP steel having finer grain sizes. The results indicate that the change in local strain distribution in the DP steels with different grain sizes affects their macroscopic mechanical properties.

It has recently been reported that nano-crystalline pearlitic steel can exhibit an extreme tensile strength of up to 7 GPa, making it one of the strongest materials known in the world. Such excellent performance can be attributed to its lamellar microstructure, resulting in many interfaces that hinder the motion of dislocations. The combination of ferrite and cementite in this microstructure is a result of the lower C solubility in ferrite as compared to the high-temperature austenitic phase. However, an atomistic picture of its decomposition into ferrite and cementite at the cooling front and therefore the nucleation of the ferrite-cementite interface is so far missing. In this work we have identified a metastable intermediate structure (MIS) which allows a proper orientation relationship with austenite, ferrite and cementite, simultaneously. Our ab initio calculations suggest that the MIS acts as a buffer layer between austenite and ferrite and has the flexibility to evolve into all three phases triggered by elastic strain, magnetic transition and C diffusion. Since the MIS is structurally equivalent to a periodic arrangement of Σ3 twin boundaries in bcc Fe, it triggers the accumulation of C and consequently the nucleation of cementite. These insights, therefore, provide the mechanism for the formation of complex microstructures such as pearlite.

Conventional martensitic steels have limited ductility due to insufficient microstructural strain hardening and damage resistance mechanisms. In this work, a Fe-9Mn-3Ni-1.4Al-0.01C (mass %) medium-Mn TRIP maraging steel is produced and heat treated under different reversion conditions to introduce well-controlled variations in the austenite-martensite nano-laminate microstructure. Uniaxial tension and impact tests are carried out and the microstructure is characterized using scanning and transmission electron microscopy based techniques and post-mortem synchrotron X-ray diffraction analysis. The results reveal that (i) the strain partitioning between austenite and martensite is governed by a highly dynamical interplay of dislocation slip, deformation-induced phase transformation (i.e. causing TRIP effect) and mechanical twinning (i.e. causing twinning-induced plasticity (TWIP) effect); and (ii) the nano-laminate microstructure morphology leads to enhanced damage resistance. The presence of both effects results in enhanced strain hardening capacity and damage resistance, hence, the enhanced ductility.

Describing grain boundary (GB) crystallography requires five independent parameters. By combining electron and optical microscopy, we have developed a new hybrid method to assess—non-destructively—the 5-D GB crystallography in polycrystalline aluminum samples. Our method allows for the direct measurement of local properties at hundreds of fully characterized GBs at once, and thus represents a powerful alternative to techniques, such as X-ray diffraction, which are more cost- and time-intensive. As a case study, we apply our method to assess GB susceptibility to intergranular corrosion and show how the data acquired can be used to build GB crystallography-property maps. These results may enable higher accuracy lifetime prediction of engineering alloys and inspire novel designs of corrosion-resistant microstructures through GB engineering.
This work was supported by the US Department of Energy, O#64259;ce of Basic Energy Sciences under award No. DE-SC0008926; and by the MISTI Seed Fund “Progetto Roberto Rocca”.

Laminate microstructures contribute to strength of materials in a wide range of systems. We review some recent results on pearlitic steels, Cu-Nb and nano-twinned metals. The Cu-Nb laminates are made by two methods, namely accumulated roll bonding (ARB) and physical vapor deposition (PVD) with lamellae thicknesses in the range 10 µm to 10 nm. Nano-twinned copper and silver are electrodeposited, for which a considerable range of texture type and strength is accessible. The pearlitic steels are self-evidently a naturally occurring nano-laminate composite. The strength of the composites, as measured by nano-indentation (and other methods) increases with decreasing lamella thickness as has long been known.
The analysis of the pearlitic steels used the viscoplastic FFT model and focused on a) the influence of the different orientation relationships that have been postulated for this system and b) the influence of the geometry of the two phases e.g. lamellar versus globular cementite. The main findings are that a) the deformation is unaffected by the choice of the orientation relationship whereas b) the properties are strongly sensitive to the phase morphology, which suggests that load transfer across the hard (cementite) phase is an important feature.
Similar strengths in the Cu-Nb composites are seen in ARB and PVD material across the length scale, with ARB strength being slightly lower. The PVD composites are tested as-deposited whereas the ARB materials are the result of substantial plastic deformation, so it is surprising that the latter are less hard. Strength varies with average layer thickness following a Hall-Petch relationship. The nano-twinned copper & silver follow a similar trend, albeit at substantially lower strengths. The more physically based confined layer slip (CLS) model describes the variations in strength across the range of thicknesses. However there are other models that also can explain the variations.

The laminates Ti6Al4V/AA2519 were fabricated by explosive welding in two versions: with and without an aluminium alloy 1050 interlayer plate. In the present communication, the focus is on interface phenomena such as diffusion and phase transformations during bonding and subsequent annealing as well as mechanical behaviour of the laminate structures. The results demonstrated that both Ti6Al4V/AA2519 and Ti6Al4V/AA1050/AA2519 composite plates exhibit good quality bonding. Scanning electron microscopy observations and energy-dispersive X-ray spectrometry (EDS) results revealed that bimetallic compounds of Ti/Al form at the bonding interface in AA 1050 interlayer. Ti/Al interlayer had a thickness of about 5 µm and consisted of nano-sized grains formed along entire bonding interface. In the samples without aluminium interlayer the thickness of bimetallic Ti/Al interlayer was significantly smaller. However, near the joint one can see the grid in the grain boundaries rich in copper. These intermetallic inclusions were often accompanied by nano-cracks. The value of microhardness decreased when the distance from the interface increased. Tensile tests revealed that in both types of joints the interfacial zone wasn&’t the weakest part of this composites.

In many interface-dominated nanostructured materials the role of interfaces during deformation is not yet completely clarified. Very fine spacing of interfaces leads to a competition between dislocation controlled and grain boundary sliding based plasticity. To improve our understanding of this competition we have to investigate the atomistic origin of deformation in the interface region.
We present the results of molecular dynamics simulations of sliding at γ/γ interfaces in lamellar TiAl alloys, which can explain existing, seemingly contradictory, experimental results on the role of interfacial sliding during creep. We suggest that the origin of the controversy lies in the pronounced in-plane anisotropy of the shear strength of the individual interfaces, which is observed in the simulations. Experimentally, the orientation of in-plane directions with respect to the loading axis has not been monitored so far.
A multi-scale concept is introduced to capture effects of both the electronic and the atomistic level. On the one hand we carried out quasi-static calculations of multi-planar generalized stacking fault energy (MGSFE) surfaces of the interface plane as well as the adjacent layers. On the other hand molecular dynamics simulations guided by ab initio GSFE-surface calculations were carried out for different bicrystal cells under different shear loading conditions. The critical stresses for shearing some of these interfaces, which were derived from these bicrystal shear simulations, are of the same order of magnitude or even lower as those for dislocation motion in a γ-single crystal, showing that these mechanisms are competitive.
In total four shear mechanisms, twin nucleation and migration/absorption, interfacial partial dislocation nucleation, rigid grain boundary sliding and grain boundary migration were observed. The comparison with the MGSFE surfaces allows to create a link between (quasistatic) physical properties such as stacking fault energies, and the (dynamic) deformation mechanisms, and hence between the results of ab-initio calculations and molecular dynamics simulations.

Magnesium and its alloys, the lightest structural materials, have a huge potential as structural materials in the aerospace and automotive industries and have been actively investigated for the past few decades. However, Mg with a hexagonal close packed structure lacks the ductility and formability of cubic materials due to scarcity of easy slip systems and the localized shear characteristic of twinning. Therefore, for their applications as structural materials, there is a strong desire to improve the deformability while maintaining the high flow strength of Mg and Mg alloys. Multilayered metallic materials containing a high density of interfaces show promising enhancements in both the yield strength and ductility of layered composites. The mechanical strength of the Mg/Nb multilayers measured according to hardness can achieve a high value of ~1.0 GPa when the layers are a few nanometers thick. Besides such high strength, we also demonstrated such layered structure can enhance fracture toughness associated with highly activated basal slip and interface shearing between Mg and Nb interfaces. We performed in situ three-point bending experiments in SEM and found that Mg/Nb interfaces could block crack propagation accompanying with blunting crack tip. Stress concentration at crack tip is significantly released by shearing interfaces parallel to Mg basal planes.

Magnesium and its alloys, the lightest structural materials, have attracted a lot of attention of the automotive industry for reducing the vehicle&’s weight to improve its fuel efficiency. However, Mg has a hexagonal close packed (HCP) structure, which lacks the ductility and formability of cubic materials due to scarcity of easy slip systems to accommodate an arbitrary plastic deformation mode. Layered composites are promising because presence of high density of interfaces in such materials that can enhance both yield strength and ductility. In this work, we present first-principles study of the energetics and atomic solubility of coherent interface in Mg/Nb multilayers. We find that solutes Zn, Cd, Sc, Ti and Zr show better solubility at Mg/Nb interface compared to the bulk Mg and the bulk Nb and would segregate at the interface. We also search for solutes that can stabilize Mg/Nb interfaces and suppress deformation twinning in Mg to improve the mechanical properties of Mg/Nb multilayer systems.

Nanolayered composites such as Cu-Ag, Cu-Nb and Zr-Nb with layer thicknesses less than 100 nm exhibit enhanced material properties such as ultra-high strength, high ductility, and radiation damage resistance. At such fine length scales, the atomic structure of the interface plays a dominant role in its response to defects and dislocations. In this work, a combined approach of experiment and atomistic and mesoscale modeling elucidates the effects of the atomic structure of bimetal interfaces on the propensity for deformation twinning. We utilize Accumulative Roll-Bonding (ARB) to process bulk Cu-Nb and Zr-Nb nanolamellar composites from 1 mm thick high-purity polycrystalline sheet into bulk nanocomposites with individual layer thicknesses below 100 nm. The ARB process imposes thousands of percent strain to refine the microstructure of ordinary coarse-grained composite metals down to submicron and nanoscales. TEM analysis showed that the stable ARB interface structure at layer thicknesses less than 50 nm maintains the conventional Kurdjumov-Sachs (KS) orientation, but joins the Cu and Nb crystals at interface planes different from those observed in Physical Vapor Deposited materials investigated previously: {112}Cu || {112}Nb and <110>Cu || <111>Nb. In addition, twin formation was experimentally observed in copper on a plane that lies 19.5 degrees from the {112} interface plane, which had not been previously observed in rolled physical vapor deposited (PVD) nanolayered Cu-Nb composites. Moreover, the ability of interfaces in the Cu-Ag system to aid in twinning processes at fine length scales is investigated and compared to those found in Cu-Nb. Cu-Ag nanoscale multilayers, in bulk cast eutectic form, have a cube-on-cube or twinned orientation relationship with interfaces consisting of common {111} planes. These differing interface types will be discussed in terms of the influence of interfacial geometry, dislocation kinetics, interfacial energetics, and atomic structure on the propensity for different deformation mechanisms.

Interfaces in materials can play an important role in determining mechanical performance of the materials through the complicated interaction between defects and interfaces. In metallic nanolayered composites, interface barrier strength (IBS), which refers to the resistance for dislocations to cross an interface, is a key factor to determine the ultrahigh strength of nanolayered composites. Although several theoretical models have been proposed to describe strengthening mechanisms in multilayers and primary contributions to the IBS, the variation of interface properties caused by element interdiffusion at atomic scales and its influence on the IBS of nanolayered composites are still not well understood. In this talk, we will present experimental and theoretical investigations of the IBS of Cu/Au multilayers, emphasizing effects of interface properties and structures. The Cu/Au multilayers with individual layer thickness ranging from 25 to 250 nm were annealed at different temperatures to modulate Cu/Au interface properties. Experimental results from nanoindentation testing show that the Hall-Petch slope in the relation between the strength and the individual layer thickness of the multilayers gradually decreases with increasing annealing temperature, indicating a decrease in the IBS. TEM and HRTEM characterization reveals that element interdiffusion between the Cu layer and the adjacent Au layer leads to a compositional gradient at the interface. A detailed analysis for effects of the compositional gradient on the resistance to dislocation crossing the interface was conducted. It is demonstrated that the interface structure of the Cu/Au multilayers has become the most important factor in governing the IBS.

The thermal stability of Mg/Ti multilayer nanofilms was investigated by examining their microstructure and nanoindentation hardness after annealing at various temperatures and time periods. The multilayers with individual layer thickness hge;5 nm exhibit excellent capability of maintaining the lamellar microstructure and high strength up to 200 °C for annealing time up to 2.0 hours. The annealed multilayer films with h=2.5 nm are still highly textured but characterized with discontinuous layer interfaces, in which the transition of atomic arrangement from hexagonal close-packed (HCP) to body-centered cubic (BCC) structure was observed at columnar boundaries. The degradation of uniform lamellar microstructure is related to the decrease of hardness with annealing temperature at this size scale. A diffusion based instability mechanism was proposed for this typical HCP-based nanoscale multilayer system.

Tri-component multilayers, such as Cu/Ni/Nb. that contain both coherent and incoherent interfaces have a greater capacity for strain hardening than a tri-component bilayer (Cu-Ni/Nb) system, in which only incoherent interfaces are present. However, the thermal stability of these systems is not as well studied as other bi-metallic multilayer systems. The current study examines both the mechanical response and the microstructure and interfacial structure of sputtered Cu/Ni/Nb tri-layers and Cu-Ni/Nb in the as-deposited and annealed conditions. In general it may be expected that coherent interfaces would provide relatively easy diffusion and homogenization and hence softening of the multilayer was expected following annealing at 573 and 773 K for several hours. However, post-annealing nanoindentation of both the bi and tri-layer films showed an increase in hardness of the tri-layer film after elevated temperature testing had been conducted, indicating the likelihood of a change in the microstructure. X-ray diffraction experiments suggest that microstructural changes are taking place in the coatings following even a modest anneal; peaks from Cu and Ni in the tri-layer system are observed to begin to merge and there is some evidence for new peaks forming. Transmission electron microscopy revealed that while the grain size is relatively stable, a Ni-Nb intermetallic forms at both the Ni/Nb, Cu-Ni/Nb and Cu/Nb interfaces. It was also observed that considerable interdiffusion of Cu and Ni took place above 700K. This formation of precipitates appears to add a strengthening component above and beyond that predicted from the confined layer slip model.

Multi-layered aluminum clad sheets are used for a variety of applications in airplane structural parts and automobile heat exchangers. Different alloys with different properties can be joined together to one sheet and it show improved properties or additional properties rater than each alloy. In this study, 6022/5023/6022 aluminum clad sheets were fabricated by roll bonding process and the mechanical properties of the annealed sheets were examined. The bonding strength of the sheets increases with increasing a thickness reduction rate of roll bonding, 6022/5023/6022 clad sheets can be fabricated successfully by roll bonding at room temperature with the thickness reduction rate per pass of 50%. After annealing at 500 ^oC for 1 hour, the sheets have a well bonded interface without a void and 2^nd phase and the microstructure of the sheets consisted of fully recrystallized grains. Stress-strain curves show same strain hardening characteristics with core Al5023 alloy sheets. Meanwhile, the elongation of the sheets was higher than that of commercial Al6022 or Al5023 alloy sheet. The best thickness combination for the elongation was 1:2:1 thickness ratio of 6022:5023:6022. In addition, the sheets had very low tensile anisotropy due to different deformation characteristics of each layer. The low anisotropy of the sheets brought with excellent drawability and the minimum earing at cupping test. It comes from a combination of different texture characteristics of each layered sheets. Numerical simulation results on the formability were corresponded well with the experimental result.

The development of grain-oriented silicon steel sheet products processing a lower core loss and higher magnetic induction has become increasingly important. Recently, an ultra-low iron loss could be accomplished by using ceramic coating such as TiN, CrN and TiC applied by the PVD or CVD method after the forsterite (Mg₂SiO₄) film of silicon steel surface was removed and polished. However, this technique suffers two major shortcomings: (i) low deposition rate, typically in the range of 0.01~0.1 mm min^-1, and (ii) low power efficiency.
Cathodic arc deposition allows the production of a highly adherent deposit due to the condensation of numerous ions on the substrate. To address the problems of sputtering and EB deposition systems, in particular low productivity, poor surface properties, we aim to make TiN, AlN, Al₂O₃ coatings using a DC arc system. Additionally, due to the ease of scaling to large areas and industrial availability, we also consider reactive magnetron sputtering system. With the arc deposition system, high deposition rates were achieved when compared to the sputtering system (> 10~ 90 times).

Due to the lack of valuable experimental techniques, epitaxial strains developed at the metal // oxide interphase can only be study using numerical approaches [1-2]. In addition, most of the techniques, i.e. Raman spectroscopy, can gives only qualitative information about the developed stress field [3]. Moreover, obtained data is burdened with high measurement error and very often difficult to interpret. Therefore influence of epitaxial strain at the metal/oxide interphase is still under investigation [4]. In this work, developed at the zirconium / zirconia interphase epitaxial strains have been study using generalized Bollmann method. Based on the SEM/FIB investigation performed on oxidized at high temperature zirconium // zirconia cross section, numerical approach have been applied to the complex hexagonal vs monoclinic (case I) and hexagonal vs tetragonal (case II) systems [5]. The objective of this work is to study the sensitivity of the “best fit criterion” to two numerical parameters. Special attention has been given to the sensitivity of the cell parameter value and the initial orientation of the metal and the oxide (theta;=0^#7506;). Theoretical strain field created between zirconium and zirconia considering both crystallographic cases have been given. This study is presented as a tool to understand the phenomena of epitaxial strain accumulation in the zirconia layer after isothermal heating at high temperature. Reported results, for the first time quantitatively describe mechanical stress field generated due to the mismatch between metallic substrate and growing oxide.
Literature:
[1]. W. Bollmann, Crystal defects and crystalline interfaces (Springer - Verlag, 1970).
[2]. L. Kurpaska, J. Favergeon, L. Lahoche, G. Moulin, Jean-Marc Roelandt, Materials Science Forum 696 (2011) 176 - 182
[3]. L. Kurpaska, J. Favergeon, L. Lahoche, G. Moulin, M. El-Marssi, Jean-Marc Roelandt, Oxidation of Metals 79 (2013) 261 - 277
[4]. I. Salles-Desvignes, T. Montesin, C. Valot, J. Favergeon, G. Bertrand, A. Vadon, Acta Materialia 48 (2000) 1505 -1515
[5]. L. Kurpaska, J. Favergeon, L. Lahoche, M. El-Marssi, Jean-Luc Grosseau-Poussard, G. Moulin, Jean-Marc Roelandt, Journal of Nuclear Materials (2015) doi:10.1016/j.jnucmat.2015.06.005

Solute segregation plays a key role in a broad range of phenomena in multiphase materials containing a high density of interfaces, yet the diverse nature of these interfaces makes quantifying and predicting segregation a difficult task. Here we report on the simultaneous segregation of Au atoms to four different interfaces in Fe/Au particles on sapphire - two distinct metal surfaces, a metal-ceramic interface, and a metal-metal grain boundary - resulting in complete encapsulation of the particles. We accessed all of these interfaces simultaneously and found substantial differences in their segregation behavior. The metal-ceramic interface exhibited the strongest segregation tendency, followed by the two surfaces, and the grain boundary. The results were analyzed quantitatively by combining experimental, theoretical and ab-initio computational methods, leading to the new synergetic insights into such systems. We then demonstrated how segregation can be directly employed to design the morphology and properties of thermodynamically-stable nanoparticles and thin films.

A Ge CVD film is expected a lower temperature of bonding, and increasing the bonding reliability and bonding strength, and can be applied to wafer-level-bonding process with a patterned wafer. In this study, the Ge CVD thin film was studied to bond the wafers and to obtain a uniform thin film between the Si wafer and Au film deposited wafer.
The Ge films were deposited on Si wafers at substrate temperature of 400°C, 450°C, 500°C and 570°C, respectively. GeH₄ flow rate was 60sccm and deposition pressure was 4x10^-4 torr for 5min. A two step-deposition-method was tried to reduce a surface roughness, first a 30 ~ 60 nm-thickness film was deposited at lower temperature of 400°C and after the second film was deposited at higher temperature of 500°C. This is possible to obtain a lower surface roughness and uniform film thickness of about 30nm or more, deposited at a lower temperature could reduce the lattice mismatch during the growth of Ge film, also which can be known that the planar growth is possible for above 100nm-thickness film (target thickness).
We have observed Arrhenius plot of deposition rate vs. deposition temperature, and the surface morphology of the thin film by atomic force microscopy (AFM). The deposition rate and a surface roughness were increased with increasing deposition temperature. The higher deposition temperature showed the higher crystallinity at deposition temperature range from 400°C to 500°C. The X-ray diffraction (XRD) peak of Ge (004) was observed at deposition temperature of 400°C.
The Ge thin film was bonded between a about 100nm-thickness Ge CVD deposited in an optimized process condition on the Si substrate wafer and an Au (200nm)/Ti (10nm)/Si wafer. The deposition of Ge films were carried out at 400°C, and the bonding were performed at temperatures of 400°C (1hr), 450°C (1hr) and 500°C (1hr), respectively. Scanning acoustic microscopy (SAM) was used to analyze the status of bonding. From results of a SAM observation and a sectional SEM sample preparation, the good adhesion between Ge and Si was observed at higher bonding temperature of 450°C. To observe the bonding strength of the bonded films and inter-diffusions of Si-Ge-Au atoms, the upper wafer of the bonded wafers was grinded by using a back-grinder tool up to from a 480mu;m-thickness (4inch wafer) to a about 10mu;m-thickness. Then the bonding element distribution was confirmed by using a secondary ion mass spectroscopy (SIMS) and Electron energy loss spectroscopy (EELS) of transmission electron microscope (TEM). In this study, we knew that the Au-Ge bond was possible at about 450°C, and the results show that the Ge CVD film can be adaptable for the application of wafer level package (WLP).

Mg-based long-period stacking ordered (LPSO) phases formed in ternary Mg-TM (transition metal)-RE (rare-earth metal) alloys have recently attracted significant attention owing to their use in the strengthening of Mg alloys and as promising components for designing advanced lightweight structural materials. In such phases, the two-dimensional stacking-fault (SF)-type interfaces containing solute atoms can play a significant role in the formation of a nanostructured state with low-energy boundaries. To understand the fundamental mechanism of the formation of LPSO phases, it is important to elucidate the controlling factors influencing the segregation and ordering of solute atoms at SF-type interfaces, i.e., the solute-solute binding, solute interaction with SFs, etc. In this study, we investigated the interaction energies and trapping effects of substitutional Zn and Y atoms in the vicinity of the SFs in Mg using first-principles calculations based on density functional theory (DFT). As no reliable empirical potential exists at the moment to reproduce faithfully the interactions of the elements, we developed an effective interaction model based on the DFT-calculated energies, which will provide the theoretical framework to find the energetically favorable configurations from the bottom-up point of view. As examples, the parameterizations for the Mg-Zn-Y systems were conducted. We applied this to atomistic and/or coarse-grained Monte Carlo modeling to predict the solute segregation and nanocluster evolution at a SF in the Mg-Zn-Y systems. We confirmed that our approach can successfully describe the formation and alignment of the Zn-Y clusters in the systems, which provides an insight into the short- and medium-range chemical order in the Mg-based LPSO structures.
This study was supported by Grant-in-Aid for Scientific Research on Innovative Area (No. 23109004) and the Elements Strategy Initiative for Structural Materials (ESISM).

A high density (asymp; 1/10 nm) of twin boundaries have been reported parallel to close-packed directions in thin FCC and HCP films made using atom-by-atom deposition processes. Such nanotwinned materials are of interest because of the potential to create materials with high strength and high ductility. However, process control of twin boundary spacing has been elusive. In this work, we show that thin Ag films can be created with a range of twin boundary spacings by varying the deposition rate in e-beam evaporation. We explain these variations using a simple model of nucleation and growth from the vapor. Results suggest that twin structure and mechanical properties can be tuned with good precision.

The motion of grain boundaries is crucial for the evolution of microstructure both in processing and in application. In experiments, the grain boundary motion is usually studied at relatively high homologous temperature during long times under small mechanical loads. But grain boundary motion is also shown to be a deformation mechanism occurring at high stresses and relative low temperatures, as for example in deformation studies of nanocrystalline metals and shock loading.
Using molecular dynamics simulations the interface motion of grain boundaries is used to elucidate the role of atomistic structure (morphology) and velocity-driving force relation (mobility) for two different driving force mechanisms: elastic anisotropy and shear coupled motion. As function of temperature and driving force, the grain boundary mobility shows a complex nonlinear behavior beyond the conventional conjecture of Arrhenius-like temperature-dependence in mobility and linear velocity-driving force relation. The velocity-limiting mechanisms range from the pinning-depinning transition at low temperature, through rare-event dynamics of critical “kink- pair” disconnection nucleation along intrinsic grain boundary dislocations, to fluctuating randomly diffusive grain boundary motion at low driving forces and high temperatures.
The GB dynamics regimes are discussed in context of grain growth at high temperatures and stress-driven coarsening in nanocrystalline metals at low temperatures in the presence of triple junctions.

Metals that are extensively used in electronics manufacturing, such as tin, zinc, and related alloys, often show electrically conductive hair-like crystalline structures on their surfaces, referred to as whiskers. These whiskers can lead to current leakage and short circuits in sensitive electronic equipment causing significant losses and, in some cases, catastrophic failure in the automotive, airspace and other industries. The mechanism responsible for whisker formation and growth remains mysterious after decades of research. Some theories attribute the whisker growth to stress relieving phenomena, but those never lead to any quantitative estimate of whisker parameters. A very recent electrostatic theory [1] provides, for the very first time, the quantitative estimates of the whisker nucleation along with their growth rates and length distributions, which is very consistent with many observations on whiskers growth and their morphology. It proposes that the imperfections on metal surfaces can form small patches of net positive or negative electric charge leading to the formation of the anomalous electric field which governs the whisker development in those areas.
In this work, which is still in progress, we report on the possibility of initiating and controlling the growth of whiskers by using surface plasmon polaritons (SPPs). As predicted in [1], localized high electric fields generated by the SPPs may play a crucial role in whisker onset and growth. The Otto and Kretschmann methods are very well established and studied attenuated total reflection geometries for implementing SPPs on the surface of metals and metal films. Generating SPPs on low-loss metal films such as Ag and Au is generally straightforward. However, generating SPPs on the surfaces of lossy metals such as Sn and Zn is challenging as the energy of the incident field is largely absorbed in the process of coupling to the SPPs. We employed a 100mW cw red light emitting laser to excite SPPs on optically smooth Zn and Sn films and then followed the evolution of the exposed surfaces by optical and electron microscopy techniques. This work has important implications: first, it would verify experimentally some aspects of the electrostatic theory and second, the controllable formation of whiskers would allow the development of accelerated failure testing methods of electronic components, which is currently not possible due to the random and unpredictable nature of whiskers formation. As a somewhat longer-term goal, we would like to explore the potential of controllable intentional whiskers growth in applications that involve microwire fabrication such as 3D electronic circuit wiring and sensors.
1. Karpov, V.G. , Phys. Rev. Applied, vol. 1, no. 4, # 044001, 2014

For over 40 years, SrRuO₃ has been the subject of extensive studies due to the coexistence of metallic conductivity and ferromagnetism.[1] Especially, as SrRuO₃ has an orthorhombic structure, there are six possible domain structures and orientations for epitaxial growth when deposited on cubic substrates (e.g. SrTiO₃). [2] It often happens that different domains simultaneously occur forming multidomain thin films under certain growth mode. Since SrRuO₃ with single domain shows optimal physical properties compared to multidomain structures, there have been many attempts to suppress the formation of multidomains.[3] Up to now, the most widely accepted claim is that vicinal SrTiO₃ substrate(ge;2°) plays a critical role in determining single domain structure originating from the growth mode change, but it is still remained controversial.[4]
In this study, we observe atomic scale structural evolution of SrRuO₃ on SrTiO₃(001) focusing on the interface. The SrRuO₃ thin films are deposited on SrTiO₃(001) substrates with various miscut angles by pulsed laser deposition technique with in situ high-pressure reflection high-energy electron diffraction (RHEED). We monitor the changes of growth mode by RHEED intensity oscillation in real time. Overall domain distribution is determined with high-resolution X-ray diffraction (HRXRD) and detailed domain composition is obtained by transmission electron microscopy (TEM) via direct imaging of microstructure. Precise atomic positions around the interface are detected by the help of annular bright field(ABF)-high angle annular dark field(HAADF) imaging technique. Based on ABF-HAADF images, we confirm that the degree of octahedral rotation is affected by domain configuration. Mechanical properties at the interface is also calculated in order to confirm the effect of interface atomic structure on interface properties. Our results demonstrate that there are close links between structure and physical property and that domain formation is strongly dependent on the substrate surface.
[1] G. Koster et al., Rev. Mod. Phys. 84, 253 (2012).
[2] J. C. Jiang et al., Appl. Phys. Lett. 72, 2963 (1998).
[3] G. Herranz et al., Phys. Rev. B. 71, 174411 (2005).
[4] Q. Gan et al., Appl. Phys. Lett. 70, 1962 (1997).

Nanometalic materials have interesting mechanical, electrical, and magnetic properties, but generally exhibit low thermal stability due to the higher interface area which acts as a high diffusivity path, and is a suitable place for grain boundary stabilization by kinetic or thermodynamic mechanisms (segregation). Specifically, nanometalic multilayers (NMMs) are an attractive solution to enhance thermal stability, as they enable local control over concentration of alloying elements, and surface energy distribution.
Low thermal stability of NMMs at high temperatures is a serious drawback for their further application. For miscible systems as Cu-Ni degradation of the multilayered structure and grain growth after annealing have been previously reported. The multilayered geometry enables grain size control (layer thickness correlates with grain size), and the alternating composition allows control over kinetic phenomena (kinetic barrier), as the initial microstructure can be tailored to be close to the equilibrium condition.
This study presents the evolution of the interlayer interface of Hf-Ti NMMs samples prepared by magnetron sputtering, and heat treated at vacuum conditions, at different annealing times and temperatures. FIB lift out was performed to prepare the samples for imaging, microstructures and grain sizes of as-sputtered and heat treated samples were analyzed by TEM, and EDX profiles were used to track Hf and Ti composition. The loss of layered structure and grain boundary segregation for various layer thicknesses is presented.

Recent studies have shown that growth of eta;-Cu₆Sn₅ on unidirectional Cu followed specific patterns with regular spatial arrangement, but the crystallographic origin and formation mechanism of such a microstructural feature is still unclear. Here we develop a phase field model incorporated with the microelasticity theory to investigate microstructure evolution during heteroepitaxial growth of eta;-Cu₆Sn₅ on unidirectional (001) Cu. The generation of eta;-Cu₆Sn₅ variants arising from symmetry reduction is determined by group theory and the microstructural evolution is simulated through phase-field modeling with the influence of epitaxial strain. Simulation results show that each eta;-Cu₆Sn₅ variant elongates along the direction of minimum atomic mismatch, which agree with experimental observations. The condition of kinematic compatibility across the interface of eta;-Cu₆Sn₅ variants is analyzed and validated by simulations. In addition, possible extension to heteroepitaxial growth on (111) Cu or other chemical systems are discussed in an analogous manner.

Recent works by Hemphill et al have shown encouraging fatigue resistance characteristics of a high entropy alloy (HEA) Al0.5CoCrCuFeNi, which compares favorably with a number of conventional alloys, including steels, titanium alloys and bulk metallic glasses. Due to the complex nature of the microstructure in HEA, the microstructural reasons to this excellent fatigue resistance is unclear at the moment. Here, we apply a novel technique in transmission electron microscopy (TEM) called scanning electron nanodiffraction (SEND) to characterize the microstructure development along the fatigue crack in Al0.5CoCrCuFeNi. SEND enables a complete structure characterization from which we can further analyze the crystal orientation and structure.
Arc-melted Al0.5CoCrCuFeNi (molar ratio) ingots were annealed at 1000 °C for 6 h, followed by water quenching and cold rolling. Four point bending fatigue testing was run at a maximum stress of 900 MPa corresponding to a stress range of 810 MPa. A sample which failed after 98,256 cycles was used in the present characterization. TEM samples were prepared by focused ion beam (FIB) technique at different locations of the fatigue crack. TEM investigation was carried out on a JEOL 2100 equipped with a LaB6 electron beam source operating at 200kV. Nanometer-sized electron beam of 2~4 nm in diameter was used to generate the diffraction patterns at each location in a serial manner. SEND patterns were acquired using a customized Gatan DM script which automates beam scanning and diffraction pattern recording. Typical exposure time for each diffraction pattern recording is 0.5 seconds.
A scanned area covering 1000x1000 nm2 using 20x20 sampling points was applied for each investigated site, where overall 400 diffraction patterns were recorded. A correlative analysis was carried out over these diffraction patterns. The regions with similar diffraction patterns was defined by a high correlation coefficient (> 0.75, 1 is the maximum). Using this technique, we are able to identify the regions in the sample that show similar diffraction patterns. The fatigue initiation region, the propagation region and the final cracking region have been all characterized. The results are compared directly with those obtained in the region far away from the fatigue crack. The microstructure evolution along fatigue crack is discussed based on the SEND results.

This talk will present our recent progress in understanding the fundamental mechanisms of plasticity and fracture in face-centered cubic metals and alloys strengthened by coherent twin boundaries (CTBs) using atomistic simulations and nanoscale experiments. The ability of CTBs in strengthening and maintaining ductility has been well documented; yet most understanding of the origin of this mechanical behavior relies on a perfect interface assumption. This presentation will focus on the important role played by intrinsic kink-like twin boundary defects in controlling strength and plasticity in nanotwinned Cu and Ag-Cu alloys. First, using simple bicrystal models, it is found that kink-like twin boundary steps have a profound impact on screw dislocation - CTB interactions, resulting in significant strengthening behavior. Second, massively-parallel molecular dynamics simulations will be used to describe new defect-dependent plastic deformation processes in columnar-grained and nanocrystalline microstructures containing inherently defective CTBs with kink-like steps. New Monte Carlo simulations will also be presented to discuss the influence of atom segregation on deformation mechanisms and twin size effects in nanotwinned Ag-Cu alloys.

Thin films having high densities of twin boundaries parallel to close-packed directions in thin FCC and HCP films made using atom-by-atom deposition processes have been reported. Such nanotwinned structures are of interest because of the potential to create materials with high strength and high ductility. However, to achieve, these properties, such structures must be stable during thermal annealing. Twin boundaries can represent significant stored energy and thus driving force for microstructural transformations. In this work we produced thin Ag films with a range of twin boundary spacings and studied the effect on structural transformations during annealing. We found that the rate and extent of grain growth scaled with the twin boundary density and showed that elimination of twin boundaries can provide both the driving force and the orientation selection mechanism for the well-known thickness-dependent (111) to (100) texture transformation. Conditions required for the stability of nanotwinned structures will be discussed.

Although nanoscale twinning is an effective means to enhance yield strength and tensile ductility in metals, nanotwinned metals generally fail well below their theoretical strength limit due to heterogeneous dislocation nucleation from boundaries or surface imperfections. Here we show that Au nanowires containing angstrom-scaled twins (0.7 nm in thickness) exhibit tensile strengths up to 3.12 GPa, near the ideal limit, with a remarkable ductile-to-brittle transition with decreasing twin size. This is opposite to the behaviour of metallic nanowires with lower-density twins reported thus far. Ultrahigh-density twins (twin thickness < 2.8 nm) are shown to give rise to homogeneous dislocation nucleation and plastic shear localization, contrasting with the heterogeneous slip mechanism observed in singlecrystalline or low-density-twinned nanowires. The twin size dependent dislocation nucleation and deformation represent a new type of size effect distinct from the sample size effects described previously.

High energy particles can introduce severe radiation damage in metallic materials especially those with low stacking fault energy. Twin boundary (TB) has recently been shown to enable the reduction of defect density in heavy ion irradiated nanotwinned Ag. However, the defect-twin boundary interaction mechanisms in nanotwinned metals remain poorly understood. This presentation will focus on the study of TB affected zone wherein time accumulative defect density and defect diffusivity are substantially different from those in twin interior. Furthermore