Lithography system and method with mask image enlargement

BACKGROUND OF THE INVENTION

In conventional lithography systems, patterns are formed on a substrate such as a semiconductor wafer or flat panel display by projecting light through a patterned mask onto the surface of the substrate which is coated with a thin film of photoresist. After the photoresist is thus exposed by the patterned light, it is developed to form the mask pattern on the surface of the substrate in the form of a pattern of hardened photoresist. The wafer or panel is then subjected to a particular desired microfabrication step such as chemical vapor deposition, ion implantation, oxidation, metal deposition, etc.

In the semiconductor and flat panel display industries, it is highly desirable to fabricate very complex circuit patterns on very small surface areas. Hence, lithography systems which can produce very small feature sizes are in great demand. To achieve small feature sizes, conventional lithography systems include an optical projection subsystem located between the mask and the substrate surface which demagnifies or reduces the image of the mask and projects the reduced image onto the surface of the substrate. By reducing the image of the mask, very small feature sizes, or, equivalently, very high resolution, can be achieved, thus enabling the system to achieve very high circuit densities.

As circuit complexities and panel sizes have increased and feature sizes have decreased, it has become commonplace to use several masks to produce a pattern on the substrate surface, with each mask forming the pattern for a single particular portion of the overall pattern. In these systems, the entire pattern is formed by a series of separate individual exposure steps, each with a different mask in a different position. That is, in a first step, a mask is installed in the system, the substrate is located such that the reduced image of the mask will be projected onto the substrate where it is desired to produce the particular associated portion of the overall pattern, and the exposure is carried out. Next, the mask is replaced with a second mask used to produce a second portion of the overall pattern. The substrate is moved to a new position where the second portion of the pattern is desired, and the second exposure is completed. This process continues until the entire pattern is exposed by all of the required masks.

There are several drawbacks to this “step-and-repeat” approach. First, the individual portions of the overall pattern must be precisely aligned with each other laterally along the surface. The process of aligning the individual portions, commonly referred to as “stitching”, can be extremely difficult, time-consuming and expensive. Features such as conductive lines can be extremely narrow, for example, on the order of 10 μm. Where such a line crosses the boundary between two portions of the pattern, it must be aligned so that continuity is not interrupted. It is a standard industry requirement that when portions of a pattern are stitched together, they must be aligned within 10% of feature widths. Hence, with a 10 μm line width, the features must be aligned within 1 μm. Such high accuracy alignment is very difficult to achieve.

Even if the individual portions can be accurately spatially stitched, other inaccuracies can also be introduced. For example, due to the sensitivity of commonly available photoresists, feature size is affected by exposure time and source luminance. Therefore, because of slight inaccuracies in the light source and/or optics of conventional systems which result in varying luminance and exposure time, the sizes of features can vary from portion to portion along the pattern. This can also degrade performance of the final circuit.

Thus, it will be appreciated that forming complex patterns on large substrates such as semiconductor wafers, multichip modules and flat panel displays can be very time-consuming and, hence, very expensive. Each wafer or panel requires several pattern layers, and each pattern layer can require several masks. The masks themselves are very expensive, as are the lithography and microfabrication processes which use them. As larger panels and/or wafers requiring higher resolution are demanded, these costs will continue to rise.

One alternative to the stitching approach has been used in some systems to produce relatively large approximately ten inch diagonal flat panel displays. In these systems, a special full-sized mask is produced with the pattern for an entire panel exposure layer. The mask is placed in close proximity to the flat panel with subaperture projection optics between them. The exposure is performed in either a line or area scanning procedure in which the exposure light is scanned over the panel in a line-by-line or in serpentine fashion until the entire pattern is exposed.

This approach also has several drawbacks. First, the special large mask is very expensive to produce. At the present time, most mask manufacturers are not equipped to produce such large masks, and those that will produce them charge very high premiums. Also, the scanning exposure process can be very time consuming, causing low panel throughput.

In other lithography applications, multiple small identical patterns can be formed on a substrate by repeated exposures, with each exposure creating one pattern or a small subset of the required patterns. These applications include small semiconductor circuits such as transistors and flat panel display pixel circuits. In either case, a typical substrate will have many such circuits formed in a repeating pattern over the entire substrate surface. Although stitching between individual patterns may not be required, this step-and-repeat process can also be very time-consuming and expensive especially where large numbers of patterns are exposed one at a time.

SUMMARY OF THE INVENTION

The present patent application describes multiple inventions in the field of lithography and substrate processing and microfabrication where the substrate can be a flat panel display, a semiconductor wafer, a multichip module, etc. One invention is directed to a lithography system and method which has very high resolution and therefore can produce very small feature sizes while producing relatively large patterns using relatively small preferably standard sized masks. The present invention is directed to a lithography system and method in which the image of the mask is enlarged, instead of reduced, as it is projected onto the surface of a substrate. The system includes a source such as a light source and illumination optics which illuminate the mask with light from the light source. Projection optics having one or more optical elements receive light transmitted through the mask and project an enlarged image of the mask onto a surface of the substrate. The projection optics achieve high resolution and other advantageous features of the invention by including at least one optical element which has a curved surface having a non-conic shape. The term conic as used herein refers to a conic of revolution, e.g., ellipsoid, hyperboloid, paraboloid and sphere.

In one preferred embodiment, the projection optics comprise a plurality of non-ellipsoidal powered mirrors. The mirrors can be a combination of elements having spherical, conic, aspheric or unusual shapes.

In one preferred embodiment, the projection optics include at least one device having an aspheric shaped surface described by a polynomial for a conic section with additional terms defining non-zero surface deformations from the conic section shape. Such surfaces can be either rotationally symmetric or non-rotationally symmetric, depending on the desired result. The optical device having such a surface can be either reflective or refractive and the projection optics can include more than one such device.

In another preferred embodiment, the projection optics include at least one device having a curved unusual surface which may be rotationally symmetric or non-rotationally symmetric, depending on the desired result. The device can be reflective or refractive, and the projection optics can include more than one such device.

The projection optics can be catadioptric, that is, they can include a combination of reflective and refractive optics, or they can be dioptric, that is, consisting of only reflective elements. One preferred catadioptric system includes plural reflective and refractive elements having surfaces with non-zero deformation from conic shape. A preferred dioptric system includes four mirrors, each having an unusual surface. One embodiment includes a combination of unusual surfaces and aspherically deformed conics.

In one embodiment, the projection optics include a distortion corrector for removing distortion from the image projected onto the mask. Also, the projection optics can preferably include an astigmatism corrector for eliminating astigmatism from the system.

The projection optics provide the system with a wide projection angle to permit high mask magnification and large mask image sizes at the surface of the substrate. The projection optics also preferably project the image telecentrically onto the substrate. Telecentricity helps eliminate distortion in the image. If the distance between the substrate and the projection optics should fluctuate, the telecentric nature of the optics prevents the magnification from introducing line separation distortion. Telecentricity also gives the system a larger depth of field. Fluctuations in the distance between the optics and the substrate are less likely to take the image out of focus.

The projection optics of the invention also provide the system with very high resolution, i.e., very small feature size and line width, while providing mask image enlargement. In one preferred embodiment, the system achieves a line width below 10 μm. One particular embodiment is used to produce a line width of 3 μm. This resolution makes the system of the invention applicable to both semiconductor fabrication and flat panel display fabrication technologies.

A typical flat panel active matrix liquid crystal display (LCD) for a lap top computer has a resolution of 640×480 pixels at a diagonal dimension of approximately 10.4 inches. At the present time, these panels are exposed in photolithographic processes by one of three standard techniques. In the first technique, each pattern layer is formed by exposing plural masks, one at a time at particular predetermined locations along the substrate, and the exposure images are stitched together. Typically, such systems include projection optics between the mask and the substrate which reduce the mask image to achieve the desired feature size and line width.

In another technique, a full-sized mask is generated with an entire layer pattern formed over the mask. In this technique, there is a one-to-one size correspondence between the mask and the panel pattern. Under this technique, the exposure can be carried out by one of two different processes. A line scanning process can be used in which a light source such as a laser beam is swept across the mask in a raster fashion side-to-side and top-to-bottom. In other systems, an area scanning process is used in which a light source and lens on opposite sides of the mask are simultaneously scanned in a serpentine fashion to sweep an exposure area over the entire mask.

In these one-to-one systems, the special large (10.4 inch diagonal) masks are extremely expensive and difficult to manufacture. In a typical display requiring 10 to 20 pattern layers, the cost of a mask set can be prohibitive. In the present invention, a standard 6 or 7 inch mask with the entire panel layer pattern formed thereon can be used. The patterned exposure light transmitted through the mask is enlarged to produce the final exposure image. Since standard readily available mask sizes can be used, a mask set for a panel is far less expensive than the special large one-to-one panels used in some systems. Also, with the image enlargement of the invention, the exposure is carried out all at once with a single brief exposure. No stitching or scanning is required.

In addition to fabricating standard size displays, the lithography system of the invention can also be used to create much larger displays such as flat panel television screens with much higher pixel resolution. The invention is applicable to fabricating high-definition television (HDTV) displays having standard pixel dimensions of 1125×882 pixels at a size of 650 mm×550 mm. Even with a standard sized mask, these pixel resolutions can be obtained at a feature size of less than 10 μm, preferably approximately 3 μm. The narrow line width permits the LCD driver circuits to be fabricated along the borders of the display panel. It will be understood that the pixel resolutions mentioned are only examples. Any number of pixels can be achieved depending upon the mask pattern.

The system of the invention can also be used in fabrication of very large low-resolution displays, such as message boards for highways and airports. Using a standard sized or larger mask and a high enlargement factor, the system can produce these larger displays with a single exposure per pattern layer in a fashion similar to that used for high-resolution applications described above.

Another invention described herein is directed to an adjustable or deformable substrate holder or chuck used to hold and shape relatively flat workpieces and a method of holding and shaping such workpieces. To achieve uniform high resolution over relatively large panel or wafer pattern sizes, the lithography system of the invention can include such an adjustable or deformable chuck. In one preferred embodiment, the chuck is controllable to be selectively deformable to flatten the panel such that the entire surface of the panel is flat to within the depth of field of the lithography system.

The deformable chuck of the invention includes a deformable top mounting plate upon which the substrate is mounted by means such as vacuum or electrostatic attraction. The deformable mounting plate is supported by a plurality of deformable supports or variable-length members distributed in a two-dimensional array along the bottom of the deformable mounting surface. In the case of vacuum hold-down, the deformable mounting plate includes a chamber accessible via a vacuum pump port. The chamber can be partially evacuated to create a negative pressure in the chamber. The top surface of the deformable mounting plate is porous or is machined with channels. When the substrate is placed on the top surface and the vacuum is activated, the negative pressure in the chamber holds the substrate to the surface.

The chamber preferably includes plural regions defined by walls within the chamber. The walls can be arranged to position the regions in a predetermined pattern to optimize the vacuum hold down function. For example, the walls can be arranged to form multiple concentric annular regions. When the vacuum is activated, the region at the center of the chamber will pump down first since it has the smallest volume. The regions will successively pump down radially from the center. This ordering results in good contact between the substrate and the top surface of the mounting plate.

Each of the supports is individually addressable and controllable by an electrical signal to change its length and, therefore, change the height of the portion of the deformable mounting plate it supports. The deformable support can be a piezoelectric transducer which changes length in relation to an applied voltage. Each support can also be a combination of a piezoelectric transducer and a controllable servo motor coupled to a shaft which supports either the mounting plate directly or the piezoelectric transducer. By controlling the servo motor, the height of the shaft can be changed to deform the mounting plate. In one embodiment, the servo is used to perform coarse adjustments in combination with the piezoelectric transducer which is used to perform fine adjustments.

The deformable chuck operates under the control of a panel shape controller. The controller receives inputs from an optical panel surface sensor which in the preferred embodiment detects the flatness of the panel. The optical sensor preferably includes a light source and optical system which directs light from the light source through a Ronchi grating onto the surface of the substrate. Light returning from the substrate surface passes back through the Ronchi grating to generate a Moire fringe pattern. The optical sensor analyzes the Moire pattern which indicates the flatness of the panel surface. The sensor generates signals indicative of the flatness of the panel and forwards the signals to the panel shape controller. The controller translates the signals from the sensor into control signals used to selectively change the length of particular supports. As the support lengths are changed, the portions of the deformable top surface and the substrate are moved relative to each other to flatten the substrate surface as desired.

The system of the invention also includes an alignment system which aligns the substrate being exposed to the deformable chuck and also aligns the projected mask image with the substrate before exposure to ensure proper registration between exposure pattern layers. The top surface of the deformable chuck includes two charge-coupled device (CCD) pixel arrays located on the top surface so as to be positioned at opposite corners of the exposure pattern. The substrate, such as a transparent glass flat panel display, can have two opaque regions with transparent circular center regions located at the opposite corners of the pattern. The substrate is first placed over the top surface of the chuck such that the opaque regions mask the CCD pixel arrays. Before the exposure, the chuck with installed substrate are located under the projection optics. A source of light having a wavelength to which the photoresist is not reactive illuminates the mask and projects light from the mask onto the substrate. The mask is formed with two transparent circular regions in opposite corners of the pattern such that when the mask pattern is aligned with the substrate, two circular patterns of light will fall within the transparent regions of the substrate located over the CCD pixel arrays of the chuck. By reading data from the pixel arrays and moving the chuck appropriately, the exposure pattern can be aligned with the substrate. To eliminate contact with the CCD or to adjust CCD detector size, the CCD and panels may be coupled by reimaging optics. After the alignment is completed, the exposure light source is activated to expose the substrate.

In the system of the invention, the exposure of a panel can be considered a two-step process. In the first step, the deformable chuck is located in a panel flattening position. A standard panel loader loads a panel onto the chuck such that the opaque regions at its corners are in alignment with the CCD pixel arrays in the chuck. Also, at this first position, the panel flattening system is activated to flatten the panel as described above. After the panel is flattened, the chuck is moved to the exposure position under the projection optics. The process of aligning the mask to the panel is then carried out with the light source emitting light to which the photoresist is not reactive. After alignment of the mask to the panel, the process is completed by exposure of the panel with exposure light through the mask.

In a preferred embodiment of the lithography system of the invention, two substrates or panels can be processed simultaneously. To that end, the preferred embodiment of the invention actually includes two deformable chucks mounted at opposite ends of a movable slide system. While a first panel is being exposed on the first chuck, a second panel is being loaded onto the second chuck and flattened. After the first panel is exposed and the second panel is flattened, the slide changes positions such that the second panel can be aligned with the mask and then exposed. The first panel is simultaneously unloaded from the first chuck, and a third panel is loaded onto the first chuck and flattened. It will be appreciated that this simultaneous flattening and exposing process greatly enhances the system's panel throughput.

The enlarging lithography system of the invention provides numerous advantages over prior traditional lithography systems. By enlarging the image of the mask, the system of the invention can produce image patterns on large substrates using standard readily available mask sizes. In other one-to-one systems, large patterns are produced from large masks. Such masks are difficult and expensive to produce and, as such, are not readily available from commercial mask suppliers.

The system of the invention also provides the further advantage of producing large patterns on substrates with a single exposure. Prior systems make multiple exposures using multiple masks for each layer and then stitch the patterns together. Other systems have one-to-one mask-to-pattern relationships; however, these systems produce the exposure by line or area scanning. This can be time consuming and also introduces irregularities in feature size because of fluxuations in the source luminance. Because the system of the invention can produce a mask pattern with a single exposure, the system of the invention can produce panels at a much higher rate than prior systems. Panels having very high resolution, i.e., very low feature size, can be produced efficiently with high yields at a very high throughput.

BRIEF DESCRIPTION OF THE DRAWINGS

The foregoing and other objects, features and advantages of the invention will be apparent from the following more particular description of preferred embodiments of the invention, as illustrated in the accompanying drawings in which like reference characters refer to the same parts throughout the different views. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating the principles of the invention.

FIG. 1

is a schematic diagram illustrating mask image enlargement in a photolithography system in accordance with the present invention.

FIG. 2

is a schematic diagram of one embodiment of the illumination optics of the present invention.

FIG. 3

is a schematic diagram of an alternative embodiment of the illumination optics of the invention.

FIG. 4

is a schematic diagram of one preferred embodiment of the projection optics of the present invention.

FIG. 5

is one alternative preferred embodiment of the projection optics of the present invention.

FIG. 6

is another alternative preferred embodiment of the projection optics of the present invention.

FIG. 7

is another alternative preferred embodiment of the projection optics of the present invention.

FIG. 8

is another alternative preferred embodiment of the projection optics of the present invention.

FIG. 9

is another alternative preferred embodiment of the projection optics of the present invention.

FIG. 10

is a schematic functional block diagram of a photolithography system in accordance with the present invention.

FIG. 11

is a schematic diagram of a preferred embodiment of the deformable chuck of the present invention.

FIGS. 12A-12C

are schematic illustrations of the substrate flattening process of the present invention.

FIG. 13

is a schematic functional block diagram of the panel flattening and panel flatness sensing processes of the invention.

FIGS. 14A and 14B

compare the Moire patterns for an irregular substrate surface and a flattened substrate surface.

FIGS. 15A and 15B

schematically illustrate the substrate and mask alignment processes of the present invention.

DETAILED DESCRIPTION OF THE INVENTION

FIG. 1

is a schematic block diagram of a photolithography system

10

in accordance with the present invention. The system

10

includes a light source

12

which directs light

14

onto illumination optics

16

. The illumination optics

16

perform various functions depending upon the application. For example, if the source

12

is a laser source, the illumination optics

16

can include an anamorphic relay, a beam enlarger, a beam homogenizer and/or a field lens for generating a uniform light beam. If the light source

12

is an incandescent source, the illumination optics

16

can comprise a condenser illumination system for producing a uniform output. The illumination optics

16

direct illumination light

18

onto a patterned mask

20

. The patterned light

22

which emerges from the mask

20

is directed onto projection optics

24

. In one preferred embodiment, the projection optics

24

generate an enlarged image of the mask and direct output light

26

onto a surface

30

of a substrate

28

such that the enlarged image of the mask is formed on the substrate

28

.

FIG. 2

is a schematic diagram of a laser source

112

and illumination optics

116

in accordance with one embodiment of the invention. The illumination optics

116

receive the light

114

from the laser

112

at an anamorphic relay

32

comprising two lenses

40

and

42

and two plates

44

and

46

. The relay

32

shapes the laser beam appropriately from rectangular to square for the photolithographic process. Light from the relay

32

is directed onto a beam homogenizing system

34

which can comprise a four-sided plate prism

48

for splitting the light beam and a lens prism

50

having a four-faceted prism surface for redirecting the beams on top of one another and thus homogenizing the laser beam.

A mirror

36

directs the homogenized light through a field lens

38

which produces the output light

118

which provides the illumination for the mask

20

. The homogenizer

34

shown in

FIG. 2

is of the type described in commonly assigned and U.S. patent application Ser. No. 08/271,029 filed on Jul. 6, 1994, now U.S. Pat. No. 5,548,444, which is incorporated herein in its entirety by reference. Other types of homogenizers may be used. In addition, depending upon the laser used, the homogenizer may be eliminated altogether.

FIG. 3

is a schematic diagram of an alternative embodiment of the source and illumination optics. In this embodiment, the source is an incandescent light

212

. Light

214

from the source

212

enters the illumination optics

216

which include three lenses

213

,

215

,

217

shaped and arranged to produce a condenser illumination system

216

. The condenser illumination optics

216

generate uniformly bright light

218

used to illuminate the mask

20

and expose the desired pattern on the substrate.

The configuration of the projection optics

24

is in general driven by the requirements of the system. The configuration is designed based upon factors such as required system resolution, exposure pattern size, system telecentricity, the wavelength of the exposure light and the numerical aperture and f/number of the system. For example, the resolution R of the system, defined as the smallest feature size which can be resolved by the system is inversely proportional to the numerical aperture (NA) of the system and directly proportional to the wavelength λ of the light at which the system is operating. Thus, at a given selected wavelength determined by the photoresist being used, improved resolution requires a high numerical aperture. Numerical aperture is inversely proportional to the f/number of the system. Therefore, good resolution requires low f/number, which can be difficult to achieve. Thus, the projection optics must be capable of providing a high numerical aperture to achieve high resolution, or, equivalently, small feature size, while also maintaining high telecentricity, a wide field, low distortion and low astigmatism.

Other photolithography systems with mask enlargement have been suggested in the prior art. For example, Japanese unexamined patent publication number 62-122126 suggests a system in which a pair of plane mirrors or a pair of lenses or even a single lens can be used to produce an enlarged image of the mask on a substrate. Also, U.S. Pat. No. 5,008,702 to Tanaka et al. suggests a system in which a single lens or a pair of ellipsoid mirrors can be used to perform the enlargement. With such simplistic enlargement optics, neither of these systems is capable of achieving the requirements which the system of the present invention achieves. For example, such systems are incapable of achieving the low resolution, wide field width, high telecentricity, low distortion and low astigmatism achieved by the present invention. These simple two-element and single-element systems using planar and conic elements cannot meet the requirements for a photolithography system dictated by the semiconductor and display industries.

FIG. 4

is a schematic block diagram of one preferred embodiment of projection optics

124

used in the photolithography system of the invention. In one preferred embodiment, these optics

124

are used to produce two-times (2×) enlargement. That is, the projection optics

124

produce an image on a substrate which is double the size of the mask pattern. In the projection optics

124

of

FIG. 4

, light

122

from the mask first passes through a refractive distortion corrector

102

which corrects for various types of distortion including pin cushion distortion, barrel distortion and other forms of distortion. A first surface

103

of the distortion corrector

102

is planar, and the second surface

104

is an aspheric shaped surface.

Light from the distortion corrector

102

is directed onto a surface of a positive aspheric powered primary mirror

106

. Light from the primary mirror

106

is directed to an air-spaced lens doublet

108

comprising a first lens

107

and a second lens

109

. The doublet

108

serves as an astigmatism corrector to improve the resolution of the system

124

. The doublet

108

includes four optical surfaces, namely, the first surface

107

b

and second surface

107

a

of lens

107

and the first surface

109

b

and second surface

109

a

of lens

109

. Surface

107

a

and surface

109

b

are both spherical surfaces. Surface

107

b

and surface

109

a

are both aspheric surfaces.

Light which passes through the lens doublet

108

strikes the negative aspheric powered secondary mirror

110

which serves as the aperture stop of the system

124

. It comprises a reflective coating

111

deposited on the surface of a shaped substrate

113

. Light from the secondary mirror

110

is reflected through the lens doublet

108

to a positive aspheric powered tertiary mirror

114

which forms the enlarged image

118

of the mask at the surface of the substrate being exposed.

Each optical surface of the optics of

FIG. 4

can be described by Equation (1) below. Equation (1) describes surfaces which are bilaterally symmetric in both X and Y but not necessarily rotationally symmetric. Equation (1) and the variables and physical quantities listed therein are taken from the Code V User's Manual and other associated literature published by and available from Optical Research Associates. Code V is an optical design software package available from Optical Research Associates. The User's Manual and all of the literature associated with Code V are incorporated herein in their entirety by reference.

\begin{matrix} Z = \frac{CUX X^{2} + CUY Y^{2}}{1 + SQRT [1 - (1 + KX) {CUX}^{2} X^{2} - (1 + KY) {CUY}^{2} Y^{2}]} + {AR [(1 - AP) X^{2} + (1 + AP) Y^{2}]}^{2} + {BR [(1 - BP) X^{2} + (1 + BP) Y^{2}]}^{3} + {CR [(1 - CP) X^{2} + (1 + CP) Y^{2}]}^{4} + {DR [(1 - DP) X^{2} + (1 + DP) Y^{2}]}^{5}; & (1) \end{matrix}

where

Z—is the sag of the surface parallel to the Z axis,

CUX, CUY—are the curvatures in X and Y, respectively,

KX, KY—are the conic coefficients in X and Y, respectively, and correspond to eccentricity in the same way as K for the ASP surface type,

AR, BR CR, DR—are the rotationally symmetric portion of the 4th, 6th, 8th, and 10th order deformation from the conic, and

AP, BP, CP, DP—represent the non-rotationally symmetric components of the 4th, 6th, 8th, and 10th order deformation from the conic

Table 1 is a file listing generated by Code V for the optical surfaces of FIG.

4

. It gives complete characteristics for all of the surfaces of all of the elements shown including substitute values for the variables in Equation (1). Table 2 is a glossary of abbreviations which defines the abbreviations used in Table 1. The abbreviations and associated physical characteristics are also defined in the Code V User's Manual and associated literature.

TABLE 1

RDM;LEN

TITLE ‘REFLECTIVE PROJ. OPTICS 2x (ASML3)’

NAO

0.097

DIM

I

WL

365.0

REF

1

WTW

1

1NI

‘GMP’

CA

APE

XOB

0.0 0.0 0.0 2.28 2.28 2.28 4.56 4.56 4.56

YOB

−3.515 0.0 3.515 −3.515 0.0 3.515 −3.515 0.0 3.515

VUX

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

VLX

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

VUY

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

VLY

0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

SO

0.0 76.1349290064

THC

0

S

−70.0176157718 −31.2106402396 REFL

CCY

0; THC 0

ASP

K

0.217500346293; KC 0

IC

Yes; CUF 0.0

A

0.964197052357e−7; B 0.734249377713e−11; C 0.601797907743e−14; D&

0.355670476138e−17

AC

0; BC 0; CC 0; DC 0

E

−0.565744496308e−20; F 0.278497772167e−23; G 0.0; H 0.0

EC

0; FC 0

J

0.0

XDE

0.0; YDE −19.3632308819; ZDE 0.0; XDC 100; YDC 0; ZDC 100

REX

11.9; REY 11.6000003815

ADY

19.435

CUM

0.0; THM 2.0

S

−31.229684918 −1.45954577544 SILICA_SPECIAL

CCY

1; THC 30

ASP

K

−8.54623113744; KC 13

IC

Yes; CUF 0.0

A

−0.310341422461e−4; B 0.311448336916e−6; C 0.211654115522e−8; D&

−0.798316365955e−10

AC

2; BC 3; CC4; DC 5

E

0.997730866037e−12; F 0.0; G 0.0; H 0.0

EC

6

J

0.0

CIR

6.15

S

551.633541885 −0.005

CCY

40; THC 31

CIR

6.15

S

−321.826568486 −1.66914076071 SILICA_SPECIAL

CCY

41; THC 32

CIR

6.15

S

−18.3450931938 −0.345104743867

CCY

7; THC 0

ASP

K

−2.17332648616; KC 14

IC

Yes; CUF 0.0

A

−0.476574734841e−4; B 0.400219027094e−06; C 0.156651391336e−7; D&

−0.193358160134e−8

AC

8; BC 9; CC 10; DC 11

E

0.805876201035e−10; F 0.0; G 0.0; H 0.0

EG

12

J

0.0

CIR

3.75

S

−24.9701983582 0.345104743867 REFL

CCY

0

STO

ASP

K

0.265464981166; KC 0

IC

Yes; CUF 0.0

A

−0.343439578675e−5; B −0.529156594997e−7; C 0.761973759698e−8; D&

−0.312374878835e−9

AC

0; BC 0; CC 0; DC 0

CIR

3.4

CUM

0.0; THM 1.0

S

−18.3450931938 1.66914076071 SILICA_SPECIAL

CCY

THC −32

ASP

K

−2.17332648616; KC 14

IC

Yes; CUF 0.0

A

−0.476574734841e−4; 0.400219027094e−6; C 0.156651391336e−7; D&

−0.193358160134e−8

AC

8; BC 9; CC 10; DC 11

E

0.805876201035e−10; F 0.0; G 0.0; H 0.0

EC

12

J

0.0

CIR

3.75

S

−321.826568486 0.005

CCY

41; THC −31

CIR

6.15

S

551.633541885 1.45954577544 SILICA_SPECIAL

CCY

40; THC −30

CIR

6.15

S

−31.229684918 29.4463993312

CCY

1; THC 0

ASP

K

−8.54623113744; KC 13

IC

Yes; CUF 0.0

A

−0.310341422461e−4; B 0.311448336916e−6; C 0.211654115522e−8; D&

−0.798316365955e−10

AC

2; BC 3; CC 4; DC 5

E

0.997730866037e−12; F 0.0; G 0.0; H 0.0

EC

6

J

0.0

CIR

6.15

S

−50.5448093835 −29.4463993312 REFL

CCY

0

ASP

K

0.210599847159; KC 0

IC

Yes; CUF 0.0

A

0.912852257642e−7; B 0.306648812717e−10; C −0.311193415968e−14; D&

0.343356863197e−16

AC

0; BC 0; CC 0; DC 0

E

−0.317533586871e−19; F 0.165343877316e−22; G 0,0; H 0.0

EC

0; FC 0

J

0.0

REX

13.95; REY 13.3000001907

ADY

−17.435

CUM

0.0; THM 2.0

S

0.0 −19.5691536203

THC

0

S

−1629.76704542 −0.5 SILICA_SPECIAL

CCY

0

ASP

K

0.0

IC

Yes; CUF 0.0

A

0.211925831612e−4; B −0.298749258785e−7; C 0.0; D 0.0

AC

0; BC 0

REX

3.15; REY 2.65000009537

ADY

−10.14

S

0.0 −2.2771100671

THC

0

REX

3.15; REY 2.65000009537

ADY

−10.14

SI

0.0 0.0

DAR

XDE

0.0; YDE −9.68161259758; ZDE 0.0; XDC 100; YDC 0; ZDC 100

THI

S6 OAL S5 . . . 7 0.0

THI

S11 OAL S10 . . . 12 0.0

GO

TABLE 2

RDM:

radius input/output mode instead of curvature

LEN:

initialize data for new lens

TITLE:

title of output

NAO:

numerical aperture in object space

DIM:

lens units (inches/centimeters/millimeters)

WL:

wavelengths

REF:

number of reference wavelength

WTW:

wavelength weights

INI:

user initials for documentation

CA:

flag for use of apertures (YES/NO/APE)

XOB:

X object height in lens units

YOB:

Y object height in lens units

VUX:

fraction of entrance pupil radius clipped off for +X direction

VLX:

fraction of entrance pupil radius clipped off for −X direction

VUY:

fraction of entrance pupil radius clipped off for +Y direction

VLY:

fraction of entrance pupil radius clipped off for −Y direction

STO:

designate stop surface

Shorthand Surface Entry:

General form:

S|Sk|Si . . . j y_rad_curv thickness

glass_name|REFL|TIRO [glass_name]

(y_rad_curv is either radius or curvature depending on

RDM designation)

ASP:

change surface type to asphere (see polynomial description)

K:

conic coefficient

IC:

seley ray intersection, with normal and ray in same direction

CUF:

Fresnel surface with designated base curve

A:

4th order aspheric coefficient

B

6th order aspheric coefficient

C:

8th order aspheric coefficient

D:

10th order aspheric coefficient

E:

12th order aspheric coefficient

F:

14th order aspheric coefficient

G:

16th order aspheric coefficient

H:

18th order aspheric coefficient

J:

20th order aspheric coefficient

XDE:

X displacement of new axis

YDE:

Y displacement of new axis

ZDE:

Z displacement of new axis

REX:

rectangular aperture X half-width dimension

REY:

rectangular aperture Y half-width dimension

ADY:

Y decenter for designated aperture

CIR:

circular aperture radius

CUM:

spherical curvature for back of 1st surface mirror

THM:

center thickness to back of 1st surface mirror

CCY:

curvature (radius) variable control code

THC:

thickness variable control code

KC:

conic constant variable control code

AC:

4th order aspheric coefficient variable control code

BC:

6th order aspheric coefficient variable control code

CC:

8th order aspheric coefficient variable control code

DC:

10th order aspheric coefficient variable control code

EC:

12th order aspheric coefficient variable control code

FC:

14th order aspheric coefficient variable control code

XDC:

X displacement variable control code

YDC:

Y displacement variable control code

ZDC:

Z displacement variable control code

DAR:

designate surface as a decenter-and-return surface

Thickness Solve: THI Sk OAL Si . . . j length

Solve center thickness on Sk to produce overall length (OAL)

between surfaces

Si and Sj equal to the value of length.

FIG. 5

is a schematic diagram of an alternative embodiment of the projection optics

150

of the present invention. Like the previously described embodiment, this embodiment also provides a 2× enlargement between the mask and the pattern image formed on the substrate. The embodiment shown in

FIG. 5

includes four mirrors, each of which has an unusual, i.e., non-conic and non-aspheric, optical surface described by a polynomial. Unusual surfaces have deformations from conic shape which cannot be described by Equation (1). Light

152

from the mask is reflected from primary mirror

154

to secondary mirror

156

which also acts as the aperture stop for the system

150

. The secondary mirror

156

reflects the light toward tertiary mirror

158

which in turn directs the light onto quaternary mirror

160

. The quaternary mirror

160

forms the 2× enlarged image

162

of the mask at the surface of the substrate. As mentioned above, each of the mirrors

154

,

156

,

158

and

160

has an unusual shape defined by a polynomial. Mirrors

154

and

158

are positive powered unusual mirrors, and mirrors

156

and

160

are negative powered unusual mirrors.

Equation (2) is the polynomial which defines unusual surfaces such as the mirror surfaces of FIG.

5

.

\begin{matrix} Z = (\frac{α - β}{{(1 - C_{z} α)}^{1 / 2} + {(1 - C_{z} β)}^{1 / 2}}) + (U_{4} γ^{2} (α^{2} - β^{2}) + U_{6} γ^{3} (α^{3} - β^{3}) + U_{8} γ^{4} (α^{4} - β^{4}) + U_{10} γ^{5} (α^{5} - β^{5})) + A_{1} y^{3} + A_{2} x^{2} y + A_{3} y^{4} + A_{4} x^{2} y^{2} + A_{5} x^{4} + A_{6} y^{5} + A_{7} x^{2} y^{3} + A_{8} x^{4} y + A_{9} y^{6} + A_{10} x^{2} y^{4} + A_{11} x^{4} y^{2} + A_{12} x^{6} + A_{13} y^{7} + A_{14} x^{2} y^{5} + A_{15} x^{4} y^{3} + A_{16} x^{6} y + A_{17} y^{8} + A_{18} x^{2} y^{6} + A_{19} x^{4} y^{4} + A_{20} x^{6} y^{2} + A_{21} x^{8} + A_{22} y^{9} + A_{23} x^{2} y^{7} + A_{24} x^{4} y^{5} + A_{25} x^{6} y^{3} + A_{26} x^{8} y + A_{27} y^{10} + A_{28} x^{2} y^{8} + A_{29} x^{4} y^{6} + A_{30} x^{6} y^{4} + A_{31} x^{8} y^{2} + A_{32} x^{10}; & (2) \\ \begin{matrix} where & α = {C_{x} (x - x_{o})}^{2} + {C_{y} (y - y_{o})}^{2}, \\ β = C_{x} x_{o}^{2} + C_{y} y_{o}^{2}, \\ γ = 1 / \sqrt{C_{x}^{2} + C_{y}^{2}} . \end{matrix} \end{matrix}

Further details regarding Equation (2) are contained in “Nonrotationally Symmetric Systems: A New Problem for Optical Design and Fabrication, Applications of Geometrical Optics,” by B. Tatian,

Proc. SPIE,

39:205, (1973) and “Testing an Unusual Optical Surface,” by B. Tatian, 1985 International Lens Design Conference,

Proc. SPIE,

554:139, (1985), both of which are incorporated herein in their entirety by reference.

Table 3 is a computer printout which lists substitute values for the variables and coefficients of Equation (2) used to describe the four optical surfaces of FIG.

5

. Table 4 is a glossary of abbreviations used in Table 3.

TABLE 3

POLY=ps4,doly=g1,PRMS=pM1,polya=ps2,t=1000;

‘CLIC’ MM=1 ;

LENS_= 17 ID=‘ASML4M.MDL’ ;

STOP= 3 TYPE= 3 OBJ1=−134.0000000 OBJ2= 1.4250725722E+03

OBJ3= 9.6500000000E−02 WV1= 0.36500000 H= 0.82090000 ;

SN= 0 ;

SN= 1

CV= 4.8155932693E−05

TH= −1.2005131554E+03

CC= −8.40817115

NM= −1.00000000 CONT=1 ;

XD= 0

YD=−150.3768207

AT= 0

BT=−3.7794594797E−01

UT= 1 MIRR=1

CX= 3.8655932693E−05

CY= 4.8155932693E−05

CZ=−3.5674739114E−04

Y0=−9.4228205652E+01

U4=−2.4870413368E−11

U6=−6.6917431136E−16

Y3= 3.0902512637E−08

X2Y= 9.4554851156E−09

Y4= 6.5125911860E−11

X2Y2= 6.9464488554E−11

X4 = 1.4696424767E−11

Y5 = 1.8131422932E−13

X2Y3= 2.8334921895E−13

X4Y= 1.1262859295E−13

Y6= 3.1157028657E−16

X2Y4= 7.6829431449E−16

X4Y2= 6.2180518423E−16

X6= 1.6373889732E−16 ;

SN= 2

CV= 3.6480239913E−04

TH= 1.5614213802E+03

CC= −0.12611114

CONT=1 ;

XD= 0

YD=−60.02104507

AT= 0

BT=−1.2555809543E−01

UT= 2 MIRR=1

CX= 3.5596823135E−04

CY= 3.6480239913E−04

CZ= 3.1879675098E−04

Y0= 8.3252661835E+01

U4= 4.4246210302E−11

U6= 5.6417338753E−18

Y3= 1.7414703278E−08

X2Y= 1.0179036312E−08

Y4=−1.9257270787E−11

X2Y2=−4.1796502011E−11

X4 =−2.0780037090E−11

Y5 = 2.9138091736E−15

X2Y3= 3.5931846661E−15

X4Y= 9.7880912390E−16

Y6=−1.9872742628E−18

X2Y4=−4.8104973423E−18

X4Y2=−4.9508308828E−18

X6=−1.7260348488E−18 ;

SN= 3

CV= 8.0712102034E−04

TH=−1.1724100918E+03

CC= 2.69611347

NM= −1.00000000 CONT=1 ;

XD= 0

YD=−49.58418071

AT= 0

BT= 4.4143975291E−01

UT= 3 MIRR=1

CX= 8.1246442550E−04

CY= 8.0712102034E−04

CZ= 2.9832108773E−03

Y0= 5.3717294465E+00

U4= 9.9422365563E−09

U6= 5.7346931839E−13

Y3= 1.4499799570E−07

X2Y= 1.0173418999E−07

Y4=−5.0846000493E−09

X2Y2=−1.0194589517E−08

X4 =−5.1112781850E−09

Y5 = 6.5151851876E−12

X2Y3= 1.3050129588E−11

X4Y= 6.5440700459E−12

Y6=−2.0087480119E−13

X2Y4=−6.0654046045E−13

X4Y2=−6.1076435589E−13

X6=−2.0491918073E−13 ;

SN= 4

CV= 5.6283787866E−04

TH= 1.5203344356E+03

CC= 1.07468593

CONT=1 ;

XD= 0

YD= 50.54751899

AT= 0

BT=−1.5817856763E−01

UT= 4

MIRR=1

CX= 5.5342621388E−04

CY= 5.6283787866E−04

CZ= 1.1677118297E−03

Y0=−1.7845331493E+01

U4=−4.5378276818E−10

U6=−5.9894313104E−16

Y3= 8.5049747506E−09

X2Y= 6.9528803749E−09

Y4= 2.1234922862E−10

X2Y2= 4.1578653373E−10

X4 = 2.0320895057E−10

Y5 = 1.9259750237E−14

X2Y3= 3.7733104534E−14

X4Y= 1.8395226634E−14

Y6= 2.0500896897E−16

X2Y4= 6.0497943131E−16

X4Y2= 5.9354293205E−16

X6= 1.9454018909E−16 ;

SN= 5 CONT=1 ;

XD= 0

YD= 5.81973751

AT= 0

BT=−2.0673315191E−01 ;

TABLE 4

CLIC:

initialize data for new lens

MM:

lens units (millimeters)

ID:

title of lens

STOP:

designate stop surface

TYPE:

object data type

OBJ1:

object height in lens units (TYPE 3)

OBJ2:

distance from object to first surface (TYPE 3)

OBJ3:

marginal ray angle (TYPE 3)

H:

maximum object height in Y direction normalized to OBJ1

WV1:

first wavelength

SN:

surface number

CV:

surface curvature

TH:

thickness to next surface

CC:

conic coefficient

NM:

index of refraction

CONT:

input data continuation flag (read in data on subsequent lines)

XD:

X displacement of new axis

YD:

Y displacement of new axis

AT:

tilt in XZ plane

BT:

tilt in YZ plane

UT:

negate tilt before proceeding to next surface

MIRR:

adjust coordinate system tilt as if mirror reflected before

proceeding to next surface

Unusual Surface Coefficients: (see attached polynomial description)

CX:

surface base curvature in X direction

CY:

surface base curvature in Y direction

CZ:

sets conic coefficient for the quadric

(CCX = −1 + CZ/CX; CCY = −1 + CZ/CY)

YO:

Y displacement of axis of base quadric from Z axis of local

coordinate system

U4:

4th order aspheric coefficient

U6:

6th order aspheric coefficient

Y3:

Y

3

aspheric coefficient

X2Y:

X

2

Y aspheric coefficient

Y4:

Y

4

aspheric coefficient

X2Y2:

X

2

Y

2

aspheric coefficient

X4:

X

4

aspheric coefficient

Y5:

Y

5

aspheric coefficient

X2Y3:

X

2

Y

3

aspheric coefficient

X4Y:

X

4

Y aspheric coefficient

Y6:

Y

6

aspheric coefficient

X2Y4:

X

2

Y

4

aspheric coefficient

X4Y2:

X

4

Y

2

aspheric coefficient

X6:

X

6

aspheric coefficient

Both the projection optics

124

of FIG.

4

and the projection optics

150

of

FIG. 5

achieve a nominal system resolution of 3 μm at an exposure light wavelength of 0.365 μm. Both systems are thus ideally suited for high density semiconductor and flat panel display applications. Hence, using a standard six inch square mask having a minimum line width of 1.5 μm, the 2× enlargement systems of

FIGS. 4 and 5

can produce a semiconductor wafer or flat panel display having a twelve inch square pattern area with 3 μm line width which can include active matrix LCD pixels as well as driver circuits along the boarder of the panel.

FIGS. 6-9

are schematic diagrams of four alternative embodiments of the projection optics of the invention.

FIG. 6

is a schematic diagram of an embodiment

200

of a 1× projection optics system. That is, the image

202

of the mask formed at the surface of the substrate is the same size as the mask pattern

201

itself. The 1× system

200

includes a primary mirror

203

which reflects light from the mask back onto a Mangin secondary mirror

204

. The Mangin mirror

204

includes a reflective coating on the back surface of a refractive optical material such as glass. The device simultaneously serves as a reflective aperture stop for the system and a refractive astigmatism corrector. Light from the Mangin mirror

204

is directed onto a tertiary mirror

205

which forms the image

202

of the mask on the surface of the substrate.

The system

200

of

FIG. 6

has numerical aperture of 0.05 and achieves a field format of 50 cm×50 cm. The system is designed to produce an image at an exposure wavelength of 0.25 μm. The diffraction blur of the system is 2.9 μm (50% energy diameter). The wavefront error is less than 0.03 waves and the distortion over the 50 cm×50 cm format is 0.0 μm. The telecentricity of the system is less than 0.31 degrees.

Regarding the optical elements themselves, the primary mirror

203

and tertiary mirror

205

are identical. Both are 1 meter×1 meter sections with identical surface prescriptions and locations. Both have sixth-order rotationally symmetric aspheric surfaces with 37 μm aspheric departure and 0.22 μm/mm aspheric departure slope. Regarding surface quality, the distortion criterion is λ/70 peak-to-valley over the entire footprint, and wavefront criterion is λ/20 peak-to-valley over the footprint at a wavelength of 0.6328 μm. The secondary Mangin mirror

204

is made of fused silica material with a reflective coating on its backside. It has a 24 cm diameter and both of its surfaces have eighth-order rotationally symmetric aspheric shape with <1 μm aspheric departure and less than 0.11 μm/mm aspheric departure slope. The surface quality is λ/20 peak-to-valley over the footprint (wavefront criterion) at a wavelength of 0.6328 μm.

FIG. 7

is a schematic diagram of a 2.5× projection optics system

250

in accordance with the invention. The system

250

provides a 50 cm×50 cm display format from a 20 cm×20 cm mask format. The system

250

is f/12.5 at the substrate and f/5.0 at the mask and is designed to operate at an exposure wavelength of 0.248 μm. The distortion of the system is <0.08 μm over 0.5 m. The telecentricity of the system at the image plane is less than 0.32 degrees and at the mask plane is less than 4.80 degrees.

The system of

FIG. 7

is similar to the 2× embodiment described above in connection with FIG.

4

. Light

251

from the mask first passes through a plano-conic distortion corrector

252

. The light then strikes a positive powered mirror of tenth-order rotational asphericity which reflects the light through an aspheric-planar astigmatism corrector

254

and onto the reflective surface of the secondary mirror and aperture stop

255

of the system. A first surface of the astigmatism corrector

254

has a sixth-order rotationally symmetric aspheric surface, and the back surface is planar. The secondary mirror

255

is a negative powered mirror having a conic shape. The light reflected by the secondary mirror

255

passes through the astigmatism corrector to tertiary mirror

255

a

which forms the image

256

of the mask on the surface of the substrate.

FIG. 8

is a schematic diagram of a 4× projection optics system

300

in accordance with the present invention. This embodiment is similar to the 2× embodiment described in connection with

FIG. 5

in that it also includes four mirror surfaces which generate the enlarged image. The RMS wavefront error of the system is <0.083 waves and the distortion is 0.025 μm over the 50 cm×50 cm field format of the system. The numerical aperture of the system is 0.0417 and the designed exposure wavelength is 0.250 μm. In the system

300

, light

301

from the mask is reflected from primary mirror

302

to secondary mirror

303

to tertiary mirror

304

to quaternary mirror

305

which forms the image

306

of the mask at the surface of the substrate. The primary mirror

302

and secondary mirror

304

are both positive powered mirrors, and the secondary mirror

303

and quaternary mirror

305

are negative powered mirrors. All four mirrors have rotationally symmetric aspheric surfaces. The primary mirror

302

is 80 cm×72 cm and has 1,172 μm aspheric departure and 3.6 μm/mm aspheric departure slope. The secondary mirror

303

is circular with a 26 cm diameter. It has 1,190 μm aspheric departure and 10.4 μm/mm aspheric departure slope. The tertiary mirror 304 is 78 cm×75 cm, has 987 μm aspheric departure and 2.7 μm/mm aspheric departure slope. The quaternary mirror

305

is 48 cm×48 cm, has 116 μm aspheric departure and 0.5 μm/mm aspheric departure slope.

FIG. 9

is a schematic diagram of a 10× projection system

350

in accordance with the present invention. Light

351

from the mask is reflected by primary mirror

352

onto and through an astigmatism corrector

353

to the secondary mirror

354

, which serves as the aperture stop of the system. Light is reflected by the secondary mirror

354

back through the astigmatism corrector

353

onto tertiary mirror

355

which reflects the light to form the image

356

of the mask on the surface of the substrate. The system is f/80 at the substrate surface and has a field format of 2.0 m×2.0 m. The distortion is less than 0.01 μm over 2.0 m, and the design wavelength is 0.248 μm. The telecentricity at the substrate plane is less than 0.29 degrees and at the mask plane is less than 13.5 degrees. The primary mirror

352

and tertiary mirror

355

both have tenth-order rotationally symmetric asphericity. The secondary mirror

354

has a conic surface, and the astigmatism corrector

353

has a conic front surface and planar back surface.

FIG. 10

is a schematic functional block diagram showing a preferred photolithography system

500

using an enlarged mask image in accordance with the present invention. The system

500

includes a source

12

of light and illumination optics

16

which receive light from the source and direct light onto the mask

20

to illuminate the mask. Light transmitted by the mask

20

is received by the projection optics

24

described in detail above.

The projection optics

24

project the enlarged image of the mask

20

onto a surface of a first substrate

28

a

mounted to a first deformable chuck

502

a

which is shown located in an exposure position by a chuck positioning structure and processor

504

. The chuck positioning structure and processor

504

also locates a second deformable chuck

502

b

in a flattening position as shown. A second substrate

28

b

is mounted on the second deformable chuck

502

b.

In operation, a substrate loader

506

loads a substrate

28

b

onto the deformable chuck

502

b.

A flatness sensor

508

of the system detects the flatness of the top surface of the substrate

28

b.

A signal indicative of the flatness of the substrate

28

b

is forwarded on line

510

to a processor

512

. The processor generates control signals and forwards them to the deformable chuck

502

b.

The control signals are used to selectively deform the chuck to flatten the surface of the substrate

28

b

to ensure that the subsequent exposure will be in focus and free of pattern distortion.

After the substrate loader

506

loads a substrate

28

b

onto the deformable chuck

502

b,

a chuck-substrate alignment system

514

is used to provide an initial coarse alignment between the substrate

28

b

and the surface of the deformable chuck

502

b.

Where the substrate

28

b

is a transparent panel for a flat panel display system, the opposite corners of the substrate are formed with opaque mask regions surrounding a circular transparent region. The surface of the chuck

502

b

includes two CCD pixel arrays located on the chuck so as to be positioned beneath the opaque masked regions and inner transparent regions of the substrate

28

b.

As a substrate

28

b

is loaded onto the chuck

502

b

by the substrate loader

506

, the substrate loader

506

is adjusted to ensure that each inner circular transparent region of the opaque masking region of the substrate

28

b

overlays a CCD pixel array.

In a preferred embodiment, while the substrate

28

b

in the flattening position is being flattened, substrate

28

a

in the exposure position can be exposed. However, before the exposure, the mask pattern is first aligned with the substrate

28

a

via the mask-substrate alignment system

516

. This fine adjustment of the substrate location is performed by activating the source

12

to illuminate the top surface of the substrate

28

a

with the enlarged mask pattern with light having a wavelength to which the photoresist on the top surface of the substrate

28

a

is not reactive. This prevents the photoresist from being exposed during the alignment process. Each mask pattern is produced to generate a circular pattern of light at opposite corners such that when the mask pattern is precisely aligned with the substrate

28

a,

the CCD pixel arrays on the deformable chuck

502

a

beneath the circular transparent regions of the opaque mask regions of the substrate

28

a

will be illuminated.

The processor in the chuck positioning system

504

reads the CCD pixel arrays during the alignment process to determine if the circular illumination spots are centered on the circular transparent regions formed on the substrate

28

a

within the opaque masking regions. The position of the deformable chuck

502

a

is adjusted according to the pixel data to center the circular illumination spots within the circular transparent regions in order to precisely align the mask pattern with the substrate

28

a

before exposure. The mask-substrate alignment process is done to ensure proper registration between pattern layers in devices requiring multiple exposure and fabrication pattern layer steps. After alignment, exposure of the substrate

28

a

with the mask pattern is implemented. The source

12

is activated to provide output light at a wavelength to which the photoresist is reactive. After the exposure, in one embodiment, a substrate unloader

518

is activated to unload the exposed substrate.

In one preferred embodiment, the chuck positioning structure and processor

504

is operable to switch the deformable chucks

502

a

and

502

b

between the exposure position and the flattening position. The system

500

simultaneously flattens one panel while it is aligning and exposing another. In operation, a substrate is first loaded onto a deformable chuck

502

located in the flattening position. The flatness sensor

508

and chuck

502

are then activated to flatten the substrate

28

. Next, the chuck positioning structure and processor

504

is activated to switch the positions of the chucks such that the substrate

28

which was just loaded is moved into the exposure position. The substrate

28

is then illuminated by the alignment pattern to align the mask pattern to the substrate. When the mask substrate alignment is complete, the substrate is exposed.

Next, in one embodiment, the substrate unloader

518

is activated to remove the substrate from the chuck while it is still located in the exposure position. In another embodiment, the chuck positioning structure and processor

504

is activated to move the substrate back to the flattening position where a substrate unloader unloads the substrate

28

and the substrate loader

506

loads the chuck with the next substrate to be exposed. Alternatively, the substrate loader and unloader can be combined into a single device which both loads and unloads substrates.

FIG. 11

is a schematic diagram of one embodiment of a deformable chuck

502

used in the photolithography system of the invention. In this embodiment, the chuck

502

comprises a base

520

on which are mounted plural controllable actuators

524

. Each actuator

524

has a variable length which is controllable by an applied control voltage. The actuators

524

can be piezoelectric transducers or a combination of a piezoelectric transducer and a stepper motor geared to a shaft, or simply a linear stepper motor. In either case, each actuator

524

is individually addressable by the controller

522

via the base

520

such that its length can be individually controlled. The controller

522

receives chuck alignment and flatness control inputs as shown in FIG.

10

and forwards appropriate actuator control outputs to the actuators

524

via the base

520

. The controller

522

flattens the top surface

29

of the substrate

28

by commanding the actuators

524

to change their lengths. This causes the mounting surface

526

to deform and hence in turn, causes the substrate

28

to deform to flatten the surface

29

.

The deformable chuck

502

includes a deformable mounting plate

526

on which the substrate

28

being patterned is mounted. The substrate

28

can be held to the mounting plate

526

via electrostatic attraction, in which case the substrate

28

and mounting plate

526

are charged at opposite polarities to attract them. Alternatively, the chamber

528

within the plate

526

can be evacuated by a vacuum system connected to a vacuum port

530

. In the case of vacuum hold down, the top surface

531

of the mounting plate

526

can be a porous metal or ceramic or it can be a metal or cermamic with machined channels such that the negative pressure generated within the chamber

528

holds the substrate

28

to the top surface

531

.

The chamber

528

in the mounting plate

526

preferably includes plural regions

519

defined by walls

521

within the chamber

528

. The walls

521

can be arranged to position the regions

519

in a predetermined pattern to optimize the vacuum hold down function. For example, the walls

521

can be arranged to form multiple concentric annular regions

519

. When the vacuum is activated, the region at the center of the chamber will pump down first since it has the smallest volume. The regions will successively pump down radially from the center. This ordering results in good contact between the substrate

28

and the top surface

531

of the mounting plate

526

.

The chuck

502

is mounted to a support structure

532

by a six-leg adjustable “hexapod”

534

. Each leg of the hexapod

534

is controllably deformable to change its length via control signals from the controller

522

. The controller

522

generates the hexapod control signals based on the alignment and flatness inputs shown in FIG.

10

. The six-leg hexapod

534

provides six degrees of freedom movement of the chuck

502

relative to the support structure

532

. Up and down, side to side, tilt and rotational movements can be implemented to position the substrate

28

relative to the projection optics as desired.

FIGS. 12A-12C

schematically illustrate flattening of a substrate

28

. In

FIG. 12A

, a substrate

28

is shown having an exaggerated bowed shape.

FIG. 12A

also shows the problems this can cause where the bow in the substrate is greater than the depth of field of the system, as illustrated. In this case, much of the pattern formed on the top surface

29

of the substrate

28

would be out of focus, resulting in failed or unreliable circuits or displays and undesirably large feature sizes.

FIG. 12B

shows the intermediate step of mounting the bowed substrate

28

on the chuck

502

and deforming the mounting plate

526

to hold the substrate

28

down. Finally, in

FIG. 12C

, the substrate

28

is shown with its top surface flattened by the deformable mounting plate

526

. As shown, the top surface

29

is flat across its entire surface within the depth of field of the system. Therefore, none of the pattern along the entire top surface

29

will be out of focus upon exposure.

FIG. 13

is a schematic diagram of the substrate flatness sensing system of the invention. A light source

546

emits light which is collimated by a lens

548

and projected onto a partially transmissive mirror

550

. Light from the mirror is directed onto a lens

552

which in turn directs the light onto a mirror

554

. The light is then directed through a Ronchi grating

540

which projects a pattern onto the surface

29

of substrate

28

mounted on the deformable chuck

502

. Light returning from the surface

29

of the substrate

28

passes through the Ronchi grating

540

, generating a Moire pattern which is reflected by mirror

554

through the lens

552

, and then through the partially transmissive mirror

550

onto the CCD detector

544

. The processor

512

reads the pixel data from the CCD detector

544

to analyze the Moire pattern. Based on the Moire pattern, the processor

512

generates and forwards flatness inputs to the chuck controller

522

. As described above, the controller

522

generates control signals used to deform the deformable actuators

524

of the chuck

502

to flatten the surface

29

of the substrate

28

.

FIGS. 14A and 14B

show two examples of portions of Moire patterns used in the flattening process.

FIG. 14A

shows a portion of a possible Moire pattern generated by a substrate surface which is not flat.

FIG. 14B

shows the same portion of the Moire pattern from the same surface after the flattening process has been completed. The processor

512

shown in

FIG. 13

reads the CCD pixels from the detector

544

and identifies the Moire pattern. As the substrate is deformed, and the pattern changes, the processor

512

continues to monitor the pattern. When the pattern approaches that of

FIG. 14B

, the flattening process is concluded.

FIGS. 15A and 15B

schematically illustrate the substrate alignment process of the invention.

FIG. 15A

illustrates a transparent substrate such as glass located over the top surface of the chuck

502

. The substrate has two opaque patterned regions

560

and

562

formed in its opposite corners. At the center of each opaque region is a transparent circular region

564

and

566

. As shown in the expanded view of the opaque region

562

, a CCD pixel array

568

is exposed through each transparent circular region when the substrate

28

is properly aligned over the chuck

502

.

FIG. 15B

shows alignment of the mask pattern with the substrate

28

. In

FIG. 15B

, the substrate

28

is illuminated with a mask pattern with light of a wavelength which does not react with the photoresist on the top surface of the substrate

28

. The mask pattern projects circular dots

570

and

572

in opposite corners of the pattern. When the pattern is properly aligned over the substrate

28

, the circular

570

and

572

should be located centrally within transparent regions

564

and

566

, respectively, as shown in the blowup insert. The CCD pixels are read by the controller

522

(

FIG. 11

) to determine the relative location between the projected circular dots

570

and

572

and their corresponding circular regions

564

and

566

, respectively, on the substrate

28

. Based on the pixel data, the controller

522

generates alignment control signals which are sent to the six-leg hexapod

534

to move the chuck

502

and hence the substrate

28

to the proper location under the mask pattern.

It should be noted that shapes other than circular can be used for the regions

564

and

566

. Also, the CCD pixel arrays can be optically coupled to the dots

570

and

572

by relay optics which can magnify the CCD elements.

While this invention has been particularly shown and described with references to preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.

Number	Name	Date
4302079	White	Nov 1981
4346164	Tabarelli et al.	Aug 1982
4391511	Akiyama et al.	Jul 1983
4475223	Taniguchi et al.	Oct 1984
4504045	Kenbo et al.	Mar 1985
4509852	Tabarelli et al.	Apr 1985
4666291	Taniguchi et al.	May 1987
4769860	Resor, III et al.	Sep 1988
4947413	Jewell et al.	Aug 1990
4974919	Muraki et al.	Dec 1990
5008702	Tanaka et al.	Apr 1991
5117255	Shiraishi et al.	May 1992
5162844	Ueda	Nov 1992
5170207	Tezuka et al.	Dec 1992
5184176	Unno et al.	Feb 1993
5220590	Bruning et al.	Jun 1993
5257139	Higuchi	Oct 1993
5285236	Jain	Feb 1994
5298939	Swanson et al.	Mar 1994
5309198	Nakagawa	May 1994
5315629	Jewell et al.	May 1994
5353322	Bruning et al.	Oct 1994
5363170	Muraki	Nov 1994
5559629	Sheets et al.	Sep 1996

Number	Date	Country
0608572A3	Dec 1993	EP
62-122126	Jun 1987	JP

Lithography system and method with mask image enlargement

Information

Patent Number

Date Filed

Date Issued

Inventors

Original Assignees

Examiners

Agents

CPC

US Classifications

Field of Search

US

International Classifications

Abstract

Description

Claims

US Referenced Citations (24)

Foreign Referenced Citations (2)

Non-Patent Literature Citations (5)

Entry
Betz, H., et al., “Possibility of enlarging x-ray exposure area with x-ray mirror or by electron beam vertical motion,” pp. 83-92.
Offner, A., “New Concepts in Projection Mask Aligners,” Optical Engineering, 14(2):130-132, (Mar.-Apr. 1975).
Bieber, M., et al., “Investigation Of X-Ray Exposure Using Plane Scanning Mirrors,” Nuclear Instruments and Methods, pp. 281-286, (1983).
Tatian, B., “Nonrotationally Symmetric Systems: A New Problem for Optical Design and Fabrication,” pp. 205-219.
Tatian, B., “Testing an unusual optical surface,” SPIE, vol. 554, International Lens Design Conference, (1985), pp. 139-147.