JP7767164B2 - Fixing jig for laminate and method for manufacturing aluminum nitride single crystal substrate - Google Patents

Description

This disclosure relates to the separation of an aluminum nitride single crystal body from a laminate.

Aluminum-based Group III nitride semiconductors made from aluminum nitride have a wide range of band gap energy values. These nitride semiconductors can be made into alloy semiconductors of any composition, and depending on the alloy composition, they can take on various band gap values.

Therefore, by using aluminum-based Group III nitride semiconductors, it is, in principle, possible to create light-emitting devices that emit a wide range of light, from infrared to ultraviolet. In particular, light-emitting devices using aluminum-based Group III nitride semiconductors are capable of emitting light at short wavelengths, such as in the ultraviolet region.

Like conventional semiconductor light-emitting devices, light-emitting devices using aluminum-based Group III nitride semiconductors can be manufactured by sequentially stacking thin films of single-crystal semiconductors (specifically, thin films that will become an n-type semiconductor layer, a light-emitting layer, and a p-type semiconductor layer) each several microns thick on a substrate. Known methods for forming such thin films of single-crystal semiconductors include crystal growth methods such as molecular beam epitaxy (MBE) and metalorganic chemical vapor deposition (MOCVD).

When manufacturing aluminum-based Group III nitride semiconductor light-emitting devices, sapphire substrates are commonly used due to their crystalline quality as a substrate, UV transparency, mass productivity, and cost. However, when an aluminum-based Group III nitride semiconductor light-emitting device is formed on a sapphire substrate, differences in lattice constant and thermal expansion coefficient between the sapphire substrate and the aluminum-based Group III nitrides that make up each layer of the semiconductor light-emitting device can cause crystal defects (misfit dislocations) and cracks, which can degrade the light-emitting performance of the device.

To solve these problems, it is desirable to use a substrate with a lattice constant and thermal expansion coefficient close to those of the aluminum-based Group III nitride semiconductor thin film that constitutes the device. Aluminum nitride single crystal substrates are suitable as substrates for growing aluminum-based Group III nitride semiconductor thin films.

When using aluminum nitride single crystal as a substrate, it is preferable for the single crystal to have a certain thickness (for example, 10 μm or more) from the perspective of mechanical strength. MOCVD is suitable for producing aluminum nitride single crystal substrates because it allows faster crystal growth than MBE. Furthermore, hydride vapor phase epitaxy (HVPE) and physical vapor transport (PVT) are known as growth methods for aluminum nitride single crystal that allow even faster crystal growth than MOCVD.

Wafer wafers are obtained by cutting aluminum nitride single crystals grown on a base substrate into plates, and aluminum nitride single crystal substrates are obtained by polishing the wafers to remove damaged layers from the surface. A wire saw is commonly used as a means of separating (cutting, slicing) aluminum nitride single crystals into plates (see Patent Document 1).

International Publication No. 2016/076270 Pamphlet

However, when cutting aluminum nitride single crystals with a wire saw, deviation from the target position in the thickness direction could occur during cutting. Also, while crystal orientation is an important factor in the manufacture of semiconductor devices, cut pieces could be produced with significant deviations in crystal orientation.

The objective of this disclosure is to suppress deviation from the target position and deviation from the orientation in the thickness direction when separating an aluminum nitride single crystal body.

After extensive research, the inventors discovered that when an aluminum nitride single crystal is fixed to a fixture with an adhesive in order to separate (cut, slice) it, the thickness of the adhesive between the aluminum nitride single crystal and the mounting surface of the fixture causes it to deviate from the target position. They further discovered that variations in the in-plane thickness of the adhesive layer lead to deviations in the crystal orientation, and completed the technology of the present disclosure by providing specific means to resolve this issue.

One aspect of the present disclosure is a fixture used when separating an aluminum nitride single crystal body from a laminate, the fixture comprising a mounting surface on which an adhesive layer that adheres the laminate body is placed, the mounting surface having at least one recess.

An adhesive layer may be provided on the mounting surface, and when the laminate is adhered to the adhesive layer, the recess may be formed at a position where at least a portion of the recess overlaps with the laminate in a plan view.

Another aspect of the present disclosure is a method for manufacturing an aluminum nitride single crystal substrate, including a preparation step of preparing a laminate including an aluminum nitride single crystal body, a placement step of placing the laminate on the placement surface of a fixing jig via an adhesive layer formed from an adhesive, and a separation step of separating the aluminum nitride single crystal body from the laminate bonded to the adhesive layer, wherein the placement step includes guiding a portion of the adhesive present between the placement surface and the laminate to a position lower than the placement surface in the thickness direction of the fixing jig.

In the separation process, the aluminum nitride single crystal may be cut into plates from the laminate using a wire saw.

In the separation process, a buffer layer is formed on the side of the aluminum nitride single crystal, and then the buffer layer is brought into contact with a wire saw before cutting the aluminum nitride single crystal.

According to the present disclosure, it is possible to suppress the degree of deviation from the target separation position and deviation in the crystal orientation of the aluminum nitride single crystal body in the thickness direction of the aluminum nitride single crystal body.

FIG. 1 is a diagram showing the flow of the manufacturing method S10 of an aluminum nitride single crystal substrate. FIG. 2( a ) is a plan view of the laminate 10 , and FIG. 2( b ) is a front view of the laminate 10 . 3A and 3B are diagrams for explaining the placing step S12, in which FIG. 3A is a plan view and FIG. 3B is a front view. FIG. 4A is a plan view of the fixture 20A, and FIG. 4B is a front view of the fixture 20A. FIG. 5A is a plan view of the fixing jig 20B, and FIG. 5B is a front view of the fixing jig 20B. FIG. 6(a) is a plan view of the fixing jig 20C, and FIG. 6(b) is a front view of the fixing jig 20C. FIG. 7A is a plan view of the fixing jig 20D, and FIG. 7B is a front view of the fixing jig 20D. FIG. 8 is a diagram illustrating the separation step S13. 9A and 9B are diagrams for explaining the separation step S13, in which FIG. 9A is a plan view and FIG. 9B is a front view. FIG. 10 is a diagram for explaining the positions at which the height of the aluminum nitride single crystal was measured.

1. Manufacturing Method of Aluminum Nitride Single Crystal Substrate A manufacturing method S10 of an aluminum nitride single crystal substrate according to one embodiment of the present disclosure includes a preparation step S11, a placement step S12, and a separation step S13, as shown in Fig. 1. Each step will be described below. Note that in the explanation of the steps, examples of fixing jigs and the like will also be described.

1.1. Preparation process S11
In the preparation step S11, a laminate including an aluminum nitride single crystal (AlN single crystal) to be separated (cut, sliced) is prepared.
Fig. 2(a) is a plan view schematically illustrating a laminate 10 including an aluminum nitride single crystal body, and Fig. 2(b) is a front view of the laminate 10. As shown in Figs. 2(a) and 2(b), the laminate 10 has a first surface 10a and a second surface 10b opposite to the first surface 10a. A thickness is defined between the first surface 10a and the second surface 10b, and a side surface 10c is formed so as to connect the outer periphery of the first surface 10a to the outer periphery of the second surface 10b. As can be seen from Figs. 2(a) and 2(b), the laminate 10 in this embodiment is disk-shaped.
The thickness of the laminate 10 is not particularly limited, but taking into consideration the cutting allowance by the wire saw, the thickness of the aluminum nitride single crystal layer is preferably 500 μm or more, and more preferably 800 μm or more.

The laminate 10 may be a single layer of aluminum nitride single crystal as shown in Figures 2(a) and 2(b), or it may be a laminate in which an aluminum nitride single crystal is laminated on another substrate.

1.2. Placement step S12
In the placing step S12, the laminate 10 including the aluminum nitride single crystal produced in the preparing step S11 is placed on the placing surface of the fixing jig 20 via an adhesive layer 11 formed with an adhesive. That is, as shown in the plan view of Fig. 3(a) and the front view of Fig. 3(b), the laminate 10 is fixed by being placed on the fixing jig 20 via the adhesive layer 11 with the second surface 10b of the laminate 10 facing the placing surface 20a of the fixing jig 20.

1.2.1. Adhesive Layer Any known adhesive can be used for the adhesive layer 11, and there are no limitations on the adhesive that can be used. Examples of adhesives that can be used include epoxy-based, cyanoacrylate-based, urea resin-based, melamine resin-based, phenolic resin-based, resorcinol resin-based, vinyl acetate resin-based, isocyanate-based, silicone resin-based, modified silicone resin-based, urethane resin-based, acrylic resin-based, polyester-based, and polyethylene-based adhesives.

There are two types of adhesives: those that use chemical reactions as the driving force behind adhesion, such as epoxy and cyanoacrylate adhesives, and those that use thermoplastic resins such as polyester and polyethylene. Of these, epoxy adhesives, which use chemical reactions as the driving force behind adhesion, are particularly easy to obtain and inexpensive, have little change in volume before and after the reaction, and allow for stable hardening. Epoxy adhesives are preferred because the resin after the reaction is highly hard, making them useful for stable cutting.

Epoxy adhesives come in a variety of types, including two-component, one-component, and heat-curing types. Two-component adhesives are more useful because they can be handled at room temperature and are less expensive. Among the two-component adhesives, Nikka Seiko's Q Bond and U Bond are particularly useful.

There are no restrictions on the viscosity of the adhesive, but the dynamic viscosity is preferably 0.01 cSt or more and 10,000 cSt or less. It is more preferable that it be 0.1 cSt or more and 1,000 cSt or less, even more preferable that it be 0.4 cSt or more and 100 cSt or less, and most preferable that it be 0.8 cSt or more and 10 cSt or less.

When applying adhesive, the surface of the fixture must be kept clean before application. If it is not clean, the adhesive's fixing ability may be reduced, or the adhesive surface may become tilted due to the inclusion of foreign matter. Effective methods for keeping the surface of the fixture clean include removing foreign matter with a blower, wiping with a clean wipe, and removing organic matter with chemicals such as acetone or alcohol.

Tools that can be used to apply the adhesive include spatulas, dispensers, bar coaters, and the like, without limitation.
When using an adhesive that hardens over time, such as a typical two-component epoxy adhesive, the viscosity may change between mixing the adhesive and bonding, so it is preferable to monitor the time to prevent variations and complete the work within a certain time.
When using thermoplastic resins, it is preferable to perform bonding when the temperature is uniform. Therefore, it is preferable to monitor the temperature using a rod thermometer, thermocouple, radiation thermometer, etc. Compared to metals, which are generally used in fixing jigs, resins have low thermal conductivity and tend not to heat up as easily.

When placing a laminate on the adhesive layer, it is preferable to apply an appropriate pressure to the laminate after placement. Applying pressure causes excess adhesive layer to flow from the outer periphery of the laminate or through holes in the fixture, resulting in uniform adhesion. There is no limit to the pressure to be applied, but the pressure is preferably 0.001 MPa to 1 MPa, more preferably 0.002 MPa to 0.5 MPa, or even more preferably 0.0025 MPa to 0.05 MPa. The time for which pressure is applied is optional, but it is preferable to keep it constant. The time for which pressure is applied is preferably 5 seconds to 24 hours, more preferably 20 seconds to 1 hour, or 1 minute to 10 minutes.
It is preferable to apply pressure to the entire surface of the laminate simultaneously. If there is a flow path that guides part of the applied adhesive to a position lower than the mounting surface in the thickness direction of the fixture, excess adhesive can be efficiently removed.

When curing adhesives, it is necessary to select the appropriate curing method depending on the type of adhesive being used. For example, with a typical two-component epoxy adhesive, curing progresses over time, so an appropriate amount of time must be allowed. With a thermosetting adhesive, a certain temperature must be applied. Furthermore, with thermoplastic resins, it is necessary to ensure that the adhesive has cooled sufficiently and that the temperature has dropped.

1.2.2 Fixing Jig In this step, during the process of placing the laminate on the mounting surface of the fixing jig, a portion of the adhesive present between the mounting surface and the laminate is guided (moved) away from between the mounting surface and the laminate in the thickness direction (vertical direction) of the fixing jig to a position lower than the mounting surface. This adjusts the amount of adhesive remaining between the mounting surface and the laminate, thinning the adhesive layer 11 and preventing displacement of the aluminum nitride single crystal body from the target position when it is separated from the laminate 10 in the separation step S13 described below.
The present invention efficiently guides excess adhesive between the laminate and the fixing jig, thereby enabling precise fixing and cutting of the laminate. The details of how the amount of adhesive remaining between the mounting surface and the laminate is adjusted by guiding (moving) a portion of the adhesive between the mounting surface and the laminate in the thickness direction (vertical direction) of the fixing jig away from the gap between the mounting surface and the laminate and guiding (moving) it to a position lower than the mounting surface are unclear, but the inventors speculate as follows: Frictional stress occurs between the adhesive, the laminate, and the fixing jig according to Newton's law of viscosity. With a conventional fixing jig, the adhesive is guided to the outer edge of the laminate, and guiding requires a work load proportional to the frictional stress and the distance the adhesive travels. Providing an opening shortens the distance the adhesive travels, reducing the required work load. As a result, the adhesive layer becomes thinner, enabling more precise fixing.

While there are no particular limitations on the means for guiding the adhesive, one specific example is to provide a recess on the mounting surface of the fixing jig 20. The shape of the fixing jig is described below.

1.2.2a. Fixture 20A
4 shows a diagram for explaining the fixing jig 20A. To distinguish it from other types of fixing jigs, it is written as fixing jig 20A, but the fixing jig 20A can be used as the fixing jig 20 described above.
Fig. 4(a) is a plan view of the fixing jig 20A, and Fig. 4(b) is an A-A' cross-sectional view of the fixing jig 20A. As can be seen from Fig. 4, the fixing jig 20A has a mounting surface 21A and a back surface 22A on the opposite side thereof, with a thickness between the mounting surface 21A and the back surface 22A, and a side surface 23A formed so as to bridge the outer periphery of the mounting surface 21A and the outer periphery of the back surface 22A. In this embodiment, the fixing jig 20A is disk-shaped, but it does not necessarily have to be disk-shaped.

The fixing jig 20A has multiple holes 24A formed in the mounting surface 21A as recesses opening into the mounting surface 21A. In this embodiment, the holes 24A are through-holes that penetrate the thickness direction through the mounting surface 21A and the back surface 22A. However, they do not necessarily have to be through-holes; they may be non-through holes (depressions) opening into the mounting surface 21A. When the holes are not through-holes, they need only have a volume large enough to transfer the amount of adhesive that should be removed when adjusting the amount of adhesive remaining between the mounting surface and the laminate. However, it is preferable that the holes have at least one through-hole that leads to the outside. This is because having a through-hole allows for efficient guidance of the adhesive. Furthermore, some adhesives, such as epoxy adhesives, require a dedicated remover to be used for removal from the jig. Having a through-hole allows the remover to penetrate effectively, enabling efficient removal.

In this embodiment, when viewed from above, the fixing jig 20A has one hole 24A at its center, with multiple holes 24A arranged on two concentric circles around this center. Four holes 24A are arranged at 90° intervals on the concentric circle closer to the center, and eight holes 24A are arranged at 45° intervals on the concentric circle farthest from the center.

The holes 24A are arranged at positions that overlap the laminate 10 when viewed from above with the laminate 10 placed on the mounting surface 21A. It is not necessary for all the holes 24A to overlap, but it is preferable that all the holes 24A be positioned so as to overlap the laminate 10 when viewed from above.
Furthermore, it is preferable that the holes 24A are arranged in point symmetry with respect to the center point of the fixing jig 20A.

The size and number of holes 24A are not particularly limited, but it is preferable that the opening rate, which is the ratio of the total area of the openings of holes 24A to the area of the mounting surface 21A, be 10% or more. Holes with sufficient opening area can efficiently guide the adhesive. Furthermore, a smaller adhesive area reduces the load associated with peeling. 20% or more is more preferable, and 30% or more is even more preferable. The opening area of the holes is preferably 60% or less of the total area of the laminate. If the area is too small, the laminate will not be sufficiently bonded. 50% or less is more preferable, and 40% or less is even more preferable.

The holes are preferably arranged so that they are as uniform as possible within the contact surface between the laminate and the fixture. Specifically, the holes are preferably arranged so that the opening length is as uniform as possible within the contact surface between the laminate and the fixture. Here, "opening length" refers to the shorter of b or c, where b is the distance from any point where the laminate and the fixture contact each other to the nearest hole opening and c is the distance to the nearest outer periphery of the laminate.

Specific examples will be listed using FIG.
Focusing on an arbitrary first point _P1 , if the distance to the opening of the nearest hole 24A (the opening located immediately to the right from the top on the concentric circle furthest from the center in the illustration) is _b1 and the distance to the nearest outer periphery of the laminate is _c1 , then since _c1 > _b1 , the opening length at the first point _P1 is _b1 .
Next, focusing on an arbitrary second point _P2 located closer to the outer periphery of the laminate than the first point _P1 , if the distance to the opening of the nearest hole 24A (the opening located two holes to the right from the top on the concentric circle farthest from the center in the illustration) is taken as _b2 and the distance to the nearest outer periphery of the laminate is taken as _c2 , then since _b2 > _c2 , the opening length at the second point _P2 is _c2 .
When the opening length is defined as above, the opening length that is the longest among the opening lengths at any point on the mounting surface is defined as the maximum opening length. In the example shown in Fig. 4, the opening length at the illustrated first specific point _Q1 is the maximum opening length _L1 . In the example shown in Fig. 4, the first specific point _Q1 is the center point of the openings of three holes 24A, consisting of the openings of two adjacent holes 24A on a concentric circle far from the center and the opening of one hole 24A on the concentric circle closer to the center and located closest to the two holes 24A.

If the maximum opening length _L1 is large, the distance that the adhesive travels increases, resulting in a thicker adhesive layer for the same amount of work, and greater variations in the thickness of the adhesive layer.
Therefore, the maximum opening length _L1 is preferably 20 mm or less. The opening length _L1 is more preferably 10 mm or less, and even more preferably 5 mm or less, or 2 mm or less. If the maximum opening length _L1 is too short, the load on the laminate increases, which may lead to damage to the laminate. Furthermore, the adhesive area between the laminate and the fixing jig decreases, which may result in insufficient fixing. The maximum opening length _L1 is preferably 0.5 mm or more, and more preferably 1 mm or more, or 2 mm or more.

The openings of the holes are preferably arranged so that the opening lengths are as uniform as possible within the contact surface between the laminate and the fixture. To achieve this arrangement, the openings are preferably arranged so as to be point symmetric or line symmetric.
The number and size of the holes in the fixing jig of this embodiment are arbitrary, but it is preferable to make the opening length as small as possible. Specifically, it is preferable to increase the number of through holes, such as hole 24A in FIG. 4, and uniformly arrange them over the entire surface that contacts the laminate and the fixing jig to make the opening length small. In such a shape, the number of holes is preferably four or more, and more preferably thirteen or more. In this case, the size of the holes is preferably 1 mm or more and 10 mm or less in diameter, and preferably 2 mm or more and 6 mm or less.

In this embodiment, the planar shape of the hole 24A is circular, but it is not limited to this and may be other shapes (triangle, square, other polygonal shapes, and ellipse).

While there are no particular limitations on the material that makes up the fixture 20A, metal is preferred from the standpoint of strength and other factors. Examples of metals include stainless steel, copper, and iron. When using oxidizable materials such as copper or iron, plating is preferred to prevent changes in size and shape due to oxidation. Any material can be used for plating, but nickel plating is preferred from the standpoint of versatility.

1.2.2b. Fixture 20B
5 shows a diagram for explaining the fixing jig 20B. To distinguish it from other types of fixing jigs, it is written as fixing jig 20B, but the fixing jig 20B can be used as the fixing jig 20 described above.
Fig. 5(a) is a plan view of the fixing jig 20B, and Fig. 5(b) is a B-B' cross-sectional view of the fixing jig 20B. As can be seen from Fig. 5, the fixing jig 20B has a mounting surface 21B and a back surface 22B on the opposite side thereof, with a thickness between the mounting surface 21B and the back surface 22B, and a side surface 23B formed so as to bridge the outer periphery of the mounting surface 21B and the outer periphery of the back surface 22B. In this embodiment, the fixing jig 20B is disk-shaped, but it does not necessarily have to be disk-shaped.

Fixing jig 20B has multiple holes 24B formed in mounting surface 21B as recesses opening into the mounting surface 21B. In this embodiment, holes 24B are through holes that penetrate through the thickness direction, passing through mounting surface 21B and back surface 22B, but they do not necessarily have to be through holes; they may be non-through holes (depressions) opening into mounting surface 21B. When the holes are not through holes, the depth of the holes need only be large enough to allow the amount of adhesive to be removed to be transferred into the holes when adjusting the amount of adhesive remaining between the mounting surface and the laminate.

In this embodiment, hole 24B is a hole shaped as if it were surrounded by arcs belonging to two concentric circles (outer arc 24Ba, inner arc 24Bb) and two straight lines (24Bc) connecting the ends of the arcs when viewed from above on fixing jig 20B, and four holes of the same shape are arranged at 90° intervals around the center.

The holes 24B are arranged at positions that overlap the laminate 10 when viewed from above with the laminate 10 placed on the mounting surface 21B. It is not necessary for all the holes 24B to overlap, but it is preferable that all the holes 24B be positioned so as to overlap the laminate 10 when viewed from above.
Furthermore, it is preferable that the holes 24B are arranged in point symmetry with respect to the center point of the fixing jig 20B.

The size and number of holes 24B are not particularly limited, but it is preferable that the opening rate, which is the ratio of the total area of the openings of holes 24B to the area of the mounting surface 21B, be 10% or more. Having holes with sufficient opening area allows for efficient guidance of the adhesive. Furthermore, reducing the adhesive area reduces the load when peeling. 20% or more is more preferable, and 30% or more is even more preferable. The opening area is preferably 60% or less of the total area of the laminate. If the area is too small, the laminate will not be sufficiently bonded. 50% or less is more preferable, and 40% or less is even more preferable.

In the example shown in FIG. 5, the opening length at the specific point _Q2 shown in the figure is the maximum opening length _L2 .
The maximum opening length _L2 is preferably 20 mm or less. The opening length _L2 is more preferably 10 mm or less, and even more preferably 5 mm or less, or 2 mm or less. If the maximum opening length is too short, the load on the laminate increases, which may lead to damage to the laminate. Furthermore, the adhesive area between the laminate and the fixing jig decreases, which may result in insufficient fixing. The maximum opening length _L2 is preferably 0.5 mm or more, and more preferably 1 mm or more, or 2 mm or more.

The number and size of holes in the fixing jig of the present invention are arbitrary, but it is preferable to make the opening length small. Specifically, it is preferable to increase the number of through holes having a cross-sectional shape such as hole 24B in Figure 5, uniformly arrange them over the entire surface that contacts the laminate and the fixing jig, and make the opening length small. In the case of such a shape, it is preferable that the number of holes is four or more. In this case, the size of the holes preferably has an area of 150 mm2 or ^more and 250 mm2 or ^less .

While there are no particular limitations on the material that makes up the fixture 20B, metal is preferred from the standpoint of strength and other factors. Examples of metals include stainless steel, copper, and iron. When using oxidizable materials such as copper or iron, plating is preferred to prevent changes in size and shape due to oxidation. Any material can be used for plating, but nickel plating is preferred from the standpoint of versatility.

1.2.2c. Fixture 20C
6 shows a diagram for explaining the fixing jig 20C. To distinguish it from other types of fixing jigs, it is written as fixing jig 20C, but the fixing jig 20C can be used as the fixing jig 20 described above.
Fig. 6(a) is a plan view of the fixing jig 20C, and Fig. 6(b) is a CC' cross-sectional view of the fixing jig 20C. As can be seen from Fig. 6, the fixing jig 20C has a mounting surface 21C and a back surface 22C on the opposite side thereof, with a thickness between the mounting surface 21C and the back surface 22C, and a side surface 23C formed so as to bridge the outer periphery of the mounting surface 21C and the outer periphery of the back surface 22C. In this embodiment, the fixing jig 20C is disk-shaped, but it does not necessarily have to be disk-shaped.

Fixing jig 20C has a hole 25C in the center of mounting surface 21C and multiple holes 24C around it as recesses opening into mounting surface 21C. In this embodiment, holes 24C and 25C are through holes that penetrate through the thickness direction through mounting surface 21C and back surface 22C, but they do not necessarily have to be through holes and may be non-through holes (recesses) opening into mounting surface 21C. When the holes are not through holes, the depth of the holes only needs to be large enough to allow the amount of adhesive that should be removed to be transferred into the holes when adjusting the amount of adhesive remaining between the mounting surface and the laminate.

In this embodiment, hole 24C can be considered similar to hole 24B in fixing jig 20B described above.

On the other hand, hole 25C is located in the center of mounting surface 21C and has a circular shape in a plan view, but is not limited to this and may have other shapes (triangle, square, other polygonal shapes, and ellipse).

The holes 24C and 25C are arranged in positions that overlap the laminate 10 when viewed from above in a position where the laminate 10 is placed on the mounting surface 21C. It is not necessary for all of the holes 24C and 25C to overlap the laminate 10, but it is preferable that all of the holes 24C and 25C be positioned so as to overlap the laminate 10 when viewed from above.
Furthermore, it is preferable that the holes 24C are arranged in point symmetry with respect to the center point of the fixing jig 20C.

There are no particular limitations on the size or number of holes 24C or the size of holes 25C, but it is preferable that the opening rate, which is the ratio of the total area of holes 24C and 25C to the area of mounting surface 21C, be 10% or more. Having openings of sufficient area allows for efficient guidance of the adhesive. Furthermore, reducing the adhesive area reduces the load during peeling. 20% or more is more preferable, and 30% or more is even more preferable. The opening area is preferably 60% or less of the total area of the laminate. If the area is too small, the laminate will not be sufficiently bonded. 50% or less is more preferable, and 40% or less is even more preferable.

In the example shown in FIG. 6, the opening length at the specific point _Q3 shown in the figure is the maximum opening length _L3 .
In the example shown in FIG. 6, the specific point _Q3 is a point located at the outermost corner (on the outer periphery of the laminate) of a hole 24C provided around the mounting surface 21.
The maximum opening length _L3 is preferably 20 mm or less. The opening length _L3 is more preferably 10 mm or less, and even more preferably 5 mm or less, or 2 mm or less. If the maximum opening length is too short, the load on the laminate increases, which may lead to damage to the laminate. Furthermore, the adhesive area between the laminate and the fixing jig decreases, which may result in insufficient fixing. The maximum opening length _L3 is preferably 0.5 mm or more, and more preferably 1 mm or more, or 2 mm or more.

The number and size of the holes in the fixing jig of this embodiment are arbitrary, but it is preferable to reduce the opening length. Specifically, it is preferable to increase the number of through holes shaped as typified by hole 24C in FIG. 6, uniformly arrange them over the entire surface that contacts the laminate and the fixing jig, and reduce the opening length. In the case of such a shape, it is preferable that the number of holes be four or more. In this case, the size of the holes 24C provided around the mounting surface 21 preferably has an area of 150 mm2 or ^more and 250 ^mm2 or less. Furthermore, it is preferable that the size of the hole 25C provided in the center is smaller than the size of the hole 24C, specifically, an area of 5 mm2 or ^more and 20 ^mm2 or less.

The material that makes up the fixing jig 20C is not particularly limited, but metal is preferable from the standpoint of strength, etc. Examples of metal include stainless steel and copper.

1.2.2d. Fixture 20D
7 shows a diagram for explaining a fixing jig 20D. To distinguish it from other types of fixing jigs, it is written as fixing jig 20D, but the fixing jig 20D can be used as the fixing jig 20 described above.
Fig. 7(a) is a plan view of fixing jig 20D, and Fig. 7(b) is a DD' cross-sectional view of fixing jig 20D. As can be seen from Fig. 7, fixing jig 20D has a mounting surface 21D and a back surface 22D on the opposite side thereof, with a thickness between mounting surface 21D and back surface 22D, and side surface 23D formed to bridge the outer periphery of mounting surface 21D and the outer periphery of back surface 22D. In this embodiment, fixing jig 20D is disk-shaped, but it does not necessarily have to be disk-shaped.

The fixing jig 20D has holes 24D at the center and near the outer periphery of the mounting surface 21D as recesses opening into the mounting surface 21D, and a groove 25D is provided in a spiral shape in a plan view connecting these two holes 24D.
In this embodiment, the hole 24D is a through hole that penetrates through the mounting surface 21D and the back surface 22D in the thickness direction, but it does not necessarily have to be a through hole and may be a non-through hole (recess) that opens on the mounting surface 21D. Also, the hole 24D does not necessarily have to be provided.
In this embodiment, groove 25D has an opening in mounting surface 21D and extends spirally around central hole 24D toward outer peripheral hole 24D. In this embodiment, groove 25D does not reach rear surface 22D in the depth direction, but this is not limiting and grooves that reach the rear surface may also be used.

The hole 24D has a circular shape in plan view, but is not limited to this and may have other shapes (triangle, square, other polygonal shapes, and ellipse).

The holes 24D and grooves 25D are positioned so that they overlap the laminate 10 when viewed in a plan view with the laminate 10 placed on the mounting surface 21D. While it is not necessary for all holes 24D and grooves 25D to overlap the laminate 10, it is preferable that all holes 24D and grooves 25D overlap the laminate 10 when viewed in a plan view.

There are no particular limitations on the size or number of holes 24D, the groove width of groove 25D, or the number of spiral turns, but it is preferable for the opening rate, which is expressed as the ratio of the total area of the openings of holes 24D and grooves 25D to the area of the mounting surface 21D, to be 10% or more. Having openings of sufficient area allows for efficient guidance of the adhesive. Furthermore, reducing the adhesive area reduces the load during peeling. 20% or more is more preferable, and 30% or more is even more preferable. The opening area is preferably 60% or less of the total area of the laminate. If the area is too small, the laminate will not be sufficiently bonded. 50% or less is more preferable, and 40% or less is even more preferable.

When arranging the holes 24D so as to make the opening length uniform, it is preferable to arrange the holes in a circular shape at equal intervals, or to arrange the holes in a shape that keeps the intervals between the holes constant, as typified by Archimedes' spiral.
In the case of the Archimedes' spiral, since there is only one hole, the number of through-holes required is reduced, allowing for greater design freedom compared to a method in which the holes are arranged in a circle and evenly spaced apart, where each hole is independent (see Figure 4). Furthermore, in the case of a method in which the holes are arranged in a circle and evenly spaced apart, the opening length depends on both distance b and distance c, so the design must be tailored to the size of the laminate, which depends on distance c. In the case of the Archimedes' spiral, it is clear that if the diameter of the spiral is expanded to a size equal to or greater than the diameter of the laminate, the distance c to the outer periphery of the laminate will always be smaller than the distance b to the opening. A fixture with such a shape is more preferable because it can be used universally and does not require redesigning the fixture even if the area or shape of the laminate changes.
When an Archimedes' spiral hole is provided, it is preferable to provide through holes at two locations, the start point and the end point.

In the case of the Archimedes' spiral shown in FIG. 7 , the maximum opening length _L4 can be defined as the distance between portions of adjacent grooves 25 in the radial direction of the fixing jig 20D. The maximum opening length _L4 is preferably 20 mm or less. The opening length is more preferably 10 mm or less, and even more preferably 5 mm or less, or 2 mm or less. A shorter maximum opening length increases the load on the laminate, which may lead to damage to the laminate. Furthermore, the adhesive area between the laminate and the fixing jig is reduced, which may result in insufficient fixation. The maximum opening length _L4 is preferably 0.5 mm or more, and more preferably 1 mm or more, or 2 mm or more.

The Archimedes spiral can be expressed by the polar coordinate equation r = aθ. Here, r is the distance from the center point of the fixture 20D to the center of the groove 25D in the width direction, θ is the angle between the line segment connecting the center point of the fixture 20D and the center of the groove 25D in the width direction and the reference line (the horizontal direction in the example shown in Figure 7), and a is a coefficient. In the shape of this embodiment, a is preferably 0.5 to 10 inclusive. It is preferably 1 to 5 inclusive, and more preferably 1.5 to 3 inclusive. In this case, if the width of the spiral is d, d is preferably 0.5 mm to 10 mm inclusive, more preferably 1 mm to 5 mm inclusive, and even more preferably 1.5 mm to 3 mm inclusive. The ratio of a to d is preferably d/a = 0.3 to 3 inclusive, more preferably 0.5 to 2 inclusive, 0.8 to 1.5 inclusive, or 0.95 to 1.1 inclusive. The Archimedes spiral can be enlarged as far as the size of the jig allows, but if the diameter of the spiral is enlarged to a size equal to or larger than the diameter of the laminate, the distance c to the outer periphery of the laminate will always be smaller than the distance b to the opening, and any laminate smaller than this can be used in this form and is more suitable.

While there are no particular limitations on the material that makes up the fixture 20D, metal is preferred from the standpoint of strength and other factors. Examples of metals include stainless steel, copper, and iron. When using oxidizable materials such as copper or iron, plating is preferred to prevent changes in size and shape due to oxidation. Any material can be used for plating, but nickel plating is preferred from the standpoint of versatility.

1.2.3. Laminate Arrangement In this process, the laminate 10 is fixed to the mounting surface of the fixing jig 20 via the adhesive layer 11 as described above. At this time, unnecessary adhesive is moved to the recesses (holes or grooves of various shapes) provided in the fixing jig 20, as described above. To achieve this, a pressing force can be applied between the laminate 10 and the fixing jig 20. This allows for smooth movement of the adhesive. The specific pressing force is not particularly limited, but is adjusted so that an adhesive layer 11 of an appropriate thickness is formed between the laminate 10 and the mounting surface of the fixing jig 20. If the adhesive layer 11 is too thick, the aluminum nitride single crystal body will deviate significantly from the target separation (cutting) position. On the other hand, if the adhesive layer 11 is too thin, the fixation will be weak and unstable, potentially preventing proper separation (cutting). The specific thickness of the adhesive layer is not particularly limited, but is preferably 0.1 μm to 50 μm. It is more preferably 0.5 μm to 30 μm, or 1 μm to 10 μm.

1.3 Separation Step In the separation step S13, the aluminum nitride single crystal is separated from the laminate 10 prepared in the placement step S12 and placed on the fixing jig 20 via the adhesive layer 11.
The separation method is not particularly limited, but may be, for example, cutting with a wire saw. That is, as shown in Fig. 8, the aluminum nitride single crystal is separated by moving a wire saw 12 from the side surface 10c of the laminate 10 along the first surface 10a and the second surface 10b as indicated by the straight arrows in Fig. 8.

In addition to the above, the separation step S13 can be configured as follows: Fig. 9 shows explanatory diagrams, in which Fig. 9(a) is a plan view and Fig. 9(b) is a front view.
9, a buffer layer 13 may be formed on the side surface 10c of the laminate 10 before cutting with the wire saw 12, and then the buffer layer 13 may be brought into contact with the wire saw 12 before cutting the aluminum nitride single crystal of the laminate 10. This can further prevent the wire saw 12 from shifting out of position.

The materials that make up the buffer layer 13 can be any adhesive, such as epoxy or cyanoacrylate, or thermoplastic resin, such as polyester or polyethylene. A different resin may be used than the one used for the adhesive layer, but from the standpoint of workability, it is preferable to use the same material. Glass, aluminum oxide, or other particles may be mixed into the resin to increase hardness. Nikka Seiko's Q Bond and U Bond are useful materials for the buffer layer.

9, the buffer layer 13 is disposed around the entire periphery of the side surface 10c of the laminate 10, but this is not necessarily limited to this, and it is sufficient if it can suppress misalignment of the wire saw 12. For this reason, it is preferable that the buffer layer 13 be disposed at least in the portion of the side surface 10c that the wire saw 12 first touches.
Furthermore, although not shown in the drawings, aluminum nitride single crystals may have convex polycrystals called crowns on their periphery depending on the crystal growth conditions. For such aluminum nitride single crystals with crowns on their periphery, it is more preferable to remove the crowns by cutting, polishing, or grinding the periphery, and then apply the method according to the above-described embodiment. In other words, the effects of the present invention are more pronounced for aluminum nitride single crystals that do not have crowns on their periphery.

2. Effects, etc. According to the present disclosure, in the process of placing the laminate on the mounting surface of the fixing jig, a portion of the adhesive present between the mounting surface and the laminate can be guided from between the mounting surface and the laminate in the thickness direction (vertical direction) of the fixing jig to a recess and to a position lower than the mounting surface. This adjusts the amount of adhesive remaining between the mounting surface and the laminate, reduces the thickness of the adhesive layer, and suppresses displacement of the aluminum nitride single crystal body from the target position when it is separated from the laminate in the separation process.

3. Examples As examples, fixing jigs were fabricated in imitation of the above-described fixing jigs 20A, 20B, 20C, and 20D, and as a comparative example, a fixing jig without a recess on the mounting surface was fabricated and tested.

Each fixture was a rectangular parallelepiped measuring 70 mm wide, 100 mm long, and 20 mm high.

The adhesive used to form the adhesive layer was the aforementioned Q Bond or U Bond manufactured by Nikka Seiko.

3.1. Example A
In Example A, a laminate 10 was prepared by laminating an aluminum nitride single crystal by the HVPE method on an AlN substrate fabricated by the PVT method. The laminate had a diameter of 2 inches and a thickness of 1400 μm. Hereinafter, the laminate 10 will also be referred to as "aluminum nitride single crystal."
The off-angle of the aluminum nitride single crystal was measured. The heights of fixtures 20A, 20B, 20C, and the fixture of the comparative example were measured using a micrometer and used as the origin. The aluminum nitride single crystal was placed on the fixture, and the height of the nitride single crystal was measured at the locations indicated by the black circles in Figure 10 (a total of 17 locations, including the center point, eight locations spaced 5 mm apart on the left and right of the center point on a line segment passing through the center of the aluminum nitride single crystal, and eight locations spaced 5 mm apart above and below the center point on a line segment perpendicular to the line segment). Similar measurements were performed at different measurement points to measure the in-plane height distribution of the aluminum nitride single crystal. After cleaning the fixture and the surface of the laminate with acetone, an adhesive was applied to the mounting surface of the fixture, and the second surface of the fabricated laminate was placed and pressed. Pressing was performed at a pressure of 0.05 MPa for 1 minute. After leaving it for 24 hours to solidify the adhesive layer, the thickness of the laminate and adhesive layer was measured using a micrometer. Similar measurements were performed at different measurement points, and the average thickness and standard deviation of the adhesive layer were calculated by calculating the difference from the height of the laminate before bonding. The off-angle of the aluminum nitride single crystal after cutting was measured, and the difference from before cutting was calculated to calculate the difference in off-angle before and after cutting. The results are shown in Table 1.

As can be seen from Table 1, compared to the fixing jig of the comparative example, fixing jig 20A, fixing jig 20B, and fixing jig 20C were able to reduce the adhesive layer thickness. This makes it possible to suppress misalignment with the target position when separating the aluminum nitride single crystal body from the laminate in the separation process.

3.2. Example B
In Example B, a laminate was prepared by laminating an aluminum nitride single crystal on a sapphire substrate by the HVPE method. The laminate had a diameter of 2 inches and a thickness of 1400 μm.
Adhesive was applied to the mounting surface of each of fixtures 20A, 20B, 20C, 20D, and the comparative fixture, and the second surface of the laminate was placed on it and pressed. In this example, the pressing force was varied and pressed for one minute at that pressure. The thickness of the adhesive layer between the second surface of the laminate and the mounting surface of the fixture was then measured. The results are shown in Table 2.

As can be seen from Table 2, even when the pressing force was changed, fixing jigs 20A, 20B, 20C, and 20D were able to reduce the adhesive layer thickness compared to the fixing jig of the comparative example. Furthermore, the standard deviation was smaller, indicating that the in-plane variation in adhesive layer thickness was reduced. It was also confirmed that the difference in off-angle before and after cutting was smaller.

REFERENCE SIGNS LIST 10 Laminate 11 Adhesive layer 12 Wire saw 13 Buffer layer 20 Fixing jig 20A Fixing jig 20B Fixing jig 20C Fixing jig 20D Fixing jig

Claims (1)

a preparation step of preparing a laminate including an aluminum nitride single crystal body;
a placing step of placing the laminate on a placing surface of a fixing jig via an adhesive layer formed by an adhesive;
A method for manufacturing an aluminum nitride single crystal substrate, comprising: a separation step of separating the aluminum nitride single crystal body from the laminate bonded onto the adhesive layer,
the placing step includes guiding a portion of the adhesive present between the placement surface and the laminate to a position lower than the placement surface in a thickness direction of the fixing jig,
In the separating step, the aluminum nitride single crystal body is cut into a plate shape from the laminate using a wire saw;
In the separation step, a buffer layer is formed on the side surface of the aluminum nitride single crystal body, and then the buffer layer is brought into contact with a wire saw before cutting the aluminum nitride single crystal body.
A method for manufacturing an aluminum nitride single crystal substrate.