Imprint mold and method for manufacturing a semiconductor device using the same, semiconductor device

Abstract

The present invention provides an imprint mold used to uniformly and stably form bump electrodes on electrode terminals in large-area semiconductor devices. [Solution] The imprint mold 50 comprises a first surface 51A, a second surface 51B facing the first surface 51A and having an uneven surface, and a plurality of pillars 52 protruding from the second surface 51B. The second surface 51B surrounding the pillars 52 has a concave shape that is recessed toward the first surface 51A. The imprint mold 50 is used to provide openings 42 corresponding to the positions of electrode terminals 11 in the resist 40 provided on the surface of the semiconductor element 10.

Description

Technical Field

[0001] The present disclosure relates to an imprint mold, a method for manufacturing a semiconductor device using the same, and a semiconductor device.

Background Art

[0002] In recent years, with the increasing demand for high-speed and high-capacity communication, high integration in semiconductor packages and multi-pinning and narrow pitching of electrode terminals in semiconductor elements such as system LSIs, GPUs, CPUs, and memories have been progressing. As a mounting technique for semiconductor elements on a mounting substrate, flip chip mounting is known.

[0003] In flip chip mounting, the electrode terminals and the connection terminals are connected and the semiconductor element is mounted on the mounting substrate by bringing the protruding electrodes (bump electrodes) formed on the electrode terminals of the semiconductor element into contact with the connection terminals of the mounting substrate and applying pressure and heat. As the bump electrodes, solder bumps are widely adopted.

[0004] However, with the narrowing of the pitch between electrode terminals, in the pressure application and heating process during flip chip mounting, bridge defects are likely to occur where the melted and deformed solder bumps are connected to adjacent solder bumps by surface tension.

[0005] Therefore, a method of forming fine metal bumps with a sharp shape made of gold, copper, or the like instead of solder bumps is known. In this method, the tip of the bump electrode is plastically deformed in the above-described pressure application and heating process, and the protruding electrode is joined to the connection terminal by solid-phase diffusion. Since the fine metal bumps are not melted during flip chip mounting, the occurrence of bridge defects caused by melting and deformation can be prevented, and it becomes easy to cope with the narrowing of the pitch between electrode terminals.

[0006] A known method for forming fine metal bumps involves coating a semiconductor element with a photoresist, creating openings in the photoresist, and then depositing metal into these openings by plating or other means. Fine metal bumps are obtained by removing the resist after the metal deposition. A method for creating sharp edges on these fine metal bumps is disclosed, for example, in Patent Document 1.

[0007] In the method disclosed in Patent Document 1, a photoresist is applied to a semiconductor substrate, and the photoresist, which has undergone exposure treatment to form an opening, is baked. Specifically, a temperature gradient is provided such that the temperature is higher on the surface side of the photoresist and lower on the semiconductor substrate side. In this way, the photoresist develops a gradient in its resistance to the developer along its thickness. After the baking treatment, when a development treatment is performed using a developer, an opening is formed in the photoresist in which the opening area on the semiconductor substrate side is larger than the opening area on the surface side. By depositing metal inside the opening by plating and removing the photoresist, a bump electrode with a sharp shape is obtained. [Prior art documents] [Patent Documents]

[0008] [Patent Document 1] Japanese Patent Publication No. 2007-73919 [Overview of the Initiative] [Problems that the invention aims to solve]

[0009] However, in the conventional method disclosed in Patent Document 1, when the area of ​​the semiconductor element is large, it becomes difficult to suppress temperature variations within the plane of the semiconductor element during baking. In this case, the developer resistance of the photoresist also varies within the plane according to the variation in baking temperature. As a result, the shape of the opening, and consequently the tip of the bump electrode, varies within the plane of the semiconductor element. When this happens, the connection state between the bump electrode and the connection terminal may not be stable during flip-chip mounting, and connection reliability may be compromised.

[0010] This disclosure has been made in view of the foregoing, and its purpose is to provide an imprint mold used for uniformly and stably forming bump electrodes provided on electrode terminals in a large-area semiconductor device, a method for manufacturing a semiconductor device using the same, and a semiconductor device. [Means for solving the problem]

[0011] To achieve the above objective, an imprint mold according to one aspect of the present disclosure comprises a first surface, a second surface facing the first surface and having an uneven surface, and a plurality of pillars protruding from the second surface, wherein the second surface around the pillars has a concave shape that is recessed toward the first surface.

[0012] A method for manufacturing a semiconductor device according to one aspect of the present disclosure comprises at least a resist forming step of covering the surfaces of a plurality of electrode terminals provided on a semiconductor device with a resist; an opening forming step of pressing the imprint mold against the resist and applying pressure to form an opening in the resist; a resist curing step of heating the semiconductor device covered with the resist after the opening has been formed and curing the resist by irradiating it with light; and a developing step of using a developer to enlarge the diameter of the opening, wherein in the opening forming step, the imprint mold is pressed against the resist and applied pressure such that the second surface abuts the surface of the resist and, when viewed from above, the tips of each of the plurality of pillars overlap the plurality of electrode terminals, and in the resist curing step, depending on the shape of the uneven surface transferred to the surface of the resist, the light penetrates into the resist, causing the solubility of the resist in the developer around the opening to change along the depth direction of the opening.

[0013] A semiconductor device according to one aspect of the present disclosure comprises at least a semiconductor element having a plurality of electrode terminals on its surface, and a resist covering the surface and having a plurality of openings, wherein, as viewed from the surface of the resist, the openings are provided so as to overlap with the electrode terminals and reach the electrode terminals, the surface of the resist is an uneven surface having protrusions provided so as to surround the openings, and in cross-sectional view, the openings have a second portion that is wider along the surface and a first portion that is narrower than the second portion, along the thickness direction of the resist. [Effects of the Invention]

[0014] According to this disclosure, in large-area semiconductor devices where the number of pins and pitch of electrode terminals are increasing, bump electrodes provided on the electrode terminals, and openings provided in the resist for forming bump electrodes, can be formed uniformly and stably. [Brief explanation of the drawing]

[0015] [Figure 1A] This is a schematic diagram illustrating the structure of a first imprint mold according to an embodiment. [Figure 1B] This is a schematic diagram illustrating the structure of the second imprint mold. [Figure 1C] This is a schematic diagram illustrating the structure of the third imprint mold. [Figure 2A] This is a schematic cross-sectional diagram illustrating the manufacturing method of the first imprint mold. [Figure 2B] This is a schematic cross-sectional diagram illustrating an alternative manufacturing method for the first imprint mold. [Figure 3A] This is a schematic cross-sectional diagram illustrating a method for manufacturing a semiconductor device using a first imprint mold. [Figure 3B] This is a schematic cross-sectional diagram illustrating the subsequent steps following the process shown in Figure 3A. [Figure 3C] This is a schematic cross-sectional diagram illustrating the subsequent steps following the process shown in Figure 3B. [Figure 4] It is a schematic cross-sectional view for explaining a resist curing process when using the first imprint mold. [Figure 5A] It is a schematic cross-sectional view for explaining a resist curing process when using the second imprint mold. [Figure 5B] It is a schematic cross-sectional view for explaining a resist curing process when using the third imprint mold. [Figure 6A] It is a schematic diagram for explaining the structure of an opening when using the second imprint mold. [Figure 6B] It is a schematic diagram for explaining the structure of an opening when using the third imprint mold. [Figure 7A] It is a schematic cross-sectional view of a bump electrode according to Modification 1. [Figure 7B] It is a schematic cross-sectional view of another bump electrode according to Modification 1. [Figure 8A] It is a schematic plan view of a resist provided with an opening according to Modification 2. [Figure 8B] It is a cross-sectional view taken along the VIIB-VIIB line in FIG. 8A. [Figure 9] It is a schematic cross-sectional view of a bump electrode according to Modification 2. [Embodiments for Carrying Out the Invention]

[0016] Hereinafter, embodiments of the present disclosure will be described based on the drawings. Note that the following description of the preferred embodiments is merely illustrative in nature and is not intended to limit the present disclosure, its applications, or its uses.

[0017] (Embodiment) [Structure of Imprint Mold] FIG. 1A is a schematic diagram for explaining the structure of the first imprint mold according to the embodiment. FIG. 1B is a schematic diagram for explaining the structure of the second imprint mold. FIG. 1C is a schematic diagram for explaining the structure of the third imprint mold.

[0018] In Figures 1A to 1C, the upper figure is a view of the imprint mold 50 from below, and the lower figure is a cross-sectional view along line AA in the upper figure. The upper figure in Figures 1A to 1C is sometimes called a plan view, and the lower figure in Figures 1A to 1C is sometimes called a cross-sectional view.

[0019] Furthermore, in Figures 1A to 1C, the number and arrangement of pillars 52 are simplified and differ from the actual number and arrangement.

[0020] Furthermore, in the following explanation, the side of the imprint mold 50 on which the first surface 51A is located may be referred to as the upper side or upper part, and the side on which the second surface 51B is located may be referred to as the lower side or lower part. Also, with respect to the semiconductor element 10 on which the resist 40 is formed (Figures 3A-3C), the side on which the resist 40 is located may be referred to as the upper side or upper part, and the side on which the semiconductor element 10 is located may be referred to as the lower side or lower part.

[0021] The first to third imprint molds 50A, 50B, and 50C shown in Figures 1A to 1C each have a plate-shaped main body 51 and a plurality of pillars 52. The main body 51 has a first surface 51A and a second surface 51B opposite the first surface. In this embodiment, the first surface 51A is the upper surface of the main body 51. On the other hand, the second surface 51B is the lower surface and is an uneven surface having a recess 51B1 and a protrusion 51B2. In the following description, the first to third imprint molds 50A, 50B, and 50C may be collectively referred to as the imprint mold 50. In this case as well, they will be referred to as the first surface 51A, the second surface 51B, the recess 51B1, and the protrusion 51B2.

[0022] The pillar 52 is a columnar member that protrudes downward from the second surface 51B and is formed integrally with the main body 51. In this embodiment, the pillar 52 has a circular cross-sectional shape perpendicular to its longitudinal direction, but is not limited to this. For example, it may be a regular polygon.

[0023] The number and arrangement of the pillars 52 on the second surface 51B correspond to the number and arrangement of the electrode terminals 11 provided on the semiconductor element 10, which will be described later.

[0024] In the first imprint mold 50A shown in Figure 1A, the recess 51B1 is a semicircle centered on the pillar 52 in cross-sectional view. In other words, the recess 51B1 is hemispherical. The convex portion 51B2 is located at the boundary between adjacent recesses 51B1.

[0025] In the second imprint mold 50B shown in Figure 1B, the protrusions 51B2 are semicircular in shape and are positioned in contact with the base end of the pillar 52 in cross-sectional view. In other words, the protrusions 51B2 are hemispherical, and two of them are provided between adjacent pillars 52. The recesses 51B1 are located at the boundary between adjacent protrusions 51B2. Also, the recesses 51B1 are located at the boundary between a protrusion 51B2 and the adjacent pillar 52.

[0026] In the third imprint mold 50C shown in Figure 1C, the recess 51B1 is a semicircle provided so as to be in contact with the base end of the pillar 52 in cross-sectional view. In other words, the recess 51B1 is hemispherical, and two of them are provided between adjacent pillars 52. The protrusion 51B2 is located at the boundary between adjacent recesses 51B1.

[0027] In other words, in each of the imprint molds 50 shown in Figures 1A to 1C, the recess 51B1 is provided for each of the multiple pillars 52 so as to surround the base end of the pillar 52.

[0028] Furthermore, recognition marks 53 are provided in all of the imprint molds 50 shown in Figures 1A to 1C. The size, shape, and number of recognition marks 53, as well as their position relative to the main body 51, are not particularly limited to the examples shown in Figures 1A to 1C. As will be described in detail later, the recognition marks 53 are used for aligning the imprint mold 50 with the semiconductor element 10.

[0029] [Method for manufacturing imprint molds] The material of the imprint mold 50 can be selected from a variety of options. For example, the material of the imprint mold 50 may be a light-transmitting resin such as acrylic resin or silicone resin, or an inorganic insulating material such as glass or silicon (Si). Alternatively, the material of the imprint mold 50 may be a metallic material such as nickel.

[0030] As will be explained later, the imprint mold 50 needs to transmit light. Therefore, from the viewpoint of light transmittance, processability, and thermal conductivity, the material of the imprint mold 50 is more preferably acrylic resin or silicone resin. In this case, the imprint mold 50 can be formed by electron beam processing or laser processing using two-photon absorption. The following will be explained with reference to the drawings.

[0031] Figure 2A is a schematic cross-sectional diagram illustrating a method for manufacturing the first imprint mold. Figure 2B is a schematic cross-sectional diagram illustrating an alternative method for manufacturing the first imprint mold.

[0032] In the method shown in Figure 2A, a light-transmitting resin 30 is applied to a base (not shown), flattened, and then irradiated with an electron beam focused by a first lens 60 for electron beams (hereinafter referred to as the focused beam). The portion irradiated with the focused beam hardens and becomes a hardened portion 31, which has high resistance to organic solvents. In other words, it becomes less soluble in organic solvents. On the other hand, the unhardened portion 32, which is through which the electron beam passes but not irradiated with the focused beam, has low resistance to organic solvents and is easily dissolved. For example, a developer solution can be used as this organic solvent.

[0033] An electron beam is irradiated onto the light-transmitting resin 30 along a predetermined trajectory so that the cured portion 31 takes the shape of the first imprint mold 50A. In this case, by moving the focal point of the focused beam relative to the thickness direction of the light-transmitting resin 30, the uneven shape of the second surface 51B and the shape of the pillar 52 can be formed.

[0034] After the formation of the cured portion 31 is complete, the light-transmitting resin 30 is impregnated with a developer solution, causing the uncured portion 32 to dissolve in the developer solution, leaving the cured portion 31. The developer solution remaining on the surface of the cured portion 31 is removed and dried, completing the first imprint mold 50A.

[0035] Furthermore, as shown in Figure 2B, a laser beam may be used instead of an electron beam. In this case, the laser beam is focused by the second lens 70, which is an optical lens, and two-photon absorption occurs only near the focal point. Typically, a femtosecond laser is used as the laser light source.

[0036] According to the method shown in Figure 2A, since the wavelength of electrons is on the order of nanometers, the first imprint mold 50A can be formed with a processing accuracy on the order of nanometers. Furthermore, as shown in Figure 2B, when utilizing two-photon absorption of a laser beam, the wavelength of the laser beam can be made to be below the wavelength of visible light, so the first imprint mold 50A can be formed with a processing accuracy on the order of submicrons. For example, when the diameter of the pillar 52 is about 1 μm to several μm, the pillar 52 of the desired diameter can be formed with high precision according to the methods shown in Figures 2A and 2B.

[0037] Although Figures 2A and 2B illustrate the first imprint mold 50A as an example, other shapes of imprint molds 50, such as the second imprint mold 50B and the third imprint mold 50C, can be formed in the same manner.

[0038] Furthermore, the method for manufacturing the imprint mold 50 made of light-transmitting resin material is not particularly limited to the method shown in Figures 2A and 2B. For example, the imprint mold 50 may be formed by preparing a master plate (not shown) on which the shape of the imprint mold 50 has been transferred, pouring the light-transmitting resin material into the master plate, curing the material, and then releasing it from the master plate. In this case, the parts of the master plate corresponding to the pillars 52 become through holes. Also, a shape with the irregularities of the second surface 51B reversed is provided on the upper surface of the master plate. Note that applying a release agent to the surface of the master plate including the through holes makes it easier to release the light-transmitting resin from the master plate.

[0039] [Manufacturing method for semiconductor device (1)] Figure 3A is a schematic cross-sectional diagram illustrating a method for manufacturing a semiconductor device using the first imprint mold. Figure 3B is a schematic cross-sectional diagram illustrating a subsequent step following the process shown in Figure 3A. Figure 3C is a schematic cross-sectional diagram illustrating a subsequent step following the process shown in Figure 3B. Figure 4A is a schematic cross-sectional diagram illustrating the resist curing process when using the first imprint mold.

[0040] <Method for manufacturing a semiconductor device provided with a resist having an opening> First, prepare the semiconductor element 10 as shown in Figure 3A(a). The semiconductor element 10 is provided with multiple electrode terminals 11 for electrical connection to an external source, such as a mounting substrate or an interposer connection terminal (neither of which are shown). The semiconductor element 10 is also provided with a recognition mark 12. This function will be described later.

[0041] Next, a seed layer 20 is formed to cover the entire surface of the semiconductor element 10, including the electrode terminals 11 (seed layer formation step). The seed layer 20 is a conductive thin film and functions as a base layer for depositing metal in the plating step described later. The seed layer 20 is preferably a metal, such as one of Cu, Ni, Zn, Au, Ag, or Cr, or an alloy containing multiple types of these. Note that if electroless plating is performed in the plating step, the formation of the seed layer 20 may be omitted.

[0042] After forming the seed layer 20, a resist 40 made of a resin material is formed on the upper surface of the seed layer 20 (resist formation step). The resist 40 hardens by at least light irradiation. However, its type is not limited to, for example, a normal photocurable type. In this embodiment, the resist 40 is a photothermal combined type that hardens by both light irradiation and heating. The resist 40 also contains a photoacid generator. The photoacid generator is a photosensitive agent that decomposes upon light irradiation and generates acid.

[0043] The resist 40 is formed on the surface of the semiconductor device 10 by methods such as spin coating, dip coating, or screen printing, so as to ensure uniform film thickness. The film thickness of the resist 40 depends on the height of the bump electrode 81, which will be described later, but is, for example, about 1 μm to 20 μm.

[0044] Next, as shown in Figure (b), the first imprint mold 50A is prepared. With the tip of the pillar 52 facing the surface 41 of the resist 40, the first imprint mold 50A and the semiconductor element 10 are aligned by an optical method using the recognition marks 53 and 12. For example, when viewed from above, the alignment is performed so that the tip of the pillar 52 coincides with the center of the electrode terminal 11. If, when viewed from above, multiple pillars 52 are positioned to overlap a single electrode terminal 11, the alignment is performed so that the tip of each of the multiple pillars 52 is at a predetermined position on the corresponding electrode terminal 11. Alternatively, if the first imprint mold 50A is made of a metal material, the area between the first imprint mold 50A and the semiconductor element 10 may be imaged from the side using a camera (not shown) to align the recognition marks 53 and 12.

[0045] Next, as shown in Figure (c), the tip of the pillar 52 is brought into contact with the surface 41 of the resist 40, and the first imprint mold 50A is pressed toward the semiconductor element 10 and pressurized (opening formation step). At the same time, the semiconductor element 10 including the resist 40 is heated. The heating temperature is, for example, in the range from room temperature to 150°C. When the pillar 52 is inserted into the resist 40, the resist 40 is pushed out and bulges upward around the pillar 52.

[0046] When inserting the pillar 52 into the resist 40, the tip of the pillar 52 may reach the surface of the electrode terminal 11, or the pressure may be stopped while a gap remains between the tip of the pillar 52 and the surface of the electrode terminal 11. If the pressure is stopped while a gap remains, the contact area between the first imprint mold 50A and the resist 40 is reduced compared to when the tip of the pillar 52 reaches the surface of the electrode terminal 11. This makes it possible to suppress the peeling of the seed layer 20 and the resist 40 during demolding, which will be described later.

[0047] (d) As shown in the figure, when the first imprint mold 50A is further pressed toward the semiconductor element 10, the resist 40 that has risen around the pillar 52 deforms its surface 41 so that it comes into close contact with the second surface 51B. As a result, the resist 40 pushed out by the pillar 52 is contained in the recess 51B1 provided on the second surface 51B (opening formation process).

[0048] If this state is maintained for a predetermined time, the resist 40 will fill the recess 51B1 at the base end of the pillar 52 without any gaps, as shown in Figure (e).

[0049] Next, as shown in Figure 3B(f), the first imprint mold 50A is withdrawn away from the semiconductor element 10 and removed from the resist 40. As a result, the area where the pillar 52 was inserted becomes an opening 42. Also, the surface 41 of the resist 40 takes on a shape that transfers the irregularities of the second surface 51B. In other words, on the surface 41 of the resist 40, the part corresponding to the recess 51B1 of the second surface 51B becomes a convex part 41A, and the part corresponding to the convex part 51B2 becomes a recess 41B.

[0050] Next, as shown in Figure (g), light is irradiated from the surface 41 of the resist 40, and the semiconductor element 10 containing the resist 40 is heated to cure the resist 40 (resist curing step). The irradiated light is, for example, UV light. The heating temperature at this time is about the same as the temperature range in the aperture formation step.

[0051] As mentioned above, since the resist 40 contains a photoacid generator, when irradiated with light, the photoacid generator decomposes and generates acid. In addition, since the resist 40 is heated while being irradiated with light, a crosslinking reaction catalyzed by the acid proceeds, and the crosslinking density increases in the area close to the surface 41.

[0052] In the area of ​​the resist 40 closest to the surface 41, the amount of incident light is high, and the amount of acid generated is also high. On the other hand, light propagating inside the resist 40 is absorbed by the resist 40 and attenuates as the propagation distance increases, so the amount of acid generated decreases. In other words, not only around the opening 42, but also light incident from the surface 41 of the resist 40 attenuates as it propagates downwards, and consequently the amount of acid generated decreases, resulting in a lower crosslinking density.

[0053] However, around the aperture 42, light that has passed through the aperture 42 reaches the surface of the semiconductor element 10 without being significantly attenuated by the resist 40. Therefore, around the aperture 42, differences in light intensity along the thickness direction of the resist 40 are less likely to occur, and consequently, differences in crosslinking density are also less likely to occur.

[0054] On the other hand, as shown in Figure 3B(g), when light is irradiated onto the resist 40, a convex portion 41A is formed on the surface 41 of the resist 40 around the opening 42. The convex portion 41A acts like a Fresnel lens for the incident light, refracting the light irradiated from directly above toward the focal point P shown in Figure 4. In this way, around the opening 42, the light irradiated from directly above is refracted or diffused by the convex portion 41A, so that the amount of light that should be incident on the lower part of the resist 40 is reduced. Consequently, as shown in Figure 4, in the upper part of the surface 41 around the opening 42, where the amount of incident light is large, the hardening of the resist 40 due to the crosslinking reaction progresses, and the shape of the area including the opening 42 stabilizes. On the other hand, in the lower part where the amount of incident light is small, the crosslinking density is lower compared to the upper part, and the hardening of the resist 40 does not progress easily.

[0055] After the resist curing process, a developer solution is supplied to the surface 41 of the resist 40, as shown in Figure (h). The method of supplying the developer solution can be appropriately selected from known methods. For example, a method of impregnating the semiconductor element 10 on which the resist 40 is formed into a tank filled with developer solution (dip) may be used. Alternatively, a method of applying developer solution to the surface 41 of the resist 40 while rotating the semiconductor element 10 on which the resist 40 is formed, and maintaining a stationary state of the developer solution using surface tension (paddle) may be used. Alternatively, a method of spraying developer solution while rotating the semiconductor element 10 on which the resist 40 is formed may be used.

[0056] Here, the developer has the effect of dissolving the resist 40. The lower the crosslinking density of the resist 40, the higher the dissolution rate by the developer. The developer may be, for example, tetramethylammonium hydroxide (TMAH) or tetraethylammonium hydroxide (TEAH).

[0057] The supplied developer flows into the opening 42, and the inner wall of the opening 42 dissolves, causing the diameter of the opening 42 to expand. Before development, the opening 42 is perpendicular to the surface 41 of the resist 40 and is formed to have the same shape across the entire surface of the semiconductor element 10.

[0058] When the developer is supplied into the opening 42, the portion of the opening 42 closest to the surface 41 (hereinafter sometimes referred to as the first portion 42A) has a high crosslink density and is sufficiently hardened, so although the diameter expands slightly, the shape before development is largely maintained. On the other hand, as the developer reaches a certain depth from the surface 41 in the opening 42, the diameter of the opening 42 expands as it approaches the semiconductor element 10 due to the decrease in crosslink density. The expanded portion is sometimes referred to as the second portion 42B.

[0059] After dissolving the resist 40 with the developer, the developer remaining inside the opening 42 is removed with a washing solution, as shown in Figure (i). The washing solution may be, for example, pure water, alcohol, ethanol, or acetone.

[0060] By following the above steps, an opening 42 can be formed in the resist 40, having a second portion 42B with a larger diameter and a first portion 42A that is continuous with the second portion 42B and located above the second portion 42B. The diameter of the first portion 42A is smaller than the diameter of the second portion 42B. As a result, a semiconductor device 90 can be obtained, consisting of a resist 40 having openings 42 that reach each of the multiple electrode terminals 11 provided on the semiconductor element 10, and the semiconductor element 10.

[0061] <Method for manufacturing a semiconductor device (2)> By further processing the semiconductor device 90 as shown in Figure 3C, a sharp-shaped bump electrode 81 (see Figure (k) in Figure 3C) can be obtained. This will be explained in detail below.

[0062] The semiconductor device 90 in the state shown in Figure 3B(i) is impregnated with a plating bath (not shown), and electroplating is performed to deposit a metal plating layer 80 inside the opening 42 (Figure 3C(j)). The electroplating bath may be, for example, a bottom-up filling capable plating bath containing a metal salt of Cu or Au. A bottom-up filling capable plating bath is suitable for filling a fine opening 42 with metal because it can increase the deposition rate from the seed layer 20 located at the bottom of the opening 42. As mentioned above, the metal plating layer 80 inside the opening 42 may also be deposited by electroless plating.

[0063] After the plating process, the resist 40 is removed with a developer or organic solvent, etc., to obtain a semiconductor device 91 in which sharp bump electrodes 81 are formed on the surface of the electrode terminals 11, as shown in Figure (k). Since the shape of the bump electrodes 81 reflects the shape of the opening 42, the upper part of the bump electrodes 81 (sometimes called the third part 81A) has a small diameter, similar to the first part 42A of the opening 42, and is perpendicular to the surface of the semiconductor device 10. On the other hand, the lower part of the bump electrodes 81 (sometimes called the fourth part 81B) has a shape in which the diameter increases as it goes downwards in cross-sectional view, similar to the second part 42B of the opening 42.

[0064] Although not shown in the diagram, after removing the resist 40, the seed layer 20 remaining between the electrode terminals 11 is removed by dry etching or wet etching. In this case, a mask to protect the electrode terminals 11 and bump electrodes 81 is formed before etching. After the seed layer 20 is removed, the mask is removed.

[0065] This prevents short circuits between the electrode terminals 11. Note that the step of removing the seed layer 20 between the electrode terminals 11 may be performed before forming the resist 40.

[0066] [Effects, etc.] As described above, the imprint mold 50 according to this embodiment comprises a first surface 51A, a second surface 51B facing the first surface 51A and having an uneven surface, and a plurality of pillars 52 protruding from the second surface 51B. The second surface 51B surrounding the pillars 52 has a concave shape that is recessed toward the first surface 51A.

[0067] According to this embodiment, when an imprint mold 50 is pressed against the resist 40 formed on the surface of the semiconductor element 10 and pressure is applied toward the semiconductor element 10, the resist 40 extruded by the pillar 52 is contained in the recess 51B1 around the pillar 52. In this way, the resist 40 around the pillar 52 has the shape of the recess 51B1 transferred to it, resulting in a convex shape. Subsequently, when the resist 40 is cured by light irradiation, the convex surface 41 of the resist 40 around the pillar 52 functions as a lens, and the amount of incident light can be increased in the part of the resist 42 that is close to the surface 41. As a result, the curing of the resist 40 progresses, and the shape of the surrounding area, including the opening 42, stabilizes. On the other hand, around the opening 42, the amount of incident light can be reduced as you move downwards, so that the resist 40 does not fully cure. As a result, in the developing step following the resist curing step, the lower part around the opening 42 is dissolved by the developing solution, and a sharp-shaped opening 42 can be obtained.

[0068] The imprint mold 50 is preferably made of a light-transmitting material. In this way, the recognition marks 53 provided on the imprint mold 50 and the recognition marks 12 provided on the semiconductor element 10 can be used to accurately align the imprint mold 50 and the semiconductor element 10 by optical means. In other words, the alignment accuracy between the multiple electrode terminals 11 provided on the semiconductor element 10 and the multiple pillars 52 provided on the imprint mold 50 can be improved. As a result, the opening 42, and thus the bump electrode 81, can be provided in a state that is accurately aligned with the multiple electrode terminals 11.

[0069] Furthermore, it is preferable that the imprint mold 50 be made of a light-transmitting resin material. In this way, for example, the light-transmitting resin 30 can be processed using a method as shown in Figures 2A and 2B to obtain the imprint mold 50. In addition, even if the diameter of the pillar 52 is on the order of microns, or if the uneven shape of the second surface 51B is complex, the imprint mold 50 can be manufactured with high processing accuracy.

[0070] The method for manufacturing the semiconductor device 90 according to this embodiment comprises at least the following steps.

[0071] In the resist formation process, the surfaces of multiple electrode terminals 11 provided on the semiconductor element 10 are covered with resist 40.

[0072] In the opening formation process, the imprint mold 50 is pressed against the resist 40 and pressure is applied to form an opening 42 in the resist 40.

[0073] In the resist curing process, after forming the opening 42, the semiconductor element 10 covered with resist 40 is irradiated with light, and then heated to cure the resist 40.

[0074] In the developing process, the diameter of the opening 42 is enlarged using a developing solution.

[0075] In the opening formation process, the imprint mold 50 is pressed against the resist 40 and pressurized such that the second surface 51B abuts against the surface 41 of the resist 40, and when viewed from above, the tips of each of the multiple pillars 52 overlap with the multiple electrode terminals 11.

[0076] In the resist curing process, depending on the shape of the uneven surface transferred to the surface 41 of the resist 40, light penetrates into the interior of the resist 40, causing the solubility of the resist 40 in the developer solution to change along the depth direction of the opening 42 around the opening 42.

[0077] According to this embodiment, the uneven surface provided on the second surface 51B of the imprint mold 50 is transferred to the surface 41 of the resist 40. Depending on this shape, light enters the interior of the resist 40, so that in the resist curing process, the surface 41 of the resist 40 functions as a lens, and the amount of incident light can be changed around the opening 42 along the thickness direction of the resist 40, that is, along the depth direction of the opening 42.

[0078] Furthermore, in the resist curing process, after irradiating the semiconductor element 10 covered with resist 40 with light, heating it promotes the crosslinking reaction and increases the crosslinking density in accordance with the amount of incident light. This makes it possible to change the curing state of the resist 40 in accordance with the amount of incident light. As a result, the solubility of the resist 40 in the developer solution around the opening 42 changes along the depth direction of the opening 42.

[0079] Furthermore, according to this embodiment, the surface 41 of the resist 40 has a convex shape around the pillar 52, and the amount of incident light can be increased in the part close to the surface 41. As a result, the hardening of the resist 40 progresses, and the shape of the surrounding area including the opening 42 stabilizes. On the other hand, the amount of incident light can be decreased as you move downwards around the opening 42. As a result, after the resist hardening process is performed, the solubility of the resist 40 in the developer around the opening 42 is higher in the part close to the surface of the electrode terminal 11 and lower in the part close to the surface of the resist 40.

[0080] As a result, during the developing process, the lower part is dissolved by the developing solution around the opening 42, and a sharply shaped opening 42 can be obtained.

[0081] Furthermore, according to this embodiment, by suppressing variations in the film thickness of the resist 40, it is possible to suppress variations in the depth of the opening 42 and, consequently, variations in the height of the bump electrode 81 to the same extent. When the resist 40 is formed by the method described above, variations in the film thickness of the resist 40 on the surface of the semiconductor element 10 can be suppressed to a few percent or less.

[0082] Furthermore, by aligning the imprint mold 50 and the semiconductor element 10 using an optical method, the pillar 52 can be precisely aligned with respect to the electrode terminals 11. As a result, the alignment accuracy of the opening 42 with respect to the electrode terminals 11 can be improved. For example, if the size of the semiconductor element 10 is 30 mm square, the misalignment between the electrode terminals 11 and the pillar 52 can be kept to about 1 μm.

[0083] Furthermore, the uneven shape of the surface 41 of the resist 40 is an important factor in determining the final diameter of the opening 42. The uneven shape of the surface 41 of the resist 40 is determined by the shape of the second surface 51B of the imprint mold 50, and the shape of the opening 42 before development is determined by the shape of the pillar 52. By processing the resist 40 using the imprint mold 50 and appropriately setting the conditions in the resist curing process and the development process, it is possible to form an opening 42 with a stable shape across the entire surface of the semiconductor element 10 and with little dimensional variation from the initial design. Examples of setting conditions in the resist curing process include the wavelength of incident light, the amount of incident light, and the light irradiation time. Examples of setting conditions in the development process include the type of developer and the development time.

[0084] Furthermore, the same effects can be achieved when using the second imprint mold 50B shown in Figure 1B and the third imprint mold 50C shown in Figure 1C.

[0085] Figure 5A is a schematic cross-sectional diagram illustrating the resist curing process when using the second imprint mold. Figure 5B is a schematic cross-sectional diagram illustrating the resist curing process when using the third imprint mold.

[0086] Similar to the case using the first imprint mold 50A, in the resist curing process, the uneven surface transferred to the surface 41 of the resist 40 functions as a lens, causing light incident on the surface 41 of the resist 40 to be refracted or scattered and enter the interior of the resist 40. In the example shown in Figure 5A, the recessed portion 41B acts like a concave lens for incident light, refracting light irradiated from directly above toward the focal point P shown in Figure 5A. In the example shown in Figure 5B, the convex portion 41A acts like a convex lens for incident light, refracting light irradiated from directly above toward the focal point P shown in Figure 5B.

[0087] As a result, as shown in Figures 5A and 5B, the amount of incident light can be increased in the area around the pillar 52, near the surface 41 of the resist 40. Consequently, the hardening of the resist 40 progresses, and the shape of the area including the opening 42 stabilizes. Furthermore, the amount of incident light can be reduced as you move downwards, increasing the solubility of the resist 40 in the developer.

[0088] Furthermore, after the development process, as shown in Figures 6A and 6B, when the second imprint mold 50B and the third imprint mold 50C are used, the lower part of the opening 42 is dissolved by the developing solution, and a sharply shaped opening 42 can be obtained. In the example shown in Figure 6B, the upper end of the first part 42A of the opening 42 has a larger diameter than its lower part.

[0089] In the opening formation process, it is preferable to heat the semiconductor element 10 covered with resist 40 while pressing the imprint mold 50 against the resist 40 and applying pressure. By doing so, the pillar 52 enters the interior of the resist 40 while the resist 40 is softened, so that the opening 42 can be easily formed.

[0090] In this process, the resist 40 extruded by the pillar 52 deforms the surface 41 so that it comes into close contact with the second surface 51B of the imprint mold 50. As a result, the resist 40 extruded by the pillar 52 is contained in the recess 51B1 provided on the second surface 51B.

[0091] The deformation of the surface 41 of the resist 40 in this manner allows the uneven shape of the surface 41 of the resist 40 to function as a lens during the resist hardening process following the opening formation process, thereby refracting or scattering incident light.

[0092] Furthermore, the method for manufacturing the semiconductor device 91 according to this embodiment includes the following additional steps compared to the method for manufacturing the semiconductor device 90 described above.

[0093] In the plating process, after the developing process, plating is performed on the semiconductor device 90 to deposit metal inside the opening 42.

[0094] In the bump electrode formation process, after the plating process, the resist 40 is removed and bump electrodes 81 are provided on the surface of each of the multiple electrode terminals 11.

[0095] In this way, sharp-shaped bump electrodes 81 can be formed with high precision and alignment for each of the multiple electrode terminals 11. Furthermore, variations in the height and diameter of the bump electrodes 81 can be suppressed. As a result, compared to using metal solder, it is possible to prevent bridging defects caused by melting and deformation when mounting the semiconductor device 91 on a circuit board using a flip-chip method, and it becomes easier to accommodate narrow pitches between electrode terminals.

[0096] Furthermore, when using electroplating in the plating process, it is preferable to include a seed layer formation step before the resist formation step in which a seed layer 20 is formed on the surface of at least a plurality of the electrode terminals 11.

[0097] Furthermore, in large-area semiconductor devices 10, as mentioned above, it is necessary to stabilize the shape of the bump electrodes 81 provided on each of the multiple electrode terminals 11. In particular, the tolerance range for the height of the bump electrodes 81 and the misalignment with the electrode terminals 11 becomes narrow.

[0098] According to this embodiment, even in such cases, sharp-shaped bump electrodes 81 that maintain the desired dimensional accuracy and positional accuracy relative to the electrode terminals 11 can be provided on the semiconductor element 10 all at once, and the number of pins and narrow pitch of the electrode terminals 11 can be easily accommodated. This improves the yield when flip-chip mounting. In particular, when the area of ​​the semiconductor element 10 is several tens of mm 2 ~several hundred mm 2 In this case, the manufacturing method of the semiconductor devices 90 and 91 according to this embodiment is particularly useful for stably obtaining sharp-shaped bump electrodes 81.

[0099] The semiconductor device 90 according to this embodiment comprises at least a semiconductor element 10 having a plurality of electrode terminals 11 on its surface, and a resist 40 covering the surface of the semiconductor element 10 and having a plurality of openings 42.

[0100] As viewed from the surface 41 of the resist 40, the opening 42 is provided so as to overlap with the electrode terminal 11 and reach the electrode terminal 11.

[0101] The surface 41 of the resist 40 is an uneven surface having protrusions 41A that surround the opening 42.

[0102] In cross-sectional view, the opening 42 has a second portion 42B that is wider along the surface 41 and is aligned with the thickness direction of the resist 40, and a first portion 42A that is narrower than the second portion 42B. The first portion 42A is continuous with the second portion 42B and is located above the second portion 42B.

[0103] In the semiconductor device 90, a sharp-shaped opening 42 that reaches the surface is provided for each of the multiple electrode terminals 11. By depositing metal to fill the inside of the opening 42, a semiconductor device 91 can be obtained that has sharp-shaped bump electrodes 81 with little variation in height and diameter.

[0104] <Example 1> Figure 7A is a schematic cross-sectional view of a bump electrode according to Modification 1. Figure 7B is a schematic cross-sectional view of another bump electrode according to Modification. For the sake of clarity, in Figure 7A and the subsequent drawings, the same reference numerals are used for parts that are the same as in the embodiment, and detailed explanations are omitted.

[0105] As shown in Figures 4 and 5A and 5B, by appropriately setting the uneven shape of the surface 41 of the resist 40, the position of the focal point P with respect to the incident light, that is, the distance to the opening 42 along the surface of the semiconductor element 10, and the depth from the surface 41 of the resist 40 can be adjusted during the resist curing process. By utilizing this, it is also possible to set the cross-sectional shape of the opening 42, and by extension the bump electrodes 82 and 83, to a desired shape.

[0106] For example, as shown in Figure 7A, the cross-sectional shape of the bump electrode 82 may be a circle with a cutout at the bottom. Also, as shown in Figure 7B, the cross-sectional shape of the bump electrode 83 may be hourglass-shaped. The shape of the bump electrode 81 provided on the semiconductor element 10 can be appropriately changed to enable the desired flip-chip mounting, by modifying the shape of the opening 42 and, consequently, the shape of the bump electrodes 82 and 83.

[0107] Furthermore, by shaping the bump electrode 83 as shown in Figure 7B, the stress applied to the connection between the bump electrode 83 and the electrode terminal 11 during flip-chip mounting can be reduced, thereby improving the reliability of the connection between the bump electrode 83 and the electrode terminal 11. In turn, the reliability of the connection between the semiconductor element 10 and the circuit board on which the semiconductor element 10 is mounted can be improved.

[0108] <Modification 2> Figure 8A is a schematic plan view of a resist with an opening according to Modification 2. Figure 8B is a cross-sectional view of Figure 8A in the VIIB-VIIB direction. Figure 9 is a schematic cross-sectional view of a bump electrode according to Modification 2.

[0109] Embodiment and Modification 1 show examples in which one opening 42 is provided on the upper surface of one electrode terminal 11, and examples in which one bump electrode 81-83 is provided on the upper surface of one electrode terminal 11. However, depending on the size of the electrode terminal 11 and the opening 42, and the requirements for joining the electrode terminal 11 and the bump electrodes 81-83, multiple openings 42 and multiple bump electrodes 81 may be provided on the upper surface of one electrode terminal 11. In the examples shown in Figures 8A and 8B, four openings 42 and four bump electrodes 81 are provided on the upper surface of one electrode terminal 11, but the number of openings 42 and bump electrodes 81 provided on the upper surface of one electrode terminal 11 is not particularly limited to this example. It can be appropriately changed according to the size of the electrode terminal 11 and the opening 42.

[0110] In other words, in the semiconductor device 90 shown in this modified example, multiple openings 42 are provided so as to overlap with and reach one electrode terminal 11 when viewed from the surface of the resist 40. In addition, in the semiconductor device 91 shown in this modified example, multiple bump electrodes 81 are provided on the upper surface of one electrode terminal 11.

[0111] By providing multiple bump electrodes 81 on the upper surface of a single electrode terminal 11, the contact area between the bump electrodes 81 on each electrode terminal 11 and the connection terminals of the mounting substrate or interposer can be increased, thereby improving the bonding strength. In addition, this improves the reliability of the connection between each electrode terminal 11 and the connection terminals of the mounting substrate or interposer, and also reduces the connection resistance.

[0112] Furthermore, even if one of the bump electrodes 81 is defective, the conductivity between the electrode terminal 11 and the bump electrode 81, and consequently between the electrode terminal 11 and the connection terminals of the mounting board or interposer, can be ensured, thereby suppressing a decrease in yield during bonding.

[0113] (Other embodiments) The components shown in Embodiment and Modifications 1 and 2 can be combined as appropriate to create new embodiments. For example, the shapes of the bump electrodes 82 and 83 shown in Modification 1 may be applied to Modification 2. In that case, the shape of the opening 42 in Modification 2 should also correspond to the shape of the bump electrode 82 or bump electrode 83.

[0114] Furthermore, in the embodiment and modified examples 1 and 2, the depth of the opening 42 generally corresponds to the thickness of the resist 40 and is in the range of approximately 0.5 μm to 20 μm. In addition, the ratio of the diameters of the first portion 42A and the second portion 42B in the opening 42 is preferably about 1:2 to 1:4. [Industrial applicability]

[0115] The imprint mold of this disclosure can uniformly and stably form openings in the resist of large-area semiconductor devices, and is useful for forming bump electrodes provided on electrode terminals that are increasingly multi-pin and narrow-pitch. [Explanation of Symbols]

[0116] 10 Semiconductor devices 11 Electrode terminal 12 Recognition Mark 20 Seed Layer 30 Light-transparent resin 31 Hardened part 32 Uncured area 40 Resist 41 Surface of resist 40 41A Convex part 41B Recess 42 Opening 42A Part 1 42B 2nd part 50 Imprint molds 50A First Imprint Mold 50B Second imprint mold 50C Third Imprint Mold 51 Main body 51A 1st page 51B 2nd side 51B1 recess 51B2 protrusion 52 Pillars 53 Recognition Mark 60 First Lens 70 Second lens 80 Metal plating layer 81 Bump electrodes 81A Part 3 81B Part 4 82, 83 Bump electrodes 90 Semiconductor Equipment 91 Semiconductor Equipment

Claims

1. Page 1 and, Opposite the first surface is a second surface which is an uneven surface, It comprises a plurality of pillars protruding from the second surface, An imprint mold characterized in that the second surface surrounding the pillar has a concave shape that is recessed toward the first surface.

2. In the imprint mold according to claim 1, The imprint mold is characterized by being made of a light-transmitting material.

3. A resist formation process in which the surface of multiple electrode terminals provided on a semiconductor device is covered with a resist, An opening formation step is to press the imprint mold described in claim 1 or 2 against the resist and apply pressure to form an opening in the resist, After forming the aforementioned opening, the resist curing step involves irradiating the semiconductor element covered with the resist with light, and then heating the semiconductor element to cure the resist. The process includes at least a developing step of expanding the diameter of the opening using a developing solution, In the opening formation step, the imprint mold is pressed against the resist and pressurized such that the second surface abuts the surface of the resist and, when viewed from above, the tips of each of the multiple pillars overlap the multiple electrode terminals. A method for manufacturing a semiconductor device, characterized in that, in the resist curing step, the light penetrates into the resist according to the shape of the uneven surface transferred to the surface of the resist, so that the solubility of the resist in the developer changes along the depth direction of the opening around the opening.

4. In the method for manufacturing a semiconductor device according to claim 3, A method for manufacturing a semiconductor device, characterized in that, after performing the resist curing step, the solubility of the resist in the developer around the opening is lower in the portion of the electrode terminal that is close to the surface, and higher in the portion of the resist that is close to the surface.

5. In the method for manufacturing a semiconductor device according to claim 3, A method for manufacturing a semiconductor device, characterized in that, in the resist curing step, the uneven surface transferred to the surface of the resist functions as a lens, causing the light incident on the surface of the resist to be refracted or scattered and to penetrate into the interior of the resist.

6. In the method for manufacturing a semiconductor device according to claim 3, A method for manufacturing a semiconductor device, characterized in that, in the opening formation step, the semiconductor element covered with the resist is heated while the imprint mold is pressed against the resist and pressurized.

7. In the method for manufacturing a semiconductor device according to claim 6, A method for manufacturing a semiconductor device, characterized in that, in the opening forming step, the resist extruded by the pillar deforms its surface so that it comes into close contact with the second surface and is housed in a recess provided on the second surface.

8. In the method for manufacturing a semiconductor device according to claim 3, After the development step, a plating step is performed to deposit metal inside the opening, A method for manufacturing a semiconductor device, further comprising: a bump electrode formation step of removing the resist after the plating step and providing bump electrodes on each of the surfaces of the plurality of electrode terminals.

9. A semiconductor element having multiple electrode terminals on its surface, The present invention comprises at least a resist that covers the aforementioned surface and has a plurality of openings, As viewed from the surface of the resist, the opening is provided so as to overlap with the electrode terminal and reach the electrode terminal. The surface of the resist is an uneven surface having protrusions that surround the opening, A semiconductor device characterized in that, in a cross-sectional view, the opening has a second portion that is wider along the surface and in the thickness direction of the resist, and a first portion that is narrower than the second portion.

10. The semiconductor device according to claim 9, wherein, as viewed from the surface of the resist, a plurality of the openings are provided so as to overlap with and reach one of the electrode terminals.

Images