Radiant energy – Irradiation of objects or material – Irradiation of semiconductor devices
Reexamination Certificate
2001-10-26
2004-04-13
Lee, John R. (Department: 2881)
Radiant energy
Irradiation of objects or material
Irradiation of semiconductor devices
C430S030000, C430S942000
Reexamination Certificate
active
06720565
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Technical Field
This invention relates to the general field of particle beam lithography and, more particularly, to real-time correction of the lithography process to compensate for proximity heating of the resist. The particles may be electrons, photons, ions or uncharged particles, but the discussion will refer particularly to electron beam lithography, which presently is the method most commonly used to manufacture lithographic masks.
2. Description of the Related Art
The production of precise patterns on surfaces is a necessary stage in the fabrication of integrated circuits, and finds applicability in many other commercial environments as well. The typical method for creating such patterns is to coat the surface to be patterned with a chemical that undergoes a chemical transformation upon exposure to energy, a ‘resist’. Positive resists undergo chemical transformation on exposure to energy leading to removal of resist from the surface in the regions so exposed. Negative resists undergo other chemical transformations, such as cross-linking, leading to removal of resist in regions not exposed to energy. After the resist in the appropriate pattern has been removed by exposure to the appropriate pattern of energy, the underlying surface may be subjected to further chemical etching or material deposition. Following surface etching or deposition, the remaining resist is removed.
The energy incident on the resist is typically either electromagnetic or a beam of particles, typically ions or electrons (‘e-beam’). The energy may be directed onto the resist in one of two general ways: 1) through a mask having both transparent and opaque regions therein permitting selective passage of the incident energy to create the desired pattern of exposure on the underlying resist, or 2) as a focused beam, guided so as to impact selectively only those areas requiring exposure.
Exposure through a mask is the most common technique presently used for producing numerous identical patterns at reduced costs. However, the mask itself must first be made, most commonly by focused beam impact. Thus, focused beam exposure of resists is a necessary step in the production of masks for lithography.
Direct beam “writing” of patterns, although generally much slower, has several advantages over printing from masks. Among these are avoiding the complications of alignment and registration of the mask, and more precise patterning accomplished by precisely focused beams.
The electronics industry is driving to reduce the feature size of components. Smaller components, which are desired because of their higher switching speeds and lower power consumption, place increasingly critical demands on pattern accuracy. Precise patterning requires precise exposure of resist: Sharp and accurately positioned boundaries are desired between exposed regions and unexposed regions for both types of resist, to permit the pattern designer to use more densely packed components without interference and overlap of imprecisely exposed adjacent pattern.
For concreteness of our description, we will consider the case of positive resists, which are removed from the underlying layer for subsequent etching or deposition where the positive resist is exposed to the incident e-beam. Completely analogous effects are present for negative resists as well understood in the art.
Precise exposure of resist requires a detailed understanding of the sensitivity of the resist to e-beam exposure. The exposure of resist to an e-beam, called the dose, is typically measured in microcoulombs per square centimeter (&mgr;C/cm
2
). The “sensitivity” of the resist means the electron dose (in &mgr;C/cm
2
) necessary to create the desired pattern in the resist upon development. This sensitivity is a function of the resist composition, the energy of the incident electron beam, the temperature of the resist, the resist development process and other factors. The changes of resist sensitivity with its temperature at the time writing occurs are a particular concern.
Changing the temperature of the resist changes its sensitivity, which may require changing the dose of electrons in order to achieve proper exposure. Failing to take into account changes in resist sensitivity with temperature may lead to overexposure of the resist, exposure of the resist in regions not intended to be exposed, and less precise patterns. As will be described in greater detail below, pattern “blooming” (illustrated more particularly in
FIG. 2
, described below) is the undesired result.
A particular concern is a phenomenon known as “medium range proximity heating.” In high voltage lithography, most of the electron beam energy passes through the resist and the underlying mask layer (typically very thin) and penetrates the substrate where most of it is deposited. (An exception occurs when thin substrates are used, typically in the manufacture of X-ray masks, where the substrate is itself a film so thin that most of the beam energy passes through it). Electron diffusion in a thick substrate deposits the heat from a single e-beam pulse or ‘flash’ in the substrate typically in a volume much larger in lateral extent (perpendicular to the e-beam direction) and also in depth into the substrate than the flash size. Subsequent thermal conduction transports a portion of this heat to the substrate surface where it heats the resist in a zone that may be tens of microns in lateral extent some microseconds following the flash, increasing to a millimeter across after many milliseconds. (Exact numbers will depend on beam energy, the composition of the substrate and its thermal properties). Such proximity heating depends on the previously written pattern and the time history of the pattern writing. It may result in temperature changes of tens of degrees. This variability makes proximity heating particularly challenging to estimate in designing a process for high accuracy e-beam writing
There are two other proximity heating effects, termed “nearby flash heating” and “global heating” which are not to be confused with this “medium range substrate heating.” These other effects may also require real time compensation but occur on very different time and distance scales and require different methods. Nearby flash heating can be compensated by simple table lookup of pre-computed temperatures. However, this lookup may have to occur very rapidly, e.g., on a nanosecond time scale. Global heating of the entire substrate may also require real-time prediction which cannot ignore the mask boundaries. It causes thermal expansion on a time scale of minutes and hours, even though the temperature rise is not large.
“Proximity heating” as used herein is not to be confused with “proximity effect” which is a term commonly used to describe the unwanted exposure of the resist by electrons backscattered from the substrate. Corrective measures are taken to compensate for this undesired exposure which, however, also is affected by heating of the resist.
FIGS. 1A and 1B
illustrate beams of low and high energy electrons, respectively, incident on a substrate. The general mode of operation of e-beam lithography makes use of a focused beam of electrons, accelerated through a voltage, typically 1000 volts (1 keV) and above. Lower voltage e-beams are more effective at exposing the resist. Higher voltage e-beams are preferred for their ability to be formed into more precisely focused beams, and with less scattering in the resist layer, resulting in more accurate lithography and the ability to fabricate smaller patterns. “High voltage” e-beams herein is commonly understood to mean e-beam energies above approximately 10 keV. Beam energies as high as 50-100 keV are used. However, high energy e-beams produce undesired heating side effects.
FIG. 1A
depicts schematically and in cross section a beam of low energy electrons (less than approximately 10 keV)
104
a
incident on a layer of resist
103
a
. Typically, resist layer
103
a
will be relatively thin, around 0.5 &mgr;m (“microns”=10
−6
meter). Resist
Babin Sergey
Innes Robert
Teitzel Robin
Veneklasen Lee
Veneklasen Mary
Applied Materials Inc.
Gill Erin-Michael
Lee John R.
Leitich Greg
Veneklasen Mary
LandOfFree
Real-time prediction of and correction of proximity resist... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Real-time prediction of and correction of proximity resist..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Real-time prediction of and correction of proximity resist... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3261746