Extraction Grid Research Articles

Particle simulation on extraction of positive and negative ions from a volume source

A two dimensional electrostatic particle simulation was done to study the extraction of positive (negative) ions from a volume plasma source. The simulation model is a rectangular system which consists of an extraction grid (left wall), a plasma grid, and a grounded wall (right wall). Upper and lower boundaries are connected by the periodic boundary condition. Full dynamics of charged particles are followed. Positive (negative) ions are extracted from the plasma region through a slit in the plasma grid to the extraction grid. Electrons are reflected by the magnetic filter and confined in the region to the right of the magnetic filter. Simulation results are compared with the Child–Langumuir law where the extracted ion current is proportional to the three-halves power of the potential of the extraction grid. In the case of the positive ion extraction, simulation results agree quite well with the Child–Langumuir law. Whereas, in the case of the negative ion extraction, simulation results agree with the law only for the lower value of extraction grid potential.

Self-aligned silicon field emission cathode arrays formed by selective, lateral thermal oxidation of silicon

This article reports on the fabrication of single crystal silicon, field emission tip arrays. The tips are formed by exposing a ‘‘capped’’ silicon pedestal to a lateral high temperature thermal oxidation. Tips formed using this process have uniform height and profile, and have small radii of curvature—typically less than 20 nm. We have fabricated tip arrays with up to 2500 elements and with the tip-to-tip spacing ranging from 1.0 to 5.0 μm. The thermal SiO2 has a nominal thickness of 600 nm and serves as insulation between the extraction grid and the substrate. The extraction grid is a self-aligned, sputter deposited Ti0.1W0.9 film. Extraction grid apertures range from 900 nm to 1.7 μm. The process sequence is flexible allowing the fabrication of tips with different heights. The vertical position of the tip apex with respect to the extraction grid aperture, therefore, is a controllable parameter. The preliminary emission measurements are presented here for arrays containing 400 elements. All testing is done in ultrahigh vacuum (typically less than 5.0×10−10 Torr). The areas of the arrays fabricated range from 40 μm×40 μm to 100 μm×100 μm. The tip heights range from 1.0 to 1.4 μm above the substrate silicon. The maximum current produced from a 400 element array is only 0.45 μA at 140 Vdc. The low array current indicates that only a small number of tips are active at a given voltage.

Atomic processes, cross sections, and reaction rates necessary for modeling hydrogen negative-ion sources and identification of optimum H− current densities

The principal electron excitation cross sections for vibrational excitation in a hydrogen discharge are reported. In the first chamber of a two-chamber hydrogen negative-ion-source system subject to the beam-line constraint of a maximum gas pressure, the density of vibrationally excited molecules reaches an asymptote for increasing discharge current or the equivalent fast electron density. Operating near this first-chamber asymptote, there exists a spatially dependent maximum negative-ion density in the second chamber. With the extraction grid placed at this maximum the optimum performance of a hydrogen-based system is determined. This optimum performance provides a criterion for the selection of differing source types for fusion applications.

Development of a dc, broad beam, Mevva ion source

We are developing an embodiment of metal vapor vacuum arc (Mevva) ion source which will operate dc and have a very large area beam. In preliminary testing, a dc titanium ion beam was formed with a current of approximately 0.6 A at an extraction voltage of 9 kV (about 18 keV ion energy, by virtue of the ion-charge state distribution) using an 18-cm-diameter set of multiaperture extraction grids. Separately, we have tested and formed a beam from a 50-cm-diameter (2000 cm2) set of grids using a pulsed plasma gun. This configuration appears to be very efficient in terms of plasma utilization, and we have formed beams with a diameter of 33 cm (FWHM) and ion current up to 7 A at an extraction voltage of 50 kV (about 100 keV mean ion energy) and up to 20 A peak at the current overshoot part of the beam pulse. Here we describe this part of our Mevva development program and summarize the results obtained to date.

Optimized eight-inch extraction system for reactive ion beam etching

A methodology of ion extraction system optimization is proposed, taking into account the influence of the ion density variation in plasma at the plasma sheath. Software tools for extraction grids design and machining have been developed, allowing compensation for the current, divergence, and deflection of a single beamlets. Simulation was used to predict the properties of the three-grid single beamlet. The measurement of the radial ion saturation current density distribution in plasma using a double Langmuir probe for an axial magnetic field up to 16 mT, vacuum chamber pressure ranging from 4×10−5 to 2×10−3 and rf power from 100 to 900 W has been performed. These distributions were used to calculate the grid compensation functions. An extraction system was optimized for the reference magnetic field in the air coil B=7.20 mT, vacuum chamber pressure p=2.0×10−4 mbar, rf power Prf=500 W, and net acceleration voltage Vnet=300 V with argon as a working gas. The ion current density distributions for optimized and nonoptimized extraction systems have been measured and compared with predictions of the model. A maximum detectable variation in the ion current density across the 8 in. target can be reduced to < 3%.

Electron attachment lineshapes, cross sections and rate constants at ultra-low energies in several halomethyl and haloethyl molecules

Electron attachment lineshapes, cross sections and rate constants are reported for the molecules CH2Br2, CF2Br2, CF3Br, CBrCl3, BrCN, C2H2Br4, CH3I, C2H5I and 1, 1, 1,-C2Cl3F3. Use is made of the krypton photoionisation method. In addition, the signal intercomparison technique is used to obtain individual channel cross sections and rate constants for Br- and CN-- formation from BrCN, and for Br- and Br2- formation from CF2Br2. Present results in 1,1,1,-C2Cl3F3 using a new ion extraction grid geometry are compared with earlier results, and differences discussed in terms of the electric field equipotentials of two extraction systems used in the present work. Approximate potential energy curves for BrCN are presented, and the temperature dependence of the Br- channel cross section discussed.

NUMERICAL SIMULATIONS TO IMPROVE THE ACCURACY OF ELECTRON-BEAM TESTING ON PASSIVATED INTEGRATED CIRCUITS

The Scanning Electron Microscope associated with a Voltage Contrast system is a contactless and non-destructive tool for the test of VLSI Circuits. Moreover, the potential of passivated tracks can be evaluated by Capacitive Coupling Voltage Contrast (CCVC). This paper presents a numerical simulation applied to improve the accuracy of CCVC measurements. Particularly, the role of the location of the beam, and the influence of the potential of the neighbouring tracks and of the extraction grid is studied.

Low power, 3.2-cm, efficient microwave electron cyclotron resonant ion source

Design improvements in a microwave (2.45-GHz) plasma disk ion source have resulted in a compact, efficient, and low power ion source capable of delivering ion beam current densities in excess of 10 mA/cm2 . The ion source has been designed for a 3.2-cm-diam double graphite extraction grid set. Ion beam extraction was studied using argon and oxygen discharges with input gas flows of 1–5 standard cubic centimeters per minute (sccm) and input powers under 150 W. Ion beam currents of 20 to over 40 mA were obtained at extraction potentials of 1600 V. Double Langmuir probe measurements were made for argon, and by using a free-fall diffusion model, extracted beam current densities were shown to be consistent with measured discharge electron and ion densities. These experiments demonstrate the ability of microwave plasma technology to produce high beam current densities for both inert and chemically reactive gases. The compact electron cyclotron resonant microwave ion source shows significant improvements in efficiency, lifetime, and ion beam current densities over earlier designs and over similar size conventional broad beam dc and rf ion sources.

Broad ion beam modeling for extraction optics optimization and etching process simulation

A model for calculating the ion–current density distribution of a broad-beam ion source at the target level is proposed. Three model parameters related to a single beamlet such as the beam divergence, the maximum ion–current density in the beamlet center, and a possible misalignment between the different grids are needed to describe the overall extraction performance. These parameters were determined partially with the axcel simulation code or experimentally via etch tests. In this work the model parameters have been determined for a capacitively coupled filamentless 13.56 MHz rf source operated with argon at low power levels ranging from 20–100 W and low acceleration voltages (100–1000 V). However, the model proposed is applicable also for other types of ion sources, for different extraction grid sets with varying geometry and a wide range of working conditions. Intensity profiles predicted by the above model show good agreement with experimental results.

Cusp width and power flow study at a high power magnetic multipole ion source

Power flow of the arc discharge on a high power magnetic multipole ion source is studied experimentally. Power flux distribution across line cusps is obtained using a probe made of a thermocouple. This confirms the fact that the full-width leakage is almost twice the size of the ion Larmor radius. It is shown that the arc discharge had strong spatial distribution in the arc chamber. The absolute value of the power flux is determined by comparing the integrated values of the distribution with heat loading measured by the coolant temperature rise. The result revealed that a very high power flux of more than 500 W/cm2 is concentrated on the anode wall at the line cusp position for a moderate arc discharge condition. It is also shown that a large portion of arc power is carried by the neutral particles. The composition of the source plasma near the grid is estimated from the power flux on the extraction grid.

Production of 35 keV, 1 A Steady-State Ion Beam

A hydrogen ion beam, (35 keV, 1 A) was extracted for 1 hour from a modified duoPIGatron ion source. The extraction grids and the plasma source were effectively cooled and the steady-state beam was extracted. When the arc plasma current was controlled by the feedback signal from the accel current, a sufficiently stable ion beam was extracted for 1 hour from 2 ms after the ignition of arc plasma. The stability of the ion beam power and the divergence angle were both less than ±0.5%.

Rf-broad-beam ion source for reactive sputtering

A broad-beam ion source to work with reactive gas has been developed. The source consists of a capacitively coupled rf-discharge plasma with multiaperture extraction grids and magnetic confinement. No hot cathode is used to avoid corrosion and achieve long time operation. To increase plasma density and long time stability we have tested different magnet plasma confinements with SmCo-magnets and coils. The dependence of the discharge and the ion current on the plasma volume dimensions will be discussed.

Particle simulation of electron beams with self-consistent magnetic fields

The design of electron guns to produce very-high-perveance ultrarelativistic low-emittance electron beams for free electron lasers or electron-beam weapons requires electron gun design programs that solve for the self-magnetic field of the beam as well as the space-charge electric field. Most gun design codes solve only for the angular component of the self-magnetic field using a straightforward application of Ampere's law. This paper reports on the development of an electron gun design code, Magun, which solves for all three components of self-magnetic field in a cylindrically symmetric system. As in other programs, the angular component is obtained from Ampere's law; the axial and radial components are found from the angular component of the vector potential, which is obtained from a solution of the vector Poisson's equation. Another feature of Magun is that the space-charge field is found by solving Poisson's equation for the electrostatic potential on a deformable finite element mesh. A discussion of the advantages of this kind of program is followed by an exposition of the code's algorithms. Finally, an application to the beam injector of the Experimental Test Accelerator (ETA) at the Lawrence Livermore National Laboratory is presented. Numerical and graphical results from our program will be compared with results from Lawrence Livermore's computer code EBQ. Emphasis is placed on the behavior of the beam injector (or gun) at extraction grid drives close to the point where a virtual cathode is formed, which provides a particularly severe test of the code.

A hemispherical retarding-field energy analyser with a microchannel plate detector and a high extraction field for voltage measurement in the scanning electron microscope

The previous hemispherical retarding-field energy analyser with a hemispherical mesh collector followed by an operational amplifier as a detector is improved in respect to the response time by using a microchannel plate (MCP) detector. The electron beam exposure (beam current*irradiation time) has been reduced to about 1/5000 of that in the previous experiment. In addition to this modification, the high extraction field is applied to the specimen without degrading the performance in the improved analyser by installing an auxiliary secondary electron extraction grid. The high extraction field can suppress the potential barrier formation in front of the fine electrode. The voltage on the 2 mu m electrode with dimensions similar to those of a current IC has been measured with the same accuracy as the previous results for the 10 mu m electrode.

Effects of internal and external magnetic fields on the characteristics of a magnetic multipole plasma source

Influences of an externally produced axial magnetic field and the field produced by the filament current on the performance of a magnetic multipole plasma source were experimentally studied using a high-proton-yield ion source developed for the JT-60 NBI. It was shown that both of the fields give significant effects on the arc discharge characteristics, the arc efficiency, and the plasma uniformity. The probe measurement showed that the change of the arc efficiency caused by the field is closely related to the change of the ion saturation current density distribution on the extraction grid. The change of the distribution can be explained by the effect of the magnetic field on the orbits of the primary electrons emitted from the filaments. On the basis of the experimental results, a permissible level of the external field in the source was evaluated to be ±5 G.

Ｋａｕｆｍａｎ型シングルグリッドイオン源の低エネルギー特性

In this present work the electrical properties of Kaufman type fine mesh single grid ion source were investigated for low energy ion irradiation during deposition of the semiconductor or superconductor films which were sensitive to defects due to high energy ion bomberdment. A 5.8 cm diameter ion source with a 1.7 cm diameter extraction grid of tungsten mesh (100 mesh/inch) was fablicated and tested. Maximum argon ion current density of 0.97 mA/cm2 over the energy range from 40 to 150 eV was obtained. The ion energy and the ion current density independently varied with the acceleration voltage and the discharge current respectively.

Ampere-order DC ion source.

A hydrogen ion beam of 1.1 A, 35 keV was extracted for 1 hour from a modified duo PIGatron ion source. The extraction grids and the plasma source were cooled effectively and the ion source was operated stably. When the arc plasma current was controlled by the feedback signal from the accel current, the sufficiently stable ion beam was extracted from 2 msec to 1 hour after the ignition of arc plasma. The stabilities of the ion beam power and the divergence angle were both less than ± 0.5%.

Improved energy analyzer for voltage measurement in electron beam probing for LSI diagnosis

To reduce the error caused by the local field effect in electron beam voltage measurement, a hemispherical retarding field energy analyzer extraction system for secondary electrons has been investigated. Based on the results of experiments, an improved energy analyzer was devised, which has a planar extraction grid combined with spherical extraction and retarding grids. It can generate a uniform extraction field of about 800 V/mm, keeping the retarding field spherical. Linearization error was reduced to ±0.15 V. Measurement error was also reduced to about one-tenth of our previously developed hemispherical analyzer. Typically error is less than ±0.3 V for electrodes more than 5 μm wide over electrode voltages from 0 to 5 V.

The significance of reactive ions and reactive neutrals in ion-beam-assisted etching

Using a custom designed beam analyzer system, the ionic species from a broad beam ion source have been measured under various etching conditions. In broad beam ion-beam-assisted etching (IBAE), it was observed that up to 25% of the ionic species are from the reactive neutral component backdiffusing into the ion source through the extraction grids depending on the ambient conditions. With the use of the constructed beam analyzer, known ionic species were also used for reactive sputtering and IBAE. The ionic species from Ar, Kr, Xe, Cl2, and CCl4 gases were directed at a Si substrate at 1 keV both with and without the presence of a chemically reactive neutral component of Cl2. In reactive sputtering, the etch yields were primarily dependent on the reactivity of the ionic species. However, in IBAE the etching mechanisms are more dramatically affected by the reactive neutral component on the surface than by the ion beam. That is, the etching component of the ion beam in IBAE is dominated more by its mass and acceleration voltage than by the reactivity of the species.

The fabrication and use of silicon and gallium arsenide ion source extraction grids

This paper describes the fabrication and use of silicon ion source extraction grids. Ion beam processing is dependent upon the extraction grid design and materials since the stability of the beam characteristics depends upon the stability of the grids. Conventional ion source extraction grids are produced using mechanical methods and have several disadvantages, including cost and distortion. A procedure is presented here for fabricating silicon grids using microelectronic fabrication methods. Designs for producing collimated, diverging or focused beams, and a fine mesh single grid configuration suitable for low energy operation are described. Beam profiles and performance are presented. Extraction grids of gallium arsenide have also been prepared, and their fabrication is described.