JP2017161807A - Manufacturing method of substrate, manufacturing method of mask blank, and manufacturing method of transfer mask - Google Patents

Abstract

PROBLEM TO BE SOLVED: To be able to obtain a substrate having high flatness without applying a local processing on a main surface of the substrate.SOLUTION: A manufacturing method of a substrate having a plurality of stages of polishing steps for simultaneously polishing two main surfaces of the substrate by using a double-sided polishing device provided with a plurality of carriers provided with more than one substrate holding holes between a surface plate and a lower surface plate includes: a step of preparing a substrate parent set configured by collecting a plurality of sheets of the substrates having two main surfaces; a first polishing step of polishing two main surfaces of the substrate of the substrate parent set in the same batch; a surface shape measuring step of acquiring surface shape data of the substrate parent set by measuring a surface shape of at least one main surface of each substrate of the substrate parent set after the first polishing step; a set dividing step of dividing the substrate parent set into two or more substrate child sets based on the surface shape data of each substrate of the substrate parent set acquired in the surface shape measuring process; and a second polishing step of polishing two main surfaces of the substrate of the substrate parent set in the same batch.SELECTED DRAWING: Figure 1

Description

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

  Conventionally, in the manufacture of a mask blank substrate, a grinding process and a polishing process are performed on a main surface of a substrate cut into a predetermined shape from an ingot. In the polishing step of polishing the main surface of the substrate, it is common to simultaneously polish the two main surfaces of the substrate using a double-side polishing apparatus as disclosed in Patent Document 1. This double-side polishing apparatus can hold a plurality of substrates with one carrier, and a plurality of carriers are arranged on a surface plate. This double-side polishing apparatus can polish the main surfaces of a large number of substrates at the same time in one polishing process (one batch). In general, the polishing process for polishing the main surface of the substrate is performed in a plurality of stages using this double-side polishing apparatus. For example, as disclosed in Patent Document 1, the multi-step polishing process includes a rough polishing process and a precision polishing process in which a main surface of a substrate is polished using a cerium oxide abrasive, and colloidal silica. An ultra-precise polishing process is performed in which polishing is performed using this abrasive.

JP2013-082612A

  As disclosed in Patent Document 1, in a conventional mask blank substrate manufacturing method, two main surfaces of a large number of substrates are simultaneously polished by a double-side polishing apparatus in a multi-step polishing process. In the multi-stage polishing process, the main surface of the substrate is subjected to one or more stages of polishing using a cerium oxide abrasive, followed by one stage of ultra-precision polishing using a colloidal silica abrasive. This is done. Further, as disclosed in Patent Document 1, the number of substrates to be polished by a double-side polishing apparatus in one process (one batch) is often the same in each stage of the polishing process. That is, a set of substrates polished in the same batch in the previous process is directly polished in the same batch in the next process. The substrate polishing conditions in the polishing step (various conditions such as the rotation speed of the upper and lower surface plates, the rotation speed of the sun gear and the internal gear, the polishing time, etc.) are usually the same between batches. After the substrate set polished in the previous step is cleaned to remove the polishing liquid in a cleaning tank or the like, it is kept in a state of being immersed in the standby tank until polishing in the next step. At this time, it is difficult to make the conditions of the washing tank and the standby tank, the time to wait in the standby tank, etc., the same between the batches. Moreover, even if the polishing conditions are the same between batches, it is difficult to make the conditions (polishing liquid, polishing pad state, etc.) between batches exactly the same. Due to these differences between batches of substrate sets, there may be relatively large differences in major surface shape trends in the set of substrates after polishing between batches. Further, there may be a large difference in plate thickness variation (in-plane difference with respect to the distance between the two main surfaces) in the set of substrates after polishing between batches.

  Because of these circumstances, a set of substrates polished in the same batch in the previous process is often polished in the same batch in the next process as it is. When the substrates of different sets with different main surface shape tendencies after polishing in the previous step are mixed and polished in the same batch in the next step, the main surface shape of each substrate after polishing in the same batch The difference in trends tends to be larger. This difference in the tendency of the main surface shape leads to variations in the flatness of the main surface, and the ratio (yield) at which a substrate with a high flatness of the main surface can be obtained in a substrate after a plurality of polishing steps is reduced. To do. Conventionally, in a multi-stage polishing process, a set of substrates polished in the same batch in the previous process is polished in the same batch as it is in the next process, so that a ratio of obtaining a substrate with high flatness is obtained to some extent. Was able to be high.

  In recent years, there has been a demand for higher flatness (for example, flatness in a rectangular region having a side of 142 mm is 0.2 μm or less) with respect to a main surface required for a mask blank substrate, and a plurality of conventional steps. It is difficult to improve the acquisition rate of such a high flatness substrate only by performing the polishing step. Measure the main surface shape of the substrate after polishing, identify a relatively convex area on the main surface from the measurement results, and perform local processing on the convex area to determine the main surface shape of the substrate. Conventionally, the surface is made to have high flatness. However, such local processing requires that processing be performed on each substrate, so that the throughput is greatly reduced. There has been a problem of increasing the ratio at which a substrate with high flatness can be obtained only by a plurality of polishing steps without performing local processing on the main surface of the substrate.

The present invention has been made in view of the above-described problems of the prior art, and its purpose is to obtain a substrate with high flatness by only a plurality of polishing steps without locally processing the main surface of the substrate. It is an object of the present invention to provide a method for manufacturing a substrate capable of increasing a possible ratio.
Moreover, the objective of this invention is providing the manufacturing method of the mask blank using the manufacturing method of the said board | substrate.
Furthermore, the objective of this invention is providing the manufacturing method of the mask for transcription | transfer using the mask blank obtained with the manufacturing method of the said mask blank.

  The inventors of the present invention conducted intensive research on a method for increasing the ratio at which a substrate with high flatness can be obtained only by a plurality of stages of polishing processes. First, the polishing allowance of the main surface of the substrate in multiple stages of polishing processes (thickness removed from the surface of the main surface by polishing) decreases from the first stage polishing process toward the final stage polishing process. It will become. The polishing process in the first half (for example, a polishing process using a cerium oxide abrasive) has a relatively high polishing rate and tends to change the macroscopic surface shape (flatness and waviness) of the main surface of the substrate. On the other hand, the tendency to improve the microscopic surface shape (surface roughness) of the main surface is small. On the other hand, in the latter stage polishing process (for example, a polishing process using a colloidal silica abrasive), the polishing rate is relatively slow, and the macroscopic surface shape (flatness and waviness) of the main surface of the substrate is changed. While the tendency is small, there is a large tendency to improve the microscopic surface shape (surface roughness) of the main surface. For this reason, when the main surface shape is greatly different between a plurality of substrates processed in one batch at the time when the polishing process in the first half is finished, the polishing process in the latter half is performed with the combination of the substrates in one batch. Even if it goes, the variation of the flatness of each substrate in the polishing process becomes large.

  As a result of further research, the present inventors measured the main surface shape of all the substrates in the set after being polished in the same batch in the previous polishing step, and measured the main surface shape of each substrate. Based on the data, one batch of substrate sets was divided into two or more substrate sets, and one batch of polishing was performed for each set of substrates divided in the subsequent polishing step.

  The present inventors considered why it is difficult to improve the acquisition rate by paying attention to the difference in action of each polishing step. In a double-side polishing apparatus used for polishing a substrate, the surface plate surface (polishing surface) of the upper surface plate is rarely fixedly connected in a state perpendicular to the rotating shaft. Usually, it is difficult to make all the thicknesses of a set of substrates, which are objects to be polished, the same, and the thicknesses are slightly different. When this set of substrates is to be polished by a double-side polishing apparatus to which the upper surface plate is fixedly connected, the load on the upper surface plate is greatly applied to a substrate having a relatively thick plate thickness in the substrate set. A substrate subjected to such a large load is likely to be cracked or scratched on the main surface, and may be cracked. For this reason, the upper surface plate of the double-side polishing apparatus is often connected to the rotary shaft so as to be swingable through a universal joint or the like.

  In such a double-side polishing apparatus, the polishing surface of the upper surface plate is slightly inclined with respect to the polishing surface of the lower surface plate so as to absorb the difference in thickness of the plurality of substrates disposed on the lower surface plate, A plurality of substrates are polished simultaneously. As a result, the deviation of the load from the upper surface plate with respect to the relatively thick substrate in the set of substrates is reduced. However, when performing double-side polishing with a double-side polishing apparatus, it is rare that a plurality of substrates are always at the same position on the polishing surface of the upper and lower surface plates. For this reason, the inclination of the upper surface plate of the double-side polishing apparatus constantly fluctuates during double-side polishing. It is difficult for the inclination of the upper surface plate to completely follow the movement of the substrates having different thicknesses. A set of substrates polished by such a double-side polishing apparatus cannot make the variation in the plate thickness between the substrates zero, but can make it smaller than the variation before polishing. However, during double-sided polishing, it is difficult to make the macroscopic surface shape of each substrate polished in the same batch into a shape with the same tendency due to the complicated movement of the inclination with respect to the lower surface plate on the upper surface plate. ing.

  On the other hand, in the case of a double-side polishing apparatus in which a plurality of substrates are arranged on one carrier, the center of each substrate is shifted from the center of the carrier. For this reason, each substrate undergoing double-side polishing revolves around the center of the carrier. Furthermore, the carrier itself revolves around the center (rotation axis) of the surface plate. These things act, and in the double-side polishing by this double-side polishing apparatus, each substrate draws a very complicated locus on the surface plate. Since it is difficult for each substrate to draw the same trajectory, it is difficult to make the macroscopic surface shape of each substrate after polishing in the same batch by this double-side polishing apparatus have the same tendency shape.

  These tendencies become conspicuous in the polishing process in the first half where the polishing allowance is relatively large. In the latter half of the polishing process, the polishing allowance is small as described above, and it is not easy to improve the macroscopic surface shape. Also, in the double-side polishing apparatus used in the latter half of the polishing process, the upper surface plate and the rotary shaft are often connected so as to be swingable. If there is a large difference in macroscopic surface shape (flatness) between sets of substrates to be polished in the same batch, the amount that the upper surface plate swings to absorb the difference increases. When the amount of rocking of the upper surface plate is large, it becomes difficult to reduce the variation in the macroscopic surface shape (flatness) between the substrates by this latter-stage polishing process.

  With these current problems in mind, we have intensively studied how to obtain a high ratio of substrates with good macroscopic surface shape (flatness) from a set of substrates polished in the same batch by the latter-stage polishing process. Went. The difference between the worst and best macroscopic surface shapes in the set of substrates polished in the same batch after the first half of the polishing process is large, but this set of substrates (substrate parent set) is, for example, If we divide into a set of substrates with relatively good macroscopic surface shape (substrate substrate set) and a set of substrates with relatively poor macroscopic surface shape (substrate substrate set), the surface shape between the substrates of each substrate child set Noticed that would be smaller. In the latter half of the polishing process, if the substrate sets divided into a plurality of pieces are polished in the same batch, the amount of oscillation of the upper surface plate can be reduced. When the latter half of the polishing process is performed on a substrate set that has a relatively good macroscopic surface shape, the microscopic surface shape (surface roughness) is greatly improved while the macroscopic surface shape remains good. Can be made. In addition, when a second-stage polishing process is performed on a substrate set having a relatively poor macroscopic surface shape, a surface shape substrate with a close difference in height of the main surface is polished in the same batch. While the amount of oscillation is small, the macroscopic surface shape can be greatly improved, and the microscopic surface shape (surface roughness) can be greatly improved.

The present invention completed as described above has the following configuration.
(Configuration 1)
Using a double-side polishing apparatus comprising an upper surface plate and a lower surface plate that rotate coaxially with each other, and a plurality of carriers provided with one or more substrate holding holes between the upper surface plate and the lower surface plate, A substrate manufacturing method comprising a plurality of polishing steps in which a substrate is disposed and the two main surfaces of the substrate are polished simultaneously,
Preparing a substrate parent set configured by collecting a plurality of substrates having two main surfaces;
The substrates of the substrate parent set are arranged in the substrate holding holes of the carrier of the first double-side polishing apparatus, and two main components in each substrate of the substrate parent set are used with a polishing liquid containing a first abrasive. A first polishing step for polishing the surface in the same batch;
A surface shape measuring step of measuring a surface shape of at least one main surface of each substrate of the substrate parent set after performing the first polishing step, and acquiring surface shape data of each substrate of the substrate parent set;
Based on the surface shape data of each substrate of the substrate parent set acquired in the surface shape measurement step, a set dividing step of dividing the substrate parent set into two or more substrate child sets;
One set of the substrate child sets is disposed in the substrate holding hole of the carrier of the second double-side polishing apparatus, and a polishing liquid containing a second abrasive is used to remove the one set of substrate child sets. And a second polishing step of polishing the two main surfaces of each substrate in the same batch.

(Configuration 2)
The set dividing step calculates a flatness within a predetermined region of the main surface from the surface shape data of each substrate, and divides the substrate parent set into the two or more substrate child sets based on the flatness. A method for manufacturing a substrate according to Configuration 1, characterized in that:
(Configuration 3)
The set division step calculates the average surface shape of each substrate from the surface shape data of each substrate, and further calculates a difference shape between the average surface shape and the surface shape of each substrate. The substrate manufacturing method according to claim 1, wherein the substrate parent set is divided into the two or more substrate child sets based on the difference shape in the substrate.

(Configuration 4)
4. The method for manufacturing a substrate according to Configuration 2 or 3, wherein the predetermined region is a region including at least a rectangular inner region having a side of 132 mm with respect to the center of the main surface of the substrate.
(Configuration 5)
5. The method of manufacturing a substrate according to any one of configurations 1 to 4, wherein a center of the substrate holding hole is shifted from a center of the carrier.

(Configuration 6)
The first abrasive is made of a material containing at least one of cerium oxide and zirconium oxide,
6. The method for manufacturing a substrate according to any one of configurations 1 to 5, wherein the second abrasive is made of a material containing colloidal silica.
(Configuration 7)
7. The substrate according to any one of configurations 1 to 6, wherein the one main surface is a main surface on a side on which a thin film for pattern formation is formed when a mask blank is manufactured using the substrate. Production method.

(Configuration 8)
A method for producing a mask blank, comprising a step of forming a pattern forming thin film on the one main surface of the substrate produced by the method for producing a substrate according to any one of Structures 1 to 7.
(Configuration 9)
A method for producing a transfer mask using a mask blank produced by the method for producing a mask blank according to Configuration 8,
A method for producing a transfer mask, comprising a step of forming a transfer pattern on the pattern forming thin film by dry etching.

ADVANTAGE OF THE INVENTION According to this invention, the manufacturing method of a board | substrate which can raise the ratio which can acquire a board | substrate with a high flatness only by a multi-step polishing process, and a mask blank, without performing a local process with respect to the main surface of a board | substrate. Can be provided.
According to the present invention, it is possible to provide a method for manufacturing a substrate capable of improving the yield while maintaining the production efficiency. Through the reduction of the production cost by improving the yield, the manufacturing cost of the substrate, the manufacturing cost of the mask blank, and the transfer The manufacturing cost of the mask can be reduced.

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.

Hereinafter, embodiments of the present invention will be described in detail.
[Double-side polishing machine]
In the substrate manufacturing method of the present invention, a double-side polishing apparatus is used.
In double-side polishing of the substrate, a planetary gear type double-side polishing apparatus is preferably used.
FIG. 1 is a cross-sectional view showing a schematic configuration of a planetary gear type double-side polishing apparatus used in a double-side polishing process, and FIG. 2 is a meshing relationship of a sun gear, an internal gear and a carrier in the double-side polishing apparatus of FIG. FIG.

  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. The carrier 50 that revolves and rotates in response to the substrate, the substrate W that is a workpiece to be polished held on the carrier 50, and the polishing pads 21 and 11 are bonded to each other, and the upper surface plate that can hold the substrate W therebetween. 20 and a lower surface plate 10, and a polishing liquid supply unit 60 for supplying a polishing liquid between the upper surface plate 20 and the lower surface plate 10. The rotation driving of the sun gear 40 (A is a rotation axis) is controlled by the sun gear rotation driving unit 31. The upper platen 20 that is swingably connected to the upper support member 22 is controlled to be lifted and rotated by an upper platen lift drive unit 24 and an upper platen rotation drive 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 upper surface plate 20 and the lower surface plate 10 rotate coaxially with each other. A plurality of carriers 50 are arranged on a surface plate. 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 substrate W is constant.

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

[Substrate manufacturing method]
As described above, the substrate manufacturing method according to the present invention includes an upper surface plate and a lower surface plate that rotate coaxially with each other, and a plurality of carriers including one or more substrate holding holes between the upper surface plate and the lower surface plate. Using the provided double-side polishing apparatus, a substrate manufacturing method comprising a plurality of stages of polishing steps in which a substrate is disposed in the substrate holding hole and the two main surfaces of the substrate are simultaneously polished,
Preparing a substrate parent set configured by collecting a plurality of substrates having two main surfaces;
The substrates of the substrate parent set are arranged in the substrate holding holes of the carrier of the first double-side polishing apparatus, and two main components in each substrate of the substrate parent set are used with a polishing liquid containing a first abrasive. A first polishing step for polishing the surface in the same batch;
A surface shape measuring step of measuring a surface shape of at least one main surface of each substrate of the substrate parent set after performing the first polishing step, and acquiring surface shape data of each substrate of the substrate parent set;
Based on the surface shape data of each substrate of the substrate parent set acquired in the surface shape measurement step, a set dividing step of dividing the substrate parent set into two or more substrate child sets;
One set of the substrate child sets is disposed in the substrate holding hole of the carrier of the second double-side polishing apparatus, and a polishing liquid containing a second abrasive is used to remove the one set of substrate child sets. And a second polishing step of polishing the two main surfaces of each substrate in the same batch (Configuration 1).

  In the substrate manufacturing method of the present invention, the plurality of steps of polishing using the double-side polishing apparatus include at least a first polishing step and a second polishing step, and a plurality of three or more including a first polishing step and a second polishing step. The case where it comprises the grinding | polishing process of is included. Below, the matter explaining about a 1st grinding | polishing process and a 2nd grinding | polishing process is similarly applicable also to the other grinding | polishing process in the multistage grinding | polishing process using the said double-side polish apparatus.

  Both of the double-side polishing apparatuses used in the first polishing process and the second polishing process of the present invention are apparatuses capable of simultaneously polishing a plurality of substrates in the same batch. The total number of substrate holding holes provided in one (one) carrier in a double-side polishing apparatus used in a multi-stage polishing process including the first polishing process and the second polishing process is one or more (one, Or multiple). In the present invention, the total number of the substrate holding holes provided in one carrier in the double-side polishing apparatus used in the first polishing step is plural, and the substrate holding provided in one carrier in the double-side polishing apparatus used in the second polishing step. The total number of holes can be one each.

  The substrate manufacturing method according to the present invention functions more effectively when the center of the substrate holding hole in the carrier is in a positional relationship shifted from the center of the carrier (carrier rotation axis) (Configuration 5). Furthermore, in the substrate manufacturing method according to the present invention, the carrier is provided with a plurality of substrate holding holes, and the positions of all the substrate holding holes are shifted from the center of the carrier (carrier rotation axis). Especially if it works. When the center of the substrate holding hole is in a positional relationship shifted from the rotation axis of the carrier, the substrate held by the substrate holding hole continues to move on the surface plate in a complicated orbit during double-side polishing. Due to the above circumstances, during polishing, the upper surface plate oscillates although it is minute, and the parallelism constantly changes with the lower surface plate. Under these conditions, a large difference in surface shape (large difference in plate thickness) can cause the load distribution from the upper surface plate to be in an appropriate state when multiple substrates move simultaneously along a complicated track. It is difficult to continue to maintain, and it is difficult to make the surface shape of each substrate after polishing high flatness. As in the method for manufacturing a substrate according to the present invention, by aligning the surface shapes of the substrates to be polished in the same batch by the double-side polishing apparatus with close ones, the substrate moves complicatedly under the upper platen during polishing. In addition, the swing of the upper surface plate can be reduced. Thereby, it becomes easy to make the surface shape of the substrate polished by the same batch have high flatness.

  In the present invention, the number of substrate holding holes (one or a plurality) possessed by one carrier, the number of carriers arranged (mounted) in one double-side polishing apparatus, and the product (one) The total number of substrates that can be arranged (mounted) in the double-side polishing apparatus can be different in the plural stages of polishing processes including the first polishing process and the second polishing process, or the same number. The total number of substrates that can be polished in one batch in a double-side polishing apparatus used in a multi-stage polishing process including the first polishing process and the second polishing process can be the same, and the latter total is the total number of the former. It can also be relatively less. The size of the double-side polishing apparatus used in the first polishing step and the second polishing step (for example, the size of the diameter of the upper and lower surface plates) can be the same or equivalent, and the latter size is relative to the former size. It can also be made smaller. The plurality of carriers can have the same configuration or different configurations.

  In the double-side polishing apparatus used in the first polishing process and the second polishing process of the present invention, the former apparatus and the latter apparatus can use different apparatuses, and the former apparatus and the latter apparatus are the same. It is also possible to use the device (use the same device twice). When separate devices are used, the basic configuration as a double-side polishing device is the same, but various sizes such as top and bottom surface plates, the number of carriers, the number of substrate holding holes, and the total number of substrates that can be treated in a batch In addition to using different devices with different numbers, there are cases where two devices of the same type having the same various sizes and various numbers are prepared and used as separate devices. The double-side polishing apparatus used in the first polishing process and the second polishing process of the present invention includes a double-side polishing apparatus dedicated to the first polishing process used exclusively in the first polishing process, and a second polishing used exclusively in the second polishing process. A double-side polishing apparatus dedicated to the process can be used.

  In the substrate manufacturing method of the present invention, the multi-stage polishing process (including the first polishing process and the second polishing process) using the double-side polishing apparatus may be performed in a multi-stage polishing process using different polishing liquids. preferable. 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 main surface when polishing the main surface of the substrate.

  The substrate manufacturing method of the present invention is suitable for polishing the main surface of a glass substrate. When finishing the main surface of a glass substrate with high flatness and reduced defects as in the present invention, polishing with different properties can be achieved by performing double-side polishing of the glass substrate in a plurality of stages of polishing processes with different polishing liquids. It is preferable to use the liquid to gradually (gradually) make the main surface of the glass substrate to the desired surface shape accuracy and quality.

  The substrate manufacturing method of the present invention is particularly suitable for manufacturing a mask blank substrate. Due to the structure of the above double-side polishing apparatus, the polishing pad on the lower surface plate side is more susceptible to clogging or deterioration of the material due to the penetration of the polishing liquid than the polishing pad on the upper surface plate side. Change with time is fast. In addition, foreign matters such as polishing scraps and hardened abrasives tend to accumulate on the lower surface plate side. For these reasons, in a multi-stage polishing process using a double-side polishing apparatus, a thin film on which a transfer pattern is formed is provided when a transfer mask is manufactured from a mask blank substrate after the polishing process is performed. It is preferable to perform polishing by arranging the main surface on the side to be on the upper surface plate side.

  The substrate manufacturing method of the present invention includes a step of preparing a substrate parent set configured by collecting a plurality of substrates having two main surfaces. Here, since the said board | substrate is a board | substrate used by a double-sided grinding | polishing process, the board | substrate which finished the pre-process of the double-sided grinding | polishing process at least is used. As a pre-process of the double-side polishing step, predetermined chamfering processing of the end surface of the substrate, grinding processing of both main surfaces of the substrate, and the like are performed. When the total number of substrates that can be polished simultaneously in the same batch using the double-side polishing apparatus is the same in the first polishing step and the previous step, the substrate parent set is set by using all the substrates polished in the same batch in the previous step as they are. It is preferable to configure. Substrates polished in the same batch at the same time tend to have a plate thickness (thickness of the substrate) and in-plane variation that are close to each other and within a certain range. Because it is suitable.

  If the total number of substrates that can be polished simultaneously in the same batch is different between the first polishing step and the previous step, and if the total number of substrates in the previous step is large, collect a plurality of substrates from the substrates polished simultaneously in the same batch in the previous step It is preferable to constitute a substrate parent set.

  In the substrate manufacturing method of the present invention, all the substrates of the substrate parent set are arranged in the substrate holding holes of the carrier of the double-side polishing apparatus, and each substrate of the substrate parent set is formed using a first abrasive. A first polishing step of polishing the two main surfaces in the same batch. Also in the first polishing step, it is preferable to perform polishing by arranging the main surface of the substrate on which the transfer pattern is formed on the upper platen side.

  The substrate manufacturing method of the present invention measures the surface shape of at least one main surface of each substrate of the substrate parent set after performing the first polishing step, and the surface shape data of each substrate of the substrate parent set. A surface shape measuring step of acquiring Here, the surface shape (surface form) of the substrate can be measured using a surface shape measuring device. In the present invention, it is preferable that the surface shape measuring step measures the surface shape of the main surface on the side where at least the transfer pattern of each substrate of the substrate parent set is formed. In the present invention, the surface shapes of both main surfaces of each substrate of the substrate parent set can also be measured.

  The measurement of the surface shape (surface form) of the substrate can be generally performed by the following method. First, measurement points are arranged in a grid shape on the surface (main surface) of the measurement object, and height information of each measurement point (a reference plane at this time is, for example, a reference plane of a measurement apparatus) is used as a surface shape. Acquired by a measuring device. Next, based on the height information of each measurement point, a surface approximated by the least square method (least square plane) is calculated and used as a reference plane. Next, the height information of each measurement point is converted into the height of each measurement point with reference to the reference plane (least-square plane), and the result is obtained as information on the surface shape at each measurement point (surface Shape data). In addition, the area | region of the measurement point used in order to calculate a least squares plane can be the whole surface of a board | substrate. The area of the measurement point used for calculating the least square plane is not necessarily the entire surface of the substrate. The region of the measurement point for calculating the least square plane can be, for example, a central region excluding more than 2 mm to 10 mm from the periphery of the substrate.

  The substrate manufacturing method of the present invention is a step of dividing a substrate parent set into two or more substrate child sets based on the surface shape data of each substrate of the substrate parent set acquired in the surface shape measurement step (set dividing step) Have In the present invention, based on the measurement data of the surface shape of each substrate of the substrate parent set, for example, by a predetermined index (PV value, flatness, etc., which is the difference between the maximum height and the minimum height in a predetermined region) Each substrate of the substrate parent set is permuted in order of size, and the substrate parent set is divided into two or more substrate child sets by a method of dividing and grouping by a numerical range group of the predetermined index.

  The division from the substrate parent set to the substrate child set can be performed in units of a predetermined number. Further, the division from the substrate parent set to the substrate child set can be performed at an arbitrary ratio (for example, an arbitrary ratio (number of sheets such as 5: 5, 6: 4, 7: 3)). The number of substrate sets need not match the number that can be polished in one batch of the second polishing step, and may be insufficient.

  One feature of the substrate manufacturing method of the present invention is that each substrate in each substrate set divided in this set dividing step is a substrate in the same substrate parent set. By adopting such a method of configuring the substrate child set, it is possible to reduce the difference in plate thickness between the substrates constituting the substrate child set. Then, when each substrate of the substrate child set is polished by the second double-side polishing apparatus (second polishing step), the swing of the upper surface plate is reduced, and the surface shape of each substrate of the substrate child set after polishing is reduced. The variation of the can be greatly reduced. As a result, it is possible to increase the ratio at which a substrate with high flatness can be obtained from the substrate set after the second polishing step.

  In the substrate manufacturing method of the present invention, in the set dividing step, the flatness within a predetermined region of the main surface is calculated from the surface shape data of each substrate, and the two or more substrate parent sets are calculated based on the flatness. It is preferable to divide the substrate into a set of substrates (Configuration 2).

  In the present invention, for example, from the surface shape data of each substrate of the substrate parent set, a PV value in a predetermined region of each surface shape (difference between the maximum height and the minimum height in the predetermined region) is calculated. Can be used as an index of flatness. The predetermined area for calculating the flatness can be, for example, a rectangular inner area having sides of 132 mm, 142 mm, 148 mm, and the like with respect to the center of the main surface. The predetermined area for calculating the flatness is preferably a rectangular inner area having a side of 142 mm with respect to the center of the main surface.

  In general, the main surface shape of a substrate is broadly divided into a shape in which the central side tends to be higher than the outer peripheral side (convex shape) and a shape in which the outer peripheral side tends to be higher than the central side (concave shape). It is done. In the case of the present invention, when the PV value (flatness) is calculated from the surface shape, for example, the PV value (flatness) in the case of the convex surface shape is represented by a positive numerical value, and in the case of the concave surface shape. It is preferable that the convex shape and the concave shape can be distinguished by expressing the PV value (flatness) as a negative numerical value. Thereby, when dividing | segmenting into a several board | substrate child set from a board | substrate parent set using PV value and flatness as a parameter | index, it becomes easy to divide | segment into the board | substrate child set with which the board | substrate with the near surface shape was gathered. In the set dividing step using the PV value and the flatness as an index, a flatness threshold value is set to 1 or more, and the substrate parent set is divided into the substrate child set based on the threshold value.

  In the present invention, from the surface shape data of each substrate of the substrate parent set, an average surface shape thereof is calculated, and further, a difference between the average surface shape and the surface shape of each substrate of the substrate parent set. It is preferable to calculate a shape and divide the substrate parent set into the two or more substrate child sets based on the difference shape (Configuration 3).

  The difference shape (difference shape data) is, for example, a shape obtained by taking a difference between the average surface shape (surface shape data) and the surface shape (surface shape data) of each substrate of the substrate parent set. Say. The differential shape is, for example, a virtual shape obtained by subtracting the other surface shape data from one surface shape data. The PV value (difference between the maximum height and the minimum height in the predetermined area) within the predetermined area of the difference shape (for example, a rectangular inner area having sides of 132 mm, 142 mm, 148 mm or the like with respect to the center of the main surface) ) Is smaller, the smaller the deviation from the average surface shape, the smaller. In the case of the PV value calculated from the difference shape, as in the case of the PV value calculated from the main surface shape, the difference shape represents either a convex shape or a concave shape with a negative value. Is preferred.

In the calculation of the difference shape, it can be appropriately determined which surface shape (surface shape data) from which the other surface shape (surface shape data) is subtracted (subtracted).
Instead of the average surface shape (surface shape data), a virtual substrate having an ideal shape is set as the main surface shape of the substrate after the first polishing step, and the ideal surface shape (surface shape) The difference shape can also be calculated using (data).

  In the substrate manufacturing method of the present invention, one set (one set) of the substrate set is disposed in the substrate holding hole of the carrier of the double-side polishing apparatus, and one set is formed using a second abrasive. A second polishing step of polishing the two main surfaces of each substrate in the substrate batch set in the same batch. Also in the second polishing step, it is preferable to perform polishing by arranging the main surface on the side where the transfer pattern of the substrate is formed on the upper surface plate side.

  In the second polishing step, when the substrate parent set is divided into the two or more substrate child sets based on the difference shape, the PV value of the difference shape (the amount of deviation from the average surface shape) The processing allowance (required removal amount) for each substrate element set in the second polishing step is obtained based on the processing conditions (especially polishing time) according to the machining allowance, and for each substrate element set. It is preferable to perform the second polishing step and control the flatness of the glass substrate surface to a predetermined reference value or less (for example, 0.2 μm or less). The machining allowance (necessary removal amount) for each substrate element set in the second polishing step can be obtained by prediction or by obtaining a correspondence relationship in advance.

  The inventors of the present invention (1) improve the acquisition ratio of a substrate with high flatness by devising a method of dividing a substrate parent set into a substrate child set, and (2) for each substrate child set (each divided group). Polishing at the optimum time improves the acquisition ratio of high flatness substrates (the polishing effect is relatively low when the polishing time is the same), and (3) irregular when dividing from the substrate parent set to the substrate child set It has been found that the acquisition ratio of high flatness substrates is improved by eliminating unnecessary substrates. In the present invention, the polishing conditions (particularly the polishing time) for each substrate element set in the second polishing step are adjusted (including fine adjustment) by prediction or by obtaining a corresponding relationship in advance according to the division mode. can do.

  In the present invention, for example, as a result of measuring the surface shape of the substrate parent set, when there are many substrates that are considered to be unfavorable to perform the post-process (for example, 50% or more), The set can be divided as a set other than the substrate set, and the first polishing step (or equivalent step) can be performed again. A substrate considered to be preferably subjected to a post-process at the time of the division is divided as a substrate child set and a second polishing step is performed. Examples of the step before the division (first polishing step) include a step III polishing (ultra-precision polishing) step and a step II polishing (precision polishing) step which will be described later.

  In the substrate manufacturing method of the present invention, the multiple-stage polishing step using the double-side polishing apparatus includes a case where the polishing liquid performs the multiple-stage polishing process. In the substrate manufacturing method of the present invention, the polishing process at each stage in the multi-stage polishing process using the double-side polishing apparatus is a mode in which polishing is performed a plurality of times using the same double-side polishing apparatus (a mode in which polishing is performed again). Is included. Further, in each of the multi-stage polishing processes using the double-side polishing apparatus, each stage of the polishing process is performed multiple times using another double-side polishing apparatus (the number of processed batches of each apparatus is the same or different). (A mode in which polishing is performed again) is included.

  In the present invention, for example, an I-stage polishing (rough polishing) process, an II-stage polishing (precision polishing) process, an III-stage polishing (ultra-precision polishing) process, and an IV-stage polishing (final precision polishing) process, which will be described later, are performed. A mode in which the same polishing step is performed twice or a plurality of times is included. Note that when the same stage polishing process is performed a plurality of times, the number of polishing processes increases, but the burden is small because it is the same stage polishing process. In the present invention, a necessary process (stage of polishing process) can be added in the process of making and finishing in stages (gradually).

  In the method for producing a substrate of the present invention, for example, double-side polishing is performed in three stages such as a first stage polishing (rough polishing) process, a second stage polishing (precision polishing) process, and a third stage polishing (ultra-precision polishing) process. It can be performed in a stage polishing process. In this case, for example, cerium oxide, zirconium oxide or the like can be used as polishing abrasive grains in the first stage polishing (coarse polishing) process, and cerium oxide, zirconium oxide or the like as polishing abrasive grains in the second stage polishing (precision polishing) process. Silica (including colloidal silica) and the like can be used. Further, silica (including colloidal silica) or the like can be used as abrasive grains in the third stage polishing (ultra-precision polishing) process. Particularly, silica (colloidal) having a small average particle diameter and capable of chemical mechanical polishing. It is preferable to use (including silica). In the polishing process from the I stage to the III stage, it is preferable to perform polishing by disposing the main surface of the substrate on the side where the transfer pattern is formed on the upper platen side.

In the method for producing a substrate of the present invention, for example, double-side polishing is performed in a first stage polishing (rough polishing) process, a second stage polishing (precision polishing) process, a third stage polishing (ultra-precision polishing) process, and a fourth stage. It can be performed by a four-step polishing process such as a polishing (final precision polishing) process. In this case, the same abrasive grains as described above can be used in the I-stage polishing to the III-stage polishing process. In the stage IV polishing (final precision polishing) process, silica (including colloidal silica) or the like can be used as the abrasive grains, but the average particle size is particularly small (in the stage III polishing (ultra-precision polishing) process). It is preferable to use silica (including colloidal silica) that can be chemically and mechanically polished, and that the average particle diameter is relatively small relative to the average particle diameter of the silica used (including colloidal silica).
In the polishing process from the I stage to the IV stage, it is preferable to perform polishing by arranging the main surface of the substrate on which the transfer pattern is formed on the upper surface plate side.

  The silica used in the present invention (particularly the stage IV polishing step) 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 substrate (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.

  In the substrate manufacturing method of the present invention, the first abrasive is made of a material containing at least one of cerium oxide and zirconium oxide, and the second abrasive is made of a material containing colloidal silica. Included (Configuration 6).

  In this embodiment, the first abrasive is an abrasive having a tendency that the main surface of the substrate is polished into a concave shape, such as cerium oxide or zirconium oxide. The second abrasive is an abrasive having a tendency that the main surface of the substrate such as colloidal silica having a relatively large particle size is polished in a convex shape. In this embodiment, for example, the first abrasive is used in the above-described stage II polishing (precision polishing) process, and the second abrasive is used in the above-described stage III polishing (ultra-precision polishing) process. An example corresponds.

  In the substrate manufacturing method of the present invention, the first abrasive and the second abrasive are made of a material containing colloidal silica, and the average particle diameter of the colloidal silica of the first abrasive is the second An embodiment in which the particle size is relatively larger than the average particle size of the colloidal silica of the abrasive is included. In the substrate manufacturing method of the present invention, the first abrasive is made of colloidal silica produced by a water glass method, and the second abrasive is made of colloidal silica produced by a sol-gel method. Is included. Colloidal silica produced by the sol-gel method is more expensive than colloidal silica produced by the water glass method, and it is more cost-effective to use colloidal silica by the sol-gel method only for final polishing. is there.

  In these embodiments, a polishing process using an abrasive made of a material containing colloidal silica is performed in two stages. In this embodiment, for example, the first abrasive is used in the above-described stage III polishing (ultra-precision polishing) process, and the second abrasive is used in the above-described stage IV polishing (final precision polishing) process. Corresponds to this example. As a method of the stage IV polishing (final precision polishing) process, for example, polishing that improves the surface roughness while maintaining the surface shape (flatness) obtained in the stage III polishing (ultra-precision polishing) process. It can be a method and polishing conditions.

  In the substrate manufacturing method of the present invention, the one main surface is preferably the main surface on the side where the pattern forming thin film is formed when the mask blank is manufactured using the substrate (Configuration 7). .

  In the substrate manufacturing method of the present invention, it is preferable to perform double-side polishing of the glass substrate by setting the main surface of the glass substrate on which the transfer pattern is formed on the upper surface plate side. Thereby, the main surface on the side where the transfer pattern of the glass substrate to be polished by contact with the polishing pad on the upper surface plate side is formed has extremely high flatness and minute defects (concave defects, convex defects). The surface shape accuracy and quality can be improved. Therefore, a thin film serving as the transfer pattern can be formed on the main surface on the side where the transfer pattern of the glass substrate finished with good surface shape accuracy and quality is formed.

  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 notch mark is formed by cutting off three surfaces of one main surface and two end surfaces forming the corner portion into an oblique cross section in one of the rectangular corner portions of the substrate having a rectangular outer shape.

In the substrate manufacturing method of the present invention, the predetermined region is preferably a region including at least a rectangular inner region having a side of 132 mm with respect to the center of the main surface of the substrate (Configuration 4).
In the substrate manufacturing method of the present invention, it is more preferable that the predetermined region is a region including at least a rectangular inner region whose side is 142 mm with respect to the center of the main surface of the substrate.

  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 a mask blank substrate used for a transfer mask to which KrF excimer laser or ArF excimer laser is applied as exposure light, in a 142 mm × 142 mm region of the main surface on the side where the transfer pattern of the substrate is formed The flatness is preferably 0.2 μm or less, particularly preferably 0.1 μm or less. Further, in the case of a mask blank substrate used for a reflective mask to which EUV exposure light is applied, the flatness in the 142 mm × 142 mm region of the main surface on the side where the transfer pattern of the substrate is formed is 0.05 μm or less. It is preferable. In the case of the mask blank substrate used for the reflective mask, the main surface opposite to the side on which the transfer pattern is formed is a surface that is electrostatically chucked when set in the exposure apparatus, but the opposite is the case. The flatness in the 142 mm × 142 mm region of the main surface on the side is preferably 0.05 μm or less.

  Further, the surface roughness of the main surface of the mask blank substrate on which the transfer pattern is formed is preferably 0.25 nm or less in terms of root mean square roughness (RMS). More preferably, it is 2 nm or less, and further preferably 0.15 nm or less. Here, the surface roughness RMS is defined in Japanese Industrial Standard (JIS) B0601 (2001). In the present invention, it is not necessary to particularly limit the lower limit value of the surface roughness RMS, and as the surface of the substrate is smoother, the effects of the present invention are more remarkable.

  In the substrate manufacturing method of the present invention, examples of the substrate include a mask blank substrate and an imprint mold substrate. The mask blank substrate is not particularly limited when used for a binary mask blank or a phase shift mask blank, as long as it has transparency with respect to the exposure wavelength to be used. A substrate (for example, soda lime glass, aluminosilicate glass, etc.) is used. Among these, the synthetic quartz substrate is particularly preferably used because it is highly transparent in an ArF excimer laser or a shorter wavelength region.

In the case of EUV exposure, the mask blank substrate is preferably in the range of 0 ± 1.0 × 10 ⁻⁷ / ° C., more preferably 0 ± 0 in order to prevent pattern distortion due to heat during exposure. A material having a low thermal expansion coefficient within the range of 3 × 10 ⁻⁷ / ° C. is preferably used, and examples of the material having a low thermal expansion coefficient within this range include SiO ₂ —TiO ₂ glass and multicomponent glass ceramics. Etc. can be used. The mask blank substrate has a rectangular shape such as a square or a rectangle. In the case of the glass substrate for the mask blank for ArF excimer laser exposure and EUV exposure described above, a 6025 substrate (about 152 mm × about 152mm, thickness 6.35mm).

  As the imprint mold substrate, 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 is not particularly limited, such as a rectangular shape such as a square or a rectangle, or a circular shape, but a circular substrate is usually used for manufacturing patterned media, and a square shape is used for manufacturing optical components and semiconductor devices. A rectangular substrate such as a rectangle is used.

  The substrate manufacturing method of the present invention provides a great effect when the shape of the main surface of the substrate, which is a workpiece to be polished, is rectangular, and a greater effect when the shape of the main surface of the substrate is square. . When the main surface is polished with the above-mentioned double-side polishing apparatus on the substrate having the shape of these main surfaces, when an abrasive having a tendency that the main surface such as cerium oxide is polished into a concave shape is used, The four corners of the main surface are relatively hard to be polished, and the height of the four corners of the main surface tends to be high. Further, when an abrasive having a tendency that the main surface such as colloidal silica is polished into a convex shape is used, the four corners of the main surface of the substrate are likely to be relatively easily polished. It tends to be low. By applying the substrate manufacturing method of the present invention, it is possible to reduce the state where the four corners of the main surface of the substrate are not easily polished or are easily polished.

[Manufacturing method of mask blank]
The present invention also provides a mask blank manufacturing method including a step of forming a pattern forming thin film on the one main surface of the substrate manufactured by the substrate manufacturing method having the above configuration (configuration 8).
In the present invention, by manufacturing a mask blank using the substrate manufacturing method having the above configuration, the manufacturing yield of the mask blank substrate can be improved, and the production cost of the mask blank substrate can be reduced.

  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. The main surface in this case is a main surface set on the upper surface plate side in the above-described double-side polishing of the glass substrate. 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), a light shielding film formed of a material containing transition metal and silicon (Si) Binary mask blank comprising: a binary mask blank comprising a light shielding film formed of a material containing tantalum (Ta); a material containing silicon (Si); or a material containing transition metal and silicon (Si) And a phase shift mask blank including a phase shift film formed by the above method.

  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.

[Transfer Mask Manufacturing Method]
The present invention is a transfer mask manufacturing method using the mask blank manufactured by the mask blank manufacturing method having the above-described configuration, and includes a step of forming a transfer pattern on the pattern forming thin film by dry etching. A manufacturing method is also provided (Configuration 9).
By producing a transfer mask using the mask blank of the present invention, the production cost of the transfer mask can be reduced through the reduction of the production cost of the mask blank substrate by improving the production yield of the mask blank substrate.

  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. Further, in the phase shift type mask blank having a structure including the phase shift film or the phase shift film and the light shielding film on the main surface of the above-described mask blank glass substrate, by patterning the phase shift film serving as a transfer pattern, A phase shift 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]
The present invention also provides a method for producing an imprint mold substrate, a method for producing a mask blank using the imprint mold substrate, and a method for producing an imprint mold using the mask blank.
The mask blank in this case has a structure in which a hard mask film is formed on one main surface of the imprint mold substrate. This hard mask film is a film that becomes an etching mask when an imprint mold substrate such as a glass substrate is dug by an etching process. An imprint mold is produced by digging the imprint mold substrate by etching. In the mask blank in this case, the thin film to be the transfer pattern is the hard mask film. The present invention provides a method for manufacturing an imprint mold for manufacturing a semiconductor device (master mold), an imprint mold for manufacturing BPM (master mold), and the like.

Hereinafter, based on an Example, this invention is demonstrated more concretely.
Example 1
This example is a specific example of a method for manufacturing a mask blank substrate. This example includes the following steps. The double-side polishing apparatus used in this example and the comparative example uses a structure in which the upper platen described above and the rotating shaft are connected by a universal joint (that is, the upper platen is very small). It is configured to be able to swing.)

(1) Stage I polishing (coarse polishing) process The end surface of a synthetic quartz glass substrate (about 152 mm × 152 mm × 6.85 mm) is chamfered, and the glass substrate that has been subjected to grinding is subjected to a double-side polishing apparatus (per carrier) The number of substrates held is 4 and 5 carriers of the same configuration are arranged on the surface plate, so the number of substrates processed in 1 batch is 20. The main surface on the side where the transfer pattern is formed is the upper surface plate side. Twenty glass substrates were arranged so that the rough polishing was performed in the same batch under the following polishing conditions. 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) Stage II polishing (precision polishing) process (first polishing process of the present invention)
A glass substrate (20 sheets processed in the same batch) after finishing the above-mentioned stage I polishing (rough polishing) was used as a substrate parent set. Each glass substrate that constitutes the substrate parent set is a double-side polishing device (4 substrates are held per carrier, and 5 carriers of the same configuration are arranged on a surface plate, so that the number of substrates processed in one batch is 20. ), 20 glass substrates were arranged so that the main surface on the side where the transfer pattern was formed was on the upper surface plate side, and precision polishing was performed in the same batch under the following polishing conditions. 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 is immersed in a cleaning tank containing a hydrofluoric acid aqueous solution having a concentration of about 0.1% (ultrasonic application), and then the cleaning tank ( It was immersed in (pure water) (applied with ultrasonic waves), washed, and dried.

(Measurement of main surface shape)
Surface shape (surface form) of the main surface (main surface on the side on which the transfer pattern is formed) of each glass substrate of the substrate parent set (20 sheets processed in the same batch) after the above-mentioned stage II polishing (precision polishing) Was measured with a surface shape measuring device (UltraFlat 200M manufactured by Corning Tropel). The area for acquiring the surface shape was 142 mm × 142 mm. In the following examples and comparative examples, the surface shape measuring device and the measurement region were the same.

(Division of substrate)
In this embodiment, the average surface shape of these substrates is calculated from the surface shape data of each substrate of the substrate parent set (20 sheets), and this average surface shape and each substrate of the substrate parent set (20 sheets) are calculated. The difference shape between these surface shapes was calculated, and the PV value of these difference shapes (the amount of deviation from the average surface shape is shown) was calculated. The PV value calculated from the difference shape is a negative value representing a numerical value with a convex difference shape. The PV values (for 20 sheets) of these differential shapes were arranged in descending order of the PV value, and 5 pieces from the top were taken and divided into 4 (4 sets) substrate sets.

(Preliminary experiment)
In addition, the same process as Example 1 is performed a plurality of times (for example, 10 times) in advance, and the average value of the PV values for each of the plurality of divided substrate sets and the third stage polishing (ultra-precision) The relationship with the polishing time (necessary removal amount) for increasing the acquisition rate of a glass substrate having a flatness of 0.2 μm or less in the (polishing) step (second polishing step of the present invention) was obtained in advance.

(3) Stage III polishing (ultra-precision polishing) process (second polishing process of the present invention)
For each substrate set of the four divided substrate sets, the glass substrate is sequentially subjected to a double-side polishing apparatus (the number of substrates held per carrier is one, and five carriers of the same configuration are arranged on a surface plate, thus 1 The number of substrates processed in a batch is 5). Five glass substrates are arranged so that the main surface on the side where the transfer pattern is formed becomes the upper surface plate side, and ultra-precision polishing is performed under the following polishing conditions. . In addition, the carrier used is a position in which the center of the substrate holding hole is shifted from the center, and the double-side polishing apparatus has an upper and lower surface plate area used in the stage II polishing process. And the same.

Four substrates (one set of 5 sheets) were processed four times in succession (4 batches), and a total of 20 glass substrates were subjected to ultra-precision polishing. The processing load and polishing time are appropriately adjusted so as to obtain the surface roughness (RMS (root mean square roughness) of 0.25 nm or less) necessary for a glass substrate used for the binary mask blank for ArF excimer laser exposure. I went. After this adjustment (after conditions are set), the processing load and polishing time between batches are usually the same (unless there is a specific reason). The polishing time was adjusted so that the acquisition rate of flatness (0.2 μm or less) was increased in consideration of the deviation amount.
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 flatness (flatness calculated on the inner side of a square with a side of 142 mm with respect to the center of the substrate) of the main surface (the main surface of the glass substrate on which the transfer pattern is formed) of the obtained glass substrate is It measured with the surface shape measuring apparatus. As a result, 19 out of 20 glass substrates had a flatness of 0.2 μm or less. Moreover, when the process similar to the present Example 1 was performed in multiple times (10 times) (200 sheets in total), the acquisition rate of the glass substrate with a flatness of 0.2 μm or less was 94% of the whole.

  In Example 1, the plate thickness was not measured after the stage II polishing (precision polishing) step (first polishing step of the present invention). However, it was confirmed that the glass substrate obtained through the stage III polishing (ultra-precision polishing) process (the second polishing process of the present invention) was within a predetermined plate thickness range. It is an advantage of the present invention that it is not necessary to measure the plate thickness when dividing the substrate parent set.

(Example 2)
Example 2 is the same as Example 1 except that the polishing times of the four substrate child sets in the stage III polishing (ultra-precision polishing) process (second polishing process of the present invention) in Example 1 are the same. It was. As a result, the number of glass substrates having a flatness of 0.2 μm or less was 16 out of 20. Moreover, when the process similar to the present Example 2 was performed in multiple times (10 times) (200 sheets in total), the acquisition rate of the glass substrate with a flatness of 0.2 μm or less was 82% of the whole.

(Example 3)
In Example 3, the substrate parent set (20 sheets in one set, the same batch process) after finishing the stage II polishing (precision polishing) process (the first polishing process of the present invention) in Example 1 was used. Thus, without calculating the difference between the average shape and the average shape, the PV value of the surface shape of each glass substrate (the PV value calculated in the inner region of a square whose side is 142 mm with respect to the center of the substrate) Simply, arrange the PV values in descending order, and take 5 pieces from the top, 4 substrate sets (5 pieces per set, 1 piece of substrate per carrier, carrier of the same configuration on the surface plate) The arrangement was the same as in Example 1 except that the number of substrates processed in one batch was five, and the number of substrates processed in one batch was five. In addition, PV value shall represent the numerical value of convex shape with the negative numerical value. As a result, the number of glass substrates having a flatness of 0.2 μm or less was 15 out of 20. Moreover, when the process similar to this Example 4 was performed in multiple times (10 times) (a total of 200 sheets), the acquisition rate of the glass substrate with a flatness of 0.2 μm or less was 76% of the whole.

(Comparative Example 1)
In Comparative Example 1, a set of 20 glass substrates (the number of substrates held per carrier is 4) after finishing the II-stage polishing (precision polishing) process (the first polishing process of the present invention) in Example 1 is fixed. A double-side polishing machine (with 4 substrates held per carrier) without dividing the substrate for 5 carriers having the same configuration on the board, so the number of substrates processed in one batch is 20. On the surface plate, five carriers with the same configuration are arranged, so the number of substrates processed in one batch is 20), and the glass substrate main surface on which the transfer pattern is formed is arranged on the upper surface plate side. The same procedure as in Example 1 was performed except that ultraprecision polishing was performed under the following polishing conditions.
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)

  As in Example 1, as a result of measuring the flatness, the number of glass substrates having a flatness of 0.2 μm or less was 13 out of 20. Moreover, when the process similar to this comparative example 1 was performed in multiple times (10 times) (total 200 sheets), the acquisition rate of the glass substrate with a flatness of 0.2 μm or less was 68% of the whole. Comparative Example 1 is an example in which a relatively good acquisition rate can be obtained in the conventional polishing process, but it is difficult to further improve the acquisition rate by simply performing multiple stages of polishing as in Comparative Example 1. is there.

(Comparative Example 2)
In Comparative Example 2, the second stage polishing (precise polishing) step (first polishing step of the present invention) in Example 1 was performed a plurality of times (10 times, 10 batches), and the same batch was obtained from a total of 200 substrates. The set of substrates was broken and reconfigured into a set of substrates with the same substrate shape and thickness. Other than that was the same as Example 1.

  As a result, the expected result was not obtained. As in Example 1, adjustment of the polishing time was attempted, but the expected effect was not obtained. From these things, it is considered better to divide with priority on board thickness and board thickness (to divide the same batch without breaking it) than to divide both board shape and board thickness. . The measurement of the plate thickness and the plate thickness variation takes time, and the process of breaking the batch from 200 substrates into a set with the same substrate shape and plate thickness also takes time. Is considered good. If the number of substrates is increased to a considerable number, the expected result may be obtained, but it is not realistic. As can be seen from Comparative Example 2, the effect of the present invention cannot be obtained if the shapes are made as uniform as possible when a plurality of substrates are divided.

(Comparative Example 3)
In Comparative Example 3, in Example 1, a set of 20 glass substrates (four substrates held per carrier) that finished the II-stage polishing (precision polishing) step (first polishing step of the present invention), For five carriers with the same configuration on the surface plate, so that the number of substrates processed in one batch is 20), the four substrates held on the same carrier are the same without measuring the surface shape of each glass substrate. (4 pieces of 5 carriers are used and 4 pieces are set in 4 pieces of substrate pieces), and 4 to 4 substrates held by the remaining one carrier. Four board child sets (one set of five, one board holding number per carrier, and five identically arranged carriers arranged on a surface plate, so that one sheet is inserted into each child set, thus one batch The number of substrates processed is 5). Except that, it was the same as in Example 1. As a result, the number of glass substrates having a flatness of 0.2 μm or less was 8 out of 20. Moreover, when the process similar to this comparative example 3 was performed in multiple times (10 times) (a total of 200 sheets), the acquisition rate of the glass substrate with a flatness of 0.2 μm or less was 40% of the whole.

(Comparative Example 4)
In Comparative Example 4, the surface of each glass substrate with respect to the substrate parent set (20 sheets in the same batch process) after finishing the second stage polishing (precision polishing) step (first polishing step of the present invention) in Example 1 Without measuring the shape, it is totally random, 4 substrate sets (5 per set, 1 substrate holding per carrier, 5 identically arranged carriers on the surface plate, thus 1 The number of substrates processed in a batch was 5). As a result, the number of glass substrates having a flatness of 0.2 μm or less was 5 out of 20. Moreover, when the process similar to this comparative example 4 was performed in multiple times (10 times) (total 200 sheets), the acquisition rate of the glass substrate with a flatness of 0.2 μm or less was 23% of the whole.

DESCRIPTION OF SYMBOLS 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

Claims (9)

Using a double-side polishing apparatus comprising an upper surface plate and a lower surface plate that rotate coaxially with each other, and a plurality of carriers provided with one or more substrate holding holes between the upper surface plate and the lower surface plate, A substrate manufacturing method comprising a plurality of polishing steps in which a substrate is disposed and the two main surfaces of the substrate are polished simultaneously,
Preparing a substrate parent set configured by collecting a plurality of substrates having two main surfaces;
The substrates of the substrate parent set are arranged in the substrate holding holes of the carrier of the first double-side polishing apparatus, and two main components in each substrate of the substrate parent set are used with a polishing liquid containing a first abrasive. A first polishing step for polishing the surface in the same batch;
A surface shape measuring step of measuring a surface shape of at least one main surface of each substrate of the substrate parent set after performing the first polishing step, and acquiring surface shape data of each substrate of the substrate parent set;
Based on the surface shape data of each substrate of the substrate parent set acquired in the surface shape measurement step, a set dividing step of dividing the substrate parent set into two or more substrate child sets;
One set of the substrate child sets is disposed in the substrate holding hole of the carrier of the second double-side polishing apparatus, and a polishing liquid containing a second abrasive is used to remove the one set of substrate child sets. And a second polishing step of polishing the two main surfaces of each substrate in the same batch.   The set dividing step calculates a flatness within a predetermined region of the main surface from the surface shape data of each substrate, and divides the substrate parent set into the two or more substrate child sets based on the flatness. The method of manufacturing a substrate according to claim 1, wherein   The set division step calculates the average surface shape of each substrate from the surface shape data of each substrate, and further calculates a difference shape between the average surface shape and the surface shape of each substrate. The substrate manufacturing method according to claim 1, wherein the substrate parent set is divided into the two or more substrate child sets on the basis of the difference shape in the inside.   4. The method of manufacturing a substrate according to claim 2, wherein the predetermined region is a region including at least a rectangular inner region having a side of 132 mm with respect to the center of the main surface of the substrate.   The substrate manufacturing method according to claim 1, wherein a center of the substrate holding hole is shifted from a center of the carrier. The first abrasive is made of a material containing at least one of cerium oxide and zirconium oxide,
The substrate manufacturing method according to claim 1, wherein the second abrasive is made of a material containing colloidal silica.   7. The substrate according to claim 1, wherein the one main surface is a main surface on a side where a pattern forming thin film is formed when a mask blank is manufactured using the substrate. Manufacturing method.   A method for producing a mask blank, comprising a step of forming a thin film for pattern formation on the one main surface of the substrate produced by the method for producing a substrate according to claim 1. A method for producing a transfer mask using a mask blank produced by the method for producing a mask blank according to claim 8,
A method for producing a transfer mask, comprising a step of forming a transfer pattern on the pattern forming thin film by dry etching.

Images