Radiation imagery chemistry: process – composition – or product th – Imaging affecting physical property of radiation sensitive... – Making electrical device
Reexamination Certificate
2000-10-13
2003-07-15
Huff, Mark F. (Department: 1756)
Radiation imagery chemistry: process, composition, or product th
Imaging affecting physical property of radiation sensitive...
Making electrical device
C430S296000, C430S312000, C430S394000, C430S396000
Reexamination Certificate
active
06593064
ABSTRACT:
FIELD OF THE INVENTION
The invention relates to imaging of etch resistant layers, also known as “resists”, in order to fabricate high resolution patterns by etching or deposition. In particular, the invention relates to the manufacturing of integrated circuits and flat panel displays and the like using an optical stepper. A stepper is an imaging device used in the semiconductor industry to project an image of a mask onto a semiconductor wafer.
BACKGROUND OF THE INVENTION
Most integrated circuits today are fabricated using selective etching or deposition according to a master pattern known as a mask, using an imaging device known as a stepper. The process of fabricating high resolution patterns, mainly on planar objects, by selective etching or deposition has been well known for centuries. In general, the layer to be shaped or patterned is covered by a protective layer known as a “resist”. The desired shapes are created in the protective layer, usually via photo-imaging. The exposed (or unexposed, if the resist is negative working) part of the image is removed, normally by using a liquid developer to expose the layer underneath. The exposed layer can now be etched through the openings in the resist layer, which protects the covered area from the etching process. Etching can be by wet chemicals or by dry plasma (a process widely used in the semiconductor industry).
Instead of etching an additive process can be used. In an additive process a material is deposited through the openings in a resist to add to the layer underneath the resist. This deposition can be done in a wet process (as in the well known “additive” process for manufacturing printed circuit boards) or in a dry process, such as a vacuum deposition by evaporation or sputtering. Another way of using a resist is in allowing chemical reactions, such as oxidation, to occur only in the areas not covered by the resist.
In general, a resist is an imagewise mask selectively controlling a chemical or physical process and limiting the process to follow the image pattern. The term “resist” should be interpreted in this broad sense throughout this disclosure and claims. Any other layer which has suitable properties and can be patterned by light or heat can be used as a resist. At the end of the process the remaining resist is normally removed, or “stripped”.
Historically most resists were photoresists, i.e. activated and imaged by the photonic action of light. Because of this photonic action most photoresists operate in the UV part of the spectrum, where the photon energy is high. Some resists are exposed by other types of radiation, such as electron-beams. All photoresists and electron beam resists share one fundamental property: they respond to the total exposure, not to the momentary illumination. In optics, exposure is defined as the integral of illumination over time. For example, a photoresist can be exposed by 100 mW/cm
2
for 1 sec to yield an exposure of 100 mJ/cm
2
(100 mw×1 sec) or it can be exposed by 1000 mW for 0.1 sec (100 mW×0.1 sec=100 mW/cm
2
) with similar results. This law is also known as the “reciprocity law” and it is the basic law governing the exposure of photoresists.
When a certain exposure is reached, a change occurs in the resist. Since exposure is a linear function of power and time, the principles of linear superposition apply. The most common resists operate by a change of solubility in a developer.
The law of reciprocity also requires a high contrast ratio and low stray light in optical systems used to expose photoresists and electron beam resists. For example, if an exposure system has a light leakage, or stray light, of 1% (e.g.: when exposure is “off”, the light level does not drop to zero but only to 1% of the “on” state) the effect of this stray light can be as large (or larger) than the main exposure if left on the photoresist for a long time.
An even larger problem is caused when trying to image high resolution features: the point spread function of the optical system causes a “spreading” of light from each feature. This causes light from one feature to overlap with adjacent features and lowers the resolution. This problem is most severe in the semiconductor industry when using steppers to image a semiconductor wafer, typically a silicon wafer.
The basic elements of a stepper are shown in
FIG. 1
, where a mask
1
, containing a pattern which has to be copied onto silicon wafer
4
is illuminated by light source
2
. Mask
1
is imaged with a reduction lens
3
to form image
5
, typically at a 5× reduction. Wafer
4
is stepped by x-y positioning system
6
and
7
and each area, known as a die, is exposed with the pattern of mask
1
. Typically wafer
4
is coated with photoresist before exposure, however in some cases a different layer which is capable of responding to light is used. Because of the extremely fine features (below
1
micron) in image
5
which are imaged on each die, lens
3
is not capable of fully resolving all detail without some distortion of features.
If a cross section of the image of the mask
1
is taken along line
8
it would look like graph
9
in FIG.
2
. If a cross section of the same image is taken at the surface of the wafer
4
along the line
8
′ (
FIG. 1
) it would look like graph
10
in FIG.
2
. For further details on microlithography in general, and operation of steppers in particular, any modern book on the subject can be consulted, such as: “Handbook of Microlithography, Micromachining and Microfabrication”, Volume 1 and 2, Edited by P. Rai-Choudhury, SPIE Press 1997, ISBN book number 0-8194-2378-5 (V.1) and 0-8194-2379-3 (V.2).
Referring now to
FIG. 2
, mask
1
is transmitting light in all areas not covered by an opaque layer. The image is made of pixels (picture elements) numbered from
1
to
16
in FIG.
2
. Graph
9
shows the light intensity distribution just under mask
1
. After passing through the lens, this light distribution is distorted by the limited resolution of the lens (item
3
in FIG.
1
). The resulting light distribution is shown in graph
10
. Photoresists are formulated to have a sharp threshold. Once the exposure level crosses the threshold a chemical change occurs. The change is normally a change in the solubility of the resist in a solvent. Because of this sharp threshold a sharp image can be produced in spite of the fact that graph
10
cannot fully reproduce the details of mask
1
. As long as all desired features cross the threshold an image will be formed. Line
11
represents the threshold.
FIG. 2
shows the use of a positive resist, which is washed away in all exposed areas. The same theory also applies to negative resists.
The resist is exposed in all areas of image
5
where a graph
10
crossed threshold
11
. In the exposed areas, the resist is washed away on the die. The features of image
5
are imaged sharply, however, their dimensions are distorted, as can be seen by comparing the pattern imaged onto wafer
4
to mask
1
in FIG.
2
. For reasons of clarity the pattern imaged onto wafer
4
and mask
1
are shown at the same size, while in most cases the pattern imaged onto wafer
4
is a reduced image. For the same reasons only a one dimensional section is shown, while the same effect is happening in the other dimension as well. All graphs are shown along the X axis (as defined in
FIG. 1
) but the identical situation also happens in the Y axis. Also, for clarity the fact that the image on wafer
4
may be inverted (depending on the optical system used) is not shown in the graphs.
Because both the optical system and the photoresist behave as linear systems (at least as far as accumulation of exposure is concerned) the principle of linear superposition will hold. This principle states that ƒ(a+b)=ƒ(a)+ƒ(b), or the response of the system to a function made up of multiple parts is equal to the sum of the responses of the system to each part when each part is applied separately. This principle is illustrated in FIG.
3
. Mask
1
can be separa
Creo Inc.
Huff Mark F.
Oyen Wiggs Green & Mutala
Sagar Kripa
LandOfFree
High resolution optical stepper does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with High resolution optical stepper, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and High resolution optical stepper will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3012307