Kurt R. Hebert

The role of viscous flow of oxide in the growth of self-ordered porous anodic alumina films.

Porous anodic alumina (PAA) films are widely used as templates for functional nanostructures, because of the high regularity and controllability of the pore morphology. However, growth mechanisms have not yet been developed that can explain quantitative relationships between processing conditions and oxide layer geometry. Here, we present a model for steady-state growth of these amorphous films, incorporating the novel feature that metal and oxygen ions are transported by coupled electrical migration and viscous flow. The oxide flow in the model arises near the film-solution interface at the pore bottoms, in response to the constraint of volume conservation. The hypothesis of viscous flow was successfully validated through detailed comparisons to observations of the motion of tungsten tracers in the film. Predictions of localized tensile stress near nanoscale ridges at the metal-film interface were supported by observations of voids at these sites. We suggest that the ordering of PAA may be explained by a mechanism in which metal-film interface motion is regulated by the combination of ionic migration in the oxide and stress-driven interface diffusion of metal atoms.

Role of Chloride Ions in Suppression of Copper Electrodeposition by Polyethylene Glycol

Chloride ions and polyethylene glycol (PEG) are used together as additives to copper damascene electroplating baths, in which they suppress deposition. When the Cl - concentration is lower than the order of 1 mM, suppression abruptly breaks down below a critical potential, around which hysteresis between active and inhibited deposition is observed. A mathematical model is presented which successfully predicts the observed Cl - concentration-dependent breakdown of PEG suppression and current-potential hysteresis. The model assumes that adsorbed Cl - ions are involved in binding of PEG to the Cu surface, and that these ions are incorporated in the deposited film. The expressions for Cl - incorporation and adsorption are consistent with experimental measurements of Cl in deposits. Hysteresis was found to depend on the high sensitivity of polymer surface coverage to the concentration of adsorbed Cl - ions, possibly because each PEG molecule has a small number of binding sites to the surface.

Chemical Mechanism of Suppression of Copper Electrodeposition by Poly(ethylene glycol)

Poly(ethylene glycol) (PEG) is an important additive to electroplating baths used for the deposition of copper interconnects on semiconductor wafers. In an earlier paper, Yokoi et al., Denki Kagaku oyobi Kogyo Butsuri Kagaku, 52, 218 (1984) found a direct relationship between the deposition rate in the presence of PEG and chloride ions with the open-circuit potential measured after plating, suggesting that the rest potential reflects the chemical state of reactive copper ions within a surface polymer film. Here, these measurements were corroborated and then interpreted in terms of a proposed mechanism of copper deposition in the presence of PEG. In this mechanism, aqueous Cu 2 + ions are reduced to an intermediate complex at the PEG-Cu interface detected earlier by Raman spectroscopy [z. V. Feng et al., J. Phys. Chem. B, 107, 9415 (2003)], in which Cu + ions associate with adsorbed Cl - ions and ether oxygen ligands of PEG. The rest potential measurements are quantitatively explained on the basis of competition for these ligands at open circuit with Cu 2 + ions absorbing from solution. The results indicate that deposition is mediated though ions partially solvated with the polymer, the concentration of which is controlled by the PEG concentration and molecular weight. PEG then behaves as a polymer electrolyte film as opposed to a passive barrier.

Modeling the Potential Distribution in Porous Anodic Alumina Films during Steady-State Growth

Porous anodic alumina (PAA) films, formed by anodic oxidation in acidic solutions, contain hexagonal arrays of parallel cylindrical pores, with pore diameter and spacing between ten and several hundred nanometers. Simulations were developed for the electrical potential distribution in the film during steady-state PAA growth, and used to calculate the rates of metal-film and film-solution interface motion. In particular, a model using the assumption of no space charge (Laplaces equation) and one based on the current continuity equation, in each case coupled with high-field ionic conduction, were evaluated with respect to the requirement that the interface profiles are time invariant. Laplaces equation, on which prior simulations of PAA growth were based, yielded unrealistic behavior with highly nonuniform interface motion, suggesting the presence of significant space charge. In contrast, interface motion predicted by the current continuity equation was uniform, except near convex ridges on the metal-film interface between pores. To fully rationalize the steady-state PAA geometry, phenomena other than conduction should be considered, which are able to provide inhibition of the oxidation rate on these ridges.

Development of Surface Impurity Segregation during Dissolution of Aluminum

Caustic dissolution, when used as a pretreatment for etching of aluminum in chloride solutions, is observed to increase the rate of pit nucleation. Rutherford backscattering spectrometry (RBS) and Auger electron spectroscopy were used to measure the composition in the near-surface region of 99.98% purity aluminum after dissolution in 1 N NaOH at room temperature. During dissolution, concentrations of impurities such as Fe, Cu, and Ga were found to accumulate continuously within a layer less than about 10 nm thick adjacent to the surface, because they dissolved more slowly than did aluminum. Impurity concentrations on the order of 1 atom percent (a/o) in this layer, much higher than equilbrium values, were found after 40 min dissolution. It is argued that the large impurity concentrations are consistent with a highly defective region in the metal near the metal/oxide interface, which has been detected using positron annihilation measurements. Dissolution produced a scalloped surface topography with typically 30 nm high ridges separated by 130 nm. A simulation of RBS measurements based on scattering from spherical particles was developed to test for the possibility of preferential impurity segregation to ridges. No such segregation was detected, suggesting that this is not the reason for the strong tendency for pit nucleation to occur on ridges, as has been observed previously.

A Model for Coupled Electrical Migration and Stress-Driven Transport in Anodic Oxide Films

Recent experimental evidence suggests that plastic flow of oxide occurs during the growth of anodic oxide films and contributes significantly to ionic mass transport. Skeldon and co-workers used tungsten tracers introduced from the metal to visualize ionic transport within porous anodic alumina films formed in acidic solutions. 1-4 The observed tracer motion deviated strongly from expectations based on electrical migration as the only transport mechanism, the authors attributing the discrepancy to plastic flow in the oxide. The tracer studies supported earlier measurements of the rate of increase in pore wall height relative to stationary reference planes. 5-7 Both experiments revealed plastic flow in the pore walls at typical velocities of 0.1‐1 nm/s. We developed a transport model of porous anodic alumina films, which validated the hypothesis of coupled electrical migration and viscous flow of oxide, through a detailed agreement with the tungsten tracer profiles. 8 The results of this study suggest that the coupled stress and potential distributions in these films regulate the interface motion during the formation of self-ordered pore arrays. The importance of viscous creep may extend beyond porous oxides to planar anodic films typically formed in neutral pH solutions. Evidence for creep in such films is suggested by several experimental studies. Leach and co-workers observed a current-dependent extension of loaded Al wires during anodizing, which they attributed to current-induced plasticity in the anodic film. 9,10 Wuthrich showed that anodic alumina films deform without cracking during anodizing. 11 Zhou et al. attributed the observations of growth and coalescence of oxygen bubbles during the passage of ionic current to the plasticity in the surrounding oxide. 12,13 An additional precedent for creep of amorphous solids at ambient temperatures is found in studies that show that Newtonian viscous flow relieves compressive stresses induced by ion irradiation. 14-17 Like the materials in these experiments, anodic alumina films are amorphous, and stresses large enough to drive significant creep 10‐100 MPa are found during the growth of both porous and planar anodic alumina. 9,11,18-21 Stresses during anodizing may arise from volume constraints at the metal‐ oxide interface, and from electrostatic forces in the oxide dielectric. 22-24 In this paper, we present a model for transport in planar anodic films by coupled electrical migration, plastic flow, and migration in the stress field. Coupling of transport processes results from the constraints of volume and charge conservation. The model is developed for the specific case of barrier-type aluminum oxide films, which have been studied extensively. However, a similar treatment may apply to amorphous anodic oxides formed on a variety of valve metals. The model is adapted from the continuum approach developed by Suo and co-workers to model coupled plastic flow and diffusional transport in metals.

Formation of Aluminum Hydride during Alkaline Dissolution of Aluminum

The role of hydrogen-containing surface species in the alkaline dissolution of aluminum was studied by secondary ion mass spectrometry SIMS and atomic force microscopy AFM. The measurements revealed quasi-periodic nucleation and dissolution of large number densities of 10–100 nm size particles, during open-circuit dissolution in 1 M NaOHD at room temperature. SIMS results using deuterated solutions, and prior Auger microprobe measurements, indicated that the particles were composed of aluminum hydride deuteride, with an aluminum hydroxide deuteroxide surface layer. The measured open-circuit potential during dissolution was close to the Nernst potential of hydride oxidation. It was concluded that AlH3 forms continuously during dissolution by reaction of cathodically generated hydrogen with the Al metal and is oxidized to aluminate ions AlOH4

Observations of the Early Stages of the Pitting Corrosion of Aluminum

The initial stages of pitting of aluminum, under constant applied current in 1N HCl at 65 o C, have been studied using millisecond current pulses and pit size distributions measured with scanning electron microscopy

Positron Annihilation Spectroscopy Study of Interfacial Defects Formed by Dissolution of Aluminum in Aqueous Sodium Hydroxide

High-purity aluminum foils were examined using positron annihilation spectroscopy (PAS) after dissolution for various times in 1 M NaOH at room temperature. Measurements of the S and W shape parameters of the annihilation photopeak at 511 keV show the presence of voids of at least nanometer dimension located at the metal-oxide film interface. The large S parameter suggests that the metallic surface of the void is free of oxide. Voids are found in as-received foils and are also produced by dissolution in NaOH, evidently by a solid-state interfacial process. Atomic force microscopy (AFM) images of NaOH-dissolved foils, after stripping the surface oxide film in chromic-phosphoric acid bath, reveal cavities on the order of 100 nm size The average cavity depth is in quantitative agreement with the PAS-derived thickness of the interfacial void-containing layer, and the dissolution time dependence of the defect layer S parameter closely parallels that of the fractional coverage of the foil surface by cavities; thus, the cavities are believed to he interfacial voids created along with those detected by PAS. The cavity distribution on the surface closely resembles that of corrosion pits formed by anodic etching in 1 M HCl, thereby suggesting that the interfacial voids revealed by AFM serve as sites for pit initiation.

Microscopic Observations of Voids in Anodic Oxide Films on Aluminum

The relationship was explored between nanoscale voids in anodic aluminum oxide films and the surface condition of aluminum samples prior to anodizing. Transmission electron microscopy (TEM) detected voids on the order of 10 nm in anodic films. Atomic force microscopy (AFM) of these films, obtained after partial oxide dissolution, revealed surface cavities; comparison of TEM and AFM suggested that the cavities were the oxide voids. AFM images after variable extents of oxide dissolution showed that the voids were distributed evenly through the inner 60% of the film thickness, indicating that they were formed at the metal-oxide interface during film growth. Both AFM and TEM showed that the void concentration in the film was sensitive to the extent of dissolution of the aluminum samples in NaOH prior to anodizing. Positron annihilation spectroscopy had previously detected voids in samples without anodic films, located in the metal near the oxide-metal interface; the quantity of these interfacial voids was controlled by NaOH dissolution. The void concentration in the inner part of the anodic films was proportional to the quantity of these pre-existing interfacial voids. It was inferred that the oxide voids were formed by incorporation, during anodizing, of interfacial metal voids into the oxide film. The uniform concentration of oxide voids in the inner film suggested that interfacial metal voids formed continuously during anodizing and that metal voids were generated repeatedly at specific interfacial sites during film growth.