GB2138163A

GB2138163A - Optical projection system

Info

Publication number: GB2138163A
Application number: GB08402807A
Authority: GB
Inventors: Akiyoshi Suzuki; Michio Kohno
Original assignee: Canon Inc
Current assignee: Canon Inc
Priority date: 1983-02-07
Filing date: 1984-02-02
Publication date: 1984-10-17
Also published as: JPS59144127A; GB2138163B; JPH059934B2; DE3404063A1; GB8402807D0; DE3404063C2

Abstract

An optical device in which the distortion of image is compensated for is provided with an imaging optical system for forming the image of an object on the image plane and for determining an imaging optical path, and a compensating thin plate having such a degree of optical thickness which substantially does not impart other influence than shifting the imaging position to the imaging performance and assuming a two-dimensionally or three-dimensionally curved shape, the thin plate being disposed in the imaging optical path, and the curvature being variable to slightly affect the magnification. Other embodiments show the use of variable thickness or variable refractive index plates.

Description

SPECIFICATION Optical device in which the distortion of image is compensated for Background of the invention Field of the invention This invention relates to an imaging optical device, and in particular to an optical device capable of adjusting the imaging magnification error. The present invention is particularly effective for modifying any lateral magnification error occurring in a precision optical system wherein semiconductor integrated circuit patterns are projected and imaged upon wafers.

Description of the prior art In recent years, the demand for making semiconductor integrated circuits minute has ben rapidly increasing and the dimension of the minute patterns in electronic circuits is close to 1 Fm. As a process for manufacturing such highly integrated elements there is the photo-process in which mask patterns are transferred onto wafers.

The so-called semiconductor printing apparatus is an apparatus used to transfer such mask patterns and there are several types of this apparatus such as, for example, the contact exposure system in which printing is effected with a mask and a wafer being brought into contact with each other, the proximity exposure system in which printing is effected with a mask and a wafer spaced apart from each other by several Fm, and the optical projection exposure system in which printing is effected by the use of a lens or a mirror.

Generally, a semiconductor integrated circuit is not of a single-layer structure but tends to have more layers as the degree of integration becomes higher. Therefore, in the photo-process, it is necessary to use the above-mentioned several printing systems properly and print different mask patterns successively on a wafer with high dimensional accuracy. On the other hand, the diameter of wafers is becoming larger to enhance the productivity of circuits. At present the diameter of five inches is becoming the main current.

However, when the image of a mask is to be transferred onto such a large wafer, a combination of various factors may sometimes cause a phenomenon that however alignment operation is effected, the mask and the wafer cannot be driven into a desired positional relation up to a minute amount of 1 Fm or less. That is, the image of the mask and the ready-made pattern on the wafer deviate from their regular relation.

The manners in which such lateral displacement occurs are multifarious, but most of them are so-called magnification error, and magnification error may be expressed as a linear function of position. However, if this magnification error enters the mask image, the image of the mask on the wafer will be transferred while being enlarged or reduced with respect to the regular dimensions as depicted in Figures 1A and 1 B of the accompanying drawings and as a result, the coupling with the ready-made pattern wil not go on successfully and a defective product will result. In the Figures, the dimensions of arrows represent the magnitude of the positional deviation.

Generally considering the causes of magnification error, a first cause is the magnification difference of the mask image which depends on the difference between the above-mentioned various projection systems, and a second cause is the magnification difference inherent to respective apparatuses even if they are of the same type. Besides these, expansion or contraction of the wafer resulting from a temperature change during the working process is an action equivalent to magnification error.

Entry of an error into the mask image is inconvenient to any system, and is more serious in the mirror projection exposure system. Thus, this system will hereinafter be described.

What is herein called the mirror projection exposure system is a system which uses an aberration-corrected, off-axis arcuate region in an optical system constituted chiefly by mirror surfaces to print a mask pattern on a wafer. In this system, the superior image region of the optical system is arcuate.

Therefore, to print a mask of a large area on a wafer, the mask M and the wafer W are caused to scan in synchronism with each other in a direction perpendicular to the direction of the arcuate region (T in Figure 2 of the accompanying drawings) projected onto the wafer.

A construction as an exmple for specifically showing the above-described construction is shown in Figure 3 of the accompanying drawings. In Figure 3, IS designates an illuminating system and AS denotes an alignment system for the mask and wafer. PS designates a mirror optical system, wherey the pattern of the mask m is transferred onto the wafer W. J denotes a frame for holding the mask M and the wafer W integrally with each other. The frame J is moved while being guided by a guide device (K.G.) such as a gas slider. D designates a driving device for moving the frame J.

The mirror optical system PS is a system symmetrical with respect to the optic axis 00' thereof, and aberrations therein are corrected at an image height R. The system is designed to transfer the minute pattern of the mask onto the wafer by the use of an arcuate region including a part of the circumference of radius R which is the superior image region. Also, as the imaging relation, if attention is paid to a light flux diverging from a point on the mask M, this light flux is reflected by a planar mirror B51, whereafter it is reflected by a concave mirror M1 and a convex mirror M2 performing the function of a stop and again by the concave mirror Ml, and has its optical path bent by a planar mirror BS2 and converges at a point on the wafer W.

The magnification errors occurring in this apparatus are divided into ones optically caused in the projection optical system and ones attributable to the mechanical feeding error when the mask and wafer are scanned, and the difficulty of the latter has already been solved as will later be described.

On the other hand, the following cause may be considered to be the cause of the former.

As previously described, wafers having a diameter of five inches are used nowadays in keeping with the tendency of the wafers toward greater diameters. To print a wafer of such size in a mirror projection exposure system, the diameter of the concave mirror M1 must be about 400 mm. To transfer a minute pattern of about 1 Ccm faithfully onto the wafer while meeting such tendency of the mirror toward a greater diameter, the distortion of the curved surface after polishing of the mirror surface must be kept at high accuracy below 1/10 of the wavelength. This is because, as indicated by dotted line in Figure 3, if surface distortion is present at the position on the mirror surface on which a light flux impinges, the lightfluxwill travel while deviating from the direction in which it should originally be reflected.As a result, the image is formed on the wafer at a position displaced from the regular position in a direction perpendicular to the optic axis and thus, a magnification error is created in the image.

In the scanning type mirror optical system, the error of this kind occurs on the wafer chiefly in a direction orthogonal to the scanning direction y, and this is schematically indicated by small arrows in Figure 4A of the accompanying drawings. In contrast, the dimensional error attributable to the second mechanical cause occurs with movement of the mask and wafer and, as depicted in Figure 4B of the accompanying drawings, it is created chiefly along the scanning direction y.

Thus, if these two magnification errors are eliminated independently, the superposition-printing accuracy of integrated circuits will be improved over the entire surface of the wafer and mixing of this system with other printing apparatus will become possilbe.

Particularly, this mirror optical system is telecentric on both the object side and the image plane side and has an optical characteristic that the imaging magnification thereof is not varied at all even if the position of the mask or the wafer which is the object in the direction of the optic axis is varied. Accordingly, the magnification error in a direction perpendicular to the scanning direction is determined by the optical system, and means for correcting this is not yet known.

On the other hand, chief ones of the mechanicai causes of magnification error are considered to be the lowness of the assembly and working accuracy of the machine. For example, when, as depicted in Figure 5 of the accompanying drawings, a sliding member K on which a mask and an wafer rest is moved along a guide surface G while being supported by static gas bearings bl and b2, if the guide surface is upwardly convex, the sliding member K will become inclined before and afterthe scanning range. Thus, the surface of the wafer also will become inclined and this will cause the error in the form shown in Figure 4B.As a method for preventing this, the pressure supplied to the bearing b1 may be increased at the front end of the scanning range and the pressure supplied to the bearing b2 may be increased at the rear end of the scanning range, whereby the pressures may be controlled so as to ensure the sliding member K to be moved horizontally.

Besides this, there is a system for eliminating the magnification error by temperature-controlling the wafer or a system for correcting the magnification error by moving an optical member in the optical system which contributes to the imaging characteristic. However, in the system using the temperature control, the-wafer expands or contracts radially relative to the center thereof and therefore, the magnification error cannot be corrected independently in the scanning direction and in a direction perpendicular thereto. Moreover, this system has a disadvantage that the utilization of heat conduction leads to consumption of much time. Also, the system for correcting the magnification error by moving the optical member in the optical system which contributes to the imaging characteristic involves a danger that the imaging performance itself of the projection system is aggravated.

Summary of the invention It is a first object of the present invention to adjust a magnification error created in an object image.

It is a second object of the present invention to compensate for a distortion created one dimensionally in a projected image.

it is a third object of the present invention to compensate for a distortion created two dimensionally in the projected image.

It is a fourth object of the present invention to vary the degree of compensation of the# distortion of the image in conformity with the resolving power required of the projected image.

To achieve these objects, in an embodiment of the present invention which will hereinafter be described, a transparent thin film member which hardly affects the other imaging performances of the projection optical system than the deviation of the imaging position is inserted in an optical path which links an bject with the image thereof, and more desirably in the vicinity of the object or the image or the intermediate imaging plane, and this portion is curved line- or point-symmetrically, whereby the imaging position in a direction orthogonal to the optic axis is modified.

The invention will become more fully apparent from the following detailed description thereof taken in conjunction with the accompanying drawings.

Brief description of the drawings Figures 1A and IB are plan views schematically showing magnification errors.

Figure2 is a plan view showing the image on a wafer.

Figure 3 is an optical cross-sectional view showing an embodiment of the present invention.

Figures 4A and 4B are plan views schematically showing magnification errors.

Figure 5 is an auxiliary illustration of a scanning mechanism.

Figure 6 is an enlarged view of the essential portions of the embodiment.

Figure 7 is a transverse optical cross-sectional view of the embodiment.

Figure 8 is a graph in which radius and thickness are variables.

Figure 9 shows a modification of the essential portions.

Figure 10A is a plan view of a constituent member.

Figure lOB is a perspective view of the constituent member.

Figure 11 is a graph showing the position and the amount of deviation.

Figures 12 and 13 are perspective views of the optical systems.

Description of the preferred embodiment An embodiment of the present invention will hereinafter be described. Most of the portions constituting the optical system of Figure 3 have already been described. That is, M designates a mask and W denotes a wafer. These are moved together in the scanning direction y. S designates a light-intercepting plate provided with an arcuate opening of the shape as shown at T in Figure 2 and secured at a position proximate to the wafer. BS1 and BS2 denote optical path changing mirrors, and a concave mirror M1 and a convex mirror M2 are disposed on the optic axis 00' with their centers of sphere deviated from each other.The central light ray from the illuminated mask M is reflected by a mirror, whereafter it is reflected by the concave mirror M1, the convex mirror M2, the concave mirror M1 and the mirror BS2 in succession and is incident on the wafer W.

Next, a member B secured to the vicinity of the wafer W is a member peculiar to the present invention and comprises a thin member having such a degree of thickness which does not affect the imaging performance and curved into a semi-cylindrical shape. The curved thin film B is disposed so that the bus line thereof is coincident with the scanning direction y, and its appearance in a direction perpendicular to the plane of the drawing sheet is as shown in Figures 6 and 7.

An optical thin film as generally represented by a pellicle can be optically regarded as a very thin parallel surface. It is known that this optical member is very thin and has no power and therefore hardly deteriorates the imaging performance. Thus, even if a transparent optical member having an inclination relative to the light ray which provides the center of the light flux (principal ray), for example, an optical thin member curved in the direction of the optic axis and having the bus line of its cylindrical surface intersecting the optic axis of the imaging system, is newly inserted into the vicinity of an object or the image plane, the degree to which this deteriorates the imaging characteristic on the image plane is small. Nevertheless, the principal ray travels while being refracted by the thin member B as shown in Figure 6 in accordance with the Snell's law.

For the reason set forth above, in a semiconductor printing apparatus, by inserting a moderately curved, transparent optical thin member into the vicinity of a mask which is the object surface in the printing optical path thereof or a wafer which is the image plane, it is possible to move the printed image on the wafer without deteriorating the imaging characteristic thereof.

The optical system of the semiconductor printing apparatus is usually caled a telecentric system and is designed such that with respect to all points in the printing picture plane, the principal ray is projected perpendicularly to the surface of the wafer. This leads to a merit that even when the wafer is defocused, the center of the light flux does not cause lateral deviation on the wafer and therefore no magnification error is caused.

In Figure 6, if the thickness of the inserted optical thin fim is d mm and the refractive index thereof is n and the direction in which it is desired to laterally deviate the printed image on the wafer is x-axis and the direction of the principal ray perpendicular thereto is z-axis, the amount of curve z mm of the inserted optical thin flim in the direction z may be expressed as a function of x.

The amount of lateral deviation Axmm in the direction x obtained at this time is

Particularly, where the optical thin member is extended in the shape of a spherical surface having a radius Rmm,

and therefore, equation (1) becomes:

When n = 1.5 and d/R is varied, there is obtained Figure 8. If this graph is used, d/R for the amount of deviation to be modified can be found. In the experiment of the embodiment, a good result was obtained by the use of a pellicle having a thickness of 10 um, but a thicker pellicle or a thinner pellicle can be suitably used in conformity with the performance of the optical system.However, in a precision optical system, it is a standard to keep the difference between the wavefront aberration before the thin member is inserted and the wavefront aberration after the thin member is inserted so as to be between X (wavelength) to -A.

Turning back to Figure 6, suppose principal rays incident on the points a, b, c, d and e on the wafer W when the thin member is absent. When the thin member B is inserted so that its concave surface faces the wafer W, each ray is refracted and imaged at a', b', d', and e'. As a result, the imaging magnification can be reduced in the direction of x-axis.

Conversely, when the thin member B is inserted so that its convex surface faces the wafer Was shown in Figure 9, the refracted rays are imaged at a", b", d" and e". Accordingly, the image can be enlarged in the direction of x-axis.

That is, if the numerical value is determined by the direction of curve of the thin member B and equation (2) above, any magnification error can be coped with.

Since the curved thin member B is formed in a semi-cylindrical shape, it acts as a mere parallel flat plate with respect to a direction perpendicular to the plane of the drawing sheet, i.e., a direction parallel to the bus line, and does not affect the magnification in this direction. In the foregoing description a case where the magnification adjusting member is inserted into the wafer side B of Figure 3 has been supposed, but just the same effect can theoretically be obtained even if the magnification adjusting member is inserted into the mask side B'. In this case, however, the relation of enlargement and reduction differs from that in the case where the magnification adjusting member is inserted into the wafer side B.That is, when the magnification adjusting member is inserted so that it is convex with respect to the wafer in B, the printed image of the wafer is enlarged, whereas when the magnification adjusting member is inserted so that it is convex with respect to the mask in A, the printed image of the wafer is reduced. Further, the thin member may be moved between the optical path changing mirror BS1 or BS2 and the convex mirror M1 and, if an intermediate imaging plane is present in the optical path, the thin member can be disposed in the vicinity thereof.

As described above, a system whereby an optical thin film member attached to a semi-cylindrical curved metal frame is inserted into the optical system PS may be mentioned as a first embodiment of the present invention. By the insertion of this member, it becomes possible to correct the delicate magnification light between numerous machines. Also, where there is expansion or contraction of the wafer due to the processing thereof, the correcting member of the present invention may be interchanged in conformity with the amount of the expansion or contraction, and if the curvature thereof is made changeable, it will be possible to freely control any change in magnification by a single member.

The correcting member including a frame carrying the optical thin member will now be described more specifically. First, there may be mentioned an example in which, as shown in Figures 10A and 10B, an optical thin film having a thickness of several microns is stretched over a metal frame curved into a semi-cylindrical shape and the curvature of the circumferential portion of the metal frame is varied. Specifically, an arcuate opening AP which will sufficiently pass a light flux therethrough is provided in a rectangular metal frame and forces are applied thereto from the opposite sides thereof to thereby make the metal frame into a semi-cylindrical shape. An optical thin film is stretched over this metal frame and, if the force applied to the metal frame is adjusted, the curvature of the circumferential portion of the semi-cylinder can be changed freely.Therefore, the magnification of the printed image on the wafer can be corrected. An example in which the thin film is stretched into other ordinary shape than a mere semi-cylindrical shape may be mentioned as a a third embodiment. That is, where the thin film is stretched into a semi-cylindrical shape, the amount of correction of the position on the surface of the wafer is a primary function of the position on the wafer as is shown in Figure 8. However, among the printed image deviations that are actually occurring, there are ones which may be represented by a secondary or further function of the position. To eliminate these image deviations, it is necessary to curve the optical member into a proper shape whih will satisfy equation (1).For example, where the image deviation is represented as a secondary or tertiary relation of the position (Figure 11), if this is to be corrected, it is seen from equation (1) that the amount of curve of the thin film must follow a tertiary or quaternary function of the position.

Further, it is also possible to curve the optical thin member three-dimensionally to thereby effect the correction of the image magnification two-dimensionally. In figure 12, a thin member formed into a semi-spherical shape indicated as B" is inserted into a projection exposure apparatus using a projection lens PL with the center of the sphere thereof being coincident with the optic axis, thereby correcting the magnification. In an apparatus using a lens, printing is effected in the area about the optic axis and therefore, the correcting means assumes a three-dimensional semi-spherical shape. The semi-spherical member can be realized by a solid material such as glass or plastics or may also be realized by stretching a film of nitrocellulose in an air-tight box and utilizing the difference in air pressure to inflate the film into a semi-spherical shape.

The technical idea of the present invention is also usableforthe correction of minute distortion of a lens optical system. For example, in a lens system, there may sometimes occur a minute magnification difference between a specific direction which is the accumulation of the working error and assembly error of glass and a a direction orthogonal thereto. Even in such a case, it is possible to adjustthe magnification difference by inserting the hitherto described optical member into the optical path and curving that member into a cylindrical shape. At this time, the bus line of the cylinder is set in a direction substantially coincident with one of two directions in which the magnification difference is occurring most greatly.This state is shown in Figure 13, wherein an optical member CL corrects anisotropic magnification. This correction may of course be effected on the side of the mask (or reticle) indicated as M.

In the above-described example, an optical thin film having a uniform refractive index and a uniform thickness has been used while being curved, but alternatively, a parallel flat plate having a refractive index continuously varying with a predetermined distribution or a flat plate having a thickness varying delicately continuously may be disposed in the vicinity of the object surface or the image plane or the intermediate imaging plane. In such case, the variation in refractive index or thickness of the flat plate imparts a refracting action similar to that of a small prism to light rays and can modify the magnification into a desired magnification.For example, to form a refractive index distribution in a parallel flat plate of glass, the use of the ion diffusing method is suitable, and to form a variation in thickness, a substance having the same refractive index as the thin film may be deposited by evaporation on the thin film.

According to the present invention hitherto described, modification of the lateral magnification is possible without reforming the conventional optical system. In the well-known semiconductor printing process, a pellicle has been stretched flatly somewhat in spaced apart relationship with a wafer to prevent dust from falling onto the wafer, and the present invention can be realized by positively utilizing the pellicle of this type.

By making the shape of the optical thin film into a predetermined shape in accordance with the present invention, it has become possible to correct expansion or contraction of the wafer resulting from a delicate difference between machines or the process. This fact has a very important meaning in the manufacture of VLSI in which alignment errors of the order of 0.1 Fm are critical. Also, where the present invention is applied to a mirror optical system, the magnification in a direction orthogonal to the scanning direction can be changed freely and this is an advantage which could not be found in the prior art.

By carrying out the present invention, the alignment error in the manufacture of LSI has been greatly reduced and production at a high yeild has become possible. The technical idea of the present invention is applicable not only to such a field but also to every field which requires minute alignment.

Claims

1. An optical device provided with: an imaging optical system for forming the image of an object on the image plane and for determining an imaging optical path; and optical means for shifting the position of the image in a direction orthogonal to the imaging optical path, said optical means having such a degree of optical thickness which substantially does not impart other influence than said position-shifting to the imaging performance and being disposed in said imaging optical path while assuming a curved shape.

2. An optical device according to Claim 1, wherein said optical means is disposed between the image plane and said imaging optical system and toward the image plane.

3. An optical device according to Claim 2, wherein said optical means is disposed closely adjacent to the image plane.

4. An optical device according to Claim 1, wherein said optical means is disposed between the object and the imaging plane and toward the object.

5. An optical device according to Claim 4, wherein said optical means is disposed closely adjacent to the object.

6. An optical device according to Claim 1, wherein said optical means is disposed substantially in coincidence with the intermediate imaging plane of said imaging optical system.

7. An optical device according to Claim 2, 4 or 6, wherein said optical means is of a cylindrical shape.

8. An optical device according to Claim 7, wherein said optical means is a pellicle.

9. An optical device according to Claim 2, 4 or 6, wherein said optical means is a thin member of semi-spherical shape.

10. An optical device according to Claim 1 ,further provided with means for integrally holding said object and a recording member for supporting said image plane and recording an image and for moving them relative to said imaging optical system and wherein said optical means is curved into a cylindrical shape having the direction of a bus line coincident with the direction of movement of said object.

11. An optical device according to Claim 10, wherein said imaging optical system is a mirror objective.

12. An optical device according to Claim 10, wherein said optical means is a pellicle.

13. An optical device according to Claim 10, wherein the thickness of said optical means is determined so that the difference between the wavefront aberration when said optical means is mounted and the wavefront aberration when said optical means is dismounted is between A (used wavelength) and -X.

14. An optical device according to Claim 10, further provided with a frame for maintaining said optical means.

15. An optical device according to Claim 13, wherein said frame is provided with a mask having an arcuate opening.

16. An optical device according to Claim 10, wherein the degree of curve of said optical means is variable.

17. An optical device provided with: an imaging optical system for forming the image of an object on the image plane and for determining an imaging optical path; and a thin plate for shifting the position of the image in a direction orthogonal to said imaging optical path, said thin plate having such a degree of optical thickness which substantially does not impart other influence than said position-shifting to the imaging performance and a continuously varying refractive index distribution and being disposed in said imaging optical path.

18. An optical device provided with: an imaging optical system for forming the image of an object on the image plane and for determining an imaging optical path; and a thin plate for shifting the position of the image in a direction orthogonal to said imaging optical path, said thin plate having a thickness continuously varying within the range of such a degree of optical thickness which substantially does not impart other influence than said position-shifting to the imaging performance and being disposed in said imaging optical path.

19. An optical device according to Claim 16 or 17, wherein the thickness of said thin plate is determined so that the difference between the wavefront aberration when said thin plate is mounted in said imaging optical path and the wavefront aberration when said thin plate is dismounted is between X (used wavelength) and -X.

20. An image forming optical system for forming an image of an object in an image plane, and means for producing a shift in the position of portions of the image in the image plane, said means producing substantially no other change in the image formed in said image plane.

21. Apparatus for forming an image of a mask on a wafer, in accordance with claim 20.

22. Apparatus for forming an image of an object according to claim 20 and substantially as hereinbefore described with reference to the accompanying drawings.