Diffusion kinetics of Au through Pt films about 2000 and 6000 Å thick studied with Auger spectroscopy

The diffusion kinetics of Au through Pt films were studied because of the frequent use of PtAu metallizations in semiconductor technology and because Au is often undesirable in active semiconductor regions etc. that are in contact with the Pt. PtAu couples with approximately 2000 Å and 6000 Å Pt films were heat treated between 250° and 350°C in 1 atm N₂ ambient. Surface composition and depth profiles were measured using Auger spectroscopy and ion milling. Au was found to diffuse initially through thin Pt films (< 6000 Å) by grain boundary migration and more than 10¹⁵ atoms cm^?2 of Au crossed the Pt film when the bulk of the Pt contained very little Au (?1 at.%). For 2250 Å Pt films on Au, the time t(X = 0.5) for half-saturation of the Pt surface with Au was given by t(0.5) = 1.2 X 10^?7X exp (0.96 eV/kT) min, where X is the fractional Au concentration in the first 7 Å of the surface. This diffusion rate is relatively fast; e.g.t(0.5)≈7 min for a 2250 Å Pt film at 350°C. At t(0.5) the bulk of the Pt contained less than 1 at.% Au and t(0.5) was proportional to Pt film thickness near 2000 to 6000 Å.     

Zinc diffusion in YBa2Cu3O7−x ceramics

The process of zinc diffusion in YBa₂Cu₃O_7?x ceramics with a porosity of 20–30% was studied in the temperature interval from 110 to 450°C using a ⁶⁵Zn radioactive tracer. The temperature dependence of the tracer diffusion coefficient is described by the relation D=5×10^?9 exp(?0.25 eV/kT) cm²/s. It is concluded that zinc migrates predominantly via pores and intergranular layers in the ceramics.     

An ultrasonic study of the diffusion of Ag in CdSe in the temperature range 20°–400 °C

The diffusion of Ag in CdSe has been studied using an ultrasonic technique. The diffusion coefficient was found to be D = 3 × 10^?4 exp (?0.53 eV/kT) cm² s^?1 Diffusion was found to be Fickian in the region up to a depth of 10^?2 cm and one diffusion mechanism was found to dominate. No significant anisotropy was found in diffusion coefficients measured parallel or perpendicular to the c-axis.     

Role of diffusion-annealing temperature on the microstructural and superconducting properties of Cu-doped MgB2 superconductors

This study deals with not only investigate the effect of the copper diffusion on the microstructural and superconducting properties of MgB₂ superconducting samples employing dc resistivity as a function of temperature, scanning electron microscopy (SEM) and X-ray diffraction (XRD) measurements but also calculate the diffusion coefficient and the activation energy of copper for the first time. Electrical-resistivity measurements indicate that both the room-temperature resistivity value and zero resistivity transition temperatures (T _c ) increase with increasing the diffusion-annealing temperature from 650 to 850?°C. SEM measurements show that not only the surface morphology and grain connectivity improve but also the grain size of the samples increases with the increase in the diffusion-annealing temperature up to 850?°C. As for the XRD results, all the samples contain the MgB₂ phase only and exhibit the polycrystalline superconducting phase with more intensity of diffraction lines, leading to the increasement in the lattice parameter a and c. Additionally, the diffusion coefficient is observed to increase from 6.81?×?10^?8 to 4.69?×?10^?7?cm²?s^?1 as the diffusion-annealing temperature increases, confirming that the Cu diffusion at lower temperatures is much less significant. Temperature dependence of the Cu diffusion coefficient is described with the aid of the Arrhenius relation D?=?3.75?×?10^?3 exp (?1.15?±?0.10?eV/k _B T) and the corresponding activation energy of copper in MgB₂ system is found to be about 1.15?eV. The possible reasons for the observed improvement in microstructural and superconducting properties of the samples due to Cu diffusion are also discussed.     

Ion back-scattering analysis of interdiffusion in Cu-Au thin films

Helium ion back-scattering has been used to examine the interdiffusion behavior of the thin film copper-gold system. Techniques for distinguishing bulk diffusion and grain boundary diffusion using back-scattering are presented. At temperatures in the range 200°–500°C, the grain boundary diffusion mechanism is shown to predominate in Cu-Au thin films. The back-scattering results suggest a model in which interdiffusion takes place by very rapid saturation of the grain boundaries in the gold film by copper and a slower filling of the copper grain boundaries by gold. The atoms in the grain boundaries then diffuse into the grains by bulk diffusion. In the parent gold film, we suggest that the amount of Cu Au ₃ formed near the grain boundaries is uniform in depth. In the parent copper film, more Cu₃Au forms near the original Cu-Au interface than further into the copper film. No evidence was found to suggest the formation of a pure layer of CuAu₃ or Cu₃Au. The interdiffusion and compound formation process is found to be characterized by an activation energy of 1.35-1.5 eV.     

Grain growth in copper and alpha-brasses

The wires of 99.999% copper and alpha-brasses containing 12, 20, 30 and 35 at % Zn have been annealed in vacuum for 30 to 240 min at 873, 923, 973 and 1023 K. The grain-growth data obtained are well encompassed by the relationD ², —D ₀ ² , =Kt exp(-H/kT), whereD is the instantaneous mean grain diameter at the time,t, of isothermal anneal andD ₀ refers to the initial mean grain diameter. In alpha-brasses the activation energy for grain-boundary self-diffusion,H, and the pre-exponential factor,K, depends on the zinc concentration,c, asH = (H ₀ — 1.1c) eV andK =K ₀ exp(-10.7c) cm² sec^–1. The values ofH ₀ andK ₀, referred to the base metal are respectively 0.87 eV and 3.0 × 10^–4 cm² sec^–1, which are in good agreement with those (0.85 eV and 3.6 × 10^–4 cm² sec^–1) found for copper.     

Low temperature interdiffusion in the Au-Pd and Au-Rh thin film couples

Interdiffusion profiles in thin polycrystalline multilayer films of Pd-Au and Ti-Rh-Au at temperatures up to 490°C have been measured by Rutherford backscattering. Room temperature grain boundary diffusion of Au into Rh was observed and analyzed to give D_B = 3.5 × 10^-17 cm² sec^-1. The Whipple analysis is applied to our data for the diffusion of Au in Pd; using the lattice diffusivity of Neukam, an activation energy for grain boundary diffusion of 0.9 eV is found. The diffusion of Pd in Au has also been analyzed using the Whipple model, which gives a grain boundary activation energy of 0.6 eV.     

Diffusion studies in Cr-Pt thin films using Auger electron spectroscopy

The diffusion of Cr through a thin film of Pt has been studied using Auger electron spectroscopy. The diffusion couples were prepared by the evaporation of 1400 Å of Cr onto Si(111) substrates followed by the deposition of 2000 Å of Pt. The substrates were not heated intentionally during the film depositions. Diffusion anneals in the temperature range 700°–850°K were carried out in an ultrahigh vacuum system in situ while Auger spectra were being recorded. Auger spectra from the Pt surface were taken periodically while the sample was held at a constant temperature and the signal from the Cr impurity was recorded as a function of time.The reciprocal of the times t resulting in equal peak-to-peak height ratios of the chromium 529 eV Auger transition to the platinum 64 eV Auger transition were plotted against the reciprocal of the temperatures T of the anneals. These results yielded an effective diffusion coefficient of Cr through Pt which is described by D′ = 1.02 × 10^-2 exp (-1.69 eV/kT)     

Low temperature interdiffusion in Cu/Ni thin films

Interdiffusion in Cu/Ni thin films was studied by means of Auger electron spectroscopy in conjunction with Ar⁺ ion sputter profiling. The experimental conditions used aimed at simulating those of typical chip-packaging fabrication processes. The Cu/Ni couple (from 10 μm to 60 nm thick) was produced by sequential vapor deposition on fused-silica substrates at 360, 280 and 25 °C in 10^-6 Torr vacuum. Diffusion anneals were performed between 280 and 405 °C for times up to 20 min. Such conditions define grain boundary diffusion in the regimes of B- and C-type kinetics. The data were analyzed according to the Whipple-Suzuoka model. Some deviations from the assumptions of this model, as occurred in the present study, are discussed but cannot fully account for the typical data scatter. The grain boundary diffusion coefficients were determined (for nickel through copper, Q_b = 33.7 kcal mol^-1 (1.46 eV), D_b⁰ = 4.2 × 10^-2 cm²s^-1; for copper through nickel, Q_b = 30.2 kcal mol^-1 (1.3 eV), D_b⁰ = 7.6 × 10^-5 cm²s^-1) allowing calculation of respective permeation distances.     

Interdiffusion and phase formation in Au/Sn thin film couples with special emphasis on substrate temperature during condensation

The interdiffusion of gold and tin thin films was studied, with special emphasis on the influence of the substrate temperature during evaporation. A special target chamber makes it possible to evaporate several films in succession and to analyse them by Rutherford backscattering at any stage in the interdiffusion process without exposing the specimens to the atmosphere. The samples can be kept at any temperature between -170 and +150°C during the evaporation, the heat treatment and the analysis. The interdiffusion of Au/Sn films is complex, with several competing processes going on at the same time. If the tin is evaporated at low temperatures, the initial stage of the interdiffusion will be dominated by grain boundary diffusion of gold into the tin film followed by diffusion into the tin grains, resulting in the formation of the AuSn₄ phase. The growth of the AuSn₄ phase was found to be linear with time. Concurrently the AuSn phase grows in a planar manner from the original interface. The thickness of this phase grows as t$\text{1}\text{2}$ and is believed to be limited mainly by the grain boundary diffusion of gold through the AuSn phase. This diffusion has an activation energy E_a = 0.59±0.06eV. When the substrate temperature during evaporation of the tin film is decreased, we observe an increase in the growth rate of the AuSn phase. This is believed to be caused by the decrease in grain size.     

Titanium diffusion in gold thin films

In the present study, diffusion phenomena in titanium/gold (Ti/Au) thin films occurring at temperatures ranging between 200 and 400 °C are investigated.The motivation is twofold: the first objective is to characterize Ti diffusion into Au layer as an effect of different heat-treatments. The second goal is to prove that the implementation of a thin titanium nitride (TiN) layer between Ti and Au can remarkably reduce Ti diffusion.It is observed that Ti atoms can fully diffuse through polycrystalline Au thin films (260 nm thick) already at temperatures as a low as 250 °C. Starting from secondary ion mass spectroscopy data, the overall diffusion activation energy ΔE = 0.66 eV and the corresponding pre-exponential factor D₀ = 5 × 10^− 11 cm²/s are determined. As for the grain boundary diffusivity, both the activation energy range 0.54 < ΔE_gb < 0.66 eV and the pre-exponential factor s₀D_gb0 = 1.14 × 10^− 8 cm²/s are obtained. Finally, it is observed that the insertion of a thin TiN layer (40 nm) between gold and titanium acts as an effective diffusion barrier up to 400 °C.     

The mechanism of elevated temperature intergranular cracking in heat-resistant alloys

Reheat or stress relief cracking phenomena have been reassessed in 2.25Cr1.5W heat-resistant alloys. During rupture test, time to intergranular failure increases with decreasing temperature and tensile stress and is shorter in the alloy containing a higher bulk content of phosphorus. Also the time to intergranular failure can be expressed by t = t₀·σ⁻ⁿ·exp(Q/RT) where t₀ is the proportional constant, n the stress exponent and Q the activation enthalpy. Matrix softening is accelerated under tensile stress and an active carbide growth occurs at grain boundaries oriented normal to the tensile stress direction. Because impurities segregate actively to dimples frequently observed at reheat intergranular fracture surfaces, the dimples are not micro-ductile fracture areas but the grain boundary carbide interfaces. The segregation concentration of the impurities is much higher at the grain boundary carbide interfaces than the carbide-free grain boundaries. The phosphorus segregation at the carbide interfaces of the alloy containing the higher bulk content of phosphorus is mainly replaced by the segregation of nitrogen, tin and tellurium in the alloy containing a lower bulk content of phosphorus. The elevated temperature intergranular cracking under tensile stress occurs finally due to the carbide-free grain boundary cracking following the decohesion of the grain boundary carbide interfaces.     

Effect of Thermal Oxidation on the Segregation of Phosphorus Implanted into Silicon

Secondary ion mass spectrometry and numerical simulation are used to investigate phosphorus diffusion and segregation in the SiO₂-Si(111) system during the thermal oxidation of phosphorus-ion-implanted silicon layers in dry and humid oxygen between 950 and 1150°C. The results demonstrate that the segregation coefficient of phosphorus in SiO₂ and Si depends not only on the oxidation temperature and environment but also on the oxidation time, which is interpreted in terms of the variation in the interfacial density of intrinsic point defects as a result of implantation-damage annealing. A model for the reaction segregation of impurities in the SiO₂-Si system is used to evaluate the equilibrium segregation coefficient of phosphorus as a function of temperature: = m _s ^* in dry O₂ and = 4.2 × 10²exp(?0.22 eV/(kT)) in humid O₂.     

Thermal desorption of silver and gold from tungsten studied by field emission microscopy

Under clean conditions thermal desorption of the last monolayer of copper, silver and gold from tungsten appears to take place in two stages having different temperature ranges. Rate measurements with silver and gold show that the higher temperature stage is a true desorption process having normal rate parameters. The lower temperature loss is detectable only if the initial coverage is less than one monolayer and, for gold, the rate parameters (E_act = 0.53 eV, pre-exponential term v = 10³ s^-1) suggest a diffusion-controlled process. It is tentatively suggested that in this stage loss of silver occurs by incorporation into the tungsten surface.The reported absence of a maximum in the work function of silver during desorption is shown to result from the presence of trace quantities of oxygen (≈ 0.01 monolayer), and the published value of desorption energy is consequently erroneous. Under clean conditions silver behaves like copper and gold; the last monolayer is readily isolated in desorption because it is more strongly bound than higher layers, having a desorption energy of 2.3±0.2 eV. It is suggested that the role of oxygen is either to create sites in the surface at which silver no longer contributes to the measured work function or to remove, from the tungsten surface, sites at which silver is normally strongly bound.     

Beryllium diffusion and influence on the luminescent and electrical properties of indium phosphide

The diffusion of beryllium in indium phosphide (InP) has been studied. The temperature dependence of the diffusion coefficient can be described by the formula D = 6.3 × 10^?5exp(?1.4/kT) [cm²/s], which yields D = 1.07 × 10^?11 cm²/s at T = 768°C. The results of measurements of the luminescence and electrical properties show that beryllium is a shallow acceptor with an activation energy of 0.03 ± 0.005 meV.     

Ag diffusion in ZnS thin films prepared by spray pyrolysis

ZnS thin films were deposited by spray pyrolysis method on glass substrates. Diffusion of Ag in ZnS thin films was performed in the temperature range 80-400 °C under a nitrogen atmosphere. The diffusion of Ag is determined with XRF, and the obtained concentration profile allows to calculate the diffusion coefficient. The temperature dependence of Ag diffusion coefficient is determined by the equation D = 8 × 10^− 9 exp(− 0.10 eV / kT). It was found that the as-grown undoped high resistive n-type ZnS thin films were converted to the p-type upon Ag doping with a slight increase in resistivity only by rapid thermal annealing at 400 °C in N₂ atmosphere. In addition, the band gap of the p-type film was decreased as compared with the undoped sample annealed under the same conditions. The results were attributed to the migration of Ag atoms in polycrystalline ZnS films by means of both along intergrain surfaces and intragrain accompanied by interaction with native point defect.     

Diffusion of promethium in silicon

The first investigations have been made on the diffusion of promethium in silicon. In the temperature range from 1100 to 1250 °C the diffusion constant of promethium increases from ∼1×10⁻¹³ cm²/s to ∼1.5×10⁻¹² cm²/s. The temperature dependence of the diffusion coefficient can be described by D = 5 x 10⁻¹ exp[-(3.3 eV/kT]cm²/s. Pis’ma Zh. Tekh. Fiz. 23, 46–50 (January 26, 1997)     

Abnormal dielectric behaviors in Mn-doped CaCu3Ti4O12 ceramics and their response mechanism

Mn-doped CaCu₃Ti₄O₁₂ (CCTO) polycrystalline ceramics have been prepared by the conventional solid state sintering. Our results indicate that 10% Mn doping can decrease the dielectric permittivity in CaCu₃Ti₄O₁₂ by about 2 orders of magnitude (from 10⁴ to 10²). The grain and grain boundary activation energies show an obvious increase from 0.054 eV to 0.256 eV, and decrease from 0.724 eV to 0.258 eV with increasing the Mn doping concentration, respectively, which may be caused by the variation of Cu and Ti valence states in the CCTO samples evidenced by the X-ray absorption spectra. The similar grain and grain boundary activation energies result in invalidation of the internal boundary layer capacitance effect for the 10% Mn-doped CCTO sample, and thus result in the dramatic decrease of dielectric permittivity.     

Self-diffusion of14C in polycrystalline β-SiC

The¹⁴C self-diffusion coefficients for both lattice (D _lc ^* ) and grain boundary (D _bc ^* ) transport in high purity CVDβ-SiC are reported for the range 2128 to 2374 K. The Suzuoka analysis technique revealed thatD _bc ^* is 10⁵ to 10⁶ faster thanD _bc ^* ; the respective equations are given by $$\begin{gathered} D_{I c}^* = (2.62 \pm 1.83) \times 10^8 exp\left\{ { - \frac{{(8.72 \pm 0.14)eV/atom}}{{kT}}} \right\}cm^2 sec^{ - 1} \hfill \\ D_{b c}^* = (4.44 \pm 2.03) \times 10^7 exp\left\{ { - \frac{{(5.84 \pm 0.09)eV/atom}}{{kT}}} \right\}cm^2 sec^{ - 1} \hfill \\ \end{gathered} $$ A vacancy mechanism is assumed to be operative for lattice transport. From the standpoint of crystallography and energetics, reasons are given in support of a path of transport which involves an initial jump to a vacant tetrahedral site succeeded by a jump to a normally occupied C vacancy.