Inspection device, bonding system, and inspection method

Abstract

This inspection device inspects a laminated substrate that is provided with a target substrate and a die to be bonded to the target substrate. The inspection device comprises: a light source; an optical system that irradiates a circular irradiation region on a surface of the laminated substrate with light emitted from the light source; a light detector that detects reflected light from the irradiation region; a drive device that changes the position of the irradiation region; and a control circuit that controls the position of the irradiation region. The control circuit sets, on the surface of the laminated substrate, a regular hexagonal lattice which is provided with a plurality of regular hexagonal regions of the same size, and sets the center of the irradiation region at each of the center of the regular hexagonal region.

Description

Inspection Device, Bonding System, and Inspection Method

[0001] The present disclosure relates to an inspection device, a bonding system, and an inspection method.

[0002] The substrate processing apparatus described in Patent Document 1 includes a mounting unit that mounts a chip on a substrate, and an inspection unit that inspects whether the bonding state of the chip to the substrate is good or bad. A chip is generally also called a die.

[0003] International Publication No. 2023 / 145558

[0004] One embodiment of the present disclosure provides a technology capable of shortening the inspection time or improving the inspection accuracy of a stacked substrate.

[0005] An inspection device according to one embodiment of the present disclosure inspects a stacked substrate including a target substrate and a die bonded to the target substrate. The inspection device includes a light source, an optical system that irradiates light emitted from the light source onto a circular irradiation region on the surface of the stacked substrate, a photodetector that detects reflected light from the irradiation region, a drive device that changes the position of the irradiation region, and a control circuit that controls the position of the irradiation region. The control circuit sets a regular hexagonal lattice including a plurality of regular hexagonal regions of the same size on the surface of the stacked substrate, and sets the center of the irradiation region at the center of each regular hexagonal region.

[0006] According to one embodiment of the present disclosure, the inspection time of a stacked substrate can be shortened or the inspection accuracy can be improved.

[0007] Figure 1 is a plan view showing a bonding system according to one embodiment. Figure 2 is a cross-sectional view showing an example of a carrier before picking up the die. Figure 3 is a cross-sectional view showing an example of a target substrate before bonding the die. Figure 4 is a cross-sectional view showing an example of a target substrate after bonding the die. Figure 5 is a cross-sectional view showing the operation of a bonding apparatus according to one embodiment. Figure 6 is a cross-sectional view showing the operation of the bonding apparatus following Figure 5. Figure 7 is a cross-sectional view showing an inspection apparatus according to one embodiment. Figure 8 is a cross-sectional view showing an example of a void. Figure 9 is a diagram showing an example of a square grid set on the non-bonded surface of the die during full-surface inspection. Figure 10 is a diagram showing an example of a regular hexagonal grid set on the non-bonded surface of the die during full-surface inspection. Figure 11 is a diagram showing an example of the relationship between inspection times T1 and T2 during full-surface inspection. Figure 12 is a diagram showing an example of the degree of overlap of the irradiation area. Figure 13 is a diagram showing an example of a square grid and a regular hexagonal grid set on the non-bonded surface of the die during thinning inspection. Figure 14 is a diagram showing an example of the relationship between the maximum radii R1 and R2 of voids that are missed during thinning inspection.

[0008] Embodiments of this disclosure will be described below with reference to the drawings. In each drawing, identical or similar components are denoted by the same reference numerals, and their descriptions may be omitted. In this specification, the X-axis, Y-axis, and Z-axis directions are perpendicular to each other. The X-axis and Y-axis directions are horizontal, and the Z-axis direction is vertical.

[0009] The X-axis direction includes the positive X-axis direction and the negative X-axis direction, which is the opposite direction to the positive X-axis direction. The Y-axis direction includes the positive Y-axis direction and the negative Y-axis direction, which is the opposite direction to the positive Y-axis direction. The Z-axis direction includes the positive Z-axis direction and the negative Z-axis direction, which is the opposite direction to the positive Z-axis direction. The positive Z-axis direction is upward, and the negative Z-axis direction is downward.

[0010] Referring to Figures 1 to 4, a bonding system 1 according to one embodiment will be described. The bonding system 1, for example, picks up a die D from a carrier E, and bonds the die D and the target substrate W with the bonding surface Da of the picked-up die D facing the bonding surface Wa of the target substrate W. For example, the bonding apparatus 33 picks up multiple dies D one by one from the carrier E in sequence and bonds them to the target substrate W.

[0011] As shown in Figure 2, the carrier E holds multiple dies D. The carrier E holds each die D with its bonding surface Da facing upward. This allows for activation and hydrophilization of the bonding surface Da of each die D. The carrier E has a carrier substrate E1 and a resin film E2 provided on the surface of the carrier substrate E1 facing the die D.

[0012] Carrier E holds multiple dies D on a resin film E2. Carrier E attaches the dies D, for example, by electrostatic attraction. By pressing the dies D against the resin film E2, the resin film E2 can be deformed to allow gas to escape from between the dies D and the resin film E2, making it possible to vacuum-attach the dies D to the resin film E2.

[0013] The carrier substrate E1 may be conductive or insulating. A first through-hole E3 is formed in the carrier substrate E1, penetrating through the substrate in the thickness direction. The die D can be detached from the carrier E by supplying gas to the first through-hole E3 or by inserting a pin (not shown) into the first through-hole E3. The number and arrangement of the first through-holes E3 are not particularly limited. One or more first through-holes E3 may be formed for each die D.

[0014] The resin film E2 is preferably composed of a flexible material, specifically a material with an elastic modulus of 2 GPa or less, more preferably 0.5 GPa or less. From the viewpoint of durability when modifying the bonding surface Da of the die D, the resin film E2 is preferably composed of, for example, polyimide or EVA (ethylene vinyl acetate copolymer). The thickness of the resin film E2 is, for example, 10 μm. In this embodiment, the resin film E2 is a single layer, but it may be a multi-layer structure. For example, the resin film E2 may have a polyolefin layer and an acrylic adhesive layer.

[0015] As shown in Figure 3, the target substrate W has a semiconductor substrate W1 such as a silicon wafer and a plurality of devices W2 formed on the semiconductor substrate W1. In this embodiment, the semiconductor substrate W1 is a silicon wafer, but it may be a compound semiconductor wafer. A glass substrate may be used instead of the semiconductor substrate W1. The target substrate W has a bonding surface Wa and a non-bonding surface Wb facing the opposite direction from the bonding surface Wa. The bonding surface Wa and non-bonding surface Wb of the target substrate W are circular, but may be square. The target substrate W has a plurality of devices W2 on the bonding surface Wa. The plurality of devices W2 are separated by a plurality of streets that are orthogonal to each other. Each device W2 includes an electronic circuit. As shown in Figure 4, a die D is electrically connected to each device W2. Then, the target substrate W is cut along the streets to separate each device W2 into individual pieces to obtain a semiconductor device. The semiconductor device includes the devices W2 and the die D.

[0016] Die D is a semiconductor substrate on which multiple devices, separate from device W2, are formed, and each device is separated into individual pieces. A glass substrate may be used instead of a semiconductor substrate. Die D has a bonding surface Da and a non-bonding surface Db facing the opposite direction from bonding surface Da. The non-bonding surface Db of die D does not need to be a surface that is not bonded to the target substrate W, and may be bonded to another die (not shown). Die D has a device on bonding surface Da. The electronic circuit of the device on die D and the electronic circuit of device W2 on the target substrate W are electrically connected. The type and number of dies D electrically connected to a single device W2 are not particularly limited. Although not shown, multiple dies D may be electrically connected to a single device W2.

[0017] As shown in Figure 1, the bonding system 1 includes a control circuit 9. The control circuit 9 is, for example, a computer. The control circuit 9 includes an arithmetic unit 91, such as a CPU (Central Processing Unit), and a storage unit 92, such as memory. The storage unit 92 stores programs that control various processes executed in the bonding system 1.

[0018] The control circuit 9 controls the operation of the junction system 1 by causing the arithmetic unit 91 to execute a program stored in the memory unit 92. A lower-level control circuit may be provided for each device constituting the junction system 1 to control the operation of that device, and a higher-level control circuit may be provided to comprehensively control multiple lower-level control circuits. The control circuit 9 may be composed of multiple lower-level control circuits and a higher-level control circuit.

[0019] The control circuit 9 includes electronic circuits such as a CPU, FPGA (Field Programmable Gate Array), or ASIC (Application Specific Integrated Circuit), and performs various control operations described in this specification by executing instruction codes stored in memory or by designing the circuit for special applications.

[0020] The bonding system 1 includes an loading / unloading station 2. The loading / unloading station 2 includes a mounting table 20. A first cassette C1, a second cassette C2, a third cassette C3, and a fourth cassette C4 are placed on the mounting table 20. The first cassette C1 contains the target substrate W before the die D is bonded. The second cassette C2 contains the target substrate W after the die D has been bonded (i.e., the laminated substrate DW composed of the target substrate W and the die D). The third cassette C3 contains the carrier E before the die D is picked up. The fourth cassette C4 contains the carrier E after the die D has been picked up.

[0021] The loading / unloading station 2 comprises a transport area 21 and a transport device 22. The transport area 21 is adjacent to the mounting table 20. The transport area 21 extends in the Y-axis direction. The transport device 22 has a transport arm. The transport arm holds and transports the target substrate W and the carrier E within the transport area 21. There may be one or more transport arms. A separate transport arm may be provided for the target substrate W and the transport arm for the carrier E. The transport device 22 has a drive unit (not shown) for moving or rotating the transport arm. The transport arm is capable of moving horizontally (both in the X-axis and Y-axis directions) and vertically (in the Z-axis direction), and rotating about the vertical axis.

[0022] The bonding system 1 includes a processing station 3. The processing station 3 includes a storage device 30. The storage device 30 is adjacent to the transport area 21. The storage device 30 is positioned on the opposite side of the loading platform 20 with respect to the transport area 21. The storage device 30 temporarily stores the target substrate W and the carrier E. The storage device 30 has a plurality of stages arranged vertically. Each stage places the target substrate W and the carrier E on it. The stage for the target substrate W and the stage for the carrier E may be provided separately.

[0023] The processing station 3 comprises a transport area 31 and a transport device 32. The transport area 31 is adjacent to the storage device 30 and extends from the storage device 30 in the positive X-axis direction. The transport device 32 has transport arms. The transport arms hold and transport the target substrate W and the carrier E within the transport area 31. There may be one or more transport arms. The transport arms for the target substrate W and the transport arms for the carrier E may be provided separately. The transport device 32 has a drive unit (not shown) for moving or rotating the transport arms. The transport arms can move horizontally (in both the X-axis and Y-axis directions) and vertically (in the Z-axis direction), and rotate about the vertical axis.

[0024] The processing station 3 includes a bonding device 33 and an inspection device 36. The bonding device 33 and the inspection device 36 are adjacent to the transport area 31 and are provided on the positive or negative Y-axis side of the transport area 31. The bonding device 33 separates the die D from the carrier E and bonds the die D to the target substrate W with the bonding surface Da of the separated die D facing the bonding surface Wa of the target substrate W. The inspection device 36 inspects the laminated substrate DW composed of the target substrate W and the die D. Details of the bonding device 33 and the inspection device 36 will be described later.

[0025] Referring again to Figure 1, the operation of the bonding system 1, that is, the bonding method, will be explained. The following processes are carried out under the control of the control circuit 9. First, the transport device 22 takes the target substrate W from the first cassette C1 and transports it to the storage device 30. Next, the transport device 32 takes the target substrate W from the storage device 30 and transports it to the bonding device 33. Also, the transport device 22 takes the carrier E from the third cassette C3 and transports it to the storage device 30. Next, the transport device 32 takes the carrier E from the storage device 30 and transports it to the bonding device 33.

[0026] Next, the bonding device 33 separates the die D from the carrier E and bonds the die D and the target substrate W with the bonding surface Da of the separated die D facing the bonding surface Wa of the target substrate W. This results in a laminated substrate DW composed of the die D and the target substrate W. After that, the transport device 32 removes the laminated substrate DW from the bonding device 33 and transports it to the inspection device 36.

[0027] Next, the inspection device 36 inspects the target substrate W and the laminated substrate DW including the dies D. The inspection device 36 detects voids present at the interface between the target substrate W and the dies D. The laminated substrate DW has multiple dies D on one side of the target substrate W. The inspection device 36 may inspect all the dies D bonded to one side of a single target substrate W, or it may inspect only some of the dies D.

[0028] Next, the transport device 32 removes the laminated substrate DW from the inspection device 36 and transports it to the storage device 30. Finally, the transport device 22 removes the laminated substrate DW from the storage device 30 and stores it in the second cassette C2. The carrier E is removed from the bonding device 33 by the transport device 32 and transported to the storage device 30, and then removed from the storage device 30 by the transport device 22 and stored in the fourth cassette C4.

[0029] The transport device 32 does not have to transport all of the laminated substrates DW to the inspection device 36; it may transport only some of the laminated substrates DW to the inspection device 36. For example, the transport device 32 may transport only the substrates that are first bonded to the die D from among the multiple (e.g., 25) target substrates W stored in the first cassette C1 to the inspection device 36. In other words, the inspection device 36 does not have to inspect all of the laminated substrates DW; it may inspect only some of the laminated substrates DW.

[0030] Referring to Figures 5 and 6, a joining device 33 according to one embodiment will be described. The joining device 33 includes a control circuit 90. The control circuit 90 controls various operations of the joining device 33. The control circuit 90 is configured in the same way as the control circuit 9 shown in Figure 1. The control circuit 90 may be a part of the control circuit 9 shown in Figure 1.

[0031] The bonding device 33 includes a housing 34. The housing 34 houses, for example, a substrate stage 40, a carrier stage 50, a pickup device 60, a pressing device 70, and a mounting device 80. The housing 34 has an input / output port 35 on its side wall. The input / output port 35 allows the target substrate W and carrier E to be loaded and unloaded.

[0032] The bonding apparatus 33 includes a substrate stage 40. The substrate stage 40 holds the target substrate W. For example, the substrate stage 40 holds the target substrate W from below with the bonding surface Wa of the target substrate W facing upwards. The substrate stage 40 is, for example, a vacuum suction chuck. The substrate stage 40 may also have an inverted structure and hold the target substrate W from above. Alternatively, the substrate stage 40 may have a vertically oriented structure and hold the target substrate W in a vertically oriented position.

[0033] The bonding apparatus 33 includes a carrier stage 50. The carrier stage 50 holds the carrier E. For example, the carrier stage 50 holds the carrier substrate E1 from below with the resin film E2 of the carrier E facing upwards. The carrier stage 50 is, for example, a vacuum suction chuck. The carrier stage 50 may also have an inverted structure, with the resin film E2 of the carrier E facing downwards and the carrier substrate E1 held from above. Alternatively, the carrier stage 50 may have a vertically oriented structure, holding the carrier E in a vertical position.

[0034] The joining device 33 includes a pickup device 60. The pickup device 60 picks up the die D from the carrier E held by the carrier stage 50. The pickup device 60 may have a structure corresponding to the structure of the carrier stage 50. The pickup device 60 may have an inverted structure. The pickup device 60 may also have a vertically standing structure.

[0035] The pickup device 60 may transport the die D. The pickup device 60 may also invert the die D during transport so that the bonding surface Da of the die D faces downward. The mounting device 80 receives the die D from the pickup device 60 and mounts the received die D onto the target substrate W held by the substrate stage 40.

[0036] The pickup device 60 has a first suction head 61. The first suction head 61 adsorbs the bonding surface Da of the die D. The first suction head 61 may be in contact with the bonding surface Da of the die D, for example, by vacuum adsorption of the die D. Alternatively, the first suction head 61 may adsorb the die D without contact so as not to contaminate the bonding surface Da of the die D.

[0037] For example, the first suction head 61 has a suction nozzle (not shown) and an injection nozzle on the surface facing the die D (e.g., the bottom surface). The suction nozzle sucks in gas, and the injection nozzle injects gas. The first suction head 61 can non-contact adsorb the die D using the gas injection pressure (positive pressure) and the gas suction pressure (negative pressure). The adsorption method is not particularly limited. Examples of non-contact adsorption methods include the Bernoulli method or the ultrasonic method.

[0038] The pickup device 60 has a first moving mechanism 69. The first moving mechanism 69 moves the die D together with the first suction head 61. The movement directions include the X-axis direction and the Z-axis direction. The movement directions may also include the Y-axis direction. The first moving mechanism 69 may also invert the die D vertically together with the first suction head 61. The bonding surface Da of the die D can be inverted vertically. The first moving mechanism 69 has, for example, a motor for each movement direction.

[0039] The control circuit 90 controls the transfer of the die D from the first suction head 61 to the second suction head 81. The roles of the first suction head 61 and the second suction head 81 can be divided. For example, while the second suction head 81 presses the die D against the target substrate W, the first suction head 61 can pick up another die D from the carrier E. Thus, the processing speed can be improved.

[0040] As shown in Figures 5 and 6, the bonding device 33 preferably includes a pressing device 70. The pressing device 70 assists in the pickup of the die D by the pickup device 60. The pressing device 70 presses the resin film E2, for example, by supplying gas to the first through-hole E3 of the carrier substrate E1, or by inserting a pin (not shown) into the first through-hole E3. The direction of pressing is the direction in which the die D is picked up (for example, the positive Z-axis direction). The resin film E2 can be deformed only in the vicinity of one of the multiple dies D, a wedge-shaped gap can be formed between the resin film E2 and the die D, and the die D can be smoothly picked up from the resin film E2.

[0041] The bonding device 33 includes a third moving mechanism 59. The third moving mechanism 59 relatively moves the pressing device 70 and the carrier stage 50 in order to change the die D to be pressed by the pressing device 70. In this embodiment, the third moving mechanism 59 moves the carrier stage 50, but the pressing device 70 may be moved instead. The third moving mechanism 59 may move both the carrier stage 50 and the pressing device 70. The third moving mechanism 59 has, for example, a motor for each moving direction.

[0042] It is preferable that the third moving mechanism 59 moves only the carrier stage 50 in the X-axis direction and the Y-axis direction among the carrier stage 50 and the pressing device 70. If the pressing device 70 does not move in the X-axis direction and the Y-axis direction, the pickup device 60 can pick up the die D at the same position every time. Therefore, the operation of the pickup device 60 can be simplified.

[0043] The bonding device 33 preferably has, for example, the following configurations (1) to (2). (1) The carrier stage 50 has a first surface 51 against which the carrier E abuts, a second surface 52 opposite to the first surface 51, and a plurality of second through holes 53 penetrating between the first surface 51 and the second surface 52. (2) The pressing device 70 has a gas supply mechanism 73 that supplies gas to the first through hole E3 via the second through hole 53.

[0044] Although not shown, the carrier stage 50 may be formed in an annular shape, and the pressing device 70 may be provided radially inside the inner circumference of the annular carrier stage 50. In this case, the pressing head 71 forms a gas supply chamber between itself and the carrier E, and the seal member 72 abuts against the carrier E (specifically, the carrier substrate E1) to seal the gas supply chamber.

[0045] However, if the pressing device 70 is arranged radially inside the inner circumference of the annular carrier stage 50, the movable ranges of the carrier stage 50 or the pressing device 70 in the X-axis direction and the Y-axis direction will be limited. The pressing device 70 can only press the die D located radially inside the inner circumference of the annular carrier stage 50. Therefore, the number of dies D that can be mounted on the carrier E will decrease.

[0046] According to the configurations of (1) to (2) above, when the third moving mechanism 59 moves the pressing device 70 and the carrier stage 50 relative to each other in the X-axis direction and the Y-axis direction, interference between the pressing device 70 and the carrier stage 50 does not occur. As a result, the number of dies D that can be mounted on the carrier E can be increased. Therefore, the frequency of carrier E replacement can be reduced, and throughput can be improved.

[0047] When the pressing device 70 presses the resin film E2, the carrier substrate E1 can be gently deformed around the pressed location. Therefore, the carrier stage 50 has a suction groove 54 that vacuum-sucks the carrier substrate E1 to the first surface 51. The carrier stage 50 can suppress the deformation of the carrier substrate E1 around the pressing device 70. The suction groove 54 is preferably provided between adjacent second through-holes 53 to locally deform the resin film E2. The suction groove 54 is formed, for example, in a lattice pattern.

[0048] The seal member 72 preferably abuts on the carrier stage 50 instead of the carrier E. The seal member 72 abuts on the second surface 52 of the carrier stage 50. Even if the seal member 72 deteriorates and particles are generated, and the second surface 52 of the carrier stage 50 becomes dirty, the first surface 51 of the carrier stage 50 is hardly soiled. Therefore, soiling of the carrier E can be suppressed.

[0049] Another effect can be obtained by the seal member 72 abutting on the carrier stage 50. When the accuracy of the parallelism between the seal member 72 and the carrier stage 50 is low, the seal member 72 can be strongly pressed against the carrier stage 50 so that gas leakage does not occur. Since the carrier stage 50 has higher rigidity than the carrier E, the carrier stage 50 will not be damaged. Therefore, the required accuracy of parallelism is low, and the parallelism adjustment work is simple.

[0050] The gas supply mechanism 73 supplies gas to the first through-hole E3 via the second through-hole 53 by supplying gas to the pressing head 71. The gas supply mechanism 73 has a supply line. The supply line forms a gas flow path. The gas supply mechanism 73 has, for example, an on-off valve and a pressure controller in the middle of the supply line. The on-off valve opens and closes the gas flow path under the control of the control circuit 90. The pressure controller controls the gas pressure under the control of the control circuit 90. The gas supply mechanism 73 may also have a leak valve in the middle of the supply line. The leak valve discharges gas.

[0051] Preferably, at the boundary between the carrier stage 50 and the carrier E, the first through-hole E3 and the second through-hole 53 are connected in a one-to-one ratio. The number of second through-holes 53 may be equal to or greater than the number of first through-holes E3, and may be greater than the number of first through-holes E3. It is sufficient that the second through-holes 53 are located where the first through-holes E3 are, and the first through-holes E3 and the second through-holes 53 are not connected in a one-to-one ratio.

[0052] At the boundary between the carrier stage 50 and the carrier E, it is preferable that the opening of the second through-hole 53 is smaller than the opening of the first through-hole E3 and positioned inside the opening of the first through-hole E3. In this case, since the gas pressure acts almost no way on the carrier substrate E1, the bending of the carrier substrate E1 can be suppressed. Only the resin film E2 can be deformed.

[0053] The pressing device 70 may use a plurality of second through holes 53 and a plurality of first through holes E3 to press a die D. For example, the pressing device 70 supplies gas to the plurality of first through holes E3 via the plurality of second through holes 53 to press a die D. A die D can be pressed at multiple points. It is also possible to divide the multiple points into multiple groups and press them sequentially.

[0054] The pressing device 70 includes a drive shaft 74. The drive shaft 74 moves the pressing head 71 in a first direction (e.g., positive Z-axis direction) and a second direction opposite to the first direction (e.g., negative Z-axis direction). When moving the pressing device 70 and the carrier stage 50 relative to each other in the X-axis and Y-axis directions to change the die D pressed by the pressing device 70, the sealing member 72 can be separated from the carrier stage 50. The drive shaft 74 includes an actuator such as a motor or cylinder.

[0055] The bonding apparatus 33 includes a mounting apparatus 80. The mounting apparatus 80 mounts the die D onto the target substrate W held by the substrate stage 40. The mounting apparatus 80 may have a structure corresponding to the structure of the substrate stage 40. The mounting apparatus 80 may have an inverted structure. The mounting apparatus 80 may also have a vertically oriented structure.

[0056] The mounting device 80 has a second suction head 81. The second suction head 81 adsorbs the die D. The second suction head 81 adsorbs the non-joint surface Db of the die D. Since it is not a problem if the non-joint surface Db is dirty, the second suction head 81 may come into contact with the die D. This can improve the suction force and suppress misalignment. The second suction head 81 adsorbs the die D, for example, by vacuum.

[0057] The mounting device 80 has a second moving mechanism 89. The second moving mechanism 89 moves the die D together with the second suction head 81 and presses the die D against the target substrate W. The second moving mechanism 89 joins the die D to the target substrate W by moving the second suction head 81 in the Z-axis direction. To improve the accuracy of the joining position, the second moving mechanism 89 may move the second suction head 81 in the X-axis and Y-axis directions, or rotate the second suction head 81 around the vertical axis. The amount of movement or rotation required to improve the accuracy of the joining position is small, and when viewed from above, the second suction head 81 hardly moves at all. The second moving mechanism 89 has, for example, a motor for each direction of movement.

[0058] The bonding device 33 may include a fourth moving mechanism 49. The fourth moving mechanism 49 moves the substrate stage 40. The fourth moving mechanism 49 moves the substrate stage 40 in the X-axis and Y-axis directions to change the bonding position of the die D on the target substrate W. If the mounting device 80 does not move in the X-axis and Y-axis directions, the operation of the mounting device 80 can be simplified. The fourth moving mechanism 49 may also move the substrate stage 40 in the Z-axis direction. The fourth moving mechanism 49 may have, for example, a motor for each direction of movement.

[0059] Next, an inspection device 36 according to one embodiment will be described with reference to Figure 7. In this embodiment, the inspection device 36 is incorporated inside the joining system 1, but it may also be installed outside the joining system 1. The inspection device 36 includes a control circuit 190. The control circuit 190 controls various operations of the inspection device 36. The control circuit 190 is configured similarly to the control circuit 9 shown in Figure 1. The control circuit 190 may also be a part of the control circuit 9 shown in Figure 1.

[0060] The inspection device 36 includes, for example, a stage 110, a light source 120, an optical system 130, a photodetector 140, and a drive device 150. Each of these components will be described below.

[0061] The stage 110 holds the laminated substrate DW. For example, the stage 110 holds the laminated substrate DW horizontally from below with the non-bonding surface Db of the die D facing upwards. The stage 110 may also hold the laminated substrate DW by suction. The stage 110 may be a vacuum chuck or an electrostatic chuck.

[0062] The light source 120 emits optical beams (LB). The light source 120 is, for example, a laser. The laser is preferably a pulsed laser. After emitting from the light source 120, the optical beams are guided to the probe 122 by, for example, an optical fiber 121, and further guided to the optical system 130.

[0063] The optical system 130 irradiates a circular irradiation area 101 (see Figures 9 and 10) on the surface of the laminated substrate DW with light LB emitted from the light source 120. Preferably, the optical system 130 irradiates the surface of the die D opposite to the target substrate W, that is, the non-bonding surface Db of the die D with light LB. Alternatively, the optical system 130 may irradiate the non-bonding surface Wb of the target substrate W, that is, the surface of the target substrate W opposite to the die D.

[0064] The optical system 130 preferably includes a galvanometer mirror 131. The drive device 150 preferably includes a motor 151. The motor 151 rotates the galvanometer mirror 131. The illumination area 101 can be moved without moving the stage 110. Therefore, the illumination area 101 can be moved at high speed. Although not shown, it is preferable to provide two or more sets of galvanometer mirrors 131 and motors 151.

[0065] The optical system 130 preferably includes an fθ lens 132. The fθ lens 132 forms a focal plane perpendicular to the Z-axis direction. While the galvanometer mirror 131 moves across the illumination region 101, the fθ lens 132 maintains the shape and dimensions of the illumination region 101. In this embodiment, the height of the illumination region 101 coincides with the height of the focal plane, but it does not have to coincide with the height of the focal plane, and it may be higher or lower than the height of the focal plane.

[0066] The photodetector 140 detects reflected light from the illumination area 101. The reflected light from the illumination area 101 is guided to the probe 122 by the optical system 130, and then to the photodetector 140 by the optical fiber 141. The control circuit 190 analyzes the waveform of the light detected by the photodetector 140 to detect voids V present at the interface between the die D and the target substrate W, as shown in Figure 8.

[0067] For example, the control circuit 190 detects a void V by analyzing the waveform of the interference wave between the light reflected from the non-bonding surface Db of the die D, the light reflected from the bonding surface Da of the die D, and the light reflected from the bonding surface Wa of the target substrate W. If a void V is present, the phase of the light reflected from the bonding surface Da of the die D and the phase of the light reflected from the bonding surface Wa of the target substrate W will be shifted.

[0068] As described above, the optical system 130 may also irradiate the non-bonding surface Wb of the target substrate W with light LB. In this case, the control circuit 190 detects the void V by analyzing the waveform of the interference wave between the light reflected from the non-bonding surface Wb of the target substrate W, the light reflected from the bonding surface Da of the die D, and the light reflected from the bonding surface Wa of the target substrate W.

[0069] The drive unit 150 changes the position of the irradiation area 101. The drive unit 150 changes the position of the irradiation area 101, preferably by rotating the galvanometer mirror 131, as described above. The drive unit 150 can also change the position of the irradiation area 101 by moving the stage 110.

[0070] Next, with reference to Figure 9, an example of a square grid 200 set on the non-bonding surface Db of the die D during full-surface inspection will be described. The square grid 200 includes multiple square regions 201 of the same size. Note that the square grid 200 can also be set on the non-bonding surface Wb of the target substrate W instead of the non-bonding surface Db of the die D.

[0071] In Figure 9, r0 is the radius of the irradiation area 101, and r1 is the radius of the circumscribed circle of the square area 201. The irradiation area 101 is set so that its center coincides with the center of each square area 201. When performing a full surface inspection of the non-bonded surface Db of the die D, r1 should be less than or equal to r0 to prevent inspection omissions. However, to shorten the inspection time, it is preferable that r1 is equal to r0.

[0072] In Figure 9, S1 is the area of ​​the square region 201, and d1 is the distance between the centers of two adjacent square regions 201. The number of irradiation regions 101 (number of irradiation points) is proportional to the reciprocal of S1. Also, the time it takes to move the optical axis between two adjacent irradiation points (hereinafter also simply called "movement time") is proportional to d1.

[0073] Next, with reference to Figure 10, an example of a regular hexagonal grid 300 set on the non-bonding surface Db of the die D during full-surface inspection will be described. The regular hexagonal grid 300 includes multiple regular hexagonal regions 301 of the same size. Note that the regular hexagonal grid 300 can also be set on the non-bonding surface Wb of the target substrate W instead of the non-bonding surface Db of the die D.

[0074] In Figure 10, r0 is the radius of the irradiation area 101, and r2 is the radius of the circumscribed circle of the regular hexagonal area 301. The irradiation area 101 is set so that its center coincides with the center of each regular hexagonal area 301. When performing a full surface inspection of the non-jointed surface Db of the die D, r0 and r2 are equal, and the irradiation area 101 is equal to the circumscribed circle of the regular hexagonal area 301.

[0075] In Figure 10, S2 is the area of ​​the regular hexagonal region 301, and d2 is the distance between the centers of two adjacent regular hexagonal regions 301. The number of illumination regions 101 (number of illumination points) is proportional to the reciprocal of S2. Also, the time it takes to move the optical axis between two adjacent illumination points (hereinafter simply referred to as "movement time") is proportional to d2.

[0076] As shown in Figure 10, if a regular hexagonal grid 300 is used instead of a square grid 200, the number of irradiation points decreases and the travel time increases.

[0077] Next, with reference to Figure 11, an example of the relationship between inspection times T1 and T2 during full surface inspection will be explained. Inspection time T1 is the time required to inspect the entire surface of the non-jointed surface Db of die D when the square grid 200 shown in Figure 9 is set as the non-jointed surface Db of die D. Inspection time T2 is the time required to inspect the entire surface of the non-jointed surface Db of die D when the regular hexagonal grid 300 shown in Figure 10 is set as the non-jointed surface Db of die D.

[0078] The examination times T1 and T2 mainly depend on the movement time, waiting time, exposure time, and number of irradiation points. The movement time is the time it takes to move the optical axis between two adjacent irradiation points, as described above. The waiting time is the time it takes to wait for the optical axis to stabilize after temporarily pausing the movement of the optical axis. The exposure time is the time it takes to irradiate the irradiation area 101 with light LB.

[0079] When illuminating with light LB while continuing to move the optical axis without pausing the movement, the inspection times T1 and T2 are governed by (movement time × number of irradiation points). Since (movement time × number of irradiation points) is dominant, (T2 / T1) is approximately 0.94, so it can be seen that the inspection time can be shortened by about 6% by adopting a regular hexagonal grid 300 instead of a square grid 200.

[0080] On the other hand, when the optical axis movement is paused and light LB is irradiated, the inspection times T1 and T2 are determined by both (movement time × number of irradiation points) and (waiting time + exposure time) × number of irradiation points. When (waiting time + exposure time) × number of irradiation points is dominant, (T2 / T1) is approximately 0.77, so it can be seen that by adopting a regular hexagonal lattice 300 instead of a square lattice 200, the inspection time can be reduced by about 23%.

[0081] As is clear from Figure 12, the inspection time can be shortened by using a regular hexagonal grid 300 instead of a square grid 200. As shown in Figure 12, by using a regular hexagonal grid 300 instead of a square grid 200, the overlap of the irradiation area 101 is reduced, and the area to be inspected overlapping is decreased. Therefore, the inspection time can be shortened by using a regular hexagonal grid 300 instead of a square grid 200.

[0082] Next, with reference to Figure 13, examples of a square grid 200 and a regular hexagonal grid 300 set on the non-jointed surface Db of the die D during thinning inspection will be described. The following will mainly explain the differences from full-surface inspection. When thinning inspection is performed, r1 and r2 are greater than r0. In Figure 13, R1 is the maximum radius of void V that will be missed in the case of the square grid 200, and R2 is the maximum radius of void V that will be missed in the case of the regular hexagonal grid 300.

[0083] Next, referring to Figure 14, we will explain an example of the relationship between the maximum radii R1 and R2 of void V that may be missed during thinning inspection. Here, we will explain the relationship between R1 and R2 when the inspection times T1 and T2 are the same. Note that if R1 and R2 are sufficiently larger than r0, (R2 / R1) is equal to (r2 / r1).

[0084] When (waiting time + exposure time) × irradiation point is dominant, the condition for inspection times T1 and T2 to be equal is, as shown in Figure 14, that the reciprocal of S1 and the reciprocal of S2 are equal. When (waiting time + exposure time) × irradiation point is dominant, (r2 / r1) and thus (R2 / R1) are approximately 0.88, so it can be seen that by adopting a regular hexagonal grid 300 instead of a square grid 200, the maximum radius of missed voids V can be reduced by about 12%, and inspection accuracy can be improved.

[0085] On the other hand, when (travel time × number of irradiation points) is dominant, the condition for inspection times T1 and T2 to be equal is that (the product of d1 and the reciprocal of S1) and (the product of d2 and the reciprocal of S2) are equal, as shown in Figure 14. When (travel time × number of irradiation points) is dominant, (r2 / r1) and thus (R2 / R1) is approximately 0.94, so it can be seen that by adopting a regular hexagonal grid 300 instead of a square grid 200, the maximum radius of missed voids V can be reduced by about 6%, and inspection accuracy can be improved.

[0086] The embodiments of the inspection apparatus, joining system, and inspection method relating to this disclosure have been described above, but this disclosure is not limited to the embodiments described above. Various changes, modifications, substitutions, additions, deletions, and combinations are possible within the scope of the claims. These also naturally fall within the technical scope of this disclosure.

[0087] This application claims priority based on Japanese Patent Application No. 2024-199887, filed with the Japan Patent Office on November 15, 2024, and the entire contents of Japanese Patent Application No. 2024-199887 are incorporated herein by reference.

[0088] 36 Inspection device 101 Irradiation area 120 Light source 130 Optical system 140 Photodetector 150 Drive unit 190 Control circuit 300 Regular hexagonal grid 301 Regular hexagonal region D Die W Target substrate DW Multilayer substrate

Claims

1. An inspection apparatus for inspecting a laminated substrate including a target substrate and a die bonded to the target substrate, comprising: a light source; an optical system for irradiating a circular irradiation area on the surface of the laminated substrate with light emitted from the light source; a photodetector for detecting reflected light from the irradiation area; a drive device for changing the position of the irradiation area; and a control circuit for controlling the position of the irradiation area, wherein the control circuit performs the following: setting a regular hexagonal grid on the surface of the laminated substrate, including a plurality of regular hexagonal regions of the same size; and setting the center of the irradiation area at the center of each of the regular hexagonal regions.

2. The inspection apparatus according to claim 1, wherein the control circuit sets the size of the regular hexagonal region such that the circumscribed circle of the regular hexagonal region has the same radius as the irradiation region.

3. The inspection apparatus according to claim 1, wherein the control circuit sets the size of the regular hexagonal region such that the circumscribed circle of the regular hexagonal region has a larger radius than the irradiation region.

4. The inspection apparatus according to claim 1, wherein the optical system irradiates light onto the surface of the die opposite to the target substrate.

5. The inspection apparatus according to claim 4, wherein the laminated substrate includes a plurality of dies on one side of the target substrate, and the control circuit controls the drive device to irradiate light only to some of the plurality of dies.

6. The inspection apparatus according to claim 1, wherein the optical system includes a galvanometer mirror, and the drive device includes a motor for rotating the galvanometer mirror.

7. A bonding system comprising: an inspection apparatus according to any one of claims 1 to 6; a bonding apparatus for bonding the target substrate and the die; a transport apparatus for transporting the laminated substrate from the bonding apparatus to the inspection apparatus; and a control circuit for controlling the transport apparatus, wherein the control circuit controls the transport apparatus so that only a portion of the laminated substrates from a plurality of laminated substrates are transported from the bonding apparatus to the inspection apparatus.

8. An inspection method comprising detecting voids present at the interface between the target substrate and the die using the inspection apparatus described in any one of claims 1 to 6.

Images