Warren K. Waskiewicz

Comprehensive analysis of electron optical design of SCALPEL-HT/Alpha

This paper describes the design of the projection optics for the SCALPEL-HT/Alpha electron beam lithography tool. We first briefly review the main design requirements. We then describe the two main simulation software packages that have been used in the electron optical design -- (1) PROJECTION, for optimizing the aberrations to fifth-order, including the electron lenses, deflectors, stigmators and dynamic focus coils, and (2) BOERSCHA, for computing the combined effects of the aberrations and discrete and global Coulomb interactions. Recently developed key features include accurate simulation of blurring caused by plasmon losses in the SCALPEL mask, and quantitative assessment of the aberrations and distortions, by plotting through-focal series of point spread functions. A symmetric magnetic doublet lens is used, with a polepiece geometry that provides both low aberrations and telocentric (normal incidence) imaging at mask and wafer. Magnetic field clamps are used to improve the symmetry in the magnetic field of the doublet lens near the SCALPEl aperture. An image adjustment device (Waskotron) is used to permit small adjustments of the magnification and image rotation. Deflectors, stigmators and dynamic focus coils are used to dynamically minimize the aberrations and landing angles along the stripe scan. The deflector and stigmator coils are wound on stacks of ferrite rings inside each lens, to enhance their sensitivities. The coil winding distributions are described, and we discuss how many deflectors, stigmators and focus coils are needed, and their inductances and drive currents are computed. An electrostatic deflector provides for high- bandwidth correction of small electron optical and mechanical position errors between the mask and wafer stages. The overall performance of the projection optics is predicted. For 100 keV beam energy and 15 (mu) A beam current, with a 400 mm mask-to- wafer distance, and a 0.25 mm square stub-field scanned over a 3 mm stripe at the wafer, using a silicon nitride membrane mask, the predicted 40 - 60% rise distance is d4060 equals 23 nm, the predicted 30% - 70% rise distance is d3070 equals 47 nm and the predicted 20% - 80% rise distance is d2080 equals 78 nm. These computed values are obtained at the optimum aperture angle of 6 mrad at the plane of best focus, which lies 11 micrometer beyond the Gaussian image plane.

Monte Carlo study of high performance resists for SCALPEL nanolithography

Abstract The semiconductor community continues to push the limits of device dimensions by exploring new high-resolution lithography technology. As part of the SCALPEL lithography resist program, our goal is to be able to print sub-100 nm structures at doses that will permit high throughput, reduce wafer heating and still maintain good process latitude. Using 100 KV exposures on a SCALPEL tool, 100 nm structures were printed at exposure dose of 5.8 μC/cm 2 (and 80 nm isolated trenches at 5.4 μC/cm 2 ) in positive resists. In negative resists, isolated 100 nm were printed at a dose of 6.8 μC/cm 2 , and 80 nm structures at 7.2 μC/cm 2 were resolved as well. These results are well below the 10 μC/cm 2 minimum dose requirement for high throughput. Monte Carlo simulations were used as means to understand energy absorption mechanisms of these e-beam optimized resists, DUV and 193 nm resists. Atomic composition was found to factor in improved resist ionization. The resin (or low-Z elements) is found to account for more than 99% of ionization events during exposure.

Writing strategy for a high-throughput SCALPEL system

Successful deployment of SCALPEL for several post-optical production lithography generations requires a unique optimum writing-strategy. Since the electron optics sub-field and the strutted mask patten segment are both smaller than the final device image area, SCALPEL utilizes a stitching approach to image-formation. A dynamic sub-field placement scheme, or writing strategy, must provide precise 2D stitching at high speed, and eliminate mask strut images on the wafer. It should also provide the extended dynamic lens field necessary for good throughput, while minimizing all non-exposure times per wafer and maintaining the time- averaged current near the instantaneous space-charge limit. The preferred writing-strategy replaces mechanical stage acceleration events with beam deflection wherever possible. The unique writing-strategy presented here also generates the required 2D seam-blending dose-profiles, which are vital to robust CD control with stitching.

Electron-optics method for high-throughput in a SCALPEL system: preliminary analysis

Abstract A likely technology to supplant optical tools for the manufacturing of sub-0.13 μm design rule ICs is one based upon SCALPEL ® (SCattering with Angular Limitation Projection Electron-beam Lithography). One serious barrier to the acceptance of any lithographic technique by the IC manufacturing community is an inability to provide economically viable wafer throughput levels. Using a simple, parametric, time-utilization model of a step-and-scan writing strategy, we have identified the areas of greatest influence on throughput in a SCALPEL system. Though issues such as stage speed, resist sensitivity, and space charge-limited beam current do constrain the problem, we have found that the effective size of the printing field is the most sensitive parameter for realizing high throughput levels in SCALPEL. In this paper we present an electron-optical method for attaining high-throughput in a SCALPEL-based exposure tool. Starting with a moderately large area beam (1 mm × 1 mm) at the mask plane and simple, telecentric reduction (4x) optics, we have investigated increasing the effective printed field size through a combination of beam deflections, image stitching, and dynamic corrections. A preliminary analysis of recent modeling results indicates that a 3 mm × 3 mm effective field size at the wafer can be achieved while maintaining beam blur within manageable limits. The extensibility of this electron-optical approach to a production-worthy level of wafer throughput is presented, including the potential impact on other system parameters.

Design of a low-brightness highly uniform source for projection electron-beam lithography (SCALPEL)

The requirements for a projection electron-beam lithography source, such as one suitable for a system based upon the SCAL- PELR (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.

SCALPEL proof-of-concept system: preliminary lithography results

We have designed, constructed, and are now performing experiments with a proof-of-concept projection electron-beam lithography system based upon the SCALPELR (scattering with angular limitation projection electron-beam lithography) principle. This initial design has enabled us to demonstrate the feasibility of not only the electron optics, but also the scattering mask and resist platform. In this paper we report on some preliminary results which indicate the lithographic potential and benefits of this technology for the production of sub-0.18 micrometer features.

Quantitative aberration assessment by a through focal analysis of pattern edge sharpness

In computing the optical properties of electron and ion beam columns, the actual beam blur needs to be obtained from the computed aberration coefficients of the column and the beam parameters. A traditional method, which has successfully been used for many years, computes aberration disks for each individual aberration and obtains overall beam blur by adding these disks in quadrature. However, this method gives no information of beam current densities to compare with experimental measurements. A study of a new simulation method for analyzing pattern edge sharpness is described in this paper. The method involves the simulation of the point spread function (PSF), which can be proven to be equivalent to the pattern edge sharpness, provided that the PSF is smaller than the pattern feature. This method provides the current density distribution of the PSF and a quantitative assessment of the aberration, defined in terms of the rise distance of the PSF, in a through-focal series of planes, thereby enabling the best focus plane to be determined. An illustrative example is presented for a typical SCALPELTM column.

CMOS compatible alignment marks for the SCALPEL proof of lithography tool

SCALPEL alignment marks have been fabricated in a SiO 2 /WSi 2 structure using SCALPEL lithography and plasma processing. The positions of the marks were detected through e-beam resist in the SCALPEL proof of lithography (SPOL) tool by scanning the image of the corresponding mask mark over the wafer mark and detecting the backscattered electron signal. Single scans of line space patterns yielded mark positions that were repeatable within 30 nm 3σ with a dose of 0.4 μC/cm 2 and signal-to-noise of 16 dB. An analysis shows that the measured repeatability is consistent with a random noise limited response. The mark detection repeatability limit, that can be attributed to SPOL machine factors, was measured to be 20 nm 3σ. By using a digitally sequenced mark pattern, the capture range of the mark detection was increased to 13 μm while maintaining 36 nm 3σ precision. The SPOL machine mark detection results are very promising considering that they were measured under electron optical conditions that were not optimized.

Critical issues for developing a high-throughput SCALPEL system for sub-0.18-um lithography generations

The potential for SCALPEL to provide economically viable production lithography capabilities for post-optical generations depends largely on achieving adequate wafer throughput. We have analyzed throughput-limiting performance attributes of the SCALPEL approach in order to identify critical design issues and develop a process for evaluating its unique parameter space. An important feature of the SCALPEL approach is that small image sub-fields are assembled to form complete device patterns. Further, electron-electron interactions result in a throughput- dependent image blur, which is a governing parameter for many inter-related performance areas of SCALPEL. Error budgets for key issues affecting critical dimension (CD) have been developed to analyze this unique design space, using models of the image-forming process including stitching on sub-field seams. These budgets assist in identifying the most critical design issues and demonstrating their inter-relationships and tradeoffs.

Global space charge effect in SCALPEL

The global space charge (SC) effect in SCALPEL electron beam lithography system is investigated. First order properties of the SC lensing action (defocus and magnification change) in SCALPEL type projection systems are analyzed using a simple analytical technique. Aberrations induced by the lenses and SC in the projection optics are evaluated numerically using a Monte Carlo code developed to calculate the combined effect of Coulomb interactions and lens aberrations in the charge particle projection systems. We found that the defocus and the magnification change induced by SC are functions of two parameters, the beam perveance and the SCALPEL aperture size, that are critical for the system performance. The strong correlation identified between the best focus plane location and the aberrations induced by SC indicates that the SC lensing action can be effectively compensated by simply adjusting either the wafer plane position or excitations of projection lenses.