JP2013082612A - Method for manufacturing glass substrate for mask blank, method for manufacturing mask blank, method for manufacturing mask, and method for manufacturing imprint mold - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a method for manufacturing a mask blank for glass substrates which is very high in flatness and reduced in microscopic defects (concave defects and convex defects).SOLUTION: A plurality of glass substrates are held between an upper surface plate 20 and a lower surface plate 10 to which polishing pads 21, 11 stick, the surface plates being arranged to be vertically opposite to each other. Then, while a polishing liquid is fed to the glass substrates from the upper surface plate 20 side, both the sides of the main surfaces of the glass substrates are polished. The glass substrates, which are rectangular, have one or more notch marks at its corner parts. Polishing both sides of the glass substrates is carried out after one of the main surfaces, on which the notch mark is not formed, is set on the upper surface plate 20 side.

Description

  The present invention relates to a method of manufacturing a glass substrate for a mask blank that is an original of a mask used for manufacturing a semiconductor device, a method of manufacturing a mask blank using the substrate, a method of manufacturing a mask, and a method of manufacturing an imprint mold About.

  Due to the recent increase in density and accuracy of VLSI devices, the demands for flatness and surface defects of glass substrates for electronic devices such as glass substrates for mask blanks are becoming stricter year by year. As a precision polishing method for reducing the flatness and surface roughness of a conventional mask blank glass substrate, for example, there is one described in JP-A-1-40267. In this precision polishing method, polishing is performed using a polishing material mainly composed of cerium oxide, and then finish polishing is performed using colloidal silica (colloidal silica).

  By the way, in order to achieve further miniaturization of the transfer pattern, the wavelength of the exposure light source tends to be shortened. For example, ArF excimer laser (wavelength 193 nm), F2 excimer laser (wavelength 157 nm), etc. are used. Further, the use of EUV light (Extreme Ultra Violet, extreme ultraviolet light: wavelength of about 13 nm) has been proposed. Along with the shortening of the wavelength of the exposure light source, requirements for the shape accuracy (flatness) and quality (defect size) of the substrate surface have become particularly strict, and the mask has extremely high flatness and no micro defects. There is a need for a blank glass substrate. The reason is that if the flatness of the substrate is poor, the dimensional accuracy of the pattern after exposure transfer is deteriorated, and if the defect size (height, depth, size) is large, it becomes a phase defect, and after exposure transfer. This is because a pattern defect of the pattern occurs. In ArF excimer laser exposure, which is being mass-produced, flatness of 500 nm or less and defect size of 100 nm or less are required.

  Further, the required value of flatness becomes strict as the pattern becomes finer. In the EUV exposure, which is the next generation exposure technology, the PV value (maximum height with respect to the reference surface) is measured in a region of 142 mm long × 142 mm wide. And the minimum height), flatness of 50 nm or less is required. For minute defects, the defect size is required to be 50 nm or less.

  In recent years, patterned media in which data tracks are magnetically separated are proposed as magnetic media for hard disk drives (HDD) that can cope with higher recording densities. The conventional technique of minimizing the width of the magnetic head to increase the density has a limit on the density increase, and the magnetic influence between adjacent tracks and the thermal fluctuation phenomenon cannot be ignored. Patterned media include discrete track recording media (hereinafter referred to as “DTR media”) in which data tracks of a magnetic disk are magnetically separated, and the DTR media is further increased in density. Bit patterned media (hereinafter referred to as “BPM”) for recording a signal as a bit pattern (dot pattern) to be developed is known.

  Such patterned media such as DTR media and BPM are intended to improve signal quality by removing magnetic material in portions unnecessary for recording (groove processing with a fine width). As a pattern transfer method for performing processing, an imprint technique (or “nanoimprint technique”) is generally used. In the imprint technique, using a master mold (also referred to as “master”) or a working mold (also referred to as “copy mold”) that is transferred by copying the master mold once or a plurality of times as an original mold, A patterned medium (for example, BPM) is produced by transferring a fine pattern of the master mold or the working mold (these master mold and working mold are hereinafter referred to as “imprint mold”) to a transfer target.

  In addition to the production of the patterned media, optical functions are achieved by fine patterns such as structural filters with nanopatterns formed on the surface, high-brightness LEDs, polarizers, photonic crystal lasers, GaN film substrates, etc. The imprint technique is also used for manufacturing optical parts with the added and semiconductor devices.

  In these patterned media, optical components, semiconductor devices, etc., there is a demand for higher density and higher accuracy of patterns. For this purpose, imprint molds used in the production of these patterned media are also fine and accurate. A high pattern is required. In particular, since imprint technology provides pattern transfer at an equal magnification, imprint molds are required to have higher accuracy.

  An imprint mold such as a master mold used in such an imprint technique is manufactured using a mold blank in which a hard mask layer and a resist layer are sequentially formed on a substrate (generally a glass substrate). Specifically, a resist pattern is formed by performing predetermined pattern exposure and development on the resist layer of the mold blank, and further, the hard mask layer and the substrate in the mold blank are etched using this resist pattern as a mask. For example, a master mold is manufactured by forming a predetermined uneven pattern on the substrate. Therefore, in order to obtain an imprint mold having a high-precision fine pattern, a glass substrate having high flatness and having no microdefects is required even in the glass substrate for the mold blank.

  In view of the background as described above, Patent Document 1 (Japanese Unexamined Patent Application Publication No. 2009-6457) describes a test substrate separately from a product substrate in order to obtain a substrate having an extremely flat and smooth surface property. If the polishing pad is not adopted in the test process, after the polishing pad is corrected or replaced until it is adopted, the test process is repeated again, and what type of polishing apparatus is used in the polishing process. Disclosed is a method for manufacturing a substrate for grasping whether there is a tendency, using a polishing pad employed in a test process, and polishing a product substrate according to a selected substrate setting direction.

JP 2009-6457 A

  However, when the method for manufacturing a substrate as disclosed in Patent Document 1 is adopted, the work load of the test process is extremely large, and the manufacturing cost increases. In addition, ArF excimer laser exposure mask blanks and EUV exposure reflective mask blanks used in generations after the semiconductor design rule 32 nm node are also regarded as problematic by microdefects of 0.1 μm size (concave defects and convex defects). Strict defect quality is required, but the substrate manufacturing method disclosed in Patent Document 1 cannot sufficiently suppress the generation of minute defects (concave defects, convex defects).

  The present invention has been made in view of the above-described problems of the prior art, and the object thereof is to manufacture a glass substrate for a mask blank that has extremely high flatness and reduced minute defects (concave defects, convex defects). It is providing the method, the manufacturing method of the mask blank obtained using this board | substrate, the manufacturing method of the mask obtained using this mask blank, and the manufacturing method of an imprint mold.

As a result of intensive studies to solve the above problems, the present inventor has found that the above problems can be solved by an invention having the following configuration.
(Configuration 1)
While sandwiching a plurality of glass substrates between an upper surface plate and a lower surface plate provided facing the top and bottom with the polishing pad attached, while supplying polishing liquid from the upper surface plate side to the glass substrate, A method of manufacturing a glass substrate for a mask blank in which both main surfaces of the glass substrate are polished on both sides, wherein the glass substrate has a quadrangular shape, and one main surface and the corner portion at the corner of the quadrangle The glass substrate has at least one notch mark formed by cutting off three surfaces of the two end surfaces into an oblique cross-sectional shape, and the double-side polishing of the glass substrate is performed on one main surface on which the notch mark is not formed. A method for producing a glass substrate for a mask blank, wherein the surface is set on the upper surface plate side.

  According to Configuration 1, one glass substrate main surface on which notch marks are not formed is set on the upper surface plate side, and double-side polishing of the glass substrate is performed. During polishing, the upper platen side has fewer such defects than the lower platen side, which is affected by clogging and material deterioration due to the penetration of the polishing liquid, and the physical properties of the polishing pad are large and easily change over time. One of the main surfaces of the glass substrate that is polished by contact with the polishing pad on the surface plate side is not very flat, and it can reduce micro defects (concave defects, convex defects). Finished with good surface shape accuracy and quality. Usually, since a thin film to be a transfer pattern is formed on one main surface of the glass substrate on which the notch mark is not formed, the thin film is formed on the main surface finished with good surface shape accuracy and quality by the present invention. Can be formed.

(Configuration 2)
While sandwiching a plurality of glass substrates between an upper surface plate and a lower surface plate provided facing the top and bottom with the polishing pad attached, while supplying polishing liquid from the upper surface plate side to the glass substrate, A method for manufacturing a glass substrate for a mask blank in which both main surfaces of the glass substrate are polished on both sides, wherein the both main surfaces are provided so as to face a first surface on which a transfer pattern is formed and the first surface. A method for producing a glass substrate for a mask blank, wherein the double-side polishing of the glass substrate is performed by setting the first surface on the upper surface plate side.

  According to Configuration 2, the first surface on which the transfer pattern of the glass substrate is formed is set on the upper surface plate side, and double-side polishing of the glass substrate is performed. During polishing, the upper platen side has fewer such defects than the lower platen side, which is affected by clogging and material deterioration due to the penetration of the polishing liquid, and the physical properties of the polishing pad are large and easily change over time. The first surface of the glass substrate to be polished by contact with the polishing pad on the surface plate side has extremely high flatness, can reduce micro defects (concave defects, convex defects), and has a good surface shape. Finished with accuracy and quality. Therefore, the thin film can be formed on the main surface finished with good surface shape accuracy and quality according to the present invention.

(Configuration 3)
The method for producing a glass substrate for a mask blank according to Configuration 1 or 2, wherein the double-side polishing is performed in a plurality of stages of polishing processes using different polishing liquids.
As in Configuration 3, by performing double-side polishing of the glass substrate in a plurality of stages of polishing processes with different polishing liquids, a glass substrate main surface is desired stepwise (gradually) using polishing liquids having different properties. Surface shape accuracy and quality can be built.

(Configuration 4)
4. The method of manufacturing a glass substrate for a mask blank according to Configuration 3, wherein in the plurality of stages of polishing, the number of glass substrates subjected to double-side polishing is the same.
When performing double-side polishing of a glass substrate in a multi-step polishing process, the same number of glass substrates to be double-side polished at each step as in Configuration 4 can improve the surface shape accuracy and quality between the substrates. The variation can be suppressed, which is preferable.

(Configuration 5)
5. The mask according to Configuration 4, wherein in the multi-stage polishing process, when the number of glass substrates to be double-side polished is insufficient, a dummy glass substrate prepared separately is inserted and the next polishing process is performed. Manufacturing method of glass substrate for blanks.
As in Configuration 5, when a shortage of the number of glass substrates to be double-side polished due to sampling inspection or the like in the middle occurs in a multi-stage polishing process, a dummy glass substrate prepared separately is introduced and the next polishing process is performed. This is preferable because variations in surface shape accuracy and quality finish between substrates can be suppressed.

(Configuration 6)
6. The mask blank glass substrate manufacturing method according to Configuration 5, wherein the dummy glass substrate has a thickness equal to or smaller than an average thickness of other glass substrates.
In a multi-stage polishing process, when the number of glass substrates to be double-side polished is insufficient, and a dummy glass substrate prepared separately is used to perform the next polishing process (Configuration 5), The thickness of the glass substrate is preferably equal to or smaller than the average thickness of the other glass substrates because it is possible to suppress variations in surface shape accuracy and quality finish between the substrates.

(Configuration 7)
A method for producing a mask blank, comprising: forming a thin film to be a transfer pattern on the main surface of the glass substrate produced by the method for producing a glass substrate for mask blank according to any one of configurations 1 to 6.
According to the present invention, a mask blank is manufactured using a glass substrate for a mask blank that has a very high flatness, reduces minute defects, and has a good surface shape accuracy and quality. Less mask blanks are obtained.

(Configuration 8)
A method for manufacturing a mask, comprising: patterning the thin film of a mask blank manufactured by the method for manufacturing a mask blank according to Configuration 7.
By producing a mask using the mask blank of the present invention, it is possible to obtain a mask having extremely good pattern dimensional accuracy after exposure transfer and having less pattern defects in the pattern after exposure transfer.

(Configuration 9)
Forming an imprint mold blank by forming a thin film to be a transfer pattern on the main surface of the glass substrate manufactured by the method for manufacturing a mask blank glass substrate according to any one of Configurations 1 to 6; Patterning the thin film of the imprint mold blank manufactured to form the pattern of the thin film; and using the thin film pattern as a mask, the glass substrate is etched by an etching process to form a digging pattern in the glass substrate And a process for forming an imprint mold.
According to the present invention, an imprint mold blank is manufactured using a glass substrate having extremely high flatness, reduced micro defects, and finished with good surface shape accuracy and quality. Further, the imprint mold blank is used. By manufacturing the imprint mold, an imprint mold having very good dimensional accuracy of the pattern formed on the glass substrate and less pattern defects can be obtained.

According to the present invention, it is possible to obtain a glass substrate for a mask blank having extremely high flatness and capable of reducing the occurrence of minute concave or convex surface defects on the glass substrate surface.
In addition, by producing a mask blank using the above glass substrate for mask blank obtained by the present invention, and further producing a mask using this mask blank, the dimensional accuracy of the pattern after exposure transfer is extremely good. In addition, it is possible to obtain a mask with few occurrences of pattern defects after exposure and transfer.
Moreover, the dimension of the pattern formed in a glass substrate by manufacturing an imprint mold blank using the said glass substrate obtained by this invention, and also manufacturing an imprint mold using this imprint mold blank. An imprint mold having very good accuracy and less pattern defects can be obtained.

It is a perspective view of the square glass substrate which has a notch mark. It is sectional drawing which shows schematic structure of the double-side polish apparatus of the planetary gear system used in a grinding | polishing process. It is a perspective view which shows the meshing relationship of a sun gear, an internal gear, and a carrier. It is a schematic sectional drawing for demonstrating the manufacturing process of an imprint mold. It is a schematic sectional drawing for demonstrating the relationship between a master mold and the working mold produced using this master mold.

Hereinafter, embodiments of the present invention will be described in detail.
[Glass substrate for mask blank]
First, the manufacturing method of the glass substrate for mask blanks concerning this invention is demonstrated.
In the present invention, as described above, a plurality of glass substrates are sandwiched between an upper surface plate and a lower surface plate that are provided with a polishing pad attached and faced up and down, and the upper surface plate side is sandwiched between the glass substrates. A method for producing a glass substrate for a mask blank, in which both main surfaces of the glass substrate are polished on both sides while supplying a polishing liquid, wherein the shape of the glass substrate is a quadrangle, and in the corner of the quadrangle, The glass substrate has at least one notch mark formed by cutting off three main surfaces and two end surfaces forming the corner portion into an oblique cross section, and the double-side polishing of the glass substrate is performed by the notch mark This is performed by setting one main surface on which the surface is not formed on the upper surface plate side.

  In the present invention, a plurality of glass substrates are sandwiched between an upper surface plate and a lower surface plate that are provided with a polishing pad attached and faced up and down, and a polishing liquid is placed on the glass substrate from the upper surface plate side. Both main surfaces of the glass substrate are polished on both sides while supplying. The glass substrate to be double-side polished in this way is a mask blank glass substrate, the outer shape of the glass substrate is a quadrangle, and two end surfaces that form one main surface and a corner at the square corner And at least one notch mark formed by cutting off the three surfaces in an oblique cross section.

FIG. 1 is a perspective view of a rectangular glass substrate having the notch mark. The notch mark is usually for distinguishing between the glass type and the front and back of the glass substrate. The glass substrate 1 shown in FIG. 1 has a notch mark 2 at one corner of the main surface 1B. However, this is merely an example, and a plurality of types of glass, for example, soda lime glass, synthetic quartz glass, low expansion glass, etc., can be distinguished by the number of notch marks, and in the case of a plurality of notch marks, the formation location. .
When the glass substrate 1 is used for a binary mask blank or a phase shift mask blank, the glass substrate 1 is not particularly limited as long as it has transparency with respect to an exposure wavelength to be used. A synthetic quartz substrate and other various glass substrates ( For example, soda lime glass, aluminosilicate glass, etc.) are used. Among them, a synthetic quartz substrate is particularly preferably used because it has high transparency in an ArF excimer laser or a shorter wavelength region.

In the case of EUV exposure, the glass substrate 1 has a range of 0 ± 1.0 × 10 ⁻⁷ / ° C., more preferably 0 ± 0. Those having a low thermal expansion coefficient within the range of 3 × 10 ⁻⁷ / ° C. are preferably used. Examples of the material having a low thermal expansion coefficient within this range include SiO ₂ —TiO ₂ glass, multicomponent glass ceramics, and the like. Can be used.
Further, the mask blank glass substrate has a square shape such as a square or a rectangle. In the case of the above-described mask blank glass substrate for ArF excimer laser exposure and EUV exposure, a 6025 substrate (about 152 mm × About 152 mm and a thickness of about 6.35 mm).

In the present invention, one glass substrate main surface on which the notch mark is not formed is set on the upper platen side of the polishing apparatus to perform double-side polishing of the glass substrate. In the case of the glass substrate 1 shown in FIG. 1, it is set in the polishing apparatus so that one main surface 1A on which the notch mark 2 is not formed is on the upper surface plate side. Usually, on one main surface of the glass substrate on which notch marks are not formed, a thin film serving as a transfer pattern is formed thereon.
The glass substrate is subjected to predetermined chamfering and grinding before the double-side polishing step.

In practice, the double-side polishing of the glass substrate is preferably performed using a double-side polishing apparatus.
2 is a cross-sectional view showing a schematic configuration of a planetary gear type double-side polishing apparatus used in the double-side polishing process, and FIG. 3 is a meshing relationship of the sun gear, the internal gear and the carrier in the double-side polishing apparatus of FIG. FIG.
A planetary gear type double-side polishing apparatus used also in a polishing process of an embodiment described later will be described with reference to FIGS.

  The planetary gear type double-side polishing apparatus meshes with the sun gear 30, the internal gear 40 arranged concentrically around the sun gear 30, the sun gear 30 and the internal gear 40, and the sun gear 30 and the internal gear 40 rotate. And a carrier 50 that revolves and rotates in response to the workpiece, and a workpiece W (for example, the glass substrate 1) held on the carrier 50, to which the polishing pads 21 and 11 are attached, respectively. A surface plate 20 and a lower surface plate 10, and a polishing liquid supply unit 60 that supplies a polishing liquid between the upper surface plate 20 and the lower surface plate 10 are provided. The rotation driving of the sun gear 40 (A is a rotation axis) is controlled by the sun gear rotation driving unit 31. The raising and lowering and rotation driving of the upper surface plate 20 fixedly supported by the upper support member 22 are controlled by the upper surface plate raising and lowering driving unit 24 and the upper surface plate rotation driving unit 23. The rotation of the lower surface plate 10 fixedly supported by the lower support member 12 is controlled by the lower surface plate rotation driving unit 13. The polishing liquid supply unit 60 supplies the polishing liquid storage unit 61 that stores the polishing liquid and the polishing liquid stored in the polishing liquid storage unit 61 to the polishing region between the upper surface plate 20 and the lower surface plate 10. The polishing liquid reservoir 61 is provided with a temperature control device so that the temperature of the polishing liquid supplied to the workpiece W is constant. ing. The polishing liquid reservoir 61 is formed in an annular shape on a horizontal plane, and is provided at a position above the upper support member 22 via a plurality of support members 63. The upper support member 22, the upper surface plate 20, and the polishing pad 21 are formed with a plurality of through holes 22a, 20a, and 21a communicating with each other, and the lower ends of the tubes 62 are connected thereto. Thus, the polishing liquid stored in the polishing liquid storage unit 61 is supplied from the upper surface plate side through the tube 62 and the through hole, and is supplied to the polishing region between the upper surface plate 20 and the lower surface plate 10. The Although not shown in the figure, the polishing liquid supplied to the polishing region is recovered in the tank via a predetermined recovery path, and then again returned to the polishing liquid via a reduction path where a pump and a filter are interposed. It is sent to the storage unit 61.

  Further, during the polishing process, a coolant flows in each of the upper surface plate 20 and the lower surface plate 10 in order to suppress warping of the surface plate due to a temperature increase of the upper surface plate 20 and the lower surface plate 10 and a temperature increase of the slurry. A refrigerant supply path is provided, and is controlled to be a constant temperature during the polishing process.

  At the time of polishing, the workpiece W held by the carrier 50 is sandwiched between the upper surface plate 20 and the lower surface plate 10, and the polishing pads 21, 11 of the upper and lower surface plates 20, 10 and the workpiece W are polished. While supplying the polishing liquid therebetween, the upper and lower surfaces of the workpiece W are mirror-polished simultaneously while the carrier 50 revolves and rotates according to the rotation of the sun gear 30 and the internal gear 40.

  In the present invention, as described above, one glass substrate main surface on which no notch mark is formed is set on the upper surface plate side, and double-side polishing of the glass substrate is performed. The reason why one main surface on which the notch mark set on the upper surface plate side is not formed by the present invention is high in flatness and has few micro defects is excellent in surface shape accuracy and quality. According to the above, it is estimated as follows.

  In the double-side polishing process using the double-side polishing apparatus as shown in FIG. 2 described above, the polishing pad on the lower surface plate side has a large amount of penetration due to the accumulated polishing liquid, which causes clogging or material deterioration. As a result, the physical properties of the polishing pad are large and easily change over time, and the flatness change accompanying the polishing process is also large. Compared to the lower platen side, the polishing pad on the upper platen side has fewer such problems, and the change in physical properties with time is small and stable, so that the glass substrate is polished by contact with the polishing pad on the upper platen side. One main surface on which no notch mark is not formed has extremely high flatness and can reduce minute defects (concave defect, convex defect), and finish with good surface shape accuracy and quality. Usually, since a thin film to be a transfer pattern is formed on one main surface of the glass substrate on which the notch mark is not formed, the thin film is formed on the main surface finished with good surface shape accuracy and quality by the present invention. Can be formed.

  The above has described the case where one glass substrate main surface on which the notch mark is not formed is set on the upper platen side of the polishing apparatus to perform double-side polishing of the glass substrate, but the present invention is not limited thereto. . That is, as another embodiment of the present invention, a plurality of glass substrates are sandwiched between an upper surface plate and a lower surface plate that are provided opposite to each other with a polishing pad attached thereto, A method of manufacturing a glass substrate for a mask blank, in which both main surfaces of the glass substrate are polished on both sides while supplying a polishing liquid from an upper surface plate side, wherein both the main surfaces are a first surface on which a transfer pattern is formed. And a second surface provided opposite to the first surface, and the double-side polishing of the glass substrate is performed by setting the first surface on the upper surface plate side. It is a manufacturing method of the glass substrate for mask blanks.

  Referring to FIG. 1, both main surfaces of the glass substrate 1 are divided into a first surface 1A on which a transfer pattern is formed, and a second surface 1B (notch provided opposite to the first surface). It is also a surface having a mark). By setting the first surface on which the transfer pattern of the glass substrate is formed on the upper surface plate side and performing double-side polishing of the glass substrate, the glass substrate to be polished by contact with the polishing pad on the upper surface plate side The first surface has a very high flatness and can reduce minute defects (concave defects, convex defects), resulting in good surface shape accuracy and quality. Therefore, a thin film serving as the transfer pattern can be formed on the first surface finished with good surface shape accuracy and quality according to the present invention.

  In the present invention, the double-side polishing is preferably carried out in a plurality of stages of polishing steps with different polishing liquids. Here, the difference in polishing liquid means that, for example, the material and particle size of the abrasive grains contained in the polishing liquid are different. The properties of the polishing liquid differ depending on the material and particle size of such abrasive grains, and affect the polishing rate and the quality of the finished surface. When finishing a glass substrate surface with high flatness and reduced defects as in the present invention, polishing liquids having different properties can be obtained by performing double-side polishing of the glass substrate in a plurality of stages of polishing processes with different polishing liquids. It is preferable that the glass substrate main surface is made to have desired surface shape accuracy and quality step by step (gradually).

  For example, the double-side polishing can be performed in three stages of polishing processes such as a first polishing (rough polishing) process, a second polishing (precision polishing) process, and a third polishing (ultra-precision polishing) process. In this case, for example, cerium oxide, zirconium oxide or the like can be used as polishing abrasive grains in the first polishing (coarse polishing) process, and cerium oxide, zirconium oxide or silica (as the polishing abrasive grains in the second polishing (precision polishing) process). (Including colloidal silica) can be used. In the third polishing (ultra-precision polishing) step, cerium oxide, silica (including colloidal silica), etc. can be used as polishing abrasive grains, especially silica having a small average particle diameter and capable of chemical mechanical polishing. (Including colloidal silica) is preferably used.

  The silica used in the present invention is preferably colloidal silica produced by a sol-gel method. For example, high-purity colloidal silica can be obtained by synthesizing a high-purity alkoxysilane from which metal impurities have been removed using a sol-gel method. Since the silica thus obtained has relatively few impurities, the generation of silica aggregates can be reduced.

The colloidal silica contained in the polishing liquid is preferably one having an average particle diameter of about 20 to 500 nm from the viewpoint of polishing efficiency.
As a solvent for the polishing liquid, colloidal silica is monodispersed and stable in an alkaline atmosphere, and therefore it is adjusted to be alkaline by adding an inorganic alkali such as NaOH or KOH or an organic alkali such as an amine. Is generally considered good, but may be adjusted to be acidic.

The content of silica in the polishing liquid is determined in consideration of the generation rate of fine protrusions and the polishing rate, and is preferably 50 wt% or less, and more preferably 10 to 40 wt%.
The temperature of the polishing liquid supplied to the glass substrate is preferably 25 ° C. or lower. The temperature of the polishing liquid is adjusted by controlling the supply temperature of the polishing liquid via a chiller while supplying the polishing liquid to the polishing machine, or by providing a cooling mechanism on the surface plate of the polishing machine to control the supply temperature of the polishing liquid. It doesn't matter. The temperature of the polishing liquid is preferably 5 ° C. or higher and 20 ° C. or lower, more preferably 5 ° C. or higher and 15 ° C. or lower.

When performing double-side polishing of glass substrates in the above-described multi-step polishing process, it is preferable that the number of glass substrates to be double-side polished in each step is the same, so that variations in surface shape accuracy and quality finish between the substrates can be suppressed. It is.
Further, in such a multi-stage polishing process, when a shortage of the number of glass substrates to be double-side polished due to a sampling inspection or the like in the middle occurs, a dummy glass substrate prepared separately may be inserted to perform the next polishing process. It is preferable because variations in surface shape accuracy and quality finish between substrates can be suppressed.
In addition, when the number of glass substrates to be double-side polished as described above is insufficient and a dummy glass substrate prepared separately is put into the next polishing step, the thickness of the dummy glass substrate is different from that of other glass substrates. It is preferable that the average plate thickness is equal to or smaller than the average plate thickness because variations in surface shape accuracy and quality finish between the substrates can be suppressed.

  By the double-side polishing, the main surface of the glass substrate on which the transfer pattern is formed is subjected to surface processing so as to have high flatness from the viewpoint of obtaining at least pattern transfer accuracy and position accuracy. For example, in the case of KrF excimer laser exposure or ArF excimer laser exposure, the flatness is preferably 0.3 μm or less in the region of the main surface 142 mm × 142 mm on the side where the transfer pattern of the glass substrate is formed, Particularly preferably, it is 0.1 μm or less. In the case of EUV exposure, the flatness is preferably 0.1 μm or less, particularly preferably 0.05 μm or less, in the main surface 142 mm × 142 mm region on the side where the transfer pattern of the glass substrate is formed. . The main surface opposite to the side on which the transfer pattern is formed is a surface that is electrostatically chucked when being set in the exposure apparatus, and has a flatness of 1 μm or less, preferably 0.8 mm in a 142 mm × 142 mm region. 5 μm or less.

  The surface smoothness required for the mask blank glass substrate is the root mean square roughness (RMS) of the main surface on the side where the transfer pattern of the glass substrate is formed. It is preferable that it is 0.3 nm or less. Here, the surface roughness RMS is defined in Japanese Industrial Standard (JIS) B0601 (2001). In the present invention, the lower limit value of the surface roughness RMS is not particularly limited, and the effect of the present invention is more remarkable as the surface of the glass substrate is smoother.

  As described above, since a glass substrate for mask blank for EUV exposure requires particularly high flatness, a local processing step and a touch polishing (finish polishing) step are further performed on the glass substrate after the double-side polishing. You may go. In the local processing step, first, the concavo-convex shape (flatness) of the glass substrate surface after the double-side polishing is measured. An optical interferometer is usually used for measuring the uneven shape on the surface of the glass substrate. Examples of the optical interferometer include a fringe observation interferometer and a phase shift interferometer. The measurement result of the concavo-convex shape measured by the optical interferometer is stored in a recording medium such as a computer. Next, the measurement result of the concavo-convex shape is compared with a predetermined reference value (desired flatness) by an arithmetic processing means such as a computer, and the difference is a predetermined area on the surface of the glass substrate (for example, 5 mm in length × It is calculated for each area of 5 mm in width. That is, the machining allowance is set according to the height of the convex portion on the surface of the glass substrate. This difference (machining allowance) is used as a necessary removal amount for each predetermined region in local surface machining.

  Next, the convex portion is locally processed for each predetermined region under the processing conditions corresponding to the machining allowance set by the arithmetic processing, and the flatness of the glass substrate surface is controlled to a predetermined reference value or less. As a local surface processing method, an MRF (Magneto Rheological Finishing) processing method is used in which a magnetic polishing slurry containing abrasive grains in a magnetic fluid containing iron is used to locally contact the glass substrate surface. Can do. In addition to the MRF processing method, a local processing method using GCIB (gas crater ion beam) or plasma etching may be used.

  In the above-described local processing step, the touch polishing is performed for the purpose of removing a surface roughness or a work-affected layer on the glass substrate surface, and the surface roughness or processing that needs to be removed on the glass substrate surface. When the deteriorated layer is not generated, the touch polishing is not particularly required. The touch polishing method may be any polishing method that can improve the surface roughness while maintaining the surface shape (flatness) obtained in the local processing step. It is preferable to carry out the same method. That is, it is preferable to perform the touch polishing by setting the main surface of the glass substrate on which the notch mark is not formed on the upper surface plate side using a double-side polishing apparatus.

[Mask blank]
This invention provides also about the manufacturing method of the mask blank which forms the thin film used as a transfer pattern on the said main surface of the glass substrate for mask blanks of the said structure. According to the present invention, a mask blank is manufactured using a glass substrate for a mask blank that has a very high flatness, reduces minute defects, and has a good surface shape accuracy and quality. Less mask blanks are obtained.

A binary mask blank is obtained by forming a light shielding film on the main surface of the glass substrate for mask blank obtained by the present invention. A phase shift mask blank can be obtained by forming a phase shift film, or a phase shift film and a light shielding film on the main surface of the mask blank glass substrate. In this case, the main surface is the main surface set on the upper surface plate side in the above-described double-side polishing of the glass substrate. For example, as described with reference to FIG. 1, the main surface on the side where the notch mark 2 is not formed. Surface 1A.
The light shielding film may be a single layer or a plurality of layers (for example, a laminated structure of a light shielding layer and an antireflection layer). Further, when the light shielding film has a laminated structure of a light shielding layer and an antireflection layer, the light shielding layer may be composed of a plurality of layers. The phase shift film may be a single layer or a plurality of layers.

  As such a mask blank, for example, a binary mask blank including a light shielding film formed of a material containing chromium (Cr), or a light shielding film formed of a material containing transition metal and silicon (Si). Binary mask blank provided, binary mask blank provided with light-shielding film formed of material containing tantalum (Ta), formed of material containing silicon (Si), or material containing transition metal and silicon (Si) For example, a phase shift type mask blank provided with the phase shift film that is used.

Examples of the material containing chromium (Cr) include chromium alone and chromium-based materials (CrO, CrN, CrC, CrON, CrCN, CrOC, CrOCN, etc.).
As the material containing tantalum (Ta), in addition to tantalum alone, a compound of tantalum and another metal element (for example, Hf, Zr, etc.), at least one of nitrogen, oxygen, carbon, and boron in addition to tantalum A material containing two elements, specifically, a material containing TaN, TaO, TaC, TaB, TaON, TaCN, TaBN, TaCO, TaBO, TaBC, TaCON, TaBON, TaBCN, TaBCON, and the like.

  As the material containing silicon (Si), a material further containing at least one element of nitrogen, oxygen, and carbon, specifically, a silicon nitride, an oxide, a carbide, an oxynitride ( A material containing oxynitride), carbonate (carbide oxide), or carbonitride (carbonitride) is preferable.

  Moreover, as the material containing the transition metal and silicon (Si), in addition to the material containing the transition metal and silicon, the material further contains at least one element of nitrogen, oxygen and carbon in addition to the transition metal and silicon. Is mentioned. Specifically, a transition metal silicide or a material containing a transition metal silicide nitride, oxide, carbide, oxynitride, carbonate, or carbonitride is preferable. As the transition metal, molybdenum, tantalum, tungsten, titanium, chromium, hafnium, nickel, vanadium, zirconium, ruthenium, rhodium, niobium, and the like are applicable. Of these, molybdenum is particularly preferred.

Moreover, a reflective type film is formed in order on the main surface of the mask blank glass substrate by reflecting a multilayer reflective film that reflects exposure light such as EUV light and a pattern forming absorber film that absorbs exposure light. A mask blank is obtained.
The multilayer reflective film is a multilayer film in which a low refractive index layer and a high refractive index layer are alternately laminated. In general, a thin film of a heavy element or a compound thereof and a thin film of a light element or a compound thereof are alternately arranged. In addition, a multilayer film laminated for about 40 to 60 periods is used. For example, as a multilayer reflective film for EUV light having a wavelength of 13 to 14 nm, a Mo / Si periodic laminated film in which Mo films and Si films are alternately laminated for about 40 cycles is preferably used. In addition, as a multilayer reflective film used in the EUV light region, Ru / Si periodic multilayer film, Mo / Be periodic multilayer film, Mo compound / Si compound periodic multilayer film, Si / Nb periodic multilayer film, Si / Mo / Examples include Ru periodic multilayer films, Si / Mo / Ru / Mo periodic multilayer films, and Si / Ru / Mo / Ru periodic multilayer films. The material may be appropriately selected depending on the exposure wavelength.

  The absorber film has a function of absorbing exposure light such as EUV light. For example, tantalum (Ta) alone or a material mainly composed of Ta is preferably used. As a material having Ta as a main component, a material containing Ta and B, a material containing Ta and N, a material containing Ta and B and further containing at least one of O and N, a material containing Ta and Si, Ta A material containing Si and N, a material containing Ta and Ge, a material containing Ta, Ge and N are used.

In general, a protective film or a buffer film can be provided between the multilayer reflective film and the absorber film for the purpose of protecting the multilayer reflective film when patterning or modifying the absorber film. In addition to silicon, ruthenium, and ruthenium compounds containing one or more elements of niobium, zirconium, and rhodium in ruthenium are used as the material for the protective film. The material for the buffer film is mainly the above-described chromium-based material. Is used.
Note that a film formation method for forming a thin film such as the light shielding film is not particularly limited. For example, a sputtering method, an ion beam deconvolution (IBD) method, a CVD method and the like are preferable.

[mask]
The present invention also provides a method for manufacturing a mask, wherein the thin film in the mask blank having the above-described configuration is patterned.
By producing a mask using the mask blank of the present invention, it is possible to obtain a mask having extremely good pattern dimensional accuracy after exposure transfer and having less pattern defects in the pattern after exposure transfer.

For example, a binary mask having a light shielding film pattern can be obtained by patterning the light shielding film in the binary mask blank described above. In addition, in the phase shift type mask blank having the structure including the phase shift film or the phase shift film and the light shielding film on the main surface of the glass substrate for the mask blank described above, the phase shift can be obtained by patterning the thin film to be the transfer pattern. A mold mask is obtained.
Moreover, a reflective mask provided with an absorber film pattern is obtained by patterning the absorber film in the above-described reflective mask blank.
As a method for patterning a thin film to be a transfer pattern in a mask blank, a highly accurate photolithography method is most suitable.

[Imprint mold]
This invention also provides the imprint mold manufacturing method which manufactures an imprint mold blank using the said glass substrate obtained by this invention, and also manufactures using this imprint mold blank.
The first surface on which the transfer pattern of the glass substrate is formed is set on the upper surface plate side in the double-side polishing of the glass substrate, and the flatness according to the present invention is extremely high, and micro defects are reduced, and a good surface is obtained. By producing an imprint mold using an imprint mold blank using a glass substrate finished in shape accuracy and quality, the dimensional accuracy of the pattern formed on the glass substrate is extremely good, and pattern defects are also generated. A few imprint molds can be obtained.

FIG. 4 is a schematic cross-sectional view for explaining an imprint mold manufacturing process, and FIG. 5 is a schematic cross-sectional view for explaining the relationship between a master mold and a working mold produced using the master mold. It is.
As shown in FIG. 4 (e) or FIG. 5 (a), for example, an embodiment 110 of the master mold which is an imprint mold obtained by the present invention has a digging pattern ( An uneven pattern) 101a is formed. In the master mold 110, the first surface on which the transfer pattern is formed is the formation surface of the digging pattern 101a.

The master mold 110 which is such an imprint mold is manufactured using an imprint mold blank (hereinafter, simply referred to as “mold blank”) 100 as illustrated in FIG.
The mold blank 100 has a structure in which a hard mask layer 102 is formed on one main surface of a glass substrate 101. The hard mask layer 102 is a layer that becomes a mask when the glass substrate 101 is dug by an etching process. In the mold blank 100, the hard mask layer 102 is the thin film that becomes the transfer pattern described above. As shown in FIG. 4A, the mold blank includes a structure having a resist layer 103 on the hard mask layer 102.

As the glass substrate 101 of the mold blank 100, the glass substrate obtained by the above-described present invention is used. Therefore, the main surface of the glass substrate 101 on which the hard mask layer 102 is formed is the first surface on which the transfer pattern is formed, and is set on the upper platen side in the double-side polishing of the glass substrate. Main surface. The main surface of the glass substrate 101 on the side where the hard mask layer 102 is not formed is a second surface provided to face the first surface.
As the glass substrate 101 for the mold blank 100, a glass substrate such as a synthetic quartz substrate is preferably used in terms of excellent flatness and smoothness. The shape of the glass substrate 101 is not particularly limited, such as a square shape such as a square or a rectangle, or a circular shape, but a circular substrate is usually used for manufacturing patterned media, and an optical component or semiconductor device is used for manufacturing. A rectangular substrate is used.

  Since the hard mask layer 102 is a layer serving as a mask when the glass substrate 101 is dug by an etching process, a material having resistance to etching of the glass substrate (usually dry etching with a fluorine-based gas) is used. In particular, when a master mold is manufactured, a conductive material is preferable from the viewpoint of preventing charge-up because a pattern is drawn with an electron beam on the resist layer.

As a material of the hard mask layer 102, for example, a material containing chromium, nitrogen and oxygen is preferably cited. In this case, it is preferable to set the composition gradient so that the nitrogen content is higher toward the glass substrate 101 and the oxygen content is higher toward the surface opposite to the glass substrate. The nitrogen has a function of suppressing oxidation in the layer, and the oxygen has a function of improving adhesion when a resist layer is formed on the surface of the hard mask layer. In addition, it is not necessary to be limited to the above chromium, and a material containing a metal such as Al, Ta, Si, W, Mo, Hf, Ti, or the like is used as long as it has resistance to etching of the glass substrate and conductivity. May be. The hard mask layer 102 may be a single layer or a plurality of layers (stacked layers).
A film forming method for forming a thin film of the hard mask layer 102 is not particularly limited. For example, a sputtering method, an ion beam deconvolution (IBD) method, a CVD method and the like are preferable. The thickness of the hard mask layer 102 is not particularly limited, but is preferably 5 nm or less from the viewpoint of forming a fine pattern.

Next, a method for manufacturing a master mold or a working mold using the mold blank 100 as described above will be described.
When producing a master mold, a fine pattern is drawn on the resist layer 103 on the mold blank 100 using an electron beam drawing machine. Then, after drawing, the resist layer 103 is developed to form a resist pattern 103a (see FIG. 4B).

  Next, the hard mask layer 102 is etched using the formed resist pattern 103a as a mask. Specifically, the mold blank 100 after the resist pattern 103a is formed is introduced into a dry etching apparatus, and for example, a hard mask layer made of CrN is dry etched with chlorine gas or a mixed gas containing chlorine gas. And patterning the hard mask layer 102 to form a hard mask layer pattern 102a (see FIG. 4C). Note that the etching end point at this time may be determined by using a reflection optical end point detector.

Subsequently, after the gas used in the above-described etching is evacuated, dry etching using, for example, a fluorine-based gas is performed on the glass substrate 101 using the hard mask layer pattern 102a as a mask in the same dry etching apparatus. Do. At this time, a digging pattern 101a is formed on the glass substrate 101 (see FIG. 4D). The digging depth of the digging pattern 101a is adjusted by the etching time.
The remaining resist pattern 103a is removed using an acid solution such as perhydrosulfuric acid (see FIG. 4D). Note that the resist pattern may be removed before the glass substrate 101 is etched.

  Thereafter, the remaining hard mask layer pattern 102a is removed. Specifically, the mold before removing the remaining hard mask layer pattern (mold in the state shown in FIG. 4D) is introduced into a wet etching apparatus. Then, wet etching is performed with a ceric ammonium nitrate solution to remove the remaining hard mask layer pattern 102a. After removing the remaining hard mask layer pattern 102a, the glass substrate 101 is cleaned as necessary. In this way, a master mold 110 as shown in FIG.

  Note that the etching of the hard mask layer 102 and the glass substrate 101 described above may employ wet etching instead of dry etching. Specifically, in the etching of the hard mask layer 102, a mixed solution of ceric ammonium nitrate solution and perchloric acid may be used as in the removal etching of the remaining hard mask layer. In the etching of the glass substrate 101, when the substrate is quartz, wet etching using hydrofluoric acid may be performed.

Next, manufacturing of the working mold will be described. The working mold is also manufactured using a mold blank as in the case of the master mold. However, the working mold differs from the master mold in that the resist pattern is formed by transferring the concave / convex pattern (digging pattern) of the master mold to the resist layer of the mold blank.
That is, the case where a resist pattern is formed not by electron beam drawing but by pattern transfer from a master mold (original mold) will be described.

As a mold blank for manufacturing a working mold, a predetermined region on the outer peripheral side of the glass substrate is removed using, for example, a photolithography process to prepare a premesa substrate 120 (see FIG. 5B) on which a predetermined pedestal structure is formed. In addition, depending on the use of a working mold, such a base structure may be unnecessary. Moreover, you may form a base structure after mold pattern formation.
Next, a hard mask layer and a resist layer similar to those in the case of the above-described mold blank for manufacturing a master mold are sequentially formed on the surface of the premesa substrate 120.

Next, a master mold 110 (see FIG. 5A), which is an original mold, is placed on the resist layer of the mold blank for manufacturing the working mold. At this time, if the resist layer is liquid, it is only necessary to mount the master mold 110. If the resist layer is solid, the master mold 110 may be pressed against the resist layer to transfer the digging pattern of the master mold 110 to the resist layer.
Thereafter, for example, in the case of optical imprinting, the photocurable resin is cured using an ultraviolet irradiation device, and the transfer pattern shape is fixed to the resist. At this time, irradiation with ultraviolet rays may be performed from the master mold side or from the glass substrate side.
In transferring the pattern, for example, a groove for an alignment mark may be provided on the substrate in order to prevent a transfer failure due to a misalignment between the master mold and the mold blank.

  After the pattern transfer, the master mold 110 is removed from the mold blank for manufacturing the working mold. At this time, a release layer may be formed on the surface of the master mold in advance so that the master mold can be easily removed. The transferred resist pattern may have a residual film unnecessary for etching the hard mask layer, but is removed by ashing using a plasma of a gas such as oxygen or ozone.

As a result, a resist pattern similar to that shown in FIG. 4B is formed. Thereafter, as in the case of the master mold manufacture, the working mold 200 as shown in FIG. 5C is completed by sequentially etching the hard mask layer and the glass substrate.
Even in the production of the working mold described above, working is performed using a mold blank that uses a glass substrate that has an extremely high flatness according to the present invention, reduces minute defects, and has a good surface shape accuracy and quality. By manufacturing the mold, a working mold is obtained in which the dimensional accuracy of the pattern formed on the glass substrate is extremely good and the occurrence of pattern defects is small.

Hereinafter, based on an Example, this invention is demonstrated more concretely.
Example 1
A present Example is a specific example of the manufacturing method of the glass substrate for mask blanks. This example includes the following steps. In the polishing steps of the following examples, comparative examples, and reference examples, the above-described double-side polishing apparatus shown in FIG. 2 was used.

(1) First polishing (rough polishing) process A glass substrate on which a notch mark is not formed in a double-side polishing apparatus using a glass substrate that has been chamfered and ground on the end surface of a synthetic quartz glass substrate (152 mm × 152 mm) 20 sheets (1 carrier, 4 sheets, 5 carriers) were set so that the main surface (the main surface of the glass substrate on which the transfer pattern was formed) was on the upper surface plate side, and rough polishing was performed under the following polishing conditions. A set of 20 sheets was continuously performed 10 times (10 batches), and a total of 200 glass substrates were roughly polished. The processing load and polishing time were adjusted as appropriate.
Polishing liquid: Aqueous solution containing cerium oxide (average particle size 2 to 3 μm) Polishing pad: Hard polisher (urethane pad)
After the polishing step, in order to remove the abrasive grains adhering to the glass substrate, the glass substrate was immersed in a cleaning tank (pure water) (ultrasonic application) and cleaned.

(2) Second polishing (precise polishing) step A glass substrate main surface on which a notch mark is not formed (a glass substrate main surface on the side on which a transfer pattern is formed) on a double-side polishing apparatus using a glass substrate that has finished the first polishing. Was set to 20 on the upper platen side (1 carrier 4 sheets, 5 carriers), and precision polishing was performed under the following polishing conditions. A set of 20 sheets was continuously performed 10 times (10 batches), and a total of 200 glass substrates were precisely polished. The processing load and polishing time were adjusted as appropriate.
Polishing liquid: Aqueous solution containing cerium oxide (average particle size 1 μm) Polishing pad: A buffer layer (a layer for controlling the amount of compressive deformation of the entire polishing pad) on a base material (for example, polyethylene terephthalate), and open on the surface Soft polisher in which a nap layer made of foamed resin having pores is formed (for example, a polishing pad in which the amount of compressive deformation of the polishing pad is 40 μm or more and the 100% modulus of the resin forming the nap layer is 14.5 MPa or more) )
After the polishing step, in order to remove the abrasive grains adhering to the glass substrate, the glass substrate was immersed in a cleaning tank (pure water) (ultrasonic application) and cleaned.

(3) Third polishing (ultra-precision polishing) step The glass substrate after the second polishing is again applied to the double-side polishing apparatus, and the glass substrate main surface on which the notch mark is not formed (the glass substrate main side on which the transfer pattern is formed) Twenty sheets (1 carrier, 4 sheets, 5 carriers) were set so that the surface) was on the upper platen side, and ultraprecision polishing was performed under the following polishing conditions. A set of 20 sheets was continuously performed 10 times (10 batches), and ultra-precision polishing of a total of 200 glass substrates was performed. The processing load and polishing time are appropriately adjusted so as to obtain the surface roughness (RMS (root mean square roughness) of 0.2 nm or less) necessary for a glass substrate used for the binary mask blank for ArF excimer laser exposure. I went.
Polishing liquid: alkaline aqueous solution (pH 10.2) containing colloidal silica
(Colloidal silica content 50 wt%) Average particle diameter: about 100 nm
Polishing pad: Super soft polisher (suede type)
After the polishing step, in order to remove abrasive grains adhering to the glass substrate, the glass substrate is immersed in a cleaning bath containing an aqueous alkaline solution (ultrasonic application), cleaned, and glass for mask blank for ArF excimer laser exposure. A substrate was obtained.

The surface roughness of the main surface of the obtained glass substrate was all good at 0.2 nm or less in RMS.
Further, the flatness of the obtained glass substrate main surface (the main surface of the glass substrate on which the transfer pattern is formed) was measured with a flatness measuring device (UltraFlat 200M manufactured by Corning Tropel). The flatness measurement area was 142 mm × 142 mm.
As a result, a glass substrate with a flatness of 0.3 μm or less was 97.5% of the whole, and a glass substrate with a flatness of 0.2 μm or less was 82.5% of the whole. This is because the main surface of the glass substrate on which the notch mark is not formed (the main surface of the glass substrate on the side where the transfer pattern is formed) is set on the upper platen side, so that the physical properties of the polishing pad vary due to slurry penetration. Is more difficult to receive than the lower surface plate side, so the flatness is considered stable.

  Moreover, the micro defect (concave defect, convex defect) of the glass substrate main surface was measured by the defect inspection apparatus (Lasertec MAGICS6640). As a result, a very good result was obtained, in which the incidence rate of concave defects of 0.1 μm size was 4% and the incidence rate of convex defects of 0.1 μm size was 5%. This is because the main surface of the glass substrate on which the notch mark is not formed (the main surface of the glass substrate on which the transfer pattern is formed) is set on the upper surface plate side. This is thought to be less susceptible to the effects of urethane debris and contaminants inside the polishing pad.

(Comparative Example 1)
In the first embodiment, the glass substrate set in the first polishing (rough polishing) step, second polishing (precision polishing) step, and third polishing (ultra-precision polishing) step is a glass substrate on which notch marks are not formed. A glass substrate for an ArF excimer laser exposure mask blank was obtained in the same manner as in Example 1 except that the main surface (the main surface of the glass substrate on which the transfer pattern was formed) was set to the lower surface plate side.

  In the same manner as in Example 1, the flatness and defects of the main surface of the glass substrate on which the transfer pattern was formed were measured. As a result, the glass substrate with a flatness of 0.3 μm or less was 77% of the whole, and the glass substrate with a flatness of 0.2 μm or less was 56% of the whole. In addition, regarding the defects, the incidence rate of 0.1 μm size concave defects was 12%, and the incidence rate of 0.1 μm size convex defects was 15%. This is because the main surface of the glass substrate on which the notch mark is not formed (the main surface of the glass substrate on which the transfer pattern is formed) is set on the lower surface plate side, so that the physical property variation of the polishing pad due to slurry penetration is large. It is considered that the flatness became unstable, and it was considered that the cause of the defect generation rate deterioration was affected by dust generated by the polishing apparatus and carrier, urethane debris of the polishing pad, and contaminants inside the polishing pad.

(Example 2, Reference Example 1)
In Example 1 described above, for the purpose of monitoring glass substrate production stability (flatness monitoring), the glass substrates that have finished the first polishing (rough polishing) step and the second polishing (precision polishing) step are extracted one by one. , A dummy glass substrate prepared separately for the number of sheets deficient due to extraction (a synthetic quartz glass substrate made of the same material as the other glass substrate, the plate thickness being the same as or smaller than the average plate thickness of the other glass substrate A glass substrate for a mask blank for ArF excimer laser exposure was obtained in the same manner as in Example 1 except that the glass substrate was charged and double-side polishing was performed. In the double-side polishing, a set of 20 sheets was continuously performed 5 times (5 batches) to produce a total of 90 glass substrates (not including a dummy substrate). For reference of Example 2, for the mask blank for ArF excimer laser exposure as in Example 2 except that double-side polishing was performed without inserting a dummy glass substrate for the shortage due to sampling. A glass substrate was obtained.

As a result, in Example 2, the glass substrate with a flatness of 0.3 μm or less is 97.8% of the whole, and the glass substrate with a flatness of 0.2 μm or less is 84.4% of the whole, which is the same as in Example 1. Good results were obtained. On the other hand, in Reference Example 1, the glass substrate with a flatness of 0.3 μm or less is 78.9% of the whole, and the glass substrate with a flatness of 0.2 μm or less is 45.6% of the whole, compared with Example 2. Declined. This is probably because polishing was performed without introducing a dummy glass substrate, so that the amount of polishing in the batch varied and the flatness deteriorated. In addition, with respect to defects, the incidence rate of concave defects of 0.1 μm size in Example 2 was 4.4%, and the incidence rate of convex defects of 0.1 μm size was 5.6%. It was. On the other hand, the 0.1 μm size concave defect occurrence rate of Reference Example 1 was 13.3%, and the 0.1 μm size convex defect occurrence rate was 17.8%, which was worse than that of Example 2. This is probably because the polishing was performed without introducing the dummy glass substrate, so that a sufficient polishing amount could not be secured due to the variation in the polishing amount, and the defect occurrence rate was deteriorated.

(Example 3)
The following are examples of a method for producing a glass substrate for a mask blank for EUV exposure. In Example 1 described above, the first polishing (rough polishing) step and the second polishing were performed in the same manner as in Example 1 except that instead of the synthetic quartz glass substrate, a low thermal expansion glass substrate made of TiO ₂ —SiO ₂ was used. A (precise polishing) step and a third polishing (ultra-precision polishing) step were performed. In addition, the above-mentioned double-side polishing performed 20 sets three times continuously (3 batches) to produce a total of 60 glass substrates.

(4) Local processing process The surface shape (surface form) of the front and back of the glass substrate which finished the 3rd grinding | polishing was measured with the flatness measuring device (UltraFlat200M by Corning Tropel). The measurement area of the surface shape (surface form) was 148 mm × 148 mm. The measurement result of the surface shape of the glass substrate surface is stored in a computer as height information with respect to a reference plane at each measurement point, and compared with the surface flatness specification 50 nm required for a glass substrate for EUV exposure mask blanks. The difference (necessary removal amount) was calculated by a computer.

  Next, the local processing conditions according to the required removal amount were set for each processing spot shape region in the glass substrate surface. Using a dummy substrate in advance, the dummy substrate is spot processed without moving the substrate for a certain period of time in the same way as in actual processing, and the shape is measured with the same measuring machine as the apparatus for measuring the surface shape of the front and back surfaces. Measure and simply calculate the spot processing volume per hour. Then, according to the necessary removal amount obtained from the spot information and the surface shape information of the glass substrate, the scanning speed for raster scanning the glass substrate was determined.

  According to the set processing conditions, the surface shape is formed by local processing so that the flatness of the front and back surfaces of the glass substrate is not more than the above reference value by the MRF (magnetofluidic fluid) processing method using the substrate finishing device with magnetic fluid. It was adjusted. The polishing liquid used was a mixture of cerium oxide and iron powder in a liquid such as water. Thereafter, the glass substrate was immersed in a cleaning tank containing an aqueous hydrochloric acid solution having a concentration of about 10% (ultrasonic application) to perform cleaning.

(5) Touch polishing process About the glass substrate which finished the local process, double-sided touch polishing was performed using the double-side polish apparatus on the conditions which improve surface roughness, maintaining the surface shape of a glass substrate main surface. In this double-sided touch polishing, a set of five sheets (one carrier, one sheet, five sheets) so that the main surface of the glass substrate on which the notch mark is not formed (the main surface of the glass substrate on which the transfer pattern is formed) is the upper surface plate side. Carrier) and touch polishing was performed under the following polishing conditions. A set of 5 sheets was continuously twelve times (12 batches) to perform touch polishing of a total of 60 glass substrates. The processing load and polishing time were adjusted as appropriate. Polishing liquid: Alkaline aqueous solution containing colloidal silica (pH 10.2) (Colloidal silica content 50 wt%) Average particle diameter: about 70 nm Polishing pad: Super soft polisher (Sweet Dype) After touch polishing, adhered to the glass substrate In order to remove the abrasive grains, the glass substrate was immersed (applied with ultrasonic waves) in a cleaning tank containing an alkaline aqueous solution and cleaned to obtain a glass substrate for a mask blank for EUV exposure.

  The surface roughness of the main surface of the obtained glass substrate was all good at Rms of 0.1 nm or less. Further, the flatness of the obtained glass substrate main surface (the main surface of the glass substrate on which the transfer pattern is formed) was measured with a flatness measuring device (UltraFlat 200M manufactured by Corning Tropel). The flatness measurement area was 142 mm × 142 mm.

  As a result, the flatness was 50 nm or less, and very good results were obtained. Moreover, the defect (concave, convex) of the glass substrate main surface was measured with the defect inspection apparatus (Lagertec MAGICS6640). As a result, the 0.1 μm size concave defect occurrence rate was 1.6%, and the 0.1 μm size convex defect occurrence rate was 3.3%. This is because the glass substrate main surface in which notch marks are not formed (the glass substrate main surface on the side on which the transfer pattern is formed) in all the double-side polishing steps of the first polishing step to the third polishing step and the touch polishing step. Is set on the upper surface plate side, so that it is considered that it is less affected by dust generated by the polishing apparatus and carrier, urethane debris of the polishing pad, and contaminants inside the polishing pad.

Example 4
This embodiment is a specific example of a method for manufacturing a binary mask blank and a binary mask. On the synthetic quartz glass substrate obtained in Example 1, a light shielding film composed of a TaN film and a TaO film was formed as follows.
A tantalum (Ta) target is used as the target, and the power of the DC power source is set in a mixed gas atmosphere of xenon (Xe) and nitrogen (N ₂ ) (gas pressure 0.076 Pa, gas flow ratio Xe: N ₂ = 11 sccm: 15 sccm). The TaN film was formed to a thickness of 44.9 nm by reactive sputtering (DC sputtering) at 1.5 kW, and subsequently, using a Ta target, a mixed gas atmosphere of argon (Ar) and oxygen (O ₂ ) ( A TaN film and a TaO film are formed by forming a TaO film with a film thickness of 13 nm with a gas pressure of 0.3 Pa and a gas flow rate ratio of Ar: O ₂ = 58 sccm: 32.5 sccm and a DC power supply of 0.7 kW. A light-shielding film for ArF excimer laser (wavelength 193 nm) composed of the above laminate was formed to prepare a binary mask blank. The optical density of the light-shielding film with respect to ArF excimer laser was 3.0, and the surface reflectance was 19.5%.

  About the obtained binary mask blank, the defect inspection was performed with the blanks defect inspection apparatus (Lagertec MAGICS M1350). As a result, a very good result was obtained, in which the incidence rate of concave defects of 0.1 μm size was 4% and the incidence rate of convex defects of 0.1 μm size was 5%.

Next, a binary mask was produced using this binary mask blank.
First, an electron beam drawing resist was applied on a binary mask blank by a spin coating method and baked to form a resist film.
Next, a mask pattern was drawn on the above resist film with an electron beam and developed to form a resist pattern.

Using this resist pattern as a mask, the TaO film was removed by etching with a fluorine-based gas (CF ₄ gas) and the TaN film was etched with a chlorine-based gas (Cl ₂ gas) to form a light-shielding film pattern.
Further, the resist pattern remaining on the light shielding film pattern was removed with hot sulfuric acid to obtain a binary mask.

(Example 5)
This example is a specific example of a method for manufacturing a mask blank for EUV exposure and a mask for EUV exposure. On the glass substrate for the mask blank for EUV exposure obtained in Example 3 above, an Si film (film thickness: 4.2 nm) and an Mo film (film thickness: 2.8 nm) were formed using an ion beam sputtering apparatus. As one period, a multilayer reflection film was formed by laminating 40 periods to obtain a substrate with a multilayer reflection film.
Next, the multilayer reflective film surface was subjected to defect inspection with a blanks defect inspection apparatus (MAGICS M1350, manufactured by Lasertec Corporation). As a result, a 0.1 μm sized concave defect occurrence rate was 0%, and a 0.1 μm sized convex defect occurrence rate was 5%.

Next, using a DC magnetron sputtering apparatus, a capping layer (film thickness: 2.5 nm) made of RuNb, a TaBN film (film thickness: 56 nm), and a TaBO film (film thickness: 14 nm) are formed on the multilayer reflective film. An absorber layer made of a laminated film was formed, and a CrN conductive film (film thickness: 20 nm) was formed on the back surface to obtain an EUV reflective mask blank.
The obtained EUV reflective mask blank was subjected to defect inspection with a blanks defect inspection apparatus (MAGICSM 1350 manufactured by Lasertec Corporation). As a result, a 0.1 μm size concave defect occurrence rate was 0%, and a 0.1 μm size convex defect occurrence rate was 8.3%.

Next, an EUV reflective mask was produced using the EUV reflective mask blank.
First, an electron beam drawing resist was applied on an EUV reflective mask blank by a spin coating method and baked to form a resist film.
Next, a mask pattern was drawn on the above resist film with an electron beam and developed to form a resist pattern.

Using this resist pattern as a mask, the absorber layer is removed by etching the TaBO film with a fluorine-based gas (CF ₄ gas) and the TaBN film with a chlorine-based gas (Cl ₂ gas) to form an absorber layer pattern on the capping layer. Formed.
Further, the resist pattern remaining on the absorber layer pattern was removed with hot sulfuric acid to obtain an EUV reflective mask.

(Example 6)
This example is a specific example of a method for manufacturing an imprint mold (master mold) for manufacturing a semiconductor device.
Using a synthetic quartz glass substrate (size 152 mm × 152 mm), double-side polishing was performed in the same manner as in Example 1 to obtain a total of 200 glass substrates. The front and back surfaces of the glass substrate are controlled by notch marks, and the main surface of the glass substrate on which the notch marks are not formed (the main surface of the glass substrate on which the transfer pattern is formed) is the upper surface plate side of the double-side polishing apparatus. Polishing was carried out by setting as described above.

The surface roughness of the main surface of the obtained glass substrate was all good at 0.2 nm or less in RMS.
Further, the flatness of the obtained glass substrate main surface (the main surface of the glass substrate on which the transfer pattern is formed) was measured with a flatness measuring device (UltraFlat 200M manufactured by Corning Tropel). The flatness measurement area was 142 mm × 142 mm.
As a result, a glass substrate having a flatness of 0.3 μm or less was 96.5% of the whole, and a glass substrate having a flatness of 0.2 μm or less was 79.5% of the whole, and a very good result was obtained.
Moreover, the micro defect (concave defect, convex defect) of the glass substrate main surface was measured by the defect inspection apparatus (Lasertec MAGICS6640). As a result, a 0.1 μm sized concave defect occurrence rate was 3.5%, and a 0.1 μm sized convex defect occurrence rate was 4.5%.

Next, the synthetic quartz glass substrate obtained as described above was introduced into a sputtering apparatus.
Then, without exposing to the atmosphere, a chromium target was sputtered with a mixed gas of argon and nitrogen (Ar: N ₂ flow rate ratio = 70: 30) to form a CrN layer with a thickness of 2.3 nm. Thereafter, the formed CrN layer was baked at 200 ° C. for 15 minutes in the atmosphere to oxidize the surface side of the CrN layer to form a hard mask layer. Thus, the mold blank of this example was manufactured.

Next, on the hard mask layer made of CrN of this mold blank, an electron beam drawing resist (ZEP520A manufactured by Nippon Zeon Co., Ltd.) was applied to a thickness of 45 nm by spin coating, and baked to form a resist layer.
Next, a desired mold pattern was drawn on the resist layer formed on the hard mask layer 12 using an electron beam drawing machine, and then the resist layer was developed to form a predetermined resist pattern.

Next, the mold blank on which the resist pattern was formed was introduced into a dry etching apparatus, and dry etching using a mixed gas of Cl ₂ and O ₂ was performed. Thus, the hard mask layer was patterned using the resist pattern as a mask to form a pattern on the hard mask layer. Then, the remaining resist pattern was removed using sulfuric acid / hydrogen peroxide (concentrated sulfuric acid: hydrogen peroxide solution = 2: 1 (volume ratio)) composed of concentrated sulfuric acid and hydrogen peroxide solution.

Further, after evacuating the gas used in the dry etching for the hard mask layer, in the same etching apparatus, using the pattern formed in the hard mask layer as a mask, dry etching using a fluorine-based gas (CHF ₃ : Ar = 1: 9 (volume ratio)), the glass substrate was dug by etching. Thus, the glass substrate was etched to form a predetermined digging pattern.

Finally, the pattern of the hard mask layer remaining on the glass substrate was removed by wet etching using cerium nitrate cerium nitrate.
The imprint mold (master mold) for manufacturing a semiconductor device of this example was manufactured through the above steps by appropriately performing a cleaning process, a drying process, and the like.

(Example 7)
This example is a specific example of a method for producing an imprint mold (master mold) for producing BPM.
In this example, a disc-shaped synthetic quartz glass substrate having an outer diameter of 150 mm and a thickness of 0.7 mm was used as the substrate. Double-side polishing was performed in the same manner as in Example 1 to obtain a total of 200 glass substrates. The front and back surfaces of the glass substrate are managed by forming a concave mark on the surface (back surface) opposite to the first surface (front surface) on which the transfer pattern is formed, and the glass on which the concave mark is not formed. Polishing was performed by setting the substrate main surface (the main surface of the glass substrate on the side where the transfer pattern is formed) to be on the upper surface plate side of the double-side polishing apparatus. In addition, since the disk-shaped substrate was used in this example, the carrier in the double-side polishing apparatus used in the polishing process was appropriately adjusted according to the substrate shape and thickness.

The surface roughness of the main surface of the obtained glass substrate was all good at 0.2 nm or less in RMS.
Further, the flatness of the obtained glass substrate main surface (the main surface of the glass substrate on which the transfer pattern is formed) was measured with a flatness measuring device (UltraFlat 200M manufactured by Corning Tropel). The flatness measurement area was 142 mm × 142 mm.
As a result, a glass substrate with a flatness of 0.3 μm or less was 90.5% of the whole, and a glass substrate with a flatness of 0.2 μm or less was 78.0% of the whole.
Moreover, the micro defect (concave defect, convex defect) of the glass substrate main surface was measured by the defect inspection apparatus (Lasertec MAGICS6640). As a result, a 0.1 μm size concave defect occurrence rate was 4.5%, and a 0.1 μm size convex defect occurrence rate was 7.0%.

Next, the disc-shaped synthetic quartz glass substrate obtained as described above was introduced into a sputtering apparatus, and a hard mask layer made of CrN was formed on the glass substrate in the same manner as in Example 6. Thus, the mold blank of this example was manufactured.
Using this mold blank, a predetermined digging pattern was formed on the glass substrate through the same steps as in Example 6 to manufacture an imprint mold (master mold) for producing BPM of this example.

DESCRIPTION OF SYMBOLS 1 Glass substrate 2 Notch mark 10 Lower surface plate 11 Polishing pad 20 Upper surface plate 21 Polishing pad 30 Sun gear 40 Internal gear 50 Carrier 60 Polishing liquid supply part 5 Upper surface plate 6 Lower surface plate 7 Polishing pad 100 Imprint mold blank 101 Glass Substrate 102 Hard mask layer (thin film)
110 Master mold 120 Pre-mesa substrate 200 Working mold

Claims (9)

While sandwiching a plurality of glass substrates between an upper surface plate and a lower surface plate provided facing the top and bottom with the polishing pad attached, while supplying polishing liquid from the upper surface plate side to the glass substrate, A method for producing a glass substrate for a mask blank in which both main surfaces of the glass substrate are polished on both sides,
The shape of the glass substrate is a quadrangle, and at least a notch mark is formed at the corner portion of the quadrangle by cutting off three surfaces of one main surface and two end surfaces forming the corner portion into an oblique cross section. Having one or more,
The method for producing a glass substrate for a mask blank, wherein the double-side polishing of the glass substrate is performed by setting one main surface on which the notch mark is not formed on the upper surface plate side. While sandwiching a plurality of glass substrates between an upper surface plate and a lower surface plate provided facing the top and bottom with the polishing pad attached, while supplying polishing liquid from the upper surface plate side to the glass substrate, A method for producing a glass substrate for a mask blank in which both main surfaces of the glass substrate are polished on both sides,
The two main surfaces have a first surface on which a transfer pattern is formed, and a second surface provided to face the first surface;
The method for producing a glass substrate for a mask blank, wherein the double-side polishing of the glass substrate is performed by setting the first surface on the upper surface plate side.   The method for producing a glass substrate for a mask blank according to claim 1 or 2, wherein the double-side polishing is performed in a plurality of stages of polishing steps with different polishing liquids.   The method for manufacturing a glass substrate for a mask blank according to claim 3, wherein in the plurality of stages of polishing steps, the number of glass substrates to be double-side polished is the same.   The said multiple polishing process WHEREIN: When the shortage of the number of the glass substrates which carry out the double-side polishing generate | occur | produces, the dummy glass substrate prepared separately is thrown in and the next grinding | polishing process is performed. A method for producing a mask blank glass substrate.   6. The method for manufacturing a glass substrate for a mask blank according to claim 5, wherein the thickness of the dummy glass substrate is the same as or smaller than the average thickness of other glass substrates.   A method for producing a mask blank, comprising: forming a thin film to be a transfer pattern on the main surface of the glass substrate produced by the method for producing a glass substrate for mask blank according to claim 1. .   A method for manufacturing a mask, comprising: patterning the thin film of a mask blank manufactured by the method for manufacturing a mask blank according to claim 7. Forming an imprint mold blank by forming a thin film to be a transfer pattern on the main surface of the glass substrate manufactured by the method for manufacturing a glass substrate for a mask blank according to any one of claims 1 to 6; ,
Patterning the thin film of the manufactured imprint mold blank to form a pattern of the thin film; and
And a step of etching the glass substrate by an etching process using the thin film pattern as a mask to form a digging pattern on the glass substrate.

Images