ELECTROCHEMICAL SCIENCE AND TECHNOLOGY TECHNICAL PAPERS
1659
S. Fouache‐Ayoub, M. Garreau, P. V. S. S. Prabhu and J. Thevenin
The kinetic and morphological properties of the surface layers formed on lithium electrode in different sulfolane‐based electrolytes have been investigated. Application of the impedance theory of the dilute binary electrolyte has allowed a determination of characteristic parameters of the surface layers, such as their cationic transference numbers, salt diffusion coefficients, and ionic concentrations. The values obtained for these parameters have demonstrated that the surface layers behave similarly to their parent liquid electrolytes but with much lower ionic concentrations. In addition, this study has shown that satisfactory cycling performances can be obtained when the values of the characteristic parameters of the layers are close to those of the parent electrolytes. Supplementary tests based on the evaluation of the layer conductivity, the Warburg constant, and the limiting current density, have been proposed for the selection of appropriate organic electrolytes and cycling procedures to be used in secondary lithium batteries.
1665
E. S. Takeuchi, S. M. Meyer and C. F. Holmes
Real‐time discharge is one of the few reliable methods available for determining capacities of low‐rate cells. The utilization of high energy density lithium batteries in low‐rate implantable applications has increased the need for more time‐efficient methods of predicting cell longevity since cells have been shown to last in excess of eight years. The relationship between heat dissipation and self‐discharge of low‐rate lithium/BCX (bromine chloride in thionyl chloride) cells was studied and allows prediction of cell life prior to the availability of real‐time data. The method was verified by real‐time cell discharge data and provided estimates of delivered capacity within 6% of the actual values.
1670
Dawn M. Bernardi
A one‐dimensional mathematical model of a porous lead‐dioxide electrode is described and used to investigate lead‐sulfate nucleation and growth during discharge. Derivation of a nucleation rate expression that is based on classical, heterogeneous nucleation theory is outlined. An electrochemical kinetic expression is derived based on a reaction mechanism involving elementary steps, and concentrated ternary electrolyte theory is used in formulating material‐transport equations. Nucleation and electrochemical kinetic parameters are estimated by comparison of model results with experimental results available in the literature. The interplay of nucleation and growth kinetics of lead sulfate is responsible for the initial minimum in the voltage‐time curve that is commonly observed during constant‐current discharge. The model simulates the voltage minimum, which is referred to as the coup de fouet, and calculates the degree of lead‐ion supersaturation, the number density of lead‐sulfate particles, and the free energy of formation as well as the size of critical nuclei. The model also predicts a disappearance of the voltage minimum with the addition of seed particles for lead‐sulfate nucleation, which is experimentally observed. The satisfactory agreement between model and experimental results confirms that the voltage dip is caused by a temporary oversaturation of lead ions during discharge and supports the proposed theoretical approach.
1681
Petr Novák and Wolf Vielstich
The properties of polypyrrole (PPy) films synthesized at very low current densities (1–250 μA/cm2) were investigated. Potentiodynamic cycling, in situ FTIR spectroscopy, and in situ differential electrochemical mass spectroscopy were employed. The use of propylene carbonate (PC) based electrolytes for the synthesis results in an incorporation of PC fragments (arising by the electro‐oxidation of PC) into the grown polypyrrole film. During subsequent cycling of PPy films in
electrolyte, the electrochemical oxidation of PC proceeds parallel with the doping/undoping process. The decrease in current density during polymer growth has the same effect as an addition of small amounts of water into the electrolyte for synthesis—the performance of the polymer in secondary lithium cells is improved.
1689
P. Cicognani, F. Gasparoni, B. Mazza and T. Pastore
This paper deals with the main aspects of the calculation of current and potential distributions and summarizes some results obtained on different systems using a computer program based on the boundary‐element method. Model results are compared to literature data for well‐known geometries, and to experimental data obtained both in laboratory testing with a suitable cell and in field testing on offshore structures (in particular platform nodes). The good predictive capability of computer modeling is shown.
1696
S. G. Real, M. R. Barbosa, J. R. Vilche and A. J. Arvia
The influence of Cl− ion concentration on the active dissolution and passivation of polycrystalline nickel rotating disk electrodes in sulfuric acid‐potassium sulfate solutions at pH 2.6 and 25°C was studied. The voltammetric data suggest that the passive layer results from a complex reaction involving the electrodissolution of the base metal, the formation of the anodic layer and its partial chemical dissolution. The influence of Cl− ions under different hydrodynamic conditions is discussed taking into account the competitive adsorption involving Cl− ions and species containing oxygen at the initial stages of the reaction yielding soluble products, and the appearance of
as a new phase. Solution stirring favors the electrodissolution reaction by increasing the transport rate of Cl− and H+ ions to the electrode surface, and simultaneously decreasing the thickness of the
layer.
1703
RaK‐Hyun Song, Su Pyun II and R. A. Oriani
The permeation of hydrogen through anodically passivated iron is measured at two values of hydrogen charging current density. The permeation transient is analyzed in terms of diffusion of hydrogen through iron in series with diffusion and migration of hydrogen ions through the anodic oxide, with local equilibrium at the metal‐oxide interface. The analysis yields the diffusion coefficient of hydrogen in the oxide, and the ratio of concentrations in the metal and in the oxide at the interface.
1706
M. Schlesinger, X. Meng, W. T. Evans, D. A. Saunders and W. P. Kampert
In this study we compared the micromorphology and magnetic properties in thin cobalt/phosphorus films of interest to the magnetic recording community. A number of film deposition parameters which control the physical properties of the end product were varied over a wide range, and a correlation between magnetic coercivity, as indicated by hysteresis curves on the one hand and the micromorphology as inferred from electron micrographs on the other, were compared. In addition, we have determined that some of the films possess a preferred c‐axis orientation perpendicular to the substrate. We conclude that for optimal results, the component crystallites should be approximately equal in size to those of the individual grains (islands). In addition, under those conditions, the films are free of any preferred crystallographic orientation, and thus have high coercivity.
1709
R. M. Williams, M. E. Loveland, B. Jeffries‐Nakamura, M. L. Underwood, C. P. Bankston, H. Leduc and J. T. Kummer
Mixed mass‐transport and kinetic control of sodium ion reduction at porous, inert electrodes on sodium beta′ alumina solid electrolyte (BASE) ceramic in a high‐temperature electrochemical cell has been observed and modeled.The high ionic conductivity of BASE and the reversibility of the liquid sodium/BASE anodic half‐cell led to assignment of potential‐dependent (nonohmic) resistances to kinetic and mass‐transport processes associated with the porous electrode. The morphology of these electrodes and typical sodium gas pressures are consistent with Knudsen, or free‐molecular, flow through the electrode. Equations for effusion of a gas through a cylindrical pore and equivalent pressure for a condensed phase evaporating irreversibly describe the pressure of sodium gas in the porous electrode. These equations and the current‐overpotential equation combine to yield equations for the cell's current‐voltage curve and its apparent charge‐transfer resistance in terms of rate and transport parameters. Inclusion of ohmic resistance losses including electrode sheet resistance via an iterative finite elements method results in a fairly complete model for the dc response of alkali metal thermoelectric converter cells.
1716
R. M. Williams, B. Jeffries‐Nakamura, M. L. Underwood, C. P. Bankston and J. T. Kummer
The exchange current, transfer coefficient, mass‐transport parameters, and electrode capacitance at the Na(g)/porous Mo/BASE phase boundary have been evaluated from 740 to 1220 K. The transfer coefficient exhibits a value close to 0.5 and the exchange current is dominated by collision frequency, with no significant activation energy. Since the porous Mo electrode adopts a fairly regular microstructure on the beta′ alumina solid electrolyte (BASE) surface, the magnitude of the exchange current of mature electrodes directly depends on the actual contact zone (the region where reaction occurs close to the porous metal/gas/electrolyte three‐phase contact) of the porous metal film with the BASE ceramic, and decreases slightly as grain growth occurs. The exchange currents and the mass‐transport parameters derived for very porous thin Mo electrodes indicate that the charge‐transfer reaction occurs at a small fraction of the interface. The capacitance calculated from the admittance of mature, clean, thin, porous Mo electrodes shows significant frequency dependence, suggesting a contribution from a Warburg impedance associated with surface diffusion. High‐frequency limiting capacitance and resistance values due to the interface show potential dependence and a value on the order of 1 F/m2 and 0.1–1.0 Ω‐cm2. If
and
are present, the interfacial capacitance is unstable with time as well as greatly increased, but after the electrodes have matured, with loss of these species, it exhibits very reproducible behavior with respect to potential, pressure, and frequency. Since the reaction area increases due to
in the pores, when they are present the apparent exchange current is also increased. Grain growth in very thin, mature, porous Mo electrodes can have a small effect on the electrode's exchange current.
1723
Y. Choquette, L. Brossard, A. Lasia and H. Menard
The hydrogen evolution reaction on the Raney nickel composite‐coated electrode is investigated in
aqueous solution at 70°C by ac impedance technique. In case of the chemically activated electrode material, the easiest pathway for the hydrogen evolution reaction (HER) is the Volmer‐Tafel mechanism, and the controlling step is the Volmer reaction with a small contribution of the Heyrovsky reaction. The electrode material becomes more electrocatalytic towards the HER after an electrochemical activation. In this case, the easiest pathway for the HER is the Volmer‐Heyrovsky mechanism, both steps being rate‐controlling.
1730
K. Ogura, C. T. Migita and H. Imura
Catalytic reduction of carbon dioxide to methanol has been achieved with a hydrogen fuel cell. This process involves a homogeneous and a heterogeneous catalysis. In the former, the catalyst consisting of a metal complex and methanol were applied, and in the latter Everitt's salt
which functions as an electron relay was used. The initial pH of the catholyte was fixed at 2, and the pH of the anolyte was required to be higher than 1.75 for the hydrogen fuel cell with
as oxidant to be feasible thermodynamically.
1733
D. Sedaries, C. Levy‐Clement and C. Godart
1739
Hideki Yabe, Keiko Ema and Yasuhiko Ito
1744
A. Bensmaili and F. Coeuret
Mass‐transfer coefficients between solid surfaces and submerged or unsubmerged liquid multijets are measured electrochemically. Various types of distributors are used; the open area
ranging from less than 5 to more than 97%. Unsubmerged and submerged multijets perform differently at high values of
when the injection Reynolds number is smaller than approximately 2000; the unsubmerged multijets interact with each other, thus forming a particular single jet which performs better than the corresponding submerged multijets.
1750
David Kohler, John Zabasajja, Frank Rose and Bruce Tatarchuk
In this study, the capacitances of stainless steel‐carbon composite electrodes in
were determined as a function of stainless steel loading, sintering temperature, and sintering time (i.e., 0.1–0.5g of stainless steel fiber/1g of carbon fiber, 1323–1423 K, and 2.5–152 min, respectively). Kinetic parameters were determined for stainless steel sintering and catalytic carbon gasification by weight loss measurements and capacitance changes using ac impedance, single potential step methods, and cyclic voltammetry. Apparent activation energies for carbon gasification and sintering were found to be 80 and 200 kJ/mol, respectively. Measured reaction rates and kinetic parameters were used to verify the importance of stainless steel‐carbon contacts toward overall electrode capacitance and to successfully predict a maximum capacitance of 45.3 F/g on a total electrode weight basis. This capacitance value yields energy densities of ca. 90 kJ/kg of electrode and does not require pressure for good electrical contact as do electrodes prepared from carbon powders. Capacitance and kinetic measurements demonstrate the ability to control conductivity and active surface area by adjusting: stainless steel to carbon loading, sintering temperature, and sintering time.
1757
Blandine Barbier, Jean Pinson, Georges Desarmot and Manuel Sanchez
Electrochemical oxidation of ω‐diamines on carbon fibers allows the bonding of these molecules to the carbon fibers. The conditions leading to this surface modification are described, and the influence of the different parameters is discussed. The structure of the bonded layer is studied by voltammetry, XPS, and SIMS. A mechanism is proposed which involves the coupling of a radical cation to the fiber. Toughness of carbon‐epoxy composites is improved when prepared from these modified fibers.
1764
Kevin J. Volk, Richard A. Yost and Anna Brajter‐Toth
Electrochemistry was used on‐line with high‐performance liquid chromatography with mass spectrometric and UV‐vis spectrophotometric detection to characterize the electrochemical oxidation pathways of 6‐thiopurine and 6‐thioxanthine. At low potentials, the electrochemical oxidation of 6‐thiopurine proceeds via one electron resulting in disulfide formation. It is proposed that purine‐6‐sulfenic acid is formed at potentials > +0.50V. Further oxidation of this unstable sulfenic acid presumably results in the formation of purine‐6‐sulfinate and purine‐6‐sulfonate. At potentials > +0.50V purine‐6‐sulfinate, purine‐6‐sulfinamide, and purine‐6‐sulfonate have been identified as the final products of 6‐thiopurine oxidation. On‐line electrochemical studies indicate that at potentials less than +0.30V, oxidation of 6‐thioxanthine results in disulfide formation. The disulfide readily disproportionates to regenerate the original thiol plus small amounts of a sulfinic acid. At potentials > +0.30V, it is proposed that the thiol group of 6‐thioxanthine is further oxidized to a sulfinic acid. Xanthine and 2‐hydroxypurine presumably form as a result of nucleophilic and electrophilic attack, respectively, on the sulfinic acid. At potentials greater than +0.40V, both the thiol group and the purine ring of 6‐thioxanthine undergo oxidation. Subsequent hydrolysis reactions produce an imine alcohol, the same intermediate which forms during uric acid oxidation, providing proof that oxidation of the purine ring occurs at potentials > +0.40V.
1772
Masashi Azuma, Kazuhito Hashimoto, Masahiro Hiramoto, Masahiro Watanabe and Tadayoshi Sakata
1778
F. Bellucci, I. Khamis, S. D. Senturia and R. M. Latanision
1785
Rachel Turoscy, Henry Leidheiser Jr. and James E. Roberts
1788
Susumu Kuwabata, Jiro Nakamura and Hiroshi Yoneyama
The electrical conductivity of polypyrrole films doped with various kinds of carboxylate anions is influenced by the basicity of the dopant anions in such a way that the greater the basicity, the smaller the conductivity. The observed dependence of the conductivity on the basicity is quantitatively discussed in terms of association of positive charges in polypyrrole with the dopant anions, the degree of which is different depending on the basicity of dopant anions. The conductivity of polypyrrole films doped with monoanions of dicarboxylic acids such as oxalic acid and maleic acid was decreased by immersing the film in such aqueous solutions as to cause dissociation of the doped monoanions into dianions, and the magnitude of the conductivity change was dependent on the degree of the dissociation of the doped monoanions.
1793
Katsuhiro Kawai, Naritetsu Mihara, Susumu Kuwabata and Hiroshi Yoneyama
1796
O. Aaboubi, J. P. Chopart, J. Douglade, A. Olivier, C. Gabrielli and B. Tribollet
1804
A. Felske, W. A. Badawy and W. J. Plieth
The electrochemical and photoelectrochemical behavior of passivated Ti electrodes in nitric acid solutions was studied and compared with the behavior of the electrodes passivated in sulfuric acid under the same conditions. The donor density and fiatband potentials were calculated from the capacitance measurements. The use of high intensity pulse lasers allowed photoelectrochemical investigations with photon energies smaller than the bandgap. The behavior of the passive films on Ti in nitric acid was not comparable with its behavior in
. The anomalous behavior of Ti passivated in nitric acid was discussed, and a model for the band structure of the passive film was suggested.
1810
D. Tafalla, P. Salvador and R. M. Benito
The photocurrent‐time dependence observed during experiments of water photo‐oxidation at an
single crystal, as a function of the applied potential and illumination intensity, has been analyzed on the basis of a kinetic model previously proposed. In a basic medium, the model assumes as a first step the photogeneration of surface‐bound
radicals from Ti coordinated OH− ions of basic character.
radicals further recombine on the
surface at a second step, leading to the formation of surface peroxo complexes (chemisorbed
species). Finally, photogenerated hydrogen peroxide species become photo‐oxidized and
evolves. At small bandbending the decay time constant, τ, characterizing the photocurrent transients is very sensitive to the applied potential, but does not depend on the illumination intensity. The photocurrent‐time dependence is dominated by
surface recombination, mainly via photogenerated
radicals. The pseudo first‐order rate constant of this reaction shows a gaussian distribution around a main value
, with
,
and
representing the bulk electron density and the bandbending respectively. This behavior indicates the existence of a heterogeneous distribution of Ti‒OH− surface sites mediating the process. In the high bandbending region τ is not sensitive to the applied external bias, but strongly depends on the light intensity. The transient is then due to the surface accumulation of positive charge at photogenerated
radicals, which produces a slow bandbending decrease and a concomitant diminution of the photocurrent. These effects are controlled by the rate,
, of hydrogen peroxide formation, which is a second‐order reaction in the surface concentration of photogenerated
radicals. A
can be estimated.
1815
M. R. Bassett and J. L. Hudson
1827
G. Tourillon, E. Dartyge, D. Guay, C. Mahatsekake, C. G. Andrieu, S. Bernstorff and W. Braun
The electronic properties and orientation of films of poly‐3‐alkylselenophenes
electrochemically deposited onto Pt have been studied by NEXAFS. The C—K edge characteristics reveal that several major effects occur when n is increased from 0 to 9: (i) In the undoped semiconducting form, a continuous decrease in the polymeric 1s → π* intensity is observed due to an overlap between the π* antibonding band from the polymer and the
band which develop when the alkyl chain length increases. (ii) The doping to the conducting state proceeds via a narrowing of the bandgap causing the appearance of metallic‐like behavior. However, the changes in the unoccupied antibonding π* band become less pronounced for a long alkyl chain length. (iii) The orientation of the selenophene ring structure switches from a “lying‐down” to an “on‐edge” configuration with the long alkyl chain being oriented perpendicular to the Pt surfaces.
1832
Daniel R. Baker and Mark W. Verbrugge
Current flow on a disk electrode can be described by means of a partial differential equation on a half‐infinite, two‐dimensional space. A numerical solution method to the steady‐state version of this problem is developed using an integral transform which reduces the problem to an integral equation on a bounded one‐dimensional space. The formulation yields approximate solutions by analytic means, which will be discussed in a future report. Simulation results are presented for the reaction profile over the disk surface (current distribution) and the corresponding surface potential and concentration profiles. The model can be used to identify conditions for which the current distribution will be uniform—a condition that facilitates the measurement of charge‐transfer rates. Simulation results are used to explain why electrodes of smaller radii (microdisks) are superior for the measurement of charge‐transfer rates.
1843
G. D. Simpson and R. E. White
A simple model for a parallel plate, zinc/bromine flow cell and associated storage tanks is presented and used to make time‐dependent predictions for various quantities in the system. The model is based on a previously published algebraic model of the cell at steady‐state and time‐dependent, first‐order differential equations for the storage tanks. The Butler‐Volmer equation is used for the electrochemical reactions, and the homogeneous reaction between bromine and bromide is included. The model predictions indicate that the charging operation of a zinc/bromine battery can be significantly improved by using a storage tank with a larger residence time for the bromine side of the system.
TECHNICAL NOTES
1846
R. E. White, F. Jagush and H. S. Burney
1848
F. A. Jagush, R. E. White and William E. Ryan
1851
Solomon Zaromb, C. Norman Cochran and Robert M. Mazgaj
1856
K. M. Abraham, D. M. Pasquariello and E. B. Willstaedt
1858
M. Schlesinger, Xianying Meng and D. D. Snyder
1859
K. Kondo, J. Ishikawa, O. Takenaka, T. Matsubara and M. Irie
SOLID-STATE SCIENCE AND TECHNOLOGY TECHNICAL PAPERS
1861
S. Mayumi, I. Murozono, H. Nanatsue and S. Ueda
The contact failure rate for double level Al‐Si‐Cu metallization was found to be much larger than that for double level Al‐Si metallization. The large failure rate in the first case was attributable to the loss of the first conductor in via holes due to pitting corrosion, which was caused at the chemical cleaning stage after the via holes had been opened. The corrosion pits were formed on the grain boundaries of Al alloys, each one accompanied by a particle in its center. The particles were identified as Cu containing precipitates with XMA. Corrosion was accelerated by increasing both Cu concentrations in the Al alloys and annealing temperatures. Some large precipitates with diameters of about 1.0 μm, which is equivalent to the film thickness, were found to exist in Al‐0.9%Si‐1.6%Cu alloys annealed at 450°C. When a large precipitate is situated in a via hole, corrosion occurs. The precipitate and Al on the surface form a galvanic cell, where the precipitate and the Al behave as the cathode and the anode, respectively, resulting in anodic dissolution of the Al. Fuming nitric acid solutions with
concentrations of more than 94% prevent the corrosion. Argon ion implantation in Al alloys was found to be effective in preventing corrosion. Annealing after the implantation enhanced the effect drastically. It was found that the Cu concentration on the surface decreases by about one‐half and that the precipitates on the surface are scattered into very small pieces when annealing takes place after implantation.
1867
J. W. Lutze, A. H. Perera and J. P. Krusius
Field oxide thinning in a modified LOCOS process is examined. Using electron beam lithography, thinning in active area spacings as narrow as 0.1 μm has been observed for the first time. Also, the cause of thinning is shown to be compressive stress in the field oxide, not two‐dimensional oxidant diffusion as has been previously asserted. A phenomenological model for field oxide thinning is given. Finally, a new figure of merit is defined which assesses the combined effects of encroachment and field oxide thinning in isolation processes for submicron integrated circuits.
1871
Z. H. Liu, P. S. Chen, Y. C. Cheng and P. T. Lai
1877
R. A. Clapper, D. G. Schimmel, J. C. C. Tsai, F. S. Jabara, F. A. Stevie and P. M. Kahora
This paper compares boron profiles measured by spreading resistance and SIMS techniques. The boron samples were prepared under conditions similar to an implanted‐diffused base of an n‐p‐n transistor. Good agreements between the two methods were obtained. Effects of sample preparation on spreading resistance measurements are discussed also.
1883
A. Anastopoulos and I. Poulios
The mathematical modeling of the capacitance characteristics of the metal‐insulator‐semiconductor‐electrolyte (MISE) configuration is attempted for thick and thin (<40 μm) semiconductor bases. The presence of energy states is accounted for and the contribution of the electrolyte double layer is examined for various electrolyte concentrations, with particular attention to the effect of the semiconductor thickness.
1887
Werner Kern
The purity of wafer surfaces is an essential requisite for the successful fabrication of VLSI and ULSI silicon circuits. Wafer cleaning chemistry has remained essentially unchanged in the past 25 years and is based on hot alkaline and acidic hydrogen peroxide solutions, a process known as “RCA Standard Clean.” This is still the primary method used in the industry. What has changed is its implementation with optimized equipment: from simple immersion to centrifugal spraying, megasonic techniques, and enclosed system processing that allow simultaneous removal of both contaminant films and particles. Improvements in wafer drying by use of isopropanol vapor or by “slow‐pull” out of hot deionized water are being investigated. Several alternative cleaning methods are also being tested, including choline solutions, chemical vapor etching, and UV/ozone treatments. The evolution of silicon wafer cleaning processes and technology is traced and reviewed from the 1950s to August 1989.
1892
A. Sherman
Chemical vapor deposition has been used to deposit titanium nitride
on silicon wafers at low pressures in a cold‐wall single‐wafer reactor. Experiments are reported for pressures in the range of 100–300 mtorr and temperatures between 450°–700°C, with titanium tetrachloride and ammonia as reactants. Both hydrogen and nitrogen are evaluated as diluents. Deposition rates as high as 1000 Å/min have been achieved. The chemical nature of the films are evaluated by Auger and RBS techniques, while the morphology is depicted by SEM. For the most part, the films are stoichiometric and contain small quantities of oxygen, chlorine, and hydrogen. Film resistivities as low as 50 μΩ‐cm are reported. Behavior of the
film as a diffusion barrier between silicon (boron doped) and aluminum, at annealing temperatures up to 550°C, is evaluated by measurements of contact resistance and diode leakage.
1898
Jun‐ichi Nishizawa, Kenji Aoki, Sohbe Suzuki and Kazuhiko Kikuchi
1904
Yicheng Lu, C. A. Paz de Araujo and T. S. Kalkur
Rapid thermal annealing (RTA) combined with proximity contact techniques has been used to activate Be‐implanted
. At the optimum RTA condition, which is 900°C for 5s, complete activation was achieved. The SIMS analyses show less dopant redistribution during RTA in comparison with furnace annealing. The proximity contact effectively suppressed arsenic loss and minimized
surface morphology degradation.
1907
Yue Kuo
Factors affecting the slope of the reactive ion etched molybdenum line have been studied with a simulation method and with experiments. Plasma chemistry and process parameters of the
mixture for the molybdenum etch have been examined. The theoretical calculation matches experimental results. Surface topography and composition of the etched molybdenum have been analyzed. A highly sloped molybdenum profile can be obtained by using the RIE method with a large process window.
1912
S. Prussin and Kevin S. Jones
The end‐of‐range defects for Si, P, Ge, and As implantations were evaluated by transmission electron microscopy over a range of implant doses. The results for P and As closely tracked those for Si and Ge, respectively. As the implant dose is increased, the defects increase slightly for Si and P but significantly for the heavier Ge and As atoms. For implant doses just above the critical value for amorphization, the defect concentration for the lighter Si and P atoms is about five times as great as for the Ge and As. The advantage of using low dose Ge postamorphization is discussed.
1914
T. Yoshida, S. Ogawa, S. Okuda, T. Kouzaki and K. Tsukamoto
A thermally stable, low‐leakage‐current, self‐aligned titanium silicide junction has now been fabricated. Ar+ sputter etching of Si substrate prior to Ti film deposition followed by two‐step rapid thermal annealing in
ambient are essential for realization of thermal stability up to 900°C in the titanium silicide layer. It is confirmed that the Ar+ sputter etching can reduce oxygen contamination at the Ti/Si interface. Consequently the titanium silicide layer could be kept extremely flat, with 60 nm thickness with specific resistivity of 19.3 μΩ · cm, even after 900°C, 30 min annealing. The junction leakage current was less than 2.1 nA · cm−2 at reverse bias of 5.0V for n+/p junction.
1918
Juri Kato
The effects of an additional fluorine ion implantation into highly doped n‐type silicon were studied. The additional F ion implantation technique reduced the diffusion enhancement of As and P, the growth of extrinsic defects, and dopant activation in n‐type silicon. F atoms were observed at the region where a high concentration of group V dopants existed. From these observations, we propose that F atoms are trapped by group V dopants in n‐type silicon due to the high electronegativity of F atoms. F atoms prevent As+‐SiI− and P+‐SiI− pairs from forming, the formation and dissociation of which cause diffusion enhancement and extended defects.
1924
S. J. Pearton, W. S. Hobson, F. A. Baiocchi and K. S. Jones
1935
R. Ramesham, S. Thakoor, T. Daud and A. P. Thakoor
Electronic neural network hardware implementations, particularly with full parallelism and self‐learning capabilities, have generated a need for compact “variable‐analog memory” elements as “synapses.” We report on electrochemically reprogrammable, solid‐state resistor elements as analog memory synapses, which are based on thermally evaporated, electrochromic tungsten oxide
thin films. The electrical resistance (connection strength) of such tungsten oxide films is reversibly modified by electrochemical cathodic injection into/anodic depletion of cations (H+, Na+, etc.) from a solid‐state electrolyte using a third control gate electrode. Two device configurations, one with chromium oxide and the other with sodium tungstate thin film electrolytes as hydrogen and sodium ion sources, respectively, are fabricated and tested. A critical component for the device operation is a silicon oxide blocking layer between the
and the electrolyte layer which provides a nonlinear barrier and thereby increases the stability of the device. The electrical resistance of the device can be modified to, and stabilized at, any desired value over a wide dynamic range of four orders of magnitude. The details of the fabrication, programming characteristics, and operation of the devices are reported in this paper.
1939
J. Schlote, K. ‐W. Schröder, P. Koy and G. Morgenstern
In connection with step coverage in CVD there are fundamental differences in behavior between top (outer edge) and bottom (inner edge) parts which do not depend on process characteristics. We propose a simple mathematical model to explain these differences.
1942
Y. ‐C. Simon Yu, Carol A. Hacherl, Evan E. Patton, Eric L. Lane, Tadanori Yamaguchi and Susan S. Dottarar
A planarized device isolution process (PLANTI) has been developed and optimized by using a deep‐trench isolation technology combined with a local oxidation of silicon (LOCOS) process for self‐aligned double polysilicon bipolar integrated circuit fabrication. To obtain a desirable trench etch profile and a uniform polysilicon etchback process, reactive ion etch processes have been optimized in terms of gas ratio, pressure, and power density. As a result of deep‐trench isolation, the collector‐to‐substrate capacitance was minimized at 9.0 fF for a bipolar transistor with a total area of
, while maintaining a transistor‐to‐transistor isolation voltage at 25V. A bipolar cut‐off frequency of 15.5 GHz and an ECL‐gate delay time of 66 ps were achieved. As proof‐of‐technology, four digital/analog circuits have been successfully fabricated. An amplifier with a bandwidth of 2.6 GHz and a gain of 20 dB, a frequency divider and a 8‐bit shift register with clock frequencies of 5.2 GHz and 2 GHz, respectively, and a dual 4‐bit analog‐to‐digital (A/D) converter with a sampling rate of 2.0 Gs/s were successfully demonstrated. A yield of 85% was achieved for the converter circuit. Furthermore, the PLANTI process was also successfully applied to an advanced BiCMOS process.
1951
W. Gao, S. ‐C. Li, D. A. Rudman, G. J. Yurek and J. B. Vander Sande
Metallic precursor ribbons with nominal compositions of
were produced by vacuum induction melting followed by melt‐spinning. Silver was added to aid in the alloy melting and metallic ribbon formation processes. These precursor ribbons were then subjected to a two‐step treatment: first oxidation and then annealing under controlled atmosphere. Oxide/Ag superconducting microcomposites with zero resistance at 104–110 K and a critical current density of 600 A/cm2 at 77 K in zero field were produced in this way with excellent reproducibility. The oxidation kinetics, oxide structure, and oxidation mechanism were studied by a combination of thermogravimetry, optical and scanning electron microscopy, x‐ray diffraction, and electron probe microanalysis. Three types of oxidation, combustion, fast and slow oxidation, were observed under different conditions, and they were found to be important for producing super‐conducting materials. While the combustion and fast oxidation produced distorted, inhomogeneous specimens, the slow oxidation produced nondeformed specimens with a relatively homogeneous mixture of oxide/Ag, from which the so‐called “2223” superconducting phase can be formed after suitable annealing.
1957
Y. Okada and S. Wagner
1960
J. M. Heddleson, M. W. Horn and S. J. Fonash
The inter‐relation between near‐surface lattice damage and hydrogen‐related effects in both the near‐surface and bulk of dry etched silicon is explored. Schottky barriers used to monitor the physical etching damage show an evolution of the near‐surface region during annealing that indicates hydrogen from the etch can be passivating some of the damage and affecting the electrical properties of contacts fabricated on the etched material. Spreading resistance measurements on corresponding samples allow this effect to be compared to the evolution of hydrogen‐related bulk effects.
1964
W. Wijaranakula and J. H. Matlock
The formation of thermal donors in silicon containing carbon at the concentration of
was studied after a 100h isochronal anneal in the temperature range between 400° and 900°C. Two types of thermal donors were observed. The first type was identified as the oxygen thermal donors formed in the temperature range between 400° and 550°C. The second type that formed in the temperature range between 500° and 800°C is hypothesized to originate from an agglomerate consisting of interstitial carbons. Agglomerates of interstitial carbons dissolved during the subsequent 1050°C anneal. The dissolution of interstitial carbon caused the carbon atoms to return to their substitutional sites. Excess silicon interstitials which were released during the dissolution condensed into interstitial‐type dislocation loops.
1970
K. Singh and S. S. Bhoga
The ionic conductivity of
glass inserted in
eutectic has been studied as a function of temperature and frequency. With such insertion, a prominent increase in the conductivity (by two orders of magnitude at low temperature) has been observed. The results have been explained in the light of dispersed phase theory, and supported by microstructural evidence. These results could provide a new approach for the development of polycrystalline solid electrolytes.
1973
Sacharia Albin, William P. Winfree and B. Scott Crews
1977
S. K. Pang, A. Rohatgi, B. L. Sopori and G. Fiegl
Dislocation‐free p‐type float‐zone (FZ), magnetic Czochralski (MCZ), and Czochralski (CZ) silicon wafers with resistivity of ∼50 Ω‐cm were investigated before and after oxidation. Prior to any heat‐treatment, FZ material showed highest minority‐carrier lifetime
with lifetimes approaching 5 ms. Lifetime in as‐grown FZ material appears to be limited primarily by point defects generated due to thermal gradient and temperature fluctuation. As‐grown CZ material showed lowest lifetimes of the order of 550 μs because of high concentration of point defects and residual impurities. This paper demonstrates for the first time that, unlike CZ, crucible‐grown MCZ material of resistivity as low as 30–60 Ω‐cm can have a lifetime as high as 2.5 ms. MCZ lifetime is limited by the residual impurities from the crucible and poly, even though the melt convection is reduced considerably. With the starting materials of comparable quality, the MCZ lifetime could be better than FZ because it has fewer defects due to stable liquid/solid interface. After oxidation, both FZ and MCZ samples gave a lifetime of ∼550 μs which could not be explained on the basis of oxidation‐induced point defects, because lifetime was limited by impurities diffused‐in during high‐temperature processing. Unlike FZ and MCZ, CZ material gave a postoxidation lifetime of only 100 μs due to high density of decorated defects, even though the impurity diffused‐in concentration was the same.
1981
Tatsuya Yamazaki, Hiroshi Minakata and Takashi Ito
We grew p+‐n+ silicon epitaxial layers, heavily doped with phosphorus and boron, continuously at 650°C using low‐temperature photoepitaxy. The n+ photoepitaxial layer with a phosphorus concentration above 1017 cm−3 grown on p− substrate shows high‐density surface pits, and as a result, poor crystal quality. However, when this n+ photoepitaxial layer is grown continuously on a heavily boron‐doped p+ photoepitaxial layer, these surface pits are drastically decreased, disappearing completely above a hole concentration of 1019 cm−3 in the p+ photoepitaxial layer. The phosphorus activation ratio and electron Hall mobility in the heavily phosphorus‐doped n+ photoepitaxial layer were also greatly improved. We investigated the cause of the surface pitting using a scanning transmission electron microscope, secondary ion mass spectroscopy, and energy‐dispersive x‐ray spectroscopy. We thus characterized the precipitation of phosphorus atoms on the crystal surface at the initial stage of the heavily phosphorus‐doped n+ photoepitaxial layer growth. Continuous p+‐n+ growth prevented phosphorus atoms from being precipitated. Using these p+‐n+ layers, we made p+‐n+ junction diodes and bipolar devices with superior electrical characteristics. The very low growth temperature enabled us to obtain extremely abrupt impurity profiles of boron and phosphorus.
TECHNICAL NOTES
1987
K. Kobayashi, Y. Inoue, T. Nishimura, M. Hirayama, Y. Akasaka, T. Kato and S. Ibuki
1990
C. P. Wu and F. Kolondra
1991
Kenneth P. Quinlan
1995
Akihiro Wakahara, Kangsa Pak, Tomohiko Sato, Hiroo Yonezu and Akira Yoshida
ACCELERATED BRIEF COMMUNICATIONS
1999
R. Cohen, Y. Lavi and E. Peled
2001
Katsumi Kuwabara, Jun Nunome and Kohzo Sugiyama
2003
Harland G. Tompkins, Mark Dydyk and P. W. Deal
2005
Michael J. Tierney and Charles R. Martin
2007
Ronald L. Cook and Anthony F. Sammells