Invited Paper
`
`The Development of Dioptric Projection Lenses for DUV Lithography
`The Development of Dioptric Projection Lenses for DUV Lithography
`
`Willi Ulrich, Hans-Jiirgen Rostalski, Russ Hudyma*
`Willi Ulrich, Hans-JUrgen Rostaiski, Russ Hudyma*
`cz SMT AG, D-73446 Oberkochen, Germany
`CZ SMT AG, D-73446 Oberkochen, Germany
`*Paragon Optics, Inc., San Ramon, CA
`*paragon Optics, Inc., San Ramon, CA
`
`ABSTRACT
`ABSTRACT
`
`Advanced dioptric projection lenses from Carl Zeiss are used in some ofthe world's most advanced deep ultraviolet
`Advanced dioptric projection lenses from Carl Zeiss are used in some of the world's most advanced deep ultraviolet
`projection lithography systems. These lenses provide a resolution of better than 100 nm across the entire field of view with
`projection lithography systems. These lenses provide a resolution of better than 100 nm across the entire field of view with
`a level of aberration control that maximizes critical dimension uniformity and lithographic process latitude. These dioptric
`a level of aberration control that maximizes critical dimension uniformity and lithographic process latitude. These dioptric
`projection lenses are currently being used for critical layer device patterning for a wide array of complex logic, memory,
`projection lenses are currently being used for critical layer device patterning for a wide array of complex logic, memory,
`and application specific integrated circuits.
`and application specific integrated circuits.
`
`Zeiss' involvement in the develop of ultraviolet lenses goes back to the year 1 902, exactly 100 years ago, when Moritz von
`Zeiss' involvement in the develop of ultraviolet lenses goes back to the year 1902, exactly 100 years ago, when Moritz von
`Rohr calculated the first monochromatic ultraviolet micro-objectives for ultra-high resolution microphotography using a
`Rohr calculated the first monochromatic ultraviolet micro-objectives for ultra-high resolution microphotography using a
`line-narrowed source. The modem dioptric projection lenses for lithography are influenced by the collective experience in
`line-narrowed source. The modern dioptric projection lenses for lithography are influenced by the collective experience in
`the field of microscopy, and the more recent experience with early step-and-repeat lenses. This paper discusses some of the
`the field of microscopy, and the more recent experience with early step-and-repeat lenses. This paper discusses some of the
`foundations of modern dioptric designs in the context of this history, demonstrating that rapid synthesis of designs is
`foundations of modern dioptric designs in the context of this history, demonstrating that rapid synthesis of designs is
`possible using combinations of monochromatic microscope objectives and early step-and-repeat lenses from the 1970's.
`possible using combinations of monochromatic microscope objectives and early step-and-repeat lenses from the 1970's.
`The problems associated with ultra high numerical aperture objectives are discussed. Specifically, it is demonstrated that
`The problems associated with ultra high numerical aperture objectives are discussed. Specifically, it is demonstrated that
`aspheres can be used effective to reduce the volume of full field projection lenses, making the mechanical implementation
`aspheres can be used effective to reduce the volume of full field projection lenses, making the mechanical implementation
`of a 0.90 NA lens feasible in production. Several contemporary dioptric projection lens designs are reviewed in detail. The
`of a 0.90 NA lens feasible in production. Several contemporary dioptric projection lens designs are reviewed in detail. The
`extension of these designs to numerical apertures greater than 1 .0 using immersion techniques is demonstrated. These
`extension of these designs to numerical apertures greater than 1.0 using immersion techniques is demonstrated. These
`immersion lenses give the potential for 40nm resolution.
`immersion lenses give the potential for 4Onm resolution.
`
`Keywords:
`Keywords:
`
`lens design, dioptric lenses, monochromat, microlithography, microscopy, ultraviolet, aspheres,
`lens design, dioptric lenses, monochromat, microlithography, microscopy, ultraviolet, aspheres,
`immersion lithography
`immersion lithography
`
`1. INTRODUCTION
`1. INTRODUCTION
`
`The manufacture of integrated circuits with smaller and smaller features demands leading-edge projection lenses with
`The manufacture of integrated circuits with smaller and smaller features demands leading-edge projection lenses with
`specifications which no one would have considered possible a few years ago . The resolution (R) of a lithographic printing
`specifications which no one would have considered possible a few years ago. The resolution (R) of a lithographic printing
`system is expressed as:
`system is expressed as:
`
`R=k1X/NA
`
`(1)
`(1)
`
`where k\ is a process dependent factor, A is the wavelength of illumination, and NA is the numerical aperture. The process
`where k1 is a process dependent factor, X is the wavelength of illumination, and NA is the numerical aperture. The process
`dependent k1 factor takes into account several factors such as partial coherence and the influence of the resolution
`dependent k\ factor takes into account several factors such as partial coherence and the influence of the resolution
`enhancement techniques like off-axis illumination and phase shift masks. Wavelength scaling, numerical aperture scaling,
`enhancement techniques like off-axis illumination and phase shift masks. Wavelength scaling, numerical aperture scaling,
`and k1 process optimization have all been used to improve resolution. For example, a projection lens with a numerical
`and k\ process optimization have all been used to improve resolution. For example, a projection lens with a numerical
`aperture of 0.70 operating at 1 93 nm can achieve a resolution of 100 nm in resist, assuming a ki-factor of 0.36.
`aperture of 0.70 operating at 193 nm can achieve a resolution of 100 nm in resist, assuming a kl-factor of 0.36.
`
`Sematech's International Technology Roadmap for Semiconductors (lTRS) shows that leading edge 130 nm design rules
`Sematech's International Technology Roadmap for Semiconductors (ITRS) shows that leading edge 130 nm design rules
`are achieved today using either 248 nm or 193 nm technology. These high numerical aperture tools are almost exclusively
`are achieved today using either 248 nm or 193 nm technology. These high numerical aperture tools are almost exclusively
`supported by dioptric projection lens technology in a step and scan mode. The ITRS predicts that 100 nm design rules can
`supported by dioptric projection lens technology in a step and scan mode. The ITRS predicts that 100 nm design rules can
`be achieved in production using 193 nm lithography, meaning that proven dioptric lens technology will likely be the
`be achieved in production using 193 nm lithography, meaning that proven dioptric lens technology will likely be the
`technology of choice at this next device node at the critical layer. But even as the industry eventually moves to adopt new
`technology of choice at this next device node at the critical layer. But even as the industry eventually moves to adopt new
`
`158
`
`International Optical Design Conference 2002, Paul K. Manhart, José M. Sasián, Editors,
`Proceedings of SPIE Vol. 4832 (2002) © 2002 SPIE · 0277-786X/02/$15.00
`
`ZEISS 1011
`
`

`

`technologies at the critical layer (e.g., 1 57 nm and extreme ultraviolet) to gain even higher resolution and smaller design
`technologies at the critical layer (e.g., 157 nm and extreme ultraviolet) to gain even higher resolution and smaller design
`rules, the demands on imaging at the sub-critical layers also increases. So the critical layer scanners oftoday will become
`rules, the demands on imaging at the sub-critical layers also increases. So the critical layer scanners of today will become
`the sub-critical layer scanners of tomorrow. Effectively this means that dioptric projection lenses will continue to be the
`the sub-critical layer scanners of tomorrow. Effectively this means that dioptric projection lenses will continue to be the
`work horse for lithography for many years to come.
`work horse for lithography for many years to come.
`
`As a company, Carl Zeiss has been involved in the development of deep ultraviolet (DUV) imaging systems for over 100
`As a company, Carl Zeiss has been involved in the development of deep ultraviolet (DUV) imaging systems for over 100
`years. The roots of this fundamental work in deep ultraviolet microscopy are often seen in today' s dioptric projection lenses
`years. The roots of this fundamental work in deep ultraviolet microscopy are often seen in today's dioptric projection lenses
`with little imagination. Line narrowing was used then and is used now to overcome problems with dispersion and lack of
`with little imagination. Line narrowing was used then and is used now to overcome problems with dispersion and lack of
`suitable materials in the deep ultraviolet. The imaging group closest to the wafer in a modem dioptric lens for lithography
`suitable materials in the deep ultraviolet. The imaging group closest to the wafer in a modem dioptric lens for lithography
`often resembles a monochromatic ultraviolet objective for microscopy with several aplanatic or near-aplanatic surfaces in
`often resembles a monochromatic ultraviolet objective for microscopy with several aplanatic or near-aplanatic surfaces in
`the final focusing group. The groups between the reticle and the aperture stop often have a series of bulges and waists,
`the final focusing group. The groups between the reticle and the aperture stop often have a series of bulges and waists,
`reminiscent of the lenses described by Glatzel over 20 years ago.
`reminiscent ofthe lenses described by Glatzel over 20 years ago.
`
`The development of modern dioptric projection lenses for deep ultraviolet lithography can be seen as an extension of this
`The development of modem dioptric projection lenses for deep ultraviolet lithography can be seen as an extension of this
`collective work. In the following sections, we discuss the historical foundation of our deep ultraviolet lens work from the
`collective work. In the following sections, we discuss the historical foundation of our deep ultraviolet lens work from the
`early 1900s starting with monochromatic high numerical aperture microscope objectives until the early 1980s with early
`early 1900s starting with monochromatic high numerical aperture microscope objectives until the early 1980s with early
`ultraviolet repeater lenses. We demonstrate how one skilled in the art is able to synthesize a high numerical aperture
`ultraviolet repeater lenses. We demonstrate how one skilled in the art is able to synthesize a high numerical aperture
`dioptric lens by combining these different design forms as a starting point for further optimizations. Since the potential of
`dioptric lens by combining these different design forms as a starting point for further optimizations. Since the potential of
`dioptric projection lenses has not yet been fully exploited, we discuss the progress with hyper numerical aperture designs
`dioptric projection lenses has not yet been fully exploited, we discuss the progress with hyper numerical aperture designs
`with numerical apertures to 0.90. A monochromatic design example is provided for use with a highly line-narrowed laser
`with numerical apertures to 0.90. A monochromatic design example is provided for use with a highly line-narrowed laser
`(0.25 pm). It is shown that the system can be designed to approach the "zero aberration" condition as required by modem
`(0.25 pm). It is shown that the system can be designed to approach the "zero aberration" condition as required by modem
`lithographic process to minimize linewidth variation across the imaging field. Dioptric projection lens can also utilize a
`lithographic process to minimize linewidth variation across the imaging field. Dioptric projection lens can also utilize a
`second material, thus forming "pseudo doublets" to improve the state of color correction at the expense of lens complexity.
`second material, thus forming "pseudo doublets" to improve the state of color correction at the expense of lens complexity.
`Design examples using this construction are also presented.
`Design examples using this construction are also presented.
`
`Finally, we provide the motivation to once again borrow from our past and explore the field of immersion lithography.
`Finally, we provide the motivation to once again borrow from our past and explore the field of immersion lithography.
`Immersion imaging has already been fully established in microscopy for more than 100 years. The resolution gain offered
`Immersion imaging has already been fully established in microscopy for more than 100 years. The resolution gain offered
`by immersion could allow the lithographic industry to continue to satisfiy Moore's law, which states that the performance of
`by immersion could allow the lithographic industry to continue to satisfiy Moore's law, which states that the performance of
`leading-edge integrated circuits doubles every 1 8 months with decreasing manufacturing cost. We provide two examples of
`leading-edge integrated circuits doubles every 18 months with decreasing manufacturing cost. We provide two examples of
`immersion lenses for both 193 nm and 157 nm lithography designed so that a numerical aperture value of 1.1 is obtained.
`immersion lenses for both 193 nm and 157 nm lithography designed so that a numerical aperture value of 1.1 is obtained.
`These lenses are capable of resolutions of 50 nm and 40 nm, respectively, suggesting that immersion could be used with
`These lenses are capable of resolutions of 50 nm and 40 nm, respectively, suggesting that immersion could be used with
`even larger numerical apertures to achieve linewidths to 30 nm usingjust basis dioptric projection lenses that are already
`even larger numerical apertures to achieve linewidths to 30 nm using just basis dioptric projection lenses that are already
`well understood.
`well understood.
`
`2. HISTORY OF ULTRAVIOLET LENS DEVELOPMENT AT CARL ZEISS
`2. HISTORY OF ULTRAVIOLET LENS DEVELOPMENT AT CARL ZEISS
`
`From the Rayleigh resolution formula, we know that the limiting resolving power of optical systems depends linearly on the
`From the Rayleigh resolution formula, we know that the limiting resolving power of optical systems depends linearly on the
`wavelength. So improvements to resolution are enabled by wavelength reduction. Ernst Abbe had already recognized this
`wavelength. So improvements to resolution are enabled by wavelength reduction. Ernst Abbe had already recognized this
`and recommended the use of UV radiation for microscopic projection as early as 1 874. However, the only two materials
`and recommended the use ofUV radiation for microscopic projection as early as 1874. However, the only two materials
`with good permeability, fluorspar and mountain crystal could not be used originally for constructing a microscope objective
`with good permeability, fluorspar and mountain crystal could not be used originally for constructing a microscope objective
`because ofthe double refraction ofthe mountain crystal'. Only in 1899 after M.Herschkowitsch2 succeeded in
`because ofthe double refraction of the mountain crystal). Only in 1899 after M.Herschkowitsch2 succeeded in
`manufacturing amorphous quartz which was sufficiently homogeneous and tension-free for optical purposes and August
`manufacturing amorphous quartz which was sufficiently homogeneous and tension-free for optical purposes and August
`Köhler at the same time made clear progress in the development of a spectrally narrowed ultraviolet light, did the first UV
`Kohler at the same time made clear progress in the development of a spectrally narrowed ultraviolet light, did the first UV
`microscope become a real possibility.
`microscope become a real possibility.
`
`August Köhler3'4 replaced the previously common light sources with a continuous spectrum by light sources with a linear
`August Kohler3,4 replaced the previously common light sources with a continuous spectrum by light sources with a linear
`spectrum. The discharge spark of a Leydener bottle jumping between metal electrodes proved particularly suitable. With
`spectrum. The discharge spark of a Leydener bottle jumping between metal electrodes proved particularly suitable. With
`mountain crystal prisms he isolated the single lines and got a series of monochromatic images of the spark. He projected
`mountain crystal prisms he isolated the single lines and got a series of monochromatic images ofthe spark. He projected
`these into the entrance pupil of the objective. He used a fluorescent image converter as a receiver. August Köhler worked at
`these into the entrance pupil of the objective. He used a fluorescent image converter as a receiver. August Kohler worked at
`first with the magnesium line at 280nm, a little later with the sharper cadmium line at 275nm. The figures 1 and 2 show the
`first with the magnesium line at 280nm, a little later with the sharper cadmium line at 275nm. The figures 1 and 2 show the
`UV-spectral apparatus and a principle diagram of the used monochromator. The refractive index of fluorspar and quartz are
`UV-spectral apparatus and a principle diagram of the used monochromator. The refractive index of fluorspar and quartz are
`so close together that the achromatic structure which has since been used for large numeric apertures failed in 1900.
`so close together that the achromatic structure which has since been used for large numeric apertures failed in 1900.
`In the Spring of 1902 the Zeiss scientist Moritz von Rohr5 discovered a totally new type of lens which, with a suitable lens
`In the Spring of 1902 the Zeiss scientist Moritz von RohrS discovered a totally new type oflens which, with a suitable lens
`combination was aplanar for any certain wavelength to be chosen within limits. In these lenses the refractions were
`combination was aplanar for any certain wavelength to be chosen within limits. In these lenses the refractions were
`
`Proc. of SPIE Vol. 4832 159
`
`

`

`distributed evenly onto the individual surfaces by following them with a series of aplanar meniscuses of diminishing
`distributed evenly onto the individual surfaces by following them with a series of aplanar meniscuses of diminishing
`strength. To improve the aplanatism, Moritz von Rohr has additionally used a dispersive meniscus as a means of
`strength. To improve the aplanatism, Moritz von Rohr has additionally used a dispersive meniscus as a means of
`correction6'7. The first three monochromates ofmelted quartz for 280nm and 275nm resulted.
`correction6,7. The first three monochromates of melted quartz for 280nm and 275nm resulted.
`
`I Pm
`
`?m
`
`rIht 7;:
`
`t
`
`Fig. 1: UV-spectral apparatus developed by
`Fig. 1: UV-spectral apparatus developed by
`AUJ!Ust Kohler in 1904
`Au,gust Köhler in 1904
`
`Fig. 2: Line-narrowed source for
`Fig. 2: Line-narrowed source for
`UV-Microphotography
`UV-Microphotography
`
`1933
`
`I 96
`
`kr-
`r— 'i
`
`I IOflOChr()fllf
`NAO35
`li5rini
`
`1
`I
`
`klonochromal
`NA125
`(a 257nm
`
`I Itlafluar
`\A12S
`( 28Onm
`
`j
`
`Fig. 3: Carl Zeiss UV-Microscope Objectives
`Fig. 3: Carl Zeiss UV-Microscope Objectives
`
`In the 1930s a new series ofmonochromates was developed for 257nm8. Further calculations were made at Carl Zeiss at the
`In the 1930s a new series ofmonochromates was developed for 257nm8• Further calculations were made at Carl Zeiss at the
`end ofthe 1950s. The figure 3 shows selected design examples from the variety ofthese UV monochromates.
`end ofthe 1950s. The figure 3 shows selected design examples from the variety of these UV monochromates.
`The use ofradiation sources with a continuous spectrum in combination with monochromators and a spectral bandwidth of
`The use of radiation sources with a continuous spectrum in combination with monochromators and a spectral bandwidth of
`approx. 5nm led to the development of achromatic UV-VIS lenses of quartz and CaF2 — one even with additional LiF -,
`approx. 5nm led to the development of achromatic UV -VIS lenses of quartz and CaF2 - one even with additional LiF -,
`which Carl Zeiss has been offering as ultrafluars since the 1960s9b0.
`which Carl Zeiss has been offering as ultrafluars since the 1960s9,JO.
`
`The development of ultraviolet lenses at Carl Zeiss found a different path starting in the late 1960s. Ultraviolet reduction
`The development of ultraviolet lenses at Carl Zeiss found a different path starting in the late 1960s. Ultraviolet reduction
`lenses intended for the production of masks and later for direction projection onto a wafer were developed at the g-line, h-
`lenses intended for the production of masks and later for direction projection onto a wafer were developed at the g-line, h(cid:173)
`line, and later i-line. These lenses were derived from photographic objectives and allowed to grow in size to become as
`line, and later i-line. These lenses were derived from photographic objectives and allowed to grow in size to become as
`relaxed as possible. These lenses required very tight control of distortion. Compared to microscope objectives, these lenses
`relaxed as possible. These lenses required very tight control of distortion. Compared to microscope objectives, these lenses
`were perhaps the first multi-element lenses requiring diffraction-limited performance with a very tight control of distortion
`were perhaps the first multi-element lenses requiring diffraction-limited performance with a very tight control of distortion
`across a large field". Since the correction ofthe large image field proved to be particularly difficult, Erhard Glatzel of Carl
`across a large field ll . Since the correction of the large image field proved to be particularly difficult, Erhard Glatzel of Carl
`Zeiss recommended designs of the distagon type, a combination ofretrofocus lens and double Gauss lens in the rear groupl2.
`Zeiss recommended designs of the distagon type, a combination of retrofocus lens and double Gauss lens in the rear group12.
`Due to the non-availability of high-index materials, he proposed the correction means of multiple "bulges"3 according to
`Due to the non-availability of high-index materials, he proposed the correction means of multiple "bulges .. 13 according to
`the recommendation of H.Slevogtl4 in order to reduce field curvature. Figure 4 shows a selection of the first Carl Zeiss
`the recommendation of H.Slevogt'4 in order to reduce field curvature. Figure 4 shows a selection of the first Carl Zeiss
`
`160 Proc. of SPIE Vol. 4832
`
`

`

`repeater lenses from the 70s. The double "bulging" of the lenses can be seen here very nicely. Gerhard Ittner'5 of Carl Zeiss
`repeater lenses from the 70s. The double "bulging" of the lenses can be seen here very nicely. Gerhard IttnerlS of Carl Zeiss
`shows in an example of a further design development ofthe S-Planar 1.6/50 how the number oflenses has to be increased at
`shows in an example of a further design development of the S-Planar 1.6/50 how the number oflenses has to be increased at
`an aperture increase from 0.28NA to 0.40NA.
`an aperture increase from O.28NA to O.4ONA.
`
`In 1980 Phillips and Buzawa described new high-resolution lenses for 365nm lithographyl6. They, as well as Glatzel and
`In 1980 Phillips and Buzawa described new high-resolution lenses for 365nm lithography16. They, as well as Glatzel and
`Braat'7, recommended avoiding balancing large high order aberrations because this usually introduces zonal imbalances.
`Braat17, recommended avoiding balancing large high order aberrations because this usually introduces zonal imbalances.
`The proposed correction means like further splitting of positive lens elements, adding a thick meniscus element near the
`The proposed correction means like further splitting of positive lens elements, adding a thick meniscus element near the
`image and adding an aspheric at the stop were well known and used in different recent optical systems like microscope
`image and adding an aspheric at the stop were well known and used in different recent optical systems like microscope
`objectives and camera lenses.
`objectives and camera lenses.
`
`Fig. 4: A selection ofthefirst Carl
`Fig. 4: A selection of the first Carl
`Zeiss repeater lensesfrom the 70s
`Zeiss repeater lenses from the 70s
`
`All later all-refractive lens designs go back to the concepts of Erhard Glatzel. In 1987, Joseph Braat demonstrated on an 18-
`All later all-refractive lens designs go back to the concepts of Erhard Glatzel. In 1987, Joseph Braat demonstrated on an 18-
`lens O.38NA design example from Cerco how the double bulging proposed by Glatzel could be logically developed. An
`lens 0.38NA design example from Cerco how the double bulging proposed by Glatzel could be logically developed. An
`additional first positive lens group simplifies the correction of the Petzval sum and at the same time enables the demand for
`additional first positive lens group simplifies the correction of the Petzval sum and at the same time enables the demand for
`telecentering on both sides to be met18. Carl Zeiss Jena already used this lens structure in 1979 to design the double-sided
`telecentering on both sides to be metl8• Carl Zeiss Jena already used this lens structure in 1979 to design the double-sided
`telecentric projection lens UM-ADR with a numerical aperture of0.4NAI9 •
`telecentric projection lens UM-AUR with a numerical aperture of O.4NA19.
`
`This history can be drawn upon as we consider the development of new lens forms that satisfy more contemporary
`This history can be drawn upon as we consider the development of new lens forms that satisfy more contemporary
`lithographic requirements. Combining scaled versions of some of these traditional design forms enables synthesis of "new"
`lithographic requirements. Combining scaled versions of some of these traditional design forms enables synthesis of "new"
`starting points that contain the waists and bulges as decribed by Glatzel, as well as the aplanatic constructions contained in
`starting points that contain the waists and bulges as decribed by Glatzel, as well as the aplanatic constructions contained in
`the early monochromatic microscope objectives. Figure 5 illustrates a fairly relaxed example that was created via the
`the early monochromatic microscope objectives. Figure 5 illustrates a fairly relaxed example that was created via the
`combination of an early photorepeater (S-Planar 50 mm, f/i .6) and a monochromatic microscope objective first calculated
`combination of an early photorepeater (S-Planar 50 mm, fIl.6) and a monochromatic microscope objective first calculated
`in the 1 930s. The photorepeater objective is turned around since its image space numerical aperture is compatiable with the
`in the 1930s. The photorepeater objective is turned around since its image space numerical aperture is compatiable with the
`required numerical aperture at the mask. After this flip, each part of the lens is scaled to give the desired reduction from
`required numerical aperture at the mask. After this flip, each part of the lens is scaled to give the desired reduction from
`mask to wafer of 4x. The cemented doublets in front ofthe stop are split and the bendings ofthe G5 group are readjusted
`mask to wafer of 4x. The cemented doublets in front of the stop are split and the bendings of the G5 group are readjusted
`somewhat to accounts for conjugates shifts. This lens has one strong waist and one weak waist near the stop. With material
`somewhat to accounts for conjugates shifts. This lens has one strong waist and one weak waist near the stop. With material
`changes and direction from one skilled in the art, this starting point could be made to take either the one-waist (two bulges)
`changes and direction from one skilled in the art, this starting point could be made to take either the one-waist (two bulges)
`or two-waist (three bulges) form.
`or two-waist (three bulges) form.
`
`Taking a Maxwellian view, if an observer stands at the aperture stop and peers toward the wafer, he sees a lens that looks
`Taking a Maxwellian view, if an observer stands at the aperture stop and peers toward the wafer, he sees a lens that looks
`like the early microscope objective. Looking the other way, the observer sees the photo-repeater. With a little imagination,
`like the early microscope objective. Looking the other way, the observer sees the photo-repeater. With a little imagination,
`many contemporary dioptric projection lenses for either KrF or ArF lithography can be thought of in this manner. What
`many contemporary dioptric projection lenses for either KrF or ArF lithography can be thought of in this manner. What
`makes these designs work is the fact that you have positive and negative lens groups working together to balance the
`makes these designs work is the fact that you have positive and negative lens groups working together to balance the
`different aberrations using their respective undercorrected and overcorrected components.
`different aberrations using their respective undercorrected and overcorrected components.
`
`Proc. of SPIE Vol. 4832 161
`
`

`

`Zeiss S-Planar 1.6/50 (1975)
`Zeiss S-Planar 1.6 /50 (1975)
`
`Zeiss microscope
`Zeiss microscope
`objective (1933)
`objective (1933)
`
`Example: NA 0.80 @248 nm
`Example: NA 0.80 @ 248 nm
`
`GI
`GJ
`
`Gl
`G2
`
`G3
`G3
`
`G4
`G4
`
`G5
`GS
`
`Fig. 5: Rapid synthesis of designs is
`Fig. 5: Rapid synthesis ofdesigns is
`possible by simply combining scaled
`possible by simply combining scaled
`versions of traditional lens forms.
`versions oftraditional lens forms.
`Here an early photo-repeater is
`Here an early photo-repeater is
`flipped, scaled, and combined with a
`flipped, scaled, and combined with a
`scaled monochromatic objective from
`scaled monochromatic objective from
`the 1930s.
`the 1930s.
`
`3. COMPACT HIGH-NA LENSES
`3. COMPACT BIGB-NA LENSES
`
`At the SPIE 200020 it was demonstrated how new refractive designs of a compact form could be produced for DUV
`At the SPIE 200020 it was demonstrated how new refractive designs of a compact form could be produced for DUV
`microlithography by using aspheres21 • The lens volume could be significantly reduced in comparison with earlier, purely
`microlithography by using aspheres21. The lens volume could be significantly reduced in comparison with earlier, purely
`spherical versions22. By further progress in design over the last two years, optical designs with a numeric aperture of
`spherical versions22 • By further progress in design over the last two years, optical designs with a numeric aperture of
`0.9NA23 and greater have been achieved whose lens volumes are even below the extrapolation curve forecast two years ago
`O.9NA23 and greater have been achieved whose lens volumes are even below the extrapolation curve forecast two years ago
`(Figure 6).
`(Figure 6).
`
`.... - .. ---.---~--.
`S.O ..-.. -.-.. ----.. -----~ ... -
`4.5 t-------.... -----~ ........................................................................ · .. i .... · .......... •
`4.0 t - - - - -.. - - - - -........................................................ ,
`
`J 3.5
`~ 3.0 t-.... ---.--------------...................................................... ", ..... _ ..... _ ......... ,,<..l_
`.~ 2.5 + .................................................... , ........ , ................... , ...... , ......................... , .. , ..................... ,,. .......... ..
`i 2.0 t" .. , .. , .. , ............ , ...... , .. " .. , .. , ....................... " ........................................ -.... -.. ' .................. , ................................ ~"' ........ , ............ "". ........ :"'~
`
`1.5 1 ....... ' ............ ' ...... ' ........ ' ............ ' .............. ' ...... -........... ·----:~-____:7'7~--
`1.0 + ............ · ............ · .................... }!/91 ...... , ..... :~""""---:-:;;..".,""'--------;
`0.5 +':::'~r.....-=::' ............... · .. · .. · ... t lJC ) ! t - - - - - - - - -
`0.0 +------,-----,-----,----.,..--
`0.60
`070
`0.70
`0.80
`0.50
`0.90
`NA
`NA
`
`2002
`
`Fig. 6: Lens volume as afunction of numerical
`Fig. 6: Lens volume as a function of numerical
`aperture. Compact high- and hyper-NA lenses
`aperture. Compact high- and hyper-NA lenses
`become feasible through the use of aspheres.
`becomefeasible through the use of aspheres.
`Upper line: spherical designs
`Upper line: spherical designs
`Lower lines: aspherical designs
`Lower lines: aspherical designs
`
`These material saving refractive designs are distinguish