JPH0750243A - Exposure method and mask used for the same as well as manufacturing method of semiconductor device using the same - Google Patents

Abstract

PURPOSE:To enable the title exposure method to be performed without fail while facilitating the pattern design by a method wherein the exposure method is performed by aligning the relative positions of the two different masks of the pattern part defining boundaries of phase inversion on transmission light and the other parttern part excluding the former pattern part. CONSTITUTION:A mask base material 9 is produced by depositing a metallic thin film 11 comprising Cr, etc., on a main surface of a transparent substrate 10 so as to be coated with an electron beam resist 12 later. Next, the whole surface is irradiated with the electron beams. Next, the glass substrate 10 is etched away in a specific depth using the Cr film 11 as a mask. Next, the Cr film 11 is removed. Furthermore, another metallic thin film 13 comprising Cr, etc., is deposited on the main surface of the transparent substrate 10 to be spin-coated with another electron beam resist 14. Besides, the coordinate bases of stepped patterns formed on the transparent substrate 10 are joined together to irradiate a specific pattern with electron beams. Furthermore, the pattern for alignment mark with wafer is exposed in the ambient part of the transfer region of the transparent substrate 10. Later, the exposed Cr film 13 is etched away to remove the electron beam resist 14.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an exposure technique used for manufacturing semiconductor integrated circuits and the like, and more particularly to a technique effective when applied to a phase shift mask for exposing by exposing a phase difference to exposure light.

[0002]

2. Description of the Related Art As semiconductor devices are mounted at high density, circuit miniaturization has advanced to a high degree, and the design rules for circuit elements and wirings have entered the submicron range. Therefore, the wavelength of the light used for exposure is the i-line (wavelength 36
5 nm). In the photolithography process of transferring an integrated circuit pattern on a photomask onto a semiconductor wafer by using light in such a wavelength range, deterioration of pattern accuracy becomes a serious problem.

FIG. 11 is a sectional view showing the structure of a mask having no phase shift, and FIG. 12 is a characteristic diagram showing the light intensity distribution of the mask. 13 is a sectional view showing the structure of the phase shift mask, and FIG. 14 is a characteristic diagram showing the light intensity distribution of the mask of FIG.

As shown in FIG. 11, a mask (or reticle) which does not require miniaturization has a chrome film 2 (metal mask) formed on one surface of the glass plate 1 so as to cover portions other than the portion to be patterned. The part you want to form a pattern is
Transparent so that light can be transmitted (P ₁ and P ₂ are light transmission areas)
And a portion other than this is a light shielding region N shielded by the chromium film 2.

In the mask having such a structure, since the pair of light transmitting regions P ₁ and P ₂ sandwiching the light shielding region N have the same phase, two light beams are generated on the semiconductor wafer. At the boundary of the, they interfere with each other and act to strengthen each other. For this reason, when the width of the light shielding region N is narrower than the exposure wavelength, the contrast of the projected image of the pattern on the semiconductor wafer is lowered as shown in FIG. 12, and the depth of focus becomes shallow, resulting in a fine pattern. It can be seen that the transfer accuracy is significantly reduced and the patterns cannot be separated and exposed.

As a means for solving such a problem, attention has been paid to a phase shift technique in which the contrast of a projected image is prevented from being lowered by changing the phase of light passing through a mask.

A specific example of this system is, for example, Japanese Patent Publication Sho 62.
As shown in FIG. 13, a phase shifter 3 (for example, a transparent glass film or the like) is provided in _{one of} a pair of light transmitting regions P ₁ and P ₂ that sandwich a light shielding region N, as shown in FIG. It has a structure, and the film thickness is adjusted so that the phases of the two lights immediately after passing through the pair of light transmission regions are mutually inverted.

As a result, as shown in FIG. 14, on the semiconductor wafer, the two lights interfere with each other at their boundaries and weaken each other, so that the contrast of the projected images of the patterns is greatly improved and the patterns are separated from each other. Exposure becomes possible. Thus, it can be seen that the phase shift mask is a mask suitable for miniaturization of semiconductor integrated circuits.

Further, in Japanese Laid-Open Patent Publication No. 62-67514, a minute light transmitting region is provided around the first light transmitting region on the mask, and the phase of two lights transmitted through the pair of light transmitting regions is set. There is disclosed a phase shift technique for improving the pattern transfer accuracy by reversing the two.

Further, as disclosed in Japanese Patent Laid-Open No. 2-140743, a phase shifter is provided in a part of the light transmission region,
A phase shift exposure technique has also been proposed in which the phases of two lights transmitted through a portion provided with the phase shifter and a portion not provided with the phase shifter are inverted to each other, for example, by bringing the light closer to the end of the light shielding region to improve the pattern transfer accuracy. ing.

[0011]

However, the circuit pattern of the integrated circuit transferred onto the semiconductor wafer is actually complicated and intricate, and the pattern is not simply drawn vertically and horizontally. Therefore, the present inventor has found that there is a problem that effective exposure cannot be performed even if the above-mentioned phase shift mask is used.

For example, in the phase shift mask shown in FIG. 11, the exposure wavelength is i-line (wavelength 365 nm),
Coherency σ = 0.3, projection optical lens characteristic NA = 0.
5, the effect was remarkably effective when the width of the light-shielding region was 0.5 μm, but when the width of the light-shielding region was increased to 0.5 μm, almost no effect was observed. This is Japanese Examiner's Sho 62-
This is an important problem of the phase shift mask described in Japanese Patent No. 50811.

Further, in the phase shift mask and the exposure method described in JP-A-62-67514 and JP-A-2-140743, the problem of arrangement of the phase shift pattern is solved, and the focus of the transfer exposure surface is eliminated. Although the depth is also improved, the present inventors have found that there is a problem that the effect of improving the resolution of the pattern is small.

Therefore, an object of the present invention is to provide a technique capable of performing reliable exposure while facilitating pattern design.

The above and other objects and novel features of the present invention will be apparent from the description of this specification and the accompanying drawings.

[0016]

Among the inventions disclosed in the present application, a brief description will be given to the outline of typical ones.
It is as follows.

That is, a phase shifter provided at a predetermined position in a light transmission region of a mask in which a plurality of patterns are formed at intervals shorter than the exposure wavelength causes a phase difference in transmitted light to form a projected image of the pattern on a sample. An exposure method for exposing, wherein, for at least one continuous pattern formed by the exposure, a pattern portion where a boundary of phase inversion occurs in transmitted light that has passed through the light transmission region and a pattern portion other than this Exposure is performed by dividing the mask into different masks and matching the relative positions of the two masks.

[0018]

According to the above means, when at least three or more circuit patterns are exposed on the semiconductor wafer by projection exposure at intervals shorter than the exposure wavelength, at least one of the plurality of patterns is exposed in the x direction or the y direction. The pattern is divided at predetermined intervals in the direction, and is divided into groups including other patterns. Corresponding to this grouping, a first mask (phase shift mask) that sequentially inverts the phase of transmitted light at predetermined intervals is formed on the mask.

Further, pattern data of an area covering the boundary area of division is created, and the pattern data is used to generate a second pattern data.
Masks (light-shielding masks) are formed, and the projected images of the patterns of the respective masks are transmitted through the projection optical system of the exposure device using the second mask and the first mask without changing the relative positions of the two. , Image on the sample.

As a result, the resolution and the depth of focus for forming an image on the photosensitive resist film on the semiconductor wafer can be greatly improved. Further, since it is possible to remove the restriction when the phase shifter is provided, the pattern design of the phase shift mask becomes easy.

[0021]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a flow chart showing an exposure method according to the present invention.

Prior to the description of the embodiment, a case where a mask M having a pattern as shown in FIG. 2 is formed by using the phase shift technique as shown in FIG. 13 will be described. In this case, according to the known technique, X1 and X3 are included in the transmissive region T (the white portion in the figure, the shaded portion is the light shielding region S) corresponding to the three patterns X1, X2, and X3. A phase shifter will be provided. However, if the width and interval of the patterns are narrow, the vertical portions where X1 and X2 are close to each other have the same phase, so that the light-shielding portion between them is not formed and a pattern short circuit occurs.

In order to solve this problem, in the present invention,
As shown in FIG. 3, the mask is divided into two pieces for exposure. That is, exposure is performed using the first mask as shown in FIG. 3A and the second mask as shown in FIG. 3B. A shadow is generated in a portion where the phase of the transmitted light is in the opposite phase in one transmission region of FIG. 3B, and the pattern is prevented by overlapping exposure of FIG. is doing.

However, since the actual circuit pattern is complicated, the use of the mask by such a combination must rely on automatic design by a computer. The specific example of the processing is shown in the flowchart of FIG. The exposure method of the present invention will be described below with reference to FIG. 1 and other drawings. In the following embodiments, at least three continuous patterns are formed by projection exposure so as to be close to each other at intervals shorter than the exposure wavelength.

The outline of the process of FIG. 1 will be explained. First, when forming a resist pattern on a semiconductor wafer,
At least one of the three patterns is divided at predetermined intervals in the x direction or the y direction, and is divided into groups including the other patterns, and the patterns are transmitted according to this grouping. A phase shift mask for sequentially inverting the phase of light at predetermined intervals is formed.

Next, pattern data of an area covering the divided boundary area is created, and a light-shielding mask is formed by the pattern data. By using this light-shielding mask and phase shift mask, a projected image of a mask pattern is formed on a photosensitive resist film on a semiconductor wafer through a projection optical system of an exposure apparatus without changing their relative positional relationship.

Next, a pattern forming method will be described with reference to the flowchart of FIG.

A pattern of a semiconductor integrated circuit or the like can be represented by a combination of rectangular figures, for example, and a combination of data in which information such as pattern width W, length H, and its center coordinates (X, Y) is described. (Step 101). The pattern data of the semiconductor integrated circuit or the like to be combined is superposed on the figures by these data (in other words,
In the case of a figure-to-figure overlap), an overlap removal process (cutout of a transfer area) is performed (step 10).
2). In the overlap removal process, for example, a figure formed by the pattern data is developed on the memory map and a logical sum (OR) process is performed. In addition, a window is provided in an area including a pattern that is adjacent to the pattern to reduce the processing time of the computer.

Then, sort processing is performed for rearranging the graphics in each of the X and Y directions (step 103). In this sort, the pattern data is grouped and rearranged in a direction having a large area ratio of adjacent patterns (for example, the X-axis direction or the Y-axis direction) at predetermined intervals (for example, wiring pitch of the semiconductor integrated circuit pattern). It is a thing.

Then, at the point where the figures overlap, the process of dividing in the X-axis direction or the Y-axis direction (pattern division) is performed again (step 104). That is, n figures P _i (i = 1
To n, where n is an integer) (where X _i ≤X <X
_{i + 1} or Yi ≦ Y <Y _{i + 1} . Then, i is initialized to i = 1 (step 105), and one figure P _i is read (step 106).

Next, the phase shifter forming process is performed on the figures included in one group that has been rearranged. Whether the figure group is an odd number or an even number (i = 2)
The subsequent processing differs depending on m, where m is an integer (step 107). In the case of an even number, the relative position coordinates of the patterns of the figures in the above group are not changed, and the transmitted light phase of the mask pattern which transmits the figure area determined by this is set to φ = 0 (step 108).

Here, when the graphic in the above group touches the boundary of the graphic of the adjacent group, the following processing is performed. Whether or not the boundaries are in contact can be determined by, for example, whether or not the x coordinate of each figure in the next group becomes a value obtained by adding the wiring pitch to the x coordinate of the figure in the previous group.

The phase of the mask pattern passing through the figure area determined by the figures of the adjacent groups is φ = π.
(180 °), and a diffractive projection image is generated due to the phase difference from the figure in the previous group. Therefore, it is necessary to eliminate the influence of this diffraction projection image. Therefore, pattern data of a figure that covers the boundary portion of the figure is created, and with this pattern data, a light-shielding mask (second
Mask). Using this mask, overlapping exposure is performed on the portion where the diffraction projection image is generated.

If it is determined in step 107 that the pattern is an even number, the phase of the mask pattern transmitted through the figure region determined by the figure in the above group without changing the relative position coordinates of the pattern is φ =. 0
(Step 109). A concrete example of the processing of steps 109 and 108 is the mask M of FIG.
FIG. 3C shows an example of forming the phase shifter.
(In the figure, T is a transmissive region, S is a light-shielding region, and F is
Is a phase shifter).

When the figure in the above group is in contact with the figure in the adjacent group, pattern data of the figure that covers the boundary portion of the figure is created, and this figure is used to form this figure. A light-shielding mask to be transferred is created, and the portion where the above-mentioned diffraction projection image is generated is overlaid and exposed.

When it is determined that the processing for n figures is completed (step 110), mask pattern formation is executed (step 112), and the semiconductor wafer is exposed (step 113). On the other hand, if (i = i + 1) is not satisfied, the data of the next group is read (step 11
1), the position with respect to the X coordinate, P _{i + 1} (X) −P _i (X) ≦ pW is determined (step 114), and if YES, the shading figure Q _i as shown in FIG. ₊₁ (ΔW, H + Δ
H, X + pW / 2, Y) is created (step 11)
5). On the other hand, if the determination is No, it is considered that no figure is generated, and the process proceeds to the next step. (However, P _i
Is the phase φ of the figure P _i = 0, P _{i + 1} is the phase φ of the figure P _{i + 1}
= Π, P _i (X) is the X coordinate of the figure P _i , P _{i + 1} (X) is the X coordinate of the figure P _{i + 1} , and pW is the X-direction pattern pitch).

Further, after the processing of steps 114 and 115, the position with respect to the Y coordinate, P _{i + 1} (Y) −P _i (Y) ≦ pH, is determined (step 116) and shown in FIG. Such a shading figure Q _{i + 1} (W + ΔW, ΔH, X, Y + pH /
2) is created (step 117), and the process proceeds to step 10
The process returns to 6 and the subsequent processes are repeatedly executed.

FIG. 4 illustrates the flow of pattern division (X-direction region division) and light-shielding figure creation processing in step 104. That is, the part where the X direction and the Y direction are connected is divided at the boundary part between the vertical part and the horizontal part,
Then, the area is divided at the middle of the horizontal portion to perform the pattern division. Furthermore, the procedure for creating a light-shielding figure having a light transmission area of Q _{i + 1} is shown.

Now, the structure of the mask shown in FIG. 3B will be described. This mask has a refractive index of 1.
About 47 transparent synthetic quartz glass is used as the base material. On the main surface of the mask, a light transmissive region (white region shown in the drawing) that is substantially transparent to the exposure light is formed with an opaque light shielding region (hatched region shown in the drawing) sandwiched therebetween. )
As shown in, the phase shift means for inverting the phase of the transmitted light is provided in one of the transmission regions.

The transmissive region is provided with a first transmissive region and at least a second transmissive region which is close to the first transmissive region and whose phase of transmitted light is inverted with respect to the first transmissive region. ing.

The above mask is a mask for transferring a predetermined integrated circuit pattern onto a semiconductor wafer made of, for example, a silicon single crystal, for example, a reticle on which an original image of an integrated circuit pattern having a size five times the actual size is formed. This integrated circuit pattern is transferred onto a semiconductor wafer through a reduction projection optical system.

Next, FIG. 5 shows another mask combination for obtaining the circuit pattern of FIG. 2, and the mask structure will be described. In the formation of the phase shifter by the simple repetition method, as described with reference to FIG. 2, when the pattern interval is narrow, a shadow is generated in the horizontal portion and the pattern is cut off.

This embodiment corresponds to the alternative example of FIG. 3, in which the horizontal portion has a mask provided with a phase shifter 4 having a narrow width in a frame shape as shown in (a) and 3 as shown in (b). A mask provided with a phase shifter 5 so that the transmissive regions of the columns are arranged in order from the left to φ = π, φ = 0, and φ = π, and by superimposing and exposing the mask, pattern exposure without shadows is performed. Will be possible. Note that (c) and (d) are cross-sectional views, and (e) and
(F) shows the amplitude waveform on the sample, and (g) and (h) show the light intensity characteristics on the semiconductor wafer.

As compared with the embodiment of FIG. 3, the size of the effective transfer pattern of FIG. 5A can be reduced and the depth of focus can be increased.

Further, FIG. 6 shows a second circuit pattern example. When a mask having such a pattern is required,
In the formation of the phase shifter according to the conventional method, if the pattern interval is narrow, the light shielding portion between the patterns in one of the horizontal direction and the vertical direction is not formed, and a short circuit occurs.

Therefore, in such a case, as shown in FIG. 7, when a mask having the sub-phase shifter 6 as shown in FIG. 7A and a mask having the phase shifter 7 as shown in FIG. A desired pattern having a shape as shown in FIG. 6 can be obtained. In (a), four sub-phase shifters 6 having a narrow width are provided so as to surround the portion Q in FIG. 6, and this is set to the region of φ = π. As a result, the size of the effective transfer pattern of FIG. 6A can be reduced and the depth of focus can be increased as compared with the embodiment of FIG.

Further, in (b), there are two transmission regions T (white portions), one on the left and two on the left side, and one on the right side, and there are three transmission regions T (white portions). Of these, the upper side on the left side is φ = π. The phase shifter 7 is provided so that the phase shifter 8 is provided so as to surround the lower half of the transmission region on the right side, and similarly, the region of φ = π is set. The other is a region of φ = 0 where no phase shifter is provided. Note that (c) and (d) are cross-sectional views of the mask, and (e) and
(F) is an amplitude waveform in the semiconductor wafer, and (g),
(H) shows the light intensity characteristic on the semiconductor wafer.

FIG. 8 is an explanatory diagram of the amplitude and intensity of light projected on the surface of a semiconductor wafer via a projection optical system (optical exposure device).

With the above configuration, as shown in FIG. 8A, the transmitted light is transmitted by the phase shifter S formed in _{one of the} transmission regions P _{1 and} P ₂ sandwiching the light shielding region N formed on the mask M. , And by utilizing the light interference between them, a pattern closer to the exposure wavelength can be imaged on the photosensitive resist film on the semiconductor wafer through the optical projection system.

The light L ₁ immediately after passing through the first transmitting region P ₁ of the mask M is the light L immediately after passing through the second transmitting region P _2.
The phase of ₂ is opposite to that of (b). The amplitude immediately after passing through the mask has a rectangular waveform, but the amplitude on the sample (semiconductor wafer) has a waveform as shown in (c). And
On the sample (semiconductor wafer), the two lights interfere with each other in the light-shielding region that is the boundary between them and weaken each other.
The light intensity waveform is as shown in (d).

In this embodiment, one first transmission region P and the other second transmission region P, which are adjacent to each other with the light-shielding region in between, are phase-inverted to each other. Thus, half the exposure wavelength
When they are separated from each other, the intensities of the transmitted lights can be mutually increased. Therefore, pattern data of a region that covers the boundary region of the division is created, a light-shielding mask is formed based on this pattern data, and a shadow portion generated by the projection of the phase shift means is overlaid and exposed.

FIG. 15 shows the entire structure of the mask (reticle) of this embodiment. As shown in FIG. 15, an opaque light-shielding region N is provided around the transfer regions A _{1 and} A ₂ used for exposure. Alignment marks B _{1 and} B _{2 of} the mask and alignment marks B _{3 and} B ₄ of the mask and the wafer which is the sample to be exposed are formed in the light-shielded region N, and the mask and the wafer are aligned with each other. Can be exposed. _{Incidentally,} B 5, B ₆ is a phase shift layer and a light-shielding layer alignment mark.

By constructing the above-mentioned phase shift mask and the light-shielding mask on the same glass substrate, during exposure, the mask blind of the exposure device shields the exposure light applied to the mask surface, thereby improving efficiency. Can be well exposed. Regarding this point, if two glasses are used, it is possible to perform the same exposure, but the exposure efficiency is reduced accordingly.

In the conventional method in which the phase difference is simply applied to one of the transmissive areas sandwiching the light-shielding area, there is a restriction on the arrangement of the fine transfer pattern, but this method does not reduce the execution throughput of the exposure, and the above restriction is imposed. Can be eliminated.

Next, a mask forming method and an exposing method will be described below.

FIG. 9 is a diagram showing a mask manufacturing method according to the present invention.

First, the mask material 9 is prepared. The mask material 9 is formed on the main surface of a transparent substrate 10 such as quartz glass by using, for example, a sputtering method and a metal thin film 1 such as Cr (chrome).
It is created by depositing 1. On the main surface of this mask material 9,
The electron beam resist 12 is spin-coated as in (a).

Next, using an electron beam drawing apparatus, the mask material 9 is irradiated with an electron beam as shown in (b). The electron beam drawing apparatus scans an electron beam by computer control based on pattern data such as graphic information of an integrated circuit pattern and position coordinates. The pattern data used here corresponds to, for example, the region (φ = π) where the phase of the transmitted light is inverted shown in step 109 of FIG.

Besides the circuit pattern, as shown in FIG.
A pattern of two alignment marks for processing a light-shielding pattern, which will be described later, is also exposed on the diagonal of the peripheral portion of the transparent substrate.
This pattern is a cross mark having a size of about 300 μm. Then, depending on whether the electron beam resist is a positive type or a negative type, the exposed portion or the unexposed portion is removed by a developing solution, and the exposed Cr film is etched as shown in (c). The Cr film can be etched using, for example, wet etching with cerium diammonium nitrate or the like.

The resist on the Cr film is usually removed in the next step of etching the transparent substrate 10. In this case, the resist is removed by oxygen plasma etching or the like, the mask is washed, and the appearance of the mask is inspected. Is desirable. The remaining defects of the minute Cr film found by this visual inspection are removed by, for example, irradiation with laser light. Also, C
The defect of the r film is spot-exposed to the defective portion by using, for example, a negative photoresist, and is developed to leave the photoresist. In this way, a Cr film pattern for forming a substantially defect-free phase shift layer can be formed.

Next, using the Cr film as a mask, the glass substrate is etched to a predetermined depth by using CF ₄ (Freon gas) plasma etching or the like. Here, the depth d of the transparent substrate 10 is the wavelength of the exposure light is λ,
When the refractive index of is n, it is expressed by the following equation. Alternatively, it is set so as to satisfy this odd multiple relationship.

D = λ / 2 (n−1) For example, the wavelength of light is 365 nm (i line), the transparent substrate 10
If the refractive index of 1. is 1.47, the depth d of the transparent substrate 10 is 3
It may be about 90 nm. In plasma etching of CF ₄ , the depth can be controlled by the etching time or the like. Note that, with respect to a defect in which a specific depth does not reach a predetermined depth by this plasma etching, the defect is irradiated by the focused ion beam method to etch the substrate. Also, instead of the transparent substrate 10, the transparent substrate 10
By depositing the transparent film and the light-shielding film having a predetermined film thickness on the top, the processing accuracy of the depth can be improved.

Next, by wet etching, the Cr film is removed as shown in (d). Further, on the main surface of the transparent substrate 10, a metal thin film 1 of Cr or the like is formed by using, for example, a sputtering method.
3 is deposited as in (e). Then, an electron beam resist 14 is spin-coated on the metal thin film 13 on the main surface as shown in (e).

Further, the position of the above-mentioned alignment cross mark is detected by using an electron beam drawing apparatus, the coordinate system of the step pattern formed on the transparent substrate 10 is aligned, and the desired pattern is electronically aligned at a predetermined position. Irradiate with rays. Here, the electron beam irradiation pattern is, for example, a pattern serving as a light-shielding region that is a positive / negative reversal region of the transmissive region (combined region of φ = π and φ = 0) shown in FIG. 1. Regarding the drawing accuracy of electron beam drawing equipment, the pattern superposition is 0.1 μm.
Since it can be set as follows, this method can be applied to, for example, a mask (reticle) of an exposure apparatus with a reduction rate of 1/5.

In addition to the circuit pattern, the transparent substrate 1
The pattern for the alignment mark with the wafer is exposed to the periphery of the 0 transfer area. The pattern of this alignment mark is specified by the reduction projection exposure apparatus. Then, depending on whether the electron beam resist is a positive type or a negative type, the exposed portion or the unexposed portion is removed by a developing solution, and the exposed Cr film is etched as shown in (f).

Finally, the above electron beam resist is removed.
Then, the appearance of the Cr film pattern is inspected. In the Cr film pattern, minute defects remaining in the Cr film are removed by, for example, irradiating laser light, and defective defects in the Cr film are irradiated by irradiating the defective portion with an organic gas added, for example, by a focused ion beam method. Thus, the defect can be corrected by forming the light shielding film of the carbon film.

In the above description, the method of etching the transparent substrate 10 is shown as the method of inverting the phase of the transmitted light, but a transparent film may be deposited on the transparent glass substrate 10. As this transparent film, for example, spin-on-glass (Spi
n On Glass) or other means can be used.

FIG. 10 is a block diagram showing the optical system of a reduction projection type exposure apparatus to which the present invention is applied.

The exposure device 15 includes a beam expander 18 and a mirror 1 on the optical path connecting the light source 16 and the sample 17.
9, a half mirror 20 (or a beam splitter), a lens 21, a mirror 22, and an objective lens 23 are provided. Further, a mirror 24, a lens 25, a mirror 26, and an objective lens 27 are sequentially arranged on the reflection optical path of the half mirror 20.

At the positions of the masks M ₁ and M ₂ , a mask provided with a phase shifter as shown in FIGS. 7A and 7B is arranged. The light source 16 is, for example, i-line
(nm) and a high pressure mercury lamp that emits light L. Further, the sample 17 (semiconductor wafer) has an X, Y stage 28.
It is accurately positioned and set on the top prior to exposure.
A photoresist that is sensitive to light is spin-coated on the surface of the sample 17.

The light L from the light source 16 is expanded by the beam expander 18, refracted by the mirror 19, and then split into _two lights L ₁ and L ₂ by the half mirror 20. After that, one of the two lights L ₁ and L ₂ (L
₁ ) goes straight on and enters the mask M ₁ , and the other light (L ₂ ) is refracted by the mirror 24 and then enters the mask M ₂ .

The light transmitted through the mask M ₁ is condensed by the lens 21, refracted by the mirror 22, and then exposed on the sample 17 by the objective lens 23. On the other hand, half mirror 2
The light L ₂ on the reflected optical path of 0 is refracted by the mirror 24.
Further, the light emitted from the mirror 24 passes through the mask M ₂ , is condensed by the lens 25, is refracted by the mirror 26, and is then exposed on the sample 17 by the objective lens 27.

In the exposure apparatus 15 of FIG. 10,
In order to prevent the interference of the _two lights L ₁ and L _2, an incoherent converter (not shown) may be inserted on the optical axis of the light L ₂ .
Further, transparent glass may be inserted so that the optical path difference between the _two lights L ₁ and L ₂ is equal to or longer than the coherence length. Further, a shutter may be provided in the latter stage of the half mirror 20 and the mirror 24 so that _{one of} the mask M ₁ and the mask M ₂ is exposed and then the other is exposed.

The invention made by the present inventor has been specifically described based on the embodiments, but the present invention is not limited to the above embodiments, and various modifications can be made without departing from the scope of the invention. Needless to say.

For example, in the above embodiment, the case where the number of masks is one and the case where the number of masks is two are shown, but the number of masks may be three or more.

[0077]

The effects obtained by the typical ones of the inventions disclosed in this application will be briefly described as follows.
It is as follows.

That is, a phase shifter provided at a predetermined position in the light transmission region of the mask in which a plurality of patterns are formed at intervals shorter than the exposure wavelength causes a phase difference in the transmitted light to form a projected image of the pattern on the sample. An exposure method for exposing, wherein a pattern portion where a boundary of phase inversion occurs in transmitted light that has passed through the light transmitting region and a pattern portion other than this are formed by being divided into different masks, and the two masks are relative to each other. Since the exposure is performed by matching the positions, it is possible to significantly improve the resolution and the depth of focus for forming an image on the photosensitive resist film on the semiconductor wafer, and eliminate the restrictions on the arrangement of the phase shifter. The pattern design of the retardation mask can be facilitated.

[Brief description of drawings]

FIG. 1 is a flowchart showing an exposure method according to the present invention.

FIG. 2 is a plan view showing an example of a mask corresponding to a circuit pattern to be realized in the embodiment of the invention.

FIG. 3 is an explanatory diagram showing a combination of masks for obtaining the circuit pattern of FIG.

FIG. 4 shows steps 104, 114 to 117 in FIG.
It is explanatory drawing which illustrated the processing content of.

5 is an explanatory diagram showing another mask combination corresponding to the circuit pattern of FIG.

FIG. 6 is a plan view showing a second circuit pattern example.

FIG. 7 is an explanatory diagram showing a combination of masks for obtaining the circuit pattern of FIG.

FIG. 8 is an explanatory diagram of amplitude and intensity of light projected on the surface of a semiconductor wafer via a projection optical system.

FIG. 9 is a diagram showing a method of manufacturing a mask according to the present invention.

FIG. 10 is a configuration diagram showing an optical system of a reduction projection exposure apparatus to which the present invention is applied.

FIG. 11 is a cross-sectional view showing the structure of a mask having no phase shift.

12 is a characteristic diagram showing a light intensity distribution of the mask of FIG.

FIG. 13 is a sectional view showing a structure of a phase shift mask.

FIG. 14 is a characteristic diagram showing the light intensity distribution of FIG.

FIG. 15 is a plan view showing an example of the shape of a general mask.

[Explanation of symbols]

 1 Glass Plate 2 Chromium Film 3 Phase Shifter 4 Phase Shifter 5 Phase Shifter 6 Sub Phase Shifter 7 Phase Shifter 8 Phase Shifter 9 Mask Material 10 Transparent Substrate 11 Metal Thin Film 12 Electron Beam Resist 13 Metal Thin Film 14 Electron Beam Resist 15 Exposure Device 16 Light Source 17 Sample 18 Beam Expander 19 Mirror 20 Half Mirror 21 Lens 22 Mirror 23 Objective Lens 24 Mirror 25 Lens 26 Mirror 27 Objective Lens 28 X, Y Stage

Claims (4)

[Claims] 1. A phase shifter provided at a predetermined position in a light transmission region of a mask in which a plurality of patterns are formed at intervals of about the exposure wavelength or shorter than the exposure wavelength causes a phase difference in the transmitted light to form the pattern on the sample. An exposure method for exposing a projected image, comprising at least one formed by the exposure.
For two continuous patterns, a pattern portion in which a boundary of phase inversion in the transmitted light passing through the light transmission region is generated and a pattern portion other than this are divided into different masks, and the relative positions of the two masks are formed. An exposure method characterized in that the exposure is performed by matching the values. 2. Three or more patterns are formed at intervals of about the exposure wavelength or shorter, and at least one pattern is formed.
A mask in which two patterns have a continuous portion in the X and Y directions and a phase shifter is provided at a predetermined position of a light transmission region to generate a phase difference in transmitted light, and a phase with respect to an adjacent pattern is substantially opposite. A first mask in which only a pattern in which the phase shifter can be provided so as to be in phase is formed, and a pattern in a boundary region in which the phase is inverted when the phase shifter is provided are formed together with the first mask. A second mask which can be aligned with the pattern of the mask. 3. The second mask is disposed in at least a part of the second mask so as to surround the periphery of the pattern of the boundary region, and a phase shifter is provided so that the phase with respect to the boundary region is approximately opposite phase. The mask according to claim 2. 4. A method of manufacturing a semiconductor device, wherein a mask provided with a phase shifter at a predetermined position of a light transmitting region is used, and a projected image of a pattern formed on the mask is exposed on a semiconductor wafer to form a wiring pattern. Then, the pattern image is formed on the negative resist film on the semiconductor wafer by using the first mask in which only the pattern capable of providing the phase shifter is formed so that the phase with respect to the adjacent pattern is substantially opposite phase. Is exposed through an optical projection system, and a second mask having a pattern covering a boundary region where the phase is inverted when the phase shifter is provided is aligned with the pattern of the first mask. In the state, the pattern of the second mask is exposed through the optical projection system, and the method for manufacturing a semiconductor device.