JP7574766B2 - Glass substrates for EUVL and mask blanks for EUVL - Google Patents

Description

This disclosure relates to glass substrates for EUVL (Extreme Ultra-Violet Lithography) and mask blanks for EUVL.

Traditionally, photolithography technology has been used to manufacture semiconductor devices. In photolithography technology, an exposure device irradiates light onto a circuit pattern on a photomask, and then reduces and transfers the circuit pattern onto a resist film.

Recently, the use of short-wavelength exposure light, such as ArF excimer laser light or even EUV (Extreme Ultra-Violet) light, has been considered to enable the transfer of fine circuit patterns.

Here, EUV (extreme ultraviolet) includes soft X-rays and vacuum ultraviolet light, and specifically refers to light with a wavelength of about 0.2 nm to 100 nm. At present, EUV with a wavelength of about 13.5 nm is mainly being considered.

EUVL photomasks are obtained by forming a circuit pattern on an EUVL mask blank.

A mask blank for EUVL has a glass substrate, a conductive film formed on a first main surface of the glass substrate, and an EUV reflective film and an EUV absorbing film formed on a second main surface of the glass substrate. The EUV reflective film and the EUV absorbing film are formed in this order.

The EUV reflective film reflects EUV. The EUV absorbing film absorbs EUV. An opening pattern, which is a circuit pattern, is formed in the EUV absorbing film. The conductive film is attracted to the electrostatic chuck of the exposure tool.

Mask blanks for EUVL require high flatness to improve the transfer accuracy of the circuit pattern. The flatness is determined mainly by the flatness of the glass substrate for EUVL. Therefore, glass substrates for EUVL also require high flatness.

The mask blank for EUVL described in Patent Document 1 has a central region and an outer peripheral region on the main surface of the conductive film opposite the glass substrate. The central region is a square region 142 mm long and 142 mm wide, excluding the outer peripheral region in the shape of a rectangular frame that surrounds the central region. In the central region, the flatness of the component having an order of Legendre polynomial of 3 or more and 25 or less is 20 nm or less.

In addition, the mask blank for EUVL described in Patent Document 2 has a difference between the maximum height and the minimum height within a calculation area of the difference data between the composite surface shape and the virtual surface shape of 25 nm or less. The calculation area is an area inside a circle with a diameter of 104 mm. The composite surface shape is obtained by combining the surface shape of the multilayer reflective film and the surface shape of the conductive film. The virtual surface shape is defined by Zernike polynomials expressed in a polar coordinate system.

Patent No. 6229807 Patent No. 6033987

As mentioned above, high flatness is required for EUVL glass substrates. Therefore, the central region of the main surface of the glass substrate for EUVL is generally subjected to polishing, local processing, and finish polishing in this order. The local processing method is, for example, the GCIB (Gas Cluster Ion Beam) method or the PCVM (Plasma Chemical Vaporization Machining) method.

In the finish polishing, the glass substrate for EUVL is pressed against the platen while the glass substrate for EUVL and the platen are rotated. The central region of the main surface of the glass substrate for EUVL is finish polished approximately axially symmetrically about its center, but is not completely axially symmetrical, and after the finish polishing, it contains both axially symmetric components and the remaining distortion components.

The distortion components include four-fold symmetric components rotated around the center of the central region. These four-fold symmetric components are caused by the finish polishing. These four-fold symmetric components are preferably expressed by Zernike polynomials rather than Legendre polynomials. This is because, unlike Legendre polynomials, Zernike polynomials are expressed in polar coordinates and are suitable for removing axially symmetric components.

However, unlike Legendre polynomials, Zernike polynomials can only express circular regions. The main surface of a glass substrate for EUVL is rectangular, and its central region is also rectangular, so the four corners of the rectangle cannot be expressed by Zernike polynomials. Therefore, in the past, it was not possible to accurately grasp the distortion components that occur during finish polishing.

As a result, it has traditionally been difficult to suppress the flatness of the central region of the main surface of a glass substrate for EUVL to less than 10.0 nm.

One aspect of the present disclosure provides a technology that suppresses the flatness of the central region of the main surface of a glass substrate for EUVL to less than 10.0 nm.

A glass substrate for EUVL according to one embodiment of the present disclosure has a rectangular first main surface on which a conductive film is formed, and a rectangular second main surface facing the opposite direction to the first main surface on which an EUV reflective film and an EUV absorbing film are formed in that order. When the coordinates of a point in a central region of a square 142 mm long and 142 mm wide excluding the peripheral region of the rectangular frame of the first main surface are expressed as (x, y, z(x, y)), the maximum height difference of the surface, which is a set of coordinates (x, y, z3(x, y)) calculated using the following formulas (1) to (3), is 6.0 nm or less.

上記座標（ｘ，ｙ，ｚ（ｘ，ｙ））において、ｘは横方向の座標、ｙは縦方向の座標、ｚは高さ方向の座標を示し、横方向、縦方向および高さ方向は互いに垂直である。

In the above coordinates (x, y, z(x, y)), x indicates the horizontal coordinate, y indicates the vertical coordinate, and z indicates the height coordinate, with the horizontal, vertical, and height directions being perpendicular to each other.

According to one aspect of the present disclosure, the flatness of the central region of the main surface of a glass substrate for EUVL can be suppressed to less than 10.0 nm.

FIG. 1 is a flowchart showing a method for manufacturing an EUVL mask blank according to one embodiment. FIG. 2 is a cross-sectional view showing a glass substrate for EUVL according to one embodiment. FIG. 3 is a plan view showing a glass substrate for EUVL according to one embodiment. FIG. 4 is a cross-sectional view showing an EUVL mask blank according to an embodiment. FIG. 5 is a cross-sectional view showing an example of a photomask for EUVL. FIG. 6 is a perspective view showing an example of a double-sided polishing machine, with a part of the double-sided polishing machine cut away. FIG. 7 is a diagram showing an example of the height distribution in the central region of the first main surface after finish polishing. FIG. 8 is a plan view showing an example of an arrangement of a plurality of points set in the central region. FIG. 9 is a diagram showing the height distribution of the component extracted from the height distribution of FIG. 7 using equation (1). FIG. 10 is a diagram showing the height distribution of the component extracted from the height distribution of FIG. 7 using equation (2). FIG. 11 is a diagram showing the height distribution of the component extracted from the height distribution of FIG. 7 using equation (3). FIG. 12 is a plan view showing the relative rotation direction of the surface plate with respect to the central region.

Below, the embodiments for carrying out the present disclosure will be described with reference to the drawings. In each drawing, the same or corresponding configurations are given the same reference numerals, and the description may be omitted. In the specification, "~" indicating a numerical range means that the numerical values before and after it are included as the lower and upper limits.

As shown in FIG. 1, the method for manufacturing an EUVL mask blank has steps S1 to S7. An EUVL mask blank 1 shown in FIG. 4 is manufactured using an EUVL glass substrate 2 shown in FIG. 2 and FIG. 3. Hereinafter, the EUVL mask blank 1 will also be referred to simply as mask blank 1. Also, the EUVL glass substrate 2 will also be referred to simply as glass substrate 2.

As shown in Figures 2 and 3, the glass substrate 2 includes a first main surface 21 and a second main surface 22 facing the opposite direction to the first main surface 21. The first main surface 21 is rectangular. In this specification, a rectangular shape includes a shape with chamfered corners. A rectangle also includes a square. The second main surface 22 faces the opposite direction to the first main surface 21. The second main surface 22 is also rectangular, like the first main surface 21.

The glass substrate 2 also includes four end faces 23, four first chamfered faces 24, and four second chamfered faces 25. The end faces 23 are perpendicular to the first main surface 21 and the second main surface 22. The first chamfered faces 24 are formed at the boundaries between the first main surface 21 and the end faces 23. The second chamfered faces 25 are formed at the boundaries between the second main surface 22 and the end faces 23. In this embodiment, the first chamfered faces 24 and the second chamfered faces 25 are so-called C-chamfered faces, but may be R-chamfered faces.

The glass of the glass substrate 2 is preferably quartz glass containing _TiO2 . Quartz glass has a smaller linear expansion coefficient and a smaller dimensional change due to temperature change than general soda-lime glass. Quartz glass may contain 80% to 95% by mass of _SiO2 and 4% to 17% by mass of _TiO2 . When the _TiO2 content is 4% to 17% by mass, the linear expansion coefficient is approximately zero near room temperature, and there is almost no dimensional change near room temperature. Quartz glass may contain a third component or impurity other than _SiO2 and _TiO2 .

In plan view, the size of the glass substrate 2 is, for example, 152 mm in length and 152 mm in width. The length and width may be 152 mm or more.

The glass substrate 2 has a central region 27 and a peripheral region 28 on the first main surface 21. The central region 27 is a square region 142 mm long and 142 mm wide, excluding the peripheral region 28, which is a rectangular frame surrounding the central region 27, and is the region that is processed to the desired flatness by steps S1 to S4. The four sides of the central region 27 are parallel to the four end faces 23. The center of the central region 27 coincides with the center of the first main surface 21.

Although not shown, the second main surface 22 of the glass substrate 2 also has a central region and a peripheral region, similar to the first main surface 21. The central region of the second main surface 22, like the central region of the first main surface 21, is a square region 142 mm long and 142 mm wide, and is the region that is processed to the desired flatness by steps S1 to S4 in FIG. 1.

First, in step S1, the first main surface 21 and the second main surface 22 of the glass substrate 2 are polished. In this embodiment, the first main surface 21 and the second main surface 22 are polished simultaneously by a double-sided polisher 9 described later, but they may be polished sequentially by a single-sided polisher (not shown). In step S1, the glass substrate 2 is polished while a polishing slurry is supplied between the polishing pad and the glass substrate 2.

The polishing pad may be, for example, a urethane-based polishing pad, a nonwoven fabric-based polishing pad, or a suede-based polishing pad. The polishing slurry contains an abrasive and a dispersion medium. The abrasive is, for example, cerium oxide particles. The dispersion medium is, for example, water or an organic solvent. The first main surface 21 and the second main surface 22 may be polished multiple times with abrasives of different materials or particle sizes.

The abrasive used in step S1 is not limited to cerium oxide particles. For example, the abrasive used in step S1 may be silicon oxide particles, aluminum oxide particles, zirconium oxide particles, titanium oxide particles, diamond particles, silicon carbide particles, etc.

Next, in step S2, the surface shapes of the first main surface 21 and the second main surface 22 of the glass substrate 2 are measured. To measure the surface shapes, for example, a non-contact measuring machine such as a laser interference type is used so as not to scratch the surface. The measuring machine measures the surface shapes of the central region 27 of the first main surface 21 and the central region of the second main surface 22.

Next, in step S3, the first main surface 21 and the second main surface 22 of the glass substrate 2 are locally processed in order to improve flatness, with reference to the measurement results of step S2. The first main surface 21 and the second main surface 22 are locally processed in order. The order in which they are processed may be either first or second, and is not particularly limited. The method of local processing is, for example, the GCIB method or the PCVM method. The method of local processing may also be a polishing method using a magnetic fluid, a polishing method using a rotating polishing tool, or the like.

Next, in step S4, the first main surface 21 and the second main surface 22 of the glass substrate 2 are subjected to finish polishing. In this embodiment, the first main surface 21 and the second main surface 22 are polished simultaneously by a double-sided polisher 9 described below, but they may be polished sequentially by a single-sided polisher (not shown). In step S4, the glass substrate 2 is polished while a polishing slurry is supplied between the polishing pad and the glass substrate 2. The polishing slurry contains an abrasive. The abrasive is, for example, colloidal silica particles.

Next, in step S5, a conductive film 5 as shown in FIG. 4 is formed in the central region 27 of the first main surface 21 of the glass substrate 2. The conductive film 5 is used to attach the EUVL photomask to the electrostatic chuck of the exposure device. The conductive film 5 is made of, for example, chromium nitride (CrN). The conductive film 5 is formed, for example, by sputtering.

Next, in step S6, an EUV reflective film 3 shown in FIG. 4 is formed in the central region of the second main surface 22 of the glass substrate 2. The EUV reflective film 3 reflects EUV. The EUV reflective film 3 may be, for example, a multilayer reflective film in which high-refractive index layers and low-refractive index layers are alternately stacked. The high-refractive index layers are made of, for example, silicon (Si), and the low-refractive index layers are made of, for example, molybdenum (Mo). The EUV reflective film 3 is formed by a sputtering method such as ion beam sputtering or magnetron sputtering.

Finally, in step S7, the EUV absorbing film 4 shown in FIG. 4 is formed on the EUV reflective film 3 formed in step S6. The EUV absorbing film 4 absorbs EUV. The EUV absorbing film 4 is formed of a single metal, alloy, nitride, oxide, oxynitride, or the like containing at least one element selected from tantalum (Ta), chromium (Cr), and palladium (Pd). The EUV absorbing film 4 is formed, for example, by sputtering.

Note that in this embodiment, steps S6 and S7 are performed after step S5, but they may also be performed before step S5.

The above steps S1 to S7 result in the mask blank 1 shown in Figure 4. The mask blank 1 has a first main surface 11 and a second main surface 12 facing opposite to the first main surface 11, and has, in this order from the first main surface 11 side to the second main surface 12 side, a conductive film 5, a glass substrate 2, an EUV reflective film 3, and an EUV absorbing film 4.

Although not shown, the mask blank 1 has a central region and a peripheral region on the first main surface 11, similar to the glass substrate 2. The central region is a square region 142 mm long and 142 mm wide, excluding the rectangular frame-shaped peripheral region surrounding the central region. The mask blank 1 also has a central region and a peripheral region on the second main surface 12, similar to the glass substrate 2. The central region is a square region 142 mm long and 142 mm wide, excluding the rectangular frame-shaped peripheral region surrounding the central region.

The mask blank 1 may include other films in addition to the conductive film 5, the glass substrate 2, the EUV reflective film 3, and the EUV absorbing film 4.

For example, the mask blank 1 may further include a low-reflection film. The low-reflection film is formed on the EUV absorbing film 4. Then, a circuit pattern 41 is formed on both the low-reflection film and the EUV absorbing film 4. The low-reflection film is used to inspect the circuit pattern 41, and has lower reflection characteristics with respect to the inspection light than the EUV absorbing film 4. The low-reflection film is formed of, for example, TaON or TaO. The low-reflection film is formed, for example, by sputtering.

The mask blank 1 may further include a protective film. The protective film is formed between the EUV reflective film 3 and the EUV absorbing film 4. The protective film protects the EUV reflective film 3 so that the EUV reflective film 3 is not etched when the EUV absorbing film 4 is etched to form a circuit pattern 41 in the EUV absorbing film 4. The protective film is made of, for example, Ru, Si, or _TiO2 . The protective film is formed by, for example, sputtering.

As shown in FIG. 5, the photomask for EUVL is obtained by forming a circuit pattern 41 in an EUV absorbing film 4. The circuit pattern 41 is an opening pattern, and is formed using photolithography and etching. Therefore, the resist film used to form the circuit pattern 41 may be included in the mask blank 1.

Incidentally, the mask blank 1 is required to have a high degree of flatness in order to improve the transfer accuracy of the circuit pattern 41. The flatness is mainly determined by the flatness of the glass substrate 2. Therefore, the glass substrate 2 is also required to have a high degree of flatness.

Then, as described above, the glass substrate 2 is subjected to polishing (step S1), local processing (step S3), and finish polishing (step S4) in this order. In the finish polishing, the glass substrate 2 is pressed against the platen while both the glass substrate 2 and the platen are rotated. In the finish polishing, for example, a double-sided polisher 9 shown in FIG. 6 is used.

The double-sided polisher 9 has a lower platen 91, an upper platen 92, a carrier 93, a sun gear 94, and an internal gear 95. The lower platen 91 is arranged horizontally, and a lower polishing pad 96 is attached to the upper surface of the lower platen 91. The upper platen 92 is arranged horizontally, and an upper polishing pad 97 is attached to the lower surface of the upper platen 92. The carrier 93 holds the glass substrate 2 horizontally between the lower platen 91 and the upper platen 92. Each carrier 93 holds one glass substrate 2 at a time, but may hold multiple substrates at a time. The carrier 93 is arranged radially outside the sun gear 94 and radially inside the internal gear 95. A plurality of carriers 93 are arranged at intervals around the sun gear 94. The sun gear 94 and the internal gear 95 are arranged concentrically and mesh with the outer peripheral gear 93a of the carrier 93.

The double-sided polisher 9 is, for example, a 4-way type, in which the lower surface plate 91, the upper surface plate 92, the sun gear 94, and the internal gear 95 rotate about the same vertical rotation center line. The lower surface plate 91 and the upper surface plate 92 rotate in opposite directions, and press the lower polishing pad 96 against the lower surface of the glass substrate 2, and press the upper polishing pad 97 against the upper surface of the glass substrate 2. At least one of the lower surface plate 91 and the upper surface plate 92 supplies polishing slurry to the glass substrate 2. The polishing slurry is supplied between the glass substrate 2 and the lower polishing pad 96 to polish the lower surface of the glass substrate 2. The polishing slurry is also supplied between the glass substrate 2 and the upper polishing pad 97 to polish the upper surface of the glass substrate 2.

For example, the lower base plate 91, the sun gear 94, and the internal gear 95 rotate in the same direction in a plan view. These rotation directions are opposite to the rotation direction of the upper base plate 92. The carrier 93 rotates on its axis while revolving. The revolution direction of the carrier 93 is the same as the rotation direction of the sun gear 94 and the internal gear 95. On the other hand, the rotation direction of the carrier 93 is determined by the product of the rotation speed and pitch circle diameter of the sun gear 94 and the product of the rotation speed and pitch circle diameter of the internal gear 95. If the product of the rotation speed and pitch circle diameter of the internal gear 95 is greater than the product of the rotation speed and pitch circle diameter of the sun gear 94, then the rotation direction of the carrier 93 and the revolution direction of the carrier 93 will be the same. On the other hand, if the product of the internal gear 95's rotation speed and pitch circle diameter is smaller than the product of the sun gear 94's rotation speed and pitch circle diameter, the direction of rotation of the carrier 93 and the direction of revolution of the carrier 93 will be opposite to each other.

The first main surface 21 and the second main surface 22 of the glass substrate 2 are finish-polished by the double-sided polisher 9 in an axially symmetric manner about their respective centers. The first main surface 21 and the second main surface 22 tend to be polished in a plane-symmetric manner based on the central plane of the glass substrate 2 in the plate thickness direction. The first main surface 21 and the second main surface 22 tend to be both polished to a convex curved surface, or both polished to a concave curved surface. Note that for the finish polishing, a single-sided polisher (not shown) may be used as described above.

Figure 7 shows an example of the height distribution of the central region 27 of the first main surface 21 after finish polishing. Here, the height distribution after tilt correction is shown. The central region 27 shown in Figure 7 is a convex curved surface in which the height at the center is higher than the heights at the four corners. The units of the values indicating the height in Figure 7 are nm, and the larger the value, the higher the height. Note that the height distribution of the central region of the second main surface 22 after finish polishing is the same as the height distribution in Figure 7, so it is not shown in the figure.

The height distribution shown in FIG. 7 was measured using an UltraFlat 200 Mask manufactured by Corning Tropel. Here, in order to eliminate the effects of gravity, the glass substrate 2 was held upright and supported so that neither the first main surface 21 nor the second main surface 22 of the glass substrate 2 was in contact with other members such as a stage, and the height distribution was measured.

As is clear from FIG. 7, the central region 27 of the first main surface 21 after the finish polishing is not completely axially symmetric, but contains axially symmetric components and the remaining distortion components. The distortion components, which will be described in detail later, contain four-fold symmetric components rotated around the center of the central region 27 as shown in FIG. 9. These four-fold symmetric components are generated by the finish polishing.

These four-fold symmetric components are preferably expressed in Zernike polynomials rather than Legendre polynomials. Unlike Legendre polynomials, Zernike polynomials are expressed in polar coordinates and are suitable for removing axially symmetric components.

However, unlike Legendre polynomials, Zernike polynomials can only express circular regions. The central region 27 is rectangular, and the four corners of the rectangle cannot be expressed by Zernike polynomials. Therefore, in the past, it was not possible to accurately grasp the distortion components that occur during finish polishing.

Therefore, in this embodiment, the coordinates of a point in the central region 27 of a square 142 mm long and 142 mm wide are expressed as (x, y, z(x, y)), and the distortion components are determined using the following equations (1) to (3).

上記座標（ｘ，ｙ，ｚ（ｘ，ｙ））において、ｘは横方向の座標、ｙは縦方向の座標、ｚは高さ方向の座標を示し、横方向、縦方向および高さ方向は互いに垂直である。

In the above coordinates (x, y, z(x, y)), x indicates the horizontal coordinate, y indicates the vertical coordinate, and z indicates the height coordinate, with the horizontal, vertical, and height directions being perpendicular to each other.

Figure 8 shows an example of the arrangement of multiple points set in the central region 27. In Figure 8, the X-axis direction is the horizontal direction, and the Y-axis direction is the vertical direction. The origin, which is the intersection of the X-axis and Y-axis, is the center of the central region 27.

As is clear from FIG. 8, z1(x, y) in formula (1) is the average value of the heights of four points that are four-fold symmetric around the origin. FIG. 9 shows the height distribution of a surface, which is a set of coordinates (x, y, z1(x, y)). The units of the values indicating height in FIG. 9 are nm, and the larger the value, the higher the height. In addition to the axially symmetric component, the height distribution shown in FIG. 9 further includes a four-fold symmetric component rotated around the origin. This four-fold symmetric component rotates, for example, counterclockwise, as shown by the dashed line in FIG. 9.

As is clear from Figure 8, z2(x, y) in formula (2) is the average value of the heights of eight points that are line-symmetrical about four reference lines L1 to L4 that pass through the origin. Reference line L1 is the X-axis, reference line L2 is the Y-axis, and reference lines L3 and L4 are diagonals of central region 27. Figure 10 shows the height distribution of the surface, which is a collection of coordinates (x, y, z2(x, y)). The units of the values indicating height in Figure 10 are nm, and the larger the value, the higher the height. The height distribution shown in Figure 10 contains only components that are approximately axially symmetric.

z3(x,y) in equation (3) is the difference between z1(x,y) in equation (1) and z2(x,y) in equation (2). Figure 11 shows the height distribution of a surface, which is a set of coordinates (x,y,z3(x,y)). The units of the values indicating height in Figure 11 are nm, and the larger the value, the higher the height. The height distribution shown in Figure 11 is the difference between the height distribution shown in Figure 9 and the height distribution shown in Figure 10, and mainly contains four-fold symmetric components rotated around the origin.

Next, with reference to FIG. 12, the reason why the height distribution shown in FIG. 11 occurs due to finish polishing will be described. The arrows shown in FIG. 12 indicate the relative rotation direction of the surface plate (e.g., the lower surface plate 91 or the upper surface plate 92) with respect to the central region 27. In other words, the arrows shown in FIG. 12 indicate the rotation direction of the surface plate in a coordinate system fixed to the central region 27.

At the four corners of the central region 27, polishing of the portion A1 on the upstream side in the rotation direction of the platen progresses easily, while polishing of the portion A2 on the downstream side in the rotation direction of the platen progresses less easily. As a result, it is believed that the height distribution shown in Figure 11 occurs due to the finish polishing.

The inventors have found through experiments and the like that if the maximum height difference Δz3 (Δz3 ≧ 0) of the surface, which is a set of coordinates (x, y, z3 (x, y)), is 6.0 nm or less, the flatness PV (PV ≧ 0) of the central region 27 can be suppressed to less than 10.0 nm.

In this disclosure, the flatness PV of the central region 27 refers to the maximum height difference of the remaining components after excluding the components expressed by a quadratic function from all components of the height distribution of the central region 27. The quadratic function is expressed by the following formula (4).

上記式（４）において、ａ、ｂ、ｃ、ｄ、ｅ、ｆは、ｚ_ｆｉｔ（ｘ，ｙ）とｚ（ｘ，ｙ）との差の二乗の和が最小になるように決められる定数であって、最小二乗法によって求められる定数である。

In the above formula (4), a, b, c, d, e, and f are constants determined so as to minimize the sum of the squares of the differences between z _fit (x, y) and z(x, y), and are constants found by the least squares method.

The components of the quadratic function are components that can be automatically corrected in the exposure device. Therefore, the components of the quadratic function do not affect the transfer accuracy of the circuit pattern 41. Therefore, when calculating the flatness PV of the central region 27, the components of the quadratic function are excluded from the total components of the height distribution of the central region 27.

In order to suppress Δz3 to 6.0 nm or less, the inventor first performed the processes of steps S1 to S4 on another glass substrate 2 in advance, and calculated the height difference z _dif (x, y) at each point in the central region 27 before and after finish polishing using the following formula (5). Then, he calculated z _{4 _ dif} (x, y) using the following formula (6).

上記式（５）において、ｚ_{ａｆｔｅｒ}（ｘ，ｙ）は仕上げ研磨後の座標（ｘ，ｙ）における高さであり、ｚ_{ｂｅｆｏｒｅ}（ｘ，ｙ）は局所加工後であって仕上げ研磨前の座標（ｘ，ｙ）における高さである。ｚ_{ａｆｔｅｒ}（ｘ，ｙ）とｚ_{ｂｅｆｏｒｅ}（ｘ，ｙ）との差分がｚ_ｄｉｆ（ｘ，ｙ）であるので、ｚ_ｄｉｆ（ｘ，ｙ）は仕上げ研磨の研磨量の分布を示す。

In the above formula (5), z _after (x, y) is the height at the coordinate (x, y) after the finish polishing, and z _before (x, y) is the height at the coordinate (x, y) after the local processing and before the finish polishing. Since the difference between z _after (x, y) and z _before (x, y) is z _dif (x, y), z _dif (x, y) indicates the distribution of the polishing amount of the finish polishing.

_{z4_dif} (x,y) in the above formula (6) is the average value of four points that are four-fold symmetric with respect to the origin. Therefore, _{z4_dif} (x,y) in the above formula (6) is a four-fold symmetric component among the distortion components, and corresponds to z3(x,y) in the above formula (3).

The present inventors have found that Δz3 can be suppressed to 6.0 nm or less by correcting the target height of each point of the central region 27 in the local processing (step S3) using the previously calculated _{z4_dif} (x,y). As a result, a glass substrate 2 with a PV of less than 10.0 nm can be obtained.

Here, the corrected target height is obtained from the difference between the target height set based on the measurement result of step S2 and the previously calculated _{z4_dif} (x,y). In other words, the corrected target processing amount is obtained from the sum of the target processing amount set based on the measurement result of step S2 and the previously calculated _{z4_dif} (x,y). The _{z4_dif} (x,y) used in these corrections is preferably the average value of multiple glass substrates 2. The average value of _{z4_dif} (x,y) is obtained for each processing condition of the finish polishing (e.g., type of abrasive, type of polishing pad, polishing pressure, and rotation speed).

The inventors also noticed that the distortion components caused by the finish polishing (step S4) are due to the relative rotation of the platen with respect to the glass substrate 2, and discovered that by reversing the rotation direction of the platen during the finish polishing, Δz3 can be suppressed to 4.0 nm or less. As a result, a glass substrate 2 with a PV of less than 8.0 nm could be obtained.

Specifically, during the finish polishing (step S4), the rotational directions of the lower platen 91 and the upper platen 92 are reversed. At the same time, the rotational directions of the sun gear 94 and the internal gear 95 are also reversed. As long as the rotational directions are reversed, the number of rotations may be the same. As mentioned above, a single-sided polishing machine may be used for the finish polishing.

If the rotation direction of the platen is reversed during the finish polishing, the direction of the arrows in FIG. 12 is reversed, and the areas where polishing is progressing are swapped with the areas where polishing is slow. In the finish polishing, the time that the platen rotates clockwise and the time that the platen rotates counterclockwise are set to be approximately the same. As a result, Δz3 can be suppressed to 4.0 nm or less.

In addition, when the rotation direction of the platen is reversed during finish polishing, the target height or target processing amount in local processing is corrected using z _{8 _ dif} (x, y) in the following equation (7) instead of z _{4 _ dif} (x, y) in the above equation (6).

上記式（７）のｚ_{８＿ｄｉｆ}（ｘ，ｙ）は、４つの基準線Ｌ１～Ｌ４を中心に線対称な８点の平均値である。４点の平均値であるｚ_{４＿ｄｉｆ}（ｘ，ｙ）の代わりに、８点の平均値であるｚ_{８＿ｄｉｆ}（ｘ，ｙ）を用いると、サンプリングの数を増加でき、誤差を低減できる。

In the above formula (7), z _{8 _ dif} (x, y) is the average value of eight points that are line-symmetrical about the four reference lines L1 to L4. By using z _{8 _ dif} (x, y), which is the average value of eight points, instead of z _{4 _ dif} (x, y), which is the average value of four points, the number of samples can be increased and errors can be reduced.

It should be noted that z _{8 — dif} (x, y), which is the average value of the eight points, does not include four-fold symmetric components as shown in Fig. 11, but this does not pose a problem because the four-fold symmetric components as shown in Fig. 11 are reduced by reversing the rotation direction of the platen during the finish polishing.

When the rotation direction of the platen is reversed during the finish polishing, the corrected target height is obtained from the difference between the target height set based on the measurement result of step S2 and the previously calculated _{z8_dif} (x,y). In other words, the corrected target processing amount is obtained from the sum of the target processing amount set based on the measurement result of step S2 and the previously calculated _{z8_dif} (x,y). The _{z8_dif} (x,y) used in these corrections is preferably the average value of multiple glass substrates 2. The average value of _{z8_dif} (x,y) is obtained for each processing condition of the finish polishing (e.g., type of abrasive, type of polishing pad, polishing pressure, and rotation speed).

The above describes the central region 27 of the first main surface 21 of the glass substrate 2, but the same applies to the central region of the second main surface 22 of the glass substrate 2. In the central region of the second main surface 22, too, if Δz3 is suppressed to 6.0 nm or less, PV can be suppressed to less than 10.0 nm.

The flatness of the first main surface 11 of the mask blank 1 is determined by the flatness of the first main surface 21 of the glass substrate 2. Therefore, if Δz3 is suppressed to 6.0 nm or less in the central region of the first main surface 11, the PV can also be suppressed to 15.0 nm or less, preferably less than 10.0 nm.

Furthermore, the flatness of the second main surface 12 of the mask blank 1 is determined by the flatness of the second main surface 22 of the glass substrate 2. Therefore, if Δz3 is suppressed to 6.0 nm or less in the central region of the second main surface 12, PV can also be suppressed to 15.0 nm or less, preferably less than 10.0 nm.

In Examples 1 to 7, steps S1 to S4 shown in FIG. 1 were performed under the same conditions except for the following conditions, a glass substrate 2 was prepared, and Δz3 and PV were measured for the central region 27 of the first main surface 21. In Examples 1 to 3, the rotation direction of the platen was reversed during the finish polishing, and the target height of the local processing was corrected using the average value of z _{8 _ dif} (x, y) obtained in advance. In Examples 4 to 5, the rotation direction of the platen was maintained in one direction during the finish polishing, and the target height of the local processing was corrected using the average value of z _{4 _ dif} (x, y) obtained in advance. On the other hand, in Examples 6 to 7, the rotation direction of the platen was maintained in one direction during the finish polishing, and the target height of the local processing was set using the measurement results of step S2 without using the average value of z _{4 _ dif} (x, y) obtained in advance. Examples 1 to 5 are working examples, and Examples 6 to 7 are comparative examples. The results are shown in Table 1.

表１から明らかなように、例１～例３では、仕上げ研磨の途中で定盤の回転方向を逆転し、且つ予め求めたｚ_{８＿ｄｉｆ}（ｘ，ｙ）の平均値を用いて局所加工の目標高さを補正したので、Δｚ３を４．０ｎｍ以下に抑制でき、ＰＶを８．０ｎｍ未満に抑制できた。また、例４～例５では、仕上げ研磨中に定盤の回転方向を一方向に維持し続け、且つ予め求めたｚ_{４＿ｄｉｆ}（ｘ，ｙ）の平均値を用いて局所加工の目標高さを補正したので、Δｚ３を６．０ｎｍ以下に抑制でき、ＰＶを１０．０ｎｍ未満に抑制できた。一方、例６～例７では、仕上げ研磨中に定盤の回転方向を一方向に維持し続け、且つ予め求めたｚ_{４＿ｄｉｆ}（ｘ，ｙ）の平均値を用いることなく、ステップＳ２の測定結果を用いて局所加工の目標高さを設定したので、Δｚ３が６．０ｎｍよりも大きくなってしまい、ＰＶが１０．０ｎｍ以上になってしまった。

As is clear from Table 1, in Examples 1 to 3, the rotation direction of the platen was reversed during the finish polishing, and the target height of the local processing was corrected using the average value of z _{8_dif} (x, y) obtained in advance, so that Δz3 could be suppressed to 4.0 nm or less, and PV could be suppressed to less than 8.0 nm. In addition, in Examples 4 to 5, the rotation direction of the platen was maintained in one direction during the finish polishing, and the target height of the local processing was corrected using the average value of z _{4_dif} (x, y) obtained in advance, so that Δz3 could be suppressed to 6.0 nm or less, and PV could be suppressed to less than 10.0 nm. On the other hand, in Examples 6 to 7, the rotation direction of the platen was maintained in one direction during the finish polishing, and _the target height of the local processing was set using the measurement results of step S2 without using the average value of z 4_dif (x, y) obtained in advance, so that Δz3 became larger than 6.0 nm, and PV became 10.0 nm or more.

Next, a mask blank 1 for EUVL was produced using the glass substrate 2 of Examples 1 to 7. First, a CrN film of 100 nm was formed as a conductive film on the first main surface 21 (the surface on which Δz3 and PV were measured) of the glass substrate 2 by ion beam sputtering. Next, a multilayer reflective film (EUV reflective film) was formed on the second main surface 22 of the glass substrate 2 by ion beam sputtering. The multilayer reflective film was formed by alternately laminating 40 periods of Si films of about 4 nm and Mo films of about 3 nm, and finally laminating a Si film of about 4 nm. Next, a Ru film of 2.5 nm was formed as a protective film on the multilayer reflective film by sputtering. Next, a TaN film of 75 nm and a TaON film of 5 nm were formed as an absorbing film (EUV absorbing film) on the protective film by sputtering. In this way, an EUVL mask blank 1 was obtained, which has a conductive film 5, a glass substrate 2, an EUV reflective film 3, and an EUV absorbing film 4 in this order.

Δz3 and PV were measured for the central region of the first main surface 11 (surface on the conductive film 5 side) of the EUVL mask blank 1 produced using the glass substrate 2 of Examples 1 to 7. The results are shown in Table 2.

表２に示すように、例１～例５では、ＥＵＶＬ用マスクブランク１の第１主表面１１の中央領域について、Δｚ３を６．０ｎｍ以下に抑制でき、ＰＶを１５．０ｎｍ以下に抑制できた。例６，例７では、ＥＵＶＬ用マスクブランク１の第１主表面１１の中央領域について、Δｚ３が６．０ｎｍより大きくなってしまい、ＰＶが１５．０ｎｍよりも大きくなってしまった。

As shown in Table 2, in Examples 1 to 5, Δz3 was suppressed to 6.0 nm or less and PV was suppressed to 15.0 nm or less for the central region of the first main surface 11 of the EUVL mask blank 1. In Examples 6 and 7, Δz3 was greater than 6.0 nm and PV was greater than 15.0 nm for the central region of the first main surface 11 of the EUVL mask blank 1.

The above describes the glass substrate for EUVL and the mask blank for EUVL according to the present disclosure, but the present disclosure is not limited to the above-mentioned embodiments. Various changes, modifications, substitutions, additions, deletions, and combinations are possible within the scope of the claims. Naturally, these also fall within the technical scope of the present disclosure.

2 Glass substrate 21 First main surface 22 Second main surface 27 Central region 28 Peripheral region 3 EUV reflective film 4 EUV absorbing film 5 Conductive film

Claims (4)

A glass substrate for EUVL having a rectangular first main surface on which a conductive film is formed, and a rectangular second main surface facing opposite to the first main surface on which an EUV reflective film and an EUV absorbing film are formed in this order,
a glass substrate for EUVL, wherein when the coordinates of a point in a central region of a square of 142 mm in length and 142 mm in width, excluding a peripheral region of the rectangular frame of the first main surface, are expressed as (x, y, z(x, y)), a maximum height difference of a surface which is a set of coordinates (x, y, z3(x, y)) calculated using the following formulas (1) to (3) is 6.0 nm or less.

In the above coordinates (x, y, z(x, y)), x indicates the horizontal coordinate, y indicates the vertical coordinate, and z indicates the height coordinate, with the horizontal, vertical, and height directions being perpendicular to each other. A glass substrate for EUVL having a rectangular first main surface on which a conductive film is formed, and a rectangular second main surface facing opposite to the first main surface on which an EUV reflective film and an EUV absorbing film are formed in this order,
a glass substrate for EUVL, in which when the coordinates of a point in a central region of a square having a length of 142 mm and a width of 142 mm, excluding a peripheral region of the rectangular frame of the second main surface, are expressed as (x, y, z(x, y)), a maximum height difference of a surface which is a set of coordinates (x, y, z3(x, y)) calculated using the following formulas (1) to (3) is 6.0 nm or less.

In the above coordinates (x, y, z(x, y)), x indicates the horizontal coordinate, y indicates the vertical coordinate, and z indicates the height coordinate, with the horizontal, vertical, and height directions being perpendicular to each other. A mask blank for EUVL having a rectangular first main surface and a rectangular second main surface facing opposite to the first main surface, and having, from a side of the first main surface to a side of the second main surface, a conductive film, a glass substrate, an EUV reflective film, and an EUV absorbing film, in this order,
a mask blank for EUVL, wherein when the coordinates of a point in a central region of a square of 142 mm in length and 142 mm in width, excluding a peripheral region of the first main surface in a rectangular frame shape, are expressed as (x, y, z(x, y)), a maximum height difference of a surface which is a set of coordinates (x, y, z3(x, y)) calculated using the following formulas (1) to (3) is 6.0 nm or less.

In the above coordinates (x, y, z(x, y)), x indicates the horizontal coordinate, y indicates the vertical coordinate, and z indicates the height coordinate, with the horizontal, vertical, and height directions being perpendicular to each other. A mask blank for EUVL having a rectangular first main surface and a rectangular second main surface facing opposite to the first main surface, and having, from a side of the first main surface to a side of the second main surface, a conductive film, a glass substrate, an EUV reflective film, and an EUV absorbing film, in this order,
a mask blank for EUVL, wherein when the coordinates of a point in a central region of a square of 142 mm in length and 142 mm in width, excluding a peripheral region of the rectangular frame of the second main surface, are expressed as (x, y, z(x, y)), a maximum height difference of a surface which is a set of coordinates (x, y, z3(x, y)) calculated using the following formulas (1) to (3) is 6.0 nm or less.

In the above coordinates (x, y, z(x, y)), x indicates the horizontal coordinate, y indicates the vertical coordinate, and z indicates the height coordinate, with the horizontal, vertical, and height directions being perpendicular to each other.

Images