KR20090093960A - Method of manufacturing a projection objective and projection objective manufactured by that method - Google Patents

Abstract

A method of manufacturing a projection objective including the steps of defining an initial design for a projection objective and optimizing the design using a merit function having a plurality of merit function components AB, IRRAD EFP, each of which reflects a particular quality parameter. One of that merit function components defines a maximum irradiance requirement requiring that a normalized effective irradiance value representing an effective irradiance AB, IRRAD EFF normalized to an effective irradiance in an image surface of the projection objective does not exceed a predefined irradiance threshold value IRR TV on each optical surface of the projection objective except for a last optical surface directly adjacent to an image surface of the projective objective. Optical surfaces positioned within caustic regions and/or critically small effective sub-apertures on optical surfaces are thereby systematically avoided.

Description

Method of manufacturing a projection objective and a projection objective manufactured by the above method

The present invention relates to a method of manufacturing a projection objective that includes defining an initial design for a projection objective and optimizing the design using an advantage function. The method is used to produce projection objectives that can be used in microlithography processes to make miniaturized devices. The invention also relates to a projection objective produced by such a method.

Microlithography processes are commonly used in the manufacture of miniaturized devices such as integrated circuits, liquid crystal devices, micro-pattern structures and micro-mechanical devices. In such a process, the projection objective serves to project a pattern of a patterning structure (typically a photo mask (mask, reticle)) onto a substrate (usually a semiconductor wafer). The substrate is coated with a photosensitive layer (resist) that is exposed to an image of the patterning structure using projection radiation.

In order to form finer structures, it is all pursued to increase the image-side numerical aperture NA of the projection objective lens and to use ultraviolet radiation having a shorter wavelength, preferably a wavelength smaller than about 260 nm. In conclusion, even higher demands are placed on the complexity of the projection objective. Projection objectives typically have at least 10 or 20 or even 25 or more multiple optical elements such as lenses, curved mirrors and the like. Each single optical element, as well as the entire structure comprising multiple optical elements arranged in some way, must be designed and manufactured with high accuracy to provide imaging of the patterning structure on the substrate within a large image field and to have a low degree of aberrations. do.

Creating a new design of a projection objective is a complex task involving the optimization of structural and quality parameters of the projection objective. Structural parameters include refractive index of the material forming the lenses, surface shape parameters of the lenses and mirrors (if applicable), the distance between the first and second surfaces of each lens, between the surfaces of different optical elements. Distance, the distance between the object plane of the projection objective lens and the incidence plane of the object-side front element of the projection objective lens, the distance between the exit face of the image-side front element of the projection objective lens and the image plane, between adjacent optical elements The refractive index of the medium disposed between the object plane and the object side front element and between the image plane and the image side front element.

The quality parameters include, for example, parameters indicating the optical performance of the projection objective lens in items such as selected aberrations, image-side numerical aperture, magnification of the projection objective lens, and the like.

Optimizing the design to meet the desired details of the optical performance and other quality characteristics of the projection objectives today involves computational methods such as tracing rays to optimize the parameters of the projection objective while observing certain boundary conditions. Include. CODE V, a lens analysis and design program sold by Optical Research Associates, Inc., is a commonly used software tool employed for that purpose. Optimization includes minimizing or maximizing a suit function appropriately selected according to the parameters of the design. Typically, the construction of the benefit function is achieved by using a number of benefit function components that can represent optical, manufacturable and other aspects that describe the optimization goals of a particular design.

Due to the high number of parameters in the design, the solution space of the optimization process has a high dimension and many local minimums in such solution spaces that are computationally trapped, resulting in results far from the design meeting the required details. And maxima exist. Accordingly, optical designers designing projection objectives for microlithography must meet careful work in order to observe certain boundary conditions based on their intuition and / or to determine new design principles suitable for a particular application. The designer will therefore embody an "initial design" that serves as a potentially successful "starting point" for computer-based optimization and then improve the design based on it by computational optimization. Typically, one or more results will still be insufficient for the desired overall details and much effort will have to be attempted until a satisfactory solution is found. Thus, in terms of computational manufacturing, the cost of the new design will be high.

Once the proper design has been found, the optical elements of the projection objective lens must be manufactured and assembled to obtain the actual product of the manufacturing process. During this manufacturing step, care must be taken to avoid as much as possible surface defects of optical surfaces such as scratches and contaminations. While low or medium aperture systems can be relatively tolerant to such artificial defects, very high, such as projection objectives designed for immersion operations in the aperture range of the image-side numerical aperture NA> 1.0. Very stringent details generally apply to surface defects in the manufacturing process of numerical aperture lithographic projection objectives.

At least one intermediate image, at least one concave mirror and at least one flat folding mirror arranged to deflect radiation coming from the concave mirror from the object surface or from the concave mirror towards the image surface Refractive index projection objectives with s are mainly applied in such applications. Representative examples are for example shown in WO 2004/019128 A2 or JP 2005-003982 A.

Surface defects on optical surfaces can seriously affect the optical performance of the projection objective by causing light loss due to different damage mechanisms. For example, relatively rough surface areas on an ideal smooth surface will typically result in stray light while negatively affecting optical performance. In addition, transparent surface defects can act as phase objects, resulting in light loss due to interference. Opaque contamination can block the area of the projection beam, thereby causing a heterogeneity of the intensity distribution in the image field. It is desired to have a projection objective that is relatively insensitive to performance degradation caused by surface defects on optical surfaces.

Figure 1 shows a detailed view of a conventional reflective refractive projection objective with exactly two mesophases and one concave mirror, where relatively small actual sub-openings of the beams occur at near-middle optical surfaces (FIG. 1A), the folding mirrors are located in the corrosive area (FIG. 1B).

FIG. 2 shows an axial schematic view of the lens surface located at the pupil plane and illuminated by dipole illumination, where contamination is present in the region of one of the poles.

FIG. 3 shows a schematic flow diagram illustrating a preferred embodiment of a manufacturing method for avoiding optical surfaces with a very large concentration of irradiance.

4 illustrates a pupil raster of pupil points for dividing the pupil plane into raster fields having substantially the same raster field area, in accordance with one embodiment of the present invention.

FIG. 5 shows the intersections of the rays of the pupil raster shown in FIG. 4 on a selected optical surface away from the pupil plane, where a region of relatively high effective irradiance (or pinhole irradiance) is shown.

FIG. 6 shows the intersections of the rays of the pupil raster shown in FIG. 4, on a selected optical surface remote from the pupil plane, in which regions with corrosion conditions are shown.

7 shows an embodiment of a refracted projection objective having two intermediate images, one concave mirror and two flat deflection mirrors, in which corrosion conditions are systematically avoided on optical surfaces close to the intermediate images. have.

FIG. 8 shows the marks of selected field points around the edge of the rectangular field on the first and second folding mirrors of the embodiment of FIG. 7.

It is an object of the present invention to provide a method of manufacturing a projection objective that allows for the production of complex projection objectives for microlithography in a cost-effective manner while maintaining high standards for optical performance.

It is another object of the present invention to provide a method of manufacturing a projection objective for producing a projection objective for microlithography having an optical performance that is relatively insensitive to surface defects present on optical surfaces in the projection objective. To provide.

It is another object of the present invention to provide a method of manufacturing a projection objective lens having a low degree of field change in intensity.

It is another object of the present invention to provide a projection objective that is relatively insensitive to surface defects caused by contaminations and other effects on optical surfaces within the projection objective.

As a solution to these and other objects, the present invention, according to one scheme, comprises a step of defining an initial design for a projection objective and optimizing the design using a benefit function. Provided is a method of making an objective lens, the method comprising:

Defining a plurality of benefit function components each reflecting a particular quality parameter, wherein one of the benefit function components is a normalized validity representative of the effective irradiance normalized to the effective irradiance at the image surface of the projection objective Define a maximum irradiance requirement requiring that irradiance values not exceed a predetermined irradiance threshold on each optical surface of the projection objective except for the last optical surface directly adjacent to the image surface of the projection objective. Doing;

Calculating a numerical value for each of the advantage function components based on the corresponding feature of the preliminary design of the projection objective;

Calculating, from the advantage function components, an overall benefit function that can be expressed as numerical items reflecting quality parameters;

Continuously changing at least one structural parameter of the projection objective lens and recalculating the resulting overall advantage function for each successive change until the resulting overall advantage function reaches a predetermined acceptable value;

Obtaining structural parameters of the optimized projection objective lens having a predetermined acceptable value for the resulting overall advantage function; And

Implementing the parameters to produce a projection objective.

The term "irradiance" generally refers to the power of electromagnetic radiation at one surface per unit area. Specifically, the term "irradiance" refers to the power of electromagnetic radiation that is incident on the surface. The term "effective irradiance" as used herein refers to the contribution to the overall irradiance incident on the optical surface, which is a contribution originating from the radiation emitted from a single object field point. Instead, the effective irradiance may be referred to as "pinhole irradiance" in the present application.

If a large value of effective irradiance is avoided on optical surfaces, the optical performance of the projection objective may become relatively insensitive to surface defects present on the optical surfaces in the projection objective.

In many cases, the "last optical surface", ie the optical surface of the projection objective closest to the image surface, is subject to special conditions that require that the last optical surface be excluded from the optimization process. The spatial distribution of the points of incidence of the radiation beams over the last optical surface is greatly influenced by the concept of the relevant projection process (dry or immersion projection) and the image-side work space (the last optical surface and image on which the substrate surface is located). Geometrical conditions in the space between the surfaces). For example, it may be useful to provide a narrow gap (typically one millimeter or more in width) between the last optical surface and the image surface in order to introduce an immersion liquid for immersion exposure. In view of the entry and exit of the immersion medium in the image space, it is often required to have a substantially flat final optical surface. Under these conditions, the intensity load of the last optical surface (eg, expressed as an effective irradiance) is basically the image side working distance, the size of the effective image field, the refractive index of the medium behind the last optical element, and the image. It is determined by the side numerical aperture NA. Therefore, the structural parameters representing the position and surface shape of the last optical surface are not free parameters and should be excluded from the optimization process.

Various conditions may result in or contribute to a local maximum of effective irradiance on optical surfaces. For example, large effective irradiance values may occur where the optical surface lies within the corrosive region. In one embodiment, such conditions can be determined by calculating the location and range of potential corrosive regions in the projection objective; And by optimizing the structural parameters of the projection objective lens so that no optical surface is located in the corrosive region.

Alternatively or in addition, optical surfaces may be important where the actual sub-openings of the ray bundles become significantly smaller. Thus, one embodiment of the method comprises the following steps: defining a number of representative field points; Computing beams originating from the field points and intersections of the beams with the optical surfaces, wherein the region of intersection of the beams with the optical surface is defined by the area of the intersection zone ( defining an actual sub-opening having a sub-aperture size; Defining a sub-opening size threshold; And for all optical surfaces of the projection objective except the last optical surface directly adjacent to the image surface of the projection objective, the actual sub-opening size for the selected field points does not fall below the sub-opening size threshold. To optimize the structural parameters of the projection objective.

Embodiments of the method may include tasks that consider one or more causes of very high effective irradiance concentrations. For example, one benefit function component may define a maximum irradiance requirement as described above, and another benefit function component may be that the size of the actual sub-openings on all optical surfaces (except the last optical surface) The minimum actual sub-opening requirement may be defined such that only the optical surfaces that lie above a given actual sub-opening size threshold automatically have the resulting optical design.

The method is useful for the design of various types of optical systems, in particular for projection objectives used in microlithography with relatively high image-side numerical apertures, such as those found mainly in projection objectives for immersion lithography that can achieve NA> 1. It can be applied in the design. Detailed analyzes of a large number of conventional projection objectives have shown that at least one concave mirror and at least one intermediate image, as well as beam beams traveling towards the concave mirror, are reflected from beam bundles reflected from the concave mirror. It has been shown that small effective and / or actual sub-openings and / or corrosion conditions may occur in particular in a refracted projection objective having at least one flat deflection mirror to separate. This type of projection objectives of the present invention does not have optical surfaces (except the last optical surface adjacent to the image surface) where corrosion conditions occur and / or very small effective and / or actual sub-openings occur. It can be designed according to the embodiments.

In some embodiments, the projection objective lens comprises a plurality of optical elements arranged to image an off-axis object field disposed on the object surface over the non-axis image field disposed on the image surface of the projection objective lens, The optical elements are:

A first refractive objective lens unit forming a first intermediate image from radiation emitted from an object surface and including a first pupil surface;

A second objective lens unit including at least one concave mirror for forming the first intermediate image into a second intermediate image and including a second pupil plane optically conjugate to the first pupil plane; And

The second intermediate image is imaged onto the image surface and forms a third refractive objective lens portion including a third pupil plane that is optically conjugate to the first and second pupil planes.

Embodiments may have exactly two mesophases.

The second objective lens portion may have exactly one concave mirror, and the projection objective lens may include a first folding mirror for deflecting the light emitted from the object surface in the direction of the concave mirror and the light emitted from the concave mirror on the image surface. It may have a second folding mirror for deflecting in the direction. The deflection mirrors may all be flat. The projection objective may be designed for immersion lithography with NA> 1.

As an introduction to some of the problems solved by the preferred embodiments of the present invention, consider a high aperture refraction-refractive projection objective suitable for immersion lithography of NA> 1. In many cases, such designs show rather small actual sub-openings on the optical surfaces. As used herein, the term "real sub-aperture" refers to the "ray bundle" originating from a specific object point (ie, a field point within the object surface) on the optical surface. Mark ”(intersection). Small actual sub-openings are refractive with at least one concave mirror and at least one intermediate image as well as at least one flat deflection mirror for separating the beam traveling toward the concave mirror from the beam reflected from the concave mirror. This can occur especially with projection objectives.

For the purposes of explanation, FIG. 1A shows a detailed view of a conventional reflective refractive projection objective taken from FIG. 2 of WO 2004/019128 A2. The projection objective lens includes a first refractive object lens unit OP1 for forming an object field from a flat object surface OS into a first intermediate image IMI1, and converts the first intermediate image IMI1 into a second intermediate image IMI2. ) And a second refraction objective lens unit OP2 including a concave mirror CM for forming an image, and a second intermediate image IMI2 on a flat image surface parallel to the object surface OS. A third refractive objective lens portion OP3 (only partially shown) is included. The first flat folding mirror FM1 deflects radiation coming from the object surface towards the concave mirror CM. A second flat folding mirror FM2 perpendicular to the first folding mirror deflects the light emitted from the concave mirror towards the image surface.

The light beam RB originating from the off-axis field point FP1 of the field of effective object located completely outside of the optical axis AX is in cross-sections of varying magnitudes depending on the position of the optical surface in the optical system. At the intersection with the optical surfaces. The intersection zones are designated as "actual sub-openings" in this application. Some representative cross sections are highlighted in bold lines in FIG. 1. The first actual sub-opening SA1 on the incidence side of the plane-parallel plate immediately after the object surface is closer to the field surface (object surface) and is relatively small. The size of the second actual sub-opening SA2 close to the first pupil plane P1 on the concave incidence side of the meniscus lens substantially corresponds to the size of the pupil at that position and is relatively large. All actual sub-openings substantially overlap at the pupil surface. The actual sub-openings, as can be seen for the actual sub-opening SA3 on the concave exit side of the positive meniscus lens immediately upstream of the first folding mirror FM1, and the first intermediate phase As can be seen by the actual sub-opening SA4 formed on the first folding mirror FM1 on the upstream side, the closer the optical surfaces are to the first intermediate image IMI1, the smaller it becomes. On the contrary, the fifth actual sub-opening SA5 on the convex surface of the negative meniscus lens just in front of the concave mirror CM is substantially equal to the size of the second pupil P2 in which the concave mirror CM is located. Has a corresponding large size.

In order to obtain an image without vignetting and pupil obscuration, a stockpile field and an image field are used in this design. When the stock field is used, the effort required to correct the imaging aberrations increases with increasing the distance between the stock object field and the optical axis so that the size of the "design object field" increases. The "design object field" includes all field points of the object surface that can be projected by a projection objective with sufficient imaging accuracy for the intended lithography process. Within the design object field all imaging aberrations must be sufficiently corrected for the intended projection purpose, whereas for field points outside of the design object field at least one of the aberrations is larger than the desired threshold. Thus, in order to facilitate correction, it may be required to keep the size of the design object field small, which also requires minimizing the offset between the optical axis and the non-axis object field. Efforts to minimize such offsets often lead to designs having lens surfaces and / or mirror surfaces relatively close to the field surface, whereby corresponding actual sub-existing on the optical surfaces close to the field surfaces are present. The openings become small. For example, the actual sub-openings SA3 and SA4 on the lens surface and the mirror surface immediately upstream of the first intermediate image IMI1 are relatively small.

Small practical sub-openings on optical surfaces generally require stringent requirements on the manufacturing process and cleanness of the resulting optical surfaces, substantially free of serious surface defects. Since the energy emitted from the selected field point is concentrated in a relatively small area in the area of the small real sub-opening, defects present in the area of the small real sub-opening may block some of the light intensity present in the light beam. And thereby generate light loss. In the case where the corresponding actual sub-openings are small, surface defects of a given size result in a greater light loss effect than in areas with large actual sub-openings, since the area of the defects leading to disturbance and the actual sub-openings This is because the ratio between the areas (marked areas) simply becomes more undesirable. Local light loss caused by defects can lead to uniformity errors with direct problems with regard to the critical dimensions of structures made with projection objectives. In particular, an undesirable change in CD (CD change) may be caused with respect to the sensitivity of the photosensitive coating (photoresist) on the substrate.

In addition, localized light loss can also result in telecentric errors resulting from the shift of the energy center of the pupil in the direction of generally undisturbed regions. As shown, FIG. 2 shows a schematic view in the axial direction of the lens surface LS located in the circular pupil plane in the projection objective. Projection exposure apparatus is operated with dipole illumination, for example, to improve depth of focus and / or contrast when printing densely packed unidirectional lines. The corresponding intensity distribution of the projected light is characterized by two "poles", the first pole PO1 and the second pole PO2 on both sides of the optical axis AX, in which the light energy is concentrated. Built. There is no light intensity on the optical axis. Surface contamination CON is present on the lens surface LS in the region of the first pole PO1. Since the contamination CON blocks a significant portion of the light energy present in the first pole PO1, the energy center of the pupil will shift towards the second pole PO2 that is not disturbed. This eccentric center of energy in the pupil corresponds to the direction in which the center of intensity of the image travels in the image space in which the wafer is located. Due to the contamination (CON), the direction of travel deviates from the optical axis and is directed at a finite angle with respect to the optical axis. This results in an oblique orientation of the aerial image, for example if the position of the substrate surface changes locally due to an uneven substrate surface and / or due to positional changes caused by the wafer stage. -May cause distortion.

Assuming that the actual sub-openings are substantially homogeneously illuminated such that there is no difference or very slight differences in local radiant energy input between the different positions within the actual sub-openings, the actual sub-as described above. The size of the openings is considered as a good indicator of which optical surfaces will be critical for surface defects.

The substantially homogeneous distribution of irradiance in the actual sub-openings may be a good approach for many purely refractive optical systems. However, there may be significant changes in local irradiance in the actual sub-openings such that serious differences in local irradiance may occur in the actual sub-openings. In order to take into account possible heterogeneity with respect to local irradiance in actual sub-openings, it has been found useful to define an "effective sub-aperture", which is the illuminance in actual sub-openings. Allows for localized changes in As mentioned above, the defects on the optical surfaces can be serious when relatively large radiant energy is concentrated in the area of the defect. Thus, in real sub-openings, such areas can be particularly severe when the local irradiance has a relatively large value compared to other parts of the real sub-openings. In order to almost avoid the problems associated with the local concentration of radiant energy, it is believed that it would be useful to analyze the area of the actual sub-opening to identify the area (or areas) where the maximum value for effective irradiance occurs. In geometrical description, the local irradiance may be characterized by a specific value for the "geometric beam density" of the rays originating from a single field point in the object plane, where the maximum value for the effective irradiance is said geometric ray It occurs where the maximum value occurs for density.

Now consider a system optimization that takes into account such local maximums of geometric ray density (or effective irradiance). Tolerance to the system layout may have a certain value assuming that each actual sub-opening is uniformly illuminated. However, the tolerance must be tighter if local concentration of radiant energy occurs within the actual sub-opening. This effect can be considered by defining a "effective sub-opening" corresponding to the sub-opening indicating the position with the maximum geometric light beam density (or maximum effective irradiance). Within this concept, the size of the "effective sub-opening" is equal to the size of the "real sub-opening" in systems where the actual sub-opening is homogeneously illuminated. However, if a local concentration of radiant energy occurs within the actual sub-opening, this is represented by the size of the effective sub-opening being smaller than the size of the actual sub-opening.

As an example, FIG. 1B shows a detailed view of another conventional reflective refractive projection objective taken from Example 5 of WO 2004/019128 A2. Identical or similar elements are denoted by the same reference numerals as in FIG. 1A. One light beam RB emerges from the selected field point FP1 at the object surface OS. The opening angle of the light beam at the object surface is determined by the object side numerical aperture NA _OBJ . The trajectory of various selected light rays, including light rays R1 and R2, is shown. As is apparent from the figure, the geometric light beam density defined by the light beams of the light beam RB extends across the optical surfaces to the area of the large positive meniscus lens ML upstream of the first folding mirror FM1. Substantially homogeneous. However, as the rays approach the first folding mirror FM1, the local density of the rays (i.e. the geometrical ray density) increases outward (farthest from the optical axis AX) as compared to the area closer to the optical axis. do. In other words, the geometric ray density is not homogeneous in the region of the first folding mirror which is a small distance upstream of the first intermediate image IMI1. In particular, the light rays R1 and R2 originating from the field point FP1 with different aperture values intersect in the region at or near the first folding mirror FM1. (Note that such intersection of single rays occurs especially in the areas of the first folding mirror FM1 and the second folding mirror FM2 optically close to the intermediate images IMI1, IMI2, respectively.) Different openings The intersection of the rays originating from the common field point to the furnaces indicates the presence of a "caustic" condition. It is evident from FIG. 1B that both the first folding mirror FM1 and the second folding mirror FM2 are located in the regions where the light beams of the light beam RB intersect, ie in the “caustic regions”.

As mentioned above, the geometric ray density corresponding to a particular field point can be considered as representing the contribution of a particular field point to the total irradiance on the optical surface. This contribution is called "effective irradiance" in this application. In thinking experiments, we consider illumination pinholes (which represent one single field point) in the object plane of the optical system. A beam of light originates from the field point. The "effective irradiance" corresponding to that field point is the irradiance contribution of the light beam originating from that field point at a selected optical surface. This contribution to the total irradiance incident on the optical surface is not only a function of the position of the pinhole in the object surface, but also a function of the position (or location) on the optical surface. It is intended that the maximum of all values of the effective irradiance (that is, the irradiance corresponding to a single object field point) should not exceed a predetermined "radiation threshold".

In corrosive areas, ie areas where corrosive conditions are present, effective irradiance (pinhole irradiance) can lead to substantial divergence and divergence of serious light energy on relatively small surface areas. In terms of ray propagation, if different light rays emitted from one object point with different numerical apertures intersect on or near one optical surface, then corrosion conditions are given on that optical surface. Surface defects on the optical surface where the corrosive areas are located affect the light rays emitted from the object point at different aperture angles, thereby potentially worsening the imaging quality substantially more than defects located outside the corrosive areas. .

In one embodiment of the method of manufacturing a projection objective according to the invention, the occurrence of corrosion conditions on the optical surfaces and / or the generation of too small effective sub-openings on the optical surfaces is such that It is systematically avoided due to the corresponding sub-routines in the software used in computer-based optical designs that lead to general deployment. If such corrosion conditions are systematically avoided, the requirements for cleanliness of the process can be relaxed without substantially jeopardizing the optical performance of the resulting projection objective.

In this embodiment, one of the benefit function components used in the computational process is the maximum for the effective irradiance IRRDD _EFF occurring on each optical surface (except the last optical surface closest to the image surface). Define a maximum irradiance requirement that requires that the value not exceed a predetermined irradiance threshold IRRTV.

As used herein, the term "irradiance" refers to the power of electromagnetic radiation on one surface per unit area. Specifically, the term "irradiance" refers to the power of electromagnetic radiation incident on the surface. SI units for irradiance are watts per square meter (W / m 2). Irradiance is sometimes called intensity, but should not be confused with radiant intensity with different units. The term “effective irradiance” (also referred to as “pinhole irradiance”) refers to the contribution to the total irradiance incident on one optical surface, originating from the radiation coming from one single object field point. In terms of ray propagation, the "effective irradiance" corresponds to the "geometric ray density" of the different rays originating from a common field point with equidistant aperture steps. Effective irradiance can be homogeneous on one surface. In general, the effective irradiance is defined to define a region of maximum effective irradiance (corresponding to a large local geometric ray density) and smaller values of effective irradiance (corresponding to a smaller geometric ray density). It may change over one surface.

If relatively large values of effective irradiance occur on the optical surface, such optical surfaces can be dangerous for surface defects such as scratches and contamination. Relatively large values of effective irradiance can be brought about, for example, by corrosion conditions on the optical surface and / or by too small effective sub-openings on the optical surface.

One embodiment of an optimization routine implemented within a software program used to calculate the optical design (layout) of a projection objective is now described with reference to the schematic flowchart shown in FIG. 3. In the first step S1 (optimization for aberrations, OPT AB), the basic layout of the optical design is optimized with respect to the aberrations to obtain the optimization design D1. To this end, conventional subroutines of an appropriate optical design program can be used to change one or more structural parameters of the projection objective and to calculate the resulting overall aberrations of the design. As a result, the optimization design D1 may or may not have optical surfaces where a large concentration of effective irradiance occurs locally on one or more optical surfaces.

In a second step S2, the normalized effective irradiance value IRRAD _EFF representing the normalized effective irradiance normalized to the effective irradiance at the image surface of the projection objective is equal to the last optical directly adjacent to the image surface of the projection objective. The optimized design is analyzed to determine whether or not the predetermined irradiance on each optical surface of the projection objective lens (excluding the surface) exceeds the threshold value IRRTV.

If the normalized effective irradiance value IRRAD _EFF does not exceed the irradiance threshold IRRTV on any optical surface (except for the last optical surface if possible), then the parameters representing the optimization design (D1) are the aberration level. An aberration determination step S3 is entered to determine whether AB is above or below a predetermined aberration threshold ATV. If the degree of aberration is below the aberration threshold, the optimization design D2 is output as a result of the optimization process. The final design D2 will be identical to the optimization design D1 which serves as an input to the irradiance determination in the second step S2.

If the irradiance determination in the second step S2 determines that the normalized effective irradiance value IRRAD of the optimization design D1 exceeds the irradiance threshold IRRTV for at least one optical surface (except the last optical surface) If so, the calculation routine proceeds to the irradiance optimization step S4 (OPT IRRAD), where the optimization design D1 is again optimized for effective irradiance to reduce the local concentration of irradiance on the critical optical surfaces. do. To this end, at least one structural parameter of the projection objective is changed and an overall advantage function is employed, including an advantage function component that defines the maximum irradiance requirement. The resulting reoptimization design D1 ′ then determines irradiance through step S1 to determine whether the normalized irradiance value is now below the irradiance threshold IRRTV for each of the considered optical surfaces. It is input to step S2. An iteration that includes two or more reoptimization steps can be used for this purpose.]

If the irradiance determination step S2 determines that the resulting reoptimization design D1 ′ has been optimized with respect to the local maximum occurrence of irradiance, then reoptimization with respect to irradiance produces one or more critical aberrations for each aberration. The optimized structure is input to the aberration determination step S3 to determine whether or not to be above the threshold. If the reoptimization design is still acceptable for aberrations, the final design D2 'is obtained and output in step S5.

If the design needs to be optimized with respect to aberrations, in order to modify the optical design (already optimized with respect to irradiance) also with respect to aberrations, the structural parameters received from step S2 are converted to step S1. ) Is entered. One or more iterations may be needed to reoptimize the design with respect to aberration and still obtain normalized irradiance values below the critical irradiance threshold IRRTV for all critical optical surfaces. The final design (D2 ') is the result of the reoptimization process.

In some embodiments, the subroutine with respect to the irradiance determining the determining step S2 allows to avoid too small sub-openings on all optical surfaces except the last optical surface closest to the image surface and all optical surfaces. It is possible to avoid the occurrence of corrosion conditions on the field.

One embodiment of a computational routine designed to systematically avoid the occurrence of corrosion conditions on optical surfaces is now described with reference to FIGS. 4-6.

In the first step of the computational part of the manufacturing method, a number of representative field points are defined. Ray tracing is performed for rays of light beams originating from the representative field points.

In a second step, a pupil raster is defined, wherein the pupil raster represents an array of raster points in the pupil plane of the projection objective, the raster points being defined within a two-dimensional array. Away from each other by distance.

In a third step, the ray trajectory of the rays originating from the representative field points and passing through the raster points of the pupil raster is calculated for each representative field point. For this purpose, commercially available design programs such as CODE V, OSLO or ZEMAX can be used.

The pupil raster typically surrounds the entire available aperture of the projection objective, where the aperture determines the starting angle of the light beam originating from each representative field point. In the embodiment of the pupil raster shown schematically in FIG. 4, the coordinates of the raster points in the pupil plane are such that neighboring raster points (ie, raster points directly adjacent to each other at a predetermined distance) have the same distance in the azimuthal direction. It is given in polar coordinates. In this embodiment, the angular step width between neighboring raster points in the circumferential direction (azimuth direction) is 10/3 degrees. The coordinates of the raster points in the radial direction (radiation direction) correspond to the opening angle contained between the respective light rays and the optical axis. This angle is also referred to as "pupil angle" in this application. The absolute value of the sine of the pupil angle is the square root function between the angle 0 and the maximum angle k _max = NA · β With i = 0, 1, ..., n, where NA is the image-side numerical aperture of the projection objective lens, and β is the magnification between the object field and the image field. By doing so, the pupil plane is subdivided into raster fields (or raster cells) having substantially the same raster field area.

In a fourth step, the intersections of the optical surfaces in the projection objective with the selected rays are calculated for each optical surface (potentially in some cases except the last optical surface immediately adjacent to the image surface).

In the following steps, pairs of directly neighboring raster points in the pupil plane are considered. Directly neighboring raster points indicate that the two raster points of the pair have the same azimuthal coordinate or the same pupillographic coordinate, while each of the other coordinates differs by one coordinate step in each coordinate direction according to the preselected step width. It features. For each pair of neighboring raster points, the following difference quotient is calculated:

Where f represents the number of each optical surface and i, j represent the index of neighboring pupil raster points. In equation (1), the variables x and k are vectors. The components of the vector x represent the coordinates of a point on the optical surfaces in real space. The components of the vector k are directional sine values representing the x, y, z coordinates of the unit vector indicating the direction of travel of the ray in the entrance pupil of the optical system (ie in pupil space). Thus, the equator quotient in Equation (1) is a measure indicating the relationship between a step having a predetermined step width (variable k) in the pupil space and a corresponding step width (variable x) in real space on the optical surface. In other words, the gradient parameter g ^f _ij defined by the quotient of equation (1) represents the degree of change of intersections on each surface for a given difference in pupil coordinates. (The difference quotient in Equation (1) is an approximation of the differential quotient indicating that finite step widths are commonly used in numerical calculations. The difference quotient in Equation (1) is zero in step width. As you approach, it becomes a derivative.)

Note that the numerical criteria given in Equation (1) above are defined using linear gradients. More precisely, the ratio between the related one in the directional space and the areas of one grid mesh element on the corresponding surface could be controlled. In practice, however, it is shown that controlling the linear gradient in nearly orthogonal directions is sufficient to avoid local peaks of light density in most cases.

In an additional step, a gradient threshold is defined that represents the minimum acceptable gradient between neighboring intersections on the optical surfaces. For example, the gradient threshold g ^f _ij (min) = 10 mm can be defined.

As further explaining the meaning of the gradient parameter, FIG. 5 shows the intersections of the rays of the pupil raster shown in FIG. 4 on the selected optical surface for the selected representative field point. In the high irradiance region (HIRAD) at the bottom right of the optical surface, the local density of the intersections is significantly larger than in other parts of the optical surface, such as in the upper left portion radially opposite to the high irradiance region. Because of this, it is evident that the distribution of intersections and consequently the distribution of irradiance across the optical surface is not uniform. However, no corrosion conditions are given on the optical surface, because in the azimuthal and radial directions (although the step width between neighboring intersections representing the amount of irradiance in each region varies greatly across the optical surface). This is because the series of intersections of is the same as in the pupil plane. The situation shown in FIG. 5 corresponds to a value of 4.7 _mm for the gradient parameter g ^f _ij .

In the next step, the minimum value of the gradient parameter is calculated for each optical surface (optionally, except for the last optical surface). If it is found that the minimum value calculated for a particular optical surface is smaller than the gradient threshold, the structural parameters of the projection objective are optimized to increase until the minimum gradient for each optical surface is greater than or equal to the gradient threshold. do. For example, the optimization process may include increasing the distance between each optical surface and neighboring field surfaces (such as object surface, mesophase surface, or image surface). Alternatively or in addition, the local tilt of the optical surface within the minimum gradient region may be changed to increase the calculated gradient for neighboring intersections.

The corresponding radius R _SUBEFF of the "effective sub-opening" can be calculated according to the following equation (2).

Where Min (g ^f _ij ) is the minimum value of the series quotient in Equation (1) and NA _OBJ is the image-side numerical aperture. Note that the area with the minimum value for the gradient parameter corresponds to the area with the maximum geometric light density (and the maximum effective irradiance).

As a final result of the optimization process, all optical surfaces of the projection objective are positioned and shaped so that the gradient parameter is equal to or greater than the gradient threshold. In terms of irradiance, this requirement is that the normalized irradiance value IRRAD, which represents the normalized irradiance for the irradiance at the image surface of the projection objective, is defined by each optical surface (projection objective of the projection objective). Corresponds to the condition that the predetermined irradiance on (except the last optical surface directly adjacent to the image surface) does not exceed the threshold value IRRTV. If the irradiance does not exceed the maximum acceptable value for normalized irradiance, then the optical performance of the projection objective will be relatively insensitive to potential defects on the optical surfaces, as described above.

For comparison, FIG. 6 shows the intersections of the rays of the pupil raster on the "virtual" system surface in the region where corrosion conditions occur with respect to the field point. At the upper left of the imaginary surface, the intersections corresponding to the pupil coordinates at the outer edge of the pupil are at the intersection of the axial ray rather than the intersections corresponding to raster points located somewhere between the optical axis and the outer edge of the pupil. It is clear that it lies closer. In other words, light rays originating from a certain field point with different numerical apertures (represented by different radial coordinates in the pupil raster) correspond to smaller aperture values where light rays corresponding to larger aperture values in the pupil plane. Intersect on or near each optical surface to intersect the optical surface closer to the axial ray than the light rays. In this description, the gradient parameter g ^f _ij obtains a minimum value of 0 mm in the region of the optical surface lying in the corrosive region CAUSTIC (upper left).

Employing this method systematically leads to an optical design in which no optical surfaces are located in corrosive areas and / or in areas with very small effective sub-openings. If such optical surfaces are avoided in the optical system, details regarding surface quality and / or contamination can be relaxed, thereby facilitating the manufacture of the optical system.

7 shows one embodiment of a refracted projection objective lens 100 that follows such conditions. The projection objective is designed for a nominal UV-operation wavelength of λ = 193 nm. The details are given in Table 1 and Table 1a. The projection objective 100 forms an image of a pattern on a reticle disposed on a flat object surface OS (object plane) on a reduced scale, precisely forming two real intermediate images IMI1, IMI2. For example, it is designed to project onto a flat image surface IS (image plane), for example 4: 1. The rectangular effective object field OF and the image field IF are off-axis completely outward of the optical axis AX. The first refractive objective lens unit OP1 is designed to image the pattern on the object surface into the first intermediate image IMI1 on an enlarged scale. The second reflection refractive index (refraction / reflection) objective lens unit OP2 forms the first intermediate image IMI1 into the second intermediate image IMI2 at a magnification close to 1: (-1). The third refractive objective lens unit OP3 forms the second intermediate image IM2 over the image surface IS at a strong reduction ratio.

The path of the chief ray CR of the outer field point of the off-axis object field OF is shown in bold in FIG. 7 so as to follow the beam path of the projection beam. For the purposes of the present application, the term "primary rays" (also known as primary rays) refers to rays which travel from the outermost field point (farthest from the optical axis) of the object field OF to be used effectively to the center of the incident pupil. Indicates. Due to the rotational symmetry of the system, the chief rays can be selected from equivalent field points within the meridional plane as shown in the figure for purposes of illustration. In the projection objective lens which is substantially telecentric on the object side, the chief ray emerges from the object surface parallel or very small with respect to the optical axis. The imaging process is also characterized by the trajectory of the ambient light. As used herein, a "peripheral ray" is a ray traveling from the object field point on the axis (field point on the optical axis) to the edge of the aperture stop. The ambient light may not contribute to image formation due to vignetting when a stockpile effective object field is used. The main and ambient light are chosen to characterize the optical properties of the projection objective. The angle included between such selected light rays and the optical axis at a given axial position is called "primary ray angle" (CRA) and "peripheral ray angle" (MRA), respectively. The radial distance between such selected light beams and the optical axis at a given axial position is called "primary ray height" (CRH) and "ambient ray height" (MRH), respectively.

Three mutually conjugate pupil surfaces P1, P2, and P3 are formed at positions where the chief ray CR crosses the optical axis. The first pupil plane P1 is formed in the first objective lens section between the object plane and the first intermediate image, and the second pupil plane P2 is formed in the second objective lens section between the first intermediate image and the second intermediate image. The third pupil plane P3 is formed in the third objective lens portion between the second intermediate image and the image surface IS.

The second objective lens unit OP2 includes a single concave mirror CM. The first flat folding mirror FM1 is disposed optically close to the first intermediate image IMI1 at an angle of 45 ° with respect to the optical axis AX such that the first flat folding mirror FM1 reflects the emission light coming from the object surface in the direction of the concave mirror CM. do. The second folding mirror FM2 having the flat mirror surface arranged at right angles to the flat mirror surface of the first folding mirror reflects the light emitted from the concave mirror CM in the direction of the image surface parallel to the object surface.

The folding mirrors FM1 and FM2 are located in the vicinity of the optical in the intermediate phase, respectively, so that the etendue (geometric flux) is kept small. The intermediate phases are preferably not located on the flat mirror surfaces, which results in a finite minimum distance between the optically closest mirror surface and the intermediate phase. This ensures that no defects in the mirror surface, such as scratches or impurities, are sharply imaged onto the image surface.

The first objective lens unit OP1 includes two lens groups LG1 and LG2 each having positive refractive power at both sides of the first pupil plane P1. The first lens group LG1 is designed to act in the manner of a Fourier lens group which performs a single Fourier transform by imaging the telecentric incident pupil of the projection objective lens on the first pupil plane P1. This Fourier transform leads to a relatively small maximum chief ray angle CRA _P1 of about 17 ° at the first pupil plane. As a result, according to the Lagrange invariant, the optical free diameter of the first pupil is relatively large, as indicated by the diameter D ₁ = 145 mm of the radiation beam at the first pupil plane.

The relatively small chief ray angle along with the large pupil diameter corresponds to the relatively large axial expansion of the pupil space PS. For the purposes of the present application, the pupil space is subject to the condition RHR <| B | for beam height ratio RHR = CRH / MRH. The ambient light height MRH is defined as an area sufficiently larger than the main light height CRH such that << 1 is satisfied. The upper limit B of the ray height ratio can be, for example, less than 0.4 or less than 0.3 or less than 0.2. If the above conditions are met, the correction applied to the pupil space will have a substantially field-constant effect. The axial expansion of the pupil space increases as the chief ray angle at each pupil decreases. In this embodiment, the condition RHR <0.3 is satisfied.

In the present embodiment, the pupil space PS comprises a first lens L1-6 (biconvex lens) of LG2 immediately downstream of the pupil plane, which is subsequently on the image side of the pupil plane. It extends to the lens L1-7 and to the biconvex positive lens L1-5 immediately upstream of the first pupil plane. Free spaces FS1 (= 41 mm) and FS2 (= 62 mm) each having an axial extension of at least 40 mm so as to place one or more thin correction elements in the free space, the pupil plane P1 in the pupil space PS On both sides of, i.e., optically close to the pupil plane. Thus, this embodiment makes it possible to interpose one or more correction elements optically close to the first pupil plane P1 in order to obtain substantially the same correction effect (correction constant in the field) for all field points of the field. do.

In the pupil space PS, the parallel plate PP is disposed on the first pupil plane P1 where the conditions RHR to 0 are satisfied. The parallel plate is part of the initial design of the projection objective and can serve as a placeholder for the correction element, which may be substantially formed as a planar parallel plate with the same thickness and material, wherein at least one The surface has an aspherical shape.

As mentioned above, in the high-aperture projection objective commonly used in microlithography, the relatively small real and / or effective sub-openings of the light beams may cause some optical surfaces, for example field planes. May occur at optical surfaces close to. For example, in the embodiment of FIG. 7, both the first folding mirror FM1 and the second folding mirror FM2 are adjacent intermediate images IMI1 such that each small actual sub-openings occur on the folding mirrors. And IMI2), respectively, in optical proximity. In general, the irradiance on the optical surface increases as the size of the sub-openings decreases. When relatively large values of irradiance occur on the optical surface, the optical surface may be critical for surface defects such as scratches and contamination. Corrosion conditions may also occur on the optical system, in particular on some optical surfaces close to the field plane. If the optical surface is in the corrosive region, the irradiance may be divergent.

Due to these effects, details about surface quality and contamination should be kept particularly strict in optical systems with optical surfaces in the region of small sub-openings and / or in the region where corrosion conditions occur. On the other hand, if such surfaces are avoided in the optical system, details about surface quality and / or contamination can be relaxed, thereby facilitating the manufacture of the optical system.

In the manufacture of the embodiment of FIG. 7, the beam path is corrected so that the first and second folding mirrors FM1 and FM2 are located far enough from the intermediate phase to obtain a relatively large sub-opening and also the two folding mirrors. Special emphasis was placed on controlling the corrosion conditions so that no corrosion conditions occurred in the fields FM1 and FM2. This is qualitatively illustrated in FIG. 8, which shows the marks of 18 selected field points around the edge of the rectangular field on the first folding mirror FM1 and the second folding mirror FM2. In each track, the beams are shown in ten equidistant opening steps. It is evident that the substantially elliptical lines corresponding to the beams of the different openings originating from the same field point do not intersect and are fitted without the intersection of the folding mirrors. This indicates that both the first and second folding mirrors are in regions free of corrosion conditions, ie in “corrosion-free” regions. In addition, a biconvex positive lens geometrically disposed in the double-path area between the folding mirrors FM1 and FM2 and the concave mirror and optically disposed relatively close to the first and second intermediate images is corroded. There is no area. As a result, the embodiment of FIG. 7 is relatively tolerant of surface defects and / or contamination on optical surfaces close to the intermediate images IMI1, IMI2.

The above description of the preferred embodiments has been given by way of example. Individual features may be implemented alone or in combination as embodiments of the invention, or may be implemented in other applications. In addition, they may represent advantageous embodiments that may be protected from the right to claim protection in the present application or to claim protection for the duration of the application as submitted. From the given specification, those skilled in the art will understand the present invention and the benefits thereof, as well as find various obvious changes and modifications to the disclosed structures and methods. Applicant therefore seeks to cover all such variations and modifications as fall within the spirit and scope of the invention as defined by the appended claims and their equivalents.

The contents of all claims form part of this specification by reference.

Claims (14)

1. A method of manufacturing a projection objective comprising: defining an initial design for a projection objective and optimizing the design using an advantage function, the method comprising: Defining a plurality of benefit function components each reflecting a particular quality parameter, wherein one of the benefit function components is a normalized validity representative of the effective irradiance normalized to the effective irradiance at the image surface of the projection objective Define a maximum irradiance requirement requiring that irradiance values not exceed a predetermined irradiance threshold on each optical surface of the projection objective except for the last optical surface directly adjacent to the image surface of the projection objective. Making; Calculating a numerical value for each of the advantage function components based on the corresponding feature of the preliminary design of the projection objective; Calculating, from the advantage function components, an overall benefit function that can be expressed as numerical items reflecting quality parameters; Continuously changing at least one structural parameter of the projection objective lens and recalculating the resulting overall advantage function for each successive change until the resulting overall advantage function reaches a predetermined acceptable value; Obtaining structural parameters of the optimized projection objective lens having a predetermined acceptable value for the resulting overall advantage function; And Implementing the parameters to produce a projection objective. The method of claim 1, Calculating the location and extent of potential corrosive regions in the projection objective; And Optimizing the structural parameters of the projection objective such that no optical surface is located in the corrosive region. The method according to claim 1 or 2, Defining a plurality of representative field points; Defining a pupil raster representing an array of raster points spaced apart from each other in the pupil plane of the projection objective; For each representative field point, calculating a ray trajectory of light rays originating from the representative field points and passing through the raster points of the pupil raster; For each optical surface, calculating intersections of the optical surface with light rays; For each optical surface, calculating a plurality of gradient parameters indicative of each gradient between intersections corresponding to neighboring raster points arranged directly adjacent to each other; Defining a gradient threshold representative of the minimum acceptable gradient between neighboring intersections; And Optimizing the structural parameters of the projection objective lens for each optical surface of the projection objective except for the last optical surface such that the gradient parameter is not below the gradient threshold. The method of claim 3, wherein Wherein the pupil raster is defined such that the pupil plane is subdivided into raster fields having substantially the same raster field area. The method according to claim 3 or 4, The pupil raster is defined in polar coordinates such that neighboring raster points have the same distance in the azimuth direction, and the pupil angle k is between 0 and k _max = NA.β. And i = 0, 1, ..., n, where NA is the image-side numerical aperture of the projection objective lens, and β is the magnification between the object field and the image field. The method according to any one of claims 1 to 5, Defining a plurality of representative field points; Calculating the luminous flux originating from the field points and the intersections of the luminous fluxes and the optical surfaces, wherein the intersection of the luminous flux with the optical surface is the actual sub-opening size defined by the area of the intersection area Defining an actual sub-opening with; Defining a sub-opening size threshold; For all optical surfaces of the projection objective except the last optical surface directly adjacent to the image surface of the projection objective, the actual sub-opening size for the selected field points is not below the sub-opening size threshold. Optimizing the structural parameters of the projection objective lens. 7. A projection objective made according to the method of any one of claims 1 to 6, comprising a plurality of optical elements configured to image a pattern provided on an object surface of a projection objective onto an image surface of the projection objective. Projection objective lens included. The method of claim 7, wherein The projection objective lens is a refraction type projection objective lens designed to form an off-axis object field disposed on an object surface with an off-axis image field disposed on an image surface, At least one concave mirror; At least one mesophase; And And at least one folding mirror arranged to deflect emission light from an object surface into the concave mirror or to deflect radiation light from the concave mirror to an image surface. The method of claim 8, The optical elements are: A first refractive objective lens portion forming a first intermediate image from radiation emitted from an object surface and including a first pupil plane; A second objective lens unit including at least one concave mirror for forming the first intermediate image into a second intermediate image and including a second pupil plane optically conjugate to the first pupil plane; And And a third refractive objective lens image forming the second intermediate image over the image surface and including a third pupil plane that is optically conjugate to the first and second pupil planes. The method of claim 9, The projection objective has exactly two intermediate images and / or the second objective has exactly one concave mirror, and the projection objective has a first folding for deflecting radiation from the object surface in the direction of the concave mirror. A projection and a second folding mirror for deflecting light emitted from the concave mirror in the direction of the image surface, and / or the projection objective is designed for immersion lithography with NA> 1. A projection objective comprising a plurality of optical elements having optical surfaces configured to image an off-axis object field disposed on an object surface into an off-axis image field disposed on an image surface of the projection objective lens. At least one concave mirror; At least one mesophase; And At least one folding mirror disposed to deflect radiation from an object surface into the concave mirror or to deflect radiation from the concave mirror to an image surface, A projection objective in which the structural parameters of the projection objective are adjusted such that no optical surface is located in the corrosive region. The method of claim 11, The optical elements are: A first refractive objective lens portion forming a first intermediate image from radiation emitted from an object surface and including a first pupil plane; A second objective lens unit including at least one concave mirror for forming the first intermediate image into a second intermediate image and including a second pupil plane optically conjugate to the first pupil plane; And And a third refractive objective lens image forming the second intermediate image over the image surface and including a third pupil plane that is optically conjugate to the first and second pupil planes. The method of claim 12, The projection objective has exactly two intermediate images and / or the second objective has exactly one concave mirror, and the projection objective has a first folding for deflecting radiation from the object surface in the direction of the concave mirror. A projection and a second folding mirror for deflecting light emitted from the concave mirror in the direction of the image surface, and / or the projection objective is designed for immersion lithography with NA> 1. The method according to claim 11, 12 or 13, And the at least one folding mirror is a flat mirror.