D.L. Beke

Investigation of the interplay of nickel dissolution and copper segregation in Ni/Cu(111) system

Abstract The effect of segregation on the dissolution of a thin Ni layer (3–14 equivalent-monolayers (eq-ML)) into a semi-infinite Cu substrate has been investigated at different temperatures (600–730 K) by Auger electron spectroscopy. Computer simulations have also been carried out to understand the details of the dissolution and a sophisticated method is proposed to analyse the experimental kinetics. The simulations indicated a step-wise character of dissolution and an interesting interference between segregation and dissolution. Because of the strong concentration dependence (there is a fast diffusion in Cu and there is practically no diffusion into Ni), the interface remains sharp and shifts proportionally with time until it does not reach the next-to-the-last Ni layer. Then the dissolution continues by the saturation of Cu in the topmost layer, according to the segregation isotherm. After the complete coverage of the surface by Cu, the final homogenisation takes place by the total dissolution of the next-to-the-last layer. According to these results, it was possible to separate the dissolution from the segregation and to determine the average speed of the interface shift, which is proportional to the intrinsic diffusion coefficient.

The temperature dependence of grain-boundary diffusion of 65Zn in polycrystalline aluminium

Abstract It is shown that the penetration plots of the grain-boundary diffusion of 65Zn in polycrystalline Al samples are curved at fixed temperatures and that this upward curvature changes with the annealing time. The penetration profiles can be interpreted in terms of sub-boundary diffusion characterized by the parameter P 1 (= D 1 K 1δ, where D 1 is the boundary diffusion coefficient, K is the segregation factor and δ is the grain-boundary width) as well as by a distribution of diffusion parameters for high-angle boundaries. The range of variation of the diffusion parameters for general large-angle grain boundaries can be characterized by the parameters P 1, P 3 and P 4. The temperature dependence of P 1, P 2 and P 3 has also been determined. It is concluded that if we approximate the penetration function by one effective slope in the B-kinetic regime of grain-boundary diffusion then the weight of different grain boundaries (with different grain-boundary diffusion coefficients) can vary greatly in the ...

Segregation inhibited grain coarsening in nanocrystalline alloys

It will be shown, in the framework of a general statistical (atomistic) treatment, that segregation inhibited grain coarsening in binary nanocrystalline alloys can be described by analytical expressions, if the effect can be attained with highly saturated grain boundaries. Relatively simple relations describing the temperature dependence of the stabilized grain-boundary fraction will be derived. The validity conditions of the relation proposed very recently by R. Kirchheim [Acta Mater. 50, 413 (2002)] are also analyzed. Our conclusions are compared with experimental results and the reliability of the few experimental data, known at present, will be discussed as well.

On the transition between the C and B kinetic regimes for grain-boundary diffusion

Abstract It is shown that, in the transitional region between the B and C kinetic regimes for grain-boundary diffusion, the product of the grain-boundary width δ, the segregation factor K and the grain-boundary diffusion coefficient D b= KδD b), as well as the grain-boundary diffusion coefficient itself, can be estimated. From penetration plots measured at different annealing times in the transitional region (0.01 < α < 10, where α = Kδ/2(Dt)1/2 is the relative grain-boundary width and D is the volume diffusion coefficient) by fitting with solutions in the B and C kinetic regimes, one can obtain the apparent values for P and D b. From the solution of the general problem of grain-boundary diffusion, correcting functions are given for the calculation of the true values of P and D b (and consequently of Kδ). This procedure is illustrated by several experimental examples published in the literature as measurements in the C kinetic regime.

On the diffusion of 59Fe into aluminium and Al Mn solid solutions

Abstract The diffusion of 59Fe into pure Al and Al Mn solid solutions has been measured by a tracer-sectioning technique. The penetration plots are fitted by a solution of the diffusion equation according to which the surface flux is controlled by a surface barrier (characterizing the dissolution of tracer atoms from the intermetallic phases formed in the near-surface layer) higher than the diffusion activation energy in the bulk. For annealing times t << τ(τ is the time required to reach the solubility limit just inside the solvent) the solution reduces to that of constant surface concentration, and at much longer t (when the surface layer is depleted) the thin-film solution can also be valid. Detailed measurements have been carried out to investigate the time dependence of the penetration plots; the diffusion coefficients calculated from the above solution were the same (within experimental error) at fixed temperature. The effect of increased grain-boundary and dislocation density in the near-surface la...

Ni diffusion in near-equiatomic Ni-Ti and Ni-Ti(-Cu) alloys

Tracer volume diffusion of 63 Ni in Ni-50.07 at.% Ti binary and Ni-48.83 at.% Ti-9.29 at.% Cu ternary alloys have been measured between 783 and 1288 K. The temperature dependence of the diffusion coefficients can be well described by a straight Arrhenius function This illustrates that a diffusional anomaly (typical in some bcc metallic alloys and related to the well-known phonon softening), if there is any in this system, is negligible. On the other hand, the small activation energy (about half the value expected from simulations for the commonly accepted mechanism with thermally activated vacancies) shows that the mechanism of diffusion is probably mediated by structural vacancies.

Diffusion of 65Zn in dilute AlCu, AlCuZn, AlSi, and AlSiZn alloys

Abstract The diffusion coefficients of 65Zn have been measured in dilute AlCu, AlCuZn, AlSi and AlSiZn alloys as a function of the concentration and temperature in the range of solid solutions. The 65Zn diffusion coefficients depend linearly on the Cu and Si concentration and the enhancement effects of the different solutes investigated are additive. The experimental values–together with previous results for 65Zn diffusion in dilute AlZn, AlMg and AlZnMg alloys (Beke et al. 1977)–are interpreted on the generalization of the model given by Le Claire (1978) and it is found that the tracer jump frequencies in the vicinity of the solutes change only by a factor less than 2. Furthermore, supposing that the jump frequencies are unaltered and the solute-solute binding are also negligible, an approximate expression can be obtained for the enhancement factor which contains only the solute-vacancy binding energy B t,v. From this relation the following approximative values were obtained B zn,v = 0.03 ± 0.01 eV, B Mg...

Nanoscale investigations of shift of individual interfaces in temperature induced processes of Ni–Si system by secondary neutral mass spectrometry

We describe a method for measurement of nanoscale shift of interfaces in layered systems by a combination of secondary neutral mass spectrometry and profilometer. We demonstrate it by the example of the investigation of interface shifts during the solid state reaction in Ni/amorphous-Si system. The kinetics of the shrinkage of the initial nanocrystalline Ni film and the amorphous Si layer as well as the average growth kinetics of the product phases were determined at 503 K. The results show that nanoscale resolution can be reached and the method is promising for following solid state reactions in different thin film systems.

Formation of CuxAu1−x phases by cold homogenization of Au/Cu nanocrystalline thin films

Summary It is shown, by using depth profiling with a secondary neutral mass spectrometer and structure investigations by XRD and TEM, that at low temperatures, at which the bulk diffusion is frozen, a complete homogenization can take place in the Cu/Au thin film system, which leads to formation of intermetallic phases. Different compounds can be formed depending on the initial thickness ratio. The process starts with grain boundary interdiffusion, which is followed by a formation of reaction layers at the grain boundaries that leads to the motion of the newly formed interfaces perpendicular to the grain boundary plane. Finally, the homogenization finishes when all the pure components have been consumed. The process is asymmetric: It is faster in the Au layer. In Au(25nm)/Cu(50nm) samples the final state is the ordered AuCu3 phase. Decrease of the film thicknesses, as expected, results in the acceleration of the process. It is also illustrated that changing the thickness ratio either a mixture of Cu-rich AuCu and AuCu3 phases (in Au(25nm)/Cu(25nm) sample), or a mixture of disordered Cu- as well as Au-rich solid solutions (in Au(25nm)/Cu(12nm) sample) can be produced. By using a simple model the interface velocity in both the Cu and Au layers were estimated from the linear increase of the average composition and its value is about two orders of magnitude larger in Au (ca. 10−11 m/s) than in Cu (ca. 10−13 m/s).

Determination of grain-boundary diffusion coefficients by Auger electron spectroscopy

Abstract Surface accumulation method, called the Hwang–Balluffi method [J.C.M. Hwang, J.D. Pan and R.W. Balluffi, J. Appl. Phys., 50(3), 1979, 1339], was applied to measure the grain-boundary diffusion of Ag at low temperatures (413 and 428 K) in a nanocrystalline Cu film. Ag atoms from the Ag layer diffused through the copper nanocrystalline film along the grain boundaries to the opposite surface (i.e. the accumulation surface) where they spread out by rapid surface diffusion and accumulated. The rate of accumulation was detected by Auger Electron Spectroscopy (AES). It was shown that the results are not sensitive to the supposition whether accumulation takes place in one or two monolayers of the surface. TEM observations have been made in the same time before and after heat treatment to check the stability of the nanostructure.