Showing papers on "Focused ion beam published in 1981"

Electron beam lithography

Abstract: Electron beam lithography means writing patterns in thin films of electron sensitive material using a finely focused (sub-micrometre diameter) electron beam. By combining electrical scanning with interferometrically monitored mechanical motion, very complex patterns can be generated with great accuracy; for example, a pattern containing one-micrometre features can extend over 100 mm with a positional accuracy of 025 μm. In the manufacture of integrated circuits this technique is used for generating masks which are then projected optically onto silicon wafers which are coated with photosensitive resists. For making circuits with sub-micrometre features the resist-coated wafer can be exposed directly with the electron beam; however this is slow because the electron beam exposure is point-by-point and there are limits to electron beam intensity and resist sensitivity. Overcoming this limit is possible using techniques which allow the exposure of many points simultaneously but such techniques are not...

Low Energy Ion Beam Etching

Abstract: Etch-rate profiles have been obtained for copper, tantalum, stainless steel and quartz using a commercial end-Hall ion source. These profiles can be used to predict uniformity and etch rates in practical etching configurations. Compared to a gridded ion source, the lower ion energy of an end-Hall ion source is offset in etching rate by its large ion-current capacity, while the lower ion energy can be a significant advantage in damage-sensitive etching applications. INTRODUCTION End-Hall ion sources are widely used for in-situ cleaning and ion assisted deposition, which are necessary in many thin-film processes. Compared to gridded ion sources with expensive and complicated high-maintenance grids, end-Hall ion sources have the natural advantages of lower cost, broad ion-beam coverage, and greater reliability. End-Hall ion sources also have much larger ion-current capabilities at low ion energies (200 eV and less) permitting useful etch rates at these low energies and reducing or avoiding the damage that would otherwise occur to the surface being etched.[1] There are, however, limited etch data for this low energy range. The argon ion beam from an end-Hall ion source was used to generate both low-energy sputter yields and etch-rate profiles for copper, tantalum, type 304 stainless steel, and quartz. The etch profiles were used to predict uniformity and etch rates for etching apparatus with a single-rotation stage or a large singlerotation substrate. Excellent uniformity can be obtained at useful etch rates with the ion-source location selected by this procedure. The practical considerations of interactions with the vacuum-chamber wall are also addressed. ION-BeAm ChARACTeRIzATION A commercial end-Hall ion source[2,3] was used for the etching described herein. The ion-beam profile and on-axis retarding potential energy analysis are shown in Fig. 1 for a discharge voltage and current of 150 V, 5.7 A. The working gas was argon and the background pressure was 2×10-4 Torr. These characteristics were obtained using an ion-beam probe, the design and use of which is described in a previous publication.[4] To minimize vignetting errors of the probe screen, the data for the profile in Fig. 1(a) were obtained with the probe facing, and kept a constant distance from, the center of the ion-source exit plane. Data from both sides of the ionbeam axis are plotted in Fig. 1(a), with the small differences in data at angles greater than zero indicating an essentially axisymmetric beam profile. The etch measurements were obtained at a flat plane oriented normal to the ion-beam axis (see Fig. 2). The profile that would be expected at such a flat plane would be reduced by the cube of the cosine of the angle from the axis, and is shown by the dashed line in Fig. 1(a). (The square of the cosine results from the inverse-square reduction due to increased distance, while an additional cosine results from the reduction due to the oblique incidence of the ions on the etched surface.) Figure 1. Ion-beam characteristics. Low-energy Ion-Beam etching James R. Kahn and Harold R. Kaufman Kaufman & Robinson Inc., Fort Collins, CO 80524 © 2006 Society of Vacuum Coaters 505/856-7188 49th Annual Technical Conference Proceedings (2005) ISSN 0737-5921 1 The retarding potential energy analysis on the ion source axis is shown in Fig. 1(b) for the same operating conditions. This plot shows the ion current reaching the probe over a range of positive probe potentials. The slope of this curve at a given potential indicates the density of the ion current with energies near that potential. (The curve would have to be differentiated to show the actual energy distribution of the ions.) The shape of the curve in Fig. 1(b) indicates large numbers of ions with energies near 25 and 150 eV, with smaller numbers distributed between. The ions near 25 eV are mostly charge-exchange ions,[4] generated when energetic beam ions pass near background neutrals, although ion-source design or operation can also contribute to the density of these low energy ions. The ions closer to 150 eV are those that provide most of the ion-beam processing capability. The mean energy can be obtained from the area under the retarding-potential curve and is about 89 eV for the curve shown by the data. If the contribution from charge-exchange ions (above the dashed line) is excluded, the mean ion energy increases to 114 eV. Figure 2. Configuration used for etch-rate profiles. eTCh-RATe PROFILeS The configuration used to measure etch-rate profiles is indicated in Fig. 2. Etch samples were placed in a plane normal to the ion-source axis at a distance of 30 cm from the source. Because of the beam symmetry shown in Fig. 1(a), etch measurements were made on only one side of the ion beam. The samples were 32 mm square with smooth surfaces. The etch measurements were calculated using electronic-scale weight measurements, the material density, and the exposed area. Step measurements with a stylus profilometer were used to verify the procedure used. The run times were adjusted to be long enough to avoid effects of surface anomalies such as native oxides and short enough to avoid any significant effect of texturing. Checks were also made to assure that sputtering from other hardware in the configuration of Fig. 2 did not affect the etch rate measurements. Figure 3. etch-rate profiles. Etch-rate profiles were obtained with copper, tantalum, 304 stainless steel, and quartz samples and are shown in Fig. 3(a). Using the current density and etch rates on axis (normal incidence), the sputter yields in atoms/ion for copper, tantalum, stainless steel, and quartz are 0.76, 0.14, 0.49, and 0.31. The mean ion energy is roughly 100 eV (between 89 and 114 eV). The same profiles were normalized to unity on the beam axis and are shown in Fig. 3(b), together with a normalized profile of ion current density obtained from the dashed line in Fig. 1(a). Referring to Fig. 3(b) and normalized parameters, the etch rates of all materials except tantalum increase above the ion current density and stay well above it as the angle increases from zero, indicating a substantial effect of the angle of incidence. The importance of angle of incidence is, of course, well known.[5] The magnitude of this effect, however, is often not available for the energies and materials of interest in a particular application. © 2006 Society of Vacuum Coaters 505/856-7188 49th Annual Technical Conference Proceedings (2005) ISSN 0737-5921 2 © 2006 Society of Vacuum Coaters 505/856-7188 49th Annual Technical Conference Proceedings (2005) ISSN 0737-5921 3 eFFeCT OF vACUUm-ChAmBeR wALL The etch profile obtained in a practical application can be affected by the proximity of the vacuum-chamber wall. A test configuration for demonstrating this interaction is shown in Fig. 4. Because the presence of the wall resulted in a nonsymmetrical etch distribution, the array of etch samples was centered relative to the ion source. The vacuum-chamber wall was simulated with a sheet of 304 stainless steel. The etch samples were made of the same material as the simulated vacuum-chamber wall to simplify the analysis of the test results. The baffles were also made of 304 stainless steel sheet, although this material is not as important because the etch samples do not “see” the surface of the baffle that is sputtered. Figure 4. Configuration used to measure wall effect. The wall-interaction test sequence consisted of three tests with the ion source operating at the same conditions as described in connection with Fig. 1. The first test was with no wall or baffle present. The second test was with only the simulated vacuum-chamber wall present. The third test was with both wall and baffles present. The etch-rate profiles for these three tests are shown in Fig. 5. The profile without wall or baffle establishes the basic etch-rate profile and is not significantly different from the 304 stainless steel profile in Fig. 3. The etch-rate profile with only the wall present shows the reduced etching of samples near the wall due to the deposition of sputtered material from the wall. The difference between these two profiles gives the deposition from the wall shown by the curve at the bottom of Fig. 5. The scatter is substantial for these deposition data because they were obtained from difference measurements. The deposition also included contamination from wall impurities, which in turn resulted in a dull, textured surface, which also affected the etch rate. Even so, the effect of wall deposition is clearly seen to extend over most of the samples. The etch-rate profile with both the wall and the baffles present shows that the sputtering from the wall is essentially avoided and the normal etch profile is obtained quite close to the wall. This profile falls off only where the samples are shadowed by the baffles. Although Fig. 5 is presented here in the context of etching, it should be apparent that it also has significance for controlling contamination in ion assist applications. Figure 5. etch-rate profiles showing the wall effect and its control. UNIFORmITY FOR SINGLe-ROTATION STAGe Single-rotation stages with multiple substrates or the singlerotation of large wafers or substrates are both widely used for thin-film processing. A numerical modeling procedure was used predict etch uniformity for such a configuration. This procedure assumed a planar etched region and divided this region into 20 radial zones and averaged over 72 equally spaced circumferential locations for each of these zones. The shape of the etch-rate profile at the planar surface being etched (fig. 3) was assumed to be of the form cosxΘ, where Θ is the angle from the beam axis and where x = n+3 for a planar surface and n is the shape parameter and Θ is the angle from the beam axis. This formalism has been used to describe the shape of end-Hall ion beams[6] and is extended here to the etch profile of such a beam. Because th

A mass‐separating focused‐ion‐beam system for maskless ion implantation

Abstract: The use of a focused ion beam for direct implantation of dopants into a semiconductor substrate results in appreciable simplification in the processing of semiconductor devices. We have demonstrated that liquid metal (LM) field‐ionization sources (based upon the electrostatic formation of an emitting cusp of liquid metal) offer the necessary high brightness to make focused ion beam microfabrication economically feasible. This paper reports upon two developments: (1) the development of eutectic‐alloy LM ion sources for the production of boron and arsenic for direct implantation of silicon devices, and (2) the development of a three‐lens variable‐energy focusing column that incorporates a mass‐separator of low aberration. Mass spectra of the ion emission of these sources show that the stoichiometric fraction of boron and arsenic is emitted. We have also demonstrated that the high vapor pressure of arsenic can be suppressed in the eutectic liquid metal, and that boron is predominantly emitted as a singly ionized species, while arsenic is emitted as both singly and doubly ionized species. A new focusing column has been developed that incorporates the new ion sources. It has the capability for focusing to sub‐micrometer dimensions with mass‐separation, a variable beam voltage of up to 150 kV, and a spot current of near 1 A/cm2. A high‐speed electrostatic‐deflection system with microprocessor control allows this machine to perform simple pattern exposures. Examples of the operation of this microfabrication system with eutectic alloy sources will be presented.

FET fabrication using maskless ion implantation

Abstract: Focused ion beams from Au–Si, Au–Be, and B–Pt liquid–metal–alloy ion sources have been used to implant GaAs and Si. An Al stopping layer on the wafers was used to trap the Au and Pt ions. Hall mobilities consistent with those in bulk materials have been obtained for B‐doped Si and Be‐doped GaAs. In addition, a 2000‐A‐diam Au–Si focused ion beam was used to implant the doped regions of GaAs metal‐semiconductor gate field‐effect transistors. The 140‐keV Si++ beam component was deflected under computer control to implant 8×50 μm active channel regions and 16×50 μm contact regions. The devices were metalized using conventional lithography. DC electrical characteristics of the 1.5‐μm‐gate‐length devices are comparable to those of conventionally processed devices of identical geometry.

Germanium selenide: A resist for low‐energy ion beam lithography

Abstract: A negative resist consisting of a layer of germanium selenide ∠1800 A thick with a surface layer of silver selenide ∠100 A thick was irradiated with 10 to 38 keV He, N, Ar, and Xe ions. While polymer resists require an incident ion to pass through the entire thickness of the resist film for exposure, germanium selenide films can be exposed with ions which penetrate only the ∠100 A thick silver selenide surface layer. Thus, germanium selenide is an attractive resist system for low energy focused ion beam lithography. Resist sensitivities of ∠1013 to 1015 ions/cm2 were measured. Both nuclear and electronic stopping contribute to exposing the resist. Submicron lines have been produced in germanium selenide with a 20 keV finely focused ion beam.

Ion beam oxidation

Abstract: We describe a new technique for controlled oxide growth using a directed low‐energy ion beam. The technique is evaluated by fabricating Ni‐oxide‐Ni and Cr‐oxide‐Ni tunneling junctions, using oxygen ion beams with energies ranging from 30 to 180 eV. High ion current densities are achieved at these low energies by replacing the conventional dual grid extraction system of the ion source with a single fine mesh grid. Junction resistance decreases with increasing ion energy, and oxidation time dependence shows a characteristic saturation, both consistent with a process of simultaneous oxidation and sputter etching, as in the rf oxidation process. In contrast with rf oxidized junctions, however, ion beam oxidized junctions contain less contamination by backsputtering, and the quantitative nature of ion beam techniques allows greater control over the growth process.

System and method for deflecting and focusing a broad plasma beam

Abstract: A method and system for deflecting a broad ion plasma beam which includes an ion source for forming an ion plasma, an extraction means for extracting a broad ion plasma beam from the ion plasma, and deflection means including a non-grounded surface located in the path of the ion plasam beam and at an angle to the path for deflecting the ion plasma beam to a target material. A grounded, screen grid is located in front of the deflecting means in the path of the ion plasma. The screen grid has openings which permit passage of the ions in the ion plasma, but block passage of the electrons. The plasma beam is deflected by the deflection means and the grounded, screen grid onto the target material for sputter cleaning, deposition and ion milling applications.

Electromagnetic instabilities in a focused ion beam propagating through a z‐discharge plasma

Abstract: A beam‐plasma system consisting of a focused light ion beam propagating through a z‐discharge plasma is analyzed for electromagnetic velocity space instabilities. In particular, the Weibel (k⋅B0=0, k⋅Vz≃0) and the whistler (k×B0≃0, k⋅Vz=0) instabilities are studied. It is found that unstable modes do not grow fast enough to affect the beam propagation in a pellet‐fusion‐reactor chamber. For k⋅Vz≳0, the instabilities convect with a velocity much less than the beam streaming velocity.

Specimen coating for high resolution scanning electron microscopy

Abstract: Thin conducting films, produced by evaporation or soft vacuum sputtering generally show cracks and grain formation, when examined under high resolution scanning electron microscopy (SEM). These artefacts can obscure surface features of coated specimens or cause confusion in the interpretation of micrographs. No such structures have been observed in films produced by ion beam deposition. Ion beam deposition equipment is described in which a cold cathode saddle field ion source, operating at low pressure (15mPa), produces a 2 mm diameter beam of energetic ions (5 keV) and neutrals. With the beam directed onto a target at 30° to glancing incidence, the sputtered material coats the specimens, which are held in a planetary system for good coverage. Conditions favouring fine grain growth are a high nucleation density and low energy transfer to the substrate by thermal conduction or radiation or by particle or photon radiation. These conditions are satisfied by ion beam deposition but evidently not by evaporation or soft vacuum sputtering. With the specimen stationary, sharp shadowing is obtained because the target acts almost as a point source, because of the small diameter of the beam and because there is little scatter at the operating pressure.

Submicron niobium tunnel junctions with reactive ion beam oxidation

Abstract: A new reactive-ion-beam oxidation technique has been applied to the fabrication of rugged, highquality niobium-lead alloy Josephson tunnel junctions. Control of critical current density over a wide range is possible, and critical current densities exceeding 10 A/cm have been obtained. In addition, a process-compatible edge geometry has been developed which allows a junction to be formed on the faceted edge of a niobium base electrode, yielding devices with areas of 10cm using I-jlm photolithography.

Electron temperature requirements for ballistically focused neutralized ion beams

Abstract: An envelope equation, based on a self-consistent equilibrium model of a charge- and currentneutralized, ballistically focused ion beam is derived. Focusing of the beam is limited by electrostatic fields within a Debye sheath at the beam edge and residual electrostatic fields within the beam interior. Numerical and approximate analytical solutions indicate that the initial temperature of the neutralizing electrons must be less than about 10 keV in heavy-ion inertial-confinement fusion systems of fewer than ten beams.

Reactive ion beam etching of silicon compounds with a saddle–field ion source

Abstract: Previous work using cold cathode ion sources of the saddle‐field type has indicated that they are suitable for the reactive ion beam etching of many materials. Using a B93 source with a nominal beam current density of 0.16 mA cm−5 (neutral beam ion equivalent for argon), targets of Si, SiO2, and Si3N4 have been etched with a beam produced by injecting the source with Ar, CF4, and CHF3. Data are presented which show the etch rate dependence on beam current density for various combinations of etchant and target material. There is evidence to indicate that silicon was etched by free radical species which were produced at low beam energy (0.6 kV anode potential). Test patterns etched in SiO2 and Si3N4 by ’’CF4 and CHF3 beams’’, which have been examined by Scanning Electron Microscopy, reveal no evidence of redeposition, trenching, or lateral etching. Lines and troughs etched in SiO2 with a ’’CHF3 beam’’ to a depth of approximately 1 μm are shown to be vertical and free of undercutting.

Ion Beam Analysis of Molecular Beam Epitaxy InAlAs / InGaAs Layer Structures

Abstract: In this paper the Rutherford backscattering technique combined with channeling is used to study the stoichiometry and perfection of molecular beam epitaxial layers of and . Corroborative evidence from SIMS and Auger profiling are used to supplement the RBS studies.

Etched Profile of Si by Ion-Bombardment-Enhanced Etching

Abstract: The basic characteristics of the ion-bombardment-enhanced etching (IBEE) of Si are investigated, using 60 keV-Ar as the bombarding ion, with regard to its application to submicron lithography. The damaged layer to be etched is determined by calculating the deposited energy distribution in the substrate and by using the threshold deposited energy. The etched profile of Si under a pattern mask is calculated using the value of 4×1023 eV/cm3 as the threshold. The calculated profile and the amount of side etching of Si are compared with the observed ones, and the agreement between them is found to be good, with an accuracy of 10 nm. The optimum dose for obtaining submicron patterns with sharp edges is found to be 0.8 –2×1015 Ar+/cm2 at 60 keV. The etched patterns made by a focused ion beam with Gaussian profile are also calculated.

Measurements of Energetic Particle Beams in a Plasma Focus

Abstract: Measurements of energetic particle beams in a plasma focus with a Mather-type device are presented. Two Faraday cups are used simultaneously to measure energetic particle beams in both the up- and down-stream directions with respect to the discharge current. In the upstream direction, only an electron beam was detected and no ion beam was observed. In the down-stream direction, an ion beam was detected first and then an electron beam followed 50–100 ns from the onset of the ion beam signal. It is shown that at a relatively high filling pressure of 4–6 Torr D2, the neutron yield increases when the electron beams are more intense in both directions.

Secondary ion mass spectrometry analysis of semiconductor layers

Abstract: The technique of Secondary Ion Mass Spectrometry (SIMS) is analyzed for its capabilities of monitoring layered structures. Primary ion beam current and secondary ion species are varied in order to minimize interface phenomena due to changes in ionization efficiences. It is noted that the choice of primary ion beam current can be particularly important when an oxygen bleed is used. Layer profiles are shown to become more complex when a secondary cluster ion is monitored instead of the monoatomic ion for species detection. With nitrogen as an example, quantitative analysis using cluster ions containing three or more atoms is shown to be extremely difficult, particularly when no oxygen bleed is used.

Multi‐aperture ion source with a deflectable focused beam for compositional control in sputter deposition

Abstract: Composition control in ion beam sputtering of thin films can be achieved by electrostatic deflection of a focused ion beam. The necessary adaptation of a multi‐aperture ion source consists in using a spherically dished screen grid and an accelerator grid built up from bent deflection plates. At a total accelerating voltage of 2 kV, a focused Ar+ beam of ∠10 mA is obtained, which can be deflected up to 11° with deflection voltages not exceeding 120 V.

Dual Electron Beam Processing System for Semiconductor Materials

Abstract: This paper describes a semiconductor processing system using two electron beams. One beam provides uniform heating of the bulk of the specimen, whilst the other performs localized heating, for example as a spot, a line beam, or as an inset scan. The annealing of localized regions in ion-implanted silicon and SOS, by local heating above the bulk temperature, illustrates the application of the system. Other applications of the systems are also considered.

The Cornell Electron Beam Ion Source

Abstract: An electron beam ion source (EBIS) for the production of low energy, multiply charged ion beams to be used in atomic physics experiments has been designed and constructed. An external high perveance electron gun is used to launch the electron beam into a conventional solenoid. Novel features of the design include a distributed sputter ion pump to create the ultrahigh vacuum environment in the ionization region of the source and microprocessor control of the axial trap voltage supplies.