JP2003021709A - Binary optics element - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for compensating a phase deviation of a light passing through regions with mutually different phase level numbers in a binary optics element with the mixed phase level numbers. SOLUTION: The widths and depths of steps are optimized in relation to an ideal shape. Especially in manufacturing the steps ranging from the eighth to the sixteenth steps, the width of the sixteenth step is shifted by 30% and its depth is shifted by -10% and the width of the eighth step is shifted by -30% so as to make the phase deviation being 0.03 λ before compensation, nearly zero.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a diffractive optical element used in a projection optical system such as a semiconductor manufacturing apparatus, that is, a binary optics element. A projection optical system using this diffractive optical element is used to project and expose a device pattern on a reticle or mask (generally referred to as a “mask” hereinafter) surface to a plurality of wafer-shaped portions by a step-and-repeat method or a step-and-scan method. , IC, LSI, CCD, liquid crystal panel and the like, which are suitable for manufacturing a device having a pattern of sub-micron or quarter-micron or less.

[0002]

2. Description of the Related Art Conventionally, most of conventional optical systems including a projection optical system of a semiconductor manufacturing apparatus have been constructed using only refractive optical elements, but recently, diffractive optical elements (Diffrac).
A large number of optical systems have been proposed, which make use of active optical elements (hereinafter also referred to as “DOE”). The diffractive optical element includes an amplitude type diffractive optical element and a phase type diffractive optical element known as a Fresnel zone plate. However, in the amplitude type, a part of light is blocked by the element, which is not preferable in terms of light utilization efficiency. on the other hand,
It is known that the phase type diffractive optical element has a diffraction efficiency of 100% when it is manufactured so as to ideally give a phase change. In particular, what is called a surface relief type is frequently used in ordinary optical systems. This is to give a light that passes through a phase change according to the position of the element surface by creating a structure in the depth direction of the element substrate, but the required depth is usually in the order of wavelength, and the thickness of the element Since various phase changes can be given by making the thickness thin and changing the position where the structure is made, the ordinary refracting optical element has a feature that the effect of making the surface aspherical can be widely realized. A function that describes the phase change given to light according to the position on the element surface is called a phase function.

A further feature of diffractive optical elements is that chromatic dispersion appears opposite to that of refractive optical elements. By using this feature, it is possible to correct chromatic aberration that occurs in the refractive optical element.

These features are suitable for a projection optical system used in a semiconductor manufacturing apparatus. The wavelength of the light used in these optical systems is the conventional Hg i-line (λ = 365 nm)
Met. Since there are a plurality of glass materials having a sufficient transmittance at this wavelength, it was possible to correct the chromatic aberration only by combining the refractive optical elements. On the other hand, SiO ₂ and CaF ₂ are the only practical glass materials having sufficient transmittance in the ultraviolet region such as the current mainstream KrF excimer laser (λ = 248 nm) and ArF excimer laser (λ = 193 nm). ₂
The laser (λ = 157 nm) has only CaF ₂ . Although the bandwidth of the laser light source is narrow, the projection optical system of the semiconductor manufacturing apparatus has extremely high required imaging performance, and chromatic aberration becomes a problem if the glass material used is limited in the optical system with only the refractive optical element. . Therefore, the bandwidth of the light source is 1p
Severe conditions such as m or less are imposed, and various devices are required to narrow the band. Also, the number of lenses required to sufficiently reduce the wavefront aberration of the optical system increases, but this leads to an increase in the total lens thickness and the surface where an antireflection film is required, and the transmittance of the entire optical system deteriorates. .
Since this also increases the absorption of the entire lens system,
It is not preferable from the viewpoint of exposure aberration. The use of the diffractive optical element is one of means for solving the problems of chromatic aberration correction, increase in the total lens thickness and lens surface, and advanced aberration correction.

Although the above-mentioned advantages of the diffractive optical element have been known for a long time, in recent years, many optical systems including them (for example, Japanese Patent Laid-Open No. 6-331941) have been proposed. This is after the idea of a binary optics element (hereinafter also referred to as “BOE”) appears.
For binary optics elements, see G.W. J. Sw
anson, Technical Report 8
54, MIT Lincoln Laborator
y, 14 Aug 1989G. J. Swanso
n, Technical Report 914, M
IT Lincoln Laboratory, 1
Details on Mar 1991 etc.

Conventionally, it is extremely difficult to directly produce an ideal shape (blazed shape) required for a diffractive optical element, that is, a shape required to correctly express a phase function, due to a problem of working accuracy and the like. there were. However, in the binary optics element, the blazed shape is not directly manufactured, but the blazed shape is approximated by the staircase shape. It is largely possible that such a staircase shape can be easily manufactured by a lithography process and a fine structure can be accurately manufactured by using a stepper as an exposure apparatus.

Hereinafter, a binary optics element will be described using an ideal lens that collects parallel light at one point. To collect parallel light (plane wave) incident on the lens at one point, use as a phase function

[0008]

[Equation 1] Should be given. However, f is the focal length, λ is the wavelength of the light used, and r is the distance from an arbitrary origin.

In a diffractive optical element, light is 2π in terms of phase.
It has the cycle of. First, r = Rm (m is an integer: R0 = 0 and sequentially counted from the origin toward the outside) where the value of the phase function φ (r) is an integral multiple of 2π is calculated, and the interval [Rm, Rm + 1 ], The value of φ (r) is [0,2π]
Phase function φ'added by an integer multiple of 2π
Make (r). In order to realize this phase function φ ′ (r), for example, the element whose surface is given a shape becomes an ideal lens by the diffractive optical element. FIG. 1 is a diagram schematically showing the surface shape at that time. Here, the annular zone interval T _m is T _m
= R _{m + 1} −R _m . The annular zone spacing is relatively large at the center (r to 0), and due to the difference in m, the difference in the annular zone spacing is large. On the other hand, in the peripheral part, the annular zone spacing is almost constant even if the value of m is different, so it can be regarded as an equidistant grid. In such a diffractive optical element, it can be generally considered that the grating is an equidistant grating except for a few annular zones at the center.

Reference numeral 101 shows the surface shape at the central portion, which has a blazed shape that completely describes the phase function. At this time, 101 is a part of the curved surface. On the other hand, reference numeral 102 indicates a blazed shape in the peripheral portion, but it can be approximately regarded as a plane.

FIG. 2 is a schematic view of the surface shape when the ideal lens is manufactured as a binary optics element, and is a shape in which the blazed shape shown in FIG. 1 is approximated by a step shape. At this time, the steps of the stairs are determined so that the phases are sampled at equal intervals. That is, if the depth in the case of the blazed shape is D, the number of phase levels used for approximation, and the number of steps of stairs in the same meaning is N, one step is D / N.
Becomes In this way, since the step is constant, 1
When the curved surface is approximated as in 03, the widths of the steps are unequal, while in 104 which approximates the plane, the steps are of equal width.

Diffractive optics as binary optics element
If an element is formed, chromatic aberration correction function or aspherical surface
Although it has an effect, it has a diffraction efficiency because it has a similar shape.
Does not reach 100%, and unnecessary diffracted light appears. Phase level
The number is N, and the diffraction order is set to satisfy the imaging condition.
When the (design order) is 1, the diffraction efficiency η of the diffraction order m
^N _mIs

[0013]

[Equation 2] Can be expressed as However, the depth of the element is the wavelength used λ
Must be optimized for. In that case, the step height d is as if the element made of a glass material having a refractive index n is placed in air (refractive index 1.0).

[0014]

[Equation 3] Is. this

[Expression 3] is an expression when light passes through the surface relief structure, and in the case of reflection, d = λ / (2 * nr * N) where nr is the refractive index of the medium on the light incident side. Become.

Usually, the bandwidth of the wavelength used in the projection optical system of the semiconductor manufacturing apparatus is about 1 pm. This is the bandwidth required because it is difficult to correct the chromatic aberration, but even if the diffractive optical element enables the correction of the chromatic aberration within an appropriate range, the bandwidth is at most if the light source used is a laser. It is 1 nm.

The depth needs to be optimized with respect to the wavelength in order to satisfy the equation (2), but the wavelength dependence of the diffraction efficiency can be almost ignored in the narrow bandwidth as described above.

[0016]

In [Equation 2], when N → ∞, η∞ ₁ = 1. Therefore, it can be seen that the diffraction efficiency is 100% in the ideal case. In general, N = 8 is often used. In that case, the diffraction efficiency is η ⁸ ₁ = 0.95, and the remaining 0.05 is directed to another diffraction order and becomes unnecessary diffracted light. Depending on the application, this 5% of unwanted diffracted light can be a problem.

The diffraction efficiency can be increased by increasing the value of N, but the size of N is limited by the manufacturing accuracy. For example, when a normal i-line stepper is used, the line width that can be manufactured is about 0.35 μm, but if a 16-step staircase shape is manufactured based on this, the annular zone spacing is 5.6 μm. In the case of N = 16, η ¹⁶ ₁ = 0.99, which can reduce unnecessary diffracted light, but the annular zone spacing is 5.6 μm.
The value of m is not enough to completely correct chromatic aberration in a projection optical system using a KrF excimer laser or the like as a light source, and a smaller ring zone interval is required. Therefore, when it is necessary to improve the diffraction efficiency, it is effective to form the phase numbers in a mixed manner on one binary optics element. Such technology is already known,
For example, SPIE volume 2577 pp. 11
4-122, A. Kathman, D.M. Hochmu
th, and D.L. Introduced on Brown.

[0018]

However, when the number of phase levels is mixed, there is a problem that light from regions having different phase levels causes a phase shift. FIG. 3 is a schematic diagram for qualitatively explaining the reason why the phase shift occurs, and shows the boundary between the phase level numbers N = 8 and N = 16. When the binary optics element is produced by a lithographic method, the element surface 52 may be considered to be in the same plane even if the number of phase levels is different, unless a special production process is performed. On the other hand, the height d

It can be expressed by the following equation, but the total depth D is D =
Since it is (N-1) d, there is a difference 51 in the total depth when the phase level is different. Therefore, when the light ray passes through the region of N = 8, the phase due to the surface relief structure and the depth difference of 51 minutes are added, and when passing through the region of N = 16, only the phase due to the surface relief structure is added. Is added.
Since these do not generally match, light from regions having different phase levels has a phase shift. According to the exact calculation, the grating pitch 53 is 10 μm and the wavelength λ2 is
48 nm, refractive index n1.51 of medium, width of each step is equal,
The depth of each step

If the light rays are incident vertically from the medium side, the light transmitted through the N = 8 region has a phase delay of about 0.03λ compared with the light transmitted through the N = 16 region. Occurs. This phase shift of the order of λ / 100 is a non-negligible amount when a very high degree of aberration correction is required, such as in a projection optical system of a semiconductor manufacturing apparatus.

In order to perform such phase correction, the method used in the phase shift diffractive optical element can be applied. For example, in Japanese Patent Laid-Open No. 5-150108, a method is adopted in which the thickness of the element in such a region is changed by the amount of phase shift. In addition, JP-A-5-157904 and JP-A-7-3
In 5912, a method of shifting the lattice is adopted.

However, the former method requires a different process for changing the thickness, and the latter method requires a different photo to produce a binary optics element in which the number of stages need not be mixed. There is a problem that a mask is needed.

An object of the present invention is to correct a phase shift generated in light transmitted or reflected in a region having a different number of phase levels in a binary optics element in which a plurality of number of phase levels are mixed for improving diffraction efficiency. In particular, the present invention is to provide a means necessary for applying the binary optics element to an optical system that requires extremely high imaging performance such as a projection optical system of a semiconductor manufacturing apparatus. .

[0022]

In order to achieve the above object, the binary optics element of the present invention is a binary optics element in which at least two phase level numbers are mixed, and the first phase level number of the binary optics element is the same. The width of each step in the region having a is determined so that the phase shift between the light transmitted or reflected in the region having the first phase level and the light transmitted or reflected in the region having the second phase level is reduced. It is characterized by being. Also,
As another method, in a binary optics element in which at least two phase level numbers are mixed, the depth of each step in the area having the first phase level number of the binary optics element is a region having the first phase level. Is determined so as to reduce the phase shift between the light transmitted or reflected through and the light transmitted or reflected through the region having the second phase level. Furthermore, in order to achieve the effect of reducing the phase shift, by combining the above two methods, in a binary optics element in which at least two phase level numbers are mixed, each of the binary optics elements in the region having the first phase level number is The width and depth of the step are determined so that the phase shift between the light transmitted or reflected in the region having the first phase level and the light transmitted or reflected in the region having the second phase level is reduced. Is characterized by.

In order to correct the phase shift of light from the areas having different phase level numbers, it is assumed that the first phase level number is larger than the second phase level number and the area having the first phase level number. The width of at least one step in
It is characterized in that it is larger than when the width of each step is made equal.

Alternatively, assuming that the first number of phase levels is larger than the second number of phase levels, the depth of at least one step in the region having the first number of phase levels is equal to the depth of each step. It is smaller than that.

Furthermore, in the region having the second number of phase levels, the width of at least one step is smaller than that when the width of each step is equal.

When the first phase level number is 16, the second phase level number is 8, and each step in the region having the first phase level number has the same width, the width is EW ₁ , When the width of each step in the area having the number of phase levels is equal to EW ₂ , the width of at least one step of the area having the first number of phase levels is 1.1 EW ₁ or more and 1.4 EW ₁ It is preferable that the width of at least one step of the region having the second phase level number is 0.6 EW ₂ or more and 0.9 EW ₂ or less.

Similarly, the depth of one step in the region having the first number of phase levels is represented by the representative wavelength λ of the wavelength range of the light used.
On the other hand, it is preferable that the ideal depth of one step is 0.85d or more and 0.95d or less.

In addition, when applying the above method, the depth or width of the step of at least the region having the first phase level number differs depending on the incident direction of the light incident on the binary optics element. It is more preferable if it is characterized.

[0029]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described in detail below with reference to the drawings.

(Embodiment 1) A case where N = 8 and N = 16 stages coexist will be described. When a 2 ^k step shape is formed by lithography, the above-mentioned Swanson
Etc., it can be manufactured using k masks. Further, in order to form a step shape of 2 ^k-1 steps to 2 ^k steps, another lithography process may be performed on the 2 ^k-1 step steps.

FIG. 4 shows a method of manufacturing a staircase shape having 8 to 16 steps. In FIG. 4A, a staircase structure 6 having 8 steps
A resist 61 is attached on each stage of No. 2. On the other hand, FIG. 4B shows a state in which the mask 63 is used to perform exposure with the exposure light 69 and development is performed, and the resist 63 remains on the 8-step structure 62. Thereafter, etching is performed to a preset depth as shown in FIG. 4C, where a 16-step structure 64 is obtained, and the resist 63 is peeled off as shown in FIG. 4D. .

At this time, the phase can be corrected by controlling the line width for etching. FIG. 5 shows a 16-step shape when the line width is narrowed when producing 8 to 16 steps. If the line width in the 16-step fabrication is narrowed, the actual width 124 becomes larger than the width 123 in the ideal shape fabrication. Hereinafter, line width (%) = (ideal shape width 123−actual width 124) / ideal shape width 123 will be used as an index. The ideal 16 stages have a phase shift of + 0.03λ compared to the ideal 8 stages, but for example, as shown in FIG. 5, the line width when manufacturing 16 stages from 8 stages is narrowed by about 30%. If 16 stages are formed by 122, it is possible to correct the phase shift of about 0.01λ from FIG. 6 and reduce it to + 0.02λ. In order to control the line width as described above, it is advisable to consider the line width of the mask 60 used for forming the 8 to 16 steps of FIG.

(Embodiment 2) Also, the phase can be corrected by controlling the etching depth. FIG. 7 shows a 16-step shape in the case where the etching depth is reduced when manufacturing 8 to 16 steps. Depth of ideal shape 12
5, using the actual depth 126, the depth (%) =
(Actual depth 126-ideal depth 125) / ideal depth 125 will be used as an index. For example, as shown in FIG. 7, the etching depth when fabricating 8 to 16 steps is shallower than the ideal shape 120 by about 10% 12
By using 1, it is possible to correct a phase shift of about 0.003λ. Such depth control can be easily performed during etching.

Needless to say, the phase can be corrected by controlling the line width and the depth at the same time.
Shows the amount of phase correction that can be achieved when both are controlled simultaneously.

When the staircase shape is changed, there is a concern that the first-order diffraction efficiency may decrease, but as shown in FIG. 10, the line width and the depth required to correct the phase shift are -30% and -10%, respectively. Even so, the reduction amount is less than 0.5%, and does not significantly affect the performance of the device.

(Embodiment 3) A correction of 0.012λ is given in the correction only when producing 16 steps, and the necessary correction amount is 0.03.
Only about half of λ can be achieved, but the required correction amount can be satisfied by changing the 8-step staircase structure. FIG. 11 is a diagram showing changes to be made to the 8-step staircase structure necessary for correcting the phase shift. Considering a structure 131 in which the step width is corrected with respect to the eight-step ideal shape 130, the line width (%) = (ideal shape width 132−actual width 1
33) / The width 132 of the ideal shape is used as an index. At this time,
If the line width (%) is + 20%, + 30%, + 40%,
It is possible to correct 0.013λ, 0.018λ, and 0.025λ, respectively. Therefore, if the line width (%) of + 40% is applied to the 8 steps, the phase shift between the 16 steps and the 8 steps can be made almost zero independently, but the decrease amount of the 1st-order diffraction efficiency is 2.2%. Therefore, it is preferable to set the line width (%) to + 30% or less and to control the width and depth when forming 16 steps.

Although the phase shift amount to be corrected is changed by changing the width of the eight steps here, it is obvious that the same control can be performed by changing the step depth.

Up to this point, using k masks, 2
^Although the case of forming a ^k- step staircase shape has been described, it is of course applicable to the case of producing k steps with k masks and the production method of Japanese Patent Laid-Open No. 11-160510, and is limited to 2 ^k steps. It is clear that there is no.

Further, since such a control of the phase has only a problem with the step shape, it can be applied not only to etching but also to other manufacturing methods such as depositing.

Next, 16 phase levels and 8 phase levels
As described with reference to FIG. 5, the phase shift of 1 occurs due to a phase mismatch provided by the surface relief structure and a phase difference of 51 minutes in depth. Here, consider a case where the incident angles are the same and only the incident directions are different. In this case, the phases provided by the surface relief structure are equal. On the other hand, there is a difference in the phase of the depth difference 51 minutes. This is the second medium 5
When vertically incident from 5, the depth difference 51 passes through at the shortest distance, but when vertically incident from the first medium 54, it is diffracted by the surface relief structure and obliquely passes through the depth difference 51. Therefore, the optical distance is different. The angle diffracted by the surface relief structure is the grating pitch 53.
The diffraction angle becomes larger as the grating pitch 53 becomes smaller, so that a phase shift may occur due to the difference in the incident direction. In order to make this correction, the width and depth of each step may be individually determined for each incident direction based on the data shown in FIG.

By performing such phase correction, it is possible to apply the present invention to an optical system that requires high precision even if the number of stages is different in each region in the binary optics element.

[0042]

As described above, according to the present invention, in the binary optics element in which the number of phase levels is mixed, the depth of each step is controlled according to the number of phase levels to correct the phase, There is an effect that it can be applied to an accurate projection optical system.

[Brief description of drawings]

FIG. 1 is a schematic diagram of a blazed shape.

FIG. 2 is a schematic diagram of a staircase shape that approximates a blazed shape.

FIG. 3 is a schematic diagram showing a boundary between the number of phase levels 8 and the number of phase levels 16

FIG. 4 is a diagram showing a method of manufacturing a stair with 16 phase levels from a stair shape with 8 phase levels through a lithography process.

FIG. 5 is a schematic diagram showing the surface shape of a binary optics element in which the width of the step is controlled to correct the phase, which is an embodiment of the present invention.

FIG. 6 is a diagram showing the amount of phase change when the step width is controlled.

FIG. 7 is a schematic diagram showing the surface shape of a binary optics element in which the depth of the step is controlled and the phase is corrected, which is an embodiment of the present invention.

FIG. 8 is a diagram showing the amount of phase change when the depth of a step is controlled.

FIG. 9 is a diagram showing the amount of phase change when the width and depth of a step are controlled simultaneously.

FIG. 10 is a diagram showing changes in first-order diffraction efficiency when the width and depth of a step are simultaneously controlled.

FIG. 11 is a schematic diagram showing the surface shape of a binary optics element in which the amount of phase shift to be corrected is changed by controlling the step width of a region having a second phase level number, which is an embodiment of the present invention.

[Explanation of symbols]

51 Difference between depth of surface relief structure having ideal shape of 8 phase levels and depth of surface relief structure having ideal shape of 16 phase levels 52 Surface of binary optics element 53 Lattice pitch (ring spacing) 54 First Medium (usually air) 55 second medium (material for binary optics element) 60 mask 61 for forming a staircase shape having 8 to 16 phase levels 61 resist 62 base 63 resist 64 16 left after exposure Stepped structure 69 Exposure light 101 Blazed shape of diffractive optical element (center part) 102 Blazed shape of diffractive optical element (peripheral part) 103 Surface relief shape of binary optics element (center part) 104 Surface relief shape of binary optics element (peripheral part) Part) 120 ideal 16 step structure 121 etching depth It was shallow 16-stage step structure 122 stage width wider (and narrow exposure line width) 16
Stepped structure 123 Ideal stepped width 124 Actual stepped width 125 Ideal stepped depth 126 Actual stepped depth 130 Ideal 8-stepped stepped structure 131 Stepped width is narrowed (exposure line width 8 steps structure 132 ideal step width 133 actual step width

Claims (9)

[Claims] 1. In a binary optics element in which at least two phase level numbers are mixed, the width of each step in the area of the binary optics element having the first phase level number is such that the area having the first phase level is A binary optics element, characterized in that it is determined to reduce the phase shift between the light transmitted or reflected and the light transmitted or reflected through a region having a second phase level. 2. In a binary optics element in which at least two phase level numbers are mixed, the depth of each step in the area having the first phase level number of the binary optics element is a region having the first phase level. A binary optics element characterized in that it is determined to reduce the phase shift between the light transmitted or reflected through and the light transmitted or reflected through the region having the second phase level. 3. In a binary optics element in which at least two phase level numbers are mixed, the width and depth of each step in the region having the first phase level number of the binary optics element is equal to the first phase level. A binary optics element, characterized in that it is determined to reduce a phase shift between light transmitted or reflected through a region having the same and light transmitted or reflected through a region having a second phase level. 4. The binary optics element according to claim 1, wherein the first phase level number is greater than the second phase level number, and at least one step is provided in the region having the first phase level number. 4. The binary optics element according to claim 1, wherein the width is larger than that when the width of each step is equal. 5. The binary optics element according to claim 1, wherein at least one step is provided in a region having the first phase level number, where the first phase level number is larger than the second phase level number. The binary optics element is characterized in that the depth is smaller than when the depths of the respective steps are made equal. 6. The binary optics element according to claim 4, wherein in the region having the second number of phase levels, the width of at least one step is smaller than the case where the width of each step is equal. A binary optics element characterized by: 7. The binary optics element according to claim 6, wherein the first phase level number is 16, the second phase level number is 8,
The width when each step in the area having the first phase level number is EW ₁ , and the width when each step in the area having the second phase level number is EW ₂ . Then, the width of at least one step of the region having the first phase level number is 1.1E.
W ₁ or more and 1.4 EW ₁ or less, and the width of at least one step of the region having the second phase level number is 0.6 EW ₂ or more and 0.
A binary optics element characterized by being 9 EW ₂ or less. 8. The binary optics element according to claim 7, wherein the depth of one step in the region having the first number of phase levels is:
A binary optics element having a depth of 0.85d or more and 0.95d or less, where d is an ideal one-step depth with respect to a representative wavelength λ of a wavelength range of light used. 9. The binary optics element according to claim 1, wherein depending on an incident direction of light incident on the binary optics element, a depth of a step of a region having at least the first phase level number is set. A binary optics element characterized by different widths.