Radiant energy – Irradiation of objects or material – Irradiation of semiconductor devices
Reexamination Certificate
2001-11-30
2004-12-14
Lee, John R. (Department: 2881)
Radiant energy
Irradiation of objects or material
Irradiation of semiconductor devices
C250S492200, C250S492230, C250S492300, C250S3960ML, C250S398000, C250S397000, C250S491100, C250S310000
Reexamination Certificate
active
06831282
ABSTRACT:
FIELD
This disclosure pertains generally to microlithography performed using a charged particle beam such as an electron beam or ion beam. Microlithography is a key technology used extensively in the fabrication of microelectronic devices such as integrated circuits, displays, thin-film magnetic-pickup heads, and micromachines. In charged-particle-beam (CPB) microlithography the pattern to be transferred to a substrate is typically defined on a segmented reticle, i.e., a reticle that is divided into multiple exposure units (termed “subfields”). Each subfield defines a respective portion of the pattern and is exposed individually. More specifically, the disclosure pertains to, inter alia, methods and apparatus for evaluating the beam blur in a CPB microlithography system.
BACKGROUND
Conventional charged-particle-beam (CPB) microlithography methods (typically using an electron beam) suffer from the disadvantage of low throughput (i.e., number of production units such as wafers that can be processed per unit time). Substantial research and development is being directed to improving the throughput so as to provide a CPB microlithography technology that is practical for mass-production of microelectronic devices. For example, electron-beam “direct-drawing” techniques currently are used mainly for manufacturing reticles, and these techniques can be used for forming patterns directly on the surface of a suitable substrate. However, with direct-drawing methods, the pattern is formed on the substrate line-by-line or feature-by-feature, which requires enormous amounts of time per wafer. Hence, there are severe limits to the maximum throughput currently achievable using conventional electron-beam direct-drawing methods.
Various approaches to CPB microlithography have been considered in an effort to increase throughput. CPB-microlithographic exposure of an entire pattern (defined on a reticle) to the substrate in one exposure “shot,” in the manner used in optical microlithography, would appear to be the best solution. However, this approach has to date been impossible for various reasons. One promising approach offering prospects of acceptable throughput is termed “divided-reticle” CPB microlithography, in which the pattern as defined on the reticle is divided into a large number of “subfields” each defining a respective portion of the pattern. Although each subfield defines only a portion of the pattern, the respective pattern portions are substantially larger than single features. Consequently, throughput is higher than with direct-drawing methods (but not as high as would be obtained if the entire reticle were exposed in one shot).
With the divided-reticle approach, achieving higher throughput principally involves configuring the subfields as large as possible so that the pattern portion projected with each respective shot is as large as possible. Increasing subfield size requires corresponding increases in the numerical aperture of the CPB optical system, which correspondingly increases the difficulty of correcting off-axis aberrations and beam blur. Hence, divided-reticle microlithography systems currently under development include respective subsystems for measuring aberrations and beam blur and for performing adjustments of the beam as required to correct the aberrations and beam blur. These adjustments typically include corrections of focal point, astigmatism, magnification, image rotation, and other parameters that impact the imaging performance of the system.
FIG. 13
is an oblique view schematically illustrating a conventional method for measuring beam blur;
FIG. 14
depicts certain details of the method in schematic block form; and
FIG. 15
is a plot of exemplary measurement results obtained using the method. This method for measuring beam blur is disclosed, for example, in Japan Kôkai Patent Document No. Hei 10-289851 (corresponding to U.S. Pat. No. 6,059,981) and in Japan Kôkai Patent Document No. 2001-203149.
Referring first to
FIG. 13
, it will be understood that an illumination-beam source and a reticle, although not shown, are located upstream of the components shown in the figure. The reticle, in addition to defining the pattern to be transferred to the substrate, also defines a pattern of measurement marks. The beamlet EB depicted in
FIG. 13
is the small electron beam produced by transmission of the illumination beam through a measurement mark on the reticle. Hence, the beamlet EB that has passed through the measurement mark has a rectangular transverse profile. The measurement mark typically is rectangular in profile. The beamlet EB is incident on a plate
100
that defines a “knife-edged” reference mark
102
. The reference mark
102
typically is rectangular in profile (the entire mark is not shown) and is configured as a respective through-hole defined by the plate
100
(see FIG.
14
). The beamlet EB is incident in a scanning manner on a knife-edge
101
of the mark
102
, wherein the knife-edge
101
is configured as in the well-known “knife-edge” test for evaluating the quality of an aerial image. An electron detector (sensor)
105
is disposed downstream of the mark
102
.
As the beamlet EB is scanned in a direction indicated by a respective arrow (labeled “SCAN”; e.g., extending to the right in FIGS.
13
and
14
), electrons of the beamlet EB incident on the plate
100
itself either are absorbed (if the plate
100
has sufficient thickness) or transmitted with forward-scattering (if the plate
100
is sufficiently thin). On the other hand, electrons incident on the reference mark
102
are transmitted through the reference mark and are detected by the electron detector
105
. As noted above, if the plate
100
is sufficiently thin (e.g., made of silicon having a thickness of 2 &mgr;m), nearly all the electrons incident on the plate
100
are transmitted with forward-scattering through the plate. (Conventionally, configuring the plate
100
sufficiently thin so as to cause forward-scattering of incident electrons is preferred. Also, a thinner plate provides a greater geometric accuracy of the knife-edged reference mark
102
.) In the following discussion, it is assumed that the plate
100
is sufficiently thin to cause forward-scattering.
In view of the above, the electrons detected by the electron detector
105
include non-scattered electrons e
1
(that have passed directly through the reference mark
102
) and electrons e
2
(that were forward-scattered during transmission through the plate
100
). The respective beam currents for the electrons e
1
and e
2
, as detected by the detector
105
, are amplified by a pre-amplifier
106
, converted by a differentiation circuit
107
(wherein the conversion is to percent change versus time) to an output waveform, and displayed on an oscilloscope
108
or analogous display. Beam blur is determined from the output waveform produced by the differentiation circuit
107
. From the determined beam blur, appropriate corrective adjustments (e.g., of focal point, astigmatism, magnification, rotation, etc.) are made to the beam. After making the corrective adjustments, another beam-blur measurement can be made to ascertain whether the corrective adjustments were appropriate.
In this conventional beam-blur measurement method, contrast of the “image” detected by the detector
105
is a function of the difference in electron scattering produced by the reference mark
102
versus the electron scattering produced by the plate
100
. Unfortunately, most of the forward-scattered electrons e
2
propagate to the detector
105
. These detected forward-scattered electrons e
2
reduce measurement contrast. Specifically, the forward-scattered electrons e
2
are a source of measurement noise, as shown in FIG.
15
. The noise produces an actual detected-current waveform W′ that is offset from an ideal detected-current waveform W (based at the “0” level). Also, leading edges of the detected waveform have a gradual slope, which results in decreased measurement accuracy.
Modern divided-reticle CPB microlithography apparatus are
El-Shammaa Mary
Klarquist & Sparkman, LLP
Lee John R.
Nikon Corporation
LandOfFree
Methods and devices for evaluating beam blur in a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Methods and devices for evaluating beam blur in a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Methods and devices for evaluating beam blur in a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3316772