JP2007194403A - Apparatus and method for manufacturing electronic device, and for inspecting underfill material filling state - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an apparatus for manufacturing a highly reliable electronic device. <P>SOLUTION: In the apparatus manufacturing the electronic device, a first component 10 and a second component 20 are mutually joined with a gap and the gap is filled with an underfill material 14. The apparatus comprises a discharge means 22 for discharging the underfill material to fill the gap with the underfill material; a detecting means 38 for detecting fillets 16, 17, 18, 19 which are formed at the side ends of the first component with the use of the underfill material; and a control means 44 for comparing detection information of the fillets with reference information of the normal fillets, and allowing the discharge means to additionally perform the discharge of the underfill material when the fillet width of the former information is smaller than that of the latter information. <P>COPYRIGHT: (C)2007,JPO&amp;INPIT

Description

  The present invention relates to a joined body in which a first component and a second component are joined with a gap (for example, a joined body between semiconductor chips or a joined body between a semiconductor chip and a mounting substrate via bumps). The present invention relates to an electronic device manufacturing apparatus and an electronic device manufacturing method, and an underfill material filling state inspection apparatus and an underfill material filling state inspection method for filling the gap with an underfill material.

  2. Description of the Related Art Conventionally, along with downsizing and integration of electronic devices such as mobile devices such as mobile phones, flip chip connection for connecting a semiconductor chip to a mounting substrate through bumps has been widely adopted. In this flip-chip connection, the stress due to the thermal expansion coefficient between the mounting substrate and the semiconductor chip is concentrated on the small bumps, leading to poor electrical connection due to cracks, etc., and therefore generally in the gap between the semiconductor chip and the mounting substrate. The reliability of the electrical connection between the semiconductor chip and the mounting substrate has been improved by infiltrating, filling and curing the sealing material (underfill material) using the capillary phenomenon. There are many reports on the filling method of this underfill material. Hereinafter, typical reports will be described.

  “The underfill filling method and the printed wiring board structure” have the following description in Patent Document 1 described later.

  In the prior art, underfill is injected from the outer peripheral portion of the bare chip LSI by a dispenser into a joint portion between the bare chip LSI flip-chip connected to the printed wiring board and the printed wiring board. Thereafter, the underfill is left to stand until it is filled between the printed wiring board and the bare chip LSI.

  In this prior art, since the underfill is filled in a natural state, air remains in the gap between the printed wiring board and the bare chip LSI, and sufficient heat buffering and connection strength cannot be obtained. It is difficult to adjust the viscosity of the underfill because the residual amount of air varies depending on the ambient temperature, atmospheric pressure, etc. In addition, since the underfill is allowed to penetrate between the printed wiring board and the bare chip LSI in a state of being left in the natural state, there is a problem that the viscosity of the underfill is limited.

  Further, in an underfill filling method for filling an underfill in a gap portion between a bare chip LSI flip-chip connected to a printed wiring board and the printed wiring board, an air vent hole is provided in a part of the bare chip LSI mounting surface of the printed wiring board, When the underfill is filled, the air remaining in the gap between the printed wiring board and the bare chip LSI is excluded from the hole, and thus, when the underfill is filled in the gap between the bare chip LSI and the printed wiring board. Since air remaining between the printed wiring board and the underfill is eliminated, the bonding strength is high, sufficient heat buffering action is obtained, and underfill filling time is shortened and viscosity restriction can be relaxed. An underfill filling method for buffering the difference in thermal expansion can be provided.

  Japanese Patent Application Laid-Open No. 2004-228561 regarding “underfill bond coating method” has the following description.

  The chip mounted on the substrate is observed with a camera to detect the positions of a plurality of corners of the chip, the underfill bond application start position is calculated by the calculation unit according to the detection result, and the calculated result is followed. Even if there is a variation in the chip size or the position of the chip, the underfill bond is applied to each chip by applying the bond by moving the bond coating head relative to the chip and applying the bond. It can be applied correctly at a position along the side to enter the gap between the chip and the substrate.

  The bond application head is connected to an air generator through a tube. By opening and closing a valve (controlled by a valve drive unit) and applying air pressure to the bond application head, an underfill bond is applied from the nozzle. Discharge.

  The following description is given in Patent Document 3 below regarding “resin sealing method of semiconductor device”.

  In a resin sealing method of a semiconductor device in which a semiconductor chip is bonded to a wiring board via bumps and an underfill molding is performed between the semiconductor chip and the wiring board, in order to seal a gap between the wiring board and the semiconductor chip. A resin supply step of supplying a predetermined amount of the sealing resin from the potting nozzle, and a first heating step of heating the sealing resin to a predetermined temperature before the sealing resin flowing down from the potting nozzle reaches the gap, The semiconductor device in which the sealing resin is filled in the gap includes a second heating step in which the sealing resin is sequentially cured toward the peripheral portion by heating from the central portion of the semiconductor chip.

  In addition, it is preferable to flow the sealing resin from above the semiconductor chip into the gap at an inclined position where the wiring board is inclined at a predetermined angle, and when the wiring board reaches below the semiconductor chip, the orientation of the wiring board is changed to a horizontal position. When the monitoring camera confirms that the sealing resin has reached the lower side of the semiconductor chip, the direction of the wiring board may be changed from the tilt position to the horizontal position. Moreover, it is preferable that a 1st heating process heats the sealing resin which flowed down from the potting nozzle within the range below room temperature and below the hardening temperature of sealing resin.

  FIG. 8 is an explanatory diagram illustrating a state in which the semiconductor device is filled with the sealing resin by the potting method, which is FIG. 1 described in Patent Document 3.

  In FIG. 8, a predetermined amount of sealing resin 85 heated to about room temperature is continuously flowed down through a potting nozzle 86 by a dispenser (not shown), and the semiconductor is held at a predetermined angle θ by a jig (not shown). Supply to device 81. While the wiring board 82 is held at an inclined position, the sealing resin 85 is caused to flow down from the potting nozzle 86 so as to fill the gap where the solder balls 83 are disposed between the wiring board 82 and the semiconductor chip 84.

  Next, the sealing resin 85 is heated to a predetermined temperature by the first heating device 87 before the sealing resin 85 flowed down from the potting nozzle 86 reaches the gap between the wiring substrate 82 and the semiconductor chip 84. Since the sealing resin 85 heated to a predetermined temperature by the first heating device 87 is not subjected to heat exchange other than room temperature, it is difficult for the curing to proceed and the temperature is slightly lowered while flowing down, but the flowability is good. It can be filled in the underfilled place while keeping the low viscosity for a long time.

  The heating temperature by the first heating device 87 is preferably at least room temperature and heated to the curing temperature of the sealing resin 85 or less, thereby reducing the viscosity of the sealing resin 85 to improve flowability and moisture. Since it is evaporated as water vapor or the sealing resin 85 is squeezed into a fine thread as it flows down, it is possible to deaerate bubbles of various sizes in the sealing resin 85, which has not been obtained in the past. It becomes possible to supply a clean sealing resin 85 having a high density.

  Next, the sealing resin 85 is poured into the gap between the wiring substrate 82 and the semiconductor chip 84 from the diagonally upper left portion 84a of the semiconductor chip 84 at an inclined position where the wiring substrate 82 is inclined at a predetermined angle θ, and the semiconductor chip 84 is tilted to the right. When the lower part 84b is reached, the direction of the wiring board 82 is changed to the horizontal position. In this way, by returning to θ = 0 just before the filling of the sealing resin 85 is completed, the sealing resin 85 in the middle of flowing out from the right lower portion 84b of the semiconductor chip 84 can be returned by the capillary phenomenon, and the right lower portion. The variation in the degree of protrusion of the sealing resin 85 of 84b can be made uniform. The orientation of the wiring board 82 may be changed by providing the monitoring camera 88 and confirming that the sealing resin 85 has reached the diagonally lower right portion 84b of the semiconductor chip 84. As a result, the amount of resin protruding from the semiconductor chip 84 can be made more uniform and contribute to improving the molding quality.

  Japanese Patent Application Laid-Open Publication No. 2003-228561 regarding “Semiconductor Device and Method for Manufacturing the Same” includes the following description.

  In a semiconductor device in which a semiconductor chip is flip-chip mounted on a substrate via an underfill resin, the length (L) of a fillet formed on the side end surface of the semiconductor chip by injection of the underfill resin is from the surface of the substrate. It is characterized by being longer than the distance (H) to the back surface of the semiconductor chip, thereby preventing the insulating substrate from warping when the underfill resin between the semiconductor chip and the insulating substrate is cured. The peeling off of the fill resin can be prevented.

  In addition, an appropriate amount of underfill resin is injected into the solder connection portion where the solder bump is interposed between the semiconductor chip and the resin substrate. By injecting the underfill resin, a fillet having a substantially right-angled triangle shape, which is a protruding portion of the underfill resin, is formed on the side end surface of the semiconductor chip. The injection amount of the underfill resin is adjusted so that the fillet length L is longer than the distance H from the surface of the resin substrate to the back surface of the semiconductor chip.

  Japanese Patent Application Laid-Open Publication No. 2003-259259, which relates to “a method and apparatus for underfilling a semiconductor device”, includes the following description.

  In order to increase the reliability of the electrical connection of the flip chip package assembly, it is common to fill the gap between the die (chip) and the package carrier (substrate) with a sealing material. Underfill with encapsulant material increases the service life of the package and reduces the reliability of the electrical connection during thermal cycling or by reducing the stress experienced by the electrical connection when the die and package carrier have a significant temperature difference Increase. The sealing material also seals the electrical connection from exposure to the surrounding environment by hermetically sealing the gap and imparts mechanical strength to the package assembly to withstand mechanical shock and bending.

  Various conventional underfill methods are used to introduce a sealing material into the gap between the die and the substrate. One conventional method relies on surface tension wetting or capillary action to induce migration of low viscosity sealing materials with high wettability from the side edges into the gap. According to this method, the sealing material is dispensed (quantitatively dispensed) as an elongated straight, L-shaped or U-shaped bead adjacent to one, two, or three successive side edges of the die, respectively. Capillary forces act to draw the sealing material into the gap. Typically, by preheating the substrate to a uniform steady temperature of about 40 ° C. to about 90 ° C. around the die before discharging the sealing material onto the substrate, the viscosity of the sealing material decreases and the flow rate increases. Thereafter, when the electrical connection is completely sealed, the underfill material is cured.

  FIG. 9 (A) is FIG. 1 described in Patent Document 5, and is a schematic diagram illustrating an underfill process in which the substrate is heated to a uniform temperature by using a capillary action according to a conventional technique. .

  FIG. 9A shows a time sequence of a typical underfill process that relies on capillary action, where an isochronal contour line 111 moves into a gap that separates the die 112 from the substrate 114. It represents the advance of the leading edge or wavefront of the stop material 110. Initially, the sealing material 110 is ejected onto the substrate 114 as L-shaped beads adjacent to the continuous side edges of the die 112 and drawn into the gap by capillary force. As time progresses, the wavefront of the encapsulant 110 advances substantially diagonally through the gap, as indicated by arrow 116. As shown by the smaller separation between adjacent pairs of contour lines 111, the drag decreases with time, and as the underfill process is nearing completion, the wavefront advance rate of the encapsulant material is dramatic. Decline.

  For larger die sizes and smaller gap dimensions, the time required to underfill using conventional capillary underfill methods due to longer liquid sealant flow paths and shear rates Becomes longer. As a result, throughput is reduced and the cost efficiency of the underfill process is reduced. One way to increase the speed of the sealing material is to perform forced underfill to increase the filling speed and filling quality, for example by relying on the use of vacuum. Underfill using vacuum draws the sealing material into the gap using the pressure differential generated across the beads of sealing material. Regardless of the underfill method, it is important to prevent voids from being formed in the sealing material. The voids can result in corrosion and undesirable thermal stresses, thereby reducing the performance of the package assembly or adversely affecting reliability. Therefore, it provides a method for underfilling the gap formed between the die and the package carrier, preventing the generation of voids between the die and the package carrier, and shortening the time required to perform the underfill process. It would be desirable to do.

  According to the principle of the invention of Patent Document 5, the gap between the die and the substrate is underfilled using a dispenser that operates to dispense (quantitatively supply) sealing material adjacent to at least one side edge of the die. An apparatus is provided. The apparatus is operative to transfer thermal energy to first and second regions of one of the die and substrate, and the first and second regions are heated to first and second temperatures, respectively. A heat source is provided. The first temperature is different from the second temperature and causes heat to be transferred non-uniformly to the encapsulant that moves within the gap between the die and the substrate.

  According to the principles of the invention of US Pat. No. 6,099,059, a method for underfilling the gap between the die and the substrate is provided. The method includes heating at least one of the die and the substrate by either conduction, convection, or radiation, thereby creating a temperature gradient in the heated one of the die and the substrate. Including doing. A sealing material is dispensed adjacent to at least one side edge of the die and then moves into the gap to seal the plurality of electrical interconnects. Thermal energy is non-uniformly transferred to the encapsulating material that moves from the heated one of the die and the substrate in a pattern determined by a temperature gradient, thereby selectively changing the flow rate of the encapsulating material that moves in the gap. In one aspect of the invention of U.S. Patent No. 6,073,075, the individual temperatures of the temperature gradient may be dynamically changed as the sealing material flows into the gap.

  The invention of Patent Document 5 is particularly advantageous for applications in which the gap between the die and the substrate is small, and applications in which a relatively large die with a large underfill space is used. In these situations, selective heating of either the die and / or substrate according to the principles of the invention of Patent Document 5 or non-uniform heating moves the underfill material into the gap and reduces the occurrence of void formation. To enhance the capillary action or forced (eg, utilizing vacuum) capillary action that would normally be relied upon to completely seal the electrical connection. This enhancement selectively reduces the viscosity of the material in different regions of temperature and changes the flow rate and direction of the material as the material moves through the gap, thereby making it more uniform before the sealing material moving forward in the gap. Give an edge or wavefront.

  The invention of Patent Document 5 improves the durability and reliability of electronic components that require an underfill sealing material in the gap between dies mounted on a substrate. The patent document 5 also reduces the time required to effectively and reliably underfill the sealing material in the gap between the die and the substrate. The invention of U.S. Pat. No. 6,053,097 increases the overall throughput of the underfill process while simultaneously addressing the flexibility required, as well as a number of different chip sizes used in the industry, reduced gap dimensions, and various Corresponds to the type of sealing material.

  FIGS. 9B and 9C show a semiconductor device package 130 including a die 132 mounted on a package carrier or substrate 134 in a flip chip mounting structure.

  FIG. 9B is a schematic perspective view of the die assembly (shown by a broken line) and the substrate package assembly in the underfill process, which is shown in FIG.

  FIG. 9C is a schematic side view of the package assembly of the die and the substrate after the underfill process, which is shown in FIG.

  As shown in FIGS. 9B and 9C, the die 132 is an area of the solder bump 136 below the die 132 that is aligned or aligned with the corresponding area array of solder pads 138 on the substrate 134. Electrically and mechanically connected to the substrate 134 via the array. In this mounting structure, a gap 140 is formed between the contact surface 141 of the die 132 and the upper surface 142 of the substrate 134.

  According to the principle of Patent Document 5, the gap 140 is filled with a sealing material 144 such as liquid epoxy. As shown in FIG. 9B, the sealing material 144 is delivered from the underfill dispenser 135 to the substrate surface, adjacent to the gap 140 and to two successive sides of the die 132. Supplied as The principle of the invention of Patent Document 5 is that a single-line sealing material 144 arranged along one side edge of the die 132 and a U-shape of the sealing material 144 arranged along three side edges of the die 132. It can be applied to any bead shape, including mold beads or other ejection patterns. The amount of encapsulant 144 in the beads 145 depends on the size of the die 132 and the desired fillet volume and under-die volume as determined by the tolerance of the solder joint height that occurs between the bump 136 and the pad 138. ) Depends on.

  The sealing material 144 flows or moves within the gap 140 under capillary action or with forced assistance, generally as indicated by the arrow 146. When the flow stops (FIG. 9C), the sealing material 144 completely seals all electrical interconnections provided by solder joints and fillets along the side edges of the die 132. 147 is formed. The sealing material 144 is cured after completion of the underfill process.

  In accordance with the principles of the invention of US Pat. No. 6,057,059, a temperature gradient is established in the die 132, the substrate 134, or both the die 132 and the substrate 134, and the gap 140 between the lower side 141 of the die 132 and the upper surface 142 of the substrate 134. Heat is transferred to the sealing material that moves in. Heat from the die 132 and / or the substrate 134 is then transferred to the sealing material that advances or moves through the gap 140. The transferred heat increases the temperature of the sealing material 144 in the gap 140 to decrease the temperature dependent viscosity, thereby increasing the uniformity of the leading edge or wavefront of the advanced sealing material 144. Non-uniform heat transfer changes the flow of the sealing material 144 by changing the flow rate and direction of movement within the gap 140.

Japanese Patent Laid-Open No. 10-261661 (paragraphs 0003 to 0012, paragraphs 0015 to 0017, FIGS. 7, 8, and 1) JP-A-11-112482 (paragraphs 0010 to 0011, paragraphs 0013 to 0014, FIGS. 1 and 2) JP 2000-82715 A (paragraphs 0007 to 0008, paragraphs 0011 to 0014, FIG. 1) JP 2000-323624 A (paragraphs 0014 to 0015, paragraphs 0020 to 0021, FIG. 1) JP-T-2005-531130 (paragraphs 0004 to 0008, paragraphs 0011 to 0014, paragraphs 0017 to 0021, FIGS. 1, 2, and 3)

  In order to improve the reliability of a semiconductor device configured by bonding between semiconductor chips via metal bumps or bonding between a semiconductor chip and a mounting substrate, improving the reliability of bonding between the two parties, Needless to say, it is essential for improving the yield of semiconductor devices. One cause of reducing the reliability of joining is the influence of an underfill material that fills the gap between two parts to be joined.

  In the manufacture of an FCBGA type semiconductor device in which a semiconductor chip is mounted on an FCBGA (Flip Chip Ball Grid Array) substrate by flip chip connection, solder bumps provided on the semiconductor chip and solder pads provided on the FCBGA substrate are soldered together. Then, a liquid underfill material is injected (filled) into the gap between the semiconductor chip and the substrate to cure it. As a result, the semiconductor chip and the substrate are mechanically coupled to relieve stress and limit thermal load and mechanical load, and protect the joint between the semiconductor chip and the substrate from external stress such as humidity, and solder The bump can be compressed to suppress solder creep.

  With the increase in the speed, area, and thickness of semiconductor chips, the number of external terminals to be joined is increased, the distance between the terminals is shortened, and the underfill filling gap is shortened. In the FBGA type semiconductor device using such a semiconductor chip, in the underfill filling process following the flip chip connection between the semiconductor chips or between the semiconductor chip and the mounting substrate, the semiconductor chip is reduced in area and thickness. As the number of external terminals increases, the distance between terminals shortens, and the underfill filling gap shortens, the amount of filling in the gap of the underfill material becomes very small. There is a problem that bubbles are easily generated.

  In order to fill the underfill material into the underfill filling gap, even if a predetermined amount of the underfill material is discharged from the discharge tip of the discharger (dispenser), the filling amount is insufficient. The following are typical reasons for this. (1) The discharge amount decreases due to variation in the discharge amount by the discharge device (the accuracy of control of the discharge amount by the discharge device is about 10%). (2) There are gap variations in the filling gap among a large number of filling objects (lots), and insufficient filling occurs when the filling gap is large. (3) The discharge amount is reduced due to clogging of the flow path of the underfill material such as the discharge tip (needle) of the discharger. (4) When the underfill material is discharged from the discharger, the underfill material easily gets on the upper surface of the semiconductor chip as the semiconductor chip becomes thinner. The generation of bubbles due to the insufficient filling amount in the filling gap hinders maintaining the reliability of the FCBGA type semiconductor device, which leads to a big problem.

  In a conventional FCBGA type semiconductor device, after underfill material is injected into a gap and cured, it is possible to check and inspect the underfill material for filling failure and voids. As an inspection method, for example, (1) Appearance inspection method for confirming presence or absence of resin by inspection of fillet part, (2) Inspection method using ultrasonic flaw detector (SAT (Scanning Acoustic Tomography)), (3) Cutting semiconductor device and cross section There are cross-sectional inspection methods for polishing and observation.

  However, the above inspection method has the following problems. In the inspection using an ultrasonic flaw detector, the semiconductor device is immersed in water for inspection. In the case of a semiconductor device using a chip-on-chip connection, the wire bond pad is contaminated, resulting in inability to connect the wire or poor connection. It cannot be used as an inspection method because it may adversely affect sex. Since the cross-sectional inspection method is a destructive inspection, it is a sampling inspection. The appearance inspection method is performed after the underfill material has been cured, so even if an underfill material is added and replenished after a lack of filling amount is found, the part that has been cured first and later are added. The refilled underfill material may not be compatible and bubbles may be generated.

  In any of the inspection methods described above, since it takes time until the inspection result is revealed, feedback to the semiconductor device manufacturing process cannot be performed in a timely manner, and a serious defect may flow out to the subsequent process. There was a big problem that there was.

  Patent Document 1 discloses a configuration in which an air vent hole is provided in a part of a bare chip LSI mounting surface of a printed wiring board, and air remaining in a gap between the printed wiring board and the bare chip LSI is excluded from the hole when the underfill is filled. However, the technique described in Patent Document 1 is used when filling an underfill material in a gap between joined parts formed by flip-chip joining between semiconductor chips having a large number of electrode terminals. In order to eliminate the air remaining in the gap between the joining parts, it is difficult to make a hole in one of the semiconductor chips in advance because the semiconductor chip has a large number of electrode terminals, so that the reliability of the semiconductor chip is maintained. It is not preferable from the viewpoint of maintaining high integration of the semiconductor chip, and cannot be applied to bonding of semiconductor chips having a very large number of electrode terminals.

  The technique described in Patent Document 2 detects a plurality of corners of a chip with a camera and determines the application start position of an underfill material, and does not use the camera for detection of a fillet portion.

  In the technique related to the resin sealing method of the semiconductor device described in Patent Document 3, the monitoring camera uses a monitoring camera to confirm that the sealing resin has arrived diagonally to the lower right of the semiconductor chip when the wiring board is returned from the predetermined angle θ to θ = 0. Checking the state of the sealing resin on the side where the sealing resin is poured, monitoring the fillet part on the side where the sealing resin is poured, and sealing via the potting nozzle A surveillance camera is not used to control the additional supply of the stop resin to the semiconductor device.

  In the technology described in Patent Document 3, the sealing resin is poured from the upper left side of the semiconductor chip, the lower right side of the semiconductor chip is monitored by the monitoring camera, and the sealing resin has reached the lower right side of the semiconductor chip. Since it is the structure to confirm, it cannot monitor with a camera whether the sealing resin poured does not contain the bubble, and whether the fillet part was normally formed in the pouring side. For this reason, there is a possibility that air bubbles may be taken into the underfill part, and when the fillet part is not formed normally on the flow-in side or the flow-out side, it is not possible to take an appropriate response quickly, and reliability There is a problem that underfill material cannot be filled.

  In the technique described in Patent Document 4, the amount of underfill resin injected is adjusted so that the fillet length L is longer than the distance H from the surface of the resin substrate to the back surface of the semiconductor chip. In many assemblies in which semiconductor chips are mounted face down, the gap between the semiconductor chip and the resin substrate is not constant and varies, so the gap between the semiconductor chip and the resin substrate is filled into each assembly. In order to adjust the injection amount of the underfill resin so that the fillet length L is longer than the distance H, it is necessary to accurately know the gap between the semiconductor chip and the resin substrate in each assembly. However, there is a problem in that it takes time to measure the gap and the manufacturing time is increased. The same problem occurs in the case of chip-on-chip connection.

  Patent Document 5 describes that a sealing material is discharged (quantitative supply) adjacent to at least one side edge of a die, but does not show a configuration for detecting a fillet portion.

  In the prior art described above, there is a problem that reliable underfill filling cannot be performed without considering the case where the fillet portion is not normally formed.

  The present invention has been made in order to solve the above-described problems, and an object of the present invention is, for example, a gap between joined bodies by bonding between semiconductor chips via bumps or bonding between a semiconductor chip and a mounting substrate. Electronic device manufacturing apparatus and electronic device manufacturing method, underfill material filling state inspection apparatus and underfill material filling state, which can improve reliability and productivity It is to provide an inspection method.

  That is, the present invention provides an apparatus for manufacturing an electronic device in which a first part and a second part are joined to each other with a gap therebetween, and the gap is filled with an underfill material, Discharging means for discharging the underfill material for filling the fill material, detection means for detecting a fillet portion formed by the underfill material at a side end of the first component, and detection information of the fillet portion And the reference information of the regular fillet part, and when the former fillet width is smaller than the latter fillet width, it has a control means for additionally discharging the underfill material by the discharge means, The present invention relates to an electronic device manufacturing apparatus.

  The present invention also relates to a method of manufacturing an electronic device in which a first component and a second component are joined to each other with a gap, and the gap is filled with an underfill material, wherein the gap is filled with the underfill. A discharge step of discharging the underfill material from a tip of a discharge means to fill the material, a detection step of detecting a fillet portion formed by the underfill material at a side end of the first component, A step of comparing the detection information of the fillet portion with the reference information of the regular fillet portion, and an additional discharge step of additionally performing discharge of the underfill material when the former fillet width is smaller than the latter fillet width The present invention relates to a method for manufacturing an electronic device.

  Further, the present invention provides an apparatus for inspecting a filling state of an underfill material in manufacturing an electronic device in which a first part and a second part are joined to each other with a gap and the gap is filled with an underfill material. And detecting means for detecting a fillet portion formed by the underfill material at a side end of the first component, comparing the detection information of the fillet portion with reference information of a regular fillet portion, and the former And a determination means for determining whether or not the fillet width of the underfill material is smaller than the latter fillet width.

  The present invention also provides a method for inspecting a filling state of an underfill material in manufacturing an electronic device in which a first part and a second part are joined to each other with a gap and the gap is filled with an underfill material. A detection step of detecting a fillet portion formed by the underfill material at a side edge of the first component, and comparing the detection information of the fillet portion with reference information of a regular fillet portion, the former And a determination step for determining whether or not the fillet width is smaller than the latter fillet width.

  According to the present invention, the detection information of the fillet part detected by the detection means for detecting the fillet part formed of the underfill material at the side edge of the first component is compared with the reference information of the regular fillet part. When the former fillet width is smaller than the latter fillet width, the underfill material is further discharged by the discharge means, so that a fillet portion having a substantially constant width can be formed, and the filling gap It is possible to provide an electronic device manufacturing apparatus that can prevent the generation of bubbles and thereby improve the reliability and productivity.

  Further, in the step of comparing the detection information of the fillet portion detected in the detection step of detecting the fillet portion formed of the underfill material at the side edge of the first component with the reference information of the regular fillet portion, the former When the fillet width is smaller than the latter fillet width, the underfill material is additionally discharged, so that a fillet portion having a substantially constant width can be formed and the generation of bubbles in the filling gap is prevented. Thus, it is possible to provide a method for manufacturing an electronic device that can improve reliability and productivity.

  In addition, the detection information of the fillet portion detected by the detecting means for detecting the fillet portion formed of the underfill material at the side edge of the first component is compared with the reference information of the regular fillet portion, and the former fillet width is compared. Underfill material can be determined whether or not the filling amount into the underfill filling gap by the previously supplied underfill material is insufficient. A state inspection apparatus and inspection method can be provided.

  In the electronic device manufacturing apparatus of the present invention, when the width of the fillet portion detected by the detection unit is smaller than a predetermined width, the control unit discharges the underfill material in an additional manner. It is preferable that the means is controlled. With this configuration, additional discharge of the underfill material is performed until the detected width of the fillet portion becomes a substantially constant width that is equal to or greater than a predetermined width, so that generation of bubbles in the fillet filling gap can be prevented, and reliability is improved. And productivity can be improved.

  Further, it is preferable to have a calculation unit that calculates an additional discharge amount of the underfill material based on the detected width of the fillet portion. With this configuration, it is possible to calculate the amount of the underfill material that has been insufficient in the gap filling by the previously performed discharge of the underfill material based on the detected width of the fillet portion. By additional discharge of the underfill material, the width of the fillet portion can be set to a substantially constant width equal to or greater than a predetermined width.

  The detection means may be configured to detect the fillet portion at a side end that is the same as or different from the side end of the first component on which the tip of the discharge means is disposed. By detecting the fillet portion at the same side end, it is possible to monitor the discharge state of the underfill material and detect the formed fillet portion with a single inspection means, which occurs when the underfill material is discharged. Abnormal conditions such as ejection of an abnormal amount of underfill material, bubbles contained in the discharged underfill material, etc. can be detected, and appropriate measures such as stopping underfill filling can be promptly taken. Can be done.

  In addition, if the width of the fillet portion at a position away from the discharge position of the underfill material is smaller than the predetermined width due to the configuration in which the fillet portion is detected at different side ends, the discharge of the underfill material previously performed is performed. By performing additional discharge of the underfill material at the position or the above-mentioned distant position, the width of the fillet portion is substantially constant over a predetermined width over the entire joined body of the first component and the second component. It can be.

  Further, it is preferable that the detection unit performs imaging of the fillet portion. With this configuration, it is possible to monitor the state of the underfill material ejected based on the image data obtained by imaging. In addition, the width of the fillet portion can be accurately detected by calculating the image data obtained by imaging, and the additional discharge amount of the underfill material can be calculated more accurately.

  Further, it is preferable that a camera used for alignment of the tip of the ejection unit is used as the detection unit. With this configuration, since the camera for positioning the discharge tip of the underfill material detects the fillet portion, the apparatus configuration is not complicated and can be configured at low cost.

  The first component and the second component are preferably joined by metal bumps. Since this metal bump can be finely configured, it is suitable as a joining member when the first component and the second component are miniaturized.

  The first component may be a first semiconductor chip, and the second component may be a second semiconductor chip or a mounting substrate. With this configuration, electronic devices for various purposes can be provided.

  In the electronic device manufacturing method of the present invention, the additional discharge step is configured to additionally discharge the underfill material when the width of the fillet portion detected in the detection step is smaller than a predetermined width. It is good. With this configuration, additional discharge of the underfill material can be performed until the width of the detected fillet portion becomes a substantially constant width equal to or greater than a predetermined width.

  Moreover, it is good to set it as the structure which has the calculation process of calculating the additional discharge amount of the said underfill material based on the detected width | variety of the said fillet part. With this configuration, the amount of the underfill material that has been insufficient in the gap filling by the discharge of the underfill material performed previously is calculated based on the detected width of the fillet portion, so the minimum number of underfill materials is calculated. By the additional discharge, the width of the fillet portion can be set to a substantially constant width equal to or larger than a predetermined width.

  In the detection step, the fillet portion may be detected at a side end that is the same as or different from the side end of the first component on which the tip of the discharge means is disposed. By detecting the fillet part at the same side edge, it is possible to monitor the discharge state of the underfill material and detect the fillet part that is formed, and detect an abnormal state that occurs during the discharge of the underfill material. Therefore, it is possible to quickly take appropriate measures such as stopping underfill filling.

  In addition, if the width of the fillet portion at a position away from the discharge position of the underfill material is smaller than the predetermined width due to the configuration in which the fillet portion is detected at different side ends, the discharge of the underfill material previously performed is performed. By performing additional discharge of the underfill material at the position or the above-mentioned distant position, the width of the fillet portion is substantially constant over a predetermined width over the entire joined body of the first component and the second component. It can be.

  In the detection step, the fillet portion may be imaged. With this configuration, it is possible to monitor the state of the underfill material ejected based on the image data obtained by imaging. In addition, the width of the fillet portion can be accurately detected by calculating the image data obtained by imaging, and the additional discharge amount of the underfill material can be calculated more accurately.

  In addition, it is preferable to have a step of performing alignment of the tip of the discharge means using a camera, and to detect the fillet portion using the camera in the detection step. With this configuration, since the camera for positioning the discharge tip of the underfill material detects the fillet portion, an inexpensive manufacturing method can be provided.

  In the present invention, the gap (or gap) between the first component (for example, the upper semiconductor chip) filled with the underfill material and the second component (for example, the lower semiconductor chip or mounting substrate) is used. A fillet portion in which the underfill material extends from the underfill portion (or filling gap), which is the lower portion of the first component, and protrudes outside the side end of the first component is referred to as a fillet portion. The distance between the position where the fillet portion contacts the second component and the air layer and the position of the side edge of the first component is referred to as the width of the fillet portion or the fillet width. A line connecting the positions where the fillet portion is in contact with the first component and the air layer is referred to as an end line of the fillet portion. The end line of the fillet portion is formed substantially parallel to the side end of the first component.

Embodiment FIG. 1 is a diagram for explaining an apparatus for manufacturing an electronic device according to an embodiment of the present invention. FIG. 1 (A) is a diagram showing the outline of the apparatus configuration, and FIG. 1 (B) is a chip portion. FIG. 1C is a cross-sectional view at the ZZ portion of the tip portion.

  As shown in FIG. 1, the upper chip 10 and the lower chip 20 that are electrically and mechanically joined via the metal bumps 12 are held by a horizontal base (fixed horizontally) 21, and then the upper chip 10. An underfill material is filled in a gap between the lower chip 20 and the lower chip 20 by using a dispenser. When the underfill material discharged from the discharger is sufficient to fill the underfill portion, a fillet portion in which the underfill material extends from the underfill portion is formed.

  The discharger includes a discharge head 24 that stores an underfill material, a discharge nozzle 22 that discharges the underfill material, a narrow tube 30, a valve 32, an air pressure generator 34, and a valve drive unit 36 that controls the opening and closing of the valve 32. . The discharge head 24 is connected to an air pressure generator 34 through a thin tube 30, and the valve 32 is opened and closed to apply air pressure to the discharge head 24, thereby discharging the underfill material from the discharge nozzle 22.

  The ejection head 24 is held by an ejection head holding unit 26, and the ejection head holding unit 26 is mounted on an ejection head moving base that can move in three directions XYZ (not shown). The ejection tip of the ejection nozzle 22 can be set to a desired position by driving the ejection head moving base by the ejection head position driving unit 28.

  The camera 38 (1) aligns the discharge tip of the discharge nozzle 22 with the upper chip 10 and the lower chip 20 by the discharge head position driving unit 28, and (2) discharges the underfill material from the discharge tip of the discharge nozzle 22. Used for imaging, (3) imaging of the fillet portion, and the like.

  The camera 38 is held by a camera holding unit 40, and the camera holding unit 40 can move in three directions XYZ (not shown) and can change the direction angle of the field of view of the camera 38. It is mounted on a moving table. The position of the camera 38 and the direction angle of the visual field can be set to a desired position and direction angle by driving the camera moving table by the camera position driving unit 42.

  In addition, the camera holding unit 40 and the discharge head holding unit 26 are mounted on a common moving table that can move in three directions of XYZ, and position drive that also serves as the camera position driving unit 42 and the discharge head position driving unit 28. Depending on the unit, the position of the discharge tip of the camera 30 and the discharge nozzle 22 and the direction angle of the field of view of the camera 30 may be set.

  In the example shown in FIG. 1, the camera 38 is disposed at a position where the first side fillet portion 16 is imaged. However, the camera 38 may be separately disposed at a position where the third side fillet portion 18 is imaged. In such a configuration, the second-side fillet portion 17 and the fourth-side fillet portion 19 may be separately arranged at positions for imaging. Moreover, you may use the illuminating device for imaging.

  The valve drive unit 36, the discharge head position drive unit 28, and the camera position drive unit 42 of the discharger are each controlled by an arithmetic control unit 44. Image data picked up by the camera 38 is stored in the memory of the calculation control unit 44 and image processing calculation is performed on the image data. The image data protrudes from the side edge of the upper chip 10 and exists on the upper surface of the lower chip 20. The fillet portion formed by the underfill material is detected.

  The discharge tip of the discharge nozzle 22 is aligned at a predetermined position from the side end of the upper chip 10 and the upper surface of the lower chip 20, and a predetermined amount of underfill material is discharged from the discharge tip, and the upper chip 10 and the lower chip are discharged. At least one fillet of first to fourth side-side fillet portions 16, 17, 18, 19 formed at the side edges of the four sides of the upper chip 10 after being filled into the gap with 20 by capillary action. The image is captured by the camera 38, and image processing is performed by the arithmetic control unit 44 to detect at least one of the widths L1, L2, L3, and L4 of the fillet portion.

  When the detected width of the fillet portion formed by filling the gap between the upper and lower chips with the previously discharged underfill material is smaller than a predetermined reference width, the detected width of the fillet portion and the predetermined width Based on the reference width, the underfill material is additionally discharged from the discharge nozzle 22 to detect the width of the fillet portion so that a fillet portion having a width equal to or larger than a predetermined reference width is formed by the additional discharge of the underfill material. This is repeated until the detected width of the fillet portion is equal to or larger than a predetermined reference width. Such additional discharge control is automatically executed by the control calculation unit 44.

  In this way, as shown in FIG. 1, the first to fourth side fillet portions 16, 17, 18, 19 and the underfill material 14 at the lower portion (underfill portion) of the upper chip 10 are continuously formed. The widths L1, L2, L3, and L4 of the first to fourth side fillet portions 16, 17, 18, and 19 are substantially constant widths that are equal to or larger than a predetermined reference width.

  After the detected width of the fillet portion exceeds a predetermined reference width, the presence or absence of generation of bubbles in the first to fourth side fillet portions 16, 17, 18, 19 is detected from above the upper chip 10. It is good also as composition to do. This bubble detection can be performed using image data obtained by imaging the fillet portion from above the upper chip with a camera. The detection of the presence or absence of the generation of bubbles can also be performed while filling the gap between the upper and lower chips with the underfill material.

  2A and 2B are diagrams for explaining a filling state of the underfill material in the present embodiment, FIG. 2A is a state in which filling is in progress, and FIG. 2B is a state in which filling is insufficient. FIG. 2C is a cross-sectional view at the ZZ portion showing a state where additional discharge is necessary, and FIG.

  As shown in FIG. 2A, a predetermined amount of underfill material is discharged onto the surface of the lower chip 20 in the vicinity of the side edge of the upper chip 10 using the discharge nozzle 22, and the discharged underfill material is 40 The gap between the upper chip 10 and the lower chip 20 is filled by capillary action.

  At this time, for example, if there is uneven flow of the underfill material in the gap, the air existing in the gap is not pushed out of the gap by the underfill material, but remains in the gap as bubbles 48 or underflow from the discharge nozzle 22. When the discharge amount of the fill material is small, the unfilled portion 50 remains and the fillet portion is not formed. Such a state in which the bubbles 48 and the unfilled portion 50 are formed is easily affected by the external environment, and becomes a major cause of reducing reliability.

  As shown in FIG. 2B, even if a predetermined amount of underfill material is discharged so that a fillet portion having a certain width is formed by filling the gap in the underfill, the upper chip 10 to be filled is filled. Insufficient filling of the underfill material because the gap in the joined body between the lower chip 20 and the lower chip 20 varies among the lots of the joined body and due to variations in the discharge accuracy of the underfill material by the dispenser. May occur, and the unfilled portion 50 remains and the fillet portion is not formed (the width of the detected fillet portion is zero). In such a case, based on the detected width of the fillet portion, the underfill material is additionally discharged from the discharge nozzle so that the width of the fillet portion is equal to or larger than a predetermined reference width, thereby FIG. As shown, the width L of the fillet portions 16, 17, 18, and 19 is substantially constant, and the filling can be in a normal state.

  FIG. 3 is a diagram for explaining the evaluation of the state of filling with the underfill material in the present embodiment.

  Even if an attempt is made to form a fillet portion having a predetermined width by discharging a predetermined amount of the underfill material and filling the gap between the joined bodies, as shown in FIG. 3, it is actually accompanied by fillet portions having various shapes. The underfill material is filled. The following is a description of the evaluation results regarding the filling state of the underfill material, in which no bubbles are formed in the gap and the mounting of the electronic device is not affected.

  The filling state shown in FIG. 3 (A) is due to insufficient filling of the underfill material due to variations in the filling gap between lots of the joined body, variations due to the discharge accuracy of the underfill material by the dispenser, and the like. 50 remains and the state of filling is poor. Such a state occurs when the filling gap is large and the discharge amount of the underfill material is small. In the state where the fillet portion is not formed, the width of the detected fillet portion is zero. The state in which the unfilled portion 50 is formed is easily influenced by the external environment and causes a large decrease in reliability, and cannot be used for mounting.

  The filling state shown in FIG. 3B is a state where a fillet portion having a width of about 100 μm is formed, and no bubbles are generated in the filling gap.

  The filling state shown in FIG. 3C is a state where a fillet portion having a width of about 150 μm is formed, and no bubbles are generated in the filling gap.

  The filling state shown in FIG. 3D is a state Δ in which the discharge amount of the underfill material is too large and the excess underfill material 52 covers the side edge of the lower chip 20. Since no bubbles are generated in the filling gap, no mounting problem occurs unless the excess underfill material 52 is excessive. However, if the excessive underfill material 52 is excessive, it becomes an obstacle when mounting such an electronic device in close contact with the mounting substrate. The state shown in FIG. 3D tends to occur as the lower chip 20 becomes smaller and the area becomes smaller.

  The filling state shown in FIG. 3E is a state Δ in which the discharge amount of the underfill material is too large and the excess underfill material 54 covers the upper surface of the upper chip 10. Since no bubbles are generated in the filling gap, there is no problem when the electronic device is chip-on-chip connected if the excess underfill material 54 is very small. However, if the excessive underfill material 54 is excessive, it becomes an obstacle when chip-on-chip connection of such an electronic device is performed. The state shown in FIG. 3E tends to occur as the upper chip 10 becomes thinner and the lower chip 20 becomes smaller and the area becomes smaller.

  The inventor of the present invention has conducted various studies on the formation state of the fillet part and the generation of bubbles in the underfill part. As a result, the formation state of the fillet part and the generation of bubbles have a strong relationship, and there is a bubble in the filling gap. It has been found that it is desirable to form a fillet portion having a width of at least 100 μm or more, more preferably 150 μm or more, in order to prevent the occurrence of the phenomenon. 3A shows a state in which the filling state is poor, FIG. 3B shows a state in which the filling state is substantially good, FIG. 3C shows a state in which the filling state is good, and FIGS. E) is a filling state that can be used except when there is a problem in terms of mounting.

  FIG. 4 is a diagram for explaining the calculation of the discharge amount of the underfill material for realizing a good filling state in the present embodiment, and FIG. 4A shows an area where the underfill material exists in each part. FIG. 4B is a partially enlarged cross-sectional view of the chip portion in the ZZ portion.

  In order to form a fillet portion having a predetermined width L by filling an underfill material into a gap between the joined bodies, the upper chip 10, the lower chip 20, the gap between the upper chip 10 and the lower chip 20, the size of the metal bump 12 In response to this, it is necessary to discharge a predetermined amount of the underfill material from the discharge nozzle 22.

  The predetermined amount of the underfill material is calculated as shown in FIG. Here, it is assumed that the shape of the metal bump 12 joining the upper chip 10 and the lower chip 20 is a cylinder, and the area of the cross-sectional shape of the fillet portion is indicated by a dotted line at the side end of the upper chip 10 in FIG. Assuming that the area of the rectangle shown in FIG. 1 is one third, the filling gap is g, the upper chip short side is D2, the long side is D1, the thickness is h, the diameter of the metal bump 12 is d, and the specific gravity of the underfill material is Let ρ.

  When the predetermined amount is expressed by the discharge volume V and the discharge weight W, the volume V1 (= L × g) of the underfill material existing in the lower portion of the upper chip 10, the first and third side fillet portions 16, 18 Volume V2 (= D2 × L × (h + g) × (2/3)), volume V3 of the second and fourth side fillet portions 17 and 19 (= D1 × L × (h + g) × (2/3) ), Volume V4 (= L × L × π × (h + g) × (2/9)) of the first to fourth corner fillets 61, 62, 63, 64, and volume V5 of the metal bump 12 (= d × d ×). Using π × g × (number of metal bumps), discharge volume V = (V1 + V2 + V3 + V4-V5), Wi = Vi × ρ (i = 1, 2, 3, 4, 5), and discharge weight W = V Xρ.

  As described above, fillet portions having a predetermined width L are always formed by the above discharge amounts (V, W) due to variations in filling gap between lots of joined bodies and variations due to discharge accuracy of the discharger. Never happen.

  FIG. 5 is a diagram showing a comparison between a calculated value and a measured value of the relationship between the discharge amount of the underfill material and the width of the fillet portion in the present embodiment.

  5A is a case where the filling gap g is 18.5 μm shown in FIG. 4, FIG. 5B is a case where the filling gap g is 23.5 μm shown in FIG. 4, and FIG. 5C is a filling gap which is shown in FIG. The result in the case of g = 28.5 micrometers is shown.

Here, the fillet width L = 150 μm, the filling gap g = 23.5 ± 5 μm, the upper chip short side D2 = 5.2 mm, the upper chip long side D1 = 6.7 mm, the upper chip thickness h = 130 μm, metal bump 12 diameter = 30 μm, number of metal bumps = 1400, lower chip long side = 8.2 mm, lower ship short side = 7.7 mm, specific gravity of underfill material ρ = 1.5 g / cm ³

  For example, when the filling gap g = 23.5 μm, W = 1.5 mg, and when the filling gap g = 28.5 μm, W = 1.7 mg.

  The measured values shown in FIG. 5 are obtained by measuring the discharge amount of the underfill material with an electronic balance and detecting the fillet portion by image processing including binarization processing of an image captured by the camera with the configuration shown in FIG. An average value of five ejection measurements for obtaining the distance (width) from the side edge of the upper chip is shown. The width when the fillet portion is not formed is zero. The calculated values and the measured values are in good agreement. The end line of the fillet part connecting the positions where the fillet parts detected in the first to fourth side fillet parts 16, 17, 18, and 19 are in contact with the lower chip and the air layer is linear, and the fillet parts of substantially the same fillet part A width was detected.

  Due to the accuracy of the apparatus for joining the upper and lower chips, the filling gap varies, that is, the filling gap varies between lots of joined bodies as g = 23.5 ± 5 μm, so that the filling gap becomes maximum g = If it is 28.5 μm, W = 1.7 mg as described above. Normally, when the discharge amount is increased, as shown in FIG. 3 (E), the excess underfill material runs on the upper surface of the upper chip 10. Therefore, the discharge is performed with an average discharge amount W = 1.5 mg corresponding to the filling gap (average value) g = 23.5 μm as a target value. When the discharge accuracy of the discharge device is ± 0.15 mg and the actual discharge amount is small (1.5−0.15 = 1.35 mg), the required discharge amount for the filling gap g = 28.5 μm W = 1.7 mg is not reached, a shortage of 0.35 mg occurs, and the worst case of poor filling with no fillet portion occurs. On the other hand, as described above, a normal product with a good filling state has a shape in which the width of the fillet portion is 100 μm or more. Moreover, even if it does not go as much as the above-mentioned worst case, the filling defect which the width | variety of a fillet part does not have 100 micrometers or more may generate | occur | produce.

  Therefore, after discharging the underfill material, as described above, the camera is moved to the side surface of the upper chip 10 to pick up a predetermined field of view, the obtained image is processed, the fillet portion is detected, and its width is If it is 0-50 μm, a control signal is sent from the arithmetic control unit 44 to the valve control unit 36 to cause 0.35 mg of the underfill material, which is the worst case, to be discharged again, and if the width of the fillet part is 50-100 μm, By performing control such that 0.17 mg of the filling material is discharged again, the width of the fillet portion can be set to about 150 μm, and an unfilled defect can be repaired.

  FIG. 6 is a diagram for explaining a method of manufacturing an electronic device in the present embodiment.

  In the memory of the arithmetic control unit 44 shown in FIG. 1, the size and thickness of each side of the upper and lower chips, the average value and variation of the filling gap of the joined body, the size of the metal bump, and the normal state A predetermined reference width that is reference information of the fillet portion, discharge accuracy of the discharger, information on a calculation formula that represents a relationship between the discharge amount of the underfill material and the width of the fillet portion illustrated in FIG. 5 and the like are stored in advance. . Hereinafter, each step will be described.

S1: Underfill Material Discharge The calculation control unit 44 discharges an average discharge amount of the underfill material corresponding to the average value of the filling gap of the joined body after control for setting the positions of the discharge nozzle 22 and the camera 38. Take control. Prior to discharging the underfill material, the underfill material and the upper and lower chips are uniformly preheated to, for example, 80 ° C. to 90 ° C. in order to reduce the viscosity, increase the flow rate, and improve the wetting. Yes. Preheating is performed by any one or combination of contact heating, infrared heating, convection heating.

S2: Detecting the width of the fillet part The camera 38 takes an image of the fillet part formed by filling the filling gap with an underfill material having an average discharge amount in a predetermined field of view, and the arithmetic control unit 44 takes the captured image. The data is processed to detect the fillet width L. At least one or more fillet portions of the first to fourth side fillet portions 16, 17, 18, 19 formed at the side edge of the upper chip 10 are imaged by the camera 38 and image processing is performed by the arithmetic control unit 44. And at least one of L1, L2, L3, and L4 is detected as the width L of the fillet portion.

  For example, as shown in FIG. 1, the discharge nozzle 22 is arranged in the vicinity of the side edge on the first side of the upper chip, and the entire side edge on the first side of the upper chip, the first side fillet portion 16 is arranged. The field of view of the camera 38 so as to include the whole, part of the side edges of the upper chip on the second and fourth sides, part of the second side fillet part 17 and part of the fourth side fillet part 19. And the width of the fillet portion in the first side fillet portion 16 is measured. At this time, you may measure the width of each fillet part from a part of fillet part in the 2nd side fillet part 17 and the 4th side fillet part 19. FIG.

  As in FIG. 1, the discharge nozzle 22 is disposed in the vicinity of the side edge on the first side of the upper chip, and the entire side edge on the third side of the upper chip, the third side fillet portion 18 is arranged. The field of view of the camera 38 so as to include the whole, part of the side edges of the upper chip on the second and fourth sides, part of the second side fillet part 17 and part of the fourth side fillet part 19. And the width of the fillet portion in the third side fillet portion 18 is measured. At this time, similarly to the above example, the width of each fillet portion may be measured from a part of the fillet portions in the second side fillet portion 17 and the fourth side fillet portion 19.

  When the end line of the fillet portion formed at the side edge of the upper chip 10 is not parallel to the side edge of the upper chip 10, the average value of the width of the fillet portion at a plurality of positions of the end line of the fillet portion is detected Suppose that the width L is set. Further, the minimum value of the fillet portion at a plurality of positions on the end line of the fillet portion may be the detected width L.

S3: Comparison between detected width L of fillet portion and predetermined width (reference width) L0 L0 is at least 100 μm, preferably 150 μm, and calculation control unit 44 determines the detected width L of the fillet portion and the predetermined width. Compare with L0. When two or more of L1, L2, L3, and L4 are detected, the average value or minimum value of the detected width is used as L.

S4: Decision branch When L <L0 is established, the process of S5 is executed, and when not established, S8 is executed.

S5: Calculation of the additional discharge amount based on the detected width L of the fillet portion The calculation control unit 44 detects the detected width L, the predetermined width L0, the discharge amount of the underfill material and the width of the fillet portion shown in FIG. The additional discharge amount to be discharged in addition to the previous discharge amount is calculated from the calculation formula representing the relationship between the first discharge amount and the previous discharge amount.

S6: Control of additional discharge of underfill material The calculation control unit 44 sends a control signal to the valve drive unit 36 so as to discharge the additional discharge amount.

S7: Execution of Additional Discharge of Underfill Material An additional discharge amount of the underfill material is discharged from the discharge nozzle 22 by driving the valve drive unit 36. The additional discharge is performed from the position where the discharge is performed first, or from the position on the side which gives the minimum value of the detected width when two or more of L1, L2, L3 and L4 are detected.

S8: Completion of additional discharge of underfill material A fillet portion having a predetermined width or more is formed, the filling gap is normally filled with the underfill material, and the filling step is finished.

S9: Curing of underfill material Following the completion of the filling process, the underfill material is heated and cured under predetermined conditions. In the present embodiment, the underfill material is cured in a single curing process.

  Thus, the filling and curing of the underfill material for one bonded body is completed, and the manufacture of the electronic device is completed.

  Note that the density of metal bumps on the bonding surface (defined as the number of metal bumps arranged per unit area) is not uniform, and the bonding surface region (for example, the region near the side edge of the upper chip, the upper part) If the density of the metal bumps varies depending on the area near the center of the chip, etc., the uniform preheating of the upper and lower chips will cause the speed of the underfill material to flow through the gap between the upper and lower chips. Depending on the density, the air bubbles are easily mixed in the underfill material filled in the gap.

  In such a case, when the upper and lower chips are uniformly preheated before discharging the underfill material, the upper and lower chips are preheated by changing the temperature according to the density of the metal bumps on the bonding surface. It is good. For example, when T1> T2, in a region where the density of the metal bumps is large (in such a region, the flow resistance is large and the underfill material is difficult to flow), preheating is performed at the temperature T1, and the density of the metal bumps is small. In the region (in such a region, the flow resistance is small and the underfill material flows easily), preheating is performed at the temperature T2 (for example, T1 = 90 ° C. to T2 = 80 ° C.). As a result, regardless of the density of the metal bumps in the portion where the underfill material flows, the underfill material flows through the gap between the upper and lower chips at a substantially uniform speed, thereby avoiding the inclusion of bubbles. In addition, fillet portions having substantially the same width can be formed at each side end of the upper chip.

  Furthermore, in order to increase the flow of the underfill material, when the discharge of the average discharge amount of the underfill material and the additional discharge are executed, the side end side of the upper chip where the underfill material is discharged becomes high. As described above, the horizontal table 21 holding the upper limit chip may be fixed at a desired inclination angle.

  According to the manufacturing method described above, at least one or more of the four side edges (sides) of the upper chip with respect to the fillet portion formed by discharging the underfill material and filling the underfill portion previously performed. By detecting the width of the fillet part, it is possible to underfill according to the condition of the fillet part formed by the previous filling without being affected by the lot-to-lot variation of the joined member of the filling gap, the discharge accuracy of the discharger, etc. It is possible to control the additional discharge of the material so that the widths of all detected fillet portions are equal to or larger than the reference width.

  Further, as described above, the underfill material that is additionally discharged by preheating the underfill material and the upper and lower chips prior to the discharge of the underfill material is the same as the underfill material previously filled in the gap. Since the underfill material filled before and after does not wet well with each other and does not form an interface between the underfill materials, the additional underfill material is also bonded to the upper and lower chip joint surfaces. Therefore, it is possible to prevent bubbles from being generated in the underfill portion.

  FIG. 7 is a view for explaining the arrangement of the discharge nozzle 22 and the camera 38 for measuring the width of the fillet portion in the present embodiment, and FIG. 7A is a plan view showing an arrangement example of the discharge nozzle 22. FIG. 7B is a diagram showing the arrangement of the camera 38.

  In FIG. 7A, as a configuration in which only the first nozzle 56 is used as the discharge nozzle 22 shown in FIG. 1, an area including the first side fillet portion 16 is used as the first camera as shown in FIG. 7B. The area including the second side view fillet portion 17 is the second camera view 67, the area including the third side fillet portion 18 is the third camera view 68, and the fourth side fillet portion 19 is included. The fourth camera field of view 69 may be used, and a camera for capturing each of these camera fields of view may be individually arranged.

  In addition, since the third camera field of view 68 includes a part of the second camera field of view 67 and the fourth camera field of view 69, the third camera field of view 68 shown in FIG. It is good also as a structure which uses only the camera for imaging. At this time, in order to detect the width of the first side fillet portion 16, a camera for imaging the first camera visual field 66 may be further used. Furthermore, it is good also as a structure which images the 1st-4th side fillet part 16, 17, 18, 19 with a single camera. The fillet portion is detected using image data obtained by imaging the fillet portion with a camera. If the fillet portion is detected at a specific position as in the technique described in Patent Document 3, it is formed on the substrate or chip at this position. If there is a wiring, there is a risk that it will be difficult to distinguish between this wiring and the fillet. In such a case, in the present invention, the most easily distinguishable place is used for detecting the fillet part. The width of the fillet portion can be accurately measured.

  The discharge nozzle 22 shown in FIG. 1 discharges the underfill material in a stationary state, but the underfill material is moved linearly along the side edge of the upper chip 10 while the discharge nozzle 22 shown in FIG. The material may be discharged. Further, instead of the discharge nozzle 22 shown in FIG. 1, the first nozzle 56 and the second nozzle 57, or the first nozzle 56 and the third nozzle 59 are used, and the two nozzles warp against the side edges of the upper chip 10. The underfill material may be discharged while moving linearly. Furthermore, the first nozzle 56, the second nozzle 57, and the third nozzle 59 may be used to discharge the underfill material while moving the three nozzles linearly along the side edges of the upper chip 10.

  According to the present embodiment described above, by providing the discharge (dispenser) device with an appearance inspection function, information on the formation of the fillet portion accompanying the filling of the underfill material can be easily transferred to the underfill filling process in real time. Since it can be fed back, it can respond quickly to problems such as clogging of the discharge nozzle (needle), produce a large number of defective products, and also feed back the appearance inspection function and the appearance inspection result to the discharge device to the discharge (dispenser) device Based on the appearance inspection result data of the fillet portion, the additional discharge amount of the underfill material necessary for forming a normal fillet portion is calculated and the discharge device (dispenser) is controlled to perform additional discharge. By having the function of It is possible to automatically prevent the occurrence of defective electronic devices that fit, productivity, can improve the yield.

  In addition, the appearance is inspected before the underfill material is cured, and if necessary, additional underfill material is discharged and refilled, so the underfill material is added to the already filled and hardened underfill material. There is no generation of bubbles between the hardened and uncured underfill materials, which is a concern in conventional restoration work of filling, and production time is not required because it does not require time to perform the second curing. Contribute to up.

  In this embodiment, unlike the configuration described in Patent Document 3 in which the lower right corner of the semiconductor chip is monitored by a monitoring camera and it is confirmed that the sealing resin has reached the lower right corner of the semiconductor chip. To monitor the state of the underfill material on the material supply (encapsulating resin) supply side, and to monitor the fillet on the underfill material supply side, to the underfill material filling gap from the discharge tip of the discharger In order to control the additional supply, the camera is used.

  With this configuration, whether or not the underfill material discharged from the discharge tip of the discharger and supplied to the filling gap does not include bubbles, which cannot be realized by the technique described in Patent Document 3, the underfill material supply side Whether the fillet portion is normally formed can be monitored by the camera. As a result, if the supplied underfill material contains bubbles, it is determined that there is a high possibility that the bubbles will be taken into the underfill portion, the underfill process is stopped, and the state of the underfill material in the discharger is reviewed. .

  In addition, if the fillet part is not formed properly on the underfill material supply side or a different side (for example, the opposite side), an appropriate response is quickly made, that is, the underfill material is added to the gap. A filling operation can be performed to form a normal fillet.

  As described above, in the present embodiment, by detecting the fillet portion, even when the amount of filling of the underfill material into the filling gap is insufficient, the underfill material is filled into the filling gap so as to form a fillet having a certain width. Since additional filling is employed, the generation of bubbles in the filling gap can be prevented, and productivity and reliability can be improved.

  For example, in a large number of assemblies in which a semiconductor chip and a substrate or between semiconductor chips are joined via bumps, even if there are variations in the gap between the semiconductor chip and the substrate or between the semiconductor chips, this gap Based on the detected width of the fillet without measuring the discharge, the discharge amount is controlled so as to form a fillet with a constant width regardless of gap variation, and an underfill material is additionally filled in the gap. The width of the part can be made substantially constant. In addition, underfill material is added to the gap by controlling the discharge amount so as to form a fillet with a constant width based on the detected fillet width without being affected by the discharge amount accuracy of the discharger. By filling, the width of the fillet portion can be made substantially constant.

  As mentioned above, although embodiment of this invention was described, this invention is not limited to the above-mentioned embodiment, Various deformation | transformation based on the technical idea of this invention is possible.

  For example, it goes without saying that the present invention is applicable to a configuration in which a mounting substrate such as an interposer substrate or a flexible substrate is used instead of the lower chip. Furthermore, it goes without saying that the present invention can also be applied to a configuration in which a semiconductor package or a module package is used instead of the upper chip, a mounting substrate is used instead of the lower chip, and the package and the mounting substrate are joined by bump electrodes.

  Further, the size of the upper and lower chips, the size and number of metal bumps, and the like can be arbitrarily changed as necessary.

  Moreover, in order to detect a fillet part more correctly, you may color and use an underfill material in the color different from an upper part and a lower chip | tip. Moreover, it is good also as a structure which mixes and uses fluorescent substance particle | grains with an underfill material, light-emits a fluorescent substance with an ultraviolet-ray, images a light emission image, and detects a fillet part.

  As described above, the present invention can provide a bonded body in which reliability of the underfill material is improved and reliability and productivity are improved in bonding between components by flip chip connection.

BRIEF DESCRIPTION OF THE DRAWINGS In the embodiment of this invention, the apparatus which manufactures an electronic device is demonstrated, (A) The schematic of an apparatus structure, (B) The top view of a chip part, (C) The sectional view in the ZZ part of a chip part. is there. Same as above, illustrating the filling state of the underfill material, (A) a state in the middle of filling, (B) a state in which filling is insufficient, (C) a sectional view in the ZZ section showing a state in which filling is normal. is there. It is a figure explaining evaluation of the state of filling of an underfill material same as the above. It is a figure explaining calculation of the discharge amount of an underfill material same as the above. It is a figure which shows the comparison with the calculated value of the relationship between the discharge amount of an underfill material, and the width | variety of a fillet part, and a measured value same as the above. It is a figure explaining the manufacturing method of an electronic device same as the above. It is a figure explaining arrangement | positioning of a discharge nozzle and a camera same as the above. It is explanatory drawing which shows the state which fills a semiconductor device with sealing resin by the potting method in a prior art. (A) Schematic diagram showing an underfill process performed using capillary action and heating the substrate to a uniform temperature, and showing a time sequence of a normal underfill process depending on capillary action, 2) A schematic perspective view of a die and substrate package assembly during an underfill process, and FIG. 2C is a schematic side view of a die and substrate package assembly after the underfill process.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Upper chip | tip, 12 ... Metal bump, 14 ... Underfill material of the lower part of an upper chip | tip, 16 ... 1st edge | side side fillet part, 17 ... 2nd edge | side side fillet part,
18 ... Third side fillet, 19 ... Fourth side fillet, 20 ... Lower tip,
21 ... Horizontal base, 22 ... Discharge nozzle, 24 ... Discharge head, 26 ... Discharge head holding part,
28 ... Discharge head position drive unit, 30 ... Narrow tube, 32 ... Valve, 34 ... Air pressure generator,
36 ... Valve drive unit, 38 ... Camera, 40 ... Camera holding unit, 42 ... Camera position drive unit,
46 ... discharged underfill material, 48 ... bubbles, 50 ... unfilled part,
52, 54 ... surplus underfill material, 56 ... first discharge nozzle, 57 ... second discharge nozzle,
59 ... 3rd discharge nozzle, 61 ... 1st corner fillet part, 62 ... 2nd corner fillet part,
63 ... third corner fillet part, 64 ... fourth corner fillet part, 66 ... first camera field of view,
67 ... second camera view, 68 ... third camera view, 69 ... fourth camera view

Claims (16)

An apparatus for manufacturing an electronic device in which a first part and a second part are joined to each other with a gap, and the gap is filled with an underfill material,
Discharging means for discharging the underfill material in order to fill the gap with the underfill material;
Detecting means for detecting a fillet portion formed by the underfill material at a side end of the first component;
The detection information of the fillet part is compared with the reference information of the regular fillet part, and when the former fillet width is smaller than the latter fillet width, the underfill material is discharged by the discharge means. An electronic device manufacturing apparatus, comprising: a control means for performing.   2. The control unit according to claim 1, wherein when the width of the fillet portion detected by the detection unit is smaller than a predetermined width, the control unit controls the discharge unit to additionally discharge the underfill material. Electronic device manufacturing equipment.   The electronic device manufacturing apparatus according to claim 2, further comprising a calculation unit that calculates an additional discharge amount of the underfill material based on the detected width of the fillet portion.   The electronic device manufacturing apparatus according to claim 1, wherein the detection unit detects the fillet portion at a side end that is the same as or different from a side end of the first component on which a tip of the discharge unit is disposed.   The electronic device manufacturing apparatus according to claim 1, wherein the detection unit performs imaging of the fillet portion.   The electronic device manufacturing apparatus according to claim 1, wherein a camera used for alignment of a tip of the discharge unit is used as the detection unit.   The apparatus for manufacturing an electronic device according to claim 1, wherein the first component and the second component are joined by a metal bump.   The electronic device manufacturing apparatus according to claim 1, wherein the first component is a first semiconductor chip, and the second component is a second semiconductor chip or a mounting substrate. A method of manufacturing an electronic device in which a first component and a second component are joined to each other with a gap, and the gap is filled with an underfill material,
A discharge step of discharging the underfill material from the tip of discharge means to fill the gap with the underfill material;
A detection step of detecting a fillet formed by the underfill material at a side edge of the first component;
Comparing the detection information of the fillet part with the reference information of the regular fillet part;
When the former fillet width is smaller than the latter fillet width, the electronic device manufacturing method includes an additional discharge step of additionally discharging the underfill material.   10. The method of manufacturing an electronic device according to claim 9, wherein the additional discharge step is performed by additionally discharging the underfill material when the width of the fillet portion detected in the detection step is smaller than a predetermined width.   The method for manufacturing an electronic device according to claim 10, further comprising a calculation step of calculating an additional discharge amount of the underfill material based on the detected width of the fillet portion.   10. The method of manufacturing an electronic device according to claim 9, wherein, in the detection step, the fillet portion is detected at a side end that is the same as or different from the side end of the first component on which the tip of the ejection unit is disposed.   The method for manufacturing an electronic device according to claim 9, wherein in the detection step, the fillet portion is imaged.   The method of manufacturing an electronic device according to claim 9, further comprising a step of aligning a tip of the ejection unit using a camera, wherein the fillet portion is detected using the camera in the detection step. An apparatus for inspecting a filling state of an underfill material in manufacturing an electronic device in which a first part and a second part are joined to each other with a gap, and the gap is filled with an underfill material,
Detecting means for detecting a fillet portion formed by the underfill material at a side end of the first component;
Inspecting the filling state of the underfill material, comprising means for comparing the detection information of the fillet part with the reference information of the regular fillet part and determining whether the former fillet width is smaller than the latter fillet width apparatus. A method of inspecting a filling state of an underfill material in manufacturing an electronic device in which a first part and a second part are joined to each other with a gap, and the gap is filled with an underfill material,
A detection step of detecting a fillet formed by the underfill material at a side edge of the first component;
An inspection of the underfill material filling state, comprising a step of comparing the detection information of the fillet part with reference information of the regular fillet part and determining whether the former fillet width is smaller than the latter fillet width. Method.

Images