The requirements for a projection electron-beam lithography source, such as one suitable for a system based upon the SCAL- PEL^R (scattering with angular limitation projection electron-beam lithography) technique, are significantly different from those of a conventional TEM, SEM, or direct write type of instrument. While high resolution imaging is still a primary goal, this must now be achieved at relatively high (1-200 (mu) A) beam currents in order to realize economically meaningful wafer throughput. Space-charge limitations considered over the entire system (not just the electron gun) lead to the use of relatively large illumination angles (approximately 0.5 mrad). Taken together with an illuminated mask area of approximately 1 mm², this means that the electron gun axial brightness needs to be only 10² to 10⁴ Acm^-2sr^-1, as compared with a value of 10⁶ to 10⁹ Acm^-2sr^-1 for a TEM. Similar considerations indicate that the source emittance must exceed 700 micrometer(DOT)mrad, which is more than an order of magnitude larger than that provided for a standard focused-beam system. Additionally, the uniformity of the illumination must be within 2% in order to ensure that the variation in printed feature size across the imaged area remains negligible. This type of source performance must be stable for extended periods of time in order to maximize the uptime of the lithography tool. In this paper we review the source built for our SCALPEL proof-of-concept system, discuss the impact of an interim modification, and then examine the potential of a further source redesign.

Accurate computer simulations of electron guns based on finite element methods for electrostatic field calculation are described. Discrete and semi analytic models for thermionic cathode emission currents are presented, the initial conditions used for the electrons at the cathode are discussed and the affects of different approximations are analyzed for an axisymmetric triode electron gun. Results from models that use local one dimensional approximations are compared to a full discrete simulation.

The design of multi-component optical systems can be supported by software tools. Designers are known to change often between different levels of abstraction, sometimes characterizing a lens only by its focal length and at other times giving a full description of the magnetic circuit. A design tool should be able to follow these changes in approach. We illustrate the use of such a tool by describing the design process of an optical system. The example shows the use of different abstraction levels and the program's ability to define constraints such as staying in focus and at a set magnification while optimizing the system's parameters. The optical system which is chosen for the example must image an object to an image plane with variable magnification and variable rotation while obeying certain requirements on resolution and distortion.

Two in-lens deflector arrangements integrated into a conical tip immersion objective lens are presented in this paper. One is of a saddle type placed on the inner-surface of the upper pole-piece close to the tip. Another is of a toroidal type integrated on the immersion objective lens by winding its coils around the upper pole-piece tip through small wire- passing holes. A nearly perfect moving objective lens (MOL) condition can be achieved with the saddle arrangement. The on- pole piece-tip toroidal deflector, although operating far from an ideal MOL condition, can also provide a wide field of view if certain parameters, such as the excitation current and the orientation angle, are chosen properly. At the optimum operation conditions, vertical landing of the probe can also be maintained.

For high-throughput electron beam lithography, projection systems using symmetric magnetic doublet lenses can produce images with zero distortion. However, the projected pattern area is limited by beam blur at large off-axis distances. If an off-axis shaped beam pattern is imaged in a projection system, the aberrations can be greatly reduced by introducing deflectors, which steer the beam through the projection lenses in a modified path. In this paper, the principle of this type of projection with in-lens deflectors is first outlined. The method for computing the optical properties of such systems, based on an extension of our previously published unified aberration theory, is then described. To provide accurate simulation of systems with such large field sizes, our new software computes both the third and fifth-order aberrations. The computation of dynamic corrections, which can not only correct deflection field curvature and astigmatism but also reduce stitching errors, is also described. A design example of an off-axis shaped beam projection system with deflectors is presented, which has been optimized by the damped least squares method. The results show that such systems can have extremely small beam blur, distortion and stitching errors. The presented design images a 0.25 mm square shot over a 3 mm square region of the wafer, with 2 mrad beam half-angle, with a beam blur less than 26 nm, and distortions and stitching errors less than 19 nm.

A new micro deflection scheme has been proposed for pixel addressing in color field emission display devices. With a pair of in-plane micro deflectors, three color pixels in an image call can be illuminated by a single field emitter array instead of three field emitter arrays in the conventional pixel addressing scheme. The feasibility of micro deflection concept is confirmed by 3-D computer simulation using a charge density method. The new scheme offers four times higher display resolution than conventional pixel addressing scheme without losing image brightness. The fabrication of micro deflectors and driving circuitry are comparable to the conventional addressing scheme.

This paper presents a structured mesh refinement technique for the finite element method in electron optics. The technique is based upon using equipotential and flux lines on a trial potential distribution to generate the refinement mesh. For rotationally symmetric electron lenses, the method provides greater continuity of the axial field distribution derivatives than other mesh refinement techniques. For three dimensional electrostatic structures, the results show that the structured mesh refinement method can be used to improve the accuracy of field computations at points close to curved metal boundaries.

In this paper we propose an improved 3-D boundary charge method for high accuracy calculation of the electric field distribution, where differentiation of the coefficient matrix element (equals the potential coefficient) with respect to each coordinate component is needed. The differentiated potential coefficient, which is called the field coefficient, is expressed as a double integral. We have found that the first integral of the field coefficient can be done analytically in much the same ways as that of the potential coefficient, thereby greatly improving the computation time of the electric field distribution without any loss of accuracy. As a practical application of the method to field analysis, we have treated the misaligned diode system of a field emission gun.

The Matrizant method for the synthesis of magnetic and electrostatic focusing systems, producing an optimal beam quality (a minimum spot size for a given beam current), is proposed. This method includes: the matrix method for the description of motion of a charged particle beam in the nonlinear approximation on the base of a general relativistic theory of charged particle beam motion along a curved optical axis, developed by one of the authors; the analytical model of the axial field or their derivatives; the method of the moments of the particle distribution function over the whole totality of the phase coordinates for finding the averaged radius of the beam; and the integral equation method to solve Laplace's equation for obtaining the parameters of the physical model which has the same axial field as the mathematical model. As an example, the developed Matrizant method is applied to minimize the beam spot size for a given beam emittance in a focusing system consisting of multiple cylinder lenses with equal diameter.

This paper describes an all-electrostatic electron-beam column which uses conventional electrode assemblies, resulting in a system which is very much smaller than a magnetic lens column. Computer modeling of the electron-optical performance of the column has shown that the concept has promise as a lithography tool. Beam broadening due to electron-electron interactions and aberrations of the final accelerating lens are both small enough to give a resolution which matches that of present-day magnetic columns.

We present a new resist and exposure strategy applicable to the fabrication of sub-micron gate-length heterostructure field-effect transistors (HFET) with T-shaped (mushroom) gate contacts. Using selected polymethylmethacrylate (PMMA) resists as well as copolymers (polymethylmethacrylate/methacrylic acid: PMMA/MAA) and by optimization of the layer thicknesses we have established an electron-beam lithography process for fabrication of 0.1 micrometer mushroom-gates. Main advantages of this new concept are the necessity of low accelerating voltages of only less than or equal to 10 kV as well as an adapted thickness of the resist stack which, furthermore, guarantees a large cross-sectional area and hence, low contact resistance. Additionally, an excellent lift-off behavior is obtained. Due to the low accelerating voltage any standard scanning electron microscope can be applied for sub-micron mushroom-gate lithography which drops costs drastically. The complete fabrication process including gate-recess etching shows an excellent reproducibility which guarantees good process control and high yield. The achieved results are comparable with well established T-gate process, thus this new concept should be directly applicable in standard process lines. It should be pointed out that all exposure parameters become almost independent of the substrate property, because the dissipation volume is completely located within the resist layer stack. Thus, the contribution of backscattered electrons to the total dose is negligible.

The testing of substrates for opens and shorts, whether multichip modules or printed circuit boards, is extremely important prior to their population with expensive active devices. As a result of the increase in the number and density of test points and the decrease of the pad sizes, conventional testing techniques are no longer possible. We describe an electron beam substrate tester which can address this problem. The method of testing with e-beams, the system itself and some of the associated technical problems which must be addressed are discussed in this article.

A table top SEM design has been presented which is based upon the use of permanent magnets. The SEM's height can be designed to be below 12 cm. Computer simulations predict that the SEM should be able to provide high spatial resolution and operate at primary beam voltages above 100 kV for thin specimens.

It is reviewed how transfer maps of arbitrary order can be calculated in a convenient way using differential algebraic (DA) techniques. Such high-order maps are of use for high- precision optical systems, as well as for the case of repetitive particle accelerators, where nonlinear effects often have a tendency to build up over many turns. It is shown how the methods can be extended to also allow a rigorous treatment of the Lagrange remainder term of the Taylor expansion. The methods can be applied for the estimation of the effects going beyond the aberration order; it is also for the study of the repetitive motion where, leaning on ideas of Lyapunov, Nekhoroshev and others, it for the first time allows a rigorous determination of guaranteed particle survival times.

Computational methods and software have been developed for the analysis and design of magnetic imaging energy filters for transmission electron microscopes. The filters considered are of the omega and alpha types, in which the electron beam is steered through three or four bending magnets, to produce a stigmatic image with energy dispersion. An energy selection slit, located at a plane conjugate with the diffraction plane, then selects an image formed by electrons of a specific energy. After briefly reviewing the theory of such filters, the initial design procedure, using the analytic SCOFF ('sharp cut-off fringing field') approximation, is described. The computer simulation of real imaging energy filters, with numerically computed fringing fields, is then described, for magnets with planar or curved polefaces. A damped-least- squares optimization procedure is used to adjust the strengths of the magnets and the positions, tilts and curvatures of the pole-faces in order to focus the beam and minimize the aberrations. External quadrupole and hexapole elements can also be introduced to improve the performance of the filter. The software can also compute the properties of filters employing conical polepieces, such as those used in the MANDOLINE filter. The computation procedure and features of the software are described and illustrated with typical examples.

Different constructions of practical electrostatic lenses are numerically investigated. The electric field in electrostatic quadrupole lenses consisting of four copper rods depends, for a given lens aperture, on the value of the rod diameter. Furthermore, the aberration coefficients of electrostatic systems depend on the axial partial derivative of the electric field which, for example, influences the beam spot size, i.e., the microprobe resolution. We apply an integral formulation to the given set of polarized conductors in vacuum and in absence of space charge. These equations are numerically solved using an accurate version of the boundary element method in order to obtain the potential and the field distributions. The axial distributions of the electric field derivatives are found for different diameters of the rods. The developed Matrizant method is then applied to minimize the beam spot size -- for a fixed beam current -- in focusing systems consisting of electrostatic quadrupole lenses, using the computed axial distributions.

Focused ion beam (FIB) technology is widely used for micromachining, which includes removal from and addition of material to a substrate with sub-micrometer beams. In order to maximize the rate at which micromachining is done it is necessary to use the highest feasible beam currents, which results in the ion beams being severely affected by spherical aberration. We have investigated the possibility of reducing the spherical aberration of a FIB focusing column by means of space charge, by introducing a negative space charge cloud into an einzel lens.

The precise understanding of the properties of large acceptance devices like modern nuclear spectrographs often requires the calculation of high-order aberrations. In many practical cases, it is necessary to treat all details of the fields of the elements, and instead of utilizing more or less simple field models, one has to rely on measured data. It is shown how aberrations of in principle unlimited orders can be obtained from measured field data; moreover, for the remaining aberrations of yet higher order, rigorous upper bounds of their influence on the motion can be found. The methods are used for the analysis and correction of the high-resolution S800 spectrograph at MSU.

Jon Projection Lithography is depended on successful realization of system project. Very critical parameter of optical image limiting permissible work field is a distortion. A peculiarity of distortion is in exceptional significance of a fiveorder aberration in optimum state oftuning, when there is a circumference of zero-distortion. This state very sensitive to any perturbations. It has worked out a conception of design taking account the main possible sources of perturbation and means of its restrictions. This means includes above all a correction by multipoles and next now the tolerances considering a presence of multipoles. Important role in the process is assigned to an ifiumination part of optical system. Keywords: Ion Projection, Distortion, Tolerances, Integral equations.

The algorithms of numerical simulation of multi dimensional electromagnetic fields and PIC models for particles have been described. A new approach to generation a boundary-fitted curvilinear mesh based on five-parameters representation of mapping the complex-shape physical domain to a set of rectangular blocks is used. This algorithm permits us to construct a fast and universal mesh generator with flexible management with mesh parameters, condensing the mesh knots near edges and in subdomain with high gradients of solution. This way is able to reduce the total number of mesh knots in many times in compare with using a regular rectangular mesh. The codes with Poisson and Maxwell solvers have been implemented, those use as finite-difference as boundary- element methods. The results of this technique are demonstrated for simulation 2D and 3D problems.

A general mesh formula is successively derived for the discretization of self-adjoint elliptic partial differential equations in irregular two-dimensional meshes. This comprises the most important cases of electron optical field calculation by means of the finite differences method.

An analytical potential model is presented that allows a fast approximate calculation of the imaging properties of einzel lenses. The axial potential of these three-electrode unipotential lenses is approximated by a linear superposition of the potentials of apertures with a circular opening. With adequately defined coefficients of the linear superposition, it is possible to approximate the axial potential of symmetric as well as asymmetric einzel lenses with 'thick' central electrodes. In the optimization process of a transfer optics consisting of five einzel lenses, a comparison with the results obtained by means of the charge density method has proven, that this rather simple potential model is sufficiently accurate to allow a reliable pre-selection of promising lens designs.

The paper deals with the problems that appear in the electron- beam processing of insulators and consist in distortion of the surface pattern created by an electron beam. There are presented the results of theoretical and experimental studies on the phenomena in insulators, which are powered by the electron-beam processing of their surfaces, and the effect of these phenomena on the precision of the processing.

Magnetic rather than electric fields are usually used to deflect charged particle streams into large angles primarily because electrostatic deflection aberrations are 2 - 3 times larger. This very mature subject has been reexamined using a ray trace program. Traditionally, beams are injected centered between deflection plates to avoid the fringe fields and scanned symmetrically. Analysis produced a simple and surprising result. Injecting the beam asymmetrically significantly reduces aberrations. This very far-off-axis solution is only effective if the beam is deflected toward the near plate. Electrostatic deflection aberrations can be reduced over 10-fold by (1) injecting the beam into the deflection plate gap with a specific off-center displacement located at the inflection point of the beam landing at the target versus injection offset curve, (2) asymmetrical scan, and (3) quadrupole upstream. These results have been partially confirmed. One application has been studied -- a 4000 by 5000 pixel CRT for digital mammography workstations. By dynamically adjusting the injection offset, the beam can be scanned 61.0 degrees with undetectable (greater than 13 fold reduction) deflection aberrations. With static offset, (offset 42% toward the attracting plate) the beam can be scanned 38.1 degrees toward and 9.5 degrees away from the near plate. Multiple discrete beams are possible.

The paper is devoted to software development for simulation, optimization, and computer-aided design of photo/thermo- emission electron optical systems and units. The first part of the paper presents the applied program package (APP) 'ELIM\DYNAMICS\ intended for computer-aided design of dynamic photo-emission image tubes with electro/magnetostatic focusing and deflection (streak tubes). The developed software allows highly precise computation of basic image quality characteristics both in static and streak modes. One of the main advantages of the new program version presented is that 'through' electron beam computation from the photocathode to image receiver is available with regard to dynamic aberrations caused by scattering fields located nearby the edges of deflecting plates. In the second part, the possibility is shown to generalize some numerical techniques being effectively applied in photo-emission imaging electron optics (namely, the (tau) -variation -- and the first kind integral equations techniques) to simulation of the thermo-emission electron beam technology units. Functions of the new APP 'CHARGE' are presented, and some numerical aspects of the self-coordinated problem are discussed.