Self-diffusion In Copper Research Articles

The Impact of Hydrogen Gas Evolution on Blister Formation in Electroless Copper Films

Electroless copper plating is a key process to provide electrically conductive layers on insulating substrates for the subsequent copper electroplating of printed circuit boards (PCBs). Recently introduced substrate materials are prone to spontaneous delamination failure (blistering) of the electroless layer during deposi-tion. A higher nickel content in electroless copper baths prevents delamination failure by increasing the internal tensile stress in the copper layer. This effect is achieved by suppressing copper self-diffusion through the incorporation of small amounts of nickel in the grain boundaries and the grain boundary junc-tions. Whether the hydrogen gas that evolves inevitably during electroless copper deposition, and that is potentially trapped in the deposit may cause blistering is still under debate. Here we test this hypothesis by quantifying the generated hydrogen gas volume during copper deposition in baths with nickel concentra-tions between 0 and 1000 ppm, and observing the rate of blister formation in the deposit. We find a con-stant molar ratio of 0.65±0.05 mol H2/mol Cu independent of the nickel content in the bath, in contrast to a systematic and substantial change of the blister probability. From this lack of a correlation between the blister rate and hydrogen evolution, we rule out hydrogen evolution as a determining factor for blister for-mation.

Interface Controlled Diffusional Creep of Cu + 2.8 at.% Co Solid Solution

High-temperature creep experiments were performed on a Cu-2.8 ат.% Co solid solution. Cylindrical foils of 18 micrometers thickness were used for this purpose. Creep tests were performed in a hydrogen atmosphere in the temperature range of about from 1233 K to 1343 K and at stresses lower than 0.25 MPa. For comparison, a foil of pure copper and Cu-20 at.% Ni solid solution were investigated on high temperature creep. Measurements on the Cu foil showed classical diffusional creep behavior. The activation energy of creep was defined and turned out to be equal 203 kJ/mol, which is close to the activation energy of bulk self-diffusion of copper. There was a significant increase in activation energy for the Cu-20 at.% Ni solid solution. Its activation energy was about 273 kJ/mol. The creep behavior of Cu-Co solid solution was more complicated. There were two stages of diffusional creep at different temperatures. The extremely large activation energy (about 480 kJ/mol) was determined at relatively low temperature and a small activation energy (about 105 kJ/mol) was found at high temperatures. The creep rate of Cu-Co solid solution was lower than that of pure copper at all temperatures. In addition, the free surface tension of Cu-2.8 ат.% Co was measured at different temperatures from 1242 K to 1352 K. The surface tension increases in this temperature range from 1.6 N/m to 1.75 N/m. There were no features on the temperature dependence of the surface tension.

Impression creep behavior of a Cu–6Ni–2Mn–2Sn–2Al alloy

The impression creep behavior of the aging-treated Cu–6Ni–2Mn–2Sn–2Al alloy and its variants with separate additions of 0.3 wt.% Cr and 0.3 wt.% Zr was investigated in the temperature range 695–795 K and under constant punch stress in the range 300–900 MPa. It was shown that simple theory, based on a steady-state power-law creep equation, has the capacity to describe the impression creep data satisfactorily. The average stress exponents of about 4 together with the activation energies of 194–196.9 kJ mol−1, which are close to 197 kJ mol−1 for lattice self diffusion of copper, suggests that in all of the tested alloys the operative creep mechanism is dislocation climb controlled by lattice diffusion. The creep resistance of the base material was better than that of the Cu–0.3Cr–0.1Ag alloy reported in the literature. Further enhancement in creep resistance was achieved by incorporating Cr and Zr into the base alloy. This was achieved by the precipitation of Cr-rich and Zr-rich particles in the Cu matrix of the Cr- and Zr-containing alloys, which act as dispersion strengtheners in the present alloys.

Determination of Copper Self-Diffusion Coefficients on the Base of High-Temperature Creep Data

Indirect method is used for determination of copper self-diffusion coefficient on the base of high-temperature creep data. These data are obtained by the use of the new measuring device and assuming that Nabarro-Herring mechanism acts.

Characterization of planar optical waveguides formed by copper–sodium, electric field assisted, ion exchange in glass

Planar waveguides have been formed by the electric field assisted diffusion of copper ions from a metallic copper film into a borosilicate glass (Corning 7740). The waveguides have been analysed using conventional prism coupling techniques and have been compared with the electric field assisted diffusion of silver into the same glass type. The measured refractive index profiles for copper guides are shown to have nearly Fermi function shapes, which are consistent with the theory of field assisted ion diffusion in glass. It is shown, however, that the guide depth formed by copper or silver ions is larger than that predicted by calculations based on the nominal composition, possibly indicating that only a fraction of the sodium ions within the glass are mobile. Annealing experiments have been performed on the guides in order to determine the copper self-diffusion coefficient in the glass. The coefficients are extracted from the data by fitting a non-linear diffusion equation model of ion motion in glass to the experimental profiles.

Mechanisms of high temperature deformation in electrolytic copper in extended ranges of temperature and strain rate

Abstract The hot deformation behavior of electrolytic copper in isothermal compression has been studied in the temperature range 300–950 °C and strain rate range 0.001–100 s −1 with a view to evaluate the rate controlling mechanisms relevant to different ranges. The flow curves observed are typical of the occurrence of dynamic recrystallization (DRX) and have exhibited single or multiple peak in the flow stress before reaching steady state, depending on the temperature and strain rate. Three different windows of temperature and strain rate have been identified in which the kinetic rate equation is obeyed by the temperature and strain rate dependence of flow stress. Both the power-law and hyperbolic sine relation yielded similar results. In the strain rate range 0.001–1.0 s −1 and temperature range 400–600 °C, the stress exponent is 7.8 and the apparent activation energy is 159 kJ/mole, and these indicate that dislocation core diffusion is the rate controlling mechanism. In the strain rate range 0.001–1.0 s −1 and temperature range 700–950 °C, the stress exponent is 7.1 and the apparent activation energy is 198 kJ/mole which matches well with that for lattice self diffusion in copper. The switching of the rate controlling mechanisms at temperatures higher than 700 °C has been attributed to the “clogging” of the dislocation pipes due to the rapid increase in the solid solubility of oxygen at higher temperatures. In the high strain rate range 30–100 s −1 and temperature range 700–950 °C, the stress exponent is 3.5 and the apparent activation energy is 91 kJ/mole which matches with that for grain boundary diffusion in copper. The apparent activation energy in electrolytic copper in all the above temperature and strain rate windows increases with increasing oxygen content and this is attributed to the increase in the back stress due to the presence of copper oxide particles in the matrix.

Mechanisms of high temperature deformation in electrolytic copper in extended ranges of temperature and strain rate

The hot deformation behavior of electrolytic copper in isothermal compression has been studied in the temperature range 300–950 °C and strain rate range 0.001–100 s −1 with a view to evaluate the rate controlling mechanisms relevant to different ranges. The flow curves observed are typical of the occurrence of dynamic recrystallization (DRX) and have exhibited single or multiple peak in the flow stress before reaching steady state, depending on the temperature and strain rate. Three different windows of temperature and strain rate have been identified in which the kinetic rate equation is obeyed by the temperature and strain rate dependence of flow stress. Both the power-law and hyperbolic sine relation yielded similar results. In the strain rate range 0.001–1.0 s −1 and temperature range 400–600 °C, the stress exponent is 7.8 and the apparent activation energy is 159 kJ/mole, and these indicate that dislocation core diffusion is the rate controlling mechanism. In the strain rate range 0.001–1.0 s −1 and temperature range 700–950 °C, the stress exponent is 7.1 and the apparent activation energy is 198 kJ/mole which matches well with that for lattice self diffusion in copper. The switching of the rate controlling mechanisms at temperatures higher than 700 °C has been attributed to the “clogging” of the dislocation pipes due to the rapid increase in the solid solubility of oxygen at higher temperatures. In the high strain rate range 30–100 s −1 and temperature range 700–950 °C, the stress exponent is 3.5 and the apparent activation energy is 91 kJ/mole which matches with that for grain boundary diffusion in copper. The apparent activation energy in electrolytic copper in all the above temperature and strain rate windows increases with increasing oxygen content and this is attributed to the increase in the back stress due to the presence of copper oxide particles in the matrix.

Influence of oxygen on rate-controlling mechanisms in hot deformation of polycrystalline copper: oxygen-free versus electrolytic grades

The kinetics of hot deformation in oxygen-free and electrolytic grades of polycrystalline copper has been studied in a wide temperature range (300–950 °C) and strain rate range (0.001–100 s −1) using isothermal hot compression tests. The flow curves are typical of the process of dynamic recrystallization with initial peak in the flow stress followed by flow softening. At lower strain rates and higher temperatures, multiple stress peaks have been observed. In the lower temperature range (400–600 °C) and strain rate range (0.001–1 s −1), both grades of copper yielded an apparent activation energy of 159 kJ/mol, suggesting that dislocation core self-diffusion is the rate-controlling mechanism. However, in the higher temperature range (700–950 °C), the electrolytic copper exhibited higher apparent activation energy (198 kJ/mol), indicating that the mechanism has switched to that controlled by lattice self-diffusion. This switching of mechanism in electrolytic copper is explained in terms of “clogging” of dislocation pipes by the interstitial oxygen atoms at higher temperatures due to a rapid increase in the solid solubility of oxygen in copper beyond 700 °C. In the higher strain rate range (3–100 s −1) and high temperature range (700–950 °C), the apparent activation was much lower (about 91 kJ/mol) irrespective of the oxygen grade and supports a mechanism involving grain boundary self-diffusion in copper.

High temperature creep behavior of in-situ TiB 2 particulate reinforced copper-based composite

The tensile creep behavior of in-situ TiB 2 particulate reinforced copper-based (TiB 2/Cu) composite and unreinforced monolithic Cu fabricated by the powder metallurgy process was investigated at 773–873 K. The creep rate of monolithic Cu exhibited an exponential dependence on the applied stress, whereas the power-law relationship prevailed for the experimental data of the TiB 2/Cu composite. The composite exhibited a stress exponent of 25.5, 20.6 and 22.1 at 773, 823 and 873 K, respectively. Furthermore, the unreinforced Cu and in-situ TiB 2/Cu composite exhibited a creep activation energy of 187 and 444 kJ/mol, respectively. It is indicated, by incorporating a threshold stress into the analysis, that for the in-situ TiB 2/Cu composite, the true stress exponent was equal to 4.8 and the true activation energy was close to that for self-diffusion of copper, indicating that the creep of the composite is controlled by the lattice diffusion.

An experimental study of copper self-diffusion in CuO, Y2Cu2O5 and YBa2Cu3O7−x by secondary neutral mass spectrometry

Download options Please wait... Article information DOI https://doi.org/10.1039/A702798B Article type Paper Download Citation J. Mater. Chem., 1997,7, 2085-2089 BibTex EndNote MEDLINE ProCite ReferenceManager RefWorks RIS Permissions Request permissions An experimental study of copper self-diffusion in CuO, Y2Cu2O5 and YBa2Cu3O7–x by secondary neutral mass spectrometry J. A. Rebane, N. V. Yakovlev, D. S. Chicherin, Yuri.d. Tretyakov, L. I. Leonyuk and V. G. Yakunin, J. Mater. Chem., 1997, 7, 2085 DOI: 10.1039/A702798B To request permission to reproduce material from this article, please go to the Copyright Clearance Center request page. If you are an author contributing to an RSC publication, you do not need to request permission provided correct acknowledgement is given. If you are the author of this article, you do not need to request permission to reproduce figures and diagrams provided correct acknowledgement is given. If you want to reproduce the whole article in a third-party publication (excluding your thesis/dissertation for which permission is not required) please go to the Copyright Clearance Center request page. Read more about how to correctly acknowledge RSC content. Search articles by author Jan A. Rebane Nikolay V. Yakovlev Dmitry S. Chicherin Yuri.d. Tretyakov Lidia I. Leonyuk Valery G. Yakunin

Experimental study of copper diffusion in YBa2Cu3O7–xpolycrystals

The self-diffusion of copper in YBa2Cu3O7–x ceramics has been studied using sputtered neutrals mass spectrometry with isotopic marking. Diffusion profiles were formed by the annealing of 63CuO thin films deposited on YBa2Cu3O7–x polycrystal substrates. The effective diffusion coefficient was calculated by an error-function approach. The parameters of the temperature dependence were found to be D0=(2 ± 0.2)× 10–1 cm2 s–1 and Ea= 240 ± 9 kJ mol–1 in the temperature range 730–880 °C. The application of Fisher's grain-boundary diffusion model made it possible to separate contributions of volume and boundary diffusion. The dependence of the copper diffusion coefficient in YBa2Cu3O7–x on P(O2) was investigated.

Point Defects in Materials Part I: Behavior and Characteristics in Different Material Classes

In this issue of the MRS Bulletin we present five papers that involve point defect phenomena in a wide variety of materials—metals, conducting ceramic oxides, semiconductors, amorphous alloys, and high Tc superconducting oxides. The unifying theme of this issue is point defects—zero-dimensional defects. Even for the high Tc oxides, where planar defects are discussed, it is the ordering of oxygen/vacancy chains that ultimately gives rise to twins in the famous YBa2Cu3O7.δ (1:2:3) oxide superconductor.Hillard Huntington, professor emeritus of physics at Rensselaer Polytechnic Institute, is an early and important pioneer in the study of point defects in metals. A theorist, he has also performed many experiments over the years; for example, he performed key early experiments on electromigration effects. Huntington's article presents a historical review of the research on vacancies and self-interstitial atoms in metals during the period that stretches from the mid-1930s to the mid-1960s. He played a crucial role in this field as a result of his seminal theoretical calculations, with Fred Seitz in 1942, on the enthalpies of formation and migration of vacancies or self-interstitial atoms in pure copper. Huntington's and Seitz's calculations indicated that diffusion occurs predominantly by a monovacancy mechanism since the enthalpy of formation of a self-interstitial atom, the [100] split form, also called the dumbbell form, is too large to be compatible with the activation enthalpy for self-diffusion in copper. It is now well established that the latter is given by the sum of the enthalpy of formation and the enthalpy of migration of a monovacancy in many face-centered-cubic (fcc), body-centered-cubic (bec), and hexagonal-close-packed (hcp) metals.

On the grain boundary internal friction peak of ?-brasses

A conventional Ke pendulum was used to investigate the temperature dependence of internal friction for thoroughly annealed samples of Cu−30% Zn. The relaxation peak observed at 330 o C at a frequency of vibration 0.60 Hz was attributed to grain boundary diffusion. The energy activating the process amounted to 1.84 eV and characterized the activation energy for self diffusion in copper. The relaxation strength showed strain amplitude dependence, however, the activation energy and peak temperature did not show any amplitude dependence. The effect of solute atom concentration, cold work and vacancies was also tested

On the grain boundary internal friction peak of α-brasses

Abstract A conventional Ke pendulum was used to investigate the temperature dependence of internal friction for thoroughly annealed samples of Cu30wt.%.Zn. The relaxation peak observed at 300 °C at a vibration frequency of 0.60 Hz was attributed to grain boundary diffusion. The energy activating the process was 1.84 eV and characterized the activation energy for self-diffusion in copper. The relaxation strength showed a strain amplitude dependence, but the activation energy and peak temperature did not. The effects of solute atom concentration, cold work and quenching were also tested.

Diffusion in nanocrystalline material

The paper reports on first investigations of the diffusion in nanocrystalline materials. The self-diffusion of the radioisotope 67Cu in nanocrystalline copper has been measured by serial sectioning with the aid of ion-beam sputtering. The values of the diffusion coefficients which were found at 353 K and 393 K are 2×10 −18 m 2/s and 1.7×10 −17 m 2/s, respectively, i.e., they are about 16 or 14 orders of magnitude larger than the bulk self-diffusion and about 3 orders of magnitude larger than the grain-boundary self-diffusion in copper. In comparison to the bulk, small values for the activation enthalpy, 0.64 eV, and the pre-exponential factor of the self-diffusion coefficient, 3×10 −9 m 2/s, have been observed.

Monovacancy and divacancy contributions to self-diffusion in face-centred cubic metals reanalysis for copper, silver, gold, nickel and platinum

Abstract Results of the Morrison analysis applied to the curved Arrhenius plots of self-diffusion in copper, silver, gold, nickel and platinum are reported. Both temperature-dependent and temperature-independent energies and entropies are taken into consideration. The analysis is applied to the most reliable diffusion coefficients measured with the aid of macroscopic (serial) and microscopic (chemical and electrochemical etching or ion-beam sputtering) sectioning techniques. The error limits of the obtained monovacancy parameters are small, and those of the divacancy parameters in copper, silver and gold are distinctly smaller than in any earlier analysis. The diffusion parameters obtained for gold differ strongly from those hitherto assumed to be valid. According to the present results, the divacancy contribution at the melting temperature amounts to about 50% in copper, silver and platinum, while in gold and nickel the contribution is about 30 and 20%, respectively. The present results also fit the isotope effect and pressure dependence of the self-diffusion coefficients in silver and gold by use of plausible values for the kinetic-energy factors and the activation volumes of diffusion.

Solute enhancement of self-diffusion of copper in copper-tin, copper-indium and copper-antimony dilute alloys

The self-diffusion coefficients of copper in copper-rich Cu-Sn, Cu-In and Cu-Sb solid solutions containing up to 3 at.% Sn, 3 at.% In and 1.7 at.% Sb, respectively, in the temperature range between 1005 and 1145 K have been determined by the serial sectioning method with radioactive tracer 64Cu. It has been found that enhancement of the self-diffusion coefficient of Cu by addition of Sn, In and Sb is expressed by a quadratic equation of the solute concentration rather than a linear one. Combining the enhancement factor of solvent diffusivity determined by the present work with the previous experimental data on the vacancy flow factor determined by the measurement of Kirkendall effect, an appropriate numerical set of the correlation factor and the jump frequency ratios in the five-frequency model have been evaluated. The correlation factor for the impurity diffusion of Sn, In and Sb in Cu thus obtained is in the range between 0.1 and 0.2 in the temperature range between 1005 and 1089 K. Such a small value of the correlation factor suggests highly correlated jumps of a solute atom to a vacancy.

Cu-SiO<SUB>2</SUB>合金単結晶の高温変形

To clarify the mechanism of the high temperature deformation of dispersion hardened Cu base alloys, the high temperature deformation behavior of Cu-0.5 and 1.0 vol%SiO2 alloy single crystals, which were produced by means of internal oxidation, were investigated. Tensile tests were performed in the temperature range from 873 to 973 K at initial strain rates between 10−5 and 10−3 s−1. In the steady-state deformation, the stress exponent, the apparent activation energy for the high temperature deformation and the internal stress were measured for these alloys. Rotation of the specimen axis was examined by the X-ray Laue method.While in the early stage of deformation the specimen axis rotates along the trace of single slip, it hardly rotates on reaching at the vicinity of the 〈100〉-〈111〉 boundary, and the steady-state deformation is realized. The steady-state strain rate is proportional to the seventh power of the applied stress. Values of the apparent activation energy for the high temperature deformation of Cu-0.5 and 1.0 vol%SiO2 alloys are 192 and 211 kJ·mol−1, respectively, and nearly equal to that for self-diffusion in copper. The internal stress is a main component of the applied stress in these alloys. These results indicate that the high temperature deformation of Cu-SiO2 alloys is controlled by a recovery process.The experimental results in Cu-SiO2 alloys were discussed on the base of the creep theory of Ansell and Weertman for dispersion hardened alloys. The stress exponent in these alloys does not agree with the one expected from the theory.

Chemical diffusion measurements in single crystalline cuprous oxide

Using electrical conductivity measurements in the temperature range 650–1100°C and for oxygen pressure greater than 10 −6atm., the variation of the chemical diffusion coefficient in cuprous oxide with temperature has been determined as: D ̃ = 1.2 10 −3 exp ( − 7800 RT ) cm 2 sec −1 . Taking into account the nature of the prevailing defects in cuprous oxide one can show that D ̃ ≅D Cu [V x Cu ] . This relation permits the results to be compared with those determined by tracer diffusivities. Using a value for the enthalpy of formation of non ionized copper vacancies in the range 12–16 kcal mol −1, the results are shown to be in agreement with the value of the activation enthalpy for self-diffusion of copper of 24 kcal mol −.

Self-diffusion in copper at “low” temperatures

A low-energy argon ion-beam sputtering system was developed and applied to serial sectioning in a radiotracer study of “low”-temperature self-diffusion of copper. Very thin uniform sections with thicknesses between several nm and a few hundred nm were removed from copper single crystals of extremely low dislocation density. Gaussian penetration curves are observed. The self-diffusion coefficient was measured over the temperature range 574 to 905 K and varies between about 3 × 10−19 and 6 × 10−13 cm2 s−1. In conjunction with the high-temperature data of Rothman and Peterson the present data are analysed in terms of mono- and divacancy contributions to the total diffusivity. In the temperature range of the present study the divacancy contribution is almost completely negligible. For the sum of formation and migration enthalpy of a mono-vacancy a best value of 2.04 eV is obtained. Es wurde eine Ionenstrahlzerstäubungsanlage entwickelt, die mit Argonionen niederer Energie arbeitet. Sie wird eingesetzt für das Abtragen dünner Schichten in einer Untersuchung der Selbst-diffusion in Kupfer bei „tiefen”︁ Temperaturen mittels radioaktiver Isotope. Sehr dünne, gleich-förmige Schichten mit Dicken zwischen einigen nm und wenigen hundert nm werden von Kupfer-einkristallen extrem niedriger Versetzungsdichte abgestäubt. Gauß'sche Eindringkurven werden beobachtet. Der Selbstdiffusionskoeffizient variiert im Temperaturbereich zwischen 574 und 905 K zwischen ungefähr 3 × 10−19 und 6 × 10−13 cm2 s−1. In Verbindung mit den Hochtemperaturdaten von Rothman und Peterson werden die vorliegenden Daten im Hinblick auf Einfach-und Doppelleerstellenbeiträge zum gesamten Diffusionskoeffizienten analysiert. Im Temperaturbereich der vorliegenden Arbeit ist der Doppelleerstellenbeitrag beinahe ganz vernachlässigbar. Für die Summe aus Bildungs- und Wanderungsenthalpie der Einfachleerstelle ergibt sich 2,04 eV als bester Wert.