Facsimile and static presentation processing – Static presentation processing – Dot matrix array
Reexamination Certificate
2000-03-28
2004-11-16
Poon, King Y. (Department: 2624)
Facsimile and static presentation processing
Static presentation processing
Dot matrix array
C358S001900, C358S003010, C358S003240, C347S253000, C347S254000
Reexamination Certificate
active
06819450
ABSTRACT:
BACKGROUND
1. Field of Invention
The invention relates to photolithography, and more particularly to gray scale photolithography used to define edges on microelectronic device patterns during integrated circuit fabrication.
2. Related Art
Photolithography is used in semiconductor integrated circuit fabrication. Direct writing photolithography is typically used to define reticle mask patterns subsequently employed during projected image photolithography printing on the integrated circuit wafer. In direct writing (two-step) lithography strategies, radiated energy, such as laser or electron beams, is directed in a controlled manner on to an energy sensitive layer (e.g., a photoresistive layer or “resist”) overlying a substrate. The energy beams expose a pattern in the energy sensitive layer which, depending on the composition of the energy sensitive layer, becomes more soluble (positive resist) or less soluble (negative resist) in a chemical developer solution. The amount of accumulated energy from one or more beams incident on a location determines the exposure amount.
The pattern to be printed is typically defined in an orthogonal array of picture elements (pixels) oriented in X (rows) and Y (columns). Thus each pixel in the array has a unique address location. A pattern is defined by exposing selected pixels in a single array, or by exposing selected pixels in two or more overlapping and/or interstitial arrays. The distance between two adjacent pixels in an array is the pixel pitch.
In some conventional printing (imaging or writing) strategies, each pixel receives either the maximum beam energy or no beam energy. In enhanced printing strategies each pixel may receive one of several intermediate (between zero and maximum) “gray level” beam energy levels (intensities or doses). In a typical gray level strategy, each gray level number represents a corresponding beam intensity that is linearly proportional to the gray level number. For example, in a gray level printing strategy using seventeen gray levels (0-16), gray level number 1 represents one-sixteenth of the maximum beam energy, gray level number 2 represents two-sixteenths of the maximum beam energy, etc. Thus the intensities corresponding to the gray level numbers linearly increase in proportion to the gray level numbers.
A typical beam energy cross section is approximately gaussian. Maximum energy per unit area occurs at the pixel center and decays with distance from the center. It is therefore helpful to consider a pixel as a particular target location rather than a defined area. To ensure proper energy sensitive layer exposure, pixels are positioned so that the energy each pixel receives from the energy beam overlaps other pixels. Thus the total accumulated energy a particular pixel receives is often the sum of beam energy directed at the particular pixel and beam energies directed at nearby pixels. When a lithographic tool exposes the energy sensitive layer by irradiating a pixel array, the edge of a printed geometric pattern is defined by the total energy received at each pixel.
FIG. 1
illustrates total energy received from an incident beam along a column of pixels
10
. The number adjacent each of the six pixels shown represents the gray level number in a strategy using, for example,
17
gray levels. Immediately above column
10
is a graph plotting energy incident at each pixel against position. As discussed above, individual energy curves
12
represent the gaussian energy cross section each of the column
10
pixels receives from the incident beam. Note that the curve representing gray level number 8 is one-half the maximum gray level number 16. No energy is incident on pixels with gray level number 0 (zero). Curve
14
is the sum of curves
12
and represents the total accumulated energy cross section received by the energy sensitive layer. Curve
14
is normalized to have 1.0 as double the intensity required to fully clear the energy sensitive layer.
Persons skilled in the art will understand that assigning various gray level numbers to pixels, and thus changing the associated energy dose directed to corresponding pixels, moves the edge position. In this way small incremental changes are made to edge positions when printing small geometric shapes. By one definition, an edge of an object to be printed occurs at 0.5 on the normalized scale described above. In this model, the slope is steepest at the 0.5 dose, thereby giving the smallest change in critical dimension (CD) or size of the feature per percent dose change. In actual practice, the maximum exposure energy used is often greater or less than the optimal dose to clear because the dose is used to make features larger or smaller. This action sacrifices CD uniformity, however, because systematic and random dose errors will generate larger CD errors than the optimal dose to clear. In
FIG. 1
, edge position
18
is shown corresponding to the position of the pixel assigned gray level number 8.
The effective grid is the grid on which pattern edges may be defined. Using multiple exposures (printing passes) and/or gray levels per pixel per exposure, the smallest effective grid of the pattern being printed that can be resolved is the pixel pitch divided by the product of the number of exposures and the number of non-zero intensity levels per exposure (assuming typically linear values for intensity levels are fixed for all exposures). On a multiple exposure printing system, such as an Etec Systems, Inc. ALTA 3500, successive exposures may have the center of the pixels shifted in X and Y. In such multiple exposure systems, the effective pixel pitch after one or more exposures would be the distance between the pixel centers. When the dominant distribution of energy per pixel spot size at half power is from 2 to 4 times wider than the effective pixel pitch, pattern edges can be defined at the desired effective grid positions.
FIG. 2
illustrates one gray level scheme similar to that used in an Etec Systems, Inc. CORE 2564 lithography tool in a single exposure, 16 non-zero gray levels per pixel mode to define the edge of an object to be printed. Alternatively,
FIG. 2
illustrates, for example, one of 8 exposures made on an ALTA 3000. Numbers shown adjacent each pixel are illustrative and represent the gray level number assigned to the pixel. During printing, an energy dose corresponding to the gray level number is then directed at the pixel. In column
2
A edge E
0
is centered on the row R
1
pixel that is assigned gray level number 8, and consequently receives 0.5 (eight-sixteenths) of the normalized maximum dose. By increasing the row R
1
gray level number by one, an action that correspondingly increases beam intensity, the edge is defined slightly higher at E
1
, as shown in column
2
B. The edge is moved in successive small increments by incrementing the row R
1
pixel gray level number until the maximum (level 16) is reached. At that point, the row R
2
pixel gray level numbers are incremented. Finally, as shown in column 2Q, edge E
16
is centered on the row R
2
pixel that is assigned gray level number 8, and is displaced by one row from the row
2
A position. Thus an edge is positioned between rows R
1
and R
2
in approximately 16 equal steps.
FIG. 3
illustrates another gray scale printing scheme used in the Etec Systems, Inc. ALTA family of lithography tools. The ALTA 3500 prints patterns using 16 non-zero gray levels per exposure, typically making 4 to 8 exposures to define a complete pattern. Each of the beam intensities associated with the 16 non-zero gray levels may be independently set in the ALTAs. The intensities are typically set in linear proportion to the gray levels. Between exposures the center of each pixel is shifted in X and Y.
FIG. 3
illustrates edge placement in one dimension and without coordinate shifting for clarity.
As shown, the scheme uses 16 gray level numbers but moves the edge by alternately incrementing row R
1
and row R
2
gray level numbers. This is implemented on the ALTA by printing row R
1
in one or more exposures and p
Hamaker H. Christopher
Jolley Matthew J.
Martyniuk Jerry
Applied Materials Inc.
Church Shirley L.
Poon King Y.
LandOfFree
Enhanced edge resolution and critical dimension linearity in... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Enhanced edge resolution and critical dimension linearity in..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Enhanced edge resolution and critical dimension linearity in... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3361811