Search

We compare the use of externally applied pulsed and steady magnetic fields for the enhancement of deposition rate in reactive High Power Impulse Magnetron Sputtering (HiPIMS), using the deposition of amorphous hafnium oxide (a-HfO2) on Si as an example. The external magnetic fields were applied by a solenoidal coil, placed above the magnetron target. In the case of a steady magnetic field, a higher voltage was required to initiate the HiPIMS discharge, a longer delay time was observed for current onset, and the films became substoichiometric. For the pulsed magnetic field, film stoichiometry was maintained under all applied external magnetic field strengths. Varying the duration and delay times of the magnetic field after the application of HiPIMS voltage pulse revealed that the afterglow of the plasma between HiPIMS pulses was actively quenched by the presence of the magnetic field. Therefore, the optimum operation with the highest plasma density was obtained by applying the external magnetic field only when the plasma was established and removing it at the end of the HiPIMS pulse. A model to explain the findings is presented in which the target poisoning by oxide formation is determined by the conditions in the afterglow. We describe an approach to achieve maximum deposition rate while maintaining film stoichiometry and high film quality. Amorphous HfO2 films with leakage current through the film of less than 5 × 10-5 A/cm² at 0.1MV/cm were obtained at the maximum deposition rate. The refractive index, at a wavelength of 500nm, of the film prepared with pulsed magnetic field was 2.05 with a very low extinction coefficient of 8 × 10-5. [ABSTRACT FROM AUTHOR]

High-power impulse magnetron sputtering (HiPIMS) is used to deposit amorphous carbon thin films with sp3 fractions of 13% to 82%. Increasing the pulse length results in a transition from conventional HiPIMS deposition to a "mixed-mode" in which an arc triggers on the target surface, resulting in a large flux of carbon ions. The films are characterized using X-ray photoelectron spectroscopy, Raman spectroscopy, ellipsometry, nanoindentation, elastic recoil detection analysis, and measurements of stress and contact angle. All properties vary in a consistent manner, showing a high tetrahedral character only for long pulses, demonstrating that mixed-mode deposition is the source of the high carbon ion flux. Varying the substrate bias reveals an "nergy window" effect, where the sp3 fraction of the films is greatest for a substrate bias around -100V and decreases for higher or lower bias values. In the absence of bias, the films' properties show little dependence on the pulse length, showing that energetic ions are the origin of the highly tetrahedral character. [ABSTRACT FROM AUTHOR]

Microstructure evolution as a function of the substrate temperature and metal content of C:Ni nanocomposite films grown by hyperthermal ion deposition is investigated. The films were grown by pulsed filtered cathodic vacuum arc on thermally oxidized Si substrates held at temperatures in the range from room temperature (RT) to 500 °C and with the metal content ranging from 7 to 40 at. %. The elemental depth profiles and composition were determined by elastic recoil detection analysis. The film morphology and phase structure were studied by means of cross-sectional transmission electron microscopy and selected area electron diffraction. For RT deposition a transition from repeated nucleation dominated toward self-organized growth of alternating carbon and crystalline nickel carbide layers is observed at a Ni threshold content of ∼40 at. %. The surface diffusion increases concomitantly with the growth temperature resulting in the formation of elongated/columnar structures and a complete separation of the film constituents into the coexisting carbon and fcc Ni phases. At the highest growth temperature (500 °C) Ni shows a tendency to segregate at the surface of the growing film and to form a continuous layer for integrated Ni contents of ≥30 at. %. A corresponding structure zone model diagram is presented, and the results are discussed on the basis of the ion induced atomic displacement, temperature activated adatom diffusion, and the metallic island coalescence processes whose complex interplay results in the observed variety of the microstructures. [ABSTRACT FROM AUTHOR]

Ti/C and TiC/C multilayers with periods ranging from 2 to 18 nm were grown by filtered high current pulsed cathodic arc. The growth was monitored in situ by ellipsometry and cantilever stress measurements. The ellipsometry results reveal that the optical properties of the carbon vary as a function of thickness. Correspondingly, the stress in each carbon layer as measured in situ exhibits two well defined values: initially the stress is low and then takes on a higher value for the remainder of the layer. Transmission electron microscopy shows that the initial growth of carbon on Ti or TiC layer is oriented with graphitic basal planes aligned parallel to the interface. After 2–4 nm of growth, the graphitic structure transforms to amorphous carbon. Electron energy loss spectroscopy shows that the carbon layer simultaneously undergoes a transition from sp2 rich to sp3 rich material. [ABSTRACT FROM AUTHOR]

The authors report molecular dynamics simulations of alumina (Al2O3) during crystallization from the melt. Using liquid quench methods, they investigate the effect of cooling rate on the structural evolution of the α, κ, and the bixbyite phases. A critical temperature window is identified, where the time spent in this window is crucial in determining the extent to which the systems approach crystallinity. A strong dependence is observed between the final structure and the quench rate, which is most pronounced for the α phase and to lesser extent for the other phases. The results show that the different phases have different tendencies to crystallize that are determined by energetics, complexity of crystal structure, and the number of metastable states. [ABSTRACT FROM AUTHOR]

The structure and magnetic properties of Fe–C films synthesized using dual cathodes and coaxial cathodes in a filtered high current pulsed cathodic vacuum arc were studied. By altering the experimental parameters such as the trigger sequence, period, and pulse length, a series of single and multilayered Fe–C films were deposited. X-ray photoelectron spectroscopy and cross-sectional transmission electron microscopy were utilized to study the composition and microstructure of the films. The magnetic properties were studied using a superconducting quantum interference device magnetometer. The hysteresis loops show that the magnetic properties are controlled by the pattern of the layers and Fe concentration in the films. The reduction in the coercive field when Fe layers are mixed with ta-C layers is a key result and confirms that magnetically soft materials can be created by the multilayer method. The relationship between the magnetic properties and the structure of single and multilayered Fe–C films is discussed. [ABSTRACT FROM AUTHOR]

Al/AlN multilayers with bilayer thicknesses ranging from 10 to 50 nm were fabricated using a filtered cathodic arc deposition system. The effects on the microstructure of using two different deposition rates and applying an 8 kV pulsed voltage (plasma immersion ion implantation or PIII) to the substrate were explored. The microstructure was found to undergo a transition in which the Al transformed from layers to an aggregated phase under some conditions. This behavior is explained by a model in which the aggregation process is limited by diffusion. High deposition rates and the application of PIII were both found to encourage the transition by increasing diffusion. The model defines a phase diagram which predicts whether a multilayer or an aggregated structure will occur. A maximum in intrinsic stress was found to occur when the average feature size was 15 nm for both layers and aggregates. [ABSTRACT FROM AUTHOR]

We examine the effect of ion impact energy on the intrinsic stress and microstructure of aluminum nitride thin films deposited using a filtered cathodic arc. The dependence of intrinsic stress on ion impact energy is studied over the range from 0 to 350 V using dc bias and up to several kV for a fraction of the ions using pulse bias. For dc bias, the stress reaches a maximum at 200 V and decreases with further increase in ion bias. The preferred orientation of the crystallites was studied by cross-section transmission electron microscopy and diffraction. We found that there is a preference for the c crystallographic axis to lie in the plane of the film under high intrinsic stress conditions (4 GPa), whereas a c-axis orientation perpendicular to the plane of the film was observed for low intrinsic stress (0.25 GPa). We carried out calculations of the expected distribution of intensity in cross-sectional electron diffraction patterns to predict the effect of rotation freedom of crystallites with the c axis pinned. The calculated patterns agreed well with experiment. © 2004 American Institute of Physics. [ABSTRACT FROM AUTHOR]

Polycrystalline and microcrystalline materials grown as thin films often exhibit a preferred crystallographic orientation. The mechanism by which this preferred orientation develops has been the subject of some debate. In this article we present the results of electron microscopy and diffraction experiments examining preferred orientation in titanium-nitride (TiN) films deposited by physical vapor deposition with bombardment energies up to 18 keV using plasma immersion ion implantation. We show that as the bombardment energy increases the intrinsic stress is reduced. For example, at 3 kV more than a 50% reduction in stress is achieved. The results are consistent with a model in which the preferred orientation is the result of minimizing the energy, which includes both surface energy and bulk strain energy terms. [ABSTRACT FROM AUTHOR]

Plasma immersion ion implantation of insulating materials is inherently problematic due to charge accumulation on the insulator surface. Surface charge can be removed by the application of an ultrathin conductive film, which is essentially transparent to the incident ions. The minimum thickness of the film is determined by its capability to effectively conduct away the implanted charge. We present a model for charge accumulation on insulators during plasma immersion ion implantation and use this to study the plasma sheath width and voltage, with and without an ultrathin metal film. Charge accumulation occurs more quickly when the plasma has a directed velocity greater than the Bohm velocity, which is the case for a cathodic arc plasma. We show that for both cases the effectiveness of plasma immersion ion implantation is improved with the application of an ultrathin conductive film. [ABSTRACT FROM AUTHOR]

Plasma immersion ion implantation is a surface modification technique in which ions to be implanted are drawn directly from surrounding plasma by a pulse biased substrate. Ion acceleration occurs in the electric sheath that forms around the substrate. The technique is most often used to implant ions, such as nitrogen, from plasmas formed by excitation of a gas. More recently it has been applied with cathodic vacuum arc plasmas in order to implant metal ions. These plasmas have a directed ion drift velocity and generally a higher density than gas based plasmas. Both of these differences influence the development of the electric sheath, which determines the implantation profile and the overall stability of the process. If the sheath expands too much during the pulse, the plasma is depleted and implantation ceases. If, however, the sheath is too thin at any point in space breakdown occurs also stopping the implantation process. The major parameters affecting the sheath formation are plasma density, ion drift velocity, and substrate curvature. The influence of each of these factors is discussed and strategies for forming optimum sheath structures are identified. © 2001 American Institute of Physics. [ABSTRACT FROM AUTHOR]

When plasma immersion ion implantation is performed in the condensable plasma stream produced by a cathodic vacuum arc, deposition as well as implantation usually occurs. In this article we describe a method of achieving pure implantation by orienting the substrate so that it is shadowed from the plasma beam. Implantation depth profiles measured in glassy carbon and CR39 polymer using Rutherford backscattering are compared to illustrate the effectiveness of the technique for conducting and insulating substrates. Charging of the insulating substrate was found to cause a reduction in implantation depth compared to a conducting substrate. The depth profiles in glassy carbon were comparable to those achieved by conventional extracted ion beam implantation. Implantation of magnesium into hydroxyapatite and alumina was carried out to improve the bone cell adhesion onto these materials for prosthetic applications. © 2000 American Institute of Physics. [ABSTRACT FROM AUTHOR]

Presents information on a study which investigated ion energy and plasma characterization in a silicon filtered cathodic vacuum arc. Experimental procedure; Results and discussion; Conclusion.

Vacuum arc or cathodic arc metal plasma sources are attractive and convenient for depositing high-quality thin metal films and metallurgical coatings. It is a common practice to use a curved magnetic filter duct to eliminate macroparticle contamination and to bias the duct wall with a positive voltage to enhance the throughput of the metal plasma. The metal plasma usually consists of several charge states and time-of-flight (TOF) experiments show that the mean charge state of the metal ions decreases with increasing bias and magnetic field applied to the filter duct. We also derive the throughput of ions with different charge states at different bias voltage and magnetic field by the particle-in-cell (PIC) method. The ion trajectory is simulated neglecting the influence of electron charge. Our results show that the simulated mean charge state displays a similar decreasing trend as the bias voltage and magnetic field strength are increased. Phenomena such as reduction of the mean charge state at high magnetic field strength and bias can be explained in terms of standalone multiply charged metal ions under the influences of magnetic and electric forces inside the filter duct. Index Terms--Ion charge states, plasma duct, vacuum arc sources.

TiCx thin films were deposited on Al2O3 substrates at 900 °C by using a multiple cathode high current pulsed cathodic arc. The Ti:C pulse ratio and, hence, the composition was varied from C rich to Ti rich. It is found that the Al2O3 substrate is decomposed and reacts with the TiCx film to incorporate significant amounts of O and Al in the growing film. When the stoichiometry is suitable, epitaxially oriented Ti2AlC MAX phase with significant O incorporated is formed. The results indicate that Al2O3 is not an ideal substrate material for the growth of transition metal carbides and MAX phase thin films. [ABSTRACT FROM AUTHOR]

Monoclinic HfO2–xNx has been incorporated into two-terminal devices exhibiting either memristor or selector operation depending on the controlled inclusion/suppression of mobile oxygen vacancies. In HfO2 memristors containing oxygen vacancies, gradual conductance modulation, short-term plasticity, and long-term potentiation were observed using appropriate voltage-spike stimulation, suggesting suitability for artificial neural networks. Passivation of oxygen vacancies, confirmed by X-ray absorption spectroscopy, was achieved in HfO2–xNx films by the addition of nitrogen during growth. Selector devices formed on these films exhibited threshold switching and current controlled negative differential resistance consistent with thermally driven insulator to metal transitions. [ABSTRACT FROM AUTHOR]

HfO2-xNx films have been deposited by high-power impulse magnetron sputtering in an Ar-O2-N2 atmosphere with a series of nitrogen partial pressures. X-ray absorption spectroscopy revealed the optimum deposition conditions required to passivate O vacancies in the HfO2-xNx films by nitrogen. Low-mobility interstitial species prevent crystallisation of nitrogen-incorporated films. These effects combine to remove leakage paths resulting in superior breakdown strengths compared to films deposited without nitrogen. The bandgap was maintained at ~5.9 eV in the films in which nitrogen passivated the oxygen vacancies. This is essential to provide sufficient band offsets for HfO2-xNx films to be used an effective gate dielectric. [ABSTRACT FROM AUTHOR]

ZnO films have been reactively deposited on sapphire substrates at 300 C using a high impulse power magnetron sputtering deposition system and characterized structurally, optically and electronically. The unintentionally doped n-type ZnO films exhibit high transparency, moderate carrier concentration (~5 x 1018 cm-3) and a Hall mobility of 8.0 cm² V-1 s-1, making them suitable for electronic device applications. Pt/ZnO Schottky diodes formed on the HiPIMS deposited ZnO exhibited rectification ratios up to 104 at ±2 V and sensitivity to UV light. [ABSTRACT FROM AUTHOR]

The surface immobilization of enzymes is important for many biomedical and industrial applications; however, the choice of enzyme and immobilization method for best performance is often difficult. This paper demonstrates that Plasma Immersion Ion Implantation (PIII) provides high enzyme coverage and rapid covalent attachment (within the first minute) of incubation, and shows how this can be used to control the composition of the attached layer from an enzyme mixture. The use of a thermophilic enzyme combined with covalent immobilization viaPIII-treatment significantly improves β-glucosidase activity. β-glucosidase is an important rate-limiting enzyme in the production of cellulosic biofuel and is useful in many other industrial applications. β-glucosidase immobilized on the PIII-treated polystyrene surface demonstrated significant improvements in reaction rate and glucose yields compared to the enzyme on the untreated surface. A thermophilic β-glucosidase immobilized with our PIII-treatment method shows an average glucose production rate more than twenty-fold higher, over five uses, compared to the commercially-available β-glucosidase immobilized by physisorption after purification. [ABSTRACT FROM AUTHOR]

A monochromator/Fizeau interferometer/intensified CCD camera system is described that was developed for the measurement of the shape of spectral lines that are rapidly time varying. The most important operating parameter that determines the performance of the instrument is the size of the entrance aperture as this determines both the light throughput and the effective interferometer wavelength resolution. This paper discusses, both theoretically and experimentally, the effect of the finite source area on the instrumental resolution to assist in optimizing the choice of this parameter. A second effect that often produces a practical limit to the quality of the spectra is drift of the interferometer plates. Measurements of the shapes of spectral lines of ions and atoms ejected from the cathode spot of continuous and pulsed cathodic arcs are presented to demonstrate the utility of this instrument. [ABSTRACT FROM AUTHOR]

In the paper, we derive an algorithm which follows from the original van der Pauw’s technique for measuring resistivity with the added advantage of allowing contacts to be positioned a distance away from the boundary. For a large sample area, we show that the resistivity calculated by our algorithm is equivalent to the resistivity calculated by the original van der Pauw’s method. In practice, this configuration is easier to achieve and can eliminate errors associated with contacts that are not placed exactly at the edge. [ABSTRACT FROM AUTHOR]

Amorphous silicon-boron-carbon-nitrogen alloys were deposited by reactive magnetron sputtering, and their bonding statistics and electronic structure were investigated using a combined approach of experiment and molecular dynamics simulations. The authors show a difference between Si-based and C-based Si–B–C–N networks, and investigate coordination numbers and behavior of individual atom types. Furthermore, the authors calculate electronic structure and photoconductivity of the materials. The authors find that both a higher Si/C ratio and an addition of hydrogen increase a band gap of the materials. [ABSTRACT FROM AUTHOR]

Particle-in-cell simulations of plasma evolution in the vicinity of a mesh electrode that is semitransparent to ions are presented. Space-charge accumulation and ion focusing have an important effect on the ion trajectories and ion energy distribution inside the mesh. The results indicate that when such an electrode is used to assist plasma immersion ion implantation of insulating targets, the initial plasma conditions and mesh geometry are key factors in achieving optimum dose uniformity and implanted ion energy. [ABSTRACT FROM AUTHOR]

Micro-arcing and breakdown of the wall plasma sheath in radio frequency (RF) plasmas is studied in a hollow cathode system, using a Langmuir probe to measure the floating potential. Micro-arcing was induced reproducibly by controlling the floating potential. By dc grounding the hollow cathode, a negative current can flow to ground resulting in a higher voltage sheath between the plasma and the earthed vacuum vessel. The wall arcing threshold of the plasma potential in this system is in the vicinity of 50 V. In the present system, the charging process to rebuild the plasma potential, which is about a few tens of milliseconds, is much slower than the microsecond discharge. The arcing frequency was found to depend strongly on the plasma potential and the pressure. We propose a mechanism for the dependence of the frequency of periodic micro-arcing based on the development of electron field emission sites. The measurement of floating potential is suggested as a useful parameter to monitor and prevent micro-arcing in RF plasmas. [ABSTRACT FROM AUTHOR]

Cathodic vacuum arcs (CVAs) are well established as a method for producing metal plasmas for thin film deposition and as a source of metal ions. Fundamental differences exist between direct current (dc) and pulsed CVAs. We present here results of our investigations into the design and construction of a high-current center-triggered pulsed CVA. Power supply design based on electrolytic capacitors is discussed and optimized based on obtaining the most effective utilization of the cathode material. Anode configuration is also discussed with respect to the optimization of the electron collection capability. Type I and II cathode spots are observed and discussed with respect to cathode surface contamination. An unfiltered deposition rate of 1.7 nm per pulse, at a distance of 100 mm from the source, has been demonstrated. Instantaneous plasma densities in excess of 1×10[sup 19] m[sup -3] are observed after magnetic filtering. Time averaged densities an order of magnitude greater than common dc arc densities have been demonstrated, limited by pulse repetition rate and filter efficiency. © 2003 American Institute of Physics. [ABSTRACT FROM AUTHOR]

Examines some effects that can occur due to the streaming plasma in an ion source configuration. Case of energetic ion beam formation; High voltage substrate in a drifting plasma stream; High voltage ion implantation using a gridless ion source.

A multicathode high current pulsed cathodic arc has been used to deposit Ti2AlC thin films belonging to the group of nanolaminate ternary compounds of composition Mn+1AXn. The required stoichiometry was achieved by means of alternating plasma pulses from three independent cathodes. We present x-ray diffraction and transmission electron microscopy analysis showing that epitaxial single phase growth of Ti2AlC has been achieved at a substrate temperature of 900 °C. Our results demonstrate a powerful method for MAX phase synthesis, allowing for phase tuning within the Mn+1AXn system. [ABSTRACT FROM AUTHOR]

The substitution of oxygen for carbon in Ti2AlC M(n+1)AXn (MAX) phase, forming Ti2AlC1-xOx, has recently been reported. In this paper we simulate the effect of oxygen incorporation on mechanical and electronic properties using ab initio calculations. While the mechanical properties are not sensitive to the change in composition, the electronic properties can be tuned by varying the oxygen concentration. As the concentration increases, the conduction changes from in plane, typical of MAX phases, to conduction also in the c-direction. The conduction along c passes from insulating to n-type and then finally to p-type. These findings reveal an anisotropic semiconducting material. [ABSTRACT FROM AUTHOR]

Time resolved optical diagnostic techniques were used to study the production of highly ionized species in aluminum plasma produced by a centered-triggered high-current pulsed cathodic arc. Controlling the spacing between cathode spots enabled a correlation between a reduction in the mean charge state and an increase in the spacing of cathode spots to be observed. As the cathode current was increased, the distances between spots were reduced and these charge states were produced for longer times. Strong cathode spot coupling is proposed as a mechanism for the production of high charge states. [ABSTRACT FROM AUTHOR]

Thin films of Ti2AlC MAX phase have been deposited using a multiple cathode pulsed cathodic arc. Evidence for substantial oxygen incorporation in the MAX phase is presented, likely originating from residual gas present in the vacuum chamber during deposition. The characteristic MAX phase crystal structure is maintained, in agreement with ab initio calculations, supporting substitutional O in C lattice positions. On the basis of these results, we propose the existence of a MAX phase-like material with material properties tuned by the incorporation of oxygen. Additionally, possible unintentional O incorporation in previously reported MAX phase materials is suggested. [ABSTRACT FROM AUTHOR]

We report experimental results for time dependent plasma properties during microarcing in a rf plasma system. A cutoff discriminator bias voltage can be defined as a parameter representing plasma potential. The cutoff voltage during the microarcing does not exactly follow the dependence of floating potential. The plasma/ion density increases significantly after microarc initiation and then reduces slowly. We propose that a significant number of electrons enter the plasma in the initial period of the microarc and gain very high energy. These high energy electrons lose their energy through a relatively slower process of collisions while generating a significantly higher plasma density. [ABSTRACT FROM AUTHOR]

The composition of the low-energy ions arising from the rise and fall time periods of the voltage pulse in plasma immersion ion implantation (PIII) are simulated by particle-in-cell (PIC) modeling. It is shown that more than 70% of the low-energy ions with an energy corresponding to less than half of the applied voltage come from the short rise time period. Although the fall time period is typically 30 times longer than the rise time, less than 25% of the low-energy ions originate from it. Based on the PIC results, the depth profile of the implanted ions is derived using the Monte-Carlo code SRIM2000 [J. F. Ziegler, The Stopping and Range of Ions in Solids (Pergamon, New York, 1985)]. The low-energy ions are found to be implanted to a much shallower depth than ions introduced during the fall time period the concentration profile which decays more sharply into the bulk. These results indicate that the most effective way to reduce or increase the surface concentration is by adjusting the rise time of the PIII voltage pulse. This will require a power supply capable of fast rise times and good matching between power supply and load. [ABSTRACT FROM AUTHOR]

We report here experimental evidence of electron oscillation within the toroidal-section magnetic duct of a filtered vacuum arc plasma source. Our results clearly demonstrate that electrons can oscillate inside the duct under the combined effects of the electric and magnetic fields. In another experiment, we observe that, under the influence of the electron motion, the trajectories of the plasma ions are more or less unchanged except in the intensity when the Bilek plate is biased. Finally, our time-of-flight experiments show that the effects due to collisional scattering between plasma ions and oscillating electrons are masked by those associated with the metal plasma flow through the duct, and collisional scattering does not give rise to an increase of the mean charge state of the plasma ions. We conclude that the application of a bias voltage to the duct not only perturbs the ions but also influences the plasma electrons. Our results demonstrate that electrons at the central axis are one of the major reasons leading to improved plasma transport through the duct. [ABSTRACT FROM AUTHOR]

Describes vacuum arc plasma transport through a magnetic duct with a biased electrode at the outer wall. Maximization of the plasma transport efficiency of the filter; Insertion of a positively biased electrode into the duct along its outer major circumference.

We have studied the time-dependent development of electric double-layers (ionic sheaths) in saline solutions by simultaneously solving the sodium and chlorine ion continuity equations coupled with Poisson's equation in one dimension. The study of the effects of time-varying electric fields in solution is relevant to the possible health effect of radio-frequency electric fields on cells in the human body and to assessing the potential of using external electric fields to orient proteins for attachment to surfaces for biosensing applications. Our calculations, for applied voltages of 10–175 mV between the electrode and the solution, predict time scales of ∼0.1–110 µs for the formation of double-layers in solutions of concentration between 0.001 and 1.0 M. We develop an empirical equation that can predict the double-layer formation time to within 10% over this wide parameter range. The method has been validated by comparing the solutions obtained, once the program has run to a steady state, with the standard non-linear Poisson–Boltzmann equations. Excellent agreement is found with the Gouy–Chapman solution of the non-linear Poisson–Boltzmann equation. Thus the method is not restricted in accuracy and applicability as is the case for the linear Poisson–Boltzmann equation. The method can also provide solutions for cases where there are orders of magnitude changes in the ion densities; this has not been the case for previous studies where small perturbation analysis has been employed. The method developed here can readily be extended to two and three dimensions using time-splitting methods. [ABSTRACT FROM AUTHOR]