JP4939313B2 - Manufacturing method of semiconductor device - Google Patents

Description

  The present invention relates to a method for manufacturing a semiconductor device, and more particularly to a dicing technique for dividing a semiconductor device having a light receiving portion from a wafer level to a chip level.

  2. Description of the Related Art In recent years, semiconductor devices have been downsized due to the trend toward high-density mounting of semiconductor devices. A semiconductor device that acquires imaging data as an electrical signal by photoelectrically converting incident light into a voltage signal corresponding to the amount of light, so-called an image sensor module, is also being mounted on a small portable device. Therefore, there is a need for further downsizing and thinning.

  FIG. 8 is a schematic configuration diagram of a conventional image sensor package. As shown in FIG. 8, the conventional image sensor package has a semiconductor substrate 1a in which a light receiving portion (image sensor) 2 is formed in a ceramic or pre-molded box-shaped package case 91. In this configuration, it is electrically connected to the external terminal 8 through the wire 92, the metal wiring 93, and the contact plug 6. Further, the semiconductor substrate 1a having the light receiving portion 2 is sealed by being covered with a glass plate 11b. The glass plate 11b is bonded to the package case 91 by the adhesive resin layer 9, and has an optical filter film 15 for selectively transmitting light in a predetermined wavelength region on the upper layer surface thereof. Further, on the semiconductor substrate 1 a, in addition to the light receiving unit 2, an element circuit 3 is provided, and an electrode pad 4 for making electrical connection thereto is provided on the surface.

  Conventionally, as shown in FIG. 8, it is common to package an image sensor chip by an LCC (Leadless Chip Carrier) package. However, as shown in FIG. 8, in this package, since it is necessary to draw out wiring in the plane direction by the wire line 92 in order to form an electrical connection between the light receiving unit 2 and the external terminal 8, the chip size (of the semiconductor substrate 1a) This results in a large package that is 1 mm or more wider in the plane direction than the (substrate surface size) (see FIG. 9). FIG. 9 is a schematic plan view of a conventional image sensor package.

  Therefore, in order to reduce the package size and reduce the thickness, for example, in Patent Document 1 below, a glass lid for protecting and sealing the image sensor portion is bonded to the semiconductor wafer in the form of a wafer, and then the individual semiconductor sensor. Is disclosed. Further, in Patent Document 2 below, a through electrode is formed on a semiconductor wafer, an electric signal is transmitted to the back side of the semiconductor device, and an external electrode terminal is provided through circuit wiring, whereby an electrode terminal is provided on the back side of the semiconductor device. A method of forming is disclosed.

JP 2005-209966 A JP 2001-351997 A

  FIG. 10 is a schematic cross-sectional structure diagram of a state in which the light receiving unit 2 and the glass member (glass wafer 11) are formed at the wafer level based on the methods described in Patent Documents 1 and 2. The semiconductor wafer 1 has a protective insulating film 5 and external terminals 8 on the back surface side, and has a light receiving portion 2 in a plurality of chip regions on the front surface side, and an adhesive resin between these adjacent light receiving portions 2. It has a layer 9. Further, the glass wafer 11 is bonded and formed on the semiconductor wafer 1 via the adhesive resin layer 9, and a certain space is formed on the upper surface of the glass wafer 11 and the light receiving unit 2 and is in a sealed state. ing. An optical filter film 15 for transmitting light in a specific wavelength range is formed on the upper surface of the glass wafer 11.

  Further, in order to form an electrical connection between the light receiving unit 2 and the external terminal 8, a back surface wiring 7 is formed on the back surface side of the semiconductor wafer 1, and electrical connection between the back surface wiring 7 and the front surface side of the semiconductor wafer 1 is performed. A contact plug 6 is formed for general connection. An electrode pad 4 connected to the contact plug 6 is formed on the surface side of the semiconductor wafer 1, and the electrical connection from the light receiving unit 2 to the external terminal 8 is achieved by electrically connecting the electrode pad 4 and the light receiving unit 2. Connection is secured.

  As shown in FIG. 10, after the light receiving unit 2 and the glass wafer 11 are formed at the wafer level, the semiconductor wafer 1 is placed on the dicing sheet 30, and the semiconductor wafer 1 is diced along the scribe line. Is performed to divide the wafer 1 into chip levels.

  At this time, as shown in FIG. 11A, it is originally desirable if dicing can be performed at once from the glass wafer 11 side to the bottom surface of the semiconductor wafer 1 along the scribe line 10 (in FIG. 11, The dicing sheet 30 is not shown). However, as shown in FIG. 10 or FIG. 11A, the wafer to be diced has a two-layer structure of a semiconductor wafer 1 and a glass wafer 11. Usually, the material of the semiconductor wafer 1 is Si, while the material of the glass wafer 11 is glass. Therefore, the wafer to be diced has a two-layer structure of different materials.

  In general, when materials with different machinability, such as glass and Si, are cut simultaneously using a single dicing blade, differences in dicing conditions, blade clogging, or blade breakage due to cutting load, etc. occur smoothly. There is a problem that it is difficult to perform dicing.

  Therefore, when cutting a wafer having a multilayer structure composed of materials having different machinability, dicing is performed using a dicing blade suitable for cutting the material constituting the layer for each layer. The method is adopted. That is, as shown in FIG. 11B, when dicing a glass wafer 11 made of glass having relatively poor machinability, dicing is performed using a dicing blade having coarse abrasive grains and a wide blade width. When dicing the semiconductor wafer 1 having better machinability than glass, dicing is performed using a dicing blade with fine abrasive grains and a narrow blade width.

  By the way, when a laminate having different hardness, elasticity, or machinability is cut by dicing as described above, the glass wafer 11 supported by the elastic adhesive resin layer 9 is bonded by vibration or impact during dicing. Chipping and cracking may occur at the interface and other surfaces. FIG. 12 is a schematic structural diagram of the state divided into chip levels. As shown in FIG. 12, the glass part 11a (obtained by dividing the glass wafer into chips) is formed with chips 94 and cracks 95 generated during dicing.

  The presence of these chippings 94 and cracks 95 lowers the adhesion between the glass member 11a and the semiconductor substrate 1a (obtained by dividing the semiconductor wafer 1 into chips), or the light receiving part 2 and its upper surface. This may make it easier for moisture to enter the space region, thereby reducing reliability. Further, when the occurrence of chipping or cracks is caused near the surface of the glass part 11a, the occurrence of chipping or cracks is also induced in the optical filter film 15 (not shown in FIG. 12) formed in the upper layer at the time of occurrence. there's a possibility that. This is because the optical filter film 15 is hard and fragile, and thus is easily cracked or cracked by receiving an impact applied to the glass portion 11a. When such a defect occurs in the optical filter film 15, there arises a problem that the optical filter film 15 is easily peeled off from the glass portion 11a later.

  Furthermore, in the state shown in FIG. 12, since the side surface of the glass part 11a is exposed, there is a concern that unnecessary light may enter the light receiving unit 2 from the side surface. When unnecessary light is incident on the light receiving unit 2, optical sensitivity and characteristics may be deteriorated. For this reason, after dividing | segmenting into a chip level, in order to prevent incidence | injection of the unnecessary light from the glass part 11a side surface, it is necessary to install the module component 95 for shielding the side surface of the glass part 11a for every chip | tip (FIG. 13). For this reason, there is also a problem that the completed image sensor module is increased in size corresponding to the width of the module component 95 rather than the size of the semiconductor substrate 1a.

  In view of the above problems, an object of the present invention is to provide a method of manufacturing a semiconductor device capable of realizing further reduction in the size of an image sensor module while maintaining high reliability.

  In order to achieve the above object, a method of manufacturing a semiconductor device according to the present invention includes forming a light receiving portion in a plurality of chip regions formed on a semiconductor wafer and then forming an adhesive resin layer between adjacent light receiving portions. And a first step of bonding and forming a glass wafer over the entire surface of the semiconductor wafer including the light receiving portion via the adhesive resin layer, and a first blade width with a first particle size after the first step. A dicing blade is used for dicing cutting on the scribe line from the glass wafer side to a depth position in the adhesive resin layer formed on the scribe line, thereby forming a groove between adjacent chips. A step, a third step of depositing a photosensitive resin layer on the entire upper surface of the glass wafer so as to completely fill the groove after completion of the second step, and a step after the completion of the third step. A fourth step of removing the photosensitive resin layer deposited on the upper surface of the glass wafer in an upper region of the light receiving unit; and after the fourth step, the second particle size is smaller than the first particle size and the second particle size is smaller than the first particle size. The photosensitive resin layer formed on the scribe line from the glass wafer side by the second dicing blade having a second blade width that is narrower than the one blade width, and the semiconductor below the scribe line And a fifth step of dividing each chip by dicing cutting to the bottom surface of the wafer.

  According to the first feature of the method for manufacturing a semiconductor device according to the present invention, after completion of the second step of cutting the glass wafer, the photosensitive resin layer is filled in the groove formed at the time of cutting, so that it is cut by any chance. Even if chipping or cracks sometimes occur, these defects are covered with the photosensitive resin layer. Therefore, even if an impact or stress is applied from the outside after that, the impact is absorbed by the filled photosensitive resin layer, and therefore the occurrence of further chipping or cracking is not induced by the presence of the defect. .

  Further, the photosensitive resin layer is covered on the side surface of the glass part included in the chip after cutting, thereby preventing the inflow of moisture from the outside. That is, it is possible to avoid a situation where the reliability is lowered due to the flow of moisture into the light receiving unit.

  Further, after completion of dicing, a photosensitive resin layer as a light shielding material is already formed on the side surface of the glass portion. For this reason, unlike the conventional configuration, it is not necessary to perform the light shielding module component attaching process for each of the divided chips, thereby simplifying the manufacturing process and improving the productivity. Furthermore, since it is not necessary to attach the module parts after the division, it can be packaged with the divided chip size, and the size can be reduced as compared with the conventional case.

  In the method for manufacturing a semiconductor device according to the present invention, after forming a light receiving portion in a plurality of chip regions formed on a semiconductor wafer, an adhesive resin layer is formed between the adjacent light receiving portions, and the adhesive resin is formed. A first step of bonding and forming a glass wafer over the entire surface of the semiconductor wafer including the light receiving portion through a layer, and a first dicing blade having a first blade width with a first grain size after the first step, A second step of forming a groove between adjacent chips by dicing cutting on the scribe line from the glass wafer side to a depth position in the adhesive resin layer formed on the scribe line; After the completion of the two steps, a third dicing blade having a third particle size in which the abrasive grains are finer than the first particle size and having a third blade width narrower than the first blade width, A third step of forming the groove to a position deeper than the upper surface of the semiconductor wafer by dicing cutting to a depth position within the range not reaching the bottom surface of the semiconductor wafer from the glass wafer side, and the end of the third step Thereafter, a fourth step of depositing a photosensitive resin layer on the entire upper surface of the glass wafer so as to completely fill the groove, and after completion of the fourth step, deposition on the upper surface of the glass wafer in the upper region of the light receiving unit. A fifth step of removing the photosensitive resin layer formed, and after completion of the fifth step, the third dicing blade or a fourth grain size with finer abrasive grains than the third grain size and a blade width smaller than the third blade width The sensation formed on the scribe line from the glass wafer side by the fourth dicing blade having a narrow fourth blade width. RESIN layer and then diced to the semiconductor wafer bottom surface of the lower layer and the second, comprising a sixth step of dividing each said chip.

  According to the second feature of the method for manufacturing a semiconductor device according to the present invention, as in the first feature, chipping and cracks are generated at the time of cutting by filling a photosensitive resin layer in a groove portion that is formed at the time of cutting. Even if it is, since these defects are covered with the photosensitive resin layer, even if an impact or stress is given from the outside afterwards, the impact is absorbed by the filled photosensitive resin layer. The presence of this does not induce further chipping or cracking.

  And when this photosensitive resin layer is covered by the side surface of the glass part with which the chip | tip after a cutting | disconnection is equipped, it has the effect which prevents inflow of the water | moisture content from the outside. Furthermore, in the case of this feature, since the photosensitive resin layer is also formed on the side surface of the semiconductor substrate constituting the cut chip, the effect of preventing moisture from flowing into the chip and the effect of mitigating external stress are further enhanced. Can do.

  In addition to the first or second feature, the method for manufacturing a semiconductor device according to the present invention has a third feature that the first blade width is 100 μm or more.

  In addition to the first to third features, the method for manufacturing a semiconductor device according to the present invention has a fourth feature that the first grain size is in the range of 300 to 1,000. To do.

  According to a fifth aspect of the present invention, there is provided a method for manufacturing a semiconductor device, wherein, in addition to any one of the first to fourth features, the photosensitive resin layer is made of an epoxy resin or a silicone resin. It is characterized by.

  According to the configuration of the present invention, it is possible to provide a method for manufacturing a semiconductor device capable of further reducing the size of the image sensor module while maintaining high reliability.

  Embodiments of a method for manufacturing a semiconductor device according to the present invention (hereinafter referred to as “method of the present invention” as appropriate) will be described below with reference to the drawings.

[First Embodiment]
A first embodiment of the method of the present invention (hereinafter referred to as “this embodiment” as appropriate) will be described with reference to FIGS.

  FIG. 1 is a schematic cross-sectional structure diagram showing a state of a semiconductor wafer before dicing. First, the light receiving part (image sensor) 2, the element circuit 3, and the electrode pad 4 for electrical connection with the light receiving part 2 and the element circuit 3 are formed in a plurality of chip regions of the semiconductor wafer 1. A protective insulating film 5, a contact plug 6, and a back surface wiring 7 are formed on the back surface side of the semiconductor wafer 1, and external terminals 8 that are electrically connected to the back surface wiring 7 are provided outside the semiconductor wafer 1. ing. Under such a state, after the adhesive resin layer 9 is formed between the adjacent light receiving portions 2, the glass wafer 11 is formed on the entire wafer surface including the upper surface of the light receiving portion 2 through the adhesive resin layer 9. An optical filter film 15 for transmitting predetermined light is formed on the surface of the glass wafer 11.

  Hereinafter, a procedure for dividing the semiconductor wafer 1 in the state of FIG. 1 for each chip will be described with reference to schematic cross-sectional views (FIGS. 2A to 2D) for each process. In addition, each schematic sectional drawing is illustrated typically to the last, and the reduced scale on drawing does not necessarily correspond with an actual reduced scale. FIG. 3 is a flowchart showing steps according to the present embodiment, and each step (steps # 1 to # 5) in the following description represents each step of the flowchart shown in FIG. To do.

  First, the back side (external terminal 8 side) of the semiconductor wafer 1 in the state shown in FIG. 1 is bonded and fixed on a stage (not shown) with a dicing tape (step # 1).

  Next, as shown in FIG. 2A, adjacent to the glass wafer 11 by using a dicing blade (hereinafter referred to as “first dicing blade”) that is optimal for glass cutting and has a large particle size and a large blade width. The scribe line 10 formed between the chip regions to be cut is diced to a depth position in the adhesive resin layer 9 (step # 2). As the first dicing blade used in this step, a blade having a particle size in the range of 300 to 1000 and a blade width of 100 μm or more is used.

  In addition, when performing alignment at the time of dicing, it is performed by visual recognition when the scribe line 10 is visible, and IR (Infrared: infrared) when the scribe line 10 cannot be visually recognized due to low transparency of the adhesive resin layer 9 or the like. ) Alignment is performed by using a transmission camera or recognizing the cutting position with a scribe mark or the like on the back surface pattern of the semiconductor wafer 1. Then, with the dicing position aligned with the center (center) of the scribe line 10, the dicing cut is performed to the depth position inside the adhesive resin layer 9. By this step, the groove 12 from the upper surface of the glass wafer 11 to the depth position inside the adhesive resin layer 9 is formed on the scribe line 10 (see FIG. 2B).

  Next, as shown in FIG. 2C, an epoxy-based or silicone-based photosensitive resin layer 13 is deposited on the entire upper surface of the glass wafer 11 (step # 3). By this step, the photosensitive resin layer 13 is also filled in the groove 12 formed in step # 2.

  Next, as shown in FIG. 2D, only the photosensitive resin layer 13 deposited above the light receiving portion 2 is removed by exposure and development processing from the photosensitive resin layer 13 deposited in Step # 3. (Step # 4).

  Next, as shown in FIG. 2E, a dicing blade (hereinafter referred to as “second dicing blade”) having a finer particle size and a smaller blade width than the first dicing blade used in step # 2 is used. Then, dicing cut is performed from the glass wafer 11 side to the bottom surface position of the semiconductor wafer 1 on the scribe line 10 (step # 5). In step # 2, the glass wafer 11 on the scribe line 10 is diced in advance to form a groove 12, and the groove 12 is filled with the photosensitive resin layer 13 in step # 3. For this reason, in the dicing process according to this step, the photosensitive resin layer 13 formed on the scribe line 10 and the underlying semiconductor wafer 1 are diced, and the glass wafer 11 is not diced. For this reason, it is possible to perform dicing using the second dicing blade having a finer particle size and a smaller blade width than the first dicing blade used in step # 2.

  Through the dicing process according to Step # 5, the photosensitive resin layer 13 formed on the scribe line 10 and the underlying semiconductor wafer 1 are diced and cut, whereby the semiconductor wafer 1 is divided into chips. It will be. As a result, as shown in FIG. 2E, each divided chip has the photosensitive resin layer 13 on the side surface of the glass part 11a (the structure part obtained by dividing the glass wafer 11 for each chip). It becomes the composition to provide. Thereby, unnecessary light can be prevented from entering the light receiving unit 2 from the side surface of the glass unit 11a.

  Thereafter, a lens component 21 having a size corresponding to the divided chip size is attached to the upper layer of the photosensitive resin layer 13 to complete the image sensor module as shown in FIG.

  In the case of the conventional configuration, after cutting for each chip, it is necessary to form a module component 95 for each chip so as to cover the side surface of the glass portion 11a included in each chip and to perform light shielding treatment on the side surface of the glass portion 11a. . However, according to the method of the present invention, since the photosensitive resin layer 13 for shielding light on the side surface of the glass portion 11a is already formed at the stage of being divided for each chip, There is no need to perform a component mounting process, the manufacturing process is simplified, and productivity is improved.

  Further, by attaching the photosensitive resin layer 13 to the side surface of the glass part 11a, even if chipping or cracks occur in the cutting process of the glass wafer 11 according to Step # 2, these are the photosensitive resin layer 13. Since it is covered, the impact is absorbed by the photosensitive resin layer 13 even if an impact or stress from the outside occurs later. Thereby, each chip (including the light receiving unit 2) can be protected from external stress and impact.

  Further, the glass portion 11a, the adhesive resin layer 9, and the semiconductor substrate 1a (obtained by dividing the semiconductor wafer 1 for each chip) are covered with the photosensitive resin layer 13, so that moisture flows into them. Therefore, the moisture resistance can be improved.

  Further, in the case of the conventional configuration, the module component 95 is attached to the side surface of the glass portion 11a after being divided for each chip. Therefore, each chip has a size larger by the width of the module component 95 than the width of the glass portion 11a. I didn't get it. However, according to the method of the present invention, since the light-blocking adhesive resin layer 9 has already been formed at the stage of being divided for each chip, the chip size can be increased by mounting a new member after the division. Absent. Further, since the adhesive resin layer 9 is formed on the side surface of the glass portion 11a, strictly speaking, the chip size is larger than the width of the glass portion 11a, but on the side surface of the glass portion 11a after the dicing process according to Step # 5. The size is larger than the width of the glass portion 11a by the thickness of the remaining adhesive resin layer 9, and the chip size is reduced as compared with the conventional configuration in which the module component 95 is mounted after the division.

[Second Embodiment]
A second embodiment of the method of the present invention (hereinafter referred to as “this embodiment” as appropriate) will be described with reference to FIGS.

  Hereinafter, a procedure for dividing the semiconductor wafer 1 in the state of FIG. 1 for each chip will be described with reference to schematic sectional views (FIGS. 5A to 5E) for each process. FIG. 6 is a flowchart showing the process according to the present embodiment, and each step (steps # 11 to # 16) in the following description represents each step of the flowchart shown in FIG. To do.

  In FIG. 5, the same components as those shown in FIGS. 1, 2, and 4 are denoted by the same reference numerals, and description thereof is omitted. Moreover, about the process which is the same as that of each step (step # 1- # 5) which concerns on 1st Embodiment, or similar, it describes so and description is simplified.

  First, as in step # 1, the semiconductor wafer 1 is bonded and fixed on the stage with a dicing tape (step # 11). Next, as shown in FIG. 5A, between the adjacent chip regions from the glass wafer 11 side, using a first dicing blade that is optimal for glass cutting and has a large particle size and a large blade width, as in step # 2. Then, the scribe line 10 formed on the substrate is diced to a depth position in the adhesive resin layer 9 (step # 12). By this step, a groove portion 12a having a relatively large diameter formed from the upper surface of the glass wafer 11 to the depth position in the adhesive resin layer 9 is formed.

  Next, the bottom surface of the semiconductor wafer 1 is reached from the glass wafer 11 side by a dicing blade having a finer particle size than the first dicing blade used in step # 12 and having a smaller blade width (hereinafter referred to as “third dicing blade”). Dicing cut is performed on the scribe line 10 to a depth position within the range not to be performed (step # 13). By this step, the groove 12 is formed by connecting the groove 12a and the groove 12b having a relatively small diameter formed from the depth position in the adhesive resin layer 9 to the depth position in the semiconductor wafer 1 ( (Refer FIG.5 (b)).

  Next, as shown in FIG. 5C, as in step # 3, an epoxy-based or silicone-based photosensitive resin layer 13 is deposited on the entire upper surface of the glass wafer 11 (step # 14). By this step, the photosensitive resin layer 13 is also filled in the groove 12 formed in steps # 12 and # 13. Since the groove 12 is formed to a depth position in the semiconductor wafer 1, the photosensitive resin layer 13 is formed not only on the side surface of the glass wafer 11 but also on the side surface of the semiconductor wafer 1 by this step. Become.

  Next, as in step # 4, as shown in FIG. 5D, the photosensitive resin layer 13 deposited above the light receiving portion 2 is removed by exposure and development processing (step # 15).

  Next, as shown in FIG. 5 (e), from the glass wafer 11 side, a dicing blade (hereinafter referred to as “fourth dicing blade”) having a finer particle size than the third dicing blade and having a smaller blade width is used. Dicing cut is performed on the scribe line 10 to the position of the bottom surface of the semiconductor wafer 1 (step # 16). Thereby, the photosensitive resin layer 13 is covered not only on the side surface of the glass part 11a but also on the side surface of the semiconductor substrate 1a. Then, by attaching the lens component 21 having a size corresponding to the divided chip size to the upper layer of the photosensitive resin layer 13, an image sensor module as shown in FIG. 7 is completed.

  Also in the method according to this embodiment, as in the case of the first embodiment, the photosensitive resin layer 13 as a light shielding member is already formed on the side surface of the glass portion 11a when it is divided for each chip. After that, there is no need to perform a process of attaching the module component 95 as a light shielding member for each chip. For this reason, as in the conventional case, the manufacturing process can be simplified and the productivity can be improved as compared with the conventional configuration while preventing the unnecessary light from being incident on the light receiving unit 2 from the glass unit 11. .

  In the case of this embodiment, the photosensitive resin layer 13 is also formed on the side surface of the semiconductor substrate 1a as compared with the first embodiment, thereby preventing the inflow of moisture into the semiconductor substrate 1a and the external The stress relieving effect from can be further enhanced.

  In step # 13, dicing cutting is performed from the bottom surface position of the groove 12a formed in step # 12 to a depth position in a range not reaching the bottom surface of the semiconductor wafer 1. That is, in this step, the semiconductor wafer 1 is mainly cut. Accordingly, the third dicing blade used in this step is preferably a dicing blade suitable for cutting the semiconductor wafer 1.

  Further, at the time immediately before the start of step # 16, the semiconductor wafer 1 has already been cut to a depth position within the range not reaching the bottom surface in step # 13. In particular, in step # 13, when the cutting is performed to the maximum depth within the range not reaching the bottom surface, in this step # 16, the photosensitive resin layer 13 embedded mainly in the groove 12 is cut. Thus, the thickness of the semiconductor wafer 1 to be cut is very small. Therefore, the fourth dicing blade used in this step is preferably a dicing blade suitable for cutting the photosensitive resin layer 13.

  On the other hand, in step # 5 in the first embodiment, the photosensitive resin layer 13 formed on the scribe line and the semiconductor wafer 1 corresponding to the initial thickness are cut. Accordingly, the second dicing blade used in step # 5 is preferably a dicing blade suitable for cutting the photosensitive resin layer 13 and the semiconductor wafer 1.

  In particular, step # 5 and step # 16 are processes for separating each chip, and in particular, a dicing blade having a finer particle size and a narrower blade width than the first dicing blade used in step # 2 or # 12 is used. It is done. Note that the fourth dicing blade used in step # 16 may have the same particle size and the same blade width as the second dicing blade used in step # 5.

  Further, step # 13 is a process for forming the groove 12b for embedding the photosensitive resin layer 13 after that, so that the dicing blade used in step # 13 has a coarse particle size enough to cut the glass wafer 11. And does not require a wide blade width. That is, in order to achieve high integration, it is preferable to make the line width of the scribe line as narrow as possible. However, in Step # 13, a third dicing blade having a finer grain size and a narrower blade width than the first dicing blade is used. On the other hand, as described above, since step # 16 is a process of mainly cutting the photosensitive resin layer 13, it is preferable to use a fourth dicing blade having a finer particle size and a narrower blade width than the third dicing blade. is there. However, as long as the blade is narrower and has a narrower blade width than at least the first dicing blade, it may be cut using a dicing blade having the same particle size and blade width in step # 13 and step # 16. Absent.

Schematic structure diagram of semiconductor wafer for image sensor before dicing Schematic sectional view for each step according to the first embodiment of the present invention The flowchart which shows the manufacturing process which concerns on 1st Embodiment of this invention. Schematic sectional view of a semiconductor device obtained by the manufacturing process according to the first embodiment of the present invention Schematic sectional view for each step according to the second embodiment of the present invention The flowchart which shows the manufacturing process which concerns on 2nd Embodiment of this invention. Schematic sectional view of a semiconductor device obtained by a manufacturing process according to a second embodiment of the present invention Schematic configuration diagram of a conventional image sensor package Schematic plan view of a conventional image sensor package Conventional schematic cross-sectional structure of a light receiving unit and a glass wafer formed at the wafer level Schematic cross-sectional structure diagram for explaining a process when dicing a wafer having a light receiving portion and a glass wafer formed at the wafer level Schematic structural diagram of a semiconductor chip in a state divided by a conventional dicing method Schematic configuration diagram of a semiconductor chip with module parts attached

Explanation of symbols

1: Semiconductor wafer 1a: Semiconductor substrate (obtained by dividing a semiconductor wafer into chips)
2: Light receiver (image sensor)
3: Element circuit 4: Electrode pad 5: Protective insulating film 6: Contact plug 7: Back wiring 8: External terminal 9: Adhesive resin layer 10: Scribe line 11: Glass wafer 11a: Glass part (glass wafer is divided into chips) Was obtained)
11b: Glass plate 12: Groove portion 12a: Groove portion with large diameter 12b: Groove portion with small diameter 13: Photosensitive resin layer 15: Optical filter film 21: Lens component 30: Dicing sheet 91: Conventional package case 92: Wire wire 93: Metallic wiring 95: Module parts

Claims (5)

The semiconductor wafer including a light receiving part formed in a plurality of chip regions formed on a semiconductor wafer, and then forming an adhesive resin layer between the adjacent light receiving parts, and including the light receiving part via the adhesive resin layer A first step of bonding and forming a glass wafer on the entire surface;
After the first step, the depth position in the adhesive resin layer formed on the scribe line from the glass wafer side by the first dicing blade having the first blade width with the first particle size. A second step of dicing up to and forming a groove between the adjacent chips;
A third step of depositing a photosensitive resin layer on the entire upper surface of the glass wafer so as to completely fill the groove after the second step;
A fourth step of removing the photosensitive resin layer deposited on the upper surface of the glass wafer in the upper region of the light receiving portion after the third step;
After completion of the fourth step, the second dicing blade having a second particle size that is finer than the first particle size and having a blade width that is narrower than the first blade width is used to form the scribe line on the scribe line. A semiconductor device comprising: a photosensitive resin layer formed on the scribe line from the glass wafer side; and a fifth step of dividing each chip by dicing cutting to a bottom surface of the semiconductor wafer below the photosensitive resin layer. Manufacturing method. The semiconductor wafer including a light receiving part formed in a plurality of chip regions formed on a semiconductor wafer, and then forming an adhesive resin layer between the adjacent light receiving parts, and including the light receiving part via the adhesive resin layer A first step of bonding and forming a glass wafer on the entire surface;
After the first step, the depth position in the adhesive resin layer formed on the scribe line from the glass wafer side by the first dicing blade having the first blade width with the first particle size. A second step of dicing up to and forming a groove between the adjacent chips;
After completion of the second step, the semiconductor is separated from the glass wafer side by a third dicing blade having a third particle size in which the abrasive grains are finer than the first particle size and having a third blade width narrower than the first blade width. A third step of forming the groove to a position deeper than the top surface of the semiconductor wafer by dicing cutting to a depth position within a range not reaching the wafer bottom surface;
A fourth step of depositing a photosensitive resin layer on the entire upper surface of the glass wafer so as to completely fill the groove after the third step;
After the fourth step, a fifth step of removing the photosensitive resin layer deposited on the upper surface of the glass wafer in the upper region of the light receiving unit;
After completion of the fifth step, the third dicing blade, or a fourth dicing blade having a fourth blade size with finer abrasive grains than the third particle size and a fourth blade width narrower than the third blade width, And a sixth step of dividing the chip on the scribe line by dicing cutting from the glass wafer side to the photosensitive resin layer formed on the scribe line and the bottom surface of the semiconductor wafer. A method of manufacturing a semiconductor device.   The method of manufacturing a semiconductor device according to claim 1, wherein the first blade width is 100 μm or more.   4. The method of manufacturing a semiconductor device according to claim 1, wherein the first particle size is in a range of 300 to 1000. 5.   The method for manufacturing a semiconductor device according to claim 1, wherein the photosensitive resin layer is made of an epoxy-based or silicone-based resin.

Images