Search

Many emerging non-volatile memories are compatible with CMOS logic, potentially enabling their integration into a CPU’s die. This article investigates such monolithically integrated CPU–main memory chips. We exploit non-volatile memories employing 3D crosspoint subarrays, such as resistive RAM (ReRAM), and integrate them over the CPU’s last-level cache (LLC). The regular structure of cache arrays enables co-design of the LLC and ReRAM main memory for area efficiency. We also develop a streamlined LLC/main memory interface that employs a single shared internal interconnect for both the cache and main memory arrays, and uses a unified controller to service both LLC and main memory requests. We apply our monolithic design ideas to a many-core CPU by integrating 3D ReRAM over each core’s LLC slice. We find that co-design of the LLC and ReRAM saves 27% of the total LLC–main memory area at the expense of slight increases in delay and energy. The streamlined LLC/main memory interface saves an additional 12% in area. Our simulation results show monolithic integration of CPU and main memory improves performance by 5.3× and 1.7× over HBM2 DRAM for several graph and streaming kernels, respectively. It also reduces the memory system’s energy by 6.0× and 1.7×, respectively. Moreover, we show that the area savings of co-design permits the CPU to have 23% more cores and main memory, and that streamlining the LLC/main memory interface incurs a small 4% performance penalty.

The role of X deficiency on the mechanical properties of MAX phases was studied by synthesizing Ti2AlN through powder metallurgy in stoichiometric and sub/extra-stoichiometric nitrogen compositions. XRD analyses and ab initio calculations indicate that nitrogen vacancies result in a lattice contraction predominantly along the c-axis. The elastic moduli and intrinsic hardness of substoichiometric Ti2AlN0.9 measured from nanoindentation tests are shown to be slightly smaller than that of Ti2AlN. The key mechanical indexes, bulk (B), shear (G) and Young’s (E) moduli as well as the hardness variation are calculated in density functional theory, and show different responses depending on the concentration of N vacancies. This joint experimental and theoretical study provides a full understanding of the energetics, chemical bonding, electronic structure, and mechanics of the N deficient MAX phases which would increase the application of nitride ceramics.

International audience; In this work, oxidation mechanisms were studied in fine-grained (FG) and coarse-grained (CG) Ti2AlC bulk samples. Results showed that the oxidation kinetics are controlled by the grain size of Ti2AlC. Bigger are the grains, faster is the oxidation. A dense and protective Al2O3 layer forms at the surface of FG-Ti2AlC samples while for the CG-Ti2AlC samples, a thick TiO2 layer forms on top of a discontinuous Al2O3. CG-Ti2AlC was observed to simultaneously transform into Ti3AlC2 and TiC instead of being directly transformed into TiC. This transformation result in the following crystallographically sandwich-like structure: (0001) Ti2AlC // (0001) Ti3AlC2 // (111) TiC. The volume shrinkage associated to this transformation produces elongated holes that are partially filled by α-Al2O3. The stress caused by the volume shrinkage generates cracks at the surface, which makes the oxygen inwards diffusion easier and thus worsens the oxidation resistance the CG-Ti2AlC bulk.

Nonvolatile memory, such as resistive RAM (ReRAM), is compatible with standard CMOS logic processes, allowing a sizable main memory system to be integrated into a CPU's die. ReRAM bitcells are fabricated within crosspoint subarrays that leave the bulk of transistors underneath the subarrays vacant. This permits placing the memory system over the CPU, improving area, parallelism, and power. Our work quantifies the impact of integrating ReRAM into a CPU's die. When integrating ReRAM over CPU logic, the best area efficiency occurs when 48% of the die is covered with ReRAM. The CPU's area increases by 18.8%, but we can recoup 35.5% of the die area by utilizing the free transistors underneath the crosspoint subarrays. When integrating ReRAM over CPU cache, up to 85.3% of the cache can be covered with ReRAM. Our work also shows that on-die ReRAM can support very high bandwidth through massively parallel memory access. At 28 nm, 4–16k independent ReRAM banks could be integrated onto the CPU die, providing 512–1024–GB/s peak bandwidth. At more advanced technology nodes, 5–10 TB/s may be possible.

Ti3AlC2 and Ti3Al0.8Sn0.2C2 MAX phase powders are densified using Spark Plasma Sintering technique to obtain dense bulk materials. Oxidation tests are then performed over the temperature range 800-1000°C under synthetic air on the two different materials in order to compare their oxidation resistance. It is demonstrated that, in the case of the Ti3Al0.8Sn0.2C2 solid solution, the oxide layers consist in TiO2, Al2O3 and SnO2. The presence of Sn atoms in

Non-volatile memory, such as resistive RAM (ReRAM), is compatible with standard CMOS logic processes, allowing a sizable main memory system to be integrated into a CPU’s die. ReRAM bitcells are fabricated within crosspoint sub-arrays that leave the bulk of transistors underneath the sub-arrays vacant. This permits placing the memory system over other logic. We propose a tileable, centralized ReRAM design over a large last level cache. This design takes advantage of ReRAMs unique characteristics while still providing flexibility to designers.

Ti3[Al1−δCuδ]C2 MAX phase solid solutions have been synthesized by sintering compacted Ti3AlC2–Cu composites produced by mechanical milling. Using X-ray and neutron diffraction techniques, it is demonstrated that the Cu mixing into the Al site is accompanied by lattice distortion, which leads to symmetry reduction from a hexagonal to a monoclinic structure. Such symmetry reduction likely results from this mixing through deviation of the A-site position from the special (0, 0, 1/4) position within the P63/mmc space group of the original Ti3AlC2 structure. Moreover, it is demonstrated that the Cu admixture into the A site can be adjusted from the composition of the reactant mixture. The lattice parameter variation of the solid solution compounds, with 10–50 atom % Cu in the A site, is found to be consistent with Vegard’s law.

The deformation mechanisms involved in the tensile creep of a Ti 3 AlC 2 specimen deformed, at 900 °C, to a 7.5% final strain are investigated through SEM and TEM observations. Tensile creep strain rate analyses on Ti 3 AlC 2 deformed at 900 °C enabled to identify a Norton's law with a n coefficient around 2, suggesting that creep mechanisms are controlled by grain boundary sliding. TEM observations revealed a highly heterogeneous microstructure consisting in both grains without any dislocations and grains highly defected. This intragranular deformation involves three different microstructural features: dislocations mainly confined in the basal planes and possibly organized in hexagonal networks, numerous stacking faults, and original lenticular non planar defects that likely play an important role in the sample deformation. The deformed microstructure observed and the estimated Norton coefficient suggest that intergranular deformation play an additional important role in the deformation mechanisms.

A new (Ti 1-e Cu e ) 3 (Al,Cu)C 2 MAX phase solid solution has been synthesized by sintering at 760 °C compacted Ti 3 AlC 2 -40 vol.% Cu composite particles produced by mechanical milling. Using XRD and TEM-EDXS, it has been demonstrated that Cu can enter the crystallographic structure of the Ti 3 AlC 2 MAX phase, whereas a Cu(Al,Ti) solid solution is also formed during thermal treatment. TEM-EELS analyses have demonstrated that Cu is mainly located on the A site of the MAX phase. The composition of the MAX phase solid solution, determined after selective chemical etching of the Cu(Al,Ti) matrix, by analyzing the filtrate and the solid phase using ICP-OES end EDXS methods respectively, is (Ti 0.93–0.97 Cu 0.07–0.03 ) 3 (Al 0.49–0.52 Cu 0.51–0.48 )C 2.

Polycrystalline samples with the Al0.693Cu0.201Fe0.106 composition, corresponding to the tetragonal P4/mnc ω-Al7Cu2Fe crystallographic structure, were synthesised by spark plasma sintering and deformed in compression under constant strain-rate conditions, e ˙ = 2 × 10−4 s−1, over the temperature range 650 K–1000 K. A brittle-to-ductile transition is evidenced between 700 K and 750 K. The stress–strain curves exhibit a yield point followed by softening or steady state conditions only. The upper yield stress, σUYS, shows a strong temperature dependence suggesting that the rate controlling deformation mechanisms are highly thermally activated. The strain-rate sensitivity of stress characterised either by stress exponents, nexp, or by activation volumes, Vexp, was measured by the load relaxation technique. High nexp values, i.e., larger than 7, associated with low Vexp, typically smaller than 1 nm3, are measured. The Gibbs free activation energy, ΔG, deduced by integrating Vexp with respect to stress varies from nearly 2 eV at 790 K to 4 eV at 1000 K. Because plasticity of the ω-Al7Cu2Fe phase takes place at temperatures at which diffusion processes are considered as dominant, the results are interpreted in the frame of dislocation climb models proposed to account for high temperature plasticity of crystalline phases.

Herein, we report—for the first time—on the additive-free bulk synthesis of Ti3SnC2. A detailed experimental study of the structure of the latter together with a secondary phase, Ti2SnC, is presented through the use of X-ray diffraction (XRD), and high-resolution transmission microscopy (HRTEM). A previous sample of Ti3SnC2, made using Fe as an additive and Ti2SnC as a secondary phase, was studied by high-temperature neutron diffraction (HTND) and XRD. The room-temperature crystallographic parameters of the two MAX phases in the two samples are quite similar. Based on Rietveld analysis of the HTND data, the average linear thermal expansion coefficients of Ti3SnC2 in the a and c directions were found to be 8.5 (2)·10−6 K−1 and 8.9 (1)·10−6 K−1, respectively. The respective values for the Ti2SnC phase are 10.1 (3)·10−6 K−1 and 10.8 (6)·10−6 K−1. Unlike other MAX phases, the atomic displacement parameters of the Sn atoms in Ti3SnC2 are comparable to those of the Ti and C atoms. When the predictions of the atomic displacement parameters obtained from density functional theory are compared to the experimental results, good quantitative agreement is found for the Sn atoms. In the case of the Ti and C atoms, the agreement is more qualitative. We also used first principles to calculate the elastic properties of both Ti2SnC and Ti3SnC2 and their Raman active modes. The latter are compared to experiment and the agreement was found to be good.

This paper presents the notion of a monolithic computer, a future computer architecture in which a CPU and a high-capacity main memory system are integrated in a single die. Such computers will become possible in the near future due to emerging non-volatile memory technology. In particular, we consider using resistive random access memory, or ReRAM, from Crossbar Incorporated. Crossbar's ReRAM is dense, fast, and consumes zero static power. Also, it can be fabricated in a standard CMOS logic process, allowing it to be integrated into a CPU's die. The ReRAM cells are manufactured in between metal wires and do not employ per-cell access transistors, so the bulk of the transistors underneath the ReRAM arrays are vacant. This means a CPU can be implemented using a die's logic transistors (minus transistors for access circuits), while the ReRAM can be implemented "in the wires" above the CPU. This will enable massive memory parallelism, as well as high performance and power efficiency. We discuss several challenges that must be overcome in order to realize monolithic computers. First, there is a physical design challenge of merging ReRAM access circuits with CPU logic. Second, while Crossbar's ReRAM technology exists today, it is currently targeted for storage. There is a device challenge to redesign Crossbar's ReRAM so that it is more optimized for CPUs. And third, there is an architecture challenge to expose the massive memory-level parallelism that will be available in the on-die ReRAM. This will require highly parallel network-on-chip and memory controller designs, and a CPU architecture that can issue memory requests at a sufficiently high rate to make use of the memory parallelism.

The mechanofusion process, a dry particle coating route, has been successfully applied to coat micrometric SiC particles with submicrometric Ni filaments. In a first step, the mechanofusion parameters were optimized to form a continuous Ni coating onto SiC particles. In a second step, the Ni-coated SiC particles were sintered by hot isostatic pressing. The temperature and pressure cycles were determined to ensure a good densification of the material. Such a densification process leads to the formation of a δ-Ni2Si bilayer at the SiC/Ni interface; the inner δ-Ni2Si layer in contact with SiC being more rich in carbon than the one in contact with the matrix. From X-ray diffraction, wavelength-dispersive X-ray spectrometry and scanning electron microscopy characterizations, a mechanism is proposed to explain the microstructure of the end-product.

Ti

The mechanical properties of the ω-Al 7 Cu 2 Fe crystalline phase have been investigated over a large temperature range (650–1000 K). Despite of its antinomic structure with the icosahedral Al–Cu–Fe quasicrystalline phase, i.e. periodic vs non-periodic, its mechanical properties are very similar to those of the quasicrystalline phase, which strongly suggest similar deformation mechanisms. Consequently, as for the quasicrystalline structure, we propose that dislocation climb might control the plastic deformation of the ω-phase. However, in the present case, the specificities of the quasicrystalline structure cannot be invoked to justify the predominance of dislocation climb, which questions the role of quasiperiodicity on dislocation mobility. We suggest that this deformation mode certainly results from specific non-planar extensions of the dislocation core.

A new submicro-layered TiC 0.61 /Cu(Al) composite has been prepared by hot-pressing a mixture of 50 vol.% Ti 3 AlC 2 and 50 vol.% Cu powders at 1150 °C and 30 MPa. It is shown that the initial reinforcement Ti 3 AlC 2 particles have, after synthesis, an unusual microstructure, which consists of submicron-thick layers of TiC 0.61 and Cu(Al) alloy. Both the width of the TiC 0.61 and Cu(Al) layers are ∼150 nm. Thus, the Ti 3 AlC 2 particles are decomposed into the TiC 0.61 phase, while the additional Al atoms provided by Ti 3 AlC 2 diffuse into the molten Cu matrix at high temperature. Compression tests were performed at constant strain rate in the temperature range 20–800 °C. The new designed TiC 0.61 /Cu(Al) composite has both a high yield stress, σ 0.2 measured at 0.2% strain offset, and a high ultimate compressive strength, σ UCS , which is attributed to strong interface bonding between TiC 0.61 and Cu(Al) phase. For instance, at 20 and 200 °C, σ 0.2 is 770 MPa and 700 MPa, while σ UCS is 1.18 GPa and 1 GPa, respectively. Plastic deformation takes place in the Cu(Al) matrix. Wavy slip lines are observed indicating that cross-slip could be the dominant deformation mechanism.

Ti, TiC, Al and AlN powders were mixed to synthesize Ti2Al(CxNy) (x + y

Herein, we report on the temperature-dependent crystal structures of Ti 3 AlC 2 and Ti 3 Al 0.8 Sn 0.2 C 2 in the 373–1273 K temperature range, as determined by Rietveld analysis of high-temperature neutron diffraction time-of-flight data. The compositions are 86(1) wt% Ti 3 AlC 2 and 14(1) wt% TiC 0.92(2) for the sample with no Sn , and 95(1) wt% Ti 3( Al 0.8 Sn 0.2) C 2 and 5(1) wt% Ti 2 AlC for the solid solution with Sn . The average linear volumetric thermal expansion is 8.0(2) × 10−6 K −1 for Ti 3 AlC 2 and 8.2(5) × 10−6 K−1 for Ti 3( Al 0.8 Sn 0.2) C 2. The average linear thermal expansion in the a and c directions, respectively, are 7.6(2) × 10−6 K−1 and 8.9(2) × 10−6 K−1 for Ti 3 AlC 2. For Ti 3( Al 0.8 Sn 0.2) C 2, the respective values are 8.0(5) × 10−6 K−1 and 8.6(6) × 10−6 K−1. In the case of the solid solution, the quadratic thermal expansion coefficients are also given. Detailed bond lengths analysis shows that the thermal expansions along the a and c directions are controlled by the thermal expansions of the Ti – C , and Ti – Al bond lengths, respectively. The atomic displacement parameters (ADPs) show that the Al and Sn atoms vibrate with a higher amplitude than the Ti and C atoms. Consistent with first-principles calculations, the ADPs of the Al/Sn site(s) in Ti 3( Al 0.8 Sn 0.2) C2 are lower than the ADPs of Al in Ti 3 AlC 2.

Reactive sintering of 3Ti:Sn:2C and 3Ti:Sn:2C:0.6Fe powder mixtures is studied in the temperature range 510°C–1200°C under argon. It is demonstrated that the recently discovered Ti3SnC2 phase is formed, provided that Fe is added to a 3Ti:Sn:2C reactant mixture within the synthesis conditions used. Using dilatometric and X-Ray diffraction analyses, the formation mechanism of Ti3SnC2 is discussed. Results show that at low temperature (about 510°C), tin is consumed to form FexSny intermetallics. At high temperature (about 1060°C), tin is newly available to form Ti3SnC2 due to the melting of FexSny. Then, the intermediate phases, TiC and Ti2SnC, and/or Ti5Sn3, TiC, C, and Ti are dissolved in the (Fe + Sn) liquid phase and Ti3SnC2 very likely precipitate from the melt. The second part of the study deals with the optimization of the Fe content in the initial 3Ti:Sn:2C reactant powder mixture to synthesize samples with larger Ti3SnC2 content by hot isostatic pressing.

A linear-response method to the density functional theory is used to derive lattice dynamics, the transport spectral function, and the electron-phonon coupling (EPC) constant of $\mathrm{T}{\mathrm{i}}_{2}\mathrm{AlC}$, a member of the very large class of nanolaminated conducting ceramics named MAX phases (where $M$ is a transition metal, $A$ is an element from groups IIIA to VIA, and $X$ is carbon and/or nitrogen). By coupling ab initio calculations with the semiclassical Boltzmann transport theory for electron-phonon scattering, the experimentally observed anisotropic electrical transport properties of $\mathrm{T}{\mathrm{i}}_{2}\mathrm{AlC}$ are rationalized. Our results indicate that in $\mathrm{T}{\mathrm{i}}_{2}\mathrm{AlC}$, because of the weak dependence of the EPC constant $\phantom{\rule{4pt}{0ex}}{\ensuremath{\lambda}}_{\mathrm{tr},\ensuremath{\alpha}}$ $(\ensuremath{\alpha}=xx$ and $zz)$ on the crystallographic direction, the anisotropy of $\ensuremath{\rho}(T)$ results from the anisotropy of the Fermi surface. These conclusions, in contrast with those obtained on $\mathrm{T}{\mathrm{i}}_{3}\mathrm{Si}{\mathrm{C}}_{2}$ (another member of the MAX phases family) using a similar approach, establish a correlation between the nanolaminated structure of the MAX phases and the origin of the anisotropy of their transport properties.

In this study, we report on the compressive behavior of Ti3AlC2 and Ti3Al0.8Sn0.2C2 MAX phases at room temperature. We found that these two phases could be classified as Kinking Nonlinear Elastic (KNE) solids. The cyclic compressive stress–strain loops for Ti3AlC2 and Ti3Al0.8Sn0.2C2 are typical hysteretic and fully reversible. At failure, both compositions fracture in shear with maximum stresses of 545 MPa for Ti3AlC2 and 839 MPa for Ti3Al0.8Sn0.2C2. Consequently, the macroshear stresses for failure, τc, are 185 MPa and 242 MPa for Ti3AlC2 and Ti3Al0.8Sn0.2C2, respectively. In addition to the grain size effects, the presence of a ductile TixAly intermetallic distributed in the grain boundaries plays an important role in the enhancement of the ultimate compressive and macroshear stresses for Ti3Al0.8Sn0.2C2. SEM observations reveal that these two MAX phases exhibit crack deflections, intragranular fractures, kink band formation and delaminations, grain push-in and pull-out.

Al-matrix material composites are produced from pure Al and 40 vol.% Ti 3 AlC 2 powders using hot isostatic pressing technique. It is demonstrated that the nanocrystallized-Ti 3 AlC 2 agglomerates, uniformly distributed in the Al matrix, form a hard continuous skeleton. The mechanical properties of the composites are evaluated over the temperature range of 20–500 °C by performing compression tests at constant strain rate. The monotonic temperature dependence of the proof stress at 0.2% plastic strain suggests that the same thermally activated mechanism controls the composite plastic deformation over the entire temperature range. The yield stress of the composite, about twice as high as that of the Al matrix in the investigated temperature range proves that Ti 3 AlC 2 particles constitute efficient reinforcement particles for Al matrix. SEM observations indicate that plastic deformation of 40Ti 3 AlC 2 /60Al composite takes place in the Al matrix while Ti 3 AlC 2 particle agglomerates undergo substantial fracture.

The HIPing of Ti-Al-1.9TiC and Ti-Al-0.2Sn-1.8TiC reac- tant mixtures yields Ti3AlC2 and Ti3Al0.8Sn0.2C2 solid solu- tions, respectively. Rietveld refinement gives a = 3.0786 ± 0.0001 Aand c = 18.589 ± 0.001 Alattice parameters for Ti3AlC2 and a = 3.0837 ± 0.0001 Aand c = 18.621 ± 0.002 A ˚ for Ti3Al0.8Sn0.2C2. Furthermore, about 2-3 vol% of TixAly intermetallics is detected in Ti3Al0.8Sn0.2C2 solid solution. The intrinsic hardness, measured by nanoindention tests, is 11.4 ± 0.7 GPa for Ti3AlC2 and 10.2 ± 0.6 GPa for Ti3Al0.8Sn0.2C2. Furthermore, hardness values, measured by microindentation tests, are compared with the ones measured by nanoindenta- tion. It is shown that microindentation and nanoindentation performed with large loads lead to underestimated hardness values as several grains are involved in the deformation pro- cess. Finally, Young's modulus of Ti3AlC2 and Ti3Al0.8Sn0.2C2 is 260 ± 10 and 250 ± 10 GPa, respectively.

Starting from a mixture of Al–Cu–Fe quasicrystalline (QC) particles and Al powder, a fully dense and almost Al–Cu–Fe ω single-phase alloy was produced by spark plasma sintering. This technique allows synthesising large samples with sizes suitable for mechanical spectroscopy experiments. Mechanical spectroscopy was selected because it is a relevant tool for detecting the presence of structural defects at both nano and microscopic scales. Young’s moduli were measured in the 15 kHz range as a function of temperature by the resonant frequency method. Young’s moduli behave similarly for typical metals and exhibit values that are comparable to those of the Al–Cu–Fe QC phase. The damping coefficient Q−1 was determined at various temperatures between room temperature and 840 K over a large frequency range, i.e. between 10−3 and 10 Hz. The results suggest that solid friction effects do occur. In addition, a relaxation peak is observed in the intermediate temperature range.

The plasticity of Ti3SnC2, a recently synthesized MAX phase, was investigated. Localized deformation was induced by nanoindentation in a polycrystalline sample, and the resulting surface topography was observed by atomic force microscopy (AFM). For several grains, buckling around the indent was observed, in agreement with the kink band deformation process often reported for MAX-phase materials. For other grains, slip lines have been revealed by AFM. The corresponding slip systems have been identified through the determination of the local crystalline orientation by electron backscatter diffraction. First- and second-order pyramidal slip systems are shown to be active for some grain orientations, as well as dislocation interactions and cross slip from one system to the other.

Ti 3 Sn (1-x) Al x C 2 MAX phase solid solutions are successfully synthesized from different reactant mixtures. Rietveld refinement allows to carefully characterize their structures and the ocathedra and trigonal prims distortion parameters as a function of the Al content. Mechanical properties of solid solutions are studied from nanoindentation experiments and dynamic resonant method. It is shown that solid solution hardening is not operative in this system. Elastic modulus is found to increase from Ti 3 SnC 2 to Ti 3 AlC 2 , and such a result is discussed in terms of Ti-A bond stiffness.

Synthesis of ω-Al–Cu–Fe single phase material is reported. Microhardness tests performed over the temperature range 293–898 K indicate a brittle-to-ductile transition between 673 K and 823 K. Fracture toughness was determined from indentation cracks at 293 K. Both the hardness and fracture toughness are comparable to that of the icosahedral Al 0.635 Cu 0.240 Fe 0.125 material.

Nanoindentation is used to measure hardness of Ti 3 SnC 2 , a recently synthesized nanolaminate carbide. Hardness cartography and optical microscopy are used to correlate hardness values and sample phases. An indentation size effect is observed, and a bulk hardness of 9.3 GPa is deduced for Ti 3 SnC 2 . Such a value is higher than most of hardness values reported for other nanolaminate carbides and nitrides. It is shown that an underestimation of the nanolaminate carbide hardness values likely results from the effect of grain boundaries.

Reactive sintering of 8Ti:Al4C3:C powder mixtures to form the ternary carbide Ti2AlC is studied in the temperature range 570–1400 °C. After sintering at 1400 °C for 1 h, only the MAX phase Ti2AlC and some TiC are produced. A series of intermediate phases, such as TiC, Ti3Al, Ti3AlC are detected during the reactive sintering process. From X-ray diffraction (XRD) and scanning electron microscopy (SEM) characterizations, a reaction path is proposed for the intermediate phases and Ti2AlC formation. Results show that reaction kinetics may play an important role in the understanding of the reaction mechanisms.

The present paper focuses on the investigation of the relationship between microstructure of Fe3Al prepared by hot isostatic pressing (HIP) and kinetics of alumina layer formation during oxidation at 900 °C, 1000 °C and 1100 °C. As prepared HIPed Fe3Al sample reveals lamellar microstructure with inhomogeneous Al distribution which originates from the preliminary mechanical activation of Fe–Al mixture. At 900 °C, Fe3Al oxidation is characterized by selective growth of very rough alumina layer containing only transient aluminium oxides. In addition to these transient oxides, α-Al2O3 stable phase is formed at 1000 °C. At the highest temperature (1100 °C), continuous and relatively smooth alumina layer mainly contains fine crystallites of α-Al2O3. The initial lamellar structure and phase inhomogeneity in as-HIPed Fe3Al samples are supposed to be the main factors that determine observed peculiarities after Fe3Al oxidation at 900 °C and 1000 °C.

A Fluidized Bed Metal Organic Chemical Vapour Deposition (FB-MOCVD) process has been successfully applied to coat Fe particles with Al. N 2 inert gas was used to transport the organometallic Al precursor (liquid triethylaluminium) inside the fluidized bed reactor whose temperature is 350 °C. XPS analyses and TEM observations have revealed the presence of a thin alumina layer surrounding the Al coating. Oxidation treatments, performed in the temperature range 350–500 °C, demonstrate that this multi-scale coating constitutes an efficient barrier to protect iron particles against oxidation. Such a treatment may be used to perform environmental barrier coatings around magnetic powders.

Metal-matrix composites are produced from Al powder and 30 vol% of icosahedral Al–Cu–Fe quasi-crystalline particles using a hot isostatic pressing technique. It is demonstrated that the initial icosahedral phase is transformed into the ω-Al70Cu20Fe10 tetragonal phase during the hot isostatic pressing (HIP) process. The mechanical properties of the composite were evaluated over the temperature range 293 to 773 K by performing compression tests at constant strain rate. The temperature dependence of the yield stress gives evidence of two temperature regimes with a transition temperature at approximately 423 K. Strain-rate sensitivity measurements support the change in rate-controlling deformation mechanisms at this temperature. It is proposed that cross-slip and/or climb mechanism control plastic flow. Finally, it is suggested that the phase transformation of the particle contributes positively to the improvement of the mechanical properties.

The electronic conductivity anisotropy of $\mathrm{T}{\mathrm{i}}_{3}\mathrm{Si}{\mathrm{C}}_{2}$ is directly evidenced from data collected on (i) a thin film epitaxially grown on a $(11\overline{2}0)$-oriented SiC single crystal and (ii) a single crystal. Density functional theory calculations, including the linear-response approach and coupled to a Bloch-Gr\"uneisen model, show that the electron-phonon interactions are mainly responsible for the observed anisotropy. Detailed analysis of the electron-phonon coupling constants allows for the rationalization of these scattering processes in terms of the $\mathrm{T}{\mathrm{i}}_{3}\mathrm{Si}{\mathrm{C}}_{2}$ nanostructure, giving insights into the possibility of modifying the electron-phonon interaction in this system by substitution effects.

TiCx–Al2O3 composites have been synthesized by pressure-assisted combustion synthesis of (Ti + C + Al2O3) reactant powder. Different alumina contents (10–40 vol%) have been investigated to study the dilution effect on TiC microstructure. A mechanical method and a mixed chemical/mechanical method have been used to obtain (Ti + C + Al2O3) powder mixtures with different alumina distributions. Scanning electron microscopy (SEM) observations of these mixtures show that alumina is distributed inside micrometric (Ti + C) aggregates for the first method whereas alumina is located around (Ti + C) aggregates for the second one. X-ray diffraction (XRD) and SEM analyses of the composites indicate that TiCx is substoichiometric in carbon and mainly consists of submicrometric grains. A distribution of Al2O3 inside (Ti + C) aggregates is more efficient to reduce TiC grain size. For the 40 vol% Al2O3 diluted (Ti + C) mixture prepared from the mechanical route, TiCx nanocrystallites have been successfully stabilized, which demonstrates that the addition of Al2O3 diluent in a (Ti + C) mixture can be efficiently used to inhibit grain growth.

International audience; Thermite reactions in the NiO-Al system have been studied. In addition to the case when the stoichiometry is set up to produce metallic nickel and alumina, we studied the case when Al is added in excess in order to react with the Ni produced through the reduction of nickel monoxide with Al to produce various nickel aluminides. As thermite reactions are highly exothermic, in order to provide a better understanding of the reactions, alumina has been added to the green mixture to reduce the reaction rate and overall exothermicity, in amounts corresponding up to 50% of the overall heat capacity of the sample.

The microstructure formation of TiC during self-propagating high-temperature synthesis (SHS) reactions is investigated thanks to the quenching method. The morphology and chemical composition of titanium carbide grains are characterized, during the growth phase, from X-ray diffraction (XRD), scanning electron microscopy (SEM) and wavelength dispersion spectroscopy (WDS) analyses. It is demonstrated that TiC x grains nucleate with a great deficit in carbon; the carbon composition evolves towards the stoichiometry during the growth phase. The TiC grain sizes are then coupled to thermal data acquired by infra red (IR) thermography. It is shown that the convolution of the temperature and of the time elapsed in the temperature range T max −1000 K constitute a pertinent parameter to give account of the variation of TiC grain sizes.

The mechanofusion process, a dry particle coating route, has been successfully applied to coat iron particles with alumina. The high-temperature oxidation resistance of Fe and coated Fe particles has been comparatively investigated in air in the temperature range 350°–900°C. The activation energy calculations demonstrate that, in the temperature range 350°–420°C, the oxidation mechanisms are the same for both particles. Scanning electron microscopy and energy-dispersive spectroscopy analyses show that oxidation products are mainly formed at the iron/alumina interface at low temperatures or at the alumina/gas interface at high temperatures. It is shown that, depending on the temperature range, two different oxidation mechanisms are predominant.

Ti 3 SiC 2 is synthesized by self-propagating high-temperature synthesis (SHS) of elemental titanium, silicon, and graphite powders. The reaction paths and structure evolution are studied in situ during the SHS of the 3Ti+Si+2C mixture by time-resolved X-ray diffraction coupled with infrared thermography. The proposed reaction mechanism suggests that Ti 3 SiC 2 might be formed from Ti-Si liquid phase and solid TiC x . Finally, the effect of the powders starting composition on the Ti 3 SiC 2 synthesis is studied. For the investigated initial mixtures, TiC x is always formed as a major impurity together with the Ti 3 SiC 2 phase.

A simple method to calculate thermal diffusivity in situ after a combustion synthesis reaction is presented. The combustion reaction was analyzed via time-resolved X-ray diffraction analysis and infrared thermography. Thermal diffusivity was estimated and used to calculate temperature profiles based on temperature profiles one second earlier. For a sample of TiC formed from Ti and C, a value of 2.00×10-6±0.20×10-6 m2 s-1 was calculated for temperatures between 1000 and 1900 K. This method is rapid and can avoid some problems associated with furnace-based measurements of thermal diffusivity, such as recrystallization and destruction of non-equilibrium phases.

Background: There appears little consensus concerning protein requirements in phenylketonuria (PKU). Methods: A questionnaire completed by 63 European and Turkish IMD centres from 18 countries collected data on prescribed total protein intake (natural/intact protein and phenylalanine-free protein substitute [PS]) by age, administration frequency and method, monitoring, and type of protein substitute. Data were analysed by European region using descriptive statistics. Results: The amount of total protein (from PS and natural/intact protein) varied according to the European region. Higher median amounts of total protein were prescribed in infants and children in Northern Europe (n = 24 centres) (infants 2-3 g/kg/day; 1-3 years of age, >2-3 g/kg/day; 4-10 years of age, >1.5-2.5 g/kg/day) and Southern Europe (n = 10 centres) (infants 2-2.5 g/kg/day; 4-10 years of age, >1.5-2 g/kg/day) and with Western Europe (n = 25 centres) giving the least (infants 2-2.5 g/kg/day, 1-3 years of age, 1.5-2 g/kg/day; 4-10 years of age, 1-1.5 g/kg/day). Total protein prescription was similar in patients aged >10 years (1-1.5 g/kg/day) and maternal patients (1-1.5 g/kg/day). Conclusions: The amounts of total protein prescribed varied between European countries and appeared to be influenced by geographical region. In PKU, all gave higher than the recommended 2007 WHO/FAO/UNU safe levels of protein intake for the general population. (C) 2015 Elsevier Inc. All rights reserved.

Mesoscopic superconducting lead nanowires with high aspect ratio and diameter ranging from 40 to 270 nm have been grown by electrodeposition inside nanoporous polycarbonate membranes. Nanowires with a diameter less than 50 nm were insulators due to a poor crystal structure. The others are shown to be type II superconductors because of their small electronic mean free path, instead of being type I which is usual for the bulk form of lead. An increase in the thermodynamic critical field H-c is observed and is attributed to the small transversal dimension leading to an incomplete Meissner effect. Finally, it is demonstrated that this enhancement agrees with numerical simulations based on the Ginzburg-Landau theory. (C) 2002 Elsevier Science B.V. All rights reserved.

Resistivity measurements have been performed on silicon powder beds with different thicknesses which were submitted to uniaxial pressure (0-640 MPa). For the smallest thicknesses of the Si pellets, the pressure dependence of conductivity may be described using models based on effective medium theory or strongest stresses network. For largest Si thicknesses, it exists an inhomogeneous distribution of pressure in the granular medium as a consequence of arching effects. It is shown that the conductivity variation with Si powder thickness may be understood using a distribution of conductivity induced by the distribution of pressure. Finally, it is shown that the conductivity of the sintered Si pellets is correlated with the uniaxial compression step.

Self-propagating high-temperature synthesis was performed on an equimolar mixture of titanium and graphite embedded in a copper tube of variable thickness. Such an experimental setup allows us to study the interaction between the reaction front and copper. Special attention was focused on copper melting, copper penetration within the sample, and quenching of the sample: it is shown that a minimum of 1 mm for the copper thickness is necessary to ensure the physical integrity of the surrounding metal. Nevertheless, reaction is not complete under these conditions. As a consequence, copper is not the most appropriate metal for field assisted combustion synthesis electrodes. As for thermal explosion and high isostatic pressing combustion synthesis, some valuable insights are provided, and a metal with slightly lower thermal diffusivity, lower electrical conductivity, and/or higher melting temperature would likely be a good candidate for such applications.

Titanium carbide was synthesized from self-propagating high temperature synthesis (SHS) performed under isostatic gas pressure ranging from 70 bars up to 700 bars. Propagation conditions of the combustion reaction have been studied as a function of gas pressure for C reactant of different diameters (20 μm and 7 μm). For coarse particles, the reaction stops upon increasing pressure, while for fine particles, reaction self-propagates up to 700 bars. It is shown that this change in propagation conditions can be explained by considering both thermal and kinetic effects. Finally, TiC formation mechanisms are discussed for SHS performed under high isostatic pressure.

Macroscopic resistivity measurements have been performed on TiC/Al2O3 random mixtures submitted to uniaxial compression (0-95 kN). Such a random mixture exhibits an insulator-conductor transition which appears at increasing force while decreasing the conductive composition of the TiC/Al2O3 mixture. It is demonstrated that the conductivity behavior may be understood in the framework of a site percolation model. Finally, the rate of creation of the contacts between conductive grains is extracted from the macroscopic resistivity measurements.

The magnetic anisotropy and domain structure of electrodeposited cylindrical Co nanowires with length of 10 or 20 μm and diameters ranging from 30 to 450 nm are studied by means of magnetization and magnetic torque measurements, as well as magnetic force microscopy. Experimental results reveal that crystal anisotropy either concurs with shape anisotropy in maintaining the Co magnetization aligned along the wire or favours an orientation of the magnetization perpendicular to the wire, hence competing with shape anisotropy, depending on whether the diameter of the wires is smaller or larger than a critical diameter of 50 nm. This change of crystal anisotropy, originating in changes in the crystallographic structure of Co, is naturally found to strongly modify the zero (or small) field magnetic domain structure in the nanowires. Except for nanowires with parallel-to-wire crystal anisotropy (very small diameters) where single-domain behaviour may occur, the formation of magnetic domains is required to explain the experimental observations. The geometrical restriction imposed on the magnetization by the small lateral size of the wires proves to play an important role in the domain structures formed.

Titanium carbide was synthesized from self-propagating high-temperature synthesis performed under isostatic gas pressure. Carbon and titanium particles of mean diameters of 7 or 20 μm and 17 μm, respectively, were used as reactants. The final product was analyzed by X-ray diffraction and scanning electron microscopy. Porosity was carefully determined by different methods. For a reactant mixture with coarse C particles, the large value of the closed porosity (22%) is shown to be compatible with the emptying core model. The origin of such a predominant mechanism in pure C and Ti mixture is tentatively discussed.

The uniaxial anisotropy energy of arrays of submicronic (30-500 hm) Ni wires synthesized by electrodeposition into cylindrical pores of track-etched polycarbonate membranes is studied as a function of temperature. At room temperature, the uniaxial anisotropy is equal to the shape anisotropy whereas an additional contribution, that reinforces the wire axis as an easy axis is evidenced at low temperature. This additional contribution is demonstrated to find its origin in magnetoelastic effects conjointly induced by the Ni and polycarbonate thermal-expansion coefficients mismatch and by the low volume fraction of Ni in Ni/polycarbonate samples.

The magnetization reversal of Ni nanowires was studied by anisotropic magnetoresistance measurements at temperatures between 15 and 300 K. The wires, synthesized by electrodeposition in a nanoporous polycarbonate membrane, are regular cylinders 22 mu m long with a diameter of 75 or 35 nm. The nucleation field was measured on individual nanowires as a function of the angle between the applied field and the wire axis. The results are quantitatively analyzed using classical magnetization reversal theories. Measurements of the nucleation field first obtained as a function of temperature evidence an extra uniaxial anisotropy induced by the contraction of the membrane at low temperature. Combining SQUID measurements and x-ray diffractometry at different temperatures, a clear picture of the large magnetoelastic effect was obtained. (C) 2000 American Institute of Physics. [S0021-8979(00)03902-5].