JPH1154397A - Aligner and method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve precision and productivity of lithography of spherical semiconductor, by installing an original sheet holding means for holding an original sheet having a pattern formed of chinoform, and an exposing means for exposing the surface of device material via the pattern formed of chinoform on the original sheet. SOLUTION: A plurality of Si balls 1 are tied up and sucked by a holder H outside an exposing machine. Positional relations between each of the Si balls 1 and right and left corresponding reference marks RM are measured by a detector outside the exposing machine. A plurality of the Si balls 1 sucked by the holder H are carried in the exposing machine, and sucked and held by a chuck C. Three-dimensional position of only the right and left reference marks RM are measured by a detector 3AS. In the state that the Si balls 1 are sucked by the holder H, the Si balls 1 are moved to under a reticle 2 by an XY stage XY/S. After the position and the attitude of the holder H in the axial direction is corrected, the Si balls 1 are exposed to a light of a light source 5 via an illumination optical system ILM.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an exposure apparatus and method, and more particularly to an exposure apparatus and method capable of suitably exposing a pattern on a spherical surface of a device such as a spherical semiconductor.

[0002]

2. Description of the Related Art Conventionally, a flat silicon wafer has been known. Recently, a spherical semiconductor device using ball-shaped silicon having a diameter of about 1 mm has been proposed. In this semiconductor device, various use forms can be changed by connecting a plurality of spherical silicons 1 with bumps B as shown in FIG.

The present applicant has previously proposed an exposure apparatus for manufacturing such a semiconductor device in Japanese Patent Application No. Hei 9-100178 (name: spherical device exposure apparatus and manufacturing method). According to this, as shown in FIG.
A pattern on the reticle SR, which is a hologram, is illuminated on the surface of the reticle SR by the illumination system 33 as reference light, and a desired circuit pattern, which is a reproduction pattern, is transferred, thereby enabling lithography of a spherical semiconductor device.

[0004]

However, in this apparatus, the reticle SR placed on the spherical silicon 1 needs to have a spherical shape.
There were some issues to be improved. For example,
The difficulty in designing the reticle SR is high, the cost of the reticle SR is high, the accuracy of the reticle SR is degraded at the time of pattern formation due to a shape error, and the like. The previous apparatus was not yet sufficient to obtain.

The present invention has been made in view of such circumstances, and an object of the present invention is to provide an exposure method capable of exposing a pattern on the surface of a spherical device material without using a spherical reticle (original plate). It is to provide an apparatus and a method.

[0006]

In order to achieve the above object, an exposure apparatus according to the present invention comprises: an original plate holding means for holding an original plate (reticle) having a pattern formed by kinoform; An exposure means for exposing the surface of the device material through a pattern formed by a foam is provided.

More preferably, the device material has a spherical shape, and the device material has a device holding means for holding at least half of the surface of the device material so as to be exposed by the exposure means. The means may expose at least a part of the non-original-side region of the surface of the device material through a parabolic mirror, or the exposing means may expose at least the non-original-side region of the device material surface. A part is exposed through an elliptical mirror.

Further, the exposure method of the present invention is characterized in that the surface of the device material is exposed to light from exposure means via a pattern formed by a kinoform on an original plate.

[0009] More preferably, the device material is spherical, and the device material is held such that at least a half of the surface of the device material is exposed by the exposure means. Further, the exposure means may expose at least a part of the non-original-side area of the surface of the device material via a parabolic mirror, or the exposure means may expose the non-original-side area of the device material surface. At least a portion may be exposed through an elliptical mirror.

Further, the kinoform on the original plate is formed such that a principal ray is perpendicular to each point on the surface of the device material, and the kinoform on the original plate is exposed to light on the surface of the device material. The illuminance unevenness can be reduced by emitting light out of the angle used for the original plate, and the kinoform on the original plate is a Genetic algorithm
Or the kinoform on the original plate is simulated annealing (Simulated Annealin
It may be optimized based on g). Further, the original plate may be a flat plate, and the kinoform may be formed on the flat original plate, and the exposure means may expose at least a part of a non-original-side region of the surface of the device material to a parabolic surface. The exposure may be performed through a mirror, and the exposure unit may expose at least a part of the non-original-side region of the surface of the device material through an elliptical mirror.

(Operation) According to the present invention, lithography of a spherical semiconductor with high accuracy and high throughput can be realized without using a spherical reticle SR.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail based on an embodiment shown in the drawings.

FIG. 1 shows an embodiment of the exposure apparatus of the present invention. In this figure, reference numeral 2 denotes a reticle formed of a kinoform as a pattern (circuit pattern) 2a.
A circuit pattern whose reproduction pattern of the kinoform is desired to be transferred to the surface of the spherical silicon 1 is created. The reticle 2 is a flat plate having an outer shape having a kinoform pattern surface 2a.

A parabolic mirror (or elliptical mirror) 3 is provided to enable transfer at a time to almost the entire surface of the spherical silicon 1, and the mirror 3 allows the reproduction light from the kinoform pattern surface 2a to be spherical. The light is guided to the front surface (back surface) of the silicon 1 on the non-reticle side.

The kinoform pattern surface 2a can be said to be a kind of a phase type hologram. However, since it is only phase information, a reproduced image including a noise component (Noisy) is obtained as it is.
Various methods for solving this problem have been proposed. For example, as described in the magazine "OplusE, November 1996, p.83", a simulated annealing method is used as an optimization algorithm.
g) Genetic algorithm Fourier iteration algorithm and the like. With these methods, theoretically 100%
A kinoform with a high efficiency becomes possible. However, some new evaluation criteria need to be added for use in this embodiment. It is Telecen condition and illumination unevenness.

First, the telecentric condition will be described. As shown in FIG. 3, one point of the spherical silicon 1 is irradiated with reproduction light of the kinoform to form a desired circuit pattern. A chief ray (light at the center of the exposure light beam) that hits one point is the spherical silicon 1 Must be perpendicular to this point (this is called the telecentric condition). When the principal ray (the light at the center of the exposure light beam) is inclined, a difference in the sensitivity of the resist on the spherical silicon 1 occurs, which causes a change in magnification with respect to a change in the outer dimensions of the spherical silicon 1.

Similarly, when the illuminance unevenness occurs, the dew conditions on the entire surface of the spherical silicon 1 are different, so that a uniform pattern cannot be formed. For this reason, for example, when the optimization is performed using a Genetic algorithm, it is necessary to perform the optimization in which the above-described telecentric condition and uneven illuminance are set as the priority evaluation items and the priority of the efficiency is reduced.

In particular, in order to satisfy the uneven illuminance, the light is diffracted by a kinoform outside the NA (opening through which the exposure light passes) used at the time of exposure, so that the entire surface (surface) of the spherical silicon 1 is completely irradiated from the reticle 2. Is optimized so that the diffracted light does not hit.

The pattern transfer exposure method has been described so far. Next, alignment will be described, and an embodiment of an actual exposure sequence will be described. This portion is the same as that disclosed in Japanese Patent Application No. Hei 9-100178.

First, the hardware for alignment in FIG. 1 will be described. In the present embodiment shown in FIG. 1, a plurality of (four in FIG. 1) spherical silicon 1 is adsorbed and supported on the tee T of the holder H. The alignment method of the present invention, which will be described below, by describing this suction state, is called a tee-up method.

The holder T is configured to be able to carry the spherical silicon 1 from outside the exposure apparatus while adsorbing the plurality of spherical silicons 1. The holder T is held on a chuck C by vacuum suction in an exposure machine as shown in the figure. A Z stage Z / S that can be driven Z (in the optical axis direction) below the chuck C along the optical axis direction of the illumination optical system ILM, and an XY stage with a laser interferometer XY / that can move along the XY plane below it. S is configured so that the spherical silicon 1 can be moved three-dimensionally together with the holder H.

In this embodiment, at least two three-dimensional detection systems 3AS are configured. Although only one is shown in FIG. 1, another is formed in the vertical direction (Y direction) of the paper surface. Practical embodiments of the three-dimensional detection system 3AS are shown in FIGS.
2 may be the same as the three-dimensional detection system 3AS of the alignment station AS described later.

In FIG. 1, reference numeral 5 denotes a light source for generating illumination light for exposure, such as an excimer laser. I
Reference numeral LM denotes an illumination optical system that illuminates the reticle 2 with exposure light from the light source 5 so as to satisfy a predetermined condition, and reference numeral 30 denotes a reticle holder serving as original plate holding means for holding the reticle 2 at a position shown in the figure.

FIG. 5 shows four spherical silicons 1 (1-1, 1).
1-2, 1-3, and 1-4) are enlarged views of the holder H that has been teeed up at intervals L, and FIG. 6 is a view of the holder H viewed from above. Reference marks R are provided on the left and right of each spherical silicon 1 in FIG.
M (RM1-1R, L; RM1-2R, L; RM1-3
R, L; RM1-4R, L). FIG. 7 is an enlarged view of the reference mark RM. Each of the reference marks RM has a mark RM-X for measuring an X shift and a mark RM-Y for measuring a Y shift.

Next, the alignment sequence will be described. FIG. 8 shows a schematic flowchart of the alignment sequence. As shown here, in this embodiment, S1) outside the exposure machine, a plurality of spherical silicons (in the schematic flowchart of FIG. 8, spherical silicon [Ball Semiconducto
r] is described as BS) 1 is tee-adsorbed to the holder H.

S2) The positional relationship between each spherical silicon 1 and the corresponding left and right reference marks RM is measured outside the exposure machine by a detector (not shown). In this measurement, the positional relationship is measured in each of the X, Y, and Z axial directions and the six axial directions of the rotational components ωx, ωy, and ωz around each axis.

S3) After the measurement of the positional relationship, the plurality of spherical silicon pieces 1 are transported to the exposing machine while being suction-adsorbed to the holder H, and are adsorbed and held by the chuck C.

S4) The exposure device measures the three-dimensional position of only the left and right reference marks RM using the detector 3AS.
Since the three-dimensional position is controlled, the above-described displacement in the six-axis direction is measured.

S5) The XY stage XY / S with a laser interferometer is moved under the reticle 2 while the spherical silicon 1 is tee-adsorbed onto the holder H, and then, based on the relative relationship information measured in steps S2 and S4. The position and orientation of the holder H in the 6-axis direction are corrected by each stage that moves the chuck C, and after this correction, the reticle 2 is illuminated with the exposure light from the light source 5 via the illumination optical system ILM to expose the spherical silicon 1. Perform

S6) By performing steps S4 and S5 on each spherical silicon 1 on the holder H, all the exposures are completed. Alignment and exposure can be performed by the following sequence.

Steps S3 through S9 shown in FIGS.
17 to 18 show the details of steps S1 to S2 shown in FIGS. FIG. 9 is a view showing the arrangement of the holder H and the exposing machine, and showing the start arrangement of the alignment.
The leftmost spherical silicon 1-1 has moved to the lower part of the three-dimensional detection system 3AS.

FIG. 10 shows the spherical silicon 1-1 at the left end.
It is a figure which shows the position detection in the Z direction (upper part in a figure) of the holder H part of the vicinity. Another three-dimensional position detection system 3AS is provided in the direction perpendicular to the paper surface, and simultaneously detects the left and right reference marks RM1-1L and RM1-1R on the holder H.

Next, as shown in FIG.
Is driven to drive the holder H to the left in the figure as shown in FIG. 12, and the leftmost spherical silicon 1-1 is positioned below the exposure machine (exposure position).

Next, as shown in FIG. 13, the holder H is driven upward (in the Z direction) in the figure to perform pattern exposure on the leftmost spherical silicon 1-1. This drive amount reflects the correction value based on the relative relationship information in step S2 as described in step S5 described above. At this time, 3
When the distance between the measurement position of the three-dimensional position detection system 3AS and the exposure position of the exposure device on the XY plane is set to be equal to the distance L of each spherical silicon 1, pattern exposure on the spherical silicon 1-1 is performed. At the same time, the left and right reference marks RM1-2L, RM1 on the holder H near the adjacent spherical silicon 1-2.
-2R three-dimensional position detection can be performed.

The allowable error between the distance between the measurement position of the three-dimensional position detection system 3AS and the exposure position of the exposure device and the distance L of each spherical silicon is as follows. If the detection range in the XY direction of the three-dimensional position detection system 3AS is larger than the sum of the distance between the three-dimensional position detection system 3AS and the exposing machine including the creation error and mounting error of the distance of the tee T in H, and the change with time, good. The point is that the reference mark RM of the next spherical silicon can be measured while the previous spherical silicon 1 is being exposed.

After the exposure of the spherical silicon 1-1 and the detection of the three-dimensional position are completed, as shown in FIG.
Is positioned below the exposure device, and the adjacent spherical silicon 1-2 is positioned below the three-dimensional position detection system 3AS. Next, FIG.
The holder H is moved to the left (X direction) in the figure as shown in FIG. 5, and the adjacent spherical silicon 1-2 is positioned below the exposure position of the exposure machine. As shown in FIG. 16, the reference exposure marks RM1-3L and RM1-3 on the right and left sides of the holder H near the adjacent spherical silicon 1-3 together with the pattern exposure on the adjacent spherical silicon 1-2.
R three-dimensional position detection is performed.

The procedure for detecting the three-dimensional position at this time is as follows.
This is performed using FIGS. By repeating the above sequence, alignment and exposure of a plurality of spherical silicon 1 can be performed.

Next, an embodiment will be described in which a three-dimensional positional relationship between the holder H and the plurality of spherical silicons 1 is determined before exposure. FIG. 19 shows a flow that describes the flow from adsorption to exposure. In this example, it is divided into three, and the suction station CS and the alignment station A are respectively in order.
S, exposure station ES.

The adsorption station CS is a place where a plurality of spherical silicons 1 are adsorbed on the tee T of the holder H. Photoresist or the like, which is a photosensitive agent, is coated on the surface of the plurality of spherical silicon 1 for exposure, placed on the tee T of the holder H, and then supported by vacuum or electrostatic suction.

After the tee-up suction, the holder H is moved from the suction station CS to the alignment station AS and from the alignment station AS to the exposure station ES. Here, the accuracy of adsorbing the plurality of spherical silicon 1 on the portion of the tee T of the holder H is such that the error of the rotational component of the three-dimensional (X, Y, Z) axis is sufficiently negligible from the pattern forming accuracy of the spherical silicon 1. It is assumed that by setting the value to be a small value, only an error in each of the X, Y, and Z axis directions occurs.

The holder H is moved from the suction station CS to the alignment station AS with the tee-up suction. The alignment station AS has a Z stage Z that can drive the holder H in a three-dimensional direction.
/ S, XY stage XY / S and at least one three-dimensional detection system 3AS.

FIGS. 17 and 18 are views for explaining the measurement of the three-dimensional positional relationship between the spherical silicon 1 and the holder H that is adsorbed by teeing up the spherical silicon 1 to the tee T of the holder H at the alignment station AS. It is.

FIG. 17 shows an alignment station AS.
FIG. 3 is a diagram showing a state in which a three-dimensional position near the top of the spherical silicon 1-1 is detected by a three-dimensional detection system 3AS in FIG. FIG. 18 shows a three-dimensional detection system 3AS with spherical silicon 1-.
Reference mark RM1-1 on the left side of the holder H near 1
FIG. 13 is a diagram illustrating a three-dimensional position detection state of L.

In the example of FIGS. 17 and 18, the three-dimensional detection system 3A
S performs three-dimensional detection as follows. The light from the halogen lamp 11 and the light passing through the reflecting mirror 12 are wavelength-selected through the wavelength selection filters 13a (long wavelength side) and 13b (short wavelength side), and thereafter enter the fiber.

The light emitted from the exit end of the fiber 14 is supplied to a condenser lens 15, an aperture stop 16, a field stop 17,
Dichroic mirror 18, illumination system lens 19, prism 2
After passing through a 0, 1/4 wavelength plate (composed of 21a and 21b) and the objective lens 101, in FIG. 17, an alignment mark (not shown but having the same shape as the reference mark RM in FIG. 11) near the top of the spherical silicon 1-1 is shown. In FIG. 18, the reference mark RM on the holder H near the spherical silicon 1-1 is shown in FIG.
Illuminate 1-1L.

The light reflected by the mark is transmitted to the objective lens 10
When the polarization state changes via the 1/4 wavelength plates 21a and 21b, the polarization state is reflected on the surfaces 20a and 20b of the prism 20, and is reflected via the relay lens 22 and the erector lens 23.
An image is formed on a two-dimensional image sensor 24 such as a CD camera. In the two-dimensional image sensor 24, the alignment mark (not shown) near the top of the spherical silicon 1-1 or the spherical silicon 1
The position shift in the X direction and the Y direction is detected from the image position of the reference mark RM1-1L in the holder H portion near -1.

A so-called image processing alignment method for photoelectrically converting and processing the image of this mark is disclosed in, for example,
This technology is proposed in Japanese Patent Application Publication No. 1820, has been actually commercialized, and its effects have been confirmed.

Next, position detection in the Z direction will be described with reference to FIGS. The light emitted from the Z-direction detection unit 25 is reflected by the dichroic mirror 18 and is provided with an illumination system lens 19, a prism 20, quarter-wave plates 21a and 21b, and an objective lens 101.
17 illuminates an alignment mark (not shown) near the top of the spherical silicon 1-1 in FIG. 17, and FIG. 18 shows a reference mark RM on the holder H near the spherical silicon 1-1.
Illuminate 1-1L.

The light reflected by the mark is transmitted to the objective lens 10
Prism 2 via 1, 1/4 wavelength plates 21a, 21b
0 are incident on the surfaces 20a and 20b. The characteristics of the surfaces 20a and 20b of the prism 20 are set as a polarizing beam splitter at the above-described wavelength for position detection in the XY directions, and as characteristics of transmitting at the wavelength for position detection in the Z direction. Then, the light transmitted through the prism 20 is reflected by the dichroic mirror 18 and
The light enters the direction detection unit 25 to perform position detection (Z direction).

When only one three-dimensional detection system 3AS is provided, the reference mark RM1-1R on the right side of the holder H in the vicinity of the spherical silicon 1-1 is similarly set on the XY stage XY.
/ S is moved in the direction perpendicular to the plane of the drawing, and the reference mark RM on the left
What is necessary is just to carry out similarly to 1-1L.

As described above, the holder H and the spherical silicon 1
The three-dimensional relative positional relationship with -1 is known. By repeating this similarly for the spherical silicon 1-2 and thereafter, the three-dimensional relative positional relationship between the plurality of spherical silicon 1 adsorbed on the holder H and the (reference mark RM of) the holder H can be determined. The holder H is moved from the alignment station AS to the exposure station ES while being stored in the memory of the connected computer and the plurality of spherical silicons 1 are sucked up in a tee-up manner, and the exposure is performed as described above.

In the case where another pair of the X, Y, and Z position detection systems 3AS are provided in the direction perpendicular to the paper of FIGS. 17 and 18, a spherical silicon 1-1 is used as the detection system.
Left and right reference marks RM1-1 in the vicinity of the holder H
L and R can be measured without moving the XY stage XY / S in the direction perpendicular to the paper surface.

In the embodiment shown in FIG. 1, an image processing system is adopted as the two-dimensional position detection system, but the present invention is not limited to this. As another method, a method of relatively scanning a laser beam or a heterodyne interference method may be used. The present invention can be similarly applied by adopting a position detection method capable of two-dimensionally measuring the relative relationship between the spherical silicon and the reference mark on the holder with or without the use of the alignment mark. Can achieve the purpose.

The same applies to the detection method in the Z direction, and is not limited to the beam shift method by oblique irradiation employed as the embodiment of the present invention in FIG.
An astigmatism method or a method using a capacitance sensor may be used. Further, it is not necessary to use the same three-dimensional position detection method in the alignment station AS and the exposure station ES.

In the exposure station ES, the three-dimensional position of each reference mark on the holder H has been measured. However, at least two reference marks may be measured by the holder H. At that time, the six-axis position control of each spherical silicon at the exposure station ES may be performed based on the measurement result of the relative relationship between each spherical silicon 1 and the reference mark on the holder H at the alignment station AS. However, the alignment accuracy at this time depends on the XY stage XY / S accuracy at the time of the alignment station AS and the XY stage XY at the exposure station ES.
Both / S precision affects.

[0056]

As described above, according to the present invention, an exposure apparatus and method suitable for manufacturing a spherical device can be provided.

[Brief description of the drawings]

FIG. 1 is a diagram showing one embodiment of the present invention.

FIG. 2 is a diagram showing an example previously proposed by the present applicant.

FIG. 3 is a diagram showing that a pattern is formed by telecentricity.

FIG. 4 is a diagram showing an example of a semiconductor device in which a plurality of spherical silicons are connected via bumps.

FIG. 5 is an explanatory view of a holder that tees up and adsorbs spherical silicon.

FIG. 6 is a diagram showing a position detection mark of a holder.

FIG. 7 is a diagram showing a position detection mark in the XY directions.

FIG. 8 is a view showing a procedure for aligning spherical silicon.

FIG. 9 is a diagram showing a start arrangement of a holder and an exposure machine.

FIG. 10 is a diagram showing a position detection state of a holder portion near the leftmost spherical silicon.

FIG. 11 is a diagram showing a state where the leftmost spherical silicon is positioned below the alignment detection system.

FIG. 12 is a diagram showing a state in which the leftmost spherical silicon is positioned below an exposure machine.

FIG. 13 is a diagram showing the state of pattern exposure on the leftmost spherical silicon and the detection of the position of the holder near the adjacent spherical silicon.

[FIG. 14] The leftmost spherical silicon is placed below the exposure machine.
The figure which shows the state which located the adjacent spherical silicon below the alignment detection system.

FIG. 15 is a diagram showing a state where an adjacent spherical silicon is positioned below an exposure machine.

FIG. 16 is a diagram showing a state of pattern exposure on an adjacent spherical silicon and detection of the position of a holder portion near the adjacent spherical silicon.

FIG. 17 shows a three-dimensional detection system in which 3 is located near the top of spherical silicon.
The figure which shows a dimension position detection.

FIG. 18 is a diagram showing three-dimensional position detection of a holder portion near spherical silicon in a three-dimensional detection system.

FIG. 19 is a conceptual diagram showing a flow of exposing spherical silicon to a holder, measuring a relative positional relationship with the holder, and performing exposure.

[Explanation of symbols]

 Reference Signs List 1 spherical silicon 2 reticle B bump 2a kinoform surface 3 parabolic mirror (elliptical mirror) H holder T tee 5 exposure light source ILM illumination optical system XY / S XY stage with interferometer Z / S Z stage C holder 11 halogen lamp 12 Reflector 13a Wavelength selection filter (long wavelength side) 13b Wavelength selection filter (short wavelength side) 14 Fiber 15 Condenser lens 16 Aperture stop 17 Field stop 18 Dichro mirror 19 Illumination lens 20 Prism 24 Two-dimensional imaging device such as CCD camera 25 Z direction detection unit Reference mark on RM holder 3AS 3D detection system

Claims (15)

[Claims] 1. An image forming apparatus comprising: an original holding means for holding an original having a pattern formed of kinoform; and an exposing means for exposing a surface of a device material through a pattern formed by kinoform on the original. Exposure equipment characterized. 2. The exposure apparatus according to claim 1, wherein the device material has a spherical shape, and has device holding means for holding an area of at least half of the surface of the device material so as to be exposed by the exposure means. 3. The exposure apparatus according to claim 2, wherein the exposure unit exposes at least a part of a region of the surface of the device material on the non-original plate side via a parabolic mirror. 4. An exposure apparatus according to claim 2, wherein said exposure means exposes at least a part of a region of the surface of said device material on the non-original plate side via an elliptical mirror. 5. An exposure method, wherein a surface of a device material is exposed to light from exposure means via a pattern formed by a kinoform on an original plate. 6. The exposure method according to claim 5, wherein the device material has a spherical shape, and the device material is held such that an area of at least half of the surface of the device material is exposed by the exposure unit. 7. The exposure method according to claim 6, wherein said exposure means exposes at least a part of a region of the surface of the device material on the non-original plate side via a parabolic mirror. 8. The exposure method according to claim 6, wherein said exposure means exposes at least a part of a region of the surface of said device material on the non-original plate side via an elliptical mirror. 9. The exposure method according to claim 6, wherein the kinoform on the original plate is formed such that a principal ray is perpendicular to each point on the surface of the device material. 10. The exposure method according to claim 9, wherein the kinoform on the original plate emits light out of an angle used for exposure to the surface of the device material to reduce uneven illuminance. 11. The exposure method according to claim 6, wherein the kinoform on the original plate is optimized based on a Genetic algorithm. 12. The exposure method according to claim 6, wherein the kinoform on the original plate is optimized based on simulated annealing. 13. The exposure method according to claim 6, wherein the original plate is a flat plate, and the kinoform is formed on the flat original plate. 14. The exposure method according to claim 6, wherein said exposure means exposes at least a part of a region of the surface of said device material on the non-original plate side via a parabolic mirror. 15. The exposure method according to claim 6, wherein said exposure means exposes at least a part of a region of the surface of said device material on a non-original plate side via an elliptical mirror.