JPH0476489B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION Field of Industrial Application The present invention relates to an exposure apparatus for devices having fine patterns, particularly semiconductor devices and the like having a submicron rule of 1 micron or less.

BACKGROUND OF THE INVENTION Semiconductor devices have recently become more densely packed, and the dimensions of fine patterns of each element are now 1 micron or less. Conventionally, alignment between a photomask and an LSI wafer during LSI manufacturing has been done by using alignment marks provided on the wafer, rotating and parallelly moving a stage on which the wafer is mounted, and aligning the marks on the photomask with the marks on the wafer. This was done by overlapping each other, but the alignment accuracy was ±
It is about 0.3 microns, and when forming submicron elements, the alignment accuracy is poor and it is not practical. Also, S. Austin [Applied physics Letters]
Vol31No.7P.428, 1977] et al. showed that the alignment method using interferometry, as shown in Figure 2,
The incident laser beam 1 is made incident on a photomask 2, diffracted by a grating 3 formed on the photomask 2, and this diffracted light is again diffracted by a grating 5 formed on a wafer 4, thereby producing diffracted light 6, Get 7, 8... When this diffracted light is expressed as a binary representation of the diffraction order on the photomask and the diffraction order on the wafer, diffracted light 6 is (0, 1), diffracted light 7 is
(1, 1), the diffracted light 8 can be expressed as (-1, 2)... This diffracted light is collected at one point by a lens and the light intensity is measured. The diffracted light has a light intensity at a position symmetrical to the incident laser beam 1. The photomask 2 and the wafer 4 can be aligned by matching the intensities of the diffracted light observed on the left and right sides. In this method, the alignment accuracy is several
It is said to be 100Å.

However, in this method, the photomask 2
Since the alignment between the photomask 2 and the wafer 4 is greatly influenced by the distance D between the photomask 2 and the wafer 4, accuracy in the distance D is required. Further, it is necessary to bring the photomask 2 and the wafer 4 close to each other and align them while maintaining the accuracy of the distance D, which complicates the apparatus and poses a problem in practical use.

In addition, there is a method for aligning elements with submicron line widths through observation using secondary electron emission from the elements, but since it cannot be handled in the atmosphere, the throughput for manufacturing LSIs becomes small and practical. There was a problem above.

In addition, the conventional example shown in Fig. 3 [IEEE, transon] ED ED-
26, 4, 1974, 723, Gijs Bouwuis], a laser beam is incident on the Fourier transform surface of a microlens shown by two lens systems L ₁ and L ₂ , and is projected onto a wafer via lens L ₂ . The formed grating is illuminated with a beam, and only the ±1st-order light diffracted from the grating by the spatial filter SF is incident on the reticle R through the lens systems L ₁ and L ₂ to form interference fringes in the vicinity of the reticle R. A configuration is shown in which a photodetector D detects the light generated and passes through a grating provided on the reticle to align the wafer W and the reticle R. In the configuration shown in FIG. 3, it is stated that the position cannot be corrected for the asymmetric grating formed on the wafer W, and this is not possible in all processes and marks in the alignment mark manufacturing method. Although it is possible, it has not yet been put into practical use.

Problems to be Solved by the Invention In view of these conventional problems, the present invention provides an accurate and easy positioning of an LSI reticle and wafer, which allows alignment of fine patterns in the atmosphere and with a simple configuration. The aim is to provide an exposure device that enables alignment with high productivity.

Means for Solving the Problems In order to achieve highly accurate positioning in a projection exposure apparatus, the present invention provides a first lens among a light beam whose wavefront is split by a grating formed on a reticle surface. A suitable light beam in the spectral plane is passed through a spatial filter, passed through a second lens system, a projection lens, and projected onto a second grating provided on the substrate. From the second grating, diffracted light is diffracted, which passes in the opposite direction into a second projection lens and is guided to a photodetector. When two beams of light are projected onto the second grating on the substrate from appropriate directions, the diffracted beams are diffracted in the overlapping direction,
Each interferes. By detecting the light intensity of this interfering diffracted light, highly accurate positioning can be achieved. Furthermore, by using separate systems for the alignment optical system and reduction projection optical system, a alignment system with high reproducibility can be realized.

Function By combining the alignment optical system and the projection lens, it is possible to separate the optical system that reduces and projects the reticle image from the optical system that performs high-precision alignment. Since exposure can be performed without removing and attaching the spatial filters, mirrors, etc. used, more accurate positioning can be achieved.

Embodiment An embodiment of the optical system according to the present invention is shown in FIG. Light emitted from the light source 11 (In order to obtain clearer coherence and deeper depth of focus, the configuration is assumed to be a laser beam, but the overall optical system is a white interference optical system, and a mercury lamp is used.) ) is incident on the entrance pupil of the first lens system 15.

In the following description, in order to briefly describe the principle of the present invention, it is assumed that the reticle is illuminated by a parallel light beam, and that the first and second lens systems 15 and 17 are Fourier transform lenses, but this is not necessarily the case. It does not have to be a Fourier transform lens.

A reticle 14 is disposed between the light source 11 and the first Fourier transform lens 15, and the image emitted from the pattern of the first grating 10 of the reticle 14 as a secondary light source is once transformed by the first Fourier transform lens 15. The light is focused, and then the image of the pattern on the reticle 14 is transferred to the wafer 18 through the second Fourier transform lens 17.
The image is projected through the reduction projection optical system 19. The diffracted light (Fourier spectrum) of the pattern of the grating 10 on the reticle is spatially distributed on the rear focal plane of the first Fourier transform lens 15, and in the present invention, a spatial filter is provided on this Fourier transform surface. 16 is arranged and filtered in the spectral plane, and the pattern of the grating 10 formed on the reticle 14 is filtered in the spectral plane, thereby generating interference fringes 20 on the surface of the wafer 18.

Second grating 21 formed on semiconductor wafer 18
From there, the diffracted light 22 is diffracted, returns in the opposite direction through the reduction projection optical system 19 and the second lens 17, and is guided to the photodetector 23 by a mirror placed at the position of the spatial filter 16.

The above is the optical system for aligning the reticle 14 and the wafer 18. On the other hand, the circuit pattern of the reticle 14 is illuminated by the projection light source 12 and the illumination optical system 13, and its projected image is projected by the reduction projection optical system 19.
The image is formed on the wafer 18 through the wafer 18. In this way, an optical system for normal circuit pattern exposure is formed.

As described above, the present invention has a light source for alignment and its optical system, and a light source for exposure and its optical system.The spatial filter is placed in the alignment optical system and used, and after alignment is completed, the reticle is removed. An image of the circuit pattern above is reduced and projected onto the wafer.

FIG. 4 shows a reticle used in the exposure apparatus of the present invention. FIG. 4a is a plan view of the reticle 14, and FIG. 4b is a sectional view taken along the line A--A in a. The reticle 14 consists of a circuit pattern section 42 and its peripheral section 43, and in the portion of the peripheral section 43 that corresponds to the scrine line, there are positioning grating patterns 41, 41' corresponding to the first grating 10 in FIG. It is formed. Incident light 44 enters the reticle 14, and as shown in FIG.
A plurality of diffracted lights are diffracted as shown below. The shield section 43 surrounding the pattern 41 is formed of a film of chromium oxide or the like, and allows the incident light 44 to pass only through the inside of the pattern 41.

In the example shown in FIG. 4, an amplitude grating pattern 41 is used to obtain the diffracted light, but this grating may be a phase grating or, if the incident light is diagonally incident, an Ethieret grating.

FIG. 5 is a diagram further explaining the principle of the exposure apparatus of the present invention. Light with wavelength λ emitted from light source 11 illuminates grating 41 on the reticle.

The phase grating pattern 41 on the reticle 14 is placed at the position x ₁ of the front focal point f ₁ of the first Fourier transform lens 15 . The pitch P ₁ of the phase grating pattern 41 and the diffraction angle θ ₁ of the diffracted light have the following relationship: P ₁ sin θ _o =nλ (n=0, ±1, ±2, . . . ). The light diffracted into a plurality of light beams in this manner enters the Fourier transform lens 15, and further forms a Fourier spectrum image corresponding to each diffracted light beam on the rear focal plane. The coordinate ζ ₆₁ corresponding to the Fourier spectrum of the first-order diffracted light is expressed as ζ ₆₁ = f ₁ sin θ ₁ P ₁ sin θ ₁ = λ, and the Fourier spectrum of the O-th order diffracted light ζ ₆₀ ζ ₆₀ = f ₁ sin θ ₀ = 0 The Fourier spectrum image is focused on the Fourier transform plane while being completely separated from the As shown in FIG. 1, a spatial filter 16 is arranged on this Fourier transform surface, and as shown in FIG. The spectrum of the folded light and the aperture pattern (excluding the zero-order light component) is passed. This diffracted light passes through the second Fourier transform lens 17 and further passes through the wafer 18W.
projected on top.

However, the reduction projection lens system 19 is omitted in FIGS. 5 and 6.

The image projected onto the wafer W approximately focuses the image of the opening (pattern 41) on the reticle, and
The ±1st-order light components of the grating pattern 41 interfere with each other to form new pitched interference fringes. Here, the pitch P ₂ of the interference fringe is given by P ₂ =λ/2sinθ ₂ . At this time, the first Fourier transform lens 1 is placed on the front focal plane of the second Fourier transform lens 17.
Since the Fourier transform surface of 5 is set, there is a relationship of f ₁ sin θ ₁ = f ₂ sin θ ₂ = ζ ₆₁ .

first and second Fourier transform lenses 15, 17,
Furthermore, between the images passed through the reduction projection optical system with reduction ratio m, P ₂ = mλ/2sinθ ₂ = mf ₂ /f ₁・λ/2sinθ ₂ = mf ₂ /f ₁
・P ₁ /2 ...There is the following relationship (1). Therefore, the pitch P ₂ of the interference fringes generated on the wafer W is half the pitch of the projected image of the grating pattern 41 on the reticle when f ₁ =f ₂ . A second grating G is formed on the wafer W by the projected image of the grating 41, and when the light beams 111 and 112 are respectively irradiated onto this grating G, they are each diffracted by the grating G that splits the wavefront. light can be obtained. Furthermore, when the two light beams 111 and 112 are irradiated onto the wafer W at the same time, interference fringes are generated, and in this case, the lights diffracted by the grating G on the wafer W interfere with each other, and the interfered light is converted into a light beam. Detector D detects the light intensity information indicating the positional relationship between the interference fringes and the grating G.

Observed light intensity I on photodetector D in Figure 6
is I=u _A ² +u _B ² +u _A ^*・u _B +u _A・u _B ^* However, u _A and u _B are the amplitude intensities of the luminous fluxes 111 and 112, respectively, u _A ^* and u _B ^* are the conjugate complex amplitudes. be.

u _A ² = A ² (sin ^N δA/2/sin δA/2) ² , u _B ² = B ² (sin ^N
δB/2/sinδB/2) ² u _A ^*・u _B +u _A・u _B ^* =2・A・Bcos | (N-1) δA − δB/2 +Kx (sinθ _A − sinθ _B ) | ×sin ^N δA/2・sin ^N δB/2/sin δA/2・sin δB/
2 (where A and B are constants, N: number of lattices, δA,
δB is the left optical path between the lights diffracted by two adjacent gratings, x is the relative positional relationship between the interference fringes of the light beams 111 and 112 and the gratings, and θ _A and θ _B are the light beams 111 and 112, respectively. 112 and the perpendicular to the wafer).

Figure 7 shows the light intensity I of the diffracted light when both the light beams 111 and 112 are simultaneously irradiated onto the grating G of the wafer.
The observation angle dependence of This is a diagram when the pitch of the generated interference fringes is 1 μm, and the pitch of the grating G is 2 μm. A sharp peak in the light intensity appears only when the pitch of the grating G is an integral multiple of the pitch of the interference fringes, as shown by the light intensity I. In Figure 6, the observation angle is −
Nine peaks appear when changing from π/2 to π/2, and the peak at θ ₂ includes the incident light 111, 112
The 0th order diffracted lights of the two overlap. The peak at θ ₄ corresponds to the peak of the first-order diffracted light when the interference fringes and the grating pitch are equal. Each peak of θ ₁ to _{θ 5} contains positional information between the interference fringe and the emitter G on the wafer.

Each peak is (-3, +5), (-2, +4),
...It is observed as the combined light intensity of the diffracted light of (+5, -3).

In particular, at θ ₁ where the (-1, +1) lights overlap, the light intensities of the diffracted lights are equal, so the contrast of the moire light to be detected is high.

Next, the operation of the exposure apparatus according to the present invention will be explained.

First, the reticle 14 is inserted into the optical path, and the wafer 18 is loaded. Reticle 14 and wafer 1
8, rough positioning is performed in advance. Next, the reticle 14 and the first grating 10 are irradiated with light emitted from the light source 11. First grating 1 on reticle 14
A plurality of diffracted lights are diffracted from zero and guided to the first lens system 15, and a spatial filter 16 is arranged so that only the ±1st-order lights selectively pass through. ±1
The next light passes through the second lens system 17 and passes through the reduction projection optical system 19 to generate interference fringes near the wafer 18 . When the wavelengths of the light source 11 and the projection light source 12 are different, the ratio of the focal lengths f ₁ and f ₂ of the first and second lens systems is changed, and the second grating 21
is set to be an integral multiple of the pitch of the interference fringes. If the projection light source 12 is a pulsed light source such as an excimer laser, a continuous wave laser, a mercury lamp, or the like can be used as the alignment light source 11. Furthermore, when an excimer laser or the like is used, matching can be achieved with a positioning light source whose wavelength is limited. Furthermore, when the resist sensitivity on the wafer 18 is limited to a certain wavelength range,
The wavelength of the alignment light source 11 can be selected to a wavelength that has no resist sensitivity, and the alignment marks can be protected.

The light diffracted from the second grating 21 is detected by the photodetector 23, and is applied to the wafer 1 according to the above-described principle.
8 and the reticle 14 with high precision. When the alignment is complete, the projection light source 1
2, the reticle image is projected by the reduction projection optical system 19, and the circuit pattern on the reticle is formed on the wafer.

In the above operation explanation, as shown in Figure 3, when a spatial filter is installed in the reduction projection optical system, the spatial filter and mirror can be removed and removed in order to block the exposure light during projection exposure. I have to do it. When the spatial filter is attached and detached in this manner, vibrations occur during attachment and detachment, and the relative positions of the lens system, wafer, and reticle shift. Also, with mirrors, it is difficult to return to the original position with good reproducibility (a position error of about 2 μm occurs).
Therefore, it is difficult to control the position of the wavefront. Furthermore, since filters, mirrors, etc. have permeability, it takes time to attach and detach them, and alignment takes longer than before.

In the present invention, as described in the operation description, the spatial filter and the mirror in the optical path are fixed to the alignment optical system, and the alignment optical system is a separate optical system from the projection optical system. Since the spatial filter and mirror do not need to be attached or detached and are fixed, the relative positions of the spatial filter and mirror system do not shift, allowing highly reproducible and highly accurate positioning. Furthermore, positioning can be performed in a short time without wasting extra time for attachment and detachment. Furthermore, as described above, it becomes possible to perform different wavelength alignment and to detect cases where the alignment mark has an asymmetrical shape, thereby realizing highly accurate alignment.

Effects of the Invention According to the present invention, a pattern on a reticle can be aligned on a wafer with high precision using interference fringes, and a pattern on the reticle can be formed on the wafer by exposure. Furthermore, since the projection optical system and the alignment optical system are configured as different optical paths, alignment of different wavelengths can be performed. Also,
Exposure can be performed without attaching or detaching a spatial filter or mirror system, and an exposure apparatus with high alignment accuracy can be configured.

[Brief explanation of the drawing]

FIG. 1 is a basic configuration diagram of alignment in an exposure apparatus according to an embodiment of the present invention, FIG. 2 is a diagram of the principle of alignment using the conventional double grating method, and FIG. 3 is a diagram of a conventional reticle and FIG. 4a is a configuration diagram of a reticle according to the present invention, FIG. 4b is a cross-sectional view of an alignment grating, and FIG. 5 is a re-diffraction optical system according to the present invention. FIG. 6 is a diagram showing the principle of the system, FIG. 6 is a detailed view of the vicinity of the wafer according to the present invention, and FIG. 7 is a diagram showing the intensity of diffracted light when two light beams are incident. 11... Alignment light source, 12... Projection light source, 13... Illumination optical system, 14... Reticle, 1
5, 17...first and second Fourier transform lenses,
16...Spatial filter, 18...Wafer, 19
...Reducing projection optical system, G...Grating, D...Photodetector.

Claims (1)

[Claims] 1 a reduction projection optical system that projects an image of a reticle onto a substrate using an illumination optical system; and a first lens system that is formed on the reticle surface, receives a light beam from a light source, splits the wavefront of the light beam, and a first grating to be made incident; a spatial filter provided near the spectral plane of the first lens system and passing a light beam having a predetermined spectrum of the wavefront-divided light beam that has passed through the first lens system; , a second lens system that makes the light beam that has passed through the spatial filter enter the reduction projection optical system; and a second lens system that is formed on the substrate and projects the light beam that has passed through the reduction projection optical system to generate and interfere with diffracted light. a second grating, the reduction projection optical system, and a photodetector for guiding the diffracted light that has passed through the second lens system and measuring the diffracted light, the light source, the first lens system,
The spatial filter, the second lens system, and the photodetector constitute an alignment optical system that is separate from the reduction projection optical system, and the first grating on the reticle surface and the alignment optical system on the substrate An exposure apparatus characterized by aligning the second grating.