JP3240641B2 - Photomask and method for manufacturing semiconductor device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a photomask using a phase shift method used in a photolithography process of semiconductor manufacturing and a method for manufacturing a semiconductor device using the same.

[0002]

2. Description of the Related Art In recent years, in the manufacture of semiconductor devices, miniaturization has been dramatically advanced. In the photolithography process,
In response to miniaturization, we are trying to increase the numerical aperture (NA) of the reduction exposure apparatus (stepper) and shorten the wavelength range in which the light source is used. In recent years, a remarkable improvement in the performance of a resist has also contributed to the miniaturization. In the half-micron process rule, the technology of the photo process will be established by the above-mentioned technology improvement.

However, as the miniaturization progresses further, a decrease in the focus margin due to the increase in the NA of the reduction exposure apparatus becomes a serious problem. Also, because it approaches the limit of miniaturization by light exposure,
There is a great possibility that it is difficult to form a resist pattern required for manufacturing a semiconductor device. The cause is due to the diffraction effect which is a characteristic of light.

FIG. 11A is a cross-sectional view of a photomask generally called a reticle generally used in a reduction exposure apparatus. The photomask made of quartz glass 80 has a plurality of chrome patterns 82 for blocking light during exposure, and a light transmitting portion 84 is formed between them. However, when the adjacent light transmitting portions 84, 84 come close to each other with the miniaturization of the pattern of the semiconductor device for the above-described reason,
Due to the diffraction effect of light, light is not completely separated on the wafer. In other words, on the wafer, FIG.
This is because the phase shown in FIG. The light intensity is the square of the light amplitude.

Therefore, when adjacent patterns come close to each other, as shown in FIG. 11 (C), light intensity is generated also in a portion which needs to be shielded by a light diffraction effect.

[0006]

Therefore, a phase shift method has recently attracted attention as a new technology corresponding to miniaturization using phase, which is another characteristic of light. FIG.
FIG. 2A is a cross-sectional view of a photomask using a phase shift method. The feature of this photomask is that a shifter 94 for changing the phase of light is provided on one of the patterns of the adjacent light transmitting portions 96a and 96b. Like a general photomask, a plurality of chrome patterns 92 are sputtered on a photomask made of quartz glass 90. As in the photolithography process of manufacturing a semiconductor device, a resist is applied on chromium and an electron beam or laser beam is applied. The resist is drawn, exposed, developed, and further etched to form a chromium pattern 92.

As shown in FIG. 12B, the phase of the light transmitted through the photomask using the phase shift method on the wafer is determined by the light transmitting portion 96a where the shifter 94 exists.
At 180 degrees. The chromium pattern 92 also generates light intensity due to the diffraction effect in the part shielded, but the light intensity is the square of the light amplitude. Therefore, as shown in FIG. , Between adjacent light transmitting portions. Therefore, the resist pattern after exposure and development is completely separated, so that a photomask using the phase shift method can easily achieve higher resolution than a general photomask.

FIG. 13 shows a method of manufacturing a photomask by a conventional phase shift method. FIG. 13A is a diagram in which a chromium pattern 92 is formed on a quartz glass 90 by the method described above. Next, as shown in FIG. 13B, SOG 98 is generally applied on the chrome pattern 92 as a material of the shifter 94 for changing the phase. The coating film thickness d is determined by the following equation which is a condition for changing the phase by 180 degrees.

D = λ / [2 × (n 1 −n 2)] λ: Exposure wavelength n 1: Refractive index of shifter n 2: Refractive index of atmosphere around mask If the ambient atmosphere of mask is air, the refractive index n 2 is almost 1. Become. When a mercury lamp g-line is used as a light source for exposure,
SOG has a refractive index of 1.45 and a g-line wavelength of 0.436.
μm, the thickness of the shifter 94 is 0.48 from the above equation.
4 μm.

Further, a resist 99 is applied on the SOG 98 applied with a film thickness d as shown in FIG. FIG. 13D is a drawing in which the resist 99 is drawn and exposed by an electron beam and then developed. Next, the SOG 98 is dry-etched using the resist 99 as a mask (see FIG. 13E), and unnecessary resist 99 is removed by using sulfuric acid, whereby a shifter 94 can be formed as shown in FIG.

Incidentally, the above-mentioned conventional system has the following problems. When a photomask is manufactured using the phase shift method, a chrome pattern 9
After forming 2, SOG98 must be applied,
Further, a large number of steps of drawing exposure, development, and etching after applying the resist 99 have been required.

The application of SOG 98 not only increases the number of steps, but also adversely affects the thickness accuracy of the shifter 94.
That is, the SOG 98 is generally spin-coated on the quartz glass 90 and the chrome pattern 92 using a spin coater. Since the area where the chrome pattern 92 does not exist becomes a step surface on the SOG 98 coating surface, the uniformity of the film thickness of the SOG 98 is significantly reduced. The change in phase due to the shifter 94 is determined by the thickness of the SOG 98, but the variation in phase caused variations in the thickness of the SOG 98. Generally, in the phase shift method, the phase shift due to the shifter is 18
It must be controlled with an accuracy of 0 degrees ± 10 degrees, but SO
At present, the phase shifter cannot be completely controlled with this accuracy in the shifter 94 formed by applying G98.

An object of the present invention is to solve the above-mentioned problems, and an object of the present invention is to provide a photomask using a phase shift method that can reduce a variation in a phase shift in a phase shifter.

Another object of the present invention is to provide an inexpensive photomask with a reduced number of manufacturing steps and an improved yield.

Still another object of the present invention is to provide a method of manufacturing a semiconductor device capable of forming a fine pattern with high accuracy by performing exposure using the above-mentioned photomask.

[0016]

According to the present invention, there is provided a photomask, comprising: a mask substrate made of a material transmitting an exposure light source light; a light shielding layer formed on the mask substrate so as to be spaced apart from the mask substrate; A mask pattern including a light transmission region formed of the mask substrate sandwiched between layers, and a thickness of the mask substrate at both ends of the light transmission region, as compared with an intermediate region sandwiched between the both ends. A phase shifter formed by providing a concave portion so as to be thin, wherein the concave portion is provided on a surface of the mask substrate opposite to a surface on which a light shielding layer is formed. I do.

As described above, the photomask according to the present invention constitutes a phase shifter by making the thickness of the mask substrate itself in the light transmission region smaller than that of the other light transmission regions, and the shifter has a thick light transmission region. The light phase is changed by giving an optical path difference to the region.

In this point, unlike the prior art, a phase shifter is formed by applying SOG on a mask substrate to form a phase shifter, which is different from the prior art in which the optical path difference is made thicker than the mask substrate. Although the principle of shifting the phase by providing an optical path difference is the same, the present invention does not require spin coating of SOG, and the number of steps is reduced accordingly. In addition, the optical path difference for inverting the phase can be made dependent on the thickness processing control of the flat mask substrate itself, as compared with the related art which depends on the film thickness accuracy of spin coating. This processing control is achieved by an etching technique in which processing accuracy higher than film thickness accuracy by spin coating is established.

Further, the photomask according to the present invention comprises a mask substrate made of a material transmitting the light source for exposure, a light shielding layer formed on the mask substrate at a distance from the mask substrate, and sandwiched between the light shielding layers. A main light transmitting region made of the mask substrate, and an auxiliary light transmitting region made of the mask substrate formed by providing a non-formed portion of the light shielding layer near an end of the light shielding layer. A mask having a mask pattern and a phase shifter formed by providing a concave portion in the auxiliary light transmitting region, wherein the concave portion is formed on a surface of the mask substrate opposite to the light shielding layer forming surface. It is characterized by being provided.

In the photomask according to the present invention, in the above photomask, the depth D of the concave portion is such that the wavelength of the light source light for exposure is λ, the refractive index of the mask substrate is n1, and the atmosphere around the mask is Where D is (100 ± 5.56)% × λ / [2 × (n1−n2)], where n is the refractive index of

By defining the depth D of the recess as described above, the optical phase can be shifted by approximately 180 degrees ± 10 degrees.

Alternatively, a method of manufacturing a semiconductor device according to the present invention is characterized in that any one of the above photomasks is used in an exposure step of forming a desired pattern on a resist formed on a layer to be etched.

Further, the method of manufacturing a semiconductor device according to the present invention
It is characterized in that i-line of a mercury lamp is used as light source light for exposure used in the exposure step.

Further, the method of manufacturing a semiconductor device according to the present invention
An excimer laser is used as light source light for exposure used in the exposure step.

[0025]

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

FIG. 1A is a sectional view of a photomask of a first reference example for comparison with the present invention. As with the photomask of the related art, a chrome pattern 22 necessary for manufacturing a semiconductor device is formed on a quartz glass 20 serving as a mask substrate. Light transmitting portions 2 are provided on both sides of the chrome pattern 22.
6a and 26b are formed. The phase shifter 24 is formed on one of the light transmitting portions 26a. Unlike the prior art, the shifter 24 of this reference example is configured by etching the quartz glass 20 itself, which is a mask substrate, instead of the shifter formed by SOG coating. Thus, the film thickness of the quartz glass 20 is changed, and the light transmitting portion 2 without the shifter is formed.
The phase is changed by providing an optical path difference with respect to 6b.

The etching amount D of the quartz glass 20 is obtained by the following equation. Here, λ is the exposure wavelength, the refractive index of the quartz glass 20 is 1.45, and the refractive index of air is 1.

D = λ / [2 × (1.45-1)] Here, g-line (wavelength: 0.436 μm) is used as light source light for exposure.
m), the etching amount D is 0.484 μm
Becomes When the i-line (wavelength: 0.365 μm) is used, the etching amount D is 0.406 μm. In any case, the etching amount D should be controlled within a range of ± 10% of the center value when the above value is set as the center value. This is because the phase can be inverted by 180 ° ± 10 ° by the shifter 24.

According to such a photomask, FIG.
FIG. 1C shows a phase as shown in FIG. 1B, and the light intensity on the wafer is as shown in FIG. 1C, and exposure corresponding to a fine pattern can be realized. Although the pattern pitch in the current technology is 0.5 μm, the pattern pitch in the next generation is 0.35 μm.
In order to realize μm, it is effective to use an i-line of a mercury lamp as a light source for exposure of a photomask employing a phase shift. Then, it is not necessary to use an excimer laser having a shorter wavelength.

FIGS. 2A to 2D are views showing a detailed method of manufacturing the photomask of this embodiment.

First, a chromium pattern 22 is formed on a quartz glass 20 just like a general photomask (FIG. 2A). As a result, light transmitting portions 26a and 26b are formed on both sides of the chrome pattern 22.

Next, a resist 28 for forming a shifter in the light transmitting portion 26a is coated with the quartz glass 20 and the chrome 2
It is applied on two patterns (FIG. 2B).

Next, the resist 28 is drawn and exposed by an electron beam drawing apparatus, and then developed to remove the resist 28 only in the region of one of the light transmitting portions 26a (FIG. 2).
(C)).

Further, using the resist 28 as a mask, the underlying quartz glass 20 is coated with Freon 14 (CF ₄ ) and Freon 23
Dry etching with a mixture of (CHF ₃ ) (FIG. 2)
(D)). In general, the etching amount is proportional to the etching time, and thus, the desired etching amount D is obtained by controlling the etching time. For example, quartz glass 20
, Etching is performed in advance to determine conditions, an etching rate is calculated from the data, and an etching time for obtaining a desired etching amount D can be set. Alternatively, a laser beam may be obliquely incident on the etched portion and the reflected light may be used to measure the film thickness in real time to detect the etching end point.

The dry etching is performed using Freon 1
A mixture of 16 (C ₂ F ₃ ) and Freon 23 (CHF ₃ ),
Alternatively, a mixture of argon (Ar), Freon 14 (CF ₄ ) and Freon 23 (CHF ₃ ), or helium (H
e), a mixture of oxygen (O ₂ ) and Freon 23 (CHF ₃ ).

Here, in the phase shift method, the phase change in the shifter must be controlled at 180 degrees ± 10 degrees. However, the photomask of the present embodiment does not need to apply the shifter material after the chromium pattern is formed. Since the dry etching having a controllability of about 3% with respect to the etching amount is used at the time of forming the shifter, the phase change by the shifter can be controlled almost completely to 180 degrees ± 10 degrees.

After the dry etching, the resist is peeled off with sulfuric acid to complete the photomask as shown in FIG.

After the quartz glass 20 is dry-etched, wet etching or dry etching may be added in order to increase the smoothness of the surface of the light transmitting portion 24. The wet etching treatment is performed, for example, by immersing in a solution in which hydrofluoric acid HF and water or a mixture of hydrofluoric acid HF and ammonium fluoride in a ratio of 1: 100 are used for 10 seconds. The processing by dry etching includes, for example, Freon 14 (CF ₄ ), Freon 23 (C
HF ₃ ), sulfur hexafluoride (SF 6), or nitrogen trifluoride (NF ₃ ).

FIG. 3 shows an exposure step using the photomask of this embodiment.

As a light source for exposure, for example, a mercury lamp 100
And i-line is used as the light source light. The light from the light source 100 is
Photomask 1 focused at 2 and having the structure shown in FIG.
04. The light transmitted through the photomask 104 is
One chip 112 in the wafer 110 is exposed through the reduction projection lens 106. The wafer 110 is the wafer stage 1
20, and can be step-driven in the X and Y directions, and exposure of all the chips 112 on the wafer 110 can be realized by step driving.

FIG. 4 shows a photolithography process of a wafer having the above exposure process.

As shown in FIG. 4A, a semiconductor substrate (S
i) A silicon oxide film (SiO ₂ ) 32 and a polycrystalline silicon film (Poly-Si) which is a layer to be etched on 30
34 are formed. A resist 36 is spin-coated on the polycrystalline silicon film 34. Next, exposure is performed using the photomask 104 as shown in FIG. 3, and by subsequent development, as shown in FIG. 4B, the resist 36 is removed only at a portion corresponding to the etching region. Here, the exposure of the resist 36 in the region corresponding to the chrome pattern 22 shown in FIG.
The edge of the resist 36 can secure a vertical edge. Thereafter, dry etching of the polycrystalline silicon film 34 is performed using the resist 36 as a mask (FIG. 4).
(C), a desired pattern of the polycrystalline silicon film 34 can be obtained by removing the resist 36 (FIG. 4).
(D)).

FIG. 5 is a sectional view of a photomask of a second reference example for comparison with the present invention. A chrome pattern 42 is formed on the quartz glass 40, and a shifter 44 is formed by etching the opposite side of the quartz glass 40 from the surface on which the chrome pattern 42 is formed.
The phase is changed by the optical path difference between 46b. The etching amount D for forming the shifter 44 and the method of manufacturing the photomask are the same as those in the reference example shown in FIG. 1, and only the etching location is different.

In the present embodiment, even when the quartz glass 40 is etched with a width larger than the width of the transmission part 46a, the exposure is blocked by the chrome pattern 42.
There is an advantage that the accuracy of the etching width can be reduced. Further, in the embodiment shown in FIG. 1, when the quartz glass 20 is etched, side etching may be performed up to the end of the chrome pattern 22, and it is necessary to consider this side etching in advance. The advantage is that you don't have to take that into account. In other words, since there is no risk of side etching of the chrome pattern 42,
A higher resolution photomask can be realized.

FIG. 6 shows a third embodiment in which the second embodiment is applied to an edge-enhanced photomask. As shown in FIG. 6A, the point that a chrome pattern 52 is formed on a quartz glass 50 and transmission portions 56 and 56 are provided on both sides of the chrome pattern 52 is the same as in the above-described embodiment. The transmission section 56 has shifters 58, 58 at both ends in contact with the chrome pattern 22. This shifter 58 is formed by etching the quartz glass 50 by a depth D, and the phase of the transmitted light is shifted by 180 degrees compared to the intermediate region 56a of the transmitting portion 56 (FIG. 6 (B)). )reference). Thereby, the edge of the transmitted light is enhanced near the shifter 58 (see FIG. 6C). Since the shifter 58 is formed by etching the quartz glass 50, the manufacturing process can be simplified as compared with a case where a shifter region is formed by adding SOG.

FIG. 7 shows a first embodiment of the present invention. In the third embodiment shown in FIG. 6, a shifter 58 is formed on the surface of the quartz glass 50 opposite to the surface on which the chrome pattern 52 is formed. An example is shown. Also in this case, even if the etching area for the shifter 58 is enlarged on the back side of the chrome pattern 52, the exposure is blocked by the chrome pattern 52, so that there is an advantage that the accuracy of the etching width can be reduced. Further, since there is no possibility of side etching of the chrome pattern 52, there is an effect that a higher resolution photomask can be realized.

FIG. 8 shows a fourth embodiment in which this embodiment is applied to an auxiliary pattern type photomask. This photomask has a chrome pattern 6 on quartz glass 60.
By forming 2, a main light transmitting portion 66 is provided on both sides of the chrome pattern 62, and a pattern non-forming portion is provided near both ends of the chrome pattern 62, and this region is used as an auxiliary light transmitting portion 64. I have. The shifter 68 has a quartz glass 60 in the region of the auxiliary light transmitting portion 64.
It is configured by reducing the thickness of the. As a result, in the shifter region 68, the transmitted light having a phase different from that of the main light transmitting portion 66 by 180 degrees is superimposed, so that the transmitted light to the chrome pattern 52 side due to the light diffraction effect is eliminated (FIG. 8 ( A), (C)). Since the shifter 68 is formed by etching the quartz glass 60 as in the case of the edge-enhanced photomask, the manufacturing process can be simplified as compared with the case where the shifter is formed by adding SOG. .

FIG. 9 shows a second embodiment of the present invention. In the fourth embodiment shown in FIG. 8, the shifter 68 is formed on the surface of the quartz glass 60 opposite to the surface on which the chrome pattern 62 is formed. ing. Also in this case, even if the etching area for the shifter 68 is enlarged behind the chrome pattern 62, the exposure is blocked by the chrome pattern 62,
There is an advantage that the accuracy of the etching width can be reduced. Furthermore, since there is no possibility of side etching of the chrome pattern 62, there is an effect that a higher resolution photomask can be realized.

FIG. 10 shows a fifth embodiment in which the present invention is applied to a so-called chromeless photomask. As shown in FIG. 10 (A), there is no chromium pattern on the quartz glass 70, and one surface thereof is etched at a depth D to form a light transmitting portion 76a in which a shifter 74 is formed. And a light transmitting portion 76b having a thickness.

The phases of the transmitted lights of the respective parts are different by 180 degrees. Since the light transmitting portion 76a has the phase shifter 74, an optical path difference can be provided to the other transmitting portion 76b, and the optical phase can be shifted by 180 degrees (FIG. 1).
0 (B)). In this case, the intensity of the light transmitted near the boundary between the light transmitting portion 76a and the light transmitting portion 76b is reduced to zero because the amplitude is canceled at the overlapping portion of the two transmitted lights having different phases, thereby forming a desired exposure pattern. Yes (Fig. 10 (C)
reference). Moreover, since the shifter 74 is formed by etching the quartz glass 70 itself without using a chromium pattern, the manufacturing process can be simplified.

[0051]

As described above, according to the photomask of the present invention, it is not necessary to spin-coat SOG, so that the mask manufacturing process is reduced by that amount, and the photomask depends on the film thickness accuracy of spin coating. Compared with the prior art, the optical path difference for inverting the phase can depend on processing control of the thickness of the flat mask substrate itself. Since this processing control is achieved by an etching technique in which processing accuracy higher than film thickness accuracy by spin coating is established, it is possible to provide a photomask useful for an exposure process of a semiconductor device in which miniaturization is advanced. In addition, according to the method of manufacturing a semiconductor device according to the present invention, exposure with high resolution can be realized, and the yield of a semiconductor device with further miniaturization can be increased.

[Brief description of the drawings]

FIGS. 1A to 1C are a cross-sectional view of a spatial modulation type photomask to which a first reference example is applied and a diagram showing a state of transmitted light.

2 (A) to 2 (D) are views showing a manufacturing process of the photomask shown in FIG. 1;

FIG. 3 is a perspective view for explaining an exposure step using a photomask of a first reference example.

FIGS. 4A to 4D are diagrams illustrating a lithography process using a photomask of a first reference example; FIGS.

FIG. 5 is a cross-sectional view of a photomask of a second reference example in which the shifter of FIG. 1 is configured by etching the back side of a mask substrate.

FIGS. 6A to 6C are a cross-sectional view of a photomask of a third reference example of an edge emphasis type to which the second reference example is applied and a diagram showing a state of transmitted light.

FIG. 7 is a cross-sectional view of the photomask of the first embodiment of the present invention in which the shifter of FIG. 6 is formed by etching the back side of a mask substrate.

FIGS. 8A to 8C are cross-sectional views of an auxiliary pattern type photomask and a diagram showing a state of transmitted light, which is a fourth reference example.

FIG. 9 is a cross-sectional view of a photomask according to a second embodiment of the present invention in which the shifter of FIG. 8 is formed by etching the back side of a mask substrate.

FIGS. 10A to 10C are a cross-sectional view of a so-called chromeless photomask to which the present invention is applied and a diagram showing a state of transmitted light.

FIGS. 11A to 11C are views showing a cross section and a state of transmitted light of a conventional photomask not using a phase shift method.

FIGS. 12A to 12C are views showing a cross section and a state of transmitted light of a conventional photomask using a phase shift method.

13 (A) to 13 (E) are views showing steps of manufacturing a conventional photomask using a phase shift method.

[Explanation of symbols]

 20, 40 Mask substrate 22, 42 Chromium pattern 24, 44 Phase shifter 26a, 26b, 46a, 46b Light transmission portion 30 Semiconductor substrate 34 Etched layer 36 Resist 50 Mask substrate 52 Chrome pattern 56 Light transmission portion 58 Phase shifter 60 Mask substrate 62 chrome pattern 64 auxiliary light transmitting part 66 main light transmitting part 68 phase shifter 70 mask substrate 74 phase shifter 76a, 76b light transmitting part 100 mercury lamp 104 photomask 106 reduction projection lens 110 wafer 120 wafer stage

Claims (6)

(57) [Claims] 1. A light source comprising: a mask substrate made of a material transmitting an exposure light source light; a light shielding layer formed on the mask substrate so as to be spaced apart from the mask substrate; and a light comprising the mask substrate sandwiched between the light shielding layers. A mask pattern including a transmissive region; and a concave portion formed at both ends of the light transmissive region such that the thickness of the mask substrate is smaller than that of an intermediate region sandwiched between the both end portions. And a phase shifter, wherein the concave portion is provided on a surface of the mask substrate opposite to a surface on which a light shielding layer is formed. A mask substrate made of a material transmitting the light source for exposure, a light shielding layer formed on the mask substrate so as to be spaced apart from the mask substrate, and the mask substrate sandwiched between the light shielding layers. A light-transmitting region, a mask pattern including an auxiliary light-transmitting region formed of the mask substrate formed by providing the light-shielding layer with a non-formed portion of the light-shielding layer; A phase shifter formed by providing a concave portion, wherein the concave portion is provided on a surface of the mask substrate opposite to a surface on which the light shielding layer is formed. . 3. The depth D of the concave portion according to claim 1, wherein the wavelength of the light source light for exposure is λ, the refractive index of the mask substrate is n1, and the refractive index of the atmosphere around the mask is n2. Wherein D = (100 ± 5.56)% × λ / [2 × (n1-n2)]. 4. A method of manufacturing a semiconductor device, comprising using the photomask according to claim 1 in an exposure step of forming a desired pattern on a resist formed on a layer to be etched. . 5. The method for manufacturing a semiconductor device according to claim 4, wherein an i-line of a mercury lamp is used as exposure light source light used in the exposure step. 6. The method for manufacturing a semiconductor device according to claim 4, wherein an excimer laser is used as light source light for exposure used in the exposure step.