JP5597059B2 - Mask blank substrate, mask blank and transfer mask manufacturing method - Google Patents

Description

  The present invention relates to a method for manufacturing a mask blank substrate, and a mask blank using the mask blank substrate and a method for manufacturing a transfer mask.

  In general, in a manufacturing process of a semiconductor device, a fine pattern is formed using a photolithography method. Further, a number of substrates called transfer masks are usually used for forming this fine pattern. This transfer mask is generally a transparent mask blank substrate provided with a fine pattern made of a metal thin film or the like. When a fine pattern is provided on a metal thin film or the like of a transfer mask, a photolithography method is used. The transfer mask is usually made from a mask blank in which a light shielding film is formed on a transparent mask blank substrate. In order to obtain a transfer mask having a fine pattern, the mask blank substrate is required to have high flatness and high parallelism. In addition, if there are defects (scratches, foreign matter, etc.) on the mask blank substrate, defects will also exist in the transfer mask manufactured using them, causing defects when forming a fine pattern. End up. Therefore, the mask blank substrate is required to be free from defects.

  As a conventional method for manufacturing a mask blank substrate, for example, in Patent Document 1, a surface plate on which a polishing cloth is pasted and a mask blank substrate are relatively moved in a state where a constant load is applied. A method for producing a mask blank substrate, comprising a step of polishing the mask blank substrate by supplying a polishing solution between the contact surfaces of the polishing cloth and the mask blank substrate, the polishing cloth pasting surface of the surface plate Discloses a method of manufacturing a mask blank substrate having a groove having a predetermined shape.

  Further, Patent Document 2 discloses an electronic device substrate, which is an electronic device substrate, and the flatness of the main surface of the substrate is more than 0 μm and 0.25 μm or less. Patent Document 2 discloses a polishing apparatus having a predetermined polishing surface plate, a polishing pad, an abrasive supply unit, and a substrate pressurizing unit as an apparatus for manufacturing an electronic device substrate having a predetermined flatness. ing.

  Patent Document 3 discloses a method for correcting dimensions of a processed quartz glass product. Specifically, Patent Document 3 discloses that a quartz glass processed product whose dimensions are to be corrected is arranged at a predetermined distance from a quartz glass welding rod, and a combustible gas flame of a burner is disposed at an end of the quartz glass welding rod. The quartz glass welding rod is radiated to heat and sublimate the quartz to form gaseous quartz, and the gaseous quartz is blown in the direction of the burner's combustible gas flame to correct the size of the processed quartz glass product. After depositing gaseous quartz on the surface, the opaque quartz deposited layer deposited and deposited is heated and welded to form a transparent quartz glass layer, and the size of the processed quartz glass product is corrected. A method for correcting dimensions of a processed quartz glass product is described.

JP 2007-54944 A Japanese Patent Laid-Open No. 2004-29735 JP 2002-326828 A

  In order to finish a mask blank substrate with high flatness, high smoothness, and no defects such as scratches, a lapping (grinding) process and a polishing (polishing) process are usually performed in a plurality of stages. The lapping process is performed for the purpose of uniformizing the processing strain layer on the main surface, adjusting it to a predetermined thickness, and improving the flatness. Also, the polishing process aims to further improve the smoothness of the main surface of the substrate while maintaining and improving the flatness obtained by the lapping step, and to remove particles adhering to the main surface of the substrate. Done. In the polishing process, polishing is performed in a plurality of stages while gradually reducing the grain size of the abrasive grains. For example, after polishing with cerium oxide abrasive grains having an average particle diameter of 0.3 to 3 μm, precision polishing is performed with abrasive grains made of colloidal silica having an average particle diameter of 300 nm or less. Thereafter, in the inspection process, the presence or absence of surface defects on the mask blank substrate is inspected.

  In recent years, due to the demand for further miniaturization of semiconductor devices, the demand for parallelism, flatness, surface roughness and surface defects of a mask blank substrate has also increased. In particular, with respect to surface defects, an allowable defect size is determined by the grade of the mask blank manufactured using the mask blank substrate, and the presence of surface defects larger than the predetermined size is not allowed. In the case of mask blank substrates used to produce state-of-the-art semiconductor devices, the size of surface defects that can be tolerated is very severe.

  On the other hand, in the case of a large mask blank substrate (330 mm × 450 mm or more) used for a liquid crystal display device (thin film transistor: TFT, etc.) or a mask for manufacturing a color filter (CF), the main surface compared to a mask blank for a semiconductor device May have a large area, and the parallel surface, flatness, and surface roughness accuracy requirements of the main surface are relatively loose. Also, the size of acceptable surface defects is relatively loose. However, in recent years, TFT patterns have been miniaturized, and demands for parallelism, flatness, surface roughness, and surface defect-freeness are increasing even in large mask blanks. In many cases, the surface defect is caused by foreign matter entering between the polishing cloth on the surface plate of the polishing apparatus and the main surface of the substrate. However, even if measures are taken to suppress the entry of foreign matter from the outside, the foreign matter produced by solidification of the abrasive grains can also be a cause of surface defects, so that no surface defects are generated during the polishing process. It is difficult to do. In the case of a mask blank substrate having a large area, the influence of the increase in manufacturing cost due to disposal when surface defects are present is larger than that in the case of a mask blank substrate having a small area.

  The mask blank substrate polished in the polishing process is subjected to defect inspection such as surface defects in the inspection process. In this inspection process, the mask blank substrate in which a surface defect having a size larger than the allowable value is found on the main surface of the substrate is usually returned to the polishing process as a rejected product, and the main surface is polished again. . Surface defects are broadly divided into convex defects and concave defects. In the polishing for removing the concave defects, in order to maintain the parallelism and flatness, the main surface must be removed not locally but entirely, so that the necessary polishing allowance increases. In addition, if the mask blank substrate has been returned to the precision polishing stage many times and becomes thinner than the specified thickness, it cannot be used as a mask blank substrate and discarded. Is done.

  Patent Document 3 describes a method for correcting a size of a processed quartz glass product using a technique of depositing and depositing gaseous quartz on a portion whose size is to be corrected. However, if the technique of depositing and depositing gaseous quartz is applied to the concave defect portion of the mask blank substrate, quartz is also deposited around the concave defect, and it is unavoidable to adversely affect flatness and surface roughness. Absent. This is because the size of the concave defect is very small in many cases on the order of μm. This necessitates a polishing step for removing the excessively deposited quartz. Concave defects are caused by foreign particles such as solidified abrasive grains entering between the polishing cloth on the surface plate and the main surface of the substrate that are rotating in a state where processing pressure is applied in the polishing process. Therefore, it often occurs by damaging the main surface. There is a problem in that a risk of occurrence of a new surface defect is generated on the mask blank substrate by performing a polishing process for removing excess deposited quartz. Furthermore, the concave defect has a very small width, and it is very difficult to deposit gaseous quartz to the bottom. Therefore, it is difficult to use the invention disclosed in Patent Document 3 in order to correct the surface defect of the mask blank substrate.

  Therefore, even if a surface defect occurs in the mask blank substrate, the present invention can easily correct the surface defect of the mask blank substrate without adversely affecting the flatness and surface roughness of the substrate. An object of the present invention is to provide a low-cost and high-yield manufacturing method of a mask blank substrate, and a manufacturing method of a mask blank and a transfer mask.

  The inventor examined various methods for correcting a surface defect of a mask blank substrate. As a result, it has been found that the surface defect of the mask blank substrate can be corrected by bringing a flame having a combustion temperature equal to or higher than the softening point of the substrate into contact, and the present invention has been completed. That is, in order to solve the above problems, the present invention has the following configuration.

  The present invention is a method for manufacturing a mask blank substrate having the following configurations 1 to 4, a method for manufacturing a mask blank having the following configuration 5, and a method for manufacturing a transfer mask having the following configuration 6.

(Configuration 1)
A mask blank substrate manufacturing method comprising a substrate having two opposing main surfaces, the step of polishing the two main surfaces of the substrate, and the surface defects present on the main surface of the substrate, And a step of correcting the surface defect by bringing a flame having a combustion temperature equal to or higher than a softening point of the substrate into contact.

(Configuration 2)
2. The method for manufacturing a mask blank substrate according to Configuration 1, wherein the surface defect is a concave defect.

(Configuration 3)
3. The method for manufacturing a mask blank substrate according to Configuration 1 or 2, wherein the substrate is made of synthetic quartz glass.

(Configuration 4)
4. The method for manufacturing a mask blank substrate according to Configuration 3, wherein the flame has a combustion temperature of 1600 ° C. or higher.

(Configuration 5)
A mask comprising a step of forming a thin film for pattern formation on a main surface of a mask blank substrate manufactured by the method for manufacturing a mask blank substrate according to any one of configurations 1 to 4 It is a manufacturing method of a blank.

(Configuration 6)
It is a manufacturing method of the transfer mask characterized by having the process of forming a transfer pattern in the thin film of the mask blank manufactured with the manufacturing method of the mask blank of the structure 5.

  Next, configurations 1 to 4 of the method for manufacturing a transfer mask of the present invention will be described.

  The present invention is a method for producing a mask blank substrate comprising a substrate having two opposing main surfaces as in Configuration 1, and a step of polishing the two main surfaces of the substrate (polishing step); And a step (surface defect correction step) of correcting the surface defect by bringing a flame having a combustion temperature equal to or higher than the softening point of the substrate into contact with a surface defect existing on the main surface of the substrate. It is a manufacturing method of a mask blank substrate.

  As in Configuration 1, in the mask blank substrate manufacturing method of the present invention, a surface defect is corrected by bringing a flame having a combustion temperature equal to or higher than the softening point of the substrate into contact with a surface defect existing on the main surface of the substrate. To do. The “substrate softening point” is the softening point of the substrate material constituting the substrate. When the viscosity of the substrate material is η (poise), the viscosity satisfying log η = 7.6 is satisfied. The temperature at which the rate is reached. Note that the viscosity (viscosity) can be measured using an intrusion method. If necessary, the viscosity can be measured using a measuring method such as a beam bending method or a ball pulling method. Further, the “main surface” of the substrate refers to the surface excluding the peripheral edge of the substrate (the end surface 72 and the chamfered surface 73) as illustrated in FIG. That is, the “main surface” of the substrate refers to a surface shown as two “main surfaces 71” facing each other in FIG.

  The present inventor softens the substrate material of the surface defect portion when a flame having a combustion temperature equal to or higher than the softening point of the substrate is brought into contact with a portion where the surface defect of the mask blank substrate exists (surface defect portion). It has been found that the substrate material can be caused to flow in the direction of correcting the surface defect portion, and as a result, the surface defect can be corrected. In the present specification, this surface defect correction method is referred to as “flame defect correction”. In the manufacturing method of the present invention, it is not necessary to attach a substrate material such as synthetic quartz glass to the surface defect portion, but by bringing a flame having a predetermined combustion temperature or more locally into contact with the mask blank substrate, the surface Planarization can be achieved by slightly flowing the substrate material around the defect. For example, when a concave defect exists as a surface defect, the substrate material can be planarized by flowing so as to fill the surface defect portion. Further, when a convex defect exists as a surface defect, the substrate material can be planarized by flowing so as to spread from the surface defect portion.

  By using flame defect correction, the area of the main surface that is affected by the correction of the surface defect portion can be minimized. In addition, the flatness change and the surface roughness change are slight even in the part affected by the flame defect correction. Therefore, the number of repetitions of the polishing process for correcting the flatness after correcting the flame defect can be reduced. That is, according to the manufacturing method of the present invention, the surface defect of the mask blank substrate can be corrected easily and without adversely affecting the flatness and surface roughness of the substrate. In particular, in the case of a mask blank substrate for FPD manufacturing, which has relatively loose restrictions on flatness and surface roughness, a predetermined specification can be satisfied without performing a polishing step after correcting a flame defect for a surface defect. There is a case. In addition, even in the case of a mask blank substrate for semiconductor device manufacturing applications, it is possible to satisfy predetermined specifications by locally performing non-contact polishing in which concave defects are less likely to occur on locations where flame defect correction has been performed. Sometimes you can. Applicable non-contact polishing includes, for example, float polishing, EEM, hydroplane polishing described in JP-A No. 2004-291209.

(Configuration 2)
Further, as described in Structure 2, in the method for manufacturing a mask blank substrate of the present invention, it is preferable that the surface defect is a concave defect.

  As in Structure 2, it is preferable that the method for manufacturing a mask blank substrate of the present invention is performed on a concave defect. The conventional method for correcting a concave defect requires a large polishing allowance when re-polishing the entire main surface, whereas the method for correcting a concave defect in the present invention does not require a polishing allowance. This is because it can be very small.

  Further, as described in Structure 3, in the method for manufacturing a mask blank substrate of the present invention, it is preferable that the substrate is made of synthetic quartz glass.

  As in Structure 3, it is preferable to use synthetic quartz glass as the substrate material. Synthetic quartz glass is chemically stable and has features such as extremely low thermal expansion coefficient compared to other industrial materials. In addition, synthetic quartz glass has high light transmittance with respect to exposure light of an ultrahigh pressure mercury lamp used in a transfer mask for FPD manufacturing. Furthermore, synthetic quartz glass has high light transmittance with respect to exposure light of a KrF excimer laser (wavelength: about 248 nm) or ArF excimer laser (wavelength: about 193 nm) used in a transfer mask for semiconductor device manufacturing applications. As can be seen from these, synthetic quartz glass can be suitably used as the material for the mask blank substrate.

  In the method for producing a mask blank substrate according to the present invention, as set forth in Structure 4, it is preferable that the flame has a combustion temperature of 1600 ° C. or higher.

  As in Configuration 4, the flame preferably has a combustion temperature of 1600 ° C. or higher. Although the softening point of synthetic quartz glass is generally 1600 ° C. or higher depending on the type, the synthetic quartz glass around the surface defects is slightly flowed by using a flame having a combustion temperature of 1600 ° C. Defects can be planarized. In order to ensure the flow of the synthetic quartz glass, the flame is preferably a combustion temperature of 1700 ° C. or higher, and more preferably a combustion temperature of 1800 ° C. or higher. The upper limit of the flame combustion temperature is preferably a temperature that does not adversely affect the synthetic quartz glass, and is, for example, 2800 ° C. or lower, preferably 2500 ° C. or lower, more preferably 2300 ° C. or lower.

  Next, the structure 5 of the manufacturing method of the mask blank of this invention is demonstrated.

  The present invention includes a step of forming a thin film for pattern formation on the main surface of the mask blank substrate manufactured by the method for manufacturing a mask blank substrate according to any one of the above configurations 1 to 4 as in the configuration 5. It is a manufacturing method of the mask blank characterized by having.

  As described in Configuration 5, in the mask blank substrate manufacturing method according to any one of Configurations 1 to 4, low-cost and high-yield mask blank substrate is obtained by easily correcting surface defects by correcting flame defects. Can be obtained. Therefore, a mask blank in which a thin film for pattern formation is formed on any one of the mask blank substrates of configurations 1 to 3 can also be manufactured at low cost and high yield. The formation of the pattern forming thin film on the mask blank substrate can be performed using a known method.

  Next, the structure 6 of the manufacturing method of the transfer mask of this invention is demonstrated.

  This invention is the manufacturing method of the transfer mask characterized by having the process of forming a transfer pattern in the thin film of the mask blank manufactured with the manufacturing method of the mask blank of the said structure 5, as it exists in the structure 6. is there.

  As in Configuration 6, the mask blank manufacturing method of Configuration 5 according to the present invention can achieve low cost and high yield. Therefore, the transfer mask in which the transfer pattern is formed on the thin film of the mask blank having the structure 5 can also be manufactured at low cost and high yield. The transfer pattern can be formed on the mask blank thin film using a known method.

  According to the present invention, even when a surface defect occurs on the mask blank substrate, the surface defect of the mask blank substrate can be corrected easily and without adversely affecting the flatness and surface roughness of the substrate. A low-cost and high-yield manufacturing method of a mask blank substrate and a manufacturing method of a mask blank and a transfer mask can be provided.

It is a cross-sectional schematic diagram which shows an example of a mode that flame defect correction is carried out to the mask blank. It is a figure which shows the process of the manufacturing method of the board | substrate for mask blanks of this invention. It is a schematic diagram which shows the board | substrate for mask blanks which can be used for the manufacturing method of this invention. It is a schematic block diagram of an example of the flame defect correction apparatus for performing a flame defect correction, observing the place which is correcting the flame defect, using an optical microscope. It is a cross-sectional schematic diagram which shows the process of manufacturing the transfer mask using a mask blank. It is a schematic diagram which shows an example of a double-side polish apparatus. It is a schematic diagram which shows an example of a single-side polishing apparatus.

<Description of Method for Manufacturing Mask Blank Substrate>
Hereinafter, the manufacturing method of the mask blank substrate 1 of the present invention will be described. Hereinafter, although the case where synthetic quartz glass is used as a material for the mask blank substrate 1 will be described as an example, the present invention is not limited thereto.

  The manufacturing method of the mask blank substrate 1 of the present invention includes the steps shown in FIG. 2, that is, the cutting step (S101), the lapping step (S102), the polishing step (S103), the inspection step (S104), and the surface defect correction step ( S105).

  In addition, between each process such as between the polishing process (S103) and the inspection process (S104), the polishing abrasives adhered to the substrate are not brought into the next process, for example, abrasive grains used in the previous process. A cleaning step for removing grains and the like can be appropriately provided.

(1) Cutting process (S101)
First, a sheet-shaped synthetic quartz glass (sheet glass) is cut out from a synthetic quartz glass ingot. Next, a rectangular substrate 1 can be obtained by cutting out from a sheet glass made of synthetic quartz glass with a grinding wheel. The rectangular dimensions vary depending on the application, but in the case of a large mask blank substrate 1, for example, 330 mm × 450 mm or more, for example, 500 mm × 800 mm, thickness 10 mm, 800 mm × 920 mm, thickness 10 mm. A shape of 1220 mm × 1400 mm, a thickness of 13 mm, a shape of 2140 mm × 2460 mm, a thickness of 15 mm, or the like can be used. Hereinafter, the processing conditions for each step will be described for a substrate shape of 800 mm × 920 mm and a thickness of 10 mm.

(2) Wrapping process (S102)
Next, a lapping process is performed on the substrate 1. The lapping process can include three processes: a rough lapping process, a shape processing process, and a fine lapping process.

  The rough lapping process aims to improve dimensional accuracy and shape accuracy. A rough lapping process is performed using a double-sided lapping apparatus. For example, the grain size of the abrasive grains can be # 400. As a specific example, alumina abrasive grains having a particle size of # 400 can be used, the load L can be set to about 100 kg, and both surfaces of the substrate 1 stored in the carrier can be finished with a predetermined accuracy.

  Next, the shape of the substrate 1 is processed by a shape processing step. Specifically, after the outer peripheral end face is ground and the external dimensions are adjusted, a predetermined chamfering process can be performed on the outer peripheral end face and the inner peripheral face.

  Next, the two main surfaces 71 of the substrate 1 are finished to a predetermined accuracy by a fine lapping process. Specifically, by changing the grain size of the abrasive grains to # 1000 and lapping the main surface 71 of the substrate 1, both surfaces of the substrate 1 can be finished with a predetermined accuracy. In order not to bring abrasive grains or the like into the next step, it is preferable to wash the substrate 1 after the fine lapping step by sequentially immersing it in each washing tank of neutral detergent and water.

(3) Polishing step (S103)
Next, the polishing step (S103) removes scratches and distortions formed during the lapping step while maintaining the flatness of the substrate 1 obtained in the lapping step (S102). The polishing step is further performed for the purpose of mirroring the main surface 71 of the substrate 1. Usually, in the polishing step, removal of scratches and distortion and mirror polishing are performed by separate polishing. Therefore, it is common to perform polishing several times in the polishing step.

  The polishing step (S103) can be performed by a known method. The polishing step can be performed, for example, using a double-side polishing apparatus for a large mask blank shown in FIG. In the double-side polishing apparatus for a large mask blank shown in FIG. 6, the distance between the rotation axis a of the upper and lower surface plates (22, 23) and the rotation axis b of the substrate 20 is that the substrate 20 rotates about the rotation axis b. Even in this case, it is preferable that any part of the substrate 20 does not pass through the part of the rotation axis a. In addition, about this grinding | polishing process, you may apply the grinding | polishing process etc. which use not only this but the double-side polish apparatus for large sized mask blanks described in Unexamined-Japanese-Patent No. 2010-76047.

(4) Inspection process (S104)
Next, in the inspection step (S104), the flatness and surface roughness of the substrate 1 obtained in the polishing step (S103), the presence / absence of the surface defect 5, and the position thereof are inspected. Although the substrate 1 obtained by the above-described steps (S101 to S103) has high flatness and high smoothness, it is known that surface defects 5 are present not often. The allowable size of the surface defect 5 varies depending on the use of the mask blank manufactured from the mask blank substrate 1 and the transfer mask. The surface defect 5 can be measured for size, shape and position by a surface inspection apparatus using an optical microscope. In addition, laser scattered light type defect inspection devices and confocal optical defect inspection devices can also be applied. For example, in the case of a mask blank substrate for use in FPD manufacturing, it is often determined that the concave surface defect 5 is not allowed to have a width larger than 1 μm. In this inspection process, as a result of inspecting the substrate 1, the main surface 71 has a flatness and surface roughness of a predetermined level or more, and the detected surface defect 5 having no unacceptable size is a good mask. The operation of selecting the blank substrate 1 is first performed. The non-defective mask blank substrate 1 is sent to the next step (the step of forming a thin film for pattern formation is performed after cleaning).

  Next, the substrate 1 in which at least one of the flatness and the surface roughness of the main surface 71 does not satisfy the predetermined value is returned to the polishing step at a correctable stage and re-polished. On the other hand, for the substrate 1 in which both the flatness and the surface roughness satisfy the predetermined conditions but the surface defect 5 having an unacceptable size exists, the type (concave defect, convex defect), position, and size of the surface defect 5 After the depth (or height) is specified, it is sent to the surface defect correction process.

(5) Surface defect correction step (S105)
With respect to the substrate 1 on which the surface defect 5 having an unacceptable size exists, it is first determined whether the surface defect 5 should be performed in the surface defect correction step (S105) or returned to the polishing step. When the surface defect 5 is a concave defect by the flame defect correction, the surface defect 5 can be corrected while maintaining the flatness of the main surface up to a depth of about 0.5 μm. However, in the case of a concave defect deeper than that, deterioration of flatness is inevitable. Therefore, the substrate 1 having the concave defect having such a depth is returned to the polishing process or discarded. In the case where a very high flatness is required for the main surface 71 of the mask blank substrate and non-contact polishing is not performed after the surface defect correction process, the surface defects 5 that can be corrected by flame defect correction The depth is desirably up to about 0.1 μm. And this surface defect correction process is performed only with respect to the board | substrate 1 determined that it is desirable to correct a flame defect. In the manufacturing method of the mask blank substrate 1 of the present invention, in the surface defect correction step, the surface defect 5 present on the main surface 71 of the mask blank substrate 1 is higher than the softening point of the material of the mask blank substrate 1. The method is characterized in that the surface defect 5 is corrected by a method (flame defect correction) in which the flame 52 at the combustion temperature is brought into contact.

  FIG. 1 shows a schematic cross-sectional view of an example of a state in which a surface defect 5 of the substrate 1 is corrected for a flame defect. The surface defect 5 of the substrate 1 is corrected by the flame 52 being heated by the flame 52 irradiated from the flame irradiation device 50. By correcting the surface defect 5 with a flame defect, the mask blank substrate 1 without the surface defect 5 having an unacceptable size can be obtained. By setting the flame 52 to an appropriate size when correcting the flame defect, it is possible to correct only the surface defect 5 while suppressing the substrate 1 from being heated more than necessary. Further, since there is no need for a large-scale facility for correcting the flame defect, the flame defect of the surface defect 5 can be corrected easily and at low cost.

  As the flame irradiation device 50 for correcting the flame defect, a burner or the like that generates the flame 52 by burning flammable gas can be used. The flammable gas used in the flame irradiation apparatus 50 is preferably selected from hydrocarbons such as methane, propane, butane and acetylene, hydrogen, and a mixed gas of oxygen and hydrogen. Since flame defect correction can be performed at a higher temperature, it is more preferable to use a flammable gas which is a mixed gas of butane or hydrogen and oxygen.

  The distance between the flame irradiation device 50 and the substrate 1 at the time of correcting the flame defect can be appropriately selected within a range in which the generation of the flame 52 is not hindered and the contact of the flame 52 with the surface defect 5 is effectively performed. . The distance between the flame irradiation device 50 and the surface defect 5 at the time of correcting the flame defect is preferably 1 mm to 50 mm from the viewpoint of effectively performing the contact of the flame 52 with the surface defect 5 without being obstructed. More preferably, it is set to ˜20 mm.

  In order to correct only the surface defect 5 while suppressing the heating of the substrate 1, by using a device that generates a relatively small flame 52 and irradiating the flame 52 locally to the place where the surface defect 5 exists It is preferable to correct the flame defect of the surface defect 5. When performing local flame defect correction, the irradiation diameter of the flame 52 with respect to the surface defect 5 is preferably 0.1 mm to 10 mm, more preferably 0.2 mm to 5 mm, and 0.5 mm to 4 mm. More preferably.

  In the flame defect correction used in the manufacturing method of the present invention, the temperature of the portion of the flame 52 varies depending on the softening point of the material of the substrate 1. When the material of the substrate 1 is synthetic quartz glass, the temperature of the flame 52 is preferably 1600 ° C. or higher, more preferably 1700 ° C. or higher, and further preferably 1700 ° C. or higher. By bringing the flame 52 in such a temperature range into contact with a local portion of the substrate 1 where the surface defect 5 exists, the temperature of the portion where the surface defect 5 exists becomes approximately the same as the temperature of the portion of the flame 52. As a result, the substrate 1 is locally softened and the flame defect correction of the surface defect 5 can be performed. The portion of the substrate 1 where the flame defect is corrected tends to change the OH group concentration distribution as compared with the other portions.

  In the manufacturing method of the present invention, it is preferable to irradiate the flame 52 while observing the surface defect 5 with the optical microscope 108 when correcting the flame defect. FIG. 4 shows a schematic configuration diagram of a flame defect correcting apparatus 100 for correcting a flame defect while observing the surface defect 5 with the optical microscope 108. In addition, as irradiation light for observation, the intensity | strength sufficient for microscope observation can be irradiated with respect to the board | substrate 1 by using the irradiation light which uses a condensing lamp etc. (not shown) as a light source. The observation light 126 from the surface defect 5 enters the optical microscope 108, and an optical microscope image of the surface defect 5 can be obtained. By providing the image processing unit 110 for identifying the surface defect 5 from the optical microscope image, it is possible to automate the detection of the presence, size, and position of the surface defect 5. If a flame defect correcting apparatus 100 as shown in FIG. 4 is used, the surface defect 5 of the substrate 1 is irradiated with flame and then observed with an optical microscope 108 to evaluate the degree of correction of the surface defect 5 by correcting the flame defect. can do. Therefore, if the flame defect correction is insufficient, a process with feedback such as further flame defect correction can be performed, so that the surface defect 5 can be reliably corrected.

  When there are a plurality of surface defects 5 on the substrate 1, the substrate 1 can be moved using the stage controller 112 and the stage 102 so that the surface defects 5 become the irradiation position of the flame 52. By using the flame defect correcting apparatus 100 provided with the control unit 114 connected to the flame control unit 104, the image processing unit 110, and the stage controller 112 for controlling the flame 52 by controlling the flow rate of the flammable gas 105 and the like. The flame defect correction can be automatically performed. Further, the position where the flame defect correction is performed can be specified from the information on the dimension, shape and position measured by the surface inspection apparatus used in the inspection process. Further, when the flame defect correcting apparatus 100 includes the stage controller 112 and the stage 102 that can automatically move, the flame irradiation apparatus is obtained by obtaining information such as the position where the surface defect 5 exists from the surface inspection apparatus. The movement for correcting 50 flame defects can be performed automatically.

  In addition, when the flatness and surface roughness of the substrate main surface do not satisfy the predetermined range due to the influence of the flame defect correction, non-contact polishing that hardly causes a concave defect is performed locally, Flatness and surface roughness can also be achieved. Applicable non-contact polishing includes float polishing, EEM, hydroplane polishing, and the like. In this case, a small particle size colloidal silica is suitable for the abrasive grains. In particular, it is effective for a mask blank substrate for use in manufacturing semiconductor devices in which the conditions of flatness and surface roughness of the substrate main surface are severe.

<Description of mask blank substrate>
It is preferable that the mask blank substrate 1 obtained by the production method of the present invention has a high flatness according to its use and further has a high parallelism. For example, the mask blank substrate 1 includes a pair of main surfaces 71 provided opposite to each other as shown in FIG. 3, two sets of end faces 72 orthogonal to the main surfaces, and the main surfaces and end faces. A square (rectangular) substrate having a chamfered surface 73 sandwiched between, the flatness of the main surface (preferably both main surfaces) of the mask blank substrate 1 (in the main surface of the mask blank substrate 1, (This means that the chamfered surface 73 is inevitably removed.) On the small mask blank substrate 1 (less than 330 mm × 450 mm), it exceeds 0 μm and 5 μm. Hereinafter, the large mask blank substrate 1 (330 mm × 450 mm or more) is a substrate having a high flatness exceeding 0 μm and not more than 30 μm.

  Further, both main surfaces 71 of the mask blank substrate 1 are mirror-finished by precision polishing, and the surface roughness is finished to an average surface roughness Ra of 0.3 nm or less. The surface roughness of the main surface 71 is preferably smaller in terms of defect detection and film surface uniformity after film formation, and Ra is 0.2 nm or less, more preferably 0.15 nm or less. Preferably it is. Further, the end face 72 and the chamfered surface 73 of the mask blank substrate 1 are preferably mirror-finished by brush polishing or the like from the viewpoint of preventing the generation of particles, and the surface roughness is an average surface roughness Ra. It is desirable that the thickness be 0.3 nm or less, more preferably 0.2 nm or less, and still more preferably 0.15 nm or less.

<Description of Transfer Mask and Transfer Mask>
Hereinafter, an embodiment of a method for manufacturing a mask blank 10 and a transfer mask according to the present invention will be described.

  The present invention is a method for manufacturing a mask blank 10 having a pattern forming thin film for forming a transfer pattern on the above-described mask blank substrate 1 of the present invention. In the manufacturing method of the present invention, a pattern forming thin film is formed on the above-described mask blank substrate 1 of the present invention. Since the surface defect 5 of the mask blank substrate 1 of the present invention is corrected, the mask blank 10 free from the surface defect 5 caused by the mask blank substrate 1 can be obtained.

  The pattern forming thin film for forming the transfer pattern is greatly different between a mask blank for semiconductor device manufacturing and a mask blank for FPD manufacturing. Details will be described later. As a method for forming the pattern forming thin film on the mask blank substrate 1, for example, a sputter film forming method is preferably exemplified, but the present invention is not necessarily limited to the sputter film forming method.

  A transfer mask can be obtained by forming a predetermined fine pattern on the pattern forming thin film of the mask blank 10 by a known photolithography method.

  In the case of a mask blank for use in manufacturing semiconductor devices, the thin film for pattern formation includes, for example, a light semi-transmissive film such as a light shielding film and a halftone type phase shift film. The light shielding film is a thin film having high light shielding performance with respect to exposure light. The light-shielding film hardly transmits exposure light (generally, the transmittance is about 0.1% or less). A mask blank to which this light shielding film is applied is also called a binary mask blank. A transfer mask produced from this mask blank is also called a binary mask.

  A halftone phase shift film, which is one type of light semi-transmissive film, transmits light having an intensity that does not substantially contribute to exposure (for example, 1% to 20% with respect to the exposure wavelength). A phase shift portion obtained by patterning this halftone phase shift film, and light having an intensity that contributes substantially to exposure without the phase shift portion being transmitted. The phase shift unit and the light transmission unit are configured such that the phase of the light is transmitted through the phase shift unit and the phase of the light is substantially inverted with respect to the phase of the light transmitted through the light transmission unit. The light that passes through the vicinity of the boundary and wraps around each other due to diffraction phenomenon cancels each other, and the light intensity at the boundary is almost zero, so that the contrast of the boundary, that is, the resolution is improved. It is intended to be.

  As a light semi-transmissive film other than the halftone phase shift film, a resist film for pattern transfer is a thin film that transmits exposure light at a predetermined transmittance that is not exposed to light. Some thin films are adjusted so as not to cause a phase difference with the exposure light that has passed through the part having no light. A mask blank to which this type of translucent film is applied is often used when an enhancer mask is manufactured. Since both the halftone phase shift film and the light semi-transmissive film have a predetermined transmittance with respect to the exposure light, when the exposure is performed on the resist film of the transfer object in the exposure apparatus, the resist is exposed. The film may be exposed. For this reason, in many cases, a film configuration is formed by laminating a light shielding film for forming a light shielding band on a halftone type phase shift film or a light semi-transmissive film.

  On the other hand, in the case of a mask blank for use in FPD production, a transfer mask produced from this mask blank is generally used in an exposure apparatus using an ultrahigh pressure mercury lamp as a light source (exposure light). The exposure light of the ultra-high pressure mercury lamp is not single-wavelength light such as ArF excimer laser, but multicolor light. The exposure light of the ultra-high pressure mercury lamp is light having particularly strong light intensity at three wavelengths of i-line (365 nm), h-line (405 nm), and g-line (436 nm). With such multicolor light, it is difficult to obtain a phase shift effect, so that the halftone phase shift film is hardly used in a mask blank for FPD manufacturing. In the case of a mask blank for FPD manufacturing, a mask blank (binary mask blank) in which a light-shielding film is applied to a thin film for pattern formation is used in the same manner as a mask blank for semiconductor device manufacturing.

  As a transfer mask for FPD manufacturing, there is a multi-tone mask having a configuration in which a transfer pattern is formed on each of a plurality of pattern forming thin films having different transmittances. Using this multi-tone mask, transfer exposure is performed on the resist film to be transferred, and development processing is performed. In the resist film, all the resist is dissolved, and a certain film thickness is dissolved. As a result, an almost undissolved region is formed. Therefore, a plurality of patterns are formed on the resist film. Such a resist pattern is formed by first patterning the thin film under the resist film using the first pattern consisting of the remaining area where the resist is completely dissolved and the remaining area where the other resist remains as a mask. Can do. Next, a process of removing the resist film by a predetermined thickness on the entire surface is performed until the resist in a region where the resist film is dissolved to some extent disappears. As a result, the region other than the region where the resist is hardly dissolved in the initial stage is a missing region where the resist is completely dissolved, and a second pattern can be formed. Using this second pattern as a mask, a pattern different from the first pattern can be formed on the thin film under the resist film. By using such a process, a single multi-tone mask can be used where two transfer masks are originally required.

  As a method of manufacturing a multi-tone mask, first, a method of completing a multi-tone mask by repeating the formation of a pattern forming thin film and the etching process of pattern formation is given. In addition, a multi-tone mask blank having a structure in which a light semi-transmissive film (one or more steps) and a light-shielding film are laminated on the mask blank substrate 1 is manufactured in advance. There is a method of completing a multi-tone mask by performing an etching process.

  Other thin films for pattern formation include an etching mask (hard mask) film and an etching stopper film.

  Further, a material containing transition metal and silicon as main components (a material containing transition metal silicide as a main component) is also applicable. In addition to transition metal silicide alone, transition metal silicide compounds in which elements such as oxygen, nitrogen, carbon, hydrogen, and boron are added to transition metal silicide include materials containing transition metal silicide as a main component. As the transition metal of the transition metal silicide, molybdenum, tantalum, tungsten, titanium, hafnium, nickel, vanadium, zirconium, niobium, palladium, ruthenium, rhodium, or the like can be applied.

  In addition, as a material applicable to the thin film for pattern formation, a material mainly composed of tantalum can be given. In addition to tantalum alone, tantalum-based materials include tantalum compounds in which elements such as oxygen, nitrogen, carbon, hydrogen, and boron are added to tantalum.

  In the case of a light shielding film, since it has a high light shielding property, the reflectance with respect to exposure light tends to be high. In order to suppress this, the light-shielding film often has a laminated structure in which an antireflection layer for suppressing reflectance is formed on a light-shielding layer having a high content of light-shielding metal (transition metal silicide). . Further, when it is necessary to suppress the reflectance of the light shielding film on the substrate side, there may be a laminated structure in which a back surface antireflection layer is further formed between the substrate and the light shielding layer. As a material for the antireflection layer and the back surface antireflection layer, a metal (transition metal silicide) compound material containing oxygen or nitrogen in a metal (transition metal silicide) is suitable. On the other hand, since the light shielding layer needs to have a high light shielding performance, a metal (transition metal silicide) or a metal compound (transition metal silicide compound) material containing as little oxygen or nitrogen as possible is preferable.

  Examples of materials applicable to the light semi-transmissive film such as the halftone phase shift film include the chromium compounds, tantalum compounds, and transition metal silicide compounds. In particular, the compound containing oxygen or nitrogen is preferable in order to adjust transmittance and the like.

  Next, the present invention will be described more specifically with reference to examples of the present invention.

Example 1
The manufacturing of the mask blank substrate 1 includes (1) a cutting process, (2) a lapping process,
(3) The polishing process, (4) the inspection process, and (5) the surface defect correction process. Hereinafter, each process will be described in detail.

(1) Cutting process The synthetic quartz glass cut out on the flat plate from the synthetic quartz glass ingot was prepared. Next, it cut out with the grinding stone from the sheet glass which consists of synthetic quartz glass, and obtained the rectangular-shaped large-sized board | substrate 1 (330 mm x 450 mm or more, specifically 800 mm x 920 mm, thickness 10 mm).

(2) Lapping process Next, the substrate 1 was subjected to a lapping process. This lapping process aims to improve dimensional accuracy and shape accuracy. The lapping process can be composed of three processes: a rough lapping process, a shape processing process, and a fine lapping process.

(2-1) Coarse lapping process The coarse lapping process was performed using a double-sided lapping apparatus, and the grain size of the abrasive grains was # 400. Specifically, alumina abrasive grains having a particle size # 400 were used, the load L was set to about 100 kg, and both surfaces of the substrate 1 housed in the carrier were finished with a predetermined accuracy.

(2-2) Shape processing step Next, the outer peripheral end surface was also ground to adjust the outer dimensions, and then the outer peripheral end surface and the inner peripheral surface were subjected to predetermined chamfering.

(2-3) Fine lapping step Next, the grain size of the abrasive grains was changed to # 1000, and the surface of the mask blank substrate 1 was lapped to finish both surfaces of the substrate 1 with a predetermined accuracy. The substrate 1 after the fine wrapping step was cleaned by immersing in a cleaning bath of neutral detergent and water sequentially.

(3) Polishing process Next, the polishing process was performed. This polishing process is intended to remove scratches and distortions remaining in the lapping process described above, and can comprise a plurality of polishing processes including a first polishing process, a second polishing process, and a third polishing process. Moreover, it can have a washing | cleaning process for performing predetermined washing | cleaning after predetermined | prescribed grinding | polishing.

(3-1) 1st grinding | polishing process It performed using the double-side polish apparatus for large sized mask blanks as shown in FIG. A hard polisher was used as the polisher, and the polishing was performed under the following polishing conditions.
Polishing liquid: cerium oxide (average particle size 2 to 3 μm) free abrasive grains with water added to abrasive grains Polishing cloth: hard polisher, thickness 2 mm
The amount of polishing liquid supplied directly to the groove only: 5 liters / minute The amount of polishing liquid supplied onto the surface plate: 2 liters / minute Upper surface plate rotation speed: 1 to 50 rpm (clockwise rotation when viewed from the top of the apparatus)
Lower surface plate rotation speed: 1 to 50 rpm (counterclockwise rotation when viewed from the top of the device)
Carrier (substrate 20) rotation speed: 1 to 50 rpm (counterclockwise rotation when viewed from the top of the apparatus)
Load (applied pressure): 80 to 100 g / cm ²
Polishing time: 30 to 100 minutes Removal amount: 35 to 75 μm

  The substrate 1 that had been subjected to the polishing step was washed by sequentially immersing it in each cleaning bath of neutral detergent, pure water, pure water, IPA, and IPA (steam drying). In addition, ultrasonic waves were applied to each cleaning tank. Further, this cleaning step can be omitted when the polishing liquid used in the next second polishing step is the same.

(3-2) Second Polishing Step Next, the second polishing step was performed using the single-side polishing apparatus for large mask blanks shown in FIG. An ultra-soft polisher was used as the polisher, and the polishing was performed under the following conditions.
Polishing liquid: Colloidal silica (average particle size 50-80 nm) Free abrasive grains with water added to abrasive grains Polishing cloth: Super soft polisher (suede type), thickness 2 mm
The amount of polishing liquid supplied directly to the groove only: 5 liters / minute The amount of polishing liquid supplied onto the surface plate: 2 liters / minute Lower surface plate rotation speed: 1 to 50 rpm (counterclockwise rotation when viewed from the top of the apparatus)
Carrier (substrate 20) rotation speed: 1 to 50 rpm (counterclockwise rotation when viewed from the top of the apparatus)
Load (applied pressure): 80 to 100 g / cm ²
Polishing time: 30-50 minutes Removal amount: 35-45 μm

(3-3) Third polishing process (mirror polishing process)
Next, using the double-side polishing apparatus for large mask blanks shown in FIG. 6, a third polishing process (mirror polishing process; final polishing) was performed. An ultra-soft polisher was used as the polisher, and the polishing was performed under the following conditions.
Polishing liquid: Colloidal silica (average particle size 50-80 nm) Free abrasive grains with water added to abrasive grains Polishing cloth: Super soft polisher (suede type), thickness 2 mm
The amount of polishing liquid supplied directly to the groove only: 5 liters / minute The amount of polishing liquid supplied onto the surface plate: 2 liters / minute Upper surface plate rotation speed: 1 to 50 rpm (clockwise rotation when viewed from the top of the apparatus)
Lower surface plate rotation speed: 1 to 50 rpm (counterclockwise rotation when viewed from the top of the device)
Carrier (substrate 20) rotation speed: 1 to 50 rpm (counterclockwise rotation when viewed from the top of the apparatus)
Load (applied pressure): 80 to 100 g / cm ²
Polishing time: 30-50 minutes Removal amount: 35-45 μm

(3-4) Cleaning Step The substrate 1 after the third polishing step was sequentially immersed in an alkaline solution (NaOH solution) and sulfuric acid for cleaning. In addition, ultrasonic waves were applied to each cleaning tank. Furthermore, it wash | cleaned by immersing in each washing tank of neutral detergent, a pure water, a pure water, IPA, and IPA (steam drying) one by one.

(4) Inspection process Next, the obtained substrate 1 was inspected for flatness, surface roughness, presence / absence of a surface defect 5 and its position. The presence or absence of the surface defect 5 was inspected with a laser scattered light type defect inspection apparatus. Here, the flatness required for the mask blank substrate 1 was 20 μm or less, and the surface roughness was 0.25 nm or less in terms of arithmetic average roughness Ra. As for the surface defect 5, it is determined that a width larger than 1 μm is not allowed. Here, in order to verify the effectiveness of the flame defect correction of the present invention, the surface defect 5 (concave defect) having an unacceptable size, although both the flatness and the surface roughness satisfy the above-described conditions for the substrate 1 after the inspection. The substrate 1 having a depth of 0.5 μm or less is selected.

(5) Surface defect correction process Flame defect correction was performed on the surface defects 5 of the substrate 1 selected in the inspection process. The flame 52 used for the correction of the surface defect 5 is as follows.
・ Flame temperature: 2200 ℃
-Distance between the flame irradiation device 50 and the mask blank substrate 1: 15 mm
・ Flame diameter: 3mm
・ Flame irradiation time: 10 minutes ・ Types of flammable gas: Mixed gas of butane and oxygen

  When the portion of the substrate 1 obtained as described above was observed with a microscope using an optical microscope, the surface defect 5 disappeared. Further, when the flatness and surface roughness of the main surface 17 were measured again with respect to the substrate 1, both of them were below the above-described determination threshold value and corrected to a level that can be used as a non-defective mask blank substrate 1. It was done.

Next, a multi-tone mask blank for use in FPD manufacturing was manufactured using the mask blank substrate 1 whose surface defect 5 was corrected and became a non-defective product. On the main surface 17 of the mask blank substrate 1, first, a light semi-transmissive film 2 made of molybdenum and silicon was formed.
Specifically, using a mixed target of molybdenum (Mo) and silicon (Si) (Mo: Si = 20 at%: 80 at%), MoSi composed of molybdenum and silicon is performed by DC sputtering in an argon (Ar) gas atmosphere. A film (light semi-transmissive film 2) was formed with a thickness of 10 nm. The translucent film 2 is adjusted so that the transmittance is 15% at the wavelength of i-line (365 nm).

Next, a light shielding film 3 made of a chromium-based material was formed on the light semitransmissive film 2.
Specifically, a chromium (Cr) target is used as a sputtering target, and a CrN layer having a film thickness of 15 nm is formed by reactive sputtering (DC sputtering) in a mixed gas atmosphere of argon (Ar) and nitrogen (N ₂ ). Then, a 65 nm thick CrCN layer is formed by reactive sputtering (DC sputtering) in a mixed gas atmosphere of argon (Ar), methane (CH ₄ ), and nitrogen (N ₂ ). A CrON layer having a film thickness of 25 nm was formed by reactive sputtering (DC sputtering) in a mixed gas atmosphere of argon (Ar) and nitrogen monoxide (NO) to form a light-shielding film 3 having a total film thickness of 105 nm. The light shielding film 3 is a film having a structure in which each layer has a composition gradient, and is adjusted to have an optical density of 3.0 at a wavelength of i line (365 nm) in a laminated structure with the light semi-transmissive film 2.
As described above, the multi-tone mask blank 10 in which the light semi-transmissive film 2 and the light shielding film 3 were laminated in this order on the mask blank substrate 1 was manufactured.

  Next, a multi-tone mask (transfer mask) was prepared using the multi-tone mask blank. FIG. 5 is a schematic cross-sectional view showing a process of manufacturing a multi-tone mask using a multi-tone mask blank. First, a resist film 4 was formed on the mask blank 10 (see FIG. 1A).

Next, the light-transmitting portion pattern of the multi-tone mask is drawn on the resist film 4 formed on the mask blank 10 by using a laser drawing device, and then developed with a predetermined developer to form a resist pattern. 4a was formed (see FIG. 4B).
Next, the light shielding film 3 was etched using the resist pattern 4a as a mask to form a light transmitting portion pattern on the light shielding film 3 (see FIG. 3C). Here, wet etching using an etching solution for Cr (an aqueous solution containing ceric ammonium sulfate and perchloric acid) was applied.

  Next, the remaining resist pattern 4a was peeled off (see FIG. 4D). Using the light transmitting part pattern formed on the light shielding film 3 as a mask, the light semi-transmissive film 2 was etched to form a light transmitting part pattern (see FIG. 5E). Here, wet etching using an etching solution for MoSi (an aqueous solution containing ammonium hydrogen fluoride and hydrogen peroxide) was applied.

Next, the resist film 4 was formed again on the mask blank 10. After drawing the light-shielding part pattern of the multi-tone mask using a laser drawing apparatus, the resist pattern 4b was formed by developing with a predetermined developer (see FIG. 5F).
Next, using the resist pattern 4b as a mask, the light shielding film 3 was etched to form a light shielding portion pattern on the light shielding film 3 (see FIG. 5F). Again, wet etching using the same Cr etching solution as described above was applied.

  Next, the remaining resist pattern 4b was peeled off to obtain a multi-tone mask (transfer mask) 15. The multi-tone mask 15 includes a transparent portion 17 having a main surface 17 exposed on the mask blank substrate 1, a semi-transparent portion 18 made of the light semi-transmissive film 2, the light semi-transmissive film 2 and the light-shielding film 3. The light shielding portion 16 having a laminated structure is formed. It should be noted that the pattern arrangement is made such that the portion where the surface defect 5 of the main surface 17 is present covers the translucent portion 17 and the semi-transparent portion 18.

  Using this multi-tone mask 15, an exposure apparatus using an ultrahigh pressure mercury lamp as a light source performed exposure transfer of a pattern to a resist film to be transferred. When development processing was performed on the resist film of the transfer object, it was confirmed that the resist pattern was normally formed including the portion where the surface defect 5 was corrected.

(Comparative Example 1)
In this comparative example 1, (4) in the inspection process, both the flatness and the surface roughness satisfy the predetermined conditions, but the surface defect 5 (concave defect) has an unacceptable size, and its depth is 0.5 μm. The process is the same as in Example 1 until the substrate 1 on which the following is present is selected. In Comparative Example 1, the surface defect 5 of the selected substrate 1 is corrected. (5) The light semitransmissive film 2 is used as the mask blank substrate 1 without performing the surface defect correction step under the same conditions as in Example 1. Then, the light shielding film 3 was formed, and the mask blank 10 was manufactured. Furthermore, using this mask blank 10, the light semi-transmissive film 2 and the light shielding film 3 were patterned in the same procedure as in Example 1 to produce a multi-tone mask 15.

  Using this multi-tone mask 20, an exposure apparatus using an ultrahigh pressure mercury lamp as a light source performed exposure transfer of a pattern to a resist film to be transferred. When the resist film of the transfer object is developed, the portion where the surface defect 5 is present and the portion where the translucent portion 17 and the semi-transparent portion 18 are exposed and transferred have a transfer defect and are normal. No pattern was formed. Even if this resist pattern is used for etching to transfer a pattern to the underlying metal film, a pattern as designed cannot be formed. Therefore, the mask blank substrate 1, the multi-tone mask blank 10, and the multi-tone mask 15 of Comparative Example 1 are not suitable for use.

1 Mask blank substrate 2 Light semi-transmissive film (pattern forming thin film)
3 Light-shielding film (thin film for pattern formation)
4 resist film 4a, 4b resist pattern 5 surface defect 10 multi-tone mask blank 15 multi-tone mask 16 light-shielding portion 17 light-transmitting portion 18 semi-light-transmitting portion 20 mask blank substrate attached to polishing apparatus 21 carrier 22 upper limit Panel 23 Lower surface plate 30 Surface plate to which polishing cloth is attached 50 Flame irradiation device 52 Flame 71 Main surface 72 Side surface 73 Chamfering surface 100 Flame defect correcting device 102 Stage 104 Flame control unit 105 Flammable gas 108 Microscope 110 Image processing unit 112 Stage controller 114 Control unit 126 Observation light

Claims (8)

Ri Do of synthetic quartz glass substrate having two opposing major surfaces, the flatness of the main surface of the substrate is an der Ru mask blank substrate manufacturing method below 30 [mu] m,
Polishing two main surfaces of the substrate;
A step of correcting the surface defect by bringing a flame having a combustion temperature equal to or higher than the softening point of the substrate into contact with a surface defect existing on the main surface of the substrate. A method for manufacturing a substrate.   The method for manufacturing a mask blank substrate according to claim 1, wherein the surface defect is a concave defect. The flame is characterized Rukoto obtained by burning inflammable gas containing oxygen, a manufacturing method of a substrate for a mask blank according to claim 1 or 2. The method for manufacturing a mask blank substrate according to claim 1, wherein the flame has a combustion temperature of 1600 ° C. or higher.   The mask blank substrate according to any one of claims 1 to 4, wherein an average surface roughness Ra of a main surface of the substrate obtained after the polishing step is 0.3 nm or less. Method.   The depth of the said surface defect is 0.5 micrometer or less, The manufacturing method of the board | substrate for mask blanks of any one of Claims 1-5 characterized by the above-mentioned. It has the process of forming the thin film for pattern formation on the main surface of the mask blank board | substrate manufactured with the manufacturing method of the mask blank board | substrate of any one of Claims 1-6 , It is characterized by the above-mentioned. Mask blank manufacturing method. A method for producing a transfer mask, comprising a step of forming a transfer pattern on a thin film of the mask blank produced by the method for producing a mask blank according to claim 7 .

Images