JP3320062B2 - Mask and pattern forming method using the mask - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a mask, a method of manufacturing the same, and a pattern forming method using the mask, and more particularly to a mask using a phase shift of light, a method of manufacturing the same, and a pattern forming method using the mask.

[0002]

2. Description of the Related Art In the process of manufacturing a semiconductor device, when a pattern such as an element or a circuit is formed on a semiconductor wafer, a pattern transfer exposure method using ultraviolet light is usually used.

The pattern transferred to a semiconductor wafer is:
It is formed depending on the presence or absence of a metal thin film provided on a glass substrate that transmits light to block light. Among those formed with a transfer pattern on this glass substrate, those provided with a pattern having the same size and number as the chip pattern transferred onto the wafer are called a full-size mask or simply a mask, for example, 5 to 5 masks.
A pattern in which a smaller number of chip patterns than the number of chips formed on the wafer by being magnified 10 times is provided is called an enlarged mask or reticle. Then, a pattern is transferred and exposed on a wafer through a resist film by parallel projection in the case of a mask or by reduction projection using a reduction lens system in the case of a reticle. In this case, it is necessary to increase the light contrast at the edge of the exposure area in order to increase the resolution when transferring a pattern that is miniaturized and highly integrated. FIG. 70 shows a conventional mask formed of an opaque layer and a transparent substrate layer having a predetermined shape, and FIG. 71 shows a case where a pattern is formed on a wafer. FIG. 70 is a configuration diagram showing a conventional photomask 450. As shown in FIG. In FIG. 70, 451
Is an opaque layer made of a material such as metallic chromium (Cr), and a predetermined pattern is formed on the transparent substrate layer 452 by well-known lithography and etching. In FIG. 71,
Light C emitted from an exposure device (not shown) is a photomask.
The opaque layer 451 does not transmit through the transparent substrate layer 452 where the opaque layer 451 is not formed. The transmitted light is
O passing through the imaging lens system 453 and coated on the wafer 454
A resist material 455 such as FPR (trade name, Tokyo Ohka Kogyo Co., Ltd.) is exposed. Thus, the same pattern as that of the photomask 450 is formed on the wafer 454 by etching.

When a pattern is formed using an optical lens system, the pattern is formed on the wafer 454 based on data of only the contrast due to the presence or absence of an opaque layer.
Therefore, pattern formation has a physical resolution limit due to the wavelength of light from the optical lens system, and it is difficult to form a pattern narrower than the wavelength of light used.

In the conventional photolithography process, chromium (Cr) is deposited on a transparent substrate such as glass or quartz.
An opaque layer such as a film is formed and patterned to be used as a reticle. FIGS. 72A and 72B show an example of such a conventional pattern forming method.

In FIG. 72A, a light source 461 is constituted by a mercury lamp or an excimer laser having filters for i-line and g-line, and light 463 generated from the light source 461 passes through an illumination system lens 462. The reticle 464 is being irradiated. The illumination system lens 462 has, for example, partial coherency σ
= 0.50 is used. The reticle 464 is formed of an opaque pattern 469 such as a Cr film on a transparent substrate 468 such as glass.
Is formed. The opaque pattern 469 on the reticle 464 is
6 An image is formed on the upper photoresist layer 467. Imaging lens
For 465, for example, one with a numerical aperture NA = 0.50 is used. In the case of the above-described pattern forming method, the resolution is K1 · λ / NA. Here, K1 is a process coefficient, usually 0.6
Take a value of ~ 0.8. λ is the light wavelength, and NA is the numerical aperture of the imaging lens. Wavelength λ of light 463 emitted from light source 461
Is, for example, about 365 nm for the i-line of a mercury lamp,
In the case of an excimer laser, the wavelength is, for example, 248 nm or 198 nm.
The numerical aperture NA depends on the imaging lens system, but is, for example, about 0.5. To improve the resolution, reduce K1 or λ, and
A needs to be increased, but K1 and NA cannot be freely selected. The wavelength λ is also limited by the light source and the optical system. When the wavelength λ of the light used for exposure, the numerical aperture NA, and the process coefficient K1 are determined, the resolution is determined, and a pattern having a resolution lower than the resolution cannot be formed.

The light 463 emitted from the light source 461 is a reticle 464
The entire surface of the opaque pattern 469 is illuminated, and the portion of the light that illuminates the opaque pattern 469 is blocked by the opaque pattern 469. For this reason, as shown in the lower part of FIG. 72A, only the light irradiated to the portion having no opaque pattern passes through the reticle 464 and is irradiated on the photoresist layer 467 by the imaging lens 465. On the photoresist layer 467, a pattern having a light intensity distribution proportional to the square of the amplitude of the irradiated light is formed, and the photoresist layer 467 is selectively exposed.

FIG. 72B shows an enlarged opaque mask having an opaque pattern. Transparent substrates such as glass and quartz
An opaque pattern 469 of a Cr pattern is formed on 468 to form a mask or reticle 464. The minimum width W of the pattern that can be exposed is limited by the resolution determined by the imaging lens 465.

With reference to FIGS. 73 (A) to 73 (D), a description will be given below of how the light intensity distribution in the case of exposing a line pattern thinned beyond the resolving power by such a conventional technique will be described. I do. In the examples of FIGS. 73A to 73D, the wavelength of the used light is 365 nm, the numerical aperture NA is 0.50, and the partial coherency σ is about 0.50.

FIG. 73A shows a light intensity distribution when a pattern having a width of 0.35 μm is formed. The light intensity distribution approaches “0” at the center position (0.0), and gradually rises on both sides. Line width approx.1.0 μ at the position where the light intensity is maximum
m or more. If the photoresist layer is developed with a light intensity of about 0.2 as a development threshold, a pattern having a designed width of about 0.35 μm can be developed.

FIG. 73 (B) shows a light intensity distribution when a pattern having a width of 0.30 μm is imaged. The width of FIG.
Compared to the case of 35 μm, the only thing that changes clearly is
This is an increase in the minimum light intensity at the center position (0, 0). There is no significant change in the width of the light intensity distribution itself.

FIGS. 73 (C) and (D) each show a width of 0.25 μm.
And a light intensity distribution when a pattern having a width of 0.20 μm is formed. As in the case of FIG. 73B, the minimum value of the light intensity at the center of the pattern gradually increases, but the pattern width itself does not show much change. That is, even if the pattern width is reduced beyond the resolving power, the pattern width of the obtained light intensity distribution does not decrease, but rather the minimum value of the light intensity at the pattern center increases. In this case, not only cannot the exposure line width be reduced, but also the black level is raised to gray. Thus, an image having a resolution lower than the resolution cannot be formed.

[0013] To this end, IBM Corporation first issued the IEEE
TRANSACTIONS ON ELECTRON DEVICES, VOL.ED-29, NO.12,
DECEMBER 1982, Improving Resolution in Photol
entitled ithography with a Phase-Shifting Mask ”, the transmitted light is obtained by setting the predetermined pattern part on the mask to have an optical path length different from that of the remaining part having other patterns. Phase of light on both wafers
A method of shifting the degree has been announced. in this way,
The interference of light between the patterns is eliminated to improve the light contrast on the wafer, and the resolution in the same exposure apparatus is improved.

However, there is a problem in that it is difficult to apply the method to a mask or a reticle having a fine pattern by the conventional method disclosed in this publication, and the time and effort for creating pattern data unique to the phase shift pattern increases.
Therefore, there is a demand for a phase shift pattern that can be more easily applied to a mask or a reticle having a fine pattern and does not cause an increase in the number of steps such as pattern data creation.

In the above-mentioned conventional phase shift mask, the phase shift pattern is formed by forming an auxiliary light transmitting pattern having a narrower width than the designed light transmitting pattern (white pattern) near the designed pattern. An organic pattern such as a resist formed through coating, exposure and development steps or an inorganic pattern formed through chemical vapor deposition and lithography steps is mounted on the auxiliary pattern as a phase shifter. For example, a phase shift mask using a negative resist pattern for a phase shifter is formed by a method described below with reference to FIGS.

First, as shown in FIG. 74A, a light-shielding film 552 is provided on a glass substrate 551, and an aperture pattern having a width of, for example, about 1.5 μm, that is, a light-transmitting film is formed by lithography using electron beam (EB) exposure. For example, a design pattern (transfer pattern) 553 composed of a region and a design pattern 553
Fine patterns 554A and 554B, each of which is an opening pattern having a width of, for example, about 0.5 μm, which is narrower than the design pattern, are formed in the vicinity area separated by about 0.5 μm as auxiliary patterns.

Next, as shown in FIG. 74 (b), on the surface of the glass substrate 551 including the inner surfaces of the opening patterns 553, 554A and 554B, a transparent thin conductive film for preventing charge-up during EB exposure. Form 555.

Next, as shown in FIG. 74 (c), a negative EB resist film 656 is formed on the glass substrate 551 so that the phase of the transmitted light is shifted by 180 degrees, and a pre-bake is performed if necessary. After that, EB exposure of a phase shift pattern is performed on the fine patterns 554A and 554B.

Here, the thickness D of the resist film 656 is obtained by the following equation (1).

D = λ / 2 (n−1) (1) λ: Wavelength of Light Used for Exposure n: Refractive Index of Shifter Material When using i-line having a wavelength of, for example, 365 nm for exposure, a resist film is used. Since the refractive index of 656 is about 1.6, the thickness D of the resist film 656 is about 304 μm.

Next, as shown in FIG. 74D, a phase shifter comprising a negative type EB resist film 656 having a film thickness of D is selectively developed on the fine patterns 554A and 554B comprising the light-transmitting regions, that is, as shown in FIG. The phase shift patterns 556A and 556B are formed.

FIG. 75 is a phase profile diagram corresponding to the pattern position of the i-line transmitted through the mask when exposure is performed with the i-line using the phase shift mask having the structure shown in FIG. 74 (d). is there.

On the other hand, when a positive resist is used, through the same steps as in the case of using the negative resist, as shown in FIG. 557 is formed. Note that each reference numeral in the figure indicates the same object as in FIG.

FIG. 77 is a phase profile diagram of transmitted light (i-line) corresponding to the phase shift mask shown in FIG.

In the phase shift masks shown in FIGS. 74 (d) and 76, as shown in the phase profile diagrams of FIGS. 75 and 77, the i-line (ia) passing through the design pattern 553 of each mask is used. And ic) and the i-line (ib and id) transmitted through the auxiliary patterns 554A and 554B are out of phase by 180 degrees. For this reason, the i-line (ia, ic) scattered laterally from the region directly below the design pattern of the resist film to be exposed is shifted 180 degrees out of phase from the region immediately below the adjacent auxiliary pattern. I
The lines are canceled by the lines (ib, id), the contrast of the end face of the exposed area is increased, and the resolution is improved.

The opening width of the auxiliary pattern is formed to be narrow enough that the amount of light for exposing to the bottom of the resist film is not obtained in the standard exposure. No pattern is transferred onto the wafer.

The prior art shown in FIGS.
For example, it has been proposed in JP-A-61-292643, JP-A-62-67514 and JP-A-62-18946.

[0028]

However, the conventional phase shift mask has the following problems.

First, in the case of a mask that does not use a phase shift, it is difficult to form a pattern thinner than the wavelength of light due to the physical resolution limit of the optical system. In order to realize a thin line width, it is necessary to make a structural change such as reducing a light wavelength to be used or increasing a numerical aperture. Therefore, future I
A fine pattern required by C or the like cannot be formed by an optical method.

Second, in the case of a mask using a phase shift, it can be applied only to a pattern having a regularity such as a so-called line and space, and cannot be applied to the manufacture of an IC including various patterns. In addition, since an opaque layer is always required, a fine pattern cannot be formed.

Third, in the case of a mask using a phase shift, since a finer auxiliary pattern than the design pattern must be patterned using an exposure technique, it is necessary to design the auxiliary pattern so as not to exceed the resolution limit. Pattern miniaturization is limited.

Fourth, in the case of a mask using a phase shift, pattern data including unique auxiliary pattern data, phase shift pattern data, and the like must be created in addition to the design pattern data. Man-hours increase.

Fifth, in the case of a mask using a phase shift, in the case of using an organic substance such as a resist for the phase shifter,
Since it is difficult to control the film quality and the film thickness that affect the refractive index, it is difficult to form a phase shift pattern with an accurate and uniform phase shift amount.

Sixth, in the case of a mask using a phase shift, since the phase shifter is made of a different material from the glass substrate, reflection occurs at the boundary between the phase shifter and the glass substrate, thereby lowering the exposure efficiency.

Therefore, according to the present invention, a mask capable of forming a narrow pattern exceeding the resolution of the related art, forming a fine pattern with improved resolution, and simplifying pattern data can be provided. It is an object of the present invention to provide a manufacturing method and a pattern forming method using a mask.

[0036]

SUMMARY OF THE INVENTION The above-mentioned object is applied to an exposure.
A transparent substrate layer that is transparent to light
A mask having an opaque layer formed thereon;
At least a part of the pattern is formed of the opaque layer and the opaque layer.
It consists of a phase shift layer provided only at the edge part,
The phase of light transmitted through the phase shift layer and the shift layer are provided.
The phase of the light transmitted through the part that is not
A shift occurs, and the phase shift layer replaces the phase shift layer.
Exclude the transparent substrate layer so as to surround the periphery of the exposed portion.
And the mask pattern has a resolution limit of 30 to 6
A portion consisting of only a phase shift layer having a size of 0%
A portion in which a phase shift layer is provided at an edge portion of the opaque layer
Achieved by a mask characterized by comprising
it can.

The above-described problem is to reduce the light transmitted through the mask
Image onto the photoresist layer on the wafer through the
A mask for forming a turn on the photoresist layer was used.
In the pattern forming method, the mask is a light used for exposure.
A transparent substrate layer that is transparent to the
Opaque layer, and at least one of the mask patterns
Part is provided only on the opaque layer and the edge of the opaque layer.
And a transparent phase shift layer.
The phase of the passed light and the portion where the phase shift layer is not provided
A predetermined phase shift occurs with the phase of light transmitted through
Wherein the phase shift layer is the transparent substrate excluding the phase shift layer.
The mask is formed so as to surround the periphery of the exposed portion of the layer.
Pattern is 30 to 60% of the resolution limit
Having only the phase shift layer and the opaque layer
And the portion where the phase shift layer is provided at the edge portion of
And a predetermined pattern is formed by the phase shift layer and the opaque layer.
A mask characterized by including a step of forming a turn.
It can also be achieved by the pattern forming method used. Said position
In the region where the phase shift layer surrounds the exposed portion of the transparent substrate layer,
The size is larger than the pattern to be formed.
May be. Further, the mask pattern is formed only from the opaque layer.
And the portion having the phase shift layer are mixed
It may be a configuration. Further, the light used for the exposure
Jarco coherency σ is set in the range of 0.3 ≦ σ ≦ 0.7
It may be.

[0038]

BRIEF DESCRIPTION OF THE DRAWINGS FIG .
A comparative example for facilitating the understanding will be described. mask
FIG. 1 shows a first comparative example . The mask 1 includes a transparent substrate layer 2 transparent to light L used for exposure, and a mask pattern layer 5 formed on the transparent substrate layer 2. The mask pattern layer 5 is composed of a phase shift layer 3a through which light L can pass.

Of the light transmitted through the mask 1, the light transmitted through the phase shift layer 3a and the light transmitted only through the transparent substrate 2 are out of phase. Therefore, at the boundary between the light transmitted only through the transparent substrate 2 and the light transmitted through the phase shift layer 3a and causing a phase shift, the light intensity is reduced due to interference. Thereby, at the time of exposure, an interference pattern smaller than the wavelength of the light L used for exposure can be formed on a wafer (not shown). Further, by adjusting the thickness of the mask pattern layer 5, the amount of phase shift of light can be adjusted to improve the resolution of the exposure pattern.

FIG. 2 shows one of the phase shift layers 3 a of the mask 1.
FIG. 4 is a diagram for explaining light intensity at one edge portion.
In FIG. 5A, the phase of light between the light transmitted through the transparent substrate 2 and the phase shift layer 3a and the light transmitted only through the transparent substrate 2 is shifted, for example, by approximately 180 degrees. Accordingly, the electric vector E and the light intensity P of the light transmitted through the mask 1 are as shown in FIGS. As is apparent from FIG. 3C, the line pattern can be exposed by utilizing the change in the light intensity at the edge of the phase shift layer 3a.

FIG. 3 is a diagram for explaining the light intensity at both edge portions of the phase shift layer 3a of the mask 1. FIG. 2, the same parts as those of FIG. 2 are denoted by the same reference numerals, and the description thereof will be omitted. In this case, when the width W of the phase shift layer 3a is sufficiently small, the light intensity P of the light transmitted through the mask 1 becomes as shown in FIG. Thus, the width of the line pattern can be controlled by controlling the width W.

FIG. 4 schematically shows an optical system used for exposure using the mask 1. The light source 6 is composed of, for example, a mercury lamp, and the mercury lamp is provided with a filter (not shown) that allows only the i-line (wavelength: 365 nm) to pass. Light from the light source 6 reaches the mask 1 located at the focal length of the lens system 8 as light L via the illumination lens system 8. Here, partial coherence (partial coherency) σ of light L
Is 0.5, but σ may be in the range of 0.3 ≦ σ ≦ 0.7.

The light transmitted through the mask 1 forms an image on a wafer 11 on which a photoresist layer 10 has been applied, via an imaging lens system 9. Here, the imaging lens system 9 comprises a 1/5 reduction lens, and the numerical aperture NA is 0.5. The wafer 11 is vacuum-sucked by a flat chuck (not shown) to keep the wafer flat.

Next, a method of forming a pattern using a mask will be described.
A first comparative example will be described with reference to FIG. 5 showing an exposure state in FIG. The light passing through the illumination lens system 8 in FIG. 4 becomes a light 7a transmitted through the mask pattern layer 5 of the mask 1 and a light 7b transmitted through a portion without the mask pattern layer 5, and has a phase difference of 180 degrees.

These lights 7a and 7b are transmitted to the imaging lens system 9
Is formed on the photoresist layer 10 on the wafer 11 through the interface, but a sharp change in light intensity occurs at a portion corresponding to the edge portion of the phase shift layer 3a due to interference. Therefore, a pattern smaller than the wavelength of light can be formed on the wafer 11.

Next, the first and second ratios of the mask manufacturing method will be described.
Comparative examples will be described with reference to FIGS. 6 and 7, respectively. 6 and 7
The same parts as those in FIG. 1 are denoted by the same reference numerals.

In FIG. 6A, the transparent substrate layer 2 is formed of a material that transmits i-rays such as quartz and glass. A resist material 3 is applied on the transparent substrate layer 2 as shown in FIG. As the resist material 3, materials such as an EB (electron beam) resist, a photo resist, and an ion resist are used. When an EB resist is used as the resist material 3, the resist pattern 4 is formed as shown in FIG. 6C by drawing and developing with an electron beam. Then, as shown in FIG. 6D, a silicon oxide constituting the phase shift layer 3a is formed on the surface of the resist pattern 4.
Sputter with a thickness of 0.388 μm. Thereafter, the EB resist is stripped with a resist stripping agent, whereby a mask pattern layer 5 of a phase shift layer (silicon oxide layer) 3a is formed on the transparent substrate layer 2 as shown in FIG.

In the comparative example shown in FIG. 7, an aluminum oxide thin film layer 12 is formed on the transparent substrate layer 2 shown in FIG. 7A and similar to FIG. 6, as shown in FIG. 7B. On the aluminum oxide thin film layer 12, silicon oxide to be the phase shift layer 3a shown in FIG.
Form 0.388 μm. Further, a resist material 3 is applied on the phase shift layer 3a as shown in FIG. And
By drawing and developing by electron beam,
A resist pattern 4 shown in (E) is formed. Thereafter, plasma etching or reactive ion etching of the phase shift layer 3a is performed using carbon tetrafluoride (CF ₄ ) gas. The resist material 3 is peeled off by ashing (ashing) using oxygen plasma to form a mask pattern layer 5 of the phase shift layer (silicon oxide layer) 3a shown in FIG. Here, the aluminum oxide thin film layer 1
Since 2 is not etched with carbon tetrafluoride, it becomes an etching stopper, and a mask pattern layer 5 of a 0.388 μm phase shift layer (silicon oxide layer) 3a can be obtained accurately.

The thickness of the phase shift layer 3a is 0.388 μm
Is a thickness that shifts the wavelength of i-line 365 nm by 180 degrees (opposite phase). The phase shift amount and the thickness of the phase shift layer 3a are represented by the following general formula (2).

(N · t / λ) − (t / λ) = S (2) In the equation (2), n is the bending rate of the phase shift layer 3 a that shifts the phase, λ is the wavelength of the used light L, S is the phase shift amount / 2π
(1/2 for the opposite phase), and t is the thickness of the phase shift layer 3a.

In the comparative examples shown in FIGS. 6 and 7, the bending ratio of the phase shift layer (silicon oxide layer) 3a is n = 1.47, the wavelength of the used light L is λ = 0.365 nm, and the phase shift amount is S.
= 1/2, the equation (2) becomes the following equation (3).

(1.47 t / 0.365) − (t / 0.365) = 1/2 (3) As a result, the thickness t of the phase shift layer 3 a becomes 0.388 μm. Here, the reason why the phase shift amount is set to 1/2 (180 degrees) is that the phase shift amount is optimal for pattern formation due to interference.

FIG. 8 is a plan view of the mask pattern layer 5 formed on the mask 1 as described above. When forming a 0.35 μm line and space pattern with a 1/5 reduction lens system, the width b should be 0.15 μm (reticle size 0.75 μm)
And the width a of the region 13 without the phase shift layer 3a is 0.55 μm.
m (reticle size 2.75 μm), the light intensity on the wafer 11 can be obtained as shown in FIG. 9B.

FIG. 9B is compared with the conventional case shown in FIG. 9A. For convenience of explanation, the conventional method uses an opaque layer
451 (FIG. 71) is made of chrome having a thickness of 50 to 80 nm and a width of 0.35 μm (reticle size 1.75 μm), and the width of the interval between the opaque film patterns is set to 0.35 μm (reticle size 1.75 μm).
As shown in FIG. 9A, the light intensity is about 50 in the conventional method.
%, And the contrast is clearly low.
In this state, a photoresist layer 455 (FIG.
It is impossible to form a pattern in 2). On the other hand, in the comparative example, as shown in FIG. 9B, the light intensity on the wafer 11 is about 80%. Therefore,
It is clear that the light intensity is significantly improved, and the contrast (resolution) is also greatly improved because there is no change in the dark part.

Next, a pattern forming method using a mask will be described.
A second comparative example will be described with reference to FIG. FIG.
1A and 1B are a plan view and a cross-sectional view of a mask 1 for forming a pattern in which a large pattern and a fine pattern are mixed on a wafer, respectively. The hatched area 14 is an opaque layer made of, for example, chrome having a thickness of 50 to 80 nm. or,
The white background is a portion where the transparent substrate layer 2 is exposed, and is divided into a large pattern region 15 and a fine pattern region 16. The phase shift layer 3a is formed in the fine pattern area 16. The large pattern region 15 is patterned by a conventional method, and the fine pattern region 16 uses the phase shift layer 3a of the present invention. Since a large pattern and a fine pattern are mixed in a pattern of an IC or the like, this comparative example is suitable for patterning an IC. As described above, the pattern formed on the wafer 11 by the mask 1 is shown in FIG.
0 (C). According to this comparative example, even a pattern including a fine pattern can be formed in a small number of steps, and the resolution can be sufficiently improved.

Next, a first embodiment of the mask according to the present invention will be described with reference to FIG. In a mask 1A shown in FIG. 11, a phase shift layer 3a is formed at an edge portion of an opaque film layer 14 formed on a transparent substrate layer 2. Opaque film layer 14
Is made of, for example, chromium having a thickness of 50 nm to 80 nm. For convenience, it is assumed that the fine pattern is isolated on the left part of the transparent substrate layer 2 in the drawing, and the fine pattern is adjacent to the right part of the drawing on the transparent substrate layer 2. 3b comprises only a phase shift layer 3a for forming a fine pattern,
A pattern is formed on the wafer 11 as described later by the effect of light interference due to the phase shift. Note that a, b, and
c will be described later.

FIG. 12 shows the phase shift layer 3a of the mask 1A.
FIG. 4 is a diagram for explaining light intensity at an edge portion of FIG. In FIG. 2A, the phase of light between the light transmitted through the transparent substrate 2 and the phase shift layer 3a and the light transmitted only through the transparent substrate 2 is shifted, for example, by approximately 180 degrees. Accordingly, the light intensity P of the light transmitted through the mask 1 is as shown in FIG. As is clear from FIG. 7B, the line pattern can be exposed by utilizing the change in light intensity at the edge of the phase shift layer 3a.

FIG. 13 shows the phase shift layer 3a of the mask 1A.
FIG. 4 is a diagram for explaining light intensity at both edge portions of FIG.
In the figure, the same parts as those in FIG. In this case, the light intensity P of the light transmitted through the mask 1A is as shown in FIG. This allows
By controlling the width W, the width of the line pattern can be controlled.

Next, a first embodiment of a pattern forming method using a mask according to the present invention will be described with reference to FIG. 14, which shows an outline of an optical system used for exposure using a mask 1A. In this figure, substantially the same parts as those in FIG. 5 are denoted by the same reference numerals, and description thereof will be omitted.

In FIG. 14, light L entering the mask 1A from the light source 6 via the illumination lens system 8 is transmitted to the opaque layer 1
4 does not pass. The phase of the light 7a transmitted through the phase shift layer 3a is shifted, and the light 7b transmitted only through the transparent substrate layer 2
The phase is inverted. The lights 7 a and 7 b are imaged on the photoresist layer 10 on the wafer 11 via the imaging lens system 9. In this case, the light 7b transmitted only through the transparent substrate layer 2 and the light 7a transmitted through the phase shift layer 3a and phase-converted interfere with each other, causing a rapid change in light intensity. Therefore,
The contrast between the portion where the opaque layer 14 is provided and the portion where the opaque layer 14 is not provided is improved.
Here, the conditions of the optical system in FIG. 14 are the same as those in FIG. Therefore, the thickness of the phase shift layer 3a is set to 0.388 μm and the amount of phase shift is set to 180 degrees (inversion angle) according to the above equation (3).

Here, the effect obtained by forming the phase shift layer 3a on the edge of the opaque layer 14 will be described with reference to FIGS. FIG. 15A shows a case where the phase shift layer 3a is not provided at the edge of the opaque layer 14, and FIG. 15B shows a case where the phase shift layer 3a has a width b = 0.15 μm.
Shows the case where m is provided. The light intensity distribution in this case is shown in FIG.
Are shown correspondingly to In the case of FIG. 16B, it can be seen that the change in the light intensity distribution is sharp at the portion of the phase shift layer 3a (around 0.0 μm on the X axis). That is, FIG.
The case (B) indicates that the ability to form a fine pattern is large. In this case, the boundary between the phase shift layer 3a and the exposed portion of the transparent substrate layer 2 is 0.0 μm on the X coordinate, and the boundary between the opaque layer 14 and the phase shift layer 3a is −0.15 μm on the X coordinate. Then, the boundary between the white background and the black character of the formed pattern is +0.1 μm, which is obtained from the light intensity distribution in FIG.

Therefore, i-line (wavelength 365 nm), numerical aperture NA
In the case of a lens of 0.5, the exposed portion of the transparent substrate layer 2 is made 0.2 μm (0.1 μm on one side) larger than the design dimension of the pattern to be finally obtained (1.0 μm for the 5 × reticle size)
(0.5 μm on one side)) and the phase shift layer 3a
When formed at 0.15 μm (reticle size 0.75 μm), a resist pattern can be formed with design dimensions. In the following description, it is assumed that a reticle is made in accordance with these rules for convenience.

It is to be noted that, when the wavelength of light or the numerical aperture of the lens changes, the rule is that the amount of dimensional conversion at the exposed portion of the transparent substrate layer 2 and the width of the phase shift layer 3a formed around the edge of the opaque layer 14 Needs to be changed. For example, K
When using a rF (krypton fluoride) laser (wavelength 248 nm) and a lens having a numerical aperture NA = 0.48, the exposed portion of the transparent substrate layer 2 is set to 0.13 μm (0.065 μm on one side) from the design dimensions.
) (5x reticle size 0.65μm (0.3 on one side)
Optimum results can be obtained by setting the width of the phase shift layer 3a around the edge of the opaque substrate layer 14 to 0.06 μm. In this case, when the white patterns approach each other, the region of the opaque layer 14 between the white patterns disappears, and the white patterns are separated only by the phase shift layer 3a on the mask.

Since it is empirically known that the resolution is improved by 20% according to the phase shift method, the amount of dimensional conversion at the exposed portion of the transparent substrate layer 2 and the width of the phase shift layer 3a are determined on the premise of this. Required. That is, when the numerical aperture NA of the lens is i = 0.5 at the i-line, the resolution limit when no phase shift is performed is represented by 0.6 × (λ / NA). Accordingly, the resolution limit when performing the phase shift is 0.6 × (λ / NA) × 0.8. When the numerical value is substituted, the width of the phase shift layer 3a becomes 0.35.
μm.

FIG. 17 shows a case where the present embodiment is applied to a so-called line and space pattern. In the figure,
The phase shift layer 3a is formed on the edge of the opaque layer 14, and the phase shift layer 3a is also formed between the four spaces where the transparent substrate layer 2 is exposed. Light intensity distribution when the dimension conversion amount a in the exposed portion of the transparent substrate layer 2 and the width b of the phase shift layer 3a are changed in a line and space pattern having a resolution limit of 0.35 μm when the phase shift is performed. Is shown in FIG. FIG. 18A shows that a = 0.45 μm and b = 0.2
FIG. 18B shows the light intensity distribution in the case of 5 μm.
FIG. 18C shows the light intensity distribution when 0.50 μm and b = 0.20 μm, FIG. 18C shows the light intensity distribution when a = 0.55 μm and b = 0.15 μm, and FIG. 18D shows a = 0.60 μm , B = 0.10
The light intensity distribution of μm is shown. In order to form a line and space pattern, it is necessary to set a + b = 0.70 μm. As shown in FIGS. 18A to 18D, when the width b of the phase shift layer 3a is large, the peak of the light intensity decreases. Further, when the width b of the phase shift layer 3a is small, the light intensity in the space part is increased, and the contrast is reduced. The contrast is optimum when b = 0.15 μm, and then a = 0.55 μm. Since the boundary between the phase shift layer 3a and the exposed portion of the transparent substrate layer 2 is shifted so as to make the exposed portion of the transparent substrate layer 2 larger than the designed dimension of 0.10 μm on one side, it corresponds to the above-described example of the edge of the large pattern. Here, when the phase shift is performed, the width b of the phase shift layer 3a is 30 to 60 (%) of the resolution limit, that is, 0.144 to 0.22.
Must be 8 μm.

In particular, a high contrast can be obtained by setting b at 40 to 50% of the resolution limit.

On the other hand, the light intensity distribution in the conventional case where the width of the exposed portion of the transparent substrate layer 452 is 0.35 μm (reticle size 1.75 μm) and the width of the opaque layer 451 is 0.35 μm (reticle size 1.75 μm) is shown in FIG. 19. In FIG. 19, the light intensity of the present embodiment is about 80% in FIG. 18C, whereas the conventional light intensity is about 55%, indicating that the contrast in the present embodiment is improved. Understand.

The width b of the phase shift layer 3a is set at 30 which is the resolution limit.
To be 60% means that the shifter pattern needs to be smaller than the design pattern by 20 to 35% (one side) of the resolution limit. For i-line lenses with NA = 0.5, 0.
070 to 0.123 μ (one side) smaller than the design pattern.
Further, when the pattern of the phase shift layer is arranged between the opaque layer and the exposed portion of the transparent substrate, the phase shift layer pattern width is 1.0 to less than the reduction width of 20 to 35% (one side).
Experience has shown that 1.5 times is better. This width is
In the case of an i-line lens with NA = 0.5, the value is 0.070 to 0.185 μm.

For example, a KrF excimer laser (wavelength 24
8 nm) and a lens with a numerical aperture NA = 0.48,
The resolution limit when performing a phase shift is 0.25 μm.
In this case, the dimension conversion amount a at the exposed portion of the transparent substrate layer 2
FIGS. 20B to 20F show light intensity distributions when the width b of the phase shift layer 3a is changed. Note that FIG.
Indicates a case where no phase shift is performed. FIG. 20 (B) shows a
= 0.35 μm, b = 0.15 μm.
FIG. 20C shows the light intensity distribution when a = 0.36 μm and b = 0.14 μm, and FIG. 20D shows the light intensity distribution when a = 0.38 μm and b
FIG. 20 (E) shows the light intensity distribution when a = 0.40 μm and b = 0.10 μm, and FIG. 20 (F) shows the light intensity distribution when a = 0.42 μm and b = 0.08. The light intensity distribution in the case of μm is shown. As shown in FIG.
When the width b of the phase shift layer 3a is 0.12 μm, an optimum contrast can be obtained. FIG. 20 shows that the width b of the phase shift layer 3a is
Is 30 to 60% of the resolution limit when the phase shift is performed.
It can be understood that it is only necessary to be within the range.

FIG. 21A shows a part of FIG. 11, in which the phase shift layer 3a is provided on both edges of the opaque layer 14.
Is formed, and the transparent substrate layer 2 is exposed between the phase shift layers 3a. FIG. 21B shows a conventional photomask 450.
And an opaque layer 451 on the transparent substrate layer 452.
In this case, a mask pattern composed of only the mask pattern is formed.

First, the width a of the transparent substrate layer 2 exposed in FIG. 21A is set to 0.55 μm (reticle size 2.75 μm) and the width b of the phase shift layer 3a is set to 0.15 μm.
m (reticle size 0.75μm). FIG. 21 (B)
The width d where the transparent substrate layer 452 is exposed is 0.35 μm (1.75 μm on the reticle). FIG. 22 shows the light intensity distributions in these cases. As shown in FIG. 22A, the light intensity in the present embodiment is about 100%, while the light intensity in the conventional example is about 65% as shown in FIG.
It can be seen that the contrast is improved in this embodiment. Therefore, as described above, a good result can be obtained by forming the portion where the transparent substrate layer 2 is exposed to be larger than the pattern to be obtained and arranging the phase shift layer 3a on the outer side of the pattern.

Here, FIG. 23 shows an example in which a black pattern is formed on a white background. FIG. 23A shows a mask for forming a 0.35 μm black pattern by using a phase shift. According to the above rules, for example, the width c is 0.15μ
m (reticle size 0.75 μm) becomes a mask pattern composed of only the phase shift layer 3a. FIG. 24A shows the light intensity of the light imaged on the wafer in this case. On the other hand, FIG.
3 (B) shows a mask when no phase shift is performed. The mask shown in FIG. 23B is made of an opaque layer 14 having a width of 0.35 μm (reticle size 1.75 μm), and the light intensity to be imaged is shown in FIG. As shown in FIG. 24, when the phase shift is performed, the fall of the light intensity is sharper, and the resolution of the black pattern is improved.

FIG. 25 shows a case where this embodiment is applied to an IC pattern. FIG. 25A is a plan view showing the shape of a part of the mask 1A. FIG. 25B is a plan view showing a pattern formed on the wafer 11. In FIG. 3A, a phase shift layer 3a is formed in a fine pattern portion adjacent to a pattern at an edge portion of the opaque layer 14 and a portion where the transparent substrate layer 2 is exposed. That is, in FIG.
Indicates an isolated pattern area, and 16 indicates a pattern adjacent area.

Next, a pattern formed on the wafer 11 using the mask 1A is shown in FIG. 26A, and the state of light interference by the mask 1A is verified. In FIG. 26A, the pattern width g is fixed at 0.35 μm, and h is 0.35 μm, 0.4 μm.
, 0.5 μm, 0.6 μm, 0.7 μm, 0.8 μm, 1.0 μm
Should be changed to In this case, according to the rule described above, when the patterns having the pattern widths h of 0.35 μm, 0.4 μm, and 0.5 μm are close to each other, the white pattern is separated only by the phase shift layer 3a as shown in FIG. . When the pattern width h is 0.6 μm, 0.7 μm, 0.8 μm
When the patterns of m and 1.0 μm are separated to some extent, as shown in FIG. 26C, the opaque layer 14 exists between the white patterns in the mask pattern.

Here, the pattern width h on the wafer is 0.35 μm.
m. First, the mask dimension a of FIG.
m (reticle size 2.75μm), b 0.15μm (reticle size 0.75μm), c 0.15μm (reticle size 0.75μm)
) Is shown in FIG. On the other hand, the conventional mask dimension e in FIG.
FIG. 27B shows the light intensity distribution when m (reticle size 1.75 μm) and f is 0.35 μm (reticle size 1.75 μm). Comparing the two, the light intensity obtained in the present example is about 85%, while the light intensity in the related art is about 55%.
%, Which indicates that the contrast is improved in this embodiment.

FIG. 28 shows the light intensity distribution when the pattern width h is 0.4 μm, FIG. 29 shows the light intensity distribution when h is 0.5 μm, and FIG. 30 shows the light intensity distribution when h is 0.6 μm.
FIG. 31 shows the light intensity distribution when h is 0.7 μm.
The light intensity distribution at 0.8 μm is shown in FIG.
FIG. 33 shows the light intensity distribution at the time of. FIG. 28-
As is clear from FIG. 33, according to the present embodiment, the light intensity is increased and the contrast is improved as compared with the conventional example.

The table shows the pattern size of the isolated pattern 15 shown in FIG. 25A with respect to the pattern width h.

[0078]

[Table 1] As is clear from the table, the pattern size change of the isolated pattern 15 due to the influence of light interference is within ± 0.01 μm.
Therefore, although it is possible to control the pattern size, in order to control the pattern size more accurately, FIG.
The values of the pattern widths a, b, c, and i in (B) and (C) may be changed.

Next, an example in which the present embodiment is applied to a contact hole pattern will be described. FIG. 34 (A)
The mask 1A in the case of forming a contact hole of 0.35 μm is shown. According to the above rules, the transparent substrate layer 2
Is 0.55 .mu.m, and the width b of the phase shift layer 3a is 0.15 .mu.m. The light intensity in this case is shown in FIG.
It is shown in (A). On the other hand, FIG. 34B shows a mask in which no phase shift is performed. The size l of the exposed portion of the transparent substrate layer 2 is 0.35 μm. The light intensity in this case is shown in FIG.
It is shown in (B). As shown in FIG. 35, the contrast of the contact hole can be greatly improved by applying this embodiment.

Next, the mask manufacturing method according to the present invention will be described .
The first and second embodiments will be described with reference to FIGS. 36 and 37, respectively. In FIGS. 36 and 37, substantially the same parts as those in FIGS. 6 and 7 are denoted by the same reference numerals, and description thereof will be omitted.

The transparent substrate layer 2 shown in FIG. 36A is made of, for example, quartz. On this transparent substrate layer 2, for example, a thickness of 5
An opaque layer 14 of 0-80 nm chromium is shown in FIG.
It is formed as shown in FIG. Then, a resist material 3 made of, for example, an EB resist is applied on the opaque layer 14, and a normal mask manufacturing process (not shown) including EB drawing, development, etching, and resist peeling is performed. afterwards,
The pattern 20 of the opaque layer 14 is formed as shown in FIG. A resist material 3 made of an EB resist is applied again on the pattern 20, and a resist pattern 4 shown in FIG. 36D is formed by EB drawing and developing steps. Further, on the resist pattern 4, for example, a thickness of 0.
A phase shift layer 3a of 388 μm silicon oxide is formed as shown in FIG. Next, resist material 3
Is stripped with a stripping agent, whereby the opaque layer 14 and the mask pattern layer 5 of the phase shift layer 3a are formed as shown in FIG.

In the embodiment of FIG. 37, FIG.
By the same process as (C), a thickness of 50 to 50 mm is formed on the transparent substrate layer 2.
Pattern 20 of opaque layer 14 of 80 nm chromium
Are formed as shown in FIG. Then, a stopper layer 21 serving as a stopper when performing a plasma etching of an oxide film described later is formed on the pattern 20 as shown in FIG.
It is formed as shown in FIG.

The stopper layer 21 is obtained by, for example, forming a thin film of aluminum oxide, which is resistant to plasma etching using carbon tetrafluoride (CF ₄ ), on the pattern 20 by sputtering or the like. On this stopper layer 21, a phase shift layer 3a made of, for example, silicon oxide is formed to a desired thickness by sputtering or the like as shown in FIG. This desired thickness is set to 0.388 μm (Equation (3)) to cause a 180 ° phase shift. On this phase shift layer 3a, a resist material 3 made of, for example, an EB resist is applied, and a resist pattern 4 shown in FIG. 37D is formed by EB drawing and development processes. Then, the phase shift layer 3 exposed by the CF ₄ plasma
a, and the phase shift layer 3 is removed by peeling the resist.
FIG. 37 shows a mask pattern layer 5 for phase shift according to FIG.
It is formed as shown in FIG.

Next, the mask manufacturing method according to the present invention will be described.
A third embodiment will be described with reference to FIG. In this embodiment, FIG.
A second embodiment of the mask according to the present invention shown in FIG. The second embodiment of the mask has a configuration in which a region corresponding to the phase shift pattern of the glass substrate is dug deeper than the design pattern to be transferred to shift the phase of light.
It has a design (transfer) pattern of μm width.

In FIG. 38A, a chromium (Cr) layer 32 having a thickness of about 500 to 1000 ° is formed as an opaque layer by sputtering or the like on a glass substrate 31 made of synthetic quartz and having a thickness of about 2 to 3 mm as usual. . The Cr layer 32 may have a thickness of, for example, 0.35 μm by lithography using normal EB exposure.
In the case of a design pattern having an m width, an opening pattern 33 having a width of 0.85 μm is formed.

Next, in FIG. 38B, a conductive layer for preventing charge-up during EB exposure is formed on the substrate 31.
Form 34. The conductive layer 34 is formed, for example, by sputtering molybdenum silicide (MoSi ₂ ) to 200 to 300
It is obtained by forming to a thickness of about Å. Then, the substrate 31
After applying a negative EB resist layer 35 thereon and performing pre-baking, a pattern having a width of 0.55 μm is EB-exposed almost at the center of the opening pattern 33 of the Cr 32. 35A
Indicates an exposed resist layer.

Next, the resist layer 35A exposed to the design pattern formation region by performing normal development is removed as shown in FIG.
And the resist layer 35A
The exposed conductive layer (MoSi ₂ layer) 34 is selectively removed by dry etching using a mixed gas of carbon tetrachloride (CCl ₄ ) and oxygen (O ₂ ) using the mask as a mask.

In FIG. 38D, the resist layer 35
Reactive ion etching (RIE) using a mixed gas of carbon tetrafluoride (CF ₄ ) and O _{2 on} the upper surface of the glass substrate 31 exposed corresponding to the phase shift pattern using the A and the Cr layer 32 as a mask. Etching by processing. On both sides of a 0.55 μm-wide pattern 36 having the same surface as the upper surface of the glass substrate 31, 0.15 μm-wide phase shift patterns 37A and 37B having lower upper surfaces are formed. The etching depth D for shifting the phase of the transmitted light of the phase shift patterns 37A and 37B by 180 degrees with respect to the light transmitted through the design pattern 36 is such that the mask has a wavelength λ = 36
It is used for exposure with i-line having a wavelength of 5 nm, and since the refractive index of glass is n = 1.54, about 0.36
μm.

Next, in FIG. 38E, the resist layer 35A is removed by a normal ashing process, and the MoSi ₂ layer 34 under the resist layer 35A is removed by a dry etching process using a mixed gas of CCl ₄ and O _2. The mask 1C according to the third embodiment of the present invention is completed.

FIG. 39 shows a mask 1C shown in FIG.
Shows the phase profile of the light transmitted through the mask 1C corresponding to FIG. In the figure, ia indicates the transmitted light of the central pattern 36, and ib indicates the transmitted light of the phase shift patterns 37A and 37B.

The mask 1C has a structure in which the phase shift pattern is dug deep by using a negative resist when patterning the phase shift pattern. However, when a positive resist is used,
FIG. 38 shows a third embodiment of a mask according to the present invention in which the center is dug deeper than the phase shift pattern, contrary to the present embodiment.
It is formed by substantially the same process.

Next, the mask manufacturing method according to the present invention will be described.
Fourth embodiment will be described with reference to FIG. In this embodiment, FIG.
A mask 1D shown in FIG. In FIG. 40, FIG.
The same parts as in FIG. 8 are denoted by the same reference numerals, and description thereof will be omitted.

In FIG. 40A, a hole pattern 33 having a width of 0.85 μm is formed in a Cr layer 32 formed on a glass substrate 31. MoS for preventing charge-up is provided on the substrate 31.
After forming the i ₂ layer 34, the positive type EB is formed on the substrate 31.
A resist layer 38 is formed. Next, this resist layer 3
The EB exposure of the design pattern is performed on the central portion of the opening pattern 33 of the Cr layer 32 of FIG. 38A indicates an exposure area.

Next, in FIG. 40B, the exposed region 38A of the resist layer 38 is removed by development, and the resist layer 3 is removed.
The exposed MoSi ₂ layer 34 is removed by dry etching with a mixed gas of CCl ₄ and O ₂ through the opening 39 of No. 8. Also, the upper surface of the glass substrate 31 exposed in the opening 39 by RIE using a mixed gas of CF ₄ and O ₂ is etched to a depth D of about 0.36 μm as in the case of FIG.

In FIG. 40C, the resist layer 38 is removed by ashing, and then the MoSi ₂ layer 34 under the resist layer 38 is removed by dry etching using a mixed gas of CCl ₄ and O _2. . Thereby, the mask 1D is completed.

FIG. 41 shows a mask 1D shown in FIG.
Shows the phase profile of light transmitted through the mask 1D. In the figure, Ic denotes transmitted light of the central pattern 40, and id denotes transmitted light of the phase shift patterns 41A and 41B.

Next, the mask manufacturing method according to the present invention will be described.
A fifth embodiment will be described with reference to FIG. FIG. 1A shows a transparent substrate 51 made of, for example, glass or fused quartz. As shown in FIG. 1B, a phase shift layer 52, for example, a silicon oxide film is formed on a transparent substrate 51. The phase shift layer 52 is formed by a vapor deposition (CVD) method,
It is formed by an on-glass (SOG) method. In the present embodiment, the phase shift layer 52 is used to invert the phase of light.
It has a thickness of 00 °.

In FIG. 42C, an opaque layer 53 is formed on the phase shift layer 52. The opaque layer 53 is made of, for example, chromium (Cr). When Cr is used, the thickness of the Cr is, for example, 700 to 1000 °.

Next, EB resist coating, EB lithography, development and etching of the opaque layer 53 are performed, and FIG.
An opaque layer pattern 53a shown in (D) is formed.

The pattern 53a of the opaque layer is used as a mask, and the phase shift layer 52 is removed by etching as shown in FIG. In this embodiment, since the phase shift layer 52 is an oxide film, RIE using a mixed gas of CF ₄ and CHF ₃ is used for etching.

Next, as shown in FIG. 42F, an unnecessary portion 53b of the opaque layer 53 is exposed by applying an EB resist layer 54, performing EB drawing and developing.

Finally, the EB resist layer 54 is removed,
By removing unnecessary portions 53b of the opaque layer 53 by etching, a mask 1E which is a fourth embodiment of the mask according to the present invention shown in FIG. 42 (G) is completed. FIG.
FIG. 42H is a plan view of the mask 1 </ b> E corresponding to FIG.

Next, the mask manufacturing method according to the present invention will be described.
Sixth Embodiment Description will be given with reference to FIG. 42, those parts which are substantially the same as those corresponding parts in FIG. 42 are designated by the same reference numerals, and a description thereof will be omitted. The steps in FIGS. 43A, 43B, and 43C correspond to the steps in FIGS. 42A, 42C, and 42D, respectively, but the phase shift layer 52 is omitted in this embodiment. ing.
In FIG. 43D, the pattern 53a of the opaque layer is used as a mask, and the transparent substrate 51 is removed by etching. As the etching, for example, CF ₄ and CHF ₃
The transparent substrate 51 is etched to a predetermined depth by using RIE using a mixed gas of In the case of inverting the phase of light, the predetermined depth is 3900 ° in this embodiment.

In FIG. 43E, an unnecessary portion 53b of the opaque layer 53 is exposed by applying an EB resist layer 54, performing EB drawing and developing.

Finally, the EB resist layer 54 is removed,
Unnecessary portions 53b of the opaque layer 53 are removed by etching to complete the mask 1C shown in FIG. FIG. 43G is a plan view of the mask 1C corresponding to FIG. 43F.

According to the fifth and sixth embodiments of the method of manufacturing the mask, the opaque layer is used as a phase shift layer or a mask for etching the transparent substrate. Can be prevented from shifting.

Next, the mask manufacturing method according to the present invention will be described.
Seventh embodiment will be described with reference to FIG. 43, those parts which are substantially the same as those corresponding parts in FIG. 43 are designated by the same reference numerals, and a description thereof will be omitted. 44A, an opaque layer 53 is formed on a transparent substrate 51, an EB resist layer (not shown) is applied on the opaque layer 53, and a pattern is formed by EB drawing. The pattern is developed and the opaque layer 53 is patterned by etching. The opaque layer 53 is made of, for example, Cr.
And is formed on the transparent substrate 51 by sputtering. After patterning, the EB resist is removed.

In FIG. 44B, the transparent substrate 51 is etched to a predetermined depth by using the pattern 53c of the opaque layer 53 as a mask. The etching condition is such that the exposed transparent substrate 51 is etched by a parallel plate electrode type high frequency excitation (13.56 MHz) RIE apparatus. The etching is performed, for example, at a power of about 0.1 W / cm ² and a pressure of 0.1 to 0.05 Torr using a mixed gas of CF ₄ or CHF ₃ and O ₂ as an etchant gas.
Control is performed while detecting with an end point detector so that the etching depth becomes λ / 2 (n−1).
When λ = 0.365 μm, λ / 2 (n−1) = 0.39 μm.

In FIG. 44 (C), after etching the transparent substrate 51, a positive resist layer 55 (for example, OFPR-800) is applied to a thickness of about 1.0 μm on the transparent substrate 51 and the pattern 53c, and prebaked at 100 ° C. I do.

In FIG. 44 (D), light from the back surface of the transparent substrate 51 to the resist photosensitive region, for example, 40 to 6 at a wavelength of 436 nm.
The entire surface is irradiated with monochromatic light of 0 mJ / cm ² to expose only the resist layer 55 on the transparent substrate 51. Then, the transparent substrate 51
Is immersed in a 2.38% aqueous solution of TMAH for 40 seconds to perform development processing, so that the resist layer 5 is formed only on the pattern 53c.
5 remains. Further, a rinsing process and a drying process are performed.

After forming the resist pattern, the opaque layer 5
The pattern 53c of FIG.
Partial removal as shown in (E). The amount of side etching is controlled by the etching time so that about 0.4 to 0.8 μm of the opaque layer 53 is removed on one side.

Finally, in FIG. 44 (F), the resist layer 55 remaining on the opaque layer 53 is ashed and removed by O ₂ ashing to complete a mask substantially the same as the mask 1D shown in FIG. 40 (C). I do.

Next, the second step of the method for manufacturing a mask according to the present invention will be described.
An eighth embodiment will be described with reference to FIG. 44, those parts which are substantially the same as those corresponding parts in FIG. 44 are designated by the same reference numerals, and a description thereof will be omitted.

The steps shown in FIGS. 45A to 45C are as follows.
The steps may be the same as the steps shown in FIGS. However, in FIG. 45B, the transparent substrate 51 is set to λ / 2 (n−
Etch to the depth of 1). When λ = 0.365 μm, λ / 2 (n−1) = 0.39 μm.

In FIG. 45 (D), as described with reference to FIG. 45 (D), the entire surface of the transparent substrate 51 is irradiated with light in the resist-exposed region to expose only the resist layer 55 on the transparent substrate 51. Removed by development. In this embodiment, the opaque layer 53 is partially exposed by over-exposure, over-development, or short-time ashing after development and drying.

Next, as shown in FIG. 45E, the exposed opaque layer 53 is removed by etching using the resist layer 55 as an etching mask.

In FIG. 45F, the resist layer 55 remaining on the opaque layer 53 is ashed and removed by O ₂ ashing, and a mask substantially the same as the mask 1D is completed.

Next, the mask manufacturing method according to the present invention will be described.
A ninth embodiment will be described with reference to FIG. 44, those parts which are substantially the same as those corresponding parts in FIG. 44 are designated by the same reference numerals, and a description thereof will be omitted.

The steps shown in FIG.
In this embodiment, the resist layer 55 is formed on the transparent substrate 51 without etching the transparent substrate 51 as shown in FIG.

In FIG. 46C, the resist layer 55 on the transparent substrate 51 is formed in the same manner as in the step of FIG.
Remove only Thereafter, in this embodiment, the resist layer 55
Is used as a mask to etch the transparent substrate 51 to a depth of λ / 2 (n−1). When λ = 0.365 μm, λ / 2
(N-1) = 0.39 [mu] m.

In FIG. 46D, a part of the pattern 53c of the opaque layer 53 is removed by side etching. The steps in FIG. 46 (E) may be the same as the steps in FIG. 44 (F).

According to the embodiment shown in FIGS. 44 to 46, there is no need to align the pattern of the opaque layer and the pattern of the phase shift layer, so that a highly accurate mask without pattern displacement can be manufactured. Further, depending on the material of the phase shift layer, it is conceivable that problems such as adhesion to the substrate and cleaning may occur. However, in these embodiments, since a part of the transparent substrate functions as the phase shift layer, the material of the phase shift layer is Does not matter. Therefore, there is no need to consider the stability and durability of the material of the phase shift layer, and a fine pattern exceeding the resolution limit of the exposure apparatus can be easily formed.

Next, a pattern forming method using a mask will be described in more detail. FIG. 47 is a diagram for explaining a pattern forming method using the edge of the phase shift layer of the mask, and corresponds to FIG. In FIG.
The same reference numerals are given to substantially the same parts as those described above, and the description thereof is omitted.

A phase shift layer 3 a having a refractive index n and a thickness t is formed on a transparent substrate 2 to form a mask 1. The pattern of the mask 1 is imaged on a photoresist layer 10 on a semiconductor substrate (wafer) 11 by an imaging lens system 9.
At this time, a desired pattern is imaged by the edge of the phase shift layer 3a.

The phase shift layer 3a transmits incident light, but imparts a phase shift to transmitted light by having a refractive index n different from that of air or vacuum. The phase shift amount S is S = (n−1) t / λ (2π (n−1) t / λ in the case of radian display). The following description is based on the assumption that the phase shift amount S is π (opposite phase).

As shown in the middle part of FIG. 47, the light transmitted through the mask 1 has its electric vector E changed in phase to the opposite phase corresponding to the pattern of the phase shift layer 3a. When light having such a distribution of the electric vector E of light is incident on the photoresist layer 10 and absorbed, the light intensity distribution at that time is proportional to E ^2, and therefore, as shown in the lower part of FIG. To form a black pattern PB having a small width. That is,
As shown in the middle of the figure, when the sign of the electric vector E of light is inverted, there is a point where the electric vector E becomes zero. Since the distribution of the obtained light intensity P is proportional to the square of the electric vector E, the minimum value of P becomes 0 at the point of E = 0. Therefore, a clear black pattern is obtained.

Generally, when the phase shift layer 3a changes the phase of transmitted light, a pattern in which the phase change is spatially distributed is obtained. When the phase change is in the opposite phase, the light intensity P at the position where the phase changes is always 0, and even when integrated over time. Therefore, the light intensity becomes 0 at the edge of the phase shift layer 3a. When the phase change is other than the opposite phase, the electric vector E has the same sign or the opposite sign depending on time, and the light intensity obtained by time integration does not become zero.

An example of image formation using such an edge portion of the phase shift layer will be described with reference to FIGS.

FIG. 48A shows a light intensity distribution when the phase shift layer gives a phase shift of π (opposite phase). In the figure, the region of the phase shift π corresponds to the region to the left of 0.0 on the horizontal axis, and is in contact with the opening of the right phase shift 0 at the position of 0.0. FIG. 48A shows the light intensity distribution at one edge portion of such a phase shift layer. At the center of the pattern corresponding to the end of the phase shift layer, the light intensity decreases to almost zero. According to such an imaging principle, it is possible to obtain a line pattern exceeding the conventional resolution.
The phase difference at the edge of the phase shift layer is 180 ° ± 30 °.
°, the same effect can be obtained.

FIG. 48B shows a light intensity distribution when the phase shift layer gives a phase shift of π / 2. When the phase shift is π / 2, there is a case where the phase of the light transmitted through the phase shift layer and a phase of the light transmitted through the opening become opposite to each other, and when the light is integrated over time. The minimum value of the light intensity does not become zero. In the case shown in the figure, the minimum value of the light intensity is about 0.5 or more, which is about half of the light intensity of about 1.0 in the uniform portion. If the development level is set to be equal to or less than the minimum value of the light intensity, the development in which the pattern in the figure is ignored can be performed. The phase difference at the edge of the phase shift layer is 90 ± 15.
°, the pattern is ignored and not formed under normal development conditions (development level 35%). If the development level is set between the minimum value and the uniform value of the light intensity, the pattern is developed at the central portion.

Next, a pattern forming method using a mask will be described.
A third comparative example will be described with reference to FIG. FIG. 49 (A),
(B) is a diagram for explaining formation of a loop-shaped pattern according to this comparative example. FIG. 49A shows a mask pattern. A mask 61 is formed on an opening 60 formed from a transparent substrate. The mask 61 has a phase shift amount corresponding to a half-wavelength optical path difference of the incident light, and gives a π phase shift to the transmitted light. The result of imaging such a mask pattern is as shown in FIG. The portion where only the opening 60 or the phase shift layer 61 is imaged has a constant light intensity and becomes white on the screen. Opening 60
A black pattern 62 is formed at the boundary between the phase shift layer 61 and the phase shift layer 61. That is, when the light transmitted through the opening 60 and the light transmitted through the phase shift layer 61 are mixed by interference, the opposite phases cancel each other, so that the light intensity becomes 0 and a black pattern is formed.

When the phase shift layer 61 is formed of a silicon oxide film and the i-line of a mercury lamp is used as light, the refractive index of the phase shift layer 61 is about 1.47, and the refractive index of air having a refractive index of about 1.00 is about It has a refractive index difference of 0.47. The thickness of the silicon oxide film forming the half-wavelength optical path difference is about 0.388 μm. The opposite phase is most effective for canceling the amplitude of the combined light due to the interference. However, even if the phase is not the opposite phase, an effective decrease in the light intensity can be obtained within a range of, for example, ± 30%. .

When a phase shift layer 61 having a phase shift π is formed on an opening having no phase shift as in the mask pattern shown in FIG. 49, a black pattern is formed at a portion corresponding to the edge of the phase shift layer 61. Is done. In this case, the black pattern has a closed loop shape.

FIGS. 50A and 50B show a mask pattern for forming an open shape (line segment) and its image forming pattern.

FIG. 50A shows the shape of a mask pattern. A phase shift layer 61 whose one side forms a target line segment is formed on the opening 60.
Adjacent to the remaining (unwanted) edge of
2 is formed. That is, one side 61 out of four sides of the phase shift layer 61 having a phase shift of π
“a” forms a boundary with the region where the phase shift is 0, and forms a black pattern 65 as shown in FIG.
Generate. The other sides 61b, 61c, 61d are
Since it is in contact with the phase shift layer 64 having a phase shift of π / 2, the phase change when traversing the edge is π / 2, and the decrease in light intensity at the edge is small. The amount of phase shift on the side around the phase shift layer 64 is also π / 2, and similarly, the decrease in light intensity is small. By adjusting the development level, only the black pattern 65 can be left. Although the case where one gray level is formed between the white and black levels has been described, two or more halftone levels may be used.

As described above, by using a plurality of types of phase shift layers having different amounts of phase shift, a pattern such as an open line segment can be formed.

FIGS. 51A and 51B show a mask pattern and an imaging pattern for forming a dot pattern.

In FIG. 51A, a phase shift layer 61 having a phase shift of π is formed on an opening 60, and a phase shift layer having a phase shift of π / 2 is formed except for one apex. 64 are surrounding. That is, the phase shift layer 61 has the opening 60 only at one vertex thereof.
Is in contact with When such a mask pattern is imaged, a black pattern 66 can be formed only in a portion where the phase shift layer 61 and the opening 60 are in contact with each other, as shown in FIG. Phase shift layer 61 and phase shift layer 64
Since the phase shift amount is π / 2 at the boundary with, the decrease in light intensity is small compared to the portion of the black pattern 66. or,
The phase shift amount of the outer peripheral portion of the phase shift layer 64 is also π / 2, and the decrease of the light intensity is similarly small. Therefore, only the black pattern 66 can be developed as a pattern, and the other halftone patterns can be treated as white patterns.

The case where one linear pattern is formed has been described above. Hereinafter, the formation of a pattern having an intersection will be described.

FIGS. 52 (A), (B) and (C) show a mask pattern and an image forming pattern for forming a pattern of intersecting lines.

FIG. 52A shows a first mask pattern. The plane is divided into four quadrants corresponding to the area divided by the intersecting lines, and an opening 60 having no phase shift is provided in a first quadrant, and a phase shift is provided in two quadrants adjacent to the opening 60. π phase shift layer 61 is provided,
In the other quadrant, another phase shift layer 67 having a phase shift of 2π is provided. That is, the boundaries between the quadrants are accompanied by a phase shift π. FIG. 52B shows a second mask pattern. As in FIG. 52A, the plane is divided into four quadrants, and the first quadrant does not have a phase shift.
A phase shift layer 61 having a phase shift of π is provided in 0 and two adjacent quadrants, and another opening 68 having a phase shift of 0 is provided in the other quadrant. Also in this case, a phase shift π occurs between adjacent quadrants.

When such a mask pattern is imaged, an image pattern as shown in FIG. 52 (C) is obtained. That is, a portion having a uniform phase is imaged as a white pattern, and a portion having a phase shift of π is a black pattern.
The image is formed as

Although the case where the straight lines intersect is illustrated and described, the intersecting line is not limited to a straight line but may be any curve.

FIGS. 53A and 53B show a mask pattern and an image forming pattern in the case of forming a T-shaped pattern in which one straight line hits another straight line and ends there. In FIG. 53A, phase shift layers 61a and 61b having a phase shift of π are formed adjacent to the opening 60 having a phase shift of 0. Also, the phase shift layer 61a
Is provided above the aperture 68 to form a boundary having a phase shift of π therebetween. Since these boundaries are accompanied by a phase shift of π, a black pattern is formed when an image is formed as shown in FIG. When the phase shift layers 61b and 68 are directly adjacent to each other, the boundary is imaged as a black pattern with a phase shift of π, so that a phase shift layer 71 having a phase shift of π / 2 is formed between them. . That is, since only a phase shift of π / 2 occurs at the boundary of the phase shift layer 71, the decrease in light intensity is relatively small. By adjusting the development threshold, such a decrease in light intensity can be developed as a white pattern. As a result, a black pattern 72 is formed as shown in FIG.

In the mask pattern shown in FIG. 53A, a case is shown in which there is a gap between the opening 68 and the phase shift layer 61b, but such a gap may be smaller than a certain value.

In a semiconductor device or the like, a wide area for making contact is provided in the middle of a wiring pattern. FIGS. 54A to 54D show a mask pattern and an imaging pattern for forming such a wiring pattern.
Shown in

In FIG. 54A, an opening 60 having a phase shift of 0 and a phase shift layer 61 having a phase shift of π are in contact with each other to form a linear boundary. An opening 74 having a phase shift of 0 is formed in the phase shift layers 75 and 61 having the shift π. Both the opening 74 and the phase shift layer 75 have a rectangular shape, and the boundary between the opening 74 and the phase shift layer 61 is linearly arranged together with the boundary between the opening 60 and the phase shift layer 61. All the boundaries shown by solid lines in the figure are boundaries accompanied by a phase change of π.

In FIG. 54B, the boundaries defining each area are formed in the same manner as in FIG. 54A. However,
The portion that was the opening 74 in FIG. 54A is replaced with a phase shift layer 76 having a phase shift of 2π. FIG.
As in FIG. 4A, a phase shift of π is formed between the upper and lower regions of the horizontal straight line. Further, a rectangular boundary of phase shift π is formed at the center.

FIG. 54C shows another mask pattern. It is largely separated into two regions by a horizontal straight line. At the upper part, an opening 60 without phase shift is arranged on the right side, and a phase shift layer 61 with a phase shift π is formed on the left side. Are in contact with each other at different lengths. Further, a phase shift layer 77 having an intermediate phase shift π / 2 between both regions is formed in the upper part in the drawing. A structure symmetrical to the upper part is formed below the center line. That is, a phase shift layer 61 having a phase shift of π is formed below the opening 60, and an opening 68 without a phase shift is formed below the phase shift layer 61, and is in contact with each other for a limited length. I have. Intermediate phase shift π / 2
The phase shift layer 77 having these regions 61 and 68
Is formed in the middle. As a whole, a boundary with a phase shift of π is formed along a horizontal straight line, and also at a central vertical line segment.

FIG. 54 (D) is similar to FIG. 54 (A), (B),
FIG. 4 shows an image forming pattern when a mask pattern as shown in FIG. That is, a black pattern 78 having a region whose width is increased at the center is formed.

Note that FIGS. 52A, 52B and 53
In the case of setting the line length to a finite length in the mask patterns of FIGS. 54A, 54A, 54B, and 54C, a region having an intermediate level phase shift such as π / 2 is set in an unnecessary portion. It may be formed.

The mask patterns for forming various image forming patterns have been described above, but what kind of light intensity profile can be obtained by numerical calculation for some of them will be described below.

In the calculation, light having a wavelength of 365 nm was used, and an imaging lens having a numerical aperture NA = 0.50 and an illumination lens having a partial coherency σ = 0.50 were used.

FIGS. 55A to 55C show examples of crossing line patterns.

FIG. 55A is a schematic diagram showing a pattern of intersecting lines and a sample link area thereof. FIG.
The mask pattern shown in (B) is adopted, and the coordinates of the X, Y, and Z axes are used in the direction shown on the right side in the figure. The region indicated by the broken line was set as a sampling region, and the light intensity in the region was calculated according to the model. FIG. 55 (B)
3 is a graph showing a light intensity profile in a sampling area shown in FIG. As can be clearly understood from the figure, a deep valley is formed at the boundary with the phase shift π, the lower half shown in FIG. 55 (A) is not shown, but has a symmetric structure. That will be obvious to those skilled in the art. When such a light intensity profile is represented by a contour line with respect to the light intensity, FIG.
As shown in FIG. That is, a region having the lowest light intensity is formed in the lower side extending in the X direction and a central portion extending in the Y direction in the drawing, and a portion where the light intensity gradually increases is formed adjacent to these regions.

Next, the light intensity of the example of the wiring pattern shown in FIG. FIG. 56A shows a mask pattern of the type shown in FIG. 54A, and its sampling region is indicated by a broken line. Also, as shown in the right part of the figure, the coordinates of the X, Y, and Z axes are used.

FIG. 56B shows the light intensity profile in the sampling area. A narrow valley is formed along the X direction, and the valley is widened near the region where X = 0.
When this profile is projected on the XY plane, FIG.
As shown in FIG. It can be clearly seen from the figure that a wide valley is formed at the center.

FIGS. 57A to 57C are diagrams showing a case where the wiring mask pattern shown in FIG. 54C is used.

FIG. 57A shows a mask pattern and its sampling area. On the horizontal axis at the center in the figure, an opening 60 with a phase shift of 0 and a phase shift layer 6 with a phase shift of π
1 is formed in contact with a part thereof, and a phase shift layer 77 having a phase shift of π / 2 is formed therebetween. The upper side of the opening 60 forms an angle θ ₁ with the horizontal direction.
Two sides of the phase shift layer (intermediate region) 77 form an angle θ ₂ . Further, the upper side of the phase shift layer 61 forms an angle θ ₃ with the horizontal direction. A phase shift layer 61 having a phase shift of π is formed symmetrically in contact with the opening 60 below the horizontal axis, and an opening 68 having a phase shift of 0 is formed below the phase shift layer 61. Further, another phase shift layer 77 is formed at a position symmetrical to the phase shift layer 77. The width of a narrow region where the central openings 60 and 68 and the phase shift layer 61 are in contact is W1. In this sample,
θ1 = θ2 = θ3 = 60 degrees and W1 = 0.2 μm. FIGS. 57B and 57C show an image forming pattern formed by using such a mask pattern. The area indicated by the broken line is the sampling area.

FIG. 57B shows the light intensity profile in the sampling area. In the figure, it can be seen that a deep valley is formed in the X direction, a valley extends in the Y direction at the center, and a shallow valley that branches further is formed.

The light intensity profile of FIG.
When projected onto a plane, the result is as shown in FIG.

FIGS. 58A to 58C show examples of mask patterns having a configuration similar to the mask pattern shown in FIG. 51A. In the figure, the opening 6 having a phase shift of 0 is shown on the right side.
0 is formed, and a phase shift layer 61 having a phase shift of π with a triangular tip on the left side is disposed, and a width W2 is provided between the two.
Contact is formed. A phase shift layer 64 having a phase shift of π / 2 is formed in an intermediate region between the opening 60 and the phase shift layer 61. The light intensity distribution was calculated for a sampling area surrounded by a broken line in the figure. Note that W
As for 2, a gap of 0.08 μm is set, and the angle formed by the two boundary lines formed by the phase shift layers 61 and 64 is 6
0 degrees.

FIG. 58 (B) shows a light intensity profile in the sampling area, and FIG. 58 (C) shows its projection on the XY plane. It will be understood that an elliptical minimum light intensity region is formed near (0.0).

FIGS. 59A and 59B show a line-and-
FIG. 3 is a diagram for explaining a space pattern.

FIG. 59A shows a line-and-space mask pattern. Open area 80 with zero phase shift
And a phase shift layer 81 having a phase shift of π are alternately formed to form a line and space pattern.
For example, when the light intensity profile formed in the imaging pattern is calculated on the assumption that the opening region 80 and the phase shift layer 81 have a width of 0.5 μm, the result is as shown in FIG. 59B. A minimum light intensity is formed in a portion corresponding to a boundary between the opening region 80 and the phase shift layer 81. The light wavelength was 365 μm, the numerical aperture NA = 0.53, and the partial coherency σ = 0.50.

FIGS. 60A and 60B show the case where mixed patterns of various sizes are formed. FIG. 60A shows a mask pattern. On the upper stage, an opaque layer (Cr layer) 85 and a phase shift layer 86 for forming a large black pattern
A phase shift layer 87 for forming an image of the phase shift layer itself is formed in an intermediate portion, and a phase shift layer 87 for forming an image pattern in a side (edge) portion is formed in a lower portion. A shift layer 88 is formed. Around the phase shift layer 88, a phase shift layer 8 having an intermediate phase shift for preventing an unnecessary side from being imaged.
9 are formed. Phase shift layers 86, 87, 88
Gives a phase shift π, and the phase shift layer 89 gives, for example, a phase shift π / 2. Numerical aperture NA for i-line = 0.4
In the case of using a lens of .about.0.6, a pattern of 0.5 .mu.m or more is formed by arranging a region 86 having a phase shift .pi. 0.3-0.
In order to form a 5 μm black pattern, as shown in the middle part, a mask is formed with a pattern of only the phase shift layer 87 having a phase shift of π. The pattern of 0.25 μm or less is formed by the boundary between the phase shift layer 88 having a phase shift of π and the opening region having a phase shift of 0. At this time, the unnecessary side of the phase shift layer 88 is surrounded by the phase shift layer 89 showing an intermediate phase shift so that an image is not formed by adjusting the development level.

FIG. 60B shows an example of an image forming pattern.
A black pattern 91 corresponding to the mask formed by the opaque layer 85 and the phase shift layer 86 is formed at the upper part, a black pattern 92 corresponding to the phase shift layer 67 is formed at the intermediate part, and a black pattern 91 is formed at the lower part. A thin black pattern 93 corresponding to the boundary between the phase shift layer 88 and the opening region 90 is formed. When an IC is manufactured using photolithography, an image of a reticle pattern is formed on a wafer using an imaging lens system. With the miniaturization of the IC pattern, the numerical aperture NA of the imaging lens system is increased to improve the resolving power, thereby coping with the miniaturization of the IC pattern. However, if the numerical aperture NA of the imaging lens system is increased in order to improve the resolution,
The depth of focus FD becomes smaller according to the following equation (4).

(Equation 1) Here, K ₂ is a process coefficient. Therefore, a pattern cannot be accurately formed on a surface having irregularities.

An embodiment capable of forming a pattern even on a surface having irregularities will be described. FIG. 61 shows a second comparative example of the mask . In the figure,
1 are denoted by the same reference numerals, and description thereof is omitted. In this comparative example, the phase shift layer 3a has a thickness D
1 and a thickness D2 (D2> D1)
And a shifter portion 3a2 having By providing the shifter portions having different thicknesses as described above, it is possible to arbitrarily control the focal position of the image at the time of exposure.

Next, a method of forming a pattern using a mask will be described.
A fourth comparative example will be described. In this comparative example, the thickness of the phase shift layer of the mask is partially varied as necessary. For convenience of explanation, it is assumed that a pattern is formed using the mask 1A shown in FIG. 11 and the optical system shown in FIG. 14, and the relationship between the thickness of the phase shift layer 3a and the shift of the focal position (hereinafter referred to as defocus) will be described. I do. Light used, phase shift layer 3
The material and the like of “a” are the same as those in the embodiment described above. (2)
When the thickness t of the phase shift layer 3a when obtaining the phase shift amounts S of 180, 120, and 240 degrees is calculated using the equation, 0.38
8, 0.259 and 0.517 μm.

First, when t = 0.388 μm and S = 180 degrees, if the defocus amount is 0, the light intensity distribution becomes as shown in FIG.
The result is as shown in FIG. However, if the direction approaching the wafer 11 along the optical axis of the lens system 9 is set to the plus (+) direction and the defocus amounts are set to +1.0 and +0.5 μm,
The light intensity distribution is as shown in FIGS. 62 (A) and (B), respectively. On the other hand, when the defocus amounts are set to -0.5 and -1.0 μm, the light intensity distributions are as shown in FIGS. 62 (D) and (E), respectively. From FIG. 62, it can be seen that the light intensity distribution changes in the same manner regardless of whether the focal position shifts in the plus direction or the minus (-) direction.

Next, when t = 0.259 μm and S = 120 degrees, the light intensity distributions obtained by setting the defocus amounts to +1.0, +0.5, 0, and −0.5 μm are shown in FIG. 63 (A). ),
(B), (C), and (D) show. FIG. 63 shows that the maximum contrast is obtained when the defocus amount is +0.5 μm.

When t = 0.517 μm and S = 240 °, the light intensity distributions obtained by setting the defocus amounts to +0.5, 0, −0.5 and −1.0 μm are shown in FIGS. 64 (A) and 64 (B), respectively. ),
(C) and (D) show. FIG. 64 indicates that the maximum contrast is obtained when the defocus amount is -0.5 μm.

Therefore, in this comparative example, the focal position can be controlled within a range of 0.5 μm by appropriately setting the thickness t of the phase shift layer.

In the IC manufacturing process, the surface of a part of the IC may be higher than the other part. For example, in a dynamic random access memory (DRAM), when a stacked capacitor is used, the memory cell section is 0.5 to 1.0 μm higher than other peripheral circuit sections.
In such a case, if a pattern is formed to the resolution limit using a lens having a large numerical aperture, the depth of focus becomes shallow as described above, and it is difficult to apply the method to the manufacture of an IC having irregularities.

However, according to this comparative example, application to a DRAM using a stacked capacitor as shown in FIG. 65 is also possible. In FIG. 65, substantially the same parts as those in FIGS. 14 and 61 are denoted by the same reference numerals, and description thereof will be omitted. In this case, a lower portion on the IC surface, such as a peripheral portion of the DRAM, is patterned using the thinner shifter portion 3a1 of the phase shift layer 3a. Therefore, D1 is made thinner than 0.388 μm. For example, the cell portion of the DRAM and the peripheral circuit portion are 1
If there is a step of about μm, set D1 = 0.259 μm. On the other hand, a high portion on the IC surface, such as a cell portion of a DRAM using a stacked capacitor, is a phase shift layer 3a.
The pattern is formed using the thicker shifter portion 3a2. Therefore, D2 is made thicker than 0.388 μm. For example, if the cell portion is about 1 μm higher than the peripheral circuit portion, D2
= 0.517 μm.

In this way, by controlling the thickness of the phase shift layer in accordance with the irregularities on the surface of the IC, it becomes possible to form a pattern by focusing on all the parts on the surface of the IC.

As described above, in the present invention and the comparative example , a pattern is formed using a phase shift layer. Therefore, an arbitrary fine pattern can be formed by using a mask in which a phase shift layer is appropriately arranged. FIGS. 66 to 69 show the respective masks (A) and the corresponding patterns (B).
Shown in 66 to 69, 90 is an opaque layer, 91 is a phase shift layer, and 92 is a window.

In the above Examples and Comparative Examples , the partial coherency σ was set to 0.5, but is not limited to this, and may be in the range of 0.3 ≦ σ ≦ 0.7.

The light used for exposure using a mask is not limited to i-line. Further, the light may be irradiated from the side of the transparent substrate on which the phase shift layer is provided or from the opposite side. The materials of the transparent substrate and the phase shift layer are not limited to those of the embodiment. For example, the transparent substrate may be transparent to light used for exposure. The phase shift layer is made of, for example, SiO ₂ , Al ₂ O ₃ , M
gF ₂ or the like.

Further, the mask in each embodiment includes a reticle. Therefore, it goes without saying that the pattern forming method using the mask according to the present invention is not limited to pattern formation of a semiconductor device, but can also be applied to pattern formation of a mask or a reticle. Although the present invention has been described with reference to the embodiments, the present invention can be variously modified in accordance with the gist of the present invention, and these are not excluded from the present invention.

[0181]

According to the present invention, a pattern having a narrow width exceeding the conventional resolution can be formed by effectively using a phase shift layer, and a fine pattern can be formed with improved resolution. This is very useful in practice because the pattern data can be simplified.

[Brief description of the drawings]

FIG. 1 is a diagram for explaining a first comparative example of a mask .

FIG. 2 is a diagram for explaining light intensity at one edge portion of a phase shift layer.

FIG. 3 is a diagram for explaining light intensity at both edge portions of a phase shift layer.

FIG. 4 is a schematic configuration diagram of an optical system used for exposure using a first comparative example of a mask.

FIG. 5 is a diagram for explaining a first comparative example of a pattern forming method using a mask.

FIG. 6 is a process diagram illustrating a first comparative example of the mask manufacturing method.

FIG. 7 is a process chart illustrating a second comparative example of the mask manufacturing method.

FIG. 8 is a plan view of a mask pattern.

9 is a diagram illustrating the light intensity distribution of FIG.

FIG. 10 is a diagram for explaining a second comparative example of the pattern forming method using the mask .

FIG. 11 is a sectional view showing a first embodiment of the mask.

FIG. 12 is a diagram for explaining light intensity at one edge portion of a phase shift layer.

FIG. 13 is a diagram for explaining light intensity at both edge portions of a phase shift layer.

FIG. 14 is a diagram illustrating a first embodiment of a pattern forming method using a mask.

FIG. 15 is a diagram for explaining the pattern in FIG. 11;

16 is a diagram illustrating the light intensity distribution in FIG.

FIG. 17 is an explanatory diagram in the case where the first embodiment of the mask is applied to a line-and-space pattern.

FIG. 18 is a diagram illustrating a light intensity distribution when the dimension conversion amount a and the width b of the phase shift layer are changed.

19 is a diagram illustrating a conventional light intensity distribution corresponding to FIG.

FIG. 20 is a diagram illustrating a light intensity distribution when a KrF excimer laser is used in FIG.

FIG. 21 is a diagram showing a part of FIG. 11;

FIG. 22 is a diagram illustrating the light intensity distribution in FIG.

FIG. 23 is an explanatory diagram when a black pattern is formed on a white background.

24 is a diagram illustrating the light intensity distribution in FIG.

FIG. 25 is an explanatory diagram in the case where the first embodiment of the mask is applied to an IC pattern format.

FIG. 26 is a diagram showing a part of the pattern in FIG. 11;

FIG. 27: Pattern width h is 0.3, 0.4, 0.5, 0.6, 0.
It is a figure explaining the light intensity distribution in the case of 7, 0.8, and 1.0 micrometer.

FIG. 28: Pattern width h is 0.3, 0.4, 0.5, 0.6, 0.
It is a figure explaining the light intensity distribution in the case of 7, 0.8, and 1.0 micrometer.

FIG. 29: Pattern width h is 0.3, 0.4, 0.5, 0.6, 0.
It is a figure explaining the light intensity distribution in the case of 7, 0.8, and 1.0 micrometer.

FIG. 30: Pattern width h is 0.3, 0.4, 0.5, 0.6, 0.
It is a figure explaining the light intensity distribution in the case of 7, 0.8, and 1.0 micrometer.

FIG. 31: Pattern width h is 0.3, 0.4, 0.5, 0.6, 0.
It is a figure explaining the light intensity distribution in the case of 7, 0.8, and 1.0 micrometer.

FIG. 32: Pattern width h is 0.3, 0.4, 0.5, 0.6, 0.
It is a figure explaining the light intensity distribution in the case of 7, 0.8, and 1.0 micrometer.

FIG. 33: Pattern width h is 0.3, 0.4, 0.5, 0.6, 0.
It is a figure explaining the light intensity distribution in the case of 7, 0.8, and 1.0 micrometer.

FIG. 34 is an explanatory diagram in the case where the first embodiment of the mask is applied to the formation of a contact hole pattern.

FIG. 35 is a diagram illustrating the light intensity distribution of FIG.

FIG. 36 is a process chart illustrating a first example of a method for manufacturing a mask.

FIG. 37 is a process diagram illustrating a second example of the method for manufacturing a mask.

FIG. 38 is a process diagram illustrating a third example of the method for manufacturing a mask.

39 is a diagram showing a phase profile of light transmitted through the mask shown in FIG.

FIG. 40 is a process chart illustrating a fourth example of a method for manufacturing a mask.

FIG. 41 is a diagram showing a phase profile of light transmitted through the mask shown in FIG.

FIG. 42 is a process diagram illustrating a fifth example of the method for manufacturing a mask.

FIG. 43 is a process diagram illustrating a sixth example of the method of manufacturing a mask.

FIG. 44 is a process diagram illustrating a seventh example of the method of manufacturing a mask.

FIG. 45 is a process diagram illustrating an eighth example of the method for manufacturing a mask.

FIG. 46 is a process diagram illustrating a ninth embodiment of a method for manufacturing a mask.

FIG. 47 is a diagram illustrating a pattern forming method using a mask.

FIG. 48 is a light intensity distribution diagram illustrating imaging using an edge portion of a phase shift layer.

FIG. 49 is a diagram for explaining a third comparative example of the pattern forming method using the mask, and is a diagram for explaining a loop-shaped pattern.

FIG. 50 is a diagram for explaining a third comparative example of the pattern forming method using the mask, and is a diagram for explaining a pattern having an open shape.

FIG. 51 is a diagram for explaining a third comparative example of the pattern forming method using the mask, and is a diagram for explaining a dot pattern.

FIG. 52 is a diagram for explaining a third comparative example of the pattern forming method using the mask, and is a diagram for explaining a pattern of intersecting lines.

FIG. 53 is a view for explaining a third comparative example of the pattern forming method using the mask, and is a view for explaining a T-shaped pattern.

FIG. 54 is a diagram for explaining a third comparative example of the pattern forming method using the mask, and is a diagram for explaining a wiring pattern.

FIG. 55 is a view for explaining a third comparative example of the pattern forming method using the mask, and is a view for explaining an intersecting pattern.

FIG. 56 is a diagram for explaining a third comparative example of the pattern forming method using the mask, and is a diagram for explaining a wiring pattern.

FIG. 57 is a diagram for explaining a third comparative example of the pattern forming method using the mask, and is a diagram for explaining a wiring pattern.

FIG. 58 is a view for explaining a third comparative example of the pattern forming method using the mask, and is a view for explaining an oval pattern.

FIG. 59 is a diagram for explaining a third comparative example of the pattern forming method using the mask, and is a diagram for explaining a line-and-space pattern.

FIG. 60 is a diagram illustrating a third comparative example of the pattern forming method using the mask, and is a diagram illustrating mixed patterns of various sizes.

FIG. 61 is a cross-sectional view showing a second comparative example of the mask.

FIG. 62 is a diagram illustrating a light intensity distribution when phase shift layers having different thicknesses are provided.

FIG. 63 is a diagram illustrating a light intensity distribution when phase shift layers having different thicknesses are provided.

FIG. 64 is a diagram illustrating a light intensity distribution when phase shift layers having different thicknesses are provided.

FIG. 65 is a diagram illustrating a fourth comparative example of the pattern forming method using the mask.

Figure 66 is a diagram showing an example of a pattern that can make form.

Figure 67 is a diagram showing an example of a pattern that can make form.

FIG. 68 is a diagram showing an example of a pattern that can make form.

Figure 69 is a diagram showing an example of a pattern that can make form.

FIG. 70 is a cross-sectional view showing a conventional mask provided with an opaque layer.

71 is a configuration diagram of an optical system that performs pattern formation using the mask of FIG. 70.

FIG. 72 is a diagram illustrating a conventional pattern forming method.

FIG. 73 is a diagram illustrating a light intensity distribution in a conventional example.

FIG. 74 is a process chart of a conventional mask manufacturing method using a negative resist having a phase shift layer.

75 is a diagram showing a phase profile of light transmitted through the mask shown in FIG. 74 (d).

FIG. 76 is a cross-sectional view of a mask having a phase shift layer manufactured using a positive resist.

FIG. 77 is a diagram showing a phase profile of light transmitted through the mask shown in FIG. 76.

──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Masao Taguchi 1015 Kamiodanaka, Nakahara-ku, Kawasaki City, Kanagawa Prefecture Inside Fujitsu Co., Ltd. 72) Inventor Yuichiro Yanagishita 1015 Uedanaka, Nakahara-ku, Kawasaki City, Kanagawa Prefecture Inside Fujitsu Limited (56) References JP-A-2-221451 (JP, A) JP-A-2-140743 (JP, A) (58) Field surveyed (Int.Cl. ⁷ , DB name) G03F 1/00-1/16

Claims (7)

(57) [Claims] 1. A transparent substrate transparent to light used for exposure.
A layer, and an opaque layer formed on the transparent substrate layer.Having
In a mask, at least a part of a mask pattern includes the opaque layer and the opaque layer.
With the phase shift layer provided only at the edge of the bright layer
And the phase of light transmitted through the phase shift layer and the shift layer
With the phase of the light transmitted through the part where no
Phase shift occurs, and the phase shift layer
Surrounding the exposed portion of the transparent substrate layer except for the transparent layer.
Formed, The mask pattern has a size of 30 to 60% of the resolution limit.
A portion consisting only of a phase shift layer having
Depending on the part where the phase shift layer is provided at the edge of the bright layer
Mask characterized by being constituted . 2. The method according to claim 1, wherein the phase shift layer is a transparent substrate layer.
The size of the area surrounding the exposed part is smaller than the pattern to be formed.
2. The structure as claimed in claim 1, wherein the shape is large.
mask. 3. The mask pattern comprises only an opaque layer.
And the portion having the phase shift layer are mixed
3. The mask according to claim 1, wherein: 4. The light transmitted through the mask is passed through a lens system.
The pattern is formed by forming an image on a photoresist layer on the wafer.
Pattern formation using a mask formed on the photoresist layer
In the method, the mask is a transparent substrate layer transparent to light used for exposure.
And an opaque layer formed on the transparent substrate layer, and at least a part of the mask pattern includes the opaque layer and the opaque layer.
A phase shift layer provided only at the edge of the opaque layer;
From it, the phase and the phase shift layer of the light transmitted through the phase shift layer
The phase of the light transmitted through the portion not provided is
Causing a phase shift , wherein the phase shift layer is formed of the transparent substrate layer excluding the phase shift layer.
The mask pattern is formed to surround the periphery of the exposed portion, and the mask pattern has a resolution of 30 to 60% of the resolution limit.
A portion consisting of only the phase shift layer having a size,
A portion where the phase shift layer is provided at an edge portion of the opaque layer;
And a predetermined pattern is formed by the phase shift layer and the opaque layer.
Forming a mask using a mask.
Turn formation method. 5. The method according to claim 1, wherein the phase shift layer is provided on the transparent substrate layer.
The size of the area surrounding the protrusion is larger than the pattern to be formed.
6. The method according to claim 4, wherein the second member is formed to be large.
A pattern forming method using the mounted mask. 6. The mask pattern comprises an opaque layer only.
And the portion having the phase shift layer are mixed
The mask according to claim 4 or 5, wherein
Pattern formation method. 7. A partial coherence of light used for the exposure.
Sea σ is set in the range of 0.3 ≦ σ ≦ 0.7.
A mask using the mask according to claim 4 or 5, wherein
Turn formation method.