JP2014110368A - Nanoimprint method and method for manufacturing patterned substrate using the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve reproducibility of a pattern feature by a simple method in the manufacture of a patterned substrate such as a template.SOLUTION: A mold 10 having a first mark 51 including a first shift amount evaluation mark 51a, and a reference substrate 20 having a second mark 52 including a second shift amount evaluation mark 52a are prepared. The first and second shift amount evaluation marks 51a, 52a, respectively, are combined to form a pair to represent a relative positional relationship to each other. The mold 10 is deformed in such a manner that each set of the corresponding first and second shift amount evaluation marks 51a, 52a is shifted to a target amount from a reference state, based on the relative positional relationship between the corresponding first and second shift evaluation marks 51a, 52a. After each set reaches the target shift state, a concavo-convex pattern 11 is transferred onto a resist 25 applied to a process target substrate 21 while the shift state is maintained.

Description

  The present invention relates to a nanoimprint method using a mold having a fine concavo-convex pattern on its surface and a method for producing a patterned substrate using the same.

  With the miniaturization of semiconductor devices, exposure apparatuses and masks of conventional photolithography (or electron beam lithography) technology are becoming expensive. Therefore, nanoimprint lithography that can form a high-resolution fine pattern at a low cost has attracted attention.

  Some semiconductor devices have a three-dimensionally complicated circuit pattern in which layers having complicated patterns are stacked. Therefore, in order to improve the yield in the manufacture, it is important to reduce an overlay error (hereinafter also simply referred to as an error) when the patterns between the upper and lower layers are aligned.

  In recent years, instead of using nanoimprint for all lithography of layers necessary for semiconductor device manufacturing, we have developed a mix-and-match method that uses conventional photolithography and nanoimprint for each layer according to the complexity of the pattern. Adoption is also being considered. Generally, a pattern formed by photolithography has a shape distortion caused by a photomask (design accuracy, distortion, etc.) and an exposure apparatus (stage alignment accuracy, mechanical deformation, optical system design accuracy, etc.). Often occurs. Therefore, in the case of forming a pattern by nanoimprint on the lower layer on which the pattern is formed by photolithography by the mix and match method, only by adjusting the overall positional relationship between the lower layer pattern and the upper layer pattern, Since it cannot cope with the distortion of the pattern shape of the lower layer, there is a problem that the error cannot be sufficiently reduced for the pattern in the region away from the alignment mark.

  For example, as a method for reducing the error, methods disclosed in Patent Documents 1 to 3 are known.

  Specifically, in Patent Document 1 or 2, a part or all of the side surface portion of the template is compressed by an actuator so that the pattern in the template (or mold) is reduced or deformed by a predetermined amount. A method of transferring a pattern by pressing a deformed template against a resist coated on a substrate is disclosed. According to this method, when the actual pattern of the template is larger than the pattern to be transferred, the error can be reduced in that the actual pattern can be corrected in a reduction direction.

  In Patent Document 3, a plurality of silicon masters having a pattern shape in which the shift amount is changed from the reference pattern shape is manufactured by photolithography, and a glass template is reproduced from each silicon master by nanoimprinting. Then, a method is disclosed in which a template whose error is less than or equal to an allowable value in relation to the pattern shape of the lower layer is selected from the templates, and nanoimprint lithography is performed on the lower layer with the selected template.

  In particular, according to the method of Patent Document 3, since a template is copied from a silicon master, it is not necessary to perform a time-consuming photolithography process many times, and nanoimprinting can be performed at low cost. Furthermore, since an appropriate template is selected from templates having different pattern shapes in accordance with the distortion of the pattern shape of the lower layer, the pattern alignment between the upper and lower layers can be performed with high accuracy.

Special table 2008-504141 JP 2012-160635 A JP 2011-258605 A

  However, in recent years when high-precision alignment is required with the miniaturization of the pattern shape, there are cases where the error cannot be sufficiently reduced even by the method of Patent Document 3.

  Specifically, in the method of Patent Document 3, pattern shapes each having a different shift amount from a reference pattern shape are formed on a plurality of silicon substrates by photolithography, and these are used as silicon masters. However, since the silicon master is more likely to be distorted and has a larger thermal expansion coefficient than the quartz master, for example, even if a template is duplicated from the same silicon master, it is transferred to the template due to differences in the holding method and temperature conditions during the duplication. In some cases, the pattern shape to be generated is different for each template. Therefore, when manufacturing a group of templates having different pattern shapes, it is still more difficult to ensure the accuracy and reproducibility of pattern shape duplication by using a plurality of silicon masters. In order to reduce such an influence by the method of Patent Document 3 and to stably supply a template that can make an error between upper and lower layers at the time of imprint less than an allowable value, a strict temperature control mechanism or An apparatus having a precise substrate holding mechanism is required, but this makes it difficult to provide an inexpensive duplicate template.

  The present invention has been made in view of the above problems, and in the manufacture of a patterned substrate such as a template, a nanoimprint method and a patterned substrate that can improve the reproducibility of replicating a pattern shape by a simple method. The object is to provide a manufacturing method.

In order to solve the above problems, a nanoimprint method according to the present invention includes:
A mold having a fine concavo-convex pattern and a plurality of first marks each including a first deviation amount evaluation mark, and a second deviation amount evaluation mark paired with the first deviation amount evaluation mark, A second mark including a second deviation amount evaluation mark that is combined with the first deviation amount evaluation mark and represents a relative positional relationship with the first deviation amount evaluation mark corresponds to a plurality of first marks. Prepare a reference board with multiple in position,
The mold and the reference substrate are arranged in correspondence with the plurality of first marks and the plurality of second marks,
Based on the relative positional relationship between the corresponding first and second deviation amount evaluation marks, each of the corresponding first and second deviation amount evaluation marks has been displaced from the reference state by a desired deviation amount. Deform the mold so that it is in the target deviation state,
After each of the groups reaches a target deviation state, the uneven pattern is transferred to a resist applied on the substrate to be processed while the state is maintained.

  That is, according to the present invention, imprinting is performed by deforming a mold so as to realize a predetermined shift amount with reference to one reference substrate. In particular, if the amount of deviation is changed variously, it is possible to manufacture (replicate) a plurality of templates having different pattern shapes using one mold.

  In the nanoimprint method according to the present invention, it is preferable that the first and second deviation amount evaluation marks that form a pair represent a positional relationship in two different directions.

  In the nanoimprint method according to the present invention, the first and second misalignment evaluation marks that make a pair are preferably a combination of a pair of moire interference marks that make a pair, or a lattice mark and a mark on the lattice mark. A combination of symbol marks indicating a position is preferable, or a combination of a scale mark and a symbol mark indicating a position on the scale mark is preferable.

  In the nanoimprint method according to the present invention, it is preferable that the first and second marks include a pair of alignment marks for alignment.

  Moreover, in the nanoimprint method according to the present invention, it is preferable to maintain the atmospheric conditions when each of the above groups reaches the target deviation state when transferring the uneven pattern.

  In the nanoimprint method according to the present invention, the first and second marks are preferably made of a metal material.

  The nanoimprint method according to the present invention is preferably performed in a helium atmosphere.

A method for producing a patterned substrate according to the present invention includes:
Forming a resist film on which a concavo-convex pattern is transferred by the nanoimprint method described above on a substrate to be processed,
By etching the substrate to be processed using the resist film as a mask, a concavo-convex pattern corresponding to the concavo-convex pattern transferred to the resist film is formed on the substrate to be processed.

  The nanoimprint method and the method for manufacturing a patterned substrate according to the present invention include a mold having a first mark including a first shift amount evaluation mark, and a second mark including a second shift amount evaluation mark. A substrate is prepared, and based on the relative positional relationship between the corresponding first and second deviation amount evaluation marks, each of the corresponding first and second deviation amount evaluation mark sets is desired from the reference state. The mold is deformed so as to be in a target deviation state, which is a deviation state by a deviation amount, and the resist applied to the substrate to be processed is maintained while maintaining the state after each of the above groups has reached the target deviation state. The uneven pattern is transferred. That is, according to the present invention, since the mold is deformed and imprinted so as to realize a predetermined deviation amount with reference to one reference substrate, the pattern shape between the templates can be obtained even if a silicon master is used, for example. The problem of variation is less likely to occur. That is, the reproducibility of the pattern shape can be improved in the manufacture of a patterned substrate such as a template.

It is a schematic sectional drawing which shows the process of the nanoimprint method of embodiment. It is a flowchart which shows the process of the nanoimprint method of embodiment. It is the schematic which shows the mark for position shift evaluation provided in a mold or a reference board | substrate. It is the schematic which shows a mode when the mark for position shift evaluation is piled up. It is the schematic which shows the mark for moire interference as a deviation | shift amount evaluation mark. It is the schematic which shows the other structure of the deviation | shift amount evaluation mark. It is the schematic which shows the relationship between the number of mark points and the pattern shape which can be identified. It is the schematic which shows the relationship between the number of mark points and the pattern shape which can be identified. It is the schematic which shows the example of arrangement | positioning of the mark for position shift evaluation. It is the schematic which shows the pattern shape of the mold before and behind a deformation | transformation. It is the schematic which shows a mode that a deformation | transformation is added to a mold on the basis of a reference board | substrate. It is the schematic which shows the determination method of the target pattern shape of a mold.

  Hereinafter, although an embodiment of the present invention is described using a drawing, the present invention is not limited to this. In addition, for easy visual recognition, the scale of each component in the drawings is appropriately changed from the actual one.

  FIG. 1 is a schematic cross-sectional view illustrating steps of the nanoimprint method according to the embodiment. FIG. 2 is a flowchart showing the steps of the nanoimprint method of the embodiment, and FIG. 3 is a conceptual diagram showing first and second marks for evaluating misalignment in the present embodiment.

  As shown in FIGS. 1 and 2, in the nanoimprint method of the present embodiment, a mold 10 having a fine uneven pattern 11 and a plurality of first marks 51 for positional deviation evaluation, and for positional deviation evaluation. A reference substrate 20 having a plurality of second marks 52 is used. The first and second marks 51 and 52 are provided on the mold 10 and the reference substrate 20 respectively so that when the mold 10 and the reference substrate 20 are overlaid, the marks 51 and 52 have a corresponding positional relationship. It has been. Each of the first marks 51 includes a first deviation amount evaluation mark 51a, and each of the second marks 52 is a second deviation amount evaluation paired with the first deviation amount evaluation mark 51a. A mark 52a is included. The first and second deviation amount evaluation marks 51a and 52a forming a pair are combined with each other so as to overlap each other, for example, and serve as an index representing a relative positional relationship. The first and second deviation amount evaluation marks 51a and 52a forming a pair are, for example, combinations of marks for forming a moire interference pair as shown in FIG. 3 in the present embodiment. In the present embodiment, each of the first and second marks 51 and 52 includes a pair of alignment marks 51b and 52b for alignment.

  In the nanoimprint method of the present embodiment, first, the mold 10 and the reference substrate 20 are associated with the plurality of first marks 51 and the plurality of second marks 52 using the mold 10 and the reference substrate 20 as described above. (FIG. 1a and STEP 1), and alignment of the mold 10 and the reference substrate 20 is performed using the alignment marks 51b and 52b (STEP 2). When the alignment of the mold 10 and the reference substrate 20 is completed, the corresponding first and second deviation amount evaluation marks 51a and 52a are provided for the respective mark positions where the first and second marks 51 and 52 are provided. With reference to the relative positional relationship, for example, the actuator 10 applies an external force F to the mold 10 to deform the mold 10 (FIG. 1b and STEP 3).

Here, for each mark position, the positional deviation between the corresponding first and second deviation amount evaluation marks 51a and 52a is in the relative positional relationship between the corresponding first and second deviation amount evaluation marks 51a and 52a. Based on this, it is evaluated whether or not the desired deviation amount (target deviation amount) has been reached from the reference state (STEP 4 to STEP 7). For example, in this embodiment, the deviation amounts ΔX _{1 to} ΔX _n of the moire interference peaks are measured at each mark position in STEP 4, and in STEP 5, the moire interference mark is calculated based on the deviation amounts ΔX _{1 to} ΔX _n from the relational expression described later. Deviation amounts Δx _{1 to} Δx _n are calculated. Then, the deviation amounts Δx _{1 to} Δx _n are compared with the target deviation amount, and it is evaluated whether or not these values match. If it is determined that the values do not match as a result of the evaluation, the process returns to STEP 3 for changing the external force application condition.

  Note that the reference state means a state serving as a reference when the change amount or the changed state is defined. That is, regarding the shift amounts of the corresponding first and second shift amount evaluation marks 51a and 52a, the reference state means a state in which there is no shift amount when they are aligned, for example, moire interference. In the present embodiment using the mark for use, it is assumed that there is no deviation of the moire interference peak. The term “reference state” is used for the individual deviation amount evaluation marks 51a, 52a and the marks 51, 52, and also means that there is no deviation amount for all of the plurality of marks. Further, the shape of the concave / convex pattern of the mold is also used in the sense that there is no deformation. The marks 51 and 52 are observed using an observation means 31 such as an optical microscope or a camera.

  If it is evaluated by the above evaluation that the positional deviation of the deviation amount evaluation marks 51a and 52a has not reached the target deviation amount, the condition for applying the external force F is changed, the mold 10 is deformed again, and the deviation amount evaluation mark is obtained. This process is repeated until the positional shifts 51a and 52a reach the target shift amount. On the other hand, when each of the corresponding pairs of the first and second deviation amount evaluation marks 51a and 52a is in a target deviation state, which is a state of deviation by the target deviation amount, that is, all the above groups are in a target deviation state. When the pattern 11 has a target pattern shape (target pattern shape), the reference substrate 20 is carried out (STEP 8).

  Thereafter, the substrate 21 to be processed is carried in (FIG. 1c and STEP 9), the concavo-convex pattern 11 is pressed and exposed to the resist film 25 applied on the substrate 21 to be processed (FIG. 1d), and the mold 10 is removed from the resist film 25. After peeling (FIG. 1e), the concavo-convex pattern 11 is transferred to the resist film 25 (STEP 10).

(mold)
The mold 10 is a quartz mold in which a concavo-convex pattern is formed on a quartz substrate by photolithography or electron beam lithography, for example. In particular, as shown in FIG. 1b, a quartz mold is preferable when the amount of displacement is measured through the mold 10. However, for example, when the amount of deviation is measured through the reference substrate 20, the mold 10 does not need to be transparent. In such a case, a mold made of silicon or a metal material can also be used. However, since silicon and metal materials are easily distorted and have a large thermal expansion coefficient, it is preferable to use a quartz mold.

  The shape of the concavo-convex pattern of the mold 10 is not particularly limited, and is appropriately selected according to the use of nanoimprint. For example, a typical pattern is a line & space pattern. And the length of the convex part of a line & space pattern, the width | variety of a convex part, the space | interval of convex parts, and the height (the depth of a recessed part) of a convex part from a recessed part bottom face are set suitably. For example, the width of the convex portion is 10 to 100 nm, more preferably 20 to 70 nm, the interval between the convex portions is 10 to 500 nm, more preferably 20 to 100 nm, and the height of the convex portion is 10 to 500 nm. Preferably it is 30-100 nm. In addition, the shape of the convex portions constituting the concavo-convex pattern may be a shape in which dots having cross sections such as a rectangle, a circle, and an ellipse are arranged.

(Release agent)
In the present invention, it is preferable to perform a mold release treatment on the surface of the mold 10 in order to improve the mold releasability between the resist and the mold 10. As a mold release agent used for the mold release treatment, Opkin (registered trademark) DSX manufactured by Daikin Industries, Ltd., Novec (registered trademark) EGC-1720 manufactured by Sumitomo 3M Co., Ltd., and the like are used as fluorine-based silane coupling agents. .

  In addition, known fluorine resins, hydrocarbon lubricants, fluorine lubricants, fluorine silane coupling agents, and the like can be used.

  For example, PTFA (polytetrafluoroethylene), PFA (tetrafluoroethylene / perfluoroalkyl vinyl ether copolymer), FEP (tetrafluoroethylene / hexafluoropropylene copolymer), ETFE (tetrafluoroethylene Ethylene copolymer).

  For example, hydrocarbon lubricants include carboxylic acids such as stearic acid and oleic acid, esters such as butyl stearate, sulfonic acids such as octadecyl sulfonic acid, phosphate esters such as monooctadecyl phosphate, stearyl alcohol and oleyl Examples thereof include alcohols such as alcohol, carboxylic acid amides such as stearamide, and amines such as stearylamine.

  For example, examples of the fluorine-based lubricant include a lubricant in which part or all of the alkyl group of the hydrocarbon-based lubricant is substituted with a fluoroalkyl group or a perfluoropolyether group.

For example, as the perfluoropolyether group, perfluoromethylene oxide polymer, perfluoroethylene oxide polymer, perfluoro-n-propylene oxide polymer (CF ₂ CF ₂ CF ₂ O) _n , perfluoroisopropylene oxide polymer ( CF (CF ₃ ) CF ₂ O) _n or a copolymer thereof. Here, the subscript n represents the degree of polymerization.

For example, the fluorinated silane coupling agent has at least 1, preferably 1 to 10, alkoxysilane groups and chlorosilane groups in the molecule, and preferably has a molecular weight of 200 to 10,000. For example, the alkoxysilane group, -Si _{(OCH 3)} 3 _{group, -Si (OCH 2 CH 3)} 3 group. Examples of the chlorosilane group, and a -Si _{(Cl) 3} group. Specifically, heptadecafluoro-1,1,2,2-tetra-hydrodecyltrimethoxysilane, pentafluorophenylpropyldimethylchlorosilane, tridecafluoro-1,1,2,2-tetra-hydrooctyltriethoxy Compounds such as silane and tridecafluoro-1,1,2,2-tetra-hydrooctyltrimethoxysilane.

(Reference board)
The reference substrate 20 functions as a ruler for measuring the deformation amount of the shape of the uneven pattern 11 from the reference state. The plurality of second marks 52 are provided at positions on the reference substrate 20 corresponding in design to the plurality of first marks 51 in the standard state of the concave-convex pattern 11. The material, shape and structure of the reference substrate 20 are not particularly limited. However, it is preferable to use quartz having excellent shape stability. The distance when the mold 10 and the reference substrate 20 are brought close to each other is preferably 100 μm or less, more preferably 10 μm or less, and particularly preferably 1 μm or less. This is because the evaluation accuracy of the shift amount is improved as the distance between the mold 10 and the reference substrate 20 is narrower. Further, these parallelism errors when the mold 10 and the reference substrate 20 are brought close to each other are preferably 0.02 degrees or less, more preferably 0.002 degrees or less, and 0.0002 degrees or less. It is particularly preferred. This is because, when the distance between the mold 10 and the reference substrate 20 is narrow, these parallelism errors may be in contact with each other and damaged.

(Mark for misalignment evaluation)
The first and second marks 51 and 52 for evaluating the positional deviation are indices for measuring the deformation amount of the shape of the concavo-convex pattern 11 from the reference state. For the first and second marks 51 and 52, a known method for forming a mark such as a vapor deposition method can be employed. Moreover, it is preferable that the 1st and 2nd marks 51 and 52 are comprised from a metal material from a viewpoint of visibility improvement. Further, for the mark on the side where the observation light is incident (observation means 31 side), it is preferable that the metal film is left only on the mark portion (left mark 51 in FIG. 3) in order to improve the light transmittance. On the other hand, for the mark on the side that reflects the observation light (the side far from the observation means 31), in order to increase the reflected light intensity, the metal film is removed from the ground of the metal film and the removed portion is used as the mark. Preferred (right mark 52 in FIG. 3). However, if the entire surface of the reference substrate is covered with a metal film, distortion due to film stress may occur, so only the peripheral area of the mark is the ground of the metal film, and the metal film in the area that has little effect on the visibility of the mark It is preferable to remove. The thickness of the metal film is preferably 10 nm or more, more preferably 30 nm or more, and particularly preferably 50 nm or more.

  The first and second marks 51 and 52 include deviation amount evaluation marks 51a and 52a indicating the deformation amount (deviation amount) of the pattern shape for each mark position as means for actually measuring the deformation amount. In the present embodiment, for example, it is a pair of moire interference marks. FIG. 4 is a schematic diagram showing a state when the first and second marks are superimposed. Specifically, the left diagram of FIG. 4 shows a state in which the first and second marks 51 and 52 are in a reference state without deviation, and the right diagram of FIG. This shows a state in which the shift amount Δx is shifted relative to the mark 52 in the x direction (for example, the horizontal direction in FIG. 4) and the shift amount Δy in the y direction (for example, the vertical direction in FIG. 4). It can be seen that interference fringes occur according to the amount of deviation as shown in the right diagram of FIG. In the present embodiment, two moiré interference marks are provided for the x-axis and the y-axis so that deviation amounts in two different directions can be efficiently measured. The two different directions do not need to be orthogonal.

When the mark for moire interference is used, the specific deviation amounts Δx and Δy are obtained as follows. FIG. 5 is a schematic diagram showing a moire interference mark as a deviation evaluation mark. For example, the upper left diagram and the upper right diagram in FIG. 5 represent the deviation amount evaluation mark 51 a provided on the mold 10, and the left middle and right middle diagrams in FIG. 5 represent the deviation amount evaluation mark 52 a provided on the reference substrate 20. The shift amount evaluation mark 51a has an array of lines with different pitches arranged vertically, and the shift amount evaluation mark 52a has an array of lines arranged vertically in the reverse order. Incidentally, each of the pitch of the array of lines and p ₁ and p _2. The lower left figure shows a state where the upper left figure and the left middle figure are superimposed, and the lower right figure shows a state where the upper right figure and the right middle figure are superimposed. At this time, since the upper right diagram and the middle right diagram are shifted by the shift amount Δx (or Δy), the interference fringe peak shifts in the lower right diagram. If the amount of deviation of the interference fringe peak is ΔX (or ΔY), ΔX (or ΔY) can be obtained by the following equation 1.

  Therefore, ΔX (or ΔY) is measured by an observation means having a length measurement function (for example, an optical microscope having a length measurement scale in the observation visual field or a camera connected to an image display means having a length measurement scale). Thus, the shift amount Δx (or Δy) can be calculated.

  In addition, regarding the deviation amount evaluation marks 51a and 52a, the case where the moire interference mark is used has been described above. However, the deviation amount evaluation mark in the present invention measures the deformation amount of the shape of the concavo-convex pattern 11 from the reference state. The mark is not particularly limited as long as the mark can be used. For example, FIG. 6 is a schematic diagram illustrating another configuration of the deviation evaluation mark. Specifically, FIG. 6a shows a combination 53 of a grid mark 53a with a scale and a symbol mark 53b indicating a position on the grid mark 53a. On the other hand, FIG. 6b shows a combination 54 of a scale mark 54a and a symbol mark 54b indicating a position on the scale mark 54a. In the case of the above-described combination 53, it is possible to efficiently measure the amount of deviation in two different directions for one combination. In addition, by providing two combinations 54 for the x-axis and y-axis, it is possible to efficiently measure different two-direction deviation amounts in the combination 54 as well. Further, the symbol marks 53b and 54b are not limited to the shapes shown in the figure, and may be other symbols such as arrows.

  There may be at least two combinations of the first and second marks for evaluating misalignment (that is, the mark positions) in order to measure the deformation amount of the concavo-convex pattern 11. For example, when the shape of the concavo-convex pattern 11 is uniformly reduced as a whole according to compression by the external force F, the degree of reduction can be grasped by the distance between two mark positions. Of course, when more complicated deformation is targeted, it is preferable to set more mark positions in order to accurately measure the deformation amount of the concavo-convex pattern 11. The number of mark positions is preferably 8 or more, more preferably 16 or more, and particularly preferably 40 or more. For example, FIGS. 7 and 8 are schematic diagrams showing the relationship between the number of mark points and the identifiable pattern shape. As shown in FIG. 7, at the four mark positions, the shapes 12d and 12e can be identified, respectively. On the other hand, the shapes 12a, 12b and 12c cannot be identified at the four mark positions. Therefore, in order to identify the shapes 12a, 12b, and 12c, for example, eight mark positions may be set. However, even if eight mark positions are set, for example, the shapes 13a and 13b as shown in FIG. 8 cannot be identified. Therefore, in order to identify the shapes 13a and 13b, for example, 16 mark positions may be set.

  The mark positions by the first and second marks 51 and 52 are preferably arranged uniformly around the concavo-convex pattern 11 in order to accurately measure the deformation of the pattern. For example, FIG. 9A shows a state in which eight mark positions are set evenly around the concavo-convex pattern 11, and FIG. 9B shows a state in which 16 mark positions are set evenly around the concavo-convex pattern 11. . FIG. 9c shows a state in which eight mark positions are set evenly around the concave / convex pattern 11 and 16 mark positions are set evenly therearound.

  In addition, when the first and second marks include alignment marks, alignment to the initial position before evaluation of the shift amount is facilitated. Furthermore, since the alignment mark and the deviation evaluation mark can be observed at the same time, there is an advantage that the alignment and the deviation evaluation can be performed simultaneously.

(External force application method)
The method of applying the external force F is, for example, a method as shown in Patent Documents 1 and 2, that is, the side surface portion of the mold 10 so that the uneven pattern 11 of the mold 10 is reduced by a predetermined amount or deformed to a predetermined shape. It is possible to employ a method in which a part or all of these are compressed by the actuator 30. For example, consider a mold 10 as shown in FIG. A pattern as shown in FIG. 10B is formed on the mold 10 in FIG. 10A, and eight first marks M1 to M8 are provided around the pattern. When it is desired to transfer a pattern in which the middle of the left side and the right side of the pattern is curved toward the center and is uniformly reduced in the vertical direction using the mold 10 as shown in FIG. Apply external force as follows. Each of the first marks M1 to M8 is made to correspond to each of the second marks R1 to R8 (that is, Mi is made to correspond to Ri, i = 1,..., 8). Thereafter, the side surface of the mold 10 is compressed by the actuator 30 while referring to the displacement amount Δi of Mi with respect to Ri. At this time, as shown in FIG. 11, an external force F is applied only to the middle of the mold portions corresponding to the left and right sides of the pattern, and the external force F is equally applied to the mold portions corresponding to the upper and lower sides of the pattern. Give. Then, when the deviation amount Δi reaches a desired value, the increase in the external force F is stopped and the strength of the external force F at that time is maintained. By performing the above operation for all i, the desired deformation of the mold 10 is completed.

(Setting of deviation)
The shift amount is set in advance for each mark position according to the pattern to be nanoimprinted. In particular, since handling becomes easy, it is preferable to set the deviation amount Δx of the x-axis component and the deviation amount Δy of the y-axis component separately in the x-axis direction and the y-axis direction as in this embodiment.

  In addition, a plurality of patterns 40a to 40l and 41a to 41l are already formed on a substrate (silicon wafer in FIG. 12) on which nanoimprinting is performed as in the manufacture of semiconductor devices, and the plurality of patterns 40a to 40l and 41a to 41l are If they should originally have the same shape but have different distortions, for example, the target pattern shape of the pattern (target deviation state of the first and second marks) is determined as follows. Is preferred. First, an average shape (common component) of the plurality of patterns 40a to 40l and 41a to 41l is calculated based on the magnification and the degree of deformation of each of the plurality of patterns 40a to 40l and 41a to 41l. On the other hand, for each of the plurality of patterns 40a to 40l and 41a to 41l, a deviation amount from the average shape is calculated as an inherent component of the deviation amount peculiar to each pattern, and it is necessary in consideration of, for example, 3σ of the inherent component. Determine the appropriate magnification. Then, a shape obtained by enlarging the average shape at the enlargement magnification is set as a target pattern shape of the pattern. By using the target pattern shape obtained by the above method, the error between the upper and lower layers at the time of imprinting can be less than the allowable value with respect to the whole of the plurality of patterns 40a to 40l and 41a to 41l. The template can be manufactured. In the above description, the average shape is calculated between a plurality of wafers. However, for example, the average shape may be calculated only within a single wafer, or the average shape may be calculated between processing batches.

(Substrate to be processed)
For silicon molds, a quartz substrate is preferred to allow exposure to the resist. The quartz substrate is appropriately selected according to the purpose without particular limitation as long as it has light transparency and a thickness of 0.3 mm or more. For example, the surface of a quartz substrate is coated with a silane coupling agent, the quartz substrate is laminated with a metal layer made of Cr, W, Ti, Ni, Ag, Pt, Au, or the like, or the quartz substrate is CrO _2. , WO ₂ , TiO _2, etc. laminated metal oxide film layers, and the laminated body surface coated with a silane coupling agent. The thickness of the metal layer or metal oxide film layer is usually 30 nm or less, preferably 20 nm or less. This is because when the thickness exceeds 30 nm, the UV transmittance is lowered and the resist is hard to be cured.

  In addition, the above “having optical transparency” specifically means that the resist film is formed when light is incident from the other surface of the substrate so as to be emitted from one surface on which the resist film is formed on the substrate. Means that the light is sufficiently cured, and at least the transmittance of light having a wavelength of 200 nm or more from the other surface to the one surface is 5% or more.

  The thickness of the quartz substrate is usually preferably 0.3 mm or more. If it is 0.3 mm or less, it is likely to be damaged by pressing during handling or imprinting.

  On the other hand, the substrate to be processed for the quartz mold is not particularly limited with respect to its shape, structure, size, material, etc., and can be appropriately selected according to the purpose. For example, in the case of an information recording medium, the shape is a disk shape. The structure may be a single layer structure or a laminated structure. The material can be appropriately selected from those known as substrate materials, and examples thereof include silicon, nickel, aluminum, glass, and resin. These board | substrate materials may be used individually by 1 type, and may use 2 or more types together. The substrate may be appropriately synthesized or a commercially available product may be used. Moreover, what coat | covered the surface with the silane coupling agent may be used. There is no restriction | limiting in particular as thickness of a board | substrate, Although it can select suitably according to the objective, 0.05 mm or more is preferable and 0.1 mm or more is more preferable. If the thickness of the substrate is less than 0.05 mm, the substrate may be bent when the resist and the mold are in close contact with each other, and a uniform contact state may not be ensured.

  Furthermore, in the present invention, a mold having a plateau shape and / or a substrate to be processed can be used. In this case, the pattern area is transferred onto the mesa portion 22 of the mesa structure (for example, FIG. 1c). In the present invention, a mold having a dig on the back surface and / or a substrate to be processed can also be used. In the present embodiment, the processing target substrate 21 has a dug 23.

(Resist)
The resist is not particularly limited, but in this embodiment, for example, a photopolymerization initiator (about 2% by mass) and a fluorine monomer (0.1 to 1% by mass) are added to a polymerizable compound. Materials can be used.

  Moreover, antioxidant (about 1 mass%) can also be added as needed. The material prepared by the above procedure can be cured by ultraviolet light having a wavelength of 360 nm. For those having poor solubility, it is preferable to add a small amount of acetone or ethyl acetate for dissolution, and then distill off the solvent.

Examples of the polymerizable compound include benzyl acrylate (Biscoat (registered trademark) # 160: manufactured by Osaka Organic Chemical Co., Ltd.), ethyl carbitol acrylate (Biscoat (registered trademark) # 190: manufactured by Osaka Organic Chemical Co., Ltd.), polypropylene glycol di In addition to acrylate (Aronix (registered trademark) M-220: manufactured by Toagosei Co., Ltd.), trimethylolpropane PO-modified triacrylate (Aronix (registered trademark) M-310: manufactured by Toagosei Co., Ltd.), etc. The compound A etc. which are represented can be mentioned.
Structural formula 1:

  The polymerization initiator may be 2- (dimethylamino) -2-[(4-methylphenyl) methyl] -1- [4- (4-morpholinyl) phenyl] -1-butanone (IRGACURE®). 379: manufactured by Toyotsu Chemiplus Co., Ltd.) and the like.

Moreover, as said fluorine monomer, the compound B etc. which are represented by following Structural formula 2 etc. can be mentioned.
Structural formula 2:

  As shown in FIG. 1, it is preferable to provide an adhesion layer 24 in order to improve the adhesion between the resist 25 and the substrate 21 to be processed.

(Imprint process)
Before bringing the mold 10 and the resist 25 into contact, it is preferable to reduce the residual gas by reducing the atmosphere between the mold 10 and the substrate 21 to a reduced pressure or vacuum atmosphere. However, in a high vacuum atmosphere, the curable resin before curing volatilizes, and it may be difficult to maintain a uniform film thickness. Therefore, it is more preferable to reduce the residual gas by setting the atmosphere between the mold 10 and the substrate 21 to a He atmosphere or a reduced pressure He atmosphere. Since He permeates the quartz substrate, the trapped residual gas (He) gradually decreases. Since it takes time to permeate He, it is more preferable to use a reduced pressure He atmosphere. The reduced pressure atmosphere is preferably 1 to 90 kPa, and particularly preferably 1 to 10 kPa.

  In the present invention, it is preferable to maintain the atmospheric conditions when each of the corresponding sets of first and second deviation amount evaluation marks reaches the target deviation state when the concavo-convex pattern is transferred. This is because when the atmospheric conditions change after each of the above groups has reached the target deviation state, the deformation state of the mold 10 changes due to expansion or contraction of the mold 10. In addition, the atmospheric conditions should just maintain the atmospheric | air pressure and temperature by an atmosphere at least.

  The mold and the substrate coated with the resist are brought into contact with each other after being aligned so as to have a predetermined relative positional relationship. An alignment mark is preferably used for alignment. The alignment mark is formed in a concavo-convex pattern that can be detected by an optical microscope, moire interferometry, or the like. The alignment accuracy is preferably 10 μm or less, more preferably 1 μm or less, and still more preferably 100 nm or less.

  The pressing pressure of the mold 10 is in the range of 100 kPa to 10 MPa. When the pressure is larger, it is easier to follow the surface shapes of the mold 10 and the substrate 2 and the flow of the resist is promoted. Furthermore, when the pressure is high, removal of residual gas, compression, dissolution of the residual gas into the resist, and transmission of He through the quartz substrate are promoted, leading to an improvement in the quality of the resist pattern. However, if the applied pressure is too strong, there is a possibility that the mold 10 and the substrate may be damaged when a foreign object is caught when the mold 10 contacts. Therefore, the pressing pressure of the mold 10 is preferably 100 kPa to 5 MPa, and particularly preferably 100 kPa to 1 MPa. The reason why the pressure is set to 100 kPa or more is that when imprinting is performed in the atmosphere, when the space between the mold 10 and the substrate is filled with liquid, the pressure between the mold 10 and the substrate is pressurized at atmospheric pressure (about 101 kPa). is there.

  After the mold 10 is pressed to form a resist film, the resist is cured by exposure with light containing a wavelength matched to the polymerization initiator contained in the resist. As a method of releasing after curing, for example, the back surface or outer edge portion of either the mold 10 or the substrate is held, and the back surface or outer edge portion of the other substrate or mold is held, and the holding portion or back surface of the outer edge is held. A method of moving the holding part relative to the direction opposite to the pressing can be mentioned.

  As described above, according to the present invention, since the mold is deformed and imprinted so as to realize a predetermined shift amount based on one reference substrate, there is a problem that the pattern shape varies between templates. Hard to happen. As a result, in the manufacture of a patterned substrate such as a template, the reproducibility of the pattern shape can be improved by a simple method. In particular, if the amount of deviation is variously changed, it is possible to duplicate a plurality of templates having different pattern shapes using one mold.

"Method for manufacturing patterned substrate"
Next, an embodiment of a method for producing a patterned substrate (template as a mold duplicate) will be described. In the present embodiment, a duplicate of the mold 10 is manufactured using the nanoimprint method described above.

  First, using the nanoimprint method, a pattern-transferred resist film is formed on one surface of a substrate to be processed. Next, dry etching is performed using the pattern-transferred resist film as a mask, and a concavo-convex pattern corresponding to the concavo-convex pattern formed on the resist film is formed on the substrate to be processed to obtain a substrate having a predetermined pattern. .

  The dry etching is not particularly limited as long as it can form a concavo-convex pattern on the substrate to be processed, and can be appropriately selected according to the purpose. For example, ion milling, reactive ion etching (RIE), sputter etching , Etc. Among these, ion milling and RIE are particularly preferable.

  The ion milling method is also called ion beam etching and introduces an inert gas such as Ar into an ion source to generate ions. This is accelerated through the grid, and collides with the sample substrate for etching. Examples of the ion source include a Kaufman type, a high frequency type, an electron impact type, a duoplasmatron type, a Freeman type, an ECR (electron cyclotron resonance) type, and the like.

  Ar gas can be used as a process gas in the ion milling method, and fluorine-based gas or chlorine-based gas can be used as an etchant for RIE.

  As described above, according to the method for manufacturing a patterned substrate of the present invention, the uneven pattern is transferred to the resist film by nanoimprinting that can improve the reproducibility of the pattern shape. It becomes possible to improve the reproducibility of the shape.

DESCRIPTION OF SYMBOLS 10 Mold 11 Uneven | corrugated pattern 20 Reference substrate 21 Processing target substrate 25 Resist film 30 Actuator 31 Observation means 51 First mark 51a for displacement evaluation First displacement evaluation mark 51b Alignment mark 52 Second for displacement evaluation Mark 52a Second deviation evaluation mark 52b Alignment mark F External force

Claims (10)

A mold having a fine concavo-convex pattern and a plurality of first marks each including a first deviation amount evaluation mark, and a second deviation amount evaluation mark paired with the first deviation amount evaluation mark, A plurality of first marks include a second mark including a second deviation amount evaluation mark that is combined with the first deviation amount evaluation mark and represents a relative positional relationship with the first deviation amount evaluation mark. Prepare a plurality of reference boards at positions corresponding to
Disposing the mold and the reference substrate in correspondence with the plurality of first marks and the plurality of second marks;
Based on the relative positional relationship between the corresponding first and second deviation amount evaluation marks, each of the corresponding first and second deviation amount evaluation marks has been displaced from the reference state by a desired deviation amount. The mold is deformed so that the target deviation state is in a state,
A nanoimprint method for transferring the concavo-convex pattern to a resist applied on a substrate to be processed while maintaining the state after each of the groups reaches the target deviation state.   The nanoimprint method according to claim 1, wherein the first and second deviation amount evaluation marks forming a pair represent a positional relationship in two different directions.   3. The nanoimprint method according to claim 1, wherein the first and second deviation amount evaluation marks forming a pair are a combination of a pair of moire interference marks forming a pair.   3. The nanoimprint method according to claim 1, wherein the first and second deviation amount evaluation marks forming a pair are a combination of a lattice mark and a symbol mark indicating a position on the lattice mark.   3. The nanoimprint method according to claim 1, wherein the first and second deviation amount evaluation marks forming a pair are a combination of a scale mark and a symbol mark indicating a position on the scale mark.   6. The nanoimprint method according to claim 1, wherein the first and second marks include a pair of alignment marks for alignment.   The nanoimprint method according to any one of claims 1 to 6, wherein when the uneven pattern is transferred, an atmosphere condition when each of the groups reaches the target deviation state is also maintained.   The nanoimprint method according to claim 1, wherein the first and second marks are made of a metal material.   The nanoimprint method according to claim 1, wherein the nanoimprint method is performed in a helium atmosphere. Forming a resist film on which a concavo-convex pattern has been transferred by the nanoimprint method according to claim 1 on a substrate to be processed;
Etching the substrate to be processed using the resist film as a mask to form a concavo-convex pattern corresponding to the concavo-convex pattern transferred to the resist film on the substrate to be processed. .

Images