Search

The factors that have lead to the current widespread use of ion beam etching are described. A brief description of the physical processes involved in the removal of material from the surface being etched is followed by a discussion of several ion etching equipments that are commercially available. The etching characteristics of one of these systems which was developed in the authors' laboratories are described in some detail. The applications of ion beam etching are then reviewed and the uses of this technique in the production of active devices, VHF quartz oscillator crystals, surface wave and integrated optical circuits, and the preparation of specimens for electron microscopy are described. Finally the place that ion beam etching will take in future technology is forecast

Microfabrication techniques are being investigated in which a focused high‐energy ion beam bombards a surface to create high‐resolution patterns of implantation doping, damage in crystal structures, or other chemical change such as molecular bond breaking. The objective is to demonstrate unique microfabrication techniques such as maskless implantation doping and the formation of self‐aligned structures. Preliminary results are presented of doped regions written in silicon by a focused boron beam. Electrical measurements on heavily doped contact regions and on a lightly doped resistive region are in good agreement with what would be expected from conventional ion implantation employing masks. The exposure of electron beam resist by a focused beam of 60‐keV helium ions is also reported. The resist sensitivity is ∼100 times greater for these ions than for electrons. The focused ion beam used in these studies was generated in a focusing arm which includes apertures, deflection plates, and an einzel lens that ...

In recent years the requirements for higher current ion beams have increased notably in a wide area of applications. To meet these requirements it has been necessary to develop better understanding of the three main components of ion beam systems, namely the plasma source in which ions of the species needed can be produced at controlled rates with spatial and temporal uniformity, the extraction electrode systems for formation of low divergence and low emittance ion beams, and the transport of the ion beams in the presence of possible high space charge forces. In this review developments in these areas are discussed with the objective of indicating the advances in understanding the basic processes which are important.

A neutral hydrogen atom beam of 1 A (equivalent), 18 keV energy, with small divergence angle (≈ 1°) has been obtained using a grounded ion source. This source, a prototype capable of delivering a total ion beam of 2 A, utilises a large-area multi-aperture three-electrode extraction system. The neutraliser (a hydrogen gas cell adjacent to the extraction system where the accelerated ions are partially converted into fast atoms by charge exchange and dissociation) was biased at the negative potential corresponding to the desired beam energy. At the exit of the neutraliser, in this set-up, there is a strong electric field which decelerates the energetic ions remaining in the beam. This field removes the unneutralised fraction from the beam; furthermore it is possible to provide for the energy recovery of this fraction. This is a very important point in view of the future needs of the thermo-nuclear fusion. This disposition also simplifies the ion-source power system and the control of the source parameters.

A compact, cryogenically cooled, supersonic molecular beam unit is described, which is suited to the neutralization of ion beams having energies as low as 2 eV. Three criteria govern the secondary beam design: it must have sufficient divergence to encompass the primary ion beam, must be sufficiently intense to give the required degree of neutralization, and must be well enough directed that a negligible number of secondary beam neutral molecules leave the crossed beam chamber, or are adsorbed on its entrance and exit holes. These criteria are formulated into four generalized design equations. A solution of these equations for a particular experimental situation is illustrated, and a supersonic beam neutralizer based on this solution is described. Characterization of this instrument shows good agreement with the design requirements.

Human and avian erythrocytes were ion beam etched in the specimen chamber of a scanning electron microscope. The cells were periodically examined during the course of etching in order to demonstrate how sub-surface features reveal themselves during ion beam etching. Nuclear and vacuolar structures and a seam separating the nucleus and cytoplasm of the avian red cell, were revealed by ion beam etching.

The application of ion beam techniques to surface machining and the surface properties of glass is discussed. Particular reference is made to the feasibility of this work with commercially available apparatus. Also included is a brief review of the relevant theory, and some experimental results. Ion beam processing can not only achieve a surface finish but can also improve surface polish.

A device is described which can be used to determine the profile, position and emittance of beams of charged or neutral particles

The extraction of an intense ion beam depends largely on the shape of the plasma boundary and ion density. If the plasma density is too high a protruding plasma cap is formed and the ion beam becomes very divergent. A method of automatically controlling the plasma density to prevent this is described.

The use of electron and ion beams in the field of microelectronics has grown rapidly over the past ten years. For example, fine-focused electron beams, which initially were employed in diagnostic tests on integrated circuits and semiconductor devices, now are being utilized in the laboratory for creating devices and circuits with resolutions considerably higher than the values achievable by photolithography, and ion beams are being employed in the production of integrated circuits. It is seen that the beam approach offers a number of advantages compared with contemporary techniques. This article describes some of the devices that can be fabricated with electron beams, thus indicating the state of the art, discusses the advantageous features of ion implantation doping, and points out the limitations on beam processes imposed by fundamental electron/ion optical effects.

The phenomena of ion trapping in electron beams are studied experimentally. For a tubular electron beam the total trapped ion charge equals that of the electron charge if no opening of the trap occurs by any negative potential electrode. When a second electron beam is introduced along the axis of the tubular beam, very high ion densities can be established, because the slow electrons created by ionizing collisions of the axial electron beam can neutralize the ionic charge on the axis.

Construction and operating characteristics of an ion source of the capillary arc type are described. The source yields 0.4 milliampere of singly charged argon ions through an orifice 1 mm in diameter, which allows, with pumping speeds of 14 liters per second, maintenance of a pressure of 5×10−5 mm Hg in the adjacent chamber where the ions are used. The ion beam is easily focused at ion accelerating voltages of 1000 to 6000 volts. The source has been used as part of an apparatus for study of controlled sputtering of metals bombarded by argon ions, and the accompanying secondary electron emission. The source can be used advantageously in a number of analogous researches of a fundamental character, and also has application in the field of engineering and biology.

A steady state, dc discharge, electrostatically focused ion beam source with permanent magnet multipole confinement is described. The ion beam is low density, nb [inverted lazy s] 108 cm−3, low divergence, θ [inverted lazy s] 1°, energetic, Vb≤5 kV, and is suitable for investigating ion beam‐plasma interactions. Ion beam speed and species are determined either by modulating the beam and analyzing the resulting free‐streaming perturbations or by a pulse and time‐of‐flight technique. The amplitude of the free‐streaming signal is enhanced in the electron cyclotron harmonic passbands due to the decreased dielectric function of the background plasma. Two beam‐plasma instabilities, the electron cyclotron drift instability, and an instability at the lower hybrid frequency are observed using this ion source.

The interaction of electron beam with amorphous semiconductor thin films has been studied with the aim of realizing an electron beam computer memory. The operational mechanism of such a memory in terms of electron beam heating for the write process and secondary electron emission for the read process are described. Some experimental verification of the write and read processes as well as the feasibility of reversible electron beam switching process are reported.

Of the various factors that affect the total efficiency available from a mercury electronbombardment ion thruster, the shape of the magnetic field and the design of the ion optical system are found to be the most important. Data are presented to illustrate that an overall power-to-thrust ratio, including all thruster and propellant system losses, of 120 kw/lb is achievable at a specific impulse of 4000 sec and a mass utilization of 90%. The performance of the thruster system is presented as a function of mass utilization and specific impulse over the range of values currently proposed for solar electric propulsion missions.

Beam profile measurements have been carried out to study the influence of the length of the ion source outlet, the source pressure, and the source magnetic field on the properties of an ion beam. From a variation of the outlet length conditions can be deduced for optimum design with respect to gas efficiency and profile shape. The profile width is found to decrease with increasing source pressure and decreasing magnetic field. This effect is attributed to a corresponding change in the energy distribution of the ions. Profile distortions occurring at high source pressure are likely to be due to ion scattering in the extraction region.

If a metal target is bombarded by the high power density electron beam, the target material are made to evaporate and the evaporated atoms will be ionized by collision with the incident electrons. As these ions have lower energy than several electron volts, the incident electron beam can trap these ions in the potential valley of space charges. The trapped ions are diffused and accelerated toward the electron cathode which have the hole in the centre (in practice the spirally wound tantalum wire was used). And the ions are taken out through this cathode hole.Ion beams got by above process have fine diameter and low aperture angle, then the brightness of ion beam become higher in itself though the total ion current is not so high.Several metals as Cr, Si, Mg-Al alloy and others were used as targets. In the case of Mg-Al target the values of normalized brightness of the ion beam reached the order of 2×10-94.6×1010 A/m2 ·rad2. And the masses of ion beam components were analyzed by using the circular pole magnet.This ion gun will be usable as a fine metal ion source and a mass analyzer of target material.

The current density actually obtained in accelerated ion beams is much smaller than would be expected on the basis of the volume in phase space occupied by the ions leaving the source. We have made a study of the trajectories of ions in a beam to find out the reason for this limitation in performance. In the cases studied the limitation could be laid entirely to the effect of the shape of the plasma emission surface. The accelerated beams had the coherence expected from the small thermal spread at the source, but the effective object resulting from the curvature of the plasma surface had an axial length of as much as 10 cm at high current with the result that only a small fraction of the accelerated current could be converged through a small aperture in image space. The effect of space charge in the beams was made negligible by neutralization. No other factor had an appreciable effect on beam optics.

For the study of electron-ion collision products, the space-charge density of a tubular electron beam can be employed in order to trap ions. This produces large ion densities on the axis of the electron beam, where a second thin beam is used for further collision processes. Design considerations for an optimal operation are developed. A few preliminary measurements are reported.

A radiofrequency ion source with extraction electrodes situated outside the discharge tube is reported. A focused ion beam with an intensity of 10–200 μA and with a H 1 + ion content of 57% was obtained using a static transverse magnetic field. The gas consumption was 2.5 cm3/h and the total electric power was 200 W (including the extraction and focusing) supplied by a small 400 c/sec generator. Excellent stability for many hundred hours and the whole mechanical construction made it possible to use this source in a 1 MeV pressurized Van de Graaff accelerator.

The technique of ion beam sputtering is described and compared with r.f. sputtering. The advantage of high vacuum (10−5−10−7 torr) during sputtering deposition is important, especially for the deposition of semiconductors and the investigation of film growth by means of in situ observations in the electron microscope. Different facilities for the deposition of films and devices have been constructed, using ion sources of the Finkelstein type and a special acceleration system. So far, mainly the growth of silicon films on spinel and on amorphous substrates has been investigated. In the temperature range 450°–550 °C a gradual transition from amorphous to polycrystalline structures was observed and above 750 °C epitaxial films with good structural properties could be obtained on clean spinel substrates, and the electronic properties could be improved gradually.

A method of spatially analyzing a positive ion beam by electrostatic means is presented. In this method, a small portion of a given ion beam passes through a low‐density gas‐filled chamber, where ionization and charge exchange products are collected and analyzed so as to reveal the original positive and neutral beam components. Results of a detailed spatial analysis of a 10‐keV argon ion beam from a Von Ardenne duoplasmatron ion source are given and discussed. Comparison of integrated beam components with theoretical beam ``blowup'' indicates that microscopic and macroscopic neutralizations are the same within limits of the measurements.

A combination of electron beam and ion beam techniques were used in conjunction with conventional planar technology to fabricate a junction field-effect microwave switch. A digital tape-controlled scanning electron beam was used to expose mask patterns in polymethyl methacrylate resist whose line widths (≤1 um) are inaccessible to conventional photolithography; ion beam sputtering was used to remove a thin gold undercoat from within the exposed patterns, thereby maintaining the good edge resolution; and ion implantation was used to dope the closely spaced interdigitated source and drain regions thus exposed by the preceding process steps in the gold contact mask.

This paper is concerned with the possibility of making selected area mass-spectrometric measurements on plane specimens of metals and rocks, with a resolution of a few micrometres, by means of a focused ion probe and a conventional mass spectrometer. The ion beam intensities expected at the detector of the mass spectrometer and the current obtainable in the primary ion beam used for bombardment of the specimen are calculated. The results obtained suggest that ratios of isotopic abundance should be measurable to a reproducibility of approximately 01% on areas of a few micrometres diameter when the mass concentration of the element in the specimen and the abundance of its isotopes are greater than about 10%.

The bombardment of a surface by 1–25 keV primary ions causes the upper atomic layers to be “sputtered” or stripped off. Most of the material leaves as neutral atoms or molecules, but a small fraction is ejected as positive or negative ions. These secondary ions can then be characterized by a mass spectrometer to provide an analysis of that portion of the material sampled by the primary ion beam. At any one instant, the excitation volume is bounded by the primary ion beam periphery (1–300 μm) and extends to a depth of 5–50 A depending on experimental conditions. By mass spectrometrically monitoring one or several masses, the analyst can obtain isotopic and concentration in-depth profiles of the sample constituents. Alternatively he can obtain a complete mass spectrum as the primary beam erodes the sample. Control and localization of the primary ion beam erosion permits chemical analysis with lateral x, y resolutions of approximately 1 μm, the examination of fractional surface monolayers, an in-depth analysis with better than 100 A resolution, and the acquisition of secondary ion images. The mass spectrometer provides elemental coverage from hydrogen to uranium, isotopic characterization and detection limits of 10-15−10-19g, depending on the element under consideration. This paper will discuss the production of secondary ions, the various types of instrumentation used, and applications of the ion probe technique.

In order to establish an upper limit on the practicality and usefulness of ion‐beam deposition of thin films, an experimental beam‐deposition system was assembled which was capable of producing an ideally focused ion beam. The system was designed under the assumptions that the final beam diameter is limited by the thermal velocity spread of the ions and the spherical aberration of the final lens of the optical column. Theoretically, this apparatus is capable of producing a 0.01‐cm‐diam 6×10−8‐A indium ion beam at 100‐eV energy. Under these conditions, the predicted deposition sweep velocity would be 0.016 cm/min to produce a 2×10−6‐cm‐thick indium film. Chromatic aberration was the factor that limited the actual measured spot diameter to 0.075 cm for a beam current of 10−8 A and ion energy of 100 eV. Indium films were deposited in this apparatus with ion energies ranging from 100 to 500 eV (total dose of 7×1017 atom/cm2). Interferometric measurements indicated that these films retained, to some extent, th...

The feasibility of depositing metal films from a well-defined beam of positively charged chromium ions has been investigated. The design and development of an ion source suitable for this work are described and problems related to the extraction, focusing and deflection of such an ion beam are considered. It is found that the desired focusing and deflection conditions cannot be achieved in practice because of certain fundamental limitations: (i) A low-energy beam of less than 500 ev is necessary to obtain a deposit and prevent damage to the substrate by sputtering. Thus space-charge effects become predominant, causing spreading of the beam. (ii) A lens system having a very short focal length is necessary to give a high beam convergence angle. This limits the deflection obtainable without loss of definition. (iii) Energy spread within the beam may also contribute to poor focusing. A minimum beam diameter of 0·5 mm has been obtained from an ion beam of energy 230 ev having a current density of 8 μA mm−2. Under these conditions the condensation coefficient was found to be approximately 25%, giving a deposition rate of 100 Å min−1.

A newly designed microwave ion source has produced with a hydrogen ion beam up to 0.5 A steady state operation with ion energy of 1 keV. The 4 cm diam extraction electrodes are multiple apertures and are arranged in an accel‐decel arrangement. Maximum divergence is 2° from the beam axis. The beam is 85% molecular ions, of which either the diatomic or triatomic component can be maximized by an adjustment of the gas fed to a discharge region. The source is technologically simple, inexpensive, and long lived. The basic concepts underlying the production of ion beams are described.

An ion-bombardment mass spectrograph has been developed with the emphasis on high mass-resolving power of 2500, yet with simple construction. It consists of a Penning ion source for primary ion production, an electrostatic lens column(an einzel lens and a quadrupole lens)for beam focusing of the primary ion, and a CEC 21-110 B double-focusing mass spectrograph without the spark ion source for analyzer. The specimen was bombarded with the Ar+ beam of 8 keV energy, 300μ beam diameter and 1 mA/cm2 current density.A comparison of the spectra produced by the rf spark and the ion bombardment, and the influence of atmosphere in the specimen chamber on the secondary ion intensity have been studied. The ion bombardment produces various polymeric and compound ions while the rf spark produces few of these ions. The rf spark produces many more multiply charged ions than the ion bombardment It is found that the secondary ion intensites of metals increase when bombarding in oxygen pressure of larger than 3 × 10-7 Torr(ultimate pressure), and that the increasing tendency varying with different metals. The secondary ion intensities of impurity elements in a metal show the same increasing tendency with the matrix element. Especially, in case of the iron sample, the secondary ion intensities of impurity elements can be maintained at high sensitivity and stability, and the relative sensitivities approach to the sensitivity of the matrix element by bombarding in oxygen pressure of 3-5×10-5 Torr. No influence of hydrogen or nitrogen pressure in the specimen chamber on the secondary ion intensity has been observed.

By applying the large-angle elastic scattering theory of Rutherford and a modified energy loss law of the Thomson-Whiddington type, expressions are obtained for the fraction of an ion beam transmitted through a target, the rate of energy dissipation with depth and the backscattering coefficient. The model is found to be in good agreement with experiment. It is shown that an etch-resisting polymer (photoresist film) can be cross-linked by an ion beam as well as by an electron beam. The exposure characteristics of KTFR and AZ photoresists for 30-50 kev argon ions are very similar to those for electron beam exposure. On the basis of a quantitative discussion of ion beam scattering it is concluded that ion beam exposure is of great value in producing semiconductors in integrated circuits, because the resolution (or smoothness) at an edge is very high due to the small back-scattering effect, and also the penetration range is considerably greater than predicted by the theory of electronic stopping.

Two high transmission electron-bombardment ion sources, based on a principle of electrical rather than mechanical ion collimation, have been developed and tested with a 6 in. radius-sector type of mass spectrometer. The new ion sources had a measured transmission efficiency of 88 and 85% as measured by two different techniques. One ion source was designed for general use for experiments in which high sensitivity is required, while the second was designed for the study of unstable species such as excited neutral molecules, products of ion-molecule reaction and free radicals. These sources are capable of measuring partial pressures of both unstable and stable species as low as 10 ftorr. Collimating slits of 0.250 in. × 0.500 in. are used throughout the general purpose ion source to give electrical rather than mechanical collimation of the ion beam. Electrical collimation increased the transmission by over two orders of magnitude with respect to mechanical collimation, whereas the resolution was reduced only by about a factor of six. The `unstable species' ion source utilizes a principle of two parallel electron beams. One electron beam is used to produce a desired species and the second electron beam ionizes the species for mass identification. An experiment in which helium is excited to the 1s2s metastable level by 23 ev electrons in the first electron beam and subsequently ionized in the second electron beam by 6 ev electrons was performed to demonstrate the operation of the source.

The aberrations, caused by the space charge, of both focused electron and ion beams, which prevent the attainment of highly resolved images in electron microscopy and which set a limit to the maximum current density in electron or ion beam machines are evaluated. Firstly, the aberrations caused by space-charge defects such as defocusing, change of magnification, as well as spherical aberration, coma, curvature of field, astigmatism and distortion analogous to Seidel's five aberrations, are deduced by means of the eiconal method by use of the potential distribution of the space charge of the beam itself in the focusing lens. It is found that, although the defocusing effect due to space charge is a most important defect, the other aberrations caused by space charge are rarely as severe as the defocusing, because all aberrations due to space charge act against the normal Seidel aberrations, i.e. so as to reduce them. According to the numerical examples, even in electron microscopy and electron beam devices with a high current, the resolving power or the minimum spot size is not yet limited by the space-charge effect, but by the spherical aberration and astigmatism caused by manufacturing inaccuracy or inhomogeneity of the material. The caustic patterns of focused ion beams are, in general, much more affected by space charge than those of electron beams of the same incident energy, because of the large mass of the ion. It is experimentally ascertained that, even though the increase of focal length is certainly caused by the defocusing effect due to space charge, a limit to the radius of least confusion of a focused ion beam with an accelerating potential of 30 kv and a beam current 500 μA is apparently not yet set by the space-charge effect, but by spherical aberration, because of the cancelling action of negative and positive aberrations.

In electron tubes utilizing long electron beams ionization of residual gases can produce a positive space charge which can partially or wholly neutralize the negative space charge of the beam. Such neutralization has, under special conditions, been successfully utilized in focusing electron beams. In certain electron devices requiring long electron beams, this type of focusing is inadequate and an axial magnetic field is employed to focus the electron beam. However, when magnetic focusing is used, the assumption is usually made that the electronic space‐charge fields of the beam are not neutralized by positive ions. Measurements have been made on a magnetically focused electron beam yielding information on (a) the longitudinal drainage of positive ions from the electron beam and (b) the degree of positive‐ion neutralization of the electron beam. It is pointed out that localized ion traps may be present in the beam if the beam diameter varies periodically along its length because of the applied magnetic field.

A simple ion beam scanner for continuous monitoring of the properties of an ion beam is described.

A brief discussion of electron and ion microscopes is presented followed by a description of a low magnification, electrostatic ion microscope. This instrument is of the emission type wherein the source of positive ions is imaged. If more than one species of ion is emitted, the images corresponding to the various atom weights may be separated by means of a transverse magnetic field and the ratio of atomic weights calculated from the displacements. The possible uses of such an instrument are pointed out.

A hot cathode arc discharge type ion source is described. The discharge is collimated in an axial magnetic field and the ion beam is extracted axially. The maximum total ion current of the source, tested with hydrogen, is 42 ma. The extraction system allows focusing of almost 100% of this current. The limiting factor of the maximum ion current is analyzed. Although strong arc discharges could allow the extraction of higher ion currents, space charge spreading causes the ion beam to hit the front surface of the extracting electrode. Secondary electrons, released there, are focused back into the arc chamber producing instabilities of the source. A convenient means of regulation or modulation of the beam current is described. A magnetic analysis of the extracted beam shows about 34% atomic ion current.

A simple ion source is described for producing a parallel beam of monoenergetic argon ions. A h-f discharge produces the ions and the beam is extracted through a lens system consisting of two apertures, with a grid in the aperture of the accelerating electrode. Beams can be produced up to 5 mA with diameter 2.5 cm and length 10 cm.

The generation of high-current (\ensuremath{\sim}${10}^{5}$ A) pulsed ion beams with ion energy in the range 0.5-10 MeV appears to be possible by modifications of present electron-beam technology.

An improved form of a new type of electrostatic ion gun, which is suitable for ion etching and ion thinning, is described. It is shown that the device has three main modes of operation, namely an `oscillating mode', a `transition mode' and a `general glow discharge mode'. An explanation of these modes, together with the intensity and profile of the emerging ion beam obtained in each case, is given. It is suggested that advantage can be taken of these effects in specific applications.

This electrostatic ion source operates as a cold cathode device. It produces a beam of energetic positive ions of, for example, 100 μA/cm2 at a pressure of 5×10−4 Torr when operated at an anode potential of 10 kV. The ion beam is reasonably well collimated and the ions have a broad energy spectrum. It is therefore most suitable for sputter-ion etching of metallic and nonmetallic specimens for electron microscopy. It is relatively simple to construct and entirely reliable in operation. A number of optical and scanning electron micrographs of etched surfaces obtained with this source are given.

The operating voltage and pressure of the twin anode electrostatic ion gun have been reduced and the range of applications extended by incorporating a thermionic source of electrons. The construction and performance of the ion gun is described. The energy of the emerging ion beam has been reduced and thus the ions will produce less structural damage to any specimen under ion bombardment. This ion gun should have applications in ultra high vacuum.