Table of contents

We prepared oxynitrides with different interfacial nitrogen concentration using RTP. Reoxidation kinetics of these oxynitrides was studied. It was found that reoxidation thickness strongly depends on the . The reoxidation kinetics can be used for the evaluation of.

A new technique has been developed to determine accurately the location and distribution of defects in layers thermally grown on Si substrates. The technique utilizes selective propane (5% in ) carbonization of the Si surface exposed by oxide pinholes to form well‐defined islands of at the interface. This propane infiltration was performed at 1300°C for times of 2–90 s at reduced pressure (5 Torr). The islands formed at the interface are circular, with a diameter directly related to the size of the oxide pinhole at each location. For a 1000 Å oxide and 15 s reaction time, the island (and hence defect) density was typically ∼700/cm².

Surface modification by suitable dyes of electrodes used in photoelectrochemical cells allows noticeable increase of their photo‐to‐electrical conversion efficiency. Several bis or trisbipyridylruthenium(II) complexes, containing different functional groups (in particular, carboxylation of the pyridine ring is necessary to obtain a strong adsorption) were studied. All of them were identified by "normal" (without any "artificial" enhancement) ex situ Raman spectroscopy. The differences (in peak position and intensity) of their spectra, typical of the pyridine rings, carboxylated or not, yields the position of the molecule with respect to the oxide surface. It is the first time that such a monolayer‐on‐oxide substrate was studied by Raman spectroscopy; the laser light must stimulate the charge transfer in the Ti‐molecule surface complex.

The electrochemical deposition of electrically conductive polyaniline films (PANI) onto the surface of porous silicon (PS) layers has been studied using kinetic measurements and infrared spectroscopy. The process of PANI deposition is sensitive to the size of the pores, as well as to the presence of a passive layer at the pore walls and bottoms. Observable polymerization takes place at PS samples processing a pore size of about 4 nm. The results are less good for PS layers with either smaller or larger pores.

New Li ion conductive solid electrolytes with a conductivity of about 10⁻³ Ω⁻¹ · cm⁻¹ at room temperature have been prepared. Li ion conductivity in the electrolytes has been demonstrated with the cycling of and solid‐state cells.

A porous, luminescent surface is produced on n‐Si by irradiation in anhydrous or aqueous hydrogen fluoride without an externally applied potential. The photoluminescence spectra of the resulting surfaces are similar to those obtained by the anodization method. A porous surface is only produced when electrical contact is made between the n‐Si disk and a metal surface such as Au or Hg, and, in the aqueous system, when dissolved is present.

The reported observations are part of an extensive study on the cathode reaction during magnesium electrolysis. The cathodic discharge process has been studied by the use of voltammetry and chronoamperometry on glassy carbon and iron substrates in pure molten at 725°C. Formation of an Mg metal phase requires a large overpotential. The potentiostatic current transients seem to depend on charge transfer, phase formation on the substrate, and diffusion of dissolved magnesium away from the electrode. The rate control varies with applied potential and time. Potentiostatic current transients on glassy carbon showed remarkable features regarding the stability of Mg nuclei.

The cathodic decomposition of 1M PC/EC electrolyte in presence of various amounts of 12 crown 4 during the first lithiation of graphite is investigated galvanostatically. The decomposition reaction is greatly reduced after addition of 12 crown 4 and reaches a limit at an optimum 12 crown 4 concentration of ∼0.35M. The reduction of electrolyte decomposition after addition of 12 crown 4 is attributed to a significant decrease in gas evolution and possibly to the formation of a more stable SEI film.

The feasibility of modifying graphite, by intercalation, to reduce galvanic reactions in carbon fiber reinforced metal matrix composites is investigated. A nickel chloride intercalation compound and a reduced nickel chloride intercalation compound were prepared, and their electrochemical properties studied in deaerated 0.1 N. The open‐circuit potential of is about 0.500 V more positive than that of natural graphite while that of is about 700 V more negative than that of natural graphite. In the latter case, the potential difference relative to aluminum is reduced by about 50%. Distinct changes in the cathodic behavior of graphite upon intercalation and reduction are also observed. Results are discussed in terms of the effects of these changes on galvanic interactions with aluminum alloy Al 6061‐T6.

Small amounts of oxide impurities in alkali chloroaluminate and fluorochloroaluminate melts can complicate markedly the electrochemical and spectroscopic behavior of other solute species in these melts. A simple method for the removal of oxides from these melts has been developed in our laboratory. This method is based on the reaction of carbon tetrachloride with oxides to convert them to chlorides. Spectroscopic techniques (UV‐visible and IR spectroscopy) have shown that addition of carbon tetrachloride results in the complete conversion of oxides to chlorides.

The galvanostatic charge and discharge of a lithium anode/solid polymer separator/insertion cathode cell is modeled using concentrated solution theory. The model is general enough to include a wide range of polymeric separator materials, lithium salts, and composite insertion cathodes. Insertion of lithium into the active cathode material is simulated using superposition, thus greatly simplifying the numerical calculations. Variable physical properties are permitted in the model. The results of a simulation of the charge/discharge behavior of the system are presented. Criteria are established to assess the importance of diffusion in the solid matrix and transport in the electrolyte. Consideration is also given to various procedures for optimization of the utilization of active cathode material.

The oxidation of phenol at the platinum electrode was studied in aqueous acidic solutions. The effects of electrode surface oxide on the oxidation reactions of phenol and on electrode passivation by reaction products were investigated using cyclic voltammetry and chronoamperometry. X‐ray photoelectron spectrometry was used to detect changes in the nature of the passive film. Phenol reacted at both the inner and outer Helmholtz layers at platinum metal electrodes. Phenol in the inner Helmholtz layer is adsorbed irreversibly and is conductive. Its oxidation involves ring cleavage with a greater than 18 eq/mol. The outer Helmholtz layer reactions are characterized by rapid simple oxidations involving minimal rearrangement of the reactant molecule. This implies that once stable oxidized products such as benzoquinone and polymers with quinone or ether structures are formed they must move from the outer to the inner Helmholtz layer to be oxidized further by ring‐cleavage reactions. We postulate that the bulk of the initial current flow during phenol oxidation is due to the simple fast outer Helmholtz reactions. This initial current continues until the buildup of unreactive products blocks further outer Helmholtz reactions and the slower inner layer reactions predominate. This electrode behavior changed if the electrode was preoxidized producing a platinum oxide coating. The inner layer reactions were greatly reduced at a platinum oxide coated electrode resulting in lower passivated electrode current flow. The onset of passivation however was delayed at the oxide coated electrode. This is attributed to a weaker adsorption of reaction products at the electrode surface requiring additional reaction to produce a stable passive film. The final resulting passive films at platinum and platinum oxide electrodes were chemically similar based on x‐ray photoelectron spectrometric analysis but differed in thickness indicating that the electrode passivation is not due simply to the thickness of the passive film.

The electrochemical reduction of under high pressures of was investigated on Ni electrodes. With an increase in pressure, the Faradaic efficiency for reduction was increased while that of evolution by water reduction was decreased. Current density at −1.8 V vs. did not depend greatly on pressure at 30°C. By increasing temperature, the Faradaic efficiency for hydrocarbon formation was increased, suggesting that thermal activation was needed, although the total reduction efficiency for reduction was decreased. The Faradaic efficiency for the hydrocarbon formation showed a maximum at −1.6 V , while those for and were increased with increasing cathodic polarization. The weight distribution of hydrocarbons formed on Fe, Co, and Ni electrodes agreed well with the Schultz‐Flory distribution, indicating that hydrocarbons were formed via a mechanism similar to a Fischer‐Tropsch reaction for thermal catalysis, i.e., polymerization of surface carbene groups produced by the reduction of which was formed by reduction.

The in situ cutting method to expose bare lithium metal, the corrosion potential‐time transients, and voltammetric measurements have been employed to study the behavior of lithium in various organic electrolyte environments. In situoptical and ex situ scanning electron microscopy techniques were used to examine the morphology of Li electrode surfaces, during exposure at open circuit and after anodic polarization. The following electrolytes have been investigated: , , , , and in unpurified THF and PC. Small amounts of water shift the corrosion potential towards more positive values (compared to purified electrolytes). During intensive anodic cycling with the presence of water leads to the breakdown of a surface film and exposure to bare lithium metal.

Potentiodynamic and surface analysis techniques were applied to investigate the passivation of metals in para‐toluene sulfonic acids (PTSA). The diffusion‐limiting current density was used to calculate the ferrous ion concentration which causes the precipitation of a pseudopassive film. From the surface analysis data, the ratio in the passive film increases in order of 430, 304, and 316 stainless steels. To analyze the differences, a correction was made to the data so that the observed difference in the oxide film composition is not attributed simply to the Cr and Fe contents of the alloys. This method also was applied to correct the ratio in the passive film. We conclude that the preferential dissolution of Ni and Fe leads to a Cr enrichment in the passive film.

Corrosion of iron in anhydrous acetonitrile solutions of some carboxylic acids containing a small amount of water was investigated by polarization measurements. Polarization curves of the iron electrode in aqueous and acetonitrile solutions of carboxylic acids (acetic, chloroacetic, dichloroacetic, and benzoic) are discussed based on the content of water and the concentration of H⁺ and carboxylate ion. A linear relationship between logarithms of the corrosion current density and proton concentration was established for iron in acetonitrile solutions. The cathodic reaction was first order with respect to the H⁺ concentration. The cathodic process of iron corrosion in the acetonitrile solutions with a small amount of water seemed to be similar to that in the aqueous solutions.

In open‐circuit potential measurements of certain Cu‐based alloys in solutions oscillations were observed in Cu‐Al and Cu‐Ag‐Al. Studies with various electrolyte concentrations and rotation speeds of the electrode led to a possible explanation of the phenomenon in terms of Al dissolution altering the main Cu dissolution/passivation process. The accumulation of aluminate ions on the interphase and its transportation to the bulk solution are the main causes of the oscillations.

The principle of equilibrium potential‐pH diagrams for two‐dimensional species adsorbed on metals³ is applied to the case of sulfur and oxygen adsorbed on nickel in water. Standard Gibbs energies of formation for sulfur and oxygen adsorbed on nickel are calculated from literature thermodynamic data for chemisorption from the gas phase. The E‐pH relations associated with the equilibria between water, the dissolved sulfur species, and adsorbed sulfur and oxygen are given at 25 and 300°C. The corresponding E‐pH diagrams are drawn for two sulfur activities (10⁻⁴ and 10⁻⁶) and superimposed on the usual Pourbaix diagrams for the system. Formation of adsorbed sulfur monolayers on the metal surface is predicted in E‐pH domains where the usual diagrams do not predict the stability of a nickel sulfide. The E‐pH diagrams for species adsorbed on metals allow one to make new predictions of the corrosion risk for metals in sulfur‐containing aqueous environments.

The anodic oxidation of zirconium was studied by in situ ellipsometry together with capacity measurements. The oxides were grown under potentiodynamic, galvanostatic, and potentiostatic conditions up to final potentials of 100 V in 0.5M solution. The refractive index of the oxides changes depending on the growth current. The films were slightly absorbing but their absorption coefficient was independent of the oxide growth conditions. Different methods of surface preparation including etching in hydrofluoric acid‐based mixtures, electropolishing and mechanical polishing were used. The surfaces and oxides were characterized by SEM examination and XPS measurements. The surface pretreatment affects both the substrate and the oxide optical constants as well as the rate of oxide growth. The density and dielectric constant of the oxides were calculated performing simultaneous ellipsometric, coulometric, and capacity measurements.

A series of thin films were formed on Pt by an MOCVD technique using Fe(III), Cr(III), and Ni(II) acetylacetonate. The corrosion resistance of the films was examined in and by measuring the film thickness using ellipsometry and the chemical analysis of test solutions with ICPS. The dissolution rate of composite films decreases exponentially with an increase in the cationic mass fraction of Cr³⁺ ions, , of the films, and at the values of larger than 0.7 it becomes two orders of magnitude lower than that of films. The same type of changes in the dissolution rate with was observed for the composite films. Therefore, the addition of to and films effectively improves the corrosion resistance. The addition of to composite films containing an adequate amount of does not bring an effective improvement in corrosion resistance. Therefore, the corrosion resistance of composite films is determined primarily by the content of the films.

The anodic dissolution and passivation of n‐ and p‐type Si in different concentrations of hydrazine solution and etching temperatures were studied. During etching, performed at 70 and 90°C, it was observed that the current‐potential characteristics for both n‐ and p‐type Si snowed a current reduction after reaching a peak value. A linear I‐V relation was observed for the room temperature etching. A mechanism, which accounts for the semiconductor energy level change in solution as under different biasing conditions, is proposed to give a qualitative explanation of the different I‐V behaviors for n‐ and p‐type semiconductors.

With the addition of 2,2'‐dipyridyl to the triethanolamine (TEA) electroless copper bath, the effects on the deposit quality were examined. While the optimum deposit quality in the ethylenediaminetetraacetic acid (EDTA) bath was obtained by adding 2,2'‐dipyridyl, the optimum quality for the TEA bath was obtained by adding higher than 2,2'‐dipyridyl. We analyzed the difference in the optimum concentrations between the EDTA and TEA baths based on the adsorption model and found that 2,2'‐dipyridyl in the TEA bath adsorbs onto the adsorbed Cu(II)‐TEA complexes. On the other hand, in the EDTA bath 2,2'‐dipyridyl adsorbs on the naked surface. The difference in the required concentration of 2,2'‐dipyridyl to attain the optimum deposit quality between the TEA bath and the EDTA bath is interpreted by the differences in the adsorption equilibrium constants of excess ligand.

In situ Raman and reflectance spectra of iron electrodes in borate buffer pH 8.4 containing 2,2'‐bipyridine (bipy) have been investigated. In the potential range where is expected to be formed the in situ Raman and reflectance spectra were identical to the resonance‐enhanced Raman and absorption spectra of the complex, respectively. The results have shown that in the presence of bipy the equilibrium is displaced to the right as a result of the formation of the complex.

By adding to buffered neutral (MEIC = 1‐methyl‐3‐ethylimidazolium chloride) ambient‐temperature molten salts, it is possible to measure directly the Li⁺/Li(s) couple at a tungsten electrode. From open‐circuit potential measurements performed in a large number of buffered melts, the Li⁺/Li(s) standard reduction potential is found to be −2.066 (±0.005) V vs. Al(III)/Al in a reference melt. Similarly, by adding to buffered neutral (DMPIC = 1,2‐dimethyl‐3‐propylimidazolium chloride), a Na⁺/Na(s) standard reduction potential of −2.097 (±0.050) V is directly determined.

The adhesion of electrolessly deposited nickel on ceramic substrates using sputtered and evaporated Ti‐Pd activator films was studied. The adhesion was measured using the direct pull‐off test and the 90° peel test. The morphology and the chemical composition of the fracture surfaces of the samples with evaporated Ti‐Pd activator film were studied with scanning electron microscopy/energy dispersive x‐ray, and static secondary ion mass spectroscopy. Failure did not occur along the metal‐ceramic interface, but mainly in the alumina, and therefore the strength of the system is determined primarily by the substrate material. Cross‐sectional transmission electron microscopy and high‐resolution transmission electron microscopy were used to study the interface structure before failure. The oxidation state of Ti at the interface was measured with x‐ray photoelectron spectroscopy. This was carried out in the (sub)monolayer range by using a Ti wedge deposited on alumina with a maximum thickness of 0.35 nm. It is concluded that the strong adhesion at the metal‐ceramic interface is caused by chemical bonding of the first Ti monolayer with substrate oxygen atoms.

We describe the formation of ultrathin tantalum oxide capacitors, using rapid thermal nitridation of the storage‐node polycrystalline‐silicon surface prior to low pressure chemical vapor deposition of tantalum oxide. The amorphous tantalum oxide film is deposited on the nitrided polysilicon surface using penta‐ethoxy‐tantalum and oxygen gas mixture at 470°C. The films are annealed at 600–900°C in dry . Densification of the as‐deposited film by annealing in dry is indispensable to the formation of highly reliable ultrathin tantalum oxide capacitors. During this densification, and desorb from the as‐deposited film, and the film crystallizes into an orthorhombic structure. The RTN treatment allows a reduction of the equivalent thickness of the capacitor dielectric layer and results in superior leakage and reliability characteristics.

Current and potential distributions, satisfying Laplace's equation and obtained by superposition of ring sources, are developed and discussed for a cylinder electrode embedded in an infinite insulating cylinder. For the primary distribution, the resistance and the coefficient describing how the current density goes to infinity at the edge of the electrode are presented as functions of the aspect ratio of the electrode. For a uniform current density on the electrode, the maximum potential variation on the electrode is presented as a function of the aspect ratio. For linear electrode kinetics, the condition for nearly uniform electrode current density is quantified, and the ratio of edge to center current densities is developed for kinetic parameters that lead toward the primary distribution. Current and potential distributions on the electrode and the adjoining insulator are presented for these three cases, and some asymptotic formulas are developed for high and low aspect ratios. This geometry is of interest in applications of cathodic protection.

The factors affecting the cathodic reduction of oxygen to hydrogen peroxide and the degradation of formaldehyde using in situ generated hydrogen peroxide are presented in this work. The current efficiency for the production of hydrogen peroxide on graphite has been indicated by the experimental results to have been significantly affected by the pH value, the current density, the temperature, and the oxygen sparging rate. The maximum current efficiency for the cathodic reduction of oxygen to hydrogen peroxide was 93.5%. The degradation of formaldehyde by the electrogenerated hydrogen peroxide was primarily governed by the current density, the pH value, and the temperature. The degradation fraction of formaldehyde was more than 99% whenever the initial concentration of formaldehyde was in the range of 250 to 1000 ppm.

The adsorption of bisulfate and sulfate anions from 0.05M solution on Pt(100) was characterized by Fourier transform infrared spectroscopy (FTIR). Subtractively normalized interfacial FTIR spectra (SNIFTIRS) indicated only weak adsorption of bisulfate at potentials below 0.4 V (vs.), in contrast to the strong adsorption seen on the (111) surface by the same technique. At potentials above 0.6 V the SNIFTIRS spectra were due entirely to diffuse layer species. At potentials between 0.4 and 0.6 V, changes in the IR bands indicated either a rearrangement (and possibly deprotonation) of adsorbed bisulfate or weak adsorption of sulfate ion.

This study examined the effect of the organic inhibitor, benzotriazole (BTA), on the oscillatory behavior of high‐rate Fe electrodissolution in concentrated chloride media. Linear sweep experiments with rotating disk electrodes showed well‐defined limiting currents for both uninhibited and inhibited iron. In the presence of BTA, the inhibition of Fe dissolution was observed in the entire potential region examined, with the onset of the limiting diffusion current occurring at lower potentials. Galvanostatic experiments showed that potential oscillations for both uninhibited and inhibited Fe occurred at applied current densities within and above the limiting diffusion current; the potential oscillations exhibited a variety of waveforms. The chaotic oscillations were characterized by power spectral densities, phase portraits, Poincaré sections, correlation dimensions, and Lyapunov exponents. The effect of BTA on the oscillatory behavior of Fe in concentrated chloride solutions was interpreted in terms of a duplex inner nonporous layer‐outer porous salt film model.

Polypyrrole grows effectively along the glass substrates pretreated with in electropolymerization from aqueous solutions, resulting in the formation of an ultrathin polypyrrole film at twin‐microband electrodes. However, no effective promotion of lateral growth of polypyrrole was observed when the substrate was pretreated with 3‐aminopropyltriethoxysilane or when the electrolysis was conducted in acetonitrile, suggesting that the hydrophobicity of the substrate is essential for promotion of the lateral growth of polypyrrole. The promotion effect is accounted for by adsorption of pyrrole monomers and selective deposition of intermediate oligomers at the hydrophobic surfaces. The polymerization anisotropy (the ratio of lateral growth rate to vertical growth rate) was ca. 25. The partial pretreatment provided a way of micropatterning with polypyrrole on an insulating substrate. Such a promotion effect at hydrophobic surfaces is amplified in the presence of a small amount of an anionic surfactant. Polyaniline also grows effectively along the hydrophobic surface with the polymerization anisotropy of >100.

The reason why an electric coupling factor is responsible for mixed conduction under dc conditions within the bulk of a porous electrode, in addition to the geometric factor, is explained qualitatively. A quantitative model for calculating the lower margin of the specific resistivity of a mixed metal‐electrolyte substance (binary mixture) is presented. The model assumes the conduction to be analyzable in terms of a conductor‐superconductor percolation problem, and uses a numerical method to calculate the mixed resistivity without the need for direct solving of Kirchhoff equations.

Electrochemical dc and ac methods were used to evaluate the corrosion behavior of stainless steel 304L in hydrazine with or without additions of . Tafel slopes and corrosion current densities were obtained from the analysis of polarization curves recorded in the vicinity of the corrosion potential after elimination of the ohmic drop. Electrochemical impedance spectroscopy was used to monitor the continuous changes of the corrosion kinetics with increasing additions of in . The results obtained with both techniques show that minute addition (up to 100 ppm) of increased the corrosion rate of the stainless steel in . Further addition produced no additional changes.

The mechanical instability of oxide layers on anodized aluminum generates a regular arrangement of straight lines parallel to the cylindrical electrode axis. This is analyzed using Euler's equation for buckling of thin cylindrical shells under compressive forces, due to volume expansion concurrent with metal‐oxide formation. The coating critical stress is determined as well as the energy dissipated by plastic deformation of the anodic coating. Most of the energy associated with compressive stress of the oxide film is dissipated by plastic deformation.

A theoretical treatment is presented for the quantitative analysis of potential‐modulated normal‐incidence reflection absorption UV‐visible spectra of solution‐phase, optically absorbing species produced at the surface of a rotating disk electrode (RDE). This novel technique is based on the application of a sinusoidal voltage of small amplitude to the RDE to generate in turn, a perturbation in the concentration profile of the absorbing species, C. Such changes introduce a modulation in the absorptivity of the solution along the axis of rotation of the RDE, and these can be monitored by (near)‐normal‐incidence UV‐visible reflection absorption spectroscopy. A mathematical analysis of the optics and hydrodynamics for the system indicates that the ratio , where is the amplitude of the ac and the magnitude of the dc components of the optical signal, is proportional to the extinction coefficient of C and to the absolute value of the integral of the time‐independent function of the oscillatory concentration profile. Excellent agreement was obtained between the approximate soloutions (in terms of the eigenfunctions, eigenvalues, and coefficients of the appropriate Sturm‐Liouville system) valid in a domain of frequencies low enough to achieve optimum sensitivity and those determined by rigorous numerical integration of the governing differential equation subject to the appropriate boundary conditions. This provides a means of extracting quantitative information from the experimental data based on a simple mathematical expression.

The production of synthesis gas, , from methane oxidation was studied in an yttria‐stabilized zirconia reactor at 950°C and atmospheric pressure. The anodic electrode was Fe and the cathode that was exposed to air was Pt. Reduced iron was more active than oxidized iron for syngas formation. The effect of two oxygen sources, gaseous oxygen and ionically transported oxygen, on methane conversion and selectivity to syngas was studied. O²⁻ transported through the electrolyte promoted formation more favorably than gaseous oxygen. The maximum selectivity and yield using O²⁻was nearly 100 and 73%, respectively. Ionically transported oxygen and gaseous oxygen, however, did not differ significantly in formation. Carbon formation occurred with both oxidants; however, less with O²⁻.

Subatmospheric chemical vapor deposited (SACVD) tetraethylorthosilicate (TEOS) oxide provides excellent deposition profiles for submicron device structures. The film properties, such as wet etch rate, shrinkage, and cracking resistance, as well as step‐coverage, depend strongly on the ratio. The SACVD oxide film quality correlates to step‐coverage and gap‐filling ability, and both can be controlled by varying the ratio.

ceramics were synthesized and their ionic conduction was investigated. These ceramics showed protonic and oxide ionic mixed conduction under fuel cell condition. While protonic conduction was predominant below 1027 K, oxide ionic conduction became significant as the temperature increased. Using these oxides as solid electrolyte, hydrogen‐air fuel cell could be constructed. exhibited the best cell performance among the electrolytes examined. The maximum short‐circuit current density was about 900 mA/cm² at 1273 K. The polarization at each electrode was low. Porous nickel could be used as anode material instead of expensive platinum and as cathode material.

A complete chromium oxide passivation technology has been developed for stainless steel surfaces for use in high purity gas‐delivery systems and process chambers. Starting with an electrochemical buffing (ECB) to add to electropolished (EP) SUS316L stainless steel material, an optimal thermal treatment was found by using a gas mixture of 10% hydrogen, 1–10 ppm oxygen and argon balance gas at 500°C for 1 h. Five‐day corrosion tests with gas (containing 1.4 ppm moisture) at 5 kg/cm² and 100°C showed no sign of corrosion on the chromium oxide passivated surface. Chemical stability tests on this surface with silane specialty gas thermal decomposition also showed a remarkable noncatalytic activity compared with conventional surfaces.

Electrical conductivity of calcium‐doped lanthanum chromites, , was determined as a function of composition, temperature, and oxygen partial pressure, , to determine its defect structure and understand its redox behavior. The conductivity was independent of and was proportional to the dopant concentration at high . The activation energy of conductivity was 0.12 to 0.14 eV and the mobility was 0.066 to 0.075 cm²/V/s in the temperature range of 900 to 1050°C, which was ascribable to small‐polaron hopping. Under reducing conditions, the conductivity decreased exponentially with decreasing and asymptotically approached a relationship. A simple point‐defect model, in which , , and were assumed as predominating defect species, was proposed to elucidate the conductivity variation with . The oxygen nonstoichiometry calculated based on the defect model was consistent with the reported thermogravimetric data, which verified the soundness of the model.

We demonstrated surface cleaning using photoexcited fluorine gas diluted with hydrogen . We found that cleaning selectively removed native Si oxide from thermal oxide without etching the bulk Si. The dangling bonds on the Si surface after cleaning were almost exclusively terminated with hydrogen atoms, with few bonds to fluorine. cleaning effectively flattened the Si surface. We applied cleaning to Si epitaxy. This cleaning allowed us to obtain single‐crystal Si film with preannealing temperatures as low as 600°C. This temperature is lower than that of conventional methods by ca. 150°C. cleaning is a good dry precleaning method for various processes which include Si epitaxy.

One contributing factor in the nonattainment of proper boron concentrations in borophosphosilicate glass films is the variability of the diborane concentration in gas mixtures used in the process. We describe an analytical procedure that allows the monitoring of small changes in diborane concentrations utilizing an internal reference standard in the gas mixture. Employing this procedure, the decomposition kinetics of diborane mixtures prepared in steel cylinders with concentrations of less than 10% was investigated. The rate law for the decomposition of diborane in these mixtures follows the rate law observed for pure diborane

where at 20°C was .

A new etch process involving three sequential etch steps was developed to measure low defect densities in a range of about 10³–10⁸ cm⁻². The main advantages of this approach as compared to earlier methods are independence from the defect density, improved simplicity, and better image contrast. Resulting structures are similar to those obtained during defect etching in bulk silicon and allow the application of automatic defect measurement tools extending measurement ranges and accuracies to the bulk technology level. Dependences of defect densities on the implanted oxygen dose and implantation energy have been studied for SIMOX wafers. The experiment shows an increase of defect density with the dose and a decrease of defect density with the implantation energy. A comparison with transmission electron microscopy results shows good agreement.

A Langmuir‐Hinshelwood growth‐rate equation is presented for the germanium‐silicon (GeSi) alloy deposition from and assuming dissociative chemisorption on a heterogeneous GeSi surface. Model parameters for the deposition kinetics have been extracted from measurements. The fit for the bond‐energy of hydrogen to a germanium surface site is 30 kJ mol⁻¹, lower compared to that of hydrogen to a silicon site. We found to a good approximation the GeSi composition of the alloy to be independent of the temperature. Moreover, the GeSi is polycrystalline down to the lowest deposition temperature we used, ie., 450°C.

The relationship between the structural properties of and deposition conditions is studied with thermal desorption spectroscopy. The principal thermally desorbed species, except for , are divided into two categories: species and ethoxyl‐group‐related species. Most part of the species, such as , OH, O and , are generated from silanol groups rather than actual water in the oxide. In contrast with O‐atom desorption, desorption is not synchronized with desorption. It is likely that desorbed has its origin in ozone included in the source gas. The ethoxyl‐group‐related species, including , , , and , arise from components that have been incorporated into the oxide atomic network due to the incompleteness of TEOS decomposition. The dependence of thermal desorption spectra on source‐gas composition and deposition temperature is described.

Pattern‐induced thermal nonuniformities, arising during rapid thermal processing, are not radially symmetric. So the pattern‐induced part of the thermal nonuniformities cannot be reduced by conventional methods such as multiple bank control, circular lamp arrangement, or wafer rotation. A new approach to the reduction of the transient thermal nonuniformities is described. The approach utilizes a digital control of the heating power in addition to the pyrometer control. The power absorbed by the wafer in any time period of processing can be controlled, limited, or ramped according to a defined recipe. Independent top and bottom heater bank control enables step by step changeable asymmetrical heating. These features allow better process design by reduction of any transient nonuniformities, especially during heating up. The additional power control, together with a new back‐radiant silicon ring design allows defect‐guarded processing.

The application of single‐wavelength ellipsometry as an in situ technique for the simultaneous measurement of the silicon wafer substrate temperature and the surface oxide film thickness in a rapid‐thermal processing system has been investigated. This technique is based on measuring the ellipsometry parameters ψ and Δ, and then calculating the wafer temperature and oxide thickness using polynomials derived from the known temperature dependence of the index of refraction of the silicon substrate and the silicon dioxide film. Temperature measured with ellipsometrey was compared to that measured with a thermocouple. The difference depends on the optical properties of the oxide and its thickness. For native‐oxide‐covered silicon wafers, the temperature measured by ellipsometry is within 20°C or less of that measured by a thermocouple for temperatures ranging from room temperature to 1000°C. For thicker oxides, similar results that agree with estimated error bounds were obtained. Temperature transients effects and technique optimization are discussed.

Gate oxide growth in ultralarge scale complementary metal oxide semiconductor technologies involves exposure to various oxidizing conditions during temperature ramp‐up and stabilization steps. These conditions are significantly different from that of the main oxidation interval. Previous studies dealing with thick oxides haying different preoxidation treatments have indicated that thermal history has a significant impact on subsequent oxide growth rate of the initial oxide. However, these trends are unclear for ultrathin oxides. In this paper, we report the results of our study on two‐step oxidation in the thin (accelerated) regime. We have used initial oxides of different thicknesses, grown under various processing conditions prior to growing the second oxide film. The results indicate that, in contrast to the case of thick oxides, the thermal history from the first step oxidation (i.e., during temperature ramp‐up and stabilization or initial oxidation) has no impact on subsequent oxide growth rate. Furthermore, we have extended the accelerated growth model by Massoud to account for two‐step oxidations.

Several PECVD processes used in VLSI applications are discussed. Films have been deposited in a 200‐mm wafer single‐chamber PECVD reactor. The processes are characterized in terms of deposition rates, uniformity across 200‐mm silicon wafers, conformality over metal lines, density, etch rates, polish rates, stress, index of refraction, and stoichiometry. PECVD nitride processes with high deposition rates and good conformality are presented along with processes with etch rates comparable to LPCVD nitride films.

A complete series of single‐phase phosphors have been synthesized by firing in an atmosphere at 1200°C. The cathodoluminescence of these compounds has been investigated at room temperature. The emission color varied from orange‐yellow to deep red as the concentration of magnesium was increased. For composition (0.1 a/o) a highly saturated red emission was obtained with chromaticity coordinates of , , which are comparable to the commercial red phosphor (, ). The quantum efficiencies of the end members of the series were the highest. As substitution was increased, the quantum efficiency rapidly decreased with the lowest values in the range. In addition, the presence of chloride, with or without phosphide, had a deleterious effect on the emission intensity; however, the chromaticity remained unchanged. Factors influencing the manganese emission are discussed in detail.

A parametric growth study was performed to determine optimum conditions for epitaxial growth of on by reactive evaporation. The growth sources were E‐beam evaporated Si and acetylene. The polycrystalline to epitaxial growth transition temperature was determined to be about 1250°C, and the optimum epitaxial growth temperature was about 1400°C. All epilayers exhibited an n‐type carrier concentration of about , independent of growth conditions, due to the high concentration of nitrogen in the acetylene. The Ti concentration ([Ti]) at the epitaxial interface was graded, due to Ti diffusion during epitaxial growth. The as‐grown [Ti] profile at the interface was stable at 500°C. However, the [Ti] profile, ion implanted into a epilayer, changed appreciably at 500°C.

Electroless deposition of Cu on Na naphthalenide‐etched poly(tetrafluoroethylene) (PTFE) leads to adherent Cu film growth through a three‐step solution process involving (i) sensitization by adsorption of Sn^II, (ii) reduction of Pd^II by adsorbed Sn^II to produce Pd⁰ nucleation sites, and (iii) reduction of Cu^II by formaldehyde for Cu film growth. Cu film thickness is determined by exposure time in the Cu^II solution. Adhesion is determined partly by the initial Sn^II sensitization step: Sn from an aqueous solution is adsorbed onto and absorbed into the rough, porous, etched surface to produce a mechanically interlocked "Sn‐base" for the subsequent steps. Patterned deposition of adherent Cu is achieved via area‐selective irradiation of the surface to produce cross‐linking and, subsequently, area selective etching. This results in major differences in adhesion strength of the resultant Cu film between the irradiated and nonirradiated areas: the Cu film on nonirradiated areas is not affected by a "Scotch‐tape" peel test while on irradiated areas XPS spectra show that the Cu film, the Sn sensitizer layer, the Pd nucleation sites, and a major fraction of the etched layer are all easily and cleanly removed. After selective peeling of the Cu film and the active layer, continuation of Cu deposition in the solution results in Cu deposition only on the remaining Cu film. Because the Sn/Pd sensitization/nucleation steps are general, the present results apply to all metals that can be electrolessly deposited.

Narrow‐pitch encapsulated Al lines are used as interconnect metallization in integrated circuits. We have measured the principal strain state of Al alloy lines passivated with silicon nitride directly as a function of temperature. We compare these results with calculations of the strain state in these lines using finite‐element modeling. The measured strain‐temperature behavior shows good fundamental agreement with finite‐element modeling, although the magnitude of the strains measured with x‐rays is less than that predicted by modeling due to voiding in the lines.

The relationship between microstructure and electron characteristics of indium tin oxide (ITO) films was investigated. The microstructure and resistivity were found to be dependent on the oxygen partial pressure in the sputtering ambient. Lower resistivity ITO films had a domain structure and a small amount of surface roughness. The small roughness and the large size domain structure caused a decrease in electron scattering, and hence an increase in mobility. Crystallinity was also affected by the oxygen partial pressure. For ITO films with a domain structure, the relative x‐ray diffraction intensity of the (440) crystal plane increased.

Atomic layer epitaxy has been found to be effective in achieving 2‐dimensional growth of highly mismatched heterostructures from the initial stages. The structural characterization of the layers has been carried out by conventional and high resolution transmission electron microscopy and high resolution x‐ray diffractometry. 60° type misfit dislocations have been observed at the heterointerface in all the structures investigated. As the layer thickness increases, neighboring 60° dislocations can react to form 90° edge type dislocations. Planar defects extending into the layers along the {111} planes have also been observed. Finally, the elastic strain relaxation has been correlated with the nature and density of the observed crystal defects.

The critical doses required to form a continuous buried stoichiometric oxide layer for 70 keV oxygen implantation either during implantation, , or after implantation and annealing, are and , respectively. The dislocation density in the silicon overlayer and the distribution and density of silicon islands in the buried layer of the annealed (70 keV) SIMOX (separated by implantation of oxygen) samples are strongly dependent on the oxygen dose (Φ) and the target temperature . Good quality thin‐film SIMOX layers with a low threading dislocation density in the silicon overlayer and low density of silicon islands in the buried layer have been produced by implantation of at 680°C.

The interactions of , , and with Si(100) have been investigated by temperature programmed desorption (TPD) with the goal of better understanding the initial stages of chemical vapor deposition for circuit metallization and wafer etching with . Coadsorption experiments with and show that under most conditions and are the main desorption products, with Ti being left behind on the surface. is a minor product. However, at sufficiently low exposures of either or , the desorption of or , respectively, is inhibited in favor of . A kinetic model involving formation of an complex at defects has been formulated which explains the results quantitatively. adsorption gives rise to the principal desorption products and , with as a minor product. The kinetic behavior can also be explained quantitatively with the proposed model. Implications for growth are discussed with reference to possible growth temperatures and source gas pressures. The mechanism for etching by is further elucidated.

Kinetics of the reaction between and to form and in a well‐mixed flow reactor at feed pressures from 0.01 to 0.5 Torr have been studied by monitoring reaction rates with a differentially pumped mass spectrometer to monitor reactant and product partial pressures. A resistively heated tungsten foil was used as the deposition substrate to provide surface temperatures up to 1500°C. This technique provides rapid measurement of gas composition in a well‐mixed flow reactor so that reaction steady states and transients could be examined. Measured film deposition rates agreed well with those determined by monitoring reaction products. Film quality was also monitored by scanning electron microscopy, x‐ray photoelectron spectroscopy, scanning Auger microprobe, and energy dispersive x‐ray analysis. Rates were found to vary from first order to negative order with respect to both and . A simple bimolecular Langmuir‐Hinshelwood rate expression, , was found to describe the steady‐state kinetics remarkably well. This equation suggests reactants and/or intermediates are competitively adsorbed. Temperature dependences of these parameters yielded heats of adsorption of reactants and reaction rate parameters. Under some conditions multiple steady states were observed vs. under nearly isothermal conditions and also vs. surface temperature for fixed reactant pressures. The bifurcation points were found to be analogous to ignition/extinction points observed in combustion systems although these deposition reactions are endothermic. The multiplicity behavior was found to be also consistent with the Langmuir‐Hinshelwood model. However, we occasionally observe as many as three stable steady states at sufficiently high pressures, with only a single steady state at lower reactant pressures.

We have used and added to radical‐beam ion‐beam etching to explore the possible benefits of H^* radicals on surface morphology. Using a novel structure composed of quarter‐wavelength stacks, etch rates of varying Al mole fraction were measured in situ by reflectance interferometry. Although H^* radicals neglibly etch , the addition of H^* enhances the chlorine (Cl^*) etch rate threefold. Contrary to our expectations, the addition of H^* to chlorine radical etching also results in a systematic increase in roughness of the surface which is directly correlated to an increase in Al content.

Semiconductor Raman lasers with cores and cladding layers are operated with the pump wavelengths of 840 and 895 nm. In contrast to previous results for the pump wavelength of 1.064 μm, it is observed that the lasing characteristics for 840 and 895 nm excitation are affected by the instantaneous temperature rise due to the excess absorption, which strongly depends on the stripe width. Two photon absorption and absorption via deep levels produced at the side interfaces are discussed, and the latter is suggested as the origin of the excess absorption.

The relative contribution of gas‐phase chemistry to deposition processes is an important issue both from the standpoint of operation and modeling of these processes. In polysilicon deposition from thermally activated silane in a cold wall rapid thermal chemical vapor deposition (RTCVD) system, the relative contribution of gas‐phase chemistry to the overall deposition rate was examined by a mass‐balance model. Evaluating the process at conditions examined experimentally, the model indicated that gas‐phase reactions may be neglected to good accuracy in predicting polysilicon deposition rate. The model also provided estimates of the level of gas‐phase‐generated associated with deposition on the cold‐process chamber walls.

In this paper we discuss the formation and prevention of residues during long‐term reactive ion etching (RIE) of thin films in a variety of fluorinated gas plasmas mixed with oxygen: , , , and . Without the addition of , all fluorinated plasmas produced significant residue for mixtures with ranging from 0 to 90%. Only and at 0% showed a low level, or the absence, of residues. The introduction of a relatively small amount of additive was observed to prevent residue formation in the etched region during the process. In etching, ∼10% was sufficient to prevent residues at all values of . A significantly larger (>10%) amount of is required to prevent residue formation in , , and plasma. For these gases, the required additive concentration increases as the oxygen decreases, reaching a maximum for 10% . Optimum RIE conditions in different fluorinated gas plasmas are presented for homojunction and heterojunction device fabrication. The effects of covering the powered electrode with graphite or Kapton sheets on residue formation and etch rates were investigated with a variety of fluorinated gas plasmas. The mechanisms of residue formation and prevention through the addition of are discussed.

Abstract not Available.

Abstract not Available.

Abstract not Available.

Abstract not Available.

Abstract not Available.